Вы находитесь на странице: 1из 1451

FRACTURE OF NANO AND ENGINEERING MATERIALS

AND STRUCTURES
Fracture of Nano and Engineering
Materials and Structures
Proceedings of the 16th European Conference of
Fracture, Alexandroupolis, Greece, July 3-7, 2006

Edited by

E. E. GDOUTOS
Democritus University of Thrace,
Dept of Civil Engineering, Xanthi, Greece
A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN-10 1-4020-4971-4 (HB)


ISBN-13 978-1-4020-4971-2 (HB)
ISBN-10 1-4020-4972-2 ( e-book)
ISBN-13 978-1-4020-4972-9 (e-book)

Published by Springer,
P.O. Box 17, 3300 AA Dordrecht, The Netherlands.

www.springer.com

Cover picture
Fracture and Delamination of Oxide: Fracture and delamination of 1µm (1x10–6 m)
SiO2 on Si with 1µm conical probe tip. Courtesy of Hysitron Inc., Minneapolis, Minnesota, USA

Printed on acid-free paper

All Rights Reserved


© 2006 Springer
No part of this work may be reproduced, stored in a retrieval system, or transmitted
in any form or by any means, electronic, mechanical, photocopying, microfilming, recording
or otherwise, without written permission from the Publisher, with the exception
of any material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work.

Printed in the Netherlands.


Contents

Editor’s Preface............................................................................................................ xliii


Organizing Committees ................................................................................................ xlv
ECF16 TRACKS......................................................................................................... xlvii
ECF16 SPECIAL SYMPOSIA/SESSIONS .............................................................. xlvix

A. INVITED PAPERS ........................................................................................... 1


Deformation and Fracture at the Micron and Nano Scales.............................................. 3
E. C. Aifantis
Statistical Mechanics of Safety Factors and Size Effect in Quasibrittle Fracture............ 5
Z. P. Bazant and S.-D. Pang
Nanoreliability – Fracture Mechanics on the Way from Micro to Nano ......................... 7
B. Michel
Fracture Mechanics and Complexity Sciences ................................................................ 9
A. Carpinteri and S. Puzzi
Failure of Composite Materials...................................................................................... 11
I. M. Daniel
Interactions of Constrained Flow and Size Scale on Mechanical Behavior .................. 13
W. W. Gerberich, W. M. Mook, M. J. Cordill and D. Hallman
Space Shuttle Columbia Post-Accident Analysis and Investigation ............................. 15
S. McDanels
The Role of Adhesion and Fracture on the Performance of Nanostructured Films....... 17
N. Moody, M. J. Cordill, M. S. Kennedy, D. P. Adams, D. F. Bahr
and W. W. Gerberich
Assessment of Weldment Specimens Containing Residual Stress ................................ 19
K. M. Nikbin
MEMS: Recent Advances and Current Challenges ....................................................... 21
R. J. Pryputniewicz
Fracture, Aging and Disease in Bone and Teeth ............................................................ 23
R. O. Ritchie and R. K. Nalla
Laboratory Earthquakes ................................................................................................. 25
A. J. Rosakis, K. Xia and H. Kanamori
A Historical Retrospective of the Beginnings of Brittle Fracture Mechanics -
The Period 1907-1947.................................................................................................... 27
H. P. Rossmanith
Dynamic Crack Propagation in Particle Reinforced Nanocomposites and Graded
Materials......................................................................................................................... 29
A. Shukla
Spatial and Temporal Scaling Affected by System Inhomogeneity: Atomic,
Microscopic and Macroscopic. ...................................................................................... 31
G. C. Sih
vi Contents

B. TRACKS
B1. Nanomaterials and Nanostructures ........................................................ 33

1T1. Fracture and Fatigue of Nanostuctured Materials .......................................... 35


Channeling Effect in Fracture of Materials with Nanostructured Surface Layers......... 35
V. E. Panin and A. V. Panin
Atomistics and Configurational Forces in Gradient Elasticity ...................................... 37
P. Steinmann and E. C. Aifantis
Tensile Behavior and Fracture of Carbon Nanotubes Containing Stone-Wales Defects 39
K. I. Tserpes and P. Papanikos
Atomic-Scale Investigation on Fracture Toughness in Nano-Composite
Silicon Carbide............................................................................................................... 41
M. Ippolito, A. Mattoni, L. Colombo and F. Cleri
Multiscale Modeling and Computer Simulation of Stress-Deformation Relationships
in Nanoparticle-Reinforced Composite Materials ......................................................... 43
L. V. Bochkaryova, M. V. Kireitseu, G. R. Tomlinson, V. Kompis
and H. Altenbach
The Mechanical Parameters of Nanoobjects (Theory and Experiment) ........................ 45
E. Ivanova, N. Morozov and B. Semenov
Advanced Manufacturing Design Concepts and Modelling Tools of the Next
Generation Nanoparticle-Reinforced Damping Materials ............................................ 47
M. V. Kireitseu, G. R. Tomlinson, R. A. Williams and V. Kompis
Fracture of Nanostructured Ionomer Membranes.......................................................... 49
Yue Zou, X. Huang and K. L. Reifsnider

1T2. Failure Mechanisms ............................................................................................ 51


Deformation and Limit States of Carbon Nanotubes under Complex Loading............. 51
A. V. Chentsov and R. V. Goldstein
Interaction of Domain Walls with Defects in Ferroelectric Materials ........................... 53
D. Schrade, R. Mueller, D. Gross, T. Utschig, V. Ya. Shur, D. C. Lupascu
Microstructure and Internal Stresses in Cyclically Deformed Al and Cu Single
Crystals........................................................................................................................... 55
M. E. Kassner
Determination of Equilibrium Configurations of Atomic Lattices at Quasistatic
Deformation ................................................................................................................... 57
S. N. Korobeynikov
Multiscale Mechanics of Carbon Nanotubes and their Composites .............................. 59
X.-Q. Feng

1T4. Fatigue and Fracture of MEMS and NEMS..................................................... 61


In-Situ Scanning Electron Microscope Indentation of Gallium Arsenide ..................... 61
C. Pouvreau, K. Wasmer, J. Giovanola, J. Michler, J. M. Breguet
and A. Karimi
Fracture of Nanostructured Lithium Batteries ............................................................... 63
K. E. Aifantis, J. P. Dempsey and S. A. Hackney
Analytical and Experimental Characterization of a Micromirror System ..................... 65
E. J. Pryputniewicz, C. Furlong and R. J. Pryputniewicz
Contents vii

A Metal Interposer for Isolating MEMS Devices from Package Stresses ..................... 67
R. J. Pryputniewicz, T. F. Marinis, J. W. Soucy, P. Hefti and A. R. Klempner
Computational Modeling of Nanoparticles in Biomicrofluidic Devices ....................... 69
R. J. Pryputniewicz, Z. Sikorski, M. Athavale, Z. J. Chen and A. J. Przekwas
Characterization of a MEMS Pressure Sensor by a Hybrid Methodology .................... 71
R. J. Pryputniewicz and C. Furlong
New Approach to Synthesis of Laser Microwelding Processes for Packaging ............. 73
R. J. Pryputniewicz, W. Han and K. A. Nowakowski
Thermal Management of RF MEMS Relay Switch....................................................... 75
R. J. Pryputniewicz

1T7. Thin Films ............................................................................................................ 77


Buckling and Delamination of Thin Layers on a Polymer Substrate ............................ 77
A. A. Abdallah, D. Kozodaev, P. C. P. Bouten, J. M. J. den Toonder
and G. De With
Carbide Coated Cutting Tool Properties Investigation by Nano-Mechanical
Measurements under 250-500°C.................................................................................... 79
B. Vasques, D. Joly, R. Leroy, N. Ranganathan and P. Donnadieu
Diamond Coating Debounding in Tool Application ...................................................... 81
D. Moulin, P. Chevrier, P. Lipinski and T. Barré
Interfacial Strength of Ceramic Thin Film on Polymer Substrate ................................. 83
M. Omiya and K. Kishimoto
Delaminate Behavior of PVD/CPVD Thin Film ........................................................... 85
S. Doi and M. Yasuoka
Experimental Study of Microhardness and Fracture of Implanted Gállium
Nitride Films .................................................................................................................. 87
P. Kavouras, M. Katsikini, E. Wendler, W. Wesch, H. M. Polatoglou,
E. C. Paloura, Ph. Komninou and Th. Karakostas

1T9. Failure of Nanocomposites ................................................................................. 89


Crack Tip Strain Field and its Propagation Characteristics in a Polymer Foam............ 89
F.-P. Chiang, S. Chang and Y. Ding
How to Toughen Ceramics – Nanocomposites .............................................................. 91
H. Awaji and S.-M. Choi
Deformation and Fracture Behaviour of Nanocomposites ............................................ 93
S. Dunger, J. K. W. Sandler, K. Hedicke and V. Altsadt
Fracture Mechanisms in Carbon Nanotube-Reinforced Composites ............................ 95
E. T. Thostenson and T.-W. Chou
viii Contents

B. TRACKS
B2. Engineering Materials and Structures .................................................. 97

2T1. Physical Aspects of Fracture .............................................................................. 99


Fractal Approach to Crack Problems with Non-Root Singularity ................................. 99
A. Kashtanov
New Method for Analysing the Magnetic Emission Signals During Fracture ............ 101
Gy. B. Lenkey, N. Takacs, F. Kun and D. L. Beke
Electromagnetic Radiation Method for Identification of Multi-Scale Fracture .......... 103
Yu. K. Bivin, A. S. Chursin, E. A. Deviatkin and I. V. Simonov
Micromechanical Modeling of Grain Boundary Resistance to Cleavage Fracture
Propagation .................................................................................................................. 105
M. Stec and J. Faleskog
Microstructure of Reactor Pressure Vessel Steel Close to the Fracture Surface.......... 107
M. Karlik, P. Hausild and C. Prioul

2T2. Brittle Fracture.................................................................................................. 109


Brittle Fracture in Heat-Affected Zones of Girth Welds of Modern Line Pipe Steel
(X100) .......................................................................................................................... 109
A. S. Bilat, A. F. Gourgues-Lorenzon, J. Besson and A. Pineau
Cleavage Fracture of Steels at Very Low Temperatures .............................................. 111
R. Rodriguez-Martin, I. Ocana and A. Martin-Meizoso
New formulation of the Ritchie, Knot and Rice Hypothesis ....................................... 113
A. Neimitz, M. Graba and J. Galkiewicz
The Effect of the Rate of Displacement on Crack Path Stability............................... 115
D. A. Zacharopoulos and P. A. Kalaitzidis
Scratching and Brittle Fracture of Semiconductor In-Situ Scanning Electron
Microscope................................................................................................................... 117
K. Wasmer, C. Pouvreau1, J. Giovanola and J. Michler
Cracks in Thin Sheets: when Geometry Rules the Fracture Path ................................ 119
P. M. Reis, B. Audoly and B. Roman
Cleavage Mechanisms in a Ship Plate Steel ................................................................ 121
R. Cuamatzi, I. C. Howard and J. Yates

2T3. Ductile Fracture................................................................................................. 123


Failure Behavior of Hybrid-Laser Welds..................................................................... 123
A. Bajric and W. Dahl
Fracture of Plastic Bodies. Deformations Concentrators............................................. 125
A. I. Khromov, A. A. Bukhanko, S. L. Stepanov and E. P. Kocherov
3D Ductile Tearing Analyses of Bi-Axially Loaded Pipes with Surface Cracks ........ 127
A. Sandvik, E. Ostby and C. Thaulow
New Model Materials for Ductile Fracture Studies .................................................... 129
A. Weck and D. S. Wilkinson
Fatigue Threshold Computation Model Based on the Shakedown Analysis .............. 131
M. A. Belouchrani, D. Weichert and A. Hachemi
Void Coalescence in Metals Involving Two Populations of Cavities .......................... 133
D. Fabregue and T. Pardoen
Contents ix

Effects of Characteristic Material Lengths on Ductile Crack Propagation.................. 135


E. Radi
Ductile Fracture by Void Nucleation at Carbides ........................................................ 137
J. Giovanola, D. Cannizzaro, R. Doglione and A. Rossol
The Significance of Maximum Load on a Load-Displacement Curve with Stable
Crack Extension ........................................................................................................... 139
J. R. Donoso and J. D. Landes
3D Visualization of Ductile Fracture using Synchrotron X-Ray Computer
Tomography ................................................................................................................ 141
L. Qian, H. Toda, T. Ohgaki, K. Uesugi, M. Kobayashi and T. Kobayashi
Non-Local Plastic-Damage Model for Failure Analysis of Sheet-Metals ................... 143
M. Brunet, F. Morestin and H. Walter-Leberre
A Novel Technique for Extracting Stretch Zone Features From Fractographs ........... 145
M. Tarafder, Swati Dey, S. Sivaprasad and S. Tarafder
Simulation of Fatigue Crack Growth by Crack Tip Blunting ...................................... 147
P. Hutar and M. Sauzay
Loading Rate Effect on Ductile Fracture ..................................................................... 149
R. Chaouadi
Experimental Investigation of Slant Crack Propagation in X100 Pipeline Steel......... 151
S. H. Hashemi, I. C. Howard, J. R. Yates, and R. M. Andrews

2T4. Nonlinear Fracture Mechanics......................................................................... 153


Esis TC8 – Numerical Round Robin on Micro Mechanical Models : Results of
Phase III for the Simulation of the Brittle to Ductile Transition Curve....................... 153
C. Poussard and C. Sainte Catherine
Closure of a Rectangular Skin Defect via the Advancement Flap............................... 155
C. Antypas, C. Borboudaki, V. Kefalas and D. A. Eftaxiopoulos
Similarity Solutions of Creep – Damage Coupled Problems in Fracture Mechanics.. 157
L. V. Stepanova and M. E. Fedina
Impact Fracture Toughness Determination of Ductile Polymers by SPB Method ...... 159
J. Wainstein , L. A. Fasce and P. M. Frontini
A Micro-Toughness Model for Ductile Fracture ......................................................... 161
K. Srinivasan, T. Siegmund and O. Kolednik

2T5. Fatigue and Fracture......................................................................................... 163


Crack Coalescence Modelling of FSW Joints.............................................................. 163
A. Ali, M. W. Brown and C. A. Rodopoulos
Fatigue Crack initiation in a Two Phase B-Metastable Titanium Alloy: Influence
of Microstructural Parameters..................................................................................... 165
A. Lenain, P. J. Jacques and T. Pardoen
Effects of Specimen Type, Size and Measurement Techniques on FCGR .................. 167
B. Kumar and J. E. Locke
The Effect of Stress Ratio on Fatigue Short Cracking ................................................. 169
C. A. Rodopoulos and S.-H. Han
Dwell-fatigue Behaviour of a Beta-Forged Ti 6242 Alloy .......................................... 171
P. Lefranc, C. Sarrazin-Baudoux and V. Doquet
Investigation into Fatigue Life of Welded Chemical Pipelines ................................... 173
x Contents

Cz. Goss and L. Sniezek


Different Analytical Presentations of Short Crack Growth under Rotation-Bending
Fatigue ......................................................................................................................... 175
D. Angelova and A. Davidkov
Variable Amplitude Load Interaction in Fatigue Crack Growth for 2024-T3
Aluminium Alloy ........................................................................................................ 177
D. Kocanda, S. Kocanda and J. Torzewski
An Investigation on the Fatigue Performance of Hydraulic Gate Wheels................... 179
D. Polyzois and A. N. Lashari
A Micromechanical Model for Crack Initiation in High Cycle Fatigue of Metallic
Materials ...................................................................................................................... 181
V. Monchiet, E. Charkaluk and D. Kondo
Comparative Analysis of Two Models for Evaluating Fatigue Data ........................... 183
E. Castillo, A. Ramos, M. Lopez-Aenlle, A. Fernandez-Canteli and R. Koller
Assessment of Damage at Notch Root of Thick Plates ............................................... 185
E. C. G. Menin and J. L. de A. Ferreira
Fatigue Strength Prediction of Spot-Welded Joints Using Small Specimen Testing... 187
E. Nakayama, M. Fukumoto, M. Miyahara, K. Okamura, H. Fujimoto
and K. Fukui
A Thermo-Mechanical Model for Random Braking of Machine Components ........... 189
F. Loibnegger, H. P. Rossmanith and R. Huber
Lifetime Calculation of Railway Wheel Steels Based on Physical Data ..................... 191
F. Walther and D. Eifler
Fatigue Crack Propagation of Super-Duplex Stainless Steel at Different
Temperatures ................................................................................................................ 193
G. Chai and S. Johansson
Transitions of Fatigue Crack Initiation From Surface, Subsurface to SNDFCO......... 195
G. Chai
Surface Fatigue of Gear Teeth Flanks.......................................................................... 197
G. Fajdiga, M. Sraml and J. Flasker
Fatigue and Fracture Processes in High Performance PM Tool Steels ........................ 199
G. Jesner, S. Marsoner, I. Schemmel and R. Pippan
Notch and Defect Sensitivity of ADI in Torsional Fatigue.......................................... 201
B. Atzori and G. Meneghetti
Multi Axial Fatigue in Welded Components ............................................................... 203
G. Mesmacque, B. Wu, C. Robin, D. Zakrzewski and X. Decoopman
Enhanced Fatigue Life by Mechanical Surface Treatments – Experiment and
Simulation .................................................................................................................... 205
H. P. Gaenser, I. Goedor, H. Leitner and W. Eichlseder
Analysis of Repaired Aluminum Panels in General Mixed-Mode Conditions............ 207
H. Hosseini-Toudeshky, M. Saber and B. Mohammadi
Effect of Strain Rate on Fatigue Behavior of Ultrafine Grained Copper..................... 209
P. Gabor, H. J. Maier and I. Karaman
Lubricant Effects on Propagation of Surface-Breaking Cracks under Rolling
Contact Loading........................................................................................................... 211
J. Lai and S. Ioannides
Contents xi

Computational Modelling of Crack Initiation in a Mixing Tee Subjected to Thermal


Fatigue Load ................................................................................................................ 213
I. Varfolomeyev
Estimation of Critical Stress Intensity Factor in Steel Cracked Wires......................... 215
J. Toribio, F. J. Ayaso, B. Gonzalez, J. C. Matos and D. Vergara
Low-Cycle Fatigue of Din 1.2367 Steels in Various Treatments................................. 217
C. C. Liu, J. H. Wu and C. C. Kuo
Impact Testing a Capable Method to Investigate the Fatigue Resistance.................... 219
K. David, P. Agrianidis, K. G. Anthymidis and D. N. Tsipas
Comparative Assessment of Fatigue-Thresholds Estimated by Short and
Long Cracks ................................................................................................................. 221
K. K. Ray, N. Narasaiahb and S. Tarafderb
Scanning Electron Microscope Measurements of Crack-Opening Stress on Fatigue
Cracks Exposed to Overloads ...................................................................................... 223
L. Jacobsson and C. Persson
Propagation Path and Fatigue Life Predictions of Branched Cracks under Plane
Strain Conditions.......................................................................................................... 225
M. A. Meggiolaro, A. C. O. Miranda, J. T. P. Castro and L. F. Martha
Short Crack Equations to Predict Stress Gradient Effects in Fatigue .......................... 227
M. A. Meggiolaro, A. C. O. Miranda, J. T. P. Castro and J. L. F. Freire
Fatigue Behaviour of Pre-Strained Type 316 Stainless Steel...................................... 229
M. Akita, M. Nakajima, K. Tokaji and Y. Uematsu
The Influence of Constraint on Fitting Fatigue Crack Growth Data ........................... 231
M. Carboni and M. Madia
Atomic Force Microscopy of Local Plastic Deformation for Tempered Martensite ... 233
M. Hayakawa, S. Matsuoka and Y. Furuya
Improvement of Fatigue Strength due to Grain Refinement in Magnesium Alloys .... 235
M. Kamakura, K. Tokaji, H. Shibata and N. Bekku
A Unified Fatigue and Fracture Model Applied to Steel Wire Ropes ......................... 237
M. P. Weiss, R. Ashkenazi and D. Elata
Correlation Between Paris’ Law Parameters Based on Self-Similarity and Criticality
Condition...................................................................................................................... 239
A. Carpinteri and M. Paggi
Thermo-Mechanical Fatigue Lifetime Assessment with Damage-Parameters,
Energy-Criterions and Cyclic-J-Integral Concepts ...................................................... 241
M. Riedler, R. Minichmayr, G. Winter and W. Eichlseder
Predicting Fatigue Crack Retardation Following Overload Cycles............................. 243
M. V. Pereira, F. A. Darwish, A. F. Camarao and S. H. Motta
Fatigue Crack Growth at Notches Considering Plasticity Induced Closure ................ 245
J. Bruening, O. Hertel, M. Vormwald and G. Savaidis
Influence of Microstructure on Fatigue Properties of Ni-Base Superalloy at
Elevated Temperature .................................................................................................. 247
Qy. Wang , Y. Matsuyama, N. Kawagoishi, M. Goto and K. Morino
Moddelling Fatigue Crack Closure using Dislocation Dipoles ................................... 249
P. F. P. de Matos and D. Nowell
Comparison Between Fatigue Crack Growth Modelled by Continous Dislocation
Distributions and Discrete Dislocations ...................................................................... 251
xii Contents

P. Hansson, S. Melin and C. Persson


Fatigue Evaluation Considering the Environmental Influence Using a Monitoring
System.......................................................................................................................... 253
R. Cicero, I. Gorrochategui and J. A. Alvarez
Thermal Fatigue Crack Initiation and Propagation Behavior of Steels for Boiler....... 255
S. Aoi, T. Marumiya, R. Ebara , T. Nishimura and Y. Tokunaga
Recent Developments in Fatigue Crack Growth ......................................................... 257
R. Jones , S. Pitt, and E. Siores
Crack Closure Effects in a Cracked Cylinder under Pressure...................................... 259
J. Zhao, R. Liu, T. Zhang and X. J. Wu
An Experimental Study of Tearing-Fatigue Interaction............................................... 261
P. Birkett, M. Lynch and P. Budden
Sif Solutions for Cracks in Railway Axles under Rotating Bending........................... 263
S. Beretta, M. Madia, M. Schode and U. Zerbst
Mechanical Characterization of Single Crystal Bars with Capacitor Discharge
Welding and Laser Cladding........................................................................................ 265
S. Chiozzi, V. Dattoma and F.o W. Panella
Fractal Dimension Analysis of Fracture Toughness Used High Strength Cast Iron.... 267
S. Doi and M. Yasuoka
Investigating Gap Effects in Fatigue Life of Spot Welded Joints ................................ 269
M. Zehsaz and S. Hasanifard
Fatigue of Pmma Bone Cement ................................................................................... 271
S. L. Evans
Influence on Thermal Barrier Coating Delamination Behaviour of Edge Geometry .. 273
H. Brodin, X. H. Li and S. Sjoestroem
Low Cycle Fatigue and Fracture of a Coated Superalloy CMSX-4 ............................ 275
S. Stekovic
Thermomechanical Fatigue of Open-Cell Aluminium Sponge ................................... 277
T. Guillen, A. Ohrndorf, U. Krupp, H. J. Christ, S. Derimay, J Hohe
and W Becker
The Influence of Alternate Block Loading on the Fatigue Lifetime............................ 279
M. Kohut and T. Lagoda
Fatigue Design and Inspection Planning of Welded Joints Based on Refined
Physical Modelling ..................................................................................................... 281
T. Lassen and N. Recho
A Mixed Mode Fatigue Crack Growth Model Including the Residual Stress Effect
Due to Weld.................................................................................................................. 283
S. Ma, X. B. Zhang, N. Recho and J. Li
Effects of Shot Peening on Fatigue Property in SICP/Al-MMC ................................. 285
Y. Ochi, K. Masaki , T. Matsumura and T. Hamaguchi
Fatigue Behaviour of Friction Stir Welded 6061-T6 Aluminium Alloy...................... 287
Y. Uematsu, K. Tokaji, Y. Tozaki and H. Shibata
Transformation of a Nonproportional Multiaxial Loading to an Equivalent
Proportional Multiaxial Loading.................................................................................. 289
A. Chamat, Z. Azari, M. Abbadi and F. Cocheteux

2T8. Polymers and Composites................................................................................. 291


Contents xiii

Acoustic Emission Monitoring of Delamination Growth in Fiber-Reinforced


Polymer-Matrix Composites ........................................................................................ 291
A. J. Brunner and M. Barbezat
Fracture Mechanics Versus Strength Concepts for Evaluation of Adhesion Quality 293
B. Lauke
Alternative Approaches for the Evaluation of the Slow Crack Growth Resistance
of Polyethylene Resins Used in the Production of Extruded Water Pipes................... 295
F. M. Peres and C. G. Schon
A Stereoscopic Method for Fractographic Investigations of Ordinary Ceramics........ 297
C. Manhart and H. Harmuth
Modelings of Fiber Deformation During Machining Aramid-FRP ............................. 299
E. Nakanishi, M. Fukumori, Y. Sawaki and K. Isogimi
Quality Control and the Strength of Glass ................................................................... 301
F. Veer, C. Louter and T. Romein
Experimental Study of Cracked Laminate Plates by Caustics ..................................... 303
G. A. Papadopoulos and E. Sideridis
Fracture of Composites in Military Aircraft ................................................................ 305
R. Pell, N. Athiniotis and G. Clark
Analysis of 7005/AL2O3/10P MMC Sheets Joined by FSW by Thermoelasticity..... 307
P. Cavaliere, G. L. Rossi, R. di Sante and M. Moretti
Surface Modification of Lightweight Aggregate and Properties of the Lightweight
Aggregate Concrete...................................................................................................... 309
T. Y. Lo and H. Z. Cui
Finite Element Based Prediction of Failure in Laminated Composite Plates .............. 311
H. Hosseini-Toudeshky, B. Hamidi, B. Mohammadi and H. R. Ovesi
An Embedded Cylindrical PZT with Electroded Imperfect Interface ......................... 313
H. M. Shodja and S. M. Tabatabaei
Characterization of Composites for the Maeslant Storm Surge Barrier ...................... 315
J. Degrieck, W. van Paepegem, L. van Schepdael, P. Samyn, P. de Baets,
E. Suister and J. S. Leendertzc
Weight Function, J-Integral and Material Forces Approach to Ceramic Multilayers.. 317
J. Pascual, C. R. Chen, O. Kolednik, F. D. Fischer, R. Danzer and T. Lube
Assessment of Matrix Fatigue Damage in CFRP ........................................................ 319
K. J. Cain and A. Plumtree
Progressive Failure of Composite Materials under Dynamic Loading........................ 321
L. Xing, X. Huang and K. Reifsnider
Aging Aircraft Transparencies: A Case History from Italian Air Force Fleet............. 323
C. M. Bernabei, D. Caucci and C. L. Aiello
Fatigue Crack Growth in Quenched Amorphous Polymers PC and PET.................... 325
M. Kitagawa and D. Nishi
Thermo-Mechanical State of Bimaterial with an Interface Crack ............................. 327
R. Martynyak, M. Matczynski and K. Honchar
Thermo-Mechanical State of Bimaterial with an Interface Crack ............................... 329
M. Tarfaoui, S. Choukri, A. Neme and M. Mliha-Touati
Indentation Response of Fibre Reinforced Composite Laminates............................... 331
P. Bourke and I. Horsfall
Analysis of Tubular Composite Cylindrical Shells...................................................... 333
R. M. Gheshlaghi and M. H. Hojjati and H. R. M. Daniali
xiv Contents

Analysis of Composite Pressure Vessels...................................................................... 335


R. M. Gheshlaghi, M. Hassan H. and H. R. M. Daniali
Gradients Influence on Damage and Cracking in Crystalline Polymers ..................... 337
S. Castagnet and J.-C. Grandidier
An Elasto-Plastic Shear-Lag Model for Single Fiber Composite ................................ 339
S. Kimura, J. Koyanagi and H. Kawada
Progressive Failure of Composite Laminates; Analysis vs Experiments .................... 341
V. Skytta, O. Saarela and M. Wallin
A Temperature Dependent Viscoelastic-Damage Model for Ceramics Failure........... 343
V. P. Panoskaltsis, L. M. Powers and D. A. Gasparini

2T11. Fracture Mechanics Analysis ......................................................................... 345


Avalanche Mechanics: Lefm vs. Gradient Model........................................................ 345
A. Konstantinidis, N. Pugno, P. Cornetti and E. C. Aifantis
Influence of Austempering on Fracture Mechanics Parameters of 65 Si 7 Steel......... 347
D. Pustai, F. Cajner and M. Lovreni
Modelling the Evolution of Elastic Symmetries of Growing Mixed-Mode Cracks .... 349
H. Schutte and K. M. Abbasi
Effect of Aging on the Microstructure and Fracture of Aluminum-Lithium ............... 351
J. M. Fragomeni
Buckling of Multicracked Columns............................................................................. 353
C. Carloni, C. Gentilini and L. Nobile
Experimental and Numerical Analysis of Interactions Between Stress
Corrosion Cracks.......................................................................................................... 355
M. Lamazouade, M. Touzet and M. Puiggali
An Improved Upper Bound Limit Load Solution for Weld Strength anisotropic
Overmatched Cracked Plates in Pure Bending ............................................................ 357
N. Kontchakova and S. Alexandrov
Fracture Parameter Estimation of Alloy Steel Reinforced with Maraging Steel ......... 359
S. Bhat, V. G. Ukadgaonker, M Jha and S. M. Nirgude
Incorporation of Length Scales in Plane Stress Fracture Analysis .............................. 361
V. P. Naumenko
Mode III Crack in a Functionally Graded Piezoelectric/Piezomagnetic Half Plane.... 363
W.-H. Hsu and C.-H. Chue
Electro-Mechanical Field of a Piezoelectric Finite Wedge under Antiplane Loading. 365
W.-J. Liu and C.-H. Chue
Sensitivity of Crack Nucleation Parameters to the Geometric Imperfection............... 367
V. P. Naumenko and Yu. D. Skrypnyk

2T13. Probabilistic approaches to Fracture Mechanics ......................................... 369


An Experimental Evaluation of a Local Approach Model for Graded Materials........ 369
B. Bezensek, J. Flasker and J. W. Hancock
A Stochastic Model for Crack Growth......................................................................... 371
C.-R. Chiang
Stochastic Evaluation of Fatigue Crack Initiation and Propagation ............................ 373
G. S. Wang
Contents xv

A Weibull-Based Method to Predict the Strength of Adhesively Bonded Joints of


Pultruded FRPS............................................................................................................ 375
T. Vallée, J. R. Correia and T. Keller

2T14. Computational Fracture Mechanics.............................................................. 377


The Lateral Constraint Index as a New Factor to Assess the Influence of the
Specimen Thickness..................................................................................................... 377
A. Fernandez-Canteli, D. Fernandez-Zuniga and E. Castillo
Analysis of Crack Propagation in Alumina-Glass Functionally Graded Materials ..... 379
V. Cannillo, L. Lusvarghi, T. Manfredini, M. Montorsi, C. Siligardi
and A. Sola
Numerical Solution of Integro-Differential Equations for Fracture Mechanics
Problems....................................................................................................................... 381
A. V. Andreev
Analytical Method of Generating DA/DN Curve for Aerospace Alloys..................... 383
B. Farahmand
Thermo-Elastic Fracture of Edge Cracked Plate under Surface ‘Shock’ Loading ...... 385
B. P. Fillery, X. Hu and G. Fisher
Failure Prediction of IC Interconnect Structures Using Cohesive Zone Modelling .... 387
B. A. E. van Hal, R. H. J. Peerlings, M. G. D. Geers and G. Q. Zhang
Non-Local Damage Simulation in Composites Using Crack Propagation and Mesh
Adaptivity..................................................................................................................... 389
F. Reusch, C. Hortig and B. Svendsen
Elastic Wave Motion in a Cracked, Multi-Layered Geological Region under
Transient Conditions .................................................................................................... 391
P. S. Dineva, T. V. Rangelov and G. D. Manolis
Wood Beam Strengthened with Glass/Epoxy Composite Sheets ............................... 393
G. E. Papakaliatakis, G-S. P. Diamantopoulos, P.A. Kalaitzidis
and E. M. Marinakis
Computation of Dynamic Stress Intensity Factors Using Enriched Finite Elements .. 395
M. Saribay and H. F. Nied
Partly Cracked Xfem Interface for Intersecting Cracks............................................... 397
J. L. Asferg, T. Belytschko, P. N. Poulsen and L. O. Nielsen
On the Evaluation of Elastic Compliance Tensor Due to Growing Mixed-Mode
Microcracks.................................................................................................................. 399
K. M. Abbasi and H. Schutte
On the Problem of Determination of Safety Factors for Machine-Building Parts
Using the Finite Element Computations ...................................................................... 401
L. B. Getsov, B. Z. Margolin and D. G. Fedorchenko
Dynamic Explicit Cell Model Simulations in Porous Ductile Metals ......................... 403
L. Siad and M. O. Ouali
Numerical Evaluation of Energy Release Rates for Bimaterials Interface Cracks...... 405
M. Belhouari, B. B. Bouiadjra, B. Boutabout and K. Kaddouri
Inclusion Effect on the Plastic Zone Size in Confined Plasticity................................. 407
M. Benguediab, M. Elmegueni, M. Nait-Abdelaziz and A. Imad
Modified Key-Curve-Method for Determination of Dynamic Crack Resistance
Curves .......................................................................................................................... 409
xvi Contents

U. Muhlich, A. Emrich and M. Kuna


A Coupled Computational Framework for Ductile Damage and Fracture .................. 411
R. H. J. Peerlings, J. Mediavilla and M. G. D. Geers
Marble Discs under Distributed Loading: Theoretical, Numerical and Experimental
Study ........................................................................................................................... 413
Ch. Markides, E. Sarris, D. N. Pazis, Z. Agioutantis and S. K. Kourkoulis
Simulation of the Mechanical Behaviour of the Lumbar Intervertebral Disc.............. 415
M. Satraki, E. A. Magnissalis, G. Ferentinos and S. K. Kourkoulis
The Pull-Out Strength of Transpedicular Screws in Posterior Spinal Fusion.............. 417
P. Chazistergos, G. Ferentinos, E. A. Magnissalis and S. K. Kourkoulis
Mechanical Behavior Simulation of Hip Prostheses Stress Distributions Analysis .... 419
M. Kadi, R. Boulahia, K. Azouaoui, N. Ouali, A. Ahmed-Benyahia
and T. Boukharouba
Dbem Analysis of Axisymmetric Crack Growth in a Piston Crown .......................... 421
T. Lucht
Residual Shear Stresses and KII Computation ............................................................. 423
W. Cheng and I. Finnie

2T15. Experimental Fracture Mechanics ................................................................ 425


Quantitative Interpretation of Crack Tip Strain Field Measurements.......................... 425
A. M. Korsunsky
Mixed Mode (I+II) Stress Intensity Factor Measurement Using Image Correlation... 427
A. Shterenlikht, P. López-Crespo, P. J. Withers, J. R. Yates
and E. A. Patterson
Fracture of Turbine Blades under Self-Exciting Modes .............................................. 429
C. A. Sciammarella, C. Casavola, L. Lamberti and C. Pappalettere
Predicting Crack Arrest Behaviour of Structural Steels Using New Procedures......... 431
C. Gallo, J. A. Alvarez, F. Gutierrez-Solana and J. A. Polanco
Mechanical Properties of Large Plastic-Mold Steel Blooms. ...................................... 433
M. Chiarbonello, D. Firrao, R. Gerosa, A. Ghidini, M. G. Ienco, P. Matteis,
G. Mortarino, A. Parodi, M. R. Pinasco, B. Rivolta, G. Scavino, G. Silva,
E. Stagno and G. Ubertalli
Non-Linear Photoelastic Method for Study Fracture Problems................................... 435
G. Albaut
Fatigue Crack Length Measurement Method with an Ion Sputtered Film .................. 437
G. Deng, K. Nasu, T. D. Redda and T. Nakanishi
Individual Fracture Events in Cellular Foods .............................................................. 439
H. Luyten, E. M. Castro-Prada, E. Timmerman, W. Lichtendonk
and T. van Vliet
Exfoliation Fracture Mode in Heavily Drawn Pearlitic Steels..................................... 441
J. Toribio and F. J. Ayaso
Investigation of Crack Closure by Using Thermoeastic Stress Analysis..................... 443
L. Marsavina, R. A. Tomlinson, E. A. Patterson and J. R. Yates
Fracture Toughness Investigations of Severe Plastic Deformed Tungsten Alloys ...... 445
M. Faleschini, W. Knabl and R. Pippan
Photoelastic Analysis of Mode I Stress Intensity Factor in Beams with Angular
Notches......................................................................................................................... 447
M. Tabanyukhova and V. Pangaev
Contents xvii

Full-Scale Experimental Investigations on Pressure Tubes Rupture of RBMK .......... 449


N. Yu. Medvedeva, A. V. Andreev, S. V. Timkin, I. A. Peshkov, V. N. Zhilko,
D. Ye. Martsiniouk and O. A. Poshtovaya
Study of Fracture Mechanism of Composite Material Buildings by Photoelasticity
and Photoelasitc Coating Methods............................................................................... 451
O. Ivanova, G. Albaut, V. Mitasov, V. Nikiforovskij and M. Tabanyukhova
Fracture Energy in Mode I and Mode II of Textile Reinforced Wood......................... 453
R. Putzger and P. Haller
Measurement Based Performance Prediction of the Europabrucke Against Traffic
Loading ....................................................................................................................... 455
R. Veit and H. Wenzel
The Effect of the Laboratory Specimen on Fatigue Crack Growth Rate..................... 457
S. C. Forth, W. M. Johnston and B. R. Seshadri
Validity of the Caustics Method for Plates with Circular Hole.................................... 459
P. Tsirigas, G. Kontos, D. N. Pazis, S. K. Kourkoulis and Z. Agioutantis
An Enhanced Normalization Method for Dynamic Fracture Toughness Testing ........ 461
S. M. Graham and D. J. Stiles
The Potential Drop Technique for Measuring Crack Growth in Shear........................ 463
V. Spitas and P. Michelis
A Modified DCB Geometry for CTOA Measurement in Thin Sheet 2024-T3
Aluminium Alloy ........................................................................................................ 465
Y. H. Tai, S. H. Hashemi, R. Gay, I. C. Howard and J. R. Yates
Could Cod Serve as Fracture Criterion in Case of Marble? ....................................... 467
A. Marinelli, S. K. Kourkoulis and I. Vayas

2T16. Creep Fracture ................................................................................................ 469


Creep Rupture of a Lead-Free Sn-Ag-Cu Solder......................................................... 469
C.-K. Lin and D.-Y. Chu
Quantitative Evaluation of Acceleration Creep in Magnesium-Aluminum
Alloys at 0.65tm........................................................................................................... 471
H. Sato
Long-Term Creep Rupture Prediction in Unidirectional Composites ........................ 473
J. Koyanagi, F. Ogawa and H. Kawada
A Computational Model for Cardboard Creep Fracture ............................................. 475
J. Schonwalder, G. P. A. G. van Zijl and J. G. Rots
Creep Fracture of Binary and Ternary Commercial Aluminum Alloys....................... 477
K. Ishikawa
Analysis of Creep Crack initiation and Growth in Laboratory Specimens.................. 479
K. Wasmer
Temperature Gradient Effects on the Creep Behaviour of Structures.......................... 481
F. Vakili-Tahami and S. Hasanifard

2T17. Environment Assisted Fracture ..................................................................... 483


A Surgical Implant Crevice-Assisted Corrosion Fatigue In-Body Failure .................. 483
H. Amel-Farzad, M.-T. Peivandi and S. M.-R. Yosof-Sani
Asymptotically Stable Growth of Delaminations under Hydrogen Embrittlement
Conditions .................................................................................................................... 485
xviii Contents

A. V. Balueva
Corrosion and Mechanical Strength of Russian Light Water Reactors ....................... 487
B. T. Timofeev
Corrosion Fatigue Characteristics of CF8A Steel Degraded at High Temperature ..... 489
S.-C. Jang, D.-H. Bae, G.-Y. Lee, and S.-Y. Baek
Modeling Environment-Assisted Fatigue Crack Propagation ..................................... 491
J.-A. Ruiz-Sabariego and S. Pommier

2T18. Dynamic, High Strain Rate, or Impact Fracture ......................................... 493


Measuring the Fracture Resistance of Composites and Adhesively Bonded Joints
at High Test Rates. ....................................................................................................... 493
B. R. K. Blackman, D. D. R. Cartie, A. J. Kinloch, F. S. Rodriquez-Sanchez
and W. S. Teo
Quasistatic and Dynamic Fracture of Pearlitic Steel.................................................... 495
B. Strnadel, P. Hausild and M. Karlik
Fragmentation in the Expanding Ring Experiment...................................................... 497
H. Zhang and K. Ravi-Chandar
Influence of Friction on Results of an Instrumented Impact Test ............................... 499
I. V. Rokach
Influence of Moisture Content on the Dynamic Behaviour of Concrete ..................... 501
I. Vegt and J. Weerheijm
Strength and Toughness Properties of Steels under Dynamic Loading ....................... 503
J. Fang
Rubber Particle Size Effect on Impact Characteristics of PC/ABS (50/50) Blends .... 505
M. Nizar Machmud, Masaki Omiya, Hirotsugu Inoue and Kikuo Kishimoto
Effect of Strain Rate on Mechanical Properties of Reinforced Polyolefins................. 507
M. Schossig, C. Bieroegel, W. Grellmann, R. Bardenheier
and T. Mecklenburg
Fracture Related Mechanical Properties of Aircraft Cast Aluminum Alloy A357...... 509
N. D. Alexopoulos
Shear Failure of TI-6AL-4V by Direct Impact and analyse of the Process of
Elastic and Plastic Wave Propagation ......................................................................... 511
P. Chwalik, A. Rusinek and J. R. Klepaczko
Evaluating of Fracture Mechanics Properties at Intermediate Strain Rates,
Transferable to Components ........................................................................................ 513
P. Trubitz, A. Ludwig, G. Pusch and H.-P. Winkler
Crack Resistance Determination From the Charpy Impact Test.................................. 515
R. Chaouadi
A Stochastic Interface Model for the Fracture of Bars ................................................ 517
S. Nagy and F. Kun
The Anti-Penetration Properties of Space Armor ....................................................... 519
Tso-Liang Teng, Cho-Chung Liang and Cheng-Chung Lu
Key Curve Methods for Dynamic Fracture Mechanics of Cast Iron ........................... 521
W. Baer
Dynamic Tensile Behavior of Aramid Frp Using Split Hopkinson Bar Method......... 523
Y. Sawaki, J. Watanabe, E. Nakanishi and K. Isogimi

2T19. Damage Mechanics.......................................................................................... 525


Contents xix

Detection of Low-Velocity Impact Damage in Carbon-Epoxy Plates using NDT ..... 525
A. M. Amaro, M. F. M. S. de Moura and P. N. B. Reis
Damage Accumulation at High Temperature Creep of a Single-Crystal Superalloy... 527
A. Staroselsky and B. Cassenti
Asymptotic Homogenisation for Heterogeneous Media with Evolving Microcracks 529
E. K. Agiasofitou, C. Dascalu and J. L. Auriault
On the Analysis of Damage Localization as Precursor of Macro-Cracks ................... 531
H. Stumpf and K. Hackl
Fatigue Assessment Based on Statistical Analysis of Theoretical Parameters ............ 533
J. Cacko
Determination of Ductile Damage Parameters by Local Deformation Fields ............. 535
M. Kuna and M. Springmann
Fracture of Concrete Due to Corrosion........................................................................ 537
N. Thanh, A. Millard, Y. Berthaud, S. Care and V. L’Hostis

2T21. Concrete and Rock .......................................................................................... 539


Experimental Study of Sprayed Concrete Strength Using Marble Aggregates ........... 539
A. Sotiropoulou and Z. G. Pandermarakis
Analysis of the Behaviour of Interface Cracks in Gravity Dam ................................. 541
B. B. Bouiadjra, A. B. Bouiadjra, M. Belhouari and B. Serier
Application of Composite Mechanics to Composites Enhanced Concrete Structures 543
C. C. Chamis and P. K. Gotsis
Initiation and Coalescence of Locals Damages on Blanco de Macael Marble............ 545
K. Mehiri, P. Vieville, P. Lipinski, A. Tidu and V. Tijeras
Influence of Concrete´s Mineralogical Components on Fracture Compressive and
Tractive ........................................................................................................................ 547
M. P. Morales Alfaro and F. A. I. Darwish
Constitutive Model for Description of High-strain Rate Behavior of Concrete ......... 549
I. R. Ionescu and O. Cazacu
Hydraylic Fracturing in Weak Rocks .......................................................................... 551
P. Papanastasiou
Application of Fracture Mechanics on Unreinforced Concrete Walls ......................... 553
T. Eck, B.-Gu Kang and W. Brameshuber
Subcritical Crack Growth in Rocks under Water Environment ................................... 555
Y. Nara, H. Kurata and K. Kaneko

2T22. Sandwich Structures ....................................................................................... 557


Stress Analysis and Prediction of Failure in Structurally Graded Sandwich Panels ... 557
A. Lyckegaard, E. Bozhevolnaya and O. T. Thomsen
Debonding and Kinking in Foam-Core Sandwich Beams ........................................... 559
D. A. Zacharopoulos, V. D. Balopoulos, Z. S. Metaxa, P. A. Kalaitzidis
and E. E. Gdoutos
Modeling Core Failure by the Tsai–Wu Criterion in the Design of Foam-Core
Sandwich Beams .......................................................................................................... 561
E. E. Gdoutos, V. D. Balopoulos, P. A. Kalaitzidis and M. Konsta
Numerical Investigation of Crack Propagation in Sandwich Structures...................... 563
E. E. Theotokoglou
xx Contents

Local Effects in Sandwich Beams: Modelling and Experimental Investigation.......... 565


M. Johannes, J. Jakobsen, V. Skvortsov, E. Bozhevolnaya
and O. T. Thomsen
Typical In-Plane Response Surfaces for Prismatic Foam-Core Sandwich Beams ...... 567
V. D. Balopoulos, P. A. Kalaitzidis, D. A. Zacharopoulos
and E. E. Gdoutos

2T23. Novel Testing and Evaluation Techniques.................................................... 569


Non-Destructive Evaluation of Yield Strength Using a Novel Miniature Dumb-Bell
Specimen-An Empirical Approach ............................................................................. 569
G. Partheepan, D. K. Sehgal and R. K. Pandey
3D Measurement of the Strain Field Surrounding Crack Tip ...................................... 571
D. Vavrik, J. Bryscejn, J. Jakubek and J. Valach
Radiographic Observation of Damage Zone Evolution in High Ductile Specimen .... 573
D. Vavrik, T. Holy, J. Jakubek, M. Jakubek and Z. Vykydal
Calibration of Fracture Parameters by Instrumented Indentation and Test
Simulation .................................................................................................................... 575
M. Bocciarelli, G. Bolzon and G. Maier
Internal Crack Detection and Analysis Using Thermoelastic Stress Analysis ............ 577
N. Sathon and J. M. Dulieu-Barton
Ultrahigh-Resolution Transversal Polarization-Sensitive Optical Coherence
Tomography: Structural Analysis and Strain-Mapping ............................................... 579
K. Wiesauer, M. Pircher, R. Engelke, G. Ahrens, G. Grutzner, R. Oster,
C. K. Hitzenberger and D. Stifter
Application of Digital Shearography in Determining Opening Mode SIF in
Edge Cracks ................................................................................................................. 581
M. Ghassemieh, A. Ghazavizadeh and N. Soltani
Finite Element Modeling of Pulse Transient IR Thermography.................................. 583
M. Krishnapillai, R. Jones, I. H. Marshall, M. Bannister and N. Rajic
A New Technique for the Machining of Natural Cracks ............................................. 585
N. P. Andrianopoulos and A. Pikrakis
Displacements Measurement in Irregularly Bounded Plates Using Mesh Free
Methods........................................................................................................................ 587
N. P. Andrianopoulos and A. P. Iliopoulos
Biaxial Strength Testing on Mini Specimens............................................................... 589
R. Danzer, P. Supancic, W. Harrer, T. Lube and A. Borger
Numerical Simulation of a Fracture Test for Brittle Disordered Materials ................. 591
T. Auer and H. Harmuth

2T26. Structural Integrity ......................................................................................... 593


Unification of the Out-of-Plane Constraint Loss in Centre-Cracked Panels ............... 593
B. Bezensek, A. Baron and J. W. Hancock
High Temperature Failure Assessment of Weldments ................................................ 595
B. Dogan, B. Petrovski and U. Ceyhan
Post-Tensioned Glass Beams ....................................................................................... 597
C. Louter, J. van Heusden, F. Veer, J. Vambersky, H. de Boer
and J. Versteegen
Contents xxi

Structural Integrity of a NPP Using the Master Curve Approach................................ 599


D. Ferreno, I. Gorrochategui, M. Scibetta, R. Lacalle, E. van Walle
and F. Gutierrez-Solana
FRP Consolidation for Masonry Arches by Using Bridged Crack Model .................. 601
G. Ferro, M. Ipperico, V. Pignata and A. Carpinteri
Structural Reliability Analysis of Pipe Subjected to Reeling ...................................... 603
H. A. Ernst, R. E. Bravo and F. Daguerre
Network Seismic Capability Assessment of Power High Voltage
Electric Equipment....................................................................................................... 605
I. Manea, C. Diaconu, C. Radu and M. Negru
FKM Guideline “Fracture Mechanics Proof of Strength for Engineering
Components” – Overview and Extension topics.......................................................... 607
B. Pyttel, I. Varfolomeyev and M. Luke
Static and Dynamic Behavior of a 3D-Periodic Structure ........................................... 609
J. Rishmany, L. Renault, C. Mabru, R. Chieragatti and F. Rezaï Aria
Environmental Effect on Pipeline Steels: A Fitness for Service Perspective .............. 611
J. A. Alvarez, F. Gutierrez-Solana and S. Cicero
Finding the Australian Railway Load Spectrum Design and Assessment of Light
Weight & Durable Railway Structural Components .................................................... 613
R. Jones and J. Baker
Structural Integrity Assessment of Componets with Low Constraint.......................... 615
S. Cicero, F. Gutierrez-Solana and J. A. Alvarez
Life Assessment of Superheater Tubes Fabricated From 2.25CR-1MO Steel ............ 617
S. Fujibayashi
Predicting Cleavage Fracture in Presence of Residual Stresses; A Numerical
Case Study.................................................................................................................... 619
S. Hadidi-Moud, C. E. Truman and D. J. Smith
A Necessary Condition for Cleavage on Laboratory Specimens and Structures......... 621
V. le Corre, S. Chapuliot, S. Degallaix and A. Fissolo
Safety Assessment of Components with Crack-Like Defects .................................... 623
Yu. G. Matvienko and O. A. Priymak
Numerical Analysis of Surface Cracks in Steam Generator Tubes ............................. 625
Z. Tonkovi, I. Skozrit and J. Sori

2T28. Mesofracture Mechanics................................................................................. 627


Tensile Simulation of Polymeric Material Considering the Meso-Scale Structure .... 627
A. Shinozaki, K. Kishimoto and I. Hirotugu
Microfracture and Strain Localization: A Computational Homogenization Approach 629
C. Dascalu, G. Bilbie and R. Chambon
Strain and Fracture at Mesoscale of Coated Materials ................................................ 631
S. Panin

2T32. Micromechanisms in Fracture and Fatigue.................................................. 633


Relating Cleavage Crack Nucleation to Cracked Carbides in A533B Steel................ 633
A. Kumar and S. G. Roberts
xxii Contents

Micro-Energy Rates for Damage Tolerance and Durability of Composite Structures 635
C. C. Chamis and L. Minnetyan
Micromechanical Observation of Fracture Process in Mortars ................................... 637
E. Schlangen and O. Copuroglu
Micro-fracture Maps in Progressively Drawn Pearlitic Steels .................................... 639
J. Toribio and F. J. Ayaso
A Brief History of Fractography.................................................................................. 641
S. P. Lynch and S. Moutsos

C. SPECIAL SYMPOSIA/SESSIONS
C1. Nanomaterials and Nanostructures ..................................................... 643

1. Fracture and Fatigue at the Micro and Nano Scales .......................................... 645
Size Effects in Lead Free Solder-joints........................................................................ 645
A. Betzwar-Kotas, G. Khatibi, A. Ziering, P. Zimprich, V. Groeger,
B. Weiss and H. Ipser
Micro-Scale Simulation of Impact Rupture in Polysilicon MEMS ............................. 647
A. Corigliano, F. Cacchione, A. Frangi and B. de Masi
Nanoindentation of CNT Reinforced Epoxy Nanocomposites.................................... 649
D. C. Lagoudas, P. R. Thakre and A. A. Benzerga
Diffusion Kinetics and Multivariant Phase Transformation in Shape Memory Alloys 651
D. R. Mahapatra and R. V. N. Melnik

3. Nanoscale Deformation and Failure..................................................................... 653


EBSD Analysis on Deformation of Nanocrystals in ECAP-Processed Copper .......... 653
H. Kimura, Y. Akiniwa, K. Tanaka and T. Ishida
The Effect of Extensional Strains on Molecular Orientation, Polymer Free Volume
Distribution and Crystallization ................................................................................... 655
H. Dong, R. Guo and K. I. Jacob
Microrotation-augmented Energy-Minimization for 3D Nanocrystalline Cu
Structures ..................................................................................................................... 657
M. A. Tschopp and D. L. McDowell
Mechanics and Electromechanics of Single Crystalline Piezoelectric Nanowires ...... 659
M.-F. Yu, Z. Wang, J. Hu and A. Suryavanshi
Multiscale Simulation for High Spped Propagation of Disordered Regions .............. 661
W. Yang, X. Li and Z. Guo
Surface-Stress-Driven Pseudoelasticity and Shape Memory Effect at the Nanoscale. 663
W. Liang and M. Zhou
Thermomechanical Behavior of Zinc Oxide Nanobelts .............................................. 665
A. Kulkarni and M. Zhou
Natural Modes of C60 Cage via Carbon-Carbon Bonding Element............................ 667
P. Zeng, X.-G. Y. and J. Du

11. Deformation and Fracture at the Nano Scale.................................................... 669


Fracture of Nanocrystalline Aluminum ....................................................................... 669
Contents xxiii

C. San Marchi, S. L. Robinson, N. Y. C. Yang and E. J. Lavernia


Wear and Fatigue in Silicon Structural Films for MEMS Applications ...................... 671
D. H. Alsem, R. Timmerman, E. A. Stach, C. L. Muhlstein, M. T. Dugger
and R. O. Ritchie
Indentation Induced Through Thickness Film Fracture on Engineering Alloys ......... 673
D. F. Bahr, K. R. Morasch and A. Alamr
Surface Nanostructured Aluminum by Severe Plastics Deformation.......................... 675
E. I. Meletis, K. Y. Wang and J. C. Jiang
Contribution of Localized Deformation to IGSCC and IASCC in Austenitic
Stainless Steels ............................................................................................................. 677
G. S. Was, Z. Jiao and J. T. Busby
A Study of Crack-Dislocations Interaction with 3D Discrete Dislocation Dynamics . 679
I. N. Mastorakos and H. M. Zbib
Numerical Simulations and Measurements of Cracks Parallel and Near Interfaces
in Graded Structures..................................................................................................... 681
I. Reimanis, K. Rozenburg, J. Berger, M. Tilbrook and M. Hoffmann
Deformation and Failure Processes Operating in Ultra-Fine Grain Metals................. 683
K. Hattar, I. M. Robertson, J. Han, T. Saif, S. J. Hearne and D. Follstaedt
Simulation of Cross-Sectional Nanoindentation in Interconnect Structures with
Cohesive Elements....................................................................................................... 685
D. Gonzalez, J. Molina, I. Ocana, M. R. Elizalde, J. M. Sanchez.
J. M. Martinez-Esnaol, J. Gil-Sevillano, G. Xu, D. Pantuso, T. Scherban,
B. Sun, B. Miner, J. He and J. Maiz
Fracture Between Two Self-Assembled Monolayers .................................................. 687
K. M. Liechti and D. Xu
Nanotube Nanoactuator................................................................................................ 689
M.-F. Yu, J. Hu, Z. Wang and A. Suryavanshi
Nanocrack Detection in Vibrating Nanowires ............................................................. 691
R. Ruoff, L. Calabri, N. Pugno, X. Chen, W. Ding and K. Kohlhaas
Fracture of atomic Layer Deposited Nanolaminate Films ........................................... 693
N. R. Moody, J. M. Jungk, T. M. Mayer, R. A. Wind, S. M. George
and W. W. Gerberich
Influence of Microstructure, Strength and Adhesion on Au Electrodeposits ............. 695
N. Yang, J. Kelly, T. Headley and C. S. Marchi
Fracture of Submicron Thin Metal Films During Cyclic Loading .............................. 697
S. Eve, D. Wang, C. Volkert, N. Huber and O. Kraft
Micromechanics of Damage Evolution in Solid Propellants ....................................... 699
N. Aravas, F. Xu and P. Sofronis
Deformation and Failure Mechanisms in Metallic Nanolayered Composites ............. 701
R. G. Hoagland, J. P. Hirth, and A. Misra
Dislocation Source Sensitivity of Plasticity and Fracture in Tungsten........................ 703
J. E. Talia and R. Gibala
Delamination of Thin Metal Films on Polymers ......................................................... 705
A. Pundt, E. Nikitin, and R. Kirchheim
Fracture Mechanics of One-Dimensional Nanostructures ........................................... 707
W. Ding, L. Calabria, K. M. Kohlhaas, X. Chen and R. S. Ruoff
Effects of Structure and Bonding at Surfaces and Interfaces on Fracture .................. 709
S. P. Lynch, S. Moutsos, B. Gable, S. Knight, D. P. Edwards
xxiv Contents

and B. C. Muddle

29. Reliability and Failure Analysis of Electronics and Mechanical Systems ...... 711
Application of the New Static Photoelastic Experimental Hybrid Method with New
Numerical Method to the Plane Fracture Mechanics................................................... 711
J.-S. Hawong , J.-H. Nam, O.-S. Kwon and K. Tche
Risk Analysis of Buried Pipeline using Probabilistic Method..................................... 713
O. S. Lee, D. H. Kim and N. H. Myoung
Reliability Estimation of Solder Joint by Accelerated Life Tests ................................ 715
O. S. Lee, N. H. Myoung and D. H. Kim
Analysis of Engineering Plastic Behaviors in Thermal Stress Condition.................... 717
S. I. Ham, D. J. Choi and S. D. Park
A Mechanistic Model for the Thermal Fatigue Behavior of the Lead-Free
Solder Joints................................................................................................................. 719
I. Kim, T.-S. Park and S.-B. Lee
Mechanical Behavior of Metallic Thin Film on Polyimide Substrate ......................... 721
D.-C. Baek, S.-Y. Kim and S.-B. Lee

31. Multiscaling in Molecular and Continuum Mechanics - Scaling in Time


and Size From Macro to Nano .................................................................................. 723
Macro-, Meso- and Micro-damage Model Based on Singularity Representation
for Anti-plane Deformation ......................................................................................... 723
G. C. Sih and X. S. Tang
Multiscaling Effects in Trip Steels............................................................................... 725
G. N. Haidemenopoulos and N. Aravas
A Hyper-Surface for the Combined Rate and Size Effects on the Material Properties 727
Z. Chen, L. Shen, Y. Gan and H. E. Fang

34. Cracks in Micro- and Nanoelectronics............................................................... 729


A New Method for Local Strain Field Analysis Near Cracks in Micro- and
Nanotechnology Applications...................................................................................... 729
B. Michel, D. Vogel, N. Sabate and D. Lieske
Experimental Investigations for Fracture Analysis of Solder Joints in
Microelectronic and MEMS Applications ................................................................... 731
H. Walter, C. Bombach, R. Dudek, W. Faust and B. Michel
Simulation of Interface Cracks in Microelectronic Packaging .................................... 733
J. Auersperg, B. Seiler, E. Cadalen, R. Dudek and B. Michel
AFM Based Fracture Analysis in Micro- and Nanomaterials...................................... 735
J. Keller, A. Gollhardt, D. Vogel and B. Michel
Simulation of Deformation and Fracture Behaviour in Microelectronic Packaging ... 737
O. Wittler, H. Walter, J. Keller, R. Dudek, D. Vogel and B. Michel

43. Interfacial Fracture in Composites and Electronic Packaging Materials ...... 739
Mixed-Mode Fracture Modeled Through a Discrete Cohesive Zone Model-DCZM.. 739
D. Xie and A. M. Waas
Signifince of K-Dominance in Delamination Cracking in Composite Laminates....... 741
C. T. Sun and Z. Yang
Contents xxv

Evaluation of Interface Toughness Between Submicron Island and Substrate............ 743


H. Hirakata, T. Kitamura, S. Matsumoto and Y. Takahashi
Three-Dimensional Thermal Stress Analysis Considering the Stress Singularity for
Bonded Structures ........................................................................................................ 745
H. Koguchi
Center of Dilatation and Penny-Shaped Crack in Viscoelastic Bimaterial .................. 747
K. T. Chau, R. C. K. Wong and Y. Z. Sun
Fracture Analysis on Popcorning of Plastic Packages During Solder Reflow ............ 749
S. W. R. Lee and D. C. Y. Lau
Delamination of PB-Free Flip Chip Underfill During 2nd Level Interconnect
Reflow.......................................................................................................................... 751
S. Chung, Z. Tang and S. Park
Reliability of Interfaces Between Components in Advanced Electronic Packages
under Solder Reflow Process ....................................................................................... 753
T. Ikeda and N. Miyazaki
Three-Dimensional Stress Intensity Factors Analyses of Interface Cracks Between
Dissimilar Anisotropic Materials ................................................................................. 755
M. Nagai, T. Ikeda, N. Miyazaki
Molecular Dynamics of Interfacial Fracture................................................................ 757
T. E. Tay, V. B. C. Tan and M. Deng

C. SPECIAL SYMPOSIA/SESSIONS
C2. Engineering Materials and Structures................................................ 759

4. Fracture and Fatigue of Elastomers..................................................................... 761


Nucleation, Growth and Instability of the Cavitation in Rubber ................................. 761
E. Bayraktar, K. Bessri and C. Bathias
Engineering Fracture Mechanics for Crack Toughness Characterisation of
Elastomers.................................................................................................................... 763
K. Reincke, W. Grellmann and G. Heinrich
Multiaxial Fatigue Crack Initiation on Filled Rubbers : Statistical Aspects................ 765
L. Laiarinandrasana, A. Bennani and R. Piques
Fracture Criteria of Rubber-like Materials under Plane Stress Loadings .................... 767
A. Hamdi, M. Nait-Abdelaziz, N. Ait Hocine and P. Heuillet
Prediction of Rubber Fatigue Life under Multiaxial Loading ..................................... 769
A. Zine, N. Benseddiq, M. Nait-Abdelaziz and N. Ait Hocine
Modeling of Biaxial Fatigue of Natural Rubber ........................................................ 771
S. Dong, C. Bathias, K. le Gorjo, F. Hourlier and J. F. Vitorri
Modeling of Crack Propagation in Elastomeric Materials Using Configurational
Forces ........................................................................................................................... 773
T. Horst and G. Heinrich
Determination of Inter-Fibre-Failure in Complex, Reinforced Composites................ 775
V. Trappe and H. Ivers
The Test Frequency Dependence of the Fatigue Behavior of Elastomers ................... 777
Z. Major, Ch. Feichter, R. Steinberger and R. W. Lang
xxvi Contents

5. Integrity of Dynamical Systems ............................................................................ 779


Nonlinear Model for Reinforced Concrete Frames Loaded by Seismic forces ........... 779
D. Kovacevic
Monitoring the Durability Performances of Concrete and Masonry Structures by
Acoustic Emission Technique ..................................................................................... 781
A. Carpinteri and G. Lacidogna
Bifurcation Control of Parametric Resonance in Axially Excited Cantilever Beam .. 783
H. Yabuno and M. Hasegawa
Adaptive Properties of Dynamic Objects..................................................................... 785
I. I. Blekhman and L. A. Vaisberg
Influence of Addendum Modification Coefficient on the Gear's Load Capacity ........ 787
I. Atanasovska and V. Nikoli-Stanojevi
Micromechanical Modelling of Fracture-induced Anisotropy and Damage in
Orthotropic Materials .................................................................................................. 789
V. Monchiet, I.-C. Gruescu, D. Kondo and O. Cazacu
Vibration Control Devices and their Application ....................................................... 791
K. Nagaya
Measurements of Dynamical System Integrity and Fracture Mechanics .................... 793
K. S. Hedrih
Modeling of the Surface Cracks and Fatigue Life Estimation..................................... 795
K. Maksimovic, S. Maksimovic and V. Nikolic-Stanojevic
Structural Damage Detection via the Subspace Identification Method ....................... 797
M. Trajkovic, D. Sumarac and M. Mijalkovic
Clock Mechanism as Base of Artillery Safety and Arming Devices........................... 799
M. Ugrci c
Twisting Deformation Evolution of Drilling Ropes .................................................... 801
N. P. Puchko
Hereditary Strain Theory of Syntetic and Steel Ropes ................................................ 803
O. O. Goroshko
Brittle and Ductile Failure in Thermoviscoplastic Solids under Dynamic Loading.... 805
R. C. Batra and B. M. Love
Some Aspects of Dynamic interfacial Crack Growth ................................................. 807
R. R. Nikolic and J. M. Veljkovic
On Stability Problems of Periodic Impact Motions ..................................................... 809
S. Mitic
Dynamical Integrity of Nonlinear Mechanical Oscillators .......................................... 811
S. Lenci and G. Rega

8. Modelling of Material Property Data and Fracture Mechanisms..................... 813


Fatigue Crack initiation and Propagation at High Temperature in a Softening
Martensitic Steel ......................................................................................................... 813
B. Fournier, M. Sauzay, M. Mottot, V. Rabeau, A. Bougault and A. Pineau
Transferability of Cleavage Fracture Parameters Between Notched and Cracked
Geometries ................................................................................................................... 815
C. Bouchet, B. Tanguy, J. Besson, A. Pineau and S. Bugat
Relation Between Crack Velocity and Crack Arrest .................................................... 817
M. Hajjaj, C. Berdin, P. Bompard and S. Bugat
Contents xxvii

Mechanisms of Damage and Fracture in Trip Assisted Multiphase Steels.................. 819


G. Lacroix, Q. Furnemont, P. J. Jacques and T. Pardoen
The Role of Sub-Boundaries in the Brittle Fracture of Polycrystalline Materials....... 821
G. Hughes, P. Flewitt, F. Sorbello, G. Smith and A. Crocker
Three-Dimensional Modelling of Fracture in Polycrystals.......................................... 823
G. Smith, A. Crocker, G. Hughes and P. Flewitt
Anti-Wing Crack Growth from Surface Flaw in Real Rock under Uniaxial
Compression................................................................................................................. 825
R. H. C. Wong, Y. S. H. Guo , L. Y. Li, K. T. Chau , W. S. Zhu and S. C. Li
Mechanical Behavior Modeling in the Presence of Strain Aging................................ 827
J. Belotteau, C. Berdin, S. Forest, A. Parrot and C. Prioul
On the Local Conditions for Cleavage Initiation in Ferritic Steels.............................. 829
J. Hohe, V. Friedmann and D. Siegele
Unified Constitutive Equations to Describe Elastoplastic and Damage Behavior of
X100 Pipeline Steel...................................................................................................... 831
T. T. Luu, B. Tanguy, J. Besson, A. Pineau and G. Perrin
Estimation of Lower Bound Engineering Fracture Toughness in the Ductile
to Brittle Transition Regime......................................................................................... 833
R. Moskovic and R. A. Ainsworth
Cleavage Fracture Micromechanisms Related to WPS Effect in RPV Steel............... 835
S. R. Bordet, B. Tanguy, S. Bugat, D. Moinereau and A. Pineau
Modelling of Fatigue Damage in Aluminum Cylinder Heads..................................... 837
R. Salapete, B. Barlas, E. Nicouleau, D. Massinon, G. Cailletaud
and A. Pineau
Local Approach to High Temperature Ductility Modeling in 6XXX Aluminium
Alloys ........................................................................................................................... 839
D. Lassance, D. Fabregue, F. Delannay and T. Pardoen

9. Micromechanisms in Fracture and Fatigue......................................................... 841


Small Fatigue Crack Growth in Steel-Compressor Disks of Aircraft Engines............ 841
A. A. Shanyavskiy and A. Yu. Potapenko
Micromechanisms of Damage in Multiaxial Fatigue of an Austenitic-Ferritic
Stainless Steel............................................................................................................... 843
A. el Bartali, V. Aubin, S. Degallaix and L. Sabatier
Multiscale Modeling of Fracture and Plasticity in Layered Structures........................ 845
A. Hartmaier, N. Brodling and H. Gao
Critical and Fracture Planes of 18G2A Steel under Non-Proportional Combined
Bending and Torsion .................................................................................................... 847
A. Karolczuk and E. Macha
Slip Processes and Fracture in Iron Crystals................................................................ 849
V. Pelikan, P. Hora, A. Machova1 and M. Landa
A Discussion of the Applicability of DK-Values to Explain the Fatigue Crack
Growth Behaviour of Short Cracks.............................................................................. 851
A. Tesch, H. Doker, K. H. Trautmann, R. Pippan and C. Escobedo
Simulation of Crack Growth under Low Cycle Fatigue at High Temperature in a
Single Crystal Superalloy............................................................................................. 853
B. Fedelich, Y. Kiyak, T. May and A. Pfennig
xxviii Contents

Fatigue Crack Growth for Different Ratios of Bending to Torsion in ALCU4MG1 .. 855
D. Rozumek and E. Macha
Ductile Damage Models Applied to Anisotropic Fracture of Al2024 T351 ............... 857
D. Steglich, W. Brocks and T. Pardoen
Fatigue and Fracture Processes in Severe Plastic Deformed Rail Steels .................... 859
F. Wetscher, R. Pippan and R. Stock
Damage Evolution in Torsion Specimens Deformed at Forging Temperatures .......... 861
G. Trattnig, R. Pippan and S. Kleber
Microstructural Effects on Short Fatigue Crack Propagation and their Modelling ..... 863
H. J. Christ, O. Duber, W. Floer, U. Krupp, C. P. Fritzen, B. Kunkler
and A. Schick
Micromechanical Aspects of Transgranular and Intergranular Failure Competition .. 865
I. Dlouhy and M. Holzmann
Defect in Ultra-fine Grained Mg-based Alloys Deformed by High-Pressure Torsion 867
J. Cizek, I. Prochazka, B. Smola, I. Stulikova, R. Kuzel, Z. Matej
and V. Cherkaska
Modelling Crack-Tip Shielding Effects in Particle Reinforced Composites .............. 869
J. Hornikova, P. Sandera and J. Pokluda
Early Stages of Fatigue Damage in 316l Steel............................................................. 871
J. Man, K. Obrtlik, J. Polak and P. Klapetek
AB Initio Study of Elasticity and Strength of Nano-Fibre Reinforced Composites .... 873
M. Cerny and J. Pokluda
Strength and Fracture of Ultra-Fine Grained Aluminum 2024 ECAP Metal .............. 875
K. B. Yoon, Y. W. Ma, J. W. Choi and S. H. Kim
Fatigue Lifetime of Bearing Steel in Ultra-High-Cycle Region .................................. 877
L. Kunz, P. Lukas, M. Cincala and G. Nicoletto
Calculation of K-Factor and T-Stress for Crack at Anisotropic Bimaterials .............. 879
M. Kotoul, T. Profant and O. Sevecek
Interaction of Microcracks with Grain Boundaries: Systematical Investigation of
the Mechanisms............................................................................................................ 881
M. Marx, W. Schaf and H. Vehoff
Dislocation Arrangements in Cyclically Strained Inconel 713LC............................... 883
M. Petrenec, K. Obrtlik and J. Polak
Crack Initiation and Fracture of Metal Matrix Composites......................................... 885
K. Unterweger and O. Kolednik
Mechanical Behaviour of Ultra-Fine Grained Austenitic Stainless Steel .................... 887
S. Brochet, A. Poulon-Quintin, J.-B. Vogt , J.-C. Glez and J.-D. Mithieux
Tribological Properties and Wear Mechanisms of Wear Resistant
Thermally Sprayed Coatings........................................................................................ 889
Sa. Houdkova, F. Zahalka and R. Enzl
Crack Propagation Resistance and Damage Mechanisms in Nuclear Graphite........... 891
A. Hodgkins, J. Marrow, P. Mummery, A. Fok and B. J. Marsden
Environment-assisted Cracking of High-Strength Magnesium Alloys WE43-T6....... 893
A. Ahmad and T. J. Marrow
Effects of Surface Finish on the Fatigue Limit in Austenitic Stainless Steels
(Modelling and Experimental Observations) ............................................................... 895
M. Kuroda, T. J. Marrow and A. Sherry
Contents xxix

Intergranular Stress Corrosion Crack Propagation in Sensitised Austenitic


Stainless Steel (Microstructure Modelling and Experimental Observation)................ 897
T. J. Marrow, L. Babout, A. P. Jivkov, P. Wood, D. Engelberg, N. Stevens, P.
J. Wither and R. C. Newman
Ideal Strength of Nanoscale Thin Films ...................................................................... 899
T. Kitamura, Y. Umeno and A. Kushima
Toughness Variability................................................................................................... 901
R. Bouchard, G. Shen and W. R. Tyson
Thermo-Mechanical Behaviour of Nanostructured Copper......................................... 903
C. Duhamel, S. Guerin, M. J. Hytch and Y. Champion
Some Insights into Fatigue Crack Initiation Stage....................................................... 905
H. Alush and Y. Katz
Fatigue Behaviour of Metallic Materials Exposed to High Pressure Hydrogen
Environments ............................................................................................................... 907
Y. Mine, S. Matsuoka, Y. Murakami C. Narazaki and T. Kanezaki
In-Situ Investigations of the Fracture Mechanisms at Various Length Scales............. 909
Z. Pakiela, W. Zielinski and K. J. Kurzydlowski

12. Interface Fracture and Behavior of Joints ........................................................ 911


Environmental Attack at Polymer/Metal Interfaces..................................................... 911
A. J. Kinloch, D. Bland, K. T. Tan and J. F. Watts
Modelling of Elastic-Plastic Peel Tests for Structural Adhesives................................ 913
A. J. Kinloch, H. Hadavinia, L. Kawashita, D. R. Moore and J. G. Williams
An Alternating Crack Growth in Adhesively Bonded Joints ...................................... 915
A. R. Akisanya
Measurements of Interface Fracture and Mechanical Properties of Low-K Dielectric
Thin Films ................................................................................................................... 917
F. Atrash and D. Sherman
Initiation of Fracture Mechanisms at the Fibre/Matrix interface................................. 919
E. Martin, B. Poitou and D. Leguillon
Effects of Plasticity and Residual Stress for Cracks Near Interfaces........................... 921
I. Reimanis, K. Rozenburg, M. Tilbrook and M. Hoffmann
Toughness of a ±45o Interface ..................................................................................... 923
L. Banks-Sills, Y. Freed, R. Eliasi and V. Fourman
Residual Stress Influence on Dissimilar Material Weld Junction Fracture.................. 925
P. Gilles and M.-F. Cipiere
Fracture Mechanisms of a Thin Elastic Plastic Laminate............................................ 927
C. Bjerken, S. Kao-Walter and P. Stahle
Crack-Tip Parameters in Polycrystalline Plates with Compliant Grain Boundaries.... 929
R. Ballarini and Y. Wang
Extended Fe Simulations of Crack Growth in Layered and Functionally Graded
Materials....................................................................................................................... 931
C. Comi and S. Mariani

13. Computational Fracture Mechanics................................................................... 933


Simulation of Plastic Fatigue Crack Growth by a Two Scale Extended Finite
Element Method........................................................................................................... 933
xxx Contents

A. Gravouil, T. Elguedj and A. Combescure


Accurate Determination of Cohesive Crack Tip Fields using Xfem and Admissible
Stress Recovery............................................................................................................ 935
B. L. Karihaloo, Q. Z. Xiao and X. Y. Liu
A New Generation of Boundary Element Method for Damage Tolerance
Assessment of Aerostructures...................................................................................... 937
M. H. Aliabadi
Robust Stress Intensity Factors Evaluation for 3D Fracture Mechanics with
X-FEM ....................................................................................................................... 939
H. Minnebo, E. Bechet and N. Moes
A Micro-Macro Partition of Unity Method for Crack Propagation ............................. 941
P. A. Guidault, O. Allix, L. Champaney and C. Cornuault
A Dynamic Crack Propagation Criteria for XFEM, Based on Path-Independent
Integral Evaluation....................................................................................................... 943
I. Nistor, S. Caperaa and O. Pantale
Truss Model as Simple Computational Tool in Fracture Mechanics ........................... 945
P. G. Papadopoulos, D. Plasatis and P. Lambrou
Finite Element Modeling of Cohesive Cracks by Nitsche’s Method........................... 947
P. Hansbo and P. Heintz
Computing Crack Growth in Quasiperiodic Alloys..................................................... 949
P. M. Mariano and F. L. Stazi
X-FEM for 3D Cracks in Shaft with Contact .............................................................. 951
S. Geniaut, P. Massin and N. Moes
Some Improvements for Extended Finite Element Methods in Fracture Mechanics .. 953
P. Laborde, J. Pommier, Y. Renard and M. Salaun

14. Cohesive Models of Fracture .............................................................................. 955


Failure Prediction of Adhesively Bonded T-Peel Joints .............................................. 955
A. Pirondi
An Approach for the Determination of Mixed Mode Cohesive Laws......................... 957
B. F. Sorensen and T. K. Jacobsen
The Use of CZM for Coupled Fatigue/Plasticity Crack Propagation Simulation........ 959
Jl. Bouvard, F. Feyel and Jl. Chaboche
Dynamic Crack Growth : Analytical and Numerical CZM Approaches ..................... 961
G. Debruyne, J. Laverne and P. E. Dumouchel
Simulation of Pre-Critical Cracking in Concrete Using 3D Lattice Model................. 963
H.-K. Man and J. G. M. Van Mier
Effect of Cohesive Law and Triaxiality Dependence of Cohesive Parameters in
Ductile Tearing............................................................................................................. 965
I. Scheider, F. Hachez and W. Brocks
Modeling Quasibrittle Material Cracking with Cohesive Cracks: Experimental
and Computational Advances ...................................................................................... 967
J. Planas, J. M. Sancho, A. M. Fathy, D. A. Cendon and J. C. Galvez
Pinwheel Meshes and Branching of Cohesive Cracks................................................. 969
P. Ganguly and K. D. Papoulia
A Dynamic Crack Growth Simulation Using Cohesive Elements .............................. 971
M. Anvari and C. Thaulow
Contents xxxi

A New Cohesive Zone Model for Mixed-Mode Decohesion ...................................... 973


M. J. van Den Bosch , P. J. G. Schreurs and M. G. D. Geers
Cohesive-Zone Modeling of Crack Growth in Specimens with Different Constraint
Conditions .................................................................................................................... 975
C. R. Chen, O. Kolednik and F. D. Fischer
Effect of Anisotropic Plasticity on Mixed Mode Interface Crack Growth .................. 977
V. Tvergaard and B. N. Legarth

16. Environment Assisted Fracture.......................................................................... 979


Characterisation of TG-SCC in Pure Magnesium and AZ91 Alloy ............................ 979
N. Winzer, G. Song, A. Atrens, W. Dietzel and C. Blawert
Hydrogen Embrittlement and Cracking of 18MN-4CR Steels .................................... 981
A. Balitskii
Transient Stress and EAC of Steam Turbine Disc Steel .............................................. 983
A. Turnbull and S. Zhou
Irreversible Hydrogen Trapping in Welded Beta-21S Titanium Alloy ........................ 985
D. Eliezer, E. Tal-Gutelmacher, C. E. Cross and Th. Boellinghaus
EAC in High Strength Steels for Gas Transportation .................................................. 987
G. Gabetta and R. Bruschi
High Temperature Fatigue Crack Growth in Titanium Microstructures...................... 989
H. Ghonem
Corrosion Damaging and Corrosion Fatigue Assessment in Three-Layered
Metallic Material.......................................................................................................... 991
I. M. Dmytrakh and V. V. Panasyuk
Simulation of Hydrogen Assisted Stress Corrosion Cracking Using a Time
Dependent Cohesive Model ......................................................................................... 993
I. Scheider, M. Pfuff and W. Dietzel
Environmental Stress Cracking of Polyethylene Pipes in Water Distribution
Networks ...................................................................................................................... 995
J. P. Dear, N. S. Mason and M. Poulton
Fatigue Crack Propagation in 2XXX Aluminium Alloys at 223K .............................. 997
C. Gasqueres, C. Sarrazin-Baudoux, D. Dumont and J. Petit
Hydrogen Assisted Cracking Paths in Oriented Pearlitic Microstructures .................. 999
J. Toribio and E. Ovejero
Effect of Residual Stress-Strain Profile on Hydrogen Embrittlement Susceptibility
of Prestressing Steel Wires......................................................................................... 1001
J. Toribio and V. Kharin
Hydrogen Embrittlement of Austenitic Stainless Steels at Low Temperatures ......... 1003
L. Zhang, M. Wen, M. Imade, S. Fukuyama and K. Yokogawa
Hydrogen Diffusion and EAC of Pipeline Steels under Cathodic Protection .......... 1005
M. Cabrini and T. Pastore
Initiation of Environmentally Assisted Cracking in Line Pipe Steel ......................... 1007
M. Elboujdaini
Fatigue Crack Growth Behaviour Depending on Environment in Magnesium
Alloys ........................................................................................................................ 1009
M. Nakajima, K. Tokaji, Y. Uematsu and T. Shimizu
xxxii Contents

Assessment of High-Temperature Hydrogen Degradation of Power Equipment


Steels .......................................................................................................................... 1011
H. M. Nykyforchyn and O. Z. Student
Stress Corrosion Cracking of 18MN-4CR Generator Rotor End-retaining Ring
Steel............................................................................................................................ 1013
N. Mukhopadhyay and U. K. Chatterjee

17. SIM, Philosophy, Instrumentation and Analysis ............................................ 1015


Non Contacting Stress Monitoring ............................................................................ 1015
W. D. Dover, R. F. Kare and N. Stone
Rapid Calculation of Stress Intensity Factors ............................................................ 1017
A. J. Love and F. P. Brennan
Variable Amplitude Corrosion Fatigue of High Strength Weldable Steel ................. 1019
S. S. Ngiam and F. P. Brennan
Crack Monitoring using ACFM................................................................................. 1021
R. F. Kare

18. Fracture of Biomaterials ................................................................................... 1023


Fatigue Behaviour of Fiber Reinforced Bone Cement .............................................. 1023
B. Kumar and F. W. Cooke
Fracture and Fatigue of Bone and Bone Cement: The Critical Distance Approach .. 1025
D. Taylor, D. Hoey, L. Sanz and P. O’Reilly
Fatigue Failure in Reconstracted Acetabula – a Hip Simulator Study....................... 1027
J. Tong, N. P. Zant and P. Heaton-Adegbile
Deformation and Fracture of Bioactive Particulate Composites Developed for
Hard Tissue Repair..................................................................................................... 1029
M. Wang
Failure of Biomaterials in Implant Fixation............................................................... 1031
P. J. Prendergast, J. R. Britton, P. T. Scannell and A. B. Lennon

19. Structural integrity Assessment in Theory and Practice ............................... 1033


Stress Analisys of High Pressure Steamlines in Thermal Power Plants .................... 1033
A. Jakovljevic
Laminar Composite Materials Damage Monitoring by Embedded Optical Fibers ... 1035
A. Kojovi, I. Zivkovi, L. Brajovi, D.Mitrakovi and R. Aleksi
SOL GEL Synthesis and Structure of Hybrid Nanomaterials with Strong Chemical
Bonds ......................................................................................................................... 1037
B. Samuneva, P. Djambaski, E. Kashchieva and G. Chernev
An Alternative Approach to Conventional Data Presentation of Fatigue.................. 1039
D. Angelova
Absorbers of Seismic Energy for Damaged Masonary Structures............................. 1041
D. Sumarac, Z. Petraskovi, S. Miladinovic, M. Trajkovi, M. Andjelkovic
and N. Trisovic
Numerical Analysis of Tensile Specimen Fracture with Crack in HAZ ................... 1043
G. Adziev, A. Sedmak and T. Adziev
Determination of JR-Curve by Two Points Method .................................................. 1045
I. Blacic and V. Grabulov
Contents xxxiii

Monitoring of Stress-Strain State of Boiler During Pessure Test ............................. 1047


J. Kurai, Z. Burzic, N. Garic, M. Zrilic and B. Aleksic
Local Variation of Crack Driving Force in a Mismatched Weld ............................... 1049
J. Predan, N. Gubeljak and O. Kolednik
Strength Recovery of Machined Alumina by Self Crack Healing............................. 1051
K. Ando, K. Takahashi, W. Nakao, T. Osada and S. Sato
Crack Initiation and Growth in HAZ of Microalloyed Steel ..................................... 1053
K. Geric and S. Sedmak
Structural Integrity at Elevated Temperatures - Residual Service Life Evaluation ... 1055
L. Milovic and S. Sedmak
The Analysis of Supporting Structure of Planetary Gear Box Satellite .................... 1057
M. Arsi, V. Aleksi and Z. Anelkovi
Failure Probability of Gear Teeth Wear ..................................................................... 1059
M. Ognjanovic
Some Aspects of Engineering Approach to Structural Integrity Assessment............ 1061
M. Kiric and A. Sedmak
Structural Integrity Assessment Applying Ultrasonic Testing................................... 1063
M. Kiric
(Crack-Healing + Proof-Test): Methodology to Guarantee the Reliability of
Ceramics..................................................................................................................... 1065
M. Ono, W. Nakao, K. Takahashi, K. Ando and M. Nakatani
Risk Based Integrity Assessment of Concrete Structures .......................................... 1067
M. Pavisic
Structural Integrity Assessment by Local Approach to Fracture ............................... 1069
M. Zrilic, M. Rakin, Z. Cvijovic, A. Sedmak and S. Sedmak
Brittle and Ductile Fracture in Service of Pressure Vessels....................................... 1071
N. Filipovic and K. Geric
Mechanisms of Fracture in Medium Carbon Vanadium Microalloyed Steels ........... 1073
N. Radovic, Dj. Drobnjak and H. Hraam
Computation and Experimental Investigations of Notched Components Fatigue
Life Estimation .......................................................................................................... 1075
S. Maksimovic, Z. Burzic and K. Maksimovic
Failure Analysis of Layered Composite Structures: Computation and Experimental
Investigation............................................................................................................... 1077
S. Maksimovic
Loading Rate Effect on HSLA Steel Welded Joints Fracture Resistance ................. 1079
V. Grabulov, I. Blai, A. Radovi and S. Sedmak
Case Study of Supporting Tubes Failure.................................................................... 1081
V. S. Zeravcic, M. Djukic, G. Bakic, B. Andjelic and B. Rajicic
Structure Integrity of Pressure Vesels Repair Welding Joints.................................... 1083
V. S. Zeravcic, G. Bakic, M. Djukic and B. Rajicic
Effect of Microalloyed Steel Welding Procedure on Fatigue Crack Growth ............ 1085
Z. Burzic, V. Grabulov, M. Burzic, M. Manjgo, V. Gliha and T. Vuherer
Fracture Resistance of High-Strength 7000 Forging Alloys...................................... 1087
Z. Cvijovic, M. Rakin and M. Vratnica

20. Critical Distance Theories of Fracture............................................................. 1089


xxxiv Contents

Does a Characteristic Crack-Tip Distance Imply Discontinuous Crack


Propagation?............................................................................................................... 1089
A. P. Kfouri
A Multiaxial Criterion for Notch Fatigue Using a Critical-Distance Method .......... 1091
A. Carpinteri, A. Spagnoli, S. Vantatori and D. Viappiani
Size Effects for Crack Initiation at Blunt Notches or Cavities in Brittle Materials... 1093
D. Leguillon, E. Martin, D. Picard and C. Putot
The Theory of Critical Distances ............................................................................... 1095
D. Taylor
Strength Analysis of Composite Pinned Joints .......................................................... 1097
H. A. Whitworth, O. Aluko and N. Tomlinson
Application of the Theory of Critical Distance to Fretting Fatigue .......................... 1099
J. A. Araujo, L. Susmel, D. Taylor and L. H. M. Lopes
The Theory of Critical Distances: Applications in Fatigue ....................................... 1101
L. Susmel
Fatigue Assessment using an Integrated Threshold Curve Method - Applications... 1103
M. D. Chapetti
Anaytical Approaches vs Atomistic Simulations in Fracture ................................... 1105
N. Pugno, A. Carpinteri, M. Ippolito, A. Mattoni and L. Colombo
A Coupled Stress and Energy Criterion within Finite Fracture Mechanics............... 1107
P. Cornetti, N. Pugno, A. Carpinteri and D. Taylor
Local Strain Energy Density and Fatigue Strength of Welded Joints ........................ 1109
P. Lazzarin, P. Livieri and F. Berto
An Implicit Gradient Application to Fatigue of Notches and Weldments .................. 1111
R. Tovo and P. Livieri
Use of JVR to Predict Static Failures in Notched Components................................. 1113
P. Livieri
Standardization of Strength Evaluation Methods Using Critical Distance Stress ..... 1115
T. Hattori, N. Nishimura and M. Yamashita
Application of Point Stress Method to Hydro-Fracturing Tectonic Stress
Measurement.............................................................................................................. 1117
T. Ito
A Unified Failure Criterion for Brittle or Quasi-Brittle Materials under Arbitrary
Stress Concentration .................................................................................................. 1119
J. Li and X. B. Zhang

22. New Investigations on Very High Cycle Fatigue of Materials ....................... 1121
Morphology of Step-Wise S-N Curves Depending on Notch and Surface
Roughness in High Strength Steel.............................................................................. 1121
H. Itoga, K. Tokaji, M. Nakajima and H. N. Ko
Very High Cycle Fatigue Behaviour under Cyclic Torsion Loading ......................... 1123
H. Mayer and S. Stanzl-Tschegg
Modelling of Fatigue Crack Growth From Exfoliation and Pitting Corrosion.......... 1125
G. Clark, P. K. Sharp and R. Jones
Does Copper Undergo Surface Roughening during Fatigue in the VH Regime?...... 1127
S. Stanzl-Tschegg, H. Mughrabi and R. Schuller
Crack Initiation Mechanism of Bearing Steel in High Cycle Fatigue ....................... 1129
Contents xxxv

T. Sakai
Very High Cycle Fatigue Behavior of High Strength Steels...................................... 1131
Y. Akiniwa, N. Miyamoto, H. Tsuru and K. Tanaka

23. Deformation and Fracture of Engineering Materials..................................... 1133


Fracture Toughness of Hydrided Zircaloy-4 Experimental and Numerical Study .... 1133
C. Langlade, P. Bouffioux and M. Clavel
Crack Growth Behavior in a Highly Filled Elastomer............................................... 1135
C. T. Liu , R. Neviere and G. Ravichandran
Crack Tip Behavior in TiAl when Approaching Grain Boundary ............................. 1137
F.-P. Chiang, S. Chang and K. Wang
Effect of Loading Rate on the Energy Release Rate in a Constrained Elastomeric
Disk ............................................................................................................................ 1139
H. K. Ching, C. T. Liu and S. C. Yen
Analyses of Progressive Damage and Fracture of Particulate Composite Materials
Using S-FEM Technique............................................................................................ 1141
H. Okada, S. Tanaka, Y. Fukui and N. Kumazawa
Fracture Mechanics on PVDF Polymeric Material : Specimen Geometry Effects.... 1143
L. Laiarinandrasana and G. Hochstetter
Fracture Toughness of Alloyed Austempered Ductile Iron (ADI)............................. 1145
O. Eric, D. Rajnovic, Z. Burzic, L. Sidjanin and M. T. Jovanovic
Prediction of Crack Growth under Random Load in Railway Wheel ....................... 1147
R. Hamam, S. Pommier and F. Bumbieler

24. Materials Damage Prognosis and Life Cycle Engineering............................. 1149


Predicting the Evolution of Stress Corrosion Cracks From Pits ................................ 1149
A. Turnbull, L. N. Mccartney and S. Zhou
Corrosion Problems in Nuclear Industry : Lessons Learned and Perspectives.......... 1151
J. M. Boursier, F. Foct, F. Vaillant and E. Walle
Aluminium Alloys Fatigue Evaluation Method......................................................... 1153
S. Rymkiewicz

25. Mixed-Mode Fracture........................................................................................ 1155


Singular Stress Fields Situations in Mode-II and Mixed-Mode Loaded Cracks ...... 1155
D. Fernández-Zúñiga, J. F. Kalthoff, A. Blázquez and A. Fernández-Canteli
Evaluation of M-Integral for Rubbery Material Problems Containing Multiple
Cracks......................................................................................................................... 1157
J.-H. Chang and D.-J. Peng
Use of a Crack Box Technique for Crack Bifurcation in Ductile Material................ 1159
D. Lebaillif, X. B. Zhang and N. Recho
Mixed Mode Fracture of Linear Elastic Materials with Cubic Symmetry ................ 1161
D. E. Lempidaki, N. P. O’Dowd and E. P. Busso
Three-Dimensional Experimental and Numerical SIFs and Crack Growth ............. 1163
D. M. Constantinescu, B. Bocaneala and L. Marsavina
An Arbitrarily Oriented Crack Near a Coated Fiber.................................................. 1165
H. M. Shodja and F. Ojaghnezhad
Simulation of the Mixed Mode Fracture of Concrete with Cohesive Models ........... 1167
xxxvi Contents

J. C. Gálvez, D. A. Cendón, E. Reyes, J. M. Sancho and J. Planas


Micromechanical Analysis of Rupture Mechanisms in Mixed Mode Ductile
Fracture ...................................................................................................................... 1169
I. Barsoum and J. Faleskog
Mode I Preloading-Mode II Fracture in Warm Pre-Stressing .................................... 1171
M. R. Ayatollahi and M. Mostafavi
Predictions of Mixed Mode I/II Fracture toughness for Soft Rocks.......................... 1173
M. R. Ayatollahi and M. R. M. Aliha
An Interface Model for Mixed-mode, Buckling-Driven Decohesion of Superficial
Layers......................................................................................................................... 1175
S. Bennati and P. S. Valvo
MXED-Mode Fracture Analyss of Orthotropc Functonally Graded Materals ......... 1177
S. Dag, B. Yildirim, D. Sarikaya
New Scheme for Fea of Mixed Mode Stable Crack Growth ..................................... 1179
S. K. Maiti , S. Namdeo and A. H. I. Mourad
Numerical Simulation of Nonlinear Crack Propagation under Mixed-Mode Impact
Loading ...................................................................................................................... 1181
T. Fujimoto and T. Nishioka
Elastic-Plastic Behaviour of Crack Propagation under Biaxial Cyclic Loading ....... 1183
T. Hoshide
Numerical Analysis of Mixed-Mode Cracking in Concrete Dams............................ 1185
Z. Shi

26. Fracture Mechanics Characterization of Wood.............................................. 1187


Species and Other Physical Effects on Parameters Describing a Wood Toughness
Test. ............................................................................................................................ 1187
B. Thibaut and J. Beauchene
Yew and Spruce Wood: Mechanical Properties and Fracture Surface Studies .......... 1189
D. Keunecke, C. Marki and P. Niemz
Critical Crack Lengths in FRP Reinforced Glulam Beams ....................................... 1191
J. Desjarlais, W. G. Davids and E. N. Landis
Failure Analysis of Engineering Wood Products ....................................................... 1193
I. Smith, M. Snow and A. Asiz
Modelization of Slow Crack Growth in Wood Considered as a Damage
Viscoelastic Material.................................................................................................. 1195
M. Chaplain and G. Valentin
Mode I Crack Propagation in Softwood, Microanalyses and Modeling.................... 1197
P. Navi and M. Sedighi-Gilani
Fracture Properties of Pine and Spruce in Mode I ..................................................... 1199
N. Dourado, S. Morel, M. F. S. F. De Moura, G. Valentin and J. Morais
Influence of the Specimen Geometry on R-Curve: Numerical Investigations. ......... 1201
C. Lespine, S. Morel, J.-L. Coureau and G. Valentin
Fracture Behaviour and Cutting of Small Wood Specimens in RT-Direction ........... 1203
S. Koponen and P. Tukiainen
Fracturing of Wood under Torsional Loading: Fracture Mechanisms and
Mechanics .................................................................................................................. 1205
E. K. Tschegg and S. E. Stanzl-Tschegg
Contents xxxvii

On the Influence of Humidity Cycling on Fracture Properties of Wood ................... 1207


S. Vasic and S. Tschegg
Determination of Cohesive Fracture Parameters for Wood ....................................... 1209
T. Astrup, J. F. Olesen, L. Damkilde and P. Hoffmeyer
The Role of Fracture Toughness in the Cutting of Wood .......................................... 1211
T. Atkins

28. Short Fatigue Crack Growth under Multi-Axial Loading Conditions ......... 1213
Short Fatigue Cracks of In-Service Fatigued Turbine Blades ................................... 1213
A. A. Shanyavskiy, M. A. Artamonov, A. L. Tushentsov
and Yu. A. Potapenko
Short Crack Growth under Cyclic Torsion with Static Tension................................. 1215
I. Ohkawa, S. Hirano, T. Negishi and M. Misumi
Resistance-Curve Method for Predicting Fatigue Thresholds under Combined
Loading ...................................................................................................................... 1217
K. Tanaka, Y. Akiniwa and M. Wakita
The Growth of Short Cracks From Defects under Multi-Axial Loading................... 1219
M. Endo and A. J. Mcevily
Short Fatigue Cracks in Notched and Unnotched Specimens under
Non-Proportional Loading ......................................................................................... 1221
O. Hertel, T. Seeger, M. Vormwald, R. Doring and J. Hoffmeyer
Microcracks Growth in Push-Pull and Reversed Torsion in Stainless Steel.............. 1223
V. Doquet and G. Bertolino
Hydrogen and Notch Effects on Torsional Fatigue of Stainless Steel ....................... 1225
Y. Kondo, M. Kubota and K. Ohguma

30. Integrity of Gears ............................................................................................... 1227


Influence of Moving Tooth Load on Gear Fatigue Behaviour................................... 1227
D. T. Jelaska and S. Podrug
Comparison of Solid Spur Gear Face Load Factors .................................................. 1229
G. Marunic
Prediction of Contact Fatigue Internal Crack Propagation in Hypoid Gears............. 1231
M. Vimercati, M. Guagliano, L. Vergani and A. Piazza
Fatigue Crack Initiation Along Inclusion Interfaces of Contacting
Mechanical Elements ................................................................................................. 1233
S. Glodez, M. Ulbin and J. Flasker
Energy Based Gear Fault Diagnostics........................................................................ 1235
S. J. Loutridis
Crack Propagation in Gear Tooth Root...................................................................... 1237
S. Pehan, B. Zafosnik and J. Kramberger
Experimental Evaluation of Stress Intensity Factors in Spur Gear Teeth .................. 1239
V. Spitas , G. Papadopoulos, Th. Costopoulos and C. Spitas

35. High Temperature and Thermomechanical Fatigue ...................................... 1241


Isothermal and Thermomechanical Fatigue Behavior of the ODS Superalloy
PM1000...................................................................................................................... 1241
W. O. Ngala, G. Biallas and H. J. Maier
xxxviii Contents

Fatigue-Creep-Environment Interactions in a Directionally-Solidified Ni-Base


Superalloy .................................................................................................................. 1243
A. P. Gordon, M. M. Shenoy, R. W. Neu and D. L. McDowell
The Effects of Microstructure, Deformation Mode and Environment on Fatigue .... 1245
S. D. Antolovich and B. F. Antolovich
Comparing Fatigue Behaviour of TI6242 and Novel TIAL Intermetallics ............... 1247
T. K. Heckel, A. Guerrero-Tovar and H. J. Christ
A TBC Failure Model Based on Crack Number Density ......................................... 1249
X. Wu, Z. Zhang and R. Liu

36. Impact Failure of Laminated and Sandwich Composite Structures............. 1251


Impact Induced Composite Delamination: State and Parameter Identification via
Unscented Kalman Filter ........................................................................................... 1251
A. Corigliano, A. Ghisi and S. Mariani
Modelling Impact Damage in Sandwich Concept Structures .................................... 1253
A. Johnson and N. Pentecote
Punch Shear Behavior of Composites at Low and High Rates.................................. 1255
B. A. Gama and J. W. Gillespie Jr.
Repeated Impact Behaviour and Damage Progression of Glass Reinforced Plastics 1257
G. Belingardi, M. P. Cavatorta and D. S. Paolino
Impact Behaviour Modelling of a Composite Leading Edge Structure..................... 1259
G. Labeas and Th. Kermanidis
Bending Strength of Sandwich Panels with Different Cores After Impact ............... 1261
W. Goettner and H. G. Reimerdes
Energy Absorbing Ability of Sandwich Composite Structures ................................. 1263
J. P. Dear, W. Maruszewska, S. T. Oh and H. Lee
Impact Behaviour of Metal Foam Cored Sandwich Beams....................................... 1265
S. Mckown and R. A. W. Mines

37. Mesofracture and Transferability .................................................................... 1267


Stress Gradient at Notch Roots Using Volumetric Method ....................................... 1267
H. Adib and G. Pluvinage
Local Approach Use at Solution of Fracture Parameters Transferability .................. 1269
L. Jurasek, M. Holzmann and I. Dlouhy
Damage in Rubber-Modified Polymers : Experimental, Modelling and
Computational Aspects .............................................................................................. 1271
N. Belayachi, F. Zaïri, N. Benseddiq and M. Naït Abdelaziz
Failure Assessment Diagrams Based on the Criterion of Average Stress ................. 1273
Y. G. Matvienko

38. Damage in Composites - Damage Development in Composite Materials


& Structures - Models of Prediction ............................................................... 1275
Material Models for Damaged Composite Laminates ............................................... 1275
J. Varna
Raman Spectroscopy Assessment of Stiffness Reduction and Residual Strains
due to Matrix Cracking in Angle – PLY Laminates .................................................. 1277
P. Lundmark, D. G. Katerelos, J. Varna and C. Galiotis
Contents xxxix

Physical Modelling of Failure Processes in Composite Materials............................. 1279


P. W. R. Beaumont
NCF Cross-PLY Laminates: Damage Accumulation and Degradation of Elastic
Properties ................................................................................................................... 1281
R. Joffe and D. Mattsson
Matrix Crack Initiation and Propagation in Laminates with Off-Axis PLIES........... 1283
N. Vrellos, S. L. Ogin and P. A. Smith
Stress Oscillation and Instability of Yielding in Polymers and Nanocomposites...... 1285
D. E. Mouzakis, G. Kandilioti, S. Tzavalas and V. Gregoriou
Prediction of Cyclic Durability of Woven Composite Laminates ............................. 1287
V. Tamuzs and K. Reifsnider

39. Aging Aerostructures......................................................................................... 1289


Repair of Corroded Aerospace Aluminium Panels Using Ultrasonic Impact
Treatment ................................................................................................................... 1289
C. A. Rodopoulos, S. Pantelakis, M. Liao and E. Statnikov
Fatigue Crack Initiation in Stress Concentration Areas............................................. 1291
C. Schwob, F. Ronde-Oustau and L. Chambon
Hydrogen Trapping: Deformation and Heat Treatment Effects in 2024 Alloy ......... 1293
H. Kamoutsi, G. N. Haidemenopoulos, V. Bontozoglou , P. V. Petroyiannis
and Sp. G. Pantelakis
An Integrated Methodology Assessing the Aging Behaviour of Aircraft Structures 1295
G. Labeas and I. Diamantakos
Numerical Investigation on the Tensile Behaviour of Pre-Corroded 2024
Aluminium Alloy ....................................................................................................... 1297
P. V. Petroyiannis, G. Labeas, Sp. G. Pantelakis, E. Kamoutsi,
V. Bontozoglou and G. N. Haidemenopoulos

40. Residual Stress and its Effects on Fatigue and Fracture................................ 1299
Assessment of Defects under Combined Primary and Residual Stresses .................. 1299
A. H. Sherry and M. R. Goldthorpe
Effect of Residual Stresses on the Crack Growth in Aluminum ............................... 1301
B. Kumar and J. E. Locke
Effect of the Cryogenic Wire Brushing on the Surface Integrity and the Fatigue
Life Improvement of the AISI 304 Stainless Steel Ground Components ................. 1303
N. B. Fredja, H. Sidhoma and C. Brahamb
Interaction of Residual Stress with Mechanical Loading in Ferritic Steels ............... 1305
A. Mirzaee-Sisan, C. E. Truman and D. J. Smith
Evaluation of Novel Post Weld Heat Treatment in Ferritic Steel Repair Welds
Based on Neutron Diffraction .................................................................................... 1307
C. Ohms, D. Neov, R. C. Wimpory and A. G. Youtsos
Surface Crack Development in Transformation Induced Fatigue of SMA Actuators 1309
D. C. Lagoudas, O. W. Bertacchini and E. Patoor
Finite Element Simulation of Welding in Pipes: a Sensitivity Analysis.................... 1311
D. Elias Katsareas , C. Ohms and A. G. Youtsos
Residual Stress Prediction in Letterbox-Type Repair Welds ..................................... 1313
L. Keppas, N. K. Anifantis, D. E. Katsareas and A. G. Youtsos
xl Contents

Effect of Reflection Shot Peening and Fine Grain Size on Improvement of Fatigue
Strength for Metal Bellows ....................................................................................... 1315
H. Okada, A. Tange and K. Ando
Viscosity Effect on Displacements and Stresses of a Two-Pass Welding Plate ........ 1317
W. El Ahmar and J. F. Jullien
Surface Integrity in High Speed Machining of TI-6WT.%Al-4WT.%V Alloy ......... 1319
J. D. P. Velasquez, B. Bolle, P. Chevrier and A. Tidu
Phase Transformation and Damage Elastoplastic Multiphase Model for Welding
Simulation .................................................................................................................. 1321
T. Wu, M. Coret and A. Combescure
The Present Sans Instrument and the New HFR-Petten Sans Facility Based
on a Cold Neutron Source.......................................................................................... 1323
O. Ucaa,B, C. Ohmsa, D. Neova and A. G. Youtsosa
Residual Stress Numerical Simulation of Two Dissimilar Material Weld Junctions. 1325
P. Gilles, L. Nouet and P. Duranton
Identification of Weld Residual Stress Length Scales for Fracture Assessment ....... 1327
P. J. Bouchard and P. J. Withers
High-Resolution Neutron Diffraction for Phase and Residual Stress Investigations 1329
P. Mikula and M. Vrana
Sensitivity of Predicted Residual Stresses to Modelling Assumptions...................... 1331
S. K. Bate, R. Charles, D. Everett, D. O’Gara1, A. Warren
and S. Yellowlees
Welding Effects on Thin Stiffened Panels ............................................................ 1333
T. T. Chau
Evaluation of Residual Stresses in Ceramic Polymer Matrix Composites Using
Finite Element Method............................................................................................... 1335
K. Babski, T. Boguszewski, A. Boczkowska, M. Lewandowska,
W. Swieszkowski and K. J. Kurzydlowski

41. Computational Modeling of Multiphysics Degrading Systems (CMMDS) .. 1337


Towards Data-Driven Modeling and Simulation of Multiphysics Degrading
Systems ..................................................................................................................... 1337
J. G. Michopoulos and C. Farhat
Mathematical Modelling of Piezoceramic Transducer Performance in the
Presence of Material Defects ..................................................................................... 1339
T. A. Christensen, N. L. Andersen, and M. Willatzen
A Continuum Approach for Identifying Elastic Moduli of Composites.................... 1341
J. G. Michopoulos and T. Furukawa
Regularized Identification of Material Constants Using Multi-Objective
Gradient-Based Method ............................................................................................. 1343
T. Furukawa and J. G. Michopoulos
Loading and Material Features Influence on Piezoelectric Material Performance.... 1345
V. G. Degiorgi and S. A. Wimmer
Modeling of Plasma Chemical Deposition and Degradation of Silicon Thin Films . 1347
V. V. Krzhizhanovskaya, P. M. A. Sloot and Y. E. Gorbachev

42. Scaling and Size Effects ..................................................................................... 1349


Contents xli

A Fractal Approach Interpretation for the Indentation Size Effect............................ 1349


A. Carpinteri and S. Puzzi
Description of Multi-Scaling Power Laws in Fracture and Strength......................... 1351
A. M. Korsunsky
The Spalling Failure Around Deep Excavations in Rock Masses ............................ 1353
A. P. Fantilli and P. Vallini
Scaling in Multiaxial Compressive Fracture.............................................................. 1355
A. S. Elkadi and J. G. M. van Mier
Fracture of Antarctic FY Sea Ice ............................................................................... 1357
J. P. Dempsey, S. Wang and D. M. Cole
Mixed Mode Fracture of Brickwork Masonry........................................................... 1359
J. C. Galvez, E. Reyes, M. J. Casati, J. M. Sancho, J. Planas
and D. A. Cendon
Geometric Scaling and Instability in FRP-Concrete Debonding ............................... 1361
K. V. Subramaniam, M. Ali-Ahmad and C. Carloni
A Simplified MCFT for Shear Capacity Scaling of R/C Beams ............................... 1363
M. T. Kazemi and V. Broujerdian
Interplay of Sources of Size Effects in Concrete Specimens..................................... 1365
M. Vorechovsky and D. Matesova
Scale Effect in Elastic and Strength Properties of Nanostructures ............................ 1367
O. S. Loboda , A. M. Krivtsov and N. F. Morozov
Fracture Toughness Assessment of a C-MN Steel Using Miniature Specimens ....... 1369
P. J. Apps, W. Geary, J. W. Hobbs and G. Wardle
Size Effect in the Bonding of Smooth and Deformed Bars: NSC versus HPC ......... 1371
P. Bamonte, D. Coronelli and P. G. Gambarova
Size Effect in the Cracking of Drying Soil ................................................................ 1373
P. C. Prat, A. Ledesma, and M. R. Lakshmikantha
Modelling of the Volume Effects Related to the Unixial Behaviour of Concrete.
From a Discontinuous to a Macroscopic Approach................................................... 1375
P. Rossi, J. L. Tailhan, J. Lombart and A. Deleurence
Size Effect and R-Curve in Quasibrittle Fracture ...................................................... 1377
S. Morel, E. Bouchaud and G. Valentin
Bifurcation and Size Effect in a Viscoelastic Non-Local Damageable Continuum... 1379
Th. Baxevanis, G. Pijaudier-Cabot and F. Dufour
Ultiscale Necessary and Sufficient Strength Criteria................................................. 1381
V. M. Kornev
Size Effects: Moving forwards................................................................................... 1383
X. Hu and K. Duan
An Experimental Study on Rapid Setting Concrete Repair Materials....................... 1385
J. P. Richards and Y. Xi

44. Multiple Cracking and Delamination .............................................................. 1387


Hierarchical Failure Modeling and Related Scale-Invariant Probability
Distributions of Strength ............................................................................................ 1387
D. A. Onishchenko
Interaction of Two Adhesively Bonded Weak Zones .............................................. 1389
I. V. Simonov and B. L. Karihaloo
xlii Contents

Multiple Cracking in Surface-Hardened Tensile Specimens and their Fracture


Mechanisms ............................................................................................................... 1391
L. S. Derevyagina, V. E. Panin, R. V. Goldstein, N. A. Antipina
and I. L. Strelkova
Fracture Criterion of Cracks Initiation and Growth................................................... 1393
M. Perelmuter
Interfacial Cracks Emanating from Partially Debonded Subsurface Circular Elastic
Inclusions ................................................................................................................... 1395
P. B. N. Prasad
Mechanics of Block Structures and its Applications to Geodynamics ...................... 1397
P. V. Makarov
Static and Dynamic Response of Multiple Delaminations ........................................ 1399
M. G. Andrews and R. Massabo
Modeling Crack Growth in Structure- Nonhomogeneous Medium under Complex
Stress State ................................................................................................................. 1401
R. V. Goldstein, Y. V. Zhitnikov and N. M. Osipenko
Nonideal Interface of a Bimaterial with Defects under Thermal Load...................... 1403
V. E. Petrova and K. P. Herrmann
Multiple Cracking Development at the Prefructure Stage of Ion Crystals ................ 1405
Y. Y. Deryugin, V. E. Panin, V. Hadjicontis, K. Mavromatou

Author Index ............................................................................................................ 1407


Editor’s Preface

This volume contains two-page abstracts of the 698 papers presented at the “16th
European Conference of Fracture,” (ECF16) held in Alexandroupolis, Greece, July 3-7,
2006. The accompanying CD attached at the back cover of the book contains the full
length papers.
The abstracts of the fifteen plenary lectures are included in the beginning of the book.
The remaining 683 abstracts are arranged in 25 tracks and 35 special symposia/sessions
with 303 and 380 abstracts, respectively. The papers of the tracks have been contributed
from open call, while the papers of the symposia/sessions have been solicited by the
respective organizers. Both tracks and symposia/sessions fall into two categories,
namely, fracture of nanomaterials and structures and engineering materials and
structures with 88 and 595 papers, respectively.
Started in 1976, the European Conference of Fracture (ECF) takes place every two
years in a European country. Its scope is to promote world-wide cooperation among
scientists and engineers concerned with fracture and fatigue of solids. ECF16 was under
the auspices of the European Structural Integrity Society (ESIS) and was sponsored by
the American Society of Testing and Materials, the British Society for Stain
Measurement, the Society of Experimental Mechanics, the Italian Society for
Experimental Mechanics, and the Japanese Society of Mechanical Engineers. ECF16
focused in all aspects of structural integrity with the objective of improving the safety
and performance of engineering structures, components, systems and their associated
materials. Emphasis was given to the failure of nanostructured materials and
nanostructures and micro and nanoelectromechanical systems (MEMS and NEMS). The
technical program of ECF16 was the product of hard work and devotion of more than
150 world leading experts to whom I am greatly indebted. The success of ECF16 relied
solely on the dedication and titanic work of the members of the Scientific Advisory
Board, the pillars of ECF16. As chairman of ECF16 I am honored to have them on the
Board and have worked closely with them for a successful conference.
Fracture mechanics analysis has been successful for many years in the prevention of
failures of engineering materials and structures. It is based on the realistic assumption
that all materials contain crack-like defects from which failure initiates. New
technological developments, however, raise new challenges for fracture mechanics
research and development. Quasi-brittle materials including concrete, cement pastes,
rock, soil, etc. are being extensively used in engineering applications. Layered materials
and especially thin film/substrate systems are becoming important in small volume
systems used in micro and nanoelectromechancial systems (MEMS and NEMS).
Nanostructured materials are being introduced in our every day life. In all these problems
fracture mechanics plays a major role for the prediction of failure and safe design of
materials and structures. Failure of materials and structures at the micro and nano scale
levels are adequately addressed at ECF16 with 93 papers referred to in this area.
xliv Editor’s Preface

More than nine hundred participants attended ECF16, while more than eight hundred
fifty papers were presented, far more than any other ECF over a thirty year period. The
participants of ECF16 came from 49 countries. Roughly speaking 66% came from
Europe, 17% from the Americas, 8% from the Far East and 9% from other countries. I
am happy and proud to have welcomed in Alexandroupolis well-known experts who
came to discuss problems related to the analysis and prevention of failure in structures.
The tranquility and peacefulness of this small town provided an ideal environment for a
group of scientists and engineers to gather and interact on a personal basis. Presentation
of technical papers alone is not enough for effective scientific communication. It is the
healthy exchange of ideas and scientific knowledge, formal and informal discussions,
together with the plenary and contributed papers that make a fruitful and successful
meeting. Informal discussions, personal acquaintance and friendship play an important
role.
I am proud to have hosted ECF16 in the beautiful town of Alexandroupolis, site of
the Democritus University of Thrace and I am pleased to have welcomed colleagues,
friends, and old and new acquaintances.
I very sincerely thank the authors who have contributed to this volume, the symposia/
sessions organizers for their hard work and dedication and the referees who reviewed the
quality of the submitted contributions. Our sponsors’ support, give in various forms, is
gratefully acknowledged. The tireless effort of the members of the Organizing
Committee as well as of other numerous individuals, and people behind the scenes is
appreciated. I am deeply indebted to the senior students of the Department of Electrical
and Computer Engineering of the Democritus University of Thrace Messrs. N.
Tsiantoulas and S. Siailis for their hard work and dedication in the preparation of the
ECF16 website in a timely and efficient manner and the organization of the conference,
and for their efforts in helping me compile this volume. Finally, a special word of thanks
goes to Mrs. Nathalie Jacobs of Springer for the nice appearance of this book and her
kind and continuous collaboration and support.

January 2006 Emmanuel E. Gdoutos

Xanthi, Greece Editor


ORGANIZING COMMITTEES

Scientific Advisory Board

Emmanuel E. Gdoutos (Chairman)

Track 1 (Nanomaterials and Nanostructures)


Awaji, H. (Japan), Bahr, D., (USA), Ballarini, R., (USA), Batra, R. (USA), Belytschko,
T. (USA), Berndt, C. (USA), Bhushan, B. (USA), Espinosa, H. (USA), Friedrich, K.
(Germany), Karimi, A. (Switzerland), Kouris, D. (USA), Lagoudas, D. (USA), Meletis,
E.I. (USA), Michel, B. (Germany), Moody, N. (USA), Plumbridge, W.J. (UK),
Pluvinage, G. (France), Ruoff, R. (USA), Sih, G.C. (China), Zhang, Z. (Germany), Zhou,
M. (USA).

Track 2 (Engineering Materials and Structures)


Akid, R. (UK), Aliabadi, M.H. (UK), Andrianopoulos, N. (Greece), Angelova, D.
(Bulgaria), Aravas, N. (Greece), Atkins, A.G. (UK), Banks-Sills, L. (Israel), Bartolozzi,
F. (Italy), Barton, J. (UK), Bartzokas, D. (Greece), Bathias, C., (France), Bazant, Z.
(USA), Beaumont, P. (UK), Beretta, S. (Italy), Blackman, B. (UK), Brocks, W.
(Germany), Bunch, J. (USA), Cacko, J. (Slovakia), Carpinteri, Alberto (Italy),
Carpinteri, Andrea (Italy), Chona, R. (USA), Daniel, I.M. (USA), Danzer, R. (Austria),
Dietzel, W. (Germany), Dover, B. (UK), Exadaktylos, G. (Greece), Fernando, U.S.
(UK), Ferro, G. (Italy), Finnie, I. (USA), Fischer, F.D. (Austria), Fleck, N. (UK), Freitas,
M.J.M. (Portugal), Gabetta, G. (Italy), Galiotis, C. (Greece), Georgiadis, H.G. (Greece),
Goldstein, R. (Russia), Gosz, M. (USA), Gubeljack, N. (Slovenia), Hawong, J.S.
(Korea), Hedrich, K. (Republic of Serbia), Hopkins, S. (USA), Ingraffea, A. (USA),
Isogimi, K. (Japan), Jelaska, D. (Croatia), Jirasek, M. (USA), Johnson, E. (Sweden),
Jones, R. (Australia), Kalluri, S. (USA), Kalthoff, J.F. (Germany), Karalekas, D.
(Greece), Karihaloo, B. (UK), Kassner, M. (USA), Kermanidis, Th. (Greece), Kienzler,
R. (Germany), Kocanda, D. (Poland), Konsta-Gdoutos, M. (Greece), Kostopoulos, V.
(Greece), Kourkoulis, S. (Greece), Landes, J. (USA), Lee, B-L (USA), Lee. O.S.
(Korea), Lee, S.B. (Korea), Liolios, A. (Greece), Liu, C.T. (USA), Luxmoore, A.R.
(UK), MacGillivray, H. (UK), Mai, Y.W. (Australia), Maier, H.J. (Germany), Marioli-
Riga, Z. (Greece), Markenskoff, X. (USA), Massabo, R. (Italy), Matczynski, M.
(Poland), Matysiak, S. (Poland), Mayer, H.R. (Austria), McEvily, A. (USA),
Michopoulos, J. (USA), Mines, R. (UK), Mitchell, M. (USA), Moscovic, R. (UK),
Murakami, Y. (Japan), Needleman, A. (USA), Neimitz, A. (Poland), Neu, R. (USA),
Nied, H. (USA), Nilsson, F. (Sweden), Nurse, A. (UK), Nykyforchyn, H. (Ukraine),
Paipetis, S.A. (Greece), Panoskaltsis, V. (USA), Pantelakis, S. (Greece), Papadopoulos,
G. (Greece), Papakaliatakis, G. (Greece), Pappalettere, C. (Italy), Patterson, E. (UK),
Pavan, A. (Italy), Petit, J. (France), Pineau, A. (France), Pokluda, J. (Czech Republic),
xlvi Organizing Committees

Pook, L. (UK), Prakash, V. (USA), Prassianakis, J. (Greece), Rajapakse, Y.D.S. (USA),


Ravi-Chandar, K. (USA), Ravichandran, R. (USA), Rodopoulos, Chr. (UK), Rosakis, A.
(USA), Rossmanith, H.P. (Austria), Saxena, A. (USA), Sciammarella, C. (Italy),
Sedmak, S. (Republic of Serbia), Shah, S. (USA), Shukla, A. (USA), Soboyejo, W.
(USA), Sotiropoulos, D. (Greece), Spyropoulos, C. (Greece), Stanzl-Tschegg, S.
(Austria), Staszewski, W.J. (UK), Steen, M. (Netherlands), Subhash, G. (USA),
Sumarac, D. (Republic of Serbia), Sun, C.T. (USA), Sutton, M. (USA), Tamuzs, V.
(Latvia), Taylor, D. (Ireland), Theotokoglou, S. (Greece), toor, P. (USA), toth, L.
(Hungary), Tsamasphyros, G. (Greece), Tvergaard, V. (Denmark), Unger, D. (USA),
van Mier, J.G.M. (Switzerland), Vardoulakis, Y. (Greece), Vodenicharov, S. (Bulgaria),
Wallin, K. (Finland), Wardle, G. (UK), Williams, J.G. (UK), Withers, P. (UK), Yates, J.
R. (UK), Youtsos, A.G. (Netherlands), Zacharopoulos, D.A. (Greece).

LOCAL ORGANIZING COMMITTEE

Emmanuel E. Gdoutos (Chairman)

Z. Adamidou, P. Kalaitzidis, M.S. Konsta-Gdoutos, G. Papakaliatakis, S. Sailis, N.


Tsiantoulas, D.A. Zacharopoulos
ECF16 TRACKS

B: TRACKS

B1: Nanomaterials and Nanostructures


1T1. Fracture and Fatigue of Nanostuctured Materials
1T2. Failure Mechanisms
1T4. Fatigue and Fracture of MEMS and NEMS
1T7. Thin Films
1T9. Failure of Nanocomposites

B2: Engineering Materials and Structures


2T1. Physical Aspects of Fracture
2T2. Brittle Fracture
2T3. Ductile Fracture
2T4. Nonlinear Fracture Mechanics
2T5. Fatigue and Fracture
2T8. Polymers, Ceramics and Composites
2T11. Fracture Mechanics Analysis
2T13. Probabilistic Approaches to Fracture Mechanics
2T14. Computational Fracture Mechanics
2T15. Experimental Fracture Mechanics
2T16. Creep Fracture
2T17. Environment Assisted Fracture
2T18. Dynamic, High Strain Rate, or Impact Fracture
2T19. Damage Mechanics
2T21. Concrete and Rock
2T22. Sandwich Structures
2T23. Novel Testing and Evaluation Techniques
2T26. Structural Integrity
2T28. Mesofracture Mechanics
2T32. Micromechanisms in Fracture and Fatigue
ECF16 SPECIAL SYMPOSIA/SESSIONS

C: SPECIAL SYMPOSIA/SESSIONS

C1: Nanomaterials and Nanostructures


1. Fracture and Fatigue at the Micro and Nano Scales (Organized by H.D. Espinosa and
I.M.Daniel)
3. Nanoscale Deformation and Failure (Organized by M. Zhou)
29. Reliability and Failure Analysis of Electronics and Mechanical Systems (O.S. Lee)
31. Multiscaling in Molecular and Continuum Mechanics – Scaling in Time and Size from Macro
to Nano (Organized by G.C. Sih)
34. Cracks in Micro- and Nanoelectronics (Organized by B. Michel)
43. Interfacial Fracture in Composites and Electronic Packaging Materials (Organized by C.T. Sun
and T. Ikeda)

C2: Engineering Materials and Structures


4. Fracture and Fatigue of Elastomers (Organized by C. Bathias and E. Bayraktar)
5. Integrity of Dynamical Systems (Organized by K. Hedrih)

8. Modelling of Material Property Data and Fracture Mechanisms (Organized by R. Moskovic)


9. Micromechanisms in Fracture and Fatigue (Organized by J. Pukluda and R. Pippan)
12. Interface Fracture and Behavior of Joints (Organized by L. Banks-Sills)
13. Computational Fracture Mechanics (Organized by T. Belytschko and A. Gravouil)
14. Cohesive Models of Fracture (Organized by W. Brocks)
16. Environment Assisted Fracture (Organized by G. Gabetta, W. Dietzel and H. Nykyforchyn)
17. SIM, Philosophy, Instrumentation and Analysis (Organized by W. D. Dover)
18. Fracture of Biomaterials (Organized by J. tong)
19. Structural Integrity Assessment in Theory and Practice (Organized by S. Vodenitsarov and S.
Sedmak)
20. Critical Distance Theories of Fracture (Organized by D. Taylor)
22. New Investigations on Very High Cycle Fatigue of Materials (Organized by H. Mayer and S.
Stanzl-Tschegg)
23. Deformation and Fracture of Engineering Materials (Organized by C.T. Liu)
24. Materials Damage Prognosis and Life Cycle Engineering (Organized by R. P. Wei, G. Harlow,
A. Ingraffea and J. Larsen)
25. Mixed-Mode Fracture (Organized by M. Gosz)
l ECF16

26. Fracture Mechanics Characterization of Wood (Organized by S. Stanzl-Tschegg)


28. Short Fatigue Crack Growth under Multi-axial Loading Conditions (Organized by Y.
Murakami and A.J. McEvily)
30. Integrity of Gears (Organized by D. Jelaska)
33. Fracture and Failure of Natural Building Stones Applications in the Restoration of Ancient
Monuments (Organized by S. Kourkoulis)
35. High Temperature and Thermomechanical Fatigue (Organized by R.W. Neu, S. Kalluri and
H.J. Maier)
36. Impact Failure of Laminated and Sandwich Composite Structures (Organized by R. Mines)
37. Mesofracture and Transferability (Organized by G. Pluvinage)
38. Damage in Composites - Damage Development in Composite Materials & Structures - Models
of Prediction (Organized by C. Galiotis)
39. Aging Aerostructures (Organized by S. Pantelakis)
40. Residual Stress and its Effects on Fatigue and Fracture (Organized by A.G. Youtsos and P.J.
Withers)
41. Computational Modeling of Multiphysics Degrading Systems (CMMDS), (Organized by J.
Michopoulos)
42. Scaling and Size Effects (Organized by Z.P. Bazant and M. Jirasek)
44. Multiple Cracking and Delamination (Organized by R. Goldstein and R. Massabo)
A. INVITED PAPERS
Invited Papers 3

DEFORMATION AND FRACTURE AT THE MICRON AND NANO SCALES

E. C. Aifantis
Laboratory of Mechanics and Materials, Polytechnic School, Aristotle University of Thessaloniki,
54124, Thessaloniki, Greece
mom@mom.gen.auth.gr

The progressively increasing demands of new science and technology to understand the behavior
of materials/components and processes at the micrometer and nanometer regimes has led in the
mid seventies to the development of micromechanics. In the mid nineties a new term
nanomechanics was used by the author to indicate the forthcoming excessive activity in this field
and point out to the need for new constitutive equations and mechanics tools to be developed in
relation to the emerging fields of nanotechnology. In fact, it was only a few years earlier, that the
first carbon nanotubes were produced in Japan - a unique example for the use of elasticity theory
at the nanoscale - and the first bulk nanopolycrystals were produced in Russia. At the same time
the first experimental observations on deformation and fracture mechanisms of nanopoly-
crystalline thin films were reported by the author and his co-workers in US. It was reported, among
other things, that plastic deformation at the nanoscale does not take place through lattice
dislocation activity but through grain boundary processes including material rotation and mass
diffusion. Moreover, fracture processes occur through nanovoid nucleation and coalescence. Some
of these experimental observations were numerically verified a few years later through molecular
dynamics (MD) multimillion atom simulations.
A first attempt to develop constitutive equations for describing deformation and fracture at the
nanoscale is outlined by the author and co-workers with two specific concepts being advanced: the
use of a mixture argument for “bulk” and “grain boundary” states and the resort to non-locality for
describing the state of stress and strain at the nanoscale. In very recent years, due to the most
promising developments in nanosciences/nanotechnologies in conjunction with the rapidly
evolved computational advances, coupled ab initio/atomistic/molecular dynamics/finite element
calculations have been employed to simulate the mechanical response of matter at the nanoscale in
thin film and bulk configurations within the so-called multiscale modeling approach. A variety of
nano-objects, including multilayered films, nanocomposites, proteins, as well as other metal or
non-metal and biological nanostructures are modeled within such multiscale modeling framework
and new nano-testing procedures and nano-apparatuses have been developed to capture this
response experimentally. HRTEM, STM, AFM, micro/nano tensile machines and micro/nano
indenters are among some of the new experimental tools for probing the mechanical response of
materials and structures at the micro/nano scale and designing MEMS/NEMS devices for a variety
of electromechanical and biomedical applications.
In an effort to have a general micro/nano mechanics framework as useful as the continuum
mechanics model that has been employed so successfully for the understanding of the mechanical
behavior at the macroscale and the design of macroscopic components and structures, a proposal
will be presented by a straightforward extension/generalization of the macroscopic continuum
model. This generalization is based on the concept of a micro/nano continuum which is capable of
exchanging mass, momentum and energy with its bounding surface. While such a model has been
suggested by the author more than twenty five years ago, it was not until recently that its
implications to elasticity, plasticity, and other continuum theories of structural defects was fully
explored. The two new basic ingredients of the model are: a) the appearance of deterministic
higher order spatial and time derivatives in the governing equations of mechanical fields; and b)
4 E. C. Aifantis

the appearance of stochastic terms due to random effects associated with the nucleation and
evolution of deformation events at the micron/nano scale regime.
Several benchmark problems are considered to illustrate the applicability of the proposed
framework, as follows: (i) Elimination of Singularities. Gradient elasticity models are shown to
eliminate the strain and stress singularities form dislocation/disclination lines and crack tips. On
the basis of these solutions new relations can be obtained for the strength, energy and interaction of
defects in nanocrystals and new fracture criteria can be derived at the nanoscale. (ii) Internal
Stress, Elastic Constants and Yield Strength: Size Effects. Gradient elasticity and gradient
plasticity are shown to produce new formulas for the determination of internal stress and elastic
moduli in micro/nano multilayers and micro/nano plates under bending. A modification of the
well-known Stoney formula is an example. Also the dependence of the elastic modulus on the
nanoplate thickness is another interesting example. In this connection, it is pointed out that the size
dependence of yield strength of micro/nano columns and the dependence of Young’s modulus/
failure stress on nanotube diameter has also been documented. (iii) Micro/Nano Indentation.
Various basic formulas that have been used for determining material properties at the macroscale
during indentation are revisited by employing gradient elasticity/plasticity with or without
stochastic terms. Displacement bursts, load-depth serrations, and size-dependent hardness are all
phenomena that are often observed during micro/nano indentation and their proper interpretation
can assist in the determination of deformation and fracture properties at these scales. (iv)
Localization of Deformation and Multiple Shear Banding. The interesting features of deformation
and fracture at the micro/nano scale are concerned with the determination of the critical grain sizes
where a plasticity transition mechanism takes place. At the nanoscale (1-100 nm) the critical grain
size determines the transition from grain rotation/sliding to massive dislocation motion, which
often manifests itself by the appearance of an inverse Hall-Petch behavior. At the ultra-fine grain
size regime (100-1000 nm) another plasticity mechanism that occurs is the so-called multiple shear
banding which often manifests itself by the appearance of a perfectly plastic behavior in the
corresponding stress-strain curve. These two plasticity mechanism transitions will be discussed
within the proposed unified material micro/nano mechanics framework.
In concluding, it will be pointed out why new techniques such as fractals, wavelets and time-
series are often necessary for capturing details and additional features of micro/nano deformation
and fracture. Information on some of the above topics can be found in references [1]-[3] and
articles quoted therein.

References
1. Aifantis, E.C., In Recent Advances in Applied Mechanics (Honorary Volume for Academician
A.N. Kounadis), edited by T. Katsikadelis, D.E. Beskos and E.E. Gdoutos, NTUA, Athens,
2000, 243-254.
2. Aifantis, E.C., Mech. Mat., vol. 35, 259-280, 2003.
3. Konstantopoulos, I., Tragoudaras, D., Mokios, G., Konstantinidis, A., Zaiser, M. and
Aifantis, E.C., Research in Progress.
Invited Papers 5

STATISTICAL MECHANICS OF SAFETY FACTORS AND SIZE EFFECT IN


QUASIBRITTLE FRACTURE

Z. P. Bazant and S.-D. Pang


McCormick Institute Professor and W.P. Murphy Professor of Civil Engineering
and Materials Science, Northwestern University,
Graduate Research Assistant
z-bazant@northwestern.edu

Throughout most of the 20th century, it was widely believed that the size effect on structural
strength has a purely statistical origin, explained by extreme value statistics based on the weakest
link model, and described by Weibull statistical theory of random strength. However, beginning
with the first suggestions made already in the early 1970s, it gradually transpired that, in
quasibrittle materials (i.e. heterogeneous brittle materials with a non-negligible fracture process
zone), the mean size effect is essentially deterministic, stemming from energy release caused by
stress redistribution in a structure prior to maximum load. The quasibrittle energetic scaling
bridges three simple asymptotic power-law scalingsthose of linear elastic fracture mechanics,
plasticity, and Weibull theory. Renormalization group transformation does not suffice to handle
the transitional nature of this quasibrittle size effect, often spanning several orders of magnitude of
size. As is now widely accepted, quasibrittle materials including concrete, rock, tough ceramics,
sea ice, snow slabs and composites exhibit major size effects on the mean structural strength that
are largely or totally deterministic in nature, being caused by stress redistribution and energy
release associated with stable propagation of large fractures or with formation of large zones of
distributed cracking.
The lecture begins by reviewing the general asymptotic properties of size effect implied by the
cohesive crack model or crack band model, and highlights the use of asymptotic matching
techniques as a means of obtaining scale-bridging size effect laws representing a smooth transition
between two power laws. Asymptotic matching is a range of diverse techniques widely used in
fluid mechanics, but overlooked in solid mechanics. Presented is a method of asymptotic matching
which is suitable for structural strength problems. The method is based on power series expansion
of the governing equation written as a function of dimensionless variables. The key idea is to
choose these variables in such a way that, at each asymptotic state, all of them vanish except one.
Attention is then focused on the size effects observed in fiber-polymer composites failing
either by tensile fracture or by propagation of compression kink bands with fiber micro-buckling.
The size effects in polymeric foams and sandwich structures are also discussed. Nonlocal
probabilistic analysis of the size effect on the statistical distribution of nominal strength of
structures is outlined and discussed from the viewpoint of the extreme value statistics. Implications
for the design of hulls, bulkheads, decks, masts and antenna covers for very large ships, and for the
design of large load-bearing aircraft fuselage panels, are pointed out.
The problem of estimating loads of extremely small failure probability, such as 10-7, required
for design is investigated. Attention is focused on the type 1 size effect, occurring in structures
failing at crack initiation, which is the only type for which material randomness affects not only the
scatter but also the mean of nominal strength. It is shown that a reform of structural reliability
concepts is necessary because of a strong effect of structure size (or brittleness) on failure
probability in the far-out tail of the cumulative probability distribution function (cdf) of structural
strength. The cdf is modeled by a chain of representative volume element (RVE) of the material,
each of which is represented by a hierarchical hybrid series-parallel coupling model. Each micro-
6 Z. P. Bazant and S.-D. Pang

element of this model simulates one of the micro-bonds within the RVE. From Maxwell-
Boltzmann distribution of thermal energies of atoms and the effect of applied stress on the
activation energy, it is deduced that the left tail of cdf of a RVE (for failure probabilities < 0.0001
to 0.01) must be a power law, while there must be a broad Gaussian core because of parallel
couplings within the RVE. The amplitude of the power law tail is obtained as a function of
temperature and load duration.
A chain-of-RVEs model is proposed to model a gradual transition of strength cdf from a
mostly Gaussian pdf with a short Weibull tail for small structures to a purely Weibull cdf for very
large structures. The equivalent number of RVEs for which the chain represents the cdf of strength
of a structure with nonuniform stress field is expressed in terms of Weibull integral according to
nonlocal Weibull theory. The center of the Weibull-Gaussian transition moves along the cdf from
left to right as a function of the equivalent number Neq of RVEs in the chain-of-RVEs model.
Matching of this model serves to calibrate a smooth analytical expression for the transitional cdf,
varying from purely Gaussian for zero size (or zero brittleness) to purely Weibull for infinite size
(or perfect brittleness) as a function of the structure size as well as geometry. The distance from the
mean to a point of a tolerable failure probability such as 10-7 or 10-6 on this transitional cdf is
shown to be strongly size and geometry dependent, and nearly doubles while passing from very
small to very large structures. To capture this major effect, it is necessary to introduce a correction
into Cornell's and Hasofer-Lind's reliability indices (known from the classical first-order reliability
method, or FORM).
To make reliability assessments realistic, it is further necessary that the 'covert' understrength
factors implied in brittle failure provisions of concrete design codes be made overt, and that a
'covert' size effect implied by excessive load factor for self-weight acting alone be eliminated.
Adaptation of the stochastic finite element method to cope with extreme value statistics of
energetic-statistical size effect is described, and its importance is demonstrated by analysis of some
famous disasters, particularly Malpasset Dam. To improve design safety and efficiency, experts in
statistical reliability and fracture mechanics will need to collaborate to tackle these problems in a
comprehensive manner
To improve design safety and efficiency, experts in statistical reliability and fracture
mechanics will need to collaborate to tackle these problems in a comprehensive manner

References
1. Bažant, Z.P., "Scaling theory for quasibrittle structural failure." Proc., National Academy of
Sciences, 101 (37), 13397-13399, 2004.
2. Bažant, Z.P., Scaling of Structural Strength. Hermes Penton Science, London, 2002. (French
translation, Hermes, Paris 2004); 2nd ed., Elsevier 2005, in press.
3. Bažant, Z.P., "Probability distribution of energetic-statistical size effect in quasibrittle
fracture." Probabilistic Engineering Mechanics, 19(4), 307-319, 2004.
Invited Papers 7

“NANORELIABILITY” – FRACTURE MECHANICS ON THE WAY FROM


MICRO TO NANO

B. Michel
Fraunhofer Micro Materials Center at IZM Berlin
Gustav-Meyer-Allee 25, D-13355 Berlin, Germany
bernd.michel@izm.fraunhofer.de

“Thermo-mechanical compatibility” (TMC) of different micro- and nanomaterials in electronic


and mechatronic systems is one mayor reason for troubles in the field of reliability and life-time of
“high-tech” applications. Problems of thermal misfit (or mismatch) have become more and more
important in the recent years because more than 60 percent of failure events in modern
microelectronics are more or less directly connected with thermal misfit problems. The interface
regions between the different materials, e.g. of a chip interconnection layer within a microsensor or
a microactuator, is very important for the reliability of the component and of the whole system.
Interface cracks, therefore, have to be dealt with in details in modern microsystem and
nanotechnology as well. The author is going to present a survey on modern crack, fracture and
reliability concepts of Microsystems in the chip interconnection region, where a lot of very
different materials are “playing the concert”. Polymers, metals, ceramic materials and composites
of very different kind will be shown to excert their specific influence on the local deformation and
crack behaviour. Fatigue, creep and moisture effects and their complicated interactions have to be
taken into account for reliable lifetime prognoses.
The author presents some new experimental techniques based on modern digital image
correlation techniques (DIC) which enable to take into account local and global effects as well.
The so-called microDAC and nanoDAC deformation analyses provide very good means for local
crack field analysis. These methods are combined with numerical field calculation and advanced
reliability and lifetime concepts and lead to very good results for time-dependent failure analysis of
Microsystems. In electronic packaging of MEMS and sensor components nanomechanics effects
have been shown to become more and more important. The author presents recent results of his
group to include focus ion beam technique (FIB), AFM analysis and advanced methods of
materials testing (e.g. nanoindentation, nanoDMA, local stress and strain analysis) into the crack
evaluation procedure in the micro-nano interface regions. Besides the above mentioned techniques
the author also is going to outline his opinion about the perspectives of modern fracture mechanics
in the field of micro- and nanotechnologies applied to the electronics, automotives and above all IT
branches.

References
1. Michel, B., Testing at Micro and Nanoscale, EuroSIME, European Conf. On Thermal,
Mechanical and Multiphysics Simulation and Experiments in Microelectronics, Berlin, 18-20
April 2005.
2. Michel, B., Keller, J., NanoDAC – A New Technique for Micro- and Nanomechanical
Reliability Analysis of Lead-free Solder Interconnects, Int. Conf. on Lead-free Soldering,
Toronto, Canada, 24-26 May 2005.
3. Walter, H., Dudek, R., Michel, B., Fracture and Fatigue Behaviour of MEMS Related Micro
Materials, Int. Conf. on Fracture (ICF 11), Turin, Italy, 20-25 March 2005.
8 B. Michel

4. Michel, B., Fracture Electronics and Thermo-Mechanical Compatibility (TMC) of


Microcomponents in High-Tech Systems, Int. Conf. Micro Materials 2000, Berlin, 17-19
April 2000.
5. Michel, B., Experimental Mechanics on the Way from Micro to Nano, Exp. Technique 29
(2005) 2, 3-5.
Invited Papers 9

FRACTURE MECHANICS AND COMPLEXITY SCIENCES

A. Carpinteri and S. Puzzi


Politecnico di Torino, Department of Structural and Geotechnical Engineering
Corso Duca degli Abruzzi 24, 10129 Torino, Italy
alberto.carpinteri@polito.it

The so-called Complexity Sciences are a topic of fast growing interest inside the scientific
community. Actually, researchers did not come to a definition of complexity, since it manifests
itself in so many different ways [1]. This field itself is not a single discipline, but rather a
heterogeneous amalgam of different techniques of mathematics and science. In fact, under the
label of Complexity Sciences we comprehend a large variety of approaches: nonlinear dynamics,
deterministic chaos theory, nonequilibrium thermodynamics, fractal geometry, intermediate
asymptotics, complete and incomplete similarity, renormalization group theory, catastrophe
theory, self-organized criticality, neural networks, cellular automata, fuzzy logic, etc.
Complex systems lie somehow in between perfect order and complete randomness –the two
extreme conditions that are likely to occur only very seldom in nature– and exhibit one or more
common characteristics, such as: sensitivity to initial conditions, pattern formation, spontaneous
self-organization, emergence of cooperation, hierarchical or multiscale structure, collective
properties beyond those directly contained in the parts, scale effects.
Aim of this paper is to provide an insight into the role of complexity in the field of Materials
Science and Fracture Mechanics. The included examples will be concerned with the snap-back
instabilities in the structural behaviour of composite structures (Carpinteri [2-3]), the occurrence of
fractal patterns and self-similarity in material damage and deformation of heterogeneous materials,
and the apparent scaling on the nominal mechanical properties of disordered materials (Carpinteri
[4,5]). Further examples will deal with criticality in the acoustic emissions of damaged structures
and with scaling in the time-to-failure (Carpinteri et al. [6]). Eventually, results on the transition
towards chaos in the dynamics of cracked beams will be reported (Carpinteri and Pugno [7]).

References
1. Garrido, M.S. and Vilela Mendes, R., Complexity in physics and technology, World
Scientific, Singapore, 1992.
2. Carpinteri, A., In Application of Fracture Mechanics to Cementitious Composites
(Proceedings of a NATO Advanced Research Workshop, Evanston, USA, 1984), edited by
S.P. Shah, Martinus Nijhoff Publishers, Dordrecht, 1985, 287-316.
3. Carpinteri, A., J. Mech. Phys. Solids, vol. 37, 567-582, 1989.
4. Carpinteri, A., Mech. Mater., vol. 18, 89-101, 1994.
5. Carpinteri, A., Int. J. Solids Struct., vol. 31, 291-302, 1994.
6. Carpinteri, A., Lacidogna, G. and Pugno, N., In Fracture Mechanics of Concrete and
Concrete Structures (Proceedings of the 5th International FraMCoS Conference, Vail,
Colorado, USA, 2004), edited by V.C. Li et al., 2004, vol. 1, 31-40.
7. Carpinteri, A. and Pugno, N., J. Appl. Mech., in print.
Invited Papers 11

FAILURE OF COMPOSITE MATERIALS

I. M. Daniel
Robert R. McCormick School of Engineering and Applied Science
Northwestern University, Evanston, IL 60208, USA
imdaniel@northwestern.edu

The failure of composites has been investigated extensively from the micromechanical and
macromechanical points of view. On the micromechanical scale, failure mechanisms and processes
vary widely with type of loading and are intimately related to the properties of the constituent
phases, i. e., matrix, reinforcement, and interface-interphase. Failure predictions based on
micromechanics, even when they are accurate with regard to failure initiation at critical points, are
only approximate with regard to global failure of a lamina and failure progression to ultimate
failure of a multi-directional laminate. For these reasons a macromechanical approach to failure
analysis is preferred.
Numerous failure theories have been proposed and are available to the composite structural
designer[1].They are classified into three groups, limit or noninteractive theories (maximum
stress, maximum strain); interactive theories (Tsai-Hill, Tsai-Wu); and partially interactive or
failure mode based theories (Hashin-Rotem, Puck). The validity and applicability of a given
theory depend on the convenience of application and agreement with experimental results. The
plethora of theories is accompanied by a dearth of suitable and reliable experimental data, which
makes the selection of one theory over another rather difficult. Considerable effort has been
devoted recently to alleviate this difficulty. The problem can be divided in two parts, one being the
prediction of failure of a single lamina and the second dealing with prediction of first-ply-failure
and damage progression leading to ultimate failure of a multi-directional laminate.
C. T. Sun [2] reviewed six failure theories and showed comparisons of theoretical predictions
with experimental results for six different composite material systems and various loading
conditions. A round robin exercise was initiated by Hinton, Soden, and Kaddour for the purpose of
assessing the predictive capabilities of current failure theories [3]. The difficulty in evaluating
failure theories is much greater in the case of a multidirectional laminate. The scope of the
proposed laminate failure analysis comprises the following [1]:
1 A selected or adapted lamina failure theory for prediction of failure initiation, i.e., first-ply
failure (FPF) in the laminate.
2 A failure mode discrimination rule and a scheme of ply discounting and failure progression
in the laminate after FPF.
3 A criterion or definition of ultimate laminate failure (ULF).
In general, a wide variation has been observed in the prediction of laminate failures by the
various theories. The divergence in the predictions is greater for FPF than for ULF; also is greater
for matrix dominated failures than for fiber dominated ones. The divergence observed may be
attributed primarily to the following factors:
1 The different ways in which curing residual stresses are introduced in the predictions,
especially in the case of first-ply-failure.
2 The concept of in-situ behavior of a lamina within the laminate which is still debated.
3 The different methods of modeling the progressive failure process and the definition of
ultimate laminate failure.
12 I. M. Daniel

4 The nonlinear behavior of matrix-dominated laminates, e.g., angle-ply laminates.


Under uniaxial loading of a laminate the deciding factor in predicting ultimate failure is
whether it is fiber or matrix dominated. In the case of matrix dominated angle-ply laminates,
predictions by the limit or interactive theories are not usually in agreement with each other and
with experimental results. Failure is governed by the lamina transverse normal stress V 2 and the
in-plane shear stress W 6 . When V 2 ! 0 , as in the case of [r45]s and the limit theories predict
higher strengths in agreement with the experiment [1]. When V 2  0 , as in the case of the
[r20]2 s laminate under tension, the Tsai-Wu criterion comes closer to the experimental results
[1]. In the more general cases of biaxial loading it is not easy to establish fiber or matrix
dominance in failure as that varies with the loading biaxiality.
In view of the multitude of failure theories, the divergence of their predictions and the lack of
definitive general conclusions regarding their applicability, a practical approach is recommended
as follows [1]:
1 Select a classical representative theory from each category, i.e., non-interactive (maximum
stress), fully interactive (Tsai-Wu), and partly interactive (Hashin-Rotem).
2 Compute and plot stress-strain relations of the laminate under representative mechanical
and hygrothermal loading.
3 Use a newly proposed failure mode discrimination rule and define ULF
4 Compute safety factors for FPF and ULF and compute and plot failure envelopes for the
selected failure theories for the two failure levels (FPF and ULF).
5 Select prediction according to degree of conservatism desired. For the most conservative
approach, limit the state of stress (loading) to within the common domain of the selected
failure envelopes.
All computations and plots can be performed by a newly developed computer program [4]. The
approach above is adequate for conservative structural design. More sophisticated theories and
approaches exist as discussed before incorporating nonlinear behavior and in-situ effects.

References
1. I.M. Daniel and O. Ishai, Engineering Mechanics of Composite Materials, Second Edition,
Oxford University Press, 2005.
2. C.T. Sun, “Strength Analysis of Unidirectional Composites and Laminates,” in
Comprehensive Composite Materials, ed. by A. Kelly and C. Zweben, Ch. 1.20, Elsevier
Science, Ltd., Oxford, UK, 2000.
3. M.J. Hinton, P.D. Soden, and A.S. Kaddour, Failure Criteria in Fibre-Reinforced-Polymer
Composites, Elsevier, Oxford, 2004.
4. J.J. Luo and I.M. Daniel, “Webcomp: Stress and Failure Analysis of Laminate Composites,”
http:www.composites.northwestern.edu/awebcomp, 2004.
Invited Papers 13

INTERACTIONS OF CONSTRAINED FLOW AND SIZE SCALE ON


MECHANICAL BEHAVIOR

W. W. Gerberich, W. M. Mook, M. J. Cordill and D. Hallman


Chemical Engineering and Materials Science Department
University of Minnesota
421 Washington Ave SE
Minneapolis, MN 55455
wgerb@umn.edu

Coupled effects between constrained flow, increased strength as a function of decreased sample
size, and resulting high stresses affect both modulus and fracture toughness. For submicron size
crystalline spheres [1,2], boxes [3], and cubes [4], we have recently shown that dislocation by
dislocation events can be followed using a combination of AFM/nanoindentation. This has led to
at least three proposed strengthening mechanisms for hardening of small constrained volumes
under compression[4]. With the increased stresses, this can produce increased moduli of elasticity
in confined volumes small in three dimensions. With increased constrained plasticity this
produces increased strength in volumes small in three, two or one dimensions.
Hardening mechanisms for the gigapascal strengths observed in small volumes are addressed.
These are explored for Si and Ti nanospheres as well as Ni, Co, and Ni80Fe20 (permalloy) thin
films in the 10-40 nm particle radius or thin film thickness[4,5] regime. Because of trapped
dislocations between the upper and lower loading surfaces being typically diamond or sapphire,
strengths approaching theoretical are often reached with deformation of only tens of nanometers or
less. Of course, this is highly dependent upon the length scale of the confining structure. For
oxide-covered nanospheres, the length scale can be taken as the volume to contact surface area as
given elsewhere[5] by (2/3)r3/a2 where r is the sphere radius and a is the contact radius at the
upper and lower platen. Alternatively, from contact geometry this length scale is (2/3)r2/G. For
deposited thin films, it was previously shown that a measured volume of plastic deformation to
contact surface area was empirically given by Dh2/a where h is the film thickness and a is the
contact radius[6]. These two measures of length scale show that the flow strength of the
nanospheres and the hardness of the thin films scale with G1/2/r and G1/4/h, respectively. Here, G is
the displacement the sphere has been squeezed or the penetration depth of the indenter into the
film. It is significant that both of these show strength to be inversely proportional to the smallest
dimension of the volume. Additionally, it is interesting that the more constrained nanosphere has
strength increasing as G1/2 compared to film hardness, small in only one dimension, increasing as
G1/4.
With this same length scale approach we have shown that the fracture toughness of a
delaminating thin film conforms to R-curve behavior with GR representing the resistance
equivalent to the strain energy release rate at fracture. This has been given by[6]

1/ 2
f
DV ys2 h § ' b ·
G ~
R
¨¨ ¸¸
E © bo ¹ (1)
where Vys/E is the strength to modulus ratio, 'b/bo is the incremental growth to initial crack size
and D is a constant on the order of 10. In the present paper we use the same approach by equating
14 I. M. Daniel

the local deformation length scale to the fracture process zone length scale. This gives the fracture
resistance for the sphere to be

6V ys2 r § t ox · 1 / 2
G RS ~ ¨ ¸ (2)
E ©G ¹
Here tox is the oxide film thickness on the sphere, and G is the vertical displacement the sphere has
been squeezed to the point of fracture. Note that the triggering event is assumed to be a crack
nucleated in the less robust oxide film around the sphere such that the defect size is tox. In both
cases for Eqs. (1) and (2), the leading term is the strain energy density times the smallest
dimension of the constrained volume.
We also discuss relationships such as Eq. (2) in terms of the overall size dependence if Vys
obeys a Hall-Petch type relation giving strength proportional to G1/2/r for the nanosphere. It is
emphasized that this is an evolutionary length scale which decreases with increasing displacement,
i.e., (2/3)r2/G. Incorporating the length scale dependence of yield strength actually gives Gg’
proportional to G1/2/r so that both strength and fracture resistance scale with G1/2/r for very small
volumes. Due to the first term in the Hall-Petch relation dominating at very large volumes for both
thicker films and larger spheres, this trend would be predicted to reverse. That is, at larger
volumes one should find an increase in fracture resistance with scale according to Eqs. (1) and (2)
as Vys becomes more nearly constant.
Most of these relationships are still in their formative stages but have some corroboration with
respect to Cu and Au film delamination studies. Regarding fracture experiments on nanospheres,
we have only recently fractured a silicon particle in situ in the transmission electron microscope.
However, this corroborates some previous indirect analysis using atomic force microscopy based
nanoindentation to determine the fracture toughness of silicon nanospheres with radii in the range
of 20 to 110 nm. In the full paper we will derive the above relationships and discuss increased
strength and fracture resistance of constrained volumes under compression or pressure as having
ramifications to friction, wear, and microelectromechanical systems.

REFERENCES
1. Gerberich, W.W., Cordill, M.J., Mook, W.M., Moody, N.R., Perrey, C.R., Carter, C.B.,
Mukherjee, R., and Girshick, S.L., Acta Mater., vol. 53, 2215-2229, 2005.
2. Gerberich, W.W., Jungk, J.M., Cordill, M.J., Mook, W.M., Boyce, B., Friedmann, T., Moody,
N.R. and Yang, D., Intern. J. Fracture, 2005 (accepted).
3. Mook, W.M., Jungk, J.M., Cordill, M.J., Moody, N.R., Sun, Y., Xia, Y., and Gerberich,
W.W., Z. Metallkd., vol. 95, 416-424, 2004.
4. Cordill, M.J., Chamber, M.M., Hallman, D., Lund, M., Perrey, C.R., Carter, C.B.,
Kortshagen, U. and Gerberich, W.W., “Plasticity Responses in Ultra-Small Confined Cubes
and Films,” 2005 (in preparation).
5. Gerberich. W.W., Mook. W.M., Perrey. C.R., Carter. C.B., Baskes. M.I., Mukherjee. R.,
Gidwani. A., Heberlein. J., McMurry. P.H. and Girshick. S.L., J. Mech. Phys. Solids, vol. 51,
979-992, 2003.
6. Gerberich. W.W., Jungk, J.M., Li, M., Volinsky, A.A., Hoehn, J. W. and Yoder, K., Intern. J.
Fracture vol. 119/120, 287-405, 2003.
Invited Papers 15

SPACE SHUTTLE COLUMBIA POST-ACCIDENT ANALYSIS AND


INVESTIGATION

S. McDanels
NASA
YA-F1 Kennedy Space Center, FL, 32899 USA
steve.mcdanels@nasa.gov

Although the loss of the Space Shuttle Columbia and its crew was tragic, the circumstances offered
a unique opportunity to examine a multitude of components which had experienced one of the
harshest environments ever encountered by engineered materials: a break up at a velocity in
excess of Mach 18 and an altitude exceeding 200,000 feet (63 KM), resulting in a debris field 645
miles/1,038 KM long and 10 miles/16 KM wide. Various analytical tools were employed to
ascertain the sequence of events leading to the disintegration of the Orbiter and to characterize the
features of the debris. The testing and analyses all indicated that a breach in a left wing reinforced
carbon/carbon composite leading edge panel was the access point for hot gasses generated during
re-entry to penetrate the structure of the vehicle and compromise the integrity of the materials and
components in that area of the Shuttle.
The analytical and elemental testing utilized such techniques as X-Ray Diffraction (XRD),
Energy Dispersive X-Ray (EDX) dot mapping, Electron Micro Probe Analysis (EMPA), and X-
Ray Photoelectron Spectroscopy (XPS) to characterize the deposition of intermetallics adjacent to
the suspected location of the plasma breach in the leading edge of the left wing, Fig.1.

FIGURE 1. Micrograph of a Left Wing Carrier Panel Slag Deposit. Cummings [1]

Fractographic and metallographic analyses of several pieces of debris, Fig. 2, were


performed to evaluate such fracture characteristics as broomstrawing and feathering of aluminum
alloys, suspected stress-assisted grain boundary oxidation (SAGBO) of Inconel components, and
intergranular high temperature fracture features observed on several A286 stainless steel spar
fittings from the left wing structure of the Orbiter. Likewise, depositional characteristics such as
composition, directionality and orientation, and the sequence and order of layering, were evaluated
to assist in re-tracing the path of the plasma flow into the wing structure.
16 S. McDanels

FIGURE 2. Interior view of aluminum drag chute canister. Parker [2]

The examination of the debris’ fracture surfaces, and of metallographically-prepared


specimens harvested from the debris, was performed via scanning electron microscope. The
resultant features and characteristics were compared to those of laboratory exemplars of similar
base materials. The observed features, along with the results from the elemental analytical testing,
helped the Space Shuttle Columbia accident investigation team reconstruct the mishap and
determine the sequence of events which ultimately led to the loss of the vehicle.
The debris from the Columbia accident now resides in the Vehicle Assembly Building (VAB)
at the Kennedy Space Center. A portion of the debris is on display, while the majority is in storage
in the VAB. A process is in place whereby universities and professional societies can request
pieces of debris for educational and research purposes, Fig.3.

FIGURE 3. Debris sample loaned out for educational purposes.

References
1. Cummings, V. J., XPS, Metallographic and SEM X-Ray elemental Dot Map Analysis of STS-
107 Debris Sample 24543-1, National Aeronautics and Space Administration, KSC-MSL-
2003-149, 2003.
2. Parker, D. S., Optical and SEM/EDS Analysis of STS-107 Debris Sample 58693-1, National
Aeronautics and Space Administration, KSC-MSL-2003-208, 2003.
Invited Papers 17

THE ROLE OF ADHESION AND FRACTURE ON THE PERFORMANCE OF


NANOSTRUCTURED FILMS

Neville Moody, Megan J. Cordill1, Marian S. Kennedy2, David P. Adams3, David F. Bahr2 and
William W. Gerberich1
Sandia National Laboratories, Livermore, CA 94550
1University of Minnesota, Minneapolis, MN 55455
2Washington State University, Pullman, WA 99164
3Sandia National Laboratories, Albuquerque, NM 87185

nrmoody@sandia.gov

Nanostructured materials are the basis for emerging technologies, such as MEMS, NEMS, sensors,
and flexible electronics, that will dominate near term advances in nanotechnology. These
technologies are often based on devices containing layers of nanoscale polymer, ceramic and
metallic films and stretchable interconnects creating surfaces and interfaces with properties and
responses that differ dramatically from bulk counterparts. The differing properties can induce high
interlaminar stresses that lead to wrinkling, delamination, and buckling in compression [1,2], and
film fracture and decohesion in tension. [3] However, the relationships between composition,
structure and properties, and especially adhesion and fracture, are not well-defined at the
nanoscale. These relationships are critical to assuring performance and reliability of
nanostructured materials and devices. They are also critical for building materials science based
predictive models of structure and behavior.
Gold films are of special interest in applications from MEMS mirrors to nanoscale
interconnects. In these applications, the need to minimize stress effects requires deposition of very
thin films. Figure 1 shows that the fracture energies for these films decrease to work of adhesion
values as the films become very thin. [2] While these fracture energies are fully capable of forming
plastic zones, there is no evidence of deformation or ductile fracture processes on the fracture
surfaces. This strongly suggests that dislocations were not emitted from the crack tips during
interfacial fracture. This is a concept supported by elastic strip and dislocation free zone models
where the only requirement for fracture is for the energy release rate to attain the true work-of-
adhesion. It also shows that adhesion controls performance and reliability of nanoscale films
through its effect on fracture.
The composition of gold-on-oxidized silicon can change markedly with increases in
temperature from any post-deposition processing. The changes can be even more dramatic for
films such as scandium, where strong reactions during deposition markedly alter composition,
structure and functionality (Figure 2). For nanoscale films, any film-substrate interactions, or when
using adhesion promoting interlayers, any film-film interactions lead to a complete change in film
composition and properties. Understanding these interactions is therefore critical for assuring
performance and reliability as there are no plastic energy dissipative processes to mitigate the
effects of low adhesive energies. These interactions and their effect on adhesion and fracture are
the focus of current work and of this presentation.
18 S. McDanels

Figure 1. (a) Steady state, *ss, and mode I, *I, fracture energies for thin gold films on sapphire.
Gold-on-oxidized silicon values are superimposed for comparison. The shaded region corresponds
to measured gold-on-sapphire work of adhesion values.

(a) (b)
Figure 2. The composition of gold-on-oxidized silicon can change markedly with post-deposition
processing as shown for (a) as deposited and (b) 300C/1hr annealed films.

References
1. R. Huang, Z. Suo, Journal Applied Physics, 91, 1135, 2002.
2. N. Moody, D. Adams, A. Mudd, M. Cordill, D. Bahr, Plasticity Effects on Interfacial Fracture
of Thin Gold Films, Proceedings ICM9, Geneva, Switzerland 2003 (SAND2003-8146)
3. T. Ye, Z. Suo, A. G. Evans, Int. J. Solids Structures, 29, 2639, 1992.

This work is supported by Sandia National Laboratories, a multiprogram laboratory operated by


Sandia Corporation, a Lockheed Martin Company for the United States Department of Energy's
National Nuclear Security Administration under contract DE-AC04-94AL85000.
Invited Papers 19

ASSESSMENT OF WELDMENT SPECIMENS CONTAINING RESIDUAL


STRESS

K. M. Nikbin
Mechanical Engineering Department, Imperial College,
Lonndon SW7 2BX., UK
k.n@ic.ac.uk

Weldments in components are regions where failures are most likely to occur either by fast
fracture, creep or fatigue. These regions could exhibit microstructural inhomogeneity as well as the
presence of micro-cracks and residual stresses. Understanding their behaviour is of major source of
interest for a range of industries. Creep and creep/fatigue crack growth models as well as residual
defect assessment codes need reliable and verifiable material properties data and validated fracture
mechanics parameters for use in their predictive methodologies. The research to develop an overall
methodology for deriving acceptable data and validated parameters for life assessment analysis has
been developing in Europe through a number of collaborative European projects. These have
covered both parent as well as weld material of a range of alloys and conditions. Although there is
substantial information and data on weld tests available, due to an absence of validated information
the industrial community cannot easily use this information with confidence. The present
standards, also, do not deal directly with testing of welded specimens and therefore. The Analysis
of the results from these projects have been continuing and furthermore have been used to assist in
the development of testing standards as well as life assessment codes of practice.
As a result a VAMAS (Versailles Agreement for Materials And Standards) Round Robin weld
testing programme has been initiated to address the issues of testing and analysis of weld related
materials. The paper presents the methodology that will lead to recommendations for a code of
practice (CoP) of welded materials. The overall aims are presented below;
Objectives
The following are the objectives of the new programme which adds emphasis on both the
Round Robin testing as well as the predictive modelling, and measuring of residual stresses in
components.
• Undertake a review of the information available on cracking of weld specimens and
components at high temperatures.
• Partners to link this initiative to their own projects
• Initiate a Round Robin testing programme of welded specimens within the TWA
collaboration based on the available information from the review. The round robin will
cover four different steels namely (347 weld, 316H stainless, P22, P91 and P92) steels
which have been offered for testing by partners in UK, Germany and Japan.
• The tests to be carried out will consist of mainly Compact Tension (CT) specimens of
parent, weld, and x-weld crack growth tests. The testing and analysis will be performed on
a Round Robin basis. Details will be set out at the kick-off meeting
• FE Modelling of residual stresses and identifying the role of stress relaxation during the
testing of the component at elevated temperatures.
• Measurement of residual stresses in a number of CT crack growth tests before and after
crack initiation using deep-hole drilling, neutron and X-ray diffraction of a number of tests
20 K. M. Nikbin

from the Round Robin. In addition collection of information on residual stress


measurements from partners involved in other research in this field.
• Identify the appropriate fracture mechanics parameters and materials and weld conditions
for different geometries, to cover the majority of cases for testing of weldments.
• Provide recommendations on weld testing and analysis plus the effects of stress relaxation
at high temperatures.
Pre-standardisation Needs
At present there is excessive conservatism in treating weldments and residual stresses.
Therefore it is envisaged that the results from this collaboration can be incorporated quickly into
existing or planned standards so that
• Relevant existing standards such as ASTM E1457 and the industrial codes of practice such
as R5, BS7910, ASME and API are likely to be improved as a result of this work.
• The recommendations from the work will improve predictive methods in High
Temperature Life Assessment procedures.
• The range of industries using this will be the power, aerospace, chemical and the electrical
industries
• Collaboration and sharing of information with active standardisation committees such as
ASTM, ‘ESIS TC11: HTMT Working Group On High Temperature Testing of
Weldments' and the 'Net European Network: Network On Neutron Techniques
Standardization For Structural Integrity'.
Deliverables
• Define the criteria and methodology for dealing with weldment testing and analysis.
• Provide relevant crack growth data from available tests of specimens and feature tests and
provide residual stress measurements and relaxation data from these tests.
• Using the appropriate fracture mechanics parameters analyse the crack growth data and
identify the effects of residual stresses.
• Make recommendations of welded component testing and analysis in the draft Code of
Practice for dealing with component creep crack growth testing and analysis of industrial
feature specimens derived from the initial phase of VAMAS TWA 25.
• Dissemination of results to the wider industrial audience and future implementation of
results in an ISO document.
Invited Papers 21

MEMS: RECENT ADVANCES AND CURRENT CHALLENGES

R. J. Pryputniewicz
Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST
rjp@WPI.EDU

Recent advances in MEMS technology have led to development of a multitude of new devices.
However applications of these devices are hampered by challenges posed by their integration and
packaging (Wei et al., 2005). Current trend in micro/nanosystems is to produce ever smaller,
lighter, and more capable devices at a lower cost than ever before. In addition, the finished
products have to operate at very low power and in very adverse conditions while assuring durable
and reliable performance (Pryputniewicz et al., 2001).
Some of the new devices were developed to function at high rotational speeds, others to make
accurate measurements of operating conditions of specific processes. Regardless of their
application, the devices have to be packaged to facilitate their use. MEMS packaging, however, is
application specific and, usually, has to be developed on a case by case basis (Pryputniewicz et al.,
2006). To facilitate advances of MEMS, educational programs have been introduced addressing
all aspects in their development (Pryputniewicz et al., 2003). This presentation will address
various aspects in a development of MEMS including, but not limited to, design, analysis,
fabrication, characterization, packaging, and testing. The presentation will be illustrated with
selected examples, Figs 1 to 5.
22 R. J. Pryputniewicz

References
1. Pryputniewicz, R. J., T. F. Marinis, D. S. Hanson, and C. Furlong, 2001, “New approach to
development of MEMS packaging for inertial sensors,” Paper No. IMECE2001/MEMS-
22906, Am Soc. Mech. Eng., New York, NY.
2. Pryputniewicz, R. J., E. Shepherd, J. J. Allen, and C. Furlong, 2003, “University – National
Laboratory alliance for MEMS education,” Proc. 4th Internat. Symp. on MEMS and
Nanotechnology (4th-ISMAN), Charlotte, NC, pp. 364-371.
3. Pryputniewicz, R. J., T. F. Marinis, J. W. Soucy, P. Hefti, and A. R. Klempner, 2006, “A
metal interposer for isolating MEMS devices from package stresses,” in press, Proc. EFC-16,
Alexandoupolis, Greece.
4. Wei, J., Wong, C. K., and Lee, L. C., 2005, “Wafer-level micro/nanosystems integration and
packaging,” Proc. 6th Internat. Symp. on MEMS and Nanotechnology (6th-ISMAN),” pp. 1-12,
Portland, OR.
Invited Papers 23

FRACTURE, AGING AND DISEASE IN BONE AND TEETH

R. O. Ritchie and R. K. Nalla


University of California, Berkeley
Department of Materials Science and Engineering, Berkeley, CA 94720, USA
roritchie@lbl.gov

Biological materials comprising mineralized tissues, such as bone and dentin in teeth, have
hierarchical structures with characteristic length scales ranging from nanometers to millimeters. In
this presentation, in vitro fracture toughness and fatigue-crack propagation properties of dentin and
human cortical bone are examined from a perspective of discerning how these properties depend
upon such microstructural hierarchies. The motivation for this is that although there is substantial
clinical interest in their fracture resistance, there is relatively little mechanistic information
available on how bone and teeth derive their resistance to cracking and how this is affected by
cyclic loads. Specifically, in vitro experiments are described that establish that the initiation of
fracture is locally strain-controlled (Nalla et al. [1]) and that subsequent crack growth
(characterized by resistance-curve behavior) is associated with a variety of extrinsic toughening
(crack-tip shielding) mechanisms, most importantly crack bridging (from individual collagen
fibrils and especially “uncracked ligaments”), macroscopic crack deflection and to a lesser extent
diffuse microcracking (Fig. 1) (Kruzic et al. [2], Nalla et al. [3]).

FIGURE 1. Schematic illustrations of some of the toughening mechanisms possible in cortical


bone: (a) crack deflection (by osteons), (b) crack bridging (by collagen fibers), (c) uncracked-
ligament bridging, and (d) constained microcracking.

Quantitative estimates for the relative contributions of these mechanisms to the overall
toughness are derived from simple micromechanical models [3]. In a manner not unlike ceramic
materials, it is shown that such extrinsic mechanisms act to toughen bone by lessening the
magnitude of stresses experienced at the tip of any cracks. Although macroscopic crack deflection
along cement lines provides a principal source of toughening in the transverse orientation, crack
bridging by intact regions in the crack wake (so-called uncracked ligaments) is the primary
toughening mechanism in longitudinal orientations; such bridges act to sustain load that would
otherwise be used to propagate the crack. In vitro fatigue experiments that seek to examine time- or
cycle-dependent crack-growth behavior, which pertain to stress fractures in bone, are also
described [4].
24 R. O. Ritchie and R. K. Nalla

Finally, we show that the role of biological aging, which causes a marked deterioration in the
fracture toughness of bone (Fig. 2), can be attributed to an age-related deterioration in the potency
of crack bridging [5], a phenomenon that we believe is associated with the role of excessive
remodeling in increasing the density of secondary osteon structures. However, the mechanistic
aspects of this age-related degradation in bone quality is additionally characterized at multiple
dimensions, including molecular (using deep UV Raman spectroscopy), sub-micron (using pico-
force atomic force microscopy) and tens of micron scale (using X-ray computed tomography). We
attempt to discriminate between possible age-related changes in the constitutive properties of the
hard tissue and age-related changes in its microstructure.

FIGURE 2. Resistance curves for stable in vitro crack extension in human cortical bone, showing
the deterioration in both the crack-initiation and crack-growth fracture toughness with age from 30
to 99 years.

References
1. Nalla, R.K., Kinney, J.H. and Ritchie, R.O., Nature Materials, vol. 2, 164-68, 2003.
2. Kruzic, J.J, Nalla, R.K, Kinney, J.H. and Ritchie, R.O., Biomaterials, vol. 24, 5209-21, 2003.
3. Nalla, R.K., Stölken, J.S., Kinney, J.H. and Ritchie, R.O., Journal of Biomechanics, vol. 38,
1517-25, 2005.
4. Nalla, R.K., Kruzic, J.J., Kinney, J.H. and Ritchie, R.O., Biomaterials, vol. 26, 2183-95,
2005.
5. Nalla, R.K., Kruzic, J.J., Kinney, J.H. and Ritchie, R.O., Bone, vol. 35, 1240-46, 2004.
Invited Papers 25

LABORATORY EARTHQUAKES

A. J. Rosakis, K. Xia1 and H. Kanamori2


Graduate Aeronautical Laboratories
1Graduate Aeronautical Laboratories and Seismological Laboratory
2Seismological Laboratory

California Institute of Technology, Pasadena, CA


rosakis@atlantis.caltech.edu

The goal of the present study is to create model laboratory experiments mimicking the dynamic
shear rupture process. We hope to use such experiments to observe new physical phenomena and
to create benchmark comparisons with existing analysis and numerics. The experiments use high-
speed photography, photoelasticity, and infrared thermography as diagnostics. The fault systems
are simulated using two photoelastic plates (Homalite) held together by friction. The far field
tectonic loading is simulated by pre-compression and the triggering of dynamic rupture
(nucleation) is achieved by an exploding wire technique. The fault forms an acute angle with the
compression axis to provide the shear driving force necessary for continued rupturing.
Earthquake dynamics and, in particular, the mechanics of dynamic shear rupture are two
relatively under-investigated sub-fields of seismology. Most efforts to date have focused on
analytical studies (Rice 2001) and on the numerical modeling of dynamic rupture processes using
finite element, finite difference, and boundary element methods (e.g., Ben-Zion and Andrews,
1997). As clearly elucidated by Rice (2001), the nature and stability of the predicted process
depends very strongly on the choice of frictional laws employed in the modeling and, as a result,
validation of the fidelity of such calculations becomes of primary importance.

FIGURE1. A homogeneous system composed of two frictionally held Homalite plates is shown.
Purely subRayleigh (D=25°, P=7 MPa) (2A) and purely supershear (D=25°, P=15 MPa) (2B)
rupture at the same time (28 Ps) after triggering.

Our goals are to investigate the dependence the characteristics of rupturing, such as rupture
speed, rupture mode on experimental conditions such as far-field biaxial compression, tilt angle of
the fault to the compression axis, as well as on the frictional properties of the fault interface. (Fig.
1.)
Results on both homogeneous and bimaterial interfaces are reported. For bimaterial interfaces,
various combination of dissimilar materials, including Homalite/polycarbonate pairs, are chosen to
26 R. O. Ritchie and R. K. Nalla

mimic wave speed mismatch conditions that are reported to exist across mature, crustal faults (Xia,
Rosakis, Kanamori and Rice 2005).
In the present lecture we concentrate on the experimental observation of the phenomenon of,
spontaneously unrelated, supershear rupture on the visualization of the mechanics of sub Rayleigh
to supershear rupture transition in such frictionally held interfaces. The results suggest that under
certain conditions supershear rupture propagation can be facilitated during large earthquakes (e.g.
the 2001 central Kunlunshan earthquake in Tibet, (Lin, Fu, Guo, Zeng, Dang, He and Zhao 2002,
Bouchon and Vallee 2002); the 2002 Denali earthquake in Alaska, (Ellsworth, Celebi, Evans,
Jensen, Nyman and Spudich 2004).

References
1. Rice, J. R., Lapusta, N., Ranjith, K., J. of the Mech. Phys. Solids, vol. 49 (9) 1865-1898,
2001.
2. Andrews, D., Ben-Zion, Y. J. of Geophysical Research, vol. 102, 553, 1997.
3. Xia, K. W., Rosakis, A. J., Kanamori, H., Science vol. 308, 681-684, 2005.
4. Lin, A. M., Fu, B. H., Guo, J. M., Zeng, Q. L., Dang, G. M., He, W. G., and Zhao,Y., Science,
296 (5575), 2015-2017, 2002.
5. Bouchon, N and Vallee, M, Science, vol. 301, 824-826, 2002.
6. Ellsworth, W. L., M. Çelebi, J. R. Evans, E. G. Jensen, D.J. Nyman, P. Spudich, Eleventh
Int’l Conference of Soil Dyn. and Earthquake Engineering, Berkeley, CA, 2004.
Invited Papers 27

A HISTORICAL RETROSPECTIVE OF THE BEGINNINGS OF BRITTLE


FRACTURE MECHANICS - THE PERIOD 1907-1947

H. P. Rossmanith
Institute of Mechanics and Mechatronics, Vienna University of Technology
Wiedner Hauptstr. 8-10/325, A-1040 Vienna, Austria
hans.peter.rossmanith@tuwien.ac.at

This contribution presents a historical retrospective of the early years of brittle fracture mechanics.
Looking back 99 years, the year 1907 is important for two reasons:
1 the German physicist Karl Wieghardt, a disciple of the famous physicist Arnold
Sommerfeld derived the complete stress field around the tip of a crack and gave the correct
order of the elastic crack tip singularity, and in solving the so-called Bach-Problem of
fracturing of roller bearings, presented the first mixed-mode fracture criterion, and
2 the Father of Engineering Fracture Mechanics, George Rankin Irwin, was born in El Paso,
Texas.
In the first part, this paper will highlight and detail the achievements in fracture mechanics by
Karl Wieghardt which for a long time – and for various reasons - were completely unknown or
disregarded by the fracture mechanics community at large. It will be shown that Wieghardt’s work
anticipated later developments by nearly fifty years [Wieghardt 1907; Rossmanith 1995a,b]. The
reasons for this neglect will be unveiled.
The second part of the paper will shed light on the very beginnings and struggles for
recognition of fracture mechanics at the time when the Griffith theory was extended to include
small scale plasticity by Irwin and Orowan. The role of research at the Naval Research Laboratory,
at North Carolina University at Chapel Hill, at several U.S. West Coast aircraft manufacturing
plants, and at Lehigh University in the founding of engineering fracture mechanics will be outlined
[Paris, P.C. 1995; Rossmanith 1997].
With the advent of a new generation of fracture mechanics researchers the author feels that the
history of fracture mechanics is not a dead branch of a huge tree but should be brought to the
attention of the young academics and professionals, and those who are interested [Cotterell B.
2002]. In fact, the future will only be mastered by looking at and understanding the past.

References
1. Cotterell, B. (2002) The Past, Present, and Future of Fracture Mechanics. Engineering
Fracture Mechanics 69:533-553.
2. Paris, P.C. (1995) Reflections on Progress in Fracture Mechanics Research. ASTM STP
1207:5-17.
3. Rossmanith, H.-P. (1995) An introduction to K. Wieghardt’s historical paper on splitting and
cracking of elastic bodies. Fatigue Fract Engng Mater Struct 12: 1367-69.
4. Rossmanith, H.-P. (1997) G.R. Irwin – The Father of Fracture Mechanics: A Biographical
Sketch. In: Fracture Research in Retrospect. An Anniversary Volume in Honour of
G.R.Irwin’s 90th Birthday. 3-36, A.A. Balkem, Rotterdam.
28 H. P. Rossmanith

5. Rossmanith, H.-P. (1997) The Struggle for Recognition of Engineering Fracture Mechanics.
In: Fracture Research in Retrospect. An Anniversary Volume in Honour of G.R.Irwin’s 90th
Birthday. 37-94, A.A. Balkem, Rotterdam.
6. Wieghardt, K. (1907) Ueber das Spalten und Zerreissen elastischer Koerper. Z. Mathematik
und Physik 55:60-103; Translation Rossmanith H.P. On splitting and cracking of elastic
bodies. Fatigue Fract Engng Mater Struct 1995, 12: 1371-1405.
Invited Papers 29

DYNAMIC CRACK PROPAGATION IN PARTICLE REINFORCED


NANOCOMPOSITES AND GRADED MATERIALS

A. Shukla
Dynamic Photomechanics Lab, Department of Mechanical Engineering
University of Rhode Island, Kingston, RI 02881, USA
shuklaa@egr.uri.edu

An experimental investigation has been conducted to evaluate the mechanical properties of novel
materials fabricated using nano and micron sized particles in polymer matrix. Experiments were
also conducted to investigate the dynamic crack propagation in theses particle reinforced materials.
High-speed digital imaging was employed along with dynamic photoelasticity to obtain real time,
full-field quantification of the dynamic fracture process. Birefringent coatings were used to
conduct the photoelastic study due to the opaqueness of these materials [1].

Unsaturated polyester resin specimens embedded with small loadings of 36 nm average


diameter TiO2 particles were fabricated using a direct ultrasonification method to study the effects
of nanosized particles on nanocomposite bulk mechanical properties [2]. The ultrasonification
method employed produced nanocomposites with excellent particle dispersion as verified by TEM.
The presence of the particles had the greatest effect on fracture toughness. Both static and dynamic
fracture toughness showed marked improvements with addition of small volume fractions (up to
1%)of nano particles. Dynamic fracture experiments were conducted with various specimen
geometries to study the complete history of dynamic crack propagation from initiation to crack
branching. Results from several of these experiments were compiled in order to establish a
relationship between dynamic stress intensity factor, KI, and crack tip velocity, a , and the
behavior of the nanocomposites is compared with that of the virgin polyester matrix. This
relationship is shown in Fig. 1. The specimens used in this study and the nano- particle distribution
is also shown in the figure. Crack arrest toughness increased by 64% in the nanocomposite relative
to the virgin polyester. Also, Crack propagation velocities in nanocomposites were found to be
50% greater than those in the virgin polyester
A detailed analytical and experimental investigation has also been conducted to understand the
behavior of a rapidly moving crack-tip in functionally graded materials (FGMs). First, an
30 A. Shukla

elastodynamic solution for a crack propagating at an angle from the direction of property variation
in an FGM is developed. Subsequently, the elastic stress, strain and displacement fields around the
crack-tip are obtained. This is followed by a comprehensive series of experiments to get more
insight into the behavior of propagating cracks in FGMs. The full-field stress data around the
propagating crack was analyzed using the stress field expansions developed in the first part of this
study. Dynamic fracture experiments were also conducted to study the behavior of a crack moving
at an arbitrary angle from the direction of property variation. It was found that when crack
propagation is inclined from the direction of property variation, crack-tip experience mixed mode
loading even if the far field loading is pure opening mode. Also, dynamic fracture experiments
were performed with different specimen geometries (modified compact tension and singe edge
notch tension) to develop a dynamic constitutive fracture relationship between the mode I dynamic
stress intensity factor (KID) and crack-tip velocity ( a ) for FGMs with crack moving in increasing
fracture toughness direction. This relationship is shown in Fig. 2.

The transient nature of crack growth in FGMs has also been investigated both analytically and
experimentally. It was concluded that not including the transient higher order terms in the analysis
of highly transient crack propagation experiments might give rise to high errors.

References
1. Der, V. K. and Barker, D. B., Mech. Res. Comm., vol. 5, 313-318, 1978.
2. Evora, V. F., Jain, N. and Shukla, A., Exp. Mech., vol. 45, 153-159, 2005.
Invited Papers 31

SPATIAL AND TEMPORAL SCALING AFFECTED BY SYSTEM


INHOMOGENEITY: ATOMIC, MICROSCOPIC AND MACROSCOPIC.

G. C. Sih
School of Mechanical Engineering,
East China University of Science and Technology, Shanghai 200237, China
Department of Mechanical Engineering and Mechanics,
Lehigh University, Bethlehem PA 18015, USA
gcs@ecust.edu.cn Fax: +86 (21) 6425-3500

The impetus of nanotechnology has shed light on the direction of new research for many fields and
continuum mechanics is no exception. The current trend is to reach down the scale such that all
disciplines would meet and benefit from one another. However, there are overwhelming
difficulties associated with the discontinuities of results from the various scales. The prevailing
gaps in materials and continuum theories are being referred to “Mesomechanics” [1].
Mesoelectronics has in fact discovered that the heat transfer behavior of small bodies is dinstinctly
different and requires fundamental studies of electronics at the subatomic scale. New physical laws
may have to be discovered to fill in the gaps. The final answer lies in multiscaling [2,3] where the
results at the smaller scales must be translated to the macroscopic level.
Modeling of multiscale material damage theories raises several basic issues. To begin with,
conditions must be invoked to connect the results observed at the different temporal and spatial
scales. Up to now, discussions seem to be confined to a very narrow range of size and time.
Furthermore, the models seldom address the effect of the initial or residual state as differentiated
from the performance sate. The former becomes increasingly more important as the size of device
is reduced to sub-microns. This is the rule rather than the exceptions in microelectronics. It is in
this nanoelectronics region that the electron transport behavior does not strictly obey quantum
mechanics nor classical physics. It has been referred to as the mesoscopic electronics region,
particularly with reference to power dissipation. Micro-chips should be kept sufficiently cool so
that they will operate in a stable manner. And yet the density of the transistors must also be high
and closely packed. The optimum balance can be achieved only by knowing the limits of how
effectively a very small device can dissipate heat. At the mesoscopic scale, the non-equilibrium
isoenergy density approach [4] can be applied. It has solved many problems with mesoscopic
phenomena [5]. The non-equilibrium theory [4] stresses in particular the power dissipation that is
derived directly by considering the mutual interaction of mechanical and thermal effects without
invoking artificial dissipation laws and/or constitutive relations. The dissipation has been shown to
depend sensitively on the temporal and spatial characters of the local deformation. Under
extension, a cooling period has been observed [5] that precedes heating for solid, liquid and gas.
This fundamental feature is not considered in classical physics. It can be very important for the
design of microchips in electronics.
The objective of this work is directed towards the development of physical models that can
relate results at different scale ranges with account for change of system homogeneity as the region
of interest is reduced in size. A corresponding increase in the time scale follows automatically. In
order to preserve the use of equilibrium mechanics in the ranges referred to as atomic, microscopic
and macroscopic, attention will be focused in the region where damage is concentrated in the form
of a singularity for the stress and energy density fields. The displacement field is required to
remain finite and continuous even though its cyclic value may become multi-valued. Cross scale
transition is made possible by imposing scale invariant criterion based on the “force” and/or
“energy” quantities.
32 G. C. Sih

The singularity representation approach [6] will be applied to illustrate how disorders in the
system at the microscopic and atomic scales can interact. The former and latter will be associated,
respectively, with micro-cracking and dislocations. Non-linear equation are solved for the coupling
of the micro-energy and dislocation-energy designated by Wmicro and W disln in normalized form,
respectively. They will be used to derive the length of inhomogeneity for the system.The
formulation entails several orders of magnitude extending from 10-11 to 10-1 cm on the lineal scale.

Examples will be presented for short time and long time effects where the Lower scale
chemical effects can have significant higher scale mechanical effects. They correspond to the
problems of solid rocket propellant explosionsand the stress corrosion of low alloy metals in high
temperature environment such as the nuclear reactors.

References
1. Prospects of Mesomechanics in the 21st Century: Current Thinking on Treatment of
Multiscale Mechanics Problems, in: G. C. Sih and V. E. Panin, J. of Theoretical and Applied
Fracture Mechanics, 37(1-3)(2001) 1-410.
2. G. C. Sih and X. S. Tang, Dual scaling damage model associated with weak singularity for
macroscopic crack possessing a micro/mesoscopic notch tip, J. of Theoretical and Applied
Fracture Mechanics, 42(1) (2004) 1-24.
3. G. C. Sih and X. S. Tang, Simultaneity of multiscaling for macro-meso-micro damage model
represented by strong singularities, J. of Theoretical and Applied Fracture Mechanics, 42(3)
(2004) 199-225.
4. G. C. Sih, Thermomechanics of solids: nonequilibrium and irreversibility, J. of Theoretical
and Applied Fracture Mechanics, 9(3) (1988) 175-198.
5. G. C. Sih, Some basic problems in nonequilibrium thermomechanics, in: S. Sienietyez and P.
Salamon, (eds.), Flow, Diffusion and Rate Processes, Taylor and Francis, New York, (1992)
218-247.
6. G. C. Sih and X. S. Tang, Singularity representation of multiscale damage due to
inhomogeneity with mesomechanics consideration, G. C. Sih, T. Kermanidis and Sp.
Pantelakis, eds., Sarantidis Publications, Patras, Greece (2004) 1-15.
B. TRACKS

B1. Nanomaterials and Nanostructures


1T1. Fracture and fatigue of nanostuctured materials 35

CHANNELING EFFECT IN FRACTURE OF MATERIALS WITH


NANOSTRUCTURED SURFACE LAYERS

V. E. Panin and A. V. Panin


Institute of Strength Physics and Materials Science, SB RAS
2/1, Academichesky pr., Tomsk, 634021, Russia
pav@ms.tsc.ru

1 The “chessboard” distribution of normal and tangential tensile and compressive stresses on
the “nanostructured surface layer – substrate” interface is revealed for metal materials (Fig.
1).

FIGURE 1. The “chessboard” model of the conjugation between nanostructured surface layer and
substrate (“+” – compressive stress, “-” – tensile stress) and channeling propagation of localized
shear.

2 A stochastic two-level model of the chessboard stress distribution on the interface is


worked out. The simulation results correlate well with appropriate experimental data.
3 In tension of metal materials with nanostructured surface layers one can observe the effects
of strain channeling in the nanostructured surface layer and fracture of the specimen as a
whole. This is related to shear localization and subsequent propagation of the main crack
along “chessboard squares” with normal tensile stresses (Fig. 2).
4 On the stage of uniform specimen elongation in the nanostructured surface layer localized
deformation bands evolve in the conjugate directions of maximum tangential stresses
W max. On the prefracture stage a localized deformation macroband in the form of an
extended neck propagates in the conjugate W max directions zigzagging along the whole
specimen length. On the stage of localized neck formation one can see the development of
two macrobands self-consistent by the scheme of a dipole or a cross (Fig. 3).
5 We have measured experimentally the evolution of main plastic shear in localized
deformation macrobands in the neck and rotational deformation modes associated with
localized shears. With non-compensated rotational deformation modes, the main crack is
generated in two localized deformation macrobands in the neck and then the specimen
fails.
6 Methods of governing the channeling effects of plastic flow and fracture of the material
with nanostructured surface layers are proposed.
36 V. E. Panin and A. V. Panin

FIGURE 2. Asymmetrical (a-d) and symmetrical (e-h) necking and fracture pattern of the tensile
cold-rolled titanium specimen with the nanostructured surface layer: optical image (a,e);
displacement-vector field (b,f); distribution of main plastic shear (c,g); fracture pattern (d,h); H =
17 %. u15.

FIGURE 3. Self-organization pattern of shears in interactions of macroscale localized-shear bands


of the dipole (a) and cross configurations (b).
1T1. Fracture and fatigue of nanostuctured materials 37

ATOMISTICS AND CONFIGURATIONAL FORCES IN GRADIENT


ELASTICITY

P. Steinmann and E. C. Aifantis


Chair of Applied Mechanics, University of Kaiserslautern, Dept. of Mechanical and
Process Engineering, P.O. Box 3049, D-67653 Kaiserslautern, Germany
Laboratory of Mechanics and Materials, Polytechnic School,
Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece
mom@mom.gen.auth.gr

The term Gradient Elasticity was introduced in the early 1990’s by Aifantis and co-workers (e.g.
Aifantis [1]) to denote a particular form of higher-order elasticity involving one extra
phenomenological coefficient in addition to the usual elastic Lamé constants. It distinguished itself
from an excessively large number of generalized elasticity and Cosserat type theories that were
advanced in the 1960’s and 1970’s in the sense that the stress remains symmetric and that the extra
gradient coefficient or internal length parameter involved may be directly related to the underlying
microstructure and determined experimentally. The theory allows for the formulation of
conveniently solvable boundary value problems by their reduction to inhomogeneous Helmholtz
equation with the source term being the solution of corresponding boundary value problems of
classical elasticity (e.g. Ru-Aifantis theorem [2]). It was shown (see also [3-6]) that the special
form of gradient elasticity can lead, among other things, to the elimination of singularities from
dislocation lines and crack tips. Motivated by this success, a large number of publications on wave
propagation, cracks, dislocations and other inhomogeneities (e.g. [7-10] and references quoted
therein) have been published recently based, more or less, on the special gradient elastic model
introduced in [1].
The aforementioned special linearized gradient elasticity model of [1] was motivated by a
simple gradient generalization of hyperelasticity theory elaborated upon by Triantafyllidis and
Aifantis [11] in relation to the “loss of ellipticity” in the governing differential equations of
nonlinear elasticity and the associated problem of localization of deformation. The gradient
hyperelastic model discussed in [11] and its linearized counterpart discussed in [1] are revisited
here in view of recent results obtained by Steinmann and co-workers (e.g. [12-14]) in relation to
atomistic modeling and configurational forces.
In relation to atomistic modeling, the Cauchy-Born rule is first extended to incorporate the
second-order deformation gradient. In addition to the Lennard-Jones and Morse interatomic
potentials a new interaction potential based on a Gaussian type nonlocal kernel of the form
(1 ʌ ʌ m 3 ) e x p [ (  r m b ) 2 ] , where mb is an intrinsic length scale (b denotes as usual the
Burger’s vector), will be used. Such nonlocal kernels can be used for the calibration of the gradient
coefficient and for obtaining corresponding new expressions for the Peierls stress. The coupling
between the molecular dynamics and the finite element method, the so-called hybrid model, will be
illustrated by considering the growth of a crack in two dimensions at the nanoscale. The difficulties
associated with the correct treatment of the transition between the atomistic and the continuum
regions and the role of higher-order gradients to settle the fundamental incompatibility between the
nonlocal character of the atomistic description and the local character of the continuum description
will be discussed.
Next, expressions are developed for the Peach-Koehler force and the force driving a crack in a
gradient elastic solid (J-integral) for the special gradient elasticity model of [1]. The notion of null
Lagrangeans and the boundary conditions are of relevance here and both topics will be addressed.
38 P. Steinmann and E. C. Aifantis

The special gradient elasticity model is presented as an example of the use of configurational
mechanics to deal with problems of defects within a linearized theory of gradient elasticity and set
up the stage for a general nonlinear treatment.
Finally, a general framework of the spatial and material settings of geometrically nonlinear
gradient hyperelasticity, allowing in particular for the consideration of material defects is
presented. Continua which are described within this constitutive class are characterized for the
spatial setting by a dependence of the stored energy density per unit volume undeformed
configuration on the first and second gradient of the spatial motion. Based on this description the
formulation of the corresponding material motion problem or rather of configurational mechanics
is developed. The material setting is particularly suited to compute the so-called material motion
problem, including the movement of material defects (dislocations, cracks, inclusions, phase
boundaries, etc.) relative to the ambient material. We highlight the duality of the spatial and the
material setting, in particular, the existence of a stored energy density per unit volume deformed
configuration depending on the first and second gradient of the material motion (i.e. the inverse
motion, deformation map); provide transition rules between them; and compare the results
obtained to those derived for continua without higher-order gradients.

References
1. Aifantis, E.C., Int. J. Engng. Sci., vol. 30, 1279-1299, 1992.
2. Ru, C.Q. and Aifantis, E.C., Acta Mechanica, vol. 101, 59-68, 1993.
3. Altan, B.S. and Aifantis, E.C., J. Mechan. Behav. Mats., vol. 8, 231-282, 1997.
4. Aifantis, E.C., J. Eng. Mater. Techn., vol. 121, 189-202, 1999.
5. Gutkin, M.Yu. and Aifantis, E.C., Scripta Mat., vol. 40, 559-566, 1999.
6. Aifantis, E.C., Mech. Mat., vol. 35, 259-280, 2003.
7. Georgiadis, H.G. and Vardoulakis, I. and Velgaki, E.G., J. Elasticity, vol. 74, 17-45, 2004.
8. Georgiadis, H.G., J. Appl. Mech., vol. 70, 517-530, 2003.
9. Lazar, M., Maugin, G.A. and Aifantis, E.C., Phys. Stat. Sol., vol. 242, 2365-2390, 2005.
10. Zhang, X. and Sharma, P., Int. J. Sol. Struct., vol. 42, 3833-3851, 2005.
11. Triantafyllidis, N. and Aifantis, E.C., J. Elasticity, vol. 16, 225-238,1986.
12. Sunyk, R. and Steinmann, P., Int. J. Sol. Struct., vol. 40, 6877-6896, 2003.
13. Steinmann, P. and Elizondo, A., RTN DEFINO Report No. 3, 2005.
14. Kirchner, N. and Steinmann, P., Phil. Mag., in press, 2006.
1T1. Fracture and fatigue of nanostuctured materials 39

TENSILE BEHAVIOR AND FRACTURE OF CARBON NANOTUBES


CONTAINING STONE-WALES DEFECTS

K. I. Tserpes and P. Papanikos


Laboratory of Structural Mechanics, Department of Rural & Surveying Engineering,
National Technical University of Athens
Department of Product and Systems Design Engineering, University of the Aegean
Zografou Campus, 9 Iroon Polytechniou St., 15780, Athens, Greece
Ermoupolis, Syros, 84100, Greece
kitserpes@hol.gr, ppap@syros.aegean.gr

The effectiveness of carbon nanotubes (CNTs) as reinforcements is designated by their mechanical


behavior as stand alone units. One of the most commonly present topological defects, whose effect
on the mechanical behavior of CNTs needs to be clarified, is the Stone-Wales (SW) defect. In this
paper, the effect of SW defect on the tensile behavior and fracture of armchair, zigzag and chiral
single-walled carbon nanotubes (SWCNTs) was investigated using an atomistic-based progressive
fracture model (PFM).
Following the concept of Li and Chou [1], CNTs are treated as space-frame structures by
assuming that the C-C bonds act as load-carrying members and the carbon atoms as joints of the
members. The PFM utilizes the FE model developed in Ref.[2] for analyzing the nanotube
structure and the Morse interatomic potential, as modified by Belytschko et al. [3], for simulating
the non-linear behavior of the C-C bonds. The FE models of the SWCNTs were created using the
ANSYS FE code. For the modeling of the C-C bonds, the 3D elastic ANSYS BEAM4 beam
element was used. The specific beam element, as all 3D beam elements of the ANSYS FE code, is
linear elastic and does not have the ability to model non-linear behavior. This restriction is
surpassed by the stepwise procedure of progressive fracture modeling. The nanotube is loaded by
an incremental displacement at one of each ends with the other end fixed. At each load step, the
stiffness of each beam element is redefined using its axial strain, as evaluated from the FE model,
and the force-strain relationship of the modified Morse potential. Ôï optimize the accuracy of
computational results, in each analysis, the number of load steps was chosen from convergence
tests. Before using the PFM to accomplish the present study, its ability to accurately simulate the
tensile behavior of defected SWCNTs was successfully verified through the comparison with the
molecular dynamics simulations of Belytschko et al. [3].
The SW defect is the 90o rotation of a bond, which transforms 4 hexagons to 2 pentagons and 2
heptagons. The modeling of the defect is performed during the creation of the FE mesh of the
nanotubes.
Fig.1 displays the predicted stress-strain curves of the SWCNTs. In all cases, SW defects
served as nucleation sites for fracture and their presence reduced the failure stress and failure strain
of the nanotubes. The percentage reduction of failure stress and strain depends on the chirality of
tubes. On the other hand, the nanotube stiffness was found to be unaffected. Fig.2 shows the
evolution of fracture in the (12,12) SWCNT. Fracture initiated from the longitudinal bond
connecting the two heptagons of the SW defect and propagated in the r45o direction. The fracture
process was completed when all bonds around the circumference failed.
40 K. I. Tserpes and P. Papanikos

FIGURE 1. Predicted stress-strain curves of the SWCNTs.

FIGURE 2. Evolution of fracture in the (12,12) SWCNT containing a SW defect.

References
1. Li, C. and Chou, T-W., International Journal of Solids & Structures, vol. 40, 2487-2499,
2003.
2. Tserpes, K.I. and Papanikos, P., Composites Part B: Engineering, vol. 36(5), 468-477, 2005.
3. Belytschko, T., Xiao, S.P., Schatz, G.C. and Ruoff, R.S., Physical Review B, vol. 65, 235430,
2002.
1T1. Fracture and fatigue of nanostuctured materials 41

ATOMIC-SCALE INVESTIGATION ON FRACTURE TOUGHNESS IN NANO-


COMPOSITE SILICON CARBIDE

M. Ippolito, A. Mattoni, L. Colombo and F. Cleri1


INFM-SLACS Sardinian LAboratory for Computational materials Science and Department of
Physics, University of Cagliari, Cittadella Universitaria, I-09042 Monserrato (Ca), Italy
1
ENEA, Unita` Materiali e Nuove Tecnologie, and INFM Centro Ricerche della Casaccia, CP
2400, I-00100 Roma, Italy
mariella.ippolito@dsf.unica.it

Ceramic materials are attractive for structural applications because of their low density, chemical
inertness, high strenght, high hardness and high-temperature stability. However they have
inherently low fracture toughness, so that plastic deformation in ceramics is found to be extremely
limited. Ceramic matrix composites (CMC) have been therefore developed to overcome the
intrinsic brittleness and lack of reliability of monolithic samples. CMC's consist of a ceramic
matrix reinforced with inclusions, such as particles, whiskers or chopped fibers (fiber
thoughening).
Although the macroscopic toughening is the result of several complex microscopic process,
such as the formation of subgrain boundaries and frontal process zone enhancement, the key issue
is represented by the interaction between the tip of possible crack front and the fiber.
In this work we present an atomistic investigation of the stress states of a crack inclusion pair
in nano-composite -SiC. We show that atomistic simulations (AS) are a powerful tool to test the
reliability of possible continuum approximations and moreover provide a predictive modelling of
stress intensification phenomena in regions where the continuum models fail to offer a unique
picture. Our simulations provide a rather general and simple constitutive equation, valid over a
range of crack-inclusion elastic mismatch and for a wide range of relative distances.
We consider two kind of coherent inclusion: carbon and silicon fibers. Because of the lattice
mismatch with respect to the matrix, both the carbon and silicon inclusions induce stress which
turns out to be, respectively, tensile or compressive.
The interatomic forces are described by the Tersoff potential for Si-C systems [1] and the load
is applied by means of the constant traction method[2].
We have compared our AS results [3] with two recent (and conflicting) continuum solutions: a
first one by Li and Chen[4], based on Eshelby equivalent inclusion approach, and a second one by
Helsing[5], showing that the available continuum models are not able to properly describe the
stress intensification phenomena at arbitrary values of distance between the crack tip and the
inclusion, and for different matrix-inclusion elastic mismatch.
The atomistic results, instead, provide a simple law for the effective variation of the crack
toughness, valid in both the silicon and carbon case, i.e. for very different matrix-inclusion
mismatch, and provides a simple and robust constitute equation for stress intensification
phenomena at any crack-inclusion distance in a ceramic composite.

References
1. J. Tersoff, Phys. Rev. B, vol. 39, 5566, 1989.
2. F. Cleri, Phys. Rev. B, vol. 65, 014107, 2002.
42 M. Ippolito et al.

3. M. Ippolito, A. Mattoni, L. Colombo, F. Cleri, submitted for publication.


4. Z. Li and Q. Chen, Int. J. Fracture, vol.118, 29, 2002.
5. J. Helsing, Engng. Fracture Mech., vol.64, 245, 1999
1T1. Fracture and fatigue of nanostuctured materials 43

MULTISCALE MODELING AND COMPUTER SIMULATION OF STRESS-


DEFORMATION RELATIONSHIPS IN NANOPARTICLE-REINFORCED
COMPOSITE MATERIALS

L. V. Bochkaryova, M. V. Kireitseu, G. R. Tomlinson1, V. Kompis2 and H. Altenbach3


National Academy of Sciences of Belarus, Lesnoe 19 – 62, Minsk 223052
1Rolls-Royce UTC and Dynamics Group, the University of Sheffield, UK
2Department of Mechanics, University of Zilina, Slovakia
3Faculty of Engineering Sciences, Martin-Luther-Universitet Halle-Wittenberg

l_silver@rambler.ru

A novel concept of nanoparticle vibration damping [1, 2] shows the effect that molecule-level
mechanism can have on the damping and that nanoparticles/fibres/tubes-reinforced materials can
provide enhanced strength and vibration damping properties. It is particularly worth noting that
carbon nanotubes can act as a simple nanoscale spring where rheological modelling can play a
significant role while understanding of mechanics of novel systems across the scale length. The
mechanisms involved in such materials need to be understood and the relevance to strength/
damping identified. Additionally adequate modelling techniques for the next generation of
vibration damping systems are technology gaps that need urgent consideration.
The computational models simulating mechanical properties range many scales starting from
nano-scale (Molec. Dynam.), investigating the atomic interaction of the nano-particles in a very
reinforcing material alone and its interaction with the atoms of matrix [3, 4], up to the description
of such composite in large structures with dimension of several meters. The mechanisms involved
in such materials need to be understood and the relevance to damping identified.
Our aim was to investigate and develop advanced computational FEM-based technique to
estimate performance and mechanical/damping properties of nanoparticle/fiber-reinforced
engineering materials and then, if successful, assemble them into viable engineering workbench.
Carbon nanotube-reinforced structures are principally used at all phases of modeling and
simulation. Thus we will form a bridge from the very basic research done by physicists and
chemists to the practical applications by engineers.
To achieve the objective we performed the development of characterization and modeling
technique of the fundamental phenomena that describe relationships between structure and
mechanical properties of the materials, formalize the set of structural mechanical approaches to
build a bridge between macro and nanoscales. Carbon-nanotube-reinforced composite material
was simulated via advanced finite element and meshfree codes, using a hollow shell representation
of the individual nanotubes. It’s noted that the recently developed meshfree techniques for shells
do not present undesirable membrane and shear locking phenomena, while these locking
phenomena are inherent in the more commonly used finite element methods. Our computational
approach is fundamentally equal simulation and modelling of materials with combined molecular
dynamics (MD) and FEM technique, where an equivalency of each other is shown by
mathematical analysis. This conclusion proves a joint application of MD and FEM principles in
2D/3D modeling of materials where stress-deformations are described in angle coordinates of sin
and cos found through application of force field methodology and structural mechanic approach [3,
4].
A continuum model for deformation of material reinforced with nanoscale hard fibers/tubes
has been developed. A model using the fast multi-pole method (FMM) is presented. The FMM
44 L. V. Bochkaryova et al.

uses the Taylor expansion of the integral equations describing the interaction of rigid inclusion
with the closest neighbors and with the flexible matrix. The FMM models reduce drastically both
computation time and storage requirements so that models, which were not possible to solve with
present computational technique, are investigated. The method enables to solve the continuum
containing up to millions of such inclusions in a computer by parallel algorithms.
In conclusion, it is worth noting that for a realistic simulation of the stability behavior of the
nanoparticle-reinforced material, the nonlinear intramolecular inter-actions between neighboring
atoms have to be taken into account. In order to reduce computational costs, it is necessary to
develop advanced homogenization technique so as to apply shell elements in the model.
Comparing to the FEM, the new technique will introduce further reduction of both computer time
and storage requirement. The results will potentially create fundamentals for investigation and
development of 3-D reinforced composite structures with high nanoscale structures volume
content, using nano-scale reinforcement architecture to reduce component mass and dimension.
Although the process of verification and validation is somewhat circular, the entry point into
this process is clearly through experiments that help determine the validity of theory and
assumptions while also helping to quantify the state variables associated with the problem. It is,
therefore, necessary that the Computational Materials approach must use experimental data to
establish the range of performance of a material and to validate predicted behaviour. Even at the
atomistic level, methods such as molecular dynamics require careful parameterization (fit) to
empirical data. Therein, it gives a challenge to Computational Materials: validation of methods
across the complete range of length and time scales. To achieve this validation requires advances
in measurement sciences as well as advances in theory and models, coupled with integrated,
interdisciplinary research. It is imperative that research laboratories maintain a focused effort to
develop new programs that provide for the simultaneous growth of all the critical elements that are
required for validation of multi-scale methods.
Research work of Dr. Kireitseu has been supported by the Royal Society in the UK and
WELCH scholarship administered through the Amer. Vac. Soc. / Int.-l Union for Vac. Sc., Tech.
and Appl. in the US and Europe. Dr. Bochkareva is continuing her research work under EU INTAS
2005-2007 postdoctoral fellowship Ref. Nr 04-83-3067. It should be noted however that the views
expressed here are those of the authors and not necessarily those of any institutions.

References
1. Rivera J.L., McCabe C., and Cummings P.T. Nanoletters, Vol. 3, No. 8, 2003.
2. Li J., Ye Q., A. Cassell, H. T. Ng, R. Stevens, J. Han, M. Meyyappan, Appl. Phys. Lett., Vol.
82 (No.15), 2003.
3. Li C. and Chou T.W. Physical Review B Vol. 68, 2003.
4. Kireitseu M., Hui D., Bochkareva L., Eremeev S. and Nedavniy I. In Proceedings of the 106th
Annual American Ceramic Society Meeting & Exp., Symposium 17 - Innovative Processing
and Synthesis of Ceramics, Glasses and Composites. - April 18-21, 2004, edited by G. Geiger.
- Indianapolis, IN, USA, 250-264.
1T1. Fracture and fatigue of nanostuctured materials 45

THE MECHANICAL PARAMETERS OF NANOOBJECTS (THEORY AND


EXPERIMENT)

E. Ivanova, N. Morozov1 and B. Semenov1


Dep. of Teoretical Mechanics, St.-Petersburg State Polytechnical University,
Politechnicheskaya str., 29, St.-Petersburg, Russia, 195251
1
Dep. of Elasticity theory, St.-Petersburg State University, Bibliotechnaya sq., 2, Staryi Petergof,
St.-Petersburg, Russia, 198904
morozov@NM1016.spb.edu

Advances in high technologies, using nanometer-size structures, requires calculation of


mechanical properties for the objects of the nanosize scale level. Majority of the theoretical
mechanical models for nanoobjects is based on the macroscopic equations of theory of elasticity.
However, a lot of researchers have noted inconsistency between the values of the elastic moduli
obtained from micro- and macroexperiments.
The presented paper is devoted to theoretical and experimental investigation of the influence of
the scale effects on the bending stiffness of a nanocrystal, which is extended in one direction and
has a limited number of atomic layers in another direction. It is theoretically shown that for small
number of atomic layers the bending stiffness of the nano-crystal substantially depends on the
number of layers and tends to its elasticity-theory value for large number of layers [1].

FIGURE 1.

The problem of the experimental determination of elastic moduli of nanoscale objects is of


present interest. The determination of the elastic moduli of thin macroscopic shells is usually based
on experiments with plates. It is known that, when grown using certain techniques, nanoobjects are
obtained only in the form of shells. Therefore, it is necessary to develop a method for determining
the elastic moduli of nanoobjects on the basis of experiments with shells. Experimental
determination of the bending stiffness of nanosize shells presents a serious problem, because for
such widespread nanoobjects as nanotubes and fullerenes under arbitrary deformation, the material
is subjected to both bending and tension. Therefore, all parameters (e.g., natural frequencies) that
can be measured directly are complicated functions of both bending and tension stiffness. In recent
years, together with nanotubes and fullerenes, nanoobjects of a more intricate configuration have
been obtained [2–4]. Nanosize cylindrical helices [2] (see Fig. 1) are of particular interest in
connection with the possible experimental determination of bending stiffness.
This is due to the fact that in helical shells under arbitrary deformation, the material is mainly
bent, so that the material tension effect can be neglected when interpreting experimental data; and
the natural oscillation shapes of helical shells are much more easily observed than those of
46 E. Ivanova et al.

cylindrical shells associated with pure bending of the material. The latter statement is illustrated in
Fig. 2, which presents the first four helical shell oscillation shapes. The analysis of helical shell
dynamics may be a theoretical foundation for experimental testing of the applicability of the
continuum theory to (a) the calculation of mechanical characteristics of nanoobjects and (b) the
experimental determination of the bending stiffness of nanoshells [5].

FIGURE 2.

Acoustical and optical methods of measuring of the eigenfrequencies of micro-objects are


based on using homodyne laser vibrometers and adaptive photodetectors. Technique for vibration
analysis of the laser vibrometer using adaptive photodetectors is based on the effect of non-steady-
state photoelectromotive force. The technique enables efficient direct conversion of high-
frequency phase modulation of speckle-like optical waves reflected from the vibrating object into
output electrical signal with concomitant setting of the optimal operation point of the
interferometer and suppression of amplitude laser noise. Methods of measuring of the
eigenfrequencies of nano-objects are based on using atomic force microscope. Technique for
vibration analysis is based on the effect of mechanical interaction of cantilever needle with nano-
object.
References
1. Ivanova, E.A., Krivtsov, A.M., Morozov, N.F., Dokl. Phys. 47, 620. 2002.
2. Golod, S.V., Prinz, V.Ya., Mashanov, V.I., Gutakovsky, A.K., Semicond. Sci. Technol. 16,
181. 2001.
3. Vorob’ev, A.B., Prinz, V.Ya., Semicond. Sci. Technol. 17, 614. 2002.
4. Prinz, V. Ya., Microelectron. Eng. 69 (2/4), 466. 2003.
5. Ivanova, E.A., Morozov, N.F., Dokl. Phys. 50, 83. 2005.
1T1. Fracture and fatigue of nanostuctured materials 47

ADVANCED MANUFACTURING DESIGN CONCEPTS AND MODELLING


TOOLS OF THE NEXT GENERATION NANOPARTICLE-REINFORCED
DAMPING MATERIALS

M. V. Kireitseu, G. R. Tomlinson1, R. A. Williams2 and V. Kompis3


National Academy of Sciences of Belarus, Lesnoe 19 – 62, Minsk 223052
1Rolls-Royce UTC and Dynamics Group, the University of Sheffield, UK
2Institute for Particle Science & Engineering, the University of Leeds, UK
3Department of Mechanics, University of Zilina, Slovakia
indmash@yandex.ru

Vibrations and noise exist in almost every aspect of our life and are usually undesirable in
engineering structures [1]. Vibrations are of concern in large structures such as aircraft either civil
(airbus A 380) or military, as well as small structures such as electronics [2]. It is now accepted
that nanotechnology can help solve vibration damping and high noise issues through the utilisation
of nanomaterials (or media) that dissipate a substantial fraction of the vibration energy that they
receive. The mechanisms involved in such materials need to be understood and the relevance to
damping identified via both computational and experimental benchmarks.
The main issue is the need to match applications to technologies/materials being developed [3-
5]. Wide-ranging application of damping materials in real-life products is one of the best ways to
ensure future development. Commercial utilisation of a damping technology depends from both
technical performance and business environment for that. There must be a business support for
extensive technology implementation during its life-cycle cost. This has often been seen to be a
limiting factor in the utilisation of novel material/technology. In this situation, a broader
understanding of the material and its potential application is of great benefit since a “secondary”
feature can make the damping system more attractive than its predecessor.
Major technical barriers also prevent greater use of damping nanoparticle-reinforced materials.
These include sensitivity to temperature (particularly viscoelastic materials based on polymer
matrix), ability to manufacture (nanoparticle is expensive to manufacture and disperse in a matrix),
increased complexity of both nanoparticle damper and a whole damping system (affecting weight
and size) as well as effort required to predict/model performance and to optimize a technology for
a specific application. The current drive for increased efficiency and reduced cost of machinery,
usually results in a requirement for mass to be minimised. To avoid vibration related problems
therefore, considerable effort has to be expended to increase the stiffness or damping of the
structure. In terms of development of materials to achieve these goals, damping has probably
received the least attention.
The focus in this paper is directed toward to the investigation into carbon-based nanoparticle/
fibre/tube-reinforced materials and coating systems and their dynamic/damping characterization.
Computational work is concentrated on hierarchical multiscale modelling of damping behaviour as
a function of frequency, amplitude and temperature. A computational model is formulated in terms
of meso- and nanoscopic ideas of damping behavior and could provide an approach to predict
vibration damping properties and optimize some manufacturing design concepts of those material
systems so as to enable the efficient synthesis of these novel damping solutions.
The novel concept of nanoparticle-based damping technology shows that a molecule-level
mechanism can considerably enhance vibration damping and dynamic of aerospace components
48 M. V. Kireitseu et al.

(fan blades) via enhanced energy dissipation because of large surface-to-volume aspects in
nanoparticle-reinforced composite material, large damping energy sources for friction and slip-
stick motion at interfaces of matrix and nanoparticle. Thus carbon nanotube can act as a simple
nanoscale damping spring in aerospace materials and is suggested for aerospace damping materials
of the next generation. The materials offer the potential to further reduce the mass and dimension,
increase performance, and reduce vibrations.
As a result the nanoparticle/tube/fibre-reinforced composite material gains advanced damping
properties compared with conventional materials reinforced through available technologies as well
as other types of commercialized damping solutions. The damping properties of the material can
be further enhanced by designing unique aerospace components based on personal manufacturer
needs of either civil or military aircrafts. Personalized set of nanoparticles can be introduced into
material matrix, for example, through the CVD-based technology combined with conventional
particle technologies (thermal spraying, PVD, etc.).
The principal conclusions are that by invoking the properties of nano-auxetics/nanostructures
it is possible to control the wave/sound/vibration propagation in the material and enhance the
energy dissipation that can assist in improving the inherent damping of materials, but an
experimental/theoretical environment is required to apply it. Nanoparticles/tubes can be used as a
reinforcement of a matrix to provide multi-functionality, and thus we need to create an
environment (knowledge) to introduce nanomaterials widely in industry. Developed computational
tools and workbench is important part of next generation aerospace design.
Topics that could yield particular success for damping in the future include 1) Micro and nano-
scaled damping materials where nanoparticle/tube/fibre reinforcement concept might give
exceptional, temperature independent damping and negligible added weight; 2) Optimisation
methods where structural mechanics/dynamics approaches could be combined with advanced
numerical FEM modeling and statistical variation to estimate dispersion of nanoparticles in a
matrix; and 3) Low-cost damping systems that is cheaper than current polymer-based coatings.
The outcome of the research work is expected to have wide-ranging technical benefits with
direct relevance to industry in areas of transportation (aerospace, automotive, rail) and civil
infrastructure development, but the goal is aerospace turbine applications.

References
1. Hollkamp J.J. and R. W. Gordon, Smart Mater. & Struct. Vol. 5(5), 1996.
2. Chung T.R. Journal of Materials Science. vol. 36, 2001.
3. Rivera J.L., McCabe C., and Cummings P.T. Nanoletters, Vol. 3, No. 8, 2003.
4. Li J., Ye Q., A. Cassell, H. T. Ng, R. Stevens, J. Han, M. Meyyappan, Appl. Phys. Lett., Vol.
82 (No.15), 2003.
5. Siegel R.W., Hu E., and Roco M.C. in Proceedings of WTEC Workshop, May 8-9, 1997
Workshop, Washington, DC, 1998.
1T1. Fracture and fatigue of nanostuctured materials 49

FRACTURE OF NANOSTRUCTURED IONOMER MEMBRANES

Yue Zou, X. Huang and K. L. Reifsnider


CT Global Fuel Cell Center
University of Connecticut, Storrs, CT 06269
Xinyu@engr.uconn.edu

Thin ionomer membranes showing nanostructure phase separation are widely used in low
temperature fuel cells as electrolytes. Examples include perfluorinated sulfonic acid (PFSA)
ionomers, better known as NafionTM produced by E. I. du Pont de Nemours and Company. Their
electrochemical performance and mechanical strength properties are determined by their molecular
morphology, which resembles water-filled hydrophilic micellae (4~5nm in size) dispersed in a
hydrophobic matrix. Temperature and humidity are two major parameters in an internal fuel cell
working environment, which have significant influence on this nanostructure and consequently
influence the material properties. The transport properties of such ionomers have been extensively
studied. In this paper, we will discuss fracture and strength of such membranes, as related to
humidity and temperature.
To understand fracture of the membrane, it is necessary to measure its strength and determine
the failure mode(s) under all relevant conditions. Measuring strength at different humidity and
temperatures for very thin membranes is difficult. A new method using optical strain measurement
(even for submerged specimens) has recently been used in the authors’ laboratory. Some example
data and fracture patterns are shown in figure 1.

FIGURE 1. Left, stress-strain response of Nafion 111 membrane under different status of
hydration and temperatures; Right, fracture patterns for fully hydrated specimens in 25 ഒ, 65 ഒ
and 80 ഒ (in (a), (b) and (c), respectively).

The material tested was NafionTM NR111. The specimens were boiled in deionized (DI) H2O
for 1 hour to fully hydrate them. With an increase of temperature, there is a significant increase of
strain to break, but a decrease of stress to break, as shown in Figure 1. Although this material is
ductile, the broken surfaces show a clean severance resembling those formed in brittle fractures.
With an increase of water content, the yield behaviour becomes less distinct. It is believed these
features are due to the evolution of the nanostructure as a function of temperature and humidity.
The micro/nanostructual nature of these phenomena is very complex. We offer some hypotheses
that qualitatively explain some of these changes.
It is widely accepted that this kind of ionic material contains two immiscible phases, i.e., the
hydrophobic fluorocarbon and the hydrophilic ionic phase, and that the ionic groups tend to
50 Yue Zou et al.

aggregate to form tightly packed regions as clusters with a size on the nanoscale [1]. It has been
claimed by G. Gebel, that these clusters can vary from spherical structures to a rod-like (phase-
inverted) structure [2] with increasing water content in the membrane. For a known hygrothermal
history of the membrane, it is possible to determine the water content in a membrane by weight[3];
therefore, the structure can be then determined in principle. With a known structure, we can talk
about strength and the failure mechanism of the membrane. For low water content, the membrane
behaves like a typical tough thermoplastic material with a very distinct yielding point, which
corresponds to the initiation of molecular chain movements over each other as the applied load
reaches a critical point that balances out the resistance to chain movement due to entanglement and
ionic “cross-linking” through cations. For high water content, we believe that the increase of water
reduces the strength of ionic “cross-linking” as the separation distance of the polar groups is
increased due to the swelling of the ionic clusters. As a result, chains sliding over each other may
occur at relatively low load levels; hence we observe a “spreading out” of the yield point. When
the temperature approaches the Tg, the semicrystalline hydrophobic backbone softens; as such we
see a reduction of initial modulus and also an increase of failure strain. The large strain observed at
high temperature may be due to the drawing out of the chains from the folded semi-crystalline
phase. Fracture occurs when the accumulated strain energy propels the complete breaking of the
primary bonds and the entanglement points.
Models in the literature used to explain the behavior of elastomers attempt to express stress-
strain in general and temperature effects in particular, but do not explain the mechanism of
fracture. For ionomeric membranes, nanostructure variations caused by changes in material
hydration play a major role. This paper will offer some interpretations of the data based on
understandings of this behavior and interpretations using multiphysics concepts and
representations.

Reference
1. Gierke, T. D., Munn, G.E. and Wilson, F.G., J. Polym. Sci., Polym. Phys. ED 19, 1687-1704,
1981.
2. Gebel, G., Polymer, vol. 41, 5829-5838, 2000.
3. Choi, P. and Datta, R., J. Electrochem. Soc., 150 (12) E601-E607, 2003.
1T2. Failure mechanisms 51

DEFORMATION AND LIMIT STATES OF CARBON NANOTUBES UNDER


COMPLEX LOADING

A. V. Chentsov and R. V. Goldstein


117526 Russia, Moscow, Vernadskogo prosp. 101 k.1, RUSSIA
chentsov@ipmnet.ru

For real experiments with nanotubes the complex and precision equipment is required. Being an
alternative method of research, the computer modelling allows to perform practically any
experiments at a level of atoms. These problems are solved by the methods of molecular dynamics.
All atoms of such model are in constant thermal motion and their interaction is described by a
known potential. At the same time the combination of geometrical scales of the objects like a
nanotube (the relation of length to radius is 100-1000, diameter to wall thickness - 15-150) allows
to model the nanotube deformation by some models of continuous medium.
The regularity of atomic structure of nanotubes enables us to replace the system of atoms with
equivalent model of elastic isotropic truss (linear or nonlinear), and the truss system in macroscale
by a continuous medium, passing to continuous model of a nanotube. The choice of model of
either the continuous cylinder or a cylindrical shell depends on a solved problem.
Just as at direct modelling of interatomic interactions in molecular dynamics, within the
framework of the deformation description for a nanotube elastic truss model the energetic
approach appears to be effective. From the computative point of view the truss model gives the big
advantage of a time of calculation.
Nanotubes possess an exclusive relationship length / radius. Tubes with such hexagonal cell
will be exposed to small deformations of atomic bonds even at significant axial deformations and
bends. Such assumption means, that as a potential of interatomic interaction it is sufficient to use
harmonic potential.
The harmonic potential corresponds to a potential strain energy of the elastic rod (spring)
connecting a pair of interacting atoms. Then for any graphene plane it is possible to build the truss
system equivalent to the atomic model at equal loading conditions. On the structures of nanometer
scale, this approach has been suggested by G.M.Odegard [1].
Covalent interactions of atoms of molecular structure can be quantitatively described by using
the methods of molecular dynamics. Forces of attraction and repulsion acting for each pair of
atoms depend on relative atomic positions and are presented by the chosen forcefield. These forces
give the contribution to full vibrational potential energy of molecular system which is equal to a
strain energy of a macroscopic body of equivalent geometry.
For nanotube modelling in most cases the only essential degrees of freedom are: tension,
change of the angle, nonbonded interactions.
Energy of torsion is small enough to be neglected. It only becomes essential in problems with
the large bending deformations. Nonbonded interactions were taken into account at interaction of
separate nanotubes and graphene planes.
To proceed from discrete atomic model to truss, it is necessary to take into account all essential
pairwise interactions in structure, and to replace with their equivalent truss.
Thus, having covered the entire hexagonal plane by two types of rods, we shall receive model
which will be strained just as discrete atoms in model of molecular dynamics.
52 A. V. Chentsov and R. V. Goldstein

For calculation of effective deformation characteristics of a nanotube model a condition is


used, that the strain energy at specifically chosen uniaxial loading depends only on one
characteristic of elasticity. Thus, to define the model property it is required to calculate only its
potential energy in the strained state and its dimensions.
By the results of numerical experiments it is possible to make a conclusion, that the Young's
modulus of a graphene plane model on chirality has a weak dependence, and is approximately
equal to experimental value for graphite.
When nanotubes are used as strengthening filler in polymers, the many atoms of nanotubes
interact with atoms of a polymeric matrix. These interactions can be covalent and Van-der-Waals
[2]. The constructed truss model enables to estimate the influence of external interactions on
elastic properties of a nanotube (nanofiber).
By the results it is possible to make a conclusion, that at non-covalent interactions with
participation of all atoms of a nanotube any significant loading transport onto a matrix does not
occur [3]. Only by introduction of covalent interactions significant improvement of elastic
properties of filler can be expected.
Modelling of the form of loss of stability was performed. The case of an axial compression was
investigated. The cases with the following boundary conditions were compared: the finite
displacements on the top side of a model, nodal compressive forces. In both cases the bottom row
of hexagons was fixed rigidly, the top flange was fixed in a plane, perpendicular to axis of a
nanotube. With tubes rather thin and long the form of loss of stability is close to classical for
compression of elastic rods. Shorter tubes showed behavior similar to a shell.
From comparison of results of calculation of loss of stability on the basis of nanotube truss
model with calculations by the theory of shells it follows, that with nanotube elongation the critical
loading decreases. However, stability rises with decrease of radius of a tube. It means, that tubes of
investigated diameters behave more likely as hollow shells, rather than as continuous rods.

References
1. Odegard G.M., Gates T.S., Nicholson L.M., NASA Langley Research Center, Technical
Memorandum NASA/TM-2002-211454, 2002.
2. Goldstein R.V., Chentsov A.V., Institute for Problems in Mechanics, Russian Academy of
Sciences. Preprint No. 739, 2003.
3. Goldstein R.V., Chentsov A.V., Mechanics of Solids, N.4, 2004.
1T2. Failure mechanisms 53

INTERACTION OF DOMAIN WALLS WITH DEFECTS IN FERROELECTRIC


MATERIALS

D. Schrade, R. Mueller, D. Gross, T. Utschig1, V. Ya. Shur2, D. C. Lupascu1


Institute of Mechanics, TU Darmstadt, Hochschulstr. 1, D-64289 Darmstadt
1Institute of Material Science, TU Darmstadt, Petersenstr. 23, D-64287 Darmstadt
2Institute of Physics and Applied Mathematics, Ural State University, 620083 Ekatarinburg,

Russia
schrade@mechanik.tu-darmstadt.de

Experimental studies suggest that domain wall movement in ferroelectric materials is strongly
influenced by the presence of certain kinds of defects. The intention of this paper is to study this
phenomenon using concepts of continuum mechanics to describe ferroelectric material behaviour.
Closely following the concept introduced in Mueller et al. [1], a 180° domain wall is modelled as a
singular surface across which a jump in the spontaneous polarization occurs. The domain wall can
be thought of as an inhomogeneity which allows for the application of configurational or driving
forces (cf. Gross et al. [2]), which can also be applied to crack problems.
The material behaviour is characterized by linear coupled constitutive equations for the stress
and the electric displacement. With the solution of the field equations for the mechanical and
electric problem it is possible to calculate the driving force on a domain wall. Motivated by
experiments [3], a linear relation between the domain wall velocity and the driving force on the
domain wall is postulated.
Using 2D finite element simulations, the influence of different kinds of defects on the kinetics
of a domain wall in ferroelectric-ferroelastic gadolinium molybdate, Gd2(MoO4)3 (GMO), is
studied. Fig. 1 shows a sketch of the model.
Three types of defects are considered. The first one is a surface defect, reflecting an imperfect
electrode. The second one is a hole in one side of a sample where the electrodes remain intact. The
third type involves a polarization defect in one domain. For a single planar domain wall, the
following fundamental results were found:

FIGURE 1. Model of a GMO sample containing 2 domains with electrodes attached at top and
bottom providing an external field E.

• When approaching an electrode defect, the total driving force on a domain wall decreases
significantly resulting in deceleration or even stopping of the domain wall.
• If the domain wall enters the electrode defect region, it is trapped in the middle of the
defect. Considerably higher external fields are necessary to move the domain wall out the
54 D. Schrade et al.

defect region. This and the latter result are in agreement with experimental findings where
a domain wall was stopped in front of an electrode defect and trapped inside it.
• A defect in form of a hole in one side at which the electrodes remain intact has little
influence on a moving domain wall. Simulations and experiments show that a domain wall
cannot be stopped at such a defect. A slight slow-down is predicted by simulation and
observed in experiments.
• For a polarization defect only numerical results are available. It was found that a domain
wall can be slowed down and even stopped in front of such a defect. The effect on the
domain wall is comparable to that of the electrode defect, however the domain wall is not
trapped in the polarization defect.

References
1. Mueller, R., Gross, D. and Lupascu, D.C., Comp. Mat. Sci., accepted for publication and in
print, 2005
2. Gross, D., Kolling, S., Mueller, R., Schmidt, I., Europ. J. Mechanics A, vol. 22, 669-692,
2003
3. Flippen, R., J. Appl. Phys., vol. 46, 1068-1071, 1975
1T2. Failure mechanisms 55

MICROSTRUCTURE AND INTERNAL STRESSES IN CYCLICALLY


DEFORMED AL AND CU SINGLE CRYSTALS

M. E. Kassner
Dept. of Aerospace and Mechanical Engineering, University of California
3650 McClintock OHE 430, Los Angeles, California 90089-1453
kassner@usc.edu

The concept of "long range internal stresses" (LRIS) is often utilized to explain various aspects of
the mechanical behavior of materials, including cyclic deformation and the Bauschinger effect.
These internal stresses are usually associated with the heterogeneous dislocation microstructure
(Straub et al. [1] and Tippelt et al. [2]). More specifically, it has long been suggested that long-
range internal stresses (LRIS) develop with plastic deformation in metals and alloys in association
with the development of a subgrain boundaries, cell walls, dipole bundles, etc. The evidence to
support the existence of LRIS has included Bauschinger experiments, dislocation radii
measurements, stress-dip tests and, especially, and more recently, asymmetry in x-ray diffraction
line profiles (XRD LPA) in association with the “composite model.” Convergent beam electron
diffraction, weak-beam dislocation dipole separation measurements and in-situ transmission
electron microscopy were all performed by the author on cyclically deformed Al and Cu single
crystals (Kassner et al. [3-5]). These provide insight into the mechanisms of cyclic plasticity but
especially indicate an absence of measurable LRIS.
More specifically, the dipole spacing measurements were performed by Kassner et al. [4,5],
and indicated that the dipole spacing and statistical distribution of spacings were independent of
the location in the heterogeneous substructure of cyclically deformed copper and aluminum single
crystals. The dipole spacings were the same in the channel and the veins and the maximum dipole
separations in both locations suggested that the stress to separate the dipoles was within a factor of
about one or two of the applied stress in both channels and veins. Independent work by Tippelt et
al. [6] on nickel also found dipole spacings that were independent of location. These experiments
suggest an absence of LRIS.
The extensive convergent beam electron diffraction measurements that were made on creep
deformed aluminum and copper, and cyclically deformed copper, by Kassner and co-workers,
showed that the lattice parameter was unchanged, at the equilibrium or stress-free value, within the
interior of the subgrains/dipole-bundles and along (within a one-beam diameter) the subgrain-
boundaries/dipole-bundles. If LRIS were present, a measurable variation in lattice parameter
between the interior of the subgrain, and near the subgrain boundary would be expected. It must be
considered possible that residual stresses were once present, but relaxed to an undetectable level
with sample thinning.
In-situ, reversed deformation of pre-cyclically deformed aluminum single crystals was
performed in the HVEM. These experiments allowed the imaging of dislocation motion during
“forward” and “reversed” deformation. Screws were imaged shuttling back and forth in the
channels, and their relaxation upon unloading was observed. However, no reversed motion or
bowing of dislocations was evident with unloading. If long-range internal stresses were present,
reversed motion of screws in the channels or reversed bowing of dipole bundle loops upon
unloading would be expected.
X-ray diffraction line profile analysis (XRD, LPA), specifically, the interpretation of
asymmetry in strain broadened Bragg diffraction peaks, has been extensively used to support
LRIS. A popular interpretation of asymmetry involves deconvoluting the asymmetric profile into
56 M. E. Kassner

two symmetric sub-profiles; one peak associated with elevated LRIS. This method follows from
the composite model and is used as evidence for the existence of LRIS (e.g. Straub et al. [1]).
However the author suggests that there is more than one reasonable explanation for x-ray line
asymmetry, and that LRIS is not required for asymmetry to be present. Computer simulation of x-
ray line profiles was undertaken by the author and co-workers in an attempt to determine what
arrangements may lead to asymmetry. The technique calculates atomic positions based on elastic
theory, then uses a kinematic scattering approximation in which reciprocal space intensities are
calculated using the squared Fourier transform (structure factor) of the real space atomic
arrangement. This technique was previously used successfully by Levine and Thomson [6]. In this
study, diffraction from screw dislocation dipoles was simulated, in an attempt to test the analytical
asymmetry predictions of Gaal [7]. These simulations were successful in replicating the behavior
predicted by Gaal, as the x-ray line profile from the polarized dipole ensemble was asymmetric,
and had a peak offset relative to the randomly polarized dipole ensemble. Other computer
modelling of x-ray diffraction from dislocated crystals is being performed using standard and
novel approaches, and these will be discussed.

References
1. Straub, S., Blum, W., Maier, H.J., Ungar, T., Borbely, A. and Renner, H., Acta Mater., vol.
44, 4337-4350, 1996.
2. Tippelt, B., Bretschneider, J. and Hahner, P., Phys. Stat. Sol. A, vol. 163, 11-26, 1997.
3. Kassner, M.E., Wall, M.A., and Delos-Reyes, M.A., Metall. Mater. Trans. A, vol. 28, 595-
609, 1997.
4. Kassner, M.E. and Wall, M.A., Metall. Mater. Trans. A, vol. 30, 777-779, 1999.
5. Kassner, M.E., Perez-Prado, M.-T. and Vecchio, K.S., Mater. Sci. Eng. A, vols. 319-321, 730-
734, 2001.
6. Levine, L. E. and Thomson, R., Acta Cryst. A, vol. 53, 590-602, 1997.
7. Gaal, I., In Proceedings of the 5th International Riso Symposium on Metallurgy
and Material Science, edited by N.H. Anderson et al., Roskilde, Denmark, 1984, 249-254.
1T2. Failure mechanisms 57

DETERMINATION OF EQUILIBRIUM CONFIGURATIONS OF ATOMIC


LATTICES AT QUASISTATIC DEFORMATION

S. N. Korobeynikov
Lavrentyev Institute of Hydrodynamics, 630090, Novosibirsk, Russia
korob@hydro.nsc.ru

Currently many researches simulate both initiation and propagation of cracks in a solid at the
atomic level making use of nanomechanics equations. There exist two commonly used approaches
to mathematical simulation. The first approach uses equations of molecular dynamics for
immediate simulation of initiation and propagation of cracks; the second one is based on solution
of problems of atomic lattice quasi-static deformation under the assumption that atomic lattice
buckling is a starter for initiation and growth of a crack. The present work is aimed at developing
the second approach to simulation of fracture of atomic lattices.
First solutions of problems of atomic lattice buckling were found by Novozhilov [1],
Thompson and Shorrock [2], Kornev and Tikhomirov [3]. However, the analytical methods for
solving the problems suggested by these authors cannot be applied for solving problems of
deformation and buckling of atomic lattices in the general case.
Numerical methods for solving general 2/D and 3/D problems of quasistatic deformation of
atomic lattices have been developed in Korobeynikov [4], Dluzewski and Traczykowski [5],
Korobeynikov [6]. In the typical case, the Cauchy problem is solved in the form

K ( U ) U R , U (0) U 0,
(1)

where U is the displacement vector of atoms in a lattice, R is the vector of internal forces acting
on lattice atoms, K is the symmetric tangential stiffness matrix of a lattice, a superimposed dot
denotes derivative of the magnitude with respect to the monotonically increasing deformation
parameter t . The matrix K is determined by summing tangential stiffness matrices of all atomic
pairs in the lattice. System (1) is solved by step-by-step integration with iteration refinement of
solution for each discrete value of the parameter t .
Atomic lattice buckling takes place when singular points on the integral curve are attained, i.e.,
at the points of the matrix K degeneracy:

det K 0. (2)
Singular points can be turning points or bifurcation points as well as turning and bifurcation points
simultaneously. At the bifurcation points there exist several continuations (branches) of solutions.
In the vicinity of singular points, the arc-length method of step-by-step integration of equations (1)
is most appropriate. In this case, the solution process continues through the turning points without
any difficulties. However when bifurcation of solution of problem (1) occurs, it is desirable that all
branches of the solution be determined. To do this, we solve the auxiliary problem on
determination of eigenvectors W :

KW 0 (3)
To determine equilibrium configurations, we propose to introduce small perturbations into
potential parameters of atom pairwise interaction in conformity with the form of eigenvector W .
58 S. N. Korobeynikov

This method of continuation of solutions by the side branch is proposed as alternative to more
exact but, at the same time, more complicated method presented by Sokol and Witkowski [7].
Presented procedures of solution continuation through singular points of integral curves
require as exact definition of a matrix K as possible in order to improve a convergence as well as
for prevent a divergence of iteration processes applied for the solution refinement. The expressions
refined in comparison with expressions given in [4,6], which account for both tension/compression
of segments connecting atomic pairs and their rotations have been proposed by Korobeynikov [8].
The approaches to solution of problems of atomic lattices deformation in the presence of
singular points on integral curves are realized in the PIONER code (cf., Korobeynikov et al. [9]).
Using this code, solutions of some problems of atomic lattice buckling are obtained, which show
the efficiency of algorithms mentioned above.
The supports from Russian Foundation for Basic Research (04-01-00191) and Integration
Project of Russian Academy of Science No. 3.11.1 are gratefully acknowledged.

References
1. Novozhilov, V.V., Prikladnaja Matematika i Mekhanika, vol. 33, 797-812, 1969 [in Russian].
2. Thompson, J.M.T. and Shorrock, P.A., J. Mech. Phys. Solids, vol. 23, 21-37, 1975.
3. Kornev, V.M and Tikhomirov, Yu.V., J. Appl. Mech. Techn. Phys., vol. 34, No. 3, 439-448,
1993.
4. Korobeynikov, S.N., Application of the FEM to solving nonlinear problems of deformation
and buckling of atomic lattices, Preprint No 1-97, Inst. Hydrodynamics, Sib. Div., Russ.
Acad.of Sci., Novosibirsk, Russia, 1997 [in Russian].
5. Dluzewski, P. and Traczykowski, P., Arch. Mech., vol. 55, 393-406, 2003.
6. Korobeynikov, S.N., Int. J. of Fracture, vol. 128, 315-323, 2004.
7. Sokol, T. and Witkowski, M., In Proceedings of the Second International Conference on
Computational Structures Technology: Advances in Non-linear Finite Element Methods,
edited by M. Papadrakakis and B.H.V. Topping, Civil-Comp Press, Edinburgh, 1994, 35-45.
8. Korobeynikov, S.N., Arch. Mech. (to be published).
9. Korobeynikov, S.N., Agapov, V.P., Bondarenko, M.I. and Soldatkin, A.N., In Proceedings of
the International Conference on Numerical Methods and Applications, edited by B. Sendov et
al., Publ. House of the Bulgarian Acad. of Sci., Sofia, 1989, 228–233.
1T2. Failure mechanisms 59

MULTISCALE MECHANICS OF CARBON NANOTUBES AND THEIR


COMPOSITES

X.-Q. Feng
Department of Engineering Mechanics, Tsinghua University,
Beijing 100084, China
fengxq@tsinghua.edu.cn

Since the discovery of carbon nanotubes (CNTs) by Iijima in 1991, interest in carbon
nanotechnology has grown very rapidly because of the unique, often enhanced, properties of
nanoscale materials. Owing to their superior mechanical and physical properties, carbon
nanotubes seem to hold a great promise as an ideal reinforcing material for advanced composites of
high-strength and low-density. However, in most of the experimental results, only modest
improvements in the strength and stiffness have been achieved by incorporating carbon nanotubes
in polymers. In this talk, the mechanical properties of carbon nanotubes and their composites will
be investigated by multiscale mechanics methods.
There are many factors that influence the overall mechanical property of CNT-reinforced
composites, e.g., the weak bonding between CNTs and matrix, the curviness and agglomeration of
CNTs. Even though the adhesion strength between the CNTs and the matrix may significantly
affect the failure behavior of composites (e.g., the ultimate tensile strength and fracture toughness),
its influence on the effective elastic modulus of composites can be negligible. Therefore, it is
thought that two most significant reasons that limit the effective elastic property of CNT
composites are the curve shape and agglomeration of CNTs due to their nanometer diameters and
large aspect ratios. In the present paper, the effects of the widely observed waviness and
agglomeration of carbon nanotubes are examined theoretically. The Mori-Tanaka effective-field
method is first employed to calculate the effective elastic moduli of composites with aligned or
randomly oriented straight CNTs. A novel micromechanics model is then developed to consider
the waviness or curviness effect of CNTs which is assumed to have a helical shape. Finally, the
influence of agglomeration of CNTs on the effective stiffness is analyzed, and analytical
expressions are derived for effective elastic stiffness of CNT-reinforced composites accounting for
the effects of waviness and agglomeration. It is established that these two mechanisms may
significantly reduce the stiffening effect of CNTs. The present study not only provides the
important relationship between the effective properties and the morphology of CNT-reinforced
composites, but also may be useful for improving and tailoring their mechanical properties.
It is of great interest to gain a deep understanding of fracture behaviors of CNTs. We use here
a hybrid atomistic/continuum mechanics method to simulation defect nucleation and facture of
CNTs under tension or torsion. Under a lower tensile strain, a CNT undergoes uniform
deformation, with the positions of atoms being determined by using the modified Cauchy-Born
rule and the Tersoff-Brenner potential. When the tensile strain reaches a critical value, defects may
nucleate as a result of the so-called Stone-Wales transformation. We use the atomistic-based
continuum mechanics theory to determine the displacements of atoms far from the defect, and an
atomistic mechanics method to calculate the positions of atoms in a local subregion around the
defect. It is found that the critical strain of defect nucleation and the subsequent fracture modes of
CNTs are sensitive to the chiral angle, and that the critical strain of a zig-zag CNT is about two
times of that of an armchair one. At low temperature, both armchair and zigzag CNTs fracture in a
brittle manner. Our numerical results on the fracture strains and tensile curves compare well with
experimental results.
60 X.-Q. Feng

The exceptional mechanical properties of CNTs make this new form of carbon an excellent
candidate for composite reinforcement. However, studies on deformation and fracture of CNTs in
composites are difficult both in experimental observation and in theoretical modeling. The above
method is also extended to analyze the fracture problem of CNTs embedded in a composite matrix.
In our hybrid atomistic/continuum method, the unit cell is divided into three zones, A, B and C,
which are dealt with in different manners according to their deformation features and the numbers
of atoms in them. The deformation of CNTs is constrained partly by the surrounding matrix.
Interaction among CNTs is considered by the Mori-Tanaka method. The calculation results of the
critical strains of defect nucleation of CNTs are given in Fig. 1. Our results show that both
armchair and zigzag CNTs in a composite are easier to fracture due to the effect of CNT-matrix
interaction.

FIGURE 1. Critical strains of defect nucleation upon the chiral angles and diameters of CNTs.

References
1. Shi, D.L., Feng, X.Q., Huang, Y., Hwang, K.C. and Gao, H., Trans. ASME, J. Eng. Mater.
Tech., vol. 126, 250-257, 2004.
2. Shi, D.L., Feng, X.Q., Jiang, H.Q., Huang, Y. and Hwang, K.C.., Int J. Fracture, (in press).
1T4. Fatigue and Fracture of MEMS and NEMS 61

IN-SITU SCANNING ELECTRON MICROSCOPE INDENTATION OF


GALLIUM ARSENIDE

Pouvreau C., Wasmer K., Giovanola J.1, Michler J., Breguet J. M.1 and Karimi A.1
Swiss Federal Laboratories for Materials Testing and Research (EMPA)
Feuerwerkerstrasse 39, 3602 Thun, Switzerland
1
Ecole Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland
cedric.pouvreau@empa.ch

Indentation is a commonly used method to evaluate the fracture toughness of brittle materials
through crack length measurements. Many models for toughness estimations are available and rely
on the assumption that surface traces of the cracks are either of the Half Penny type, Lawn and
Evans, [1], Lawn, et al., [2], or of the radial type, Niihara, et al., [3], Niihara, [4]. However, it has
been shown that, especially at low loads, the crack morphology depends strongly on material and
thus this dependence affects calculated toughness values, Cook and Pharr, [5].

Figure 1: SEM micrographs taken during an in-situ indentation experiment showing the cracking
sequence : a) Half-load (250 mN); b) Maximum load (500 mN); c) Half unload (250 mN); d) Full
unload.

This paper reports on the results of indentation experiments on Gallium Arsenide (GaAs) with
two types of indenters (conical with 60° or 120° apex angles and cube corner) using two novel
experimental techniques, (1) in-situ indentation in the Scanning Electron Microscope (SEM) while
recording indentation load and displacement and (2) sectioning of indent zones by cleavage to
reveal the subsurface crack morphology. The first technique, described by Rabe, et al., [6] serves
to establish the cracking kinetics by correlating surface cracking observations with load-
displacement histories. The second technique is used to establish accurately crack shapes,
orientations and sizes. In this cleavage technique, a starter crack is generated by scratching the
sample along the [110] direction with an indenter. Then, five series of five indents each are made
with a Nanoindenter XP at a given load level, each series shifted by 10 µm with respect to the
preceding one, perpendicularly to the scratch direction in the indentation plane. Finally, the starter
crack is propagated through the whole set of indents with a home made cleaving device.
The indenter geometries were chosen to study the influence on crack generation of
axisymmetric and singular stress fields, respectively. Moreover, for the wedge and conical
indenters, 60° and 120° apex angles have been used to evaluate the indentation fracture
dependence on indenter geometry. One edge of the cube corner indenter was aligned with the [10]
direction.
The following observations were made combining the two novel techniques.
62 C. Pouvreau et al.

The cracking sequence of an in-situ indentation with the 60° conical indenter is presented on
Figure 1a) to d).
For conical indentations, radial cracks are initiated at loads lower than 50 mN. Manufacturing
imperfections on the conical indenter strongly influence the crack initiation site and conditions. At
500 mN, chipping out occurs at the end of the loading cycle when at loads lower than 250 mN it
generally occurs at the final stage of the unloading. Generally, this chipped region has the shape of
a Maltese cross. This pattern is commonly observed in glasses and is generally associated with the
residual stress field.
During indentations with the cube corner, important cracks emanate from the indenter corner
during the loading part of the cycle, followed by secondary cracks from the indenter faces. These
cracks have an angle of 45° with the [110] direction and could be the traces of cracks propagating
in a {110} plane. Additionally, traces of slip bands are easily visible at the surface of the indented
material as previously observed (Fujita, et al., [7]).
These results demonstrate that the combination of in-situ SEM indentation and transverse
cleaving through indents opens new perspective for crack investigation in brittle single crystals
since crack morphologies, cracking sequence and cracking conditions can all be accurately
determined and correlated, for a broad range of indentation parameters.

References
1. Lawn, B. R. and Evans, A. G., J Mater Sci, 12,2195-2199, 1977
2. Lawn, B. R., et al., J Am Ceram Soc, 63,574-581, 1980
3. Niihara, K., et al., Journal of Materials Science Letters, 1,13, 1982
4. Niihara, K., Journal of Materials Science Letters, 2,221, 1983
5. Cook, R. F. and Pharr, G. M., J Am Ceram Soc, 73,787-817, 1990
6. Rabe, R., et al., Thin Solid Films, 469-70,206-213, 2004
7. Fujita, S., et al., Philos Mag A, 65,131-147, 1992
1T4. Fatigue and Fracture of MEMS and NEMS 63

FRACTURE OF NANOSTRUCTURED LITHIUM BATTERIES

K. E. Aifantis, J. P. Dempsey1 and S. A. Hackney2


Univ. of Cambridge & Univ. of St Petersburg
Wilberforce Road, Cambridge CB3 0AL, UK
1
Clarkson Univ., 8 Clarkson Avenue, Potsdam, NY 13699-5710, USA, jdempsey@clarkson.edu
2
Michigan Technological Univ., 1400 Townsend Dr., Houghton, MI 49931, USA,
hackney@mtu.edu
kaifantis@mom.gen.auth.gr, ka279@damtp.cam.ac.uk

Lithium-ion batteries have gained considerable attention during the past twenty years due to their
favourable properties over nickel-cadmium and nickel-metal-hydride batteries. Unlike the
aforementioned material systems Li is non-toxic and it is the third lightest element. Moreover, the
high energy density of Li batteries reduces their weight by half and their volume by 20% to 50%
compared to the same capacity of nickel-cadmium and nickel-metal-hydride batteries. However,
the use of pure Li as a negative electrode is not feasable due to safety and reliability concerns.
Therefore, extensive research is being performed to find suitable material canditates that form
alloys with Li. Unfortunately, many materials that form Li alloys with the required high volumetric
density of Li also show over 100% volume change during the discharge and charge of the battery,
leading to fracture of the electrode material and associated loss of capacity. Thus, materials designs
are being explored which can stabilize the mechnical integrity of the electrode material while still
providing large volumetric Li density.
In particular, battery developers have suggested that composite materials which contain
domains of electrochemically active and inactive materials may provide resistance to mechanical
degradation while still providing capacities between 900 and 4000 mAh g-1, Beaulieu et al. [1].
Graphitic carbon, which is the material that is presently being used for the anode by Li ion battery
manufacturers, gives a much lower Li capacity of 372 mAh g-1, Graetz et al. [2]. These active/
inactive composites typically comprise of a metal inclusion (such as Si, Sn, Al, Bi) that forms rich
Li compounds and is surrounded by an inert ceramic or glass, as shown in Fig. 1.

FIGURE 1. Idealized geometry of the electrode: Li-insertion particles (shaded) embedded in a


glass (blank) matrix. A unit cell is defined by a circle of radius b surrounding a circular particle of
radius a.

However, these composite materials designs have not been optimized for fracture resistance
and Li capacity. We consider an idealized analytical model to consider some major design
parameters for this problem. In particular, to model fracture of the anodes, the unit cell of Fig. 1 is
divided in three zones: the active site, the damage zone that develops at the active site/matrix
interface, and the undamaged matrix (see Fig. 2). It should be noted here that experimental studies
suggest that if the active sites are of nanometre size and are surrounded by a ceramic, the surface of
64 K. E. Aifantis et al.

the active site does not fracture, and hence the damage zone consists only of ceramic nanoparticles.

FIGURE 2. Damage in unit cell of anode: a and b denote the radii of the active site and matrix,
respectively, ' denotes the free expansion the active site would undergo during charging, and U-a
denotes the crack length.

The present authors have modelled the internal stresses that develop during the charge and
discharge of the battery, Aifantis and Hackney [3], and they have also predicted the stability of
these systems, Aifantis and Dempsey [4]. The purpose of this study is to determine the critical
crack length at which the electrode will fracture. Furthermore, the effect that the size of the active
site has to overall capacity of the battery will explored and finally, predictions will be made for the
values of a and b that will result in no cracking.

References
1. Beaulieu, L. Y., Eberman, K. W., Turner, R. L., Krause, L. J., Dahn, J. R., Electrochem.
Solid-State Lett., vol. 4 (9), A137, 2001
2. Graetz, J., Ahn, C.C., Yazami, R., Fultz, B., Electrochem. Solid-State Lett., 6(9), A194, 2003
3. Aifantis, K.E., Hackney, S.A., J. Mech. Behav. Matls, 14, 413, 2003
4. Aifantis, K.E., Dempsey, J.P, J. Power Sources, 143, 203-211, 2005
1T4. Fatigue and Fracture of MEMS and NEMS 65

ANALYTICAL AND EXPERIMENTAL CHARACTERIZATION OF A


MICROMIRROR SYSTEM

E. J. Pryputniewicz, C. Furlong1, and R. J. Pryputniewicz1


Current address: Department of Homeland Security, Science and Technology, Alexandria, VA
1Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST

rjp@wpi.edu

Continued demands for miniaturization and increased functionality of microelectromechanical


systems (MEMS) require construction of complex three-dimensional structures. Recent
developments at Sandia National Laboratories allow construction of such structures from parts
fabricated in a plane of a wafer, by actuating them to extend hundreds of microns in the direction
normal to the fabrication plane. One such structure is a hinged positionable micromirror system
actuated by an electrostatic microengine, Fig. 1. It should be emphasized that this structure is
batch fabricated with no piece part assembly required and that it is actuated using only the on-chip
microengine; the torque delivered by the microengine is amplified by the transmission, which, in
turn, pushes a linear rack that positions the hinged micromirror, Fig. 2. In the configuration
shown, the micromirror can be used to reflect light beam capable of triggering, or activating,
sensors and other circuitry when it is in its elevated position. Therefore, repeatability of a reflected
beam impinging on a target, changes in performance of the micromirror over time, and distortion
of its surface during activation are just some of the characteristics that must be determined with
high accuracy and precision. Until recently, this characterization was hindered by lack of suitable
methodologies. However, building on advances in photonics, electronics, and computational
analyses, we have developed a new methodology Pryputniewicz et al. [1], using fiber-based laser
optoelectronic holography system, Fig. 3, to quantitatively characterize micromirror systems in
motion, Figs 4 and 5. In addition, in order to determine forces acting on various components of the
system, we have also developed analytical models to study kinematics and kinetics of these
components, based on vector calculus Pryputniewicz [2], Pryputniewicz [3]. According to this
model, magnitudes of forces, acting on the smallest gear (60 Pm diameter) in the system, range
from 4 nN to 27 Pm, as a function of rotational speed. The corresponding displacements of the
micromirror are 113 Pm.
In this paper, the methodology for analytical and experimental characterization of a
micromirror system is described and its use is illustrated with a representative case study. By
characterizing performance of the micromirror system, we can make specific suggestions for their
future improvements and we can verify the effect of these improvements.
The micromirror systems used in this study were fabricated at and provided by Sandia National
Laboratories. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed
Company, for the United States Department of Energy under Contract DE-AC04-94AL85000.
66 E. J. Pryputniewicz et al.
1T4. Fatigue and Fracture of MEMS and NEMS 67

A METAL INTERPOSER FOR ISOLATING MEMS DEVICES FROM


PACKAGE STRESSES

R. J. Pryputniewicz, T. F. Marinis1, J. W. Soucy1, P. Hefti, and A. R. Klempner


Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST
1Draper Laboratory, Electronic Packaging Department, Cambridge, MA

rjp@wpi.edu

Many classes of MEMS devices, such as those with resonant structures, capacitive readouts, and
diaphragm elements, are sensitive to stresses that are exerted by their surrounding package
structure. Such stresses can arise as a result of changes in temperature, ambient pressure, or
relative humidity. We have demonstrated a dramatic reduction in scale factor bias over
temperature for a tuning fork gyroscope by mounting it on an interposer structure within a
conventional chip carrier, Fig. 1. Optimization of a MEMS sensor package for high performance
subject to various constraints cannot be accomplished by analysis alone Hanson et al. [1]. There
are too many unknown parameters, e.g., material properties, process conditions, and components/
package interface conditions, to make this feasible. Extensive performance evaluation of packaged
sensors is also prohibitively expensive and time consuming. However, recent advances in
optoelectronic laser interferometric microscope (OELIM) methodology Furlong and
Pryputniewicz [2] offer a considerable promise for effective optimization of the design of
advanced MEMS components and MEMS packages. Using OELIM, sub-micron deformations of
MEMS structures are readily measured with nanometer accuracy and very high spatial resolution
over a range of environmental and functional conditions. This greatly facilitates characterization
of dynamic and thermomechanical behavior of MEMS components, packages for MEMS, and
other complex material structures. In this paper, the OELIM methodology, which allows
noninvasive, remote, full-field-of-view measurements of deformations in near real-time, is
presented and its viability for development of MEMS is discussed. Using OELIM methodology,
sub-micron displacements of sensors can be readily observed and recorded over a range of
operating conditions, Fig. 2. In addition, detailed mapping of deformation fields due to process
conditions can also be made, Fig. 3. In this case, the OELIM results clearly show that the proof
masses are not flat, but rather exhibit curvature, with the maximum deviation from planarity of
1.05 Pm. This curvature, which affects performance of the sensor, is due to residual stresses
generated during fabrication of the device. Where applicable, the OELIM measurements are
coupled with the corresponding analytical and computational modeling results, in order to validate
and refine quantitative models of packages and complex material structures Przekwas et al. [3] and
Pryputniewicz et al. [4] as well as to develop operational relationships for MEMS structures,
which offer considerable promise for effective optimization of design of advanced sensor packages
and their fabrication processes.
68 R. J. Pryputniewicz et al.

References
1. Hanson, D. S., T. F. Marinis, C. Furlong and R. J. Pryputniewicz, “Advances in optimization
of MEMS inertial sensor packaging,” Proc. Internal Congress on Applied Mechanics for
Emerging Technologies, Portland, OR, pp. 821-825, 2001.
2. Furlong, C., and R. J. Pryputniewicz, “Absolute shape measurements using high-resolution
optoelectronic holography methods,” Opt. Eng., 39:216-223, 2000.
3. Przekwas, A. J., M. Turowski, M. Furmanczyk, A. Hieke, and R. J. Pryputniewicz
“Multiphysics design and simulation environment for microelectromechanical systems,”
Proc. Internat. Symp. on MEMS: Mechanics and Measurements, Portland, OR, pp. 84-89,
2001.
4. Pryputniewicz, R. J., P. Galambos, G. C. Brown, C. Furlong, and E. J. Pryputniewicz, “ACES
characterization of surface micromachined microfluidic devices,” Internat. J.
Microelectronics and Electronic Packaging (IJMEP), 24:30-36, 2001.
1T4. Fatigue and Fracture of MEMS and NEMS 69

COMPUTATIONAL MODELING OF NANOPARTICLES IN BIOMICROFLUIDIC


DEVICES

R. J. Pryputniewicz, Z. Sikorski1, M. Athavale1, Z. J. Chen1 and A. J. Przekwas1


Worcester Polytechnic Institute, ME/CHSLT-NEST, Worcester, MA USA
1CFD Research Corporation, Huntsville, AL, USA

rjp@wpi.edu

Focused laser beam can be used to trap and manipulate small particles without mechanical contact.
Particles trapped by laser tweezer can be used as local force probes in biomicrofluidic
environment. Optical tweezer forces are in the pico-Newton range, which makes them suitable for
study of physics of biological objects like cells, viruses, bacteria, and research in genetics.
Theoretical models of optical forces acting on particles much smaller than wavelength of light
(Rayleigh regime) are based on dipole model of interaction of light with a dielectric particle. Total
optical force can be split into gradient force, which drags particle toward maximum of laser beam
intensity in the focal region, and scattering force, which pushes particle in the direction of Poynting
vector of laser light. Ratio of scattering force to gradient force increases with a particle radius.
Stable trap requires gradient force to be larger than scattering force. If potential well produced by
optical force field is higher than other forces acting on particle (e.g., viscous, Brownian, biological
interactions) it remains in the trap. Optical manipulation of nano-beads attached to a complex
biomolecule allows to rotate, stretch, and cut it, measure kinetic constants of binding, and study
behavior of molecular motors.
This paper presents a model of optical manipulation of nanoparticles in a scanning laser beam
Sikorski et al. [1] Laser beam is fast scanned in one direction and slow scanned in the other.
Optical field time averaged over the fast scan builds an optical trap. Slow scan in the second
direction allows dragging particles by this elongated optical trap. Formulas for optical forces are
derived for this case. CFDRC multiphysics solver ACE+ has been used for transient simulation of
particle manipulation in biomicrofluidic devices. Figure 1 shows an H-filter which is 1 mm high.
In this filter, the channel width is 20 Pm and beads of 70 nm radius, made of material with index of
refraction of 1.56, are used. The slow scan velocity, transverse to the channel, is 10-5 m/s.
Observed is particle rotation and sweeping for different shapes of particles. Device heating and
influence of temperature distribution on particle and fluid movement are presented. Figure 2
shows an enlarged snapshot of the H-filter separation region, after 1500 time steps of 50 ms each.
Although the laser beam is not shown in Fig. 2, its effects are visible: beads change from RHS
stream to the LHS stream. If the laser were not present, the beads would continue in the RHS
stream. Since this effect depends on the size and properties of the beads (i.e., particles, molecules,
etc.) it can be used to separate them. Continued work will result in additional examples.
70 R. J. Pryputniewicz et al.

References
1. Sikorski, Z., M. Athavale, Z. J. Chen, A. J. Przekwas, and R. J. Pryputniewicz, “Modeling of
optical trapping and manipulation of nanoparticles in biomicrofluidic devices,”, Proc. 4th
Internat. Symp. on MEMS and Nanotechnology (4th-ISMAN), Charlotte, NC, pp. 350-357,
2003.
1T4. Fatigue and Fracture of MEMS and NEMS 71

CHARACTERIZATION OF A MEMS PRESSURE SENSOR BY A HYBRID


METHODOLOGY

R. J. Pryputniewicz and C. Furlong


Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST
rjp@wpi.edu

Recent advances in surface micromachining and microelectromechanical systems (MEMS)


technology have led to development of a MEMS pressure sensor, Fig. 1a. In this sensor, pressure
changes are detected by deformations of a diaphragm, Fig. 1b. This diaphragm is about 400 Pm
long, 100 Pm wide, and about 2 Pm thick, made of several different materials, each having
different properties, especially different coefficients of thermal expansion (CTE). As a result, as
the sensor is exposed to the environment, where in addition to pressure changes temperature
changes also, its measurements will be affected by thermomechanical response.
The objective of this paper is to develop a hybrid, computational and experimental,
methodology for characterization of thermomechanical behavior of MEMS pressure sensors. The
computational development is based on a finite element method (FEM). The experimental
development is based on the state-of-the-art optoelectronic laser interferometric microscope
(OELIM) method Pryputniewicz et al. [1].
Accurate and precise pressure measurements by MEMS sensors are limited by the effects that
environmental temperature variation has on their performance. As the sensor is exposed to a
changing pressure, the diaphragm deforms. Deformations of the diaphragm cause changes in
resistance of the bridge circuit, which is an integral part of the MEMS sensor. In this study the
sensors were subjected to a differential pressure of 0.2 MPa and a temperature ranging from 10qC
to 50q C. The analyses were performed for the combined pressure and thermal loads.
Representative results are shown in Figs 2 and 3. Figure 2 shows typical fringe patterns due to
changes in the differential pressure. Ideally, there should be no fringes when the sensor is at rest,
i.e., at atmospheric pressure. However, because of residual stresses due to fabrication, the
diaphragm is deformed by about 40 nm. Also, deformation nonlinearity, due to loadings by a
positive and a negative pressure differences of the same magnitude, is vividly displayed by the
corresponding fringe patterns; the negative pressure difference yields the maximum deformation of
792 nm, while the positive one yields the magnitude of 813 nm, resulting in a difference of 21 nm,
which is significant while interpreting the results. Computationally determined deformations and
stresses due to the positive pressure difference are shown in Fig. 3. Comparison of computational
and experimental results indicates good correlation and shows that the hybrid methodology is very
effective for characterization of MEMS pressure sensors.
72 R. J. Pryputniewicz and C. Furlong
1T4. Fatigue and Fracture of MEMS and NEMS 73

NEW APPROACH TO SYNTHESIS OF LASER MICROWELDING


PROCESSES FOR PACKAGING

R. J. Pryputniewicz, W. Han and K. A. Nowakowski


Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST
rjp@wpi.edu

Microelectronics and packaging industry has ever-increasing requirements for miniaturized


features with high edge acuity and negligible thermal damage zone. Laser micromachining is a
unique way of processing materials with less thermal distortion and minimum metallurgical
damage to the workpiece, therefore it is a good alternative to conventional micromachining
processes. Typical applications of laser micromachining include laser bonding of a wafer, or of a
microsystem chip, to one or two parts Mescheder et al. [1], and laser micromachining of 3D
microchannel systems in chemical, biomedical, DNA, and environmental science analyses Qin and
Li [2].
In laser micromachining processes, such as laser welding, drilling, and cutting, the materials
experience heating, melting, evaporating, and re-solidifying stages. As a result, the laser
micromachined components can be affected by a number of factors such as laser beam properties,
system-cooling conditions, and surface roughness/reflectivity of the areas exposed to laser light.
This paper presents synthesis of the laser microwelding processes for packaging. More
specifically, various parameters are evaluated individually and then together to determine their
influence on the finished product. This evaluation is done analytically, computationally, and
experimentally Han and Pryputniewicz [3]. Analytical and computational results yield
temperatures in the heat affected zones (HAZs), Fig. 1. In order to make these results valid, they
are correlated with time dependent temperature measurements, Fig. 2; to facilitate this correlation,
the results shown were normalized. Thermal gradients developed during laser microwelding cause
deformations of surfaces exposed to laser beam. Successful development of advanced packaging
depends on accurate knowledge of these deformations. In order to quantify surface deformations
due to laser microwelding, we have developed an optoelectronic methodology based on the use of
CCD cameras for high-resolution imaging of the affected areas. Figure 3 shows a representative
fringe pattern recorded for one of the HAZs. Interpretation of this fringe pattern produces detailed
spatial distribution of deformations, Fig. 4; the maximum deformations shown are 2.5 Pm, which
may lead to fractures. In fact, in order to understand the influence that a laser beam has on the
materials, we make these measurements before, during, and after the laser microwelding, based on
which we will optimize the processes to obtain minimal deformations.
Comparison of preliminary analytical, computational, and experimental results shows good
correlation and indicates viability of the approach we have developed for synthesis of laser
microwelding processes for packaging.
74 R. J. Pryputniewicz et al.
1T4. Fatigue and Fracture of MEMS and NEMS 75

THERMAL MANAGEMENT OF RF MEMS RELAY SWITCH

R. J. Pryputniewicz
Worcester Polytechnic Institute, Mechanical Engineering Department / CHSLT-NEST
rjp@wpi.edu

Radio frequency (RF) switches are one of many MEMS devices that make it possible to
communicate, sense, and measure while using minimal amount of space and very low power. The
RF microswitches have either capacitive or resistive configuration. The capacitive switches use a
flexible membrane design, in which capacitance between two electrodes is induced via electrical
voltage; reaching threshold capacitance activates the switch, which enables transmission of a
signal. The resistive switches, on the other hand, make direct metal to metal contact. Such design
usually uses a cantilever beam that bends as voltage is applied to the two electrodes.
In the RF MEMS contact switch configuration, a moving component is a cantilever beam with
a shorting bar at its free end, Fig. 1. The shorting bar comes into contact with the traces on the
substrate as the beam deflects under the electrostatic force loading. As the shorting bar makes
contact with the traces, i.e., as it closes the path, electric signals pass from one trace to the other.
Because of finite internal resitances of the components as well as resistance of the contact
interfaces, Joule heat is generated, affecting performance of the switch.
In this paper, we investigate the Joule heating effects computationally using thermal analysis
system (TAS) software. TAS models the structure to be analyzed by geometrically simple finite
elements, Fig. 2, convert these elements into resistors representing all three modes of heat transfer:
conduction, convection and radiation, and then solves the resistor network using the finite
difference solver, which performs heat balance at each node of the model Pryputniewicz et al. [1].
This entails calculating a node temperature based on the resistances and the temperatures of all
nodes attached to the node in question. During TAS model execution, temperature, or time
dependencies are interpolated for each computation time step.
Representative results obtained for 300 mA current are displayed in Figs 3 and 4. Figure 3
shows that the shorting bar reaches temperatures from 680qC to 717qC, for the geometry,
dimensions, and the material properties considered in this case; for other set of parameters the
results would be different. Figure 4 shows that temperatures in the traces range from 25qC to
45qC. It should be observed that the 665qC temperature difference between the components shown
in Figs 3 and 4 is due to effective heat transfer between the individual components of the switch as
well as with its environment.
Results generated in this study show that TAS provides an effective approach for thermal
management of RF MEMS switch designs.
76 R. J. Pryputniewicz
1T7. Thin films 77

BUCKLING AND DELAMINATION OF THIN LAYERS ON A POLYMER


SUBSTRATE

A. A. Abdallah, D. Kozodaev1, P. C. P. Bouten2, J. M. J. den Toonder2 and G. de With


Laboratory of Materials and Interface Chemistry, Eindhoven University of Technology
Den Dolech 2, 5600 MB Eindhoven, The Netherlands
1
NT-MTD Co./NTI-Europe. Arnhemseweg 34 d, 7331 BL Apeldoorn, The Netherlands
2
Philips Research Laboratories, Prof. Holstlaan 4, 5656 AA Eindhoven, The Netherlands
a.abdallah@tue.nl

Flexible displays are multi-layered structures consisting of a polymer-based substrate sandwiched


between a numbers of functional layers that act as a gas barrier layer, e.g. inorganic thin silicon
nitride layer (Si3N4), and a transparent conducting oxide layer, e.g. indium tin oxide layer (ITO).
For the realization of flexible displays, a study of the mechanical integrity of multi-layered
structures and their reliability is required.
High compressive stresses in the thin layer in combination with insufficient adhesion at the
interface can cause the layer to buckle and to delaminate from the substrate (see Fig.1). These
buckling phenomena are undesirable from a functional and reliability point of view and therefore
should be avoided.

FIGURE 1: Atomic force microscope image obtained for buckling and delamination of a 400 nm
thin Si3N4 layer on a polymer substrate. Circular buckles with layer cracking due to uniaxial-
compressive external stress.

In the framework of buckling theory for a beam clamped on both edges, numerous buckle
morphologies have been studied for various thin layer structures on substrates, often neglecting the
effect of the substrate’s deformation. Recently, attention is paid to buckling on compliant
substrates [1, 2, 4]. When the substrate deformation is taken into account, significant effects of the
energy release rate on buckles are described, leading to lower critical strains for buckling onset.
Bouten and Van Gils [4] present buckling maps for compliant substrates; buckle sizes (height and
width) are related to adhesion energy and internal strain (prior to buckling). Yu and Hutchinson [2]
define a characteristic length l, adjacent to the buckle, contributing to the energy release upon
buckling. This length is given as

2 hf
l=
1D (1)
78 A. A. Abdallah et al.

where hf is the layer thickness, D is the elastic mismatch (Dundurs’ parameter [3]).
Assuming that only the elastic energy release of the buckling layer contributes to the buckling
process, the energy balance for buckling can be written as [4]

2bGss=2(b  l )(Go  Gc) (2)


where Go is the elastic energy stored in a thin layer per unit width, Gc is the remaining energy in
the buckled portion of the layer and b is characteristic buckle size. The right hand term describes
the energy release in the thin layer upon buckling as a function of the effective buckle length (b +
l), which depends on the layer thickness and the elastic mismatch. As a consequence significantly
more elastic energy is available for buckling formation. The steady-state energy release rate of the
advancing buckle Gss is an important parameter to describe layer buckling and delamination. The
left hand term represents the energy required for delamination.
In this paper attention is paid to the influence of loading mechanism, layer thickness, and level
of adhesion on buckle initiation and propagation. Based on the results of the experimental work, a
numerical model has been formulated to describe buckling and delamination phenomena in these
structures using the J-integral and cohesive zone elements. Different types of buckle profiles are
presented and the buckle evolution mechanism associated with the applied stress is discussed.

The results showed that the buckle onset strain defined at a given quality of adhesion for a
compliant substrate is lower than the one predicted by the model for a rigid substrate [5]. Apart
from that, the critical buckling strain Hc is found to be reduced for a stiff layer on a compliant
substrate due to the rotation and the displacement at the buckle edges. The buckle size is found to
increase with the layer thickness and the uniaxial-compressive strain.

References
1. B. Cotterell and Z. Chen, Int. J. Fracture 104, 169-179, 2000.
2. H.H. Yu and J.W. Hutchinson, Int. J. Fracture 113, 39-55, 2002.
3. J. Dundurs, J. App. Mech. 36, 650-652, 1969.
4. P.C.P. Bouten and M.A.J van Gils, In Proceedings of MRS, Boston, 2004. T4.9
5. J.W. Hutchinson and Z. Suo, Adv. Appl. Mech. 29, 63-191, 1992.
1T7. Thin films 79

CARBIDE COATED CUTTING TOOL PROPERTIES INVESTIGATION BY


NANO-MECHANICAL MEASUREMENTS UNDER 250-500°C.

B. Vasques, D. Joly, R. Leroy, N. Ranganathan and P. Donnadieu1,2


LMR, Productique Department, University François Rabelais de Tours,
7 rue Marcel Dassault, 37204 Tours, France
1
SAFETY Production SA, rue Henri Garih, 37230 Fondettes, France
2
CEROC Cutting Tool Research Centre, rue Henri Garih, 37230 Fondettes, France
brigitte.vasques@groupe-safety.com, rene.leroy@univ-tours.fr, ranganathan@univ-tours.fr

Since most mechanical properties are temperature dependent. This one of the reasons that usual
mechanical tests at room temperature are insufficient to characterize advanced PVD coatings
designs. Many of these coating inserts are exposed to high temperatures in either processing or
working environments such as high speed machining or dry cutting operations [1, 2, 3, 4]. Wear
reaction between tool and work piece takes place at high temperature [5]. We need to study these
materials under such environment to really understand and predict there behavior. That’s why wear
resistance of PVD hard coatings are investigated with a particular attention to their critical
mechanical properties at higher temperature.
These properties are measured by a Nano Test pendulum set up specially equipped for high
temperature measurements. The hot stage consists of a thermally insulating ceramic block that is
attached to the specimen holder. A thermal shield is placed between the pendulum and the stage to
prevent thermal instability of the probe (Fig. 1).

Figure 1: Nano Test pendulum modified for high temperature measurements.

Indentations are performed at room temperature, 250 and 500°C using a Berkovich indenter.
The following parameters are used for the measurements: maximum load (50-250mN), loading
rate (2-5 mN/s, dwell time at maximal load (10-30 s). Scratch and impact experiments are
employed at the same room temperature to analyze critical load to adhesion failure and fatigue
properties allowing identification of operating limits. Microstructure analysis of film is done by
Scanning Electronic Microscopy (Jeol JSM6480).
Investigation of thin films and surface mechanical properties at high temperature will allow
optimizing the “Couple Tool Material “projects in metal cutting.
Keywords: Berkovich indenter, coated inserts, PVD films, High Temperature, Nano Test
techniques
80 B. Vasques et al.

References
1. Beake D., Garcia M. and Smith F., Thin Solid Films, 398-399 (2001) 438-443.
2. Beake D. and Ranganathan, ICMAT, Singapore, 3-8 July 2005.
3. Karimi A., Wang Y., Morstein M., Thin Solid Films, 420-421 (2002) 275-280.
4. Gong J., Miao H., Peng, Acta Materialia, vol. 52, 785-793, 2004.
5. Morant C., Prieto, Forn A., Surface and Coated Technology, 180-181 (2004) 512-518.
1T7. Thin films 81

DIAMOND COATING DEBOUNDING IN TOOL APPLICATION

David Moulin, Pierre Chevrier, Paul Lipinski and Thierry Barré1


Laboratoire de Physique et Mécanique des Matériaux (LPMM) /
Ecole Nationale d’Ingénieurs de Metz (ENIM)
ENIM, ile du Saulcy, 57045 Metz, France
1Balzers Luxembourg, ZI Hanoesbesh, Luxembourg

d.moulin@enim.fr, chevrier@enim.fr, lipinski@enim.fr, thierry.barre@balzers.com

Machining of recent engineered complex materials such as aeronautical ones always require new
tools to lower production cost and in the same time increase quality. One promising solution is the
use of Chemical Vapour Deposition (CVD) diamond coatings on carbide tools, which is studied
here. Diamond is used because of its extreme hardness, which made it a perfect candidate to reduce
wear of the cutting edge. It is synthesized in a batch at low pressure and medium temperature
directly on the tools. Deposition process on tungsten carbide tool bound with cobalt, which is the
main industrial material for tool manufacture, and especially the adhesion of the diamond layer to
the substrate has to be improved, and many research are focused on this point [1]. The whole
process of deposition is described elsewhere [2]. This paper proposes an approach to study the
decohesion of the diamond layer via a combination of experiments and simulation.
Many techniques to study adhesion of the diamond layer were developed, such as indentation
[3], sand blasting [4]. However, these techniques have a high limitation: they are only qualitative,
and don’t take into account the physic of adhesion. Results obtained with these techniques depend
of the test procedure and of the machine used to apply the load. Results of a series of tests cannot
be compared to others. A better comprehension of the physic involved can lead to a quantitative
test, with a good repeatability and a measured value, which do not depend of the test process.
Experimentally, it can be observed that diamond delamination occurs because of a stress along
the interface between carbide and diamond (Fig. 1).

Figure 1 : Diamond fracture on rake face of a turning insert

This stress is of different natures. When coating flaked during machining, pressure and
temperature induced by the machining process, which are external loads, are responsible of the
degradation of the interface. However, debounding can occurs into the batch, without any external
loading, because high thermal stresses during cooling (see Fig. 2). These high thermal stresses are
due to by the difference of thermal dilatation coefficients and can be calculated with a finite
element model. This stress has to be taken into account together with that due to the external
loading. Observation of the insert after decohesion (fig 1) confirms that fracture occurs at the
interface of the two materials.
82 D. Moulin et al.

Figure 2 : Normal stresses induced by thermal cooling along the interface Carbide – diamond
depending on thickness and edge radius (centred on the edge)- Results on FEM calculation

In this work, adhesion is studied by completing experimental loading of the system by


simulations with a Finite Element Model (FEM) in order to obtain the stress state responsible for
the de-bounding of the diamond layer. A criteria is then written to characterize the behavior of the
system based on a combination of the normal and shearing stress responsible for de-cohesion.
Different loading configurations are imagined to obtain the form of the criteria: 3 or 4 points
flexion tests, impact tests, and variation of the radius edge. Flexion tests induce pure traction-
compression stress in both the coating and the substrate but with two different levels, leading to
shearing of the interface and to debounding of the coating.
Indentation and impacts tests are more complex loads, respectively quasi-static and dynamic in
nature.
Deposition on different edge radii leads to different thermal residual stresses, which play the
role of the load. A critical radius can be found, which is the smallest radius before debonding.
Stress state can be calculated according to this radius.
These different techniques, combined with the simulation, give a better understanding of the
phenomenon that control the adhesion. They lead to a debounding criteria useful to characterize
adhesion.

References
1. S. Kamiya, U.H. Takahashi, R. Polini and E. Traversa, Diamond Relat. Mater. , vol 9, 191-
194, 2000
2. H.G. Prengel, W.R.Pfouts, and A.T. Santhanam, Surf. Coat. Technol. , Vol. 102, 183 – 190,
1998
3. G.Jörgensen, M.Lahres, and J.Karner, Surf. Coat. Technol. , vol 97, 238-243, 1997
4. Friedrike Deuerler, Heiko Gruner, Michael Pohl and Ladji Tikana, J. Mater. Process.
Technol. , Vol. 99, 266 – 274, 2000
1T7. Thin films 83

INTERFACIAL STRENGTH OF CERAMIC THIN FILM ON POLYMER


SUBSTRATE

M. Omiya and K. Kishimoto


Department of Mechanical and Control Engineering,
Tokyo Institute of Technology,
2-12-1, O-okayama, Meguro-ku, Tokyo, 152-8552, Japan
oomiyam@mep.titech.ac.jp

In this paper, we aim to evaluate the interfacial adhesion strength between ITO (Indium Tin Oxide)
coating layer and PET (Poly-Ethylene Terephthalate) substrate. To evaluate the interfacial
adhesion strength, we focused on the buckling phenomena of the coating layer during simple
tensile tests. During the tensile tests of PET/ITO specimen, the buckling induced delamination
occurred on the coating layer. These phenomena are related to the interfacial adhesion strength and
we evaluated the interfacial adhesion strength by considering the energy balance during the
buckling induced delamination and cracking.
During tensile tests, two types of cracks are formed on the coating layer: vertical cracks and
parallel cracks. Parallel cracks are induced by the buckling of thin coating film on the polymer
substrates and these phenomena are strongly related to the interfacial adhesion strength. Therefore,
we estimate the strain energy stored in the coating film by modified shear-lag model[1] and
considering the energy balance to evaluate the interfacial adhesion strength between thin film and
substrate.
Due to the tensile loading, the coating layer is assumed to be segmented as shown in Fig.1. In
the coating layer, the compressive loads are induced by the interfacial shear force. From modified
shear-lag model, when the compressive stress in the substrate is V c , the compressive stress in the
coating layer is,

(1)

where, E f , Es are Young’s modulus for coating layer and substrate, respectively. h, H are the
thickness of the coating layer and the substrate. L is the segment length of the coating layer
andăis the load transfer length. The energy balance before and after buckling induced
delamination and cracking is,

We Wd  Wec  W f
(3)

where the strain energy before buckling crack is W e , the delamination energy W d , the strain
energy of undelaminate coating layer is W and the energy for crack formation is W f . Note that
ec
the change of the strain energy in the substrate is small because the compressive stress is
continuously acting on the substrate from the applied load and it constrains the elastic recover of
the substrate due to the delamination and cracking. Then, the interfacial strength, * d , can be
described as,
84 M. Omiya and K. Kishimoto

(4)

where is the fracture toughness of the coating layer, l d is the delamination length and b is the
width of the segmented film. In this study, the dimensions needed to evaluate the interfacial
strength from by Eq.(4) are measured by atomic force microscope. From this observation, we
obtained the delamination length, l d , and evaluate the interfacial adhesion strength, * d . The
2
average value of the interfacial strength for PET/ITO specimen is 19.6 J/m and PET/ITO(UV)
specimen is 9.6J/m2. The obtained results agree well with the peel test results[2] as shown in Fig.2.
During the buckling and delamination process of the coating layer, the phase angle continuously
changes. Therefore, the obtained results are considered to be the averaged interfacial strength for
several phase angles. However, this simple approach is useful for the first estimation of the
interfacial strength and it is valuable for quick interfacial evaluation at manufacturing premises.

FIGURE 1. Buckling induced delamination and cracking.

FIGURE 2. Comparison with the results measured by Multi-stages Peel Test.

References
1. Yanaka, M., Kato, Y., Tsukahara, Y. and Takeda, N., Thin Solid Films, Vol.355-356, 337-
342, 1999.
2. Omiya M., Inoue H., Kishimoto, K., Yanaka, M. and Ihashi, N., Journal of the Society of
Materials Science, Japan, Vol.52, 856-861, 2004.
3. Omiya M., Inoue H., Kishimoto, K., Yanaka, M. and Ihashi, N., Key Engineering Materials,
Vol.297-300, 2284-2289, 2005.
1T7. Thin films 85

DELAMINATE BEHAVIOR OF PVD/CPVD THIN FILM

S. Doi and M. Yasuoka


700 Dannoharu Oita-shi,JAPAN
doi@cc.oita-u.ac.jp

Accompanied with sophistication of the technology of creating surface, multi-layered structures of


film tend to increase for improving the accuracy with a large degree of the freedom playing an
important role. In particular, various kinds of methods for producing film have been proposed and
the application range of film has become wider. In this research, focusing on coating film materials
fabricated using PVD, CVD and a combined method of PVD+CVD, grasping their surface
morphology is aimed. On the other hand, the reliability of such a new material with a different
property between the surface function and the interface function is required. Accordingly, different
techniques from conventional methods of evaluating functions of laminated material are needed.
Therefore, in this research, so-called textured materials with new structure coated by a series of
carbon film or titanium nitride film as described in the following, were selected. The effect of
applied impact loading on the adhesion and the interface quality of their films was investigated by
observation using laser confocal microscope and fractal dimensional analysis of cracks or
exfoliation surfaces.
The molecular structure of DLC attracting attention recently differs from one of diamond and
graphite and constitutes a crystal structure with no aligning. This structure is said to be a kind of
amorphous structure with a similar characteristic to one of diamond showing high hardness and a
partially similar to one of graphite. The basic arrangement of the carbon includes the atomic
arrangement of SP‚R and this causes a different property of hardness of DLC. Also, TiN has a
structure with substrate of M35, that is to say, a two-layered structure and this leads to a crystalline
textured structure constituting the interface and the surface, respectively. As a method of
examining surface characteristics in fractal dimensional analysis, the box count method was
adopted because this is an adequate method for observing the local change.
86 S. Doi and M. Yasuoka

Fig.2 Relationship fractal dimension and cyclic impact numbers


1T7. Thin films 87

EXPERIMENTAL STUDY OF MICROHARDNESS AND FRACTURE OF


IMPLANTED GÁLLIUM NITRIDE FILMS

P. Kavouras, M. Katsikini, E. Wendler1, W. Wesch1, H. M. Polatoglou, E. C. Paloura, Ph.


Komninou and Th. Karakostas
Physics Department, Aristotle University of Thessaloniki
541 24 Thessaloniki, Greece
1Institut für Festkörperphysik, Friedrich-Schiller-Universität Jena, Max-Wien-Platz 1, Jena D-

07743, Germany
pkavo@skiathos.physics.auth.gr, katsiki@auth.gr, wendler@pinet.uni-jena.de,
Werner.Wesch@uni-jena.de, hariton@auth.gr, paloura@physics.auth.gr, komnhnoy@auth.gr,
karakost@auth.gr

The investigation of radiation effects in III-V semiconductors is of current interest because of the
potential application of ion implantation in the production of electronic and photonic devices. Of
the plethora of III-V semiconductors, gallium nitride (GaN) has attracted keen interest, during past
decade, as wide gap semiconductor for numerous applications, including high-power or high
frequency devices and high-power switches. Although there is considerable interest in determining
the influence of ion implantation on the mechanical properties of GaN films, the matter has
received only scant attention, see for example Kavouras et al. [1]. Indeed, studies of the processes
controlling hardness, contact damage and cracking of epitaxially grown GaN films have significant
technological importance.
In this work, we present the results of the application of the Static Indentation (SI) technique,
using Vickers and Knoop indenter geometries, in the investigation of implantation induced effects
on the mechanical properties and fracture of epitaxially grown GaN films. More specifically, we
compare a number of physical characteristics that govern the elasto-plastic behavior before and
after implantation. The implanted species used to produce implantation induced damage were Au,
Xe, Ar and O ions.

TABLE 1. Structure and microhardness of as-grown and implanted GaN films. The errors
represent the standard deviations.
Material Name Structure Microhardness / GPa
As-grown GaN GaN - 13.73r0.12
Au-implanted GaN:Au Amorphous 12.28r0.16
Xe-implanted GaN:Xe Amorphous 10.19r0.10
Ar-implanted GaN:Ar Amorphous 9.41r0.14
O-implanted GaN:O Heavily damaged 14.53r0.26

A Knoop indenter was utilized to obtain the microhardness values (Table 1). The structure of
GaN implanted epilayers was obtained by X-ray Absorption Fine Structure (XAFS) technique. It
was found that in the case where implantation produces heavily damaged structure microhardness
is increased, while implantation induced amorphisation lowers microhardness value. Additionally,
the influence of the indentation load on the microhardness value was studied, i.e. the Indentation
Size Effect (ISE) curves were obtained in all cases. It was observed that GaN and GaN:O showed a
normal ISE behavior, i.e. microhardness value increases with decreasing indentation load, while
GaN:Au, GaN:Xe and GaN:Ar showed a Reverse ISE (RISE), i.e. microhardness value decreases
with decreasing indentation load. The magnitude of the post-indentation healing of indentation
88 P. Kavouras et al.

prints was evaluated by measuring (a) the relative length of the two normal diagonals of Knoop
indentation prints and (b) the depth of residual indentation prints by means of Atomic Force
Microscopy (AFM). The above information can give an estimation of the elastic behavior and
elastic modulus, as it was firstly recognized by Marshall et al. [2].

FIGURE 1. Optical micrographs, under transmitted illumination, of fracture formation sequences


for GaN (a, b and c) and GaN:O (d, e and f) films.

Fracture sequence was also observed in all cases, utilizing Vickers indenter geometry
according to the formalism elaborated by Cook and Pharr [3]. GaN did not show initiation of a
specific microcrack type. Only cumulative fracture events and/or film detachment were observed
for indentation loads higher that #1.5 N. In all other cases, fracture occurred in lower loads,
indicating that implantation has an effect analogous to embrittlement. Fig. 1 shows characteristic
sequences of fracture formation in the cases of GaN and GaN:O.

References
1. Kavouras, P., Katsikini, M., Kehagias, Th., Paloura, E.C., Komninou, Ph., Antonopoulos, J.,
Karakostas, Th., J. Phys. Cond. Mat., vol. 14, 12953-12959, 2002.
2. Marshall, D.B., Noma, T., Evans, A.G., J. Amer. Cer. Soc., vol 65, C175-C176, 1982.
3. Cook, R.F. and Pharr, G.M., J. Amer. Cer. Soc., vol 73, 787-817, 1990.
1T9. Failure of nanocomposites 89

CRACK TIP STRAIN FIELD AND ITS PROPAGATION CHARACTERISTICS IN


A POLYMER FOAM

Fu-pen Chiang, Sheng Chang and Yi Ding


Department Of Mechanical Engineering, Stony Brook University
Fu-Pen.Chiang@sunysb.edu, fchiang@notes.cc.sunysb.edu

In recent years there has been an increasing interest in using foam composites as shipbuilding
materials. A sandwich panel built of fiberglass face sheets and a foam core has strong rigidity and
bending strength. It’s advantage for ship construction is obvious. Furthermore when a small
amount of nanoparticles are added to the material, it tends to increase its stiffness and retards fire
as well. In this paper we employ a unique micro/nano speckle technique to investigate the crack tip
deformation field of two different polymer foams at a length scale that has never been studied
before. The principle of the technique is described in [1]. Figure 1 shows the micrograph recorded
by a scanning electron microscope of a crack propagating through a NEAT foam specimen under
uniaxial tension. The void in front of the crack tip was kind of spontaneously generated as a result
of the load. The main crack then tends to link itself towards the void as demonstrated in the
sequence of pictures shown.

Fig 1. Crack propagation in a NEAT foam specimen under uniaxial tension

While in the macro scale the crack will largely show propagation in the mode I characteristics
at a micro/nanoscale the crack path is far from being a straight line. It tends to circle around a foam
cell and then advance. But under the right circumstance it will break across the cell-to-cell
interface. The nature of the propagation characteristics is revealed in the next example. Figure 2
shows the two micrographs (again recorded by an scanning electron microscope) at 50X and 100X
magnifications, respectively, of a nano-phased foam beam specimen under 3-point bending with a
single edge crack. The crack is perpendicular to the longitudinal axis of the beam and the tip of the
crack is visible in the SEM micrograph at 50X.
Figure 3 shows the displacement contours surrounding the crack tip at 50X magnification. The
deformation field is rather complicated and it bears almost no relationship to the classical
displacement field in the region around a crack tip. At a higher magnification, the deformation
field is even more complicated, as shown in Figure 4.
90 F.-P. Chiang et al.

Fig 2. SEM micrograph of a nano-phased PVC foam beam under 3 point bending with a single
edge crack

Fig 3. u and v displacement fields at 50x. Contour constant = 0.002Pm

Fig 4. u and v displacement fields at 100x. Contour constant 0.001Pm, Nano-phased PVC foam
specimen with an edge crack under 3 point bending.

As can be seen this is not a classical crack tip deformation field (for an isotropic and
homogeneous material) at all. Of particular interest is the dense contour lines (at 1 nm/contour)
surrounding a foam cell in front of the crack tip, indicating a stress concentration region. This
implies that the cell-to-cell interface is strong and the crack tends to propagate around the foam cell
as observed.
Acknowledgement
The author would like to thank Dr. Yapa D.S. Rajapakse, Manager of the Solid Mechanics
Program, US Office of Naval Research for supporting this work through grant #N000140410357
and Professor Hazzan Mahfuz of Tuskepee University for providing the foam materials.

Reference
1. Chiang F.P., 2003a, “Evolution of white light speckle method and its application to micro
nanotechnology and heart mechanics”, Optical Engineering Vol. 42 no.5, 1288-1292.
1T9. Failure of nanocomposites 91

HOW TO TOUGHEN CERAMICS – NANOCOMPOSITES

H. Awaji and S.-M. Choi


Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya, 466-8555 Japan
awaji@nitech.ac.jp

Ceramic-based nanocomposites with intra-type nano-structure have high strength with moderate
fracture toughness. In this research, the toughening and strengthening mechanisms of the ceramic-
based nanocomposites will be explained by dislocation activities even in brittle materials. Based
on these mechanisms, we fabricated toughened and strengthened alumina-nickel nanocomposites.
Fracture toughness of the annealed nanocomposites was 7.6 MPam1/2, which was two times higher
than that of monolithic alumina.
Dislocation activities
In intra-type nano-structure, nano-sized second-phase particles are embedded within the matrix
grains. The highest strength or fracture toughness is mostly achieved when only a few volume
percent of the second-phase particles are dispersed in ceramic materials. The structural
characteristic of these nanocomposites results in the generation of thermally induced residual
stresses around the dispersed particles in matrix grains [1]. We analyzed residual stresses using a
simplified model consisting of a spherical particle within a concentric matrix sphere with axial
symmetry [2]. Residual stresses numerically calculated on the particle-matrix boundary for
alumina/nickel nanocomposites are shown in Table 1, where the symbols with suffix p indicate the
properties of the particle (nickel) and the symbols with suffix m are the properties of the matrix
(alumina). It is noted that there is a large maximum shear stress on the boundary.
Figure 1 shows the temperature dependence of the residual shear stress on the boundary of
alumina/nickel, and critical resolved shear stresses for basal and prism plane slips in a single á-
alumina crystal measured by Lagerlöf et al. [3] This figure indicates that dislocation movements
are possible in the alumina grains at temperatures ranging from 600 to 1400 ºC, suggesting that this
temperature range is quite important in creating dislocations in the alumina matrix during the
cooling process.
Strengthening mechanism
The grains of sintered alumina contain tensile residual stresses resulting from anisotropic
thermal expansion, etc. Therefore, it is conceivable that the large crack along a grain boundary
created by the synergetic effect of both residual stresses and processing defects, will be equivalent
to the grain size of the materials and that the weakest crack generated along a boundary in the
specimen will dominate the strength of the specimen. Nanocomposites, however, will yield
dislocations around the particles, and the dislocations release residual stresses in the matrix.
Consequently, the defect size along the grain boundaries is reduced in nanocomposites.

TABLE 1. Residual stress along the particle-matrix boundary in alumina-nickel nanocomposites.

System Į m /Į p ×10 -6 E m /E p IJ m ax
(K -1 ) (GPa) (GPa)
Al 2 O 3 -Ni 8.8/13.3 380/207 1.0
92 H. Awaji and S.-M. Choi

Also, the dislocations are difficult to move in ceramics at room temperature, serve as origins of
small stress concentrations, and create nano-cracks around the propagating crack tip. These nano-
cracks reduce the strength of the alumina matrix slightly and change the fracture mode from
intergranular fracture in monolithic alumina to transgranular fracture in nanocomposites.
Toughening mechanism
To improve the intrinsic fracture toughness of ceramics, the fracture energy consumed in the
frontal process zone (FPZ) must be increased [4], which is called the FPZ toughening mechanism.
Because ceramics are brittle, the FPZ of ceramics is considered to be constructed by many nano-
cracks rather than dislocations in metals. Therefore, we must consider how to create many nano-
cracks in the FPZ.
Assuming that we obtained the nano-structure with dispersed dislocations within the matrix
grains after annealing, these dislocations become sessile dislocations at room temperature. In this
situation, when the tip of a propagating large crack reaches this area, these sessile dislocations will
operate as nano-crack nuclei in the vicinity of the propagating crack tip. The highly stressed state
in the FPZ is then released by nano-crack nucleation, and the nano-cracks expand the FPZ size,
enhancing the intrinsic fracture toughness of the materials.

References
1. T. Matsunaga, et al., J. Ceram.Soc., Japan, vol. 113, 123-125, 2005
2. H. Awaji, et al., Mechanics of Materials, vol. 34, 411-422, 2002.
3. K. P. D. Lagerlöf, et al., J. Am. Ceram. Soc., vol. 77, 385-397, 1994.
4. S-M. Choi and H. Awaji, Science & Tech. of Advanced Mater., vol. 6, 2-10, 2005.
1T9. Failure of nanocomposites 93

DEFORMATION AND FRACTURE BEHAVIOUR OF NANOCOMPOSITES

S. Dunger, J. K. W. Sandler, K. Hedicke and V. Altsadt


University of Bayreuth, Department of Polymer Engineering, FAN A - Universitatsstrase 30
95448 Bayreuth, Germany
Siljana.Dunger@uni-bayreuth.de

Polymer nanocomposites are of great scientific interest due to the potential of improving the
resulting physical and mechanical properties of components, which cannot be modified using
conventional reinforcements. However, the successful development and industrial implementation
of such novel materials, especially of structural components, pose unique challenges.
Nanocomposites with oriented nanoparticles and good adhesion between both phases can offer a
stiffness and strength increase comparable to conventional fibre-reinforced composites [1 - 3];
however, only moderate filler weight fractions are possible due to the large specific surface area of
the particles [4]. Nevertheless, at the same time, nanoscale fillers can maintain or even increase the
composite thoughness [5-7]. For certain applications such as delicate structures the use of
nanoscale reinforcements therefore offers a unique potential [8].
For a given matrix and nanoscale reinforcement system, processing leading to a good
dispersion and distribution as well as orientation of the nanophase is crucial. The dispersion of
nanoparticles especially depends on a magnitude of processing parameters, e.g. shear rate and
processing temperature, as well as on the selection of appropriate materials. In addition,
interactions between the nanoscale reinforcement and the polymer matrix during processing
significantly influence the orientation of individual particles and the molecular morphology in the
vicinity of the particles.
The presentation will highlight the resulting deformation and fracture behaviour of different
polymer nanocomposites, taking into account variations in the matrix morphology as a function of
nanofiller type, geometry, surface chemistry and processing. Given that nanofillers have been
shown to lead to distinct variations in the resulting interphase, an investigation of the fracture
mechanics is especially important in order to understand the commonly occurring transition from a
ductile to a brittle behaviour of such nanocomposites. In order to analyse the influence of the
matrix ductility on the resulting nanocomposite fracture behaviour, polystyrene, polyamide 6, and
polystyrene-polybutadiene-polystyrene block copolymer were chosen as matrix materials. The
influence of the nanofiller geometry is investigated by using silicate clays, silicium dioxide
nanoparticles and tubular clay nanostructures. Moreover, the resulting interphase properties are
modified by different silicate surface treatments. The mechanical properties of the nanocomposites
as well as of comparative glass fibre-reinforced composites were characterised by tensile tests, KIc
tests and crack propagation tests. The fractured surfaces of the specimens were extensively
analysed by SEM. In order to investigate the fracture behaviour during deformation, TEM analyses
help to observe and explain the crack propagation of the nanocomposites as presented in fig 1.
Finally, a comparison between traditional fibre-reinforced systems and the novel nanocomposites
is made.
94 S. Dunger et al.

Fig. 1: Crack propagation and deflection in polyamide 6 nanocomposites

References
1. Fornes, T. D., Paul, D. R.; Polymer 44, (2003), 4993-5013
2. Galgali, G., Agarwal, S., Lele, A.; Polymer 45, (2004), 6059-6069
3. Fisher, F. T., Bradshaw, R. D., Brinson, L. C.; Composite Science and Technology 63,
(2003), 1689-1703
4. Luo, J.-J., Daniel, I. M.; Composite Science and Technology 64, (2003), 1607-1616
5. Zerda, A. S., Lesser, A. J.; Journal of Polymer Science: Part B 39, (2001), 1137-1146
6. Chen, L., Wong, S.-C., Liu, T., Lu, X., He, C.; Journal of Polymer Science: Part B 42, (2004),
2759-2768
7. Gojny, F. H., Wichmann, M. H. G., Kˆpke, U., Fiedler, B., Schulte, K.; Composite
8. Science and Technology 64, (2004), 2363-2371
9. Nam, P. H., Maiti, P., Okamoto, M., Kotaka, T.; Polymer Engineering and Science 42,
(2002), 1907-1918
1T9. Failure of nanocomposites 95

FRACTURE MECHANISMS IN CARBON NANOTUBE-REINFORCED


COMPOSITES

E. T. Thostenson and T.-W. Chou


Department of Mechanical Engineering and
Center for Composite Materials
University of Delaware
Newark DE 19716 USA
chou@me.udel.edu

Carbon nanotubes have been targeted for potential applications ranging from the next generation of
computers and flat-panel displays to structural and functional materials. In addition to their well-
known stiffness (> 1 TPa) and strength (~30 GPa) properties, carbon nanotubes also possess
exceptionally high electrical and thermal conductivities, with the axial thermal conductivity near
that of crystalline diamond. The unique mechanical and physical properties of nanotubes offer
tremendous opportunity for the development of multi-functional composites [1, 2]. Full
understanding of the thermo-mechanical behavior of nanotube-based composites, requires
knowledge of the elastic and fracture properties of carbon nanotubes as well as interactions at the
nanotube/matrix interface. Although this requirement is no different from that in conventional
fiber composites [2], the scale of the reinforcement phase diameter has changed from micrometer
(e.g. glass and carbon fibers) to nano-meter. The change in reinforcement scale poses new
challenges in the development of processing techniques for these composites as well as
characterization techniques and methodologies to measure their elastic and fracture behavior.
A fundamental knowledge of the process/structure/property relationships is required to enable
the design of multi-functional materials by structuring at the nanoscale. A novel technique to
produce continuous nanocomposite ribbons of aligned multi-walled carbon nanotubes has been
developed [3]. This model nanocomposite system serves as a basis for the investigation of
structure/property relationships through characterization of their elastic and fracture behavior. The
elastic and fracture behavior of the model nanotube composites indicate the anisotropy in the load
transfer and confirm that nanotubes are able to carry load that is transferred via shear stresses at the
nanotube/matrix interface and through characterization of this model system a fundamental
knowledge of their structure/property relations has evolved [4]. The tensile fracture behavior of
carbon nanotube composites show similar mechanisms as in traditional fiber composites including
nanotube fracture, pullout, and crack bridging [3]. For compressive deformation, critical
nanoscale buckling behavior of carbon nanotubes was observed where small diameter nanotubes
deform through global bending analogous to Euler-type buckling and large diameter nanotubes
show locally sharp kinking [5]. These deformation behaviors suggest a critical diameter may exist
for the change in buckling modes and could have significant implications on the nanoscale design
of composite compressive properties.
Recent research by Gojny et al. [6] has shown that very low concentration of double-walled
carbon nanotubes (0.1 wt%) can result in substantial improvements in fracture toughness. In order
to evaluate the influence of multi-walled carbon nanotubes on the fracture toughness of epoxy
nanocomposites we fabricated composites with nanotube contents ranging between 0.1 wt% and 1
wt% in an EPON 862 epoxy matrix. The nanotubes were first dispersed in the epoxy resin and the
curing agent (Epi-Cure W) was added. The nanocomposites were then placed in a mold and cured
for 6 hours at 130oC. Fracture toughness measurements were conducted using the single-edge-
notch bending (SENB) method. Specimens were notched with a tapered diamond blade and a pre-
crack was introduced by tapping with an ultra-sharp carbon steel razor blade.
96 E. T. Thostenson and T.-W. Chou

The fracture toughness of the epoxy nanocomposites was significantly improved as compared
to the unreinforced resin. This indicates that nanotubes provide a reinforcing effect in improving
the fracture toughness through crack deflection or nanotube fracture and pullout. SEM
micrographs of the composite fracture surface show a change in the micron-scale surface
roughness and also the presence of nanotube pullout.

References
1. Thostenson, E.T., Li C.Y. and Chou T.W. Compos. Sci. Technol., vol. 65, 491-516, 2005
2. Thostenson, E.T., Ren Z.F. and Chou T.W. Compos. Sci. Technol., vol. 61, 1899-1912, 2001
3. Thostenson, E.T. and Chou T.W. J Phys D: Appl. Phys. vol. 35, L77-L80, 2002
4. Thostenson, E.T. and Chou T.W. J Phys D: Appl. Phys. vol. 36, 573-582, 2003
5. Thostenson, E.T. and Chou T.W. Carbon vol. 42, 3015-3018, 2004
6. Gojny F., Wichmann M and Kopke U, et al. Compos. Sci. Technol., vol. 64, 2363-2371, 2004
B. TRACKS

B2. Engineering Materials and Structures


2T1. Physical aspects of fracture 99

FRACTAL APPROACH TO CRACK PROBLEMS WITH NON-ROOT


SINGULARITY

A. Kashtanov
St.-Petersburg State University
Universitetsky pr. 28, 198504 St.-Petersburg, Russia
arsen@ak1340.spb.edu

Fractal approach to the problems of fracture mechanics with non square root singularity of the
stress field is discussed. Fractal generalization of Griffith energy balance equation is proposed and
analysis of the fracture process at the sharp angular notch in a plate is performed. Good
correspondence between suggested approach and experiments is observed.
It is obvious that propagation of fracture surface is much more complex process than a simple
spreading of rectilinear crack with smooth faces and usually the crack surface has a lot of
irregularities of different sizes. The roughness of fracture surface can be accounted with the help of
fractal correction for the calculation of specific properties suggested by Mandelbrot [1]. It allows
us to receive more precise model of fracture process and to solve the problems of crack mechanics,
which do not have an adequate solution within the frameworks of traditional theory. One way is
the simulation of the crack by fractal with dimension determined from some theoretical reasons at
the fixed scale level. Then, solving a specific problem for simulated “fractal” crack it is possible to
determine some physical magnitudes, which values can be easily checked experimentally, for
example, the value of critical loading.
In particular it is suggested to evaluate the fractal dimension by sufficing to the energy balance
equation. In order to construct this equation for the case of “fractal” cracks the following
generalized energy balance concept was proposed by Kashtanov and Petrov [2]:
• The work of crack opening ' W is the integral parameter of stated problem and is
determined at the macroscopic scale level.

• The surface energy of crack '3 is calculated using the crack fractal length L to take into
account the microstructure of crack surface.
• The fractal dimension D of simulated crack is defined from the validity of energy balance
equation ' W '3 at the macroscale.
• At the macroscopic scale level the fracture process is characterized by the size of an
elementary fracturing cell

2 K 12c
d
S V c2 (1)

where V c is the ultimate strength of defectless material and K 1c is the static fracture toughness.
Therefore it is convenient to use the fractal low at the macroscale in the form

L l d D d (2)

where l is the crack length at the macroscale.


100 A. Kashtanov

As an example of such approach the plane problem about an angular notch can be considered.
This problem is characterized by the non square root singularity of the stress field and it has no
solution in Griffith–Irwin theory of fracture. Nevertheless, if we suppose that the crack formed in
the notch vertex is a fractal then the connection between the value of fractal dimension of modeled
crack and actual structural parameters of fracture process can be constructed analytically from the
generalized energy balance equation. Then the critical load required for fracture at the crack tip can
be found and good coincidence with experiments is observed (Fig. 1).

FIGURE 1. The dependence between the critical loading and the hole length.

At the Fig. 1 the dependence between fracturing loading and the notch angular is displayed.
The solid line corresponds to the solution of generalized energy balance equation; the dotted line
displays the same dependence calculated from Neuber – Novozhilov fracture criterion; and
experimental data are pointed by circles. All the experiments have been conducted in
St.-Petersburg University by I. Bugakov and I. Demidova.

References
1. Mandelbrot, B. B., The Fractal Geometry of Nature, Freeman, Berlin, New York, 1983.
2. Kashtanov, A.V., Petrov, Y.V., Int. J. of Fracture, vol. 128, N 1, 271-276, 2004
2T1. Physical aspects of fracture 101

NEW METHOD FOR ANALYSING THE MAGNETIC EMISSION SIGNALS


DURING FRACTURE

Gy. B. Lenkey, N. Takacs1, F. Kun2 and D. L. Beke3


Bay Zoltán Foundation for Aplied Research, Institute for Logistics and Prod. Systems, Miskolc,
Hungary
1
Metalelektro Co. Budapest, Hungary
2
Department of Theoretical Physics, University of Debrecen, Debrecen, Hungary
3
Department of Solid State Physics, University of Debrecen, Debrecen Hungary
lenkey@bzlogi.hu, takacsn@delfin.klte.hu, feri@dtp.atomki.hu, dbeke@delfin.klte.hu

In dynamic fracture testing the precise determination of the onset of crack initiation is crucial in
order to obtain characteristic quantities of the material. This task can be solved easier in case of
brittle fracture since brittle fracture is usually accompanied by a sudden drop in the force signal.
But in case of ductile fracture or if stable crack propagation occurs before unstable one, the instant
of crack initiation cannot be determined directly from the force signal. In these cases additional
measurement techniques should be applied. The magnetic emission technique has been proved for
detecting brittle crack initiation of ferromagnetic materials [1-3].
Two physical phenomena contribute to the magnetic emission signal [1]: (a) mechanically
induced Barkhausen signals appear when the internal magnetic structure changes during loading,
and (b) a propagating crack causes the internal magnetic field to emerge from the solid into the gap
between the two crack surfaces, thereby changing the external magnetic field. These field
variations can be observed locally by a magnetic transducer which basically consists of a coil. The
transducer's output voltage is the magnetic emission (ME) signal which is proportional to the
derivative of the magnetic field (MF).
The aim of the present work was to apply new statistical methods [4-9] to analyze the magnetic
emission signal to extract more information about the fracture process.
Magnetic emission spectra recorded in dynamic fracture experiments on ferromagnetic
materials are composed of more or less well-separated voltage peaks which provide direct
information about the microscopic processes involved in fracture. It will be illustrated, that before
crack initiation sudden movements of domain walls under external mechanical loading results in
low peaks. After crack initiation the opening of the growing crack provides the dominating
contribution for the voltage signals and is responsible for the higher peaks of the spectrum. After
removing a background voltage level the ME spectrum is characterised by the distribution of the
height, area, and width of peaks, furthermore, correlations among these quantities. One example
for the evaluated peak parameters is shown in Fig. 1.
While for the “low peaks” part (before crack initiation) no typical functional form can be
extracted, for the height distribution of peaks in the crack propagation regime the distributions
show power law behavior over a range of one order of magnitude. The analyses of magnetic
emission signals recorded during brittle and ductile failure showed that the value of the exponent
of the power law regime is characteristic for the failure mode: an exponent significantly higher
than for brittle failure characterizes ductile failure. The distribution of the peak areas has a similar
overall character, however, the power law regime spans practically two orders of magnitude and
the area distribution is universal in the sense that it does not change with the impact velocity: it
depends solely on the failure mode.
102 Gy. B. Lenkey et al.

FIGURE 1. Area distribution of peaks p(A) for different impact velocities and failure modes.

References
1. Gy. B. Lenkey, S. Winkler.: Fatigue and Fracture of Engineering Materials and Structures,
Vol. 20., No. 2., pp. 143-150., 1997
2. Equation Section 1 Gy. B. Lenkey, L. Tóth: Mat.wiss. und Werkstoffmech., 32, 1-6, pp. 1-6.,
2001
3. Gy. B. Lenkey: ASTM STP 1380, T. Siewert and M. P. Manahan, Sr., Eds., American Society
for Testing and Materials, West Conshohocken, PA, pp. 366-381., 1999
4. M. C. Miguel, A. Vespignaid, S. Zapperi, J. Weiss and.J. Grasso, Nature 410, 667, 2001
5. A. Maes, C. Van Moffaert, H. Frederix, and H. Strauven, Phys. Rev. B 57, 4987, 1998.
6. A. Guarino, S. Ciliberto, A. Garcimartin, M. Zei, and R. Scorretti, Eur. Phys. Jour. B 26, 141,
2002
7. R. A. White and K. A. Dahmen, Phys. Rev. Lett. 91, 085702, 2003
8. L. I. Salxninen, A. I. Tolvanen, and M. J. Alava, Phys. Rev. Lett. 89, 185503, 2002
2T1. Physical aspects of fracture 103

ELECTROMAGNETIC RADIATION METHOD FOR IDENTIFICATION OF


MULTI-SCALE FRACTURE

Yu. K. Bivin, A. S. Chursin, E. A. Deviatkin and I. V. Simonov


Institute for Problems in Mechanics of RAS
Vernadskogo av. 101-1, 119526 Moscow, Russia
simonov@ipmnet.ru

The electromagnetic radiation (EMR) as well as acoustic emission (AE) is the statistical
phenomenon in fracture. Measurements of EMR can serve the purpose of identification. The
experimental data on this point are known for ion crystals, metals, rocks. The explanation of
physical mechanism has been given ideally, for example, in the case of shock wave in an “ideal”
metal.
Aim of the present work is to develop this promising method of EMR and to show its
possibilities for registration of multi-scale fracture mainly within non-conductors, structural
materials. A review of experimental results includes the following events: the fast micro-fracture
in ice, the single crack propagation in glass, the velocity decay during ball penetration in a ground
medium, the single free fiber break. The EMR signals were recorded by an antenna over radio
range of frequency spectrum. It is noteworthy that they reflect adequately the mentioned dynamic
fracture processes. So, when a single crack propagates with a constant or varying velocity, the
EMR signal behaves in like manner. Moreover, the high frequencies appear also due to micro-
cracking on the crack surfaces, and even their vibrations with low frequencies after the crack stop
is designed clearly on the oscillogram.
Series of tests on edge splitting of plates has been also conducted. Sizes of specimens from
polymetilmetacrilate (PMMC), glass-epoxy (GE) and carbon-carbon composite (CCC) (the last
two are the unidirectional fiber-reinforced composites) were approximately uu
mm³. The artificial initial 10-15 mm length cut along the central plane was sharpened. Each
specimen was loaded step by step with the incremental normal force on the split plate edge. The
determination of fracture parameters was performed by a compliance calibration technique, in
particular, for the purpose of validation of the tests. The theoretical predictions base on the
principles of linear fracture mechanics and the two conventional methods are demonstrated in the
tests. To determine the fracture toughness and Young’s module, the limit load and edge
displacement were measured at the crack start. These values are in agreement with the tabular
ones. Another concerns the calculation of new crack surface and the loss of work per the "loading-
unloading" cycle on the "force-displacement" quarter-plane during a crack jump. Then the energy
release rate can be determined. The acoustic emission and the outward electric field near to crack
tip were simultaneously recorded.
The load and displacement as a mechanical field, along with the AE and EMR, say, the
physical fields, exhibit different feature before the crack start and during its galloping motion. The
AE and EMR appear at the sharp increasing the load when a damage (micro-cracking, fiber break)
apparently occurs nearby the stationary crack front. The downward load jump corresponds to the
crack start and a maximum of the physical fields activity. The analysis of correlation of these
fields and comparison their amplitude and frequencies for different structures with each other are
the new elements in the failure/fracture. It was revealed that the GEC gives maximum the AE and
EMR activity among the materials tested. This reflects the intense micro-cracking of GEC which
has been proved by visual inspection. The minimum accounts for the CCC that is rather more
“perfect” structure than GEC. The EMR pulses turn out to possess the good sensitiveness and
104 Yu. K. Bivin et al.

correlate with AE data in the main. But the electric activity minimum in the case of CCC is also
explained by good carbon conductivity.
It can be concluded that the EMR record is the very simple, non-destructive, temperature
independent, and cheaper method for detection of different dynamic fracture stages into structural
materials and determination of limit loads. To use this method in combination with the AE
technique will lead to essential increasing reliability for the event identified, especially, in the
cases when the noise influence is expected. The future experimental data will also be presented.
This work was partly supported by the Program N13 of the Russian Academy of Sciences and
the Russian Foundation of Basic Researches, N 05-01-00628.
2T1. Physical aspects of fracture 105

MICROMECHANICAL MODELING OF GRAIN BOUNDARY RESISTANCE TO


CLEAVAGE FRACTURE PROPAGATION

M. Stec and J. Faleskog


KTH Solid Mechanics, Royal Institute of Technology
Osquars Backe 1, SE-100 44 Stockholm, SWEDEN
mateusz@hallf.kth.se, jonasf@hallf.kth.se

Failure of polycrystalline structures due to cleavage fracture can be summarized in three critical
steps: (i) fracture of a brittle particle, e.g. a carbide, cause rapid growth of a micro-crack, (ii) which
must generate sufficient energy release to propagate into the much tougher surrounding matrix
material. In order to become critical, (iii) the micro-crack must propagate over grain boundaries
into neighbouring grains. An initiated microcrack may arrest at several instances. In a
micromechanical analysis by Kroon and Faleskog [1] step (ii) was examined and they found that a
micro-crack can arrest before the first grain boundary is reached. Based on a weakest link concept
Anderson et al. [2] studied cleavage crack growth across hexagonal grains and showed that grain
boundary resistance have a direct influence on the fracture toughness threshold. Moreover, they
also showed that if the microcrack arrests, it will most likely arrest at the first grain boundary
encountered. In this study the conditions for a microcrack to arrest at the first encounter with a
grain boundary is investigated using micromechanical analysis.
The cleavage planes in ferritic steels is one of the three {100} planes (Miller index notation).
At a grain boundary, the orientation of cleavage planes typically changes, as illustrated in Figure
1(a), where the cleavage crack has propagated from grain A into grain B. The relative difference in
orientation between adjacent cleavage planes can be characterized by two angles: a tilt-angle and
a twist-angle , see Figure 1(a). A grain usually has about 5-6 neighboring grains. The distance
along the grain boundary, where grain A meets primary cleavage planes of grain B, is here on the
average taken as w, see Figure 1(c). Experimental estimations of w can be found in a series of
fracture tests carried out by Qiao and Argon [3]. The micromechanical analysis is in the present
study inspired by the experimental observations made by Qiao and Argon [3] and based on an
idealized model of two grains, where grain A is modeled as a circular cylinder entirely surrounded
by grain B, see Figure 1(b). Furthermore, the primary cleavage planes in grain A and grain B are
assumed to intersect each other with a periodicity of w, as is illustrated in Figure 1(c). In order to
understand to what extent the grain boundary can act as a barrier and obstacle, an effective energy
release rate for a microcrack that has propagated to radius R ( t R G B ) an be formulated as

w
*e K 2 * A  (1  K 2 ) * B f B (\ , M )  K 2 * GB f GB (\ , M )
R GB . (1)

Here, *A , *B and *GB are critical energy release rates associated with grain A, grain B, and the
grain boundary, K RGB R (see Figure 1(a)), and f B and f GB are functions of the
misalignment angles ( ȥ , ij ) and is of order unity.
Both grain A and grain B are modeled as elastic viscoplastic materials. The crack growth is
modeled using cohesive surfaces, where the tractions are governed by a modified exponential
cohesive law in order to control the initial slope. The micromechanical model was numerically
analyzed by use of finite element modeling. The analysis consisted of two phases. In the first one,
true (Cauchy) tractions were applied quasi-statically on the remote boundary (see Fig. 1(b))
106 M. Stec and J. Faleskog

corresponding to the axial and radial stresses Ȉ Z and Ȉ R , respectively. In the following dynamic
phase both Ȉ Z and Ȉ R were held constant. The cleavage crack was then initiated in grain A and
started to grow in the radial direction until it encountered the grain boundary. The crack plane
either penetrated the grain boundary and continued into grain B or arrested. The main question in
the analysis was: given the overall stress state—how large can grain A be in order for the cleavage
crack not to penetrate grain B and possible cause catastrophic failure. The critical stress state
required to propagate the cleavage microcrack across grain boundary into grain B and further
depends primarily on: stress ratio [6R 6Z ] and grain boundary features
[ ȥ , ij , w * G B ( R G B * B )].It should also be pointed out that the characteristics of the cohesive
surfaces and the elastic viscoplastic behavior affects the results.

Figure 1. Grain boundary model characterization: (a) orientation of cleavage planes, (b) idealized
micromechanical model and (c) crack plane definition at grain boundary.

References
1. Kroon M., Faleskog J., J. Mech. Phys. Solids, vol. 53, 171-196, 2005
2. Anderson T., Stienstra, D. and Dodds, Jr. R. H., In Fracture Mechanics: 24th volume, ASTM
STP 1207, edited by J.D. Landes, et al. ASTM, Philadelphia, 1994, 186-214.
3. Qiao Y., Argon A.S., Mech. Mat., vol. 35, 313-331, 2003.
2T1. Physical aspects of fracture 107

MICROSTRUCTURE OF REACTOR PRESSURE VESSEL STEEL CLOSE TO


THE FRACTURE SURFACE

M. Karlik, P. Hausild and C. Prioul1


Czech Technical University, Faculty of Nuclear Sciences and Physical Engineering,
Department of Materials, Trojanova 13, 120 00 Praha 2, Czech Republic
1
Ecole Centrale Paris, LMSS-Mat,Grande Voie des Vignes, 92295 Châtenay–Malabry, France
Miroslav.Karlik@fjfi.cvut.cz, Petr.Hausild@fjfi.cvut.cz, Prioul@mssmat.ecp.cz

A transmission electron microscopy study of microstructure in the vicinity of fracture surface was
carried out in French tempered bainitic nuclear reactor pressure vessel steel 16MND5 (equivalent
to the American standard A508 Class 3). Cross-section thin foils from nickel electroplated Charpy
V-notch specimens fractured in impact and quasi static loading at -30°C (ductile to brittle
transition temperature region) were prepared in the ductile tearing and cleavage regions of the
fracture surface (Fig. 1).

FIGURE 1. Slice cutting in the central part of a Charpy V-notch sample and the TEM disc
positions in cleavage – C and ductile – D fracture zones.

In the ductile tearing zone, the microstructure was very heterogeneous. Dislocation cells, shear
bands, and fine heavily deformed subgrains were found. The deformation in the impact specimen
was often localized only in the vicinity of the fracture surface, where long thin cells formed due to
dynamic recovery (Fig. 2). In the quasi static three-point bend specimen, the localization was
found also in deeper areas under the fracture surface. There were shear bands (bundles of long thin
cells) mostly aligned in parallel to the fracture surface. Numerous areas of the shear band
intersections (at ~45° to the main shear band direction) were also observed.
Comparing the microstructure in the ductile tearing zone of the Charpy impact specimens with
the corresponding part of the fracture of the CT25 fractured in quasi-static loading in previous
research [1], there are two similar features. The deformation is very heterogeneous - heavily
deformed material with very small and dislocation arranged in cells. On the other hand, there were
no twins, neither shear bands found in Charpy specimens. The shear bands were transformed by
dynamic recovery into long thin subgrains.
108 M. Karlík et al.

FIGURE 2. Microstructure of the steel in the ductile tearing region - very thin subgrains (50 to 100
nm) formed due to the local increase in temperature leading to dynamic recovery of dislocations
during rupture of impact specimen.

In the cleavage zones of both types of specimens, only an increased dislocation density was
found, no twins were observed. The reason for the absence of twins could be that the material of
the much smaller Charpy specimen was plastically deformed (bending of the test specimen) before
the cleavage crack propagation and the twins could not form due to an increased density of
dislocations.

References
1. Karlík, M., Nedbal, I., Siegl, J., Mater. Sci. Eng. A 357, 2003, 423-428.
2T2. Brittle fracture 109

BRITTLE FRACTURE IN HEAT-AFFECTED ZONES OF GIRTH WELDS OF


MODERN LINE PIPE STEEL (X100)

A. S. Bilat, A. F. Gourgues-Lorenzon, J. Besson and A. Pineau


Centre des Matériaux, Ecole des Mines de Paris,
UMR CNRS 7633, B.P.87, F-91003 Evry cedex
anne-sophie.bilat@mat.ensmp.fr, anne-francoise.gourgues-lorenzon@ensmp.fr

Gas field development requires a cost reduction of gas transport. For that purpose, one solution is
to increase gas pressure inside pipes. High strength steels of API grade X100 (yield stress above
690 MPa) are potentially good candidates for these new applications [1], but the toughness of
X100 welded joints involves characterization. The aim of this project is to improve the accuracy of
prediction tools of the toughness and defect acceptability in such a GMAW welded joint, by
building a quantitative prediction tool of the risk of brittle fracture, adapted to the girth welds of
new X100 steels, and by determining and validating a crack initiation criterion in the most critical
zones of this welded joint. This study is conducted according to the local approach to fracture
procedure [2]: first, metallurgical and mechanical characterization of the girth weld, then, thermal
welding cycle simulation of the most critical zones and characterization of their fracture properties
and, finally, finite element modelling of brittle fracture initiated at these critical zones.
Hardness and microstructures were thoroughly characterized in each zone (Fig. 1). The weld
consists of three main zones: the base metal (BM), the weld metal (WM) and the Heat-Affected
Zone (HAZ). The BM microstructure is a textured bainitic matrix with ferrite and some secondary
phases; the WM is acicular ferrite and the Heat-Affected Zone is upper bainite. The HAZ is itself
divided in three zones: the coarse-grained (CG)HAZ (austenite grain size around 30 µm, spread
over 100 µm) near from the WM, the fine-grained (FG)HAZ (austenite grain size around 10 µm,
spread over 100 µm), and the intercritical (IC)HAZ (MB tempered between Ac1 and Ac3 spread
over 1 to 2 mm).
The mechanical strength of the joint was characterized by using tensile and impact tests (Fig.
2). Specimens were cut from the BM, and the WM, but the HAZ is too small to machine
homogeneous specimens, so that it is necessary to reproduce some HAZ microstructures. Two
microstructures (CG and FG) are reproduced by applying a simulated weld thermal cycle on base
metal blanks with a thermal-mechanical simulator (Gleeble). The welding thermal cycles are
determined experimentally by inserting a thermocouple in the heat affected zone of the joint during
a real welding operation and the Rykaline model [3]. All these specimen are then machined and
mechanically tested in tension at various temperatures. The weld is matched in yield stress
(YSWM=YSBM), and overmatched in tensile strength (TSWM>TSBM), because the base metal of
the pipe exhibits little work-hardening. The impact tests show low fracture energy at –20 °C, when
the specimen is taken from the inner subsurface part of the pipe wall and the notch is located on the
fusion line (FL i). The brittle fracture of these specimens is systematically localized in the CG
zone, which consists of upper bainite with Martensite-Austenite constituents.
The mechanical properties of the various zones of the weld will be used in a numerical finite
element model together with a statistical brittle fracture criterion to describe the behaviour of this
“multi-material” assembly.
110 A. S. Bilat et al.

Acknowledgments: Technical support and constant interest from Serimer-Dasa, Gaz de France
and Europipe as well as financial support from CEPM are gratefully acknowledged.

References
1. Glover, A., In Proceedings of the international Pipe Dreamer’s Conference, co-organized by
M.Toyoda and R. Denys, Yokohama, Japan, 2002, 33-52.
2. Besson, J., Local approach to fracture, Presses de l’Ecole des Mines, Paris, France, 2004.
3. Rykaline, N. N., Calcul des processus thermiques de soudage, Soudage et Techniques
connexes, 1961 (in French)
2T2. Brittle fracture 111

CLEAVAGE FRACTURE OF STEELS AT VERY LOW TEMPERATURES

R. Rodriguez-Martin, I. Ocana and A. Martin-Meizoso


CEIT, Centro de Estudios e Investigaciones Técnicas de Guipúzcoa and
TECNUN, Escuela Superior de Ingenieros, University of Navarra
Paseo Manuel de Lardizábal, 15, 20018 San Sebastián, Spain
rarodriguez@ceit.es, iocana@ceit.es, ameizoso@ceit.es

The mechanical strength of polycrystalline metals is frequently associated with the propagation of
defects present in their structure as cracks or dislocations. Nevertheless, when these defects are
absent, the crystal lattice becomes decisive and the strength is determined by the stress at which the
lattice losses its stability. This stress is known as the ideal strength, Clatterbuck et al. [1], Krenn et
al. [2], and it is especially relevant in experimental situations where there are few mobile defects.
This paper proposes that, in bcc metals, brittle fracture at very low temperatures is caused
when an ideal cleavage strength is reached in their structure. At such low temperatures, the
movement of the defects is frozen and the material will behave as if it were free from defects.
Under these conditions, a common bcc metal would cleavage on (100) planes under an uniaxial
tensile stress. The mechanism proposed lies in the initial breakage of the grain having the best
oriented (100) plane with respect to the applied load. Subsequently, other grains with a favourable
orientation take its load and may break, causing the macroscopic failure of the test-piece or
component. Figure 1 shows an example of a random grain distribution with their respective
orientation for their (100) planes.

FIGURE 1. Schematic of a fatigue pre-cracked body with a random grain distribution at the fatigue
crack front. The orientation of a (100) plane is showed for some grains.

The value of the ideal cleavage strength on (100) has been investigated for many bcc metals on
the basis of ab-initio electronic structure calculations, [1-2], Friak et al. [3]. From these
calculations, cleavage fracture of the high strength steels of which chemical compositions are
listed in Table 1, has been analysed.
112 R. Rodriguez-Martin et al.

TABLE 1. Chemical composition of steels (%).

Previously precracked compact tensile specimens extracted from these materials have been
used in fracture tests. These tests were conducted at very low temperatures (liquid nitrogen).
Subsequent fractographic analyses showed the presence of multiple initiation sites situated close to
the fatigue crack in the fractured specimens of all the materials. Also a low-scatter in the obtained
results suggested that a similar mechanism causes the brittle failure for the three materials studied.
Taking into account all these experimental observations, the possibility of ideal cleavage
strength to control the brittle fracture makes sense. In order to predict this fracture behaviour at low
temperatures, a simulation program has been developed in Matlab£ code. The program generates
a random grain distribution for a precracked body. The crack is assumed to blunt under load,
following the behaviour proposed by McMeeking [4]. From this point forward, the program
calculates the applied load required for the failure of the random grain distribution by cleavage.
This is achieved by taking into account that the ideal cleavage strength is reached in the grain with
the best oriented (100) plane and subsequently transmitted to other grains with a favourable
orientation.

References
1. Clatterbuck, D.M., Chrzan, D.C. and Morris, J.X., Acta Mater., vol. 51, 2271-2283, 2003.
2. Krenn, C.R., Roundy, D., Morris, J.W., Cohen, Marvin L., Mat. Sci. Eng. A319-321, 111-114,
2001.
3. Friak, M., Sob, M. and Vitek, V., Phil. Mag., vol. 83, 3529-3537, 2003.
4. McMeeking, R.M., J.Mech.Phys.Solids., vol. 25, 357-381, 1977.
2T2. Brittle fracture 113

NEW FORMULATION OF THE RITCHIE, KNOT AND RICE HYPOTHESIS

A. Neimitz, M. Graba and J. Galkiewicz


Kielce University of Technology
Al. 1000 lecia P.. 7, 25-314 Kielce, Poland
neimitz@tu.kielce.pl

Several authors adopted a local fracture criterion to assess the constraint influence on fracture
toughness. They assumed that the cleavage fracture to happen requires that the opening stress
reaches critical value, Vc at certain distance from the crack tip, rc or within the certain volume in
front of the crack tip.
O’Dowd at al [1] adopted modified small strain HRR solution, henceforth called the OS
model, to derive simple, approximate formula in order to predict the influence of in-plane
constraint on fracture toughness of structural element. According to O’Dowd’s model the critical
conditions must be satisfied independently the level of constraint, which was quantified by the
actual value of the Q-stress, utilizing the OS theory.
O’Dowd’s final formula to compute the actual value of fracture toughness is as follows:
n 1
§ Q ·
JC J IC ¨¨ 1  ¸¸
© V c /V 0 ¹ (1)
where J is the J-integral, Vo is a yield stress, subscript c or Ic denotes critical state, Q is a parameter
“measuring stress level in the modified OS model [1]
Eq. 1 follows from the Ritchie, Knot, Rice (RKR) hypothesis provided one assumes that the
stress field in front of the crack is singular (small strain assumption). In such a case the opening
stress is greater than critical one over the distance rc from the crack front. However, the critical
stress Vc is simply the parameter which adjusts theoretical prediction to experimental results.
Almost all other theories concerning the local approach to fracture use the stress distribution
in front of the crack characteristic for the finite strain. In real elastic-plastic materials the crack tip,
originally sharp, is blunted by plastic deformation. This process modifies the small-strain HRR
solution in front of the crack. Instead of singular stress field at the crack tip the opening stress
decreases towards the blunted crack tip after reaching a maximum at the normalized distance
r=MJ/V0 from the crack tip. Here, M is a function of constraint level, and the Ramberg-Osgood
power exponent n at given external loading.
In such a case the RKR hypothesis should be based on another assumption.
Hypothesis: It is assumed that the cleavage fracture may happen if the opening stress in front
max
of the crack exceeds the critical level over the distance l greater than lcrit ( V 22 >Vc and ltlc ).
max
If V 22 >Vc but l<lc fracture may proceeds due to void’s growth and coalescence, provided
the plastic strains are sufficient large. From simple geometrical consideration it follows:

V max  V c
l # 2 rmax
V max  EV 0 (4)
where EVo is equal to the opening stress component at the blunted crack tip.
114 A. Neimitz et al.

In order to apply this model to estimate fracture toughness at arbitrary in-plane constraint level
we assume that lc is characteristic length for material of interest.
To apply proposed hypothesis detailed numerical investigations of the stress field in front of
the blunted crack have been performed. They lead to general observations:
The maximum, value of the opening stress component saturates to a constant level which does
not depend on external loading and depends weakly on the in-plane characteristic specimen size a/
W. It depends on the deformation properties of the material, n and V0/E. It happens when plastic
zone covers an essential part of the ligament in front of the crack tip. However, for a very short
crack, e.g. a/W=0.05 or less the maximum stress at the crack front becomes more sensitive to the
change of the a/W ratio. In this case the maximum value decreases slightly with decreasing a/W.
The maximum, saturated value of the opening stress is more sensitive to the change of the a/W
ratio for the larger values of the n exponent.
The stress maximum is located at the distance r=MJ/V0 from the crack tip. It can be shown that
the coefficient M is a function of n, V0/E and the in-plane constraint measure Q when the maximum
stress saturates to a constant value. In fact, this function does not change with external loading but
changes with the ratio a/W. When the stress maximum still changes with external loading the
max
coefficient M depends on the V 22 also.
Basing on above observations the analytical formulas were derived to assess the influence of
the in-plane constraint on fracture toughness of structural element. One of them which satisfies
deformation pattern in front of the crack in the Sumpter and Forbes experiments [2] is as follows:
1 n
ª 1 1 º
«§ 1 · 1 n Q ª E º 1 n »
JC J IC «¨ ¸  ~ « » » MQ 0
«¨© M Q 0
¸
¹ V 22 ¬ DV 0 I n ¼ »
«¬ »¼
(5)
or
(1 n )
§ Q ·
JC J IC ¨¨1  max ¸¸
© V 22 / V o ¹ (6)
max
Where MQ=0, or V 22 must be computed numerically. Above equations fit experimental results
very well. More details of the numerical results as well as another equations will be presented in
the paper and presented during the conference.

References
1. O’Dowd, N.P. and Shih, C.F., J. Mech. Phys. Solids, vol. 49, 939-963, 19
2. Sumpter, J.D.G. and Forbes, A.T., Black, P.A., In Proceedings of TWI/EWI/IS Int. Conf.
Shallow Crack Fracture Mechanics Tests and Applications, Cambridge, U.K., paper 7, 1992.
2T2. Brittle fracture 115

THE EFFECT OF THE RATE OF DISPLACEMENT ON CRACK PATH


STABILITY

D. A. Zacharopoulos and P. A. Kalaitzidis


School of Engineering
Democritus University of Thrace
GR 671 00, Xanthi, Greece
dzachar@civil.duth.gr, pkalait@civil.duth.gr

The search for methods for forecasting the crack path and its type, finally became an issue of
particular interest due to the increasing use of advanced materials like composites or coated
materials (substances) for improvement of their thermomechanical properties. With the increasing
use of structural adhesives in construction, aerospace, and automotive industries the need for an
estimate of the locus of failure and the crack path propagation is essential in ordrer to improve the
durability of the bonded joints. In addition, the prediction of the crack path can be beneficial in the
design of safe structures and gives answers on the possible initial conditions of loading in the case
of a destructive fracture.
We will try to clarify the concepts of ‘stable’ and ‘unstable’ crack path in the fracture of a solid
or a structure.
Part of the path of the propagated crack can be characterized as ‘stable’ if and only if this part
resulting from repeated experiments onto bodies with the same geometry and under the same
loading conditions, appears with identical geometrical characteristics. Specifically, when a stable
crack path situation prevails at the spread of a fracture in specimen and leads to the catholic
destruction of the specimen, the corresponding broken pieces have the same shape.
The verification of a theory or a method for the approach of a problem is usually based on the
experimental process. However, if the results are exceptionally sensitive on the initial and on
general conditions that exist during the experiment, then a scattering of the results is very likely.
The evaluation of this scattering in phenomena of "instability" can lead to erroneous conclusions.
In other words, in each experimental process we should compare the scattering of the results owed
to the problem’s endogenous instability factors with the magnitude of the experimental conditions’
divergence, like the constrained displacements, or the geometry of the specimens. Observations on
experimental results show that the propagation trajectory of an extended crack depends on the
material properties, the geometry of the specimen, the rate of loading or displacement , the
dynamic loading and the temperature. Furthermore, the control of the load’s increase or the
displacement’s increase on the specimen by the loading machine also plays an important role. On
the interface, the propagation trajectory also depends on the tensile strength and fracture toughness
of the bonded materials. Approaching the multifactor problem of stability for a part of a crack path
we could verify that all the above factors participate in the configuration of a situation that will
determine the stability of the crack path, adding the history of the crack path and the magnitude of
the applied forces.
A more developed study on the method, determining the crack path stability can be found at
[1]. For any two-dimensional problem with an arbitrary geometry of the body and arbitrary
constraints, the stress and strain fields can be calculated by developing the displacement and
loading results, by using a finite elements program. Using that data, we can draw the contour map
of strain energy density with a graphical program. This map has common features with a
geographic map: the point where the failure begins, according to the basic hypotheses, is the hilltop
116 D. A. Zacharopoulos and P. A. Kalaitzidis

of the topographic map respectively and it is surrounded from closed contours that may have U
shape. The higher contours are always enclosed by lower ones.
In the present work, we will demonstrate how we can estimate the crack path stability with the
elaboration of this map, when a very slow rate of displacement is imposed on the specimen. When
the critical locus on the plane body is the point O, which can be a crack tip, according to the above
hypotheses, the beginning, the initial direction of crack growth as well as the crack path of
propagation will occur. According to the strain energy density criterion (see [2,3]), the initiation of
r >
fracture from the O, takes place along the direction OL, (OL)= c , when the dW / dV max min L @
reaches the critical value (dW/dV)c. The predicted crack path during propagation is the curve that
starts from point L passes the points with the maximum gradient of (dW/dV) and ends up at point
G, where the global minimum value of (dW/dV) develops.
On the map, the crack path is indicated by V shaped contours of the strain energy density. If
the apicals of the Vs points are joined by a line, then the resulted plot curve, starts from the peak O
and arrives in the vicinity of the point G. In the geographic map the curve OG, represents a gorge
or a riverbed, which starts from the hilltop O. The drawing of this gorge can give additional
information for the estimation of the crack path’s stability according to following hypothesis:
The stability of the crack path can be deduced from the degree of the sharpness with which the
curve of the "gorge" is drawn.
By collecting the results for different geometric characteristics of the specimens, we create a
classification diagram of crack path stability.
Our experiments were performed on double cantilevering beam (DCB) specimens made from
PMMA material. The tests were performed by using a testing machine with controlled
displacement at a rate of 0.5mm/min or less. These slow rates were chosen to reveal the specimens
behavior while transiting across regions with different crack path stability of the classification
diagram.

References
1. Zacharopoulos, D.A., “Stability Analysis of Crack Path Using the Strain Energy Density
Theory”, Theoretical and Applied Fracture Mechanics, vol. 41, 327-337, 2004.
2. Sih G.C., “Introductory chapters in Mechanics of Fracture”, Vols. I to VII, ed. G.C. Sih,
Martinus Nijhoff: The Netherlands, (1972-1982).
3. Gdoutos, E.E., “Fracture mechanics”, Kluver Academic Publishers, 1993.
2T2. Brittle fracture 117

SCRATCHING AND BRITTLE FRACTURE OF SEMICONDUCTOR IN-SITU


SCANNING ELECTRON MICROSCOPE

K. Wasmer, C. Pouvreau1, J. Giovanola1 and J. Michler


Swiss Federal Laboratories for Materials Testing and Research (EMPA)
Feuerwerkerstrasse 39, 3602 Thun, Switzerland
1
Swiss Federal Institute of Technology Lausanne (EPFL), ME A1 400, Station 9, 1015 Lausanne,
Switzerland
kilian.wasmer@empa.ch

Mastering the brittle facture (also called cleavage) of semiconductors is important when separating
complex devices produced on epitaxially grown multi-layers or when processing III-V optical
devices. Cleavage is usually performed in two stages. The surface is first micro-scratched with a
sharp diamond tip, which induces structural defects. Secondly, by applying a load to the sample,
fracture propagation is initiated, starting from the defect area. However, the mechanisms leading to
the preparation of high quality cleavage surface are often not fully understood. The main objective
of this paper is to present the fundamental phenomena occurring during the scribing and
subsequent fracturing process of brittle semiconductors via in-situ Scanning Electron Microscope
(SEM).
In the literature, nano-indentation has been largely investigated also for semiconductors,
Bradby et al. and Grillo et al. [1-2], but very little information is available in comparison in the
field of scratching, Ballif et al. and Gassilloud et al. [3-4]. In this paper, the complete sequence of
scratching is dissected via in-situ SEM observations. In order to perform these observations, a
miniaturized microscratch device has been built for use inside a scanning electron microscope,
Rabe et al. [5]. Also performed are Transmission Electron Microscope (TEM) investigations to
visualise the network of dislocations. It has been found that the scratching operation can be divided
into 5 distinct regions which are dependent of several parameters such as the shape of the diamond
tip, the scratching direction and the applied loads. The first one is the elastic regime which cannot
be observed in normal condition due to the total recovery after the withdrawal of the load. The
second is the plastic regime where dislocation nucleation and / or slip bands are generated. This is
followed by the subsurface cracking regime where median cracks (MC) along crystallographic
planes are created. Then, cracks such as radial and lateral appear at the surface. The last regime
takes place when the radial and lateral cracks join together since at this point a large amount of
chipping out occurs. In order to obtain atomically flat cleavage surfaces, only the median crack,
which depth will influence significantly the crack initiation and propagation, is required. Based on
our experiments, it has been also seen that the depth of the MC can be assumed to be either linear
or to the power 3/2 to the applied load as described by Eq. (1). This latter expression has already
been reported to be valid for the nano-indentation Lawn [6]. Furthermore, the influence of the
different epilayer on GaAs wafer is perceived to influence the onset of the crack median as well as
its depth for loads ranging between 10mN and 50 mN.

(a) P v c ˜ d and (b) P v c ˜ d 3 / 2 (1)


The second part of the paper scrutinise the fracture process in brittle semiconductors such as GaAs.
This includes the crack initiation of a defect in GaAs wafer, GaAs + epilayer wafer and laser diode
processed on GaAs wafer with a thickness of around 150 m. The defect is assumed to be semi-
elliptic with a length of 400 m and a depth in ranging from 5 to 20 m depending on the applied
load. The wafer is then subjected to bending stress through two apparatus allowing in-situ cleavage
118 K. Wasmer et al.

of devices. The first one, integrated into the SEM chamber, has a stroke of several hundred of
micrometers and is "displacement-controlled". Its displacement control is better than 100 nm and
the force applied by the cleaving rod is acquired simultaneously. The second apparatus is very
similar to the first one with a major advantage which is to be able to follow the crack during its
propagation. With these tools, it has been demonstrated that although, semiconductors such as
GaAs are considered as very brittle, it is possible to control the crack propagation when a big
enough starter crack is present. For uncontrolled crack growth, features are visible on the cleaved
surface which is consistent with the dynamic fracture mechanics principle Sauthoff et al. [7].
Finally, the effects due to the different epitaxial layers are in line with those observed on the depth
of the median crack.

FIGURE 1: SEM pictures of a (a) scratch with a Cube Corner diamond tip; b) cleave showing
crack deviation

References
1. Bradby, J. E., Williams, J. S., Wong-Leung, J., Swain, M.W., MV, and P. Munroe, J. Mater.
Res. Vol. 16, pp: 1500-, 2001.
2. Grillo, S. E., Ducarroir, M., Nadal, M., Tourni, E. and Faurie, J.-P., J. Phys. D: Appl. Phys.
Vol. 36, pp: 5-, 2003.
3. Ballif, C., Wasmer, K., Gassilloud, R., Pouvreau, C., Rabe, R., Michler, J., Breguet, J.-M.,
Solletti, J.-M., Karimi, A. and Schulz, D. Advanced Engineering Materials, June 2005.
4. Gassilloud, R., Ballif, C., Michler, J. and Schmuki, P., Submitted at Journal of Material
Research.
5. Rabe, R., Breguet, J.-M., Schwaller, P., Stauss, S., Patscheider J. and J. Michler, Thin Solid
Films, Vol. 469-470, pp: 206-213, 2004.
6. Lawn, B., "Fracture of Brittle Solids", 2nd Ed., Cambridge University Press, 1997.
7. Sauthoff, K., Wenderoth, M., Heinrich, A. J., Engel, K. J., Reusch, T. C. G. and Ulbrich, R.
G., Phys. Rev. B, Vol. 60, pp: 4789-95, 1999.
2T2. Brittle fracture 119

CRACKS IN THIN SHEETS: WHEN GEOMETRY RULES THE FRACTURE


PATH

Pedro M. Reis, Basile Audoly1 and Benoit Roman2


Levich Inst., City College of New York,
140th St. & Convent Av., New York, NY 10027, USA
1
LMM, UMR 7607 CNRS/UPMC, 4 place Jussieu, case 162, 75252 Paris cedex 05, France
2
PMMH, UMR 7636 CNRS/ESPCI, 1O rue Vauquelin 75231 CEDEX 5 Paris, France
preis@levdec.engr.ccny.cuny.edu

We study both experimentally and theoretically the propagation of brittle fracture coupled to large
out-of-plane bending, as when a brittle elastic thin sheet is torn by a rigid moving object [1-3].
Taking into account the separation of the film’s bending and stretching energies and using classical
fracture theory we have shown that such cracks can propagate according to a simple set of 2D
geometrical rules in the limit of vanishing thickness, h [3]. Numerical integration of our
geometrical model accurately reproduces both the shape of the fracture pattern and the detailed
time evolution of the propagation of the crack tip.
In our experiments, a rigid object, the cutting tool, is forced through a thin polymer film and
tears through the material as it advances. A schematic diagram of the process is presented in Fig.
1(a,b). The film is clamped at its lateral boundaries imposing no initial tension. The cutting tool is
oriented perpendicularly to the horizontal film and is driven through the material at constant
velocity, v, in a direction parallel to the film’s major length. We have used biaxially-oriented
polypropylene (BOPP) and acetate films (25<h<127Pm) which, being brittle, undergo negligible
irreversible deformation besides fracture.

FIGURE 1. Schematic of an oscillatory path, P, obtained when a cutting tool, I, is driven at


constant velocity through a thin polymer film clamped at two of its lateral boundaries. L and A are
the wavelength and amplitude of the pattern. (b) 2D projection of the cutting process. (c)
Photograph of an experimental crack path. 30mm BOPP film, v=1.6mms-1 , rectangular cross
section tool (w=6.35mm). Videos of the fracture process can be found in [4].

This setup yields surprisingly complex crack propagation: for tools considerably wider than
the film's thickness the crack tip, T, follows a highly reproducible non-sinusoidal oscillatory path
as shown in Fig. 1(c). Each single period of this path consists of two smooth curves separated by a
120 P. M. Reis et al.

kink, at which there is a sharp change in the direction of curvature. Propagation is primarily quasi-
static but is interrupted by periodic bursts of dynamic propagation immediately after each kink. By
decreasing the size of the cutting tool down to widths comparable to the film thickness, the crack
path eventually becomes straight, as reported in [1], but here we focus on the oscillatory behaviour
far above threshold. In this regime, the crack morphology is independent of v (as long as this
remains much smaller than the speed of sound in the material), h (provided h>>w) and the film's
width, D. The only relevant lengthscale in this scenario appears to be the width of the cutting tool,
w, with which both the pattern's amplitude, A, and wavelength, O, scale linearly [1].

FIGURE 2. Motion of the crack tip from the model compared to experiments for a cylindrical
cutting tool (w=20.4mm, h=50mm, v=1.2mms-1). (a) Time-series for the crack tip coordinates x(t)/
w and y(t)/w as a function of dimensionless tool advance vt/w. Note the periodic gaps in crack
position, corresponding to dynamic propagation. (b) Final pattern in the film plane (x,y).

We shall show that the classical equations for elastic plates and for crack propagation can be
reduced to a simple set of 2D geometrical rules, which explain the experimentally observed crack
behaviour. Following a common procedure in fracture theory our approach is as follows: i) we
first calculate the elastic energy of the system, taking into account the possible large out-of-plane
deformations of the film induced by the cutting tool, ii) we then apply Griffith's criterion for crack
propagation and iii) finally establish the direction of propagation of the crack tip, E tip according
to the principle of local symmetry. In our model, the mechanical properties of the film are captured
by the two physical parameters, D >* /(EL)@ and EaSwhere * is an effective toughness of
the film (dimensions of surface tension), E is the Young’s Modulus and L is a typical length scale
near the crack tip. We have numerically integrated our geometrical model and have found
excellent agreement with experiments, as shown in Fig. 2. Both the kinks and the subsequent
bursts of dynamic propagation arise from a simple set of rules and are therefore intrinsic feature of
the tearing process. To our knowledge this is the first example where a complex crack motion is
entirely ruled by geometry.

References:
1. Roman B., Reis P.M., Audoly B., De Villiers S., Viguié V., Vallet D., C.R. Mecanique 331
811 (2003).
2. Ghatak A. and Mahadevan L., Phys. Rev. Let. 91, 215507 (2003).
3. Audoly B., Reis P.M. and Roman B., Phys. Rev. Lett., accepted, in press (2005);
4. Audoly B., Reis P.M., Roman B., http://www.lmm.jussieu.fr/platefracture/
2T2. Brittle fracture 121

CLEAVAGE MECHANISMS IN A SHIP PLATE STEEL

R. Cuamatzi, I. C. Howard and J. Yates


Mechanical Engineering Department, Sheffield University,
Mappin Street, Sheffield, S1 3JD, UK
R.Melendez@sheffield.ac.uk

The proportion of Grade A ship plate steel for the construction of merchant ships is huge, however
there is an increasing concern about the susceptibility of Grade A steel plate to brittle fracture. A
recent ship failure [1] illustrate the importance of the concern. At low enough temperature, the
fracture of structural steels is often dominated by cleavage. McMahon and Cohen [2] demonstrated
that the cracking of carbide particles located at ferrite grain boundaries represents a primary
initiation mechanism in mild steels. Further investigation led to the Smith model [3] which
assumed that a cleavage crack could initiate when a grain boundary carbide is fractured by
impingement of a dislocation pile-up, and the final fracture is controlled by the carbide-sized crack
propagation into the neighbouring ferrite grain. A number of studies [4-8] have used notched
specimens to investigate the mechanism controlling cleavage fracture and the fracture criterion. In
the present work, an investigation into the micro-mechanisms for cleavage fracture has been
undertaken using notched specimens. The material used for the investigation of cleavage fracture
is a Grade A ship plate steel. The chemical composition is given in table 1. The microstructure
reveals the ferrite matrix and bands of pearlite.

Table 1. Chemical composition of Grade A ship plate steel, 20 mm plate (in wt %)

A set of three blunt-notch four-point bend laboratory specimens were tested at test temperature
of -60oC, which is in the lower transition region close to the lower shelf, on a servo-hydraulic
Instron 8501 test machine under displacement control of 0.01mm/s to analyze the consistency of
the observations in each test. It is well known that in double-notched 4PB specimens, when
cleavage initiates from one notch, fracture will certainly propagate in that notch and the surviving
notch must be very close to the critical state due to the same stress condition developed in that
notch. The surviving notches were sectioned for subsequent examination using scanning electron
microscopy for the identification of microcracks and possible particles that induce those
microcracks. Metallographic samples were prepared and etched with 5% nital and numbered.

Figure 1 – Typical microcracks found about 50 m ahead of the notch root.


Observations were made of microcracks developed in front of the notch root of each slice.
Typical micrographs of microcracks are shown in Figure 1. Numbers were used to identify
122 R. Cuamatzi et al.

microcracks nucleated in lamellar pearlite microstructure and capital letters were used to identify
microcracks nucleated in the pearlite boundary. Figure 1 was taken from slice two 50m ahead of
the notch root; this figure shows microcracks nucleated in both, the pearlite lamellar
microstructure and the pearlite boundary microstructure. Microcrack 1 initiated in the lamellar
pearlite microstructure, grew and crossed the pearlite microstructure and finally stopped in the
boundary of the ferrite matrix. Microcrack 2 was nucleated in the centre of the pearlite
microstructure and grew about 5m. Microcrack A originated from pearlite microstructure and
stopped in the ferrite boundary. These two mechanisms of cleavage nucleation also appeared in
slices 3, 4 and 5. For most microcracks, cracked carbides are clearly visible. A microcrack of about
210m size was found in a pearlite colony, the distance of the tip of this microcrack to the notch root
is 184m, that crack was nucleated in the pearlite lamellar microstructure and grew along a pearlite
boundary. Once the tip reached the ferrite grain, it stopped. More microcracks closer to the notch
root were developed. For all these microcracks, cracked carbides are identified as the cleavage
nucleation site.
A JEOL Scanning Electron Microscope was used to obtain more detailed resolution to provide
evidence of cracked carbides as the controlling mechanism for cleavage initiation. Carbides with a
diameter of approximately 0.4m were found. In order to identify which of either mechanism is the
dominant one for nucleating microcracks, the microcracks were counted and it was found that the
domain mechanism for cleavage nucleation is that of microcracks nucleated in the lamellar pearlite
microstructure and that cracked carbides are the controlling mechanism for cleavage initiation
which are probably the critical event of cleavage for this steel. The critical length of the
microcracks for cleavage extension is related to the size of pearlite colonies, suggesting that the
weakest link for cleavage to take place is the pearlite microstructure. Cracks adjoining the initial
microcracks determine the catastrophic event.

References
1. Transportation Safety Board of Canada, Marine Investigation Report, Report Number
M02L0021, Hull Fracture Bulk Carrier Lake Carling Gulf of St. Lawrence, Quebec. March
2002.
2. C. J. McMahon and M. Cohen, Initiation of cleavage in Polycrystalline Iron, Acta Metall.,
1965. 13: p. 591-604.
3. E. Smith, The nucleation and growth of cleavage microcracks in mild steels, Proceedings of
the conference on physical basis of yield and fracture, Institute of Physics and Physical
Society, London, 1966: p. 36.
4. Hendrickson J.A., Wood D. S. and Clarke, D. S., The cleavage fracture of mild steel,
Transaction ASM, 1958(50): p. 656-676.
5. J. F. Knott, Some effects of hydrostatic tension on the fracture behavior of mild steel, Trans.
Iron steel institute, 1966(204): p. 104-111.
6. Eric M. Taleff, John J. Lewandowski, and Bamdad Pourladian, Microstructure-property
relationships in pearlitic eutectoid and hypereutectoid carbon steel, Metallurg. Trans.
2002.(6): p. 25-30.
7. G. Z Wang and J. H. Chen, A statistical model for cleavage fracture in notched specimens of
C-Mn steel, Fatigue Fract. Engng. Mater. Struc, 2001(24): p. 451-459.
8. S. R. Bordet, A. D. Karstensen, D. M. Knowles, C.S. Weisner, A new statistical local
criterion for cleavage fracture in steel. Part II: application to an offshore structural steel,
Engineering fracture mechanics, 2005(72): p. 453-474.
2T3. Ductile fracture 123

FAILURE BEHAVIOR OF HYBRID-LASER WELDS

A. Bajric and W. Dahl


Department of Ferrous Metallurgy, Aachen University, Intzestr. 1, 52072 Aachen, Germany
aida.bajric@iehk.rwth-aachen.de

In recent years the failure behavior of power beam welds have been investigated by means of
Finite Element Analysis. In order to determine the local stress state around a crack tip before the
crack initiation occurs, the J-Q-M concept has been applied [1]. A Q-factor is added to the crack
driving force, expressed as a J integral and account is taken of the effect of strength mismatch by
the M factor. As a results the recommendations for the application of constraint corrections have
been derived. Additionally the stable crack growth and crack path deviation have been simulated
by using the GTN damage model.
Based on these results the further numerical and experimental investigations have been carried
out on the hybrid laser welds. The objective is to analyze the failure behavior of these welds, with
different geometry and mismatch, which lie between the conventional and pure laser weldments.
First the local stress state in front of the crack tip are quantified by stress triaxiality and Q+M
in dependence of material properties, thickness, mismatch, weld seam geometry and type of
loading. Hence the conclusions are drawn, which can be used for safety assessment of the
component with hybrid laser welds.
For simulation of the stable crack growth the damage models are applied The GTN damage
model and cohesive zone model are compared with regard to prediction of the crack resistance
curves and crack path. The identified parameter sets are verified on fracture mechanics and
component like specimens with different in-plane and out-of-plane constraints. The transferability
of GTN damage parameters have already been demonstrated in [2]. The FIGURE 1 shows
measured (Exp.) and calculated (GTN) load versus diameter reduction curves for round bar
specimens with different notch radii U and net diameter DN. The damage parameters are
determined both for base and weld metal. As regards the crack growth and load-deformation
behavior of CT specimen, the well agreement is achieved between experiments (Exp.) and
calculations (GTN_Model A) (see FIGURE 2). The model A represents due to symmetry one
fourth of the specimen. For the opening displacement greater then 3 mm the crack growth is
underestimated due to the advanced distortion of the elements. This problem can be avoided, when
one half of the specimen is modelled (GTN_Model B). Since the course of crack path depends
besides on the mismatch also on the weld seam geometry, the variation of the weld seam width
across the specimen thickness will be taken in account.
124 A. Bajric and W. Dahl

FIGURE 1. Load versus diameter reduction curves for the notched round bar specimens of
homogenous base metal

FIGURE 2. Load and crack growth versus opening displacement curves for 0.4CT specimens with
hybrid laser welds
References
1. Heyer, J., Lokale Beanspruchung in angerissenen strahlgeschweißten Stahlbauteilen, Berichte
aus dem Institut für Eisenhüttenkunde, vol. 9/2004, Shaker Verlag, August 2004.
2. Nègre, P., Steglich, D., Brocks, W., Crack extension in aluminium welds: a numerical
approach using the Gurson-Tvergaard-Needleman model, Engineering Fracture Mechanics,
Nr.7, S. 2365-2383, 2004.
2T3. Ductile fracture 125

FRACTURE OF PLASTIC BODIES. DEFORMATIONS CONCENTRATORS

A. I. Khromov, A. A. Bukhanko, S. L. Stepanov and E. P. Kocherov


Institute of Machining and Metallurgy Far Eastern
Branch of the Russian Academy of Sciences
1, Metallurgov Str., Komsomolsk-on-Amur, 681005, Russia
khromov@imim.ru, bukhanko@imim.ru

On the basis of the theory of ideal rigid-plastic body the approach to definition of invariant
tensorial characteristics of fracture is formulated on the basis of standard mechanical tests on
uniaxial tensile of flat and cylindrical samples. Instead of experimentally determined
characteristics of materials fracture (relative extension and narrowing of a sample at fracture), two
are entered invariant tensorial characteristics of a degree of sample deformation. They correspond
to the moment of macrocrack initiation and critical deformation in the crack vertex determining
process of crack propagation.
One of problems of the theory of ideal rigid-plastic bodies is nonuniqueness of position, mode
of plastic area and together with it nonuniqueness of a field of displacement velocities which
define change of a body geometry. For practical use of theoretical solutions it is offered: criterion
of a choice of a preferable instant field of displacement velocities and the criteria determining
change of a velocities field in time (change of plastic area). Suggested criteria are based on
extreme principles of nonequilibrium thermodynamics.
Within the framework of the theory of ideal rigid-plastic body the approach to investigation of
areas of abrupt change of the form as strains concentrators is offered. Definition of strains fields in
the neighbourhood of the concentrator is reduced to integration of the ordinary differential
equations. The number of analytical solutions for concentrators as V-notched bar is received.
Dependence of strains fields on form change and plastic area position is investigated during plastic
current. The number of solutions of a problem about fracture of V-notched rigid-plastic bar is
offered.
The approach to investigation of fracture processes for more complex bodies models is
formulated on the basis of the obtained solutions: elasto-plastic, strengthened plastic bodies, etc. It
is offered to consider the material in the neighbourhood of vertex crack as rigid-plastic material. It
allows to carry out the analytical description of strains fields in the neighbourhood of vertex crack
and to apply new deformation and power fracture criteria.
2T3. Ductile fracture 127

3D DUCTILE TEARING ANALYSES OF BI-AXIALLY LOADED PIPES WITH


SURFACE CRACKS

Andreas Sandvik, Erling Ostby1 and Christian Thaulow


Department of Engineering Design and Materials, The Norwegian University of
Science and technology, N-7491 Trondheim, Norway
1Department of Applied Mechanics and Corrosion, SINTEF Materials and Chemistry, N-7465
Trondheim, Norway
asand@statoil.com, erling.ostby@sintef.no, christian.thaulow@sintef.no

This paper concerns 3D ductile tearing analyses of outer surface cracked pipes subjected to tension
loading and internal pressure.
Many new offshore development projects are in ultra-deep water depths with high pressure and
temperature reservoirs. Consequently, the pipelines may be exposed to extreme loadings under
service conditions and/or laying. Today, the tensile side often limits the allowable strain, and
therefore might limit the material utilization. It is believed that existing procedures/standards are
rather conservative with regard to fracture assessment for scenarios with global plastic strain
occurring. As a consequence, more knowledge is needed about the governing fracture response
parameters.
In this work we present FEM analyses of 3D outer surface cracked pipes loaded in tension with
and without internal pressure. The outer diameter, D, is 400[mm] and thickness, t, 20[mm]. Three
different crack depths and crack lengths are chosen. All the dimensions of the cracks are chosen
such that they are representative examples for possible girth weld defects. The canoe type defect
geometry assumed is illustrated in Fig.1.
The analyses are performed with a material isotropic power law hardening on the form:
n
§ H ·
Vi V 0 ¨¨1  p ¸¸
© H0 ¹ (1)
where V i is the flow stress, V 0 is the stress at the proportional limit, H p is the plastic strain,
and n is the hardening exponent. H 0 = V 0 e E , is the strain at the proportional limit, and E is the
Young’s modulus. If V  V 0 the material behaviour is linear elastic. Further, n 7 and
V0 460 [MPa] are assumed in the current analyses.
Ductile tearing is taken into account using the so-called Gurson-Tvergaard-Needleman model,
[1,2]. This model accounts for void growth and void coalescence. The void coalescence is
modelled with the Tvergaard and Needleman’s [3] “effective void volume fraction”. The value for
the initial void volume fraction parameter was fitted to represents a realistic material resistance
found from X65 pipeline-steel.
128 A. Sandvik et al.

FIGURE 1. (a) The pipe geometry with an external circumferential constant-depth surface flaw.
(b) Details of the canoe type defect with arc length, 2c, depth, a, and end radius, r, equal to the
crack depth, a.
The FEM model was solved using Abaqus Explicit [4], which originally was developed to
solve dynamic events. However, it is possible to retrieve quasi-static solutions with this solver, as
long as you prevent significant mass effects.
The results show how variations in crack lengths and crack depths, in addition to the effect of
internal pressure, influences the crack driving force and the strain capacity of the pipe.

References
1. Gurson, A.L., Continuum theory of ductile rupture by void nucleation and growth: Part I –
Yield criteria and flow rules for porous ductile materials J. of Eng. Materials, vol. 99, 2-15,
1977.
2. Tvergaard V., Influence of voids on shear band instabilities under plane strain conditions, Int.
Journal of Fracture, vol. 17, 389-407, 1981.
3. Tvergaard, V, Needleman, A., Analyses of the cup-cone fracture in a round tensile bar”, Acta
Metallurgica, vol. 32, 157-169, 1984.
4. Abaqus User’s manual, Version 6.5, Abaqus Inc., 2004.
2T3. Ductile fracture 129

NEW MODEL MATERIALS FOR DUCTILE FRACTURE STUDIES

A. Weck and D. S. Wilkinson


Department of Materials Science and Engineering, McMaster University
1280 Main St. West, Hamilton, ON, L8S 4L7, Canada
weckag@mcmaster.ca

Ductile fracture of metals involves a sequence of overlapping processes that include the
nucleation, growth and coalescence of voids. Of these, the last is the most important because it
dictates the ductility of metals but is however the less understood. As the coalescence is a
stochastic event occurring over very short strains, it is really difficult to capture experimentally.
Attempts to fabricate model materials that would simplify the analysis of the ductile fracture
process have already been made by Babout et al.[1], Gammage et al.[2], Magnusen et al.[3], Jia
and Povirk [4] and Nagaki et al.[5]. However, they are of limited help when one wants to study the
coalescence event in detail because of their complicated microstructures, of there limited ductility,
of the controlling effect of the nucleation event, of the lack of constraint (2D approaches instead of
3D) or simply because they do not reproduce the key features of the microstructure (like the void
size for instance).
The present work describes the fabrication and characterization of two new model materials
for ductile fracture studies that overcome the previously described limitations.
The first consists of a controlled two dimensional array of laser drilled holes in metallic sheets
(2D approach) and the second consists of a controlled three dimensional array of laser drilled holes
in the bulk of metallic samples (3D approach). To drill the holes, a femtosecond laser is chosen for
the small heat affected zone that is created around the hole and the small hole size (10microns) that
can be produced.
The first model material consists of an array of laser drilled holes in 100microns thick metallic
sheets (Fig.1).

The samples are then tested in-situ in a Scanning Electron Microscope to capture the
coalescence event (Fig.2). Different arrays of holes can be used and the preliminary results are that
different modes of coalescence are obtained depending on the array of holes.
For the second model material, the laser holes are drilled in thin (typically 15 microns in
thickness) metallic sheets and the sheets are then diffusion bonded to obtain the 3D structure
(Fig.3). These samples are then tested in situ in the X-Ray Computed Tomography set-up at the
synchrotrons in France (ESRF) and in Japan (SPring-8).
130 A. Weck and D. S. Wilkinson

The results from the tomography experiments (Fig.4) confirm the successful fabrication of the
model material. Furthermore, the holes growth and coalescence can be easily followed allowing
valuable experimental data to be collected to support subsequent modeling efforts.

References
1. L. Babout, E. M., J.Y. Buffière and R. Fougères, Acta Mater., 49, 2055-2063, 2001.
2. J. Gammage, D. S. Wilkinson, J. D. Embury and Y. Brechet, Acta Mater., 52, 5255-5263,
2004.
3. Magnusen, P. E., E. M. Dubensky, et al., Acta Metallurgica, 36, 1503-1509, 1988.
4. S. Jia, G. F. R., G.L. Povirk, International Journal of Solids and Structures, 39, 2517–2532,
2002.
5. Nagaki, S., Y. Nakayama, et al., International Journal of Mechanical Sciences, 40, 215-226,
1998.
2T3. Ductile fracture 131

FATIGUE THRESHOLD COMPUTATION MODEL BASED ON THE


SHAKEDOWN ANALYSIS

M. A. Belouchrani, D. Weichert1 and A. Hachemi2


Laboratoire Génie des Matériaux, E.M.P. BP 17C Bordj El Bahri Alger Algérie
1Insa de Rouen, Place Emile Blondel, BP 8 76131, Mont-Saint-Aigan France
2Institut für Allgemeine Mechanik, RWTH-Aachen, D-52056, Aachen Germany

nbelouch@yahoo.fr

The fatigue thresholds remain very significant parameters, helping the designer and the
manufacturer in their decision to reform a mechanical structure or its dimensioning.
These last years, several models were advanced in order to propose a method of their
determination or the study of their influence factors such as the microstructure or the load ratio.
The established fatigue threshold models can be classified in two groups, the theoretical
models and the models based on the experimentation. The majority, not to say all the models, agree
on the fact that the fatigue threshold increases with the size of the grain and depends strongly on
the yield stress of the material.
The fatigue threshold is interpreted physically as the sum of two components, one
i c
microstructural 'Kth and the other physics 'K th caused by the crack closing effect. So that:

' K th ' K thi  ' K thc (1)


For a structure with a regular distribution of the grains and with various orientations, the relation
allowing to connect the values of the microstructural threshold with the microstructure dimension
[ (grain size) combined with the definition of the crack propagation force related to the
deformation energy located at the crack tip, provided the expression of the fatigue threshold
depending on R (load ratio) [1, 2]:

i c 1
' K th ' K th  ' K th ( C 1 (1  R )  )V FL max S[
2 (2)
For materials verifying the law of Petch, the fatigue threshold can be written like:

' K th C 2  C 3[ 0 . 5 (3)
Expression formulated in experiments by the majority of the researchers, who consider
nevertheless the values thresholds for propagation velocities of about 10 -8 mm/cycle [3].
In this work, we will present a synthesis of these models which will be compared with our
model based on the shakedown analysis [4]. This analysis consider an elastic-plastic structure,
subjected to variable mode I loading P(t) and occupying a volume : with a surface * consisting of
disjoint parts *V and *u, where statical and kinematical conditions are prescribed, respectively.
The values of P(t) vary arbitrarily with time t, but remains between a prescribed loads Pmin and
Pmax. One then looks for the maximum value of the load factor D, such that the structure shake
down under the loads DP(t). This load factor called shakedown load factor DSD is determined as
solution of the following optimisation problem:
132 M. A. Belouchrani et al.

D SD max$ D
D ,U
(4)

U ij$ , j 0 in :
(5)

n j U ij$ 0 on *V
(6)

F (DV c ( P )  U $ , V s )  0 in :  P  >P min , Pmax @ (7)


m
ªm 1 1 D 1 º ( m 1)
a lim a0  « ³
U ij$ L ijkl U kl$ d : »  ac
¬ m K D 1 : 2 ¼
(8)

Here, F is a convex yield surface, V s the yield stress, a0 and ac are, respectively, the initial
crack length and the critical one for which unstable crack propagation occurs, when alim is the
limit crack length which can be attained when the shakedown state is reached V c (t ) is the time-
dependent stress state for a purely elastic comparison problem, differing from the original problem
only by the fact that the material reacts purely elastically with the same elastic moduli as for the
elastic part of the material law in the original problem and U $ time-independent state of residual
stress.
With the shakedown load factor D SD computed for a cracked structures loaded in mode I. We
can compute the fatigue threshold corresponding to the shakedown state 'K th , given by :

'K th D SD .P. Sa (9)

The computation of 'Kth with different grain size shows that for materials verifying the law of
Petch, the fatigue threshold can be written like [5]:

'K th (C4  C5[ 0.5 )V s


(10)
Where C1, C2, C3, C4, C5, m and K are material constants. The comparison of all these models
shows that the shakedown analysis constitutes an effective tool in the determination of the fatigue
thresholds. It has the advantage of determining them at the initiation of the crack propagation and
correctly translated the effect of the yield stress of the material by the influence of the residual
stresses and the grain size.
References
1. Wasen J., Hamberg K. and Karlsson B., Mater. Sci. Engng. A, 102, 217-226, 1988
2. Ravichandran K. S. and Dwarakadasa E. S., Acta Metall. Mater., vol. 39, N° 6, pp. 1343-1357, 1991
3. Nakai Y., Tanaka K. and Nakanishi T., Engng. Fract. Mech., vol. 15, N° 3-4, pp. 291-302, 1981.
4. Belouchrani M. A., Weichert D., Int. J. Mech. Sci., vol. 41, N° 2, pp. 163-177, 1998.
5. Belouchrani M. A., Weichert D. and Hachemi A., Mechanics. Resea. Comm., vol. 27, N° 3, pp. 287-293,
2000
2T3. Ductile fracture 133

VOID COALESCENCE IN METALS INVOLVING TWO POPULATIONS OF


CAVITIES

D. Fabregue and T. Pardoen


Unité de Matériaux et des Procédés (IMAP), Université Catholique de Louvain (UCL)
Place Sainte Barbe 2, B-1348 Louvain-la-Neuve, Belgium
fabregue@imap.ucl.ac.be, pardoen@imap.ucl.ac.be

Void coalescence is the final stage in the failure of ductile materials. It consists in the localization
of the plastic deformation in the intervoid ligament between neighbouring voids. Several
experimental evidences obtained from fractographic analyses of broken samples or metallographic
analyses of polished samples strained near fracture have shown that a second population of cavities
nucleated on small particles significantly affect the damage process controlled by the first
population of cavities nucleated on larger particles. Although a second population of voids is
considered for a while in the literature as very detrimental for the ductility (e.g. Marini et al. [1]),
only a limited number of studies have been devoted to the modelling of this phenomenon
(Tvergaard [2], Brocks et al. [3], Faleskog and Shih [4], Perrin and Leblond [5], Enakousta et al.
[6]). The aim of this work is to develop a constitutive model for the nucleation, growth and
coalescence of voids that incorporate the effect of this second population on the onset of
coalescence.
Firstly, FE unit cell simulations have been performed with a first population of cavities
explicitly modelled in the mesh while the second population is introduced through the use of the
Gurson response for the matrix surrounding the first population (Gurson [7]) such as in Brocks et
al. [3]. The underlying assumption of this approach is thus that the second population is made of
much smaller voids than the first population, with a volume fraction sufficient to smear out their
effect within the ligament between the big voids. Calculations have been carried out for different
volume fractions of the first population, different stress triaxialities, different volume fractions of
the second population and different nucleation conditions for the second population. Fig. 1 shows
that the onset of coalescence is about 0.76 for a first population of 1.5% without second population
whereas the coalescence occurs for a macroscopic strain of 0.46 if a second population is present
for the same overall amount of initial porosity. Cell calculations show that the presence of the
second population significantly cut down the strain at the onset of coalescence. The effect of the
nucleation strain is also exhibited in Fig. 1.
The cell calculations show that the presence of a second population essentially affects the onset
of coalescence and not the evolution of the first population before coalescence. The void
coalescence model of Thomason (Thomason [8]) has been modified to take into account the
presence of the second population of voids through its softening effect on the strength of the matrix
material. Cell calculations show that the onset of coalescence is dictated by the local softening
caused by the second population inside the ligament very close to the large void. This model
provides very accurate predictions of the onset coalescence when all the evolutions of the variables
(relative void spacing, void aspect ratio of first population and current yield stress near the large
void) are extracted from the FE cell calculations. As a second step, the coalescence model is
coupled to an enhanced Gurson model for the description of the first population (Pardoen and
Hutchinson [9]). The difficulty relies on approximating the local strains near the large voids to
solve the Gurson model locally in order to calculate the evolution of the second population of
voids introduced into the void coalescence conditions. This approach which assumes that the
second population mostly affect the void coalescence conditions must be contrasted with another
134 D. Fabregue and T. Pardoen

assumption which treat the second population as an extra contribution to growth rate of first
population by repetitive coalescence between the big and small cavities (Enakousta et al. [6]).

FIGURE 1. Influence of the presence of the second population in cell calculations (f1=fraction of
1st population, f2=fraction of 2nd population, ånucl=strain at nucleation of the 2nd population)

References
1. Marini, B., Mudry, F, Pineau, A., Engng. Fracture Mech., vol.22, 989-996, 1985
2. Tveergard V., Int. J. Solids Structures, vol.35, 3989-4000, 1998.
3. Brocks, W., Sun, D.Z., Hönig, A., International Journal of Plasticity, vol.11, 971-989, 1995.
4. Faleskog, J., Shih, F.C., Journal of the Mechanics and Physics of Solids, vol.45, 21-50, 1997.
5. Perrin, G., Leblond, J.B., International Journal of Plasticity, vol.6, 677-699, 1990.
6. Enakousta, K., Leblond, J.B., Audoly, B., In Proceedings of the 11th International
Conference on Fracture, CD-Rom, 2005.
7. Gurson, A.L., Journal of Engineering Materials and Technology, vol.99, 2-15, 1977.
8. Thomason, P., Ductile Fracture of Metals, Pergamon Press, Oxford, 1990.
9. Pardoen, T., Hutchinson, J.W., Journal of the Mechanics and Physics of Solids, vol.48, 2467-
2512, 2000.
2T3. Ductile fracture 135

EFFECTS OF CHARACTERISTIC MATERIAL LENGTHS


ON DUCTILE CRACK PROPAGATION

E. Radi
Dipartimento di Scienze e Metodi dell’Ingegneria, Università di Modena e Reggio Emilia
Viale Allegri, 13. I-42100 Reggio Emilia, Italy
eradi@unimore.it

Classical plasticity theories fail in characterizing the constitutive behavior of ductile materials at
the micron scale, which is necessary to define in order to investigate the stress and strain fields
near a propagating crack tip. Experimental observation on the macroscopic fracture toughness and
atomic work of separation of an interface between a ductile crystal of niobium and a sapphire
single crystal performed by Elssner et al. [1] found that the interface between the two materials
remained sharp and not blunted up to the atomic scale. Moreover, the stress level required to
produce atomic decohesion of the lattice turns out to be about 10 times the tensile yield stress,
whereas fracture mechanics analyses based on classical plasticity theories (Drugan et al., [2])
provide a maximum stress level near a crack tip not larger than 4–5 times the tensile yield stress.
Classical continuum theories are also unable to predict the size effect arising at small scales, due to
the lack of a length scale. Therefore, in order to describe the stress and deformation fields very near
the crack-tip during its propagation, it become necessary to adopt enhanced incremental
constitutive models, which account for the non linear behaviour of the material as well as for the
microstructure and the presence of dislocations, by incorporating one or more characteristic
lengths, typically of the order of few microns for ductile metals.
The couple stress (CS) flow theory of strain gradient plasticity has been introduced, in a first
attempt, by Fleck and Hutchinson [3]. Their model involves an intrinsic material length " and an
elastic length scale "e arbitrarily introduced in order to partition the deformation curvature rate
tensor into its elastic and plastic parts. The incremental model of CS plasticity developed by
Ristinmaa and Vecchi [4] and Ottosen et al. [5] is based on the Koiter [6] theory of couple stress
elasticity and introduces two distinct intrinsic material length, namely " and "c, whose effect on the
propagating crack tip field is almost unexplored.
Several analyses, carried out to investigate the effects of microstructure in fracture mechanics,
showed that the incorporation of couple stress and rotation gradients in the constitutive description
of ductile materials improves considerably the estimation of the stress traction level ahead of the
crack-tip. In particular, for the problem of mode I steady-state crack propagation, Wei and
Hutchinson [7], analysed the effects of stretch gradients (SG plasticity) in a numerical simulation
and found a sensible amplification of the traction level ahead of the crack-tip. It must be observed
that former investigation of crack propagation in linear hardening ductile materials described by
classical plasticity theories (Ponte Castañeda, [8]; Bigoni and Radi, [9, 10]) predicted an extremely
weak stress singularity for the typically small values of the strain hardening coefficient.
Recently, Radi [11] performed an asymptotic analysis of the same problem adopting the CS
flow theory of plasticity developed in [3] and found that the couple stress field is dominant near to
the crack-tip and produces a remarkable increase of the stress singularity, even for low hardening.
However, these analyses showed that the contribution of the elastic strain gradients strongly affects
the asymptotic crack-tip fields, through the elastic length scale "e. The further analysis performed
by Radi and Gei [12] indicated that the stress singularity increase at the crack-tip also for the Mode
III crack problem, due to the sole effects of rotation gradients provided by the CS theory of
plasticity, with no need to take stretch gradients into consideration.
136 E. Radi

In the proposed paper, the effects of strain rotation gradients on steady-state crack propagation
are investigated by performing an asymptotic analysis of the crack-tip fields derived from the flow
theory of CS plasticity with two characteristic material lengths. Rotation gradients are expected to
become significant at a distance from a crack-tip smaller with respect to these characteristic
lengths, and negligible at larger distances, with a gradual transition in the intermediate region.
According to the results obtained for a single characteristic material length " in [11, 12], the
couple stress and the skew-symmetric stress field are expected to dominate the asymptotic field
under Mode I and Mode III loadings conditions, respectively, and to produce an increase of the
stress singularity at the crack tip, also for a small hardening coefficient. The role of both
characteristic lengths " and "c will be examined in detail and their influence on the stress
singularity will be explored by numerical investigations of the asymptotic crack tip fields.
The performed asymptotic analysis will provide predictions on the level of traction ahead of
the propagating crack-tip more realistic then the classical solution obtained for the J2-flow theory
or for a single characteristic length, allowing the detailed mechanisms by which fracture may grow
and propagate in ductile metals to be understood in more depth, up to the micron scale.

References
1. Elssner, G., Korn, D. and Ruehle, M., Scripta Metall. Mater. vol. 31, 1037-1042, 1994.
2. Drugan, W.J., Rice, J.R. and Sham, T.L., J. Mech. Phys. Solids, vol. 30, 447-473, 1982.
3. Fleck, N.A. and Hutchinson, J.W., J. Mech. Phys. Solids, vol. 41, 1825-1857, 1993.
4. Ristinmaa, M., and Vecchi, M., Comp. Meth. Applied Mech. Engnrg., vol. 136, 205-224,
1996.
5. Ottosen, N.S., Ristinmaa, M. and Ljung, C., European J. Mech. - A/Solids, vol. 19, 929-947,
2000.
6. Koiter, W.T., Proc. Roy. Netherlands Acad. Sci. vol. B67, 17-44, 1964.
7. Wei, Y. and Hutchinson, J.W., J. Mech. Phys. Solids, vol. 45, 1253¯1273, 1997.
8. Ponte Castañeda, P., J. Mech. Phys. Solids, vol. 35, 227-268, 1987.
9. Bigoni, D. and Radi, E., Int. J. Solids Struct., vol. 30, 899-919, 1993.
10. Bigoni, D. and Radi, E., Int. J. Fracture, vol. 77, 77-93, 1996.
11. Radi, E., J. Mech. Phys. Solids, vol. 51, 543-573, 2003.
12. Radi, E., Gei, M., Int. J. Fracture, vol. 130, 765-785, 2004.
2T3. Ductile fracture 137

DUCTILE FRACTURE BY VOID NUCLEATION AT CARBIDES

J. Giovanola, D. Cannizzaro, R. Doglione1 and A. Rossoll


Ecole Polytechnique Fédérale de Lausanne
LCSM-IPR-STI, EPFL, Station 9, 1025 Lausanne, Switzerland
1Politecnico Torino

jacques.giovanola@epfl.ch

Predicting ductile fracture of initially undamaged structures still represents a challenge in most
applications such as for instance crashworthiness evaluations. This paper focuses on ductile
fracture by the process of void nucleation at tightly bonded and closely spaced inclusions in an
otherwise homogeneous ductile matrix. Particular attention is given to loading conditions
involving only low stress triaxialities (0 to 0.5).
As model material for the investigation, we selected a Vacuum Arc Remelted (VAR) steel:
DIN Standard 39NiCrMo4, AMS 6414, austenitized at 860 °C for 1 h, oil quenched, tempered at
704 °C for 1h. and air cooled. This heat treatment resulted in a material with a ferritic matrix (yield
strength of 700 MPa, ultimate strength 800 MPa, reduction of area 37.5%) and a dense distribution
of fine carbides (volume fraction of about 6% with a size distribution ranging from 0.1 to 1 Pm,
with an average size of 0.23 Pm).
We developed a simple torsion-tension test using a thin-walled tubular specimen (Fig. 1) that
allowed us to investigate void nucleation at carbides under low triaxiality loading (0 – 0.5, axial
load of 0, 1/3 and 2/3 of the limit load Plim). We were able to measure local strains and to interrupt
the test during the softening stage of deformation and thus gain access to the damage history in the
specimen.

FIGURE 1. (a) Thin-walled cylindrical specimen for low triaxiality tension-torsion tests. (b)
Torque and axial displacement versus twist angle curves.

By means of grid deformation and high speed photography techniques, we demonstrated that
in the investigated quenched and tempered VAR steel, two competing failure mechanisms lead to
fracture: shear localization, which dominates at very low triaxialities and micro-void nucleation
and coalescence, as soon as some significant tensile load is superimposed on shear deformation.
We modeled the observed void nucleation process by means of a critical interfacial stress
condition. We used a published interfacial stress model based on dislocation theory, the Brown
Stobbs [1] Kwon Asaro [2] model (BSKA) to calculate an average interfacial stress acting in a
138 J. Giovanola et al.

representative volume element from the continuum stress and strain fields applied to this element.
Using the BSKA model to estimate the critical interfacial stress from experimental data obtained in
low-triaxiality tension-torsion tests, as well as in notched tension tests with higher triaxialities, we
estimated a critical interfacial stress of 3000 MPa and were able to fit a single void nucleation
threshold curve over a range of triaxialities from 0 to 1.5 (Fig. 2)

FIGURE 2. Void nucleation threshold curve: fit of BSKA model to experimental data for a critical
interfacial stress value of 3000 MPa.

We provided a first qualitative validation of the predictive capabilities of the model by


combining stress and strain results of finite elements calculations with the BSKA model to
determine the location in various torsion and torsion-tension specimens, where voids nucleate first
i.e the location of the maximum interfacial stress, The results of these simulations indicate that, as
axial loading is increased, the site of first void nucleation moves from the outside edge toward the
center of the specimen section. These predictions were confirmed by detailed fractographic
observations of tested thin-walled cylindrical specimens.

References
1. Brown, L. M. and Stobbs, W. M. , Phil. Mag., 34, 351-372, 1976.
2. Kwon, D. and Asaro, R.J., Metallurgical Transaction A, 21A, 117-134, 1990.
2T3. Ductile fracture 139

THE SIGNIFICANCE OF MAXIMUM LOAD ON A LOAD-DISPLACEMENT


CURVE WITH STABLE CRACK EXTENSION

J. R. Donoso and J. D. Landes


Universidad Técnica Federico Santa María, Valparaíso, CHILE
University of Tennessee, Knoxville, TN, USA
juan.donoso@usm.cl, landes@utk.edu

The ASTM standard method for the measurement of fracture toughness, E 1820 [1], covers
procedures and guidelines for the determination of this material property in metallic materials
using the parameters K, J or CTOD. The fracture toughness may be measured as a point value, or
as a complete fracture toughness resistance curve. In the latter option a J- or CTOD-based
resistance curve may be obtained from a single specimen fracture test, in which the crack length is
measured from compliance changes, and later verified by optical measurements. The single
specimen J-R curve construction procedure defined in E1820 involves several steps, which go
from obtaining and plotting the raw J-'a data, to calculating an interim value of J, termed JQ, to
finally qualifying JQ as JIc, a size-independent value of fracture toughness.
Recently, Donoso, Zahr and Landes proposed an alternative way of obtaining the specimen J-R
curve using the Concise and Common (C&C) Formats [2]. The Concise Format [3] and the
Common Format [4] developed by Donoso and Landes are calibration functions that relate the
load, P, to the displacement, v, and the crack length, a, of a fracture toughness specimen. The
Concise Format stands for the elastic regime, in which v = vel [3], whereas the Common Format
deals with the plastic component of the displacement, vpl [4].
The Common Format Equation, CFE, was originally proposed by Donoso and Landes [4] as an
extension of the load separation concept [5] to describe the load-plastic displacement relationship
for a blunt-notch fracture specimen. As such, it relates the load P with two variables representing
the non-linear deformation of a fracture specimen with a stationary crack: vpl/W, the plastic
component of the load-line displacement, normalized by the specimen width W, and b/W, the
normalized ligament size (ligament b in lieu of the crack length a). The CFE also includes a term
that denotes the out-of-plane constraint, :*, and is written as:

P = :*˜B˜C˜W˜(b/W)m ˜V*˜(vpl/W) 1/n (1)


where B is the specimen thickness; C and m are the geometry function parameters, and V* and n
are material properties, obtained from a stress-strain curve.
In order to obtain the specimen J-R curve using the Concise and Common (C&C) Formats,
Donoso, Zahr and Landes [2] proposed a “crack growth law” to account for the relation between
stable crack extension, 'a, and vpl. The assumed crack growth law is a two-parameter power law
equation relating the change in crack length, 'a, with normalized plastic displacement, vpl/W, that
has the form

l1
'a § v pl ·
l 0 ¨¨ ¸
¸
W © W ¹ (2)
140 J. R. Donoso and J. D. Landes

The crack extension, 'a, may also be written in terms of the change in ligament size, that is, 'a =
bo – b, where bo is the initial ligament size (equal to W minus the initial crack length, ao). Thus,
Eq. (2) gives the following expression for the current ligament size, b:

l1
b b § v pl ·
o
 lo¨ ¸
W W
©W ¹ (3)
Substitution of Eqs. (2) and (3) into Eq (1) yield the following expression in terms of only the
plastic displacement when there is stable crack growth, with D being the product of the parameters
V* and :*:

m
§
¨ bo § v pl ·l 1 ·¸ § v pl
1/ n
·
P DBCW ¨  lo ¨ ¸ ¸ ¨
¨W
¸
¸
¨W ¨ W ¸ ¸ © ¹
© © ¹ ¹
(4)
The use of Eq. (4) not only has made it possible to generate Jpl as a first step to obtaining the
complete J-R curve [2], but also to predict maximum load on a load vs. displacement (P-v) curve
for a specimen that undergoes stable crack growth [6]. The purpose of this paper is to show how
the maximum load on a P-v curve obtained with a 1T-C(T) fracture specimen showing crack
extension relates to the value of J at initiation of ductile cracking, as defined by JQ following the
construction procedure of E1820. Following this, the potential use of the maximum load on a P-v
curve to determine directly a value of JQ will be explored and discussed.

References
1. E1820-99, Standard Test Method for the Measurement of Fracture Toughness, Annual Book
of Standards, Vol. 03.01.
2. Donoso, J.R., Zahr, J., and Landes, J.D., Second International ASTM/ESIS Symposium on
Fatigue and Fracture Mechanics, Tampa, FL, USA, November 2003.
3. Donoso, J.R. and Landes, Fatigue and Fracture Mechanics: 32nd Volume, ASTM 1406, R.
Chona. Ed., American Society for Testing and Materials, West Conshohocken, PA, 2001,
261-278.
4. Donoso, J.R. and Landes, J.D., Engineering Fracture Mechanics, vol. 47, No. 5, 619-628,
1994.
5. Ernst, H.A, Paris, P.C., and Landes, J.D., Fracture Mechanics; Thirteenth Conference, ASTM
STP 743, Richard Roberts, Ed., American Society for Testing and Materials, 1981, 476-502.
6. Donoso, J.R. and Landes, J.D., An Instability Analysis for a Crack Growth Situation Based on
the Common Format, ECF-15, Stockholm, 2004.
2T3. Ductile fracture 141

3D VISUALIZATION OF DUCTILE FRACTURE USING SYNCHROTRON X-


RAY COMPUTER TOMOGRAPHY

L. Qian, H. Toda, T. Ohgaki, K. Uesugi1, M. Kobayashi and T. Kobayashi


Department of Production Systems, Toyohashi University of Technology,
Toyohashi, Aichi, 441-8580, Japan
1Japan Synchrotron Radiation Research Institute, Sayo, Hyogo, 679-5198, Japan

qian@sp-mac4.pse.tut.ac.jp (L. Qian)

The initiation, growth and coalescence of voids during ductile fracture can be evidenced by means
of traditional methods such as optical microscope or SEM. However, the obtained information is
limited by the surface observations in that the surface stress-strain state differs from that of the
interior material. In addition, false images may arise from sample surface preparation. Recently,
synchrotron X-ray CT has been applied to investigate fatigue crack problems [1,2]. The purpose of
the present work is to use this advanced technique to visualize three-dimensionally the fracture
process in a practical ductile material, hence gaining a new insight into the ductile fracture.
Uni-axial tensile tests were conducted on a notched specimen made of Al-Si alloy using a test
rig specially designed for the X-ray tomography. The in-situ observations were performed during
tensile loading using a high resolution X-ray tomography at the X-ray imaging beamline BL47XU
of SPring-8. The reconstructed 3D images, with a resolution of 0.474Pm, of the damaged material
were quantitatively visualized.
It is found that at a smaller load, voids are nucleated in front of the notch. The voids occur
mainly within the eutectic region and are closely related to the eutectic particles. With increasing
load, the size and number of voids are increased, and some adjacent voids coalesce with each
other. Sectional images indicate that the number, size and shape of the voids differ so much from
slice to slice. Some slices demonstrate crack length longer than 60 Pm while other slices show only
a few small voids.
Fig. 1(a) and (b) presents the reconstructed 3D images after two loading stages, showing the
void distributions by removing other material components. On the whole, the high density of voids
is mainly concentrated near the notch, and increasing load leads to an increase in number and size
of the voids. Quantitative analyses indicate that the generated voids almost lie within a region of
about 150 Pm near the notch tip. Both the number and size become even larger with further
increasing the load. It is also indicated that more and larger voids are generated in the internal
region than on the side surfaces.
142 L. Qian et al.

FIGURE 1. Void distribution in front of notch after loading to (a) 51.6 N and (b) 55.0 N. Arrows
indicate crack growth direction.

In conclusion, the high resolution X-ray CT has been proved to be a feasible way to visualize
and quantify the ductile fracture, and provide much more information than traditional methods.
Voids are nucleated more easily in notch front than in remote regions. The distribution of voids is
not uniform in a real material, and the initiation, growth and coalescence of voids are not
simultaneous along the notch front. In addition, three-dimensional architecture of particles packing
in the eutectic region, and the 3D morphology of D-phases are also visible, which can be applied as
input data for numerical simulations of fracture behavior.
Finally, the synchrotron radiation experiments were performed at the BL47XU in the SPring-8
with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal No
2004B0457-NI-np).

References
1. Toda, H., Sinclair I., Buffiere J. –Y., Maire E., Connolley T., Joyce M., Khor K.H., Gregson
P., Philos. Mag., vol. 83, 2429-2448, 2003.
2. Toda, H., Sinclair I., Buffiere J. –Y., Maire E., Khor K.H., Gregson P., Kobayashi T., Acta
mater., vol. 52, 1305-1317, 2004.
2T3. Ductile fracture 143

NON-LOCAL PLASTIC-DAMAGE MODEL FOR FAILURE ANALYSIS OF


SHEET-METALS

M. Brunet, F. Morestin and H. Walter-Leberre


LaMCoS, Institut National des Sciences Appliquées,
20 Avenue A. Einstein, 69621 Villeurbanne, France.
Michel.brunet@insa-lyon.fr

A new damage model which takes account of void shape effect and anisotropy of the matrix
material is integrated into the dynamic explicit finite element framework to predict the damage
evolution which occurs under crash or stamping process. For the strain localization and failure, the
pathological mesh dependence has been overcome by a non-local approach where the evolution
equation for the porosity and the equivalent plastic strain is modified by an additional term
containing a characteristic internal length. The non-local plastic-damage potential is written as:

q2 § NV H ·
) C  2q1 f cosh¨ ¸  (1  q f 2 )
3 0
p
V y2 (H loc , ’ 2H p ) ¨ V (H p , ’ 2H p ) ¸
© y loc ¹ (1)
The damage model can take into account the three main phases of damage evolution: growth,
nucleation and coalescence. To determine the critical porosity f c , the void coalescence failure
mechanism by internal necking is considered by using a modified Thomason’s plastic limit-load
model on the reference volume element such as:

­ N M ½ ª
° § RZ · §R · ° V1 º V1
®F / ¨¨ ¸
¸  G /¨ X ¸ ¾ An «1  T »d
°̄ © X  RX ¹ © X ¹ °¿ ¬ « V y »¼ Vy
(2)
Consistent with (2), the plastic-damage potential (1) is used to calculate the void and matrix
geometry changes using the current strain, void volume fraction f and shape factor S. Once the
inequality (2) is satisfied, the void coalescence starts to occur and the void volume fraction at this
point is the critical value f c .

To evaluate the non-local variables (f , H p ) and their gradients at a Gauss point, the partial
differential equation of the implicit gradient formulation as:

p
’ 2H p ( x)  P 2H p ( x)  P 2H loc ( x)
(3)
is solved using the integral equation method at the beginning of each time step employing the
overlay mesh defined by the Gauss integration points of the underlying shell-element mesh
restricted to be in the same plane in order to solved a two-dimensional problem.
Once the damage model has been calibrated by an inverse identification technique using the
non-local FEM analysis on tensile tests, it can be used to predict the appearance of tearing in any
stamping process. In particular, the comparisons between experimental and numerical FLDs
calculated with the proposed non-local damage model is shown in Figure 1. The occurrence of
ductile rupture has been recognised following the satisfaction of the modified Thomason’s
coalescence criterion while the necking (load instability) has been determined when the calculated
punch force reaches a maximum. For this Ni-based alloy, the material parameters found are:
144 M. Brunet et al.

anisotropy: r0 = 1.20; r45 = 1.12; r90 = 0.85, isotropic hardening: V 0 547 Mpa; Qf 1891 Mpa;
b 1.84, void growth: f 0 0.001 ; S0 0.12 ; nucleation: A 0 0.08 , void coalescence: F=0.31;
G=1.23 and O0 1.15 .

Figure 1: Experimental and numerical FLD’s (INCO 718)

References
1. Pardoen, T., and Hutchinson, J.W.: An extended model for void growth and coalescence, J.
Mech. Physics of Solids, Vol. 48(2000), 2467-2485.
2. Brunet M., Morestin F. and Walter H.: Damage Identification for Anisotropic Sheet-Metals
Using a Non-local Damage Model, International Journal of Damage Mechanics, Vol.
13(2004), 35-57.
2T3. Ductile fracture 145

A NOVEL TECHNIQUE FOR EXTRACTING STRETCH ZONE FEATURES


FROM FRACTOGRAPHS

M. Tarafder, Swati Dey, S. Sivaprasad and S. Tarafder


National Metallurgical Laboratory, Jamshedpur-831007, India
mt@nmlindia.org, swati@nmlindia.org, shiva@nmlindia.org, star@nmlindia.org

The crack tip blunting process leaves an imprint in the form of stretch zone on fracture surfaces
during the event of ductile fracture. A schematic representation of the stretch zone in a fractured
specimen and a corresponding fractographs is shown in Fig. 1. A typical stretch zone has two
components, stretch zone width (SZW) and stretch zone depth (SZD). The SZW is basically the
virtual crack extension 'a and SZD is half of the CTOD, G. Stretch zones can be easily identified,
when observed under the SEM, since they have visually identifiable boundaries in between the
fatigue precracked region and the ductile fracture region

.
FIGURE 1. (a) Elevation profile of crack tip zone ( b ) SEM image of fracture surface
Previous authors [1-4] have estimated SZW and SZD and have attempted to establish their
correlation with the initiation fracture toughness of materials. A typical correlation of Ji with SZW
[1] is as following

Ji
SZW
2 k V f tan T

where Vf is the flow stress, 2T is the crack blunting angle and k is a material constant. Sivaprasad
et. al [2] have devised a correlation to compute SZD using

SZT  SZW cos D


SZD
sin D (2)
where SZT is the projection of the SZW after rotating the specimen at an angle Dand further used
SZD to obtain the fracture toughness of materials. They have estimated SZW and SZT by manual
estimation procedure, which is tedious, time consuming and lacks reproducibility and
repeatability. This paper reports an image analysis (IA) technique for on-line automatic
measurement of stretch zone on ductile fracture surfaces, and applies the technique on Cu-
strengthened HSLA steel for the determination of ductile fracture toughness.
146 M. Tarafder et al.

Fig. 2 shows a schematic diagram of the placement of fracture surface in the SEM at 0o and
tilted conditions with four key points identifying the locations of interest. Fig. 2 also includes a
typical image signal profile after processing. The schematic is an idealisation, which does not show
local uneven topographical features that are present in real specimens. By carrying out simple
addition and subtraction operations of the image signals, the spatial co-ordinates of the key points
P1, P2, P3 and P4 can be computed.

FIGURE 2. Schematic diagram of 0o and tilted positioning of fracture surface under SEM and the
identified key points in the image signals
Considering f0(x,y) and fT(x,y) as the post processed image signals of the fracture surface at 0o
and tilted conditions respectively and g0(x) and gT(x) are the corresponding vertical density
histograms, can be written as
y2 y2
g 0 ( x) ¦
y y1
f 0 ( x , y ), g T ( x ) ¦
y y1
f 45 ( x , y )
(3)
Assuming no translational shifts in x and z directions after the specimen is tilted, mathematically
the following conditions can be stated to identify the key points.

P 4 : Max ^A ( x , y ) g 0 ( x )  g T ( x )`
P 1 : ^S ( x , y ) g T ( x )  g 0 ( x ) ~ 0`
P 2 , P 3 : Min ^S ( x , y ) g T ( x )  g 0 ( x )` (4)
After identification of the stretch zone boundaries, linear profile in terms of pixels (between P1 and P2 for
SZW and P1 and P4 for SZT) are estimated using
SZW=F.Zand SZT= F.W
where F is the conversion factor from pixel to micron meter and Z and W is the width of the stretch zone in 0o
and tilted condition in terms of the number of pixels. It was found that this technique generates reliable data
which can be used for prediction of initiation fracture toughness of the HSLA material.
References
1. Bassim M.N., J. Material Processing Technology, vol 54, 109-112, 1995
2. Sivaprasad S., Tarafder S., Ranganath V.R., Das S.K. and Ray K.K., Met. Matallurgical and
Materials Transactions A, vol 33A, 2002.
3. Srinivas M., Malakondaiah G. and Rama Rao P., Acta Metall. Matter, vol 41:4, 1301-1312,
1993.
4. Rogerio de O. Hein L., Ammann J.J. and Nazar A.M.M., Material Characterization, vol 43,
21-30, 1999
2T3. Ductile fracture 147

SIMULATION OF FATIGUE CRACK GROWTH BY CRACK TIP BLUNTING

P. Hutar and M. Sauzay


Commissariat a l’Energie Atomique (CEA)
Bat.455, 91 191 Gif-sur-Yvette, cedex France
hutar@ipm.cz

The crack tip blunting and re-sharpening on the crack tip is one of the basic mechanisms for
fatigue crack growth in ductile metals and alloys, e.g. Laird [1], Neumann [2]. Previous numerical
studies of crack tip blunting made by Mc Meeking [3], Needleman and Tvergaard [4] have been
carried out for monotonic loading. Subsequently, this kind of analysis has been made for cyclic
loading (Gu and Ritchie, Tvergaard and Hutchinson [5]), for the load ratio R=0. It is shown that is
possible to obtain fatigue crack growth, which can be comparable with Paris law. Perfect plasticity
and a finite element mesh with a small initial radius of the crack tip were used. Because of strong
mesh distortion they were applied only three full cycles.

FIGURE 1. Initial mesh of finite elements used for the calculations

Recently, Tvergaard [6] made a numerical simulations of the fatigue crack growth for 200 full
cycles, using remeshing at several stages of the cyclic plastic deformation, for three different
values of the loading ratio.
This study is mainly focused on the possibility of modelling this phenomenon in the case of
long and physically short cracks. The numerical modelling is based on elastic-plastic finite element
analysis with the code CASTEM. The possibility of comparing experimentally obtained results of
the fatigue crack growth with the numerical simulation of crack blunting in the first few cycles is
studied.
Because of practical interest, many authors are interested in finding the ratio between CTOD
and fatigue crack growth e.g. Tomkins [7]. Tvergaard [5], [6] proposed a relationship between the
cyclic change of CTOD and the growth of fatigue crack in this form:

da
ȕ .' CTOD,
dN (1)
148 P. Hutar and M. Sauzay

where ȕ # 0.33 for the chosen initial radius at the crack tip B and for the material model of
perfect plasticity (non-hardening material). The initial radius depends on the applied loading and
the yield stress of the material. Our simulation accurately studies this kind of relation, and shows
the limits of this ratio.
Technical materials have more complex behaviour than can be described by a material model
of perfect plasticity. Therefore, in our computations different models of material hardening were
used and the effect of the hardening on the crack tip blunting was discussed.
For illustration initial mesh of finite elements used for analysis can be seen in Fig.1. The
displacement fields are approximated in terms of plane strain 8-noded isoparametric elements. A
plate with a small edge crack with initial radius B is used. The crack length is approximately ten
5
times smaller than other dimensions of the specimen. The ratio a/B 10 is used. Around the
crack tip was a very fine mesh of finite elements (the size of the smallest elements was
approximately a/ 2. 10 7 ).
The results presented can contribute to a better understanding of one of the basic mechanisms
controlling fatigue crack growth in the case of ductile materials.

References
1. Laird, C., In Fatigue Crack Propagation, ASTM STP 415, ASTM, Philadelphia, 1967, 131
2. Neumann, P., Acta Metallurg., vol. 17, 1219, 1969
3. McMeeking, R. M., J. Mech. Phys. Solids, vol. 25, 357-381, 2004
4. Needleman, A., Tvergaard, V., ASTM STP 803, Philadelphia, PA, 1983, 80-115
5. Tvergaard, V., Hutchinson, J.W., Fatigue 2002, vol. 1/5, EMAS, UK, 2002, 107-116
6. Tvergaard, V., J. Mech. Phys. Solids, vol. 52, 2149-2166, 2004
7. Tomkins, B., Philosophical Magazine, vol. 18, 1041-1066, 1968
2T3. Ductile fracture 149

LOADING RATE EFFECT ON DUCTILE FRACTURE

R. Chaouadi
SCK-CEN
Boeretang 200, 2400 Mol, Belgium
rchaouad@sckcen.be

The Charpy impact test is widely used to monitor the quality requirements of industrial processes.
It was also adopted by engineers and scientists to monitor material embrittlement resulting from
environmental effects like for example irradiation. In this work, we investigated the effect of
loading rate, namely quasi-static versus impact loading, on the ductile fracture behavior. Two low
alloyed steels used in the reactor pressure vessel industry were selected, namely A533B and
20MnMoNi55. These steels were extensively characterized from the flow, Charpy impact and
fracture toughness properties [1-2]. Figure 1 shows how the loading rate affects the ductile to
brittle transition curve for both materials. As can be seen, the major effect of loading rate is located
in the fully ductile regime where quasi-static loading requires significantly less energy to full
fracture than dynamic (impact) loading. Two temperatures, namely 25°C and 290°C were selected
to investigate these loading rate effects. At both temperatures, tensile and crack resistance
measurements were performed at both quasi-static and dynamic loading. For the A533B steel, at
25°C, the fracture is not fully ductile and therefore only tests at 290°C were considered for this
material. The results are shown in Table 1 for the various materials and conditions. These result
clearly show that empirical correlations [3] relating fracture toughness to Charpy impact energy
are not applicable without an in-depth analysis. Moreover, the loading rate effects on the crack
resistance cannot be solely attributed to the strain rate sensitivity of the material.

TABLE 1. Ductile crack initiation toughness.


150 R. Chaouadi

FIGURE 1. Effect of loading rate on the ductile-to-brittle Charpy transition curve.

References
1. Chaouadi, R. and Fabry, A., In From Charpy to Present Impact Testing, edited by D. François
and A. Pineau, Elsevier, 2002, 103-117.
2. Chaouadi, R., J. Test. and Eval., Vol. 32, No. 6, 469-475, 2004.
3. Rolfe S.T. and Barsom, J.M., Fracture and Fatigue Control in Structure – Application of
Fracture Mechanics, Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1977.
2T3. Ductile fracture 151

EXPERIMENTAL INVESTIGATION OF SLANT CRACK PROPAGATION IN


X100 PIPELINE STEEL

S. H. Hashemi, I. C. Howard1, J. R. Yates1, and R. M. Andrews2


The University of Birjand, Department of Mechanical Engineering, Birjand, IRAN
1The University of Sheffield, Department of Mechanical Engineering, Sheffield, UK
2Advantica, Loughborough, UK

shhashemi@birjand.ac.ir

Failure information from full-thickness burst experiments on long distance gas transportation
pipelines has shown that unstable fractures propagating in the pipeline axial direction are
dominated by ductile slant shearing [1-3]. Different test samples, e.g. Charpy impact [4], drop
weight tear test (DWTT) [5] and double cantilever beam (DCB) [6] have been proposed to study
the ductile shear crack growth in pipeline steels in a laboratory scale experiment. Because of the
design geometry of these specimens, slant crack growth is often preceded by flat fracture in the
specimen un-cracked ligament.

In this research the slant fracture characteristics of 52cc O.D x 21mm W.T pipeline made
from high-toughness steel of grade X100 was investigated using modified compact tension C(T)
$
specimens. Two sets of test samples were used in this study; smooth and novel side-grooved 45
slant notch C(T) specimens, see Fig. 1. The latter restraint specimen was used to maintain the
propagating crack in its original 45$ plane.

FIGURE 1. Design geometry of slant C(T) samples: a) smooth, and, b) side-grooved specimen

All test specimens were loaded in tensile opening mode I. In each test a crack initiated from an
initial 45$ through-the-thickness machined slit, and propagated in slant mode fracture through the
specimen ligament. This allowed simulation of the slant crack growth in the laboratory scale close
to the real structure.
The test data monitored in each experiment was load and crack mouth opening displacement
(CMOD). Crack length measurements were conducted using two methods; the current direct
potential drop (CDPD), and an optical technique. In the CDPD test the input and output leads were
located at the positions such that the Johnson's formula [7] for the relationship between crack
length and electric potential was applicable. In the latter a fine square mesh was scored on the side
surface of the test sample. The crack length measurements then were monitored using high-
resolution digital video camera. Fig. 2 is a photograph of the experimental set up
152 S. H. Hashemi et al.

.
FIGURE 2. Experimental set up used in slant C(T) testing of X100 pipeline steel

From the test results it appeared that the PD method (and Johnson formula, initially developed
for crack growth measurement in standard C(T) and 3PB specimens with straight cracks) could be
used for slant crack growth estimation. This was validated by comparing the final crack length of
the broken specimens in liquid nitrogen with those predicted by the Johnson formula, and with
those monitored by the video imaging system used.
The load-displacement plots of all test specimens showed substantial load drop after small
amounts of ductile slant crack propagation. The optical observation of broken test samples
revealed that the fracture surface of both smooth and side-grooved slant C(T) specimens contained
traces of quasi-cleavage fracture even though the test temperature was above the material transition
temperature. The sharp load drop was associated with this mode of fracture in the specimens
tested. From the test data the specific slant fracture energy (in terms of J / mm2 ) was estimated for
the X100 steel tested, and shown to be similar to that measured on DCB like specimen for the same
class of steel.

References
1. Rothwell, A. B., In Proceedings of Pipeline Technology, Edited by R. Denys, Elsevier
Science, 2000, 387-405.
2. Demofonti, G., Mannucci, G., Spinelli, C. M., Barsanti, L. and Hillenbrand, H. G., In
Proceedings of Pipeline Technology, Edited by R. Denys, Elsevier Science, 2000, 509-520.
3. Berardo, G. , Salvini, P., Mannucci, G. and Demofonti, G. , In Proceedings of the 2000
International Pipeline Conference, Vol. 1. New York, ASME, 2000, 287-294.
4. Leis, B. N., Eiber, R. J., Carlson, L. and Gilroy-Scott, A., In Proceedings of the International
Pipeline Conference, Vol. II, ASME, 1998, 723-731.
5. Wilkowski, G. M., Rudland, D. L., Wang, Y. Y., Horsley, D., Glover, A. and Rothwell, B., In
Proceeding of IPC’02 4th International Pipeline Conference, Alberta, Canada, 2002, 1-7.
6. Shterenlikht, A., Hashemi, S. H., Howard, I. C., Yates, J. R. and Andrews, R. M., Engineering
Fracture Mechanics, Vol. 71, 1997-2013, 2004.
7. Johnson, H. H., Materials Research & Standards, Vol. 5, 442-445, 1965.
2T4. Nonlinear fracture mechanics 153

ESIS TC8 – NUMERICAL ROUND ROBIN ON MICRO MECHANICAL


MODELS : RESULTS OF PHASE III FOR THE SIMULATION OF THE
BRITTLE TO DUCTILE TRANSITION CURVE

C. Poussard and C. Sainte Catherine


Commissariat à l’Energie Atomique, DEN/DMN/SEMI
CEA Saclay, Bat. 625P, F-91191 Gif Sur Yevtte Cedex, France
Christophe.poussard@cea.fr, Claude.sainte-catherine@cea.fr

This paper presents the results of a round robin organized by CEA Saclay within the ESIS
(European Structural Integrity Society), TC8 (Technical Committee n°8) dedicated to the
numerical analysis and the comparison with experimental data. The objective of this round robin
(phase III) was to model the ductile to brittle transition curve of the 22NiMoCr37 German RPV
(Reactor Pressure Vessel Steel) using elastoplastic damage models. The principle of the round
robin is that initial sets of damage parameters are first imposed. Then, contributors are free to
adjust and justify their parameters so that a good latch to experimental data may be obtained.
A good participation was received for this round robin given the substantial technical efforts
that were required to contribute. Ten laboratories including nine disseminated in Europe as well as
one laboratory in Asia have contributed. Seven finite element codes have been used and three
participants performed full 3D fracture mechanics analyses accounting for ductile crack growth.
Six participants contributed to the three steps that were proposed in the specifications.
Differences have been obtained between the contributions for which explanations have been
suggested. The first step was dedicated to the prediction of a 1TCT specimen at -20°C. For that
step, no reference was made to experimental results and the purpose was to compare the codes with
each other. The mesh, boundary conditions and critical damage parameters were imposed to the
participants and resulted from previous phases of the round robin. Differences have been obtained
which have been primarily explained by the fact that participants used different elements, either 8
nodes reduced integration elements or 4 nodes fully integrated elements. For a given element size,
this appears to play a major role when predicting ductile crack growth. For the prediction of
cleavage fracture, differences were obtained for the plastic volume, due probably to the definition
of the threshold used to define the plastic strain but this had no influence when Weibull stresses
were compared. The differences were smoothed in step two when comparing the predictions to
experimental results and when adjusting critical damage parameters was made. Most contributors
agreed to the fact that the initial ductile damage parameters that were supplied to the participants
did not allow to predict the ductile crack growth behavior observed in the tests. Larger elements at
the crack tip and along the crack growth path and a smaller initial void volume fraction were
identified from the computation of 1TCT specimens. The mandatory Beremin parameters used
also in preceding phases of the round robin did not lead to satisfactory predictions of the failure
probabilities at higher temperatures (-40 and -20°C). The correlation was slightly improved when
accounting for a well identified ductile damage model but the results clearly showed that the
critical cleavage stress had to be diminished in order to satisfactorily predict the experimental data.
Finally, it was found that these models could lead to very good prediction of the transition
curve. The correlation was found initially poor when using the initially imposed mandatory ductile
and cleavage damage parameters but when identified parameters were used, a good correlation
could be obtained within the temperature range covered by the test matrix. The agreement between
the computation and the observation was further improved when using a linear dependant critical
cleavage stress with temperature. Figure 1 is an example of the quality of the results that can be
154 C. Poussard and C. Sainte Catherine

obtained with the local approach mechanical models over the transition. Between -150°C and -
20°C, a perfect match to the experimental data is obtained. At 0°C, the prediction is found to be
over conservative since in this example, only a 2D computation was made whilst a full 3D model
was required to model correctly ductile crack growth prior cleavage.

FIGURE 1. Prediction of the transition assuming a linearly dependant critical cleavage stress.

The results will now be used in order to support an ESIS guideline document entitled Guidance on
local approach of rupture of metallic materials, document that describes the state of the art to
apply the local approach to crack components.

Acknowledgements
The authors are very grateful to the round robin participants for the high quality level of the work
they have produced for that round robin and the useful discussions that arose when comparing the
results. Their patience for obtaining this synthesis is also gratefully acknowledged. The authors
also whish to express their gratitude to the GKSS staff in Hamburg and in particular to Wolfgang
Brocks, Juergen Heerens, Karl Schwalbe and Ingo Scheider not only for providing the data used in
that round robin but also for their continuous and fruitful collaboration within ESIS TC1 and TC8
over the past decade.
2T4. Nonlinear fracture mechanics 155

CLOSURE OF A RECTANGULAR SKIN DEFECT VIA THE ADVANCEMENT


FLAP

C. Antypas, C. Borboudaki, V. Kefalas and D. A. Eftaxiopoulos


Department of Mechanics, School of Applied Mathematical and Physical Sciences,
National Technical University of Athens
Theocaris Building, Zografou Campus, 15773 Athens, Greece
eftaxiop@central.ntua.gr

The article is concerned with the closure of a rectangular skin defect, via the enlongation of the
advancement local skin flap. Local skin flaps are skin islands that are stretched, rotated or
transposed in order to cover a skin defect, which was created due to the removal of a diseased or
injured skin portion. The paper consists of two parts, namely the experimental part and the
computational part. In the experimental part, in – vitro tension experiments conducted on piglet
skin strips are described. The experimental results yielded the nonlinear stress – strain curves of
skin. In the computational part, a stress – strain curve obtained in the experimental part, is used in
a finite element model for the closure of a rectangular skin defect via the advancement flap. The
skin, the subcutaneous fatty tissue and the underlying muscle tissue are included in the model. The
muscle and the subcutaneous tissue mechanical responses, are obtained from the literature.
Nonlinear finite element analysis with large deformations is performed in the model. Several stress
distributions, within the skin layer, are obtained. Finite element results indicate that stress
concentrations at the points where the stitches are done, are created. On the rest of the area of the
advancement flap the stress distribution is almost uniform.
Borboudaki [1] did in vitro tension experiments on piglet skin strips and the nominal stress –
engineering strain curves for the skin were obtained. Antypas [2] did a nonlinear finite element
analysis of the closure of a rectangular skin defect, via the advancement skin flap. The skin, the
subcutaneous tissue and the underlying muscle were considered as separate materials. The article
is based on the work performed in [3] and [4].
For the tension experiments on skin, a male, two month old Duroc piglet, weighing 15 kg, was
sacrificed. Immediately after the animal sacrifice, a large single skin portion was removed from the
abdominal and the thoracic area of the animal. This skin island, was kept in Ringer’s solution. The
fatty tissue was removed from the skin island by using a surgical knife. Then four pairs of
dumbbell specimens, i.e. eight specimens overall, were cut from the skin island by using a
dumbbell cutter D412 of the ASTM standard. The specimens were kept in Ringer’s solution. Each
pair consisted of two strips, taken from symmetric positions, left and right from the two columns of
nipples, located in the chest and the abdomen of the animal. Two pairs consisted of longitudinal
specimens (long dimension along the spinal cord direction) and the other two pairs consisted of
transverse specimens (long dimension perpendicular to the spinal cord direction).
Eight simple tension experiments were performed, one for each specimen. The experiments
were performed by using an INSTRON 1140 machine. Prior to each experiment, the specimen was
placed flat on the working bench, and was regularly rinsed with Ringer’s solution. Two black
marker lines were ascribed on each specimen, perpendicular to the longitudinal axis of the
specimen. The lines were ascribed by using a drawing ink marker with 0.1mm thick tip and were
placed well inside the neck of the dumbbell specimens. The length between the two markers was
measured. Afterwards, eight rectangular pieces of emery cloth were glued on the ‘bell’ area of the
specimen. The moving grip of the INSTRON 1140 machine was removed and the specimen was
attached to the grip. The grip together with the specimen was attached to the machine and the other
156 C. Antypas et al.

end of the specimen was attached to the other grip. Then the upper grip was displaced such that the
length between the marker lines, became equal to the length between them as measured on the
working bench. Afterwards, the tension experiment was executed, with a grip displacement
velocity of 10mm/s. The experiment was terminated when the specimen slipped out of the grips. A
digital video recorder captured the displacement of the marker lines during the experiment and a
video editor was used for breaking down the movie into individual frames. By using the number of
pixels between the marker lines, the strain was measured at specific times. The strain was
corresponded to the tensile force, recorded by the testing machine at specific times. In turn the
force was converted to nominal stress via a division by the undeformed cross sectional area of the
specimen. For one pair of longitudinal and one pair of transverse specimens, the stress – strain
curves were quite close to each other. For the other two pairs of longitudinal and transverse
specimens, the stress – strain curves were not close to each other.
In the subsequent finite element analysis, a simplified flat three – dimensional model of the
advancement skin flap was constructed, by using the commercial package ANSYS 8.0. The skin
layer was meshed with shell elements, the subcutaneous tissue was meshed with one dimensional
nonlinear truss elements and the underlying muscle tissue was meshed with pentahedral solid
elements. The Multilinear Elastic (MELAS) material model of ANSYS was used, where points of
the experimental nonlinear stress - strain curve were introduced into the finite element model. Via
the MELAS option, only isotropic materials could be modeled. Large displacements were also
incorporated in the model. Appropriate displacement constraint boundary conditions were applied,
such that the distance between selected nodes on the edge of the skin flap and on the surrounding
skin edge, became equal to zero, i.e. these nodes were merged. In this way the stitching process for
the closure of a rectangular skin defect during plastic surgery, was simulated. In an area close to
the edges of the flap and the surrounding skin, the link elements were removed, for the
undermining of skin during plastic surgery to be simulated. Computational results indicated that
stress concentrations, at the points where the stitches were placed, were created. On the rest of the
area of the advancement flap the stress distribution was almost uniform.

References
1. Borboudaki C., Experimental and computational study of a skin strip under tension, Diploma
Thesis, School of Civil Engineering, National Technical University of Athens, 2004.
2. Antypas C., A computational study of skin as a composite material – The advancement flap,
Postgraduate Thesis, Interscholar Programme of Postgraduate Studies on ‘Computational
Mechanics’, National Technical University of Athens, 2005.
2T4. Nonlinear fracture mechanics 157

SIMILARITY SOLUTIONS OF CREEP – DAMAGE COUPLED PROBLEMS IN


FRACTURE MECHANICS

L. V. Stepanova and M. E. Fedina


Department of Continuum Mechanics, Samara State University
Akad. Pavlov str., 1, 443011, Samara, Russia
lst@ssu.samara.ru

The asymptotic solution to Mode III and Mode I crack problems in a creeping solid in the
framework of Continuum Damage Mechanics is presented. The kinetic law of damage evolution is
the Kachanov – Rabotnov equation [1]. The damage parameter is incorporated into the power-law
creep constitutive equations. Thus the coupled system of damage mechanics – creep theory
equations is considered. It is to be expected during such a coupling process that the damage
involved gives a great effect on the stress field near the crack tip and when the damage parameter
reaches its critical value a totally damaged zone near the crack tip may occur [2-4].
The present contribution is an attempt to obtain asymptotic fields of stress, creep strain rate and
continuity near Mode III and Mode I cracks in creeping damaged materials as functions of the
similarity variable initially proposed by Riedel [5] under the assumption that the totally damaged
zone in the vicinity of the crack tip does really exist. The shape and the characteristic length of the
totally damaged zone are not a priori and should be obtained as a part of the solution. The
instantaneous response of the materials characterised by the creep power-law constitutive
equations and the power-law damage evolution equation is non-linear viscous and the evolution of
damage at short times after load application can be analysed under the remote boundary conditions
that the stress field must approach the Hutchinson – Rice – Rosengren field of non-linear fracture
mechanics. If the HRR-field represents the initial conditions and the remote boundary conditions
then dimensional analysis shows that the damage mechanics equations must have similarity
solutions [5]. Since the classical continuum mechanics equations can not be valid inside the totally
damaged zone the asymptotic solution to the problem is sought at large distances from the crack tip
(at large distances as compared with the characteristic length of the totally damaged zone but at yet
still small distances as compared with the crack length, with the characteristic length of the cracked
body).
The asymptotic expansions of the continuity (integrity) parameter \ and the effective stress
Vij /\ (the stress referred to the surface that really transmits the internal forces) for large
distances R o f from the crack tip, where R is the similarity variable, have the following form

\ ( R , M ) 1  R J g ( 0 ) (M )  R J g (1) (M )  ... ( J , J 1  0 ),
1

V ij
( R , M ) R s f ij( 0 ) (M )  R s f ij(1) (M )  ... ( s , s1  0 ).
1

\ (1)
The two-term asymptotic expansion for the effective stress and the three-term asymptotic
expansion for the continuity (integrity) parameter are obtained. The construction technique of the
far-field stress asymptotics is elucidated. The asymptotic fields allow to find the geometry of the
totally damaged zone. The geometry of the totally damaged zone for different values of the
material constants is given and analysed (Figures 1 and 2). It is found that the HRR-solution for the
creep constitutive equations can not be used as the remote boundary conditions and the actual
stress field at infinity is found.
158 L. V. Stepanova and M. E. Fedina

FIGURE 1. The geometry of the totally damaged zone in the crack tip region of a Mode I crack
under plane strain conditions: 1 – the configuration given by the two-term asymptotic expansion of
the continuity parameter, 2 – the configuration given by the three-term asymptotic expansion of the
continuity parameter.

FIGURE 2. The geometry of the totally damaged zone in the crack tip region of a Mode I crack
under plane stress conditions.

References
1. Kachanov L.M., Introduction to Continuum Damage Mechanics, Martinus Nijhoff,
Dordrecht, Boston, 1986.
2. Zhao J., Zhang X., Engn. Fracture Mechanics, vol. 50, 131-141, 1995.
3. Zhao J., Zhang X., Int. J. of Fracture, vol. 108, 383-395, 2001.
4. Astafjev V.I., Grigirova T.V., Pastuchov V.A., In Mechanics of Creep and Brittle Materials,
edited by A.C.F. Cocks, A.R.S. Ponter, Elsevier, London. 1991, 49 -61.
5. Riedel H., Fracture at High Temperature, Springer, Berlin, 1987.
2T4. Nonlinear fracture mechanics 159

IMPACT FRACTURE TOUGHNESS DETERMINATION OF DUCTILE


POLYMERS BY SPB METHOD

J. Wainstein , L. A. Fasce and P. M. Frontini


Instituto de Investigaciones en Ciencia y Tecnología de Materiales –INTEMA-,
Univ. Nac. de Mar del Plata - CONICET;
Juan B. Justo 4302, B7608FDQ, Mar del Plata, Argentina.
pmfronti@fi.mdp.edu.ar

Fracture toughness of ductile materials is often characterized by the J parameter that was
developed from the J-integral concept. The determination of a critical value of the J-integral is
generally performed through the construction of the resistance curve J-'a of the material. The
commonly used method for this purpose is the multiple specimen technique [1] in which several
specimens are loaded to obtain different amounts of crack growth. However, this method is very
difficult to apply at high rate conditions because of the need of interrupting the test at different
crack growth levels. Recently, a new extremely simple single specimen method has been
developed from the separation parameter, Spb, and successfully applied in fracture toughness
characterization of metals and ductile polymers [1,2]. It consists on the assumption that the load
can be separated into two multiplicative functions: the geometry (G) and the deformation (H)
functions. The separation parameter, Spb, is defined as the load ratio of a sharp and a blunt notched
specimen of the same material, geometry and constraint. Assuming that the load falls in
correspondence with stable crack propagation, the load drop after maximum load can be taken as a
symptom of crack growth. From the Spb expression (Eq. 1) and counting with at least two
calibration points (in order to assess the m parameter in Eq. 1), a simple relationship between the
load and the crack growth length can be simply obtained and then the J-'a curve can be evaluated.
m
§ ap · § ap ·
G p ¨¨ ¸¸ ¨¨ ¸¸
P a ,Q
m
p p ©W ¹ ©W ¹ § ap ·
S pb ¨¨ ¸¸
P a ,Q §a · § ab ·
m
© ab ¹
b b Q Gb ¨ b ¸ ¨ ¸ Q
©W ¹ ©W ¹
Q (1)
The Spb method may appear very similar to the Normalization method; though, Spb method has the
appealing advantage of requiring the assumption of only one hypothesis, i.e. a geometry function,
which indeed is well known for several specimen configurations. In the proposed Spb method, the
load separation principle is applied to total displacement with the advantage of non-subtracting the
elastic component of the displacement, which results difficult to evaluate from the load-
displacement impact records.
In this work the capability of the Spb method in determining the impact fracture toughness of
ductile polymers is studied. The methodology is assessed for two commercial grade polymers that
exhibit ductile impact fracture behavior: acrylonitrile-butadiene-styrene terpolymer ABS (Lustran
ABS-640 HR 850) and polypropylene block copolymer PPBC (Tipplen CS 2-8000). Stes
containing two records: one of a sharp-notched specimen and one of a blunt notched specimen are
employed in the application of the Spb method. Fracture tests are performed at 1m/s in a Fractovis
Ceast Falling Weight type machine and they are conducted up to the point of complete failure of
the specimen so that the final crack length is not available as a calibration point. In order to
determine the “m” parameter in Eq. 1, two points of the load displacement record for which crack
lengths are known are used. The first one corresponds to the initial crack length where the Spb
160 J. Wainstein et al.

parameter is constant (solid arrow in Fig. 1-b and c); the initial crack length can be measured on
the fracture surface of the broken specimen. The second point is taken under the assumption that
when the sharp specimen achieves the same crack length as the blunt notched specimen, both are
bearing the same load and the Spb parameter is equal to one [3] (dash arrow in Fig. 1-a and c). The
steps of the proposed Spb method are illustrated in Fig1.

FIGURE 1. Steps in Spb method for PPBC at 1m/s. a) Load displacement records of sharp and
blunt notched samples; b) Spb parameter; c) Estimated crack length.

J-R curves arisen from the application of the Spb method for triplicate set of samples are
shown in Figure 2. Plane strain fracture toughness values (JIc) derived from the fitted average
curves are 7.5 N/mm and 5.4 N/mm for ABS and PPBC, respectively. These fracture parameters
will be compared with the fracture toughness parameters evaluated by the essential work of
fracture approach (EWF)

.
FIGURE 2. J-R curves determined by the Spb method for a) ABS and b) PPBC.

The appealing of the Spb method is basically its simplicity. It seems to be applicable for many
polymers and it will result helpful for that cases in which the multiple specimen technique is hard
to performed such as at high loading rates, high temperatures or aggressive environments.

References
1. Wainstein J., de Vedia L., Cassanelli A.. Eng. Fracture Mechanics, 70, 2489-2496, 2003.
2. Wainstein J., Frontini P., Cassanelli A., Polymer Testing, 23, 591-598, 2004.
3. Kobayashi T., Yamamoto I., Niinomi M., JTEVA, 21, 145-153, 1993.
2T4. Nonlinear fracture mechanics 161

A MICRO-TOUGHNESS MODEL FOR DUCTILE FRACTURE

K. Srinivasan, Thomas Siegmund and Otmar Kolednik1


School of Mechanical Engineering
Purdue University, West Lafayette, IN 47907, U.S.A.
1Erich Schmid Institute of Materials Science, Austrian Academy of Sciences, A-8700 Leoben,

Austria
ksriniv@ecn.purdue.edu, siegmund@ecn.purdue.edu, kolednik@unileoben.ac.at

Ductile fracture occurs through growth and coalescence of micro-voids that originate at the
location of inclusions and precipitates. The plastic work dissipated in these micro-separation
processes leading to the creation of a unit fracture surface area is a measure of the micro-toughness
of the material. Furthermore, void growth and coalescence processes are accompanied by plastic
deformation of the material surrounding the voids. The energy dissipated by the micro-separation
processes of void growth and coalescence, and the plastic deformation in the bulk material
surrounding the voids together contribute to the overall fracture toughness of the material.
Conventional fracture toughness tests fail to individually measure these two very different
contributions. As a result, there is limited transferability of fracture toughness test data from the
laboratory to an actual structure.
To overcome this problem there is a need for measurements that allow for the determination of
the micro-toughness in ductile fracture. The transferability of material toughness and crack growth
resistance can then be solved with a non-linear analysis model like a cohesive zone model (CZM).
If the micro-toughness can be determined experimentally then only the cohesive strength remains a
free parameter, and the CZM can then be easily connected to experiments. Furthermore,
measurements of micro-toughness are of fundamental interest in the development of materials with
improved crack growth resistance.
A micro-toughness model using measurements of the ductile fracture surface topology to
estimate the plastic work dissipated in the process of the formation of a dimple fracture surface was
proposed in [1]. This model provides the micro-toughness, * , (also called the cohesive energy) as
* SV H where V is a flow stress, H is the dimple height, and S is a shape factor [1]. The
dimple height can be obtained by topographic measurements performed by taking stereo image
pairs in a scanning electron microscope and analyzing them using a digital image analysis system
[2]. Detailed investigations of void growth processes reveal that the model of [1] underpredicts the
micro-toughness in ductile fracture. This issue is overcome by the present work by developing a
new micro-toughness model.
The micro-toughness involves the use of accurate expressions for the effective plastic strain,
H , fields and a numerical integration procedure to calculate the micro-toughness. The model
incorporates the initial spherical expansion of the void as well as the subsequent lateral void
expansion stage. A criterion for the transition from the spherical void growth to a lateral void
expansion is incorporated [3], and a criterion on final void link-up based on a critical inter-
ligament plastic strain, H , is used. The micro-toughness is then given from the following
parameters

* * H , ri , V Y , n, H c
(1)
162 K. Srinivasan et al.

where ri is the inclusion size, V Y yield strength, n hardening. Figure 1(a) depicts predicted
values of the micro-toughness in dependence of the dimple height. Good agreement between the
model and corresponding FEM simulations are obtained for the individual stages of void growth,
* 1 and * 2 as well as for the total micro-toughness. Practical examples of the application of the
micro-toughness model are discussed for two high strength steels (St37 and V720), as well as an
aluminium metal matrix composite [4].
The micro-toughness model is also applied to study the effects of dimple size on micro-
toughness. For dimple sizes approaching the material internal length scale, l, of plasticity (set
mainly the dislocation spacing) micro-toughness is calculated through a strain-gradient plasticity
model [5] in combination with the micro-toughness model. The material flow stress is then given
as V f (H )  lK with f (H ) the hardening function, and K the strain gradient. Figure 1(b)
demonstrates that for small dimples (dimple radius R0 < 10 Pm) such strain gradient effects need
indeed be accounted for in order to accurately predict the micro-toughness. The implications of
the finding to inter-granular ductile fracture are discussed.

(a) (b)
Figure 1: (a) Micro-toughness in dependence of dimple size. (b) Dimple size dependence of
micro-toughness, comparison of conventional plasticity and strain gradient plasticity.

References
1. Stüwe, H.P., Eng. Fract. Mech., vol. 13, 231-236, 1980.
2. Stampfl, J., Scherer, S., Gruber, M. and Kolednik, O., Appl. Phys., vol. A63, 341-346, 1996.
3. Pardoen, T. and Hutchinson, J.W., J. Mech. Phys. Solids, vol. 48, 2467-2512, 2000.
4. Miserez, A., Rossoll, A. and Mortensen, A., Acta Mater. vol. 52, 1337-51, 2004.
5. Huang, Y., Qu, S., Hwang, K.C., Li, M., Gao, H., Int. J. Plast., vol. 20, 753-782, 2004.
2T5. Fatigue and fracture 163

CRACK COALESCENCE MODELLING OF FSW JOINTS

A. Ali, M. W. Brown1 and Chris A. Rodopoulos2


Department of Mechanical and Manufacturing
Engineering Faculty, Putra University Malaysia, 43400, Serdang Selangor , Malaysia
1Department of Mechanical Engineering

University of Sheffield , Sir Frederick Mappin Building, Mappin Street


Sheffield S1 3JD,United Kingdom
2Structural Materials and Integrity Research Centre,

Materials and Engineering Research Institute, Sheffield Hallam University, City Campus, Howard
Street, United Kingdom
aidy@eng.upm.edu.my, M.Brown@sheffield.ac.uk, C.Rodopoulos@shu.ac.uk

In the present work, Friction Stir Welds (FSW) of 2024-T351 aluminium alloys is characterised in
terms of macrostructure, microstructure, hardness, precipitate distribution, and weld residual
stress. Cyclic properties and fatigue endurance of the FSW joints are also investigated and
discussed. Critical areas for natural fatigue crack initiation in FSW are pinpointed. The fatigue
mechanism in FSW is identified to follow a multiple crack coalescence nature (Figure 1). The
number of cracks participate in coalescence and the resulting crack growth rate is governed by the
dynamic distance between the crack tips from crack initiation to coalescence. The above represents
a complex condition for modelling.
Based on crack growth and characterisation of FSW joints, a modified version of the Hobson-
Brown [1-3] is adopted. The good correlation achieved between the experimental data and the
model predictions is shown in Figure 2. The model can successfully handle multiple cracking and
coalescence as well as modelling crack growth within residual stress fields.
The model follows the damage tolerance design philosophy, which is widely accepted in
aircraft manufacturing. The damage tolerance approach incorporates knowledge of fatigue crack
detection, initial crack length, propagation direction, fatigue crack growth rate and maximum crack
length that leads to inspection time intervals.

Figure 1. Replicas images for LCF of 2024-T351 AL Alloy FSW polished mirror specimen at
300MPa applied stress with load ratio R=0.1 showing the coalescence of cracks.
164 A. Ali et al.

Figure 2. Life prediction of FSW 2024-T351 Al Alloy polished mirror.

References
1. P. D. Hobson, M.W. Brown., E. R. de Los Rios, Two phases of short crack growth in a
medium carbon steel. Behaviour of shorth fatigue crack(Edited by K. J. Miller & Eduardo),
1986: p. 441-459.
2. Shaikh, Z., Initiation, Propagation and Coalescence of Short Fatigue Crack in AISI 316
Stainless Steel at 200C and 5500C. PhD. Thesis, Sheffield University, 1991.
3. Gao, N., M. W. Brown, Short Crack Coalescence and Growth in 316 Stainless Steel
Subjected to Cyclic and Time Dependent Deformation. Fatigue and Fracture Engineering
Material Structure, 1995. 18(12): p. 1423-1441.
2T5. Fatigue and fracture 165

FATIGUE CRACK INITIATION IN A TWO PHASE %-METASTABLE TITANIUM


ALLOY: INFLUENCE OF MICROSTRUCTURAL PARAMETERS

A. Lenain, P.J. Jacques and T.Pardoen


Département des Sciences des Matériaux et des Procédés, Université catholique de Louvain,
IMAP, Place Sainte Barbe 2, B-1348, Louvain-la-Neuve, Belgium
lenain@imap.ucl.ac.be, jacques@imap.ucl.ac.be, pardoen@imap.ucl.ac.be

A new generation of DEtitanium alloys with enhanced performance over density ratio is currently
receiving lots of attention from the aerospace industry. One of the critical parameter is the fatigue
resistance in the both low and high cycle fatigue (LCF and HCF) regimes, in particular the crack
initiation. Fatigue initiation is strongly influenced by several characteristics of the microstructure
such as grain size, grain orientation and misorientation (e.g Jin et al. [1]). Up to now, the initiation
of fatigue cracks in the new low cost beta (LCB Ti alloys) has been characterised only in the case
of a fully E microstructure (Hu et al. [2], Krupp et al. [3], Floer et al. [4]). Furthermore, hardly
anything can be found in the literature on the fatigue crack initiation in multiphase Ti-based
microstructures.
This study aims at understanding the couplings between cracking initiation and microstructural
features in the DE TIMETAL LCB titanium alloy under cycling loading.
Fatigue tests were carried out on notched specimens presenting either equiaxed (Fig. 1 (a)) or
acicular (Fig. 1(b)) morphologies. The first microstructure is obtained by hot rolling and
subsequent annealing at 760°C and the second one is obtained by continuous cooling from 810°C
at a speed rate of 2°C/min. These two microstructures were first characterized by SEM, XRD,
EBSD and TEM.
Fatigue tests are regularly stopped and sheet specimens are observed with the FEG-SEM.
Results, obtained on the equiaxed microstructure (Fig. 1 (a)), show that slip lines appears, mostly
in the E grains (Fig. 2 (a)). The density increases with the number of cycles. These slip lines
constitute privileged damage sites. The orientation of the E grains in which slip lines first appear
has been determined. Fig. 2 (b) shows that there is a misorientation profile across two gliding lines
in a E grain which is explained by the lattice rotation due to a deformations gradient into the grain.
166 A. Lenain et al.

References
1. 1. Jin O., Mall S., Materials Science and Engineering A, vol 359, p. 356-367, 2003.
2. Hu Y.M., Floer W., Krupp U., Christ H.-J., Materials Science and Engineering A, vol. 278, p.
170-180, 2000.
3. Krupp U., Floer W., Lei J., Hu Y., Christ H.-J., Philosophical Magazine A, vol. 82, No 17/18,
p. 3321-3332, 2002.
4. Floer W., Krupp U., Christ H.-J., Schick A., Fritzen C.-P, In Proceedings of the Eighth
International Fatigue Congress, edited by A.F. Blom, Stockholm, 2002, p. 2369-2376.
2T5. Fatigue and fracture 167

EFFECTS OF SPECIMEN TYPE, SIZE AND MEASUREMENT TECHNIQUES


ON FCGR

B. Kumar and J. E. Locke


National Institute of Aviation Research & Wichita State University
1845. N Fairmount St, Wichita KS 67260
brijesh.kumar@wichita.edu, James.locke@wichita.edu

The literature on fatigue and fracture is extremely extensive, mainly due to the large number of
variables involved in the process. It is not possible to look at all the possible variables involved in
the fatigue process. The present study is focussed on some of the aspects involved in generating
fatigue crack growth data namely: specimen type, size, measurement techniques and test methods.
Schjive [1] mentioned several reasons for conducting fatigue test programs such as:
comparative experiments; compiling basic material data; verification of prediction models and
research on fatigue and fracture phenomena. The choice of specimen type is usually determined
by specimen production; ease of carrying out experiments, reproducibility, accuracy of test results,
and comparability to other test programs.
The specimen types that have been and are being investigated can be classified into the
following types:
1 Symmetric specimens
a. Middle crack tension specimen M(t)
b. Double edge notch tension specimen
2 Asymmetric specimens
a. Compact tension C(t) specimens
b. Eccentrically loaded single edge coupon ESE (t)

The specimen width’s being considered are 2.00” & 3.00” for both the thin and thick sheet
materials. The thicknesses range from 0.040” to 0.25”. The material’s being investigated are are
2024-T3 & T3511, 7050-T7451, 7075-T6 &T7351 and 7475-T7351 Aluminum alloys. They were
so chosen as they are widely used by the aircraft industry.
The test data has been generated using several different measurement techniques: optical
microscope, clip gage, Fractomat™ based Krak gages, and the ACPD system to monitor crack
growth. All these instruements provide crack growth resolution higher than required by ASTM
647E. The figures 1 & 2 show the crack length measurement data for 7475 and 7050 Al alloys.
Though there seems to be some variation in the reading between the optical microscope and the
clip-gage. We are testing some thicker specimens to see if we still see a difference between them.
There was very good agreement with the Krak Gage and the optical reading. The differences as a
result of the measurements are being analyzed and the results will be presented in da/dn versus
Kmax or K effective.
168 B. Kumar and J. E. Locke

Figure 1 & 2: Crack length versus number of cycles for Al 7475 and 7050 (t = 0.10”) ESE (t)
specimen using clip gage, optical microscope and Krak Gages.

There are several crack initiation techniques recommended, and their subsequent effect on
crack propagation data is still a source of concern and debate. Initial testing using the guidelines
provided in ASTM 647E, which recommend crack initiation to occur in no less than 1.0 x105
cycles has been followed. Crack growth data shows considerable variability near the threshold
regime Marci [2 & 3], this is because of the effects of the R-ratio, testing methods and specimen
types. It would seem that the differences in the data generated is dependent on of the above
mentioned variables. Though all the variables are beyond the scope of this investigation we are
will conduct some limited testing to see the effects of the testing methods. Most of the testing used
in this investigation is based on the constant Pmax loading with two stress ratio’s R =0.5 and R
=0.1. The K reducing method to determine FCGR is also planned.
Finally results from the testing of the M(t) , C(t), ESE(t) and the DENT double edge notch
specimens using different techniques will be reviewed and presented.

References
1. J. Schjive, “Fatigue Specimens for Sheet and Plate Material”, Fatigue & Fracture of
Engineering Materials & Structures, vol 21, pp 347-357, 1998.”
2. G., Marci, D. E., Castro, V., Bachmann, “Fatigue Crack Propagation Threshold”, Journal of
Testing and Evaluation JTEVA, vol. 17, No. 1, Jan. 1989.
3. G., Marci, “Non-propagation conditions (Kth) and Fatigue Crack Propagation Threshold
(KT)”, Fatigue Fracture Engineering Mater. Struct. vol. 17, No 8, pp 891-907, 1994.
2T5. Fatigue and fracture 169

THE EFFECT OF STRESS RATIO ON FATIGUE SHORT CRACKING

C. A. Rodopoulos and S.-H. Han


Structural Materials and Integrity Research Centre, Materials and Engineering Research Institute,
Sheffield Hallam University, City Campus, Howard Street, United Kingdom
C.Rodopoulos@shu.ac.uk
Structural Safety Group, Korea Institute of Machinery & Materials,
171 Jang-dong, Yusung-Gu, Daejeon, 305-343, Republic of Korea
seungho@kimm.re.kr

Short or Stage I crack growth represents a unique case where the Paris-Erdogan Linear Elastic
Fracture Mechanics model fails to predict crack growth and hence fatigue life. The propagation
rate of short cracks has been found in numerous works and for a variety of materials to be
significantly higher than that predicted by the aforementioned model [1,2]. Yet, there is a strong
relation between the material and the extent of short cracking. In general, aluminium alloys exhibit
a much larger short cracking potential to high strength steels. In addition, stress ratio has also been
found to play an important role on such phenomenon [3]. In brief, the extent of short cracking has
been found to decrease with the stress ratio and even disappearing after a particular value [4]. In a
number of papers Rodopoulos and co-workers argued that the phenomenon depends on the ratio
between the fatigue limit and the cyclic yield stress [5], as shown in Figure.1.

FIGURE.1: Extent of fatigue short cracking (iint) versus the ratio between fatigue limit and cyclic
yield stress.

Such argument was based on examining the point where the transition line between a Stage I
and a Stage II (long crack) crosses the line representing transition between a propagating and a
non-propagating crack (Kitagawa-Takahashi diagram), see Figure 2.
170 C. A. Rodopoulos and S.-H. Han

FIGURE.2: Condition for evaluating the extent of short cracking. The solid line represents the
transition between short and long crack growth and the dashed line represents conditions for crack
arrest.

In this work, a sensitivity analysis is presented after the above transition lines have been
modified to include the effect of stress ratio. The work concludes with the development of material
maps versus stress ratio and the tendency towards short cracking.

References
1. Pearson, S. Engng., Fract. Mech. Vol. 7, 235-247, 1975.
2. Morris, W. L. Metall. Trans., A84, 589-596, 1976.
3. Edwards, P. R., Newman, J.C. AGARD R-767, 1990.
4. C. A. Rodopoulos, E. R. de los Rios, Facta Universitätis, Series Mechanics, Vol. 3, 13, 647-
655, 2003.
5. C. A. Rodopoulos and E. R. de los Rios, Inter. J. of Fatigue, 24, 719-724, 2002.
2T5. Fatigue and fracture 171

DWELL-FATIGUE BEHAVIOUR OF A BETA-FORGED TI 6242 ALLOY

P. Lefranc, C. Sarrazin-Baudoux and V. Doquet2


LMPM – UMR CNRS 6617, ENSMA, 86961 Futuroscope-Chasseneuil Cedex, France
2LMS, Ecole Polytechnique, Palaiseau, France

lefranc@lmpm.ensma.fr, baudoux@lmpm.ensma.fr, doquet@lms.polytechnique.fr

For over three decades, gas turbine industries have been confronted to the phenomenon of ambient
temperature dwell fatigue sensitivity in titanium alloys. This effect consists of a reduction of
fatigue lives due to the introduction of dwell periods at the peak stress of the cyclic loading. Up to
now, no real mechanism can explain such behaviour.
During the last three decades, many studies [1] were performed to investigate the influence of
parameters such as microstructure, temperature, dwell period's length and hydrogen content on this
dwell-fatigue behaviour. Characteristic features are a significantly increase in dwell sensitivity
with a coarse lamellar microstructure or with an increase in hydrogen content and the
disappearance of dwell susceptibility in the temperature range of 150°C and 200°C.
This paper deals with a study of the influence of microstructural anisotropy on the dwell-
fatigue behaviour and with the influence of the dwell period introduction at the peak stress of the
cyclic loading on the crack initiation. So, cyclic and two minutes dwell tests were performed at
room temperature on a near alpha Ti-6242 forged above the beta transus. This kind of forging plus
heat treatments lead to a coarse lamellar microstructure either aligned in prior beta grains or basket
weaved microstructure.
Cylindrical specimens used in this study have a utile length of fifteen millimetres and a
diameter of four millimetres. Tests were conducted at different stress levels leading to lives
ranging between 104 to 105 cycles. Some specimens were instrumented with acoustic detection
gages in order to detect initiation.
The results show that a same dwell life debit ranging between 3:1 and 5:1 can be noticed for
the both specimen orientations. But one of the specimen orientations has shown cyclic and dwell
lives two times greater than the other one. Specimens with the weaker cyclic and dwell lives have
also shown up a lot of cracks along its external surface.
Finally, MEB observations have been performed in order to precise the cracking process. The
fracture surface morphologies for cyclic and dwell loading are respectively characterised with a
particular attention to the localisation of the ignition sites. Then, these morphologies are related to
the crack growth rate as determined from striations spacing and using relationships from the
literature which relate the crack propagation rate to the stress intensity factor or the local plastic
deformation for short and long cracks.
The mechanisms controlling the dwell effect are finally discussed on the basis of these results
and observations confronted to the available literature [1-6].

References
1. M. R. Bache. A Review of Dwell Sensitive Fatigue in Titanium Alloys : The Role of
Microstructure, Texture and Operating Conditions, International Journal of Fatigue, Vol. 25,
pp 1079-1087, 2003.
172 P. Lefranc et al.

2. D. Eylon,, J. A. Hall., Fatigue behaviour of E-processed titanium alloy IMI 685, Metallurgical
Transactions, Vol. 8A, pp. 981-990, 1977
3. W. J. Evans, C. R. Gostelow. The effect of hold time on the fatigue properties of D/E-
processed titanium alloy, Metallurgical Transactions, Vol. 10A, pp .1837-1846, 1979.
4. J. E. Hack and G. R. Leverant. The influence of microstructure on the susceptibility of
titanium alloys to internat hydrogen embrittlement, Metallurgical Transactions, Vol. 13A, pp.
1729-1738, 1982.
5. M. R. Bache. M. Copet, H. M. Davies, W. J. Evans and G. Harrison Dwell sensitive fatigue in
a near D titanium alloy at ambient temperature, International Journal of Fatigue, Vol. 19, Sup.
No l, pp S83-S88, 1997.
6. M. E. Kassner, Y. Kosaka and J. A. Hall. Low-cycle dwell-time fatigue in Ti-6242,
Metallurgical and Materials Transactions, Vol. 30A, pp. 2383-2389, 1999.
2T5. Fatigue and fracture 173

INVESTIGATION INTO FATIGUE LIFEOF WELDED CHEMICAL PIPELINES

Cz. Goss and L. Sniezek


Military University of Technology
00-908 Warszawa, ul. Kaliskiego 2, Poland
l.sniezek@wme.wat.edu.pl

Industrial pipelines, in particular the ground-based and overhead ones within any chemical works,
demand special attention owing to hazards to the employees, population in the surrounding area,
and environment. Hazards are greatest while transferring dangerous media between manufacturing
departments and during the storage thereof. Hence, there is a permanent need to deal with the
problems of pipeline strength [1-2].
Failures/damages are usually attributable to, e.g. corrosion, design errors, flaws, and welding
notches. The latter ones are closely related with any discontinuities and changes in the shape of the
pipeline’s cross-section, which often become potential locations of crack initiation due to stress
concentrations.
The states of strain and stress in the bottom of a welding notch, i.e. the point of fatigue
cracking initiation, are determined with account taken of the fact that the material can get
plasticized even if the rated stresses (beyond the notch) are lower than the yield point of the steel.
Such being the case, one should aim at defining the effort of the material in the region of the
welding notch [3-4].
In the course of dimensioning the welded joints within the range of low-cycle fatigue strength,
the so-called ‘notch-bottom method’ is used. In the bottom of the notch there are stress and strain
concentrations occurring at the same time. They can be produced owing to some change(s) in the
shape of the joint itself and the shape of the weld; also, due to welding defects in the forms of
undercuts, complete fusion, no weld penetration, and to residual stress that remains after
manufacturing processes needed to fabricate the structure, e.g. after cutting, bending, welding, etc.
The effort in the bottom of the notch (in a transverse weld this would be the so-called line of
fusion, i.e. the native material and that of the weld penetrating each other) is practically found by
means of computations only.
Estimated were pipelines made of the 1H18N9T steel: those after thirty years’ service, and the
‘new’ ones. Chemical compositions of the steel used to construct these pipelines are shown in
Table 1. Table 2 presents static and cyclic mechanical properties of the steel.

TABLE 1. Chemical composition of the 1H18N9T steel.


Pipeline C Si Mo Ti Cr Mn Ni
% by weight
„old” 0,05 0,64 0,21 0,63 18,5 1,2 9,8
„new” 0,04 0,8 0,4 0,4 17,88 1,8 8,89
174 Cz. Goss and L. Sniezek

TABLE 2. Material properties.

The computational model presented in the paper provides capability to determine the effort of
the material (i.e. of the steel used to manufacture the chemical pipeline) in the area of the welding
notch. On the grounds of both the literature data and experimental work carried out by the Authors,
analytical dependences have been formulated to calculate then the elastic-plastic range of strain in
welding notches in pipelines operated for thirty (30) years, and in the ‘new’ ones. Computations
were made with the strain criterion applied, i.e. the effort of the material in the welding-notch area
was determined. A comparison between the range of strain in the welding notch (with the range of
nominal stresses assumed) and the destructive strain of steel characterised with the computational
range of strain has been accepted as the basis to calculate load capacity of welded joints for the
low-cycle range.
Some formulae have been suggested in the paper to calculate computational, low-cycle fatigue
strength expressed with the range of total strain for a pre-set number of loading cycles. On these
grounds formulae have been introduced to facilitate computations of service lives of pipelines. The
service lives are defined with the number of loading cycles Nf up to the crack initiation, with the
range of nominal stresses in the pipeline coating assumed to be 'VN. The calculated values of
fatigue lives for pipelines after many years’ operational use, with R = 0,1 assumed, are lower by 60
- 80% as compared to ‘new’ pipelines.

References
1. DeWolf G. B.: Process Safety Management in the Pipeline Industry: Parallels and
Differences Between the Pipeline Integrity Management (IMP) rule of the Office of Pipeline
Safety and the PSM/RMP Approach for Process Facilities. Journal of Hazardous Materials.
vol. 104, 169-192, 2003.
2. Papadakis G. A.: Major Hazard Pipelines: a Comparative Study of Onshore Transmission
Accidents. Journal of Loss Prevention in the Process Industries. vol. 12, 91-107, 1999.
3. Samuelsson J.: Design and Analysis of Welded High Strength Steel Structures. EMAS.
Stockholm, Sweden, 2002.
4. Maddox S.J.: Fatigue Strength of Welded Structures. Second edition. Albington Publishing,
Albington, 1991.
The testing work has been done and will be continued under the Research Project No. 5 TO7B 031
25, with financial support of the Committee for Scientific Research (KBN) in the years 2003-2006.
2T5. Fatigue and fracture 175

DIFFERENT ANALYTICAL PRESENTATIONS OF SHORT CRACK GROWTH


UNDER ROTATION-BENDING FATIGUE

D. Angelova and A. Davidkov


Donka Angelova – Professor, University of Chemical Technology and Metallurgy-Sofia (UCTM),
8 Kl. Ohridsky Blvd., 1756 Sofia, Bulgaria, e-mail: dona_ang@techno-link.com
Aleksander Davidkov – Research Fellow, Institute of Metal Science, Bulgarian Academy of
Sciences, 67, Shipchensky Prohod Blvd.,1574 Sofia, Bulgaria, e-mail: lda14042001@yahoo.com

Although the enormous progress in fatigue investigations and the understanding achieved during
the last years, fatigue phenomenon stays as an important problem concerning strength of metals,
their life and structural integrity of engineering constructions. One method which is very useful,
informative and easy for application is that of investigation of short fatigue-crack growth by
replica monitoring of surface crack propagation from the initiation to failure. In our case, this
method includes short-crack rotation-bending (R-B) experiments and length measuring of
propagating crack a at some cycles N on smooth hour-glass specimens subjected to different
frequency and symmetric cycling loading, Table 1. Material under investigation is a rolled low-
carbon, low alloyed steel (RLCLAS), mostly used for off-shore applications and in shipbuilding,
marked as 092 according to the Bulgarian Construction Steel Standard. Experimental conditions
under in-air R-B are the following: specimen 1 (Stress range , MPa – 620, Frequency f, Hz – 11),
and respectively: specimen 2 (620, 11), specimen 3 (580, 11), specimen 4 (620, 6.6).
The experiments are carried out on a table-model machine for R-B fatigue, FATROBEM–
2004, newly designed, constructed and assembled in the Laboratory of Plastic Deformation of
UCTM. The machine is shown on Fig. 1 and its work described in Davidkov [1].

FIGURE 1. Test apparatus scheme: electric engine 1, driving belt 2, ball-bearing unit 3, leading
shaft 4, corrosion box 5, specimen 6, leaded shaft 7, device for circulation and aeration of
corrosion agent 8, working box 9, device for loading and load changing 10, counter 11.

A model “short crack growth-rate, da/dN – crack length, a” under cyclic R-B fatigue – Eq. (1)
– comprises parabolic-linear presentation of the three regimes of crack propagation (I SFC – short
fatigue crack, II PhSFC – physically small and III LFC – long crack), and analytical determination
of microstructural barriers d1 and d2, at which cracks are slowing down and stop [1]; a0 and af –
the initial and final crack size, A, B, C – coefficients depending on metal nature

da
I SFC {a(a0 , d1]} and II PhSFC {a[d1 , d2]}: Aia 2  Bia  Ci , i=1, 2;
dN
176 D. Angelova and A. Davidkov

da
III LFC { a[d2 , af )}: A3 a B3 (1)
dN

The basic model functions from (1) are corrected for each regime, considering the highest crack
growth rates, and presented in Fig. 2 as a parabolic-linear family 1.

FIGURE. 2. Different presentation of fatigue data: a. Corrected dependences “Crack growth rate,
da/dN – crack length, a” o Family 1; and b. “Crack growth rate, da/dN – surface energy,
'W ” o Line 2

An alternative approach to already described, classical way of treating fatigue comprises a new
function W, and a new presentation of experimental data in parameters dN / da , 'W . The function
'W discussed for the first time in Angelova [2, 3] has a dimension of surface energy per second
and unity of crack size, and leads to a linear presentation of fatigue data, da/dN– 'W o Line 2
shown as thick line in Fig. 3. The scatter band is indicated by fore thin lines at a condition of
excluding two distant points from Family 1; the inner thin lines correspond to ½ and 2 folds of da/
dN, and the outside ones - to 0.4 and 2.5. Such a precision makes it possible to use the presentation
da/dN– 'W at a lesser number of the measurements (recordings) made under every fatigue test,
especially at high correlation coefficient which in our case is fcor>0.8. Afterwards a transition to
the presentation “da/dN – a” showing much larger scatter would not be difficult.

References

1. Davidkov A., Angelova D., In Proceedings of the 2-nd International Conference


“Deformation, Processing and Structure of Materials”, Belgrade, May 2005, 173-178.
2. Angelova, D., In CD Proceedings of ECF13, Abstracts, San Sebastian, September 2000, 128.
3. Angelova, D., In Proceedings of the ECF14, Vol. 1, Cracow, September 2002, 89-97.
2T5. Fatigue and fracture 177

VARIABLE AMPLITUDE LOAD INTERACTION IN FATIGUE CRACK


GROWTH FOR 2024-T3 ALUMINIUM ALLOY

D. Kocanda, S. Kocanda and J. Torzewski


Military University of Technology
Kaliskiego 2 Str., 00-908 Warsaw, Poland
d.kocanda@wme.wat.edu.pl

Special needs that concern fatigue damage of aircraft components require undertaking the
investigations of fatigue crack growth in aerospace alloys and finding the correlation between the
images of fracture surface and the accumulated damage due to applied load. Importance of
fractography supporting the analysis of failure cases of real structures was pointed out among
others by J. Schijve in his recently published book [1]. In order to check the capability of the
reconstruction of load-time history on the basis of fractographic analysis extensive research
program was developed for 2024-T3 Aclad aluminium alloy sheet subjected to different variable
amplitude loads [2]. The paper presents the results both of fatigue crack growth response and the
microfracture analysis for the CCT specimens (400x100x3 mm) under two similar loads LHL
(low-high-low) and FAF (flight-after-flight) (Fig. 1a and b). These loads are employed when
simulating the flight loads of the lower skin aircraft wing structure.

FIGURE 1. Loading of type LHL (a) and type FAF (b)

Load sequence LHL contains 2400 cycles whereas FAF sequence counts 240 cycles of the
stress levels just the same the LHL load. However, these loads differ either from the number of
cycles in the particular load blocks, thereby the number of overloads, or the order of the load
blocks in the program. Employing these two load programs we can learn the influence of the load
shape on the crack growth rate and the lifetime of a component as well. Crack growth rate
behaviour in 2024-T3 alloy was analyzed either on the surface or in the depth of the samples and
performed on the diagrams. The extension of crack depth was estimated on the basis of
fractographic analysis with the help of SEM or TEM microscopes. In many cases the effect of
multiple overloads that intersperse the baseline cycles is not clearly visible on the typical
experimental plot da/dN as function of 'Keff. More details of crack retardation and acceleration
associated with the overloads provide the microfracture analysis done with the help of TEM
microscope. Figure 2 performs the micrographs which illustrate the systems of fatigue striations on
the fracture surface under LHL and FAF load sequences. Microfracture analysis revealed a big
variation of crack growth rate within the particular load sequences. The courses of local crack
growth rate, affected by one load program of LHL type (Fig. 3a) and FAF type (Fig. 3b) against
crack length are shown in Figure 3. Crack growth rates were estimated on the basis of fatigue
striations spacing referred to the particular blocks of loading. Further analysis proved that under
LHL program the period of crack propagation covers less than 10% of load program duration
178 D. Kocanda et al.

whereas in the case of FAF program 26% of total time is devoted to crack propagation. Well, most
time of the load program duration is associated with the crack arrest or crack growth at very low
rate. This crack behaviour results from the plastically induced crack closure effect as well as the
crack penetration through the plastic zones associated with the overloads.

FIGURE 2. TEM (a) and SEM (b) micrographs illustrate the fatigue striation systems on the
fracture surface that are associated with LHL program (a) and FAF program (b).

FIGURE 3. Variation of crack growth rates in microscopic scale for 2024-T3 alloy within one load
sequence of LHL (a) and FAF (b) loadings against crack length, respectively.

For prediction the crack growth under VA loading and estimation the fatigue life of a
component was elaborated a deterministic approach. The crack retardation model is based on the
Wheeler model.

References
1. Schijve , J., Fatigue of structures and materials, Kluwer Academic Publishers, 2001.
2. Kocanda D., Kocanda S. and Torzewski J., Archive Mech. Engineering, vol. LI, No 3, 361-
376, 2004.
2T5. Fatigue and fracture 179

AN INVESTIGATION ON THE FATIGUE PERFORMANCE OF HYDRAULIC


GATE WHEELS

D. Polyzois and A. N. Lashari


Department of Civil Engineering, University of Manitoba
15-Gillson St., Winnipeg, MB, R3T 5V6, Canada
polyzoi@cc.umanitoba.ca

Manitoba Hydro currently owns fourteen hydro power generating stations with a total capacity of
over 7500 MW. Both emergency intake gates and spillway gates are used in each. These are fixed-
wheel gates with wheels mounted on both sides which roll on roller paths. Environmental
corrosion along with high wheel loads cause differences in the profile of the roller path surface.
Combined with the relatively high torsional stiffness of the gate end girders, a condition of wheel
load redistribution occurs where some wheels are relieved of load while others are loaded beyond
their maximum design values. These loads can be as high as two to three times larger as the
original design loadings. Failure of one wheel could jeopardize the overall operation of the gate.
Currently, the design of gate wheels and roller paths do not consider the fatigue life of these
elements. Metal fatigue is a process resulting in failure or damage of a component subjected to
repeated loading. Although failure by progressive fracture is often associated with localized tensile
stresses, fatigue cracks can also occur on the plane of maximum shearing stresses. In the case of
wheels or wheels used in spillway or control gates, contact stresses on or somewhat beneath the
surface of the contact surface can cause failure of one or both of the bodies. In this case, since the
contact point changes as the gates open and close, the contact stresses are repeated over many
times, a situation that could eventually lead to fatigue failure. Usually, failure starts as a localized
fracture just below the surface of contact where the shearing stress is high and progresses
outwardly under the influence of the repeated wheel loads. Contact stresses can also cause pitting
at the surface of contact, as shown in Fig. 1.
An experimental investigation was carried out at the University of Manitoba in Winnipeg,
Canada, which involved the testing of three wheels and six roller path plates under cyclic loading.
The wheels were 838.2 mm in diameter while the roller path plates were 381 x 177.8 x 50.8 mm.
One of the wheels was made of cast iron while the other two were made of forged steel. The
material in two of the roller path plates was ASTM C 1040 with no heat treatment. The material in
the other plates was ASTM C 1045 with heat treatment. A unique test set up, shown in Fig. 2, was
designed and constructed for this special fatigue type of testing. The average Brinell Hardness
Number (BHN) for plates P1 and P2, which were not heat treated, was 291, whereas, for the rest of
the roller path plates, which were heat treated, the average BHN was 364. The hardness for the cast
iron wheel varied from 391 BHN at the rolling surface to 219 BHN at 38 mm away from the
rolling surface while hardness for forged steel wheel (R2) varied from 373 BHN at the rolling
surface to 326 BHN at 63 mm away from the rolling surface. The hardness for the other forged
steel wheel (R3) varied from 473 BHN at the rolling surface to 428 BHN at 38 mm away from the
rolling surface.
180 D. Polyzois and A. N. Lashari

The wheels were subjected to a radial load of, approximately, 825 kN that remained fairly
constant while the wheels were “rolled” over the roller path for up to one million cycles. The
tensile strain on the cast iron wheel ranged from 103 µ to 2057 µ while the tensile strain on the
forged steel ranged from 4 µ to 206 µ. Tensile strains were observed in almost all the strain gauges
installed on all roller path plates.
Roller path plates P1 and P2, which were not heat treated, exhibited a maximum indentation of
1.48 mm and 1.21, respectively, after one million cycles. Roller path plates P3 and P4 recorded a
very low indentation depth of 0.03 mm and 0.11 mm, respectively. The other four roller plates
were heat treated and suffered a much smaller surface indentation which ranged from 0.02 to 0.11
mm after 400,000 cycles. Clearly the test results demonstrated that cast iron wheels performed
very poorly under fatigue loading while heat treated forged steel wheels performed well.
2T5. Fatigue and fracture 181

A MICROMECHANICAL MODEL FOR CRACK INITIATION IN HIGH CYCLE


FATIGUE OF METALLIC MATERIALS

Vincent Monchiet, Eric Charkaluk and Djimedo Kondo


Laboratoire de Mécanique de Lille, UMR CNRS 8107, Université Lille 1, France
Vincent.Monchiet@ed.univ-lille1.fr, eric.charkaluk@univ-lille1.fr, kondo@univ-lille1.fr

Most of structural components resisting to high cycle fatigue are subjected to a multiaxial state of
stress. Since fatigue cracks generally initiate and propagate in a plane of maximal shear stress
(stage I), the first approaches of Crossland and Sines considered the octahedral plane and their
criteria are based on the amplitude of the second invariant of the deviatoric stress tensor. In order
to take into account the mean stress, these authors postulated also a linear combination of J 2 and
the hydrostatic part of the stress tensor. It is now widely recognized that under high cycle loadings
conditions, metals fracture is the result of the strain localization in some unfavorably oriented
grains undergoing plasticity. Dang Van, followed by Papadopoulos and others, had proposed a
multiscale framework which leads to a sufficient condition for non nucleation of cracks under high
cycle loading. This condition is ensured by elastic shakedown at the grain scale (see the
representive elementary volume on Fig. 1). Nevertheless, microplastic activity alone cannot
explain the role of pressure on the fatigue limit and a generalized multiaxial fatigue limit
depending on the hydrostatic pressure is due. So far, this is the proposal already suggested by
Papadopoulos in his generalized multiaxial fatigue limit.

FIGURE 1 : Representative Elementary Volume (RVE)

To support this type of modelling, we propose in the present study to incorporate some
observed damage mechanisms (reported in literature for materials involving faced-centred-cubic
structures) in the multiscale approach of Dang Van. The characteristics of the study lies in the
consideration of the scale of Persistent Slips Bands (PSB) at which appear the damage
micromechanisms. Actually, the PSB constitute preferential sites for fatigue cracks nucleation.
The outline of the study is as following :
1 we first propose a simple model for the grain behavior under high cycle loading :
microplasticity and microdamage mechanisms (induced by micro-cavities growth) are
incoprorated in this grain level model. The monocrystal plastic behavior is described by
the a Schmid's law with a linear isotropic hardening and non linear kinematics hardening
law, the later being based on the effective plastic slip. On the basis of experiments by
Antonopoulos et al. (1976) and Esseman et al. (1982), damage along PSB is associated to
the production of vacancies (which is responsible of irreversible volume change) and non
spherical microvoids nucleation and growth into these localized bands. The model is
completed by adopting the crack nucleation condition proposed by Antonopoulos et
182 V. Monchiet et al.

al.(1976) according to which a crack nucleates at the PSB-matrix interface when the total
strain along the PSB reaches a critical value. At the grain level, a crack nucleation criterion
is then derived and allows to determine the local condition for fatigue limit.
2 a second part of the study deals with non linear homogenization techniques allowing to
translate the local crack nucleation criterion into a macroscopic fatigue criterion. Different
non linear homogenization schemes are considered : Lin-Taylor model, Sachs scheme and
Kroner's accomodation law. The consideration of such simple schemes is justified by the
fact that in high cycle fatigue context, plasticity is localized in one or few grains.
An illustration of the micro-macro approach in the case of macroscopic affine loadings is
presented on Fig. 2 ; this clearly shows the capability of the model to take into account the effect
of the pressure. Besides, the fatigue limit appears to be independent of the mean shear stress.

Ta
FIGURE 2 : Affine torsion-tension loadings. Amplitude of the normalized shear stress as
W0
function of the normalized mean pressure. Comparison of the predictions of two homogenization
schemes (Sachs, Eshelby-Kroner).

References
1. J. G. Antonopoulos, L. M. Brown, A. T. Winter. Vacancy dipoles in fatigued copper.
Philosophical Magazine, Vol. 34, No. 4, pp. 549-563, 1976.
2. M. Berveiller, A. Zaoui. An extension of the self-consistent scheme to plastically flowing
polycrystals. J. Mech. Phys. Solids, Vol. 26, pp. 325-344, 1979.
3. U. Essmann, U. Gosele, H. Mugrhabi. A model of extrusions and intrusions in fatigued
metals. I. Point-defects production and the growth of extrusions. Philosophical Magazine A,
Vol.44, No. 2, pp. 405-426, 1991.
2T5. Fatigue and fracture 183

COMPARATIVE ANALYSIS OF TWO MODELS FOR EVALUATING FATIGUE


DATA

Enrique Castillo, Antonio Ramos2, Manuel Lopez-Aenlle2, Alfonso Fernandez-Canteli2 and


Roland Koller3
ETSICCP, University Cantabria, Avda. de los Castros s/n, 39005 Santander, Spain.
2EPSIG, University Oviedo, Campus de Viesques, 33203 Gijón, Spain.
3EMPA, Überlandstr. 129, 8600 Dübendorf, Switzerland.

castie@unican.es, afc@uniovi.es, Roland.Koller@empa.ch

The optimization of fatigue programs comprising planning, testing and results evaluation is an
issue of paramount importance for material and testing laboratories and a recursive subject in the
literature which research groups dealing with fatigue design are long concerned with. There is a
general belief that no specific model fulfils in a satisfactory way all these functions, i.e. a model to
be considered acceptable by the community. Nevertheless, the up-and-down method, despite its
limitations, finds a good recognition by both, research groups related to the academy as well as to
the testing laboratories. This can be due to the simple test strategy and evaluation technique
implied in the practical application of the up-and-down method, which is easy to understand and to
apply in practice. The limited information supplied in this methodology, consists in the
determination of the probabilistic stress range for a certain limit number of cycles, here denoted as
pseudo endurance limit, which is, generally, considered, although unjustifiably, sufficient for
practical design [1,2].
Among the inconsequences of the up-and-down method we find the assumption of a normal
(or log-normal) distribution function for the variation of the stress range for given number of
cycles, which is not suppported by statistical analysis of real data [3]. Further, data results
corresponding to the long life fatigue region, i.e., that properly involved in the up-and-down
method, is to be completed by additional results in the medium life fatigue region if the whole S-N
field is required to be considered in the current fatigue life design.
In this paper, an alternative fatigue test strategy and the subsequent parameter estimation from
data results is proposed for use. The procedure is based on a regression Weibull model comprising
the whole S-N field, as developed by Castillo et al. [4]. The model enables us to incorporate
specimens of different lengths into the analysis [5]. Firstly, the two models considered here, that
based on the up-and-down methodology and the regression model proposed by the authors, are
introduced and discussed. The possibility of considering a Weibull distribution for the stress range
in the up-and-down methodology, leading to the so-called Weibull-up-and-down model, is
justified and proposed. The experimental program carried out at the EMPA (Swiss Federal
Laboratories for Materials Testing and Research) is also presented (see Fig. 1). Thereafter, a
comparative analysis of the fatigue results obtained in the frame of this program is performed using
both methodologies.
The applicability and reliability resulting from both procedures is then analyzed and the
advantages and shortcomings of both methods are discussed. Finally, the conclusions of this
investigation are presented.
184 E. Castillo et al.

FIGURE 1. Adjusting fatigue data of 42CrMo4 steel using the proposed regression model .

References
1. Deubelbeiss E., Materialprüfung, Vol. 16, 240-244, 1974
2. Hück, M., Schütz, W., Zenner, H., Industrieanlagen Betriebsgesellschaft mbH, ReportB-TF-
742B, 1974.
3. Castillo, E., Fernández Canteli, A., Esslinger, V., Thürlimann, B.,IABSE Periodica 1/1985,
IABSE Proceedings 82/85, 1-40, 1985.
4. Castillo, E., Fernández Canteli, A., Int. Journal of Fracture Nr. 107, 117-137, 2001.
5. Castillo, E., López-Aenlle, M., Ramos, A., Fernández-Canteli, A., Esslinger ,V., Kieselbach,
R., Int. J. of Fatigue. Submitted to evaluation.
2T5. Fatigue and fracture 185

ASSESSMENT OF DAMAGE AT NOTCH ROOT OF THICK PLATES

E. C. G. Menin and J. L. de A. Ferreira


University of Brasília - UnB, Mechanical Engineering Department, GAMMA Research Group
Campus Univ. Darcy Ribeiro, Asa Norte, Brasília, Brazil,
70940-910; 55-61-3073706 R. 213
edumenin@unb.br, jorge@unb.br

Heavily loaded structural components may present local yield at stress concentrators, such as,
notches and geometrical discontinuities [1]. In cases involving cyclic loads, the presence of local
plasticity could lead to the nucleation and propagation of fatigue cracks and complete fracture of
the components [2, 3]. In such cases, the strain life approach should be used to evaluate fatigue
damage, making it necessary to perform an elastoplastic analysis of stresses and strains at the
proximities of these stress concentrators. Due to reduced computational effort requested,
approximated models are widely used with that purpose. The most used are the ones proposed by
Neuber [4], Seeger et al. [5], Glinka [6] and Ye et al. [7].
One factor that complicates the use of these approaches consists in the evaluation of stress
concentration problems, by most part of the specialized literature, as plane approximation
problems that do not take into account the variations in the stress-strain field configuration along
the thickness of the components, as these become larger in magnitude. To corroborate the
importance of such 3D analysis, several numerical and experimental studies have been published
indicating that the stress field near the geometrical discontinuities varies along the notch root, as
components become thicker [8]. In such cases, 2D values of Kt and stress/strain plane
approximations should be carefully evaluated when associated to local strain models in order to
correctly estimate local levels of solicitations at the notch root of thick components.
In that sense, this study presents examples of the difficulties associated to the use of local strain
approximated models when evaluating cases involving notched components presenting expressive
values for the ratio between thickness and notch root radius, or dimensionless thickness, t/r. A
preliminary linear elastic behavior analysis for the variation of Kt and the displacement constraint
level of points along the notch root was performed using 3D FEM. After that, elastoplastic
analyses were performed by means of the most used local strain models, as well as, finite element
3D models and 2D approximation. These simulated the local behavior at the root of different
geometries and types of steel associated to two cases of study, exemplified on Tables 1 and 2.

TABLE 1. Notch geometries and materials.

TABLE 2. Mechanical and fatigue properties


186 E. C. G. Menin and J. L. de A. Ferreira

The results obtained for the linear elastic domain analysis of the cases of study made it possible
to evaluate the stress constraint level of points at the notch root of the stress concentrators, as well
as, the behaviour of Kt for components with different values of t/r tending to zero and to determine
the reliability and coherence of the use of 2D values of this factor to describe the stress
concentration phenomena.
The results presented for the elastoplastic analysis of the two cases of study made evident the
difficulties associated to the use of local strain approach in cases involving components with
expressive values of the dimensionless thickness. Among these are the ones associated to the
determination of correct values of Kt and the constitutive curve constants to be used along with the
local strain models, the determination of the points subjected to the most severe conditions, as well
as, the type of bi-dimensional approximation to be used with FEM.
The analysis results suggested the necessity of 3D FEM evaluation to determine in a precise
way the local strain and fatigue life of components with expressive values of t/r. Moreover, the
results made evident that, in the impossibility to use such practice, one should be careful when
using plane stress and strain approximations and 2D values of Kt because these could present
divergences with respect to the real conditions evaluated for the models.

References
1. Peterson, R.E. (1997), Stress Concentration Factors, Ed. John Wiley & Sons, EUA.
2. Filippini, M. (2000), “Stress gradient calculations at notches”, Int. Journal of Fatigue, 22,
397-409.
3. Visvanatha, S.K., Straznicky, R.L. e Hewitt, R.L. (2000), “Influence of strain estimation
methods on life prediction using the local strain approach”, Int. Journal of Fatigue 22, 675-
81.
4. Neuber, H. (1961), “Theory of stress concentration for shear-strained prismatical bodies with
arbitrary nonlinear stress-strain law”, Journal of Applied Mec., 28.
5. Seeger, T.H. e Heuler, P. (1980), “Generalised application of Neuber’s rule”, Journal of
Testing and Evaluation, 8, 199-204.
6. Glinka, G. (1985), “Energy density approach to calculation of inelastic strain-stress near
notches and cracks”, Eng. Fract. Mechanics, 22(3), 485-508.
7. Ye, D., Matsuoka, S., Susuki, N. e Maeda, Y. (2003), “Further investigation of Neuber’s rule
and the equivalent strain energy density (ESED) method”, Int. J. of Fatigue, 22, 675-681.
8. Zhenhuan L., Wanlin, G. e Zhenbang, K. (2000), “Three-Dimensional elastic stress fields
near notches in finite thickness plates”, Int. Journal of Solids and Structures, 37, 7617-7631.
2T5. Fatigue and fracture 187

FATIGUE STRENGTH PREDICTION OF SPOT-WELDED JOINTS USING


SMALL SPECIMEN TESTING

Eisuke Nakayama, Manabu Fukumoto, Mitsuo Miyahara, Kazuo Okamura, Hiroki Fujimoto and
Kiyoyuki Fukui1
Corporate Research & Development Laboratories, Sumitomo Metal Industries, Ltd.,
1-8, Fuso-Cho, Amagasaki, Hyogo 660-0891, Japan
1Steel Sheet, Plate, Titanium & Structural Steel Company, Sumitomo Metal Industries, Ltd.,

1-8-11, Harumi, Chuo-ku, Tokyo 104-6111, Japan


nakayama-eis@sumitomometals.co.jp

It is well known that fatigue strength of spot weld of high strength steel sheet is not improved,
compared with that of mild steel sheet. In this study, the governing factors and the effects of steel
grade on fatigue strength of spot weld is investigated. Firstly, small specimens with total length of
less than 3mm are taken from the spot weld of mild steel sheet (270MPa-grade) and high strength
steel sheet (590MPa-grade). And then, tensile and high cycle fatigue properties are individually
evaluated by newly-developed testing technique. Secondly, finite element analyses of tensile-shear
specimen of spot-welded joints under cyclic loading are carried out and fatigue limit of the joints
are predicted, using the above-mentioned local material strength properties and considering
welding residual stresses around spot weld. Predicted results are nearly equal in both steels, which
coincides with experimental results. It is found that fatigue strength of HAZ, which is the crack
initiation site in joint, of 590MPa-grade steel is higher than that of 270MPa-grade steel. However,
residual stress in 590MPa-grade steel is also higher and this may be one of the reasons why
590MPa-grade steel exhibits little improvement in fatigue strength of the joint over 270MPa-grade
steel.

FIGURE 1. Shape and dimensions of fatigue specimen (unit:mm).

FIGURE 2. S-N diagrams of HAZ specimen.


2T5. Fatigue and fracture 189

A THERMO-MECHANICAL MODEL FOR RANDOM BRAKING OF MACHINE


COMPONENTS

F. Loibnegger, H.P. Rossmanith1 and R. Huber


Vienna University of Technology
Institute for Testing and Research in Materials Technology, Karlsplatz 13, A-1010 Vienna
1Institute of Mechanics and Mechatronics, Wiedner Hauptstr. 8-10/325, A-1040 Vienna

friedrich.loibnegger@tuwien.ac.at, hans.peter.rossmanith@tuwien.ac.at,
richard.huber@tuwien.ac.at

This contribution presents a hybrid analytical-numerical treatment of the thermo-mechanical


model for the analysis of the braking process of rotating machine parts such as wheels, etc. The
machine parts have the shape of a thick disk. These disks are repeatedly frictionally disk-braked at
random sequences of time instants. As the loading is rather severe, the stresses during braking and
subsequent cooling reach the plastic limits in compression and tension, respectively. Hence, the
loading may cause conditions similar to those encountered in shake-down phenomena.
The thick disk is assumed to contain initial compressive and tensile stresses as a result of
inappropriate heat treatment after forging and due to press-fitting the disk on the axis. In addition,
the disk is allowed to develop and extend radial surface cracks within the annular brake regions.
The shapes of these surface cracks initially are semi-elliptical, but vary largely, depending on the
loading sequences and the existing residual stress fields after cyclic straining [1].
Under certain loading conditions and random sequences of brake phenomena, the
circumferentially distributed and radial oriented cracks may exhibit accelerated crack growth in
axial direction and in a few cases cracks may even emerge on free rim surfaces.
A thermo-elasto-plastic fracture mechanical analysis was performed, based on a finite element
analysis. The crack was locally modelled as a semi-circular surface crack which was allowed to
propagate at variable velocities along the surface and into the material. Linear elastic fatigue type
crack advance was assumed during the cooling phase after braking.
It was detected, that the distribution of residual stresses had a decisive influence on the stability
behaviour of fatigue cracks. It could be shown that the extension of cracks on the surface along the
circumferential direction was controlled by the residual stresses, ultimately leading to crack arrest.
The large tensile circumferential stresses within the brake region are balanced by the residual
compressive stresses in the rim and hub regions.
Results indicate that the actual load is depending on the randomness and characteristics of the
braking events. The speed of directional crack advance also depends on the style and the loading
sequence.
Safe life – time diagrams could be constructed for the use of the rotating machine parts [2].
These safety diagrams clearly demonstrate the effect of the initial crack size on the arrest
behaviour of the observed crack and the remaining life time of the structure.
This paper discusses in detail the various steps of the very complex analysis [3] and at each
stage indicates alternative routes and avenues for improvement of the analysis while still keeping
the amount of numerical calculations at bay. A practical example will be demonstrated.
190 F. Loibnegger et al.

References
1. Rolfe S.T. & Barsoum, J.M., Fracture and Fatigue Control in Structures - Applications of
Fracture Mechanics, Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1977.
2. Anderson T.L., Fracture Mechanics - Fundamentals and Applications. CRC Press, Boca
Raton, USA, 1991.
3. Gdoutos, E.E., Rodopoulos, C.A. & J.R. Yates : Problems of Fracture Mechanics and Fatigue
- A Solution Guide, Kluwer Academic Press Publishers, 2003.
2T5. Fatigue and fracture 191

LIFETIME CALCULATION OF RAILWAY WHEEL STEELS


BASED ON PHYSICAL DATA

F. Walther and D. Eifler


Institute of Materials Science and Engineering
University of Kaiserslautern, P.O. Box 3049, D-67653 Kaiserslautern, Germany
walther@mv.uni-kl.de, eifler@mv.uni-kl.de

Higher speeds in passenger traffic and higher axle loads in freight traffic cause a substantial
increase of the mechanical load of railway wheels. To develop and verify fatigue criteria and
lifetime calculation methods, a reliable fatigue data base of highly loaded wheels steels, in
particular under near-service loading conditions, is required.
Fatigue tests with variable maximum stresses were performed at ambient temperature on
servohydraulic testing systems with specimens of the wheel steels R7 (SAE 1055) and B6 (SAE
1070) according to UIC standards. The specimens were taken from different depth positions of the
rim of original railway wheels. Due to the industrial heat treatment and the size of the components
microstructural gradients are unavoidable [1, 2]. In addition to the plastic strain amplitude Ha,p
from mechanical stress-strain-hysteresis measurements, changes of the specimen temperature 'T
and the electrical resistance 'R during cyclic loading were considered. The integral temperature
and resistance data can be used to evaluate continuously the fatigue behaviour under variable
amplitude loading. Besides the specimen geometry and temperature, the resistance depends on the
specific resistance of the material and changes in a characteristic manner as a function of the defect
density in the material. To describe proceeding fatigue damage, the resistance can be measured as
a reference value at the beginning and during a fatigue test as well as in load-free states [1, 2].

FIGURE 1. Variable amplitude loading with constant amplitude (measuring) sequences.

The basic idea of the new testing procedure developed at the Institute of Materials Science and
Engineering is to combine near-service load spectra with periodically inserted measuring
sequences with a constant amplitude below the endurance limit. With this method (Fig. 1) reliable
information about proceeding fatigue damage under near-service loading can be obtained on the
basis of Ha,p, 'T and 'R. In representative cyclic ‘deformation’ curves the mean values of these
physical quantities are plotted as a function of the number of cycles N*. N* was calculated by
multiplication of the frequency and the time of experiment. The random signal was generated by a
combination of a White Noise and a Gauss distribution. After normalising the random signal to
values between -1 and 1 it was multiplied with the maximum stress.
Cyclic softening and hardening processes were observed for random loading experiments with
maximum stresses of 400 MPa d Vmax d 800 MPa (Fig. 2), comparable to single-step tests. An
192 F. Walther and D. Eifler

increasing maximum stress leads to higher measured values in an unique manner and consequently
shorter lifetimes. The physical quantities Ha,p, 'T and 'R are directly correlated with
microstructural details. The test with Vmax = 400 MPa was stopped without specimen failure.

FIGURE 2. Representative cyclic Ha,p-, 'T- and 'R-curves for random loading, R7.

Furthermore, the new testing procedure allows to establish lifetime calculation methods for
near-service loading. To derive a physically-based lifetime calculation (PHYBAL), 'T and 'R
were used equivalently to Ha,p in Morrow and Manson/Coffin curves. Using the mathematical
descriptions of these curves, S,N curves according to the Basquin equation can be calculated. The
lifetimes calculated on the basis of the plastic strain amplitude, the temperature and the resistance
agree very well with the experimentally determined lifetimes [3].

References
1. Walther, F., Eifler, D., Mater. Sci. Eng. A 387-389, 481-485, 2004.
2. Walther, F., Eifler, D., In Proceedings of the 11th International Conference on Fracture,
edited by A. Carpinteri, Turin, No. 3944, 2005, 1-6.
3. Starke, P., Walther, F., Eifler, D., PHYBAL - A new method for lifetime prediction based on
strain, temperature and electrical measurements, Int. J. Fatigue, 2005, in press.
2T5. Fatigue and fracture 193

FATIGUE CRACK PROPAGATION OF SUPER-DUPLEX STAINLESS STEEL


AT DIFFERENT TEMPERATURES

G. Chai and S. Johansson


Sandvik Materials Technology, R&D Centre
811 81 Sandviken, Sweden
Linköping University, Department of Mechanical Engineering
SE 581 83 Linköping, Sweden
guocai.chai@sandvik.com, stejo@ikp.liu.se

The fatigue crack propagation behaviour in an austenite-ferrite two phase super-duplex stainless
steel (SDSS) (UNS S32750) has been investigated at temperatures from –50ºC to 150qC. Two
material conditions (as delivered condition-AD and aged condition (475 ºC/4h)) and two stress
ratio values (R=0,1 and 0,5) were used. The fatigue crack propagation rate, threshold value,
closure and fracture behaviour have been studied.
The fatigue crack propagation of this alloy at –50ºC shows a tendency of lower crack
propagation rates at small stress intensity factor range, but higher rates at higher stress intensity
factor range, and a higher crack propagation threshold value comparing with those at RT. The
crack propagation rates at 150ºC are similar to those at RT, and the crack propagation threshold
value is also higher than that at RT, but smaller than that at –50ºC. The ageing of the alloy or an
increase in R value increases crack propagation rate, but decreases the threshold value (Figure 1).

FIGURE 1. Influence of temperature, material and stressconditions on crack propagation


behaviour of a DSS material

This alloy shows rather high fatigue crack propagation threshold values 'Kth (Table 1). It was
found that it is mainly due to the fact that this two phase material has large crack closure,
especially at low and high temperatures (Table 1). The following two phenomena observed may
give some explanations. One is type of “crack-bridges” occurred in the softer phase (Figure 2a).
Compressive residual stresses in this phase were measured after the fatigue testing. This may
indicate that fatigue crack closure can be induced by residual stresses. The other is type of
“fracture mismatch” observed near the threshold regime. The fracture changes its pattern or
directions when a crack propagates from one phase to another (Figure 2b). This may increase the
surface roughness induced closure.
194 G. Chai and S. Johansson

TABLE 1. Influence of temperature on the closure and effective threshold values.

*: from da/dN measurements; **: from closure measurements. '.cl: closure value; 'Keff th :
effective threshold value.

FIGURE 2. (a). Residual stress induced closure; (b). Fracture mismatch from austenite phase Jto
ferrite phase D

The higher effective threshold value 'Keff th at 150qC, which can not be explained by the
classic theories, was further investigated. It may be attributed to dynamic strain ageing.
A transition from cleavage fracture to facet fracture was observed near the threshold values in
the as delivered material at -50qC and in the aged material at RT. The possible mechanisms were
discussed.
Key words: Super duplex stainless steels, Fatigue crack propagation behaviour, Fatigue crack
closure, Temperature, Fracture.
2T5. Fatigue and fracture 195

TRANSITIONS OF FATIGUE CRACK INITIATION FROM SURFACE,


SUBSURFACE TO SNDFCO

G. Chai
Sandvik Materials Technology, R&D Centre,
811 81 Sandviken, Sweden
guocai.chai@sandvik.com

As known, fatigue crack initiation mainly starts at surface defects at high stress or strain
amplitudes, but shifts to pre-existing subsurface defects such as inclusions or pores at low stress
amplitudes or in the very high cycle fatigue regime [1]. Recently, it was reported that fatigue crack
initiation could occur in non-defect existing areas and form a subsurface non-defect fatigue crack
origin (SNDFCO) [2]. Figure 1 shows a summary of these three phenomena.

FIGURE 1. Three types of fatigue crack initiation; (a). From a surface defect; (b). From a
subsurface inclusion; (c). From subsurface non-defect fatigue crack origin (SNDFCO).

This paper gives a discussion on the transition mechanisms of fatigue crack initiation from
surface defects, subsurface defects to subsurface non-defect fatigue crack origins in some two
phase or multiphase steels. It was found that surface crack initiation mainly is caused by extrusion
and intrusion of slip bands at the specimen surface, where the applied stresses are higher than the
elastic limit of the material (Figure 2a). The occurrence of subsurface crack initiation is caused by
localised cyclic plastic deformation or dislocation slipping processes around the subsurface defects
due to stress concentrations (Figure 2b). The formation of SNDFCO in some two phase alloys is
also a cyclic plastic deformation process. Although the stresses applied are lower than the elastic
limit of the material, but they can be higher than the elastic limit of the softer phase. This leads to
the occurrence of cyclic plastic deformation in this softer phase, which causes the damage and then
the formation of micro or short cracks in the softer phase (Figure 2c) or at the grain boundaries or
corners due to the stress concentration by dislocation pileups.
196 G. Chai

FIGURE 2. Transition mechanisms of the fatigue crack initiation from the surface, subsurface to
subsurface non-defect fatigue crack origin; (a). Extrusion and intrusion of slip bands at the
specimen surface; (b). Slip band formed near the subsurface defect; (c). Short cracks formed in the
softer phase.

The formation of SNDFCO is a crack propagation process, which is controlled by


microstructure mechanics. The crack propagation transition from stage I to stage II leads to the
formation of SNDFCO. The size and morphology of the SNDFCO strongly depend on the applied
stresses. The size of the SNDFCO increases and their fracture morphology changes from more
ductile to facet when the applied stresses decrease.
Key words: Fatigue crack initiation, Subsurface non-defect fatigue crack origin (SNDFCO),
Fracture, Two phase alloys, Dislocation slip band.

References
1. Murakami, Y., Metal Fatigue: Effect of small defects and non-metallic inclusion, Elsevier
Science Ltd, 2002.
2. Chai, G., In Proceedings of the Third International Conference on Very High Cycle fatigue,
edited by T. Sakai and Y Ochi, Kusatsu, 2004, 24-31 and 374-381.
2T5. Fatigue and fracture 197

SURFACE FATIGUE OF GEAR TEETH FLANKS

G. Fajdiga, M. Sraml and J. Flasker


University of Maribor, Faculty of Mechanical Engineering
Smetanova ulica 17, SI-2000 Maribor, Slovenia
gorazd.fajdiga@uni-mb.si

Mechanical behaviour of various machine elements, such as gears, brakes, clutches, rolling
bearings, wheels, rails, and screw and riveted joints, are influenced by interaction between contact
elements and surfaces. Surfaces in rolling and/or sliding contact are exposed to material contact
fatigue. Contact fatigue can be defined as a kind of damage caused by changes in the material
microstructure which results in crack initiation followed by crack propagation, under the influence
of time-dependent rolling and/or sliding contact loads. The contact fatigue process can in general
be divided into two main parts:
• initiation of micro-cracks due to local accumulation of dislocations,high stressesat local
points, plastic deformation around inhomogeneous inclusions or other imperfections in or
under the contact surface;
• crack propagation, which causes permanent damage to a mechanical element.
In this paper the pitting phenomenon of gears is addressed and the developed numerical model
is used for determination of pitting resistance, i.e. the service life of gear teeth flanks. The
initiation of fatigue cracks represents one of the most important stages in the pitting process. The
position and mode of fatigue crack initiation depends on the microstructure of the material, the
type of the applied stress and micro- and macro-geometry of the specimen, Cheng [1]. The crack
initiation periods can be very different and cracks can be initiated either on or under the surface,
depending on a different combination of rolling and sliding contact conditions, Kaneta [2].
In general, gear teeth pitting may be initiated on sub-surface or surface. Sub-surface pitting
initiation can be observed in high quality gears made of alloy steel, with smooth contact surfaces
and good lubrication, where the large shearing sub-surface stresses due to contact loading initiate
substantial dislocation motion, which governs the crack initiation process, Glodež [3]. Surface
pitting initiation is common in industrial gears, which have rougher surfaces and are made of
ordinary construction steels. The surface pitting is strongly influenced by surface roughness and
other surface defects, like machining marks, large notches, inclusions, etc. The surface cracks may
also appear as a consequence of thermal treatment of the material due to residual stresses.
This paper considers only the second mechanism of gear pitting, i.e. surface pitting. The
process of surface pitting can be visualized as the formations of small surface initial cracks grow
under repeated contact loading. Eventually, the crack becomes large enough for unstable growth to
occur, which causes the material surface layer to break away. The resulting void is a surface pit
(Fig. 1).
The number of stress cycles N required for pitting of a gear teeth flank to occur can be
determined from the number of stress cycles Ni required for the appearance of the initial crack in
the material and the number of stress cycles Np required for a crack to propagate from the initial to
the critical crack length, when the final failure can be expected to occur:

N Ni  N p
(1)
198 G. Fajdiga et al.

FIGURE 1. Typical surface pits on gear tooth flank.

This paper describes a computational model for contact fatigue crack initiation and crack
propagation in the contact area of gear teeth flanks. The purpose of the present study is to present,
firstly, a model for prediction of contact fatigue initiation, which is based on continuum
mechanics, real cyclic contact loading and specific material fatigue parameters. The material
model is assumed as homogeneous, without imperfections such as inclusions, asperities,
roughness, residual stresses, etc., as often occur in mechanical elements. A moving contact load is
often used for simulation of the cyclic loading in fatigue crack initiation and propagation analyses
on mechanical elements (simulation of meshing of gears) Fajdiga [4]. The second part of the
computational model is a crack propagation model based on appropriate short fatigue crack growth
theories Navarro [5], Sun [6]. The model attempts to account for the different parameters
influencing the crack propagation process (Hertzian contact pressure, friction between contacting
surfaces, fluid trapped in the crack, meshing of gears, etc,) leading to pitting, starting from the
initial surface fatigue crack to the critical crack length, when the occurrence of a surface pit is
expected. The results of the computations can be used to predict the time required for the
development of pits, i.e. the service-life of gears with regard to pitting [4].

References
1. Cheng W., Cheng HS., Mura T., Keer LM. Micromechanics modelling of crack initiation
under contact fatigue. ASME Journal of Tribology; 116: 2-8, 1994.
2. Kaneta M., Yatsuzuka H., Murakami Y. Mechanism of crack growth in lubricated rolling/
sliding contact. ASLE Transactions; 28: 407-414, 1985.
3. Glodež S., Ren Z., Flašker J. Simulation of surface pitting due to contact loading.
International Journal for Numerical Methods in Engineering; 43: 33-50, 1998.
4. Fajdiga G., Flašker J., Glodež S. and Ren Z. Numerical simulation of the surface fatigue crack
growth on gear teeth flanks, Journal of Mech. Engineering 46(6), 359-369, 2000.
5. Navarro A. and Rios E.R. Short and long fatigue crack growth- a unified model,
Philosophical Magazine, 57, 15-36, 1988.
6. Sun Z., Rios E.R. and Miller K.J. Modelling small fatigue cracks interacting with grain
boundaries, Fatigue Fract. Engng Meter, 14, 277-291, 1991.
2T5. Fatigue and fracture 199

FATIGUE AND FRACTURE PROCESSES IN HIGH PERFORMANCE PM


TOOL STEELS

G. Jesner, S. Marsoner1, I. Schemmel2 and R. Pippan3


Materials Center Leoben Forschung GmbH
Franz-Josef-Strasse 13, 8700 Leoben, Austria
gerhard.jesner@mcl.at
1Materials Center Leoben Forschung GmbH

Franz-Josef-Strasse 13, 8700 Leoben, Austria


marsoner@mcl.at
2Boehler Edelstahl Ges.m.b.H

Mariazeller Strasse 25, 8605 Kapfenberg, Austria


ingrid.schemmel@bohler-edelstahl.at
3Erich Schmid Institute for materials science of the Austrian academy of sciences

Jahnstrasse 12, 8700 Leoben, Austria


pippan@unileoben.ac.at

For cold forging applications often high performance PM – tool steels were applied. The tools are
cyclically repeated loaded to very high stresses by the forging process. The stresses are often
significantly larger than the yield stress. This will result in a local plastic deformation, crack
initiation, crack propagation and finally to the failure of the tools.
It is important to know the fatigue behaviour of these types of steels to get information about
the life time of such tools.
From microstructural point of view the investigated steels can be considered as MMC with
primary carbides with the size of a few microns and a martensitic matrix consisting of secondary
carbides. By heat treatments the strength of the matrix can be varied over wide range.
Microstructures with low and high primary carbide content and an extreme variation in
hardness – i.e. a large variation in the matrix microstructure are investigated.
The fatigue crack propagation curve and the threshold of stress intensity range fatigue crack
growth curves were determined for different stress ratios in such tool steels with different
microstructures and heat treatment. Fig. 1 shows as an example the micrographs of two selected
steels.
The different fatigue and fracture mechanism and the influence of the design of the
microstructure as well as the mechanical properties of the matrix will be discussed.

FIGURE 1. Microstructure of two selected high performance tool steels.


2T5. Fatigue and fracture 201

NOTCH AND DEFECT SENSITIVITY OF ADI IN TORSIONAL FATIGUE

B. Atzori and G. Meneghetti


Department of Mechanical Engineering, University of Padova
Via Venezia, 1 – 35131 Padova (Italy)
giovanni.meneghetti@unipd.it

Recently Atzori et al. [1,2] analysed the sensitivity to defects and to standard notches in axial
fatigue of metallic materials commonly adopted in manufacturing engineering components and
structures, like steels and aluminium alloys. As a result of the proposed approach, the effect of any
kind of geometrical discontinuity on the fatigue limit can be treated, such as small defects, cracks,
sharp notches or crack-like notches and standard rounded notches characterised by an arbitrary
notch tip radius and notch opening angle. Two material parameters are needed, i.e. the material
fatigue limit and the threshold value of the Stress Intensity Factor.

FIGURE 1. plot of the torsional fatigue limit for components weakened by V-notches.

In the present work such an approach has been extended from axial fatigue to torsion fatigue
and can be summarised by means of the diagram presented in Fig. 1, where the fatigue limit in
terms of nominal shear stress range referred to the gross section 'Wg,th is reported as a function of
an effective dimension of the component aeff defined as:

1
§ K 3 · 1O3
aeff ¨ ¸
¨W ¸
© g ¹ (1)
being K3 the Notch-Stress Intensity Factor for mode III loading evaluated for the same
component but imposing a notch tip radius equal to zero, Wg the nominal shear stress evaluated on
the gross section and O3 is the degree of singularity of the local stress field, which depends on the
notch opening angle. Three regions can be schematically singled out divided by two length
parameters, namely a 0V t , *V
which is a material parameter, and a t , which depends also on the
V
elastic stress concentration factor Ktg for torsion. If aeff < a 0 t, then the presence of a (small) notch

does not lower the fatigue limit with respect to the material torsional fatigue limit. If a 0V t<aeff
202 B. Atzori and G. Meneghetti

*V
<a t , the fatigue limit is determined by the condition that the N-SIF equals the threshold value,
*V
which is thought of as a material property, i.e. 'K3='K3,th. If aeff >a t , then the fatigue limit can
be predicted by applying the classic Notch Mechanics criterion, i.e. 'Wg,th='W0/Ktg. The smooth
V *V
transition zones around a 0 t and a t (see figure 1) characterize the sensitivity to defects and
notches, respectively, and in this work an austempered ductile iron was studied.

TABLE 1. Geometries of the tested series

Torsion fatigue tests have been conduced on both smooth and notched specimens according to
the geometries listed in Table 1. A total number of 6 geometries were taken into account for a total
number of 72 fatigue tests. The adopted load ratio, defined as the ratio between the minimum and
the maximum applied torque, was equal to –1. Results have been summarised in the diagram,
which describes the sensitivity of the considered material to defects and notches in torsional
fatigue. Additionally, observations conduced by means of the Scanning Electron Microscope
enabled us to investigate the early crack propagation phase. As an example, fig. 2 shows small
cracks initiated after 2˜106 cycles from a graphite nodule at the tip of a 5-mm-deep notch having a
root radius equal to 0.1 mm.

References
1. Atzori, B., Lazzarin, P. and Meneghetti, G., Fatigue Fract. Engng Mater. Struct, vol. 26, 257-
267, 2003.
1. Atzori, B., Lazzarin, P. and Meneghetti, G., Int. J. Fract., vol. 133, 61-87, 2005.
2T5. Fatigue and fracture 203

MULTI AXIAL FATIGUE IN WELDED COMPONENTS

G. Mesmacque, B. Wu1, C. Robin1, D. Zakrzewski1 and X. Decoopman


University of Science and Technology, Lille, France
1Ecole des Mines, Douai, France

gerard.mesmacque@univ-lille1.fr

In order to study the fatigue behavior of 6000 aluminium alloy welded structures and to take into
account the small size of the affected zone, a specific heat treatment was developped to get a
structure equivalent to that of the heat affected zone (simulated HAZ). Such an equivalent structure
was determined by comparing their hardness, microstructure and tensile properties. A new
cumulative damage law has been developed taking in account the loading history .

TABLE 1 Mechanical characteristics for simulated HAZ

FIGURE . Simulated Heat affected Zone Choice of Heat treatment

Several existing multiaxial fatigue criteria (Dang Van K.[1], Findley[2], Matake [3],
McDiarmid [4]) have been evaluated under Castem 2000. The misalignment, the local geometry
and the residual stress have been taken in account in the numerical model.
204 G. Mesmacque et al.

FIGURE 2. Numerical analysis and fatigue tests-Geometry of specimen real -simulated HAZ
Specimen were tested in the High Cycle Fatigue regime, under uniaxial and in-phase and out-
of-phase biaxial loadings. The results of all simulations were compared to the test data. The
position of the critical plane given by the simulations was also compared to the test results. Some
criteria (Dang Van, K. & McDiarmid) were found to be conservative. The criterion of McDiarmid
was found to be the best whatever the loading, while the criteria of Dang Van and Matake were
found to give out life predictions not enough precise, the proportion between the test results and
the life assessments being sometimes larger than 10. The criterion of Findley did not agree well
with the test results.
The cumulative damaging law [4] developed by G. Mesmacque and al [4] was used in blocks
loading history. Some results are presented in the frame of this work.
A general discussion in the light of the multi axial criterion in biaxial loading (in phase and out
phase) is made and compared to the IIW recommendations.
Keywords : (HCF) High Cycle Fatigue, Multiaxial Fatigue Criterion, Heat Affected Zone,
Aluminium Welded Joints

References
1. Dang Van K., Sciences et techniques de l'Armement, 1973, N° 47, 3ème fascicule
2. Findley W.N. Transaction of the American Society of Mechanical Engineers, 1957, Vol . 79.
3. D.L. Mc Diarmid Journal, July 1974,PP. 325-329.
4. G. Mesmacque International Journal of fatigue Volume 27, Issue 4, Pages 455-469 April
2005
2T5. Fatigue and fracture 205

ENHANCED FATIGUE LIFE BY MECHANICAL SURFACE TREATMENTS –


EXPERIMENT AND SIMULATION

H. P. Gaenser, I. Goedor, H. Leitner and W. Eichlseder


Montanuniversität Leoben
A-8700 Leoben, Austria
{Hans-Peter.Gaenser|Istvan.Godor|Heinz.Leitner|Wilfried.Eichlseder}@mu-leoben.at

Although mechanical surface treatments are commonly used in engineering practice for their
beneficial effect on fatigue life, only few quantitative data are available from literature [1, 2]. This
contribution presents first results of a long-term program that aims at developing quantitative
design guidelines for the influence of mechanical surface treatment on fatigue life of components.
To this purpose, experimental work and numerical simulations are performed in parallel, giving a
detailed understanding of the effects of mechanical surface treatments on surface roughness,
residual stress distribution, and resulting fatigue strength.
Four-point rotating bending tests were performed on notched specimens of a quenched and
tempered low-alloy steel (34CrNiMo6) as well as of an aluminum wrought alloy (AlCuBiPb). The
specimens had a minimum diameter of 7.5 mm and a notch stress concentration factor of 1.8,
giving a generic geometry and stress gradient typical for usual rotating machine parts, cf. Fig. 1.
The heat-treated and machined specimens were subjected to two different mechanical surface
treatments: shot peening and deep rolling. In what follows, some preliminary results representative
for the underlying mechanisms and effects of surface treatments are presented.
Compared to the untreated specimens, deep rolling gives a very smooth surface and a fatigue
strength enhancement of as much as 27 %, cf. the S/N curve for 34CrNiMo6 in Fig. 1. Shot
peening results in a rough surface, but still in a fatigue strength enhancement of 18 %. In order to
test whether this discrepancy was due to the surface roughness of the shot peened specimens (Rmax
= 30 µm), some experiments with peened and subsequently polished specimens were conducted.
Contrary to the expected increase in fatigue strength due to polishing, and also contrary to the
findings reported in [2], the endurance limit of the peened and polished specimens was within the
scatter band of the peened, unpolished samples.

FIGURE 1. Rotating bending experiments (34CrNiMo6):


specimen (left), normalized S/N curves (right)

In order to gain a deeper understanding of the underlying mechanical processes, finite element
(FE) simulations of both surface treatments – deep rolling and shot peening – were performed. For
deep rolling, the entire specimen and process were modeled (Fig. 2). For shot peening, a
206 H. P. Gaenser et al.

representative volume element (RVE) was subjected to an impact of eight rigid particles, giving
approximately the same shot distribution as in experiment (Fig. 3).
For an assessment of the resulting residual stress distributions (Figs 2-3), the axial (rsp. in-
plane) stresses, compensating the bending stresses in rotating bending loading, are evaluated. The
maximum compressive residual stress is reached in a depth of approx. 0.7 mm in deep rolling of
aluminum, compared to 0.05 mm in shot peening of steel. One observes that, for the steel
specimens, the depth of the maximum residual stress due to shot peening is of the same order of
magnitude as the surface roughness, 0.03 mm. If the latter is removed by polishing, the residual
stress maximum will be reduced in order to restore overall stress equilibrium. In the present case,
the increase in fatigue strength due to a finer surface finish and the decrease due to a reduction of
the residual compressive stress are precisely compensating each other.

FIGURE 2. Finite element simulation of deep rolling (AlCuBiPb): FE mesh of specimen with tools
(left) and normalized resulting residual stress distribution in the notch ground (right)

FIGURE 3. Finite element simulation of shot peening (34CrNiMo6): representative volume


element with resulting surface structure (left) and normalized residual stress distribution (right)

References
1. Torres, M.A.S, and Voorwald, H.J.C., Int. J. Fatigue, vol. 24, 877-886, 2002
2. Schreiber, R., Untersuchungen zum Dauerschwingverhalten kugelgestrahlter Einsatzstähle –
Abschlussberich (Experimental investigation of the fatigue limit of shot-peened case-
hardened steels – final report) (in German), Universität Karlsruhe, 1975
2T5. Fatigue and fracture 207

ANALYSIS OF REPAIRED ALUMINUM PANELS IN GENERAL MIXED-MODE


CONDITIONS

H. Hosseini-Toudeshky, M. Saber2 and B. Mohammadi3


Associate Professor, 2 & 3 Post Graduate Student
Aerospace Engineering Department, Amirkabir University of Technology, No. 424, Hafez Ave.,
Tehran, Iran
hosseini@aut.ac.ir

In the real applications, most of the cracked components are subjected to cracks in general mixed-
mode condition. A typical geometry and loading of single-side repaired panels in mixed-mode
conditions is shown in Fig. 1. A single-sided repair produces out-of-plane bending, which causes
stress variations over the thickness of the cracked panel. One of the most challenging aspects of
bonded composite repair technology has been the stress analysis of repaired panels and the
calculations of subsequent fracture parameters. The stress variations over the thickness of a
cracked plate in un-symmetric repair present a greater challenge in modeling due to the existence
of out-of-plane bending and developing highly complicated three dimensional stresses.
Non-uniform crack propagation along the panel thickness is expected due to the above reasons.
In the previously published works [1], the authors assumed that the crack-front remains
perpendicular to the panel surfaces during the crack propagation of the repaired panels in mixed-
mode conditions. But, experimental results show that crack growth rates are not uniform along the
thickness and the actual crack-front becomes curved shape as shown in Fig. 2. This figure also
shows that the crack surfaces are three dimensional. Therefore, fracture parameters analysis of the
repaired panels in general mixed-mode conditions are required to predict the crack re-initiation and
propagation lives and crack trajectories.
In this paper, three-dimensional finite element analysis of the repaired aluminum panels (AL
2024 T3) in general mixed-mode conditions using glass/epoxy composite patches with different
lay up configurations are performed. Theses analyses are based on linear elastic fracture mechanics
assumptions. Variation of stress intensity factors in repaired panels with composite patch along the
thickness in the mixed modes conditions is studied. The difficulty of this work arises from the
complexity of modeling and meshing. The sub-modeling technique is used and then a fine mesh is
generated at the crack-front area to calculate the fracture parameters.
The computational fracture analysis is based on the calculation of separated energy release
rates (SERRs) by the aid of the modified virtual crack closure technique (MVCCT) in order to
calculate the local SERR distributions along the crack-front. MVCCT has the advantageous of: 1-
ability to decouple of three modes of fracture 2- simplicity of formulation, but it imposes a regular
mesh around the crack-front.
Variation of stress intensity factors versus distance from the patch surface along the initial
crack-front has been shown in Fig. 3 for a typical patch lay up configuration. The variations and
magnitudes of the stress intensity factors in different modes show that none of them can be
neglected in crack re-initiation and propagation lives and crack-front configuration. These
behaviors are considerably changed with the changes of patch layers orientation and lay up
configuration which will be discussed in the full paper.
208 H. Hosseini-Toudeshky et al.

FIGURE 1. Typical geometry and loading of the repaired panels

FIGURE 2. Crack-front and crack surface configuration obtained from experiments

FIGURE 3. Variations of stress intensity factors along the crack-front (E=45°)

References
1. Hosseini-Toudeshky H., Mohammadi B. and Daghyani HR., “Mixed-Mode fracture analysis
of aluminum repaired panels using composite patches”, Accepted for publication in the
journal of composites science and technology (special issue dedicated to ETDCM-6), 2005.
2T5. Fatigue and fracture 209

EFFECT OF STRAIN RATE ON FATIGUE BEHAVIOR OF ULTRAFINE


GRAINED COPPER

P. Gabor, H.J. Maier and I. Karaman1


Lehrstuhl für Werkstoffkunde, Universität Paderborn,
Pohlweg 47-49, 33098 Paderborn, Germany
1Department of Mechanical Engineering, Texas A&M University,

College Station, TX 77843, U.S.A.


hmaier@zitmail.uni-paderborn.de

In recent years, ultrafine grained (UFG) materials have received a great deal of attention, as they
exhibit superior combinations of high strength and enhanced ductility as compared to conventional
large grained alloys, c.f. Valiev et al. [1], Love and Valiev [2]. Until recently, most studies that
focused on rate sensitivity and work hardening behavior of UFG materials at different strain rates
were conducted under monotonic test conditions, e.g. Wang and Ma [3] and Li et al. [4]. However,
for many of the envisaged structural applications of these materials, cyclic properties and their
dependence on loading parameters will be a key issue.
The experiments were performed on oxygen-free high conductivity copper with 99.995%
purity acquired from Hitachi Cable Ltd..The UFG microstructures were obtained by equal channel
angular extrusion (ECAE). Small dog-bone type samples were electro-discharged machined from
the ECAE billets, which underwent up to 16 ECAE passes. In the low-cycle fatigue tests, strain
rate was varied between 1.2-1 s-1 and 1.2-5 s-1, and a total strain amplitude ('t/2) of 1 x 10-3 was
employed in all LCF tests. In addition to the as-ECAEd material, samples that had received an
additional heat treatments at 200 °C for 3 minutes and 170 °C for 2 hours, respectively, were
studied.
The ultrafine grained copper was found to be strain rate dependent at room temperature in all
tested microstructural conditions. The highest sensitivity to strain rate changes was observed for
the as-ECAEd material (Figure 1), where the stability of the microstructure is expected to be the
lowest. Interestingly, the stress amplitude and LCF life both increase at higher strain rate for all
specimens. The detrimental effect of lower strain rates on fatigue life is attributed to changes in the
deformation mechanisms.
210 P. Gabor et al.

FIGURE 1. Stress amplitude as a function of strain rate for LCF tests conducted at a total strain
amplitude of 1 x 10-3.

In order to shed light on the effect of the strain rate on the evolution of fatigue damage,
additional in-situ fatigue tests were performed using a small-scale load frame placed within an
environmental scanning electron microscope. The evolution of the surface features was inspected
periodically at given cycle intervals. The observed operating damage mechanisms were correlated
with the effect of strain rate on macroscopic stress-strain response and on resulting fatigue life.
Optimized heat treatments and ECAE process routes are discussed with respect to microstructural
stability and resulting fatigue life.

References
1. Valiev, R.Z., Kozlov, E.V., Ivanov, Y.F., Lian, J., Nazarov, A.A., Baudelet, B., Acta Metall.
Mater., vol. 42, 2467-2475, 1994.
2. Lowe, T.C., Valiev, R.Z., In: Investigations and Applications of Severe Plastic Deformation,
NATO Science Series, v.3 80, Kluwer Publishers, Dordrecht, 2000.
3. Wang, Y.M., Ma, E., Mater. Sci. Eng., vol A 375–377, 46–52, 2004.
4. Li, Y.J., Zeng, X.H., Blum, W., Acta Mater., vol. 52, 5009–5018, 2004.
2T5. Fatigue and fracture 211

LUBRICANT EFFECTS ON PROPAGATION OF SURFACE-BREAKING


CRACKS UNDER ROLLING CONTACT LOADING

J. Lai and S. Ioannides


SKF RDC B.V., P.O. Box 2350, 3430 DT Nieuwegein, The Netherlands
Junbiao.Lai@skf.com

It has been well recognized that lubricants or fluids may have significant effects on propagation of
surface-breaking cracks under rolling contact loading. Various models have been proposed in the
literature [1,2] to describe different possible cracking mechanisms associated with the effects of
lubricant inside the crack, including fluid-lubricated crack faces, fluid seepage into the crack by
load, and fluid entrapment. To date, the solution to one fundamental problem is still lacking, i.e.
the amount of lubricant forced into the crack by load, and /or the amount of lubricant entrapped in
the crack. This is a problem of special technical relevance for bearings and gears, which operate
under oil- or grease-lubricated condition, since the viscosity of lubricant is largely dependent on
pressure and temperature. The lubricant pressure exerted on the crack faces depends, among
others, on the amount of lubricant seeping into or entrapped inside the crack. The amount of
lubricant seeping into the crack depends on the pressure at the crack mouth and the pressure-
dependent viscosity of the lubricant.

FIGURE 1. Schematic illustration of lubricant seeping into a slit by Hertzian pressure.

In the present paper an attempt is made to determine the amount of lubricant forced into a
surface-breaking crack by contact pressure based on the solution for viscous fluids in laminar flow.
Considering a 2D configuration as shown in Fig. 1, a close-form solution is derived of the lubricant
seepage depth L inside the slit under a moving Hertzian pressure p with speed V,

2 2b p( x)
L( x) B
3V ³0 P ( x)
dx
(1)
in which b is the half width of the Hertzian pressure, B is the opening displacement of the crack
mouth, and P is the viscosity of the lubricant. The temperature- and pressure-dependent viscosity
of fluids is represented by the Roelands relation. Equation (1) relates the amount of lubricant
seeping into a slit/crack to the load, rolling speed and viscosity of lubricant. Numerical
(integration) results show that the lubricant seepage depth decreases with load, owing to the
significant increase of the viscosity of the lubricant with pressure.
212 J. Lai and S. Ioannides

The seepage model (1) is implemented in finite element calculations to simulate lubricant
seeping, entrapment and pressurization in the crack during rolling of a roller over the crack. The
amount of lubricant seeping into the crack determines if the crack mouth will be closed or if the
lubricant will be entrapped in the crack. Traction on rolling contact surface and friction between
crack faces are taken into account in the FE calculations. Fig. 2 shows the FE simulation of
lubricant entrapment, pressurizing and releasing from the crack when a load is moving over the
crack. Stress intensity factors (KI and KII) at the crack tip are calculated, based on which crack
propagation in mixed mode is discussed.

FIGURE 2. Snapshots from FE simulation of lubricant entrapment in crack.

References
1. Bower, A.F., J. Trobology. vol. 110, 704-711, 1988.
2. Kaneta, M. and Murakami, Y., J. Tribology, vol. 113, 1991
3. R. Byron Bird, Warren E. Stewart and Edwin N. Lightfoot: "Transport Phenomena", John
Wiley & Sons, 1960, New York.
2T5. Fatigue and fracture 213

COMPUTATIONAL MODELLING OF CRACK INITIATION IN A MIXING TEE


SUBJECTED TO THERMAL FATIGUE LOAD

I. Varfolomeyev
Fraunhofer-Institut für Werkstoffmechanik, Wöhlerstr. 11, 79108 Freiburg, Germany
igor.varfolomeyev@iwm.fhg.de

Thermal fatigue phenomenon is an important subject in the life-time management of piping


components of nuclear power plants. Especially in T-joints, high frequent temperature fluctuations
as a result of mixing of the hot and cold fluids may cause material degradation and thus affect the
component fatigue life. The European research project THERFAT was initiated to study the
significance of thermal fatigue issues and to improve analysis tools for structural integrity
assessments under thermal fatigue conditions, Metzner et al. [1].
Within the THERFAT consortium involving 16 research teams from 8 European countries,
several key aspects of the thermal fatigue evaluation were investigated. These included collation of
field experience, load determination, material characterization, integrity assessment, verification
tests. The overview and general approach to solving the problem are presented in [1], whereas
some results of investigations are provided in [2, 3].
This paper concerns with numerical computations of the stress and strain evolution in an
austenitic mixing tee subjected to thermal fluctuations, as produced by two mass flows with
different temperature and flow rates. Based on the stress and strain analysis results, predictions are
then made for fatigue crack initiation in the component.
Input data for the thermal-mechanical analyses were obtained through temperature history
measurements performed by another project partner (SPG, Dresden) on thin-walled models. Figure
1 gives an example of records for the fluid temperature at the inner surface in the mixing zone and
at the outer surfaces of the pipe. The temperature difference 'T in those experiments was about
90°C. In view of difficulties to define exact values of the heat transfer coefficient (D) and the
frequency of thermal fluctuations (f) for in-service conditions, the influence of D and f on the stress
state in a pipe was examined in a parameter study.
To investigate possible component behaviour, a fictitious mixing tee with the wall thickness of
10 mm, the outer diameter of 60 mm and the temperature difference up to 'T = 250°C was
investigated. Since no direct measurements of the thermal-hydraulic data for in-service conditions
are available, the above experimental results were scaled and used as boundary conditions in the
stress analyses.
Calculations revealed that local plastic deformations in the mixing area would be possible at
certain combinations of the thermal load parameters D, f and 'T. To properly describe elastic-
plastic stress and strain behaviour under cyclic loading, a Chaboche type material model after
Jiang [4] was employed taking into account combined isotropic and kinematic strain hardening, as
well as strain rate effects. The model has previously been adopted (Sester and Chauvot [5]) to
describe the ratchetting phenomena in a pipe subjected to thermal stratification.
Based on the results of the stress-strain calculations (Fig. 2) and making use of engineering
models describing short crack behaviour (see [5]), the crack initiation in the component can be
predicted.
214 I. Varfolomeyev

FIGURE 1. Fluid and outer wall temperature in the mixing zone.

FIGURE 2. Example of calculated stress versus strain evolution.

References
1. Metzner K.J., Braillard, O., Faidy, C., Marcelles, I., Solin, J., Stumpfrock, L., In Proceedings
of the 30th MPA-Seminar, Stuttgart, 2004, 17.1-17.12.
2. Ertem-Müller, S., Gerstner, L., Schippers, M., In Proceedings of the 3rd International
Conference on Fatigue of Reactor Components, Seville, 2004.
3. Paffumi, E., Nilsson, K.-F., Taylor, N.G., In Proceedings of the 2nd International Symposium
on Fatigue and Fracture Mechanics, Tampa, ASTM/ESIS, 2003.
4. Jiang, Y., Ph. D. Thesis, University of Illinois, Urbana-Champaign, 1993.
5. Sester, M., Chauvot, C., In Proceedings of the 26th MPA-Seminar, Stuttgart, 2000, 3.1-3.9.
2T5. Fatigue and fracture 215

ESTIMATION OF CRITICAL STRESS INTENSITY FACTOR IN STEEL


CRACKED WIRES

J. Toribio, F. J. Ayaso, B. Gonzalez, J. C. Matos and D. Vergara


Department of Materials Engineering, University of Salamanca
E.P.S., Campus Viriato, Avda. Requejo 33, 49022 Zamora, Spain
Tel: (34-980) 54 50 00; Fax: (34-980) 54 50 02
toribio@usal.es

In the framework of damage tolerance analyses in structural engineering, a problem of major


concern is the formulation of a fracture criterion useful for engineering design against catastrophic
failure. This is particularly important in the case of high-strength structural members in the form of
bars, wires, strands, tendons and cables (e.g., high-strength steels for reinforcing and prestressing
concrete) which are axially loaded under very severe loads and frequently work in harsh
environments, so that they can suffer fatigue, stress corrosion cracking or corrosion-fatigue, thus
increasing the risk of catastrophic failure and reducing the structural life [1-3]
Two fracture criteria for prestressing steel cracked wires are considered. Firstly, a global
fracture criterion [4-6] is formulated on the basis of energetic considerations (strain energy release
rate concept) or compliance measurements. It only requires a single-parameter approach
accounting for the crack depth (related to the compliance), or an average value of the energy
release rate along the crack front, as follows:
G* = 12s-s+sKI2E'ds (1)
where s is the curvilinear coordinate over the crack line. Thus the SIF is:

KI* = [0.473 – 3.286 (a/D) + 14.797 (a/D)2]1/2 [(a/D) – (a/D)2]–1/4 V Sa (2)
where Vis the axial remote stress and a the crack depth.
A local fracture criterion [6-8] is also considered, according to which fracture takes place when
the maximum SIF along the crack line reaches a critical value, i.e., a local instability at a certain
point produces the catastrophic fracture of the overall flaw:

S up * K I (s) = K IC (3)
where * is the crack line. The SIF is in this case:
4 3
KI** = ¦ ¦
i 0 j 0
Cij (a/D)i (a/b)j V Sa 
iz1
(4)
where the coefficients Cij are given in [9] and a, b are the dimensions of the crack ( elliptical
shape).
In this paper an engineering estimation of the critical SIF is proposed for eutectoid steel
cracked wires subjected to axial loading. The method for estimating the critical SIF consists of
computing it by considering, in addition to the fatigue precrack, the extension of the subcritical
crack propagation before the final fracture situation of the catastrophic type.
Fig. 1 shows the critical SIF for the fracture tests in air using fatigue precracked samples taken
from eutectoid steel wires with different levels of cold drawing, where it is seen that cold drawing
216 J. Toribio et al.

clearly improves the fracture behaviour of the steels, in addition to the improvement of classical
mechanical properties (yield strength) which is the final aim of manufacturing.

FIGURE 1. Critical stress intensity factor for the fracture tests in air.

References
1. Elices, M., Llorca, J. and Astiz, M.A., in Handbook of Fatigue Crack Propagation in Metallic
Structures (edited by A. Carpinteri) pp. 191-220, Elsevier, Amsterdam (1994).
2. Barsom, J.M. and Rolfe, S.T., Fracture & Fatigue Control in Structures, Prentice-Hall, NJ,
(1987).
3. Toribio, J. and Valiente, A., Proceedings of the 13th European Conference on Fracture (ECF-
13), San Sebastián, Spain, 2000.
4. Elices M., in Fracture Mechanics of Concrete: Structural Application and Numerical
Calculation (edited by G.C. Sih and A. DiTommaso), pp. 226-271, Martinus Nijhoff,
Dordrecht (1985).
5. Athanassiadis A., Boissenot J.M., Brevet P., Francois D. and Raharinaivo A., Int. J. Fracture
vol. 17, 553-566, 1981.
6. Valiente, A., Ph. D. Thesis, Polytechnic University of Madrid, 1980.
7. Astiz M.A., Elices M. and Valiente A., in Fracture Control of Engineering Structures/ECF6
(edited by H.C. Van Elst and A. Bakker), pp. 65-74, EMAS, West Midlands (1986).
8. Bui H.D. and DangVan K., J. Mécanique Appliquée, vol. 3 (2), 205-225, 1979.
9. Astiz M.A., Int. J. Fracture, vol. 31, 105-124, 1986.
2T5. Fatigue and fracture 217

LOW-CYCLE FATIGUE OF DIN 1.2367 STEELS IN VARIOUS TREATMENTS

C. C. Liu, J. H. Wu2 and C. C. Kuo


Department of Mechanical Engineering, Chien Kuo Technology University
No. 1, Jieshou N. Rd., Changhua City, Changhua County 500, Taiwan, R.O.C.
2Automotive Research and Testing Center
2No. 6, Lugong S. 7th Rd., Lugang Town, Changhua Country 505, Taiwan, R.O.C.

ccl@ctu.edu.tw, JuiHung@artc.org.tw, cckuo@ctu.edu.tw

DIN 1.2367 (X38CrMoV5) is currently one of the most often used chromium-molybdenum tool
steels, but which contains a higher molybdenum content. The higher molybdenum content gives
DIN 1.2367 better hot hardness and temper resistance compared to similar-type AISI H11 and H13
tool steels. The steel DIN 1.2367 has become of increasing importance for various applications,
such as gears, shear blades, mandrels punches and tooling dies, due to its high hot hardness and
toughness, and good resistance to softening, wear and thermal fatigue.
The mechanical, friction and wear properties of DIN 1.2367 tool steels have been extensively
investigated as a function of tempering treatment and associated microstructure [1,2]. Little work
has been done on the strain-controlled low-cycle fatigue (LCF) behaviour of this steel. As some
applications of DIN 1.2367 tool steels are subjected to high cyclic plastic deformation which
causes micro chipping and fracture in service, it is important to characterize the LCF behaviour of
this steel in order to better predict the fatigue lifetime for components made from this material.
The objective of the present study is, macroscopically and microscopically, to characterize the
effects of tempering and plasma-nitriding treatments on the LCF behaviour of DIN 1.2367 tool
steels by systematic experiments. Appropriate fatigue life prediction approaches are proposed to
correlate the LCF lifetime data obtained for this DIN 1.2367 steels under various treatment
conditions. Literatures related to this work are listed in references 1-7.
The LCF behaviour has been investigated for DIN 1.2367 tool steel in seven different
conditions including tempering and plasma-nitriding treatments. The LCF results indicated that
the fatigue life for the three tempered conditions took the following sequence: 580oC temper >
540oC temper > 620oC temper. For the three tempered conditions with plasma-nitriding, the
differences of fatigue life were significantly decreased and the strain-life curves almost merged
together. Furthermore, due to the formation of brittle Fe4N compounds on the specimen surface
easily resulting in crack initiation, the fatigue life was apparently decreased after plasma-nitriding
for each tempered condition. Fractography observations indicated that all of the fatigue fracture
surfaces for the plasma-nitrided specimens exhibited a common fracture mode dominated by brittle
cracking with no obvious crack initiation sites. In addition, the LCF lifetime data generated under
seven different treatment conditions for the given DIN 1.2367 tool steel are well correlated with a
yield-strength-normalized Smith-Watson-Topper (SWT) parameter in a log-log linear relationship,
proposed by Lin and Chu [7], for which the correlation coefficient, r2, is up to 0.83.

References
1. Mebarki, N., Delanges, D., Lamesle, P., Delmas, F. and Levaillant, S., Mater. Sci. Eng. A,
vol. 387-389, 171-175, 2004.
2. Barrau, O., Boher, C., Gras, R. and Rezai-Aria, F., Wear, vol. 255, 1444-1454, 2003.
218 C. C. Liu et al.

3. ASTM E606-80, Annual Book of ASTM Standards, vol. 3.01, 609-615, 1991.
4. Bannantine, J. A., Comer, J. J. and Handrock, J. L., Fundamentals of Mental Fatigue
Analysis, Prentice-Hall Press, New Jersey, 1990.
5. Nakagawa, H. and Miyazaki, T., J. Mater. Sci., vol. 34, 3901-3908, 1999.
6. Mar, S., Li, Y. and Xu, K., Surf. Coat. Tech., vol. 137, 116-121, 2001.
7. Lin, C. K. and Chu, C. C., Fatigue Fract. Eng. Mater. Struct., vol. 23, 545-553, 2000.
2T5. Fatigue and fracture 219

IMPACT TESTING A CAPABLE METHOD TO INVESTIGATE THE FATIGUE


RESISTANCE

K. David, P. Agrianidis, K. G. Anthymidis1 and D. N. Tsipas2


Department of Mechanical Engineering, Technological University of Serres,
Terma Magnesias str., GR 62124, Serres, Greece
1Serres Applied Research Center, Terma Magnesias str., GR 62124, Serres, Greece
2Department of Mechanical Engineering, Aristotle University of Thessaloniki,

54006, Thessaloniki, Greece


kanth@teiser.gr

The modern power generation steam turbines are being designed to have higher efficiencies and to
meet the stringent environmental regulations, ensuring plant reliability, availability and
maintainability without compromising cost. High efficiencies can be achieved by higher
temperatures. Therefore, the operating temperature is expected to rise from 550oC to 650oC and
from the material perspective to implement turbine components protected by spallation and
oxidation resistant coatings.
To guarantee the reliability of coated steam turbines components used in power plants, the
lifetime assessment of the coatings and their failure prediction become very important.
Microhardness, scratch adhesion and pin-on-disc sliding tests are commonly used for rapid
evaluation of the mechanical properties of coatings. However, they do not model the dynamic
cyclic impact fatigue. Recently the impact test method has been introduced as a convenient
experimental technique to evaluate the fatigue strength of coatings being exposed in alternate
impact loads, Bouzakis et al. [1]. According to this method a coated specimen is exposed to a
cyclic impact load. The superficially developed Hertzian pressure induces a complex stress field
within the coating, as well as in the interfacial zone. Both these stress states are responsible for
distinct failure modes such as a cohesive or adhesive one. The exposure of the layered compounds
against impulsive stresses creates the real conditions for the appearance of coating fatigue
phenomena based upon structural transformation, cracking generation and cracking growth, which
are responsible for the gradual microchipping and the degradation of the coating.
The objective of this experimental study was to investigate the influence of the impact stress
fields on the performance and fatigue strength of thermal spray HVOF coatings. Furthermore, the
overall aim of the current research was to prove the reliability of the impact test, as a new testing
method, to assess the coating lifetime against fatigue, to interpret the failure modes of coatings,
and thereby to exam its capability, whether this non-standard test can be used as a useful method
for the development and optimisation of fatigue resistant coatings working under impact loading.
Figure 1 shows the impact test rig where the experiments have been conducted.
220 K. David et al.

FIGURE 1. The impact test rig.


The stress strain problem related to the impact test is the Herzian contact, which develops
between the spherical indentor (carbide ball) and the examined layered space. The contact load
leading to coating fatigue fracture was recorded in fatigue-like diagrams (endurance strength
curves) versus the number of impacts. Gradual intrinsic coherence release and coating
microchipping or abrupt coating fracture and consequent exposure of the substrate material
designate the coating failure. The coating failure mode and its extent were assessed by SEM
observations and EDX analysis. In case of relatively tough coating microstructure with high wear
resistance as the WC-CoCr thermal spray coating is, the coating layer sustains the cyclic impacts
without any sign of cohesive delamination failure. Instead of that, only superficial abrasive wear
and spalling failure has been observed (Fig 2). This behaviour can be attributed to the improved
fracture toughness of this coating.

FIGURE 2. WC-CoCr coating failure initiation (cohesive failure mode) and microhardness
measurement of the layered compound.

References
1. Bouzakis, K., Vidakis N. and David, K., J. of Thin Solid Films, vol. 351, 1-8, 1999.
2T5. Fatigue and fracture 221

COMPARATIVE ASSESSMENT OF FATIGUE-THRESHOLDS ESTIMATED BY


SHORT AND LONG CRACKS

K. K. Ray, N. Narasaiahb and S. Tarafderb


Department of Metallurgical and Materials Engineering, Indian Institute of Technology,
Kharagpur-721302, India,
b National Metallurgical Laboratory, Jamshedpur-831007, India.

kkrmt@metal.iitkgp.ernet.in, nn@nmlindia.org, star@nmlindia.org

The safe life design philosophy of structural components is based on the knowledge of the fatigue
threshold for long cracks ('Kth). The emergence of the concept of short cracks, its numerous
experimental verification and the continued studies on its growth behaviour over the past years
have brought forward challenges to the basis of the safe life design philosophy. However, short
cracks interestingly exhibit either single or multiple-thresholds. But prior investigations related to
the determination and examinations of fatigue threshold for short cracks ('Kthsc) in structural
materials are limited in number. In an earlier communication [1], the largest value amongst the
detected thresholds associated with short cracks has been referred as “near long crack fatigue
threshold” (NLFTH) and the magnitude of NLFTH has been hypothesized as close to the fatigue
threshold obtained from long crack experiments as illustrated using Fig.1. This investigation aims
to explore the magnitudes of 'Kthsc and 'Kth for a few steels, to determine the NLFTH values for
these materials and finally to make a comparative assessment of NLFTH with their corresponding
'Kth values.

Fig.1 Schematic view of the critical crack length at the transition between short and long crack and
their fatigue thresholds. The points B and C are the thresholds for short and long cracks. The point
D (in part a) or B and C (in part b) indicates the transition length [1].
Experimental investigations on short crack growth behaviour and measurement of long crack
fatigue thresholds have been carried out using some recently developed specimens [1-3] which can
be coupled to rotating bending machine for the convenience of carrying out fatigue tests at
relatively high frequency. Schematic configurations of these specimens are shown in Fig.2.

Fig.2 Configuration of specimens used for (a) short crack and (b) long crack growth experiments.
222 K. K. Ray et al.

The experiments have been carried out on four plain carbon steels and the generated data of
crack growth rates against stress intensity factor range have been analyzed to examine the
threshold values, 'Kth, 'Kthsc and NLFTH. The magnitudes of 'Kth have been estimated by load
shedding procedure whereas the magnitudes of 'Kthsc have been evaluated from the short crack
growth behaviour. An alternative set of experiments has been conducted to examine the
appropriateness of the 'Kth values obtained using the new specimen configuration by comparison
with standard conventional tests using compact tension specimens. Some typical results on 'Kth,
'Kthsc and NLFTH are depicted in Fig.3. The fatigue crack growth studies have been
supplemented with characterization of the microstructure, hardness and tensile property of the
selected steels. In addition, the effect of the associated microstructures on the investigated short
crack paths has been carefully examined.
The transition of short to long crack in the investigated steels was found to be dependent on the
nature of the short cracks. But, the maximum fatigue threshold values obtained from short crack
growth experiments on various specimens were found to be independent of the nature of the short
cracks. This maximum values of 'Kthsc, referred here as NLFTH for short cracks are interestingly
found to be in good agreement with their corresponding 'Kth values. The short cracks are
considered to be un-influenced by the crack closure effect, whereas the long cracks considered in
these experiments are observed to have little effect due to crack closure at R =-1. These aspects are
considered to lead the above agreement. The current observations can be supported by some earlier
reports of James and Knott [4], Sadananda and Vasudevan [5] and Taylor [6].

References
1. Ray K. K., Narasaiah N., and Sivakumar R., Mater. Sci. Engg. A, vol. 372, 81-90, 2004.
2. Narasaiah N., Initiation and growth of cracks near fatigue threshold in plain carbon steels,
Ph.D Dissertation, IIT, Kharagpur, India, 2004.
3. Narasaiah N. and Ray K. K, Role of microstructure on short crack propagation and its
threshold, Paper No. 5075, ICF11, Turin , Italy, 2005
4. James M. N. and Knott J. F., Fat. Fract. Eng. Mater. Struct., vol.8, 177-183, 1985
5. Sadananda K. and Vasudevan A. K. , Int. J. Fatigue. vol.19, S99-103
6. Taylor, D., Fatigue Thresholds, Butterworths and Co., London, p-135 1989.
2T5. Fatigue and fracture 223

SCANNING ELECTRON MICROSCOPE MEASUREMENTS OF CRACK-


OPENING STRESS ON FATIGUE CRACKS EXPOSED TO OVERLOADS

L. Jacobsson and C. Persson


Materials Engineering, Lund University, Sweden
Box 118, 221 00 Lund, Sweden
lars.jacobsson@material.lth.se

A fatigue crack exposed to an overload change the characteristics of propagation due to the
enlarged plastic zone at the crack tip. The residual stresses in the material surrounding the crack tip
lead to a crack closure phenomenon, and when the global applied stress gives a zero stress at the
crack tip, the crack start to open. The enlarged plastic zone changes the crack closure and opening
stresses that affect the effective stress intensity factor that is a measure of the stresses acting at the
crack tip.
To measure the global stress when the crack tip starts to open, different techniques are used.
The use of a travelling microscope gives visual observations of the crack, and the observations can
be done to detect when the crack opens. The compliance-offset method is used to measure the
crack closure stress from the reload compliance curve that is produced from an extensometer
measuring the displacement in the crack mouth. Variations in the potential drop signal throughout
a load cycle are used by Andersson et al. [1] to identify opening and closure stresses. Shimojo et al.
[2] uses an electrochemical technique to detect the creation of fresh surface within the crack. They
discovered that the plastic zone expand more in the region with plane stress condition close to the
surface after an overload, and when material was removed from the surface both the retardation of
the crack propagation rate and crack closure decreased. During in-situ fatigue crack experiments
Halliday et al. [3] observed differences in crack closure stresses and near tip COD for the load
conditions R = 0.05 and -1. They dismiss the influence of mechanisms such as crack tip blunting
and strain hardening.
The aim of this study were to make more exact measurements of crack opening and closure
stresses and crack shape very local around the crack tip to find relations to the crack propagation
rate after an overload. This was done from high-resolution scanning electron microscope images
that were analyzed with an image analysis program. Experiments with different R-values and
overload ratios were performed.
To detect the crack shape and the crack opening and closure stresses, an image analyzing
technique is used together with in-situ fatigue crack propagation experiments within a scanning
electron microscope. Throughout the load cycle, high-resolution images are taken of the crack. A
selected spot in one image can be re-found with a cross-correlation algorithm in images at other
stress conditions. With this technique the displacement field around the crack is determined
throughout the load cycle and the stresses when the crack opens and closes can be accurately
measured. Also, the crack shape and compliance curve were determined. In fig. 1 a) the
compliance curve is plotted where displacements are measured a few micrometers from the crack
tip and fig. 1 b) show the crack shapes, from the crack tip to 0.1 mm, for increasing loads.
224 L. Jacobsson and C. Persson

FIGURE 1. a) Compliance curve where the displacements are measured at the crack tip. (o)
increasing load, () decreasing load. b) Crack shapes for different applied loads.

References
1. Andersson, M., Persson, C., Melin, S., ICF 9, 2003
2. Shimojo, M., Chujo, M., Higo, Y., Nunomura, S., Int. J. Fatigue, vol. 20, no. 5, 365-371,1998
3. Halliday, M.D., Zhang, J.Z., Poole, P., Bowen, P., Int. J. Fatigue, vol. 19, no. 4, 273-282,
1997
2T5. Fatigue and fracture 225

PROPAGATION PATH AND FATIGUE LIFE PREDICTIONS OF BRANCHED


CRACKS UNDER PLANE STRAIN CONDITIONS

M. A. Meggiolaro, A. C. O. Miranda1, J. T. P. Castro and L. F. Martha2


Dept. Mech. Eng., 1Tecgraf, 2Dept. Civil Eng., Pontifical Catholic University of Rio de Janeiro
Rua Marquês de São Vicente 225, Gávea – Rio de Janeiro, RJ, Brazil 22453-900
meggi@mec.puc-rio.br, amiranda@tecgraf.puc-rio.br, jtcastro@mec.puc-rio.br, lfm@tecgraf.puc-
rio.br

Fatigue cracks can significantly deviate from their Mode I growth direction due to the influence of
overloads, multi-axial stresses, micro structural inhomogeneities such as grain boundaries and
interfaces, or environmental effects, generating crack kinking or branching (Lankford and
Davidson [1]). The stress intensity factors (SIF) associated to branched fatigue cracks can be
considerably smaller than that of a straight crack with the same projected length, causing crack
growth retardation or even arrest, as discussed by Suresh [2]. This mechanism can quantitatively
explain retardation effects even when plasticity induced crack closure cannot be applied, e.g. in
high R-ratio fatigue problems under plane strain conditions. Analytical solutions have been
obtained for the SIF of some branched cracks, however numerical methods such as the ones
presented in Miranda et al. [3] are the only means to predict the subsequent curved propagation
behaviour.
In this work, a specialized Finite Element program is used to calculate the propagation path
and associated SIF of bifurcated cracks. The numerical calculations are validated through
experiments on 4340 steel ESE(T) specimens. A total of 6,250 FE calculations are used to fit
empirical equations to the process zone size and crack retardation factor along the curved crack
branches. The bifurcation simulations include several combinations of bifurcation angles, branch
asymmetry ratios, crack growth exponents, and even considers interaction between crack
branching and other retardation mechanisms such as crack closure, assuming the crack opening
level is well known.
It is shown that very small differences between the lengths of the bifurcated branches are
sufficient to cause the shorter one to eventually arrest as the longer branch returns to its pre-
overload conditions, see Fig. 1. The process zone size is found to be smaller for lower bifurcation
angles and for branches with greater asymmetry, in both cases due to the increased shielding
effects on the shorter branch. Higher crack closure levels also result in smaller process zones,
because the shorter branch is more easily arrested due to the reduction in its stress intensity range.
However, a competition between smaller process zone sizes and lower growth rates of the longer
branch takes place to determine the real effect of combined bifurcation and closure.
The proposed equations can be readily used to predict the propagation behaviour of branched
cracks in an arbitrary structure, as long as the process zone is small compared to the other
characteristic dimensions. From these quantitative results, it is shown that crack bifurcation may
provide a sound alternative mechanistic explanation for overload-induced fatigue crack retardation
on structural components.
226 M. A. Meggiolaro et al.

FIGURE 1. Bifurcated crack propagation behaviour.

References
1. Lankford, J., Davidson, D.L. Advances In Fracture Research, vol. 2, 899-906, 1981.
2. Suresh, S., Fatigue of materials, Cambridge University Press, U.K., 1998.
3. Miranda, A.C.O., Meggiolaro, M.A., Castro, J.T.P., Martha, L.F., Bittencourt, T.N., Engng.
Fracture Mechanics, vol. 70, 1259-1279, 2003.
2T5. Fatigue and fracture 227

SHORT CRACK EQUATIONS TO PREDICT STRESS GRADIENT EFFECTS


IN FATIGUE

M. A. Meggiolaro, A. C. O. Miranda1, J. T. P. Castro and J. L. F. Freire


Dept. Mechanical Engineering and 1Tecgraf, Pontifical Catholic University of Rio de Janeiro
Rua Marquês de São Vicente 225, Gávea – Rio de Janeiro, RJ, Brazil 22453-900
meggi@mec.puc-rio.br, amiranda@tecgraf.puc-rio.br, jtcastro@mec.puc-rio.br,
jlfreire@mec.puc-rio.br

It is well known that the notch sensitivity factor q can be associated with the presence of non-
propagating cracks. Such cracks are present when the nominal stress range 'Vn is between 'V0/Kt
and 'V0/Kf, where 'V0 is the fatigue limit, Kt is the geometric and Kf the fatigue stress
concentration factors of the notch. Therefore, in principle it is possible to obtain expressions for q
if the propagation behaviour of small cracks emanating from notches is known.
Several expressions have been proposed to model the dependency between the threshold value
'Kth of the stress intensity range and the crack size a for very small cracks, see Chapetti [1]. Most
of these expressions are based on length parameters such as El Haddad-Topper-Smith’s a0 [2],
estimated from 'Kth and 'V0, resulting in a modified stress intensity range

2
'K I 'V ʌ( a  a 0 ) 1 § 'K th ·
, a0 ¨ ¸ (1)
S © 'V 0 ¹
which is able to reproduce most of the behaviour shown in the Kitagawa-Takahashi plot. Yu et al.
[3] and Atzori et al. [4] have also used a geometry factor D to generalize the above equation to any
specimen, resulting in

2
1 § 'K th ·
'K I D ˜ ' V ʌ( a  a 0 ) a ¨ ¸
, 0 S © D ˜ 'V 0 ¹ (2)

Peterson-like expressions are then calibrated to q based on these crack propagation estimates.
However, such q calibration is found to be extremely sensitive to the choice of 'Kth(a) estimate.
In this work, a generalization of El Haddad-Topper-Smith’s equation, which better correlates
with experimental crack propagation data collected from Tanaka et al. [5] and Livieri and Tovo
[6], is proposed:

1 / n
' K th ª § a ·n / 2 º
«1  ¨ 0 ¸ »
'K 0 «¬ © a ¹ »¼
(3)
where 'K0 is the threshold stress intensity factor for a long crack with load ratio R = 0. In the
above equation, n is typically found to be between 1.5 and 8.0, see Fig. 1. Clearly, Eqs. (1) and (2)
are obtained from Eq. (3) when n = 2.0.
228 M. A. Meggiolaro et al.

FIGURE 1. Ratio between long and short crack propagation thresholds as a function of a/a0.
Equation (3) is used in a Finite Element program to evaluate the behavior of cracks emanating
from circular and elliptical holes. For several combinations of notch dimensions, the smallest stress
range necessary to both initiate and propagate a crack is calculated, resulting in expressions for Kf
and therefore q. It is found that the q estimates obtained from this generalization better correlate
with experimental crack initiation data. Expressions for the maximum admissible flaw sizes at a
notch root are also obtained.

References
1. Chapetti, M.D., Int. J. of Fatigue, vol. 25, 1319–1326, 2003.
2. El Haddad, M.H., Topper, T.H., Smith, K.N., Engng. Fract. Mech., vol. 11, 573-584, 1979.
3. Yu, M.T., DuQuesnay, D.L., Topper, T.H., Int.J.Fatigue, vol. 10, 109-116, 1988.
4. Atzori, B., Lazzarin, P., Meneghetti, G., Fatigue Fract. Engng. Mater.Struct., vol. 26, 257-
267, 2003.
5. Tanaka, K., Nakai, Y., Yamashita, M., International J. of Fracture, vol.17, 519-533, 1981.
6. Livieri, P., Tovo, R., Fatigue Fract. Engng. Mater.Struct., vol. 27, 1037-1049, 2004.
2T5. Fatigue and fracture 229

FATIGUE BEHAVIOUR OF PRE-STRAINED TYPE 316 STAINLESS STEEL

Masayuki Akita, Masaki Nakajima1, Keiro Tokaji2 and Yoshihiko Uematsu2


Faculty of Eng., Gifu Univ., 1-1 Yanagido, Gifu 501-1193, Japan
1 Dept. of Mech. Eng., Toyota National College of Technology,

2-1 Eisei-cho, Toyota 471-8525, Japan


2 Dept. of Mech. and Systems Eng., Gifu Univ., 1-1 Yanagido, Gifu 501-1193, Japan

akita@cc.gifu-u.ac.jp

Type 316 stainless steel has excellent corrosion resistance, which has been widely used for
machine components and structures. Materials are usually subjected to plastic deformation by
processing or forming. Therefore, it is very important to understand the effect of pre-strain on
fatigue behaviour (Radhakrishnan and Baburamani [1], Yokotsuka and Ikegami [2]). The present
paper describes the fatigue behaviour of pre-strained 316 stainless steel. Rotating bending fatigue
tests have been performed using specimens subjected to tensile pre-strains and the fatigue
behaviour was discussed.
The material used is a 316 stainless steel with a diameter of 16 mm. The chemical composition
(wt.%) is C: 0.04, Si: 0.23, Mn: 1.33, P: 0.33, S: 0.03, Ni: 10, Cr: 16.9, Mo: 2.01. The material was
solution treated at 1353 K for 1 h. After solution treatment, two different tensile pre-strains of 5 %
and 15 % were given to the material from which hourglass-shape fatigue specimens with a
diameter of 5 mm were machined. Vickers hardness was 137HV for the virgin specimen, 177HV
for the 5% pre-strained specimen and 214HV for the 15% pre-strained specimen. Before fatigue
test, specimens were mechanically polished by emery paper.

FIGURE 1. S-N diagram.

Fatigue tests were performed using cantilever-type rotating bending fatigue testing machine
operating at a frequency of 53 Hz. Crack initiation and small crack growth were monitored with
replication technique. After experiments, fracture surfaces were examined in detail by a scanning
electron microscope (SEM).
The S-N diagram is shown in Fig.1. It can be seen that fatigue strength increases with
increasing pre-strain level. Only a slight increase in fatigue strength can be seen in the 5% pre-
strained specimens, while a very large increase is attained in the 15% pre-strained specimens. The
fatigue limits of the virgin, 5% pre-strained and 15% pre-strained specimens are 300 MPa, 320
230 M. Akita et al.

MPa and 380 MPa, respectively. Fatigue tests were continued to 108 cycles, but no failure took
place.
Based on the observation of crack initiation and growth, the increase of fatigue strength due to
pre-strain was attributed to increased crack initiation resistance, because small crack growth was
not affected significantly by pre-strain.
Figure 2 illustrates the results of stress-incremental fatigue tests. Experiments were started
from the stress level of 20 MPa lower than the fatigue limit for each specimen. When the
specimens were not fractured until 107 cycles, then the stresses were raised by 20 MPa. As seen in
the figure, the fatigue limits are increased significantly in the virgin and 5% pre-strained specimens
(27% and 25% increase, respectively), while no increase is seen in the 15% pre-strained specimen,
suggesting that the coaxing effect depends on the initial condition of specimens. The hardnesses of
fractured specimens were increased remarkably in the virgin and 5% pre-strained specimens, while
a slight increase in the 15% pre-strained specimen. Therefore, the coxing effect is strongly related
to the ability of work hardening during stress cycling.

FIGURE 2. Stress-incremental fatigue test results.

References
1. 1. Radhakrishnan, V.M. and Baburamani, P.S., Mater. Sci. Eng., 17, 283-288, 1975.
2. 2. Yokotsuka, T. and Ikegami, K., J. Soc. Mat. Sci., Japan, 48, 38-43, 1999.
2T5. Fatigue and fracture 231

THE INFLUENCE OF CONSTRAINT ON FITTING FATIGUE CRACK


GROWTH DATA

M. Carboni and M. Madia


Politecnico di Milano
Via La Masa 34, I-20156 Milano, Italy
carboni@mecc.polimi.it, mauro.madia@mecc.polimi.it

Fracture mechanics, in terms of crack propagation, is one of the most popular approaches for life
prediction of components and structures subjected to fatigue loads. Experimental tests (as an
example for a structural steel see Fig. 1) are carried out in order to quantify crack growth rates and
thresholds.

FIGURE 1. Typical experimental results obtained from fracture mechanics tests on a structural
steel.

The obtained results, together with dedicated numerical analyses, are used in the calibration
process of crack propagation analytical models, such as the Strip-Yield model (Newman [1]) able
to keep into account the crack closure phenomenon. These kind of empirical models, initially
introduced for aeronautical applications, are included in widespread life prediction softwares, such
as AFGROW [2] and NASGRO [3], which adopt, as propagation law, the so-called “NASGRO
Equation”
p
§ ' K th ·
1
º ¨© ' K ¸¹
n
da ª§ 1  f ·
C «¨ ¸ 'K » q
dN ¬© 1  R ¹ ¼ § K max ·
¨1  ¸
© K crit ¹
(1)
where the parameter “f” is named “Newman’s closure function” and is a function a the constraint
factor “D” [1] used to introduce in the calculations the real 3D stress field at the crack tip.
The existing software incorporate numerical procedures for fitting parameters of Eq. (1) by
using da/dN data without any explicit reference to test conditions. On the other hand, fatigue crack
growth data obtained with different experimental conditions (specimen type, maximum applied
stress) are also differing in D at the crack tip and so the direct fitting of Eq. (1) is not correct.
In this paper we examine a set of crack growth data obtained onto a mild structural steel with
different specimens (SE(B) and C(T)) from the point of view of constraint factor at the crack tip. In
particular, constraints have been analysed with a series of detailed FEM analyses (Fig. 2).
232 M. Carboni and M. Madia

Results have allowed us to show how apparently conflicting data (especially in terms of 'Kth)
can be re-analysed considering D factors at the crack tip in the different tests.

FIGURE 2. Analysis of the constraint factor at the tip of a SE(B) specimen

References
1. Newman, J. C. Jr., ASTM STP 748, 53-84, 1981.
2. Anonymous. 2001, NASA Technical Report JSC-22267B, Website: www.nasgro.swri.org,
2001.
3. Harter, J. A., U.S. Air Force Research Laboratory Technical Report AFRL-VA-WP-TR-2002-
XXX, Website: http://afgrow.wpafb.af.mil, 2002.
2T5. Fatigue and fracture 233

ATOMIC FORCE MICROSCOPY OF LOCAL PLASTIC DEFORMATION FOR


TEMPERED MARTENSITE

Masao Hayakawa, Saburo Matsuoka1 and Yoshiyuki Furuya


Materials Information Technology Station, National Institute for Materials Science (MITS/NIMS),
1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan
1Faculty of Engineering, Kyushu University, Fukuoka, Japan

HAYAKAWA.Masao@nims.go.jp

An AFM technique established for observing two types of fine and complicated tempered
martensite structures was developed into a technique for quantitatively evaluating local plastic
deformation near tensile yield points. Surface steps with 5-40 nm were observed in locally
deformed martensite blocks near prior austenite grain boundaries.
The material used was JIS-SCM440 steel (containing of 0.40%C, 0.25%Si, 0.79%Mn,
1.12%Cr, 0.17%Mo; mass%). The 7.75 mm-diameter bar was austenized for 720 s at 1323 K, and
then rolled at 1048-1053 K to the total area reduction was 50% as a modified-ausforming treatment
[1]. The bar was tempered for 10 s at 803 K by induction heating. (This was called MAQT;
Modified-Ausformed-Quenced and Tmepered.) Another bar was austenized for 0.9 ks at 1153 K
and then tempered for 1.8 ks at 673 K. (This was called CQT; Conventional Quenched and
Tempered.) Tensile strength showed the same value of 1580 MPa.
AFM observation was conducted under the tapping mode in the atmosphere. The detail of an
AFM observation method is described in the reference [2]. Briefly, the method is: 1) taking AFM
images of an electropolished surface (see Figs. 1(a) and 2(a)); 2) taking AFM images of the same
location after the tensile test (tensile plastic strain: 0.2, 0.4, 0.6%, respectively) after finding the
position using a micro-Vickers indent as a reference mark (see Figs. 1(b) and 2(b)); 3) Finally,
prior austenite grain boundaries in the chemically corroded surface were identified.
Figures 1(a) and 2(a) show AFM images of the electropolished surfaces for MAQT and CQT
before the tensile test, respectively. The black and white contrast in the images show the level
differences on the surface attributable to differences in crystal misorientation with high-angle
boundaries. Therefore, each strip of uniform brightness corresponds to a martensite block that has
a high-misorientation-angle boundary. The mean block widths were 0.38 and 0.49 m for MAQT
and CQT, respectively. Cementite particles, which project out from the base metal, are shown as
white spots.
Figures 1(b) and 2(b) are images created by superimposing prior austenite grain boundaries on
the AFM images of the electropolished surfaces for MAQT and CQT after the tensile tests,
respectively. The horizontal direction is the tensile loading direction. The local plastic deformation
occurred in relative large blocks near prior austenite grain boundaries. The surface steps caused by
local deformations in MAQT were lower than in CQT, since the martensite blocks in MAQT were
finer than in CQT. The values of the surface steps corresponded to the number of pile-up
dislocation at block boundaries.
234 M. Hayakawa et al.

FIGURE 1. AFM images of the electropolished surface before and after the tensile test in (a)
plastic strain 0% and (b) 0.6% for MAQT. Prior Ȗ grain boundaries were shown in (b).

FIGURE 2. AFM images of the electropolished surface before and after the tensile test in (a)
plastic strain 0% and (b) 0.6% for CQT. Prior grain boundaries were shown in (b).

References
1. Yusa S., Hara T., Tsuzaki K. and Takahashi T., CAMP-ISIJ vol.12, 1296, 1999
2. Hayakawa M., Matsuoka S. and Tsuzaki K., Mater. Trans., vol. 43, 1758-1766, 2002
2T5. Fatigue and fracture 235

IMPROVEMENT OF FATIGUE STRENGTH DUE TO GRAIN REFINEMENT IN


MAGNESIUM ALLOYS

Mitsutoshi Kamakura, Keiro Tokaji1, Hideaki Shibata2 and Norikatsu Bekku3


Graduate Student, Gifu Univ., Yanagido, Gifu 501-1193, Japan
1Dept. of Mech. Systems Eng., Gifu Univ., Yanagido, Gifu 501-1193, Japan
2Gifu Prefectural Research Institute, Oze, Seki 501-3265, Japan
3Honda Material Industries, Inc., Yokohira, Ooi-machi, Ena 509-7201, Japan

kamakura@metal.rd.pref.gifu.jp

Magnesium (Mg) alloys are very attractive materials for structural applications because of
excellent specific strength. However, their absolute strengths are insufficient, thus it is necessary to
further improve the mechanical properties, particularly fatigue strength. One of the methods for
improving mechanical properties is grain refinement, but studies on grain refinement in Mg alloys
are very limited (Yamashita et al. [1], Kumar et al. [2]). The purpose of the present study is to
achieve grain refinement due to controlled extrusion and associated improvement of fatigue
strength in wrought Mg alloys. First, grain refinement due to extrusion was studied and then the
fatigue behaviour of the extruded materials was discussed.
Billets (grain size: 200-250Pm) of AZ61A and AZ31B were extruded at an extrusion ratio of
67. The most important parameter influencing grain refinement is working temperature that was
controlled to be low (L), middle (M) and high (H) temperatures. The microstructures on the cross
section in AZ61A and AZ31B are shown in Figs1 and 2, respectively. In both alloys, grain size
decreases with decreasing working temperature and grain refinement is much more remarkable in
AZ31B. The average grain sizes are 12.1Pm, 12.7Pm and 5.8Pm for AZ61A-H, AZ61A-M and
AZ61A-L, 7.4Pm, 2.9Pm and 2.1Pm for AZ31B-H, AZ31B-M and AZ31B-L, respectively.

FIGURE 1. Microstructures in AZ61A: (a) AZ61A-H, (b) AZ61A-M, (c) AZ61A-L.

FIGURE 2. Microstructures in AZ31B: (a) AZ31B-H, (b) AZ31B-M, (c) AZ31B-L.


236 M. Kamakura et al.

FIGURE 3. S-N diagram.

FIGURE 4. Grain size dependence of proof stress and fatigue strength.

Rotary bending fatigue tests were performed in laboratory air using smooth specimens of the
extruded materials in both alloys. The S-N diagram is shown in Fig.3. In AZ61A, the grain size
dependence of fatigue strength is less remarkable, but the fine-grained material (AZ61A-L)
exhibits slightly higher fatigue strength in long life regime. On the other hand, in AZ31B, fatigue
strength increases with decreasing grain size. Based on observation of crack initiation and growth,
such grain size dependence of fatigue strength in AZ31B resulted from both improved crack
initiation resistance and small crack growth resistance. Proof stress and fatigue strength at 107
cycles are represented in Fig.4 as a function of grain size, where both strengths were expressed
properly with the Hall-Petch relation in AZ31B, but not in AZ61A.

References
1. 1. Yamashita, A., Horita, Z. and Langdon, T.G., Mater. Sci. Engng, A300, 142-147, 2001.
2. 2. Kumar, N.V.R., Blandin, J.J., Desrayaud, C., Montheillet, F. and Suéry, M., Mater. Sci.
Engng, A359, 150-157, 2003.
2T5. Fatigue and fracture 237

A UNIFIED FATIGUE AND FRACTURE MODEL APPLIED TO STEEL WIRE


ROPES

M. P. Weiss, R. Ashkenazi and D. Elata


Faculty of Mechanical Engineering, Technion IIT
Haifa 32000, Israel
mweiss@technion.ac.il, matron@netvision.net.il, elata@technion.ac.il

The behavior of short and very short fatigue cracks, emanating from so called “smooth” specimens
with stress concentrations has been an intriguing research topic for a long time. It is well known
that micro-cracks are embedded in the surface of any smooth specimen and are created during the
manufacturing process. A few of these micro-cracks will eventually propagate due to fatigue loads
and will become major cracks. Due to a combination of high local stresses and stress
concentrations, the short cracks are the first to become the major cracks and eventually cause
failure.
A quantitative two-term model for a step by step simulation of crack propagation from very
short cracks to fracture has been proposed by one of the authors, Weiss [1] and Weiss and
Hirshberg [2]. The model is based on the Fatigue Diagram that segregates the whole fatigue and
fracture domain into 6 unique zones, and relates each zone to a known fatigue and/or fracture
regime. Zone 4 in the diagram is the most prevalent in industry, as it is so bounded that the stress
amplitudes are higher than the endurance stress, but lower than the elastic limit, and the stress
intensity factor ranges are higher than the threshold and lower than the fracture toughness. Most
metal parts and structures that are loaded by alternating loads, finally break down in this zone,
either by gross yielding or by critical crack propagation. In this zone the superposition of two
fatigue crack propagation rules is used.
In the current study, the model is described, discussed and enhanced and is shown to be
applicable to fatigue failure in single wires of steel wire ropes in general and specially in non-
rotating tower crane wire ropes.
In its service life a wire rope is subjected to fluctuating tension loading and to bending over
sheaves. The bending stresses along the individual double helix wire are strongly dependent on the
sheave-to-wire diameters ratio and on the arrangement of the reeving system. Predicting the
fatigue life of a wire rope is considered as an essential objective for users and manufacturers.
Generally, the failure of the individual wire is caused due to an initial crack that propagates to a
critical length. In order to improve the safety level in hoisting appliances, the development of a
model that will reliably predict the fatigue life and the safe service life of the rope, is needed.
Moreover, a reliable model based on all the stress affecting factors, will considerably reduce
complex and expensive experimental study.
In this study a new approach to simulate the fatigue life of individual wires within a wire rope,
that is subjected to tension-tension and bending over sheave is presented. The model is based on
the two-term fatigue propagation model [1] and on a new analysis of the stresses that simulates
bending and tension stresses that are generated along the double helix wire within a rope, as shown
by the authors, Elata et al. [3]. It is assumed that initial flaws that were generated during the cold
drawing manufacturing process exist in the individual wire. Several of these flaws (initial cracks)
are located at critical sites where tension and bending stresses are maximal. Once the hoisting
cycle is defined by means of the rain-flow method, it is converted to an “equivalent” loading cycle
with a completely reversed loading (R=-1) by employing the Gerber parabolic line. The growing of
the assumed initial crack is then simulated according to the appropriate zone on the fatigue
238 M. P. Weiss et al.

diagram [1]. Material parameters are calibrated according to experimental fatigue data that was
measured on eutectoid steel wires.
An un-lubricated rope is assumed. Accordingly, relative displacements between adjacent wires
are not permitted. The stress simulation model fully considers the configuration of the straight and
bent double helix wire. The fatigue life of individual wires within an 18u7 non rotating rope are
simulated for various reeving systems, D/d ratios and tension loads. This type of rope is mainly in
use in tower crane applications.
The theoretical results were tested on regular, industrial wires, using a specially built tension
apparatus, that is free to rotate on one side, so that the twisting moments and deformations, that are
generated in tension, can be measured. The results of the stresses were found to be close to the
analytical predictions. Fatigue life was found to be in good agreement with Feyrer 's [4; 5]
experimental prediction formula. Moreover, simulation results predict extensive fatigue
deterioration of the wires within the inner strand layer, under free end attachment using a swivel,
just as is customary in many tower cranes.
The model quantitatively emphasizes the strong dependence of the fatigue life on the quality of
the surface finish of the wires, namely on the residual micro-cracks that were generated during the
deep drawing plastic deformation, in the diameter reduction process of the single wires in
manufacturing.
The new approach may be applied to predict the fatigue life of a complicated cross section of a
rope within a hoisting system. Moreover, the model will extensively reduce expensive
experimental study and can also be used to design better wire configurations. The demonstrated
application in wire ropes is just one example for the use of the two-term model in fatigue life
prediction of complicated mechanical parts underd different stress combinations and histories.

References:
1. Weiss M.P., Int. J. of Fatigue, Vol. 14, No. 2, pp. 91-96, 1992
2. Weiss M.P. and Hirshberg Z., Fatigue & Fracture of Engineering Materials & Structures,
Vol. 19, No. 2\3, pp. 241-249, 1996
3. Elata D., Eshkenazy R and Weiss M.P., International Journal of Solids and Structures, 2004,
Vol 41 /5-6, pp. 1157-1172
4. Feyrer K., Drahtseile, Springer-Verlag Berlin, Heidelberg, 1994
5. Feyrer K., Wire, vol. 45 (2) pp.99-103, 1995
2T5. Fatigue and fracture 239

CORRELATION BETWEEN PARIS’ LAW PARAMETERS BASED ON SELF-


SIMILARITY AND CRITICALITY CONDITION

A. Carpinteri and M. Paggi


Department of Structural and Geotechnical Engineering, Politecnico di Torino
Corso Duca Degli Abruzzi 24, 10129 Torino, Italy
alberto.carpinteri@polito.it, marco.paggi@polito.it

Fatigue crack growth data are usually presented in terms of the crack growth rate, d a / d N , and
the stress-intensity factor range, ' K . The typical fatigue crack propagation curve is shown in
Fig.1, where Region I is referred to as the near-threshold region, Region II as the power-law region
and Region III as the rapid crack propagation region where K max o K IC and crack growth
instability occurs. In Region II the Paris’ equation (Paris and Erdogan [1]) provides a good
approximation to the majority of experimental data:

da
C ('K )m
dN (1)
where C and m are empirical constants usually referred to as Paris’ law parameters.

FIGURE 1. Scheme of the typical fatigue crack propagation curve. Point CR corresponds to the
onset of crack growth instability.

From the early 60’s, research studies have been focused on the nature of the Paris’ law
parameters, demonstrating that C and m cannot be considered as material constants. In fact, they
depend on the testing conditions, such as the loading ratio R (Radhakrishnan [2]), on the size of the
specimen (Barenblatt [3]), and, as pointed out very recently, on the initial crack length (Spagnoli
[4]).
However, an important question regarding the Paris’ law parameters still remains to be
answered: are C and m independent of each other or is it possible to find a correlation between
them based on theoretical considerations? Concerning this point, it is important to take note of the
controversy about the existence of a correlation between C and m in the Literature (Radhakrishnan
[5], Cortie [6], Bergner and Zhouar [7]). However, a very consistent empirical relationship
between the Paris’ law parameters is usually represented by the following formula [5]:
240 A. Carpinteri and M. Paggi

C AB m
, (2)
where parameters A and B depend on the material being studied and are obtained from a best-fit
procedure on experimental data.
In the present paper, the correlation existing between the Paris’ law parameters is derived on
the basis of theoretical arguments. To this aim, both self-similarity concepts [3] and the condition
that the Paris’ law instability corresponds to the Griffith-Irwin instability at the onset of rapid crack
growth (see point CR in Fig.1) are profitably used. Comparing the functional expressions derived
according to these two independent approaches, a relation between the Paris’ law parameters C and
m formally similar to Eq.(2) is proposed. Parameter A is found to be dependent on the ultimate
size of the inelastic zone ahead of the crack tip and on the Carpinteri’s brittleness number
(Carpinteri [8]). On the other hand, parameter B turns out to be dependent on the loading ratio and
on the critical stress-intensity factor. The main consequence of these relations is that only one
macroscopic parameter is needed for the characterization of damage during fatigue crack growth.
An experimental assessment of this new correlation is proposed for a wide range of materials
including steels, Aluminium alloys, epoxy resins with liquid-filled urea-formaldehyde
microcapsules and polymer/silica interfaces. The effect of the loading ratio on the parameter C is
also deeply discussed. A good agreement with experimental data is achieved, showing the
effectiveness of the new proposed correlation.

References
1. Paris P.C. and Erdogan F., J. Basic Engineering, vol. 85D, 528-534, 1963.
2. Radhakrishnan V.M., Engineering Fracture Mechanics, vol. 11, 359-372, 1979.
3. Barenblatt G.I., Scaling, self-similarity and intermediate asymptotics, Cambridge University
Press, Cambridge, 1996.
4. Spagnoli A., Mechanics of Materials, vol. 37, 519-529, 2005.
5. Radhakrishnan V.M., Engineering Fracture Mechanics, vol. 13, 129-141, 1980.
6. Cortie M.B., Engineering Fracture Mechanics, vol. 30, 49-58, 1988.
7. Bergner F. and Zouhar G., International Journal of Fatigue, vol. 22, 229-239, 2000.
8. Carpinteri A., Materials and Structures, vol. 14, 151-162, 1981.
2T5. Fatigue and fracture 241

THERMO-MECHANICAL FATIGUE LIFETIME ASSESSMENT WITH


DAMAGE-PARAMETERS, ENERGY-CRITERIONS AND CYCLIC-J-
INTEGRAL CONCEPTS

M. Riedler, R. Minichmayr, G. Winter and W. Eichlseder2


University of Leoben, Christian Doppler Laboratory for Fatigue Analysis
Chair Mechanical Engineering2
Franz-Josef-Str. 18, A-8700 Leoben
Martin.Riedler@notes.unileoben.ac.at

The simulation of the thermo-mechanical fatigue (TMF) behaviour of cylinder heads is an


important design step in the automotive industry. The steady rise of engine power and the demand
of lightweight construction with a concurrent enhanced reliability require an optimised
dimensioning process. The goal of this paper is to apply classical damage parameters, plastic and
total energy criterions and cyclic J-integral concepts for a thermo-mechanical lifetime assessment
of aluminium and cast iron alloys.
Many of the empirical models are strain based criterions like the Manson-Coffin [1-2] criterion
with numerous modifications. Criterions based on damage parameters try to find a correlation
between the number of cycles to failure and loading parameters. The fracture mechanical view
allows e.g. with cyclic J-integrals a description of the TMF lifetime. Cumulative models, as the
Chaboche [3] models, try to cumulate the damage for each cycle, therefore they need a lot of
computing time for complex structures. The use of submodels is one possibility of using them for a
TMF lifetime assessment of cylinder heads. Another method is the accumulation of the specific
damage parts (pure fatigue, oxidation, creep), as e.g. Miller [4] (accumulation of crack propagation
rates) or Neu/Sehitoglu [5] (accumulation of damage rates) do it. Microstructural methods have a
physical background, but are often difficult to use for a TMF lifetime estimation.
It is shown in previous work (Riedler [6]), that energy based criterions are qualified for TMF
lifetime approaches. Due to the interplay of strain and stress values at a TMF loading they are able
to account for more specific influences as well as their interactions. The general energy based
lifetime criterion is as follows:

'W e ˜ f ('We )  p ˜ f ('W p ) Nf A ˜ 'W  B


, (1)
The specific energy is the sum of an elastic and plastic energy function, whereas both are
characterised with strain and stress parameters, as e.g. amplitude, maximum, minimum or effective
values. A power law is used to describe the dependency of the lifetime.
The plastic energy criterion W1 [6] is best suited for the very ductile Al-alloy AlCuBiPb to
account for the influences of a varying maximum temperature, dwell time, mean strain, pre-ageing
and ageing in service time. The classical damage parameter according to Smith-Watson-Topper [7]
gives the best result for the ductile aluminium cast alloys AlSi7MgCu0.5 and AlSi8Cu3. For
inductile materials like alloy AlSi6Cu4 a criterion based on fracture mechanics in the manner of a
cyclic J-integral approach according to Tomkins [8] provides the best TMF lifetime estimation
result. A TMF lifetime assessment for the different materials investigated shows a reasonable
scatter of about 2.5, see Fig. 1. Since only the Tomkins approach as a sum of an elastic and plastic
energy part [8]:
242 M. Riedler et al.

2
§ S ˜ V max 2S ˜ V max ˜ 'H p · B
'J a ˜¨  ¸ N § 'J ·
¨ ¸ f A˜¨ ¸
© E 1  n' ¹ , © a ¹ (2)
which is a special type of Equ. (1), never gives a rank worse than five (of ten investigated
criterions) for the different materials researched, a standard deviation minimisation process to
determine an appropriate energy definition based on Equ. (1) is executed. This allows us to
describe the TMF life of different materials and material properties with an improved accuracy.

FIGURE 1. Simulation of the TMF lifetime.

References
1. Manson, S.S, Behaviour of materials under conditions of thermal stress, NACA Report No.
1170, 1954.
2. Coffin, L.F, Trans. ASME, vol. 76, 931-950, 1954.
3. Lemaitre, J. and Chaboche, J.L., Mechanics of solid materials, Cambridge Univ. Press, 1990.
4. Miller, M.P., McDowell, D.L., Oehmke, R.L.T. and Antolovich, S.D., In Proceedings of
Thermo-Mechanical Fatigue Behaviour of Materials, ASTM STP 1186, edited by H.
Sehitoglu, Philadelphia, 1993, 35-49.
5. Neu, R.W. and Sehitoglu, H., Metals Trans. A, vol. 20A, 1989, 1755-1767 and 1769-1783
6. Riedler, M.: PhD Thesis, University of Leoben, 2005.
7. Smith, K.N., Watson, P. and Topper, T.H., J. of Materials, vol. 5, No. 4, 767-778, 1970.
8. Tomkins, B., Sumner, G. and Wareing, J., In Proceedings of International Symposium on Low
Cycle Fatigue Strength and Elasto-Plastic Behaviour of Materials, edited by K.-T. Rie and E.
Haibach, DVM, Berlin, 1979, 495-508.
2T5. Fatigue and fracture 243

PREDICTING FATIGUE CRACK RETARDATION FOLLOWING OVERLOAD


CYCLES

Marcos V. Pereira, Fathi A. Darwish1, Arnaldo F. Camarao2 and Sérgio H. Motta3


Department of Materials Science and Metallurgy / Catholic University of Rio de Janeiro
Rua Marquês de São Vicente 225, Rio de Janeiro / RJ, CEP 22453-900, Brazil
1Fluminense Federal University

Rua Passo da Pátria 156, Niterói / RJ, CEP 24210-240, Brazil


2GEP / ArvinMeritor CVS

Av. João Batista 825, Osasco / SP, CEP 06097-900, Brazil


3Technology Division / Brasilamarras

Rua Eng. Fábio Goulart 40, Niterói / RJ, CEP 24050-090, Brazil
marcospe@dcmm.puc-rio.br, fadarwish@poscivil.uff.br, arnaldo.camarao@arvinmeritor.com,
shamotta@brasilamarras.com

Dating back to the beginning of the seventies, a number of models have been proposed to predict
fatigue crack growth rate under variable amplitude loading. This effort was motivated by earlier
observations that the application of an overload is followed by crack growth retardation over a
crack length increment. The model of Willenborg, which belong to the group of yield zone models,
incorporate interaction effects and is characterized by introducing crack tip plasticity. Although
interaction models are generally considered to be applicable for high strength alloys with limited
ductility, empirical verification of the predictions made by these models is rather limited.
Accordingly, the present study was initiated in an effort to test the validity of the Willenborg
model for predicting fatigue crack growth retardation in an R3 grade structural steel following the
application of overload cycles. As this grade steel is used for fabricating offshore mooring chains,
the study was extended to also include flash welded joints taken from the chain links.
CT specimens were machined from both the base metal as well as from the welded joints and a
number of these specimens were subjected to a heat treatment that involved quenching and
tempering. The CT specimens, both heat treated and untreated, were fatigue tested under constant
amplitude (CA) loading in order to establish the typical da/dN versus K curves. During CA fatigue
loading, some specimens in different material conditions, were subjected to single and multiple
overloads applied at a given crack length, and crack growth rate da/dN was followed as a function
of K, evidencing the retardation in crack propagation over an interval of crack length. Crack
propagation rate da/dN within the delay period was predicted by Willenborg model and then
compared with the experimental data. Finally, the results are presented and discussed focusing on
the comparison between the predictions made by the model in light of the experimental data.
244 M. V. Pereira et al.

FIGURE 1. Variation of crack size (a) with the number of cycles (N) for flash welded joint after
overloading.
2T5. Fatigue and fracture 245

FATIGUE CRACK GROWTH AT NOTCHES CONSIDERING PLASTICITY


INDUCED CLOSURE

Jutta Bruening, Olaf Hertel, Michael Vormwald and Georgios Savaidis


Technische Universität Darmstadt, Materials Mechanics Group
Petersenstraße 12, D-64287 Darmstadt, Germany
bruening@wm.tu-darmstadt.de
hertel@wm.tu-darmstadt.de
vormwald@wm.tu-darmstadt.de
Aristotle University Thessalonica, Division of Mechanical Engineering, Laboratory of Machine
Elements and Machine Design
GR-54124 Thessalonica, Greece
gsavaidis@meng.auth.gr

Classical fatigue analyses discriminate between technical crack initiation (crack length of about
1mm) and crack propagation stages. The stage of crack initiation, however, is itself dominated by
the growth of short fatigue cracks. Based on the assumption that a fatigue life to initiate a micro-
structurally short crack of dimensions in the order of 10Pm may be neglected Dankert et al. [1]
have proposed a so-called unified elastic-plastic model for fatigue crack growth evaluation which
describes the whole fatigue life (technical crack initiation and stable crack growth) of notched and
unnotched components by integrating an appropriate crack growth law. In the meantime this model
has entered a guideline for the proof of the strength of components [2].
The present paper reports on the results of an investigation on this model which led to an
improvement of the model’s assessment of effective ranges of the crack driving force (i.e. the J-
Integral). Furthermore, a generalization enabling its application for arbitrary notch situations is
presented. Additionally, new results are shown of comparisons of experimentally determined and
calculated crack growth curves as well as life curves for both the initiation of a technical crack
(crack depth 1mm) and the final failure .
The basic modules of the model are the following:
The range of the cyclic J-integral is taken as the relevant crack driving force. This is essential
because generally short fatigue cracks grow in a notch field where elastic-plastic conditions prevail
in the uncracked situation. Therefore, the limits for the application of the linear fracture mechanics
are violated. Nevertheless, the special case of the linear fracture mechanics based calculation
(negligible cyclic plastic deformation) is incorporated. Approximation formulas for the J-integral
of semi-elliptical and quarter-elliptical surface cracks as well as through-the-thickness cracks in
components with elliptical notches are presented.
A procedure for assessing cracks in arbitrarily notched components is outlined such that this
problem can be reduced to the solved case with elliptical notches.
Effective ranges are considered exclusively. This means that only the range between the crack
closure level and the peak level of a cycle is taken to calculate the J-integral range.
The approximation formulas for the crack opening level are taken as proposed in the literature
by Dankert et al. [1] and Savaidis and Dankert [3].
however, the reference load for cyclic loading is taken to arrive at appropriate effective J-
integral ranges instead of its monotonic counterpart.
A Paris-type crack growth equation is formulated in terms of this effective J-integral range.
Integration of this equation leads to the crack growth curves and fatigue lives.
246 J. Bruening et al.

The integration starts at a suitably chosen initial crack length. The relevant criterion is that the
integration starting at this very initial crack length will result in the material’s strain-life curve for
the special case of an unnotched material specimen.
Variable amplitude loading is taken into account by identifying cycles according to the
rainflow algorithm.
Saal [4] has performed well documented experiments on the initiation (technical crack size
1mm) and growth of fatigue cracks in notched plates of low alloyed steel. The comparison of the
life curves with the corresponding curves calculated by applying the improved unified model
reveals the reasonable accuracy of the model. The differences of the new results compared to those
of the original model are exemplified.

References
1. M. Dankert, S. Greuling, T. Seeger. A Unified Elastic-Plastic Model for Fatigue Crack
Growth at Notches Including Crack Closure Effects.Symposium on Advances in Fatigue
Crack Closure Measurement and Analysis, San Diego, Calif., 1997, ASTM, 1999, STP 1343,
S.480
2. FKM-Richtlinie ”Bruchmechanischer Festigkeitsnachweis”, 2. Auflage, VDMA Verlag
GmbH, Frankfurt am Main, to be published 2005
3. G. Savaidis, M. Dankert, T. Seeger, An analytical procedure for predicting opening loads of
cracks at notches. Fatigue and Fracture of Engineering Materials and Structures, 1995,
18(4), S. 425-442
4. Saal, H., Einfluß von Formzahl und Spannungsverhältnis auf die Zeit- und Dauerfestigkeit
und Rißfortschreitungen bei Flachstäben aus St52, Heft 17, Institut für Stahlbau und
Werkstoffmechanik, Technische Universität Darmstadt, 1971
2T5. Fatigue and fracture 247

INFLUENCE OF MICROSTRUCTURE ON FATIGUE PROPERTIES OF NI-


BASE SUPERALLOY AT ELEVATED TEMPERATURE

QY. Wang , Y. Matsuyama1, N. Kawagoishi1, M. Goto2 and K. Morino3


Sichuan University, Chengdu 610065, China
1Kagoshima University, 1-21-40 Korimoto, Kagoshima, Japan
2Oita University, 700 Tannoharu Oita, Japan
3Tokuyama College of Technology, 3538 Kume Takajyo Shunan, Japan

wang@imr.ac.cn, hiro@mech.kagoshima-u.ac.jp, masagoto@cc.oita-u.ac.jp,


morino@tokuyama.ac.jp

Ni-base superalloys are used under sever conditions like a high temperature or sever corrosive
environment because of their superior properties on corrosion, creep, static strength at high
temperatures and so on. Therefore, many studies on creep and fatigue properties of the alloy have
been carried out, e.g. Brown and Hicks [1]. Recently, it is very important to know the fatigue
properties in long life region, because machines and structures are used for long-term from the
point of view of environmental and economical demands. However, most of the studies on fatigue
were very limited within the short life up to107cycles. It is reported that fracture occurs from an
internal inclusion in long life region in many high strength steels and surface treated steels, while
surface fracture occurs in short life region, e.g. Murakami et al. [2]. Moreover, the fatigue
properties may be influenced by microstructure. However, the mechanism and the evaluation
method for fatigue life of internal fracture are not clarified.
In the present study, fatigue properties of Inconel 718 and the influence of grain size on the
properties were investigated in the wide region of fatigue life until 108 cycles at room temperature
and 500 under rotating bending. Fatigue tests were carried out using materials with two kinds of
grain sizes, about 20m and 100m. Table 1 shows the mechanical properties of the materials. Prior
to fatigue tests, all the specimens were electro-polished about 20m from the surface layer.
Observations of the change in the surface state of a specimen due to stress repetitions and fracture
surface were carried out under an optical microscope using plastic replication technique or under a
scanning electron microstructure (SEM) directly.
Fig. 1 shows S-N curves. Fatigue strength is high in the fine grain at both temperatures,
especially at 500. Although, all the fracture originated from the surface except for the fine grain at
500 in long life region, resulting duplex S-N curve.
Fig. 2 shows crack growth curves. There is no or little differences in the crack initiation life
248 QY. Wang et al.

and the growth rate of a larger crack between the fine grain and the coarse grain at both
temperatures. However, the growth rate of a crack smaller than about 200m is higher in the coarse
grain than in the fine grain at room temperature. This is caused that the crack length initiated is
larger in the coarse grain than in the fine grain. On the other hand, a definite suppression of a crack
growth at 500, when a crack is small. The suppression is caused by oxide induced crack closure at
500. These are main reasons for the difference in the influence of grain size on the fatigue strength
at both temperatures.

References
1. Brown, C.W. and Hicks, M.A., Int. J. Fatigue, vol. 4, 73-81, 1982
2. Murakami, Y., Yokoyama, N. N. and Nagata, J., Fatigue Fract. Engng. Mater. Struct., vol.
25, 735-746, 2002
2T5. Fatigue and fracture 249

MODDELLING FATIGUE CRACK CLOSURE USING DISLOCATION


DIPOLES

P. F. P. de Matos and D. Nowell


University of Oxford, Oxford, UK
paulo.dematos@eng.ox.ac.uk, david.nowell@eng.ox.ac.uk

Over the past 30 years the study of fatigue crack closure has been one of the most challenging
research topics in fatigue crack growth. This phenomenon was first noticed by Elber [1] and then
has been documented by several researchers, both experimentalists and analysts. Different closure
mechanisms have been observed, although the primary mechanism of closure seems to be
plasticity-induced. One of the reasons why fatigue crack closure has been so extensively studied is
because of the need of predicting fatigue lives of structural components, e.g. aircraft industry. It
has been proved that fatigue crack closure plays an important role in crack growth rates together
with load ratio and load history.
To quantify the crack closure effect, different techniques can be used, e.g. experimental,
numerical and analytical. Usually experimental techniques are not used for practical purposes as
they are time consuming and because of the high cost of the experimental work. Numerical
techniques such as FEM modelling have been successfully used to model crack closure. However
they are time consuming since elasto-plastic models are required to accurately model the plastic
deformation ahead the crack tip and the plastic wake. Usually FEM results are dependent on the
level of mesh refinement, crack-tip node release scheme etc. Concerning the analytical techniques
different approaches can be used. A novel technique was developed by Newman [2], his model is
based on the Dugdale model but was modified to leave deformed material on the wake of the
advancing crack. This model is dependent on an empirical constraint factor which is used to
establish the link between the plane stress and plane strain condition. Furthermore, yield strip
models are only suitable to model cracks under plane stress; therefore the use of this constraint
factor for modelling plane strain is only an approximation since the deformation mechanism is
completely different.
In the present paper a quadratic programming technique is used to model fatigue crack closure.
The model herein used was developed by Nowell [3] for modelling fatigue cracks under plane
stress. In the present paper we are going to show how this model can be changed to model crack
closure under plane strain conditions. In the original model, Nowell [3] used displacement
discontinuity opening dipoles (see FIGURE 1) to model both crack and plastic region ahead of the
crack tip. For modelling crack closure under plane strain this model was changed in order to take
into account the different deformation mechanism. Sliding dislocation dipoles were collocated
along the planes of maximum shear stress to quantify the plastic deformation along these planes
and at the tip of the crack. Initial results compare well with those given by Kanninen and
Atkinson’s [4] superdislocation model for the case of a static crack, FIGURE 2. The aim of the
present paper is to investigate the deformation of the crack tip in both deformation modes plane
stress and plane strain. The results are compared with different methods, e.g. FEM and Newman
model [2]. The case of more general loading conditions under plane strain will be discussed.
250 P. F. P. de Matos and D. Nowell

FIGURE 1. a) Opening and sliding dislocation dipoles; b) Dipole displacement field.

FIGURE 2. Crack tip opening displacement. Present model vs Kanninen and Atkinson [4].

References
1. Elber W., Engineering Fracture Mechanics, vol. 2(1), 37-44, 1970
2. Newman Jr. J.C., ASTM STP, vol. 748, 53-84, 1981.
3. Nowell D., Fatigue & Fracture of Engineering Materials & Structures, vol. 21(7), 857-871,
1998.
4. Kanninen M.F. and Atkinson C., International Journal of Fracture, vol. 16(1), 53-69, 1980.
2T5. Fatigue and fracture 251

COMPARISON BETWEEN FATIGUE CRACK GROWTH MODELLED BY


CONTINOUS DISLOCATION DISTRIBUTIONS AND DISCRETE
DISLOCATIONS

P. Hansson, S. Melin and C. Persson1


Division Mechanics, 1Division of Materials Engineering
Lund Institute of Technology, Box 118, SE-22100 LUND
Per.Hansson@mek.lth.se

Short fatigue cracks are known to have a growth behaviour different from that of long cracks, the
latter well predicted by linear elastic fracture mechanics. Short cracks can grow at high rates at
load levels well below the threshold value for long cracks, before entering into the long crack
region, or arrest and become nonpropagating cracks. The growth behaviour of short cracks is
strongly influenced by the microstructure of the material, such as grain boundaries and direction of
slip planes within the grains, as well as of local plasticity around the crack tip. Microstructurally
short cracks, typically shorter than a few grains, grow in a single shear mechanism along specific
slip planes within the grains, cf. Suresh [1], leading to a zigzag crack path, cf. Fig.1.

FIGURE 1. Zigzag shaped edge crack.

For low growth rates, in the order of a few burgers vectors per cycle, it is important to account
for the discrete dislocations within the material. Studies taking the nucleation, movement and
annihilation of discrete dislocations along specific slip planes into account have been performed by
Riemelmoser et al. [2] to study the cyclic crack tip plasticity for a long mode I crack. A similar
model have also been developed by Bjerken and Melin [3] to study the influence of grain
boundaries on a short propagating mode I crack, subjected to fatigue loading. Another approach
was used by Krupp et al. [4] to study the propagation of short cracks in a duplex steel. In this
model, the plastic zone was described by dislocation elements distributed along the slip planes, and
the crack growth rate was assumed to be determined by the crack tip shear displacement.
In this study, the propagation of a short edge crack, situated within one grain in a bcc material
subjected to fatigue loading is modelled using a dislocation formulation. The external boundary,
defined as the free edge and the crack itself, is modelled by dislocation dipole elements, consisting
two glide dislocations and two climb dislocations, allowing the crack surfaces to both open and
shear. The plasticity is modelled by a continuous distribution of dislocation elements along the slip
planes, allowing the slip planes only to shear. The grain boundary is treated as an impenetrable
hinder for the dislocations, and the crack is assumed to be the only dislocation source in the
initially dislocation free material. A schematic description of the model is seen in Fig. 2.
252 P. Hansson et al.

As the applied load exceeds a certain critical value, cf. [5], dislocations will nucleate at the
crack tip or a corner point of the crack, forming plasticity along the slip planes. As the load is
increased, the plastic zone is increased until maximum load is reached and the load is reversed.
During unloading dislocations will move back towards the crack and annihilate, resulting in crack
growth in the corresponding direction under the assumption that no healing of the crack surfaces is
allowed. To describe the plasticity with continuous dislocation distributions instead of with
discrete dislocations as used in earlier studies by Hansson and Melin [5], creates a more efficient
model as regards to computational time, which, in turn allows for simulations at higher growth
rates and growth through additional grains. Though, the approach results in decreasing accuracy in
describing the plastic zone and loss of physically correct dislocation distribution as compared to
the approach in [5].

FIGURE 2. Schematic description of the model.

References
1. Suresh, S. Fatigue of Materials, second edition. University Press, Cambridge, 1998.
2. Riemelmoser, F.O, Pippan, R. and Kolednik, O. Cyclic crack growth in elastic plastic solids:
a description in terms of dislocation theory. Computational Mechanics, 20:139-144, 1997.
3. Bjerkén, C. and Melin, S. A study of the influence of grain boundaries on short crack growth
during varying load using a dislocation technique. Engineering Fracture Mechanics,
71(15):2215-2227, 2004.
4. Krupp, U., Düber, O., Christ, H.-J. and Künkler, B. Application of the EBSD technique to
describe the initiation and growth behavior of microstructurally short fatigue cracks in a
duplex steel. Journal of Microscopy., 213(3):313-320, 2003.
5. Hansson, P. and Melin, S. Dislocation-based modeling of the growth of a microstructurally
short crack by single shear due to fatigue loading. Int Jnl of Fatigue, 27:347-356, 2005.
2T5. Fatigue and fracture 253

FATIGUE EVALUATION CONSIDERING THE ENVIRONMENTAL


INFLUENCE USING A MONITORING SYSTEM

R. Cicero, I. Gorrochategui and J. A. Alvarez


University of Cantabria
E.T.S. Ingenieros de Caminos, Av/Los Castros s/n, 39005, Santander, Spain
ciceror@unican.es

The Nuclear Power Plants were designed for an initial life of 40 years, however nowadays, is under
study the possibility of its life extension. This process requires, among other issues, the analysis of
all the time limited aging calculations to the life period of the original design.
A typical case of this kind of calculations are the fatigue evaluations. In early designs fatigue
damage was not considered in the Nuclear Power Plants. Once it was introduced, the fatigue
damage evaluation was carried out based uniquely on stress considerations, being necessary to
demonstrate that the fatigue damage was lower than the unity.
Along the years, new considerations have been introduced like the variation of the material
properties with the temperature, new computer applications and programs to calculate the stresses
in components with higher accuracy, as well as the possibility of implementing monitoring
systems to register all the real transients that occur in the Nuclear Power Plants, and therefore,
allowing a more realistic evaluation of the fatigue damage. With all these new elements, it has
been being demonstrated that the fatigue damage is always lower than one even considering a life
of 60 years.
In the last years, it has been observed the influence of the environment where the components
are submerged on the fatigue damage [1] and methods have been developed to estimate this effect.
A first analysis of fatigue, considering the environmental effect [2], extrapolating the number of
cycles linearly and applying correction factors for the environmental effect, can lead to the result
that the fatigue damage in some components for 60 years of life is higher than one, and then they
are in a critical situation of fatigue failure.
For these reasons, nowadays it becomes necessary to focus the resources on the development
of new more precise evaluation methods of the environmental influence. These new methods have
to take into account the changes that occur during the transients (temperature, O2 concentration...),
the composition of the affected materials, the loading process...
One possible solution for the fatigue damage evaluation could be the implementation of a
monitoring system that registers the environmental real conditions and the type of loading in each
moment [3, 4]. In this way, it would be possible to establish an accurate approach with the
possibility of a life extension till 60 years, as for this degradation mechanism is referred.
This work illustrates this methodology through its application to a real case.

References
1. NUREG/CR-6260, Application of NUREG/CR-5999 Interim Fatigue Curves to Selected
Nuclear Power Plant Components.
2. Mehta, H.S. and Gosselin, R.S., Environmental factor approach to account for water effects in
pressure vessel and piping fatigue evaluations. Nuclear Engineering and Design 181, 1998
254 R. Cicero et al.

3. Stevens G.L., Deardorff F.A. and Gerber. A.D., Fatigue Monitoring for Demonstrating
Fatigue Design Basis Compliance.
4. D. Pando, J. A. Alvarez and I. Gorrochategui, On the use of a monitoring system for fatigue
usage calculations, Engineering Failure Analysis. 2004. Vol. 11, 765-776.
2T5. Fatigue and fracture 255

THERMAL FATIGUE CRACK INITIATION AND PROPAGATION BEHAVIOR


OF STEELS FOR BOILER

S. Aoi, T. Marumiya, R. Ebara, T. Nishimura and Y. Tokunaga


Department of Advanced Materials Science, Kagawa University
2217-20Hayashi-cho, Takamatsu-city,Kagawa,761-0396,Japan
ebara@eng.kagawa-u.ac.jp
Miura Institute of Research & Development,
7 Horie-cho,Matsuyama-city,Ehime,799-2696, Japan
Miura Co.,Ltd., 7 Horie-cho,Matsuyama-city,Ehime,799-2696, Japan

Thermal fatigue tests were conducted for SB410 steel and SUS310S steel for boiler. A laboratory
made thermal fatigue testing apparatus was used. This apparatus consisted principally of a heating
device using oxygen and LPG gas, a temperature control device for the heated zone and a rapid
cooling device for the specimen. During thermal fatigue tests the heating and cooling cycles were
repeatedly loaded on the specimen placed on the specimen holder. City water was used as a
cooling medium. The heating temperatures were 673K, 773k and 873K,and cooling water was
sprayed on to the specimen surface through a nozzle. As it was difficult to measure the surface
temperature of the specimen, small holes for a thermocouple were prepared to measure the
temperature of the notch during the thermal fatigue tests. Thus the measured temperatures were
used as the testing temperatures. The plate specimens with an electric discharged notch at

Figure 1.Thermal fatigue test specimen


the bottom of the mechanical U notch. The thicknesses of the specimens were 24mm and
12mm.The thermal fatigue crack initiation tests and thermal fatigue crack propagation tests were
conducted up to 100 cycles for specimens with 24mm thick and 300 cycles for specimens with 12
mm thick. Figure 1 shows the figure and the size of the thermal fatigue test specimen. The thermal
fatigue crack length was measured at every ten cycles by use of a viewing microscope with
magnification of 20 after interrupting thermal fatigue tests. The thermal fatigue cracks were
examined by an optical microscope and thermal fatigue fracture surfaces were examined by a
JEOL scanning electron microscope (JSM5500S). Figure2 shows crack propagation curves for
SB410 steel with 12mm thick. The higher the testing temperature the smaller the number of heat
cycles is. The higher the testing temperature the faster the crack propagation rate is. The same
phenomenon was observed for SUS310S. The thermal fatigue crack propagation rate of SB410
steel was faster than that of SUS310S steel. Plural thermal fatigue cracks were observed on the
bottom of the electric discharged notch for SB410 steel. The crack branching was observed on the
256 S. Aoi et al.

specimen tested at 873K. Striation was observed on fracture surfaces of SB410 steel. Striation was
predominant on fracture surfaces of SUS310S. The striation spacing per cycle ,S obtained from the
measured striation spacing S versus K curve was well coincident with the da/dN K curve in the
high crack propagation rate. It can be concluded that thermal fatigue crack of boiler steels
propagate in association with striation.

Figure2. Thermal fatigue crack propagation curves of SB410 steel with 12mm thick
2T5. Fatigue and fracture 257

RECENT DEVELOPMENTS IN FATIGUE CRACK GROWTH

R. Jones , S. Pitt1, and E. Siores2


Cooperative Research Centre for Railway Engineering, Department of Mechanical Engineering,
P.O. Box 31, Monash University, Victoria, 3800, Australia. Rhys.jones@eng.monash.edu.au
1Air Vehicles Division, Defence Science and Technology Organisation, 506 Lorimer Street,

Fishermans Bend 3207, Australia, Susan.pitt@dsto.defence.gov.au


2Centre for Materials Research and Innovation, Bolton University,Deane Road, Bolton , BL3 5AB,

England, Esiores@bolton.ac.uk

Recent research [1 - 3] has shown that the Frost Dugdale law [4] can be applied to a wide class of
engineering problems, and that for complex load spectra the crack growth rate per loading block
da/dB is essentially a linear function of a, see Figure and [1] for more details. These findings
together with the realisation that “in the threshold regime, there is something missing either in the
model”, see [5], led to the conjecture [1-3] that in Region I the crack growth rate can be expressed
in the form:

da/dN = C ( a/a*)(1-m/2) ('K) m (1)


which for small cracks integrates to give the Frost-Dugdale relationship, viz:

a = ao e EN (2)

Figure 1: Crack growth rates for A7 wing, from [1, 3].

where N is the number of cycles, E is a parameter that is geometry, material and load dependent, a
is the crack depth and, a0 is the initial size of the defect. This paper presents a large number of
examples to show that this relationship is also true for both over and underloads, see Figs 2 and 3
which present the experimental results presented by Liu and Dittmer [6] and Schijve and Broek [7]
respectively.
This methodology has been linked with NASTRAN and is used, in conjunction with a finite
element model of the F-111 wing, to predict crack growth in the 1969 F111 wing test, which was
performed in the US as part of the F-111 certification program.
258 R. Jones et al.

References
1. Barter S., Molent L., Goldsmith N. and Jones R., 2004, "An Experimental Evaluation Of
Fatigue Crack Growth", Journal Engineering Fracture Mechanics, 2004.
2. R. Jones, S. Barter, L. Molent and S. Pitt, "Crack growth at low K's and the Frost Dugdale
law", Journal Chinese Institute of Engineers, Special Issue in Honour of Professor G. C. Sih,
27, 6, 869-875, 2004.
3. L. Molent, R. Jones, S. Barter, and S. Pitt, “Recent Australian Developments In Fatigue Life
Assessment”, Submitted International Journal of Fatigue, April 2005.
4. N. E. Frost and D. S. Dugdale, “The propagation of fatigue cracks in test specimens”, Journal
Mechanics and Physics of Solids, 6, pp 92-110, 1958.
5. J.C. Newman Jr., A. Brot, C. Matias, “Crack-growth calculations in 7075-T7351 aluminum
alloy under various load spectra using an improved crack-closure model”, Engineering
Fracture Mechanics 71 (2004) 2347–2363, 2004.
6. F. Liu and D. F. Dittmer, “Effect of multi-axial loading on crack growth”, AirForce Flight
Dynamics Laboratory, Air Force Systems Command, AFFDL-TR-78-175, Volumes I – III,
September 1978.
7. J. Schijve and D. Broek, “Crack Propagation Based on a Gust Spectrum with Variable-
Amplitude Loading”, Aircraft Engineering, 34, pp. 314-316, 1962.
2T5. Fatigue and fracture 259

CRACK CLOSURE EFFECTS IN A CRACKED CYLINDER UNDER


PRESSURE

J. Zhao, R. Liu, T. Zhang1 and X. J. Wu2


Department of Mechanical and Aerospace Engineering, Carleton University, 1125 Colonel By
Drive, Ottawa, Ontario, Canada K1S 5B6, rliu@mae.carleton.ca
1Department of Mathematics, Northeastern University, Shenyang, Liaoning, P.R. China 110004,

ztmath@163.com
2Institute for Aerospace Research, National Research Council Canada, 1200 Montreal Road,

Ottawa, Ontario, Canada K1A 0R6, Xijia.Wu@nrc-cnrc.gc.ca

The circumferential stress varies through the wall thickness when a cylinder is subjected to internal
or external pressure. Internal pressure will induce tensile circumferential stress while external
pressure compressive stress, in terms of the thick-walled cylinder theory [1]. The resultant stress
due to the synergetic contribution of internal and external pressure may be tensile at the inner
surface and compressive at the outer surface or vice versa in some cases, depending on the load
levels, which would lead to crack-face closure at the compressive edges when the cylinder contains
an axial crack. Historically, crack problems in shells were formulated in terms of either the
classical theory [2] or the transverse shear theory [3], which were all based on the linearized
shallow shell theory [4-6]. However, one deficiency of these solutions is that the crack face
interpenetration or overlap was allowed at the compressive edge when a bending load was
involved, which is physically unrealistic. In reality, crack-face closure on the compressive edge
may occur when a shell or plate containing a through-the-thickness crack is subjected to bending
load. The present research is aimed to develop a formulation for the determination of stress
intensity factor for a cylinder containing an axial crack, which incorporates the effect of the crack-
face closure.
According to the formulation developed by Delale and Erdogan [7-9], the problem of a cracked
shell subjected to membrane force and bending moment can be reduced to a pair of coupled
singular integral equations,

1GW
1 dW  1>k K,W G W  k K,W G W @dW 2SF1 K ,
³ ³ 11 1 12 2
1W K 1 (1a)

1Q 2 G2 W
1 1
h
4 ³
dW  ³>k21 K,W G1 W  k22 K,W G2 W @dW 2S F2 K ,
O0 1 W K 1
a
(1b)

³ G W dW
i 0,
1 (i = 1, 2) (1c)

where K y / a , -a < y < a, and G1 and G2 denote the derivatives of the crack-face normal
displacement, u(0+, K), and the crack face rotation, E(0+, K), respectively,
260 J. Zhao et al.

G1 K
w
wK

u 0 ,K G2 K
w
wK

E 0 ,K .
and (2)
The simulation of the crack-face closure is achieved by introducing a contact force at the
compressive edge so that in the closure region the normal displacement of the crack face at the
compressive edge must remain equal or greater than zero, that is,

h

u 0  ,K r 2a

E 0  ,K t 0
(3)
Due to the curvature effect of shell crack-face closure may not always occur on the entire
length of the crack, depending on the geometry of the shell and the nature (direction) of bending
load. The closure regions will be determined as a mixed-boundary value problem through an
iterative process such that either the normal displacement at the compressive edge is equal to zero
or the contact pressure is equal to zero. The stress intensity factor at the crack tip is due to the
synergetic contribution of membrane and bending effects, which is evaluated by Delale and
Erdogan [8],

E a ª z º
k I z  «¬ g 1 1  a g 2 1 »¼
2 (4)
which shows that the stress intensity factor varies through the shell wall thickness. The results
demonstrate that due to the curvature effect the crack closure behavior in shells differs from that in
flat plates. Full-length closure always occurs in flat plates, but it is not a case in shells. Partial
closure may occur in shells, depending on the shell geometry and the bending load direction. It is
found that crack-face closure greatly reduces the bending component but increases the membrane
component of the stress intensity factor. The crack-face closure prevents the penetration of the
crack faces at the compressive edges physically when a cracked shell is subjected to bending load,
which leads to a zero value of the stress intensity factor at the closure edge.

References
1. Benham, P.P., Crawford, R.J. and Armstrong, C.G., Mechanics of Engineering Materials,
Longman Group Limited, Singapore, 1996.
2. Reissner, E., in Proceedings of the First Symposium on Naval Structural Mechanics, 1958,
74-113.
3. Sih, G.C. and Hagendorf, H.C., Plates and Shells with Cracks, Sih, Noordhoff Int Pub, 1977.
4. Reissner, E., J. Math. Phys., vol. 25, 80-85, 1946.
5. Reissner, E., J. Math. Phys., vol. 25, 279-300, 1946.
6. Nagdi, P.M., Quart. Appl. Math., vol. 14, 331-333, 1956.
7. Delale, F. and Erdogan, F., Quart. Appl. Math., vol. 37, 37:239-258, 1979.
8. Delale, F. and Erdogan, F., Int. J. Solids Struct., vol. 15, 907-926, 1979.
9. Delale. F. and Erdogan, F., Eng. Fract. Mech., vol. 18, 529-544, 1983.
2T5. Fatigue and fracture 261

AN EXPERIMENTAL STUDY OF TEARING-FATIGUE INTERACTION

Peter Birkett, Michael Lynch and Peter Budden1


Serco Assurance, Birchwood Park, Warrington, Cheshire, WA3 6AT, UK
1British Energy Generation Ltd., Barnett Way, Barnwood, Gloucestershire, GL4 3RS, UK

peter.birkett@serco.com

Structural integrity assessments of defects in components under cyclic loading typically add the
individual contributions to crack extension from tearing and fatigue, calculated using the J-
resistance (JR) curve and fatigue crack propagation data, respectively. This approach, with the
limiting crack size determined by an instability analysis, is employed in the widely used R6 defect
assessment procedure [1]. However, it is likely that under conditions of variable or severe cyclic
loading, the fatigue crack growth rate can be retarded or enhanced by tearing occurring within the
fatigue cycles. Similarly, crack extension due to tearing can be affected by prior fatigue crack
growth. An experimental and analytical programme is in progress designed to explore tearing-
fatigue interaction, and to refine and validate R6 advice, using load-history transients typical of
those considered in nuclear reactor pressure vessel (RPV) assessments.

FIGURE 1. A typical load-crack mouth opening displacement trace for a tearing fatigue test.

In the programme, transients with tearing and fatigue components were applied to standard
fracture toughness test specimens and the resulting crack extension behaviour was compared with
predictions made using the R6 methodology. A ferritic RPV steel was used, but an austenitic steel
was also used for some tests. Baseline fatigue crack growth rate data were determined at relevant
R-ratios and temperatures. The baseline JR curve was defined at ambient temperature and 288ºC
using standard test procedures. Fifty-eight tearing-fatigue (TF) tests were performed, consisting of
a series of pre-defined fatigue stages separated by monotonic loading stages. The parameters
varied included: R-ratio (-1.1 to +0.8); temperature (ambient, 288ºC,); ‘’ and ‘’ factors (ratios of
the maximum and minimum cyclic loads to the most recent tearing load); specimen thickness
(15mm to 40mm); and the numbers of cycles and stages. Fig. 1 shows a typical load-displacement
curve for a test loaded to P1 followed by fatigue cycling between DP1 and EP1, a further tear to P2,
and so on. The tests were analysed to estimate the crack length at each stage of the test. Fig. 2
shows a plot of the estimated and measured crack growth values at the end of each test. The
majority of points fall above the 1:1 reference line, indicating that estimated values are
conservative. However, a significant number of points fall below the line and were found to relate
mainly to thinner specimens. The analyses were refined by (i) increasing the estimated tearing
crack growth incrementally, rather than assuming that tearing recommences at the start of the JR
262 P. Birkett et al.

curve, and (ii) using an enhanced fatigue crack growth law at high K values. These refinements
decreased conservatism. Sensitivity studies were also carried out in order to examine how the
estimated crack growth was affected by changes in the limit load solution, fatigue crack growth
law and JR curve. Further analysis used crack driving force

FIGURE 2. Comparison of estimated versus measured crack growth for tearing-fatigue tests.

diagrams to carry out ductile tearing analyses, from which maximum allowable crack sizes at
instability were obtained. Three alternative J-estimation schemes were used, R6 Options 1 and 2
and the GE-EPRI approach. The results indicated that: (i) the R6 guidance on tearing-fatigue
generally gave conservative estimates of tolerable crack size compared with the measured crack
sizes from the TF tests, with those based upon the GE-EPRI scheme being the least conservative;
(ii) simple addition of tearing and fatigue crack growth components did not always give
conservative estimates of crack growth, compared with the crack sizes measured; (iii) the use of
the tolerable crack size from EPRI as a limiting crack size value for the simple addition of tearing
and fatigue crack growth components was conservative for the thicker specimens, but may not be
for the thinner specimens and (iv) the simple addition results were sensitive to relatively small
changes in the definitions of the JR curve and Paris Law.

References
1. R6 – Revision 4, Assessment of the Integrity of Structures Containing Defects, British Energy
Generation Ltd, Amendment 3, 2004.
2T5. Fatigue and fracture 263

SIF SOLUTIONS FOR CRACKS IN RAILWAY AXLES UNDER ROTATING


BENDING

S. Beretta, M. Madia, M. Schodel and U. Zerbst1


Politecnico di Milano, Dipartimento di Meccanica,
Via La Masa 34, 20156 Milano
1GKSS Research Center Geesthacht GmbH

Max Planck Straße, 21502 Geesthacht, Germany

Railway axles are designed for infinite life. However, in order to correctly manage the few failures
detected in service [1] for this safety component, there has been an increasing attention to damage
tolerance analysis and determination of inspection intervals for railway axles [2]. Among the
different input parameters of a damage tolerance analysis, a wide knowledge of SIF solutions for
railway axles is not available because of the presence of an effect induced by rotating bending
upon the SIF at the surface crack tip of a growing crack (Fig. 1.a and 1.b): K is higher in rotating
bending than in plane bending and in particular there exists an angle T0 such that
K(T0)=max{K(T)}. The only solutions available for this kind of problem are the ones by Carpinteri
et al. [3,4] who obtained combined SIF solutions obtained under plane bending at T = 0° and T =
90° for calculating SIF in smooth bars.

Figure 1. Schemes adopted for the analyses: a) scheme of the crack; b) scheme of the rotating
bending and explanation of the superposition of the effects (M’X and M’Y).

In order to analyze this problem, a series of SIF solutions for some typical notches (named as
T-notch, U-notch and S-transition) [5] of railway axles under plane and rotating bending were
carried out. In particular the effect of rotating bending upon cracks in axle body and near the
press-fittings is investigated.
Considering the three different notches, K values have been calculated for different crack
depths and shapes, taking into account different rotation angles for the axle. In particular analyses
for T = 0° and T = 90° resulted to be useful to determine the behavior of K for different rotation
angles. By this point of view, the results show that SIF of surface cracks onto axle body, under
rotating bending, can be obtained as a simple superposition of the effects of the two bending
moments (M’X and M’Y) acting on the crack (Fig. 1.b).
This simple approximation works well also for cracks near press-fittings (Fig. 2.a). It is very
interesting to compare SIF for cracks at S-transition obtained with press-fitting and with a tied
connection (Fig. 2.b). While there is no significant effect for tip A, SIF at point B with press-fitting
is approx. 25% higher than the one obtained with the tied connection: this fact seems to be due to a
264 S. Beretta et al.

different stress distribution at the notch root. It is of some importance to remark that this result
shows that the evaluation of SIF near (or at) the press-fittings has to be carried out considering the
effective load transfer and SIF solutions for axle body cannot be directly applied.

Figure 2. Results in terms of SIF at S-transition for a crack with a/D = 0.1267 and a/b = 0.736
(Snom = 107 MPa): a) superposition of the effects vs. FEM in case of press-fitting; b) comparison
of FEM results for tied-contact and press-fitting.

References
1. Smith, R. A. and Hillmansen, S., “A brief historical overview of the fatigue of railway axles”,
Proc. Instn Mech. Engrs, 218 (2004).
2. Zerbst, U. Vormwald, M. Andersch, C. Mädler, K. and Pfuff, M., “The development of a
damage tolerance concept for railway components and its demonstration for a railway axle”,
Engng Fract. Mech., 72, 209-239 (2005).
3. Carpinteri, A. and Brighenti, R., “Part-through cracks in round bars under cyclic combined
axial and bending loading”, Int. J. Fatigue, 18, 33-39 (1996).
4. Carpinteri, A. Brighenti, R. and Spagnoli, A., “Surface flaws in cylindrical shafts under rotary
bending”, Fatigue Fract. Engng Mater. Struct., 21, 1027-1035 (1998).
5. EN 13261, “Railway applications – Wheelsets and bogies – Axles – Product requirements”,
CEN, September 2003.
2T5. Fatigue and fracture 265

MECHANICAL CHARACTERIZATION OF SINGLE CRYSTAL BARS WITH


CAPACITOR DISCHARGE WELDING AND LASER CLADDING

Samanta Chiozzi, Vito Dattoma and Francesco W. Panella


Dipartimento di Ingegneria dell’Innovazione, Università degli Studi di Lecce
Via per Arnesano, 73100 Lecce, Italy
samanta.chiozzi@unile.it, vito.dattoma@unile.it, francesco.panella@unile.it

The efficiency of gas turbine systems used for energy production and for aeronautical engines
construction can be improved through the elevated thermo-mechanical capabilities of advanced
materials employed for blades production. It is well known in fact, the importance to increase the
turbine inlet temperature; this temperature is limited by the highest working temperature of steels
and super-alloys of which blades and rotor are made. Therefore it is necessary to study the
employment of new materials more resistant to high temperatures.
Actually two different ways are being employed to reach the above results: the application of
stators and rotors cooling systems or the substitution of ordinary steels with more resistant high
temperature superalloys. Both methods are valid, but the first one is less preferred by aeronautical
industries because it needs the blade profile modification to facilitate the inside cooler circulation,
affecting the blade efficiency and increasing the total weight [1].
Among the innovative materials currently under study, multicrystal and Single Crystal Nickel-
base superalloys allow the theoretical achievement of 1300°C without damage and with the ability
to withstand high loads if compared with the ordinary materials and they are being strongly
considered for the future aeronautical engines.
Single Crystal superalloys are of particular interest since they have the advantage to avoid the
presence of microstructural discontinuities originated by the grain edges and the sites of
precipitates concentration, which in multicrystal superalloys generally represent the zone of crack
initiation and propagation [2,3].
In the present work the static behavior of Single Crystal Nickel-base superalloy CMSX-4 has
been analyzed, performing deep mechanical and microstructural characterization, necessary to
establish in detail the superalloy behaviour at the expected service stresses. Tests have been
performed both at room and working temperature (about 800°C), comparing base material, CD
welded and Laser Clad cylindrical bars.
The microstructure has been successively analyzed through investigations performed with the
optic microscope and the scanning electron microscope (SEM) observations. The micrographic
analysis concerned base material and as-welded specimens; this way the microstructural
modifications and the Heat Affected Zones have been studied as well as the interaction between
the weld material layers and parent metal. The mechanical and microstructural characterization for
“as welded” material is needed because of the requirement to develop advanced technologies for
the repair of damaged blades, particularly in the case of blisks, constituted of blading and rotor in
one piece.
The Capacitor Discharge Welding (CDW) technique is particularly suitable because it allows
the removal of the damaged part and the subsequent substitution of a new blade without the
replacement of the whole block and additional machining; it allows to achieve extremely thin
welded joints with narrow Heat Affected Zones, in order to avoid stress concentration effects at the
weld toe and to reach good material continuity [4-7]. Actually the CDW process for aeronautical
applications is under analysis, since it is not yet employed in production lines because of the lack
266 S. Chiozzi et al.

of enough experimental tests on advanced materials with the involved welding parameters. Laser
cladding is an highly flexible repair technique which is already applied for the aero engine industry
[8].
In this work the first welding trials on Single Crystal specimens have been analyzed in order to
establish the mechanical properties decay of a single crystal welded component, but also the
microstructural modifications of the parent material after the welding process in relation to the
used technological parameters had to be investigated.
Static test results reveal acceptable structural abilities of the welded joints, but it is not possible
to compare them with other data since in literature very few works about CMSX-4 can be found.
Laser clad specimens show excellent tensile strength and higher yield behavior than those obtained
with CD welded joints; the welding parameters and the heat treatments have not yet optimized and
the welded material distribution and microstructure shell improove in the next experiments.

References
1. Gell, M., Duhl, D.N. and Giamei, A.F., The development of Single Crystal superalloy turbine
blades, Pratt & Whitney Aircraft Group, Superalloys, 1980, 205-214.
2. Hemmersmeier, U. and Feller-Kniepmeier, M., Element distribution in the macro- and
microstructure of nickel base superalloy CMSX-4, Material Science and Engineering A374,
2004.
3. Wagner, A., Shollock, B.A. and McLean, M., Grain structure development in directional
solidification of nickel-base superalloys, Material Science and Engineering A256, 1998.
4. Chiozzi, S., Dattoma, V. and Panella, F.W., First results on the mechanical behaviour of
CDW welded superalloys, 5th International Conference on Fatigue and Fracture, Bari-Italy,
9-10 May 2005.
5. Casalino, G., Dattoma, V., Ludovico, A. and Panella, F.W., Numerical model for Capacitor
Discharge Welding, 13th DAAAM International Symposium, Vienna-Austria, October 2002.
6. Alley, R.L., ASM Handbook, Vol. 6: welding brazing and soldering Capacitor Discharge
Study Welding”, American Welding Society, 1991.
7. Dattoma, V. and Panella, F.W., Studio della resistenza su saldature CDW di barre in acciaio
AISI 304, XXXII Convegno AIAS, Salerno-Italy, 3-6 September 2003.
8. Richter, K.H., Orban, S. and Nowotny, S., Laser Cladding of the Titanium Alloy Ti6242 to
restore damaged blades, XXIII International Congress on Applications of Lasers and Electro-
Optics, 2004.
2T5. Fatigue and fracture 267

FRACTAL DIMENSION ANALYSIS OF FRACTURE TOUGHNESS USED


HIGH STRENGTH CAST IRON

S. Doi and M. Yasuoka


700 Dannoharu Oita, Japan
doi@cc.oita-u.ac.jp

The rotating bending fatigue tests were done using ADI and PDI with ion carburizing. The plain
specimen with ion carburizing about 30MPa than the virgin PDI, and the fatigue strength of the
carburized ion notched specimen was given an equivalent to the virgin PDI.
Especially, the ion carburized notched specimen is effective material they have a better effect
than lower that the notched PDI. In the recapitulation, on the element parts demanded a shallow
notch, the clear distinction appeared whether ion charge or not, respectively. An important point is
that carburized PDI presented two bending S-N curve not clear in high cycle fatigue. If the
hardness is constantly holed to an internal point, the shape of S-N curve is unclearly appeared. But,
its phenomena appeared from ion boundary layer by observation of SEM photograph. The ion
Carburizing can control the charged depth by the sputtering time. Consequently, this method
warrants consideration as a non-polluting process and economical use of energy.
1 Pcn S-N curve by ion-carburized effect was controlled to affect on the crack propagating in
our researched meso-scopic structure.
2 Typical of quasi-fish eye in the surface vicinity can be decided by a/b. Where, a: equivalent
small defect radius, b: equivalent cleavage radius (yield area). Accordingly, it is expressed
by life cycle of S-N curve.
3 The plastic zone of ADI is shorter than that of PDI at the fracture. The fracture toughness
of PDI is larger than that ADI. In other words, there is not relation toughness and
elongation too much.

Fig.1 Profile and dimensions Fig.2 Fracture surface ADI,PDI


268 S. Doi and M. Yasuoka

Fig.3 Fracture toughness and plastic zone


2T5. Fatigue and fracture 269

INVESTIGATING GAP EFFECTS IN FATIGUE LIFE OF SPOT WELDED


JOINTS

M. Zehsaz and S. Hasanifard


Faculty of Mechanical Engineering, University of Tabriz, Tabriz, Iran
zehsaz@tabrizu.ac.ir, hasanifard@tabrizu.ac.ir

The use of aluminium alloys in vehicle bodies are ever increasing especially when the main
manufacturing process of these components is spot welding.
The gap effects in 3-D finite element analysis are generally neglected. This effect in spot
welded joints is important in terms of the mechanical behaviour of them. In this research, the gap
effects on the fatigue life of the joints has been studied using strain-based approach to obtain the
fatigue life.
The numbers of spots are two in a row, joining two sheets of 5182-0 aluminium alloy. To do
so, the problem is physically modeled using ANSYS FEM based software. The three-dimensional
mesh models, the gap effects and a non-linear analysis have been used for the joints of sheets with
the gap of 0.06mm, 0.12mm, 0.18mm, as shown in Fig. 1. Because of symmetry, only one of the
hot spot-welded joints is modelled.

FIGURE 1. 3-D meshes and the boundary conditions in two different views of a hot spot weld

The elastic-plastic stresses at edges of spot welds were considered and Manson-Coffin formula
has been applied to obtain fatigue life.
The comparison between the experimental results and numerical approach shows that the
existence of gap between sheets increases the amount of the stresses near the roots of nuggets.
Therefore, the fatigue life of spot welded sheets decreases with increasing the gap distance.
This is illustrated in Fig. 2.
270 M. Zehsaz and S. Hasanifard

FIGURE 2. L-N curve for different sheet gaps

References
1. Ni, K., Mahadevan, S., Strain-based probabilistic fatigue life prediction of spot-welded joints,
International Journal of Fatigue 26(2004) 763-772.
2. Adib, H., Gilgert, J., Pluvinage, G., Fatigue life duration prediction for welded spots by
volumetric method, International Journal of Fatigue 26(2004) 81-94.
3. Radaj, D., Sonsino, C.M., Fatigue Assessment of Welded Joints by Local Approaches,
woodhead publishing Ltd.1998.
4. SHARP, M.L., NORDMARK, G.E., MENZEMER, C.C., Fatigue design of aluminium
components & structures, McGraw-Hill, 1996.
5. GEAN, A., WESTGATE, S.A., KUSZA, J.C., EHRSTROM, J.C., Static and Fatigue
Behaviour of Spot-Welded 5182-0 Aluminium Alloy Sheet, 1999.
2T5. Fatigue and fracture 271

FATIGUE OF PMMA BONE CEMENT

S. L. Evans
School of Engineering, Cardiff University
The Parade, Cardiff CF24 3AA, UK
EvansSL6@cardiff.ac.uk

The use of PMMA bone cement is the most successful way of fixing implants such as hip and knee
replacements, and has many other uses in orthopaedic surgery. However, in the long term
mechanical failure of the cement is common and may lead to clinical failure. Despite a large
volume of published research, the fatigue of bone cements is not well understood and much more
research is needed before fatigue failure can be reliably predicted and avoided.
Most of the published studies of cement fatigue have used simple S-N tests on unnotched
specimens. The presence of stress concentrations that act as sites for crack initiation has a critical
effect on the fatigue life and so many of these studies have identified porosity as an important
factor. However, it is not clear whether porosity is as important in the clinical setting where there
are many other stress concentrations and perhaps cracks due to curing shrinkage. Here crack
propagation may be more important, and there are many complications such as variable amplitude
loading, slow crack growth and microstructural effects.
PMMA has been widely used as a model for studies of fatigue and fracture in polymers, but
there is still a limited understanding of the effects of factors such as variable amplitude loading.
Bone cements have a more complicated structure, with previously polymerised beads in a softer
matrix which cures on implantation, and other components such as particles of barium sulphate or
zirconia to make the cement visible in radiographs. This microstructural complexity means that
the cement may behave very differently from pure PMMA.
In smooth specimens, cracks initiate from pores, at the stress concentrations formed between
previously polymerised beads, and there is a correlation between porosity and fatigue life. The
stress concentration Kt at a spherical pore depends only on the Poisson’s ratio and for cement it is
about 2.06. The size of the pores does not affect Kt, unless the pores are large enough to
significantly reduce the cross- section. This reduction in cross- sectional area can account for the
observed reduction in fatigue life for specimens with greater porosity. Arguably the effect of
porosity is therefore an artefact of the specimen design, rather than a change in the properties of the
material. Apart from this effect, porosity may only be important when it increases the likelihood of
interaction between pores, which can cause higher stress concentrations. There is evidence for
crack initiation due to shrinkage stresses developed during curing, although this does not always
occur. In the clinical setting, this may be more relevant than initiation from pores by fatigue
processes.
Crack propagation has not been widely studied, and there are several complications such as
slow crack growth and microstructural differences in the crack path. Physiological loading varies
widely in amplitude and frequency content and the effects of these variables have not been fully
investigated. Cracks typically propagate through the beads, but in some cases it is found that the
beads pull out of the matrix. It is not clear why this occurs. There does not appear to be poor
adhesion between the beads and the matrix, and it may be that the presence of residual compressive
stresses in the beads due to curing shrinkage of the surrounding matrix causes the crack to deviate.
Since the residual stresses relax over periods of weeks or months, this may explain some of the
inconsistencies between different studies, and the change at different 'K levels that has been
observed. The crack deviation that is caused by this phenomenon leads to considerable
272 S. L. Evans

disturbance of the crack front, with some shielding by intact ligaments that bridge the crack. This
may account for much of the increased resistance to crack propagation in cements as compared to
pure PMMA. Even when the crack propagates through the beads, the fracture surface is much
rougher than in pure PMMA.
Clinical failure of the cement occurs over long time periods, and this implies that the crack
growth rate is very low, perhaps as low as 10-12 m/cycle. Many studies have assumed a threshold
at a much higher growth rate, but when load shedding is correctly carried out it appears that there
may not be a threshold and the Paris Law is obeyed to 10-11 m/cycle or below. However, crack
arrest can occur at higher 'K, perhaps for microstructural reasons.
Preliminary measurements of the effects of variable amplitude loading raise more questions
than they answer. In pure PMMA, considerable retardation occurs, particularly at lower 'K. 30%
overloads every 100 cycles resulted in crack growth rates as much as two orders of magnitude
lower than constant amplitude loading. At higher 'K the effect was much less pronounced. In
cement, however, little or no retardation occurred as a result of individual overloads. Single
overloads applied at a crack growth rate of around 10-9 m/cycle produced an acceleration of
subsequent crack growth by up to a factor of five, and in some cases the acceleration persisted for
25,000 cycles or more. Single overloads at this level produced no retardation whatsoever. The
mechanisms by which these effects occur are unknown at present, but it is likely that plasticity
induced crack closure may play a role. Other mechanisms such as extended crazing ahead of the
crack tip or breakage of ligaments that bridge the crack may also be relevant; in pure PMMA in
particular there can be extensive bridging at the sides by shear- yielded material in the plane stress
region and this may be affected by overloading.
Despite extensive research the fatigue behaviour of bone cements is not well understood. The
effect of porosity has been highlighted in many studies, but it is not clear how relevant it is to
clinical cement failure. Crack propagation behaviour is very different to that of pure PMMA due
to the complex microstructure of cements. The effects of slow crack growth under constant load
may be important, and there is evidence that variable amplitude loading may produce dramatic
effects including both acceleration and retardation of subsequent crack growth. As yet the
mechanisms involved are unknown and much further research is needed before it is possible to
predict fatigue failure with any accuracy.
2T5. Fatigue and fracture 273

INFLUENCE ON THERMAL BARRIER COATING DELAMINATION


BEHAVIOUR OF EDGE GEOMETRY

H. Brodin, X. H. Li1 and S. Sjoestroem1


Department of Mechanical Engieering, Linköping University,
SE-581 83 LINKÖPING, Sweden
1SIEMENS Industrial Turbomachinery AB, SE-612 83 FINSPÅNG, Sweden

hakan.brodin@siemens.com, soren.sjostrom@siemens.com

Ceramic thermal barrier coatings are commonly used in gas turbine hot components (e.g.,
combustor liners/buckets and guide vane platforms). In components that are only partially coated
or have cooling-air outlets, coating-end stress singularities may lead to the spallation of the
coating.
Depending on the geometry of the transition from coated to uncoated material, the severity of
the stress singularity will vary. Basic references for the analysis of such stress singularities are, for
instance, Bogy [1], Bogy [2], Bogy [3], and Dundurs [4], where it is shown that the severity of the
stress singularity depends on the chamfer angle \ and that making this angle \ > 0 decreases the
singularity order at the coating end, see Fig. 1.

FIGURE 1. Schematic drawing of a TBC system. A chamfer angle \ is defined in the figure.

In the present study, a thin thermal barrier coating system has been studied. Bond- and top
coats have been sprayed to a thickness of 150µm and 350µm, respectively. Vacuum-plasma-
spraying technology was used, and the test specimens were rectangular (30x50x5mm) coupons of
a nickel-based superalloy, Haynes 230. A NiCrAlSiY bond coat and an Y2O3 stabilised ZrO2 top
coat were used.
In order to achieve well-defined chamfers, sprayed components were ground on the edges with
SiC grinding paper to desired geometry. By inspections of cross-sections that had not undergone
thermal fatigue cycling, it was ensured that no damage was introduced into the system.
Mechanical testing was done in a thermal cyclic test rig where specimens are heated in a
furnace and cooled with compressed air. The thermal cycle data were:
Tmin=100°C, Tmax=1100°C, dwell time at max temperature Tdwell=40 min and cycle time
Tcyc=60 min.
FE modelling of the system has been done aiming at supporting the findings from thermal
fatigue tests. A parametric study including variation of the chamfer angle \has been done and the
stress state near the chamfer evaluated.
274 H. Brodin et al.

Evaluation of fatigue damage can be done visually for observation of coating failure
(macroscopic observation on coating surface). 20% area with complete spallation was considered
as thermal barrier coating failure. For evaluation of damage development, additional light
microscopy investigations of cross-sections have been done.
Results show that the fatigue life benefits from introduction of a chamfer angle at the coating
end during thermal fatigue cycling. This is seen in Fig. 2, where fatigue life is plotted against
chamfer angle \

FIGURE 2. Influence on fatigue life on chamfer angle \.

Reference
1. Bogy D.B., J. Appl. Mech., vol. 35, 460-466, 1968
2. Bogy D.B., Int. J. Solids Struct., vol. 6, 1287-1313, 1970
3. Bogy D.B., J. Appl. Mech., vol. 38, 377-386, 1971
4. Dundurs J., J. Appl. Mech., vol. 36, 650-652, 1969
2T5. Fatigue and fracture 275

LOW CYCLE FATIGUE AND FRACTURE OF A COATED SUPERALLOY


CMSX-4

S. Stekovic
Division of Engineering Materials, Linköping University
58183 Linköping, Sweden
svjst@ikp.liu.se

The single crystal CMSX-4 is the second-generation rhenium containing nickel-base superalloy
developed by Cannon-Muskegon Corporation, Davis [1]. The alloy is widely used because of its
good mechanical properties such as long-time strength and toughness at high temperatures, Simms
et al. [2]. Alone, the alloy has limited oxidation and corrosion resistance and to solve this problem
it is protected with coatings.
The objective of this study is to examine and establish fatigue and fracture behaviour under
low cycle fatigue test conditions of the coated single crystal superalloy CMSX-4. For this purpose
three different coatings have been chosen, an overlay coating AMDRY997, an aluminide diffusion
coating RT22 and an innovative coating called IC1. Cylindrical solid specimens were cyclically
deformed with fully reversed tension-compression loading with total strain amplitude control at
two temperatures, 500oC and 900oC. The tests were also done at the uncoated specimens taken
from the same batch. The empirical relationship between plastic strain amplitude and the number
of reversals to fatigue failure was determined by Coffin-Manson, Klesnil and Lukas [3], se eq. (1).
The results will be given in the full paper.

'H p
H '(2 N f )c
2 (1)

At 500oC the coatings had detrimental effects on the fatigue life of CMSX-4 relative to the
uncoated specimens while at 900oC the coated specimens exhibited longer life than the uncoated
specimens at the same temperature. AMDRY997 gave the longest life compared to RT22 and IC1.
The coated superalloy exhibited hardening and higher stress level at higher applied strains and
lower temperature. At 900oC softening occurred together with lower stress response level. The
coatings lowered the stress response level from about 12% to 31% compared to the uncoated
specimens under the same test conditions. Most of the observed cracks initiated at the coating
surface, one example is shown in Fig. 1. Majority of the cracks was arrested at the interface
between the coating and the superalloy. Surface roughness or rumpling was found in AMDRY997
with some cracks initiated from the rumples.

FIGURE 1. SEM micrograph showing surface crack initiation observed at the surface of
AMDRY997 at 500oC
276 S. Stekovic

The fatigue behaviour of CMSX-4 can be affected by precipitation of topologically close


packed (TCP) phases such as round shaped P phase and acicular shaped V phase by two damage
mechanisms, embrittlement damage mechanism related to the brittle nature of those phases and
softening of J matrix by depletion of strengthening elements to form TCP phases, Simonetti and
Caron [4]. The phases were observed under and in the interdiffusion zone, although no
microcracks initiated from them were found.
The cracks found in the coatings grew more or less perpendicular to the load axis. The initial
crack path is flat along the plane with high tensile stress, normal to the stress axis. The cracks
began to fluctuate then with visible fatigue striations.

References
1. Davis, J.R., Heat-resistant materials, ASM specialty handbook, Materials Park, ASM
International, The USA, 1997.
2. Simms, C.T., Stoloff, N.F. and Hagel, W.C., Superalloys, Wiley, New York, 1987.
3. Klesnil M. and Lukas P., Fatigue of Metallic Materials, Elsevier, Amsterdam, The
Netherlands, 1992.
4. Simonetti, M. and Caron, P., Materials Science and Engineering A, vol. A254, 1-12, 1998.
2T5. Fatigue and fracture 277

THERMOMECHANICAL FATIGUE OF OPEN-CELL ALUMINIUM SPONGE

T. Guillen, A. Ohrndorf, U. Krupp, H. J. Christ, S. Derimay1, J Hohe2 and W Becker1


Institut für Werkstofftechnik, Universität Siegen
Paul-Bonatz-Str. 9-11, 57068 Siegen, Germany
Phone +49 271 2184
Fax +49 271 2545
ohrndorf@ifwt.mb.uni-siegen.de
1Institut für Mechanik, TU Darmstadt
2Fraunhofer-Institut für Werkstoffmechanik, Freiburg

While closed-cell metal foams have been intensively studied during the last ten years, only little
attention has been put to the group of open-cell metal sponges. Due to their low density and good
homogeneity metal sponges are promising candidates for structural and at the same time functional
applications. For the successful design of new components using metal sponges a sound
understanding of the behaviour under monotonic and cyclic loading conditions is required. The
aim of the present study is to investigate the behaviour of the open-cell aluminium foam under this
kind of loading conditions placing special attention to thermomechanical fatigue and creep
characteristics.
The studied metal sponge was precision-cast AlSi7Mg obtained by m-pore GmbH, Dresden,
having a porosity of 8 ppi and a relative density of 5.1%. Figure 1 shows the macroscopic cell
structure of the studied material which has similar mechanical properties than the PORMET
material (AlSi9Cu sponge) that was studied in an earlier work [1].

FIGURE 1. Macrograph of the AlSi7Mg sponge.

Rectangular specimens (25mmx25mmx70mm) were glued to a special specimen mount and


subjected to monotonic loading and fatigue at room temperature. The tests were carried out by
means of a servohydraulic testing machine (MTS 810).
Since an induction heating system could not be used to heat the inhomogeneous sponge
specimens, thermomechanical fatigue tests were carried out in a self-designed temperature
chamber, where the specimens can be heated by a hot air fan and cooled by compressed air (see
Fig. 2). For this tests the specimen’s ends were cast-infiltrated with Zn serving as specimen mounts
in order to a allow the performance of tests at temperatures higher than 200°C.
278 T. Guillen et al.

Results of monotonic tests revealed that the plateau stress in compression lies in a range of
0,7- 0,75 MPa.
With respect to the tensile behaviour it was observed that the amount of plastic deformation at
fracture is strongly influenced by the ductility of the cell strut material. This could be confirmed
within the present study resulting in strain of fracture levels of 1,5%.

FIGURE 2. Temperature chamber for the realisation of TMF tests on open-cell aluminium sponge.

Thermomechanical fatigue behaviour of the aluminium sponge has been compared to that
under isothermal fatigue of room temperature and at elevated temperatures. The fatigue damage
mechanisms for each loading condition was analysed by optical and scanning electron microscopy.
The experimental results were compared with a micromechanical model that is based on the
idealization of the Kelvin model foam and subsequent examination of the effective mechanical
properties to be used within an energetic homogenisation scheme [2].

References
1. A. Ohrndorf, U. Krupp, H.-J. Christ, Proc,. Fatigue 2002, A.F. Blohm (ed) pp.3093-3098,
2. J. Hohe, W. Becker , Int. J. Mech. Sci., 45, 2003, 891-913.
2T5. Fatigue and fracture 279

THE INFLUENCE OF ALTERNATE BLOCK LOADING ON THE FATIGUE


LIFETIME

M. Kohut and T. Lagoda


Technical University of Opole
ul. Mikoajczyka 5, 45-271 Opole, Poland
tlag@po.opole.pl

The paper presents the results of fatigue tests of cylinder specimens made of duralumin PA6 under
alternate block loading for bending and torsion. Each block contains n and n cycles with sinusoidal
course. Based on the fatigue curves amplitudes suitable for 105, 3˜105 and 106 number of cycles to
failure were determined. The loading was applied in alternate two-amplitudes blocks by 104, 3˜104
or 105 each one (10% of the fatigue life for the given loading level) until failure of the specimen
(fig.1).

Fig.1. Part of the stress course

The obtained experimental results were described with the criterion of normal and shear
energy density parameter in the critical plane [1]

W eq (t) ȕW Șs (t)  țW Ș (t)


(1)

where ȕ, ț – constant dependent on S-N description for bending and torsion of


suitable m