Regidly Framed Earth Retaining Structures

Springer Series in Geomechanics & Geoengineering
Walid Aboumoussa
Magued Iskander
Rigidly Framed
Earth Retaining
Structures
Thermal Soil Structure Interaction
of Buildings Supporting Unbalanced
Lateral Earth Pressures
Springer Series in Geomechanics
and Geoengineering
Series editors
Wei Wu, Universität für Bodenkultur, Vienna, Austria
e-mail: wei.wu@boku.ac.at
Ronaldo I. Borja, Stanford University, Stanford, USA
e-mail: borja@stanford.edu
For further volumes:

http://www.springer.com/series/8069
About this Series
Geomechanics deals with the application of the principle of mechanics to geoma-

terials including experimental, analytical and numerical investigations into the me-
chanical, physical, hydraulic and thermal properties of geomaterials as multiphase
media. Geoengineering covers a wide range of engineering disciplines related to
geomaterials from traditional to emerging areas.
The objective of the book series is to publish monographs, handbooks, workshop
proceedings and textbooks. The book series is intended to cover both the state-of-
the-art and the recent developments in geomechanics and geoengineering. Besides
researchers, the series provides valuable references for engineering practitioners and
graduate students.
Walid Aboumoussa · Magued Iskander
Rigidly Framed
Earth Retaining
Structures
Thermal Soil Structure Interaction
of Buildings Supporting Unbalanced
Lateral Earth Pressures
ABC
Walid Aboumoussa Magued Iskander
Ventrop Engineering Consulting Group, NYU Polytechnic School of Engineering
PLLC New York
New York USA
USA
ISSN 1866-8755 ISSN 1866-8763 (electronic)

ISBN 978-3-642-54642-6 ISBN 978-3-642-54643-3 (eBook)
DOI 10.1007/978-3-642-54643-3
Springer Heidelberg New York Dordrecht London
Library of Congress Control Number: 2014933215
c Springer-Verlag Berlin Heidelberg 2014

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer.
Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations
are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of pub-
lication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any
errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect
to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

To our children
Lauren Iskander
Katherine Aboumoussa
Christopher Iskander
NO WARRANTY IS EXPRESSED OR IMPLIED
The intent of this document is to supply information not render

engineering, consulting, or other professional services. The
information contained in this document is deemed reliable.
Nevertheless, neither the publisher nor the authors guarantee
the completeness, accuracy, or usefulness of any information
contained herein.
The publisher and the authors shall not be held responsible for
any errors, omissions, or damages arising from the use of the
material contained herein
Acknowledgements
A large number of professional colleagues contributed formally and informally to

the success of this effort. In particular, we are grateful to Mr. Robert Antonucci,
PE, for his support, advice, and encouragement throughout this work. We are also
grateful to Mr. Pierre Gouvin of Geo Instruments, Inc., and Mr. Frank Calfa, PE,
of Structural Contracting Services, Inc., for their assistance in instrumentation.
Farah Massoud, Andrew Dimond, Alexey Sidelev, Ziad Karnaby, and Saumil
Parikh made important contributions to the research included in this book. Mehdi
Omidvar, Roula Maloof, Zhibo “Chris” Chen, Weihua Jin, and Stan Roslyakov
reviewed the entire manuscript. This work could not have been finished without
the assistance of Nikolaos Machairas and Bhujang Patel who provided substantial
editorial support.
We are grateful to the United States National Science Foundation for its finan-
cial support through NSF grants No. CMS 9733064, DGE 0337668, and DGE
0741714.
We would like to thank Dr. Thomas Ditzinger for adopting this manuscript into
the Springer Series on Geomechanics and Geoengineering. We are also grateful to
the series editors: Prof. Wei Wu from Universität für Bodenkultur in Austria and
Prof. Ronaldo Borja from Stanford University for recommending publication of
this manuscript. Credit is also due to Holger Schaepe from Springer for his assis-
tance during the production of this work.
Last but not least, we would like to express our gratitude to our families. This
work could not have been finished without their support and encouragement.
Notation
The following symbols and notations are generally used in the main text of this
monograph, as well as in figures and tables. Other specific notations not listed
herein is defined in the main text.
δ si Lateral drift at i’th story

Ao Equivilant area of RFERS
b Total number of bays of RFERS
E Elastic modulus
G Shear modulus
i Story from the top where deflection is computed (i= 0 at top)
Ib Moment of inertia of individual beam
Ic Moment of inertia of individual column
K0 Coefficient of lateral earth pressure at rest
Ka Coefficient of active lateral earth pressure
Kp Coefficient of passive lateral earth pressure
lb Length of beam
lc Length (Height) of column
Ma Moment value at the end of backfilling (initial)
Mccx Moment value at the end of contraction cycle x
Mecx Moment value at the end of expansion cycle x
RFERS Rigidly Framed Earth Retaining Structure
s Total number of stories of RFERS
Sb Beam Stiffness, Ib/Lb
Sc Column Stiffness, Ic/Lc
Uh Horizontal movement
Uha Horizontal movement at the end of backfilling (initial)
Uhccx Horizontal movement at the end of contraction cycle x
Uhecx Horizontal movement at the end of expansion cycle x
Uhnx Horizontal movement of frame with x bays
XII Notation
W Total acting force on RFERS

α, λ, β, υ Calibration factors
γ Shear strain
κ, ψ, χ Geometric factors
ν Poisson’s ratio
σ' Lateral earth pressure
σ'a Lateral earth pressure at the end of backfilling (initial)
σ'active Coulomb's active earth pressure
σ'at rest Lateral earth pressure at rest
σ'ccx Lateral earth pressure at the end of contraction cycle x
σ'ecx Lateral earth pressure at the end of expansion cycle x
σhnx Lateral earth pressure behind frame with x bays
φ Soil internal angle of friction
Selected Unit Conversions
Conversions from American (British) to SI Units
Length 1 in. = 2.54 cm

1 ft = 0.3048 m
Area 1 in.2 = 6.452 cm2

1ft2 = 929.03 cm2
Volume 1 in.3 = 16.387 cm3

1 ft3 = 28.317 x10-3 m3
Force 1 lb = 4.448 N
1 kip = 4.448 kN
1 US Ton = 8.896 kN
Stress 1 psi = 6.895 kN/m2

1 psf = 47.88 N/m2
1 tsf = 95.76 kN/m2
1 kip/ft2 = 47.88 kN/m2
Unit Weight 1 pcf = 0.1572 kN/m3
Moment 1 lb-ft = 1.3558 N.m

1 lb-in. = 0.11298 N.m
Energy 1 ft-lb = 1.3558 J
Conversions from SI Units to American (British) Units
Length 1 m = 3.281 ft
1 cm = 0.3937 in
Area 1 m2 = 10.764 ft2

1 cm2 = 0.155 in.2
Volume 1 m3 = 35.32 ft3

1 cm3 = 0.061023 in.3
XIV Selected Unit Conversions
Force 1 N = 0.2248 lb
1 kN = 0.2248 kip
1 metric ton = 2204.6 lb
Stress 1 kN/m2 = 20.885 psf

1 kN/m2 = 0.01044 tsf
1 kN/m2 = psi
Unit Weight 1 kN/m3 = 6.361 pcf
Moment 1 N.m = 0.7375 lb-ft

1 N.m = 8.851 lb-in.
Energy 1 J = 0.7375 ft-lb

Contents
Acknowledgements ....................................................................................... IX
Notation ......................................................................................................... XI
Selected Unit Conversions ............................................................................ XIII

Conversions from American (British) to SI Units .................................... XIII
Conversions from SI Units to American (British) Units .......................... XIII
About the Authors ........................................................................................ XXIII
1 Introduction to Rigidly Framed Earth Retaining Structures

(RFERS) .................................................................................................. 1
1.1 Rigidly Framed Earth Retaining Structures (RFERS) ..................... 1
1.2 Initial Lateral Earth Pressure ........................................................... 2
1.3 The Effects of Thermal Movements ................................................ 3
1.4 Scope of Contribution to the Behavior of REFRS .......................... 4
1.4.1 Closed Form Expressions for Lateral Deflection of
RFERS ................................................................................ 4
1.4.2 Experimental Study ............................................................. 4
1.4.3 Numerical Analysis ............................................................. 5
2 Classical Earth Pressure Theory Related to Framed Structures ....... 7

2.1 Introduction ..................................................................................... 7
2.2 The Development of Earth Pressure Theory ................................... 8
2.2.1 Inadequacies of Classical Solutions .................................... 9
2.2.2 The Stiffness of the Retaining Wall .................................... 10
2.3 The Case of the Integral Bridge Abutment...................................... 10
2.4 The Case of Rigidly Framed Earth Retaining Structures ................ 14
3 Closed-Form Expressions for Lateral Deflection of Rigid

Frames ..................................................................................................... 15
3.1 Introduction ..................................................................................... 15
XVI Contents
3.2 Approximate Drift Analysis of Buildings ....................................... 16

3.3 Simplified Expression for Lateral Deflection of Rigidly
Framed Structures ........................................................................... 16
3.3.1 Derivation of Equations for Lateral Deflection, δs .............. 18
3.3.2 Derivation of Equivalent Area, Ao ...................................... 22
3.3.3 Numerical Parametric Modeling to Calibrate Derived
Equations ............................................................................. 26
3.3.4 Calibration Factors for Derived Equations .......................... 26
3.4 Confidence in the Derived Equations .............................................. 29
3.4.1 Weibull Statistical Analysis ................................................ 32
3.4.2 Confidence in the Derived Equations .................................. 34
3.5 Effect of Input Parameters on Expression Accuracy ....................... 36
3.6 Examples ......................................................................................... 38
3.6.1 Symmetric Rigidly Framed Structure Subject to
Uniform Loading ................................................................. 39
3.6.2 High Rise Building Subject to Wind Load .......................... 39
3.6.3 Rigidly Framed Earth Retaining Structure Subject to
Earth Pressure...................................................................... 40
3.7 Unknown Earth Pressure Distributions ........................................... 41
3.8 Earth Pressure from a Known Deflection........................................ 42
3.9 Case of In-Filled Frames ................................................................. 42
3.10 Limitations of the Developed Equations ......................................... 44
3.11 Conclusions ..................................................................................... 44
4 Case Study of a Full Scale RFERS in Service ...................................... 45

4.1 Introduction ..................................................................................... 45
4.1.1 Mechanism of Failure.......................................................... 49
4.1.2 Objectives of the Instrumentation and Monitoring
Program ............................................................................... 49
4.2 Instrumentation ............................................................................... 50
4.2.1 Selection of Instruments ...................................................... 50
4.2.2 Instrumentation Details ....................................................... 51
4.2.2.1 Vibrating Wire Displacement Transducers
(VW) .................................................................... 51
4.2.2.2 Electrolytic Tiltmeters .......................................... 54
4.2.2.3 Data Collection and Management ........................ 56
4.2.2.3.1 Atomatic Data Acquisition .................. 56
4.2.2.3.2 Manual Data Acquisition..................... 57
4.2.2.3.3 Data Management ............................... 57
4.2.2.4 Instrumentation Limitations ................................. 57
4.3 Monitoring Results of Instrumented RFERS .................................. 57
4.3.1 Electrolytic Tilt Sensors Data.............................................. 58
4.3.1.1 Electrolytic Tilt Sensors on Level A .................... 59
4.3.1.2 Electrolytic Tilt Sensors on Level B .................... 59
Contents XVII
4.3.1.3 Electrolytic Tilt Sensors on Level C .................... 65

4.3.1.4 Electrolytic Tilt Sensors on Level D .................... 69
4.3.2 Vibrating-Wire Displacement Transducers ......................... 71
4.3.2.1 Thermal Study of Structure: Sensors Normal
to Expansion Joint ................................................ 72
4.3.2.1.1 Annual Range of Movements .............. 72
4.3.2.1.2 Seasonal Behavior ............................... 75
4.3.2.2 Sensors Parallel to Expansion Joint
(North-South Direction) ....................................... 78
4.3.2.2.1 Roof Level Sensors ............................. 78
4.3.2.2.2 Level B Sensors ................................... 81
4.4 Conclusions ...................................................................................... 83
5 Relationship between Temperature and Earth Pressure

for RFERS............................................................................................... 85
5.1 Introduction ..................................................................................... 85
5.2 Building Description ....................................................................... 86
5.3 Geotechnical Properties of the Retained Soil .................................. 86
5.4 Instrumentation Program ................................................................. 89
5.5 Apparent Thermal Coefficient of Expansion of PG-1 ..................... 90
5.6 Lateral Displacement of Building Parallel to Earth Pressure .......... 91
5.6.1 Measured Lateral Displacement .......................................... 91
5.6.2 Baseline Correction Due to Thermal Movement
of PG-2 ................................................................................ 96
5.6.3 Correction of Lateral Displacement Due to Thermal
Movement of PG-1 .............................................................. 96
5.6.4 Accuracy of Computed PG-1 Movements .......................... 97
5.7 Relationship between Lateral Deflection and Earth Pressure.......... 99
5.8 Earth Pressure Causing Lateral Deformation .................................. 102
5.9 Limitations of This Study................................................................ 104
5.10 Conclusions ..................................................................................... 105
6 Numerical Analysis of Instrumented RFERS ...................................... 107

6.1 Introduction ..................................................................................... 107
6.2 The Finite Element Model............................................................... 108
6.2.1 The Structural Frame ........................................................... 108
6.2.2 The Backfill Soil ................................................................. 109
6.2.3 The Analysis Procedure ...................................................... 110
6.3 Numerical Analysis ......................................................................... 110
6.3.1 Thermal Analysis of Rigid Frame (Part 1) .......................... 110
6.3.1.1 Description of Analysis Procedure ....................... 110
6.3.1.2 Numerical Analysis Results (Part 1) .................... 111
6.3.2 Thermal Analysis of Rigid Frame with Mohr-Coulomb
Backfill (Part 2) ................................................................... 113
XVIII Contents

6.3.3 Thermal Analysis of Rigid Frame with Hardening-Soil
Backfill (Part 3) ................................................................... 117
6.3.4 Comparison of Numerical Analysis Results ........................ 121
6.4 Conclusions ..................................................................................... 123
7 Parametric Study of Earth Pressure behind RFERS at Backfill

Stage ........................................................................................................ 125
7.1 Introduction ..................................................................................... 125
7.2 Parametric Numerical Analysis ....................................................... 126
7.2.1 Finite Element Analysis Model Details ............................... 128
7.2.2 Single Story Rigidly Framed Earth Retaining
Structures............................................................................. 130
7.2.2.1 Backfill Soil with 30º Internal Friction Angle...... 130
7.2.2.2 Effect of Lateral Frame Stiffness on the
Mobilizations of Active Earth Pressure................ 133
7.2.2.3 Effect of Staged Construction Calculation on the
Mobilizations of Active Earth Pressure................ 135
7.2.3 Two-Story Rigidly Framed Earth Retaining Structures ...... 137
7.2.4 Three-Story Rigidly Framed Earth Retaining Structures .... 141
7.2.5 Four-Story Rigidly Framed Earth Retaining Structures ...... 145
7.2.6 Five-Story Rigidly Framed Earth Retaining Structures....... 148
7.2.7 The Case of Frames Braced against Lateral Sway .............. 151
7.2.8 Analysis of a Single Story 6-Bay Shear Wall Structure ...... 152
7.2.9 Analysis of a Two Story 15-Bay Shear Wall Structure ....... 152
7.3 Conclusions ..................................................................................... 153
8 Analysis of Single Story RFERS Subject to Temperature

Variations ................................................................................................ 155
8.1 Introduction ..................................................................................... 155
8.2 Numerical Parametric Analysis ....................................................... 155
Contents XIX
8.3 Analysis of Single Story Rigidly Framed Earth Retaining

Structures ........................................................................................ 157
8.3.1 Backfill Soil with 30º Internal Friction Angle ..................... 157
8.3.1.1 Single Bay Frame (Bay Length, Lb, 10 Feet,
Column to Beam Stiffness Ratio, Sc/Sb, of 1) ..... 158
Column to Beam Stiffness Ratio, Sc/Sb, of 4) ...... 161
8.3.1.5 10-Bay Frame (Bay Length, Lb, 10 Feet,
8.4 Conclusions ..................................................................................... 187
9 Multi-story RFERS Subject to Temperature Variation ..................... 189

9.1 Introduction ..................................................................................... 189
9.2 Numerical Parametric Analysis ....................................................... 189
9.2.1 Three Story Rigidly Framed Earth Retaining
Structures............................................................................. 191
9.2.1.1 Frames with Bay Length, Lb, 10 Feet, and
Column to Beam Stiffness Ratio, Sc/Sb, of 1 ........ 191
9.2.2 Five Story Rigidly Framed Earth Retaining Structures ....... 209
9.3 Conclusions ...................................................................................... 219
XX Contents
10 Conclusions and Recommendations ..................................................... 221

10.1 Summary and Conclusions .............................................................. 221
10.1.1 Instrumentation and Monitoring of in Service
RFERS ................................................................................ 221
10.1.2 Numerical Analysis of In-Service Structure ........................ 223
10.1.3 Approximate Expressions for Lateral Deflection of
Frames ................................................................................. 225
10.1.4 Relationship between Temperature and Earth
Pressure ............................................................................... 225
10.1.5 Numerical Analysis of Earth Pressure at Backfill
Stage .................................................................................... 226
10.1.6 Thermal Parametric Analysis of Single Story RFERS ........ 226
10.1.7 Thermal Parametric Analysis of Multi-story RFERS .......... 229
10.2 Recommendations ........................................................................... 231
10.2.1 Analysis of RFERS at the Initial Backfill Stage.................. 231
10.2.1.1 Recommended Procedure for Single Story
RFERS ................................................................. 232
10.2.1.2 Recommended Procedure for Multi-story
RFERS ................................................................. 232
10.2.2 Analysis of RFERS Subject to Temperature Variations...... 233
Appendix A .................................................................................................... 235

A.1 Thermal Soil Structure Interaction of Single Story
RFERS ........................................................................................... 235
A.1.1 Backfill Soil with 30º Internal Friction Angle ..................... 235
A.1.1.1 Three-Bay Frame (Bay Length, Lb, 10 Feet,
A.1.1.5 Six-Bay Frame (Bay Length, Lb, 10 Feet,
A.1.1.6 Six-Bay Frame (Bay Length, Lb, 10 Feet,
A.1.1.7 15-Bay Frame (Bay Length, Lb, 10 Feet,
A.1.1.8 15-Bay Frame (Bay Length, Lb, 10 Feet,
Contents XXI
A.1.2 Backfill Soil with 40º Internal Friction Angle ..................... 260
A.1.2.1 Frames with Bay Length, Lb, 10 Feet, and
Appendix B .................................................................................................... 279

B.1 Thermal Soil Structure Interaction of Multi-story RFERS ............. 279
B.1.1 Two Story Rigidly Framed Earth Retaining
Structures............................................................................. 279
B.1.1.1 Frames with Bay Length, Lb, 10 Feet, and
B.1.2 Four Story Rigidly Framed Earth Retaining
Structures............................................................................. 297
References ..................................................................................................... 307
Subject Index................................................................................................. 313

About the Authors
Walid Aboumoussa, PhD, PE, is an adjunct

Professor at NYU Polytechnic School of En-
gineering (formerly Polytechnic Institute of
New York University, Polytechnic University
or Brooklyn Poly), where has been the lead
instructor of the Cap Stone design class for
nearly a decade. He received B.Sc., M.Sc.,
and Ph.D. degrees from Polytechnic Univer-
sity. Dr. Aboumoussa specializes in
structural design and rehabilitation. His re-
search interests include concrete structures,
underground excavations and earth retaining
structures. Previously he was a partner with
Antonucci and Associates, Architects and
Engineers, LLP. Presently, he is President of
AVS Inc., and a principal at Ventrop Engi-
neering Consulting Group, PLLC. He can
be reached at walidam@avsnyc.com.
Magued Iskander, PhD, PE, F.ASCE is

Professor of Geotechnical Engineering
and Head of the Civil and Urban Engi-
neering Department at New York Univer-
sity (NYU). He specializes in foundation
engineering, soil structure interaction,
and modeling with transparent soils. He
has over 130 publications, including 3 au-
thored books, 9 edited books, and 65
journal articles and special publications.
He conducted over $9 million in grants
and contracts and graduated 35 doctoral
and masters’ students. He also served as a
reader on the dissertations of 19 doctoral
graduates at several universities. He can
be reached at Iskander@nyu.edu
Chapter 1
Introduction to Rigidly Framed Earth Retaining
Structures (RFERS)
Abstract. Structures placed on hillsides often employ the building frame to retain
earth on one side only and derive their resistance to lateral earth pressure from rig-
id frame action, without the presence of any other restraining elements or forces
against lateral displacement. The relationship between temperature and earth pres-
sure acting on rigidly framed earth-retaining structures (RFERS), subject to wide
temperature variations, is explored in this book through a 4.5 year monitoring
program of an instrumented RFERS, as well as numerical analyses and parametric
studies. The study demonstrates the important role of thermal cycles on earth pres-
sure and structural integrity.
1.1 Rigidly Framed Earth Retaining Structures (RFERS)
Unfavorable land features coupled with financial constraints often result in limiting
the options that design professionals may have when preparing plans for real estate
improvements. For example, structures placed on hillsides often present a number of
challenges and a limited number of economical choices for site design. Fig. 1.1 illus-
trates a hypothetical situation where a building structure is located on a hillside,
where, in addition to gravity and other environmental loads, the structure is subject-
ed to lateral earth pressure applied by the retained soil. The resistance of such struc-
tures against earth pressure is typically derived from lateral load resisting elements
such as rigid frames, braced frames, shear walls, or combinations thereof.
We therefore define a rigidly framed earth retaining structure (RFERS) as a
structure that retains earth on one side and derives its resistance to lateral earth
pressure from rigid frame action, without the presence of any other restraining
elements or forces against lateral displacement.
Although the mechanisms of analysis and design of free standing rigidly
framed structures are well documented in the literature and in building codes, the
behavior of RFERS structures retaining soil received little or no attention. Of par-
ticular interest in the behavior of RFERS are (1) the magnitude of initial lateral
earth pressure exerted by the backfill, and (2) the subsequent effects of soil-
structure interaction due to temperature expansion and contraction cycles on the
strength and serviceability requirements of RFERS.
W. Aboumoussa and M. Iskander, Rigidly Framed Earth Retaining Structures, 1

Springer Series in Geomechanics and Geoengineering,
DOI: 10.1007/978-3-642-54643-3_1, © Springer-Verlag Berlin Heidelberg 2014
2 1 Introduction to Rigidly Framed Earth Retaining Structures (RFERS)
AREA TO BE
BACKFILLED
EARTH PRESSURE
AREA TO BE
EXCAVATED
AREA TO BE
BACKFILLED
Fig. 1.1 Example of a Rigidly Framed Earth Retaining Structure
1.2 Initial Lateral Earth Pressure
Whether a rigidly framed earth retaining structure is erected behind an existing

soil mass or the retained soil is backfilled behind the completed structure, the mo-
bilization of stresses in the retained soil is of interest. Almost all lateral earth pres-
sure theories aspire to describe the magnitude of forces exerted by the retained soil
on structures such as flexible, rigid, braced, or anchored retaining walls. The au-
thors are not aware of any other published research on the behavior of soil re-
tained by rigidly framed structures with unrestrained lateral displacement. RFERS
are a special case of retaining structures, given the potential complexity in config-
uration and mechanism of resistance to lateral load compared to simple retaining
walls. Additionally, since RFERS resist earth pressure forces through a larger
number of structural elements than is the case for simple retaining walls, a sub-
stantially conservative design of these elements have greater impact on the econ-
omy and viability of the structure, and conversely an unsafe design may lead to
great dangers to the building and its occupants.
Consequently, the study of the mobilization of earth pressures behind rigidly
framed earth retaining structures is needed, otherwise design procedures will re-
main dependent on idealizations and simplifications that may not accurately pre-
dict the forces RFERS are required to resist. The assumption that full active earth
pressure is mobilized may not be accurate for stiff rigid frames, and the postula-
tion that the frame movements are such that the soil mass remains at rest may
prove too unrealistic.
Several major building codes adopted in the United States of America pre-
scribe the magnitude of soil lateral loads on structures, as shown in Table 1.1.
1.3 The Effects of Thermal Movements 3
These codes, however, do not separate between different types of retaining struc-
tures. The two codes, ASCE 7-10 and IBC 2000, include provisions for unyield-
ing walls, where the magnitude of earth pressure is increased from active toward
earth pressure at rest. The provision is provided in footnotes that read “…For
relatively rigid walls, as when braced by floors, the design lateral soil load shall
be increased… Basement walls extending not more than 8 feet below grade and
supporting light floor systems are not considered relatively rigid walls.” Conse-
quently, no guidance can be found in the major building codes for the determina-
tion of the magnitude or shape of the lateral earth pressure behind rigidly framed
structures with unrestrained lateral displacement.
In order to advance the state of the art related to the lateral soil loading on rig-
idly framed earth retaining structure, a study of the effects of the lateral stiffness
of RFERS and the properties of the retained soil on the mobilization of lateral
earth pressure is conducted within the framework of classical earth pressure
theories and presented herein.
Table 1.1 Selected Soil Lateral Loads in Model Codes and Standards
Design Lateral Soil Loada

Unified (psf per foot of depth)
Description of Backfill Material Soil
ASCE BOCA SBC IBC
Class.
7-10 1999 1999 2000
Well graded clean gravel, gravel-sand mix GW 35 30 30 30
Poorly graded clean gravel, gravel-sand mix GP 35 30 30 30
Silty gravels, poorly graded gravel-sand mix GM 35 41 45 40
Well graded clean sand, gravelly sand mix SW 35 30 30 30
Poorly graded clean sands, sand gravel mix SP 35 30 30 30
Silty sands, poorly graded sand-silt mix SM 45 41 45 45
Sand-silt clay mix with platic fines SM-SC 85 44 45 45
Clayey sands, poorly graded sand-clay mix SC 85 48 60 60
Mixture of inorganic silt and clay ML-CL 85 44 60 60
a. ASCE 7, IBC 2000—Design lateral loads are given for moist conditions for the specified soils
at their optimum densities.
1.3 The Effects of Thermal Movements
A more complex and significant aspect of the behavior of rigidly framed earth re-
taining structures is the soil-structure interaction resulting from volumetric strains
undergone by the structure due to temperature variations. This soil-structure inter-
action may conceivably be negligible for structures subjected to slight temperature
variations, but for a RFERS where thermal movements are large, such as open
4 1 Introduction to Rigidly Framed Earth Retaining Structures (RFERS)
parking structures, the effects of temperature cycles on the movements and stress-
es in the structure and retained soil would be substantially more pronounced. In
some respects, this behavior of rigidly framed earth retaining structures is similar
to bridges with integral abutments where temperature movements of the bridge
deck cast integrally with the bridge abutment result in cyclic interaction between
the soil and the abutment structure. In the latter case, however, the soil-structure
interaction is less complex than is the case for RFERS, given that generally inte-
gral bridges can be mostly treated as single story rigid structure restrained against
lateral movements by soil at both ends of the bridge, while RFERS may be multi-
story, multi-bay structures with lateral displacements unrestrained at one end.
Nonetheless, extensive research has been conducted recently documenting the
soil-structure interaction of integral bridge abutments.
1.4 Scope of Contribution to the Behavior of REFRS
The relationship between temperature and earth pressure acting on rigidly framed
earth-retaining structures (RFERS), subject to wide temperature variation, is ex-
plored in this book. After a review of relevant literature on earth pressure is
presented in Chapter 2, three distinct studies are presented, as follows:
1.4.1 Closed Form Expressions for Lateral Deflection of RFERS

A simplified closed form analytical expression is formulated in Chapter 3 for cal-
culating the lateral drift of low rise rigidly framed structures subjected to hydro-
static, uniform, seismic, or semielliptical loading. Additionally, the general form
of the equations can be used to predict the magnitude of the lateral force even if
the shape of the earth pressure is unknown, with a reasonable degree of accuracy.
A statistical analysis determined that the expression had better than 80% probabil-
ity to yield deflections that are within 25% of the value computed using finite
element analysis (FEA).
1.4.2 Experimental Study

An extensive review of the literature resulted in the conclusion that no published
research on the soil-structure interaction of rigidly framed earth retaining struc-
tures was available. For this reason, an experimental study of a full scale RFERS
in service was conducted to understand and document its behavior.
The structure, described in Chapter 4, is a four-story reinforced concrete park-
ing garage, open to the elements on three sides and subjected to large temperature
variations. A visual survey of the building revealed severe distress in several
structural elements and the failure of one column on the top level. Furthermore,
the building structure underwent large lateral deformations nearing 3 inches on the
same level. The magnitude of these displacements could not be validated through
1.4 Scope of Contribution to the Behavior of REFRS 5
a three-dimensional finite element analysis of the structure with a classical earth

pressure load. The back analysis indicated that a linear soil load with a coefficient
of lateral earth pressure, K, of 2 was required to reproduce the displacements
measured in the field. This large value of K could not be justified for the backfill,
and it became apparent that a more complex soil-structure interaction was respon-
sible for the deformations and severe structural distress in the building.
After some repairs, movement of the building was monitored, and recorded
hourly for a period of four and half years. The monitoring revealed complex
temperature-dependent soil structure interaction.
In Chapter 5, the closed form expressions derived in Chapter 3 are used to
obtain the relationship between temperature and earth pressure for the structure
monitored in Chapter 4. The data indicates that the coefficient of earth pressure
behind the monitored RFERS has a strong linear correlation with temperature.
During the cold season the building contracts, and the retained soil follows. Dur-
ing the hot season, the building is unable to overcome the earth pressure, thus it
expands away from the soil, resulting in a cumulative annual displacement. The
coefficient of lateral earth pressure changed by approximately 0.005/°C varying in
the range of 1.25 to 1.5, depending on the season. The study demonstrates that
thermal cycles, rather than lateral earth pressure, caused some of the structural
elements to fail.
1.4.3 Numerical Analysis

A numerical analysis of the instrumented structure studied in Chapter 4 was also
performed, simulating the backfill stage and subsequent temperature cycles. The
resulting deformations and stresses in the retained soil and structural elements are
presented in Chapter 6. Next, the results of parametric numerical studies on a mul-
titude of rigidly framed earth retaining structures with variable sizes and proper-
ties are presented. First, earth pressure behind RFERS at backfill stage is explored
in Chapter 7. Second, analysis of single story RFERS subject to temperature varia-
tion is investigated in Chapter 8. Finally, multi-story RFERS subject to tempera-
ture variations are analyzed in Chapter 9. Recommendations and conclusions are
presented in Chapter 10.
Chapter 2
Classical Earth Pressure Theory Related to
Framed Structures
Abstract. The development of earth pressure theory as it relates to rigidly framed

earth retaining structures is chronicled in this chapter. Inadequacies of classical
theories are explored along with studies related to earth pressure acting against in-
tegral bridges, which resemble, in some ways, a single story rigidly framed earth
retaining structure.
2.1 Introduction
Earth retaining structures have been used for centuries to support vertical or near
vertical soil and rock faces, as well as slopes steeper than their angle of repose.
This class of structures has taken the numerous forms, constructed of various ma-
terials, such as wood, masonry, stone, steel or concrete, or a combination thereof,
with diverse geometries and stability mechanisms.
A very common type of retaining structure is the backfilled retaining wall, of-
ten constructed of timber, masonry, or concrete. Early retaining walls consisted
mostly of stone, masonry or mass concrete, where they derived their resistance to
the retained soil loads from their own weight. These walls are known as gravity
walls. More recently, however, and with the introduction of reinforced concrete,
other types of retaining walls became more dominant, such as the cantilever, semi-
gravity, and counterfort walls, sometimes referred to as structural walls. Unlike
their earlier counterparts, these walls relied on the weight of the retained backfill,
as well as on their own weight, for stability.
Alongside the extensive use of structural walls, other retaining wall systems
were also growing in popularity, examples of which are the crib walls, and the re-
inforced-earth walls. These structures derived their stability from their own weight
and from the strength of the soil retained within them.
Where deep excavations were proposed, in temporary works or on permanent
basis, the earth retention systems of choice took the form of sheet pile walls, dia-
phragm walls, or bored-pile walls. The latter structures were also built with vari-
ous materials, such as wood, steel or concrete, and were often installed in place
prior to the earthwork. They are generally built as cantilevered, anchored, or
braced walls.

8 2 Classical Earth Pressure Theory Related to Framed Structures
Irrespective of the earth retaining system used, two requirements must be satis-
fied: The stability of the retention system as a whole (external stability) and the
strength and stability of the retaining structure itself (internal stability).
2.2 The Development of Earth Pressure Theory
The earth pressure problem dates back to the beginning of the 18th century. Gau-
tier (1717) lists five areas requiring research, one of which was the dimensions of
gravity-retaining walls needed to hold back soil. A number of engineers such as
Bullet (1691), Couplet (1726, 1727, 1728), Belidor (1729), and Rondelet (1812),
appear to have worked on the problem, and published their findings. It was Cou-
lomb, in a paper read to the Academie Royale des Sciences in Paris on the 10th of
March and the 2nd of April 1773, who was to make the first lasting impression in
the field.
Coulomb (1776) introduced two ideas essential in soil mechanics, when he
separated the strength of materials into two components, namely cohesion and
friction. This concept introduced by Coulomb, and later refined by Terzaghi to in-
clude the effective stress concept, remains the basis of soil-strength theory today.
Coulomb also considered the case of a rigid soil mass sliding upon a shear
failure surface, which formed the basis for his equations to calculate the lateral
earth-pressure on retaining walls (Heyman 1997).
In 1808 Mayniel extended the work of Coulomb (1776) and others namely
Woltmann (1794) and Prony (1802) to give a general solution for a frictional,
non-cohesive soil, with wall friction.
Müller-Breslau (1906) expanded further on Mayniel’s work to give a general
solution for a frictional cohesionless soil that allows for sloping backfill behind
frictional retaining walls. Müller-Breslau’s equation took the following form:
1 f1
Qa = γ H 2 (2.1)
2 sin α.cos δ
sin 2 (α + ϕ ).cos δ
where, f1 = 2
sin(ϕ + δ )sin(ϕ -β )
sinα.sin(α -δ ) 1+
sin(α -δ )sin(α + β )
Figure 2.1 illustrates Müller-Breslau’s solution for a frictional cohesionless soil
in active state. This solution could also be obtained for the passive state.
The previous solutions, however, were all developed in term of total stress for a
rigid soil mass that was defined on a critical discrete planar shear surface. In 1857
Rankine extended on this earth pressure theory in his paper “On the stability of
loose earth” by deriving a solution for a complete soil mass in a state of failure. In
his analysis, however, Rankine assumed that the resultant force on the vertical
place acts parallel to the ground surface.
2.2 The Development of Earth Pressure Theory 9
α γ, φ
Fig. 2.1 Müller-Breslau’s Solution for frictional cohesionless sloping soil and a frictional
wall
Engineers remained interested in the topic and several published monographs

have been dedicated to earth pressure including Ketchum (1907, 1911, 1919), and
Cain (1916).
In 1915 Bell extended Rankine's solution to allow for the effect of soil cohe-
sion. Next, following the work of Terzaghi in the early 1920s and the introduction
of the concept of effective stress as the controlling influence on strength and com-
pression, coefficient of earth pressure have been defined in terms of effective
stress as a function of the shear strength parameters of the soil.
2.2.1 Inadequacies of Classical Solutions

Work continued in the late 1920s and 1930s on the refinement of the earlier ana-
lytical solutions for earth pressure. Observation of excavation reported by Meem
(1908) and Moulton (1920) were found to contradict the lateral earth pressure im-
plied by Coulomb solution, namely a hydrostatically increasing pressure with
depth, which signifies that the wall undergoes rotation about its base.
If a wall is constructed before the soil is placed, or if a wall can be placed with
minimal disturbance to the soil, then before the wall is allowed to move, the pres-
sure on the wall will be the earth pressure at rest. For the soil in the wedge adja-
cent to the wall to reduce the earth pressure load on the retaining wall, the soil
must move from the at-rest condition to the active condition. This change in nor-
mal stress accompanied by an increase in shear stress implies a volumetric in-
crease in the soil. Therefore, to achieve the active condition implied by Coulomb,
the wall must rotate about its base.
Ohde (1938) and Terzaghi (1941) proposed two relatively complex analyses,
neither of which is in much use due to their complexity, which demonstrated dif-
ferent results for wall rotating about the top or for translating walls. These ideas
were expressed by Terzaghi (1943) in idealized relationships between the average
yield of the wall and the coefficient of lateral earth pressure depending on the type
of wall movement.
Other assumptions in both the classical Coulomb and Rankine earth-pressure
theories have been criticized by Terzaghi (1936) and others. Despite these criti-
cisms, for engineering purposes the Coulomb solution remains sufficiently accu-
rate, and the Rankine solution remains popular and relevant for many loading
cases, particularly reinforced concrete cantilevered retaining walls with a soil
mass above the heel.
2.2.2 The Stiffness of the Retaining Wall

The classical solutions involve rotation of a rigid wall about its base. In practice,
however, not all structures behave in an approximately rigid manner, such as sheet
piles, which are substantially more flexible than reinforced concrete gravity walls.
Rowe (1952) pointed out that the deflected shape of the wall below the excavation
level affects the bending moments in anchored sheet piles, and that reduction in
bending moment is a function of the flexibility of the wall relative to the soil.
Rowe carried out several tests on model walls of differing metal thickness sup-
porting various soils in loose and dense conditions, and was able to use the results
of his tests and similitude analysis to develop design charts that are widely used in
practice to size sheet pile walls.
2.3 The Case of the Integral Bridge Abutment
Deterioration of concrete bridges has often proved a serious problem in several

countries where water, carrying de-icing salts, penetrates through the bridge deck
joints. Consequently, jointless bridges would offer major benefits to durability and
reduce disruptive and costly maintenance efforts.
With jointless bridges that include integral abutments, however, it has been
recognized that the thermal deck movements are accommodated by soil structure
interaction between the supporting abutment and its foundation and the surround-
ing strata.
The movement of integral bridge abutments, especially due to thermal expansion
and contraction of the bridge deck, can create passive and active soil conditions in
the backfill. The soil reaction is nonlinear and varies with depth. The earth pressures
are dependent on the stiffness of the soil and the amount and nature of the wall dis-
placement, which can be a translation and/or a rotation. This interdependency of the
nature and amount of displacements both in the soil and the structure to the stresses
created by this process defines the soil-structure-interaction problem.
2.3 The Case of the Integral Bridge Abutment 11
Solutions to this problem would require iterative analysis where the soil reac-
tions are adjusted according to the amount and mode of deformations behind the
abutment where the deformations depend on the relative stiffness of the abutment
wall, bridge superstructure and the soil itself.
Passive pressures that develop behind the integral bridge abutment depend on
the soil density, soil to wall friction angle, mode of wall displacement, effect of
backfill confinement, and repeated loading.
The value of the lateral earth pressure coefficient increases with time towards
the passive limit with progressive cycles; thus yielding of soil and hence plastic
deformations can occur.
Broms and Ingelson (1971, 1972) measured the lateral earth pressure acting on
the abutments of several jointless bridges of 150 and 110 m length in Sweden. The
measurements were performed during backfill placement and compaction, and af-
ter the completion of the bridges. They concluded that the earth pressure was ap-
proximately constant immediately after the placement of the fill. The pressure at
the center of the abutment, however, was found to be larger than at the top and
bottom. The modulus of horizontal subgrade reaction increased with the distance
below the ground surface. The pressure distribution was hydrostatic when the
abutment was displaced laterally and parabolic when it rotated about its base. It
was concluded that the earth pressure increases due to self-compaction during the
cycles of thermal expansion and contractions and the maximum earth pressure can
reach up to passive earth pressure level.
Jorgenson (1983) took field measurements from a six-span concrete integral
bridge monitored for approximately one year, and the abutment pile behavior was
observed. Jorgenson reported that the movements at the two ends of the bridge
due to the change in length were not equal. No vertical movements in the abut-
ments were recorded. The pile stress calculated from the maximum movement
measurement was big enough to initiate a yield stress but not enough to generate a
plastic hinge.
Greimann et al. (1986) developed a nonlinear finite element procedure for
evaluating pile-soil interaction in integral bridge abutments. Piles were represent-
ed by beam-column elements and the soil was idealized as nonlinear springs.
Based on the analysis results, they concluded that thermal expansion introduced
some additional vertical loads leading to reduction in the vertical load carrying
capacity of the piles.
Girton et al. (1991) worked on the verification of design procedures for piles in
integral abutment bridges. They used experimental data collected for two years
from two skewed bridges in Iowa. The data consisted of air and bridge tempera-
ture, bridge displacements and pile strains. They recommended coefficients of
thermal expansion and proposed design methodologies.
Sandford and Elgaaly (1993) collected field measurements on a 20 degree, 165
ft span skewed bridge in Maine to investigate the soil pressures behind skewed
bridge abutments and the skew effects on these pressures. The results were found
to be similar to those by Broms and Ingelson (1971, 1972). They, however, rec-
orded earth pressure magnitudes at the obtuse side of the abutment twice those at
the acute side. The ratio was reported to reach 4 at the extremes of the abutments.
They observed that the effect of skew diminishes with time, indicating no signs of
cyclic stiffening of the retained soil. Furthermore, the average earth pressure at the
girder level was found to be 5 times higher than the earth-pressure at rest.
Springman and Norrish (1994) performed centrifuge tests on a 1/60 scale
model to examine the behavior of full height abutments with spread base and
piles. They concluded that large shear strain cycles could cause increase in the
bending moments of the wall, axial deck loads and severe slumping of the back-
fill. The failure mechanism was observed to be a clear wedge behind the wall.
The bending moments increased by 32% after the 75th cycle of 30 mm horizontal
displacement.
Fang et al. (1994) carried out experiments to investigate the earth pressure
against a rigid retaining wall that moved towards a mass of dry sand with a free
horizontal top boundary. The experiments revealed that the earth pressure distribu-
tion is significantly dependent on the mode of the displacement and on the loca-
tion of the point of rotation. The passive earth pressure was linear when the wall
underwent horizontal translation. With wall rotation about the base, however, the
passive earth pressure was found to be nonlinear and the maximum stress was rec-
orded at about mid-height of the wall.
Thippeswamy et al. (1995) analyzed five jointless bridges in service by means
of 2-dimensional frame models using one dimensional beam theory and concluded
that the earth pressure, even in the passive case, produces negligible stresses in the
bridge abutments.
Additional centrifuge tests conducted by Springman et al. (1996) indicated that
the horizontal stresses resulting from cyclic expansions and contractions of the
bridge deck remain approximately constant to depths of up to 6 m, and the pres-
sure distribution acting on the abutment is similar to the classical compaction
stress distribution with typical magnitudes ranging between 25 and 50 kPa. This
observation suggests that the use of a constant soil stiffness value with depth (for a
given strain) is reasonably realistic. According to test results the passive pressure
increases throughout the first thermal cycles, but the change becomes less pro-
nounced after the 20th cycle.
Carder and Card (1997) investigated several methods of avoiding the develop-
ment of high lateral pressures on bridge abutments due to cyclic movements of the
abutments. They recommended the usage of a low stiffness, stress absorbing com-
pressible elastic layer between the backfill soil and the abutment wall. Some of the
materials found considered to be suitable for this purpose were polystyrene prod-
ucts, polyethylene foam, rubbers, geo-composites, and geo-foam (Horvath 2004).
Ting and Faraji (1998) studied two and three-dimensional numerical models of
a 45.7 m long jointless bridge with pile supported abutments. The nonlinear soil
behavior was accounted for through the use of nonlinear springs having stiffness
values varying with depth. A 44.4°C thermal loading range was selected and the
soil compaction levels were varied. The analysis results indicated that the axial
forces in the deck, and moments in the abutment, doubled in the case of dense
backfill when compared to loose backfill. The earth pressure distribution behind
2.3 The Case of the Integral Bridge Abutment 13
the abutment wall was slightly nonlinear for a displacement of about 0.01 m, but
was expected to be more pronounced for larger displacements. The full passive
soil resistance was nearly achieved near the ground surface, but at greater depths
the pressure was approximately half of the passive value.
Lehane et al. (1999) developed a simple elastic model where an equivalent
abutment height with a single translational spring is used to simulate the soil struc-
ture interaction. Calculations using a frame analysis program were compared to
finite elements analysis and found to be in good agreement.
Thomson (1999) performed full-scale tests on integral bridge abutments found-
ed on spread footings and piles. The abutments were passively displaced into the
backfill and earth pressures and deflections were measured. Thomson reported
that the use of un-compacted sand cushion behind the abutment helped reduce lat-
eral earth pressure but that the sand has a tendency to compact after one cycle of
abutment movement.
Thomson also stated, based on his experimental work and a literature survey,
that for wall rotation about the top, the distribution of lateral pressures on the wall
is parabolic with the maximum value at the bottom. For wall rotation about the
bottom, the distribution of lateral earth pressure is also parabolic with the maxi-
mum value accruing towards the upper half of the wall. For wall translation, the
distribution of lateral earth pressure is linear having the maximum value at the
bottom. The magnitude and the distribution in this case are closest to the classical
theory. The exact location of the resultant earth pressure force exerted on the
abutment also depends on the wing-wall geometry where the location of
the resultant force moves upward as the angle between the wing-wall and the
abutment changes from a parallel position to a perpendicular position.
England et al. (2000) investigated integral bridges having lengths of 60 m, 120
m, and 160 m attached to stiff abutment walls with pinned bases. Several 1:12
scale-model tests were conducted as well as numerical analyses the results of
which indicate that the stress change in the backfill escalate quickly to reach a
hydrostatic state and remains at its peak value, while settlements continues to
change with further cycles.
As a result of controlled cyclic loading on granular soils and laboratory triaxial
tests, Carder and Hayes (2000) concluded that granular soils are likely to perform
adequately under the strain limits that are likely to develop under the spread
footing foundations of integral bridge abutments. Clayey soils, however, may
experience yielding and reduction in bearing capacity.
Barker and Carder (2000) obtained field measurements on two full height integral
bridge abutments for bridges with approximately 40 m spans. The abutment wall
and deck loads, moments, changes in the deck length, as well as abutment move-
ments were recorded. A year after the completion of construction, during the sum-
mer season, the magnitude of the lateral pressure on the abutment of one bridge was
found to be slightly larger that the earth pressure at rest. Stress escalation behind the
abutment was reported to occur after several cycles.
Lawver et al. (2000) investigated a 66 m three-span integral bridge. The bridge

was monitored during construction, and for several seasons. The bridge components
were found to have performed within design limits. The bridge abutments were
found translating laterally as a reaction to expansion and contraction of the deck.
The effect of temperature changes on the bridge was determined to be as large as
the live-load effects.
Barker and Carder (2001) evaluated the behavior of a 50 m integral bridge dur-
ing construction and subsequently over the period of three years in service. The
abutment movement and corresponding earth pressure, as well as the strains and
temperature of the deck were recorded. The observation indicated that the earth
pressure behind the abutment at the backfill stage was comparable to the earth-
pressure at rest. This pressure increased slightly above the at-rest value during the
expansion cycles, with relatively higher pressures recoded towards the top of the
abutment.
Xu et al. (2003) performed a numerical analysis of an embedded 12 m high in-
tegral bridge abutment to investigate the distribution and magnitude of lateral
earth pressure developed during thermal cycles. A ±10 mm cyclical horizontal
movement of the top of the abutment was simulated for several cycles. This analy-
sis indicated that the bending moments increased with the increase in the number
of cycles, but said increase became negligible after 20 cycles. The lateral earth
pressure behind the abutment increased approximately 20% during the same
cycles.
2.4 The Case of Rigidly Framed Earth Retaining Structures
The authors could not find any studies on the case of rigidly framed earth retain-
ing structures (RFERS) such as those presented in Chapter 1. This class of struc-
tures resembles that of the integral bridge abutment since it can also be subject to
cyclical thermal loading, however, three fundamental differences exist between
RFERS and integral bridges. First, RFERS generally retain soil on one side and
are free to move on the other, unlike integral bridges, which are braced by soil at
both ends. Second, while integral bridge abutments are generally analogous to
single story frames, RFERS are typically multistory structures. Finally, depending
on the aspect ratio of the RFERS, and given that soil is present on one end only,
the stiffness of RFERS can vary greatly.
Chapter 3
Closed-Form Expressions for Lateral Deflection
of Rigid Frames
Abstract. Determining the magnitude of the lateral deflection of the building

frame of low-rise rigidly framed structures, due to external loads, is needed in
order to meet appropriate serviceability and design code requirements. In this
chapter, a simplified rational closed form analytical expression is formulated for
calculating the lateral deflection of low rise rigidly framed structures subjected to
different lateral force distributions varying with the height of the frame.
3.1 Introduction
Low-rise rigidly framed structures, such as small office buildings, apartment com-
plexes, and parking garages are some of the most widely built structures in the
world. Determining the magnitude of the lateral deflection of the building frame
of these structures, due to external loads, is needed in order to meet appropriate
serviceability and design code requirements. Classic methods of determining the
deflection of a low-rise rigidly framed structure can be challenging and/or time
consuming; especially when the building frame has many members. Therefore, the
lateral deflection of such structures is typically obtained using the finite element
method (FEM), which requires tedious data entry and is subject to human error.
The availability of a simple method that can be used to determine the lateral de-
flections of rigidly framed structures would simplify the design process and
provide an efficient means to verify computer aided calculations.
In this chapter, a simplified rational closed form analytical expression is formu-
lated for calculating the lateral deflection of low rise rigidly framed structures sub-
jected to different lateral force distributions varying with the height of the frame,
as shown in Fig. 3.1.
(a) (b) (c) (d)

Fig. 3.1 Pressure Distributions Used in this Study (a) Hydrostatic (Earth Pressure), (b) Tri-
angular (Wind), (c) Uniform, and (d) Semi-Elliptical

16 3 Closed-Form Expressions for Lateral Deflection of Rigid Frames
3.2 Approximate Drift Analysis of Buildings

A generalized method for estimating the drift of high-rise structures has been pro-
posed by Heidebrecht and Smith (1973, 1974). The building was modeled as a
combination of flexure and shear vertical cantilever beams interconnected by a
number of rigid members that transmit horizontal forces between both cantilevers.
The method was extended to asymmetric structures (Rutenberg and Heidebrecht
1975), and generalized to include the analysis of braced frames, rigid frames, and
coupled shear walls (Smith et al. 1984). Although hand calculations can be used
with this approach, the equations are tedious, so the finite strip method is some-
times used for decoupling frame elements to simplify the calculations
(Swaddiwudhipong et al. 1988).
In recent years there has been increasing interests in approximate methods to
predict the lateral drift under seismic loading. For example, Miranda (1999) pre-
sented an approximate method similar to Heidebrecht and Smith (1973) to com-
pute the drift of buildings responding to earthquakes in the building’s fundamental
mode and expanded the method to buildings with non-uniform stiffness (Miranda
and Reyes (2002).
Alternatively, Bang and Lee (2004) proposed an energy based analytical ap-
proach for tall rigidly framed structures. The effect of the shear deformation of
the wall and the flexure deformation of the frame were shown to be important for
tall or slender buildings. Although Bang and Lee’s approach yield good results,
the equations cannot be easily solved manually.
When a rigidly framed structure is subject to lateral loading, it deflects in a
flexure mode near the top and in a shear mode near the bottom. For tall buildings,
the deflection mode is a hybrid of both mechanisms. The available approximate
methods are geared towards taller buildings, and account for both mechanisms us-
ing a variety of approaches. For shorter structures an approximation based on
shear deflections only may yield simpler equations. These expressions would be
particularly beneficial for analysis of rigidly framed earth retaining structures
(RFERS), which are commonly used in urban and suburban locations to maximize
land use in hilly sites. The expressions would also be useful for analysis of short
structures under any loading condition.
3.3 Simplified Expression for Lateral Deflection of Rigidly

Framed Structures
In a rigid frame, the lateral deflection of one floor relative to the floor below (sto-
ry drift) due to lateral pressure results from a combination of shear and bending
deformation of the beams and columns, as shown in Fig. 3.2. Bending moment
causes the greatest deflection in a long beam (length > depth). Shear forces cause
the greatest deflection in short beams (depth > length). A beam is considered
short when the span-to-depth ratio is less than 8, depending on the material
(Young and Budynas, 2004).
3.3 Simplified Expression for Lateral Deflection of Rigidly Framed Structures 17
δ s0 ,1
δ so ,n
Original Shape
Deflected Shape
w(z )
Fig. 3.2 Deflected Frame due to Lateral Pressure Distribution
δ s0 ,n δ s0,n
i=0
i =1
z
i=2
Ao
i=n
(a) (b)
Fig. 3.3 (a) Deflected Frame, and (b) Equivalent Cantilever Beam for Frame Analysis
The fundamental premise of this work is that a rigid frame can be represented
by an equivalent cantilever beam (having an equivalent area, Ao) to derive an ana-
lytical expression for the deflection, δs shown in Fig. 3.3. The frames used for
analysis have a height-to-length ratio less than or equal to 5, and may therefore be
represented as short beams and, as such, the beam bending deflection can be
neglected under transverse loads. Errors caused by this approximation can be
calibrated using FEM.
Many low-rise rigidly framed structures are designed using repeating fixed
beam and column configurations. For these structures, Iskander et al. (2012a)
stipulate that the lateral deflection at any elevation within a low rise rigid frame
is dependent on ten variables: (1) modulus of elasticity, E; (2) Poisson’s ratio, ν;
(3) total acting force, W; (4) moment of inertia of the individual columns, Ic; (5)
moment of inertia of the individual beams, Ib; (6) the height of the columns, lc;
(7) the length of the beams, lb; (8) the total number of stories, s; (9) the total
number of bays, b; and (10) the story from the top where the deflection is
desired, i.
3.3.1 Derivation of Equations for Lateral Deflection, δs

The slope of the deflection curve of a beam due to shear force, dδS /dz, is approx-
imately equal to the shear strain, γ, of any beam (Gere & Timoshenko 1984). As-
suming that the material is linear homogenous and isotropic, and applying
Hooke’s law, the shear strain, γ, is expressed as follows:
dδs Q
=γ = (3.1)
dz GAo
where, Q is the total applied shear force, and Ao is the equivalent area over which
the shear stress is acting. The lateral deflection can be obtained by integration:
Q
δs =  d δ s = 
z=H
dz (3.2)
z=0
GA0
where, H is the total height of the frame, and z is the distance from the top of
the frame to the point where the lateral deflection is to be determined. At the bot-
tom of the frame, z=H and at the top of the frame z=0 (Fig. 3.4). The total shear
force, Q, can be expressed as a function of the shear force distribution along the
height of frame, as follows:
Q=  q ( z) ⋅ dz (3.3)
An equation for the shear force as a function of the distance, z, from the top of
the frame, can be obtained by evaluating the expression for the applied load distri-
bution shown in Fig. 3.1. For example, a hydrostatic pressure, with a magnitude
ab1 , I b1
ac1 , I c1 lc1
lc 2
n
H =  l ci
i =1
a cn , I cn l cn
w( z )
lb1 lb 2 l bm
m
L =  l bj
j =1
Fig. 3.4 Legend for Stories, Bays, and Beam and Column Attributes of Rigidly Framed
Structure
of w(H) at the bottom of the frame and zero at the top, the shear force q(z) is
expressed as:
z
q ( z) = w ( H ) (3.4)
H
If W is the total acting force and q(z) is the hydrostatic pressure distribution
along the height of the frame, z, then:
2W
w(H ) = (3.5)
H
By substituting Eq. (3.4 and 3.5) into Eq. (3.3) and integrating, an expression
for the total shear force due to the hydrostatic pressure is found:
2W z 2 W 2
 q ( z) ⋅ dz = 
2W
Q= z ⋅ dz = = z (3.6)
H2 H2 2 H2
Substituting Eq. (3.6) into Eq. (3.2) and simplifying the results Eq. (3.7) is
found:
WH  z3 

z=H W W
δsh = z 2
dz = ⋅ (H 3
− z 3
) = ⋅ 1−  (3.7)
z=0
GA0 H 2 3GAo H 2 3GAo  H 3 
If the total number of stories of the frame is s, and i is the ith story from the top
of the frame (Fig. 3.3), while lc is the height of the columns (Fig. 3.4), then by
substituting H=s.lc and z=i.lc into Eq. (3.7), we obtain:
W ⋅ slc  ( ilc )  W ⋅ slc  i3 

3
δ =
h
⋅ 1− = ⋅ 1−  (3.8)
3GAo  ( slc ) 3  3GAo  s 3 
s
Eq. (3.8) is rewritten with calibration factors, α, β, in order to account for its
approximate nature, as follows:
W h ⋅ slc   i  
3
δsh = α ⋅ 1−  β   (3.9)

3GAo   s  
A similar process was used to find the deflection equation for each of the pres-
sure distributions shown in Fig. 3.1. The lateral deflection due to triangular
wind/seismic pressure δ sw is expressed with calibration factors, α, λ, β as:
W w ⋅ slc   i  i 
2 3
δsw = α ⋅ 2 − 3  λ  +  β   (3.10)
3GAo   s   s  
For uniform pressure, the lateral deflection δsu is expressed as:
W u ⋅ slc   i 2 
δ =α
u
⋅ 1−  λ   (3.11)
  s  
s
2GAo
Finally, for a semi-elliptical pressure distribution, the expression for the total
shear force due to the semi-elliptical pressure distribution is a function of a
complex number. Integrating the equation would also result in a solution with
complex numbers, and the equation for lateral deflection would not be simple.
Therefore, the standardized form of Eq. (3.9–11) is adopted for the semi-elliptical
pressure distribution and the calibration constants α, λ, β are obtained through
regression analysis, as follows:
W ⋅ slc   i  i 
2 3
δsi = α ⋅ ψ − χ ⋅  λ  +  β   (Eq. 3.13)

κ ⋅ GAo   s   s  
30
A0 = (Eq. 3.25)
 
 3 1 
lc  + 
 ( ) ( )
 ( b +1) c
I
lc
( b) b 
I
lb 
κ =3 α = 1.17
Hydrostatic Pressure Distri-
ψ =1 λ =0
bution (Eq. 9)
χ =0 β = − 78
κ =3 α = 0.96
Seismic Earth Pressure Dis-
ψ =2 λ = 9 10
tribution (Eq. 10)
χ =3 β =1
κ =2 α = 1.05
Uniform Earth Pressure Dis-
tribution (Eq. 11)
ψ =1 λ = 910
χ =1 β =0
Semi-Elliptical Earth Pres-

sure Distribution (Eq. 12) κ =2 α = 1.02
ψ =1 λ = 7 11
χ =1 β = − 23
κ =2 α = 1.03
Undefined Earth Pressure
ψ =1 λ = 45
Distribution
χ =1 β = − 25
δ si Lateral drift at i’th story; W= Total Load; s= No. of Stories; i= Story from the top (i= 0 at
top) top; lc = Length of Column; G = Shear modulus; Ao = Equivilant Area
Fig. 3.5 Approximate Story Drift of Low Rise Rigidly Framed Structure when H/L ≤ 5
(Iskander et al. 2012b)
W e ⋅ sl c   i  i 
2 3
δ =α
e
⋅ 1 − 1 ⋅  λ  +β   (3.12)
2 ⋅ GAo
s
  s   s  
The general form of Eq. (3.9– 3.12) is:
W ⋅ slc   i  i 
2 3
δs = α ⋅ ψ − χ ⋅  λ  +  β   (3.13)
κ ⋅ GAo   s   s  
where, the variables κ, ψ, and χ depend on the loading conditions shown in

Fig. 3.1.
The simplified analytical equations for lateral drift are summarized in Fig. 3.5,
where δs is the lateral deflection calculated for the ith story from the top; lc is the
height of the individual columns; α, β, and λ are calibration factors; s is the num-
ber of stories in the frame; and i is the number of stories from the top to the point
of analysis. W is the total acting lateral pressure force of the distributions shown
in Fig. 3.1, G is the shear modulus of elasticity, and Ao is the equivalent area over
which the shear stress is acting.
3.3.2 Derivation of Equivalent Area, Ao
An expression is formulated to determine the equivalent area, Ao, of the rigid

frame in a manner similar to the method used in structures to approximate a
built-up column by a homogeneous column (Iskander et al. 2012b). Fig. 3.6 illus-
trates a typical frame section under shear displacement. The equivalent shear
stiffness, GAo, is found from equating the internal and external work done by the
frame. The external work is equal to the shear force (Q) multiplied by displace-
ment (Δ). The internal work is equal to the sum of moment (M) multiplied by the
curvature (M/EI) for all members. Using (1) the equivalence of external and inter-
nal work, and (2) the cantilever method for internal virtual work, the deflection
Δis obtained. For any monolithic intersection of beams and columns, the work
done is given by:
x
M ⋅m n
 dx =  i i
Ah
Q⋅Δ = (3.14)
0 EI i=1 Ei I i
where Ai is the area of the moment diagram of the real force, and hi is the
height to the centroid of the virtual force moment diagram. Using Eq. (3.14)
and Fig. 3.6 the total work of the frame due to an external virtual force Q = 1 is
calculated:
4  Q ⋅ lc lc 1   2 lc  2  Q ⋅ lc lb 1   lc 2  (3.15)
1⋅ Δ =  ⋅ ⋅  ×  ⋅  +  ⋅ ⋅  ×  ⋅ 
EI c  4 2 2   3 4  EI b  2 2 2   2 3 
Q  lc3  Q  lc2lb 
Δ=  +  
EI c  24  EI b  12 
(3.16)
where, as shown in Fig. 3.4, lc is the distance between the centroid of the beams
(the length of the column) and lb is the distance between the centroid of the col-
umns (the length of the beam). Ic and Ib are the centroidal cross-section moment
of inertias of the columns and beams, respectively. The shear modulus of elastici-
ty G is expressed in terms of Young’s modulus, E, and Poisson’s ratio, ν. The de-
flection, Δ, can also be obtained from the shear strain γ as Δ = γ⋅lc (Fig. 3.6).
Then:
Qlc3 Ql 2l
Δ= +  c b = γ ⋅ lc =
Q
⋅ lc (3.17)
24EI c 12EI b GAo
l2
=  c + c b
1 ll (3.18)
GAo 24EI c 12EI b
The shear modulus of elasticity G can be expressed in terms of, E, Young’s

modulus, and Poisson’s ratio, ν, as follows:
E
G= (3.19)
2(1 + v )
A Poisson’s ratio of 0.17 is used in this study, which is appropriate for concrete
retaining structures. The expression for the shear modulus of elasticity then
becomes:
E
G= = 0.427 E ≈ 0.4 E (3.20)
2(1 + 0.17)
Fig. 3.6 (a) Shear Block (b) Moment Diagram of Shear Frame (c) Moment and Diagram
and Deflection of Shear Frame Section (d) Real and Virtual Force Moment Diagram for
Columns and beams (Iskander et al. 2012b)
The equivalent area, Ao, is obtained by substituting the expression for shear
modulus, (Eq. 3.20), into the expression for equivalent shear stiffness, Eq. (3.14),
and simplifying:
1
0.4 EAo =
l c2 ll
 24 EI +  12cEIb (3.21)
c b
Rearranging the equation:
1
Ao = 2
l l c lb
 60 I c
+
30 I b
(3.22)
c
30
Ao = (3.23)
 
 1 1 
lc  u + 
  c
I
 Ib 
 l c lb 
The coefficient, u, in the first term of the denominator in Eq. (3.23) is intro-
duced as a correction factor for the type of end connection. The correction factor,
u, is typically taken as 0.5 for a fixed beam (Bazant & Cedolin, 2003). The model
used for derivation is a rigid frame represented by an equivalent cantilever beam.
Therefore, a variable value of u is introduced in this model. The actual value of u
that best describes the behavior will need to be determined during regression anal-
ysis. Finally, Ao is written in terms of the number of bays, b, as follows:
30 (3.24)
A0 =
 
 1 1 
l c u + 
 (b + 1) I c  (b ) I b  
  lc   l b  
The relationship in Eq. (3.24) is used to determine the equivalent area of rig-
idly framed structures that are relatively short and thick, with a height-to-length
ratio less than or equal to 5.
3.3.3 Numerical Parametric Modeling to Calibrate Derived

Equations
Numerical parametric studies using finite element analysis were conducted to pro-
vide data to calibrate the expressions derived in Eq. (3.13, 3.24) and test their ac-
curacy. Lateral deflections were obtained by performing FEM analyses, using
Staad Pro, on 28,000 different frame and load combinations in order to calibrate
the derived expression. The modeled frames have a varying amount of stories,
bays, size and location of the columns and beams as shown in Table 3.1. Eight of
the ten parameters were varied while the modulus of elasticity, E = 3150 ksi, and
Poisson’s ratio, ν = 0.17, were held constant at values suitable for concrete.
Frames varying from one to five stories and from one to ten bays were analyzed.
Three combinations of column height and beam length, and twenty-five combina-
tions of the moments of inertia were analyzed as shown in Table 3.1. Two differ-
ent values of total applied earth load, W, were analyzed for each of the four
pressure distributions (Fig. 3.1) as shown in Table 3.2. Three different combina-
tions of column height and beam length, and twenty-five different combinations of
the moment of inertias of the beams and columns were analyzed. Finally, 14,000
additional load frame combinations were analyzed in order to verify the principal
of superposition for the derived expressions. Lateral deflections were determined
from the FEM analysis at every story level of every analyzed frame, resulting in
120,000 data sets.
3.3.4 Calibration Factors for Derived Equations

Multivariable nonlinear regression analysis (MNRA) was performed to calibrate
Eq. (3.13) using the regression analysis software DataFit (Oakdale 2002). In
MNRA, best-fit parameters for a model were obtained by minimizing the differ-
ence between all 120,000 of lateral frame deformations calculated using Eq. (3.13)
(model), and those obtained using FEM (data). The measurement of agreement of
the model and data is called the merit function, and is arranged so that small val-
ues represent close agreement between the data and the model. In MNRA the de-
pendence of the unknown calibration functions (u, α, β, and λ) is non-linear with
respect to the model, and the process of the merit function minimization is an
iterative approach. DataFit uses the Levenberg-Marquardt algorithm to adjust the
parameters. Applying the MNRA to Eq. (3.24), finds a calibration factor u = 3.
The equivalent area, Ao, is thus rewritten as:
30 (3.25)
A0 =
 
 3 1 
lc  + 
 (b + 1) I c  (b ) I b  
  lc   lb  
The calibration parameters α, β and λ for Eq. (3.13) are summarized in figure
3.5 for all loading conditions. Inspection of Eq. (3.13) indicates that load shape
plays a secondary effect on deflection, which is governed by the total lateral load,
W. For example, the maximum deflection (i = 0) is controlled by the shape effi-
ciency term, η:
α ⋅ψ
η= (3.26)
κ
η ranges between 0.39 (hydrostatic) to 0.64 (wind). The hydrostatic and triangular
wind/seismic distributions provide the minimum and maximum values of η due to
the opposite nature of these distribution shapes, while the values for the other
shapes fall in the middle of the range.
Table 3.1 Geometric and Material Properties used for Analysis (Iskander et al. 2012A)
No. of Column Beam No. of Combinations of area moments
stories Height Length bays of inertias used in FEM
lc lb Ic Ib
S (in) (in) b (in4) (in4)
2073.6 2073.6
120 120
2073.6 3110.4
1 2073.6 3110.4
120 180
2073.6 4147.2
2073.6 4147.2
180 120
2073.6 5184
2073.6 6220.8
120 120
2073.6 8294.4
2073.6 10368
2 120 180
2073.6 20736
1, 2073.6 41472
180 120 2, 2073.6 41472
3,
2073.6 103680
120 120 4,
3110.4 2073.6
5,
6, 3110.4 2073.6
3 120 180
7, 4147.2 2073.6
8, 4147.2 2073.6
180 120
9, 5184 2073.6
10 6220.8 2073.6
120 120
8294.4 2073.6
10368 2073.6
120 180
4 20736 2073.6
41472 2073.6
180 120 41472 2073.6
10368 2073.6
120 120
5
120 180
Table 3.2 Total Loads used for Analysis

Col. Beam Total Loads Values Analyzed for Values of Combination of Loads
No. of
Height length Each Pressure Distribution Used to Test Superposition
stories
(in) (in) (Kips) (kips)
Wh + Wh + Wt + Wu +
S lc lb Wh Ww Wu We
Ww Wu We We
15 & 15 & 75 & 39 & 50 + 50 + 100 +

120 120 50 + 50
50 50 100 79 100 79 79
15 & 15 & 39 & 50 + 50 + 100 +

1 120 180 75 & 100 50 + 50
50 50 79 100 78.54 79
34 & 34 & 112.5 & 59 & 75 + 75 + 150 +

180 120 75 + 75
75 75 150 118 150 118 118
60 & 60 & 150 & 79 & 100 + 100 + 200 +

120 120 100 + 100
100 100 200 157 200 157 157
2 60 & 60 & 150 & 79 & 100 + 100 + 200 +

120 180 100 + 100
100 100 200 157 200 157 157
135 & 135 & 225 & 118 & 150 + 150 + 300 +
180 120 150 + 150
150 150 300 236 300 236 236
135 & 135 & 225 & 118 & 150 + 150 + 300 +
120 120 150 + 150
150 150 300 236 300 236 236
135 & 135 & 225 & 118 & 150 + 150 + 300 +
3 120 180 150 + 150
150 150 300 236 300 236 236
304 & 304 & 338 & 177 & 225 + 225 + 450 +
180 120 225 + 225
225 225 450 353 450 353 353
240 & 240 & 300 & 157 & 200 + 200 + 400 +
120 120 200 + 200
200 200 400 314 400 314 314
240 & 240 & 300 & 157 & 200 + 200 + 400 +
4 120 180 200 + 200
200 200 400 314 400 314 314
540 & 540 & 450 & 236 & 300 + 300 + 600 +
180 120 300 + 300
300 300 600 471 600 471 471
375 & 375 & 375 & 196 & 250 + 250 + 500 +
120 120 250 + 250
250 250 500 393 500 393 393
5
375 & 375 & 375 & 196 & 250 + 250 + 500 +
120 180 250 + 250
250 250 500 393 500 393 393
3.4 Confidence in the Derived Equations 29
3.4 Confidence in the Derived Equations
In order to analyze the statistical validity of each equation as a supplement or re-

placement for FEM, a comparative statistic was necessary to compare the deflec-
tions provided by FEM (δFEM) and that obtain from the derived equations (δEq.13).
Because each point of deflection provided by FEM has a corresponding point of
deflection under similar conditions from the derived equations, the percent error
between the two data points was used, as follows:
δ Eq.13− δFEM
% error = ×100 (3.27)
δFEM
Using % error, a single population of data points representing the relationship

between the model (equation data) and the FEM analysis results, which are pre-
sumed to be the correct values (data) was created and analyzed. The % error for
each equation is graphed as a histogram to give a general idea of the distribution
of the population. As shown in Fig. 3.7, the % error takes into account the direc-
tion of error, creating a symmetric normal distribution with a mean error of -4.97
to 6.17%. However, the mean error is a misleading indicator of error because pos-
itive and negative errors cancel out. Therefore, to better assess the accuracy of the
derived equations, the direction of error must be negated. The absolute value of
the % error is:
δ Eq.13− δFEM
% error = ×100 (3.28)
δFEM
The resulting population distributions are shown in Fig. 3.8 for each equa-
tion. As expected taking the absolute value of the % error results in a folded
normal distribution. A folded normal distribution is an approximate exponential
distribution. However, because it is only an “approximate” exponential distri-
bution, the distributions are analyzed as Weibull distributions. Weibull distri-
butions provide more flexibility in distributional shape, and can approximate an
exponential, normal, or skewed distribution. Weibull distributions are popular
among engineers for analyzing strength and failure rates of materials
(Vardeman & Jobe 2001). For the purpose of this population, the % error acts
as a failure rate.
Uniform Pressure Hydrostatic Pressure
14000 14000
12000 12000
Occurence
Occurence
10000 10000
8000 8000
6000 6000
4000 4000
2000 2000
0 0
-100
-80
-60
-40
-20
20
40
60
80
-100
-80
-60
-40
-20
20
40
60
80
100
100
% Error % Error
Triangular Wind Pressure Semi-Elliptical Pressure
6000 12000
5000 10000
Occurence
Occurence
4000 8000
3000 6000
4000
2000
2000
1000
0
0 -100
-80
-60
-40
-20
20
40
60
80
100
-100
-80
-60
-40
-20
20
40
60
80
100
% Error
% Error
Uniform + SE Pressure Hydrostatic + Uniform Pressure
3500 3500
3000 3000
Occurence
Occurence
2500 2500
2000 2000
1500 1500
1000 1000
500 500
0 0
-100
-80
-60
-40
-20
20
40
60
80
-100
-80
-60
-40
-20
20
40
60
80
100
100
% Error % Error
Wind + SE Pressure Wind + Hydrostatic Pressure
3000 3000
2500 2500
Occurence
Occurence
2000 2000
1500 1500
1000 1000
500 500
0 0
-100
-80
-60
-40
-20
20
40
60
80
-100
-80
-60
-40
-20
20
40
60
80
100
100
% Error % Error
Fig. 3.7 Percent Error Histogram for each Equation

10000 14000
9000
12000
8000
Occurence
Occurence
7000 10000
6000 8000
5000
4000 6000
3000 4000
2000
2000
1000
0 0
0
10
20
30
40
50
60
70
80
90
100
10
20
30
40
50
60
70
80
90
100
% Error % Error
5000 12000
4500
4000 10000
Occurence
Occurence
3500 8000
3000
2500 6000
2000
1500 4000
1000 2000
500
0 0 0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
% Error % Error
Uniform + SE Pressure Hydrostatic + Uniform Pressure
3000 3000
2500 2500
Occurence
Occurence
2000 2000
1500 1500
1000 1000
500
500
0
0
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
% Error % Error
Wind + SE Pressure Wind + Hydrostatic Pressure
2500 3000
2000 2500
Occurence
Occurence
2000
1500
1500
1000
1000
500 500
0 0
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
% Error % Error
Fig. 3.8 Absolute Value of Percent Error Histograms for Each Equation
3.4.1 Weibull Statistical Analysis

The Weibull probability analysis involves fitting a data set with the cumulative
probability function:
 0 x<0

F(x) =   x β
− 
(3.29)
 1− e  α  x≥0
where, α > 0 is a scaled parameter of a curve measuring the spread of the data, and
β is the shape parameter of the curve, indicating whether the failure rate is increas-
ing, remaining constant, or decreasing (Vardeman & Jobe 2001). Derivation of the
Weibull parameters (α and β) to fit the data (|% error|) to the aforementioned cu-
mulative distribution, using Weibull probability plots, was performed using the
procedure provided by Dorner (1999) on excel. The Weibull procedure can be
summarized as follows: First, the |% error| is arranged in ascending order. Second,
each |%error| is assigned a corresponding rank, which ranges from 1 to n, where n
is the total number of data points. The median rank, MR, is found for each data
point by dividing i by n. A Weibull probability plot is a graph of ln(|% error|) vs.
  1  .
ln ln  
  1 − MR  
A regression analysis is run to find an approximate linear equation to fit the
Weibull plot. Figure 3.9 shows the Weibull plots and corresponding regression
linear equation for each derived deflection equation.
The Weibull cumulative distribution function can be transformed to represent a
linear equation as shown below:
 x β
− 
F(x) =1− e α 
(3.30)
β
x
ln (1− F(x)) = −   (3.31)
α 
 1   x β
ln  =  (3.32)
 1− F(x)  α 
  1  x
ln  ln   = β ln   (3.33)
  1− F(x)  α 

y = 1.1874x - 3.253 y = 1.1063x - 2.85
2
2
0
0
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 -15 -10 -5 -2 0 5
ln(ln(1/(1-MR))
ln(ln(1/(1-MR))
-2
-4
-4
-6
-8
-6
-10
-8 -12
-14
- 10
-16
- 12 -18
ln([% Error]) ln([% Error])

y = 1.2444x - 3.4762 y = 1.0176x - 2.8148
4 2
2
0
0 -10 -5 0 5
ln(ln(1/(1-MR))
ln(ln(1/(1-MR))
-10 -5 0 5 10 -2
-2
-4 -4
-6 -6
-8
-8
-10
-10
-12
-14 -12
Uniform+SE Pressure Hydrostatic + Uniform Pressure

y = 1.0926x - 2.9694 y = 1.1511x - 3.0598
4 4
2 2
0
ln(ln(1/(1-MR))
0
ln(ln(1/(1-MR))
-15 -10 -5 -2 0 5 10
-10 -5 0 5 10
-2 -4
-4 -6
-8
-6
-10
-8 -12
-14
-10
-16
-12
ln([% Error])
ln([% Error])
Wind+SE Pressure
Wind + Hydrostatic Pressure
y = 1.1101x - 3.086 y = 1.1742x - 3.1764
4
4
2 2
0
ln(ln(1/(1-MR))
0
ln(ln(1/(1-MR))
-10 -5 0 5 10 -10 -5 -2 0 5 10
-2
-4
-4 -6
-6 -8
-10
-8
-12
-10 -14
-12 -16
Fig. 3.9 Weibull Probability Plots

  1 
ln  ln   = β ln x − β ln α (3.34)
  1− F(x) 
When comparing Eq. (3.34) to the linear equation, Y = mX+b, it is evident that
the left side represents the Y, the β corresponds to the slope m, and -β ln α corre-
sponds to the intercept, b. Therefore, the value for the Weibull parameters α, β
can be obtained from the plots as:
b
− 
β 
β=m and α =e (3.35)
Table 3.3 shows Weibull parameters for each equation found from the above
Weibull probability plot. The Weibull cumulative distribution model is plotted
along with the cumulative distribution of |% error| in Fig. 3.10. The Weibull
Model does a remarkable job in approximating the data.
Table 3.3 Calculated Weibull Probability Parameters from Fig. 3.8
Weibull Parameters
Equation β b α
Uniform 1.187 -3.253 15.495
Hydrostatic 1.106 -2.850 13.147
Triangle Wind 1.244 -3.476 13.335
Semi-Elliptical (SE) 1.018 -2.815 15.883
Uniform + SE 1.093 -2.969 15.125
Unifom+Hydrostatic 1.151 -3.060 14.276
Wind + SE 1.110 -3.086 16.122
Wind+Hydrostatic 1.174 -3.176 14.959
3.4.2 Confidence in the Derived Equations

The probability for a specified percent error is found for each equation using the
cumulative distribution function:
P [ S ≤ x ] =1− e−( x/α )

β
(3.36)
such that S represents the statistical unit, in this case |% error|, x represents
the magnitude of the unit, and α and β are the Weibull parameters for each
equation.

100 100
80 80
Percent Error
Percent Error
60 60
40 40
20 20
0 0
0 0.5 1 0 0.5 1
Cumulative Probability Cumulative Probability

100
100
80
80
Percent Error
Percent Error
60
60
40
40
20
20
0
0
0 0.5 1
0 0.5 1
100
Uniform+SE Pressure Hydrostatic+Uniform Pressure
100
80
80
Percent Error
Percent Error
60
60
40 40
20 20
0 0
0 0.5 1 0 0.5 1
100
Wind+SE Pressure Wind+Hydrostatic Pressure
100
80
80
Percent Error
Percent Error
60
60
40 40
20 20
0
0
0 0.5 1
0 0.5 1
Fig. 3.10 Weibull Cumulative Probability Distribution (smooth curve) Superimposed Over
Cumulative Probability Distribution of |%error|
Inspection of Fig. 3.10 shows that derived equations have similar cumulative
probabilities, thus the probability of error for each equation was similar. Table 3.4
provides a range of desired percent errors and their associated probability for each
equation. There is approximately a 50% probability that the |% error| of deflec-
tion between all equations and FEM is less than or equal to 10%. Furthermore,
the equations had better than 80% probability to yield a deflection that is within
25% of the value computed using FEM. Each equation has close to a 90% proba-
bility of output yielding less than 30% error. More importantly, there is close to a
99% certainty that each equation will produce an output that is less than or equal
to a 50% difference from FEM.
Table 3.4 Confidence in Derived Equations
Probability that error is less than or equal to indicated percentage

P[%error≤X]
Percent Error (%) X=10% X=15% X=20% X=25% X=30% X=35% X=40% X=45% X=50%
Uniform 0.4576 0.6326 0.7584 0.8448 0.9022 0.9394 0.9630 0.9778 0.9868
Triangle Wind 0.4190 0.5929 0.7233 0.8164 0.8807 0.9238 0.9521 0.9704 0.9819
Hydrostatic 0.5223 0.6856 0.7962 0.8695 0.9172 0.9479 0.9674 0.9798 0.9875
Semi-Elliptical 0.4732 0.6231 0.7314 0.8092 0.8648 0.9044 0.9325 0.9525 0.9666
Uniform+Hydrost. 0.4852 0.6532 0.7711 0.8514 0.9048 0.9397 0.9622 0.9765 0.9855
Uniform +SE 0.4702 0.6282 0.7420 0.8225 0.8788 0.9177 0.9444 0.9626 0.9750
Wind+Hydrost. 0.4639 0.6334 0.7550 0.8393 0.8661 0.9337 0.9582 0.9739 0.9838
Wind + SE 0.4450 0.6028 0.7194 0.8037 0.8637 0.9061 0.9356 0.9561 0.9702
3.5 Effect of Input Parameters on Expression Accuracy
The Weibull statistical analysis gives a general estimate of the absolute error in
the derived equations. Therefore, the effect of various input parameters on the an-
ticipated |% error| is investigated in Fig. 3.11, where |% error| for each equation is
plotted against the following parameters:
1. The building aspect ratio (H/L)

2. Ratio of column inertia to beam inertia (Ic/Ib)
3. Number of stories per frame (s)
4. Number of bays per frame (b)
5. Location of story from the top, i, normalized by number of stories s, (i/s)
6. Aspect ratio of each rigidly framed cell (ratio of column length to beam length
(Lc/Lb)
3.5 Effect of Input Parameters on Expression Accuracy 37
100 100
10 10
|Error %|
|Error %|
1 1
Earth Earth
Wind Wind
Uniform Uniform
Semi-Elliptical Semi-Elliptical
0.1 0.1
0.1 1 10 0.1 1 I /I 10
H/L c b
100 100
10 10
|Error %|
|Error %|
1 1
Earth
Wind Earth
Wind
Uniform
Uniform
Semi-Elliptical Semi-Elliptical
0.1 0.1
0 1 2 3 4 5 6 0 2 4 6 8 10 12
No. of Stories, s No. of Bays, b
100 100
10 10
|Error %|
|Error %|
1 1
Earth Earth
Wind
Wind
Uniform
Uniform
Semi-Elliptical
Semi-elliptical
0.1 0.1
0 0.2 0.4 0.6 0.8 1 0.5 0.75 1 1.25 1.5 1.75
i/s L /L
c b
Fig. 3.11 |% Error| vs. Selected Building Properties

A second-degree polynomial equation is fitted in the |% error| for each parame-

ter. The polynomial fit represents the average |% error| trend for the available com-
parisons. Polynomials, for the most part, show |% error|, under 20%, with many
comparisons showing |% error| below 10%. The polynomials for all equations are
clustered together indicating that all equations yield similar accuracy.
The cell properties do not appear to have an effect on the |% error|. For exam-
ple the error trend for Ic/Ib and Lc/Lb is flat. Similarly the number of bays, b, has
no effect on the error trend. The building aspect ratio has a small effect on the
computed error with buildings having H/L = 2-4 showing a slightly smaller error
trend. Likewise the location where the lateral drift is being computed (i/s) has a
small effect on the |% error|. The number of stories, s, has a noticeable effect on
|% error|, with one story buildings exhibiting the most error. This is not surprising
since these structures may exhibit some flexure deflection that is not accounted for
in the formulation. In any case, the error is not large, and there are convenient
methods of estimating the deflection of single story frames.
3.6 Examples
Several examples are presented to aid the reader in applying the derived equations in
everyday situations (Table 3.5). The examples are chosen such that the parameters
employed are different from those used to calibrate the equations.
Table 3.5 Parameters used in Examples
Example 1 Example 2 Example 3

Units
Loading Type Uniform Wind or Seismic Earth
Maximum Lateral Stress Psf 250 100 1480
No. of Bays, b … 4 6 10
No. of Stories, s … 4 20 4
Height of Column, Lc ft 10 10 9
Length of Beam, Lb ft 10 12.5 15
Column Dimensions (bxd) ft 12 x 12 20 x 20 16 x 16
Wall Dimensions (txd) ft … … 12 x 180
Beam Dimensions (bxd) ft 12 x 12 16 x 24 12 x 20
Inertia of Column, Ic ft4 0.0833 0.643 0.2633
Inertia of Wall, Iwall ft4 … … 1.25
4
Inertia of Beam, Ib ft 0.0833 0.888 0.3858
Shear Modulus, G ksf 216000 216000 216000
Frame Spacing ft 10 12.5 15
3.6 Examples 39
3.6.1 Symmetric Rigidly Framed Structure Subject to Uniform

Loading
Example 1 is a 4 story, 4 bay structures subject to uniform loading, the column
length and bay width are both 10ft. Columns and beams are 12x12 in, and the in-
ertia of the columns and beams Ic= Ib= 0.0833 ft4. Frames are spaced (10 ft on
center. The loading is uniformly distributed 250 psf, therefore the 4-story frames
are subject to a two dimensional total load, W = 100 kips. The Equivalent Area, Ao
is computed according to Eq. 3.25 as follows:
30
A0 = = 0.0294 ft 2 (3.37)
 
lc  
3 1
+
 (
 ( 4 +1) 0.0833
)
10 ( 4 )
0.0833
(10 ) 

The deflection is computed according to Eq. 3.13 and the parameters shown in
Fig. 3.5 as follows:
100 × 4 ×10   i 
2
i 
3
δs ( ft) = 1.05 × 1−1× 0.9 ×  + 0 ×   (3.38)

2 × 216000 × 0.0294   4  4 
δs ( ft) = 0.3304 × [1− 0.050625i 2 ]
The deflections computed at each floor are shown in Table 3.6. The deflection
was also compared with deflections computed using 3 commercially available Fi-
nite Element packages (ETABS (CSI 2009) SAP2000 (CSI 2011) and STAAD
(REI 2002). The |Error %| are less than 5% except for the first floor, where the
error was 11%. This example illustrates the ease of use of the derived expres-
sions, and that they yield reasonable answers when used within their intended
range.
3.6.2 High Rise Building Subject to Wind Load

Example 2 is a 20 story, 6 bay structures subject to wind loading. The number of
stories exceeds the recommendations of this study, but the example is presented in
order to illustrate errors that can take place if the expressions are applied to tall
structures. The dimensions of the structure are shown in Table 3.5. Wind loading
is applied linearly varying from 100 psf at the top to zero at the bottom, resulting
in a total load, W = 100 kips. The Equivalent Area, Ao and the lateral drift, δ, are
computed according to Eq. 3.25 and 3.13 as:
30 (3.39)
A0 = = 0.333 ft 2
 
lc  
3 1
+
 (
 (6 +1) 0.643
)
10 (6)
0.888
(
12.5 ) 

100 × 20 ×10   i  
2
i 
3
δs ( ft) = 0.96 × 2 − 3 × 0.9 ×  + 1×  

3 × 216000 × 0.333   20   20  
(3.40)
δs ( ft) = 0.08897 × [2 − 6.075 ×10−3 × i 2 + 1.25 ×10−4 × i 3 ]
The deflections computed using the equations are shown in Table 3.6 along
with deflections computed using FEM, every fifth floor. As expected, the |Error
%| are much larger than in Example 1, and increase from 18% at the top to 66% in
the lower quarter.
3.6.3 Rigidly Framed Earth Retaining Structure Subject to Earth

Pressure
Example 3 is a 4 story, 10 bay rigidly framed earth-retaining structure. The di-
mensions of the structure are shown in table 3.5. A retaining wall spans the width
of the structure to resist earth pressure. Earth Pressure loading is applied linearly
varying from 1480 psf at the bottom to zero at the top, resulting in a total load, W
= 400 kips. The earth pressure corresponds to a soil density of 123 pcf and a coef-
ficient of lateral earth pressure of 0.33. The Equivalent Area, Ao and the lateral
drift, δ, are computed according to Eq. 3.25 and 3.13 as follows in Eq. 3.41 and
3.42 respectively:
(3.41)
400 × 4 × 9   i  7 i 
2 3
δs ( ft) = 1.17 × 1− 0 × 0.9 ×  + − ×  

3 × 216000 × 0.3074   4  8 4 
δs ( ft) = 0.0846 × [1− 0.010467 × i 3 ] (3.42)
The deflections computed using the equations are shown in Table 3.6 along
with deflections computed using FEM at every floor. The |Error %| was reasona-
ble, being ≤ 7% again illustrating that the expressions yield reasonable results
when used within their intended range.
3.7 Unknown Earth Pressure Distributions 41
Table 3.6 Comparison of Deflections Computed using FEM and Approximate Formulas
Location ETABS SAP STAAD Average Equation |Error %|

FEM (Fig. 3.5)
in in in in in
EXAMPLE 1 (4 Stories 4 Bays subjected to Uniform Pressure)
Roof 3.88 3.88 3.87 3.87 3.97 2.58 %

3rd Floor 3.71 3.71 3.71 3.71 3.77 1.62 %
2nd Floor 3.27 3.27 3.27 3.27 3.16 3.36 %
1st Floor 2.43 2.43 2.43 2.43 2.16 11.11 %
EXAMPLE 2 (20 Stories 6 Bays subjected to Wind/Seismic Pressure)
Roof 1.81 1.81 1.81 1.81 2.13 17.7 %

15th Floor 1.59 1.59 1.59 1.59 1.99 25.2 %
10th Floor 1.18 1.18 1.18 1.18 1.62 37.3 %
5th Floor 0.68 0.68 0.68 0.68 1.13 66.2 %
EXAMPLE 3 (4 Stories 10 Bays subjected to Earth Pressure)
Roof 1.13 1.12 0.89 1.05 1.02 2.92 %

3rd Floor 1.14 1.15 0.88 1.05 1.00 4.76 %
2nd Floor 1.09 1.07 0.85 1.00 0.93 7.00 %
1st Floor 0.83 0.85 0.63 0.77 0.73 5.72 %
3.7 Unknown Earth Pressure Distributions
The shape of the earth pressure behind RFERS is usually unknown. It is therefore
advantageous to establish a relationship between the observed deflection and the
total applied force. Therefore, Eq. 3.13 is employed for calculating the deflection
due to an earth pressure with an undefined shape. The variables κ ,ψ , and χ
were taken similar to the uniform loading and the parameters α, λ, β are obtained
through regression analysis, as shown in Fig 3.5.
Inspection of Eq. 3.13 indicates that load shape plays a secondary effect on de-
flection, which is governed by the total lateral load, W. For example, the maxi-
mum deflection (i=0) is controlled by the shape efficiency term, η:
α ⋅ψ
η= (3.43)
κ
The hydrostatic and seismic distributions provide the minimum and maximum
values of η due to the opposite nature of these distribution shapes, while the
values for the other shapes fall in the middle of the range. The range of values is
between 0.39 (hydrostatic) to 0.64 (seismic). For total load due to maximum de-
flection at the top (Eq. 25), the effect of distribution shape is a function of
κ / αψ , simply the inverse of η. Again, the hydrostatic and seismic distributions
provide the minimum and maximum values, such that the range of values for all
distributions is 1.56 (hydrostatic) to 1.96 (seismic).
The preceding discussion clearly demonstrates that for any magnitude of earth
pressure, shape of the earth pressure distribution plays a secondary role to the
geometric properties of RFERS on the computed deflection of RFERS.
3.8 Earth Pressure from a Known Deflection

Using the calibration factors tabulated in Fig. 3.5, the magnitude of the total earth
force from known deflection is found by re-arranging Eq. 3.13 as follows:
κδ sGA0
W= (3.44)
  i  i 
2 3
α slc ψ − χ  λ  +  β  
  s   s  
3.9 Case of In-Filled Frames

A large number of structures such as parking garages and office buildings have
bare frames only, where the equations are applicable. In cases where these struc-
tures are laterally loaded, say by earth pressure, failure may occur if the effect of
earth pressure is not accounted for (Iskander et al 2001). Additionally, masonry is
rarely used in modern North American construction, and “sheetrock” partitions
that are typically used do not provide considerable restrain to lateral deformation.
Nevertheless, there are cases where frames may be in-filled. The in-fill pattern of
buildings vary drastically based on many factors, including architectural require-
ments, local norms, frame type, structure purpose, date of construction, local regu-
lations, and economic factors. Nevertheless, a large number of residential and
office buildings have open spaces in first floors to accommodate lobbies, com-
mercial spaces, and parking garages. Therefore, it is reasonable to assume that the
vast majority of structures have open first floors even if the higher floors are fully,
or partially in-filled. Example 1 and 3 presented previously were re-analyzed with
100 mm (4 in.) wide masonry infill having E=4.83 GPa (700 psi) using SAP2000
v.12 FEM software. These examples were chosen because they are within the in-
tended usage range of the equations. FEM simulations show that absence of infill
in first floor significantly increases lateral deformation of the frame, over frames
that are in-filled in the first floor (Fig. 3.12, 3.13). In comparison, the estimated
lateral deformation of the buildings calculated by the proposed formula is more
reasonable than assuming an in-filled first floor.
3.9 Case of In-Filled Frames 43
3
Floor Level
Formula
1
No infill
Infill in all bays no 1st floor
Infill in 1st bay all floors
0
0 20 40 60 80 100 120
Lateral Deflec on (mm)
Fig. 3.12 Effect of Infill on Lateral Deformation of Frame Shown in Example 1 (4 stories, 4
bays, subjected to uniform pressure)
3
Floor Level
No infill
1 Formula
Infill in all bays no 1st floor
Infill in 1st bay all floors
0
0 5 10 15 20 25 30 35 40
Lateral Deflec on (mm)
Fig. 3.13 Effect of Infill on Lateral Deformation of Frame Shown in Example 3, with infill
between all bays (4 stories, 10 bays, subjected to earth pressure. A retaining wall spans
entire length).
It is evident, that for tall and slender buildings the effect of infill will be signifi-
cant, but the developed expressions are not intended for these cases. In any case,
all similar approximate closed-form expressions have some limitation, and should
be used with consideration of their boundary conditions. Additionally, closed-
form formulas cannot be expected to accommodate variations in structural
systems, while maintaining ease of use.
3.10 Limitations of the Developed Equations
The derived expressions yield good results for low-rise structures. For example
the lateral drift of short symmetric structure can be predicted with nearly 98%
confidence. However, as the structure becomes taller, flexure deformation that is
neglected in our formulation, become dominant. The error for tall structures in-
creases exponentially as seen from example 2. Additionally, the expressions are
not valid for determining the lateral drift at the ground level.
3.11 Conclusions
Closed form equation that can be used to determine the lateral deflections of rigid-
ly framed structures were derived based on principles commonly employed in
structural analysis. The derived formulas provide a simple and accurate method to
approximate the lateral deflections of a low-rise rigid structure with height equal
to or less than 20% of their length. Using the general derived equation for deflec-
tion (Fig 3.5), if one value of displacement or pressure is known (or assumed) the
other can be computed. Furthermore, the deflection of a structure under multiple
loads can be found using superposition of the deflection equations. Consequently,
for the small deflections encountered for low rise rigid frame structures, these
equations provide reasonable accuracy for preliminary design.
Inspection of the equations illustrates that for any magnitude of lateral earth
pressure force the shape of the earth pressure distribution is less important than the
geometric and material properties of RFERS when calculating the deflection. In
fact statistical analysis of the data reveals that there is an 80% probability of cal-
culating a deflection that is within 25% of that computed using FEM, even when
the shape of the earth pressure is unknown. As a result, the derived formulas
provide simple and reliable method for predicting the relationship between lateral
displacement and earth pressure for RFERS. These formulae can serve as a
tool for (1) preliminary design (2) validation of numerical codes, and (3) deter-
mining the earth pressure acting on the frame of distressed structures, as shown in
Chapter 5.
Chapter 4
Case Study of a Full Scale RFERS in Service
Abstract. A distressed Rigidly Framed Earth Retaining Structure (RFERS) open

concrete garage that retains 11 m (36 ft) of soil was instrumented. After some
repairs, movement of the building was monitored and recorded hourly. The
monitoring revealed complex temperature-dependent soil structure interactions,
which are reported in this chapter.
4.1 Introduction
A structural condition survey of a partially underground (hillside) car parking

structure revealed the presence of lateral deformations on the order of a few
centimeters that led to severe structural distress. The structure, shown in Fig. 4.1 and
labeled PG-1 hereafter, is a four-story reinforced concrete building including a full
basement, with a rectangular footprint measuring 52.42 m by 71.32 m (172 ft by 234
ft). The building’s structural system, shown in Fig. 4.2, consists of reinforced
concrete waffle slabs supported by rectangular reinforced-concrete columns.
The north side of the building is a reinforced concrete retaining wall cast
against earth at full height with a thickness of 457 mm (18 in) at the base tapering
down to 305 mm (12 in) on top. On the southern side, a one-story high, 305-mm
(12-in) thick, cast-in-place concrete wall provides enclosure for the basement. The
floor slabs are cast monolithically with the wall and steel reinforcement extended
in the slab and wall. Resistance to lateral loads is provided by (1) the northern
retaining wall, (2) irregularly placed concrete walls at the eastern edge, and (3) the
lateral stiffness of the reinforced concrete frame.
The building is openon three sides and is subject to large temperature
variations.Fig. 4.3 illustrates a longitudinal section of PG-1.
An optical survey revealed that the structure underwent a maximum lateral drift
of 76 mm (3 in) in the north-south direction, and a drift of 25 mm (1 in) in the east-
west direction, at the roof level, during its 25 years of service. The amount of
deflection measured decreased from the highest to the lowest level, which is similar
to the deflection of rigid frames subjected to lateral loads. The large movements in
both building directions indicate the presence of torsion, which is associated with
the lack of symmetry in the lateral load resisting elements. This lateral-torsional
movement induced severe cracking in several columns on the fourth-story level and
the failure of one column at the same level (Fig. 4.4).

46 4 Case Study of a Full Scale RFERS in Service
INSTRUMENTED BUILDING
STONE RETAINING WALL
71.323 m
GRADE EL. 80.162 m

GRADE EL. 88.544 m
EXTERIOR PARKING AREA
FOUR STORY
PARKING GARAGE
52.425 m
(BUILDING PG-1)
EXPANSION JOINT
RE
IN 2
A P FO R 7 S T
AR C O
TM ED RY ADJACENT
37.287 m
EN C O
T B NC PARKING GARAGE
UI L RE
DIN TE (BUILDING PG-2)
G
GARAGE ENTRANCE
RAMP
N
Location of
Retaining
Wall
PG-1
PG-2
Fig. 4.1 Site Plan and Building Layout: Schematic Plan(Top), aerial view(Bottom)
4.1 Introduction 47
Fig. 4.2 Structural Plan of PG-1 and PG-2
Adjacent to building PG-1 is another four-story parking-structure (PG-2) of

similar construction, but different footprint, which houses the concrete ramps
providing access to PG-1. The two buildings are separated by an expansion joint
along their lengths. PG-2 was not subject to earth pressure at the northern side due
to the presence of a multi-story building with an excavated cellar extending the
full height of PG-2 above the basement, at its northern edge (Fig. 4.1). No signs
of structural distress or lateral movements were observed in PG-2.
Boreholes taken behind the northern wall of PG-1 indicated that the top 6 m (20
ft) to 8.5 m (28 ft) of soil retained behind the northern wall of the building consist
of miscellaneous uncontrolled fill composed of shot rock intermixed with brown
medium to fine sand with varying amount of silt and gravel. Large boulders were
also common. The fill was underlain by a layer of medium to fine sand nearly 3 m
(10 ft) thick overlying sound rock. The ground water table was below the
building’s foundations.
Fig. 4.3 Longitudinal Section of Building PG-1 (Section A-A in Fig. 4.2)
Fig. 4.4 Photograph of Column Failure

4.1 Introduction 49
4.1.1 Mechanism of Failure

To assess the safety of the building and devise efficient repair schemes and
procedures, it was necessary to determine the cause of excessive lateral
deformations undergone by the structure and the internal stresses in the structural
elements. Preliminary structural analysis using the finite-element method
indicated that the structural deformations of PG-1 were primarily caused by a
large lateral pressure applied to the north wall of the building. The analysis
consisted of a three-dimensional model of the structure, subjected to a pressure
distribution acting on the north wall. The actual magnitude and shape of the
pressure could not be determined using traditional earth pressure theories, due to
the uncertainty about (1) the initial magnitude of backfill compaction stresses in
the retained soil, (2) the construction sequence, and (3) the magnitude of
relaxation and redistribution of earth pressure due to structural movements.
Accordingly, a simplified triangular (hydrostatic) pressure distribution was
utilized, and its magnitude gradually increased until the measured and calculated
building movements were matched. The final magnitude of the pressure
corresponded to an equivalent lateral earth pressure coefficient, k, equal to 2,
which includes the effect of hydrostatic pressure, compaction stresses, and friction
at the interface between the soil and wall. This coefficient is substantially larger
than the active or at rest earth pressure coefficients calculated using classical earth
pressure theories.
4.1.2 Objectives of the Instrumentation and Monitoring Program

At the onset of the project, the safety and stability of the structure were
questionable, and the rate of movement was unknown. In the absence of definitive
knowledge on the amount of movement, the structure was in immediate need of
stabilization and additional strengthening. Failed and severely cracked columns
were strengthened using 100–150 mm (4–6 in) reinforced concrete jackets. The
design of the stabilization scheme against additional lateral movement was based
on estimates of earth pressure obtained from the finite element analysis discussed
earlier. On the other hand, the possibility of relaxation of earth pressure due to the
large movements and the potential redistribution of the forces within the elements
of the concrete structure warranted more investigation into the behavior of the
building in an effort to minimize the cost of structural retrofit. Thus,
instrumentation and monitoring of the building became a viable and economical
alternative.
The monitoring program was carried out for four and a half years to measure
the building movements, assess the safety of the building, and determine the need
for additional strengthening and stabilization. Data from the monitoring program
revealed that the behavior of the building involved a complex soil-structure caused
by large thermal movements of the building since the building elements were
subjected to large temperature variations. The instrumentation plan and

installation procedures are presented in this Chapter.
4.2 Instrumentation
4.2.1 Selection of Instruments

The primary purpose of the instrumentation was to monitor movements of PG-1 in
the directions parallel (N-S) and perpendicular (E-W) to the applied earth
pressure. The movements were monitored in the E-W direction because the
structural survey and subsequent FEM analysis indicated the presence of torsion.
The expansion joint separating the two buildings was therefore utilized to mount
four pluck-type, vibrating-wire, displacement transducers (VW), used to monitor
the displacement of PG-1 in the direction of the earth pressure at two levels. This
was based on the assumption that building PG-2 is stationary relative to PG-1 in
the N-S direction since (1) it was not subject to a lateral earth pressure, (2) it
exhibited no signs of distress or lateral drift, and (3) the targeted PG-1 movements
were on the order of centimeters. The expansion joint was also used to mount
four additional VW transducers, used to monitor the relative displacement
between the two buildings in the E-W direction.
Tiltmeters were selected for monitoring the tilt of the northern wall in the
direction of earth pressure. Additional tilt meters were also installed along the
building perimeter to monitor for torsional movements. A tiltmeter is a precision
bubble-level that is sensed electrically as a resistance bridge. It is used to monitor
changes in the inclination of a structure, which can be used to calculate
displacement and curvature using trigonometric rules. This information can be
used to approximate the magnitude of lateral earth pressure based on the solution
of the fourth order differential equation relating the deflected shape of the wall to
the applied load. Tiltmeters were selected over other instruments, such as
inclinometers and optical surveys, due to their accuracy, ease of integration within
the facility and with the datalogger, and relatively low cost.
All VW transducers and tiltmeters were equipped with temperature sensors,
which recorded temperature along with displacements and tilts.
No signs of a general shear failure in the retained soil or translation of PG-1
were observed during the initial structural survey. The deflected shape of the
northern retaining wall also supported the assumption that movements were not
caused by a general shear failure. Therefore, since the northern retaining wall is
exposed and its tilt can be measured directly, and locating a general shear
failure surface is not required, plans to use an inclinometer were delayed until
other measurements warranted its use. Additionally, several attempts to
measure the in-situ lateral earth pressure using dilatometer and pressuremeter
testing were unsuccessful due to the often abundant presence of large boulders.
Chapter 5 presents the in-situ earth-pressure tests performed and the resulting
measurements.
4.2 Instrumentation 51
4.2.2 Instrumentation Details

4.2.2.1 Vibrating Wire Displacement Transducers (VW)
The VW displacement transducer uses a vibrating wire to detect displacements,

and can be used to monitor the movement of joints or cracks in structures.
Anchors are installed on opposite sides of the joint and the transducer is mounted
across the anchors. When movement occurs in the joint, a change in the distance
between the anchors induces a variation in the frequency signal produced by the
transducer when excited by the readout device. The readout device processes the
signal, applies calibration factors, and displays a reading in the required
engineering units. Displacement is calculated by comparing the current reading to
the initial reading.
Locations: Clusters of VW transducers (also known as jointmeters or

crackmeters) were mounted at four locations across the expansion joint separating
PG-1 and PG-2. Each cluster contained two transducers (Fig. 4.5) measuring
movement parallel and perpendicular to the joint. The clusters were mounted on
the underside of the third and roof level slabs (Levels B & D), with one cluster on
each of the northern and southern sides of the joint at each level.
Fig. 4.5 Photograph of VW-Transducer Cluster
Specifications: The VW displacement transducers used in the instrumentation

have a measuring range of 100 mm (4 in), with a resolution of 0.025% of full
scale, or 0.025 mm, and a precision is ±0.5% of full scale or ±0.5 mm. The
transducers minimum and maximum lengths are 375 mm (14.75 in) and 475 mm
(18.75 in) respectively. They are also equipped with a built-in temperature sensor
with a range of – 45 °C to 100 °C. The transducers are connected to data-logger

installed on site via shielded cables with a 22-gauge tinned copper conductors and
polyurethane jacket.
Fig. 4.6 VW Displacement Transducer across Joint
Installation Procedure: To install the VW displacement transducers across the

joint, the underside of the concrete slab was drilled at two locations across the
joint to a depth of 56 mm (2.5 in), and a high-strength fast-setting non-shrink
epoxy (Hilti HY150) was placed in the cleaned holes. No. 5 reinforcing bars
fabricated with a threaded hole on the exposed side were then introduced in the
holes and held in place by means of wood wedges. The transducer was then
mounted on the anchors by means of ball joints connected to each end, and
screwed to the threaded part of the bars. Fig. 4.6 illustrates the VW displacement
transducer installed across the construction joint.
A different setup was used to install the VW displacement transducer parallel to
the joint (Fig. 4.7). A 152 mm (6 in) long equal-legs steel angle with 50.8 mm (2
in) legs and a thickness of 12.7 mm (0.5 in) was mounted across the joint by
means of 12.7 mm (0.5 in) diameter expansion bolts anchored to the underside of
the concrete slab on each side of the joint, and bolted to the horizontal leg of the
angle. The vertical leg of the angle was threaded on one side of the joint to receive
one end of the transducer. A no. 5 reinforcing bar is also placed at the same side of
the joint. The transducer is then mounted in a position by means of a ball point
screwed to the anchored rebar and by a direct screwed connection to the vertical
leg of the steel angle. As shown in Fig. 4.7.
Fig. 4.7 VW Displacement Transducer Parallel to Joint
Each VW displacement transducer was uniquely calibrated before installation, and

a calibration record for individual transducers was produced. Upon excitation by
the datalogger, the change in frequency produced by the transducer was read in Hz
and converted into millimeters using the calibration factors and Eq.4.1 below.
D = Ax2 + B x + C (4.1)
whereA, B, and C are the manual calibration factors. Once the frequency readings
are converted in millimeters, the joint displacement was then calculated using
Eq.4.2 below.
ΔD = Di - Dd (4.2)
The displacement value obtained from Eq.4.2 was then corrected for
temperature. First, the change in temperature was found using Eq.4.3, and
converted into millimeters using the temperature coefficient obtained from sensor
calibration factors as shown in Eq.4.4.
ΔT = Ti – Td (4.3)
Tcorrected= CTxΔT (4.4)
The temperature correction was then applied by subtracting it from the value
obtained for displacement in Eq. 2, as shown in Eq. 5 below.
Dcorrected = ΔD –Tc (4.5)
where D denotes displacement (mm), T is the temperature (°C), CT is a

temperature coefficient, the indices i and d indicate subsequent measurements.
4.2.2.2 Electrolytic Tiltmeters
The electrolytic tilt sensors monitor changes in the inclination of a structure. A tilt
sensor is a precision bubble-level that is sensed electrically as a resistance bridge.
The bridge circuit outputs a voltage proportional to the tilt of the sensor. The
sensor is housed in a compact enclosure and is installed on a ballpoint mounted on
an anchor and adjusted afterwards. After an initial reading is taken, change in
inclination is found by comparing the current reading to the initial reading.
Locations: A total of 24 tiltmeters were installed. Twelve sensors were mounted

on the northern retaining wall of the building (Fig. 4.8). Each level was monitored
using three tiltmeters installed near the western, eastern, and center portion of the
wall. These sensors were connected to the datalogger and recorded hourly. The
remaining twelve sensors were installed on the exterior columns along the eastern
and southern sides of the building, and were used to monitor the tilt of the
structure in the east-west direction and verify the readings of the tiltmeters
installed on the northern retaining wall, respectively. These twelve sensors were
read at intermittent intervals using a manual readout unit.1
Specifications: The tiltmeters are capable of measuring inclinations within a ±40

arc-minutes range, with a resolution of 1 arc-second or better when read with the
datalogger or 2 arc-seconds using the manual readout unit. The sensors
measurement repeatability is ±3 arc-seconds. They operate in a temperature range
of –20 °C to +50 °C and are equipped with built-in temperature sensors. The
tiltmeters were connected to the datalogger via 24-gauge shielded cables.
1
Signal aliasing of manually read tilt sensors occurred due to insufficient reading
frequency and were discarded.
Fig. 4.8 Photograph of Electrolytic Tiltmeter
Fig. 4.9 Electrolytic Tiltmeter Details
Installation Procedure: Fig. 4.9 illustrates the installation details of the tiltmeters.
To install each sensor, a 127 mm (5 in) horizontal hole was drilled into the
concrete wall and filled with fast-setting high-strength non-shrink epoxy (Hilti
HY150). A 12.7 mm (0.5 in) diameter stainless steel threaded anchor was then
placed in the cleaned hole. The horizontal anchor was then adjusted to form a
90-degree angle with the face of the wall and to ensure vertical and horizontal
levelness of the ballpoint hardware mounted on the exposed end of the anchor.
After setting and hardening of the epoxy inside the hole, the tilt sensor was affixed
to the ballpoint and adjusted for vertical levelness and a zero initial reading (±5°).
Finally, the tilt sensor was connected to the datalogger installed on site.
4.2.2.3 Data Collection and Management
Data collection from the sensors was performed using two data-acquisition
systems. A datalogger was installed on the fourth level of the building and
mounted on the northern retaining wall 23 m from the northwestern corner, and
was connected to 8 VW displacement transducers and 12 tilt sensors all of which
were installed on the retaining wall. The datalogger was programmed to read the
sensor data at 15 second intervals and store one reading at 1 minute after the hour
in the memory. The data was then accessed and downloaded through a personal
computer via a remote modem installed on site and connected to the datalogger.
Twelve additional tilt sensors were read manually at weekly intervals, for a period
of one year, using a multi-purpose readout unit capable of displaying and
recording the sensors inclination and temperature data. The manual data were then
downloaded into a personal computer. A brief description of the data acquisition
systems is provided below for completeness.
4.2.2.3.1 Automatic Data Acquisition

The automatic dataacquisition system was comprised of eight VW displacement
transducers and twelve tilt sensors connected to two AM416 relay multiplexers
(by Campbell Scientific, Inc.) used to increase the datalogger capability of
scanning a larger number of sensors. The datalogger is capable of controlling
several multiplexers positioned between the sensors and the datalogger.
Mechanical relays are used to switch four wires simultaneously to each of the 16
channels that can be scanned.
The multiplexers were then connected to a CR10X datalogger (by Campbell
Scientific, Inc.) with non-volatile memory used to scan the sensors every 15
seconds and store their displacement or inclination and temperature every minute
after the hour. An AVW100 vibrating wire interface employing an amplifier and
filter circuit is placed between the datalogger and the multiplexer connected to
VW displacement transducers. The datalogger and the vibrating wire interface
were powered using a PS12LA lead-acid battery connected to a main AC power
supply available on site through an AC adapter. Communication with the
datalogger was established via telephone lines and a 9600-baud modem,
employing the Hayes AT command set, installed on site and connected to the
datalogger. A computer program was written to automate data retrieval and
4.3 Monitoring Results of Instrumented RFERS 57
control the datalogger. Communication was initiated by a personal computer

equipped with the PC208 datalogger support software (by Campbell scientific,
Inc.) used for automated data retrieval from the storage modules and for editing
the datalogger program.
4.2.2.3.2 Manual Data Acquisition

A manual readout unit capable of storing 8,000 time-and-date stamped readings
was employed to retrieve the data from twelve tilt sensors on a weekly basis.
Manual readings were selected in order to eliminate the high cost of wiring
required to connect remote instruments to the datalogger. The data was then
downloaded into a personal computer equipped with software used as an interface
between the readout unit and the PC. The manual unit has a 20-bit resolution and
an accuracy of ±0.03° of reading and ±0.02° of range.
4.2.2.3.3 Data Management

Data collection from the sensors was performed using two data-acquisition
systems. The datalogger programmed to record data hourly was connected to
eight VW displacement transducers, and the 12 tiltmeters installed on the retaining
wall. Manual readings were also integrated with the datalogger data.
Approximately 675,000 data records have been collected in the four and a half
years monitoring period. Each record consisted of 148 parameters. The data was
stored in a Microsoft Access database, which was queried as necessary.
4.2.2.4 Instrumentation Limitations

The temperature measured by each sensor is that of the sensor itself, and it
approximates the ambient air temperature. Consequently, there is a thermal lag
between the measured sensor temperatures and those of the concrete elements.
Due to their placement in a fully functional facility, the sensors were
susceptible to tampering or vandalism in-spite of continuous effort to notify the
building users to refrain from unauthorized interference with the equipment. This
led to loss of some data, as well as a few spurious data points, that were manually
deleted from the record.
4.3 Monitoring Results of Instrumented RFERS

The movements of the vibrating-wire displacement transducers and electrolytic tilt
sensors were recorded hourly for a period of nearly four and half years, starting
from the month of May, 1999, and ending with the month of October, 2003.
The measurements obtained from the vibrating-wire displacement transducers
(VWDT) were limited to providing information on the relative movement between
the two structures (PG-1 and PG-2) along and across the construction joint.
Consequently, to simplify the interpretation of the VWDT data, a study of the

thermal behavior of the building was performed using the measurements obtained
from the VWDT installed across the expansion joint. These sensors showed a
particularly close correlation between movement and temperature. The
coefficients of thermal expansion of the structure in service corresponding to
yearly, seasonal, and diurnal movements were determined and used to further
interpret the measurements obtained from the VWDT installed along the expansion
joint.
The data recorded by the tilt sensors will be presented and discussed first,
followed by the study of the thermal behavior of the building and VWDT data.
4.3.1 Electrolytic Tilt Sensors Data

The tilt recorded by the electrolytic tilt sensors (ETS) installed on the northern wall
of the building and connected to the datalogger are presented. Measurement
obtained from the manually read ETS installed at the southern and eastern ends of
the building were found to be unreliable due to signal aliasing and are not
presented.
Fig. 4.10 shows the locations and designations for later reference of the tilt
sensors installed at the northern building wall. The first two letters of the sensor
designation (EL) refers to Electrolytic tilt sensors. The third letter (A, B, C or D)
refers to the level at which the sensors are installed. The fourth letter (N) refers to
the Northern wall of the building. The fifth letter (W, C or E) refers to West,
Center or East side of the wall. As an example, a sensor designation EL-CNE,
denotes the tilt sensor located at level E on the northern wall, at the eastern side of
the wall. All sensors are installed approximately 30 cm (12 in) from the bottom of
the slab.
Fig. 4.10 Electrolytic Tilt Sensors Installed on Northern Wall

4.3.1.1 Electrolytic Tilt Sensors on Level A
Data collected from the three tilt sensors installed on level A (or Basement) are
shown in Fig. 4.11. The top graphic refers to the sensor at the western side of the
wall, followed by the center and eastern sensors. The temperature and angle-
change trends are respectively represented by the solid and dashed lines obtained
from linear regression analysis. A positive angle-change value indicates a
movement in the north direction, or towards the retained soil mass, associated
with an increase in temperature and related expansion movement of the building.
The data in Fig. 4.11 indicate that the tilts measured by the sensors installed at
the basement are in close correlation with the measured temperatures. The wall
movements are therefore due to the volumetric strains induced by temperature
variations in the structure. Nevertheless, while the temperature trends are nearly
constant, the tilt trends indicate a slight but constant increase in tilt.
A linear regression analysis was performed to determine the equations
describing the relationship between the tilt (α), as the dependent variable, and
temperature (T), as the independent variable. The graphics of Fig. 4.12 present the
results of said analysis, where the linear correlation coefficients of the curve-fits
ranged from 91 to 95 percent. The sensors at the western and eastern side of the
wall had nearly identical relationships, with the eastern sensor showing a closer
correlation between tilt and temperature with a linear correlation coefficient
higher than 95 percent. All sensors indicated zero tilt when the temperature neared
approximately 20ºC.
4.3.1.2 Electrolytic Tilt Sensors on Level B
Data collected from tilt sensors at level B are shown in Fig. 4.13, and unlike their
counterpart at level A, the tilt and temperature measurements of the sensors at
level B are not as closely related, although evidently the tilt movements are
induced by temperature variations.
The data from the western sensor (EL-BNW), for instance, show a constant
temperature trend, but a linear angle-change trend noticeably steeper than its
corresponding sensor at the basement level. Starting with a zero tilt and a
temperature of 15.5ºC at installation, sensor EL-BNW showed little response to the
initial rise in temperature in the spring and summer of 1999, and a continuous
increase in tilt away from the soil mass during the fall and winter seasons of the
same year. At the next temperature cycle, starting with the increase in temperature
during the spring of 2000, sensor EL-BNW recorded a tilt movement toward the
retained soil, but the sensor never reached its initial zero tilt when the temperature
reached the installation temperature of 15.5ºC. In fact, with an additional increase
in temperature of nearly 10ºC beyond the installation temperature, the sensor did
not revert to its initial position with zero tilt.
2 30
Angle Change (minutes)
1 25
Temperature ( C)
o
0 20
-1 15
-2 10
-3 5
EL-ANW Temperature EL-ANW Data
-4 0
2 30
1 24
Temperature ( C)
o
0 18
-1 12
-2 6
EL-ANC Temperature EL-ANC Data
-3 0
/ / / / / / / / / / / / / / / / / /
3 30
2 25
Temperature ( C)
o
1 20
0 15
-1 10
-2 5
EL-ANE Temperature EL-ANE Data
-3 0
5/1/99 11/22/99 6/14/00 1/5/01 7/29/01 2/19/02 9/12/02 4/5/03 10/27/03
Fig. 4.11 Electrolytic Tilt Sensors Data – Level A

5
4 α = -3.5994 + 0.16743T R= 0.91904
Angle Change, α (Minutes)
EL-ANW Data
3
2
1
0
-1
-2
-3
-4 Only 5% of the data is shown for clarity
-5
5
4 α = -2.379 + 0.12683T R= 0.95403
EL-ANC Data
3
2
1
0
-1
-2
-3
-4
Only 5% of the data is shown for clarity
-5
5
4 α = -3.1315 + 0.16797T R= 0.95752
EL-ANE Data
3
2
1
0
-1
-2
-3
-5
0 5 10 15 20 25 30
o
Temperature, T ( C)
Fig. 4.12 Angle Change vs. Temperature for Tilt Sensors atLevel A
1 40
0
EL-BNW Temperature EL-BNW Data
-1 27.5
Temperature ( C)
o
-2
-3 15
-4
-5 2.5
-6
-7 -10
1 40
EL-BNC Temperature EL-BNC Data
0 27.5
Temperature ( C)
o
-1 15
-2 2.5
-3 -10
2 40
EL-BNE Temperature EL-BNE Data
1 30
Temperature ( C)
o
0 20
-1 10
-2 0
-3 -10
5/1/99 11/22/99 6/14/00 1/5/01 7/29/01 2/19/02 9/12/02 4/5/03 10/27/03
Record Date
Fig. 4.13 Electrolytic Tilt Sensors Data – Level B

3
2
(Minutes)
EL-BNW Data = -4.2008 + 0.072653T R= 0.46535

1
0
-1
-2
Angle Change,
-3
-4
-5
-7
2
1.5
(Minutes)
EL-BNC Data = -1.0425 + 0.011656T R= 0.18258

1
0.5
0
-0.5
Angle Change,
-1
-1.5
-2
-2.5 Only 5% of the data is shown for clarity
-3
2
1.5 EL-BNE Data = -2.0969 + 0.084545T R= 0.92657
(Minutes)
1
0.5
0
-0.5
Angle Change,
-1
-1.5
-2
-2.5 Only 5% of the data is shown for clarity
-3
-5 0 5 10 15 20 25 30 35
o
Temperature, T ( C)
Fig. 4.14 Angle of Change vs. Temperature for Tilt Sensors – Level B
The observed trend of sensor EL-BNW continued with the remaining

temperature cycles causing a cumulative increase in tilt in the direction away from
the retained soil mass, resulting in a steep trend in angle-change when compared
to the temperature trend. At the last recorded temperature cycle, the measured tilt
corresponding to the installation temperature of 15.5ºC was nearly -4 minutes.
It is consequently apparent that the presence of the retained soil mass behind
the northern wall of building PG-1 resulted in the modification of the expansive
and contractive movement of the structure. Without the soil restraint present on
only one side of the building, the volumetric strains induced by temperature
variations of the rigid frame would be expected to be equal in magnitude at
corresponding temperatures, and repetitive with repeating temperature cycles. In
the case of sensor EL-BNW, the measured tilts were not equal at corresponding
temperatures, although they were fairly repetitive within individual temperature
cycles.
Examination of the data of the center sensor (EL-BNC) reveals a behavior
largely similar to that of the western sensor, with a few exceptions. The range of
movements of EL-BNC (approximately 0.5 to -2.25 minutes) was smaller than BNW
(approximately 0 to -6.2 minutes), which can be noted as well for the center and
western sensors installed at the basement level. Moreover, EL-BNC virtually
reverted to its initial position during the second temperature cycle with a
corresponding installation temperature of 15.5ºC. Nevertheless, the sensor
subsequently recorded a permanent increase in tilt during the remaining
temperature cycles, and in general, the measurements of sensor EL-BNC illustrate a
tilt trend rather steeper than the constant temperature trend, and a cumulative
movement of the northern wall away from the retained soil mass.
Contrasting with the western and center sensors data, the tilt measured by the
eastern sensor (EL-BNE) demonstrate a relatively stronger relationship between
temperature and tilt, roughly similar to the sensors installed at the basement level.
The range of movement of EL-BNE (approximately 1 to -2.5 minutes) was smaller
than BNW (approximately 0 to -6.2 minutes), and larger than BNC (approximately
0.5 to -2.25 minutes), which can similarly be noted for the sensors installed at the
basement level. The comparable difference in the range of tilts recorded by the
western, center and eastern sensors on both levels A and B could be attributed to
several factors, including (1) a difference in stiffness within the soil mass, (2) the
fact that at the western side of the wall, the soil mass is retained in part by the
northern wall of an underground walkway leading to the basement of the adjacent
apartment building, and terminates in a stable slope extending into the excavated
cellar of the latter building, thus resulting in less confinement of the soil mass at
the western end of the PG-1, (3) the presence of reinforced concrete walls acting
as shear walls at the eastern end of PG-1 on levels A and B, running the entire
length of PG-1, and (4) the stiffness distribution of the lateral load resisting
elements of PG-1.
The difference in the behavior recorded by sensor EL-BNE can also be attributed
to the presence of the concrete wall at the eastern end of the building. The wall is
part of the lateral load resisting elements of PG-1, and its presence is likely to
provide a significant additional restraint against the thermal and non-thermal
lateral movements of PG-1, resulting in the discrepancy of measurements between
the eastern sensor and its western and center counterparts.
A linear regression analysis of the data collected by level B sensors was
performed similarly to that completed for level A sensors. As shown in Fig. 4.14,
the results of the analysis indicate a rather weak relationship between the
temperatures and tilts recorded by the western and center sensors, but a stronger
relationship for the eastern sensor. The graphic of EL-BNW data indicates 5
distinct sets of linearly related tilt and temperature data, corresponding to the five
temperature cycles recorded by the sensor. Consequently, the linear correlation
coefficient of the curve-fit for the data is nearly 46%. Similarly, distinct sets of
data can be noted on the EL-BNC graphic corresponding to a linear correlation
coefficient of approximately 18%, which indicates the presence of other factors
affecting the tilt measurements of the sensor. On the other hand, the regression
analysis of the eastern sensor data demonstrates that the tilt recorded by the sensor
can be closely related with the cyclic temperature variations.
4.3.1.3 Electrolytic Tilt Sensors on Level C
Data collected from tilt sensors at level C are shown in Fig. 4.15. The
measurements of all three sensors show that the wall tilts at level C follow the
cyclic temperature variations similarly to the sensors at the lower two levels,
however the tilt and temperature data are not closely related.
The data recorded by the western sensor (EL-CNC) were generally similar to
those shown for its counterpart on level B. The initial rise in temperature after
installation did not cause an expected tilt movement towards the soil mass. The
subsequent temperature-rise cycles induced a smaller range of tilt movement
compared with the temperature-drop cycles. This indicates that the soil restraint
behind the wall impedes the ability of the structure to expand, but does not have a
similar effect to the contraction movement of the building.
Furthermore, EL-CNW did not revert to its initial position after installation, but
recorded a cumulative increase in tilt throughout the monitoring period. The
constant temperature trend was also contrasted with a linear tilt trend, similar to
the western sensor installed at level B.
The measurements of the center sensor were similar to those recorded by its
counterparts EL-CNW and EL-BNW.
The behavior of sensor EL-CNE was also in line with that of the other level C
sensors, but showed a slightly stronger relationship between temperature and tilt,
although not quite as strong as its eastern counterpart on level B. It should be
noted, however, that the length of the concrete wall at the eastern end of the
building at level C is only 14.5 m (47.6 ft) or 20% of the length of the concrete wall
at level B. This indicates that the presence of the different length walls at levels B
and C acting as shear walls resulted in additional modification of the thermal
movement of the structure at the eastern end. This modification however is clearly
dependent on the length or lateral stiffness of the shear walls.
0 40
-0.6 30
Temperature ( C)
o
-1.2 20
-1.8 10
-2.4 0
EL-CNW Temperature EL-CNW Data
-3 -10
1 40
0.5 30
Temperature ( C)
0 20
o
-0.5 10
-1 0
-1.5 -10
EL-CNC Temperature EL-CNC Data
-2 -20
1 40
0 30
Temperature ( C)
o
-1 20
-2 10
-3 0
EL-CNE Temperature EL-CNE Data
-4 -10
5/1/99 11/22/99 6/14/00 1/5/01 7/29/01 2/19/02 9/12/02 4/5/03 10/27/03
Record Date
Fig. 4.15 Angle of Change vs. Temperature for Tilt Sensors – Level C
The ranges of tilt movements recorded by level C sensors were smaller than
those recorded by level B sensors. This is typical of rigid frames where the
volumetric strains due to temperature cause the largest tilt in the vertical frame
elements spanning between the foundation (restraint) and the first level, as shown
in Fig. 4.16. In this case level B is the first unrestrained level.
Fig. 4.17 presents the results of the regression analysis of the level C sensors’
data. As expected, the linear correlation coefficients of the curve-fit for the data of
the western and center censor indicate a rather weak relationship between tilt and
temperature.
The data of the eastern sensor showed a stronger relationship, with a linear
correlation coefficient of nearly 52%. Nevertheless, all tilt data recorded on this
level had weaker relationship with temperature compared to the data obtained
from the sensors installed on the level below.
Furthermore, the graphics of Fig. 4.17 illustrate the distinct sets of linearly
related tilt and temperature data, corresponding to the temperature cycles recorded
by the sensors.
EXPANSION CONTRACTION
Fig. 4.16 Typical Volumetric Expansion and Contraction of Rigid Frames

2
α = -1.8871 + 0.019121T R= 0.28353
1 EL-CNW Data
-1
-2
-3
-4
2
α = -0.83365 + 0.013046T R= 0.26446
EL-CNC Data
1
-1
-2
-3
2
EL-CNE Data α = -2.1534 + 0.041912T R= 0.52366
-1
-2
-3
-4
-5 0 5 10 15 20 25 30 35
o
Temperature, T ( C)
Fig. 4.17 Angle Change vs. Temperature for Tilt Sensors atLevel C
4.3.1.4 Electrolytic Tilt Sensors on Level D
Data collected from tilt sensors installed on level D are shown in Fig. 4.18. The
tilt measurements of all three sensors followed the cyclic temperature variations
similarly to the sensors at the lower two levels, but with a wider range of
movement. This may be due to the fact that the soil restraint at the roof slab level
is significantly smaller than that at the lower levels, given that the top of the roof
slab is nearly at the elevation of the top of the backfill soil.
4 40
2 30
Temperature ( C)
o
0 20
-2 10
-4 0
-6 -10
EL-DNW Temperature EL-DNW Data
-8 -20
/ / / / / / / / / / / / / / / / / /
4 40
EL-DNC Temperature EL-DNC Data
2 30
Temperature ( C)
o
0 20
-2 10
-4 0
-6 -10
6 40
EL-DNE Temperature
4 30
Temperature ( C)
Temperature (oC)
2 20
0 10
-2 0
o
-4 -10
EL-DNE Data
-6 -20
5/1/99 11/22/99 6/14/00 1/5/01 7/29/01 2/19/02 9/12/02 4/5/03 10/27/03
Record Date
Fig. 4.18 Electrolytic Tilt Sensors Data – Level D

4
α = -3.2088 + 0.015915T R= 0.082011
2 EL-DNW Data
-2
-4
-6
-8
α = -3.0098 + 0.065325T R= 0.46408

4 EL-DNC Data
-2
-4
-6
6
α = -3.1206 + 0.10949T R= 0.68051
4 EL-DNE Data
-2
-4
-6
-5 0 5 10 15 20 25 30 35
o
Temperature, T ( C)
Fig. 4.19 Angle Change vs. Temperature for Tilt Sensors at Level D
Additionally, the tilt data indicate a larger diurnal scatter, with tilt varying more
than 4 minutes within a 24 hour period in some instances. The scatter was present
in all three sensor measurements with similar diurnal magnitude, and was more
pronounced during the higher temperatures. This behavior, unique to the top level
sensors, may be the result of direct exposure of the top of the uninsulated concrete
roof slab to direct sunlight and the elements, creating a temperature gradient
between the top and bottom of the slab. Consequently, the actual temperature of
the concrete slab may be different than the ambient air temperature measured by
the sensors installed inside the structure. Diurnal fluctuations in temperature, in
addition to direct exposure to sunlight, may cause a more pronounced fluctuation
of the concrete slab temperature and thus a larger scatter in daily movements.
The wall tilts at level D also show a weak relationship with temperature, and a
cumulative tilt ranging between 4 and 7 minutes at the end of the monitoring
period.The results of the regression analysis for level D sensors are shown in Fig.
4.19. The relationship between tilt and temperature for the western sensor is
clearly the weakest of all sensors, with linear correlation coefficient of less than
10%. The western sensor is the farthest away from the 14.5 m shear wall installed
at the eastern side of the building. The center and eastern sensors showed a
relatively stronger relationship between tilt and temperature, with a linear
correlation coefficient of 46% for the center sensor, and 68% for its eastern
counterpart.
4.3.2 Vibrating-Wire Displacement Transducers

The data collected by the eight vibrating-wire (VW) displacement transducers was
limited to measuring the relative movements between the floor slabs of buildings
PG-1 and PG-2 at levels B and D. In the north-south direction, the data is
interpreted based on the premise that building PG-2 is not restrained by soil at its
northern edge and is therefore free to undergo thermally induced volumetric
strains, while the retained soil for building PG-1 will provide some restraint on
thermal movements.
The two structures, PG-1 and PG-2 have identical geometrical and structural
framing schemes in the vicinity of the expansion joint where the VW sensors are
mounted, and would as a result undergo similar temperature movements.
Therefore, assuming that both structures are free to expand and contract, there
would be no relative movement between the two structures at the expansion joint
in the longitudinal (north-south) direction, and the VW sensors would record no
movements. However, given that PG-1 is not free to move as PG-2 is, the amount
of expansion and contraction movement will differ between the two structures
depending on the magnitude of restraint imposed by the retained soil on building
PG-1. The recorded movement of the VW sensors would thus equal the amount of
movement restrained during expansion, and the amount of relative movements
between the structures during contraction.
To help quantify the thermal movements of building PG-1, a thermal study was
performed using the data recorded by the sensors installed normal to the
expansion joint, in order to arrive at an apparent value of the coefficient of thermal
expansion (CTE) of the structure in service, rather than applying the average
value of the coefficient of thermal expansion of concrete (Abomoussa and
Iskander 2003). This latter value of the CTE is derived from laboratory tests on
unrestrained concrete specimens with various aggregates, cement pastes and steel
reinforcement (Callan 1952, Zoldners 1968, Berwanger 1968), and may prove to
be overestimated when applied to concrete in service, which typically has some
amount of restraint in the case of rigidly framed structures. The results of the
thermal study are presented first, followed by the data obtained from the VW
sensors installed in the north-south direction.
4.3.2.1 Thermal Study of Structure: Sensors Normal to Expansion Joint
The study of the thermal movements of the structure shown herein was performed
using the expansion joint movements recorded by the two sensors installed at the
southern side of the building at level C and roof slabs for a period of one year. The
data consisted of approximately 7,180 displacement and corresponding
temperature readings for each sensor, resulting in a total number of utilized
measurements of 28,720.
Table 4.1 Range of Annual Movements
Sensor Side South South

Sensor Level Roof C
Number of Readings 7181 7180
Displacement Range, mm 14.01 15.51
Max. Displacement, mm 3.97 6.5
Min. Displacement, mm -10.04 -9.01
Mean Displacement, mm -1.72 -0.29
Displ. Std Deviation, mm 3.08 3.24
Max. Temperature, °C 34.85 34
Min. Temperature, °C -12.38 -15.1
Mean Temperature, °C 13.92 13.14
Temp. StdDeviation, °C 10.25 9.47
4.3.2.1.1 Annual Range of Movements

The yearly movement and temperature measurements obtained from the VW
transducers are shown in Fig. 4.20, and summarized in Table 4.1. A positive
displacement indicates joint expansion, and a negative displacement indicates
contraction. The roof level sensor installed on the south side of the building
indicated a range of movement of 14.01 mm (0.55 in), over a temperature range of
47°C (85°F), and level C sensor installed on the south side underwent a range of
movement of 15.51 mm, (o.61 in) over a temperature range of 49°C (88°F).
To calculate the coefficients of thermal expansion, we define the expansion

length of the concrete as half the sum of the lengths perpendicular to the
expansion joint of the slabs on both sides of the joint. This definition is adopted to
conveniently calculate the A p p a re n t Co e ff ic ie n t s o f T h e r ma l E x p a n sio n
(ACTE) for the structure in service, and express said coefficients in temperature
units to allow their comparison. The assigned expansion length of concrete was
47.35 m (155.3 ft) for the southern sensors.
6 40
Temperature
3 30
Displacement (mm)
Displacement
Temperature ( C)
o
0 20
-3 10
-6 0
-9 -10
Roof Level Sensor, South Side
-12 -20
6 40
3 30
Displacement (mm)
Temperature ( C)
o
0 20
-3 10
Temperature
-6 0
Displacement
-9 -10
Level C Sensor, South Side
-12 -20
Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May
Record Date
Fig. 4.20 Southern VW Sensors Data (Normal to Expansion Joint)

Given the measured movements and related temperatures of the expansion

joint, we define the annual A p p a re n t Co e ff ic i e n t o f T h e r ma l E x p a n sio n
( ACT E) of thejoint, α A , as the slope of the linear regression line describing (at
95% upper confidence level) the relationship between joint movement and
temperature, divided by the expansion length of concrete.
10
Total Number of readings = 7181
Range = 14.1 mm Std. Deviation Lines
5
Displacement, mm
Std. Deviation = 0.778 mm
Correlation Coeff. R = 0.968

0
Slope = 0.29324
Mean = - 1.7 (mm)
-5

-10
Neg. Displ. = Joint Contraction
Pos. Displ. = Joint Expansion
-15
10
Total Number of readings = 7180 Std. Deviation Lines
Range = 15.51 mm
5
Displacement, mm
Std. Deviation = 0.715 mm
0
Correlation Coeff. R = 0.976 Mean = - 0.272 (mm)
Slope = 0.33579
-5
Level C Sensor, South Side

-10
-15
-20 -10 0 10 20 30 40
o
Temperature, C
Fig. 4.21 Annual Joint Movements vs. Temperature
The plots of displacement versus temperature for the entire monitoring period
of one year are shown in Fig. 4.21, along with the linear regression lines. The
annual Apparent Coefficients of Thermal Expansion (ACTE), reported in units of
percent per 100°C, are shown in Table 4.2.
The values of α A obtained from the southern sensors measurements are nearly
70% of the average Coefficient of Thermal Expansion (CTE) of concrete. Said
values are comparable to the coefficients of thermal expansion found in an
unpublished report prepared in 1943–1944 by structural engineers of the Public
Buildings Administration, which presents the expansion joint movement in nine
federal buildings for a period of one year (ACI, 1995). The report demonstrates
that the dimensional changes in the upper level of the buildings correspond to
values of the apparent coefficient of thermal expansion between 0.036 to 0.09%
per 100°C (0.02 to 0.05% per 100°F). Ndon and Bergeson (1995) reported that
values of the field coefficient of thermal expansion for the Boone River and Maple
River bridges were 0.072 and 0.085% per 100°C respectively.
Table 4.2 Annual Apparent Coefficients of Thermal Expansion
Sensor Level Roof C
Regression Line Slope, mm/°C 0.29324 0.33579
Expansion Length, mm 44,350 44,350
Annual Apparent Coefficient of Thermal Exp, %/100°C 0.0661 0.0757
Ratio of CTE to ACTE 1.51 1.32
4.3.2.1.2 Seasonal Behavior

The seasonal behavior of the expansion joints was also examined. To do so, the
data was grouped into three periods corresponding to the winter, summer, and
fall/spring seasons. However, since temperature values may overlap between
adjacent seasons, the data was grouped according to temperature ranges, and not
according to the seasons limiting calendar dates, in an effort to increase their
statistical accuracy and reliability. The seasonal temperature ranges were
determined by dividing the total temperature range measured by the sensor into
three equal intervals, resulting in the dividing temperatures shown in Table 4.3. It
should be noted, however, that the data for the spring and fall seasons are not
collected consecutively, but are separated by the summer season.
The plots of seasonal displacement versus temperature are shown in Fig. 4.22
along with the linear regression lines. The seasonal Apparent Coefficients of
Thermal Expansion, α s , are also calculated, similarly to their annual counterparts,
and are shown in Table 4.4.
Except for the south sensor at the roof level, where measurements across the
summer season show a wider scatter of displacement vs. temperature, the
calculated values of α s fall within a close range of each other, and of the values
of αA. The average values over the four seasons for both sensors are
approximately 46% smaller than the average value of the average coefficient of
thermal expansion of concrete. This demonstrates that the seasonal values of
thermal movements of open concrete buildings may be adequately determined
using the average CTE of concrete, provided no additional mass and restraint from
concrete walls is imposed.
More information about the effect of structural restraint on the observed annual
and seasonal ACTE is available in Aboumoussa and Iskander (2003) and Iskander
et al. (2012c).
Table 4.3 Ranges of Temperature for analysis ofSeaonal Movements

Sensor Level Roof Level C
Summer Season 35.0°C to 19.1°C 24.0°C to 17.6°C
Spring/Fall Season 19.0°C to 3.3°C 17.5°C to 1.2°C
Winter Season 3.2°C to –12.5°C 1.1°C to –15.1°C
Table 4.4 Seasonal Apparent Coefficients of Thermal Expansion

Sensor Level Roof Level C
Summer 0.0466 0.0709
Spring / Fall 0.0794 0.0751
Winter 0.0733 0.0657
Average 0.0681 0.0693
4
2 Summer Data
Displacement, mm
0 Slope = 0.20673
-2 Winter Data
-4
Slope = 0.32516
-6
-8
-10 Neg. Displ. = Joint Contraction
-12
4
2
Displacement, mm
0 Fall and Spring Data

Slope = 0.35239
-2
-4
-6
-20 -10 0 10 20 30 40
o
Temperature, C
10
Level C Sensor, South Side Summer Data
Neg. Displ. = Joint Contraction Correlation Coeff. R = 0.881
Displacement, mm
5 Pos. Displ. = Joint Expansion Slope = 0.31453
0
Winter Data
-5 Slope = 0.29180
-10
2
Displacement, mm
0 Fall and Spring Data

Slope = 0.33333
-2
-4 Level C Sensor, South Side

-6
-20 -10 0 10 20 30 40
o
Temperature, C
Fig. 4.22 Seasonal Joint Movements vs. Temperature

4.3.2.2 Sensors Parallel to Expansion Joint (North-South Direction)
The movements of building PG-1 relative to PG-2 in the direction restrained by

the soil mass can be obtained from the measurements of the VW sensors installed
parallel to the expansion joints at the underside of the roof and level C slabs. Fig.
4.23 illustrates the temperatures and corresponding movements for the sensors
installed at the roof level.
Fig. 4.23 Measurements of VW Sensors Parallel to Joint at Roof Level
4.3.2.2.1 Roof Level Sensors

The data of the northern roof sensor show an initial relative movement of nearly 2
mm after installation, corresponding to a rise in temperature of approximately
20ºC. This movement indicates shortening of the sensor while the two structures,
PG-1 and PG-2, underwent expansion due to rise in temperature.
Fig. 4.24 demonstrates the physical behavior of the two buildings at the
expansion joint. The soil-restrained structure undergoes expansion movements
smaller than its unrestrained counterpart, resulting in positive measurements
confirming shortening of the VW sensor. The magnitude of the movement
recorded by the sensor is equal to the amount of relative movement undergone by
the two structures.
Fig. 4.24 Illustration of Typical VW Sensor Movements
We hypothesize that this recorded movement is the sum of two lateral

movement components in their vector forms. The first is the amount of movement
restrained by the soil mass, and the second is the lateral movement undergone by
the structure in order to achieve force and compatibility equilibrium due to rise in
earth pressure. Thus, at every expansion increment, an increase in lateral earth
pressure induces a lateral movement of the building away from the soil mass. The
behavior of the structure is therefore nonlinear, and is a function of the expansion
properties and lateral stiffness of the structure, in addition to the stiffness and
lateral pressure-displacement characteristics of the retained soil.
During the next temperature drop cycle of nearly 40ºC, the northern sensor
recorded elongation movements leading to a relative movement at the end of the
cycle of nearly 0.2 mm. Both structures underwent contraction and reverted to
nearly the same position. Subsequent temperature cycles indicate that PG-1 moved
cumulatively relative to PG-1 in the direction away from the soil mass, with a total
relative movement recorded at the end of the monitoring period of 5 mm.
This movement trend was similar to that recorded by the electrolytic tiltmeter at
level D.
The data of the southern sensor at the roof level, shown in Fig. 4.23, also show
relative movements between PG-1 and PG-2. During the period of temperature
rise, after installation, the southern sensor recorded a shortening of about 1.8 mm,
indicating that the expansion movements of building PG-1 were larger than PG-2
at the southern side. The additional expansion recorded for PG-1 is the sum of the
magnitude of the restrained expansion at the northern edge and the lateral
movement resulting from any increase in earth pressure.
During the temperature drop cycle of nearly 46ºC (6ºC larger than the
temperature drop recorded by the northern sensor), the relative movements
recorded by the southern sensor indicate that the two structures, PG-1 and PG-2,
assumed the same position corresponding to a zero reading, followed by an
elongation of the sensor of approximately 2.8 mm. This elongation signifies that
PG-1 underwent larger movements in the direction of contraction when compared
to PG-2. This may be due to the fact that the contraction of the structure induced a
relaxation of earth pressure leading to a total or partial elimination of the lateral
movements undergone due to increase in earth pressure.
The data discussed thus far indicate that the presence of the soil mass retained
by the rigidly framed building induce a complex soil-structure interaction
dependent on many factors including soil stiffness, lateral structural stiffness,
thermal movement properties of the structure, displacement and earth pressure
relationship, among others.
In an effort to quantify the movements of PG-1, we make use of the thermal
study presented earlier where several values the Apparent Coefficient of Thermal
Expansion (ACTE) of the structure in service were derived. The values obtained
for the seasonal ACTE would be most appropriate to determine the magnitude of
movements undergone by PG-1. To do so, we calculate the magnitude of thermal
movements undergone by PG-2 based on the seasonal values of the ACTE, and
add the recorded movement of the sensor (in vector form) from the calculated
PG-2 movements.
The calculated displacements of building PG-1, δPG-1, derived from the data
collected by the roof level sensors are shown in Fig. 4.25, along with the change
in temperature, ΔT. During the first temperature rise cycle, the northern end of
PG-1 underwent an expansion of about 1 mm, while its southern end expanded by
nearly 4.5 mm. Through the subsequent temperature drop cycle, PG-1 displaced
approximately 5 mm away from the soil mass at the northern end, and 9 mm in the
opposite direction at the southern end. A similar trend in movements could also be
seen for the remaining temperature cycles, where the thermally induced
movements of PG-1 are consistently smaller than those of PG-2. Additionally, the
displacements of both the northern and southern ends of building PG-1 indicate a
linear movement trend away from the soil mass.
2 30
(mm)
Sensor ID: VW-DNP (Roof Level, North Side, Parallel to Joint)

0 20
PG-1
PG-1 Displacement, δ
-2 10
ΔT, ( C)
o
-4 0
-6 -10
-8 -20
ΔT PG-1 Displacement
-10 -30
/ / / / O / / / / / / / / O / /
12 30
(mm)
Sensor ID: VW-DSP (Roof Level, South Side, Parallel to Joint)

9 20
PG-1
PG-1 Displacement, δ
6 10
ΔT, ( C)
o
3 0
0 -10
-3 -20
ΔΤ PG-1 Displacement
-6 -30
Apr/26/99 Jan/25/00 Oct/25/00 Jul/26/01 Apr/26/02 Jan/25/03 Oct/27/03
Record Date
Fig. 4.25 Calculated Displacements of PG-1 at Roof Level
4.3.2.2.2 Level B Sensors

The data recorded by the sensors installed at level B (the underside of level C
slab) is shown in Fig. 4.26. The trends of relative movements of buildings PG-1
and PG-2 at level B were qualitatively similar to their counterparts at the roof
level, except for a slightly steeper trend indicated by the data of the southern
sensor at level B. The range of relative movements recorded by the northern
sensor at level B, however, was nearly one third that recorded by the northern roof
sensor. In contrast, the range of movements of both southern sensors at levels B
and roof were nearly equal. Both level B sensors measured a relative movement of
nearly 2 mm at the end of the monitoring period.
The displacements of building PG-1 were calculated at this level similarly to

their counterpart at the roof level. The seasonal values of the Apparent Coefficient
of Thermal Expansion of the structure at this level were applied to determine the
thermal movements of PG-2. Fig. 4.27 shows the calculated displacement of
building PG-1 at level B sensors.
2.1 40
Sensor ID: VW-BNP (Level B, Noth Side, Parallel to Joint)
1.75 30
Displacement (mm)
Temperature ( C)
o
1.4 20
1.05 10
0.7 0
0.35 -10
o
T ( C) Δ (mm)
0 -20
5 40
Sensor ID: VW-BSP (Level B, South Side, Parallel to Joint)
4 32.5
Displacement, Δ (mm)
Temperature, T ( C)
3 25
o
2 17.5
1 10
0 2.5
-1 -5
-2 Δ (mm) o
T ( C)
-12.5
-3 -20
Apr/26/99 Jan/25/00 Oct/25/00 Jul/26/01 Apr/26/02 Jan/25/03 Oct/27/03
Record Date
Fig. 4.26 Measurements of VW Sensors Parallel to Joint at Level B
The movements of PG-1 at the northern end are in-line with temperature
variations, unlike the relative movements shown in Fig. 4.24 at the same end. The
building expansion during the first cycle after installation of the sensors was
nearly 3 mm into the soil mass at the northern end, compared with approximately
5 mm away from the soil mass at the southern end. During the first contraction
cycle, PG-1 underwent a total movement of 7 mm away from the soil at the
northern end, versus about 11 mm at the southern end towards the soil. This trend
of unequal movements continued throughout the entire monitoring period,
indicating a behavior similar to that previously presented at the roof level.
Further analysis of the VW sensor data is presented in Chapter 5 to investigate
the relationship between the building temperature and the earth pressure acting
on it.
4.4 Conclusions 83
Fig. 4.27 Calculated Displacements of PG-1 at Level B
4.4 Conclusions
From the study of the measurements of electrolytic tiltmeters and vibrating-wire

displacement transducers presented in this chapter, it is clear that the monitored
rigidly-framed four story structure restrained by soil on one side undergoes a
complex soil-structure interaction induced by volumetric strains resulting from
large temperature variations.
During the period of rise in temperature, the structure undergoes limited
expansion movements into the soil mass at the restrained end, causing larger
expansion movements, and stresses, at the other end. The movements of the
structure toward the retained soil induce an increase in earth pressure, and
possibly in soil stiffness, causing the rigid-frames to deflect in the direction away
from the soil mass to maintain the required force equilibrium, while still
undergoing thermal expansion movements. This behavior results in a nonlinear
interaction between the structure and the soil affected by several factors including,
but possibly not limited to, the soil stiffness characteristics, the lateral earth
pressure-displacement relationship, the lateral stiffness and volumetric-strain
characteristics of the structure.
During the period of drop in temperature, the structure undergoes

asymmetrical contraction movements at its ends, and a movement of the soil
into the gap formed between the soil mass and the contracted form of the
retaining structure. This soil movement prevents the structure from reverting to
its position before contraction at the next expansion cycle, causing a cumulative
lateral movement of the structure away from the soil over several temperature
cycles. This continuous movement induced large strains that resulted in stresses
that caused severe structural distress in the building elements, and the failure of
one column at level D.
Consequently, a better understanding of the soil-structure interaction in
rigidly-framed earth retaining structures subject to large temperature variations
is necessary if structurally adequate designs should be accomplished.
Chapter 5
Relationship between Temperature and Earth
Pressure for RFERS
Abstract. The relationship between temperature and earth pressure acting on a

Rigidly Framed Earth Retaining Structure (RFERS) subject to wide temperature
variation was explored in this chapter. The open concrete garage RFERS present-
ed in Chapter 4 was instrumented and monitored for a period of four and a half
years. The structure retains 11 m (36 ft) of soil on one side only. The measured
displacements were used to calculate the earth pressure coefficient using closed
form equations that were developed in Chapter 3. The data indicated that the coef-
ficient of earth pressure behind the monitored RFERS had a strong linear correla-
tion with temperature. The study also reveals that thermal cycles, rather than
lateral earth pressure, caused some of the structural elements to fail.
5.1 Introduction
Structures placed on hillsides often present a number of challenges and a limited

number of economical choices for site design. An option often employed in de-
sign is to use the building frame as a retaining element, comprising a RFERS.
Earth pressure behind RFERS typically begins as a hydrostatic distribution fol-
lowing the conventional Rankine distribution. However, its shape and magnitude
change with every thermal cycle (Aboumoussa 2009).
Broms and Ingelson (1971) Sandford and Elgaaly (1993), Barker and Carder
(2001) measured the lateral earth pressure acting on the abutments of several in-
tegral (jointless) bridges. They concluded that the earth pressure changed from
its initial hydrostatic distribution to a parabolic distribution with thermal cycles.
This is not surprising because Integral bridges resemble simple one story RFERS.
These studies point to the complex nature of the soil structure interaction for
RFERS.
An instrumentation program was undertaken in Chapter 4 to monitor the
movements of a four-story reinforced concrete RFERS, exhibiting signs of large
deformation and severe structural distress including the failure of a column
(Iskander et al 2001). In this chapter movement of the structure presented in

86 5 Relationship between Temperature and Earth Pressure for RFERS
Chapter 4 is correlated to earth pressure, according to the following methodology.

First, measured movements were corrected for thermal expansion and contraction
of the structure, to obtain lateral displacement due to earth pressure only. Next,
the resulting lateral displacement due to earth pressure only was correlated to
earth pressure using the calibrated closed form expressions derived in Chapter 3.
The results confirm the strong dependence of earth pressure on temperature for
rigidly framed earth retaining structures.
5.2 Building Description
The structure, shown in Fig. 4.1–4.2 and labeled PG-1, is a four-story parking gar-
age including a full basement, with a rectangular footprint measuring 172 ft by
234 ft. The building’s structural system consists of reinforced concrete waffle
slabs supported by rectangular reinforced-concrete columns. The north side of the
building is a reinforced concrete retaining wall cast against earth at full height
with a thickness of 16 in at the base tapering down to 12 in on top. On the south-
ern side, a one-story high, 12 in thick, cast-in-place concrete wall provides enclo-
sure for the basement. Resistance to lateral loads is provided by (1) the retaining
walls, (2) irregularly placed concrete shear-walls, and (3) the flexural resistance of
the concrete columns. The building is open on three sides and is subject to large
temperature variations.
At the beginning of the project, an optical survey revealed that PG-1 underwent
a maximum lateral drift of 3 in. in the north-south direction, and a drift of 1 inch
in the east-west direction, at the roof level, during its 25 years of service assuming
that PG-1 was built square and plumb.
Adjacent to building PG-1 is another four-story parking-structure (PG-2) of
similar construction, but different footprint. The two buildings are separated by an
expansion joint along their lengths. PG-2 was not subject to earth pressure at the
northern side due to the presence of a multi-story building with an excavated
basement at its northern edge. No signs of structural distress or lateral movements
were observed in PG-2.
5.3 Geotechnical Properties of the Retained Soil
The soil retained by PG-1 had been filled as part of the site development, but the
surface profile prior to the building construction was unknown. A subsurface
soil investigation was conducted in an effort to obtain the properties for the back-
fill soil for use with this analysis. Boreholes taken behind the northern wall of
PG-1, by an independent testing agency, indicated that the top 20–28 ft of soil re-
tained behind the northern wall of the building consist of miscellaneous uncon-
trolled fill composed of shot rock intermixed with brown medium to fine sand
5.3 Geotechnical Properties of the Retained Soil 87
with varying amount of silt and gravel. Large boulders were also common. The
fill was underlain by a layer of medium to fine sand nearly 10 ft thick overlying
sound rock. The unit weight of the retained soil was estimated as 100 pcf. The
ground water table was not encountered in the borings and was believed to be
below the foundations.
To obtain further information about the backfill soil, a number of in-situ dila-
tometer and pressuremeter tests were attempted at several locations behind the
structure, by a specialized in-situ testing company. The common presence of
large and hard boulders encountered during the advancement of the test probes
resulted in severely damaging several dilatometers and pressuremeter probes
(Fig. 5.1). Numerous trials were repeatedly performed but were unsuccessful
in yielding much useful information. The in-situ testing program was hence
abandoned.
The only successful pressuremeter test is shown in Fig. 5.2. It was conducted
65.5 ft behind the center of PG-1 at a depth of 31.7 ft. The test was conducted on
an exceptionally cold day on February 2001. This location is outside the zone of
influence of the structure (Fig. 5.3). The pressuremeter test was used to obtain the
in-situ coefficient of lateral earth pressure, and a value of 1.4 was computed ac-
cording to Briaud (1992). The coefficient of lateral earth pressure was considered
very high, which could have resulted from a high compaction effort or interaction
with PG-1.
Fig. 5.1 Damaged Dilatometer Cell and Pressuremeter Probe

3500
70000
σv = 152 KPa (3170 psf)
σh = 215 KPa (4490 psf)
3000
K = / = 1.4 60000
0 h v
2500
50000
Corrected Pressure (kPa)
Corrected Pressure (psf)

2000
40000
1500
30000
1000 20000
500 10000
0 0
0 5 10 15 20 25 30
dR/R0 (%)
Fig. 5.2 Result of the Pressuremeter Test

Fig. 5.3 Geotechnical Profile Showing Location of Pressuremeter Test

5.4 Instrumentation Program 89
5.4 Instrumentation Program
Movement of the expansion joint was monitored at four locations using clusters of
vibrating wire (VW) displacement sensors. Each VW cluster contained two trans-
ducers measuring movement parallel and perpendicular to the joint (Fig. 4.5).
Sensors were anchored on the underside of the roof and second level slabs (Levels
B & D), with one cluster on each of the northern and southern sides of the joint at
each level (Fig. 5.4). Movements parallel to the expansion joint (N-S) provide in-
formation regarding the influence of earth pressure on the structure. Movement
perpendicular to the expansion joint (E-W) provides data on the Apparent Coeffi-
cient of Thermal Expansion (ACTE) of the structure. The VW transducers have a
measuring range of 4 in, with a resolution of 0.001 in, a precision of ± 0.0.02 in,
and a calibration accuracy of ± 0.004 in. All VW transducers were equipped with
temperature sensors, which were recorded along with displacements. The built-in
temperature sensors have a range of –49 °F to 212 °F. The expansion joint move-
ments along with temperature were collected hourly. Details of the instrumenta-
tion program are available in Chapter 4.
The two structures, PG-1 and PG-2 have identical geometrical and structural
framing schemes in the vicinity of the expansion joint where the VW sensors were
mounted, and would as a result undergo similar temperature movements. There-
fore, assuming that both structures are free to expand and contract, there would be
no relative movement between the two structures at the expansion joint in the lon-
gitudinal (north-south) direction, and the VW sensors would record no move-
ments. However, given that PG-1 is not free to move as PG-2 is, the amount of
expansion and contraction movement will differ between the two structures de-
pending on the magnitude of restraint imposed by the retained soil on building
PG-1.
59.12 m (194 ft)

N 7.3 m (24 ft) Sensors S
ROOF LEVEL (LEVEL D)
LEVEL C
LEVEL B
LEVEL A
BASEMENT
Fig. 5.4 Typical Framing along Column Line

5.5 Apparent Thermal Coefficient of Expansion of PG-1
To help quantify the thermal movements of building PG-1, a thermal study was
performed using the data recorded by the sensors installed normal to the expan-
sion joint, in order to arrive at an apparent value of the coefficient of thermal
expansion (ACTE) of the structure in service, rather than applying the reported
average value of the coefficient of thermal expansion of concrete.
The expansion length of the concrete is defined as half the sum of the lengths
perpendicular to the expansion joint of the slabs on both sides of the joint. The as-
signed expansion lengths of concrete were 115.3 ft for the northern sensors, and
155.3 ft for the southern sensors. The expansion coefficient can thus be computed
from the measured displacements of VW sensors perpendicular to the expansion
joint, as follows:
1  δ max − δ min  m
α =  = (5.1)
L  t max − t min  L

where, α is the ACTE; L is the expansion length of concrete; δmax and δmin are the
maximum and minimum displacement for a given calculation period; and tmax and
tmin are the corresponding maximum and minimum temperature for the same cal-
culation period; and m is the slope of the displacement versus temperature regres-
sion line.
Numerical values for annual, seasonal, and daily Apparent Coefficient of Ther-
mal Expansion (ACTE) of the building were calculated from the movements’ per-
pendicular to the expansion joint (E-W) using 4.5 years of data (Iskander et al
2011A). In general, values of the ACTE were dependent primarily on degree of
restraint and to a lesser extent on temperature. The measured ACTE ranged be-
tween 0.000586% and 0.0793% per 100°C (0.00032 – 0.044% /100°F), depending
on temperature and imposed structural restraint. For the southern partially re-
strained side, the seasonal values of ACTE shown in Table 5.1, were used in this
study to compute the thermal movement of PG-1 parallel to the expansion joint.
Table 5.1 Seasonal Apparent Coefficient of Thermal Expansion (ACTE) for Building
Temperature Range Roof Level C

°F %/100°F %/100°F
Summer 95.0°F to 66.4°F 0.0328 0.0378
Spring / Fall 66.4°F to 37.9°F 0.0371 0.0440
Winter 37.9°F to 9.5°F 0.0323 0.0450

Average 0.0341 0.0423
5.6 Lateral Displacement of Building Parallel to Earth Pressure 91
5.6 Lateral Displacement of Building Parallel to Earth Pressure
5.6.1 Measured Lateral Displacement

Measured displacements parallel to the expansion joint and earth pressure are
shown along with temperature in Fig. 5.5–5.8. The four joint meters measured the
relative movements between the floor slabs of buildings PG-1 and PG-2 at levels
B and D. In the north-south direction, the data is interpreted based on the premise
that building PG-2 is not restrained by soil at its northern edge and is therefore
free to undergo thermally induced volumetric strains, while the retained soil for
building PG-1 will provide some restraint on thermal movements.
Complex temperature dependent soil structure interaction is evident. For ex-
ample, during the first 3 months, the data of the northern roof sensor show an ini-
tial relative movement of nearly 2 mm after installation, corresponding to a rise in
temperature of approximately 68ºF (Fig. 5.5). This movement indicates shortening
of the sensor while the two structures, PG-1 and PG-2, underwent expansion due
to rise in temperature. PG-1 undergoes expansion movements smaller than its
unrestrained counterpart, resulting in shortening of the VW sensor (positive read-
ings). During the next cycle the temperature drops by nearly 104ºF and the north-
ern sensor recorded elongation movements leading to a relative movement at the
end of the cycle of nearly 0.2 mm. Both structures underwent contraction and
reverted to nearly the same position. Subsequent temperature cycles indicate that
PG-1 moved cumulatively relative to PG-2 in the direction away from the soil
mass, with a total relative movement recorded at the end of the monitoring period
of approximately 0.2 in (5 mm). Similarly, the data of the southern sensor at the
roof level, shown in Fig. 5.6, also show relative movements between PG-1 and
PG-2. During the period of temperature rise, after installation, the southern sensor
recorded a shortening of about 0.07 in, indicating that the expansion movements
of building PG-1 were larger than PG-2 at the southern side. The additional ex-
pansion recorded for PG-1 is the sum of the magnitude of the restrained expansion
at the northern edge and the lateral movement resulting from any increase in earth
pressure. During the temperature drop cycle of nearly 115ºF (10ºF larger than the
temperature drop recorded by the northern sensor), the relative movements rec-
orded by the southern sensor indicate that the two structures, PG-1 and PG-2,
assumed the same position corresponding to a zero reading, followed by an elon-
gation of the sensor of approximately 0.11 in. This elongation signifies that PG-1
underwent larger movements in the direction of contraction when compared to
PG-2. This may be due to the fact that the contraction of the structure induced a
relaxation of earth pressure leading to a total or partial elimination of the lateral
movements undergone due to increase in earth pressure.
Hot Season Cold Season -10

0
Displacement Temperature
Measured δ (mm)
Temperature (C)
1
M
2 10
4 20
5 30
6 40
Baseline Correction δ (mm)
-6 -10
B
Temperature (C)
-4
-2 10
0 20
2 30
4 40
-8 -10
Corrected δ (mm)
temperature (C)
-6 0
C
-4 10
-2 20
0 30
2 40
-10
-1
Rod Model δ (mm)
0
Temperature (C)
-0.9
R
10
-0.5
20
0
30
0.5
40
-8 -10
(mm)
-6 0
Temperature (C)
EP
Earth Pressure δ
-4 10
-2 20
0 30
2 40
99-Apr 00-Apr 01-Apr 02-Apr 03-Apr 04-Apr
Fig. 5.5 Roof Level Sensor, North Side, Parallel to Joint


0
Measured δ (mm)
Temperature (C)
1
M
2 10
4 20
5 30
6 40
-10
-4
B
Temperature (C)
-2
10
0
20
2
30
4
40
-8 -10
Corrected δ (mm)
Temperature (C)
-5
C
-3 10
0 20
3 30
40
-12 -10
Rod Model δ (mm)
0
Temperature (C)
-8
R
-4 10
0 20
4 30
8 40
(mm)
-10
-5 Displacement Temperature
0
Temperature (C)
EP
-4
Earth Pressure δ
10
-2
20
0
30
2
40
Fig. 5.6 Roof Level Sensor, South Side, Parallel to Joint

Hot Season Cold Season

-2 -10
Measured δ (mm)
Temperature (C)
-1
M
0 10
1 20
2 30
3 40
-10
-3
B
Temperature (C)
-2
10
0
20
2
30
3
40
-5 -10
Corrected δ (mm)
Temperature (C)
-3
10
C
-2
0 20
2 30
40
-10
Rod Model δ (mm)
-1.2
0
Temperature (C)
-0.6
R
10
0
20
0.6
30
1.2
40
-10
(mm)
-3
0
Temperature (C)
-2
EP
Earth Pressure δ
-1 10
0 20
1 30
40
Fig. 5.7 Level B Sensor, North Side, Parallel to Joint


-2
Measured δ (mm)
-1 0
Temperature (C)
0
M
10
1
20
2
3 30
4 40
-10
-4
B
Temperature (C)
-2 10
0 20
2 30
4 40
-6 -10
Corrected δ (mm)
Temperature (C)
-3
10
C
3 20
6 30
40
-12 -10
Rod Model δ (mm)
-8 0
Temperature (C)
-4
R
10
0
20
4
8 30
12 40
-10
(mm)
-5
0
Temperature (C)
EP
-3
Earth Pressure δ
10
0
20
3 30
5 40
Fig. 5.8 Level B Sensor, South Side, Parallel to Joint

A similar trend can be observed in all sensors, where the relative movement
undergone by the two structures is heavily dependent on temperature and earth
pressure. Therefore measured displacements must be corrected for the thermal
expansion of PG-1 and PG-2, in order to obtain the net movement due to earth
pressure alone.
5.6.2 Baseline Correction Due to Thermal Movement of PG-2

Movement of PG-2 represents an error in the baseline reference used to obtain
movements of PG-1. We calculate the magnitude of the baseline movement (Cor-
rection) δb undergone by PG-2 based on the seasonal values of the ACTE, αs
shown in Table 5.1, as follows:
Baseline Correction = δ b = α s L ΔT (5.2)
where, L is the expansion Length shown in Fig. 5.9 as L1 or L2, for North and South
sensors respectively. The same value of L1 or L2 was used for roof level or Level-B
sensors. The baseline corrections are added to the measured movement of the
sensors (in vector form) to compute the corrected δC shown in Fig. 5.5 – 5.8.
Location of ROD MODEL

Joint Meters 
SOIL RESTRAINT
L3 = 7.3 m N L3
ΔL1
L1 = 28.5 m
ΔL3
L4 = 59.1 m
L4
L2 = 23.3 m
ΔL2 PG-2 PG-1 ΔL4
Expansion Joint
Fig. 5.9 Expansion Lengths and Rod Model of Expansion
5.6.3 Correction of Lateral Displacement Due to Thermal

Movement of PG-1
A fixed end rod model is used to calculate the thermal movement of PG-1, assum-
ing the soil provides infinite restraint. This is obviously a questionable assump-
tion for the roof level sensors, where the granular fill lacks sufficient strength to
resist building movements. Nevertheless, we calculate the magnitude of the rod

model displacement δR undergone by PG-1 based on the seasonal values of the
ACTE, αs shown in Table 5.1, as follows:
Rod Model Displacement = δR = α s LΔT (5.3)
where, L is the expansion Length shown in Fig. 5.9 as L3 or L4, for north and south
sensors, respectively. The same value of L3 or L4 was used for roof level or level
B sensors. The computed rod model displacements were added to the corrected δC
of the sensors, in vector form, to compute the Earth Pressure Displacement δEP
shown in Fig. 5.5 – 5.8
The computed δEP for all sensors showed an increasing trend, although, as ex-
pected, the temperature trend remained constant. The sensors showed PG-1 mov-
ing away from the retained soil at an annual rate of 0.01 – 0.04 in/year. The rate of
movement was not constant. The building moved away from the retained soil dur-
ing the cold season, and attempted to return to its original position during hot sea-
son, but was unable to return to its original position. This was not surprising since
the energy needed to overcome the soil’s passive resistance was vastly larger than
the energy stored in the building when it contracts under active earth pressure
conditions.
5.6.4 Accuracy of Computed PG-1 Movements

A measure of the accuracy of the computed δEP can be obtained by comparing the
displacement computed at different locations to each other (Fig. 5.10). Assuming
temperature is uniform within the structure, sensors on the same level should ex-
hibit a linear correlation. The sensors on Level-B show an excellent fit between
north and south level sensors. The small deviation from the 1:1 fit is due to the
presence of some shear walls, which cause the building to twist. The roof level
sensors exhibit a poor fit for several reasons. First, the assumption of soil restraint
(rod model) is obviously poor. Second, exposure to the sun at the roof level may
cause a difference between the actual temperature of the structure and the meas-
ured temperature of the sensor. Third, temperature is not uniform within the
structure, as mentioned previously. Finally, the rod model correction probably
contributes to an error because it applies a negative correction, while the actual
correction should be a positive value. The magnitude of temperature (shown in 3
different colors/ranges) does not appear to have an important effect on any of the
comparisons.
Sensors on the same side should have a non-linear correlation. Comparison be-
tween south level sensors is acceptable because the rod model correction on the
roof level is applied in the same direction as the actual movement of PG-1. Com-
parison of North side sensors is less satisfactory because the rod model correction
captures the actual movement of PG-1 at Level-B, but is applied in a direction

opposite to the probable movement of PG-1 at the roof level.
The preceding discussion indicates that the rod model assumption is valid for
level B, but as expected is questionable near the roof. Therefore the measured da-
ta from level B can be given more weight in analyzing the building performance
than data from the roof level.
An attempt was made to compare the statistical properties of the δEP compari-
sons of various sensors by applying a correction to δEP on Level-B, and applying
no correction to the roof level readings (Table 5.2). Although roof level sensors
exhibit a somewhat better R2 when no correction is applied (18.1 vs. 5.2), all other
comparison metrics show less favorable R2. In addition to temperature and earth
pressure, the behavior of the roof is also influenced by sunlight, which was
not recorded. Therefore, we will continue to apply the rod-model correction for
further analysis of the data.
Fig. 5.10 Comparison of Earth Pressure Displacement for all Four Sensors
5.7 Relationship between Lateral Deflection and Earth Pressure 99
Table 5.2 Statistical Correlation of Calculated displacements at Various Locations
COMPARISON METRIC Rod Model Correction No Correction

2 2
R F P-Value R F P-Value
Roof Level Sensors 5.2 1462 0 18.1 5870 0
Level-B Sensors 88.1 19666 0 88.1 19666 0
North Side Sensors 24.2 8468 0 7.4 2142 0
South Side Sensors 76.0 84175 0 34.7 14141 0
5.7 Relationship between Lateral Deflection and Earth Pressure
A simple and reliable method for predicting the relationship between lateral dis-
placement and earth pressure for RFERS was developed in Chapter 3. Closed form
equations were derived such that if one value of displacement or earth pressure is
known (or assumed) the other can be computed for hydrostatic, seismic, uniform,
and semi-elliptical earth pressure distributions (Fig 3.5). In RFERS, the lateral de-
flection of one floor relative to the floor below (story drift) due to lateral earth pres-
sure results from a combination of shear and flexure (bending) deformation of the
beams and columns. Ignoring flexural deformations, a RFERS can be represented
by an equivalent cantilever beam, having an equivalent area, Ao, to derive an analyt-
ical expression for the deflection, γs. An expression is formulated to determine the
equivalent area, Ao, by equating internal and external work done by the RFERS and
equivalent cantilever beams supporting the same load as follows:
30
A0 = (5.4)
 
 1 1 
l c u + 
 (b + 1) I c  (b ) I b  
  lc   lb  
where Ic , Ib are moment of inertia of the individual columns and beams (Fig. 5.11
(a)); , lc and lb are the height of the columns and the length of the beams, respec-
tively; b is the total number of bays, and the coefficient, u, is introduced as a
correction factor for the type of end connection taken as 3.0. A0 = 5.86, for use in
future analysis, is computed for the entire PG-1 building using dimensions and
properties shown in Table 5.4, Fig. 5.4, and Fig. 4.2-4.3, as follows:
30
A0 = = 5.855 (5.5)
 
3 1
9 


(
23 .83 )(
9
+ 18 .6
9
)(
+ 10 . 1
9
) (
+
102
18.5.
)(
+ 357
28
)

δ s0,1 δ s0, n δ s0, n

i=0
z
i=1
w(z)
i=2
Ao
i=n
(b) (c)
I b1
lc1
n
H =  l ci
i =1
lc 2
I cn l cn
w( z )
lb1 lb 2 l bm
m
L =  lbj
j =1
(a)
Fig. 5.11 (a) Legend for Stories, Bays, and Beam and Column Attributes of RFERS,
(b) Deflected Frame, and (c) Equivalent Cantilever Beam for Frame Analysis
The calibration factors shown in Fig 3.5 were obtained from FEM analysis of
42,000 different RFERS configurations, using multivariate non-linear regression
between the derived expressions and FEM (Iskander et al 2011b, c). The calibrat-
ed expression for earth pressure (hydrostatic) loading is:
W ⋅ sl c   i  
3
δ si = α ⋅ 1 +  β   (5.6)
3GAo   s  
 
where, W is the total force acting on the RFERS, i is the story where the lateral
drift is computed, s is the number of stories, lc is the length of the column, G is the
5.7 Relationship between Lateral Deflection and Earth Pressure 101
Table 5.3 Member Properties of PG-1

Member Cross Sectional I Total Available Inertia Average Inertia
Designation on Dimensions Number of about Axis of Ro- for Computing
Figure. 5.6 (bxh) Columns tation for Entire Ao
Building
inches ft4 ft4 ft4
±
W1 2040x12 14.16 1 14.16
23.83
W2 ± 2040x16 33.5 1 33.5
C1 12x28 1.05 17 17.85
18.61
C2 14x28 1.23 17 20.91
C3 28x12 0.19 46 8.74
10.12
C4 28x14 0.307 46 14.12
Floor Slabs 0.7273 ft2/ft 0.3 ft4/ft 1 51 51
±
Length of Building = 51.85 m (170 ft).
shear modulus taken as 250,000 ksf for concrete, Ao is the Equivalent area calcu-
lated using Eq. 5.5, α and β are the factors derived from multivariate calibration
analysis taken as 1.17 and -7/8, respectively (Fig. 3.5).
The change in the value of the earth pressure coefficient, ΔK, can be obtain by
(1) assuming a hydrostatic earth pressure distribution (W = 0.5 γ H2 L ΔK), where
H is the total height of PG-1 (H = slc) , L = is the width of the building (170 ft),
and (2) rearranging Eq. 5.6 as follows:
3GAo
ΔK = δ si (5.7)
  7i  3 
( ( ))
1.17 0.5γH L sl c 1 −   
2
  32  
For PG-1 the change in earth pressure coefficient, ΔK, due to change in its tem-
perature can be obtained from the displacement due to earth pressure, δEP by as-
suming δEP = δsi as follows in Eq. 5.8 and 5.9:
3 × 250,000 × 5.858
ΔK ( Roof Level ) = δ 0 = 9.47 × δ EP
0
 3  EP
( )  7×0 
1.17 0.5 × 0.1× 36 2 × 170 (4 × 9 )1 −   
  32  
(5.8)
3 × 250,000 × 5.858
ΔK ( Level − B ) = δ EP
2
= 10.33 × δ EP
2
  7× 2  
( )
3
1.17 0.5 × 0.1× 36 2 × 170 (4 × 9)1 −   
 
32  

(5.9)
5.8 Earth Pressure Causing Lateral Deformation
The Earth Pressure Displacements are used as the independent parameter in Eq.
5.8 and 5.9 in order to compute the change in earth pressure due to the expansion
and contraction of PG-1. δEP computed using the rod model (Fig. 5.9) was used
for all sensors. The change in the coefficient of lateral earth pressure, K0 , is plot-
ted for data obtained from all 4 sensors in Fig. 5.12, assuming a hydrostatic earth
pressure distribution shape. Data from Level-B sensors is given more weight in
the foregoing discussion for reasons discussed previously.
PG-1 experiences a change in the acting earth pressure coefficient of approxi-
mately 0.25 throughout the annual temperature cycle. Assuming that K0 = 1.4
measured using the pressuremeter is valid within the zone of influence of PG1,
then the earth pressure behind PG-1 fluctuates in the range K0 = 1.25 –1.5. ΔK
shows a small decreasing trend with time, as the building gradually drifts away
from the retained soil, but it is difficult to ascertain that trend using all 4 meas-
urement locations. The reduction of ΔK with each cycle is consistent with data
reported by England et al (2000) for upper bound wall reaction ratio (passive earth
pressure) behind integral bridges. The magnitude of K0 is very high, but it is the
successive thermal cycles that caused some of the structural members of PG-1 to
fail, in order to release the built in stresses due to excessive deformations that
were not designed for. The building was repaired prior to its instrumentation, but
it is likely to fail again if the structural system is not modified to accommodate the
thermal cycles, or if the earth pressure is not reduced by using a light-weight fill.
The measurements clearly indicate that PG-1 undergoes a complex soil-structure
interaction induced by volumetric strains resulting from large temperature varia-
tions. During the warm season, the structure undergoes limited expansion move-
ments into the soil mass at the restrained end, causing larger expansion movements,
and stresses, at the other end. The movements of the structure toward the retained
soil induce an increase in earth pressure, and possibly in soil stiffness, causing the
rigid-frames to deflect in the direction away from the soil mass to maintain the re-
quired force equilibrium, while still undergoing thermal expansion movements.
Through the cold season, the structure undergoes asymmetrical contraction move-
ments at its ends, and a movement of the soil into the gap formed between the soil
mass and the contracted building. This soil movement prevents the structure from
reverting to its position before contraction at the next expansion cycle, causing a
cumulative lateral movement of the structure away from the soil over several
temperature cycles.
The change in earth pressure coefficient, ΔK, is plotted against temperature for
all four measuring locations (Fig. 5.13). All sensors exhibit a good correlation be-
tween temperature and earth pressure except the north roof level sensor, for rea-
sons discussed earlier. At level B, PG-1 experiences a ΔK of approximately
0.005/°C. At the roof level the change in earth pressure is 0.002–0.004/°C,
reflecting the effect of diminished soil restraint. The data in Fig. 5.13 provides
convincing evidence of the proportional relationship between structural tempera-
ture and earth pressure. Indeed, the coefficient of earth pressure behind PG-1 is
linearly proportional to the temperature of the retaining RFERS.
5.8 Earth Pressure Causing Lateral Deformation 103

0.1
0.05
0
-0.05
ΔK
-0.1
-0.15
Roof Level, North Level B, North Side
-0.2
-0.25
Record Time
-0.2
-0.1
0
ΔK
0.1
Roof Level, South Side Level B, South Side
0.2
0.3
Record Time

0.1
0.05
0
-0.05
ΔK
-0.1
-0.15
Roof Level, North Side Roof Level, South Side
-0.2
-0.25
Record Time
0.15
0.1
0.05
0
ΔK
-0.05
-0.1
-0.15 Leve B, North Side Level B, South Side
-0.2
Record Time
Fig. 5.12 Change in Pressure Coefficient for Sensor at Four Locations

0.05 0.1
Roof Level Sensor Roof Level Sensor
0 North Side 0.05 South Side
-0.05 0
-0.1 -0.05
ΔK
ΔK
-0.15 -0.1
-0.2 -0.15 y = -0.064049 + 0.0041269x

y = -0.12886 + 0.0021248x
R2= 0.19321 R2= 0.84971
-0.25 -0.2
-10 0 10 20 30 40 -20 -10 0 10 20 30 40
Temperature (C) Temperature (C)
0.1 0.15
Level B Sensor Level B Sensor
North Side 0.1 South Side
0.05
0.05
0
0
ΔK
ΔK
-0.05 -0.05
-0.1
-0.1
y = -0.099945 + 0.0047072x -0.15 y = -0.094126 + 0.0049824x
R2= 0.83625 R2= 0.76715
-0.15 -0.2
-10 0 10 20 30 40 -20 -10 0 10 20 30 40
Fig. 5.13 Change in Lateral Earth Pressure Coefficient vs. Temperature
5.9 Limitations of This Study
A large number of simplifying assumptions were employed in order to achieve a

numerical relationship between the change in building temperature and earth pres-
sure. The main assumptions include (1) treating the soil as a linear elastic body,
(2) ignoring its hysteretic behavior and densification of the fill with repeated
thermal cycles. (3) Additionally, as discussed earlier, assuming that the back fill
provides adequate restraint near the surface according to the rod model is probably
invalid. (4) Finally, the shape of the earth pressure distribution was assumed to be
hydrostatic which is questionable. To address these limitations we investigated
other expansion models as shown in Table 5.2. The first and second assumptions
provide an important avenue for future research. The third and fourth assumptions
have a cumulative effect on the order of 25% of the numerical answer, but they do
not affect the shape of the trend lines of the data. The numerical value of change
5.10 Conclusions 105
in the earth pressure coefficient with temperature change is valid for the instru-
mented building (PG-1) only. Therefore, the insights derived from this case histo-
ry are applicable to all rigidly framed earth retaining structures, including jointless
(segmental) bridges.
5.10 Conclusions
This case history demonstrates that rigidly framed earth retaining structures un-
dergo complex temperature-dependent soil-structure interaction. Displacements of
a four-story rigidly-framed structure that retained 36 ft of fill on one side only
were monitored along with temperature for a period of 4.5 years. During periods
of temperature decrease, the structure contracts, and the soil follows it. During
periods of rise in temperature, the structure undergoes limited expansion move-
ments into the soil mass at the restrained end, causing larger expansion move-
ments, and stresses, at the other end. This is not surprising since the energy needed
to overcome the soil’s passive resistance is vastly larger than the energy stored in
the building when it contracts under active earth pressure conditions. Expansion
of the structure toward the retained soil induce an increase in earth pressure,
and possibly in soil stiffness, causing it to deflect away from the soil mass to
maintain the required force equilibrium, while still undergoing thermal expansion
movements.
The observed building displacements were correlated to the coefficient of lat-
eral earth pressure, K. K was found to be linearly dependent on the building tem-
perature; it changed by approximately 0.005/°C varying in the range of 1.25 to
1.5, depending on the season. A residual translation away from the restrained soil
is observed at the end of each thermal cycle. It is the thermal cycles rather than
the high earth pressure that caused some of the structural elements of the building
to distress and fail, in order to release some the built in pressure.
The measurements presented in this book demonstrate that after the first cold
cycle, RFERS contract and the retained soil typically follows the structure. As
temperature increases, the structure attempts to expand, but is restrained by the
soil, causing it to drift away from the retained soil. Repeated cycles of contraction
and expansion may cause distress of the structure.
Chapter 6
Numerical Analysis of Instrumented RFERS
Abstract. This chapter presents the finite element analysis of the building
monitored in Chapters 4 and 5. A plane strain model was employed through
dividing the column properties by the tributary width of the frame, and utilizing
the equivalent area and moment of inertia per foot of length of the waffle slab.
The analysis confirms many of the results gleaned from the discrete
instrumentation measurements. In particular, the lateral earth pressure exerted on
the rigid frame developed during the thermal expansion cycles is considerably
larger than the lateral earth pressure at rest.
6.1 Introduction
A 2-dimensional plain-strain finite element analysis of the longitudinal rigid frame

located along column line D (see Fig. 4.2) was performed. The frame is located
approximately at the center of the structure, and has a tributary width for soil load
of nearly 9.5 m. The finite element model included the northern wall of the
structure, and the nine columns and waffle-slab elements composing the rest of the
rigid frame. The structural elements were modeled to accommodate the plane
strain limitations of the model, through dividing the column properties by the
tributary width of the frame, and utilizing the equivalent area and moment of
inertia per foot of length of the waffle slab. This modeling of the waffle slab
assumes that the effective width of the slab engaged in resisting the lateral loads is
equal to the sum of half the distances between the frame along line D and the two
adjacent frames on opposite sides. This allows acceptable modeling of the lateral
stiffness of the moment frame, coupled with accurate modeling of the area of the
slab subject to volumetric temperature strains. Furthermore, to account for the
long-term load duration effects on the concrete elements and the presence of
cracked regions along the length of said members, the section properties of the
wall, columns and slab were reduced by 75 percent. This reduction is smaller than
that recommended by the Building Code Requirements for Structural Concrete
(ACI 318), but the actual field survey of the structural elements does not warrant a
further reduction.

108 6 Numerical Analysis of Instrumented RFERS
The backfill soil was modeled using an elastic-plastic Mohr Coulomb model
and a Hardening Soil model that accounts for stress-dependency of stiffness
modulus of the backfill soil.
6.2 The Finite Element Model
The geometry and section properties of the actual building frame, along with the
equivalent properties used in the finite element model are presented herein. The
backfill soil properties for the Mohr-Coulomb and Hardening-Soil models are also
discussed.
6.2.1 The Structural Frame

The actual geometry of the rigid frame along column line D, and the section
properties of its elements were obtained from the original structural drawings of
the building and verified through field measurements. The frame is a four-story
reinforced concrete structure retaining earth at full height on one end, and is
unrestrained against lateral movement on the other end except at the basement
level. The finished grades at both ends of the frame serve as car parking lots. Fig.
6.1 illustrates the geometrical features of the 9-bay frame, as well as the general
structural framing. Wall members are marked with the letter W followed by a
suffix indicating varying properties, while column elements are similarly marked
with the letter C. The gross properties of the marked frame elements are shown in
table 6.1, along with the adjusted properties integrated into the numerical analysis
model.
The frame supports are modeled with fixities in the vertical and horizontal
directions, given that the building foundation bears on rock. No rotational fixity
was introduced at the supports.
Fig. 6.1 FEM Frame Model along Column Line D

6.2 The Finite Element Model 109
6.2.2 The Backfill Soil

The backfill soil was modeled using an elastic-plastic Mohr-Coulomb model and a
Hardening Soil model to comparatively investigate the effect of temperature
movements on the structural frame.
A subsurface soil investigation was conducted in the field in an effort to obtain
properties for the backfill soil for use with this analysis. Investigatory borings with
Standard Penetration Test (SPT) were conducted by an independent testing agency
commissioned by the building owner and supervised by the authors prior to the
start of this research study. The boring information indicated that the top 20 to 28
feet of backfill soil consist of shot rock intermixed with brown medium to fine
sand with varying amount of silt and gravel, where large boulders were common.
An approximately 10-ft deep layer of medium to fine dense sand was found to
underlay the top layer, and overlay a sound rock stratum.
To obtain further information about the backfill soil for the purpose of this
study, a number of in-situ dilatometer and pressuremeter tests were attempted at
several locations behind the structure. The common presence of large and hard
boulders encountered during the advancement of the test probes resulted in
severely damaging several dilatometer cells and pressuremeter probes. Numerous
trials were repeatedly performed but were unsuccessful in yielding any reliable
information. The in-situ testing program was hence subsequently abandoned.
The actual highly heterogeneous soil properties are thus difficult to simulate
using published constitutive soil models. Consequently, the backfill soil was
modeled using a drained relatively dense sand layer with negligible cohesion with
the properties shown in table 6.2.
Table 6.1 Rigid Frame Member Properties
Member Actual Cross Sectional Numerical Model Properties

Mark Properties (inches) Variation (E = 485,590 kip/ft2)
EA (kip/ft) EI (kip.ft2/ft)
W1 12 in thick wall 485,590 30,350
W2 16 in thick wall 647,455 71,940
C1 28 x 12 36,353 12,370
C2 28 x 14 42,413 14,438
C3 12 x 28 36,353 2,272
C4 14 x 28 42,413 3,608
A = 0.7273 ft2/ft
Floor Slabs 353,170 110,297
I = 0.3 ft4/ft
Table 6.2 Backfill Soil Model Properties
Dry Internal Friction Dilatancy Elasticity Elasticity Elasticity

Soil Model Density Angle Angle Modulus Modulus Modulus
(pcf) (Degrees) (Degrees) Eref (ksf) Eoed (ksf) E50 (ksf)
Mohr Coulomb 110 38 8 648 720 N/A
Hardening Soil 110 38 8 N/A 720 800
Poisson’s ratio ν = 0.2; power m (HS model) = 0.5; Eur = 2160 ksf; K0 = 0.384
6.2.3 The Analysis Procedure

The numerical analysis of the full-scale structure is composed of three parts. The
first part consists of a plain strain analysis of a selected frame along column line D
under temperature loading without the presence of soil backfill on either side of
the structure. This analysis is used as a benchmark to compare the strains
developed in the frame due to thermal stresses versus the strains developed for the
same structure retaining the backfill soil.
The second part of the analysis includes the frame with the backfill soil
idealized using the elastic-plastic Mohr-Coulomb constitutive model. The retained
soil is added at the initial stage of the analysis using a staged construction
simulation, followed by the application of 50ºF thermal expansion load, and
several 90ºF contraction and expansion load-cycles subsequently.
The third analysis procedure repeats the second part described above, with the
exception that the backfill soil is modeled using the Hardening Soil model.
6.3 Numerical Analysis
The three analysis procedures are further introduced and presented herein.
6.3.1 Thermal Analysis of Rigid Frame (Part 1)

6.3.1.1 Description of Analysis Procedure
The rigid frame with the properties shown in Fig. 6.1 is analyzed independently of
any soil loading or restraint for thermal strains. The first temperature cycle
consists of simulating an increase in temperature of 50ºF assuming the
construction of the reinforced concrete structure was concluded towards the end of
the spring season. The second and subsequent temperature loading cycles simulate
an alternating decrease and increase in temperature of 90ºF. These thermal loads
are in-line with the average temperature changes in the building locale, and are
obtained from averaging the range of temperatures measured on-site for nearly
fifty-four months and presented earlier in Chapter 4.
6.3 Numerical Analysis 111
The numerical analysis was performed using the commercially available finite
element analysis code for soil and rock applications Plaxis published by Plaxis
BV. The structural frame members were modeled using the software beam
elements defined by 5 nodes with three degrees of freedom per node (two
translational and one rotational). The beam elements formulation is presented
further in Chapter 7.
Temperature loading is not provided for in the Plaxis code, or user interface,
and is thus applied using equivalent thermal forces applied at the beam nodes. The
temperature forces were only applied in the horizontal direction to simulate the
horizontal thermal strains. The magnitude of the force is found as follows:
F T = As EsαΔT (6.1)
where, As is the horizontal beam (slab) cross sectional area (ft2)

Es is the horizontal beam (slab) modulus of Elasticity (kip/ft2)
α is the coefficient of thermal expansion (0.055% per 100ºF)
ΔT is the change in temperature in degrees Fahrenheit
6.3.1.2 Numerical Analysis Results (Part 1)
The results of the finite element analysis of the rigid frame subjected to
temperature cycles in the absence of the soil are presented herein. Figs. 6.2 and 6.3
illustrate the horizontal movements of the retaining wall and end column,
respectively, representing the two outermost boundaries of the frame. Note that
the retaining wall designation is used in this analysis for consistency with the
remainder of the numerical analysis, although no soil is being retained in this case.
36 Roof
o
U (+ 50 F)
hec1
Level D
27 U o
(- 90 F)
hcc1
o
U (+ 90 F)
hec2
z, ft
Level C
18 No backfill soil present
Level B
9
Level A
0
-0.1 -0.05 0 0.05 0.1
Horizontal Retaining Wall Movement, U , ft
h
Fig. 6.2 Horizontal Retaining Wall Movements with Temperature (No Backfill)
36 Roof
o
U (+ 50 F)
hec1
Level D
27 U o
(- 90 F)
hcc1
o
U (+90 F)
hec2
z, ft
Level C
Level B
9
Level A
0
-0.1 -0.05 0 0.05 0.1
Horizontal End Column Movement, U , ft
h
Fig. 6.3 Horizontal End Column Movements with Temperature (No Backfill)
36 Roof
o
M (+ 50 F)
ec1
Level D o
27 M (- 90 F)
cc1
o
M (+ 90 F)
ec2
z, ft
Level C
Level B
9
Level A
0
-60 -40 -20 0 20 40 60
Retaining Wall Moment, M, kips-ft
36 Roof
o
M (+ 50 F)
ec1
Level D o
27 M (- 90 F)
cc1
o
M (+ 90 F)
ec2
z, ft
Level C
Level B
9
Level A
0
-20 -15 -10 -5 0 5 10 15 20
End Column Moment, M, kips-ft
Fig. 6.4 Bending Moments in Retaining Wall and End Column (No Backfill)
The horizontal movements of the frame extremities are not equal given the
varying lateral stiffness of the structural members. The lateral stiffness of the
retaining wall is substantially larger than the end column, and the rest of the frame
columns, thus creating more restraint against temperature movement at the wall
side. The horizontal wall displacement from its initial position with a 50ºF
increase in temperature is nearly 0.018 ft (0.21 in), while with 90ºF temperature
change the wall movement is approximately 0.033 ft (0.4 in). The end column, on
the other hand, is displaced nearly 0.04 ft (0.5 in) with a 50ºF rise in temperature,
and approximately 0.072 ft (0.86 in) with a 90ºF temperature change. The total
range of movement during the 90ºF expansion and contraction cycle is 0.066 ft
(0.8 in) for the retaining wall and 0.144 ft (1.7 in) for the end column.
The bending moments in the retaining wall and end column are shown in Fig.
6.4. The magnitude of the flexural stresses developed in the retaining wall is
substantially larger than those found in the end column. This is due to the larger
lateral restraint exhibited by the retaining wall due to its larger lateral stiffness.
The end column moments in all except the first level are nearly equal, with the
maximum moment occurring at the top of the first level with a magnitude of 11
kip-ft. The maximum retaining wall moment occurs at the same location and is
equal nearly 50 kip-ft.
The temperature movements and bending moments in the extreme frame
members will be further discussed in the succeeding sections and compared with
those obtained from the numerical analysis that includes the effect of the backfill
and with the results of the monitoring data obtained on site.
6.3.2 Thermal Analysis of Rigid Frame with Mohr-Coulomb

Backfill (Part 2)
The frame shown in Fig. 6.1 is analyzed with the Mohr-Coulomb backfill soil.
The Mohr-Coulomb constitutive law is a well-known soil model and is used
herein as a first approximation of the soil-structure behavior, particularly as a
qualitative study of the effect of backfill soil on the behavior of the structure under
thermal loading. The model is defined using the parameters listed in table 6.2.
The numerical analysis is performed starting with the unloaded frame as an
initial phase, followed by the addition of the backfill soil as a staged construction
phase, then by temperature loading cycles similar to those presented in Part 1.
The results of the numerical analysis for the rigid frame with backfill soil are
discussed herein. The retaining wall and end column movements and bending
moments are presented for the initial backfill stage as well as for each temperature
cycle. The lateral earth pressure developed behind the structure is also presented
for all analysis phases.
Fig. 6.5 illustrates the horizontal movement of the retaining wall. At the initial
backfill stage, the wall exhibits a maximum horizontal displacement on top of
nearly 0.04 ft (0.5 in), followed by an expansion movement of approximately 0.01
ft during the first expansion cycle simulating a 50ºF rise in temperature. The
maximum wall contraction during the subsequent cycle of 90ºF decrease in
temperature is nearly 0.05 ft (0.6 in), followed by an expansion movement into the
soil mass of about 0.0275 ft at the next 90ºF temperature increase cycle. The wall
subsequent expansion and contraction movements then remain nearly equal for the
following two cycles. This of course is predicted for the Mohr-Coulomb’s model,
which is not formulated to simulate change in loading direction or the loading and
reloading cycles. Moreover, the range of movement of the wall during the 90ºF
temperature cycles was nearly 0.0275 ft (0.33 in), or less than half the range of
movements found from the analysis in part 1 of the frame without the backfill soil.
36 Roof
U
ha
Level D U
27 hec1
U
hcc1
U
z, ft
Level C hec2
18
U
hcc2
U
Level B hec3
9
Level A
0
-0.1 -0.05 0 0.05 0.1
h
Fig. 6.5 Horizontal Retaining Wall Movements with Temperature (Mohr-Coulomb

Backfill)
Moreover, the range of wall movements found from the numerical analysis is
also quite comparable to its counterpart determined from field measurements and
shown in Fig. 4.23 of Chapter 4 to be nearly 0.0215 ft.
The end column movements are presented in Fig. 6.6 below. While the initial
movement at the backfill stage is nearly equal to that of the retaining wall, the
movements due to temperature are substantially larger at the end column. This
result is in line with the information obtained from the field measurements of the
building movements presented in Chapter 4. The expansion and contraction
movements of the end column are found larger in the numerical analysis for two
primary reasons. The first reason is that the lateral stiffness of the retaining wall
was found to place some restraint on the thermal movement at the wall end of the
structure, thus increasing the movement at the end column, as found from the
analysis in part 1. The second reason is the presence of the backfill soil at the wall
side, imparting a substantially larger restraint to the frame movement at the wall
side. The magnitude of the movement away from the wall at the end column
are, however, substantially larger in the presence of the soil restraint. In particular,
the maximum movement of the end column at the top of the frame with backfill
soil is nearly 0.157 ft (1.88 in) compared with 0.072 ft (0.86 in) for the unloaded
frame.
36 Roof
U
ha
Level D U
27 hec1
U
hcc1
z, ft
Level C U
18 hec2
U
hcc2
Level B U
hec3
9
Level A
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
h
Fig. 6.6 Horizontal End column Movements with Temperature (Mohr-Coulomb Backfill)
The range of movement at the end column found from the numerical analysis is
considerably larger than its counterpart found from field measurements, unlike the
movements at the retaining wall end. Fig. 4.23 of Chapter 4 indicates that the
range of movement at the southern end of the structure is nearly 0.033ft.
Nevertheless, the magnitude of relative expansion movements between the
retaining wall and end column, found from the analysis, is reasonably comparable
to its counterpart determined from field measurement. Figs. 6.5 and 6.6 indicate
that the end column movement is nearly twice that of the retaining wall. This
result can also be found in Fig. 4.23 for the sensors installed at the northern and
southern ends of the structure, where the expansion movement at the southern end
is nearly twice that at the northern end (the retaining wall end). The analysis
results for contraction movements, on the other hand, do not seem to reasonably
agree with the field measurements, thus producing a large different in the range of
the thermal movements.
The bending moments in the retaining wall and end column elements are
presented in Fig. 6.7. The maximum retaining wall moment for the frame with
backfill soil is found to be over 70 percent larger than its counterpart found in part
1 for frames without backfill, located nearly at the same position for both frames.
The magnitude of the bending moments in the upper two levels of the retaining
wall, however, are found to be similar for both frames.
Additionally, for the frame with backfill soil, the magnitude of the bending
moment at level B in the end column is 3.6 times larger than its counterpart in the
frame without backfill, and twice as large for the remainder of the wall height.
Furthermore, the retaining wall and end column bending moment generated
during the temperature cycles are approximately 4 times larger than the bending
moments resulting from the addition of the backfill soil at the initial stage of the
analysis. This therefore indicates that the effect of large thermal movements
coupled with the presence of the backfill soil behind the structure result in
substantially larger flexural stresses in the extreme vertical elements of the rigid
frame.
36 Roof
M
a
Level D M
27 ec1
M
cc1
z, ft
Level C M
18 ec2
M
cc2
Level B
M
ec3
9
Level A
0
-60 -40 -20 0 20 40 60 80 100
36 Roof
M
a
Level D M
27 ec1
M
cc1
z, ft
Level C M
18 ec2
M
cc2
Level B
M
ec3
9
Level A
0
-20 -10 0 10 20 30 40 50 60
Fig. 6.7 Bending Moments in Retaining Wall and End Column (Mohr-Coulomb Backfill)
The lateral earth pressure developed behind the structure is also found to vary
substantially with temperature movements as shown in Fig. 6.8. The lateral earth
pressure at the backfill stage, as well as at the thermal contraction cycles, is found
to be comparable to the earth pressure determined using Coulomb’s classical
active earth pressure theory. During the expansion cycles, however, the lateral
earth is found to be substantially larger than the earth pressure at backfill, and is
larger that the lateral earth pressure at rest.
Consequently, the building frame is subjected to a large range of lateral earth
pressure loads, further emphasizing the effect of temperature movements on the
behavior of the rigid frame and the soil structure interaction.
The results presented herein help therefore support the observation of
substantial structural distress observed throughout the operational full scale
parking structure described in Chapter 4. The design of rigidly framed earth
retaining structures subjected to large temperature variations must therefore
account for the effects of thermal movements and the soil-structure interaction
during temperature cycles.
36 Roof
σ'
a
Level D σ'
27 ec1
σ'
cc1
σ'
z, ft
Level C ec2
18
σ'
cc2
σ'
Level B ec3
9 σ'
Active
σ'
At rest
Level A
0
0 0.5 1 1.5 2 2.5
σ' , kips/ft
2
h
Fig. 6.8 Lateral Earth Pressure Behind Rigid Frame (Mohr-Coulomb Backfill)
6.3.3 Thermal Analysis of Rigid Frame with Hardening-Soil

Backfill (Part 3)
The numerical analysis presented below is similar to its counterpart discussed

above in part 2, with the exception that the backfill soil is simulated using the
Hardening-Soil Model.
The Hardening-Soil (HS) model is an enhanced version of the popular
Hyperbolic Soil model presented by Duncan and Chang, 1970. Both models use
the hyperbolic stress-strain relationship found from drained triaxial tests between
the axial strain and the deviatoric stress. The HS model, however, improves on the
Hyperbolic-Soil model in several areas, namely the use of the theory of plasticity
rather than the theory of elasticity, the inclusion of soil dilatancy and the
introduction of a yield cap.
In contrast to the elastic perfectly-plastic Mohr-Coulomb model used earlier,
the yield surface of the HS model is not fixed in the principal stress space, but can
expand due to plastic straining. Additionally, both shear and compression

hardening are included in the formulation of the HS model, and the stress
dependency of the soil stiffness is also accounted for. For more information on the
Hardening Soil model, the reader is referred to Schanz (1998), and to the Material
Models Manual accompanying the analysis software Plaxis.
The horizontal retaining wall and end-column movements obtained from the
analysis are presented firstly in Fig. 6.9. The analysis included 4 expansion cycles
and 3 contraction cycles, after the initial backfill stage, or an additional two
cycles compared to the analysis presented in part 2 earlier.
36 Roof
U
ha
U Level D
hec1
27
U
hcc1
Uhec2
z, ft
Level C
18
Uhcc2
U
hec3
Level B
9 U
hcc3
U
hec4
Level A
0
-0.05 0 0.05 0.1 0.15
Horizontal Retaining Wall Movement, U h, ft
36 Roof
U
ha
Level D U
hec1
27
Uhcc1
Uhec2
z, ft
Level C
18
U
hcc2
Uhec3
Level B
9 U
hcc3
U
hec4
Level A
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
Horizontal End Column Movement, U h, ft
Fig. 6.9 Horizontal Retaining Wall and End Column Movements (Hardening-Soil Backfill)
Fig. 6.9 indicates that the maximum end-column horizontal movement of 0.2 ft
at the end of the fourth expansion cycle is in-line with the displacement obtained
from field surveys of the actual structure. The retaining wall and end column
movements with temperature change are also qualitatively similar to the
movements measured in the field, and discussed in chapter 4, where both the wall
and column are found to progressively displace away from the retained soil with
every full cycle of temperature change. The retaining wall movements indicate a
larger increase in displacement at the end of each temperature cycle compared to
the end column. This is in-line with the field measurements presented in Fig. 4.23.
The total change in displacement for the retaining wall at the top of the structure is
approximately 0.0756 ft (2.3 mm) between the second and fourth temperature
cycle, almost equal to the change in displacement obtained from field
measurements. Moreover, the range of movements of the retaining wall with
temperature cycles is smaller than its counterpart for the end column, which is also
found the monitoring of the actual structure. Overall, the general displacement
behavior of the rigid frame extremities, as well as the magnitude of said
displacement, is in agreement with the measured movements obtained in the field
and presented in Chapter 4.
The bending moments in the retaining wall and end column found from the
analysis are shown is Fig. 6.10 below.
36 Roof
Ma Mcc2
Level D M M
27 ec1 ec3
Mcc1 Mcc3
z, ft
Level C Mec2 Mec4

18
Level B
9
Level A
0
-50 0 50 100 150
36 Roof
Ma Mcc2
Level D M M
27 ec1 ec3
M M
cc1 cc3
z, ft
Level C M M
18 ec2 ec4
Level B
9
Level A
0
-20 -10 0 10 20 30 40 50 60
Fig. 6.10 Retaining Wall and End Column Bending Moments (Hardening-Soil Backfill)
The top graphic in Fig. 6.10 indicates that the retaining wall moment increases
during the contraction cycles (rise in temperature) and decreases after expansion.
The variation between the bending moments developed during the different
temperature cycles is most pronounced at the first unrestrained level (level B) and
is nearly equal to 250 percent. Furthermore, the maximum bending movement
developed during the third contraction cycle is 70 percent larger than its
counterpart found at the end of the initial backfill stage.
On the other hand, the bottom graphic in Fig. 6.10 indicates that the end
column moment increases during the expansion cycles (rise in temperature) and
decreases after contraction. The variation between the bending moments
developed during the different temperature cycles is also most pronounced at the
first unrestrained level (level B) and is nearly equal to 500 percent, nearly twice
the corresponding variation found for the retaining wall. Additionally, the
maximum bending moment developed in the end column at the last contraction
cycle is approximately three times larger than the corresponding moment found at
the end of the backfill stage.
Fig. 6.11 illustrates the variation of lateral earth pressure exerted on the
retaining wall with the change in temperature.
36 Roof
σ' σ'
a ec3
σ' σ' Level D
ec1 cc3
27
σ'cc1 σ'ec4
σ'ec2 σ'Active
z, ft
Level C
18
σ' σ'
cc2 At Rest
Level B
9
Level A
0
0 0.5 1 1.5 2 2.5 3 3.5 4
2
σ' , kips/ft
h
Fig. 6.11 Lateral Earth Pressure behind Rigid Frame (Hardening Soil Backfill)
At the initial backfill stage, the lateral earth pressure developed in the
hardening soil model is approximately equal to the lateral earth pressure at rest,
substantially larger than the active earth pressure determined from Coulomb’s
classical earth pressure theory. With increase in temperature causing the
expansion of the rigid frame, the lateral earth pressure increases substantially to
nearly similar values irrespective of the number of cycles, and decreases
considerably after contraction movements. These results are in general qualitative
concordance with those found from the analysis presented in Part 2 earlier. A
comparison of the analysis results presented in the three formed parts is drawn
below.
6.3.4 Comparison of Numerical Analysis Results

The maximum displacement and bending moments developed in the retaining wall
and end-column found from the analysis discussed earlier are compared herein.
36 Roof
U
h-No Backfill
Uh-MC Backfill Level D
27
Uh-HSS Backfill
z, ft
Level C
18
Level B
9
Level A
0
-0.05 0 0.05 0.1 0.15
Maximum Horizontal Retaining Wall Movement, U h, ft
Fig. 6.12 Maximum Horizontal Retaining Wall Movements
Fig. 6.12 presents the maximum horizontal displacement obtained for the
retaining wall from the numerical analysis of the rigid frame without backfill, with
backfill soil simulated using the Mohr-Coulomb soil model, and with soil backfill
simulated using the Hardening Soil model. Evidently, the maximum wall
movements vary with each condition, with the smallest movement occurring
during temperature decrease (contraction) cycles for the frame without any
retained soil. A 240 percent larger wall movement is found for the frame retaining
Mohr-Coulomb backfill, and a 340 percent larger movement is found for the
frame retaining the Hardening Soil backfill.
36 Roof
Uh-No Backfill
Level D Uh-MC Backfill
27
U
h-HSS Backfill
z, ft
Level C
18
Level B
9
Level A
0
-0.05 0 0.05 0.1 0.15 0.2 0.25
Maximum Horizontal End Column Movement, U h, ft
Fig. 6.13 Maximum Horizontal End Column Movements

Similar results could be found for the movements of the end column shown in
Fig. 6.13. It is therefore obvious that the effect of thermal movements on the
displacements of the rigid frame is far more pronounced for a structure restrained
by backfill soil then for a free structure.
Fig. 6.14 Maximum Retaining Wall and End Column Moments
The maximum retaining wall and end column moments from the previous three
analyses are presented in Fig. 6.14. The comparison reveals similar results for the
bending moments as for the displacements, where the moments developed in the
restrained frame structure are substantially larger than those found for the free
frame. However, while the bending moments developed in the retaining wall
differed quite notably with the two soil models for backfill, the corresponding
moments for the end column show close agreement for the both backfill models.
Finally, the maximum lateral earth pressures found from the analysis using the
Mohr-Coulomb and Hardening Soil models are presented in Fig. 6.15.
6.4 Conclusions 123
36 Roof
σ'MC-Backfill
27 σ' Level D
HSS-Backfill
σ'
Active
σ'
z, ft
Level C
18 At Rest
Level B
9
Level A
0
0 0.5 1 1.5 2 2.5 3
2
Max σ' , kips/ft
h
Fig. 6.15 Maximum Lateral Earth Pressure
The maximum lateral earth pressure exerted on the rigid frame developed
during the thermal expansion cycles and is considerably larger than the classical
active earth pressure and the lateral earth pressure at rest. The hardening soil
model developed larger magnitudes for the lateral earth pressure than did the
Mohr-Coulomb’s model.
6.4 Conclusions
Finite element modeling of the instrumented structure presented in Chapter 4 and

5, confirms many of the results gleaned from the discrete instrumentation
measurements. In particular, the lateral earth pressure exerted on the rigid frame
developed during the thermal expansion cycles is considerably larger than the
lateral earth pressure at rest. The hardening soil model simulated larger
magnitudes for the lateral earth pressure than did the Mohr-Coulomb’s model,
perhaps due to its ability to incorporate the stress dependency of the soil stiffness.
It could also be that the lateral pressure increased due to yield of the soil in
Hardening model in each cycle, whereas the Mohr Coulomb model kept elastic
conditions. Additionally, the Mohr Coulomb has elastic behavior before the failure
criterion is achieved, which may produce less overall residual strain compared
with the Hardening model due to less permanent strains.
Chapter 7
Parametric Study of Earth Pressure behind
RFERS at Backfill Stage
Abstract. This chapter presents the results of parametric finite element analyses
performed to explore the relationship between earth pressure and the stiffness of
Rigidly Framed Earth Retaining Structures (RFERS). A plane strain model was
employed. The stages of construction were incorporated in the analysis to simulate
an initial stage where a structural frame is first completed followed by the addition
of backfill soil in several stages. The displacement of the structures, as well as the
earth pressure and resultant load developed behind them were obtained to examine
the relationship between the stiffness of the retaining structure and the
development of lateral earth pressure in the retained soil mass.
7.1 Introduction
All retaining structures undergo some form of displacement after the application
of lateral earth pressure before reaching a state of equilibrium. The lateral part of
the displacement transforms the state of stress in the ground depending on the
amount and type of displacement (FHWA, 1976 pp 32). An infinitely rigid
retaining wall, for instance, may undergo lateral displacements in the form of pure
translation, rotation about the bottom, rotation about the top, or a combination
thereof. A flexible wall, on the other hand, may undergo the same type of
displacements as a rigid wall, in addition to displaying flexure under lateral loads.
Moreover, flexible retaining walls which are braced or tied may also undergo
similar displacement types but with different flexural deformation characteristics.
The state of stress in the retained soil mass is generally dependent on the type
of displacement and the shape of the displaced retaining structure (Winterkorn and
Fang, 1975 pp 216-218). Fig. 7.1 (Finn, 1963) presents elastic solutions for lateral
pressures on walls produced by two types of lateral displacement namely
translation and rotation about the bottom. The resultant lateral load is shown to
vary with the type of wall displacement for the same material.
Despite the existence of a voluminous amount of literature and research work
on the subject of lateral earth pressure for various types of retaining structures and
soil material, nevertheless, the authors are not aware of any research or published

126 7 Parametric Study of Earth Pressure behind RFERS at Backfill Stage
information on lateral earth pressure behind rigidly framed earth retaining

structures (RFERS). For this reason, parametric numerical studies of RFERS
retaining cohesionless backfill were performed to qualitatively and quantitatively
study the mobilization of earth pressure behind RFERS.
Fig. 7.1 Resultant Lateral Earth Pressure vs. Wall Movements
7.2 Parametric Numerical Analysis
Two-dimensional plain strain finite element analysis of rigidly framed earth

retaining structures of varying length, height, number of bays, number of levels,
and member stiffness were performed using the commercial finite element
analysis software Plaxis. The structural members composing the rigid frames were
modeled using an elastic material model, while the backfill soil was modeled
using a Mohr Coulomb model with varying parameters. The individual structural
frames were assumed to have a beam stiffness and a column stiffness that are
respectively equal throughout the structure. The structural wall retaining the
backfill was assumed to be of concrete construction. The structures were assumed
to have infinitely rigid foundation strata.
7.2 Parametric Numerical Analysis 127
To accommodate the plain strain limitations of the finite element model, the
cross sectional properties of the structural elements were determined by dividing
the column and beam properties by the tributary width of the frame selected as 10
feet, and by utilizing the equivalent area and moment of inertia per foot of length
of the floor slab. Given that gravity loads on the structural elements are not of
concern, this method of resolving the cross sectional properties of the rigid frame
elements allows an acceptable modeling of the lateral stiffness of the rigid frame
under plane strain conditions.
Table 7.1 Finite Element Analysis Parameters
Parameter Variation
Number of Bays for Structural Frames, n 1, 3, 6, 10, 15 and 20
Bay Lengths for Structural Frames, Lb (ft) 10 and 20
Number of Stories for Structural Frames, s 1 thru 5
Story Height, h (ft) 10
Structural Beam Properties
Cross Sectional Area, Ab (in2) 144

Moment of Inertia, Ib (in4) 2073 and 1036
Structural Columns Properties
Cross Sectional Area, Ac (in2) 144 and 432

Moment of Inertia, Ic (in4) 2073 and 41460
Structural Wall Properties
Cross Sectional Area, Aw (in2/ft) 144

Moment of Inertia, Iw (in4/ft) 2073
Modulus of Elasticity of Structural Members, Es (ksi) 3150
Poisson Ratio of Structural Members, νs 0.25
Density of Structural Members, γs (pcf) 150
Backfill Soil Properties
Internal Friction Angle, φ (Degrees) 30 and 40

Dilatancy Angle, ψ (Degrees) 0 and 10
Cohesion, c (psi) 0
Modulus of Elasticity, E (ksi) 41500 and 144000
Density, γ (pcf) 100 and 120
Strength Reduction for Interfaces 0.75
φ
ψ
γ
γ ν
Fig. 7.2 Finite Element Analysis Parameters
The stages of construction were incorporated in the analysis to simulate an

initial stage where a structural frame is first completed followed by the addition of
backfill soil in several stages. This method of construction of multistory RFERS is
generally more common than casting a retaining wall and the remainder of the
structural frame against the soil mass. Nevertheless, parametric models with the
latter means of construction were analyzed as well.
The displacement of the structures, as well as the earth pressure and resultant
load developed behind them were obtained to examine the relationship between
the stiffness of the retaining structure and the development of lateral earth pressure
in the retained soil mass.
The parameters varied for the finite element analysis are listed in table 7.1 and
shown on Fig. 7.2. Overall, a total number of 240 rigid frames with varied
stiffness characteristics were analyzed with two backfill soil properties.
7.2.1 Finite Element Analysis Model Details

A typical finite element model of a rigidly framed earth retaining structure
consists of two parts. The first is the rigid frame composed of beams, columns and
the wall retaining the soil backfill. These structural elements were modeled using
three-noded beam elements with three degrees of freedom per node (two
translational degrees of freedom and one rotational degree of freedom). The beam
elements are based on Mindlin’s beam theory, which allows for element
deflections due to bending and shear deformations. The elements stiffness are
determined using their flexural rigidity (EI) and axial stiffness (AE), from which
an equivalent thickness, deq, can be obtained from equation 7.1 as follows:
EI
d eq = 12 (7.1)
AE
Bending moments and axial forces in the beam elements are determined at
stress points located in pairs at two locations at a distance above and below the
centerline of the beam element equal to:
d eq 3 (7.2)
The beam elements are connected with rigid joints throughout the frame
structure except at the supports where the bending moments were released at the
joints and horizontal and vertical fixities were applied.
The second part of the model consists of the backfill soil, which extends the full
height of the structure a minimum distance equal to the largest of either 3 times
the total height or the length of the rigid frame. The backfill soil was modeled
using 6-noded triangular elements, which provide a second-order interpolation for
displacement. The stiffness matrix of the soil elements is evaluated by numerical
integration using a total of three gauss points.
Fig. 7.3 Finite Element Analysis Model Details
Interface elements, with zero thickness, defined by three pairs of nodes were
introduced between the rigid frame and backfill soil to enhance flexibility of the
finite element mesh and prevent non-physical stress results for soil-structure
interaction. The stiffness matrix for interface elements is obtained using three
Newton-Cotes integration points. An elastic-plastic model using Coulomb shear
stress criterion and a tension cut-off criterion is used to describe behavior of the
interfaces. The strength properties of the interface elements are linked to strength
properties of backfill soil, with an associated strength reduction factor selected as
0.75 for analysis.
7.2.2 Single Story Rigidly Framed Earth Retaining Structures

7.2.2.1 Backfill Soil with 30º Internal Friction Angle
The analysis results of single story rigid frames are presented herein. The results
are reported in Fig.7.4 illustrating the frame movements at the retaining wall for
various frame configurations. The top two graphics in said figure illustrate the
lateral frame movements, Uh, for structures of varying number of bays, with 10
feet long bays and a column to beam stiffness ratios of 1 and 4. The number of
bays is indicated using the subscripts shown in the legend. Similarly, the bottom
two graphics illustrate the frame movements for structures with 20 feet long bays.
The lateral displacements of the structures indicate that the larger the lateral
stiffness of the rigid frame, the smaller the lateral movement due to the exerted
earth pressure. For the rigid frames with a column to beam stiffness ratio of 1 and
bay length of 10 ft, the lateral displacements at the top of the structure varied
between 0.019 ft for a single bay frame, and 0.0023 ft for a 20-bay frame. On the
hand, for the frames with similar bay length but with a column to beam stiffness
ratio of 4, the lateral displacements at the top of the structure varied between
0.0026 ft for a single bay frame, and 0.0002 ft for a 20-bay frame.
Furthermore, for the rigid frames with a bay length of 20 ft and a column to
beam stiffness ratio of 1, the lateral displacements at the top of the structure varied
between 0.038 ft for a single bay frame, and 0.0065 ft for a 10-bay frame. The
displacement of the frames with similar bay length and a column to beam stiffness
ratio of 4 were approximately one order of magnitude lower than their latter
counterparts.
The total amount of movement obtained in this analysis indicate that only 5 out
of the 18 rigid frames underwent movements near or larger than the displacement
of 0.001 times the height of the structure necessary to develop active earth
pressures (Winterkorn and Fang, 1975 pp 405).
The lateral earth pressures developed at the end of the backfill stage, simulated
in this analysis using staged construction calculations, are presented in Fig. 7.5.
The earth pressures for each group of rigid frames is compared to the Coulomb
active earth pressure for a level backfill with a friction angle between the wall and
backfill soil of 22.5º, and the prescribed lateral earth pressure values found in the
major building codes adopted in the United States presented in table 1.1 of
Chapter 1. The magnitude of the lateral earth pressure at rest is nearly 43% larger
than the lateral soil loads prescribed by ASCE 7-98 and is not included herein.
The examination of the plots in Fig. 7.5 reveals that the magnitude of the lateral
earth pressures developed during the backfill stage varies with the lateral stiffness of
the rigid frames. The top plot indicates that the classical Coulomb active earth
pressure, and the earth pressure loads prescribed by BOCA (1999), NBC (1999) and
SBC (1999) adequately predict the lateral soil loads exerted on the single story
frames at the end of the backfill stage, while ASCE 7-98 tends to slightly
overestimate the lateral earth pressure. Additionally, the top plot indicates that the
range of earth pressures developed behind this group of frames is quite limited for
the best part of the height of the structure, diverging slightly towards the base of the
frames.
10
8
U
hn1
6 U
z, ft
hn3
L = 10 ft U
b hn6
4
φ = 30
o U
hn10
U
2 hn15
S /S = 1
c b U
hn20
0
0 0.005 0.01 0.015 0.02
10
8
U
hn1
6 U
z, ft
hn3
L = 10 ft U
b hn6
4
φ = 30
o U
hn10
U
2 hn15
S /S = 4
c b U
hn20
0
0 0.0005 0.001 0.0015 0.002 0.0025 0.003
10
6
z, ft
L = 20 ft U
4 b hn1
φ = 30
o
U
hn3
2
S /S = 1 U
c b hn10
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
10
6
z, ft
L = 20 ft U
b hn1
4
φ = 30
o
U
hn3
2
S /S = 4 U
c b hn10
0
0 0.001 0.002 0.003 0.004 0.005
h
Fig. 7.4 Frame Movements at Retaining Wall for Single Story Frames (φ = 30º)
10
σ
hn1
8 σ
hn3
σ
hn6
6
σ
z, ft
hn10
L = 10 ft σ
4 b hn15
σ
φ = 30o hn20
Coulomb
2
S /S = 1 ASCE 7-98
c b
BOCA, SBC, IBC
0
10
σ
hn1
8 σ
hn3
σ
hn6
6 σ
z, ft
hn10
L = 10 ft σ
hn15
4 b
σ
φ = 30o hn20
Coulomb
2
ASCE 7-98
S /S = 4
c b BOCA, SBC, IBC
0
10
σ
hn1
8 σ
hn3
σ
hn10
6
z, ft
Coulomb
ASCE 7-98
L = 20 ft
4 b BOCA, SBC, IBC
φ = 30o
2
S /S = 1
c b
0
10
σ
hn1
8 σ
hn3
σ
hn10
6
z, ft
Coulomb
ASCE 7-98
Lb = 20 ft BOCA, SBC, IBC
4
φ = 30o
2
S /S = 4
c b
0
0 0.1 0.2 0.3 0.4 0.5
σ , kips/ft
2
h
Fig. 7.5 Lateral Earth Pressure Developed behind Single Story Frames (φ = 30º)
On the other hand, the second plot of Fig. 7.5 shows a larger magnitude of
lateral earth pressure compared with the top plot, with the Coulomb active earth
pressure underestimating the backfill loads, whereas ASCE 7-98 reasonably
predicting the developed loads. The columns to beams stiffness ratio for the
frames in the second plot is four (4) times larger than its counterpart for the frames
in the top plot, thus producing substantially larger lateral stiffness and
considerably smaller lateral displacements as illustrated in Fig. 7.4. These smaller
magnitudes of lateral displacements reduce the soil ability to mobilize the full
active earth pressure.
The lateral earth pressure developed behind the rigid frames with bay lengths of
20 ft are shown in the bottom two graphics of Fig.7.5. For frames with columns to
beam stiffness ratio of 1, the Coulomb active earth pressure as well as the soil
loads prescribed by BOCA (1999), SBC (1999) and IBC (2000) appear to
reasonably predict the earth pressure resulting behind the rigid frames, while
ASCE 7-98 seems to slightly overestimate the backfill loads. Conversely, for
frames with column to beam stiffness ratio of four (4), Coulomb, BOCA, SBC and
IBC tend to underestimate the backfill loads, except for the single bay frame,
whereas ASCE 7-98 appears to adequately predict the lateral earth pressure.
7.2.2.2 Effect of Lateral Frame Stiffness on the Mobilizations of Active

Earth Pressure
As presented above, the lateral earth pressure developed in the backfill soil is
largely dependent on the lateral stiffness of the rigid frames receiving said
backfill. To further understand the relationship between the frame stiffness and the
development of earth pressure, we introduce the following expression for the
lateral stiffness of rigid frames subjected to triangular loads based on equations
developedin Chapter 3:
 sl    7i  
3
1
= 1.17 c  1 −    (7.3)
KL  3GA0    8s  
where, KL is the lateral frame stiffness
s is the number of stories
i is the number of the story from top
lc is the length of the columns
E is the elasticity modulus of the frame members
ν is the poisson’s ratio of the frame members
Ic is the moment of inertia of the columns
Ib is the moment of inertia of the beams
lb is the length of the beams
nb is the number of bays in the frame
E
G=
2(1 + ν )
30
A0 =
 
 3 1 
lc  + 
 (nb + 1) I c  nb  I b  
  lc   lb  
The lateral stiffness of the single story frames analyzed are calculated based on
Eq. 7.3 and presented in Table 7.2 below.
Table 7.2 Lateral Stiffness of Single Story Frames (per foot of width of structure)
Frame Mark Number of Bays Bay Length (ft) Sc/Sb KL (kip/ft/ft)

F1 1 10 1 405
F2 3 10 1 914
F3 6 10 1 1455
F4 10 10 1 2061
F5 15 10 1 2761
F6 20 10 1 3436
F7 1 10 4 1718
F8 3 10 4 4705
F9 6 10 4 9135
F10 10 10 4 15031
F11 15 10 4 22395
F12 20 10 4 29759
F13 1 20 1 227
F14 3 20 1 574
F15 10 20 1 1472
F16 1 20 4 1031
F17 3 20 4 2925
F18 10 20 4 9439
As shown in Table 7.2, the lateral stiffness (per foot of structure) of frames F9
through F12, and F18, are substantially larger than the rest of the structures. Those
frames have also shown the largest earth pressure developed in the backfill soil as
illustrated in Fig.7.5. In fact, the lateral earth pressure developed behind said
frames is slightly larger than the soil load prescribed by ASCE 7-98, for at least
half the height of the structure.
It can therefore be concluded that for RFERS with lateral frame stiffness per
foot of width of structure less than nearly 5000 kip/ft/ft calculated based on
equation 7.3, the classical Coulomb active earth pressure theory may reasonably
predict the lateral earth pressure developed in the backfill soil at the end of the
backfill stage.
On the other hand, for RFERS with lateral frame stiffness per foot of width
larger than 5000 kip/ft/ft, the lateral earth pressure is better predicted based on the
provisions of ASCE 7-98.
7.2.2.3 Effect of Staged Construction Calculation on the Mobilizations of

Active Earth Pressure
To examine the effect of the staged construction calculation used to analyze the
rigid frames for the development of lateral earth pressure at the backfill stage,
frames F1 through F12 are reanalyzed with the elimination of the staged
construction step. In other words, the RFERS and the backfill soil are assumed to
be present simultaneously, i.e. “wished in place.” In the latter case, the analysis is
initiated with the calculation of the in-situ state of stress of the soil mass, followed
by the analysis of the full soil-structure problem.
10
σ σ
hn1 hn20
8 σ Coulomb
hn3
σ ASCE 7-98
hn6
6 σ BOCA, SBC, IBC
z, ft
hn10
L = 10 ft σ K
0
hn15
4 b
φ = 30 o
Staged Construction Not Included
2
S /S = 1
c b
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
σ , kips/ft
2
h
Fig. 7.6 Lateral Earth Pressure for Single Story Frames - Staged Construction not included
Fig.7.6 presents the lateral earth pressure resulting from the analysis of frames
F1 through F7. The figure indicates that the lateral earth pressures developed in
the retained soil mass when staged construction is not included in the numerical
simulation are substantially larger than their counterparts obtained from
simulating a backfill stage. The earth pressures predicted using Coulomb’s active
earth pressure theory, along with the loads stipulated by the major building codes,
all tend to underestimate the lateral earth pressure developed behind the stiffer
rigid frames, namely frames F3 through F7. The pressure distribution is such that
active pressure conditions prevail for the top half of the structure, followed by a
pressure increase along the bottom half, reaching at rest conditions at the base of
the frame.
The lateral earth pressure determined from the numerical analysis of the stiffer
backfill soil are presented in Fig.7.7 for frames with bay length of 10 feet, and
Fig.7.8 for frames with 20-ft bays.
10
σ σ
hn1 hn20
8 σ Coulomb
hn3
σ ASCE 7-98
6 hn6
z, ft
σ BOCA, SBC, IBC

hn10
L = 10 ft σ K
4 b
hn15 0
φ = 40 o
2
S /S = 1
c b
0
10
σ σ
hn1 hn20
8 σ Coulomb
hn3
σ ASCE 7-98
6 hn6
z, ft
σ BOCA, SBC, IBC

hn10
L = 10 ft
4 b
σ K
0
hn15
φ = 40o
2
S /S = 4
c b
0
0 0.1 0.2 0.3 0.4 0.5 0.6
σ , kips/ft
2
h
Fig. 7.7 Lateral Earth Pressure behind Single Story Frames (Lb= 10 ft, φ = 40º)
Fig.7.7 indicates that Coulomb active earth pressure can reasonably predict the
lateral earth pressure developed behind the frames with a lateral stiffness less than
5000 kip/ft/ft, whereas BOCA (1999), SBC (1999), IBC (2000) and ASCE 7-98
tend to overestimate the lateral soil load.
On the other hand, the lateral earth pressure predicted using Coulomb’s
classical theory is significantly smaller than the pressure developed behind rigid
frames with lateral stiffness larger than 5000 kip/ft/ft, especially for the bottom
half of the structure where the lateral soil load increases to the magnitude of the
lateral earth pressure at rest.
Fig.7.8 below indicate similar results to those presented in its earlier

counterpart, where the lower stiffness of the frames result in nearly a full
mobilization of the coulomb active pressure in the retained backfill, whereas
higher frame stiffness tend to result in substantially larger lateral earth pressure
loads of on the structure.
10
σ
hn1
8 σ
hn3
σ
hn10
6 Coulomb
z, ft
ASCE 7-98
L = 20 ft BOCA, SBC, IBC
4 b
K
φ = 40o 0
2
S /S = 1
c b
10
σ
hn1
8 σ
hn3
σ
hn10
6
z, ft
Coulomb
ASCE 7-98
L = 20 ft
4 b BOCA, SBC, IBC
φ = 40o K
0
2
S /S = 4
c b
0
0 0.1 0.2 0.3 0.4 0.5
σ , kips/ft
2
h
Fig. 7.8 Lateral Earth Pressure behind Single Story Frames (Lb= 20 ft, φ = 40º)
7.2.3 Two-Story Rigidly Framed Earth Retaining Structures

The analysis results of two-story rigid frames are presented herein. Fig.7.9 below
illustrates the lateral earth pressure developed behind rigidly framed structures of
various bay numbers with a bay length of 10 feet. The top graphic shows the
lateral earth pressure behind frames with 1 to 20 bays, and a column to beam
stiffness ratio of 1, while the bottom graphic presents the pressure behind frames
with column to beam stiffness ratio of 4. Additionally, the lateral soil loads
prescribed by the several national building codes and the lateral earth pressure at
rest are shown in each graph for comparison purposes.
20
σ σ
hn1 hn20
σ Coulomb
15 hn3
σ ASCE 7-98
hn6
σ BOCA, SBC, IBC
z, ft
hn10
10
L = 10 ft σ K0
b hn15
φ = 30o
5
S /S = 1
c b
0
20
σ σ
hn1 hn20
σ Coulomb
15 hn3
σ ASCE 7-98
hn6
σ
z, ft
BOCA, SBC, IBC

10 hn10
L = 10 ft
b
σ K
0
hn15
φ = 30 o
5
S /S = 4
c b
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
σ , kips/ft
2
h
Fig. 7.9 Lateral Earth Pressure behind Two Story Frames (Lb= 10 ft, φ = 30º)
Fig.7.9 indicates that the lateral earth pressures developed behind the top group
of frames are a close match to the magnitude and distribution of earth pressures
determined using Coulomb’s active earth pressure theory. The soil loads stipulated
by BOCA, SBC and IBC are also a close match to the earth pressure loads found
from the analysis, whereas the soil loads specified by ASCE 7-98 seem to
overestimate the lateral pressure. Moreover, the magnitude of the earth pressure
developed behind the frames is notably smaller than the lateral earth pressure at
rest.
For the second group of substantially stiffer frames, however, where the
column to beam stiffness is increased four times, the lateral earth pressures
developed behind the frames with 10, 15 and 20 bays seem to be closely
correlated with the lateral loads prescribed by ASCE 7-98, and are about 20% to
25% larger than the earth pressures developed behind the same frames with the
smaller column to beam stiffness ratio. Nevertheless, the largest earth pressure
found from the analysis remains notably smaller in magnitude than the lateral
earth pressure at rest.
Fig.7.10 illustrates the analysis results for earth pressures behind frames with
20-ft bay length. The results shown in said figure are reasonably comparable to
those presented in the previous figure. The lateral earth pressures developed
behind the stiffer frames tend to closely correlate with the lateral soil loads
stipulated by ASCE 7-98, while the magnitudes of lateral earth pressure developed
behind the frames with lower stiffness are reasonably close to that determined
using Coulomb’s active earth pressure theory, and to the soil loads specified by
BOCA, SBC and IBC.
20
σ
hn1
15 σ
hn3
σ
hn10
z, ft
Coulomb
10
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
φ = 30o K
5 0
S /S = 1
c b
0
20
σ
hn1
15 σ
hn3
σ
hn10
z, ft
10 Coulomb
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
φ = 30 o
5 K
0
S /S = 4
c b
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
σ , kips/ft
2
h
Additional numerical analysis of two story frames is performed for the stiffer
backfill soil with an internal angle of friction of 40º which properties are shown in
table 7.1. The results of the analysis for frames with 10-ft bay length and varying
column to beam stiffness are presented in figure 7.11.
The top graphic of said figure indicates that the lateral earth pressures
developed at the end of the backfill stage behind the group of frames with the
lower lateral stiffness are reasonably comparable to the earth pressures determined
from Coulomb’s active earth pressure theory, and from the lateral soil loads
specified by BOCA, SBC and IBC. Moreover, the lateral load calculated based on
the provisions of ASCE 7-98 is fairly larger than the earth pressures found from
the analysis, and the magnitude of the lateral earth pressure at rest is also
significantly larger.
The second graphic in Fig.7.11 indicates that lateral earth pressure behind
stiffer rigid frames, such as those with 10, 15 and 20 bays, and with a column to
beam stiffness ratio of 4, is notably larger than the same group of frames with
lower column to beam stiffness ratio. The lateral earth pressure distribution behind
the stiffer frames is such that Coulomb’s active earth pressure conditions prevail
for top half of the height of the structure, increasing subsequently to reach the
magnitude of ASCE 7-98 soil load and finally the magnitude of the lateral earth
pressure at rest.
20
σ σ
hn1 hn20
σ Coulomb
15 hn3
σ ASCE 7-98
hn6
σ
z, ft
BOCA, SBC, IBC

10 hn10
L = 10 ft σ K
b hn15 0
φ = 40 o
5
S /S = 1
c b
0
20
σ σ
hn1 hn20
15 σ Coulomb
hn3
σ ASCE 7-98
hn6
z, ft
10 σ BOCA, SBC, IBC

hn10
L = 10 ft
b
σ K
0
hn15
φ = 40o
5
S /S = 4
c b
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
σ , kips/ft
2
h
Fig.7.12 illustrates the analysis results for two story frames with bay length of
20 feet retaining the backfill soil. The top graphic in said figure indicates that the
lateral earth pressure developed behind the frames with the lower stiffness is
reasonably comparable to the lateral pressure determined using the classical
Coulomb active pressure theory. The lateral soil loads stipulated in the major
national building codes seem to overestimate the lateral earth pressure with
increasing degrees of magnitude.
The bottom graphic in Fig.7.12 indicates that for the frames with the larger
lateral stiffness, such as the frame with 10 bays, the lateral earth pressure
developed behind the structure is similar to Coulomb’s active earth pressure for
nearly the top half of the retained soil, increasing almost linearly to meet the
magnitude of the lateral earth pressure at rest at the bottom of the structure. The
frames with lower lateral stiffness showed similar results to their counterpart
discussed described earlier, where the lateral earth pressure is found to be inline
with the results obtained using the Coulomb active earth pressure.
20
σ
hn1
σ
15 hn3
σ
hn10
Coulomb
z, ft
10 ASCE 7-98
b
K
φ = 40o 0
5
S /S = 1
c b
0
0 02 04 06 08 1
20
σ
hn1
σ
15 hn3
σ
hn10
Coulomb
z, ft
10 ASCE 7-98
b
K
φ = 40o 0
5
S /S = 4
c b
0
0 0.2 0.4 0.6 0.8 1
σ , kips/ft
2
h
7.2.4 Three-Story Rigidly Framed Earth Retaining Structures

The analysis results of three-story rigid frames are presented herein. Fig.7.13
below illustrates the lateral earth pressure developed behind the rigid frames with
bay length of 10.
The top graphic, which shows the lateral earth pressure behind frames with
column to beam stiffness ratio of 1, indicates that the lateral earth pressure
developed behind this group of frames is reasonably comparable to the Coulomb

active earth pressure, and to the soil load prescribed by SBC, BOCA and IBC. The
load stipulated in ASCE 7-98, on the other hand, slightly overestimates the lateral
earth pressure behind the frames, which is consistently the case from previous
analysis results for this group of smaller stiffness frames.
30
σ σ
hn1 hn20
25
σ Coulomb
hn3
20 σ ASCE 7-98
hn6
σ
z, ft
BOCA, SBC, IBC

15 hn10
L = 10 ft σ K
b hn15 0
10 φ = 30o
5 S /S = 1
c b
0
30
σ σ
hn1 hn20
25
σ Coulomb
hn3
20 σ ASCE 7-98
hn6
z, ft

hn10
L = 10 ft
b σ K
0
hn15
10 φ = 30o
5 S /S = 4
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
Fig. 7.13 Lateral Earth Pressure behind Three Story Frames (Lb= 10 ft, φ = 30º)
The bottom graphic in Fig.7.13 shows that the lateral load developed behind the
stiffer frames, such as those with more than 6 bays and a column to beam stiffness
ratio of 4, is more comparable to the magnitude of the soil load prescribed by
ASCE 7-98. In no case presented in Fig.7.13 does the lateral earth pressure loads
obtained from the numerical analysis correspond to the magnitude of the lateral
For the frames presented in Fig.7.14 below, the magnitude of the lateral earth
pressure obtained from the analysis is found to be mostly comparable to
Coulomb’s active pressure in the case of frames with lower stiffness, and to the
load stipulated in ASCE 7-98 for the stiffer frames.
30
σ
hn1
25
σ
hn3
20 σ
hn10
z, ft
Coulomb
15 ASCE 7-98
L = 20 ft
b BOCA, SBC, IBC
10
φ = 30o K
0
5 S /S = 1
c b
0
30
σ
hn1
25
σ
hn3
20 σ
hn10
z, ft
Coulomb
15 ASCE 7-98
L = 20 ft
b BOCA, SBC, IBC
10
φ = 30o K
0
5 S /S = 4
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
The analysis results of three-story rigid frames with the stiffer backfill soil are
presented herein. Fig.7.15 below illustrates the lateral earth pressure developed
behind the rigid frames with bay length of 10 feet.
The top graphic in said figure indicates that for the frames with 1 and 3 bays,
the lateral earth pressure is substantially lower than Coulomb’s active earth
pressure. This in fact is due to the inadequate stiffness of the frames to retain the
backfill soil, resulting in the collapse of the retained soil body. For the remainder
of the frames analyzed, the results shown in Fig. 7.15 indicate that the lateral earth
pressure developed behind the frames with lower lateral stiffness is comparable to
the lateral earth pressure determined using Coulomb’s active earth pressure
theory, or using the lateral soil load prescribed by BOCA, SBC IBC. For the
stiffer frames, such as those with 15 and 20 bays in the bottom graphic of Fig.
7.15, the lateral earth pressure is found to be reasonably in-line with Coulomb’s
active earth pressure for the top half of the structure, then increasing linearly to
reach the magnitude of the lateral earth pressure at rest towards the bottom of the
retained soil mass.
30
σ σ
hn1 hn20
25
σ Coulomb
hn3
20 σ ASCE 7-98
hn6
σ
z, ft
BOCA, SBC, IBC

15 hn10
σ K L = 10 ft
hn15 0 b
10
φ = 40o
5 S /S = 1
c b
0
30
σ σ
hn1 hn20
25
σ Coulomb
hn3
20 σ ASCE 7-98
hn6
σ
z, ft
BOCA, SBC, IBC

15 hn10
σ K
0
L = 10 ft
b
hn15
10
φ = 40o
5
S /S = 4
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
30
σ
hn3
25
σ
hn10
20 Coulomb
ASCE 7-98
z, ft
15 BOCA, SBC, IBC

L = 20 ft
b K
0
10
φ = 40o
5 S /S = 1
c b
0
30
σ
hn1
25
σ
hn3
20 σ
hn10
z, ft
15 Coulomb
L = 20 ft ASCE 7-98
b
10 BOCA, SBC, IBC
φ = 40o K
0
5 S /S = 4
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
Similar results were found for frames with bay lengths of 20 feet. Fig.7.16
illustrates the lateral earth pressure obtained from the analysis of 5 frames. The
results for the rigid frame with a single bay and a column to beam stifness ratio of
1 are not shown given that the numerical calculation could not be completed due
to the collapse of the soil body.
7.2.5 Four-Story Rigidly Framed Earth Retaining Structures

The analysis results of four-story rigid frames are presented herein. Fig.7.17 below
illustrates the lateral earth pressure developed behind frames with bay length of 10
feet. The analysis results shown are consistent with their counterpart discussed
earlier for three-story frames, where the lateral earth pressure developed behind
the frames with lower lateral stiffness is found to be inline with the Coulomb
active earth pressure, while the earth pressure developed behind the stiffer frames
was more comparable to the slightly larger soil load stipulated by ASCE 7-98. In
all cases, however, the lateral earth pressure found from the analysis was
substantially smaller than the lateral earth pressure at rest.
40
σ σ
35 hn1 hn20
30 σ Coulomb
hn3
σ ASCE 7-98
25 hn6
z, ft

hn10
L = 10 ft
15
b σ K
0
hn15
φ = 30o
10
5 S /S = 1
c b
0
40
σ σ
35 hn1 hn20
30 σ Coulomb
hn3
σ ASCE 7-98
25 hn6
z, ft

hn10
L = 10 ft
15
b σ K
0
hn15
φ = 30o
10
5 S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. 7.17 Lateral Earth Pressure behind Four Story Frames (Lb= 10 ft, φ = 30º)
The results for frames with bay length of 20 feet are shown in Fig.7.18. The
lateral earth pressure developed behind the majority of the frames in the figure
was reasonably comparable to Coulomb’s active earth pressure, and to the soil
load prescribed by SBC, BOCA an IBC.
40
σ
35 hn1
σ
30 hn3
σ
25 hn10
z, ft
Coulomb
20 ASCE 7-98
L = 20 ft
15
b BOCA, SBC, IBC
φ = 30o K
0
10
5 S /S = 1
c b
0
0 0 1 1 2 2 3
40
σ
35 hn1
σ
30 hn3
σ
25 hn10
z, ft
Coulomb
20 ASCE 7-98
L = 20 ft
15
b BOCA, SBC, IBC
φ = 30o K
0
10
5 S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. 7.18 Lateral Earth Pressure behind Four Story Frames (Lb= 20 ft, φ = 30º)
The analysis results for frames retaining the backfill soil with 40º internal friction
angle are shown in Fig.7.19. The relative magnitude and distribution of the lateral
earth pressure developed behind the four-story frames are similar to their
counterparts presented earlier for the same group of three-story frames.
Said results are also generally in-line with the remainder of the analysis
presented.
40
σ Coulomb
35 hn3
σ ASCE 7-98
30 hn6
σ BOCA, SBC, IBC
hn10
25
σ K
z, ft
hn15 0
20
σ L = 10 ft
hn20
b
15
φ = 40o
10
5 S /S = 1
c b
0
40
35 σ σ
hn1 hn20
30 σ Coulomb
hn3
σ ASCE 7-98
25 hn6
σ
z, ft
BOCA, SBC, IBC

20 hn10
σ K
0
L = 10 ft
b
15 hn15
10
φ = 40o
5 S /S = 4
c b
0
40
σ
35 hn3
σ
30 hn10
Coulomb
25 ASCE 7-98
z, ft
BOCA, SBC, IBC

20 K
0 L = 20 ft
b
15
φ = 40o
10
5 S /S = 1
c b
0
40
35 σ
hn1
30 σ
hn3
25 σ
hn10
z, ft
Coulomb
20
L = 20 ft ASCE 7-98
b
15 BOCA, SBC, IBC
φ = 40o K
0
10
5 S /S = 4
c b
0
0 0.5 1 1.5 2 2.5
σ , kips/ft
2
h
Fig. 7.19 Lateral Earth Pressure behind Four Story Frames (φ = 40º)
7.2.6 Five-Story Rigidly Framed Earth Retaining Structures

The analysis results of four-story rigid frames are presented herein. Fig.7.20 below
illustrates the lateral earth pressure developed behind frames with bay length of 10
feet. The results shown are in general agreement with the four story frames
presented earlier, where the lateral earth pressure is found to vary between the
Coulomb active earth pressure for the lower stiffness frames and the soil load
prescribed by ASCE 7-98 for the stiffer structures.
50
σ σ
hn1 hn20
40 σ Coulomb
hn3
σ ASCE 7-98
30 hn6
z, ft
σ BOCA, SBC, IBC

hn10
L = 10 ft
20 b σ K
0
hn15
φ = 30o
10
S /S = 1
c b
0
50
σ σ
hn1 hn20
40 σ Coulomb
hn3
σ ASCE 7-98
30 hn6
z, ft
σ BOCA, SBC, IBC

hn10
L = 10 ft
20 b σ K
0
hn15
φ = 30o
10
S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3 3.5
σ , kips/ft
2
h
Fig. 7.20 Lateral Earth Pressure behind Five Story Frames (Lb= 10 ft, φ = 30º)
Similar results are also found for frames with bay length of 20 feet. Fig.7.21
illustrates the lateral earth pressures developed behind this group of rigid frames,
which is also in-line with the results discussed hitherto.
50
σ
hn1
40 σ
hn3
σ
hn10
30
z, ft
Coulomb
ASCE 7-98
L = 20 ft
20 b BOCA, SBC, IBC
φ = 30o K
0
10
S /S = 1
c b
0
50
σ
hn1
40 σ
hn3
σ
hn10
30
z, ft
Coulomb
ASCE 7-98
L = 20 ft
20 b BOCA, SBC, IBC
φ = 30o K
0
10
S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3 3.5
σ , kips/ft
2
h
Fig. 7.21 Lateral Earth Pressure behind Five Story Frames (Lb= 20 ft, φ = 30º)
The analysis results for the five-story frames retaining the backfill soil with an
internal angle of friction of 40º are shown in Fig.7.22.
In the top most graphic of said figure, the results of the single and 3-bay frames
are not shown given that their lateral stiffness were not adequate to carry the
retained soil, thus causing the soil mass to collapse. A similar behavior is also
indicated for the single bay frame with bay length of 20 feet and a column to beam
stiffness ratio of 1. For the remainder of the frames, the results are similar in
general to those found earlier for the various number of stories, which generally
show the lateral earth pressure increasing with increasing frame stiffness.
50
σ Coulomb
hn6
40 σ ASCE 7-98
hn10
σ BOCA, SBC, IBC
hn15
30
σhn20
z, ft
K
0
L = 10 ft
20 b
φ = 40o
10
S /S = 1
c b
0
50
σ σ
hn1 hn20
40 σ Coulomb
hn3
σhn6 ASCE 7-98
30
σ
z, ft
BOCA, SBC, IBC

hn10
σ K L = 10 ft
20 hn15 0 b
φ = 40o
10
Sc/Sb = 4
0
50
σ
hn3
40 σhn10
Coulomb
30 ASCE 7-98
z, ft
BOCA, SBC, IBC

K
0 L = 20 ft
20 b
φ = 40o
10
Sc/Sb = 1
0
50
σ ASCE 7-98
hn1
hn3
σ K
hn10 0
30
z, ft
Coulomb
Lb = 20 ft
20
φ = 40o
10
S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. 7.22 Lateral Earth Pressure behind Five Story Frames (φ = 40º)
7.2.7 The Case of Frames Braced against Lateral Sway

For all practical purposes, the lateral displacement undergone by shear wall or
braced frame structures is generally significantly smaller than rigidly framed
structures. Consequently, the preceding results presented from the analysis of
rigidly framed earth retaining structures could not be extended to represent the
lateral earth pressure loads developed in the backfill soil retained behind structures
braced against lateral sway.
Thus, for comparison and completeness purposes, two cases of rigid frames
with bay length of 10 feet were modified to include a 12 inch thick concrete shear
wall spanning a single bay to assess the effects of bracing the frames on the
development of lateral earth pressure in the backfill soil.
10
U
hRF
8 U
hSW
6 6 bays
z, ft
L = 10 ft
b
4
φ = 30o
2
S /S = 1
c b
0
0 0.002 0.004 0.006 0.008 0.01
h
10
6 bays σ
aRF
8
L = 10 ft σ
aSW
b
6 φ = 30 o Coulomb
z, ft
ASCE 7-98
S /S = 1 K
4 c b 0
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
σ' , kips/ft
2
h
Fig. 7.23 Displacement and Lateral Earth Pressure behind a 6-Bay Shear Wall Structure
7.2.8 Analysis of a Single Story 6-Bay Shear Wall Structure

The results of the analysis of a 6-bay frame with a column to beam stiffness ratio
of 1 and the backfill soil with an internal angle of friction of 30º modified to
include a shear wall in the third bay are shown in Fig.7.23. The top graphic shows
the lateral displacement of the rigid frame (RF) and the shear wall (SW) structures
where a substantially smaller deflection is found for the SW structure compared to
the rigid frame. The lateral earth pressure developed behind the shear wall
structure is more than 20 percent larger compared to the rigid frame.
7.2.9 Analysis of a Two Story 15-Bay Shear Wall Structure

The results of the analysis of a two story 15-bay frame shown in Fig.7.24 indicate
very similar results to those obtained from the analysis of the single story 6-bay
frame presented above. The lateral movement of the rigid frame modified to
include a shear wall is approximately 10 times smaller than the original rigid
20
15
15 bays
z, ft
10
L = 10 ft
b
φ = 30o
5 U
hRF
U S /S = 1
hSW c b
0
0 0.005 0.01 0.015 0.02
h
Fig. 7.24 Lateral Displacement of 15-Bay Shear Wall Structure
20
15 bays σ
aRF
15 L = 10 ft σ
aSW
b
φ = 30 o Coulomb
z, ft
10 ASCE 7-98
S /S = 1 K
c b 0
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 7.25 Lateral Earth Pressure behind a 15-Bay Shear Wall Structure
7.3 Conclusions 153
frame. The lateral earth pressure developed behind the shear wall structure (Fig
7.25) is more than 25% larger than its counterpart found behind the rigid frame.
7.3 Conclusions
The parametric numerical analysis of rigidly framed earth retaining structures

(RFERS) with varying lateral stiffness, bay lengths, number of bays, number of
stories simulating the addition of the backfill soil with varying properties using
staged construction computation indicates that the magnitude of lateral earth
pressure developed behind the rigid frames varies with the magnitude of the lateral
stiffness of the frames.
In general, it was found that for rigid frames with relatively lower stiffness, the
magnitude and distribution of the lateral earth pressure developed behind the
frames at the end of the backfill stage is comparable to the lateral earth pressure
obtained from the Coulomb’s active earth pressure theory, and is also in line with
the soil loads stipulated by some of the national building codes adopted in the
United States, such as the International Building Code and the BOCA code. The
soil loads prescribed by ASCE 7-98 were found to slightly overestimate the
pressure developed behind the more flexible frames.
On the other hand, for the stiffer frames, the lateral earth pressure was found to
be reasonably comparable to Coulomb’s active earth pressure for the top half of
the retained soil height, and increasing linearly thereafter to reach the magnitude
of the lateral earth pressure at rest at the bottom of the retained height.
This conclusion may not be extended however to encompass framed structures
braced against lateral sway such as shear wall or braced frame structures where
the lateral earth pressure was found to be even larger than that for rigid frames.
Chapter 8
Analysis of Single Story RFERS Subject to
Temperature Variations
Abstract. This chapter presents the results of numerical parametric analysis of

single-story structures with varying geometries and properties. The primary
purpose of this analysis is to investigate the effect of thermal movements of
Rigidly Framed Earth Retaining Structures (RFERS) on (1) the displacement of
the rigid frames, (2) the stresses developed in the structural elements, and (3) the
lateral earth pressure developed in the soil mass. The results are reported for 1,
10, 20 bay frames in this Chapter and for 3, 6, 15 bay frames in Appendix A for φ
= 30°. Results for φ = 40° are also presented in Appendix A.
8.1 Introduction
The soil-structure interaction governing the behavior of a full scale in-service
rigidly framed earth retaining structures (RFERS) subject to large temperature
variations was analyzed and discussed in earlier chapters and found to have
substantial effects on the stresses and strains developed in the retained soil mass
and the structural elements of the RFERS.
The soil-structure interaction problem involving thermal movements of the
RFERS is further examined herein through a numerical parametric analysis of
structures with varying geometries and properties. The primary purpose of this
analysis is to investigate the effects of thermal movements of the RFERS on the
displacement of the rigid frames, the stresses developed in the structural elements,
and the lateral earth pressure developed in the soil mass.
Additionally, the results of the numerical parametric analysis presented in
Chapter 7 for the earth pressure at backfill stage, combined with the parametric
analysis presented in this chapter will be used to draft general guidelines and
recommendation for the analysis and design of single story rigidly framed earth
retaining structures subjected to large temperature variations.
8.2 Numerical Parametric Analysis

Two-dimensional plane strain finite element analysis of single-story rigidly
framed earth retaining structures of varying number of bays, bay lengths, and

156 8 Analysis of Single Story RFERS Subject to Temperature Variations
member stiffness were performed using the commercial finite element analysis
software Plaxis. The structural members composing the rigid frames were
using an elastic-plastic Mohr Coulomb model with varying parameters. The
individual structural frames were assumed to have beam stiffness and column
stiffness that are respectively equal throughout the structure. The structural wall
retaining the backfill was assumed to be of concrete construction. The structures
were assumed to have infinitely rigid foundation strata.
To accommodate the plane strain limitations of the finite element model, the
elements allows an adequate modeling of the lateral stiffness of the rigid frame
Table 8.1 Finite Element Analysis Parameters (Single Story RFERS)
Parameter Variation
Number of Stories for Structural Frames, s 1
Moment of Inertia, Ib (in4) 2073 and 1036
Structural Columns Properties
4
Moment of Inertia, Ic (in ) 2073 and 41460
4
Moment of Inertia, Iw (in /ft) 2073
Internal Friction Angle, φ (Degrees) 30 and 40
Dilatency Angle (Degrees 0 and 10
Cohesion, c (psi) 0
Modulus of Elasticity, E (ksi) 41500 and 144000
Density, γ (pcf) 100 and 120
8.3 Analysis of Single Story Rigidly Framed Earth Retaining Structures 157
φ
ψ
γ
γ ν
Fig. 8.1 Typical Finite Element Analysis Parameters

initial stage where a structural frame is first completed and later receiving backfill
soil in several stages. After the addition of the backfill soil, an increase in
temperature of 60ºF was applied to the structure, followed by a temperature
decrease of 100ºF and subsequent cycles of 100ºF increase and decrease in
temperature. The displacement of the structures, stresses in the structural
elements, as well as the lateral earth pressure developed in the retained soil were
obtained for each temperature cycle.
The parameters varied for the finite element analysis are listed in Table 8.1 and
shown on Fig. 8.1. Overall, a total number of 48 rigid frames with varied stiffness
characteristics were analyzed with two backfill soil properties. Additional details
of the analysis models are also presented in Chapters 6 and 7. Numerical
parametric analysis of multi-story RFERS is presented in Chapter 9.
8.3 Analysis of Single Story Rigidly Framed Earth Retaining

Structures
8.3.1 Backfill Soil with 30º Internal Friction Angle

The analysis results of single story rigid frames are presented herein and Appendix
A. The results are reported for 1, 10, 20 bay frames (Chapter 8) and 3, 6, 15 bay
frames (Appendix A) under separate sections corresponding to the parameters
varied in Table 8.1. The results reported for each temperature cycle, in addition to
the initial backfill stage, are (1) the displacement of the retaining wall; (2) the
lateral earth pressure developed in the retained soil mass; (3) the retaining wall
bending moment; (5) the displacement of the end column; and (6) the end column
bending moment.
8.3.1.1 Single Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam
Stiffness Ratio, Sc/Sb, of 1)
The analysis results for a single bay frame with a bay length of 10 feet and a
column to beam stiffness ratio of 1 are presented in Fig. 8.2 for the retaining wall,
and 8.4 and 8.5 for the end column. The lateral earth pressure developed in the soil
mass is shown in Fig. 8.3.
Fig. 8.2 Analysis results for Retaining Wall (Single Bay, Lb = 10 ft, Sc/Sb = 1)
The results of Fig. 8.2 for the retaining wall indicate that temperature variations
induce relatively small changes in the wall displacement, bending moment and shear
force. The largest impact of thermal movement is found on the retaining wall
bending moment which is found to be 8 percent larger during the expansion cycles
compared with the initial backfill stage, and 5 percent larger during said cycles
compared with the bending moments developed during the contraction cycles.
The thermal movements undergone by a 10 ft long single-bay frame are
relatively smaller than longer multi-bay framed structures, and are therefore less
effective in producing larger stresses in the structure or the retained soil when
compared with the initial backfill stage. Additionally, a single story, single bay,
rigid frame possesses a smaller lateral stiffness compared with multi-bay frames
of same element properties, and will therefore undergo larger lateral displacement
and smaller expansion into the soil mass, thus having smaller effects on the lateral
earth pressure developed in the soil as illustrated in Fig. 8.3.
The horizontal end column displacements are shown in Fig. 8.4. The soil
restraint present at the wall side of rigid frame curtails the expansion movements
of the structure at the wall end, thus producing larger displacement at the end
column.
The displacement away from the retained soil mass at the top of the end column
during the expansion cycle is nearly 12.5 percent larger than its counterpart at the
initial backfill stage, and nearly 30 percent larger than the displacement occurring
during the contraction cycle.
The bending moment and shear force are both largest during the expansion
cycle. The moment at the top of the end column is nearly 12 percent larger during
the expansion cycle compared with the initial backfill stage, so is the shear force.
10
1 bay, L = 10 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 1 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 8.3 Retaining Wall Earth Pressure (Single Bay, Lb = 10 ft, Sc/Sb = 1)
Fig. 8.4 Horizontal End Column Displacement (Single Bay, Lb = 10 ft, Sc/Sb = 1)
10
1 bay, L = 10 ft
b
8
φ = 30 o
M
a
6 S /S = 1 M
c b
z, ft
ec1
M
cc1
4 M
ec2
M
cc2
2
M
ec3
0
-2 -1 0 1 2 3 4 5
10
1 bay, L = 10 ft
b Q
a
8 φ = 30o Q
ec1
S /S = 1 Q
6 cc1
c b
z, ft
Q
ec2
4 Q
cc2
Q
ec3
2
0
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3
End Column Shear Force, Q, kips
Fig. 8.5 End Column Bending Moment and Shear Force (Single Bay, Lb = 10 ft, Sc/Sb = 1)
The single bay frame with 10-ft bay length and a column to beam stiffness ratio of
4 possesses a substantially larger lateral stiffness than its counterpart presented
above (see Table 7.2). The analysis results for are shown in Fig. 8.6 through 8.9.
10
1 bay, L = 10 ft
b
8
φ = 30
o
Uha
6 S /S = 4 Uhec1
c b
z, ft
Uhcc1
4 Uhec2
Uhcc2
2
Uhec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
1 bay, L = 10 ft
b
8 φ = 30o M
a
S /S = 4 M
6 c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
1 bay, L = 10 ft
b
8
φ = 30o Q
a
S /S = 4 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Retaining Wall Shear Force, Q, kips
Fig. 8.6 Analysis Results for Retaining Wall (Single Bay, Lb = 10 ft, Sc/Sb = 4)
The horizontal movement of the retaining wall shown in Fig. 8.6 is

considerably smaller than its counterpart presented in Fig. 8.2 earlier for frames of
lesser lateral stiffness. The retaining wall moments and shear forces, however, are
of nearly similar magnitudes. The variation of the wall bending moment with the
temperature cycles in the stiffer frame (Fig. 8.6) is noticeably larger than its
counterpart shown in Fig. 8.2. This result is also similar but less pronounced for
the wall shear force.
The lateral earth pressure developed behind the retaining wall with the varying
temperature cycles is shown in Fig. 8.7 and appears to be nearly 35% larger for
the bottom half of the structure than its counterpart shown in Fig. 8.3.
10
1 bay, L = 10 ft
b σ'
a
8 φ = 30o σ'
ec1
S /S = 4 σ
'cc1
6 c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
10
1 bay, L = 10 ft U
b ha
8
φ = 30o U
hec1
U
S /S = 4 hcc1
6 c b
z, ft
U
hec2
U
4 hcc2
U
hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
Fig. 8.8 Horizontal End Column Displacement (Single Bay, Lb = 10 ft, Sc/Sb = 4)
The variation of the earth pressure magnitude with temperature cycles,

however, is relatively small, yet larger than its counterpart found behind the frame
with lesser stiffness as presented in Fig. 8.3.
The horizontal movements of the end column, shown in Fig. 8.8, are noticeably
smaller than the movements found for the end column of the frame with lesser
lateral stiffness. The range of movement, nevertheless, is nearly of the same
magnitude for both single bay frames.
The bending moment and shear forces developed in the end column, however,
are found to be markedly larger for the stiffer frame as shown in Fig. 8.9.
10
1 bay, L = 10 ft
b
8 φ = 30o M
a
6 S /S = 4 Mec1
c b
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-2 -1 0 1 2 3 4 5 6
10
1 bay, L = 10 ft
b
8 φ = 30o
Q
a
6 S /S = 4 Q
c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2
The analysis results for the single bay rigid frame with 20-ft bay length and a
column to beam stiffness ratio of 1 are shown in Fig. 8.10 through 8.12.
10
U
ha
8
U
hec1
U
6 hcc1
z, ft
U
hec2
1 bay, L = 20 ft
4 U b
hcc2
U φ = 30o
hec3
2 S /S = 1
c b
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
1 bay, L = 20 ft
b
8
φ = 30 o
M
a
6 S /S = 1 M
c b ec1
z, ft
M
cc1
4 Mec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
1 bay, L = 20 ft
b
8
φ = 30o Q
a
S /S = 1 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. 8.10 Analysis Results for Retaining Wall (Single Bay, Lb = 20 ft, Sc/Sb = 1)
The horizontal movements of the retaining wall shown in Fig. 8.10 are the
largest of all three frames presented thus far. This is predictable since the single
bay frame with a bay length of 20 feet and a column to beam stiffness ratio of has
the smallest stiffness of all three frames as indicated in Table 7.2.
Additionally, Fig. 8.10 indicates that the magnitude of the maximum bending
moment in the wall is relatively smaller than its counterpart presented previously.
An inflection point is found in the retaining wall bending moment diagram at
approximately one-fifth to one-quarter the height from the top of the wall, unlike
the two preceding rigid frames presented earlier. The location of the maximum
bending moment is also lower for the retaining wall shown in Fig. 8.10, compared
to those shown in Fig. 8.2 and 8.6.
The lateral earth pressure developed behind the retaining wall is presented in
Fig. 8.11, where the magnitude of the pressure is shown to be in-line with its
counterparts presented previously for the frames with 10-ft bay length.
10
1 bay, L = 20 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 1 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
The end column horizontal movements are shown in Fig. 8.12 to be noticeably
larger than the movements of the end column for the 10-ft bay frames.
The end column bending moment, shown in Fig. 8.12, is also larger compared
to the 10-ft bay frames, whereas the range of bending moment change with
temperature cycles is relatively smaller.
10
1 bay, L = 20 ft
b U
ha
8
φ = 30o U
hec1
S /S = 1 U
6 c b
hcc1
z, ft
U
hec2
4 U
hcc2
U
hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
1 bay, L = 20 ft
b
8
φ = 30 o
M
a
6 S /S = 1 M
c b
z, ft
ec1
M
cc1
4 M
ec2
M
cc2
2
M
ec3
0
-2 0 2 4 6 8
10
1 bay, L = 20 ft
b Q
a
8 φ = 30o Q
ec1
S /S = 1 Q
6 cc1
c b
z, ft
Q
ec2
4 Q
cc2
Qec3
2
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2
The analysis results for the single bay rigid frame with 20-ft bay length and a
column to beam stiffness ratio of 4 are shown in Fig. 8.13 through 8.15.
10
1 bay, L = 20 ft
b
8
φ = 30
o
Q
a
S /S = 4 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
10
1 bay, L = 20 ft
b
8
φ = 30o M
a
6 S /S = 4 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
1 bay, L = 20 ft
b
8 φ = 30o U
ha
6 S /S = 4 U
c b hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2
U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
Fig. 8.13 Analysis results for Retaining Wall (Single Bay, Lb = 20 ft, Sc/Sb = 4)
The horizontal retaining wall movements for the 20-ft bay frame with a column
to beam stiffness ratio of 4 are substantially smaller than their counterparts of the
smaller stiffness frame shown in Fig. 8.8. The range of movements with
temperature cycles is also larger when compared to the results found for the three
rigid frames presented earlier. According to table 7.2, the lateral stiffness of the
frame presented in this section is approximately 2.5 times larger than the first rigid
frame presented (F1), and 40% smaller than the second frame (F7), and 4.5 larger
than the third frame (F13). This indicates that the frame possesses a relatively
large lateral stiffness that is responsible for smaller horizontal movements found
from the analysis. Additionally, length of the rigid frame is twice as long as the
10-ft bay frames thus resulting in larger horizontal thermal strains due to
temperature changes, which explains the larger range of movements for retaining
wall shown in Fig. 8.13.
The magnitude of the bending moment in the wall is similar to its counterpart
shown in Fig. 8.8. In this case, however, the moment developed in the wall during
the expansion cycles is nearly 50% larger than the moments found during
contraction cycles and at the initial backfill stage. The bending moment
distribution indicates the retaining wall assumed a double curvature shape.
The lateral earth pressure developed behind the structure, shown in Fig. 8.14, is
nearly equal at the initial backfill stage and during the contraction cycles, but
slightly larger during the expansion cycles. The longer expansion length of the 20-
ft bay frame combined with the larger column to beam stiffness ratio resulted in
the largest increase in lateral earth pressure during the expansion cycles of all
single-story single bay frames presented thus far.
10
1 bay, L = 20 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 4 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
The horizontal movements of the end column, shown in Fig. 8.15, indicate a
large increase in horizontal displacement away from the soil mass during the
expansion cycles compared with the initial backfill stage. The shear force and
bending moment in the end column corresponding to the last expansion cycle are
nearly 32% larger than their counterpart developed at the backfill stage. The
difference in displacement, shear and bending moment in the end column between
the initial backfill stage and the last expansion cycle is found to be more pronounced
for the stiffer frames presented thus far. This indicates that temperature movements
have a more considerable effect on stiffer single story RFERS.
10
1 bay, L = 20 ft Q
b a
8
φ = 30o Q
ec1
Q
6 cc1
S /S = 4
z, ft
c b Q
ec2
4 Q
cc2
Q
ec3
2
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2
10
1 bay, L = 20 ft
b
8
φ = 30o U
ha
S /S = 4 U
6 c b
hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
1 bay, L = 20 ft
b
8
φ = 30o
Ma
6 S /S = 4 Mec1
c b
z, ft
Mcc1
4
M
ec2
Mcc2
2
Mec3
0
-2 0 2 4 6 8 10 12
8.3.1.5 10-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam Stiffness
Ratio, Sc/Sb, of 1)
The analysis results for a 10-bay frame with a bay length of 10-ft and a column to
beam stiffness ratio of 1 are presented herein.
Fig. 8.16 shows the retaining wall movements for the 10-bay frame. The
relatively lower frame stiffness is evident from the wall movements during the
expansion cycles, where while the frame length is the largest thus far, the frame
expansion into the soil is rather smaller than that for stiffer frames (see Appendix
A for additional configurations).
10
10 bays
8
L b = 10 ft U ha
6 φ = 30 o U hec1
z, ft
U
hcc1
Sc /Sb = 1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
Fig. 8.16 Retaining Wall Horizontal Movements (Ten-Bay, Lb = 10 ft, Sc/Sb = 1)
The retaining wall bending moment diagram is shown in Fig. 8.17. The
magnitude of the maximum moment at about mid-height of the wall developed
during the last expansion cycle is found to be nearly 2.5 times larger than its
counterpart developed during the initial backfill stage.
The shear force diagram, also shown in Fig. 8.17, indicates that the maximum
shear force occurs in the wall during the latter expansion cycles with a magnitude
nearly 4 times larger than any other cycles.
The lateral earth pressure developed in the retained soil behind the rigid frame
is shown in Fig. 8.18.
10
10 bays
8
Lb = 10 ft M
a
φ = 30o M
6 ec1
z, ft
S /S = 1 M
cc1
c b
4 Mec2
Mcc2
2 Mec3
0
-8 -6 -4 -2 0 2 4 6
10
10 bays
8
L b = 10 ft Qa
6 φ = 30 o Qec1
z, ft
Qcc1
S /S = 1
c b
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. 8.17 Retaining Wall Shear and Moment Diagrams (Ten-Bay, Lb = 10 ft, Sc/Sb = 1)
10
10 bays σ'
a
8
σ'
ec1
L = 10 ft
b
σ'
6 cc1
φ = 30
o
z, ft
σ'
ec2
4 S /S = 1
c b
σ'
cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h
Fig. 8.18 Retaining Wall Earth Pressure (Ten-Bay, Lb = 10 ft, Sc/Sb = 1)

The variation in the magnitude and distribution of earth pressure is found to be

large between the various temperature cycles. During the last expansion cycle, the
magnitude of the earth pressure at the top region of the wall is nearly 9 times
larger than during the initial backfill stage or contraction cycles.
10
10 bays U
ha
8 U
L = 10 ft hec1
b
U
φ = 30
o hcc1
6
z, ft
U hec2
S /S = 1
c b U hcc2
4
U hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
Horizontal End Column Movement, Uh, ft
Fig. 8.19 End Column Horizontal Movement (Ten-Bay, Lb = 10 ft, Sc/Sb = 1)
The horizontal end column movements, shown in Fig. 8.19, indicate a relatively
large displacement during the expansion cycle. The maximum displacement
magnitude at the top of the end column during the expansion cycle is nearly 8
times larger than its counterpart found at the end of the initial backfill stage. The
end column bending moment diagram, shown in Fig. 8.20, indicate that the
maximum moment occurs at the top of the end column during the last expansion
cycle with a magnitude more than times larger than its counterpart developed at
the end of the initial backfill stage.
10
10 bays
8
L = 10 ft
b Ma
φ = 30
o
6 Mec1
z, ft
S /S = 1 Mcc1
c b
4 M
ec2
M
cc2
2
M
ec3
0
-2 -1 0 1 2 3 4 5
Fig. 8.20 End Column Bending Moment (Ten-Bay, Lb = 10 ft, Sc/Sb = 1)
Ratio, Sc/Sb, of 4)
beam stiffness ratio of 4 are presented herein. This frame has the largest lateral
stiffness of all frames presented thus far, as indicated in Table 7.2.
The retaining wall horizontal movements are shown in Fig. 8.21. The
expansion and contraction movements appear nearly symmetrical with respect to
the small initial movements undergone at the backfill stage, despite the presence
of the retained soil restraint behind the wall. The movement of the frame into the
original position of the retained soil mass is the largest found thus far, but is
similar however to the 6-bay frame with a column to beam stiffness ratio of 4
(Appendix A). It is therefore apparent that frames with relatively large lateral
stiffness are capable of overcoming the soil restraint and undergo large expansion
movement into the soil mass.
10
10 bays
8
L = 10 ft U
b ha
φ = 30
o
6 U
hec1
z, ft
U
S /S = 4 hcc1
c b
4 U
hec2
U
hcc2
2
Uhec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
Fig. 8.21 Retaining Wall Horizontal Movements (Ten-Bay, Lb = 10 ft, Sc/Sb = 4)
The bending moment diagram for the retaining wall is shown in Fig. 8.22,
along with the shear force diagram. The magnitude of the moment developed
during the temperature cycles are substantially larger than the other frames
presented hitherto. The maximum bending moment found at the top of the
retaining wall at the end of the last expansion cycle is found nearly 16 times larger
than its counterpart developed at the end of the initial backfill cycle.
The variation in shear force magnitude is also found relatively large between
the various stages of analysis. The maximum shear force at the last expansion
cycle is nearly 7 times larger than its counterpart found at the end of the backfill
stage.
10
8
M
10 bays a
6 M
ec1
z, ft
Lb = 10 ft M
cc1
φ = 30
4 o
M
ec2
Sc/Sb = 4 M
cc2
2
M
ec3
0
-15 -10 -5 0 5 10 15 20
10
10 bays
8 L = 10 ft
b Qa
φ = 30
o
6 Qec1
z, ft
Sc /Sb = 4 Qcc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-6 -4 -2 0 2 4 6 8
Fig. 8.22 Retaining Wall shear and Moment Diagram (Ten-Bay, Lb = 10 ft, Sc/Sb = 4)
10
10 bays σ'
a
8
σ'
L = 10 ft ec1
b
σ'
6 φ = 30 o cc1
z, ft
σ'ec2
S /S = 4 σ'cc2
4 c b
σ'ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h

The lateral earth pressure developed behind the retaining wall at the various
analysis cycles is shown in Fig. 8.23. The earth pressure developed during the
initial backfill stage and the contraction cycles is found to be substantially smaller
than the pressure found during the expansion cycles. Furthermore, the earth
pressure distribution during the expansion cycles is not triangular in shape as
stipulated by classical earth pressure theories, given that the wall movements into
the soil mass are larger at the top of the wall than at its bottom.
The end column horizontal movements, shown in Fig. 8.24, are in line with the
horizontal wall movements at the other end of the frame. This is attributed to the
relatively small effect of the backfill soil restraint at the wall end of the rigid
frame, evident from the frame movements into the soil mass at the various
temperature cycles.
10
10 bays
8 L = 10 ft
b
U
φ = 30
o ha
6 U hec1
z, ft
S /S = 4 U hcc1
c b
4 U hec2
U hcc2
2
U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
10 bays
8 L = 10 ft
b
M
φ = 30o a
6 M
z, ft
ec1
S /S = 4 M
c b
cc1
4
Mec2
Mcc2
2
Mec3
0
-20 -15 -10 -5 0 5 10 15 20
Fig. 8.24 End Column Analysis Results (Ten-Bay, Lb = 10 ft, Sc/Sb = 4)

The bending moment in the end column during the expansion and contraction
cycles has a magnitude substantially larger than the remainder of the RFERS
presented thus far. The maximum bending moment developed in the end column
during the last expansion cycle is nearly 10 times larger than its counterpart
developed at the end of the initial backfill stage.
Ratio, Sc/Sb, of 1)
beam stiffness ratio of 1 are presented herein. The frame is the longest structure
presented thus far, but has the approximately 10 times the lateral stiffness of the
preceding 10 bay frame with Lb, 10 feet and Sc/Sb, of 4,as indicated in Table 7.2.
The horizontal retaining wall movements, shown in Fig. 8.25, indicate that the
temperature movements during the expansion and contraction cycles are
disproportionate, in contrast with the frame formerly presented. While the
structure presented herein has the longest potential expansion (or contraction)
length, thus the largest potential movement due to the temperature change, the
frame possesses a relatively smaller lateral stiffness insufficient to overcome the
retained-soil restraint and pressure.
The variation in retaining wall bending moment with the temperature cycles is
also shown in Fig. 8.25. The maximum moment occurring at about mid-height of
the wall during the last expansion cycle is nearly 4 times larger than the moment
developed during the contraction cycles or the initial backfill stage. The maximum
shear force also occurs at the last expansion cycle, and is more than 3 times larger
than its counterpart found at the initial backfill stage or during the contraction
cycles.
The retaining wall earth pressure, shown in Fig. 8.26 for the various analysis
cycles, indicate a substantial increase in the magnitude of the pressure towards the
top of the structure developing during the expansion cycles. The earth pressure
distribution is an inverted triangular shape with the maximum pressure occurring
near the top of the wall.
The end column horizontal movements are shown in Fig. 8.27. The magnitude
of horizontal displacements at the top of the end column, and the overall range of
movement during the various analyses cycles are the largest of all frames
presented thus far.
The large range of movements is attributed to the presence of backfill soil
behind other end of the frame, thereby restraining the expansion movement of the
structure at the soil side, which is consequently undertaken at the end column. The
maximum bending moment in the end column occurs at the last expansion cycle
and is nearly twice its counterpart developed at the end of the initial backfill stage.
10
10 bays
8 L = 20 ft
b
φ = 30
o
6
z, ft
S /S = 1
c b
4 U U
ha hec2
U hec1 Uhcc2
2
U hcc1 Uhec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
10 bays
8
Lb = 20 ft
Ma
φ = 30o Mec1
6
z, ft
S /S = 1 Mcc1
c b
4 M
ec2
M
cc2
2
M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
10 bays
8
L = 20 ft Qa
b
φ = 30
o
Qec1
6
z, ft
Qcc1
S /S = 1
c b
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. 8.25 Retaining Wall Analysis Results (Ten-Bay, Lb = 20 ft, Sc/Sb = 1)

10
8
σ'
10 bays a
6 σ'ec1
z, ft
L = 20 ft
b σ'cc1
4 φ = 30o σ'ec2
S /S = 1 σ'cc2
2 c b
σ'
ec3
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h
10
6 10 bays
z, ft
Lb = 20 ft
4 U U
φ = 30o ha hec2
Uhec1 U hcc2
2 Sc/Sb = 1
Uhcc1 U hec3
0
-0.04 -0.02 0 0.02 0.04 0.06 0.08
10
M
a
8
M
ec1
6 10 bays M
cc1
z, ft
M
L = 20 ft ec2
b
4 Mcc2
φ = 30 o
Mec3
2 S /S = 1
c b
0
-8 -6 -4 -2 0 2

Ratio, Sc/Sb, of 4)
beam stiffness ratio of 4 are presented herein. The frame is of the same total length
as the latter structure presented, but has the second largest lateral stiffness of all
frames presented thus far, as indicated in table 7.2. The lateral frame stiffness is
also comparable to that of a six-bay frame with 10-ft bay length and a column to
beam stiffness ratio of 4 (Appendix A).
Fig. 8.28 presents the horizontal retaining wall movements obtained from the
various analyses cycles. The displacement at the top of the wall during the
expansion and contraction cycles is found nearly symmetric about the wall
displacement at the initial backfill stage. The relatively large lateral stiffness of the
RFERS coupled with the greater expansion length resulted in substantial
movements into the soil mass during the expansion cycles.
10
8
U
ha
6 10 bays U
hec1
z, ft
U
Lb = 20 ft hcc1
4 U
φ = 30o hec2
U
hcc2
2 Sc/Sb = 4 U hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
Fig. 8.28 Horizontal Retaining Wall Movements (Ten-Bay, Lb = 20 ft, Sc/Sb = 4)
The bending moment and shear force diagrams are shown in Fig. 8.29. The
maximum moment occurs at the top of the wall during the last expansion cycle,
and its magnitude is nearly 4 times larger than its counterpart found at the end of
the initial backfill stage, and approximately twice the moment developed during
the contraction cycles. On the other hand, the maximum shear force is nearly 6
times larger at the last expansion cycles compared with the initial backfill stage.
The retaining wall earth pressure is shown in Fig. 8.30, where the pressure
during the expansion cycles is found to be substantially larger than during any
other cycle particularly for the one half of the wall height.
The analysis results for the end column are, in general, similar to those for the
retaining wall and are shown in Fig. 8.31. The horizontal displacements of the end
column during the temperature cycles are also found to be nearly symmetrical
about the column movements during the initial backfill stage. The magnitude of
the bending moment during the last expansion cycles are comparable to those
found for the retaining wall when compared to the moments developed at the end
of the initial backfill stage and during the contraction cycles.
10
8
M
10 bays a
6 M
ec1
z, ft
Lb = 20 ft
M
cc1
4 φ = 30o Mec2
Sc/Sb = 4 Mcc2
2
Mec3
0
-15 -10 -5 0 5 10 15 20 25
10
8
Q
a
10 bays
6 Q
ec1
z, ft
L = 20 ft Q
b
cc1
4 φ = 30o Q
ec2
S /S = 4 Q
cc2
2 c b
Q
ec3
0
-4 -2 0 2 4 6 8
Fig. 8.29 Retaining Wall Moment and Shear Diagrams (Ten-Bay, Lb = 20 ft, Sc/Sb = 4)
10
8
σ' a
6 σ' ec1
z, ft
σ'
cc1
4 10 bays, L = 20 ft
b
σ'
ec2
φ = 30
o
σ'
cc2
2
S /S = 4
c b
σ'
ec3
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
2
σ' , kips/ft
h

10
8
U
ha
10 bays
6 U
hec1
z, ft
L = 20 ft U
b hcc1
4 φ = 30
o
U
hec2
Uhcc2
2 S /S = 4
c b
Uhec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
10
6 10 bays
z, ft
L = 20 ft
b
4 M M
a ec2 φ = 30o
M M
ec1 cc2
2 S /S = 4
M M c b
cc1 ec3
0
-25 -20 -15 -10 -5 0 5 10 15
Ratio, Sc/Sb, of 1)
beam stiffness ratio of 1 are presented herein. The frame has the third largest
lateral stiffness of all frames presented thus far in this chapter and the largest
expansion length.
The retaining wall analysis results are shown in Fig. 8.32. The horizontal
retaining wall movements indicate a horizontal displacement occurring at end of
the initial backfill stage larger than the movements found for stiffer frames during
the same analysis stage. Furthermore, thermal movements of the frame at retaining
wall are smaller during expansion cycle compared with the contraction
movements. The rigid frame stiffness is therefore smaller than that required to
offset the soil restraint.
The maximum bending moment is found at mid-height of the wall occurring at

the end of the last expansion cycle. Said moment magnitude is more than 4 times
larger than its counterpart developed during the initial backfill stage, and nearly
twice that developed during the contraction cycles.
10
8
U ha
6 20 bays U hec1
z, ft
U
L = 10 ft hcc1
b
4 U
φ = 30
o hec2
U
hcc2
2 S /S = 1
c b U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
8
M
20 bays a
6 M
ec1
Lb = 10 ft
z, ft
M
cc1
4 φ = 30o Mec2
Sc/Sb = 1 Mcc2
2 Mec3
0
-8 -6 -4 -2 0 2 4 6
10
20 bays
8 L = 10 ft
b
φ = 30
o
6
z, ft
Sc /Sb = 1
4 Q Q
a ec2
Q Q
ec1 cc2
2
Q Q
cc1 ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. 8.32 Retaining Wall Analysis Results (Twenty-Bay, Lb =10 ft, Sc/Sb = 1)
The maximum shear force is also found to occur during the last expansion
cycle, but at the top of the wall. The magnitude of said force is nearly 80 times
larger than the shear developed during the contraction cycles, and approximately 7
times larger than its counterpart found at the end of the initial backfill stage.
The lateral earth pressure developed behind RFERS during the various analysis
cycles is shown in Fig. 8.33. The results are found to be similar to other frames
presented earlier with large expansion length and relatively large lateral stiffness.
The horizontal movements and bending moments for the end column of the 20-
bay frame are shown in Fig. 8.34. The column displacements with the temperature
changes are found to be fairly larger during the expansion cycles in the direction
away of the retained soil mass. This is practically a mirror image of the horizontal
displacements found at the retaining wall end of the frame where the retained soil
poses sufficient restraint to resist part of the expansion movements. The maximum
bending moment occurs at the top of the end column at the end of the last
expansion cycle, with a magnitude nearly 6 times larger than its counterpart found
at the end of the initial backfill stage, and approximately twice as large at the
moment developed during the expansion cycles.
10
20 bays,
8 L = 10 ft
b
φ = 30o
6
z, ft
S /S = 1
c b
4
σ' σ'
a ec2
σ' σ'
2 ec1 cc2
σ' σ'
cc1 ec3
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 8.33 Retaining Wall Earth Pressure (Twenty-Bay, Lb =10 ft, Sc/Sb = 1)
10
20 bays
8 L = 10 ft
b
φ = 30o
6
z, ft
S /S = 1
c b
4
U U
ha hec2
2 U
hec1
U
hcc2
U U
hcc1 hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
20 bays
8 Lb = 10 ft
φ = 30o
6
z, ft
S /S = 1
c b
4 M M
a ec2
M M
ec1 cc2
2
Mcc1 Mec3
0
-2 -1 0 1 2 3 4 5
Fig. 8.34 End Column Analysis Results (Twenty-Bay, Lb =10 ft, Sc/Sb = 1)
Ratio, Sc/Sb, of 4)
beam stiffness ratio of 4 are presented herein. The frame has the largest lateral
stiffness of all frames presented thus far and the largest expansion length. This is
also the last frame in this series to be presented.
The analysis results for the retaining wall are shown in the Fig. 8.35. The
horizontal wall displacements are qualitatively similar to the 15-bay frame with
similar element stiffness (Appendix A), where the initial displacements are found
relatively small, and the subsequent temperature related expansion and contraction
movements are found to be nearly symmetric about the initial wall displacement.
The magnitude of the horizontal thermal movements is, however, larger for the
20-bay frame compared with its 15-bay counterpart.
The bending moment diagram indicates that magnitude of the maximum
moment occurring during expansion or contraction cycles is nearly the same. The
shear force, however, is substantially larger at top of the wall during the expansion
cycles, with a magnitude nearly 2.7 times its counterpart developed during
contraction cycles, and more than 11 times the shear found at the end of initial
backfill stage.
10
8
U
ha
6 20 bays U
hec1
z, ft
U
hcc1
L = 10 ft
4 b
U
hec2
φ = 30o U
hcc2
2
S /S = 4 U
c b hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
8
M
a
6 20 bays M
z, ft
ec1
L = 10 ft M
b cc1
4 M
φ = 30 o ec2
M
2 cc2
S /S = 4 M
c b
ec3
0
-40 -30 -20 -10 0 10 20 30 40
10
20 bays
8 L = 10 ft
b
6 φ = 30o
z, ft
S /S = 4
c b
4 Q Q
a ec2
Q Q
ec1 cc2
2
Q Q
cc1 ec3
0
-6 -4 -2 0 2 4 6 8
Fig. 8.35 Retaining Wall Analysis Results (Twenty-Bay, Lb =10 ft, Sc/Sb = 4)
The lateral earth pressure developed behind the structure is shown in Fig. 8.36
for the various analysis cycles. The distribution of the pressure during the last
expansion cycle is similar for other frames presented earlier. The magnitude of the
maximum earth pressure, however, developed behind this 20-bay frame is the
largest found thus far.
10
6
z, ft
20 bays
4 L = 10 ft
b
σ' σ'
a ec2
φ = 30o σ' σ'

2 ec1 cc2
S /S = 4
c b σ' σ'
cc1 ec3
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 8.36 Retaining Wall Earth Pressure (Twenty-Bay, Lb =10 ft, Sc/Sb = 4)
10
8
20 bays U
ha
6 U
hec1
z, ft
L = 10 ft
b U
hcc1
4 φ = 30o U
hec2
S /S = 4 U
hcc2
2 c b
U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
8
M
a
6 20 bays
M
z, ft
ec1
Lb = 10 ft M
cc1
4
φ = 30 M
o
ec2
M
cc2
2 Sc/Sb = 4
M
ec3
0
-40 -30 -20 -10 0 10 20 30 40
Fig. 8.37 End Column Analysis Results (Twenty-Bay, Lb =10 ft, Sc/Sb = 4)
8.4 Conclusions 187
The end column analysis results are shown in Fig. 8.37. The horizontal end
column movements are shown as nearly a mirror image of the retaining wall
movements presented in Fig. 8.35. This indicates minimal restraining effect of the
soil retained behind the structure during the temperature cycles. The bending
moment diagrams show that the maximum magnitude of flexural stresses occur
during the last expansion cycle, albeit only slightly larger than their counterparts
found during the contraction cycles.
8.4 Conclusions
The effects of large temperature changes on the displacement of, and stresses
developed in RFERS, and the lateral earth pressure developed in the retained the
soil mass were investigated. Two-dimensional plane strain finite element analysis
of single-story rigidly framed earth retaining structures of varying number of bays,
bay lengths, and member stiffness were performed using the commercial finite
element analysis software Plaxis. The analysis of single-story single-bay rigid
frames indicates that the lateral earth pressure developed behind RFERS due to
temperature variations is larger when the lateral stiffness of the RFERS is larger.
For frames with similar stiffness, the expansion length also plays an important
role. The lateral earth pressure found during the expansion cycle of multi-bay
frames are substantially larger than those of the single-bay frames, and smaller
during the contraction cycle for the same frame. Similarly, RFERS undergo larger
displacements, particularly at the free end when the lateral stiffness decreases, and
develop higher stresses, when the frames lateral stiffness and length increase.
Furthermore, the rigid frames are shown to move into the original plane of the
retained soil mass during expansion, unlike the less stiff RFERS analyzed.
For relatively stiff RFERS, the lateral earth pressure developed in the backfill
soil during the various expansion cycles are much larger compared with the
contraction cycles and the initial backfill stage. The pressure distribution in some
cases is nearly an inverted triangle with the maximum pressure occurring near the
top of the structure.
The magnitudes of the maximum shear force and bending moment are
substantially affected by thermal cycles, particularly for stiffer frames.
Additionally, walls assume a variety of curved shapes due to temperature. These
results indicate that a reinforced concrete wall designed for the flexural stresses
produced from the application of the backfill soil will be inadequately reinforced
to resist the stresses produced during the expansion cycles.
A large number of parametric studies for 3, 6, 15 bay frame structures for φ =
30° is also included in Appendix A as well as analyses for φ = 40° in order to aid
with identifying the effect of thermal soil structure interaction on structures, which
might be of interest to the reader.
Chapter 9
Multi-story RFERS Subject to Temperature
Variation
Abstract. This chapter presents the results of numerical parametric analysis of

multi-story structures with varying geometries and properties. The primary
purpose of this analysis is to investigate the effects of thermal movements of
Rigidly Framed Earth Retaining Structures (RFERS) on (1) the displacement of
lateral earth pressure developed in the soil mass. The results are reported for three
and five story structures in this Chapter and for two and four story structures in
Appendix B.
9.1 Introduction
The soil-structure interaction problem involving thermal movements of the multi-
story RFERS is examined herein through a numerical parametric analysis of
structures with varying geometries and properties, similar to the single story
RFERS presented in Chapter 8. The purpose of this analysis is to investigate the
effects of thermal movements of the RFERS on the displacement of the rigid
frames, the stresses developed in the structural elements, and the lateral earth
pressure developed in the soil mass.
9.2 Numerical Parametric Analysis

Two-dimensional plane strain finite element analysis of multi story rigidly framed
earth retaining structures of varying length, height, number of bays, number of
levels, and member stiffness were performed using the commercial finite element
analysis software Plaxis. The structural members composing the rigid frames were
using a Mohr Coulomb model with varying parameters. The individual structural
frames were assumed to have beam stiffness and column stiffness that were
respectively equal throughout the structure. The structural wall retaining the
backfill was assumed to be of concrete construction. The structures were assumed
to have infinitely rigid foundation strata.

190 9 Multi-story RFERS Subject to Temperature Variation
To accommodate the plane strain limitations of the finite element model, the
elements allows an adequate modeling of the lateral stiffness of the rigid frame
initial stage where a structural frame was first completed and later received
backfill soil in several stages. After the addition of the backfill soil, an increase in
temperature of 60ºF was applied to the structure, followed by a temperature
decrease of 100ºF and subsequent cycles of 100ºF increase and decrease in
temperature. The displacement of the structures, stresses in the structural
Table 9.1 Finite Element Analysis Parameters
Parameter Variation
Number of Stories for Structural Frames, s 2, 3, 4, 5
4
Moment of Inertia, Ib (in ) 2073 and 1036
Structural Column Properties
Moment of Inertia, Ic (in4) 2073 and 41460
Moment of Inertia, Iw (in4/ft) 2073
Internal Friction Angle, φ (Degrees) 30
Dilatancy Angle, ψ (Degrees) 0
Cohesion, c (psi) 0
Modulus of Elasticity, E (ksi) 41500
Density, γ (pcf) 100
9.2 Numerical Parametric Analysis 191
elements, as well as the lateral earth pressure developed in the retained soil were
obtained for each temperature cycle.
The parameters varied for the finite element analysis are listed in table 9.1 and
shown on Fig. 9.1. Overall, a total number of 96 rigid frames with varied stiffness
characteristics were analyzed with one backfill soil properties. Additional details
of the analysis models are also presented in Chapters 6 and 7. Numerical
parametric analysis of single-story RFERS was presented in Chapter 8.
φ
ψ
γ
γ ν
Fig. 9.1 Typical Finite Element Analysis Parameters
9.2.1 Three Story Rigidly Framed Earth Retaining Structures

All analyses presented in this chapter and in Appendix B were performed
assuming an angle of friction φ of 30° for the retained soil.
9.2.1.1 Frames with Bay Length, Lb, 10 Feet, and Column to Beam Stiffness
Ratio, Sc/Sb, of 1
The horizontal movements of the retaining wall portion of the three story RFERS
with a bay length of 10 feet and column to beam stiffness ratio of 1 are presented
in Fig. 9.2 and 9.3 for the initial backfill stage and the last expansion cycle,
respectively. The retaining wall displacements are found to vary slightly between
the backfill stage and the last expansion cycle. This is indicative of the RFERS
inability to expand into the soil mass during temperature rise cycles. The wall
movements are essentially found to have increased at the end of the last expansion
cycle albeit only slightly.
30
25
U
20 hn1
U
z, ft
hn3
15 U
L = 10 ft hn6
b
U
10 φ = 30o hn10
U
hn15
5 S /S = 1
c b U
hn20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Fig. 9.2 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The lateral earth pressure developed behind the rigid frames at the end of the
last expansion cycle is shown in Fig. 9.4. The pressure magnitude and distribution
in the soil mass retained by the single bay, and the 3, 6 and 10-bay frames, is
found to be comparable to the Coulomb’s active earth pressure, and the soil loads
prescribed by the national American building codes.
30
25
U
20 hen1
U
z, ft
hen3
15 U
L = 10 ft hen6
b
U
10 φ = 30 o
hen10
U
hen15
5 S /S = 1
c b U
hen20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Fig. 9.3 Retaining Wall Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The pressure is subsequently found to be larger behind the remainder of the

RFERS, increasing in magnitude with the increase in frame length and number of
bays. The 15-bay frame is found to develop a pressure slightly larger than the load
prescribed by ASCE 7-98, but smaller than the lateral earth pressure at rest, while
the pressure behind the 20-bay frame is found to be yet larger, nearly equal to the
lateral earth pressure at rest. The pressure distributions remain nearly triangular as
predicted by classical lateral earth pressure theories, and are qualitatively in line
with the results of the single RFERS discussed in Chapter 8 and two story RFERS
in Appendix B.
30
σ σ
hen1 hen20
25
σ Coulomb
hen3
20 σ ASCE 7-98
hen6
σ BOCA, SBC, IBC
z, ft
hen10
15
L = 10 ft σ K0
b hen15
10 φ = 30 o
5 S /S = 1
c b
0
0 0.5 1 1.5 2 2.5
σ , kips/ft
2
h
Fig. 9.4 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
The retaining wall bending moments found at the end of the initial backfill
stage and the last expansion cycle are shown in Fig. 9.5 and 9.6, respectively.
The magnitude of the bending moment in the retaining wall portion of the
single, 3 and 6-bay frames are found to vary slightly between the two stages of
analysis, with the larger variation obtained for the frames with the larger number
of bays. The bending moments at the first frames level of the 10, 15 and 20-bay
frames, on the other hand, are found to decrease at the end of the last expansion
cycle, compared with their counterparts at the initial backfill stage.
30
M L = 10 ft
bn1 b
25
M
bn3 φ = 30o
20 M
bn6
S /S = 1
M c b
z, ft
bn10
15
M
bn15
10 M
bn20
0
-100 -80 -60 -40 -20 0 20
Fig. 9.5 Retaining Wall Moment at Backfill Stage (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
30
M L = 10 ft
en1 b
25
M φ = 30o
en3
20 M
en6
S /S = 1
M c b
z, ft
en10
15
M
en15
10 M
en20
0
-100 -80 -60 -40 -20 0 20
Fig. 9.6 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The end column horizontal movements are presented in Fig. 9.7 and 9.8 for the
initial backfill stage and the last expansion cycle, respectively.
The variation between the horizontally displaced positions of the end columns
of the RFERS at the end of the initial backfill stage and the last expansion cycle
are found to be nominal for the single bay, and the 3 and 6-bay frames. For the
larger frames, on the other hand, the horizontal deflections of the end columns at
the end of the last expansion cycle are substantially larger, compared to their
counterparts found at the initial backfill stage.
30
25
U
20 hn1
U
z, ft
hn3
15 U
L = 10 ft hn6
b
U
10 φ = 30 o
hn10
U
hn15
5 S /S = 1
c b U
hn20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Fig. 9.7 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
30
25
U
20 hen1
U
z, ft
hen3
15 U
L = 10 ft hen6
b
U
10 φ = 30o hen10
U
hen15
5 S /S = 1
c b U
hen20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Fig. 9.8 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The bending moment diagrams found for the end column at the end of the
initial backfill stage and the last expansion cycle are shown in Fig. 9.9 and 9.10
respectively.
The results indicate that the bending moments in the end column of the single,
3 and 6-bay frames are nearly identical at the end of both analysis stages. The
magnitude of the bending moments at the first framed level in the end columns of
the 10, 15 and 20 bay frames, on the other hand, are found to decrease at during
the last expansion cycle compared to their counterparts at the end of the initial
backfill stage.
30
L = 10 ft M
b
bn1
25
φ = 30 o
M
bn3
20 M
S /S = 1 bn6
c b
M
z, ft
bn10
15
M
bn15
10 M
bn20
0
-20 -10 0 10 20 30 40
Fig. 9.9 End Column Moment at Backfill Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 1)
30
L = 10 ft
b M
25 en1
φ = 30o M
en3
20 M
S /S = 1 en6
c b
z, ft
M
15 en10
M
en15
10 M
en20
0
-20 -10 0 10 20 30 40
Fig. 9.10 End Column Moment at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 1)
Ratio, Sc/Sb, of 4
in Fig. 9.11 and 9.12 for the initial backfill stage and the last expansion cycle
respectively.
30
25
U
20 hn1
U
z, ft
hn3
15 U
L = 10 ft hn6
b
U
10 φ = 30o hn10
U
hn15
5 S /S = 4
c b U
hn20
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2
h
Fig. 9.11 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 4)
30
25
U
20 hen1
U
z, ft
hen3
15 U
L = 10 ft hen6
b
U
10 φ = 30o hen10
U
hen15
5 S /S = 4
c b U
hen20
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2
h
Sc/Sb = 4)
The retaining wall horizontal movements for this group of frames with column
to beam stiffness ratio of 4 are substantially smaller than their counterparts for the
frames of lesser stiffness presented latterly. The 10, 15 and 20 bay frame are found
to displace into the soil mass during the last expansion cycle, while the horizontal
movements of the single, 3 and 6-bay frames are found to vary only slightly
between the initial backfill stage and the last expansion cycle. This behavior has
also been found for the single and two story RFERS with column to beam stiffness
ratios of 4. (Chapter 8 and Appendix B)
last expansion cycle is shown in Fig. 9.13. The pressure magnitude and
distribution in the soil mass retained by the single bay, and the 3 and 6-bay
frames, is found to be comparable to the Coulomb’s active earth pressure, and the
soil loads prescribed by the national American building codes.
30
σ σ
hen1 hen20
25
σ Coulomb
hen3
20 σ ASCE 7-98
hen6
σ BOCA, SBC, IBC
z, ft
hen10
15
σ K0 L = 10 ft
hen15 b
10
φ = 30o
5 S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. 9.13 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 4)
The pressure is subsequently found to be larger behind the remainder of the

RFERS, increasing in magnitude with the increase in frame length and number of
bays. The 10-bay frame is found to retain a pressure slightly smaller than the earth
pressure at rest, while the pressure behind the 15-bay frame is found to be slightly
larger. The magnitude of the lateral earth pressure developed behind the 20-bay
frame, on the other hand, is more than twice the Coulomb’s active earth pressure,
and nearly 30% higher than the lateral earth pressure at rest. The pressures behind
these 3 frames are also larger than their counterparts presented latterly.
stage and the last expansion cycle are shown in Fig. 9.14 and 9.15, respectively.
30
L = 10 ft M
b
bn1
25
φ = 30o M
bn3
20 M
S /S = 4 bn6
c b
M
z, ft
bn10
15
M
bn15
10 M
bn20
0
-40 -30 -20 -10 0 10 20 30 40
30
L = 10 ft M
b en1
25
φ = 30 o
M
en3
20 M
S /S = 4 en6
c b
M
z, ft
en10
15
M
en15
10 M
en20
0
-40 -30 -20 -10 0 10 20 30 40
Fig. 9.15 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 4)
The magnitude of the bending moment in the retaining wall portion of the
single, 3 and 6-bay frames are found to vary slightly between the two stages of
analysis, with the larger variation obtained for the frames with the larger number
of bays. The bending moments at the first frames level of the 10, 15 and 20-bay
frames, on the other hand, are found to increase substantially at the end of the last
expansion cycle, compared with their counterparts at the initial backfill stage.
The end column horizontal movements are presented in Fig. 9.16 and 9.17 for
the initial backfill stage and the last expansion cycle, respectively.
30
25
U
20 hn1
U
z, ft
hn3
15 U
L = 10 ft hn6
b
U
10 φ = 30 o
hn10
U
hn15
5 S /S = 4
c b U
hn20
0
0 0.05 0.1 0.15 0.2
h
Fig. 9.16 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 4)
30
25
U
20 hen1
U
z, ft
hen3
15 U
L = 10 ft hen6
b
U
10 φ = 30 o
hen10
U
hen15
5 S /S = 4
c b U
hen20
0
0 0.05 0.1 0.15 0.2
h
Fig. 9.17 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 4)
The horizontal displacements of all end columns are found to increase at the
end of the last expansion cycle. The magnitude of said increases vary with the
frame length or number of bays, with the largest change found for the 20-bay
frame to be nearly 8 times larger at the end of the expansion cycle compared to the
initial backfill stage. The end column horizontal displacement for the 20-bay
frame at the end of the expansion cycle is even larger than the 6-bay frame.
The end column bending moment diagrams developed at the end of the initial
backfill stage and the last expansion cycle are shown in Fig. 9.18 and 9.19,
respectively. Said figures indicate that the magnitude of the bending increases
during the last expansion cycle for all RFERS presented. This increase is found to
be small for the single and 3 bay frame, and substantial for the 15 and 20-bay
frames. The bending moment at the first framed level in the end column of the 20-
bay frame, for instance, is found to increase nearly 900% during the last expansion
cycle.
30
L = 10 ft M
bn1
25 b
M
φ = 30 o bn3
20 M
bn6
S /S = 4 M
z, ft
c b bn10
15
M
bn15
10 M
bn20
0
-20 0 20 40 60 80 100
30
L = 10 ft M
en1
25 b
M
φ = 30 o
en3
20 M
en6
S /S = 4
z, ft
c b M
15 en10
M
en15
10 M
en20
0
-20 0 20 40 60 80 100
Sc/Sb = 4)
Ratio, Sc/Sb, of 1
in Fig. 9.20 and 9.21 for initial backfill stage and the last expansion cycle
respectively.
30
25
20
z, ft
15
L = 20 ft
b
U
10 hn1
φ = 30o U
hn3
5 S /S = 1 U
c b hn10
0
0 0.2 0.4 0.6 0.8 1
Horizontal Retaining Wall Movement, Uh, ft
Fig. 9.20 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb = 20 ft,
Sc/Sb = 1)
30
25
20
z, ft
15
L = 20 ft
b
U
10 hen1
φ = 30o U
hen3
5 S /S = 1 U
c b hen10
0
0 0.2 0.4 0.6 0.8 1
Sc/Sb = 1)
30
σ
hen1
25
σ
hen3
20 σ
hen10
z, ft
15 Coulomb
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
10
φ = 30o K0
5 S /S = 1
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
Fig. 9.22 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb = 20 ft,
Sc/Sb = 1)
The retaining wall horizontal movements for this group of frames are larger
than the three story RFERS presented earlier, and the effect of temperature
movements are relatively less pronounced, given the presence of soil restraint.
The lateral earth pressures developed at the end of the last expansion cycle
behind the three RFERS presented in this section are shown in Fig. 9.22. The
magnitude and distribution of the pressure found behind the single-bay frame is
in-line with Coulomb’s active earth pressure, and the soil loads prescribed by
three of the major building codes adopted in the United States.
The lateral earth pressure magnitude behind the 3 and 10-bay frames, on the
hand, are found to be slightly larger, and comparable with soil pressure loads
prescribed by ASCE 7-99.
The bending moment diagrams for the retaining wall portion of the frames are
shown in Fig. 9.23 and 9.24, for the initial backfill stage and end of the last
expansion cycle, respectively.
30
M
bn1
25
M
bn3
20 M
bn10
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 1
c b
0
-80 -60 -40 -20 0 20 40
Fig. 9.23 Retaining Wall Moment at Backfill Stage (φ = 30º, All Bays, Lb = 20 ft, Sc/Sb = 1)
30
M
en1
25
M
en3
20 M
en10
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 1
c b
0
-80 -60 -40 -20 0 20 40
Fig. 9.24 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb = 20 ft,
Sc/Sb = 1)
The magnitude of the bending moments developed at the end of the last
expansion cycle are found to slightly increase from their counterparts found at the
initial backfill stage. This increase is more pronounced, however, for the 10-bay
frame than for the remaining two RFERS. The bending moment distributions and
magnitudes are also comparable to those found for the three story frames shown in
Fig. 9.5 and 9.6.
The end column horizontal movements are shown in Fig. 9.25 and 9.26. The
single bay frame column is found to have nearly the same final horizontal
deflection at the end of both the initial backfill stage and the last expansion cycle.
30
25
20
z, ft
15
L = 20 ft
b
10 U
φ = 30o hn3
U
hn10
5 S /S = 4
c b
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2
h
Fig. 9.25 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb = 20 ft,
Sc/Sb = 1)
30
25
20
z, ft
15
L = 20 ft
b
U
10 hen1
φ = 30o U
hen3
5 S /S = 1 U
c b hen10
0
0 0.2 0.4 0.6 0.8 1
Fig. 9.26 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb = 20 ft,
Sc/Sb = 1)
The horizontal deflection of the end column of the 3-bay frame, on the other
hand, is found to increase by nearly 10% due to thermal expansion movements,
while the horizontal deflection of its counterpart of the 10-bay frame is found to
nearly double at the end of the last expansion cycle. By comparing the horizontal
movements of the retaining wall portion of the three RFERS shown in Fig. 9.20
and 9.21 to the end column movements, it is apparent that the entire potential
thermal expansion strains of all three frames occur at the free end, in the direction
away from the soil mass.
The bending moment diagrams of the end column at the initial backfill stage
and the last expansion cycle are shown in Fig. 9.27 and 9.28, respectively.
The magnitude of the bending moments is found to increase at the end of the
last expansion cycle, particularly at the first and second frame levels. The relative
increase in moment magnitude is found to be largest for 10-bay frame, and
smallest for the single bay frame.
30
M
bn1
25
M
bn3
20 M
bn10
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 1
c b
0
-20 -10 0 10 20 30
30
M
en1
25
M
en3
20 M
en10
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 1
c b
0
-20 -10 0 10 20 30
Sc/Sb = 1)
Ratio, Sc/Sb, of 4
respectively. Only the 3 and 10-bay frames are presented given that the single bay
analysis could not be completed due to failure of the soil mass during the
numerical simulation.
Sc/Sb = 4)
30
25
20
z, ft
15
L = 20 ft
b
10 U
φ = 30o hen3
U
hen10
5 S /S = 4
c b
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2
h
Sc/Sb = 4)
The horizontal movements of the retaining wall for this group of frames are
substantially smaller than their counterparts presented latterly for the lesser
stiffness RFERS. Additionally, the 10-bay frame is found to undergo thermal
expansion movement at the end of the last expansion in the direction of the soil
mass and into the retained soil.
The lateral earth pressure developed behind the two RFERS at the end of the
last expansion cycle is shown in Fig. 9.31. The pressure shown in said figure for
the single bay frame is that developed at the end of the first expansion cycle, or
the second stage of analysis before the collapse of the soil mass. The magnitude
and distribution of the pressure are comparable to those obtained using Coulomb’s
active earth pressure theory, and the soil loads prescribed BOCA, the SBC and the
IBC.
30
σ
hen1
25
σ
hen3
20 σ
hen10
z, ft
Coulomb
15
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
10
φ = 30o K0
5 S /S = 4
c b
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
Sc/Sb = 4)
The lateral earth pressure diagrams for the 3 and 10-bay frames are those found
at the end of the last expansion cycle. The magnitude of the pressure developed
behind the 3-bay RFERS is similar to the load stipulated by ASCE 7-98, while the
pressure found behind the 10-bay frame is more than 20% larger than the lateral
The bending moment diagrams for the retaining wall portion of the 3 and 10-
bay RFERS are shown in Fig. 9.32 and 9.45 for the initial backfill stage and the
last expansion cycle, respectively. A substantial increase in the moment magnitude
is found for both frames due to thermal movements.
The bending moment at the first frame level of the 3-bay RFERS is shown to
increase nearly 800% at the end of the last expansion cycle, while its counterpart
for the 10-bay frame was nearly 20 times larger due to thermal expansion.
30
M
25 bn3
M
bn10
20
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 4
c b
0
-20 -10 0 10 20 30
30
M
25 en3
M
en10
20
z, ft
15
L = 20 ft
b
10 φ = 30o
5 S /S = 4
c b
0
-20 -10 0 10 20 30
Sc/Sb = 4)
The end column horizontal movements are shown in Fig. 9.34 and 9.35. The
horizontal deflection of the end column of the 3-bay frame at the end of the last
expansion cycle is shown to be nearly 10% larger than its counterpart found at the
end of the initial backfill stage, while the deflection of the end column of the 10-
bay frame is found to be approximately 4 times larger, and even comparable to the
deflection of the end column of the 3-bay frame.
These results denote a complexity in the prediction of the lateral deflection of
the free end of the RFERS due to thermal movements, given that two frames with
substantially different lateral stiffness and expansion length were to assume nearly
the same laterally deflected position.
30
25
20
z, ft
15
L = 20 ft
b
U
10 hn3
φ = 30o U
hn10
5 S /S = 4
c b
0
0 0.05 0.1 0.15 0.2
h
Sc/Sb = 4)
30
25
20
z, ft
15
L = 20 ft
b
U
10 hen3
φ = 30o U
hen10
5 S /S = 4
c b
0
0 0.05 0.1 0.15 0.2
h
Sc/Sb = 4)
The end-column bending moment diagrams found at the end of the initial
backfill stage and the last expansion cycle are shown in Fig. 9.36 and 9.37. The
magnitude of the maximum bending moment developed in the end column of the
3-bay frame is found to be nearly 75% larger at the end of the last expansion cycle
compared to the initial backfill stage. On the other hand, the maximum moment in
the end column of the 10-bay frame is found to be more than 50 times larger at the
end of the last expansion cycle.
30
M L = 20 ft
b
bn3
25
M
bn10
φ = 30o
20
S /S = 4
c b
z, ft
15
10
0
-20 -10 0 10 20 30 40 50 60
30
M L = 20 ft
b
en3
25
M φ = 30o
en10
20
S /S = 4
c b
z, ft
15
10
0
-20 -10 0 10 20 30 40 50 60
Sc/Sb = 4)
9.2.2 Five Story Rigidly Framed Earth Retaining Structures

The analysis results of five story rigid frames supporting backfill Soil with 30º
internal friction angle are presented herein, similar to their counterparts presented
earlier.
Ratio, Sc/Sb, of 1
The horizontal movements of the retaining wall portion of the five story RFERS
respectively.
The horizontal wall-deflections for all six frames are found to be comparable at
the end of both the initial backfill stage and the last expansion cycle, indicating
that the horizontal frame-movements at the wall are not significantly affected by
temperature changes. Nevertheless, the horizontal deflections of the frames are
significantly larger than would be practically acceptable for structures with
serviceability concerns.
50
40
U
hn1
30 U
z, ft
hn3
L = 10 ft U
b hn6
20
U
φ = 30o hn10
U
10 hn15
S /S = 1
c b U
hn20
0
0 0.5 1 1.5 2 2.5 3
Fig. 9.38 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb
= 1)
50
40
U
hen1
30 U
z, ft
hen3
L = 10 ft U
b hen6
20
U
φ = 30 o
hen10
U
10 hen15
S /S = 1
c b U
hen20
0
0 0.5 1 1.5 2 2.5 3
Sc/Sb = 1)
The lateral earth pressures developed behind the rigid frames at the end of the
last expansion cycle are shown in Fig. 9.40. The magnitudes of said pressures are
found to vary between the lower bound pressure distribution determined in
accordance with Coulomb’s active earth pressure theory, and the upper bound soil
pressure stipulated in ASCE 7-98.
Note here, that the results presented for the single and 3-bay frames on this
group of RFERS are based on a numerical analysis carried to the first expansion
cycle only, since failure of the soil mass occurs subsequently.
50
σ σ
hen1 hen20
40 σ Coulomb
hen3
σ ASCE 7-98
hen6
30
σ BOCA, SBC, IBC
z, ft
hen10
L = 10 ft σ K0
20 b hen15
φ = 30o
10
S /S = 1
c b
0
0 0.5 1 1.5 2 2.5 3 3.5
σ , kips/ft
2
h
Sc/Sb = 1)
50
M M
bn1 bn10
40 M M
bn3 bn15
M M
bn6 bn20
30
z, ft
L = 10 ft
20 b
φ = 30o
10
S /S = 1
c b
0
-300 -240 -180 -120 -60 0 60 120
50
M M
en1 en10
40 M M
en3 en15
M M
en6 en20
30
z, ft
L = 10 ft
20 b
φ = 30o
10
S /S = 1
c b
0
-300 -240 -180 -120 -60 0 60 120
Sc/Sb = 4)
The retaining wall bending moment diagrams are presented in Fig. 9.41 and
9.42 for the initial backfill stage and the last expansion cycle, respectively. The
wall moments are found to be comparable during both analysis stages, indicating
that temperature expansion movements have negligible effects on the restrained
end of the RFERS in this group. This is due to the presence of the retained soil
mass providing sufficient restraint to minimize the amount of expansion
movements undergone by the frame at the retaining wall end. The balance of
thermal expansion strains will therefore be manifested through additional
horizontal movements at the free end of the frame as shown subsequently.
50
40
U
hn1
30 U
z, ft
hn3
L = 10 ft U
b hn6
20
U
φ = 30 o
hn10
U
10 hn15
S /S = 1
c b U
hn20
0
0 0.5 1 1.5 2 2.5 3
Sc/Sb = 4)
50
40
U
hen1
30 U
z, ft
hen3
L = 10 ft U
b hen6
20
U
φ = 30o hen10
U
10 hen15
S /S = 1
c b U
hen20
0
0 0.5 1 1.5 2 2.5 3
Sc/Sb = 4)
The end column horizontal movements at the end of the initial backfill stage
and the last expansion cycle are shown correspondingly in Fig. 9.43 and 9.44. The
effect of temperature change on the end column horizontal movements for the
single, 3 and 6-bay frames are minor, but are more pronounced for the larger
RFERS. The end column of the 15-bay frame, for instance, is found to undergo
approximately a 21% increase in horizontal movements at the end of the last
expansion cycle, compared with the movements found at the end of the initial
backfill stage.
The end column bending moment diagrams are presented in Fig. 9.45 and 9.46
for the initial backfill stage and the last expansion cycle, respectively.
Temperature effects on the flexural stresses developed in the end column of the
single, 3 and 6-bay frames are again found minor. The magnitude of the bending
moment at the first frame level for the 10 though 20-bay frames increased nearly
20% to 60% respectively.
50
M
bn1
40 M
bn3
M
30 bn6
z, ft
M
bn10
L = 10 ft M
20 b bn15
φ = 30o M
bn20
10
S /S = 1
c b
0
-100 -75 -50 -25 0 25 50 75 100
50
M
en1
40 M
en3
M
30 en6
z, ft
M
en10
L = 10 ft M
20 b en15
φ = 30o M
en20
10
S /S = 1
c b
0
-100 -75 -50 -25 0 25 50 75 100
Sc/Sb = 1)
Ratio, Sc/Sb, of 4
The horizontal movements of the retaining wall portion of the five story RFERS
respectively.
50
40
U
hn1
30 U
z, ft
hn3
L = 10 ft U
b hn6
20
U
φ = 30o hn10
U
10 hn15
S /S = 4
c b U
hn20
0
-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
h
Sc/Sb = 4)
50
40
U
hen1
30 U
z, ft
hen3
L = 10 ft U
b hen6
20
U
φ = 30o hen10
U
10 hen15
S /S = 4
c b U
hen20
0
-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
h
Sc/Sb = 4)
The wall movements are found to be substantially smaller than their

counterparts presented earlier for the frames with column to beam stiffness ratio of
1. Additionally, the magnitudes of the horizontal movements are within acceptable
practical limits for all frames except the single and 3-bay RFERS.
The effect of the temperature at the last expansion cycle is found to be
relatively significant, where the retaining walls of the 15 and 20-bay frames are
shown to undergo horizontal movements into the retained soil mass. The wall of
the 20-bay frame is additionally found to move beyond its original position in the
direction of the soil.
The lateral earth pressures developed at the end of the last expansion cycle for
this group of frames are shown in Fig. 9.49. The pressures found behind the
frames, except for the 15 and 20-bay frames, are found to vary between the active
pressure distribution determined in accordance with Coulomb’s theory, and the
soil load prescribed by ASCE 7-98.
50
σ σ
hen1 hen20
40 σ Coulomb
hen3
σ ASCE 7-98
hen6
30
σ
z, ft
BOCA, SBC, IBC

hen10
L = 10 ft σ K0
20 b hen15
φ = 30 o
10
S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3 3.5
σ , kips/ft
2
h
Sc/Sb = 4)
The magnitudes of the pressures developed behind the 15 and 20-bay frames
were found to be yet larger, nearly approaching the lateral earth pressure at rest at
the top of the structure.
The bending moment diagrams developed in the retaining walls corresponding
to the wall movements and lateral earth pressures presented latterly are shown in
Fig. 9.50 and 9.51 for the initial backfill stage and the last expansion cycle,
respectively.
The bending moment magnitudes for the wall of the single, 3 and 6-bay frames
are found to be comparable at both analysis stages. The moments for the walls of
the 15 and 20-bay frames, in contrast, are found to increase by more 2.5 times in
magnitude due to the expansion movements. This amount of increase is too large
to be neglected in the analysis or design of the walls.
50
M
bn1
40 M
bn3
M
bn6
30
M
z, ft
bn10
M L = 10 ft
20 bn15 b
M
bn20 φ = 30o
10
S /S = 4
c b
0
-120 -100 -80 -60 -40 -20 0 20 40 60
50
M
en1
40 M
en3
M
en6
30
M
z, ft
en10
M L = 10 ft
20 en15 b
M
en20 φ = 30o
10
S /S = 4
c b
0
-120 -100 -80 -60 -40 -20 0 20 40 60
Sc/Sb = 4)
50
40
U
hn1
30 U
z, ft
hn3
L = 10 ft U
b hn6
20
U
φ = 30o hn10
U
10 hn15
S /S = 4
c b U
hn20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Sc/Sb = 4)
50
40
U
hen1
30 U
z, ft
hen3
L = 10 ft U
b hen6
20
U
φ = 30o hen10
U
10 hen15
S /S = 4
c b U
hen20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
h
Sc/Sb = 4)
The horizontal movements of the end column at the free end of the frames are
shown in Fig. 9.52 and 9.53, corresponding to the results obtained at the end of
the initial backfill stage and the last expansion cycle.
The end columns of the single, 3 and 6-bay frames are shown to undergo little
change during the expansion cycle, while the end column of the 15-bay frame, in
contrast, is shown to undergo additional horizontal movements during the
expansion cycle equal to more than twice their counterparts found at the initial
backfill cycle. The end column of the 20-bay frame is also shown to undergo
substantial change in horizontal movements during the expansion cycle attaining
nearly 3 times the movements developed during the initial backfill stage.
Furthermore, the laterally deflected shapes of the end column of the 10, 15 and 20-
bay frames are found to be nearly identical at the end of the last expansion cycle.
50
L = 10 ft M
b bn1
40 φ = 30 o M
bn3
M
bn6
30 S /S = 4
c b M
z, ft
bn10
M
20 bn15
M
bn20
10
0
-40 0 40 80 120 160 200 240
50
L = 10 ft M
b en1
40 M
φ = 30 o
en3
M
en6
30 S /S = 4
c b M
z, ft
en10
M
20 en15
M
en20
10
0
-40 0 40 80 120 160 200 240
Sc/Sb = 1)
The end column bending moments corresponding to the horizontal movements

presented above are shown in Fig. 9.54 and 9.55 for the initial backfill stage and
the last expansion cycle, respectively. The bending moment magnitudes at the first
frame level for all the frames, except the single bay RFERS, are found to undergo
relatively large increases in magnitude during the last expansion cycle. The
moment magnitude in the end column of the 3-bay frame at the first frame level,
for instance, is shown to increase by nearly 15% at the end of the last expansion
cycle from the magnitude developed at the initial backfill stage.
The magnitude of the bending moment in the end column of the 15 bay-frame,
for instance, found at the end of the last expansion cycle at the first frame level is
nearly 3 times larger than its counterpart developed at the initial backfill stage.
9.3 Conclusions 219
9.3 Conclusions
The soil-structure interaction problem involving thermal movements of multistory

RFERS was examined through a numerical parametric analysis of structures with
varying geometries and properties. Similar to single story frames, it is clear that
the longer the frame the larger the earth pressure magnitude developed behind it.
Furthermore, frames of comparable length and frames with higher stiffness
developed larger earth pressures.
By comparing the lateral displacement and corresponding earth pressures as
well as structural stresses developed at the backfill stage in the absence of
temperature changes, and those after the application of thermal movements, it can
be concluded that structural frames analyzed and designed without the inclusion of
the soil-structure interaction effects due to temperature changes will likely develop
undesirable behavior while in service.
A large number of parametric studies for two and four story structures is also
included in Appendix B in order to aid with identifying the effect of thermal soil
structure interaction on structures, which might be of interest to the reader.
Chapter 10
Conclusions and Recommendations
Abstract. This chapter summarizes the methodology and typical results of the (1)
experimental investigations, (2) numerical analyses, and (3) parametric studies
conducted to investigate the thermal soil structure interaction of Rigidly Framed
Earth Retaining Structures (RFERS). Recommendations for the design of such
structures are also presented.
Comparison of the lateral displacement and corresponding earth pressures as
well as structural stresses developed at the backfill stage in the absence of
temperature changes, and those after the application of thermal movements,
suggests that structural frames designed without the inclusion of the soil-structure
interaction effects due to temperature changes will likely undergo undesirable
response in service.
Preliminary recommendations regarding the analysis and design of RFERS at
back fill stage and due to exposure to thermal cycles are also made in this section.
10.1 Summary and Conclusions
The general behavior of structures that retain earth and derive their lateral force
resistance from rigid frame action (RFERS) was analyzed and presented in this
manuscript. Of particular interest are structures that are subject to a relatively
large variation in temperature and the corresponding lateral thermal movements.
Although the mechanisms of analysis and design of rigidly framed structures
are well documented in the literature and in building code, the behavior of such
structures retaining soils received little or no attention. Of particular interest in the
behavior of RFERS are (1) the magnitude of initial lateral earth pressure exerted
by the backfill, and (2) the subsequent effects of soil-structure interaction due to
temperature expansion and contraction cycles on the developed earth pressure as
well as the strength and serviceability requirements of RFERS.
10.1.1 Instrumentation and Monitoring of in Service RFERS

A hillside four-story reinforced concrete car parking structure retaining earth over
its entire height on one side and exhibiting significant signs of structural distress

222 10 Conclusions and Recommendations
was instrumented and monitored for a period of nearly four and half years starting
in May, 1999, and ending in October, 2003, in Chapter 4. The building’s lateral
force resisting system consisted primarily of rigid frames in the direction of earth
pressure. The instrumentation program included the use of four pluck-type,
vibrating-wire displacement transducers (VW) mounted parallel to the north-south
direction, the earth pressure direction, on two levels via expansion joints between
said building and an adjacent structure of similar construction not subjected to
earth pressure. Four additional VW’s were installed normal to the expansion joints
adjacent to the aforementioned sensors to monitor the relative building movements
in the direction perpendicular to the earth pressure.
The changes in inclination of the building wall retaining the soil backfill were
also monitored at twelve locations using three electrolytic tiltmeters installed at
nearly one-quarter points along the length of the wall on all four building levels.
All the instruments were equipped with temperature sensors, and were
connected to two relay multiplexers linked to a datalogger with non-volatile
memory used to scan the sensors every 15 seconds and store the displacement or
inclination and temperature every minute after the hour. Communication with the
datalogger was initiated through a personal computer equipped with support
software used for automated data retrieval from the storage modules and for
editing the datalogger program.
From the study of the measurements of electrolytic tiltmeters and vibrating-
wire displacement transducers presented chapter 4, it is clear that the rigidly-
framed structure undergoes a complex soil-structure interaction induced by
volumetric strains resulting from large temperature variations.
During the period of rise in temperature, the structure undergoes limited
expansion movements into the soil mass at the restrained end, causing larger
expansion movements, and stresses, at the other end. The movements of the
structure toward the retained soil induce an increase in earth pressure, and
possibly in soil stiffness, causing the rigid-frames to deflect in the direction away
from the soil mass to maintain the required force equilibrium, while still
undergoing thermal expansion movements. This behavior results in a nonlinear
interaction between the structure and the soil affected by several factors including,
but possibly not limited to, the soil stiffness characteristics, the lateral earth
pressure-displacement relationship, the lateral stiffness and volumetric-strains
characteristics of the structure.
Furthermore, through the period of drop in temperature, the structure undergoes
asymmetrical contraction movements at its ends, and a movement of the soil into
the gap formed between the soil mass and the contracted form of the retaining
structure. This soil movement prevents the structure from reverting to its position
before contraction at the next expansion cycle, causing a cumulative lateral
movement of the structure away from the soil over several temperature cycles.
This continuous movement induced large strains that resulted in stresses that
caused severe structural distress in the building elements, and the failure of one
column at the topmost level.
10.1 Summary and Conclusions 223
10.1.2 Numerical Analysis of In-Service Structure

A 2-dimensional plane-strain finite element analysis of a longitudinal rigid frame
located approximately at the center of the structure was performed in Chapter 6.
The frame has a tributary width for soil load of nearly 10.5 m. The finite element
model included the retaining wall of the structure, and the nine columns and slab
elements composing the rest of the rigid frame. The actual geometry of the rigid
frame and the section properties of its elements were obtained from the original
structural drawings of the building and verified through field measurements.
The backfill soil was modeled using an elastic-plastic Mohr Coulomb model
and a Hardening Soil model that accounts for stress-dependency of stiffness
modulus of the backfill soil. A subsurface soil investigation was conducted in the
field indicating that the top 20 to 28 feet of backfill soil consist of shot rock
intermixed with brown medium to fine sand with varying amount of silt and
gravel, where large boulders were common. An approximately 10-ft deep layer of
medium to fine dense sand was found to underlay the top layer, and overlay a
sound rock stratum.
The numerical analysis of the full-scale structure was composed of three parts.
The first part consisted of a plane strain analysis of the selected frame temperature
loading without the presence of the soil backfill on either side of the intramural
structure. This analysis was used to compare the thermal strains developed in the
frame and the relating member stresses to the same structure retaining the backfill
soil. The first temperature cycle consisted of simulating an increase in temperature
of 50ºF and the second and subsequent temperature loading cycles simulate an
alternating decrease and increase in temperature of 90ºF. The second part included
the frame with the backfill soil idealized using the elastic-plastic Mohr-Coulomb
constitutive model. The retained soil was added at the initial stage of the analysis
using a staged construction simulation, followed by the application of 50ºF
thermal expansion load, and several 90ºF contraction and expansion load-cycles
subsequently. The third analysis procedure repeated the second part described
above, with the exception that the backfill soil was modeled using the Hardening
Soil model.
Part 1 of the numerical analysis indicated that the horizontal thermal
movements of the frame were not equal at both extremities given the varying
lateral stiffness of the structural members. It was therefore found that the end
column would undergo lateral deflections over twice as large as the building wall
at the other end.
Figure 10.1 (repeated from 6.14), presents the maximum horizontal
displacement obtained for the retaining wall from the numerical analysis of the
rigid frame without backfill, with backfill soil simulated using the Mohr-Coulomb
soil model, and with soil backfill simulated using the Hardening Soil model.
Evidently, the maximum wall movements vary with each condition, with the
smallest movement occurring during temperature decrease (contraction) cycles for
the frame without any retained soil. A 240 percent larger wall movement was
found for the frame retaining Mohr-Coulomb backfill, and a 340 percent larger
movement was found for the frame retaining the Hardening Soil backfill.
Similar results could be found for the movements of the end column shown in
Fig. 6.14. It is therefore obvious that the effect of thermal movements on the
displacements of the rigid frame is far more pronounced for a structure restrained
by backfill soil then for a free structure.
The maximum retaining wall and end column moments from the previous three
analyses are presented in Fig. 6.10. The comparison reveals similar results for the
bending moments as for the displacement, where the moments developed in the
restrained frame structure are substantially larger than those found for the free
frame. However, while the bending moments developed in the retaining wall
differed quite notably with the two soil models for backfill, the corresponding
moments for the end column show close agreement for both backfill models.
Finally, the maximum lateral earth pressures found from the analysis using the
Mohr-Coulomb and Hardening Soil models are presented in Fig, 10.2 (reproduced
from Fig. 6.15). As shown in the figure, the maximum lateral earth pressure
exerted on the rigid frame developed during the thermal expansion cycles and is
36 Roof
U
h-No Backfill
Uh-MC Backfill Level D
27
Uh-HSS Backfill
z, ft
Level C
18
Level B
9
Level A
0
-0.05 0 0.05 0.1 0.15
Maximum Horizontal Retaining Wall Movement, U h, ft
Fig. 10.1 Maximum Horizontal Retaining Wall Movement of in Service Building
36 Roof
σ'MC-Backfill
27 σ' Level D
HSS-Backfill
σ'
Active
σ'
z, ft
Level C
18 At Rest
Level B
9
Level A
0
0 0.5 1 1.5 2 2.5 3
2
Max σ' , kips/ft
h
Fig. 10.2 Maximum Lateral Earth Pressure
considerably larger than the classical active earth pressure and the lateral earth
pressure at rest. The hardening soil model developed larger magnitudes for the
lateral earth pressure than did the Mohr-Coulomb’s model.
10.1.3 Approximate Expressions for Lateral Deflection of Frames

Closed form equations that can be used to determine the lateral deflections of
rigidly framed structures were derived, in Chapter 3, by treating the structure as a
cantilever and neglecting the bending deflection (Fig. 3.5). The derived formulas
were calibrated using 42,000 FEM analyses. The equations provide a simple and
accurate method to approximate the lateral deflections of a low-rise rigid structure
with height equal to or less than 20% of their length. Comparison of the
deflections computed using the equations and FEM reveals that there is an 80%
probability of the equations calculating a deflection that is within 25% of that
computed using FEM, even when the shape of the earth pressure is unknown.
Inspection of the equations illustrates that for any magnitude of lateral earth
pressure force, the shape of the earth pressure distribution is less important than
the geometric and material properties of RFERS when calculating the deflection.
As a result, the derived formulas provide simple and reliable method for
predicting the relationship between lateral displacement and earth pressure for
RFERS.
0.1 0.15
Level B Sensor Level B Sensor
North Side 0.1 South Side
0.05
0.05
0
0
ΔK
ΔK
-0.05 -0.05
-0.1
-0.1
y = -0.099945 + 0.0047072x -0.15 y = -0.094126 + 0.0049824x
2
R = 0.83625 R2= 0.76715
-0.15 -0.2
-10 0 10 20 30 40 -20 -10 0 10 20 30 40
Fig. 10.3 Change in Coefficient of Lateral earth Pressure Behind Monitored Structure at
Level-B
10.1.4 Relationship between Temperature and Earth Pressure

Temperature and displacement data for the monitored in-service RFERS was used
in Chapter 5 to obtain a relationship between the structure’s temperature and the
coefficient of lateral earth pressure, K, of the retained soil. K was found to be
linearly dependent on the building temperature; it changed by approximately
0.005/°C (Fig. 10.3) varying in the range of 1.25 to 1.5, depending on the season.
Although the numerical value of change in the earth pressure coefficient with
temperature change is valid for the monitored building only, the insights derived
from this case history are applicable to all rigidly framed earth retaining
structures, including jointless (segmental) bridges.
10.1.5 Numerical Analysis of Earth Pressure at Backfill Stage

Given that the state of stress in the retained soil mass is generally dependent on
the type of displacement and the shape of the displaced retaining structure
(Winterkorn and Fang, 1975), a parametric numerical study of RFERS retaining
cohesionless backfill was performed to qualitatively and quantitatively study the
mobilization of earth pressure behind this type of retaining structures.
Two-dimensional plane strain finite element analysis of rigidly framed earth
retaining structures of varying length, height, number of bays, number of levels,
and member stiffness were performed using the commercial finite element
analysis software Plaxis, in Chapter 7. The stages of construction were
incorporated in the analysis to simulate an initial stage where a structural frame is
first completed and later receiving backfill soil in several stages. The displacement
of the structures, as well as the earth pressure and resultant load developed behind
them were obtained to examine the relationship between the stiffness of the
retaining structure and the development of lateral earth pressure in the retained
soil mass.
In general, it was found that for rigid frames with relatively lower stiffness, the
magnitude and distribution of the lateral earth pressure developed behind the
frames at the end of the backfill stage is comparable to the lateral earth pressure
obtained from the Coulomb’s active earth pressure theory, and is also in line with
the soil loads stipulated by some of the national building codes adopted in the
United States, such as the International Building Code and the BOCA code. The
soil loads prescribed by ASCE-7 were found to slightly overestimate the pressure
developed behind the more flexible frames.
On the other hand, for the stiffer frames, the lateral earth pressure was found to
be reasonably comparable to the Coulomb’s active earth pressure for the top half
of the retained soil height, and increasing linearly thereafter to reach the
magnitude of the lateral earth pressure at rest at the bottom of the retained height.
This conclusion may not be extended however to encompass framed structures
braced against lateral sway such as shear wall or braced frame structures where
the lateral earth pressure was found to be even larger than that for rigid frames.
10.1.6 Thermal Parametric Analysis of Single Story RFERS

The soil-structure interaction problem involving thermal movements of RFERS
was further examined through a numerical parametric analysis of structures with
varying geometries and properties. The primary purpose of this analysis was to
investigate the effects of large temperature changes on the displacement of the

rigid frames, the stresses developed in the structural elements, and the lateral earth
pressure developed in retained the soil mass.
Two-dimensional plane strain finite element analysis of single-story rigidly
framed earth retaining structures of varying number of bays, bay lengths, and
member stiffness were performed using the commercial finite element analysis
software Plaxis, in Chapter 8 and Appendix A. The structural members composing
the rigid frames were modeled using an elastic material model, while the backfill
soil was modeled using a Mohr Coulomb model with varying parameters. The
individual structural frames were assumed to have beam stiffness and column
stiffness that are respectively equal throughout the structure. The structural wall
retaining the backfill was assumed to be of concrete construction. The structures
were assumed to have infinitely rigid foundation strata.
10
1 bay, L = 10 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 1 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
10
1 bay, L = 10 ft
b σ'
a
8 φ = 30o σ'
ec1
S /S = 4 σ
'cc1
6 c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 10.4 Retaining Wall Earth Pressure (Single Bay, Lb = 10 ft)
The analysis of single-story single-bay rigid frames indicated that the lateral
earth pressure developed behind the RFERS due to temperature variations is larger
when the lateral stiffness of RFERS is larger. Similar results were found for the
lateral displacement of the frames and the stresses developed in the structural
members.
Fig. 10.4 and 10.5 illustrate the lateral earth pressure at different analysis
stages, starting with the backfill stage (σ’a) and followed by thermal cycles of
expansion and contraction, σ’ec and σ’cc, respectively. The two frames with larger
stiffness (Sc/Sb = 4) were found to mobilize larger lateral earth pressures during
the expansion cycles compared to the initial backfill pressures and those during
the contraction cycles.
10
1 bay, L = 20 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 1 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 02 04 06 08 1 12
10
1 bay, L = 20 ft σ'
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 4 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. 10.5 Retaining Wall Earth Pressure (Single Bay, Lb = 20 ft)
These results were more pronounced with frames of even larger stiffness and
longer length of expansion and contraction, such as multi-bay frames. The lateral
earth pressure found during the expansion cycle of the three and ten-bay frames
shown in Fig. 10.6 were substantially larger than those of the single-bay frames.
The lateral stiffness of the two frames, determined from approximate methods,
was comparable, while the thermal properties of the frames, particularly the frame
length, was substantially larger for the 10-bay frame. Additionally, Fig. 10.6
clearly shows that the earth pressure at the backfill stage, σ’a, was comparable for
both the 6-bay and the 10-bay frames, while the lateral earth pressure during the
expansion cycle was substantially larger for the longer frame, and smaller during
the contraction cycle for the same frame.
The horizontal displacements of the rigid frames, as well as the moment and
shear diagrams of the retaining wall and end columns, all indicate similar results.
The RFERS undergo larger displacements, particularly at the free end, and
develop higher stresses, when the frames lateral stiffness and length increase.
10
3 bays, L = 20 ft σ'
b a
8
φ = 30o σ'
ec1
σ'
6 S /S = 4 cc1
z, ft
c b σ'
ec2
4 σ'
cc2
σ'
ec3
2
10
8
σ'
a
6 σ'
ec1
z, ft
σ'
cc1
4 10 bays, L = 20 ft
b
σ'
ec2
φ = 30o
σ'
cc2
2
S /S = 4 σ'
ec3
c b
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
2
σ' , kips/ft
h
Fig. 10.6 Retaining Wall Earth Pressure (Three and Ten Bay Frames, Lb = 20 ft)
10.1.7 Thermal Parametric Analysis of Multi-story RFERS

A numerical parametric analysis of multi-story RFERS was also performed in a
similar manner to single story frames presented earlier (Chapter 9 and Appendix
B). Fig. 10.7 below illustrates the variation in lateral earth pressure developed
behind the two-story frames’ retaining wall at the end of the expansion cycles. The
indices in the legend indicate the number for bays, and the coulomb, ASCE 7,
BOCA, SBC, IBC, and lateral earth pressure at rest, are all shown in the figure for
comparison.
Similar observations to the single story frames can be made from Figure 10.7,
where it is clear that the longer the frame the larger the earth pressure magnitude
developed behind it. Furthermore, frames of comparable length, but higher
stiffness developed larger earth pressures.
20
σ
hen1
σ
15 hen3
σ
hen6
σ
z, ft
hen10
10
L = 10 ft σ
b hen15
σ
φ = 30o hen20
5 Coulomb
ASCE 7-98
S /S = 1 BOCA, SBC, IBC
c b
K0
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
20
σ
hen1
σ
hen3
15
σ
hen6
σ
z, ft
hen10
10
L = 10 ft σ
b hen15
σ
φ = 30o hen20
5 Coulomb
ASCE 7-98
c b
K0
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
Fig. 10.7 Retaining Wall Earth Pressure for two-Story RFERS
For those frames with column to beam stiffness ratio of 1, with more than 6
bays in length, the earth pressure magnitude is found to exceed those prescribed
by the national building codes, while for the frames with the larger stiffness ratio,
this result is observed for even shorter frames. Furthermore, the longest two
frames with Sc/Sb of 1 were found to develop earth pressure magnitudes larger
than the lateral earth pressure at rest, while four of their stiffer counterparts show
the same result.
Similar qualitative results can be found for the lateral displacement of the two
story frames, as well as for the stresses developed in the structural elements. By
comparing the lateral displacement and corresponding earth pressures as well as
structural stresses developed at the backfill stage in the absence of temperature
changes, and those after the application of thermal movements, it can be
concluded that structural frames analyzed and designed without the inclusion of
the soil-structure interaction effects due to temperature changes will likely
undergo undesirable response in service.
10.2 Recommendations 231
The articulation of these results, however, decreases with the increase in the
number of stories or the height to length ratios of the rigid frames. They
nevertheless remain significant and a complete soil-structure interaction analysis
should be performed regardless.
10.2 Recommendations
The analysis of Rigidly Framed Earth Retaining Structures is demonstrated to

involve a soil-structure interaction at both, the initial construction (backfill) stage
and the subsequent in-service stage. Some preliminary recommendations
regarding the analysis and design of RFERS at the stages identified earlier are
made in this section. Further studies are required, however, to obtain a clearer
quantitative understanding of the soil-structure interaction problem.
10.2.1 Analysis of RFERS at the Initial Backfill Stage

The analysis of RFERS for retained soil loads requires an adequate estimate of the
magnitude and distribution of lateral earth pressures mobilized behind these
structures at the initial construction stage. For RFERS not subject to large
temperature variations, and for which cyclical thermal movements are relatively
negligible, this initial earth pressure represents the soil load that the structure must
be designed to support during its lifespan.
Several methods exist to help estimate the soil loads on retaining structures,
such as the popular classical methods introduced by Coulomb in the late 18th
century, and by Rankine in mid 19th century. Other estimates of lateral soil loads
are also prescribed in national building codes in the United States of America such
as the International Building Code, the BOCA code, the Southern Building Code,
the National Building Code, and the American Society of Civil Engineers,
“Minimum Design Loads for Buildings and Other Structures,” ASCE 7.
The results of the numerical parametric analysis of the lateral earth pressure
problem developed behind RFERS were presented in Chapter 7, and compared to
the provisions of the national building codes and the results obtained from
classical methods. It was generally found that the distribution of lateral earth
pressure behind RFERS is in agreement with the classical theories, but the
magnitude of said pressure was found to vary with the lateral stiffness of the
RFERS.
The authors therefore recommend a simple procedure for the determination of
lateral soil loads suitable for the preliminary analysis and design of RFERS in
professional engineering practice. The procedure makes use of the classical lateral
earth pressure theories and the provisions of U.S. national building codes, but
requires the determination of the lateral stiffness of rigid frames. Final design
should employ numerical analyses able to account for soil-structure interaction.
The lateral stiffness of RFERS can be estimated using approximate methods,

such as equation 7.3, suitable for manual computations, or using engineering
mechanics principles. This equation has been developed by rearranging the
equations developed in Chapter 3 for frames subject to triangular (hydrostatic)
loading. Computer aided analysis and design software is widely available in
engineering practice and can be used to determine the lateral stiffness of rigid
frames.
10.2.1.1 Recommended Procedure for Single Story RFERS
For single story RFERS, the recommended procedure to estimate the soil load can
therefore be stated as follows:
Classical Theories If KL<5000 Kip/ft/ft
Determine Soil Load from
ASCE-7 If KL≥5000 Kip/ft/ft
10.2.1.2 Recommended Procedure for Multi-story RFERS
As presented in Chapter 7, the shape and magnitude of the lateral earth pressure at
backfill stage developed behind multi-story RFERS ranged between the lower
bound determined using the classical theories as well as BOCA, SBC, and IBC,
and the upper bound determined using ASCE-7.
Given the aspect ratio of the frames (height/length), the section displacement of
the RFERS retaining walls generally follows a convex curve given that the lateral
deflection is governed by shear deformations. The ratio of the top-of-the-wall
deflection (δtw) over the wall height (hw), however, could be used to determine a
simple upper bound limit for which the ASCE 7 recommended earth pressure
should be used.
From the results of the parametric analysis of the retaining wall deformation
and corresponding lateral earth pressures developed at the backfill stage, we
recommend that a simple structural analysis of the RFERS be performed for the
ASCE 7 loading case, and subsequently determine the appropriate pressure as
follows:
For δtw/ hw < 1/1000 Use ASCE 7 recommendations
For δtw/ hw ≥ 1/1000 Use Classical Earth Pressure Theories

10.2 Recommendations 233
The authors do not recommend this simple method for structural braced frames
or frames with shear walls. The analysis of two such frames, presented in Chapter
7, indicated that the lateral earth pressure behind them can exceed the
recommended ASCE-7 loading.
10.2.2 Analysis of RFERS Subject to Temperature Variations

As shown throughout this manuscript, the response of RFERS subject to large
temperature variation is an important condition that requires investigation for the
proper design and detailing of this class of structures.
The magnitude of the lateral earth pressure developed during the expansion
cycles (rise in temperature) was found to vary greatly depending on the RFERS
geometry, overall aspect ratio, and structural stiffness. Earth pressure magnitudes
reaching those equivalent to the passive earth pressure, even higher, were found to
develop (theoretically) behind a number of frames.
Given the non-linear aspects of the cyclical behavior, and number of
parameters affecting the overall response of the soil-structure system, it is a
difficult task to determine simple procedures for the analysis of these structures, or
even analytical methods for the estimation of some of the parameters that are
important in the design of RFERS.
More research is therefore recommended in an effort to develop simplified
methods and procedure to aid in the analysis and design of RFERS subjected to
large temperature variations.
In the interim, the large number of parametric analyses presented in Chapters 8
and 9 as well as Appendices A and B, can serve as a preliminary guidance to
estimate the magnitude of earth pressure loads developed during expansion cycles,
and the corresponding deformations and stresses in the RFERS structural
members.
As a general rule, longer and stiffer frames tend to develop larger stresses in the
structural members as well as larger earth pressure, compared to the shorter and
more flexible counterparts.
Appendix A
Abstract. This Appendix extends on the results of numerical parametric analysis

of single-story structures presented in Chapter 8. The results are reported for 1, 10,
20 bay frames in Chapter 8 and for 3, 6, 15 bay frames in this Appendix for φ =
30°, along with results for φ = 40°. The primary purpose of this analysis is to in-
vestigate the effects of thermal movements of RFERS on (1) the displacement of
lateral earth pressure developed in the soil mass.
A.1 Thermal Soil Structure Interaction of Single Story RFERS
A.1.1 Backfill Soil with 30º Internal Friction Angle

The analysis results of single story rigid frames are presented under separate sec-
tions corresponding to the parameters varied in Table 8.1. The results reported for
each temperature cycle, in addition to the initial backfill stage, are (1) the dis-
placement of the retaining wall; (2) the lateral earth pressure developed in the re-
tained soil mass; (3) the retaining wall bending moment; (5) the displacement of
the end column; and (6) the end column bending moment.
A.1.1.1 Three-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam

The analysis results for a three-bay frame with a bay length of 10-ft and a column
to beam stiffness ratio of 1 are presented herein. This frame has 3 times the overall
expansion (contraction) length of its single 10-ft-bay counterparts, and a lateral
stiffness 225% larger than the single bay frame with a column to beam stiffness
ratio of 1, but smaller than the stiffer single bay frame.
Fig. A.1 presents the analysis results for the retaining wall portion of the rigid
frame. The horizontal movements of the retaining wall are nearly half those of the
single bay frame of similar properties, but with a wider range of variation during
the temperature change cycles. This of course is consistent with the results found
earlier for stiffer frames with longer expansion (contraction) length.
236 Appendix A
10
3 bays, L = 10 ft
b
8 φ = 30 o
U
ha
S /S = 1 U
6 c b hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
3 bays, L = 10 ft
b
8 φ = 30o M
a
6 S /S = 1 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
3 bays, L = 10 ft
b
8 φ = 30 o
Q
a
S /S = 1 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. A.1 Analysis Results for Retaining Wall (Three-Bay, Lb = 10 ft, Sc/Sb = 1)
The bending moment diagram also show a wider variation in the magnitude
with varying temperatures, as does the shear force diagram.
The lateral earth pressure developed behind the strucutre, shown in Fig. A.2,
indicates the largest variation in magnitude with change in temperature.
A.1 Thermal Soil Structure Interaction of Single Story RFERS 237
The horizontal movements of the end column shown in Fig. A.3 indicate that
the displacement of the top of the frame during the last expansion cycle is nearly
twice its counterpart produced during the initial backfill stage. The displacement
range of the end column due temperature variation is larger than the single bay
frames.
The bending moment diagram presented in Fig. A.4 show a relatively
small moment magnitude developed in the end column in comparison with single
bay frames analyzed. The moment developed at the end of the last expansion
cycle is approximately 50% larger than its counterpart found at the initial backfill
stage.
The shear force in the end column also is of a smaller magnitude than the single
bay frames presented earlier, with a magnitude range similar to the single bay
frame of smaller stiffness.
10
3 bays, L = 10 ft
b σ'
a
8 φ = 30o σ'
ec1
S /S = 1 σ'
cc1
6 c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. A.2 Retaining Wall Earth Pressure (Three-Bay, Lb = 10 ft, Sc/Sb = 1)
10
3 bays, L = 10 ft U
b
ha
8 φ = 30 o
hec1
U
U
6 S /S = 1 hcc1
c b
z, ft
U
hec2
U
4 hcc2
U
hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
Fig. A.3 End Column Horizontal Movements (Three-Bay, Lb = 10 ft, Sc/Sb = 1)

238 Appendix A
10
3 bays, L = 10 ft
b
8 φ = 30o
M
a
6 S /S = 1 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2
M
ec3
0
-2 -1 0 1 2 3 4 5
10
3 bays, L = 10 ft
b Q
a
8 φ = 30o Q
ec1
S /S = 1 Q
6 cc1
c b
z, ft
Q
ec2
4 Q
cc2
Q
ec3
2
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2
Fig. A.4 End Column Moment and Shear Diagrams (Three-Bay, Lb = 10 ft, Sc/Sb = 1)

to beam stiffness ratio of 4 are presented herein. This frame has 3 times the overall
expansion (contraction) length of its single 10-ft-bay counterparts, and a lateral
stiffness 275% larger than the single bay frame with a column to beam stiffness
ratio of 4. The lateral stiffness of this frame is also nearly 5 times larger than its 3-
bay counterpart presented earlier.
Fig. A.5 presents the analysis results for the retaining wall portion of the rigid
frame. The retaining wall horizontal movement at backfill stage is the smallest of
all frames presented thus far, and the range of movement between the last expan-
sion cycle and the backfill stage is the largest. Furthermore, the rigid frame is
shown to move into the original plane of the retained soil mass during expansion,
unlike the less stiff RFERS analyzed earlier.
10
3 bays, L = 10 ft
b
8 φ = 30o
U
ha
S /S = 4 U
6 c b hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
3 bays, L = 10 ft
b
8 φ = 30o
Q
a
6 S /S = 4 Q
c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
10
3 bays, L = 10 ft
b
8 φ = 30o M
a
6 S /S = 4 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
The magnitudes of the maximum shear force and bending moment are substantial-
ly larger than their counterparts presented formerly, with a significant range of varia-
tions with temperature. The retaining wall moment diagram indicate that the bending
moment at the top of the wall developed during the expansion cycles is nearly 3 times
240 Appendix A
larger than its counterpart produced at the initial backfill stage. Additionally, the wall
assumes a double curvature shape during expansion resulting in stress reversal for the
top 35% of the wall height, associated with large bending moments.
These results indicate that a reinforced concrete wall designed for the flexural
stresses produced from the application of the backfill soil will be inadequately re-
inforced to resist the stresses produced during the expansion cycles. The shear
force at the top of the wall is also substantially larger during the expansion cycles
compared with the initial backfill stage or even the contraction cycles.
The lateral earth pressure developed behind the rigid frame is presented in Fig.
A.6 for the various loading cycles. As shown in chapter 7 (Fig. 7.5), the lateral
pressure developed behind the frame at the initial backfill stage is equal to the
Coulomb’s active earth pressure for the top one-half of the wall height and ex-
ceeds Coulomb’s pressure for the bottom half. This justifies the magnitude of lat-
eral pressure developed during the contraction cycle, which is approximately
equal to the pressure found at the backfill stage for the top one-half of the struc-
ture, but is smaller than said pressure for the bottom one-half of the frame and
approaching the active earth pressure determined from Coulomb’s theory.
The lateral pressure developed during the expansion cycle behind the top one-half
of the structure’s height is substantially larger than the pressures developed during
the initial backfill stage and the subsequent contraction cycles. The pressure, howev-
er, decreases to a slightly larger magnitude than the initial backfill pressure for the
bottom half of the frame height. This pressure increase during the expansion cycles
is the largest found from the analyses presented hitherto, with a pressure magnitude
developed at expansion reaching twice its initially developed counterpart.
10
3 bays, L = 10 ft
b σ'
a
8 φ = 30o σ'
ec1
S /S = 4 σ'
cc1
6 c b
σ'
z, ft
ec2
σ
4 'cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
The end column horizontal movements shown in Fig. A.7 are slightly larger
than their counterparts for the wall at the opposite end of the frame. This of course
is due to the frame ability to expand into the soil at the wall end, as shown in Fig.
A.5, and thus reducing the amount of restrained expansion transferring to the end
column at the free end of the structure.
10
3 bays, L = 10 ft U
b
ha
8 φ = 30o U
hec1
U
6 S /S = 4 hcc1
c b
z, ft
U
hec2
U
4 hcc2
U
hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
M
a
8 M
ec1
M
cc1
6
M
z, ft
ec2
M 3 bays, L = 10 ft
4 cc2 b
M φ = 30o
ec3
2
S /S = 4
c b
0
-8 -6 -4 -2 0 2 4 6 8
10
3 bays, L = 10 ft
b Q
a
8 φ = 30 o
Q
ec1
S /S = 4 Q
6 c b cc1
z, ft
Q
ec2
4 Q
cc2
Q
ec3
2
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2
242 Appendix A
The end column bending moment and shear force diagrams are shown in Fig.
A.8. The magnitude and range of bending moment values developed in the end col-
umn at the various analysis stages are substantially larger than their counterparts
found for the remainder of the rigid frames presented yet. The shear force magni-
tude, on the other hand, is comparable to some of the frames presented earlier, but
the range of variation of shear force values with temperature is the largest thus far.

to beam stiffness ratio of 1 are presented herein. This frame has a substantially
smaller lateral stiffness than the three-bay frames presented earlier. The lateral
stiffness of this frame is only about 40% larger than the single 10-ft bay frame
with a column to beam stiffness ratio of 1 (Table 7.2).
The horizontal retaining wall movements shown in Fig. A.9 indicate that mag-
nitude of displacement at the top of the rigid frame are comparable to the lower
stiffness frames presented earlier, but a larger range of movements with varying
temperature cycles. Unlike the three-bay frame presented latterly, the rigid frame
presented herein is not displaced into the retained soil mass during the latter ex-
pansion cycles, indicating that additional horizontal displacement occurs due to
temperature movements. This displacement is nearly 20% larger than its counter-
part produced at the end of the initial backfill cycle, and should thus be accounted
for in the design of RFERS subject to lateral drift limitations.
10
U
ha
8 U
hec1
U
hcc1
6
z, ft
U
hec2
U 3 bays, L = 20 ft
b
4 hcc2
U φ = 30o
hec3
2 S /S = 1
c b
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
Fig. A.9 Horizontal Retaining Wall Movement (Three-Bay, Lb = 20 ft, Sc/Sb = 1)
The bending moment and shear force diagram for the wall portion of the rigid
frame are shown in Fig. A.10. The magnitude of the bending moment and shear
force are found to be smaller than the latterly presented frame, and more inline
with the lower stiffness three-bay frame presented earlier, with the exception of
the moment diagram showing an inflection point approximately two-third the
height of the wall from its bottom.
The lateral earth pressure developed behind the rigid frame, shown in Fig.
A.12, indicate a significant increase in lateral earth pressure in the retained soil
mass during the expansion cycles for the top one-half on the structure’s height.
This result is inline with that found for the frame presented latterly.
The horizontal movements of the end column are found to be notably larger
than nearly all the frames presented hitherto, particularly during the latter expan-
sion cycles. The range of movement of the end column is also large compared
with the other structures discussed formerly.
This can be attributed to the lower lateral stiffness of the frame, which results in
larger lateral displacement and a reduced ability of the structure to resist the
applied loads with relatively small displacements. A stiffer frame, such as the
latterly presented RFERS, is capable of expanding into the soil given its greater
ability to resist the applied forces while undergoing expansion movement due to
temperature change.
10
3 bays, L = 20 ft
b
8
φ = 30o M
a
6 S /S = 1 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
10
3 bays, L = 20 ft
b
8
φ = 30o
Q
a
6 S /S = 1 Q
c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. A.10 Retaining Wall Moment and Shear Diagrams (Three-Bay, Lb = 20 ft, Sc/Sb = 1)
244 Appendix A
10
U
ha
8 U
hec1
U
hcc1
6
z, ft
U
hec2
U 3 bays, L = 20 ft
4 hcc2 b
U
hec3 φ = 30o
2
S /S = 1
c b
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
b
a
8 φ = 30o σ'
ec1
σ'
6 S /S = 1 cc1
c b
σ'
z, ft
ec2
σ'
4 cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
The bending moment diagram of the end column, shown in Fig. A.13,
is similar to many of the frames presented previously with respect to the magni-
tude and range of bending moments developed during the various temperature
cycles.
10
M
a
8 M
ec1
M
cc1
6
z, ft
M
ec2
M 3 bays, L = 20 ft
4 cc2 b
M φ = 30o
ec3
2
S /S = 1
c b
0
-2 -1 0 1 2 3 4 5 6
10
Q
a
8 Q
ec1
Q
cc1
6
z, ft
Q
ec2
Q 3 bays, L = 20 ft
4 cc2 b
Q
ec3 φ = 30o
2
S /S = 1
c b
0
-1 -0.8 -0.6 -0.4 -0.2 0 0.2

to beam stiffness ratio of 4 are presented herein. This frame is 5 times stiifer than
its counterpart with Sc/Sb, of 1 (Table 7.2).
The horizontal movements of the retaining wall portion of the rigid frame,
shown in Fig. A.14, indicate a magnitude and range of displacement comparable
to the three 10-ft-bay frame with a column to beam stiffness ratio of 4.
The displacement at the top of the wall indicates small initial lateral movement,
and an expansion movement into the original position of the retained soil mass.
The bending moment and shear force diagram for the wall are also shown in
Fig. A.14. The magnitude of the moment at the top of the wall is the largest found
yet, and the magnitude of change in moment at the end of the expansion cycle
is nearly an 86% increase compared with the moment developed at the initial
backfill stage.
246 Appendix A
10
3 bays, L = 20 ft
b
8 φ = 30o U
ha
S /S = 4 U
6 c b hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
8
M
a
6 M
ec1
z, ft
M
3 bays, L = 20 ft cc1
4 b
M
ec2
φ = 30 o
M
cc2
2
S /S = 4 M
c b ec3
0
-4 -2 0 2 4 6 8
10
3 bays, L = 20 ft
b
8
φ = 30o Q
a
S /S = 4 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
The shear force diagram also show the largest range of shear developed in the
retaining wall yet, with the magnitude of the shear force developed during the last
expansion cycle reaching a value nearly 280% larger than its counterpart found at
the end of the backfill cycle.
The lateral earth pressure distribution behind the rigid frame is shown in
Fig. A.15. The effect of temperature movement on the magnitude of earth pressure
developed behind the structure is vastly evident in said figure, where the magni-
tude of earth pressure (for the top one half of the wall height) during the last
expansion cycle is nearly 5 times larger than the initial backfill pressure.
The analysis results for the end column are shown in Fig. A.16. The end col-
umn horizontal displacements are similar to their counterparts found earlier for the
three 10-ft bay frame with a column to beam stiffness ratio of 4.
The bending moment magnitudes and range of values for the end column
are the largest presented thus far, in this Appendix, as are the magnitudes and
variations in the shear force with temperature.
10
b a
8
φ = 30 o σ'
ec1
σ'
6 S /S = 4
cc1
z, ft
c b σ'
ec2
4 σ'
cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h

248 Appendix A
10
3 bays, L = 20 ft
b
8 φ = 30o U
ha
S /S = 4 U
6 c b
hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
8
M
a
6 M
z, ft
ec1
3 bays, L = 20 ft M
b cc1
4
M
φ = 30 o ec2
M
2 cc2
S /S = 4 M
c b
ec3
0
-2 0 2 4 6 8 10 12
10
3 bays, L = 20 ft
b Q
a
8 φ = 30o Q
ec1
S /S = 4 Q
6 c b cc1
z, ft
Q
ec2
4 Q
cc2
Q
ec3
2
0
-1.2 -1 -0.8 -0.6 -0.4 -0.2 0
Fig. A.16 Analysis Results for End Column (Three-Bay, Lb = 20 ft, Sc/Sb = 4)
A.1.1.5 Six-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam Stiffness
Ratio, Sc/Sb, of 1)
The analysis results for a six-bay frame with bay length of 10-ft and column to
beam stiffness ratio of 1 are presented herein. This frame is relatively stiff with KL
= 1455 kip/ft/ft (Table 7.2 Frame F3).
The retaining wall horizontal movements, shown in Fig. A.17, indicate a rela-
tively large initial displacement into the soil mass during the first expansion cycle.
The wall position assumed at the subsequent expansion cycles was nearly identical
to original wall position after initial backfill stage. The bending moment diagram,
also shown in Fig. A.17, indicates a significant difference in the moment magni-
tude developed during the final expansion cycles compared with the initial backfill
stage and the rest of the temperature cycles. The maximum moment found at the
backfill stage was nearly 50% its counterpart developed at the last expansion
cycle.
10
6 bays, L = 10 ft
b
8 φ = 30o U
ha
S /S = 1 U
6 c b hec1
z, ft
U
hcc1
4 U
hec2
U
hcc2
2 U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
h
10
6 bays, L = 10 ft
b
8 φ = 30 o
M
a
6 S /S = 1 M
c b ec1
z, ft
M
cc1
4 M
ec2
M
cc2
2 M
ec3
0
-8 -6 -4 -2 0 2 4 6
Fig. A.17 Analysis Results for Retaining Wall (Six-Bay, Lb = 10 ft, Sc/Sb = 1)
250 Appendix A
The retaining wall shear force diagram, presented in Fig. A.18, show the max-
imum shear force magnitude at top of the wall during last expansion cycle to be
approximately 240% larger than its counterpart found during the initial backfill
stage.
10
6 bays, L = 10 ft
b
8 φ = 30o Q
a
S /S = 1 Q
6 c b ec1
z, ft
Q
cc1
4 Q
ec2
Q
cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. A.18 Retaining Wall Shear Force (Six-Bay, Lb = 10 ft, Sc/Sb = 1)
The lateral earth pressure developed behind the rigid frame is shown in Fig. A.19.
The pressure magnitude developed during the latter expansion cycles is found sub-
stantially larger than its counterparts developed during the initial backfill and the
subsequent contraction cycles. The largest increase in lateral earth pressure is shown
to be distributed over nearly the top one-half of the retaining wall height.
10
6 bays, L = 10 ft
b σ'
a
8 φ = 30o σ'
ec1
S /S = 1 σ'
6 cc1
c b
z, ft
σ'
ec2
4 σ'
cc2
σ'
ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
σ' , kips/ft
2
h
Fig. A.19 Retaining Wall Earth Pressure (Six-Bay, Lb = 10 ft, Sc/Sb = 1)
The end column horizontal movements, shown in Fig. A.20, indicate a relative-
ly large range of movements related to temperature change. The horizontal dis-
placement of the end column during the latter expansion cycles is nearly 4 times
larger the displacement underwent during the initial backfill stage.
10
6 bays, L = 10 ft
b
8 φ = 30
o
Ma
S /S = 1
6 c b Mec1
z, ft
Mcc1
4 M
ec2
M
cc2
2
M
ec3
0
-2 -1 0 1 2 3 4 5
10
6 bays, L = 10 ft
b Q
a
φ = 30
8 o
Q
ec1
S /S = 1 Q
6 c b cc1
z, ft
Q
ec2
4 Q cc2
Q ec3
2
0
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3
Fig. A.20 End column horizontal Movements (Six-Bay, Lb = 10 ft, Sc/Sb = 1)
10
6 bays, L = 10 ft U
b
ha
8 φ = 30 o
U
hec1
U
6 S /S = 1 hcc1
c b
z, ft
U
hec2
U
4 hcc2
U
hec3
2
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
Fig. A.21 End Column Moment and Shear Diagrams (Six-Bay, Lb = 10 ft, Sc/Sb = 1)
252 Appendix A
The shear force and bending moment diagrams developed in the end column
during the various analysis cycles and corresponding to the horizontal end column
movements presented in Fig. A.20 are shown in Fig. A.21.
The bending moment is found to have a large variation in magnitude. At the lat-
ter expansion cycles, for instance, the maximum moment at the top of the end col-
umn is found nearly 233% larger than its counterpart developed at the end of the
backfill stage. Given that no external transverse load is applied onto the end col-
umn, the shear force developed in the end column of the rigid frame is equal to the
maximum moment divided by the total length of the end column. Therefore, the
variation in the magnitude of shear force is identical to the variation of the maxi-
mum bending moment.
Consequently, the shear force diagram will not be presented for the end column
for the subsequent rigid frames.
A.1.1.6 Six-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam Stiffness
Ratio, Sc/Sb, of 4)
The analysis results for a six-bay frame with a bay length of 10-ft and a column to
beam stiffness ratio of 4 are presented herein. This frame has the largest lateral
stiffness of all frames presented thus far as indicated in Table 7.2. In fact, the lat-
eral stiffness of the said frame is nearly twice its counterpart for the stiffest frame
presented thus far in this Appendix.
The horizontal retaining wall movements, shown in Fig. A.22, indicate a rela-
tively small wall displacement at the initial backfill stage, followed by a relatively
large wall movement into the retained soil. This movement into the soil mass and
through the original position of the wall is the largest found hitherto.
Associated with the wall movements is a large variation in bending moment
developed during the various stages of analysis, where the retaining wall assumes
a double curvature during expansion and single curvature during contraction. The
retaining wall’s largest maximum moment occurs at the last expansion cycle at top
of the wall with a magnitude nearly 8 times its counterpart found during the initial
backfill stage. The shear force developed in the retaining wall was also found to
have a large variation with temperature cycles. The maximum shear force corre-
sponds to the maximum moment developed during the last expansion cycle, and is
nearly 4.5 times larger than its counterpart found during the initial backfill stage.
The lateral earth pressure developed in the backfill soil during the analysis cy-
cles is shown in Fig. A.23. The pressure at the initial backfill stage is found simi-
lar to that developed during contraction cycles, while the earth pressure developed
during thermal expansion of the RFERS was found substantially larger, particular-
ly for the top half of the wall height. For instance, the magnitude of the lateral
earth pressure at mid-height of the retaining wall is nearly 3 times larger during
expansion, and approximately 8 times larger at about four-fifth the wall height.
The end column horizontal movements are similar in magnitude to the wall
movements at the other end of the frame, as shown in Fig. A.24. This is attributed
to the relatively large lateral stiffness of the rigid frame coupled with a relatively
long length capable of overcoming the retained soil restraint, thus leading to nearly
10
6 bays
8 L = 10 ft
b U
ha
φ = 30
o
U
6 hec1
z, ft
U
S /S = 4 hcc1
c b
4 U
hec2
U
hcc2
2 Uhec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
10
6 bays
8 Lb = 10 ft M
a
φ = 30o M
6 ec1
z, ft
S /S = 4 M
cc1
c b
4 M
ec2
M
cc2
2
M
ec3
0
-10 -5 0 5 10
10
6 bays
8 L = 10 ft
b Qa
φ = 30
o
6 Qec1
z, ft
Sc /Sb = 4 Qcc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. A.22 Analysis Results for Retaining Wall (Six-Bay, Lb = 10 ft, Sc/Sb = 4)
254 Appendix A
10
6 bays
σ'
a
8 L = 10 ft
b σ'
ec1
φ = 30
o
σ'
6 cc1
z, ft
σ'ec2
S /S = 4
c b
4 σ'cc2
σ'ec3
2
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h
Fig. A.23 Retaining Wall Earth Pressure (Six-Bay, Lb = 10 ft, Sc/Sb = 4)
10
8
U
ha
6 U hec1
z, ft
U hcc1
4 6 bays, Lb = 10 ft
U hec2
φ = 30o U hcc2
2
S c/S b = 4 U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
10
8
M
a
6 M
ec1
z, ft
M
6 bays, Lb = 10 ft cc1
4 Mec2
φ = 30
o
Mcc2
2 Sc/S b = 4 Mec3
0
-10 -5 0 5 10 15
Fig. A.24 End Column Analysis Results (Six-Bay, Lb = 10 ft, Sc/Sb = 4)

equal thermal movements at both ends of the frame. The maximum end column
bending moment, also shown in Fig. A.24, varies widely with the temperature cy-
cles. The maximum moment developed at the top of the end column during the
last expansion cycle, for instance, is nearly 6.5 times larger than its counterpart
developed during the initial backfill cycle.
A.1.1.7 15-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam Stiffness
Ratio, Sc/Sb, of 1)
beam stiffness ratio of 1 are presented herein. The frame is relatively stiff having
KL=2761 kip/ft/ft, as indicated in Table 7.2, and the second largest expansion
length at 150 feet.
The horizontal movements of the retaining wall are shown in Fig. A.25. Said
movements are qualitatively comparable to those of the wall of the 10-bay frame
with same element stiffness. The wall appears to expand into the soil mass for a
larger distance than its 10-bay counterpart, but the effect of the soil restraint re-
mains evident from the asymmetry of the temperature related displacements.
10
15 bays Uha
6
z, ft
Uhec1
L = 10 ft
b Uhcc1
4
φ = 30
o
U
hec2
Uhcc2
2 S /S = 1
c b
U
hec3
0
-0.02 -0.01 0 0.01 0.02 0.03 0.04
Fig. A.25 Retaining Wall Horizontal Movements (Fifteen-Bay, Lb =10 ft, Sc/Sb = 1)
The bending moment and shear force diagram developed in the wall during the
various analysis cycles are shown in Fig. A.26. The maximum bending moment
occurs at approximately mid-height of the wall at the end of the last expansion cy-
cle, and has a magnitude nearly 3.5 times larger than the moment developed dur-
ing the initial backfill cycle. Similarly, the maximum shear force occurs at the top
of the wall with a magnitude 6 times larger during the last expansion cycle
compared to the initial shear force developed during the backfill stage.
The lateral earth pressure developed behind the 15-bay frame is shown in Fig.
A.27. The results are similar to the frames of relatively large lateral stiffness and
long expansion length, where the pressure developed during the expansion cycles
is found substantially larger than its counterpart found at the end of the initial
backfill stage and the contraction cycles.
256 Appendix A
10
8
M
15 bays a
6 M
ec1
Lb = 10 ft
z, ft
M
cc1
4 φ = 30o Mec2
Sc /Sb = 1 Mcc2
2 Mec3
0
-8 -6 -4 -2 0 2 4 6
10
8
15 bays Qa
Qec1
6 L = 10 ft
z, ft
b Qcc1
φ = 30
o
4 Q
ec2
S /S = 1 Q
c b cc2
2 Q
ec3
0
-3 -2 -1 0 1 2 3 4 5
Fig. A.26 Retaining Wall Moment and Shear Diagrams (Fifteen-Bay, Lb =10 ft, Sc/Sb = 1)
10
15 bays, Lb = 10 ft
8
φ = 30o
6
Sc /Sb = 1
z, ft
4 σ' σ'
a ec2
σ' σ'
ec1 cc2
2
σ'cc1 σ'ec3
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h
Fig. A.27 Retaining Wall Earth Pressure (Fifteen-Bay, Lb =10 ft, Sc/Sb = 1)
The end column horizontal movements, shown in Fig. A.28, indicate larger ex-
pansion movements compared with the movements undergone during the contrac-
tion cycles. The maximum bending moment in the end column, which occurs dur-
ing the last expansion cycle, is nearly 4 times larger than the moment developed
during the initial backfill stage and the contraction cycles.
10
15 bays
6
z, ft
Lb = 10 ft
4
φ = 30o U ha Uhec2
U hec1 Uhcc2
2 S c/Sb = 1
U hcc1 Uhec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
10
6 15 bays
z, ft
L b = 10 ft
4 M M
φ = 30o a ec2
M M
ec1 cc2
2 S /S = 1
c b M M
cc1 ec3
0
-2 -1 0 1 2 3 4 5
Fig. A.28 End Column Analysis Results (Fifteen-Bay, Lb =10 ft, Sc/Sb = 1)
A.1.1.8 15-Bay Frame (Bay Length, Lb, 10 Feet, Column to Beam Stiffness
Ratio, Sc/Sb, of 4)
beam stiffness ratio of 4 are presented herein. The frame has the largest lateral
stiffness of all frames presented thus far in this Appendix and the second largest
expansion length of all analyzed frames. Moreover, this frame has 8 times the lat-
eral stiffness of the 15-bay structure presented earlier.
The horizontal wall movements, shown in Fig. A.29, indicate relatively small
horizontal displacements at the initial backfill stage, and expansion and contraction
258 Appendix A
movements nearly symmetrical about the initial displacement. This behavior is at-
tributed to the relatively large lateral stiffness of the frame, coupled with a long
length leading to the nearly full expansion movements of the frame into the backfill
soil during the expansion cycles. Similar behavior was also observed earlier in
frames with similar stiffness and strength characteristics.
10
8
U
15 bays ha
6 U
hec1
z, ft
L = 10 ft
b U hcc1
φ = 30
o
4 U hec2
S /S = 4 U hcc2
c b
2 U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04
Fig. A.29 Horizontal Retaining Wall Movements (Fifteen-Bay, Lb =10 ft, Sc/Sb = 4)
The bending moment and shear force diagrams for the retaining wall are shown
in Fig. A.30. The bending moments developed in the wall during the temperature
change cycles are clearly substantially larger than their counterpart found during
the initial backfill stage. However, the magnitude of the maximum bending mo-
ment in this case is approximately identical during both the expansion and con-
traction cycles, indicating no apparent effect of the backfill soil on the develop-
ment of flexural stresses in the retaining wall.
The shear force diagram, on the other hand, illustrates a maximum shear
force occurring at the top of the retaining wall during the last expansion cycle,
with a magnitude nearly 3 times its counterpart developed during the contraction
cycles, and approximately 6 times the shear found at the end of the initial backfill
stage.
The lateral earth pressure developed in the backfill soil during the various anal-
ysis cycles is shown in Fig. A.31. The diagram for the pressure associated with the
last expansion cycle indicate that substantially large forces develop during the ex-
pansion movement of the structure compared with the contraction cycles and the
initial backfill stage. The pressure distribution is nearly an inverted triangle with
the maximum pressure occurring near the top of the structure.
Given the linear variation of the vertical effective stress in the soil mass, the
earth pressure magnitude and distribution indicate that the coefficient of lateral
earth pressure varies along the height of the structure in a non-linear fashion.
10
15 bays
8 L = 10 ft
b M
a
φ = 30
o
M
6 ec1
z, ft
Sc/Sb = 4 M
cc1
4 Mec2
Mcc2
2 Mec3
0
-40 -30 -20 -10 0 10 20 30 40
10
15 bays
8 L = 10 ft
b Qa
φ = 30
o
6 Qec1
z, ft
Sc /Sb = 4 Qcc1
4 Q
ec2
Q
cc2
2
Q
ec3
0
-6 -4 -2 0 2 4 6 8
Fig. A.30 Retaining Wall Moment and Shear Diagrams (Fifteen-Bay, Lb =10 ft, Sc/Sb = 4)
10
6
z, ft
15 bays
4 L = 10 ft σ' σ'
b a ec2
φ = 30
o σ' σ'
ec1 cc2
2 S /S = 4
c b σ' σ'
cc1 ec3
0
0 0.2 0.4 0.6 0.8 1 1.2
2
σ' , kips/ft
h
Fig. A.31 Retaining Wall Earth Pressure (Fifteen-Bay, Lb =10 ft, Sc/Sb = 4)
260 Appendix A
The end column analysis results are shown in Fig. A.32. The horizontal end
column movements are nearly equal to those of the retaining wall at the other end
of the rigid frame, given the small effect of the soil restraint on the thermal
movements of the structure.
10
8
U
ha
6 15 bays U hec1
z, ft
L = 10 ft U hcc1
b
4 U hec2
φ = 30
o
U hcc2
2 S /S = 4
c b U
hec3
0
-0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05
h
10
15 bays
8 L = 10 ft
b M
a
φ = 30
o
M
6 ec1
z, ft
M
Sc/Sb = 4 cc1
4 Mec2
Mcc2
2 Mec3
0
-40 -30 -20 -10 0 10 20 30 40
Fig. A.32 End Column Analysis Results (Fifteen-Bay, Lb =10 ft, Sc/Sb = 4)
The end column bending moments developed during temperature cycles are
also nearly symmetrical about the moments found at initial backfill stage. The
maximum moment magnitude is found at top of the column during last expansion
cycle slightly larger and opposite to its counterpart developed during contraction
cycles.
A.1.2 Backfill Soil with 40º Internal Friction Angle

The analysis results of single story rigid frames retaining the backfill soil with the
40º internal friction angle are presented herein. The results are presented in four
groups of frames with the same configurations as those presented earlier, for the
initial backfill stage and the last expansion cycle only. The reason for presenting
the analysis results differently than in the preceding sections is conciseness.
A.1.2.1 Frames with Bay Length, Lb, 10 Feet, and Column to Beam Stiffness
Ratio, Sc/Sb, of 1
The horizontal movements of the retaining wall portion of the single story RFERS
with bay length of 10-ft and column to beam stiffness ratio of 1 are presented
in figure A.33 and A.34 for the initial backfill stage and last expansion cycle,
respectively.
10
8
U
hn1
6 L = 10 ft U
b
z, ft
hn3
φ = 40o U
hn6
4 U
hn10
S /S = 1
c b U
2 hn15
U
hn20
0
0 0.005 0.01 0.015 0.02
h
Fig. A.33 Retaining Wall Movements at Backfill Stage (φ = 40º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The retaining wall displacements at the end of the backfill stage, shown in Fig.
A.33, are found to decrease with the increase in the lateral stiffness of the frame.
The displacement at the top of the wall of the single bay frame, for instance, is
nearly 7 times larger than its counterpart found for the 20-bay frame.
Fig. A.34, on the hand, shows the deflected shape of the wall at the end of the
last expansion cycle. The wall of the single bay frame is found near the position it
initially assumed at the end of the backfill stage, while many other frames are
found to have moved substantially. The deflected shape of the 20-bay frame indi-
cates a large curvature is induced in the wall due to the relatively large thermal
expansion force exerted at the wall top couple with the lateral earth pressure load
exerted by the backfill soil.
262 Appendix A
10
8
U
hen1
L = 10 ft U
6 b hen3
z, ft
φ = 40o U
hen6
4 U
S /S = 1 hen10
c b
U
hen15
2 U
hen20
0
0 0.005 0.01 0.015 0.02
he
Fig. A.34 Retaining Wall Movements at Expansion Cycle (φ = 40º, All Bays, Lb =10 ft,
Sc/Sb = 1)
10
6
z, ft
L = 10 ft
4 b σ σ
en1 en10
φ = 40o σ σ
en3 en15
2 σ σ
S /S = 1 en6 en20
c b
0
0 0.5 1 1.5 2 2.5
σ , kips/ft
2
h
Fig. A.35 Lateral Earth Pressure at Expansion Cycle (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 1)
The lateral earth pressure developed behind the frames at the initial backfill
stage was presented previously in the top graphic of Fig. 6.7 in Chapter 6. The
earth pressure developed at the end of the expansion cycle is shown in Fig. A.35.
The pressure developed at the last expansion cycle is clearly substantially larger
for the frames with large expansion length compared with the initial backfill pres-
sure. The distribution of the lateral earth load is also significantly different at the
expansion cycle, with the largest pressure magnitude occurring near the top of the
wall.
The maximum moments found in the wall during the initial backfill stage and
the last expansion cycles are shown in Fig. A.36 and A.37 respectively. For the
single and 3-bay frames, the moments shown in both figures are nearly comparable
in magnitude. The moments for the longer frames, however, developed at the end
of the last expansion cycle are evidently substantially larger than their counterpart
found at the initial backfill stage. The wall of the 15-bay frame, for instance, devel-
oped a maximum moment magnitude at mid-height during the expansion cycle ap-
proximately 9 times than the moment found at the initial stage.
10
M
bn1
8 M
bn3
M
bn6
6
z, ft
M
bn10
M L = 10 ft
4 bn15 b
M φ = 40o
bn20
2
S /S = 1
c b
0
-20 -15 -10 -5 0 5
Fig. A.36 Retaining Wall Moments at Backfill Stage (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 1)
The horizontal end column movements for the 6 frames analyzed are shown in
Fig. A.38 for initial backfill stage and the last expansion cycle. The two graphic in
said figure indicate than temperature movements have neatly negligible effects on
the movements of the end column for the single bay frame, and increasingly larger
effects on column movements with the increasing number of bays in the frame.
This behavior is clearly expected since increasing the number of bays results in an
10
Men1
8 M
en3
M
en6
6
z, ft
M
en10
M L = 10 ft
4 en15 b
M
en20 φ = 40o
2
S /S = 1
c b
0
-20 -15 -10 -5 0 5
Fig. A.37 Retaining Wall Moments at Expansion Cycle (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 1)
264 Appendix A
increase of the expansion and contraction length of the frame. However, the end
column horizontal displacement is also related to the lateral stiffness of the frame
and the stiffness of the retained soil mass as was shown earlier for structures
retaining the backfill with a 30º internal friction angle.
10
8
U
hn1
6 U
hn3
z, ft
L = 10 ft U
b hn6
4 U
φ = 40 o
hn10
U
hn15
2 S /S = 1
c b U
hn20
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
h
10
8
U
hen1
6 U
hen3
z, ft
L = 10 ft U
b hen6
4
φ = 40o U
hen10
U
S /S = 1 hen15
2 c b
U
hen20
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
h
Fig. A.38 End Column Horizontal Movements (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 1)
The end column moment developed at the initial backfill stage and the last ex-
pansion cycle are shown in Fig. A.39. The top and bottom graphics of said figure
show a large variation in the moment between the two stages of analysis with the
variation of the rigid frame length and stiffness.
10
8
M
bn1
6 M
z, ft
bn3
M
bn6 L = 10 ft
4 M
b
bn10
φ = 40o
M
2 bn15
M S /S = 1
bn20 c b
0
-10 -5 0 5 10
10
8
M
en1
6 M
en3
z, ft
M
en6 L = 10 ft
4 M b
en10
φ = 40o
M
en15
2
M S /S = 1
en20 c b
0
-10 -5 0 5 10
Fig. A.39 End Column Moments (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 1)
Ratio, Sc/Sb, of 4
in Fig. A.40 for the initial backfill stage and the last expansion cycle.
The top graphic in Fig. A.40 shows that the horizontal retaining wall move-
ments during the initial backfill stage are substantially smaller for this series of
frames compared with their counterparts presented previously in Fig. A.33. Addi-
tionally, frames with more than three bays appear to undergo a horizontal dis-
placement into the retained soil mass during the last expansion cycle unlike the
movements of their previous counterparts shown in Fig. A.34.
266 Appendix A
10
8
Uhn1
6 U
hn3
z, ft
U
hn6 Lb = 10 ft
4 U
hn10
φ = 40o
U
hn15
2
Uhn20 S /S = 4
c b
0
-0.03 -0.02 -0.01 0 0.01 0.02
h
10
8
U
hen1
6 U
z, ft
hen3
U
hen6 Lb = 10 ft
4
U
hen10
φ = 40o
U
2 hen15
U S /S = 4
hen20 c b
0
-0.03 -0.02 -0.01 0 0.01 0.02
h
Fig. A.40 Horizontal Retaining Wall Movements (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 4)
10
8
σen1
6 σ
z, ft
en3
L = 10 ft
b σ
en6
4 φ = 40o σ
en10
S /S = 4 σ
en15
2 c b
σen20
0
0 1 2 3 4 5
σ , kips/ft
2
h
Fig. A.41 Lateral Earth Pressure at Expansion Cycle (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 4)
The earth pressure developed behind the rigid frames at the last expansion cy-
cle is shown in Fig. A.41. The earth pressure distribution and magnitude vary
widely with the length and stiffness of the frames, and is also substantially larger
for these frames compared with their counterparts of the same length shown in
Fig. A.35. These results illustrate the significant contribution of the lateral frame-
stiffness to the development of lateral earth pressures at expansion.
The retaining wall moments developed at initial backfill stage are shown in Fig.
A.42. The moment magnitudes vary slightly between the 6 frames analyzed, and are
smaller than moment magnitudes developed in the retaining wall portion of the
frames with column to beam stiffness ratio of 1 presented earlier in Fig. A.36. The
magnitudes of the retaining wall moments developed during the last expansion cycle
near mid-height of the wall are comparable or slightly larger than their counterparts
for frames with lesser lateral stiffness as illustrated in Fig. A.43 and A.37. On the
other hand, however, the moment magnitudes at the top of the retaining wall, shown
in Fig. A.43, indicate that substantially larger moments developed during the expan-
sion cycle in the frames with the larger column to beam stiffness ratio.
10
8
M
bn1
6 L = 10 ft M
z, ft
bn3
b
M
4 φ = 40 o bn6
M
bn10
S /S = 4 M
c b bn15
2
M
bn20
0
-20 -10 0 10 20 30 40
Fig. A.42 Retaining Wall Moments at Backfill (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 4)
10
8
Men1
6 L = 10 ft M
z, ft
b en3
M
φ = 40 o
en6
4 M
en10
S /S = 4
c b M
2 en15
M
en20
0
-20 -10 0 10 20 30 40
Fig. A.43 Retaining Wall Moments at Expansion (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 4)
268 Appendix A
The end column horizontal movements are presented in Fig. A.44 for the initial
backfill stage and the last expansion cycle. The movements shown are substantially
smaller than their counterparts for the frames shown in Fig. A.38 for both the initial
backfill stage and the expansion cycle. The movements of the end column of the 10-
bay frame shown in Fig. A.44, for instance, are nearly 4 times smaller than its coun-
terpart of the frame with lesser lateral stiffness for both stages of analysis.
This large difference in frame behavior is strictly attributed to the change in the
lateral stiffness of the frame, given that all other soil and structural properties re-
mained constant. The 10-bay frame with the larger stiffness underwent a larger
expansion movement at the retaining wall side, thus decreasing the amount of
expansion at the end column and resulting in substantially different structural
deflections for the rigid frame due to temperature induced movements.
10
8
U
hn1
6 U
L = 10 ft hn3
z, ft
b
U
4 φ = 40o hn6
U
hn10
S /S = 4 U
c b hn15
2
U
hn20
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
h
10
8
U
hen1
6 U
hen3
z, ft
L = 10 ft
b
U
hen6
4 φ = 40o
U
hen10
S /S = 4 U
c b hen15
2
U
hen20
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
h
Fig. A.44 End Column Horizontal Movements (φ = 40º, All Bays, Lb =10 ft, Sc/Sb = 4)
The end column moments developed during the initial backfill stage and
the last expansion cycle are shown in Fig. A.45. The magnitude of the end column
moments during the initial backfill stage are reasonably comparable to their
counterpart developed in the end column of the frames with the smaller column to
beam stiffness shown in Fig. A.39. Conversely, however, the magnitude of the end
column moments shown in Fig. A.45 at the last expansion cycle are substantially
larger than their counterparts presented in Fig. A.39.
10
8
M
bn1
6 M
z, ft
bn3
M
L = 10 ft bn6
4 b
M
φ = 40o bn10
M
2 bn15
S /S = 4 M
c b bn20
0
-40 -30 -20 -10 0 10 20 30 40
10
8
M
en1
6 M
z, ft
en3
L = 10 ft M
en6
4 b
M
φ = 40 o en10
M
2 en15
S /S = 4 M
c b en20
0
-40 -30 -20 -10 0 10 20 30 40
Ratio, Sc/Sb, of 1
The final wall displacements at the end of initial backfill stage and last expan-
sion cycle are substantially larger than their counterparts of frames with 10-ft bay
length. Additionally, difference in horizontal displacements of the wall between
expansion cycle and initial backfill stage are substantially larger than the frames
with 10-ft bay length, and are shown to increase with increasing length of the
structure.
270 Appendix A
10
6
z, ft
L = 20 ft
4 b U
hn1
φ = 40o U
hn3
2 U
S /S = 1 hn10
c b
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
h
10
6
z, ft
L = 20 ft
4 b U
hen1
φ = 40o U
hen3
2 S /S = 1 U
hen10
c b
0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
h
10
6
z, ft
L = 20 ft σ
4 b
en1
φ = 40o σ
en3
2 σ
Sc/Sb = 1 en10
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
σ , kips/ft
2
h
Fig. A.47 Lateral Earth Pressure at Expansion Cycle (φ = 40º, All Bays, Lb =20 ft,
Sc/Sb = 1)
last expansion cycle is shown in Fig. A.47. As established from earlier results, the
pressure is found to increase with the increase in the number of bays. The maxi-
mum pressure magnitude, however, found for the 10-bay frame with a length of
200 feet, is smaller than its counterpart developed for the frames with similar
length discussed previously. This frame also has the smallest lateral stiffness of all
frames with the same length presented thus far (for the backfill soil with the 40º
internal friction angle) as indicated in Table 7.2 of Chapter 7. Additionally, the
earth pressure magnitude is found to peak only close to the top of the wall and de-
crease rapidly towards the bottom, more rapidly than the frames with 10-ft bay
length.
The retaining wall moments at backfill stage are shown in Fig. A.48. The
moment magnitude at the top of the wall for all three frames is larger than its
counterpart found for the frames with 10-ft bay length presented earlier.
10
6
z, ft
L = 20 ft
4 M
b
bn1
φ = 40o
M
bn3
2
M S /S = 1
bn10 c b
0
-10 -5 0 5
Fig. A.48 Retaining Wall Moment at Backfill Stage (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 1)
10
6
z, ft
L = 20 ft
4 M
b
en1
φ = 40o
M
en3
2
M S /S = 1
en10 c b
0
-10 -5 0 5
Fig. A.49 Retaining Wall Moment at Expansion (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 1)
272 Appendix A
Fig. A.49, on the other hand, illustrates the retaining wall moments found at the
last expansion cycle, where the moment diagram for the single and 3-bay frames
are found reasonably comparable to their counterparts developed at the initial
backfill stage, while the 10-bay frame moment is found to be substantially larger
at mid-height of the wall.
The horizontal end column movements at the initial backfill stage and the last
expansion cycle are shown in Fig. A.50. The top graphic of said figure indicates
that the column displacements during the backfill stage are fairly larger for the 20-
ft bay frames compared with their counterparts presented earlier. The bottom
graphic is Fig. A.50, show substantially larger horizontal movements during the
expansion cycles.
10
6
z, ft
L = 20 ft
4 b U
hn1
φ = 40o U
hn3
2 U
S /S = 1 hn10
c b
0
0 0.02 0.04 0.06 0.08 0.1
h
10
6
z, ft
L = 20 ft
b U
4 hen1
φ = 40o U
hen3
2 U
S /S = 1 hen10
c b
0
0 0.02 0.04 0.06 0.08 0.1
h
Fig. A.50 Horizontal End Column Movements (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 1)
The end column moments resulting from the analyses of the three 20-ft bay
frames are shown in Fig. A.51. The moment magnitudes developed are shown to
be more comparable to the 10-ft bay frames with a column to beam stiffness ratio
of 1, rather than to the stiffer frames, for both stages of analyses.
10
6
z, ft
L = 20 ft
4 M
b
bn1
φ = 40o
M
bn3
2
M S /S = 1
bn10 c b
0
-10 -5 0 5 10
10
6
z, ft
L = 20 ft
4 M
b
en1
φ = 40o
M
en3
2
M S /S = 1
en10 c b
0
-10 -5 0 5 10
Ratio, Sc/Sb, of 4
The three frames of this series possess lateral stiffness substantially larger than
their counterparts presented latterly of smaller column to beam ratio, as indicated
in Table 7.2 of Chapter 7. The horizontal retaining wall movements vary with the
stiffness of the frames. During the backfill stage, for instance, the movements are
found to be relatively small. At the end of the expansion cycle, the 10-bay frame is
shown to undergo large expansion movements into the soil mass.
274 Appendix A
10
6
z, ft
L = 20 ft Uhn1
4 b
φ = 40o U
hn3
2 Uhn10
S /S = 4
c b
0
-0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02
h
10
6
z, ft
L = 20 ft
b
4 Uhen1
φ = 40o U
hen3
2 S /S = 4 Uhen10
c b
0
-0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02
h
10
6
z, ft
Lb = 20 ft
4 σ
hn1
φ = 40o σ
hn3
2
Sc/Sb = 4 σ
hn10
0
0 0.5 1 1.5 2 2.5 3 3.5
σ , kips/ft
2
h
Fig. A.53 Lateral Earth Pressure at Expansion Cycle (φ = 40º, All Bays, Lb =20 ft,
Sc/Sb = 4)
The lateral earth pressure developed behind the frames at the end of the last ex-
pansion cycle is shown in Fig. A.53. The magnitudes and distributions of the
backfill pressures on the retaining wall are found to increase with the increasing
length and stiffness of the frames, and are reasonably comparable to their counter-
part found behind the 10-ft bay frames with the column to beam stiffness ratio of
4, but substantially larger than for the frames presented latterly.
The retaining wall moments resulting from the analysis at the initial backfill
stage are shown in Fig. A.54. The moment diagrams shown in said figure indicate
that the flexural stresses developed in the three frames are reasonably comparable,
similar to what has been found for the 10-ft bay frames with a column to beam
stiffness ratio of 4. The retaining wall moments at the end of the expansion cycle,
on the other hand, vary with the lateral stiffness and length of the frames. The
magnitude of said moments are substantially larger than their counterparts devel-
oped in the 20-ft bay frames presented latterly, and are fairly comparable to the
10-bay frames particularly at mid-height.
10
6
z, ft
L = 20 ft
4 b M
bn1
φ = 40o M
bn3
2
S /S = 4 M
c b bn10
0
-20 -15 -10 -5 0 5 10 15 20
Fig. A.54 Retaining Wall Moments at Backfill (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
10
6
z, ft
L = 20 ft
4 b M
en1
φ = 40o M
en3
2 M
S /S = 4 en10
c b
0
-20 -15 -10 -5 0 5 10 15 20
Fig. A.55 Retaining Wall Moments at Expansion (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
276 Appendix A
The horizontal end column movements undergone at the initial backfill cycles
are shown in Fig. A.56. The movements are in-line with those found at the other
end of the frame, and are relatively small compared to the 20-ft bay frames pre-
sented latterly. The movements of the end column at the end of the last expansion
cycle, however, vary largely with the length and stiffness of the frames, as shown
in Fig. A.57.
10
6
z, ft
L = 20 ft
4 b U
hn1
φ = 40o U
hn3
2 U
S /S = 4 hn10
c b
0
0 0.01 0.02 0.03 0.04 0.05
h
Fig. A.56 End Column Movements at Backfill (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
10
6
z, ft
L = 20 ft
4 b U
hen1
φ = 40o U
hen3
2 U
S /S = 4 hen10
c b
0
0 0.01 0.02 0.03 0.04 0.05
h
Fig. A.57 End Column Movements at Expansion (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
Additionally, said figure shows the horizontal displacement for the column of
the 10-bay frame to be nearly twice that of the three bay frame, while the expan-
sion length of the 10 bay frame is more than 30 times larger. This of course is due
to the relatively large expansion movements undergone by the 10-bay frame at the
retaining wall end, thus decreasing the expansion demand at the free end.
10
6
z, ft
L = 20 ft
4 b
M
bn1
φ = 40o
M
bn3
2
M S /S = 4
bn10 c b
0
-30 -20 -10 0 10 20
Fig. A.58 End Column Moments at Backfill (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
10
6
z, ft
L = 20 ft
4 M
b
en1
φ = 40o
M
en3
2
M S /S = 4
en10 c b
0
-30 -20 -10 0 10 20
Fig. A.59 End Column Moments at Expansion (φ = 40º, All Bays, Lb =20 ft, Sc/Sb = 4)
The end column moments at the end of the backfill stage, shown in Fig. A.58,
are fairly comparable to their counterpart of the 20-ft bay frames with a column to
beam stiffness ratio of 1. The end column moments at expansion, however, are
substantially larger than their counterparts of the frames presented latterly. This of
course indicates that in addition to the expansion and contraction length of the
frames, the effect of lateral stiffness is also fairly pronounced in the development
of flexural stresses in the frame members due to the thermal movements.
Appendix B
Abstract. This Appendix extends on the results of numerical parametric analysis

of multi-story structures presented in Chapter 9. Results are reported for three and
five story structures in Chapter 9 and for two and four stories in this Appendix.
The primary purpose of this analysis is to investigate the effects of thermal
movements of Rigidly Framed Earth Retaining Structures (RFERS) on (1) the dis-
placement of the rigid frames, (2) the stresses developed in the structural elements,
and (3) the lateral earth pressure developed in the soil mass.
B.1 Thermal Soil Structure Interaction of Multi-story RFERS
B.1.1 Two Story Rigidly Framed Earth Retaining Structures

The analysis results of two story rigid frames are presented herein. The results are
reported for 1, 3, 6, 10, 15 and 20 bay frames under four separate sections corre-
sponding to the varied parameters shown in Table 9.1. The results reported for
each temperature cycle, in addition to the initial backfill stage, are (1) the dis-
placement of the retaining wall; (2) the lateral earth pressure developed in the re-
tained soil mass; (3) the retaining wall bending moment; (4) the displacement of
the end column; and (5) the end column bending moment.
B.1.1.1 Frames with Bay Length, Lb, 10 Feet, and Column to Beam Stiffness
Ratio, Sc/Sb, of 1
The horizontal movements of retaining wall portion of the two story RFERS with a
bay length of 10 feet and column to beam stiffness ratio of 1 are presented in Fig.
B.1 and B.2 for the initial backfill stage and last expansion cycle, respectively.
The retaining wall displacements at the end of the backfill stage, shown in Fig. B.1,
are found to decrease with the increase in the lateral stiffness of the frame. The
displacement at the top of the wall of the single bay frame, for instance, is over 12
times larger than its counterpart found for the 20-bay frame. The displaced walls
at the end of the expansion cycle, shown in Fig. B.2, indicate a slight change in
their displacements found at the end of the backfill stage.
280 Appendix B
The lateral earth pressure developed at the end of expansion cycle is shown in
Fig. B.3. The earth pressure for each the of rigid frames is compared to the Cou-
lomb active earth pressure for a level backfill with a friction angle between the
wall and backfill soil of 22.5º (0.75 φ), and prescribed lateral earth pressure values
found in major building codes adopted in the United States presented in table 1.1
of Chapter 1.
20
L = 10 ft U
b hn1
U
15 φ = 30 o
hn3
U
hn6
S /S = 1
z, ft
c b U
10 hn10
U
hn15
U
hn20
5
0
0 0.05 0.1 0.15 0.2
h
Fig. B.1 Retaining Wall Movements at Backfill Stage ( φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
20
L = 10 ft U
b hen1
15 φ = 30 o
U
hen3
Uhen6
S /S = 1
c b
U
z, ft
hen10
10
U
hen15
U
hen20
5
0
0 0.05 0.1 0.15 0.2
h
Fig. B.2 Retaining Wall Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
B.1 Thermal Soil Structure Interaction of Multi-story RFERS 281
20
σ
hen1
σ
15 hen3
σ
hen6
σ
z, ft
hen10
10
L = 10 ft σ
b hen15
σ
φ = 30o hen20
5 Coulomb
ASCE 7-98
c b
K0
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
σ , kips/ft
2
h
Fig. B.3 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
Fig. B.3 indicates that the lateral earth pressure developed in the soil mass at
the end of the last expansion cycle increases with the increase in the number of
bays (or length of the rigid frame). The earth pressure found behind the single and
3 bay frames, for instance, is found to be comparable to the lateral soil pressures
prescribed by the major building codes and the pressure determined using Cou-
lomb’s active pressure theory. The pressure found behind the 15 and 20 bay
frames, on the other hand is found to exceed the lateral earth pressure at-rest for
the predominance of the backfill height.
The bending moment diagrams for the retaining wall at the end of the backfill
stage and last expansion cycle are shown in Fig. B.4 and B.5.
20
M L = 10 ft
b
bn1
M φ = 30o
15 bn3
M
bn6 S /S = 1
c b
M
z, ft
bn10
10
M
bn15
M
bn20
5
0
-35 -30 -25 -20 -15 -10 -5 0 5
Fig. B.4 Retaining Wall Moment at Backfill Stage (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
282 Appendix B
20
M L = 10 ft
en1 b
15
M
en3
φ = 30o
M
en6 S /S = 1
c b
M
z, ft
en10
10
M
en15
M
en20
5
0
-35 -30 -25 -20 -15 -10 -5 0 5
Fig. B.5 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
The magnitudes of the moments for the single, 3 and 6 bay frames at both anal-
ysis stages are found to be nearly equal. The change in moment values appears to
increase, however, with the increasing number of bays from 10 to 20. A complete
reversal in the moment direction is found at the mid height of the wall (or the first
floor level) for the 20-bay frame. The moments at the top of the wall, however, are
nearly equal for all frames at both analysis stages.
The horizontal end column displacements are presented in Fig. B.6 and B.7.
The movements found at the backfill stage are found to vary slightly for shorter
frames from their counterparts found at the end of the last expansion cycle and
substantially for longer frames.
20
L = 10 ft U
b hn1
U
15 φ = 30 o
hn3
U
hn6
S /S = 1
z, ft
c b U
10 hn10
U
hn15
U
hn20
5
0
0 0.05 0.1 0.15 0.2
h
Fig. B.6 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
20
L = 10 ft U
hen1
b
U
15 φ = 30 o hen3
U
hen6
S /S = 1 U
z, ft
c b hen10
10
U
hen15
U
hen20
5
0
0 0.05 0.1 0.15 0.2
h
Fig. B.7 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 1)
The horizontal displacement of the end column in the 20-bay frame, with the
largest expansion length, is found to exceed that of the 3 bay frame for the majori-
ty of the height of the column. The restraining effects of the retained soil mass are
therefore clearly pronounced for the longer frames.
The end column moment diagrams are shown in Fig. B.8 and B.9. The
moments found at the end of the backfill stage single and 3-bay frame are shown
to only vary slightly from their counterparts found at the end of the expansion
cycle.
20
M
bn1
M
15 bn3
M
bn6
M
z, ft
bn10
10
M L = 10 ft
bn15 b
M φ = 30o
bn20
5
S /S = 1
c b
0
-10 -5 0 5 10
Fig. B.8 End Column Moment at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 1)
284 Appendix B
20
M
en1
M
15 en3
M
en6
M
z, ft
en10
10
M L = 10 ft
en15 b
M φ = 30o
en20
5
S /S = 1
c b
0
-10 -5 0 5 10
The longer frames, on the other hand, show an increase in the magnitude of the
end column bending moment occurring at the first floor level (or mid-height), with
the end column moment of the 20-bay frame reaching a magnitude at the end of
the expansion cycle over 5 times larger than that at the backfill stage.
Ratio, Sc/Sb, of 4
The horizontal movements of the retaining wall portion of the two story RFERS
in Fig. B.10 and B.11 for the initial backfill stage and the last expansion cycle,
respectively.
20
L = 10 ft U
b
hn1
φ = 30o U
15 hn3
U
S /S = 4 hn6
c b
U
z, ft
hn10
10
U
hn15
U
hn20
5
0
-0.04 -0.02 0 0.02 0.04 0.06
h
Fig. B.10 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb =10 ft,
Sc/Sb = 4)
20
L = 10 ft
b U
hen1
15
φ = 30o U
hen3
S /S = 4 U
hen6
c b
z, ft
U
10 hen10
U
hen15
U
hen20
5
0
-0.04 -0.02 0 0.02 0.04 0.06
h
Sc/Sb = 4)
The displacement response of the retaining wall for this group of frames with
the larger column to beam stiffness ratio vary considerably from its counterpart
presented earlier for the frames with a column to beam stiffness ratio of 1.
While the horizontal displacements of the retaining walls in the latter frames
were nearly comparable at end of the backfill stage and the last expansion cycle,
said displacements are found to only be similar for the single and 3-bay frames
with the column to beam stiffness ratio of 4. The remainder of the frames with the
latter stiffness are found to expand into the soil mass at varying amounts depend-
ing on their lengths.
The lateral earth pressure developed in retained soil mass at the end of the ex-
pansion cycle for the group of frames is shown in Fig. B.12. The results indicate
that earth pressure developed behind the single and 3 bay frames is comparable to
Coulomb’s active earth pressure and to the soil pressures prescribed by the major
national codes adopted in the United States of America.
20
σ
hen1
σ
hen3
15
σhen6
σhen10
z, ft
10
Lb = 10 ft σhen15
σhen20
φ = 30
o
5 Coulomb
ASCE 7-98
c b
K0
0
0 0.5 1 1.5 2
σ , kips/ft
2
h
286 Appendix B
Conversely, the magnitude of the earth pressure developed behind the 6-bay
frame is found to be comparable to the magnitude of the lateral earth pressure at
rest, for the top one-half of the wall height. Additionally, the earth pressure is
found to increase as the frame length increases, reaching relatively large magni-
tudes. The pressure developed behind the 15-bay frame is shown to have a magni-
tude nearly twice that of the lateral earth pressure at rest.
The moment diagrams for the retaining wall portion of the two-story RFERS
are shown in Fig. B.13 and B.14.
The magnitudes of the retaining wall moment at the end of the backfill stage
and the last expansion cycle for the single and 3-bay frames is found to be similar.
As the number of bays increases, the magnitude of the retaining wall moment de-
veloped at the first floor level are found to increase accordingly. The moment for
the 20-bay frame, for instance, developed at the end of the last expansion cycle is
shown in to be nearly 25 times larger than its counterpart found at the end of the
initial backfill stage.
20
M L = 10 ft
bn1 b
M
15 bn3 φ = 30o
M
bn6
S /S = 4
M c b
z, ft
bn10
10
M
bn15
M
bn20
5
0
-30 -20 -10 0 10 20 30
20
M L = 10 ft
en1 b
M
15 en3 φ = 30o
M
en6
S /S = 4
M c b
z, ft
en10
10
M
en15
M
en20
5
0
-30 -20 -10 0 10 20 30
The horizontal movements of the end column of the frames are shown in Fig.
B.15 and B.16. The horizontal movements of the end columns in this group of
frames are substantially smaller than their counterpart shown in Fig. B.6 and B.7
for the frames with smaller column to beam stiffness ratios.
The effect of temperature change and soil restraint on the frame movements are
evident when comparing the two figures below to the horizontal movements of the
retaining wall at the other end of the frame shown in Fig. B.10 and B.11. The end
column of the 15-bay frame, for instance, is shown to have a final horizontal
movement twice its counterpart at the other end.
The end column moment diagrams are shown in Fig. B.17 for the backfill
stage, and Fig. B.18 for the last expansion cycle.
While the magnitude of the bending moment in the end column of the single
bay frame is shown to merely vary slightly in the two figures below, the end col-
umns of the remainder five frames are found to develop bending moment at the
20
L = 10 ft U
b hn1
U
15 φ = 30 o
hn3
U
hn6
S /S = 4
z, ft
c b U
10 hn10
U
hn15
U
hn20
5
0
0 0.01 0.02 0.03 0.04 0.05
h
20
L = 10 ft U
b
hen1
15
φ = 30 o
U
hen3
U
S /S = 4 hen6
c b
U
z, ft
hen10
10
U
hen15
U
hen20
5
0
0 0.01 0.02 0.03 0.04 0.05
h
Fig. B.16 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb =10 ft, Sc/Sb = 4)
288 Appendix B
end of the expansion cycle substantially larger than their counterparts found at the
backfill stage. The end column of the 20-bay frame is found to have a bending
moment developed at the first level at the end of the expansion cycle nearly 40
times larger than the moment developed at the initial backfill stage.
20
L = 10 ft
b M
bn1
φ = 30o M
15 bn3
S /S = 4 M
c b bn6
z, ft
M
10 bn10
M
bn15
M
bn20
5
0
-10 0 10 20 30 40 50
20
M
en1
M
15 en3
M
en6
M
z, ft
en10
10
M L = 10 ft
en15 b
M
en20 φ = 30o
5
S /S = 4
c b
0
-10 0 10 20 30 40 50
Ratio, Sc/Sb, of 1
in Fig. B.19 and B.20 for the initial backfill stage and the last expansion cycle
The horizontal retaining wall movements for this group of frames are found to
be generally larger than their counterpart presented earlier. The horizontal position
of the wall for all three frames presented herein is similar in the two figures shown
below, with a small change in horizontal movement found at the end of the last
expansion cycle away from the retained soil mass.
20
15
z, ft
10
L = 20 ft
b
U
hn1
φ = 30o U
5 hn3
S /S = 1 U
c b hn10
0
0 0.05 0.1 0.15 0.2
h
Fig. B.19 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb =20 ft, Sc/Sb = 1)
20
15
z, ft
10
L = 20 ft
b U
hen1
φ = 30o U
5 hen3
S /S = 1 U
c b hen10
0
0 0.05 0.1 0.15 0.2
h
Sc/Sb = 1)
The lateral earth pressure developed in the retained soil mass at the end of the
last expansion cycle is shown in Fig. B.21. The top graphic in Fig. 7.10 of chapter
7 presents the lateral earth pressure for the same group of frames found at the end
of the backfill stage.
Fig. B.21 indicates that the lateral earth pressure at the end of the expansion cy-
cle developed in the soil mass retained by the single and 3-bay frames is nearly
equal to the Coulomb’s active earth pressure, and the earth pressure loads pre-
scribed in national building codes. The earth pressure developed behind the 10-
bay frame, on the other hand, is comparable to lateral earth pressure at rest.
Comparing the lateral earth pressure found behind the 10-bay frame with a
length of 200 feet shown in Fig. B.21, to its counterpart developed behind the
290 Appendix B
20-bay frames (with the same length) shown in Fig. B.3 and B.12 earlier, we find
that the pressure shown in the latter two figures is substantially larger for both 20-
bay frames. Additionally, the horizontal wall movements for the latter frames are
substantially smaller than those of the 10-bay frame. These results indicate that the
magnitude of lateral earth pressure developed due to thermal expansion depends
of both the length and the lateral stiffness of the RFERS.
20
σ
hen1
15 σ
hen3
σ
hen10
z, ft
Coulomb
10
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
φ = 30o K0
5
S /S = 1
c b
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
σ , kips/ft
2
h
stage and the last expansion cycle are shown in Fig. B.22 and B.23, respectively.
The bending moments developed in the wall portion of the single and 3-bay
frames are found to be nearly equal for both analysis stages shown in the two fig-
ures below. The bending moment diagram for the wall of the 10-bay frame, on the
other hand, is found altered at the end of the last expansion cycle at the first floor
level. The magnitude of the bending is actually found to decrease by nearly 100%
at the first floor level, while remaining unchanged at the second floor.
The end column horizontal movements are presented in Fig. B.24 and B.25 for
the backfill stage and last expansion cycle, respectively.
The horizontal end column displacements at the end of the backfill stage shown
in Fig. B.24 are in-line with their counterparts found at the other end of the frames
and shown in Fig. B.19. At the end of the last expansion cycle, on the other hand,
the end column horizontal position is found to vary from that of the retaining wall
at the other end of the frame. The position of the end column at the end of the last
expansion cycle of the 10-bay frame, for instance, is found to be similar to the end
column of the 3-bay frame, despite the frame differences.
The end column bending moments developed at the end of the backfill stage
and the last expansion cycle are shown in Fig. B.26 and B.27 respectively.
20
M L = 20 ft
b
bn1
M φ = 30o
15 bn3
M
bn10 S /S = 1
c b
z, ft
10
0
-35 -30 -25 -20 -15 -10 -5 0 5
Fig. B.22 Retaining Wall Moment at Backfill Stage (φ = 30º, All Bays, Lb =20 ft,
Sc/Sb = 1)
20
M L = 20 ft
b
en1
M φ = 30o
15 en3
M
en10 S /S = 1
c b
z, ft
10
0
-35 -30 -25 -20 -15 -10 -5 0 5
Fig. B.23 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb =20 ft,
Sc/S b = 1)
The magnitudes of the bending moment along the height of the end column of
the single bay frame are nearly identical for both analysis stages. On the other
hand, the bending moment in the end column of the 3-bay frame is shown to in-
crease by nearly 40% due to thermal expansion, while the moment in the end col-
umn of the 10-bay frame is shown to increase by nearly 300% at the end of the
expansion cycle.
292 Appendix B
20
15
z, ft
10
L = 20 ft
b
U
hn1
φ = 30o U
5 hn3
S /S = 1 U
c b hn10
0
0 0.05 0.1 0.15 0.2 0.25
h
20
15
z, ft
10
L = 20 ft
b
U
hen1
φ = 30o U
5 hen3
S /S = 1 U
c b hen10
0
0 0.05 0.1 0.15 0.2 0.25
h
20
M
bn1
M
15 bn3
M
bn10
z, ft
10
L = 20 ft
b
φ = 30o
5
S /S = 1
c b
0
-15 -10 -5 0 5 10 15
20
M
en1
M
15 en3
M
en10
z, ft
10
L = 20 ft
b
φ = 30o
5
S /S = 1
c b
0
-15 -10 -5 0 5 10 15
Fig. B.27 End Column Moment at Expansion Cycle (φ = 30º, All Bays, Lb =2 ft, Sc/Sb = 1
Ratio, Sc/Sb, of 4
in Fig. B.28 and B.29 for the initial backfill stage and the last expansion cycle.
The behavior of the retaining walls in this group of frames varies substantially
from their counterpart in the previous group of frames presented. The horizontal
wall movements are found to be substantially smaller at the backfill stage, and the
thermal expansion movements are more pronounced, especially for the 10-bay
frame.
20
15
z, ft
10
L = 20 ft
b
U
hn1
φ = 30o U
5 hn3
S /S = 4 U
c b hn10
0
-0.05 0 0.05 0.1 0.15
h
Fig. B.28 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb =20 ft, Sc/Sb = 4)
294 Appendix B
The lateral earth pressure developed in the retained soil mass at the end of the
last expansion cycle is shown in Fig. B.30. The bottom graphic in Fig. 7.10 of
Chapter 7 presents the lateral earth pressure for the same group of frames found at
the end of the backfill stage.
Fig. B.30 indicates that the lateral earth pressure developed behind the single
bay frame is comparable to the Coulomb’s active earth pressure and the earth
pressure loads prescribed by the national American building codes.
20
15
z, ft
10
L = 20 ft
b
U
hen1
φ = 30o U
5 hen3
S /S = 4 U
c b hen10
0
-0.05 0 0.05 0.1 0.15
h
Sc/Sb = 4)
The lateral earth pressure developed behind the 3-bay frame, on the other hand,
is substantially larger, approaching the magnitude of the lateral earth pressure at
rest for the majority of the height of the wall. This behavior is clearly significantly
different from the retained soil behavior presented for the latter group of frames
with the same bay length but with a column to beam stiffness of 1.
The lateral earth pressure for the 10-bay frame in this group of frames is also
noticeably larger than its counterpart found for the group of frames presented lat-
terly, and is more in line with the lateral earth pressure developed behind the
20-bay frame with 10-ft bays and a column to beam stiffness of 4, shown in
Fig. B.12.
The bending moment diagrams of the retaining wall portion of the RFERS pre-
sented in this section are shown in Fig. B.31 and B.32, at the end of the backfill
stage and the last expansion cycle respectively. The magnitudes of the moments
along the wall height are found to be substantially smaller than their counterpart
found for the frames of lesser column to beam stiffness ratio. The effect of tem-
perature, however, is considerably more pronounced for the stiffer frames, where
the maximum moment at the last expansion cycle for the 10-bay frame is found to
be 11 times higher than its counterpart at the initial backfill stage.
20
σ
hen1
15 σ
hen3
σ
hen10
z, ft
10 Coulomb
L = 20 ft ASCE 7-98
b
BOCA, SBC, IBC
φ = 30 o
5 K0
S /S = 4
c b
0
0 0.4 0.8 1.2 1.6
σ , kips/ft
2
h
20
M
bn1
M
15 bn3
M
bn10
z, ft
10
L = 20 ft
b
φ = 30o
5
S /S = 4
c b
0
-20 -10 0 10 20 30
20
M
en1
M
15 en3
M
en10
z, ft
10
L = 20 ft
b
φ = 30o
5
S /S = 4
c b
0
-20 -10 0 10 20 30
296 Appendix B
20
15
z, ft
10
L = 20 ft
b
U
hn1
φ = 30o U
5 hn3
S /S = 4 U
c b hn10
0
-0.05 -0.025 0 0.025 0.05 0.075 0.1
h
20
15
z, ft
10
L = 20 ft
b
U
hen1
φ = 30o U
5 hen3
S /S = 4 U
c b hen10
0
-0.05 -0.025 0 0.025 0.05 0.075 0.1
h
The end column horizontal movements at the initial backfill stage and the last
expansion cycle are shown in Fig. B.33 and B.34, respectively.
While the horizontal displacements of the end column are considerably smaller
for this group of frames compared to their counterparts presented latterly, the ef-
fect of temperature movements on said displacements is notably more pro-
nounced. The horizontal displacement at top of the end column of the 10-bay
frame at end of the last expansion cycle, for instance, is found to be equal to that
of single bay frame.
The bending moment diagrams of the end columns of the three rigid frames
presented in this section are shown in Fig. B.35 and B.36, for the initial backfill
stage and the last expansion cycle, respectively.
The magnitude of the bending moment in the end columns is found to increase
due to the expansion movements for the majority of the height of the columns.
This is shown to occur for all three RFERS, with increasing degrees depending on
the number of bays. The end column of the 10-bay frame, for instance, is found to
have a maximum moment at the end of the expansion cycle nearly 12 times larger
than its counterpart found at the end of the initial backfill stage.
20
M
bn1
M
15 bn3
M
bn10
z, ft
10
L = 20 ft
b
φ = 30o
5
S /S = 4
c b
0
-20 -10 0 10 20 30 40 50 60
Fig. B.35 End Column Moment at Expansion Cycle (φ = 30º, All Bays, Lb = 20 ft, Sc/Sb = 4)
20
M
en1
M
15 en3
M
en10
z, ft
10
L = 20 ft
b
5
φ = 30o
S /S = 4
c b
0
-10 0 10 20 30 40 50 60
B.1.2 Four Story Rigidly Framed Earth Retaining Structures

The analysis results of four story rigid frames are presented herein, similar to their
counterparts presented earlier.
Ratio, Sc/Sb, of 1
The horizontal movements of the retaining wall portion of the four story RFERS with
a bay length of 10 feet and column to beam stiffness ratio of 1 are presented in Fig.
B.37 and B.38 for the initial backfill stage and the last expansion cycle respectively.
298 Appendix B
40
35
30
U
hn1
25
U
z, ft
hn3
20 U
Lb = 10 ft hn6
15 U
φ = 30o hn10
10 U
hn15
S /S = 1
5 c b U
hn20
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
h
Fig. B.37 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 1)
40
35
30
U
hen1
25
U
z, ft
hen3
20 U
L = 10 ft hen6
b
15 U
φ = 30o hen10
10 U
hen15
S /S = 1
5 c b U
hen20
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
h
Sc/Sb = 1)
The horizontal wall deflections for all six frames are found to be comparable at
the end of both the initial backfill stage and the last expansion cycle, indicating
that the horizontal frame-movements at the wall are not significantly affected by
temperature changes.
The lateral earth pressures developed behind the RFERS at the end of the last
expansion cycle are shown in Fig. B.39. The magnitudes and distributions of the
earth pressures developed behind the six frames are found to vary between the
lower bound pressure determined in accordance with Coulomb’s active earth pres-
sure theory, and in accordance with the soil loads prescribed by BOCA, the SBC
and the IBC, and the lateral earth pressure at rest shown as the upper bound pres-
sure. Two frames only (the 15 and 20-bay frames), however, are found to exceed
the pressure stipulated by ASCE 7-98.
40
35 σ σ
hen1 hen20
30 σ Coulomb
hen3
25 σ ASCE 7-98
hen6
z, ft

hen10
L = 10 ft
b σ K0
15 hen15
φ = 30o
10
5 S /S = 1
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. B.39 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 1)
The bending moment diagrams for the retaining wall are shown in Fig. B.40
and B.41 for the initial backfill stage and the last expansion cycle, respectively.
The magnitude and distribution of the bending moments at both analysis stages are
found comparable, with little noticeable change in moment due to temperature
movements except for the 20-bay frame, which shows a decrease in the moment
magnitude at the first frame level.
Note here that the results presented thus far for the single and 3-bay frames in-
clude only the results of the numerical analysis up to the first expansion cycle,
given that failure of the retained soil mass occurs subsequently. Nevertheless, the
retaining wall horizontal movements presented earlier and the bending moment
diagrams discussed herein for this group of frames indicate that the effect of the
temperature movements at the expansion cycle is negligible on the overall re-
sponse of the restrained end of the RFERS.
40
35 M
bn1
30 M
bn3
25 M
bn6
z, ft
M
20 bn10
M L = 10 ft
bn15 b
15
M φ = 30o
bn20
10
5 S /S = 1
c b
0
-240 -200 -160 -120 -80 -40 0 40 80
300 Appendix B
40
35 M
en1
30 M
en3
25 M
en6
z, ft
M
20 en10
M L = 10 ft
en15 b
15
M φ = 30o
en20
10
5 S /S = 1
c b
0
-240 -200 -160 -120 -80 -40 0 40 80
Fig. B.41 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 1)
The end column horizontal movements at the initial backfill stage and the last
expansion cycle are presented in Fig. B.42 and B.43, respectively.
The results show that the effects of temperature on the deflection of the free
end of the single and 3-bay frames are nearly negligible, but said effects become
more pronounced for the remaining frames.
The horizontal deflection of the end column 10-bay frame, for instance, found
at the end of the last expansion cycle is larger by nearly 20% than its counterpart
found at the end of the initial backfill cycle, while the horizontal deflection of the
free end of the 20-bay frame at the end of the last expansion cycle is more than 1.5
times its counterpart found at the initial backfill stage.
40
35
30
U
hn1
25
U
z, ft
hn3
20 U
L = 10 ft hn6
b
15 U
φ = 30o hn10
10 U
hn15
S /S = 1
5 c b U
hn20
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
h
Fig. B.42 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 1)
40
35
30
U
hen1
25
U
z, ft
hen3
20 U
L = 10 ft hen6
b
15 U
φ = 30o hen10
10 U
hen15
S /S = 1
5 c b U
hen20
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
h
Fig. B.43 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 1)
The bending moment diagrams developed in the end column of the RFERS at
the initial backfill stage and the last expansion cycle are shown in Fig. B.44 and
B.45, respectively.
Similar to its horizontal movements, the bending moment magnitudes of the
end column of the single and 3-bay frames are found to be nearly identical at the
both the initial backfill stage and the last expansion cycle. The magnitude of the
maximum bending moment in the end column of the 15 and 20-bay frames, on the
other hand, are found to increase substantially due to the thermal expansion.
The bending moment at the first level of the end column of the 20-bay frame,
for instance, is found to double at the end of the last expansion cycle compared to
its counterpart found at the initial backfill stage.
40
35 M
bn1
30 M
bn3
25 M
bn6
z, ft
M
20 bn10
M L = 10 ft
bn15 b
15
M φ = 30o
bn20
10
5 S /S = 1
c b
0
-80 -60 -40 -20 0 20 40 60 80
Fig. B.44 End Column Moment at Backfill Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 1)
302 Appendix B
40
35 M
en1
30 M
en3
25 M
en6
z, ft
M
20 en10
M L = 10 ft
en15 b
15
M φ = 30o
en20
10
5 S /S = 1
c b
0
-80 -60 -40 -20 0 20 40 60 80
Ratio, Sc/Sb, of 4
The horizontal movements of the retaining wall portion of the four story RFERS
in Fig. B.46 and B.47 for the initial backfill stage and the last expansion cycle re-
spectively.
The results indicate that the horizontal deflection of the restrained end of the
single, 3 and 6-bay frames is not greatly affected by temperature change. For the
15 and 20-bay frames, however, the retaining wall is found to undergo horizontal
expansion movements into the retained soil mass.
40
L = 10 ft
b
35
φ = 30o
30
U
hn1
25 S /S = 4
c b U
z, ft
hn3
20 U
hn6
15 U
hn10
10 U
hn15
5 U
hn20
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
h
Fig. B.46 Retaining Wall Movements at Backfill Stage (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 4)
40
L = 10 ft
35 b
30
φ = 30o
U
hen1
25 S /S = 4
c b U
z, ft
hen3
20 U
hen6
15 U
hen10
10 U
hen15
5 U
hen20
0
-0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3
h
Sc/Sb = 4)
The lateral earth pressures corresponding to the expansion movements of the

retaining wall shown above are presented in Fig. B.48. Note that in this group of
frames, the results shown for the single bay RFERS are those obtained from the
numerical analysis carried up to the first expansion cycle only, since failure of the
soil mass occurs afterward.
40
σ σ
35 hen1 hen20
σ Coulomb
30 hen3
σ ASCE 7-98
hen6
25
σ BOCA, SBC, IBC
z, ft
hen10
20
L = 10 ft σ K0
b hen15
15
φ = 30 o
10
5 S /S = 4
c b
0
0 0.5 1 1.5 2 2.5 3
σ , kips/ft
2
h
Fig. B.48 Lateral Earth Pressure at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 4)
The earth pressure magnitudes and distribution are found to be bound by Cou-
lomb’s active earth pressure and the lateral earth pressure at rest. The majority of
the frames are found to develop earth pressure loads larger than those prescribed
by the major building codes adopted in the United States, including ASCE 7-98.
304 Appendix B
The bending moment diagrams developed in the retaining wall portion of the
RFERS at the end of the initial backfill stage and the last expansion cycle are cor-
respondingly shown in Fig. B.49 and B.50.
40
35 M
bn1
30 M
bn3
25 M
bn6
z, ft
M
20 bn10
M L = 10 ft
bn15 b
15
M
bn20
φ = 30o
10
5 S /S = 4
c b
0
-120 -80 -40 0 40 80
40
35 M
en1
30 M
en3
25 M
en6
z, ft
M
20 en10
M L = 10 ft
en15 b
15
M
en20
φ = 30o
10
5 S /S = 4
c b
0
-120 -80 -40 0 40 80
Fig. B.50 Retaining Wall Moment at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft,
Sc/Sb = 4)
The effects of temperature movements on the magnitude of the bending mo-

ments are largest at the restrained end of the 10, 15 and 20-bay frames. The mo-
ment in the retaining wall at the first level of the 10-bay frame is increased more
than 10 times due to temperature expansion movements, while its counterpart in
the 15 and 20-bay frames were found to increase by nearly 600% during the last
expansion cycle, compared with the initial backfill stage.
The end column horizontal movements at free end of the RFERS are presented
in Fig. B.51 and B.52 for the initial backfill stage and the last expansion cycle, re-
spectively.
The horizontal deflections of the end columns for the single and 3-bay frame
are found to undergo little change between the initial backfill stage and the last
expansion cycle.
40
35
30
U
hn1
25
U
z, ft
hn3
20 U
L = 10 ft hn6
b
15 U
φ = 30 o
hn10
10 U
hn15
S /S = 4
5 c b U
hn20
0
0 0.05 0.1 0.15 0.2 0.25 0.3
h
Fig. B.51 End Column Movements at Backfill Stage (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 4)
40
35
30
U
hen1
25
U
z, ft
hen3
20 U
L = 10 ft hen6
b
15 U
φ = 30 o hen10
10 U
hen15
5 S /S = 4 U
c b hen20
0
0 0.05 0.1 0.15 0.2 0.25 0.3
h
Fig. B.52 End Column Movements at Expansion Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 4)
On the other hand, the horizontal deflections of the end column of the remain-
ing RFERS are found to increase relatively substantially due to temperature in-
crease. The end column of the 6-bay frame, for instance, is shown to have under-
gone an approximate increase of 45% at the end of the last expansion. The
horizontal deflection of the end column of the 20-bay frame, in contrast, is 4 times
larger at the end of the last expansion cycle than at the initial backfill stage.
The bending moment diagrams in the end column of the RFERS presented in
this group are shown in Fig. B.53 and B.54.
306 Appendix B
The figures indicate, that with the exception of the end column of the single bay
frame, all the end columns developed substantially larger moment magnitudes at
the end of the last expansion cycle compared to those found at the end of the ini-
tial backfill stage. The bending moment at the first level of the end column of the
6-bay frame, for instance, is found to increase nearly 3 times due to temperature
movements, while the bending moment of the end column of the 20-bay frame at
the first level increased by more than 2000%.
The analysis results for the frames with 20-ft-bay frames are omitted since
several unsuccessful numerical simulations lead to a reduced number of frames
analyzed.
40
35 L = 10 ft
b M M
bn1 bn10
30 φ = 30o M M
bn3 bn15
25 S /S = 4 M M
bn6 bn20
c b
z, ft
20
15
10
5
0
-20 0 20 40 60 80 100 120 140
Fig. B.53 End Column Moment at Backfill Cycle (φ = 30º, All Bays, Lb = 10 ft, Sc/Sb = 4)
40
L = 10 ft M M
35 b
en1 en10
30 φ = 30o M M
en3 en15
25 M M
S /S = 4 en6 en20
c b
z, ft
20
15
10
5
0
-20 0 20 40 60 80 100 120 140
References
1. Aboumoussa, W.: On the behavior of rigidly framed earth retaining structures. Ph.D.
Dissertation, Polytechnic Institute of NYU (2009)
2. Aboumoussa, W., Iskander, M.: Thermal movement in concrete: Case study of
multistory underground car park. J. Materials in Civil Engineering, ASCE 56(6),
545–553 (2003)
3. American Concrete Institute (ACI) Committee 224, Joints in concrete construction.
ACI Report 224.3R-95 (1995)
4. American Concrete Institute (ACI) Committee 504, Guide to sealing joints in concrete
structures ACI Report 504R-90 (1990)
5. Arsoy, S., Duncan, J., Barker, M.: Performance of piles supporting integral bridges.
TRR 1808 7, 162–167 (2002)
6. ASCE, ASCE Standard: Minimum Design Loads for Buildings and Other Structures
ASCE/SEI 7-10, Structural Engineering Institute of the American Society of Civil
Engineers (2010)
7. ASCE ACI Committee 445. Recent Approaches to Shear Design of Structural
Concrete. Journal of Structural Engineering, ASCE 124(12), 1375–1417 (1998)
8. Bakir, P., Boduroglu, H.: Mechanical Behavior and Non-Linear Analysis of Short
Beams using Softened Truss and Direct Strut and Tie Models. Engineering Structures,
Elsevier 27, 639–651 (2005)
9. Bang, M., Lee, J.: An Analytical Model for High-rise Wall-frame Building Structures.
In: Proc. Council for Tall Buildings and Urban Habitat, CTBUH 2004, Seoul, Korea,
October 10-13 (2004)
10. Barker, K.J., Carder, D.R.: Performance of Two Integral Bridges forming the A62
Manchester Road Overbridges, Technical Report 436. Transport Research Laboratory,
England (2000)
11. Barker, K.J., Carder, D.R.: Performance of an Integral Bridge over the M1-A1 Link
Road at Bramham Crossroads. Technical report 52, Transport Research Laboratory,
England (2001)
12. Bazant, Z., Cedolin, L.: Stability of Structures: Elastic, Inelastic, Fracture, and Damage
Theories. Dover Publications, Inc. (2003)
13. de Belidor, B.F.: La science des ingenieursdans la conduite des travaux de fortification
et d’architecture civil, Paris (1729)
14. Bell, A.L.: The Lateral Pressure and Resistance of Clays, and the Supporting Power of
Clay Foundations. In: Century of Soil Mechanics, ICE, London, pp. 93–134 (1915)
15. Berwanger, C.: The modulus of concrete and the coefficient of concrete and reinforced
concrete at below normal temperatures. Temperature and Concrete, SP25. American
Concrete Institute (1968)
308 References
16. Bowles, J.E.: Foundation Analysis and Design. McGraw Hill, New York (1988)
17. Briaud, J.: The Pressuremeter. Balkema, Roterdam (1992) ISBN 90-6191-125-7
18. Broms, B., Ingelson, I.: Earth Pressure Against the Abutment of Rigid Frame Bridge.
Geotechnique 21(1), 15–28 (1971)
19. Broms, B., Ingelson, I.: Lateral earth pressure on a bridge abutment. European
Conference on Soil Mechanics and Foundation Engineering 1, 117–123 (1972)
20. Bullet, P.: L’architecturepratique, Paris (1691)
21. Cain, W.: Earth pressure, retaining walls and bins. John Wiley and Sons, New York
(1916), http://books.google.com (last accessed July 3, 2011)
22. Callan, E.: Thermal expansion of aggregates and concrete durability. Journal of the
American Concrete Institute, Part 2 (1952)
23. Carder, D.R., Card, G.B.: Innovative Structural Backfills to Integral Bridge
Abutments, Project Report 52, Research Laboratory, England (1997)
24. Carder, D.R., Hayes, J.P.: Performance under Cyclic Loading of the Foundations of
Integral Bridges, TRL Report 290, Transport Research Laboratory, England (2000)
25. Coulomb, C.A.: Essaisurune application des regles de maximisetminimis a
quelquesproblemes de statique, relatifs a l’architecture. Memoires de Mathematiques et
de Physique presentes a l” Academie Royale de Sciences 7, 343–382 (1776)
26. Couplet, P.: De la poussee des terres contre leurs revestements, et de la force des
revestement qu’on leur doit opposer. Histoire de l’Academie Royale des Sciences
(1726, 106) (1727, 139) (1728, 113) (1726-28), Paris
27. CSI. ETABS Computer Software Technical Manual, Computers & Structures, Inc.
(2009), http://www.csiberkeley.com/
28. CSI, SAP 2000 Computer Software Technical Manual, Computers & Structures, Inc.
(2011), http://www.csiberkeley.com/
29. Dorner, W.: Using Microsoft Excel for Weibull Analysis. Quality Digest (1999),
http://www.qualitydigest.com/jan99/html/body_weibull.html
(last accessed January 11, 2010)
30. Emanuel, J., Hulsey, L.: Prediction of the thermal coefficient of expansion of concrete.
ACI Journal 74(4), 149–155 (1977)
31. Emerson, M.: Thermal movements in concrete bridges: Field measurements and
methods of prediction, ACI SP70. American Concrete Institute (1981)
32. England, G.L., Tsang, N.C.M., Bush, D.I.: Integral Bridges: A fundamental approach
to the time-temperature loading problem. Thomas Telford Ltd. (2000)
33. Fang, Y.S., Che, T.J., Wu, B.F.: Passive earth pressure with various wall movements.
Journal of Geotechnical Engineering, ASCE 1270, 1307–1323 (1994)
34. Gautier, H.: Dissertation sur L’Epaisseur des Culées des Ponts . . . sur L’Effort et al
Pesanteur des Arches . . . et sur les Profiles de Maconnerie qui Doivent Supporter des
Chaussées, des Terrasses, et des Remparts. Cailleau, Paris (1717)
35. Gere, J., Timoshenko, S.: Mechanics of Materials, 2nd edn. PWS Publishers (1984)
36. Girton, D.D., Hawkinson, T.R., Greinman, L.F.: Validation of design
recommendations for integral abutment bridges. Journal of Structural
Engineering 117(7), 2117–2134 (1991)
37. Greinman, L.F., Yang, P.S., Wolde-Tinsae, A.M.: Nonlinear analysis of integral
abutment bridges. Journal of Structural Engineering, ASCE 112(10), 2263–2280
(1986)
38. Heidebrecht, A., Smith, B.: Approximate Analysis of Tall Wall-Frame Structures.
Journal of Structural Engineering Division, ASCE 99(2), 199–221 (1973)
References 309
39. Heidebrecht, A., Smith, B.: Closure to Approximate Analysis of Tall Wall-Frame
Structures. Journal of Structural Engineering Division, ASCE 100(7), 1524–1525
(1974)
40. Heyman, J.: Coulomb’s memoir on statics: An essay in the history of civil engineering.
Imperial College Press, London (1997)
41. Horvath, J.: Integral-Abutment Bridges: A Complex Soil-Structure Interaction
Challenge. GSP (126), 460–469 (2004)
42. Hsu, T.: Unified Approach to Shear Analysis and Design. Cement and Concrete
Composites, Elsevier 20, 419–435 (1998)
43. IBC, International Building Code 2000, International Code Council, Falls Church,
Viginia (2000)
44. Iskander, M., Aboumoussa, W., Gouvin, P.: Instrumentation and monitoring of a
distressed multi-story underground parking garage. Journal of Performance of
Constructed Facilities, ASCE 15(3), 115–123 (2001)
45. Iskander, M., Masood, F., Parikh, S., Dimond, A., Aboumoussa, W.: Closed Form
Expressions For Lateral Deflection of Low Rise Rigidly Framed Concrete Structures.
Canadian Journal of Civil Engineering 39 (2012a), doi:10.1139/L11-104
46. Iskander, M., Dimond, A., Aboumoussa, W., Masood, F.: Approximate Deflection of
Rigidly Framed Earth Retaining Structures Due to an Unknown Earth Pressure
Distribution. International Journal for Numerical and Analytical Methods in
Geomechanics (2012b), doi:10.1002/nag.1025
47. Iskander, M., Parikh, S., Aboumoussa, W.: Apparent Thermal Coefficient of
Expansion of A Concrete Building, with Restraint. ACI J. of Materials 109(1)
(January-February 2012c), Title no. 109-M07
48. Jorgenson, J.L.: Behavior of abutment piles in an integral abutment in response to
bridge movements. Transportation Research Record 903, 72–79 (1983)
49. Ketchum: The Design of Walls, Bins and Grain Elevators, 1st edn. 1907, 2nd edn.
1911, 3rd edn. 1919. McGraw Hill, New York (1907, 1911, 1919)
50. Kunin, J., Alampalli, S.: Integral abutment bridges: Current practices in United States
and Canada. Journal of Performance of Constructed Facilities, ASCE (2000)
51. Lawver, A., French, C., Shield, C.K.: Field performance of integral abutment bridge.
Transportation Research Record 1740, 108–117 (2000)
52. Lehane, B.M., Keogh, D.L., O’Brian, E.J.: Simplified elastic model for restraining
effects of backfill soil on integral bridges. Computers and Structures 73, 303–313
(1999)
53. Mayniel, K.: Traite experimental, analytique et pratique de la poussee des terre et des
murs de revetement, Paris (1808)
54. Meem, J.C.: The bracing of trenches and tunnels. Trans. ASCE LX, 1 (1908)
55. Mehta, P., Monterio, P.: Concrete:Structures, Properties and Materials. Prentice-Hall
Inc., Englewood Cliffs (1993)
56. Miranda, E.: Approximate Seismic Lateral Deformation Demands in Multistory
Buildings. Journal of Structural Engineering, ASCE 125, 417–425 (1999)
57. Miranda, E., Reyes, C.: Approximate Lateral Drift Demands in Multistory Buildings
with Nonuniform Stiffness. Journal of Structural Engineering, ASCE 128, 840–849
(2002)
58. Moulton, H.G.: Earth and rock pressures. Trans. Am. Soc. Min. and Metall. Eng.
(1920)
59. Muller-Breslau, H.: Erddruck auf Stutzmauert. Alfred Kroner, Stuttgart (1906)
310 References
60. Naik, T., Kraus, R., Kumar, R.: Influence of Type of Coarse Aggregate on the
Coefficient of Thermal Expansion of Concrete. J. Materials in Civil Engineering,
ASCE 23(4), 467–472 (2011)
61. National Academy of Sciences, Expansion joints in buildings. Technical Report No.
65, Washington, DC (1974)
62. Ndon, U.J., Bergeson, K.L.: Thermal Expansion of Concretes: Case Study in Iowa.
Journal of Materials in Civil Engineering 7(4), 246–251 (1995)
63. Ohde, J.: Zur Theorie des Erddruckes unter besonderer Berlicksichtigung der Erddruck
verteilung, Die Bautechnik, Germany (1938)
64. Oakdale, DataFit for Windows Version 8.0 User’s Manual, Oakdale Engineering
(2002), http://www.oakdaleengr.com/index.html
65. Portland Cement Association (PCA), Building Movements and Joints, Rep.
EB086.01B, Skokie, Ill (1982)
66. de Prony, R.: Recherchessur la poussee des terres, et sur la forme et les dimensions a
donner aux murs de revetement. Bull. Societe Philomatique (1802)
67. Rankine, W.J.M.: On the stability of loose earth. Phil. Trans. Roy. Soc.
London 147(2), 9–27 (1857)
68. REI, STAAD Pro, Technical Manual, Research Engineers International, Research
Engineers, CA (2002)
69. Rondelet, J.: Traite theorique et pratique de l’art de batir, 5th edn., Paris (1812)
70. Rowe, P.W.: Anchored sheet pile walls. Proc. ICE 1(1), 27–70 (1952)
71. Rutenberg, A., Heidebrecht, A.: Approximate Analysis of Asymmetric Wall-Frame
Structures. Building Science 10, 27–35 (1975)
72. Schanz, T.: Zur modellierung des mechanishen verhaltens von reibungsmaterialen.
Habilitation. Stuttgard University (1998)
73. Schanz, T., Vermeer, P.A., Bonnier, P.G.: The hardening soil model: formulation and
verification. In: Beyond 2000 in Computational Geotechnics: 10 Years of PLAXIS
International; Proceedings of the International Symposium beyond 2000 in
Computational Geotechnics, Amsterdam, The Netherlands, March 18-20. Taylor &
Francis (1999)
74. Sandford, T.C., Elgaaly, M.: Skew effects on backfill pressures at frame bridge
abutments. In: Transportation Research Record TRR 1415, pp. 1–11. National
Academy Press, Washington DC (1993)
75. Smith, B.S., Kuster, M., Hoenderkamp, J.C.D.: A Generalized Approach to the
Deflection Analysis of Braced Frames, Rigid Frame and Couples Wall Structures.
Canadian Journal of Civil Engineering, 230–240 (1981)
76. Smith, B.S., Kuster, M., Hoenderkamp, J.C.D.: Generalized Method for
Estimating Drift in High-Rise Structures. Journal of Structural Engineering,
ASCE 110(7), 1549–1562 (1984)
77. Springman, S.M., Norrish, A.R.M.: Soil-structure interaction: Centrifuge modeling of
integral bridge abutments. In: Proc. of Henderson Colloquium, Toward Joint Free
Bridges, pp. 251–263 (1994)
78. Springman, S.M., Norrish, A.R.M., Ng, C.W.W.: Cyclic Loading of Sand Behind
Integral Bridge Abutment, Technical Report 146, UK Highways Agency (1996)
79. Swaddiwudhipong, S., Lim, Y.B., Lee, S.L.: An Efficient Finite Strip Analysis of
Frame-Shear Wall Tall Building. Computers and Structures 29(6), 1111–1118 (1988)
80. Terzaghi, K.: A Fundamental Fallacy in Earth Pressure Computations. J. Boston
Society of Civil Engineers 23(2), 71–88 (1936)
81. Terzaghi, K.: General wedge theory of earth pressure. Trans. ASCE 106, 68–97 (1941)
References 311
82. Terzaghi, K.: Theoretical Soil Mechanics. Wiley (1943)

83. Thippeswamy, H.K., GangaRao, H.V.S.: Analysis of in- service jointless bridges.
Transportation Research Record, TRR1476, 162–170 (1995)
84. Thomson Jr., T.A.: Passive Earth Pressures behind Integral Bridge Abutments. PhD
thesis, University of Massachusetts Amherst, Massachusetts (1999)
85. Timoshenko, G.J., Stephen, P.: Mechanics of Materials, 2nd edn. PWS Engineering
(1984)
86. Ting, J.M., Faraji, S.: Streamlined Analysis and Design of Integral Abutment Bridges.
Final research report UMTC-97- 13, University of Massachusetts Transportation
Center, Massachusetts (1998)
87. Vardeman, S., Jobe, J.: Basic Engineering Data Collection and Analysis. Brooks/Cole,
Pacific Grove (2001)
88. Wasserman, E.: Jointless bridge decks. AISC Engineering Journal (1987)
89. Williams, A., Clements, S.: Thermal movements in the upper floor of a multi-story car
park. Technical Report 539, Cement and Concrete Association, England (1980)
90. Woltmann, R.: Hydraulische Architectur, vol. 3 (1794)
91. Xu, M., Bloodworth, A.G., Marcus, M.K.L.: Numerical analysis of embedded
abutments of integral bridges. In: IABSE Symposium, Structures for High-Speed
Railway Transportation, Antwerp, vol. 87, pp. 10–111 (2003)
92. Yang, S., Kim, N., Kim, J., Park, J.: Experimental Measurement of Concrete Thermal
Expansion. J. of Eastern Asia Society of Transportation Studies 5, 1035–1048 (2003)
93. Young, W., Budynas, R.: Roark’s Formulas for Stress and Strain, 7th edn. McGraw-Hill
(2004)
94. Zoldners, N.: Thermal properties of concrete under sustained elevated temperatures.
In: Temperature and Concrete SP25, pp. 1–31. American Concrete Institute (1971)
Subject Index
A B
abutment 4, 10–14, 85 backfill 1, 2, 5, 7, 8, 10–14, 49, 69, 86,
accuracy 2, 4, 10, 26, 29, 36, 38, 44, 87, 108–118, 120–122, 125, 126,
50, 57, 75, 89, 97, 107, 225 128–130, 133–137, 139, 140, 143,
Acquisition 56, 57 146, 149, 151–159, 168, 170, 172,
ACTE 73–76, 80, 89, 90, 96, 97 173, 175, 176, 179, 181–185, 187,
Active 2, 3, 133, 135 189–205, 207–214, 216–219,
air 11, 57, 71 221–224, 226–228, 230–232, 235,
algorithm 26 237, 238, 240, 242, 245, 247, 249, 250,
alternating 110, 223 252, 255, 257, 258, 260–265, 267–269,
ambient 57, 71 271–273, 275–277, 279–290, 293, 294,
amplifier 56 296–306
anchor 2, 7, 10, 51, 52, 54, 55, 56, 89 baseline 96
angle 7, 11, 13, 52, 56, 59, 64, 130, basement 3, 45, 47, 59, 64, 86, 108
139, 146, 149, 152, 191, 209, 260, battery 56
264, 271, 280 Bazant 25
annual 5, 73, 74, 75, 76, 90, 97, 102 bearing 13
apparent 5, 64, 71, 74, 90, 173, 204, benchmark 110
258 best-fit 26
approximate 5, 10–14, 16, 18, 20, 22, bolt 52
29, 32, 34, 36, 44, 50, 57–59, 64, 65, borings 87, 109
72, 76, 78, 80, 82, 91, 102, 105, 107, boulders 47, 50, 87, 109, 223
109, 113, 114, 116, 119, 120, 130, braced 1, 2, 3, 7, 14, 16, 125, 151,
152, 165, 168, 176, 179, 183, 208, 153, 226, 232
213, 223, 225, 228, 231, 305, 237, bubble-level 50, 54
240, 242, 250, 252, 255, 258, 263
arc-minutes 54
C
arc-second 54
asymmetric 16, 83, 102, 222
at-rest 9, 14, 281 cables 51, 54
automatic 56, 57, 222 calendar 75
average 10, 12, 38, 71, 74, 76, 90, 110 calibration 18, 20, 22, 26, 27, 38, 42,
axial 12, 117, 128, 129 51, 53, 54, 86, 89, 100, 101
314 Subject Index
cantilever 7, 10, 16, 17, 22, 25, 99, 225 counterfort 7

cap 117 cracked 45, 49, 107
car 45, 108, 221 crackmeters 51
casting 4, 45, 86, 128 cracks 51
cell 36, 38, 47, 64, 87, 109 curvature 22, 50, 168, 240, 252, 261
cement 71 curve-fit 59, 65, 67
centerline 129
centroid 22, 23 D
channels 56
circuit 54, 56 daily 71, 90
clay 3 data acquisition 56, 57
cluster 38, 51, 89 database 57
code 1, 2, 3, 15, 44, 111, 130, 135, 137, datalogger 51, 54, 56, 57, 58, 222
140, 153, 192, 197, 202, 221, 226, deck 4, 10, 12–14
230, 231, 280, 281, 285, 289, 294, 303 dense 10, 12, 109, 223
coefficient 5, 9, 10, 11, 25, 40, 49, 54, densification 104
58, 59, 65, 67, 71, 72, 74–76, 80, 82, density 11, 40
85, 87, 90–102, 105, 111, 225, 258 depth 3, 9, 10, 12, 13, 16, 52, 87
cohesion 8, 9, 109 deviatoric 117
cohesionless 8, 9, 126, 226 diaphragm 7
cold 5, 87, 97, 102, 105 dilatancy 117
collapse 143, 145, 149, 206 dilatometer 50, 87, 109
compaction 11, 12, 49, 87 distress 4, 45, 47, 50, 84, 85, 86, 105,
compressible 12 117, 221, 222
compression 9, 117 distressed 44, 45
concrete 4, 7, 10, 11, 23, 26, 45, 47, 49, diurnal 58, 70
52, 55, 57, 64, 65, 70, 71, 72, 74, 76, drift 4, 16, 21, 22, 38, 39, 40, 44, 45,
85, 86, 90, 101, 107, 108, 110, 126, 50, 86, 99, 100, 102, 105, 242
151, 156, 187, 189, 221, 227 durability 10
confidence 29, 34, 36, 44, 74
confinement 11, 64 E
consistency 111
consistent 81, 102, 142, 145, 235 earth-pressure 8, 10, 12, 14, 50
constant 11, 12, 20, 26, 32, 59, 64, 65, earthquakes 16
97, 268 earth-retaining 1, 4, 40
constitutive 109, 110, 113, 223 eastern 45, 54, 58, 59, 64, 65, 67, 71
constraints 1 efficiency 15, 27, 41, 49
contraction 1, 10, 14, 65, 71, 72, 80, elastic 12, 13, 104, 117, 123, 125, 126,
82, 83, 86, 89, 91, 102, 105, 110, 156, 189, 227
113–116, 118–121, 159, 168, 172, elasticity 18, 22, 23, 26, 117, 133
173, 175, 176, 179, 181–184, 187, elastic-plastic 108–110, 129,
221–223, 228, 235, 238, 240, 250, 156, 223
252, 255, 257, 258, 260, 264, 277 electrolytic 54, 55, 57, 58, 80, 83, 222
corrected 54, 86, 96, 97 elongation 80, 91
correction 25, 54, 96, 97, 98, 99 enclosure 45, 54, 86
coulomb 108, 109, 110, 113, 114, end-column 118, 121, 209
116, 117, 120–123, 126, 129, 130, energy 16, 97, 105
135–139, 141, 143, 145, 148, 153, epoxy 52, 55, 56
156, 189, 192, 197, 206, 211, 215 error 15, 18, 29, 32, 34–36, 38, 39, 44,
223, 226, 227, 229, 231 96, 97
Subject Index 315
excavated 7, 9, 10, 47, 64, 86 four-story 4, 45, 47, 85, 86, 105, 108,
excitation 53 145, 146, 148, 221
expansion 1, 10, 11, 14, 47, 50–52, 58, fourth 45, 50, 56, 58, 104, 118
59, 71–76, 78, 79, 80, 82, 83, 84, 86, fourth-story 45
89–91, 96, 97, 102, 104, 105, 107, frequency 51, 53, 54
110, 111, 113–116, 118–120, 123, friction 8, 9, 11, 49, 130, 139, 146,
159, 168, 170, 172, 173, 175, 176, 149, 152, 191, 209, 260, 264,
179, 181–184, 186, 187, 191–218, 271, 280
221–224, 228, 229, 233, 235, 237,
238, 239, 240, 242, 243, 245, 247, 249, G
250, 252, 255, 257, 258, 260–265,
267–269, 271–273, 275–277, 279–291,
gap 83, 102, 222
293, 294, 296–306
garage 4, 15, 42, 45, 85, 86
experimental 4, 11–13, 221
geometric 27, 42, 44, 71, 89, 108, 225
exponential 29, 44
granular 13, 96
exposure 70, 71, 97, 221
gravel 3, 47, 87, 109, 223
expression 4, 5, 15–20, 22, 23, 25, 26,
gravel-sand 3
39, 40, 44, 86, 99, 100, 133
gravity 1, 7, 8, 10, 127, 156, 190
external 8, 15, 22, 23, 99, 252
F H
factor 20, 22, 25, 26, 42, 51–54, 64, 65, hardening 56, 108–110, 117, 118, 120,
80, 83, 99–101, 129, 222 123, 225
failure 4, 8, 12, 29, 32, 42, 45, 49, 50, heterogeneous 109
84, 85, 123, 205, 211, 222, 299, 303 hillside 1, 45, 85, 221
Fill 11, 42, 47, 55, 86, 87, 96, 102, 104, histogram 29
105, 125, 126, 128, 155–157, 159, homogeneous 18, 22
172, 173, 175, 176, 189–191, 193, hydrostatic 4, 9, 11, 18, 19, 27, 42, 49,
194–199, 201, 202, 221–224, 226, 85, 99–102, 104, 232
227, 228, 230, 232 hyperbolic 117
filter 56 hysteretic 104
fine 47, 86, 87, 109, 223
five 8, 12, 26, 65, 149, 189, 209, 210, I
214, 279, 287
five-story 149 inclination 50, 54, 56, 222
fixed 18, 25, 96, 117 inclinometer 50
fixities 108, 129 infill 42–44
flexibility 2, 10, 29, 125, 129, 153, 226, inflection 165, 242
233 in-situ 50, 87, 109, 135
flexural 16, 38, 44, 86, 99, 113, 116, install 7, 49–59, 64, 65, 67, 69, 71,
125, 128, 187, 213, 240, 258, 275, 277 72, 78, 80, 81, 82, 90, 91 115, 222
footing 13 instrumentation 1, 5, 45, 49, 50, 51,
footprint 45, 47, 86 57, 85, 89, 102, 105, 107, 123, 222
foundation 10, 13, 47, 67, 87, 108, 126, integral 4, 7, 10, 11, 13, 14, 85, 102
156, 189, 227 interaction 1, 3, 4, 5, 10, 11, 13, 80,
four 4, 5, 26, 45, 47, 49, 50, 51, 56, 57, 83, 84, 85, 87, 91, 102, 105, 117,
76, 83, 85, 86, 89, 91, 102, 105, 108, 129, 155, 187, 189, 219, 221, 222,
110, 133, 138, 145, 146, 148, 189, 226, 230, 231
219, 221, 222, 230, 252, 260, 279, interpolation 129
297, 302
316 Subject Index
J north 45, 49, 59, 71, 72, 86, 89, 91, 97,
102, 222
jacket 49, 51 north-south 45, 71, 72, 86, 89, 91, 222
joint 10, 47, 50–53, 57, 58, 71–74, 75,
78, 86, 89, 90, 91, 129, 222 O
jointless 10, 11, 12, 85, 105, 226
jointmeters 51
observation 11, 12, 14, 41, 47, 50, 64,
L 76, 86, 96, 105, 117, 229, 230, 258
obtuse 11
laboratory 13, 71 one-story 45, 86
lag 57 optical 45, 50, 86
lifespan 231 output 36, 54
light-weight 102
linear 5, 12, 13, 18, 32, 34, 39, 40, 59, P
65, 67, 71, 74, 75, 81, 85, 97, 102,
104, 105, 141, 143, 153, 225, 226, parabolic 11, 13, 85
258 parallel 8, 13, 50, 51, 52, 78, 89, 90,
live-load 14 91, 222
load 1, 2, 3, 5, 7, 9, 11–13, 15, 18, 26, parameter 1, 5, 9, 26, 27, 32, 34, 36,
27, 39, 40–42, 44, 45, 50, 64, 65, 86, 38, 39, 41, 57, 102, 113, 125, 126,
99, 107, 110, 117, 125, 127, 128, 128, 153, 155–157, 187, 189, 191,
130, 133–143, 145, 146, 148, 151, 219, 221, 226, 227, 229, 231–233,
153, 156, 190, 192, 197, 202, 206, 235, 279
207, 215, 223, 243, 226, 231–233, parking 4, 15, 42, 45, 47, 86, 108, 117,
252, 261, 262, 289, 294, 298, 303 221
loose 8, 10, 12 passive 8, 10–13, 97, 102, 105, 233
low-rise 15, 18, 44, 225 perpendicular 13, 50, 51, 72, 89, 90,
222
M plain-strain 107
manual 16, 53, 54, 56, 57, 58, 231 plastic 11, 117
masonry 7, 42 polynomial 38
mobilize 2, 3, 126, 133, 137, 226, 228, precision 50, 51, 54, 89
231 pressuremeter 50, 87, 102, 109
modem 56 probable 4, 32, 34, 36, 44, 98, 225
modulus 11, 18, 21–23, 25, 26, 101, proportional 54, 102
108, 111, 133, 223
monitor 1, 5, 11, 14, 45, 49–51, 54, 57, Q
65, 71, 74, 80–83, 85, 89, 91, 105,
107, 113, 119 222, 225 qualitative 81, 113, 118, 120, 126, 184,
monolithic 22, 45 193, 226, 230, 255
multiplexer 56, 222 quantify 71, 80, 90, 126, 226, 231
multistory 4, 14, 128, 219 quarter 40, 165, 222
multivariate 100, 101 queried 57
N R
negligible 3, 12, 14, 109, 212, 231, 263,
299, 300 ramps 47
nonlinear 10, 11, 12, 26, 79, 83, 97, rate 29, 32, 49, 97
100, 222, 233, 258 rebar 52
Subject Index 317
rectangular 45, 86 slender 16, 44

regression 20, 25, 26, 32, 41, 59, 65, soil 1–5, 7–14, 40, 45, 47, 49, 50, 59,
67, 71, 74, 75, 90, 100 64, 65, 69, 71, 78, 79–89, 91, 96, 97,
reinforced 4, 7, 10, 45, 49, 64, 85, 86, 102, 104, 105, 107, 108, 109–111,
108, 110, 187, 221, 240 113–123, 125, 126, 127, 128, 129,
reinforced-concrete 45, 86 130, 133–143, 145, 146, 148, 149,
reinforcement 45, 72 151–153, 155–159, 168, 170, 173,
reinforcing 52 175, 176, 179, 181, 183, 187–192,
relaxation 49, 80, 91 197, 202, 204–206, 209, 211, 212,
relay 56, 222 215, 219, 221–227, 230–233, 235,
reliable 44, 75, 99, 109, 225 238, 240, 242, 243, 245, 249, 252, 255,
remote 56, 57 258, 260, 261, 264, 265, 268, 271, 273,
repair 5, 45, 49, 102 279, 280, 281, 283, 285, 287, 288, 289,
repeatability 54 294, 298, 299, 302, 303
repeated 11, 104, 223 span 11, 13, 14, 16, 40, 43, 67, 151
repose 7 span-to-depth 16
residual 105, 123 spring 11–13, 59, 75, 76, 90, 110
retrofit 49 SPT 109
rock 7, 47, 86, 87, 108, 109, 111, 223 strain 3, 11–14, 18, 23, 59, 64, 67, 71,
rod 96, 97, 98, 102, 104 83, 84, 91, 102, 107, 110, 111, 117,
rod-model 98 123, 125–127, 155, 156, 168, 187,
roof 45, 51, 69, 71, 72, 75, 78, 80, 81, 189, 190, 204, 212, 222, 223, 226,
82, 86, 89, 91, 96, 97, 98, 102 227
rotate 9, 10, 11, 12, 13, 125 stress 2, 4, 5, 10, 12, 13, 38, 49, 83, 84,
102, 105, 110, 113, 116, 155, 157,
S 159, 187, 189, 190, 213, 219,
221–223, 227, 229, 230, 233, 235,
salts 10 240, 258, 275, 277, 279
sand 3, 12, 13, 47, 86, 87, 109, 223 subgrade 11
season 5, 13, 14, 58, 59, 75, 76, 80, 82, summer 13, 59, 75
90, 96, 97, 102, 105, 110, 225 sun 97
segmental 105, 226 sunlight 71, 98
semielliptical 4, 20, 99 survey 4, 13, 45, 50, 86, 107, 118
sensor 50, 51, 54–59, 64, 65, 67, symmetric 29, 44, 45, 173, 179, 184,
69–82, 89–98, 102, 103, 115, 222 258, 260
serviceability 1, 15, 210, 221
shear 1, 8, 9, 12, 16, 18, 19–25, 38, 50, T
64, 66, 71, 86, 97, 99, 101, 117, 128,
129, 151–153, 159, 160, 162, 163, thermal 1, 3, 5, 10–12, 14, 49, 57, 58,
166, 168, 169, 170, 171 173, 174, 65, 66, 71–76, 80, 82, 83, 85, 86,
176, 179, 180, 183, 184, 187, 226, 89–91, 96, 102, 104, 105, 107, 110,
229, 232, 236, 237, 238, 239, 240, 241, 111, 113–117, 122, 123, 155, 159,
242, 245, 247, 250, 252, 255, 258 168, 181, 184, 187, 189, 204, 206,
shear-walls 86 207, 208, 212, 219, 221–224, 226,
silt 3, 47, 87, 109, 223 228–231, 235, 252, 255, 260, 261, 277,
similitude 10 279, 290, 291, 293, 301
skew 11, 12, 29 thick 10, 25, 45, 47, 52, 86, 87, 109,
slab 45, 51, 52, 58, 69, 71, 72, 73, 78, 128, 129, 151
81, 86, 89, 90, 91, 107, 111, 127, three-dimensional 5, 12, 49
156, 190, 223
318 Subject Index
tilt 50, 54, 56, 57, 58, 59, 64, 65, 67, V
69, 70, 71
tiltmeter 50, 54, 57, 80, 83, 222 vibrating 51, 56, 89
time 11, 12, 15, 57, 102 vibrating wire 50, 51, 56, 57, 71, 83,
torsion 45, 50 89, 222
transducer 50–53, 56, 57, 71, 72, 83, voltage 54
89, 222 volumetric 3, 9, 59, 64, 67, 71, 83, 91,
translation 10, 12, 13, 50, 105, 111, 102, 107, 222
125, 128 volumetric-strains 222
trend 38, 59, 64, 65, 80, 81, 82, 96, 97, VWDT 58
102, 104
W
triangular 20, 27, 49, 129, 133, 175,
176, 192, 232 waffle-slab 45, 86, 107
triaxial 13, 117 weight 7, 87, 98, 102
tributary 107, 127, 156, 190, 223 western 54, 59, 64, 65, 67, 71
trigonometric 50 width 39, 40, 101, 107, 127, 134, 135,
twelve 54, 56, 57, 222 156, 190, 223
twenty 26 wind 20, 27, 39
twenty-five 26 wing-wall 13
twist 97 wire 56, 57
U Y
uncertainty 49 yield 4, 10, 11, 13, 16, 36, 38, 39, 40,
uniform 4, 15, 16, 20, 21, 34, 36, 38, 44, 87, 109, 117, 123
39, 41, 43, 97, 99
unrestrained 2, 3, 4, 67, 71, 78, 91, Z
108, 120
zero 19, 39, 40, 56, 59, 80, 91, 129
unyielding 3
zone 87, 102

Regidly Framed Earth Retaining Structures

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Regidly Framed Earth Retaining Structures

Загружено:

Авторское право:

Доступные форматы

Springer Series in Geomechanics & Geoengineering

For further volumes:

Geomechanics deals with the application of the principle of mechanics to geoma-

ISSN 1866-8755 ISSN 1866-8763 (electronic)

Library of Congress Control Number: 2014933215

c Springer-Verlag Berlin Heidelberg 2014

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

The intent of this document is to supply information not render

A large number of professional colleagues contributed formally and informally to

δ si Lateral drift at i’th story

W Total acting force on RFERS

Conversions from American (British) to SI Units

Length 1 in. = 2.54 cm

Area 1 in.2 = 6.452 cm2

Volume 1 in.3 = 16.387 cm3

Stress 1 psi = 6.895 kN/m2

Unit Weight 1 pcf = 0.1572 kN/m3

Moment 1 lb-ft = 1.3558 N.m

Energy 1 ft-lb = 1.3558 J

Conversions from SI Units to American (British) Units

Area 1 m2 = 10.764 ft2

Volume 1 m3 = 35.32 ft3

Stress 1 kN/m2 = 20.885 psf

Unit Weight 1 kN/m3 = 6.361 pcf

Moment 1 N.m = 0.7375 lb-ft

Energy 1 J = 0.7375 ft-lb

Selected Unit Conversions ............................................................................ XIII

About the Authors ........................................................................................ XXIII

1 Introduction to Rigidly Framed Earth Retaining Structures

2 Classical Earth Pressure Theory Related to Framed Structures ....... 7

3 Closed-Form Expressions for Lateral Deflection of Rigid

3.2 Approximate Drift Analysis of Buildings ....................................... 16

4 Case Study of a Full Scale RFERS in Service ...................................... 45

4.3.1.3 Electrolytic Tilt Sensors on Level C .................... 65

5 Relationship between Temperature and Earth Pressure

6 Numerical Analysis of Instrumented RFERS ...................................... 107

6.3.2.1 Description of Analysis Procedure ....................... 113

7 Parametric Study of Earth Pressure behind RFERS at Backfill

8 Analysis of Single Story RFERS Subject to Temperature

8.3 Analysis of Single Story Rigidly Framed Earth Retaining

9 Multi-story RFERS Subject to Temperature Variation ..................... 189

10 Conclusions and Recommendations ..................................................... 221

Appendix A .................................................................................................... 235

Appendix B .................................................................................................... 279

References ..................................................................................................... 307

Subject Index................................................................................................. 313

Walid Aboumoussa, PhD, PE, is an adjunct

Magued Iskander, PhD, PE, F.ASCE is

1.1 Rigidly Framed Earth Retaining Structures (RFERS)

W. Aboumoussa and M. Iskander, Rigidly Framed Earth Retaining Structures, 1

Fig. 1.1 Example of a Rigidly Framed Earth Retaining Structure

1.2 Initial Lateral Earth Pressure

Whether a rigidly framed earth retaining structure is erected behind an existing

Design Lateral Soil Loada

1.3 The Effects of Thermal Movements

1.4 Scope of Contribution to the Behavior of REFRS

1.4.1 Closed Form Expressions for Lateral Deflection of RFERS

1.4.2 Experimental Study

a three-dimensional finite element analysis of the structure with a classical earth

1.4.3 Numerical Analysis

Abstract. The development of earth pressure theory as it relates to rigidly framed

W. Aboumoussa and M. Iskander, Rigidly Framed Earth Retaining Structures, 7

2.2 The Development of Earth Pressure Theory