Вы находитесь на странице: 1из 7


discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/281098625

Structural Connections of Steel Building

Frames under Extreme Loading

Article in Advanced Materials Research · July 2015

DOI: 10.4028/www.scientific.net/AMR.1111.223


3 146

5 authors, including:

Dinu F. Dan Dubina

Polytechnic University of Timisoara Polytechnic University of Timisoara


Calin Neagu Ioan Petran

Polytechnic University of Timisoara Universitatea Tehnica Cluj-Napoca


Some of the authors of this publication are also working on these related projects:

Structural conception and collapse control performance based design of multistory structures under
accidental actions View project

Experimental validation of the response of a full scale frame building subjected to blast load -
FRAMEBLAST View project

All content following this page was uploaded by Ioan Marginean on 13 December 2015.

The user has requested enhancement of the downloaded file.

Advanced Materials Research Vol 1111 (2015) pp 223-228 Submitted: 2015-02-16
© (2015) Trans Tech Publications, Switzerland Revised: 2015-03-30
doi:10.4028/www.scientific.net/AMR.1111.223 Accepted: 2015-03-31

Structural Connections of Steel Building Frames under Extreme


DINU Florea1,2,a*, DUBINA Dan1,2,b, MARGINEAN Ioan1,c NEAGU Calin1,d

and PETRAN Ioan3,e
Politehnica University Timisoara, Dept. of Steel Structures and Structural Mechanics, Romania
Laboratory of Steel Structures, Romanian Academy, Timisoara Branch, Romania
Technical University of Cluj Napoca, Dept. of Structures, Romania
a b c d
florea.dinu@upt.ro, dan.dubina@upt.ro, ioan.marginean@upt.ro, calin.neagu@upt.ro,

Keywords: Steel frame, robustness, T-stub, beam-to-column joint, experiment, catenary, failure

Abstract. Resistance to progressive collapse under extreme loading is a measure of the structural
robustness, and relies primarily on resistance of key elements, continuity between elements and
ductility of elements and their connections. In case some hazards occur simultaneously or
consecutively in a very short period of time, e.g. fire after explosion or impact, the capacity of the
members and connections can be exceeded and this can initiate the progressive collapse of the
structure. The paper presents the results of a research program that focused on the ultimate capacity
of connection macro-components under large deformation demands and different loading rates. The
specimens were extracted from extended end plate bolted beam-to-column connections with
different strength and stiffness ratios to the beams.

Capacity of multi-storey steel frame buildings to resist extreme loading may depend on the
performance of beam-to-column joints to provide continuity across the damaged area, and thus to
allow the development of alternate loads paths (AP). The AP method, with its emphasis on
continuity and ductility, is similar to current seismic design practice ([1]). However, there are
specific problems which need to be considered when key elements, particularly columns, are lost,
i.e. development of the catenary forces in the beams and admissibility criteria to be considered in
the design considering the interaction between axial loads and bending moments. It is therefore of
interest to study if today design procedures ([2]) can provide enough robustness for connections
under extreme loading conditions.
Yang and Tan ([3]) tested experimentally the performance of common types of bolted steel
beam-to-column joints under a central column removal scenario. The results showed that tensile
capacities of beam-to-column joints after undergoing large rotations usually control the failure mode
and the formation of catenary action. Tests performed by Gong [4] on double-angle connections
confirmed that the ductile behaviour required for the development of catenary action should be
assured by robustness design of the connection in accordance with the capacity design. Dubina et al.
[5] also investigated the role of beam-to-column joints in mitigating the progressive collapse of
multi-story steel frame buildings in case of column loss. Experimental and numerical tests on
connection under quasi-static and dynamic loading performed by Rahbari et all [6] showed that the
failure mode of web cleat connections is not influenced by the loading rate due to the flexibility of
the connection. They also recommended to extend the research to other types of connections.
A large experimental program on connection components, joints and sub-assemblies is in
progress within the framework of the CODEC research program at PU Timisoara [7]. Connection
components included weld details and bolted T-stubs, and were tested under cvasi-static and

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans
Tech Publications, www.ttp.net. (ID:,12:58:47)
224 Structural Integrity of Welded Structures XI

dynamic loading, at room and elevated temperatures. The paper presents the results of the
experimental tests on T-stub connections and weld details at room temperature and the validation of
numerical models against experimental data.

Specimens and test setup

The experimental program included tests on bolted T-stubs and weld details. The T-stub
configurations have been designed to fail in mode 1 and 2 [8]. From the different possible
configurations, the following typologies have been selected for the experimental program: T-10-16-
100; T-10-16-120; T-10-16-140; T-12-16-100; T-12-16-120; T-12-16-140. First letter represents the
T-stub, second term represents the thickness of the end plate, followed by the diameter of the bolt,
and then the distance between the bolts, all distances in mm. Bolt class was 10.9 and the steel
grades were S235 for flanges (end-plate) and S355 for webs. The 10 mm webs were welded to the
end plates using a 7 mm throat fillet welded. Weld details were made from two 15 mm thick webs
welded on a 25 mm thick end plate. The width of the webs was reduced near the end plate. W-∆, W-
V and W-Y are the typologies labels, where first letter represents the weld, and the next symbol
represents the type of weld, i.e. fillet weld (∆), single bevel but-weld (V) and respectively double
bevel but-weld (Y). To note that metal active gas (MAG) process and 1.2 mm wire were used for
welding. In order to avoid any influence of the steel material properties, specimens of the same
series were fabricated from the same steel plate. The results of the tensile test on steel plates and
bolts are given in Table 1. The ultimate elongation shows the steel elements have a good ductility.
Web details have been monitored using advanced image correlation in order to obtain the specific
deformations. One specimen of each T-stub configuration and two specimens of each weld details
were tested at room temperature and two strain rates. The low strain rate test (called static) had an
imposed displacement of 0.05mm/sec (suffix C). The high strain rate test (called dynamic) had an
imposed displacement of 10 mm/sec (suffix CS).




a) b) c)
Fig. 1 Bolted T-stub (a), weld detail (b), and testing machine with the specimen (c)

Table 1. Average characteristic values for the steel plates and bolts
fy fu Agt At
N/mm² N/mm² % %
T-stub web, t = 10 mm 390 569 18.7 26.5
t = 10 mm 310 408 22.5 34.7
T-stub flange
t = 12 mm 305 445 23.3 32.7
Weld detail web, t = 15 mm 299 402 15.71 24.11
Weld detail flange, t = 25 mm 261 441 16.43 22.92
Bolt, M16 965 1080 5.0 6.5
Note: fy -yield strength; fu -tensile stress; Agt - total elongation at maximum force; At - total elongation after fracture
Advanced Materials Research Vol. 1111 225

Test results
Fig. 2 shows the force-displacement curves for static and dynamic tests on bolted T-stubs. It may
be seen that the increase of the end plate thickness increases the resistance but reduces the ultimate
deformation capacity. When the distance between bolt rows increases from 100 to 120 mm, there is
a reduction of the resistance, but the deformation capacity increases. When bolt row distance has
been increased to 140 mm, the deformation capacity increases without any further reduction of the
resistance. In all cases the failure is ultimately attained due to the fracture of the bolts, see Fig.3. No
failure of the welding was recorded. For the most elongated specimens (with 140 mm bolt row
distance), some cracks were observed near the welding, in the heat affected zone, see Fig. 4. There
is a small influence of the loading rate, in terms of ultimate resistance, deformation capacity and
failure mode.


Force, kN
Force, kN

100 100 T-12-16-100-C

T-10-16-100-C T-12-16-100-CS
T-10-16-100-CS T-12-16-120-C
T-10-16-120-C 50
50 T-12-16-140-CS
T-10-16-140-C T-12-16-140-C
T-10-16-140-CS T-12-16-140-CS
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Displacement, mm Displacement, mm
Fig. 2. Force-displacement curves for T-stub tests


a) b) c) d)
Fig. 3. T-stub specimens after failure: a) T-10-16-100, b) T-10-16-140 c) T-12-16-100, d) T-12-16-

a) b) c) d) e)
Fig. 4. Views with welding for the specimens with largest deformations: a) T-10-16-140-C; b) T-
12-16-140-C; c) d) and e) fillet welds before test, at the end of static test and at the end of dynamic
226 Structural Integrity of Welded Structures XI

In the weld detail tests, the elongations were measured using as reference the reduced section
length. Fig. 5 presents the force-displacement curves for the weld detail specimens, tested under
static and dynamic loading. None of the specimens failed in weld or in the heat affected zone (Fig.
6), and the failure occurred due to fracture of the web away from the weld. The maximum force is
higher for specimens tested under dynamic loading, and in most cases is accompanied by a small
decrease of ductility.

400 400 400

300 300 300

Foce, kN

Foce, kN

Foce, kN
200 W-∆-C- test1 200 W-V-C- test1 200 W-Y-C- test1
W-∆-C- test2 W-V-C- test2 W-Y-C- test2
100 W-∆-CS- test1 100 W-V-CS- test1 100 W-Y-CS- test1
W-∆-CS- test2 W-V-CS- test2 W-Y-CS- test2
0 0 0
0 20 40 60 0 20 40 60 0 20 40 60
Displacement, mm Displacement, mm Displacement, mm
Fig. 5. Force-displacement curves for weld detail

a) b)
Fig. 6.Weld detail specimens at failure: a) W-∆-C, W-V-C and W-Y-C under static test; b) W-∆-
CS, W-V-CS and W-Y-CS under dynamic test

Numerical analysis
A numerical analysis program has been developed using ABAQUS computer program to validate
the numerical models for T-stub components and weld details (Fig. 8). The geometry of the models
was based on the precise measurements of the specimens that were done before testing. Material
properties (plates and bolts) were in accordance with the results of the experimental tests on
materials (see Table 1). For all components, the type of finite element was solid element C3D8R (8-
node linear brick, reduced integration), and the analysis was dynamic explicit. The mesh of the
elements was done using linear hexahedral elements. The general contact type was used between
elements: the tangential component is defined by frictionless formulation while the normal
component is defined by a “hard” contact pressure-overclosure. Based on the observations of the
failure modes, no special requirements for weld were necessary and the weld was modelled using
the same properties as the base material.

Fig. 7. Numerical FE model of the T-stub and weld detail

Advanced Materials Research Vol. 1111 227

Loading was applied in displacement control, similar with the test. The influence of the strain
rate on the material properties was computed as follows [8]:
f ysr 21 ε fusr 7.4 ε
= 1 + log , = 1+ log
fy fy ε0 fu fy ε0
where : ε = strain rate, ε0 = 10−4
f y , f u = yield and tensile strength in quasi − static conditions, ε0 = 10−4
f ysr , f usr = yield and tensile strength at strain rate ε
When the numerical simulations are compared with the experimental test, the results indicate that
the FE models follow with high accuracy the actual behaviour of the specimen, see Fig. 8. The type
of failure was also very well replicated. Fig. 9 plots comparatively the experimental and numerical
force-displacement curves for T-stub series T-10-16, while Fig. 10 plots the curves for weld detail
series W-Y.

Fig. 8. Specimen T-10-16-120-C, experimental and numerical (the equivalent plastic strain PEEQ)

Foce, kN

Foce, kN

T-10-16-100-C T10-16-100-CS numeric
T-10-16-100-C-numeric T-10-16-120-CS
50 T-10-16-120-C 50 T10-16-120-CS numeric
T-10-16-120-C-numeric T-10-16-140-CS
T-10-16-140-C T10-16-140-CS numeric
T-10-16-140-C-numeric T-10-16-140-C
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Displacement, mm Displacement, mm
a) b)
Fig. 9. Experimental vs numerical force-displacement curves for T-stubs: a) static; b) dynamic

Foce, kN

200 W-Y-C
150 W-Y-C numeric
100 W-Y-CS
50 W-Y-CS numeric
0 10 20 30 40 50
Displacement, mm
Fig. 10. Experimental vs numerical force-displacement curves for weld details

Fig. 11 displays the strain obtained numerically and using VIC 3D system. It may be seen that
there is a good correlation between the FEM results and the data obtained from 3D image
monitoring with respect to specific strain maps.
228 Structural Integrity of Welded Structures XI

44.0mm 40.3mm 0.61
46.0mm 41.5mm 0.64 0.78

0.0 0.0
0.0 0.0mm 5.0mm
2.5mm 5.0mm 0.0
a) b) c) d)
Fig. 11. Numerical model and image correlated data before failure a) vertical displacement W-Y-C
b) maximum strain W-Y-C; c) vertical displacement W-Y-CS; d) maximum strain W-Y-CS
The behaviour of T-stub components and weld details was investigated with respect to their
influence on the capacity of beam-to-column joints to resist the loss of a column. Experimental
results on bolted T-stub specimens showed more ductile configurations can be obtained by reducing
the end plate thickness or by increasing bolt row distance. The loading rate has a reduced impact on
the ultimate capacity and failure mode on both T-stubs and weld details. Numerical models have
been validated against the experimental data obtained in static and dynamic tests. The numerical
models will be used for validation of full scale structural models under different loading conditions.
Partial funding for the research was provided by the Executive Agency for Higher Education,
Research, Development and Innovation Funding, Romania, under grant PCCA 55/2012 (2012-2016)
and by the strategic grant POSDRU/159/1.5/S/137070 (2014) of the Romanian Ministry of
Education, co-financed by the European Social Funds - Investing in People, within the Sectorial
Operational Programme Human Resources Development 2007-2013.
[1] NISTIR 7396, Best Practices for Reducing the Potential for Progressive Collapse in Buildings,
National Institute of Standards and Technology, Oakland, CA, 2007
[2] EN1993-1-8, Eurocode 3, Design of steel structures, Part 1-8: Design of joints, CEN, 2005.
[3] Yang B., Tan K.H., Experimental tests of different types of bolted steel beam–column joints
under a central-column-removal scenario, Engineering Structures, 54, 112–130, 2013.
[4] Gong Y., Ultimate tensile deformation and strength capacities of bolted-angle connections,
Journal of Constructional Steel Research, 100 (2014) 50–59.
[5] Dubina D., Dinu F., Marginean I., Petran I., Collapse prevention design criteria for moment
connections in multi-story steel frames under extreme actions, 4th IRF Conf., Funchal, Portugal,
23-27 June 2013.
[6] Rahbari R., Tyas A., Davison J. B., Stoddart E., Web shear failure of angle-cleat connections
loaded at high rates, Journal of Constructional Steel Research, 103 (2014) 37–48.
[7] Dinu F., Dubina D., and Marginean I., Improving the structural robustness of multi-story steel-
frame buildings, Struct. and Infrastr. Eng., 2014, http://dx.doi.org/10.1080/
[8] Kaneko H., Influence of strain-rate on yield ratio, Kobe Earthquake Damage to Steel Moment
Connections and Suggested Improvement. JSSC Technical Report No.39, 1997.

View publication stats