You are on page 1of 14

Journal of Constructional Steel Research 170 (2020) 106068

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Block shear failure of S275 and S690 steel angles with single-line
bolted connections
Binhui Jiang a,b, Michael C.H. Yam b,c, Ke Ke d,e,⁎, Angus C.C. Lam f, Qingyang Zhao b,d
a
School of Civil Engineering, Central South University, China
b
Department of Building and Real Estate, The Hong Kong Polytechnic University, Hong Kong, China
c
Chinese National Engineering Research Centre for Steel Construction (Hong Kong Branch), The Hong Kong Polytechnic University, Hong Kong, China
d
Hunan Provincial Key Laboratory for Damage Diagnosis of Engineering Structures, Hunan University, Changsha, China
e
Key Laboratory of New Technology for Construction of Cities in Mountain Area, School of Civil Engineering, Chongqing University, Chongqing, China
f
Department of Civil & Environmental Engineering, University of Macau, Macau, SAR, China

a r t i c l e i n f o a b s t r a c t

Article history: Block shear is a common failure mode for bolted-steel shear connections. However, limited research has been
Received 18 September 2019 conducted on the block shear failure of bolted high-strength steel (HSS) angle members. In this study, a test pro-
Received in revised form 28 February 2020 gramme was conducted using eight single-line bolted angle specimens made of grade S690 steel and five S275
Accepted 30 March 2020
counterparts, to examine their block shear behaviour. The test parameters included steel grades, bolt arrange-
Available online 21 April 2020
ments (bolt number, pitch, and edge distance), and connection legs. Eleven of the specimens failed by block
Keywords:
shear, and the remaining two by shear-out. Finite element models were developed to study the behaviour of
High-strength steel the test specimens and to investigate the effects of connection legs on the block shear resistance. Laboratory
Block shear tests and numerical analyses indicated that the block shear resistance increased only slightly with increasing
Test lengths of the angle legs, and the relatively lower ductility of HSS did not significantly influence the block
Finite-element model shear resistance of the HSS angles. It was also found that the predictions made by AISC 360–16 and Eurocode 3
Design method were conservative, whereas CSA S16–14 provided unconservative predictions. The predictions obtained from de-
sign equations proposed in previous studies were also evaluated for their accuracy.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction [11] examined a test database of tension angles and determined the pa-
rameters that affect the block shear resistance most significantly.
Steel sections such as angles, channels, tees, and I-sections are com- Esptein [12] examined the block shear resistance of tension angles
monly used as tension members in buildings and bridges. They are often with various double-line bolt arrangements (e.g. staggered and non-
joined with other structural components (e.g. a gusset plate) via bolted staggered bolted connections) through an experimental programme,
connections. Owing to the high resistance of high-strength (HS) bolts and proposed design recommendations. Cunningham et al. [13] evalu-
[1], a bolted tension member and the connecting component may fail ated the accuracy of AISC design provisions for predicting the block
by block shear, in which a block of the member or the connecting com- shear resistance of tension members using the available test data in
ponent is torn out locally. The block shear resistance of connection ele- North America. Kulak and Grondin [14] assessed the adequacy of design
ments [2–10] and tension members [1,11–16] has attracted specifications from North America, Europe, and Japan for quantifying
considerable interest from the research community. A bolted steel the block shear resistance of steel angles and tees using test results in
angle that fails by block shear is characterised by a combination of frac- the literature, and they developed a modified design model. Topkaya
ture at the tensile plane perpendicular to the axial force and shear fail- [15] developed a series of design expressions for predicting the block
ure (either significant shear yielding or shear rupture) along a shear shear resistance of tension members based on finite element (FE) anal-
plane, as illustrated in Fig. 1. Orbison et al. [1] examined the behaviour ysis results. More recently, Samimi et al. [16] examined the block shear
and strength of tension members composed of steel angles and tees, resistance of hot-rolled steel channel members with staggered bolted
with a focus on their block shear resistance. Madugula and Mohan connections, through an experimental and numerical study. The scope
of all these studies was limited to tension members made of normal
⁎ Corresponding author at: Hunan Provincial Key Laboratory for Damage Diagnosis of
steel (NS) with a nominal yield stress below 460 MPa.
Engineering Structures, Hunan University, Changsha, China; Key Laboratory of New In recent years, the application of high-strength steel (HSS) with a
nominal yield stress equal to or greater than 460 MPa has been

https://doi.org/10.1016/j.jcsr.2020.106068
0143-974X/© 2020 Elsevier Ltd. All rights reserved.
2 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

Fig. 2. Split block shear failure mode.

block shear equations confirmed the improved accuracy of the pro-


posed Eqs. [31–33] over the current design provisions [34–36]. Al-
though the findings from these studies provided insightful
information on the block shear behaviour of steel tension sheets/plates
with higher yield stress, more experimental evidence is needed to fur-
ther validate the proposed design equations and to offer an in-depth un-
derstanding of block shear resistance and behaviour of HSS tension
members, especially those subjected to eccentric loading.
This study has two primary objectives, namely, to provide more ex-
perimental evidence for quantifying the block shear resistance of HSS
bolted angles and to examine the effect of the lower ductility of grade
S690 HSS material on the block shear resistance of bolted tension an-
Fig. 1. A bolted steel tension angle failing in block shear. gles. Thirteen (13) tension angle specimens with single-line bolted con-
nections were tested. The considered parameters included steel grades,
bolt arrangement, and connections legs. A detailed FE model including
attracting attention from the structural engineering community. The the simulation of the damage evolution was developed and validated
use of HSS members can significantly reduce the size and weight of a against the test results. The validated FE model was used to study the
structure, providing significant savings in material/energy consumption block shear behaviour of the test specimens and to confirm the experi-
and construction costs. The load demand on foundations can also be re- mental observations. Further FE analyses covering a broader range of
duced by using a lighter superstructure constructed of HSS members. angle sections and connection legs were conducted to examine their ef-
The attractiveness of HSS construction has stimulated research on HSS fects on the block shear behaviour and resistance. The design equations
material properties [17–20], HSS-based connections/components found in the current design specifications and the literature were
[21–24], and HSS-based structural systems [25–28]. revisited, and their accuracy in predicting the block shear resistance of
Unfortunately, the higher yield stress of HSS is generally realised at HSS angles with single-line bolted connections was examined through
the expense of ductility [29]. For example, the ultimate strain (i.e. the comparisons against the test results.
engineering strain at the tensile strength) of S690 steel material is ap-
proximately 6–8%, and the ratio of the tensile strength to the yield stress
is close to unity [20,21]. The lower ductility of HSS material compared 2. Test programme
with NS material (e.g. grade S275 steel) may not allow a sufficient stress
redistribution to occur in a bolted HSS tension member, and thus may 2.1. Specimens
cause premature failure of a critical section. This phenomenon may be
aggravated when only part of the section is connected (e.g. one leg of A total of 13 bolted angle specimens were tested in the experimental
a single tension angle connected to a gusset plate), causing load eccen- programme. Eight of them were made of grade S690 steel with a nom-
tricities. Studies on the block shear performance and resistance of ten- inal yield stress of 690 MPa. The remaining five specimens were made of
sion members with a higher yield stress than conventional NS have grade S275 steel for comparison purposes. All steel angles were fabri-
been conducted by some researchers. Gross et al. [30] examined the cated from two steel plates joined on one side at right angles, through
block shear behaviour of 10 steel angle specimens with a yield stress groove welding. The average measured thicknesses of the S690 and
of 430 MPa. The potential limitation of the design specifications for eval- S275 plates were 6.03 mm and 5.88 mm, respectively. Each angle spec-
uating the block shear resistance of tension angles with a relatively high imen was connected to a 10-mm gusset plate with Grade 12.9 M22 high
yield stress was confirmed. However, the applicability of the research strength bolts on both ends, as depicted in Fig. 3.
findings [30] to tension angles with an even higher yield stress remains Three angle sections, having dimensions of 85 × 65 × 6, 65 × 65 ×
unclear. Teh and Clements [31] experimentally examined the block 6, and 125 × 65 × 6 (unit: mm), were included in the specimen ma-
shear performance of bolted cold-reduced steel sheets with measured trix, enabling the examination of the influence of the connection
yield stresses over 500 MPa and proposed that an ‘active shear plane' lo- legs. Other parameters considered were the bolt number, pitch
cated between the net shear plane and the gross shear plane indicated (p), and edge distance (e). The latter two are defined in Fig. 3. The
in Fig. 1 be used for computing the shear resistance area. With more specimen configurations are summarised in Table 1. The label of
test data, Teh and Yazici [32] proposed a design equation for determin- each specimen starts with either ‘A’ or ‘B’, referring to S275 or
ing the block shear resistance of bolted cold-reduced steel sheets failing S690 steel, respectively. The number following the first letter indi-
by ‘split block shear’, as schematically shown in Fig. 2. Based on the cates the angle dimensions, with ‘1’, ‘2’, and ‘3’ referring to 85 ×
block shear tests of NS and HSS plates described in the literature, Teh 65 × 6, 65 × 65 × 6, and 125 × 65 × 6 (unit: mm) sections, respec-
and Deierlein [33] proposed a slightly altered design equation for calcu- tively. The following letter ‘S’, ‘L’, or ‘E’ indicates a short-leg, long-
lating the block shear resistance of hot-rolled steel tension members. leg, or equal-leg connection, respectively. The number of bolts at
Comparisons between the available test database and the various each connection is the value following ‘Bt’. The designations ‘e1’
B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 3

50 60 60

Gusset plate

80
D=26

120
a p 100 360

80

5
5
80
1(9)

4 5 6
400

80

e
80

5
80 D1

Centreline
120

Angle specimen
80

3(11)

5
Strain gauge

78
2(10)
8
10

Gusset plate Critical section Mid-length section

Fig. 3. Typical specimen.

Table 1
Specimen configurations, test results and FE predictions.
a
Specimen Dimension (mm) Steel grade Bolt No. e (mm) p (mm) a (mm) D1 (mm) Failure mode Ptest (kN) Ptest/PFE

A1S-Bt2-e1-p1 85 × 65 × 6 S275 2 26.5 50.1 28.7 23.0 BS 120 0.90


A1S-Bt3-e1-p1 85 × 65 × 6 S275 3 26.4 49.9 28.6 23.6 BS 183 1.02
A1L-Bt2-e1-p1 85 × 65 × 6 S275 2 27.2 51.0 29.4 23.7 BS 139 1.03
A1L-Bt2-e2-p1 85 × 65 × 6 S275 2 38.9 51.3 29.7 23.8 S 182 1.02
A1L-Bt3-e1-p1 85 × 65 × 6 S275 3 27.7 50.0 30.2 23.7 BS 204 1.10
Mean 1.01
CoV 0.071
B1S-Bt2-e1-p1 85 × 65 × 6 S690 2 26.5 49.0 30.9 23.7 BS 216 0.99
B1S-Bt2-e1-p2 85 × 65 × 6 S690 2 28.3 60.1 29.8 23.9 BS 224 0.90
B1S-Bt3-e1-p1 85 × 65 × 6 S690 3 29.9 50.2 30.3 23.8 BS 291 0.92
B1L-Bt2-e1-p1 85 × 65 × 6 S690 2 27.4 49.9 31.5 23.7 BS 212 0.94
B1L-Bt2-e2-p1 85 × 65 × 6 S690 2 39.6 49.9 30.5 23.8 S 256 0.90
B1L-Bt3-e1-p1 85 × 65 × 6 S690 3 27.7 50.7 29.9 23.6 BS 282 0.93
B2E-Bt2-e1-p1 65 × 65 × 6 S690 2 28.1 49.9 29.3 23.8 BS 201 0.89
B3L-Bt2-e1-p1 125 × 65 × 6 S690 2 27.4 49.3 30.0 23.9 BS 232 0.99
Mean 0.93
CoV 0.042
a
BS = block shear failure; S = shear-out.

and ‘e2’ correspond to the designed edge distances of 28 mm and 40 The material properties of the steel plates were determined through
mm, respectively, while ‘p1’ and ‘p2’ correspond to the designed tensile coupon tests according to ASTM A370 [37] and are summarised
bolt pitches of 50 mm and 60 mm, respectively. The measured di- in Table 2. The applied load was recorded using a load cell of the testing
mensions are shown in Table 1, and the connected leg is underlined machine, and the engineering strain was obtained using an extensome-
in the angle section. ter with a gauge length of 50 mm mounted on the coupon. It is worth

Table 2
Material properties.

Material Elastic modulus (GPa) Yield stress fy (MPa) Tensile strength fu (MPa) Ultimate strain εu fu/ fy

S690 angle plate 185 640 715 0.061 1.12


S690 gusset plate 195 705 730 0.060 1.04
S275 angle plate 195 310 470 0.163 1.52
4 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

600 S275 900 S690


Engineering stress (MPa)

Engineering stress (MPa)


800
500
700
400 600
500
300
400
200 Coupon A 300 Coupon A
Coupon B 200 Coupon B
100 Coupon C
Coupon C 100
0 0
0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.05 0.10 0.15 0.20
Engineering strain Engineering strain

Stress S275
Stress S690
Tensile
Tensile strength
strength

Strain Strain
Ultimate strain Ultimate strain

Fig. 4. Coupon test results.

noting that although the measured yield stress fy of the S690 steel plate Table 2 and represent typical material properties of S690 and S275
used to fabricate the angle specimens is lower than the nominal value, steels.
the measured ultimate strain and the ratio of the measured tensile
strength to the measured yield stress (fu/fy) are typical for S690 steel. 2.2. Test set-up and loading procedure
Fig. 4 shows the engineering stress–strain responses of the S275 and
S690 steel coupons. The figure also defines “ultimate strain”, which is Fig. 5 illustrates the test setup with a SATEC testing machine. The
the engineering strain at the tensile strength of the material. The aver- gusset plates were connected to the end fixtures through M24 Grade
age tensile strength and ultimate strain of the materials are shown in 12.9 high strength bolts. The end fixtures were gripped by the

Fig. 5. Test arrangement.


B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 5

crossheads of the testing machine, and they were conservatively de- ultimate shear-out strength of a bolted connection can be reached with-
signed to ensure that they would deform elastically during the loading out fracture due to geometric changes in the downstream bolt hole.
of the angle specimens. All the bolted connections were snug-
tightened using a wrench. The applied load and elongation of the test 3.3. Applied load versus strain curves
specimens were measured using the load cell and the built-in displace-
ment transducers of the testing machine. Strain gauges with a gauge Representative strain distributions of the NS specimens (A1L-Bt2-
length of 10 mm were used to monitor the longitudinal strains at the e1-p1 and A1S-Bt2-e1-p1) and the HSS specimens (B1L-Bt2-e1-p1
specimen's critical and middle-length section. The arrangement of the and B1L-Bt3-e1-p1) are shown in Fig. 8. It can be seen that the general
strain gauges is schematically shown in Fig. 3. At the beginning of trends of the strain development of the HSS and NS specimens were
each test, a small preload was applied to align the test specimen and similar. The locations of the numbered strain gauges at both the critical
to reduce the slip in the connection (i.e. the gap between the bolt hole and the mid-length sections of the angle are indicated in Fig. 3. To dem-
and the bolt shank), after which a stroke rate of 0.2 mm per minute onstrate the yielding behaviour of the section, the yield strains derived
was applied to tension the specimen. from the coupon test results of S690 and S275 steel are also indicated
in the figure. In general, owing to the combined effect of shear lag and
secondary bending, a non-uniform strain distribution was recorded for
3. Test results and discussions
both the critical and mid-length sections.
At the critical section, the material around the connected leg was
3.1. General
generally in tension, and tensile yielding was triggered adjacent to the
bolt hole as a result of strain concentration. However, the top edge of
Eleven specimens failed by block shear, as indicated in Table 1 and
the outstanding leg at the critical section (strain gauges 3 and 11) was
shown in Fig. 6a. It is evident from the photographs that the tensile frac-
initially subjected to a slight compression due to secondary bending
ture of each specimen preceded any shear fracture. This observation,
caused by the load eccentricity. As the specimen continued to extend,
which is consistent with the finding of Teh and Clements [31] for bolted
the compression of the outstanding edge was alleviated and gradually
connections in flat sheets, is particularly clear from the photographs of
reversed to tension due to the rotation of the connected end, which re-
specimens A1S-Bt3-e1-p1, A1L-Bt2-e1-p1, B1S-Bt3-e1-p1, and B3L-
duced the load eccentricity. However, the top edge of the outstanding
Bt2-e1-p1 (underlined in Fig. 6a). The remaining two geometrically
leg at the mid-length section (strain gauge 8) generally stayed in com-
identical specimens (A1L-Bt2-e2-p1 and B1L-Bt2-e2-p1), each of
pression until the ultimate load was reached.
which had a block with a relatively low aspect ratio (i.e. the ratio of
the shear plane length to the tensile plane length), were governed by
3.4. Discussions of test results
shear-out failure as shown in Fig. 6b. The expected failure mode was
in line with the contention made by Teh and Clements [31]. However,
As mentioned, the inelastic deformation of the HSS angle specimens
the HSS specimen B1L-Bt2-e2-p1 failed by pure shear-out, while the
was not as pronounced as that of their NS counterparts (Fig. 7). This phe-
normal steel specimen A1L-Bt2-e2-p1 failed by combined shear-out
nomenon was attributed to the different mechanical properties of S690
and end-split, as indicated in Fig. 6b. The end-split failure of the down-
and S275 steels. Specifically, the coupon test results illustrated in
stream bolt hole was associated with a tensile fracture due to normal
Table 2 show that the tensile strength to yield stress ratio of S275 steel
strains perpendicular to the loading direction. The results for the two
was 1.52, whereas this ratio for S690 steel (angle) was only 1.12. In addi-
geometrically identical specimens are consistent with the findings of
tion, the ultimate strain of S275 steel was 16.3%, which was much larger
Teh and Uz [38], in that (a) less ductile steel was unable to redistribute
than that of S690 steel (i.e. 6.1%). In this context, the large ductility and
shear stresses along the shear failure path; and (b) material ductility
the relatively higher tensile strength to yield stress ratio of S275 steel
had a considerably lower effect on tensile failure.
readily allowed the stress redistribution near the ‘shear block’ after the in-
ception of initial yielding. Similar observations were also characterised for
3.2. Applied load versus elongation curves the two specimens failing by shear-out failure. The effect of the edge dis-
tance e on the block shear strength of the angle specimens can be ob-
Fig. 7 presents the load versus elongation curves of all the test spec- served from Table 1. As expected, the block shear strength increased
imens. In general, the figure shows that the elastic deformation range of with increasing edge distance as the net tensile area of the block in-
the HSS angle specimens is wider compared to those of the NS angle creased. However, the failure mode was shifted from a block shear pattern
specimens because of the high yield stress of HSS materials. On the to shear-out failure when the edge distance was increased. The test re-
other hand, the load versus elongation curves of the NS specimens sults given in Table 1 also show that the block shear strength generally in-
show a longer range of inelastic deformation compared to those of the creased with increasing bolt pitch and number of bolts as the shear area of
HSS specimens. Once the ultimate load of each block shear specimen the block increased. However, it should be noted that there is a limit to the
was reached, the resistance dropped abruptly due to the fracture of effect of an increase in bolt pitch, and a continued increase in the bolt
the tensile plane of the block. With further displacement, a plateau pitch will eventually lead to a change from block shear failure to plate
with a slight increase in the applied load was formed (except for speci- bearing failure or net section fracture. Similarly, increasing the number
mens B1S-Bt3-e1-p1, A1L-Bt2-e1-p1, and B3L-Bt2-e1-p1). This behav- of bolts will ultimately lead to a change from block shear failure to net sec-
iour is consistent with the finding of Teh and Uz [39] that full or almost tion fracture. The test results suggest that the variation in the leg lengths
full shear strain hardening was generally achieved at the ultimate limit did not significantly affect the block shear strength of the angle speci-
state of the block shear failure. The present test results are also consistent mens, echoing the findings by Orbison et al. [1] and Topkaya [15]. Due
with the results obtained by Gross et al. [30]. The load-displacement to the limited test data, FE analyses were performed to further examine
graphs plotted in Fig. 7 do not show significant differences in the connec- the effects of the leg configurations, as presented in the next section.
tion ductility between the NS and HSS specimens that failed in block
shear. This result is interesting considering the significant difference in 4. Finite element analysis
the ultimate strain between the two steel grades, as shown in Table 2.
The resistance of the two shear-out specimens (A1L-Bt2-e2-p1 and 4.1. Numerical modelling techniques
B1L-Bt2-e2-p1) only dropped precipitously well after the ultimate limit
state, indicating that fracture took place in the unloading stage. This ob- The commercial package ABAQUS [41] was adopted to model the
servation echoes the finding of a recent investigation [40] that the structural responses of the test specimens for a further interpretation
6 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

Fig. 6. Failure modes of specimens: (a) block shear failure and (b) shear-out failure.
B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 7

250 250
B1S-Bt2-e1-p2
200 200 B3L-Bt2-e1-p1
B1L-Bt2-e1-p1
(kN)

(kN)
LLooaad (kN)
Loaadd (kN)

150 150
B1S-Bt2-e1-p1
100 100 B2E-Bt2-e1-p1
A1L-Bt2-e1-p1
A1S-Bt2-e1-p1
50 50

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30 35
Elongation (mm)
Elongation (mm) Elongation (mm)
Elongation (mm)
(a) (b)

350 350
300 B1S-Bt3-e1-p1 300 B1L-Bt3-e1-p1

250 250
(kN)
(kN)

d (kN)
LLooad (kN)

200 200
Load

150 150
A1L-Bt3-e1-p1
A1S-Bt3-e1-p1
L

100 100
50 50
0 0
0 5 10 15 20 0 5 10 15 20 25 30 35
Elongation (mm)
Elongation (mm) Elongation (mm)
Elongation (mm)
(c) (d)

300
B1L-Bt2-e2-p1
250
200
Load (kN)

150
A1L-Bt2-e2-p1
100
50
0
0 5 10 15 20 25 30
Elongation (mm)
(e)

Fig. 7. Load versus elongation curves of test specimens.

of the block shear test results. A typical FE model is schematically shown the ‘penalty friction’ behaviour was assigned to all contacting surfaces.
in Fig. 9. Eight-node solid elements with reduced integration and hour- For the gusset plate connecting to the end fixture, all the degrees of free-
glass control (C3D8R) were used to model all the major components of dom in the connection were coupled using ‘kinematic coupling’, and
the specimens, i.e. the steel angle, bolts, and gusset plates. To simulate only translational movement along the U3 direction (Fig. 9) was
the interaction among these components, a ‘hard’ normal contact with allowed. Linear elastic springs were adopted to simulate the minor
8 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

180 A1L-Bt2-e1-p1 (Critical section) 180 A1L-Bt2-e1-p1 (Mid-length section)


150 3 150 8 7 6 5
2 10 1
11 4
(kN)
120 9 120

(kN)
LLooaadd(kN)

LLooaad (kN)
90 90
Yield strain Yield strain
60 60 8
3(11)
7
30 1(9) 2(10) 30 4 5 6
0 0
0 1000 2000 3000 4000 5000 -500 0 500 1000 1500 2000 2500
Microstrain
Microstrain Microstrain
Microstrain

(a)
150 A1S-Bt2-e1-p1 (Critical section) 150 A1S-Bt2-e1-p1 (Mid-length section)
3 10
120 11 2 120 8 7 6 5 4
kN)

9
LLooaadd ((kN)

(kN)
90 90

LLooaadd (kN)
1
60 60
Yield strain 3(11) 8
30 30
1(9) 2(10) Yield strain
4 56
7
0 0
0 1000 2000 3000 4000 5000 -500 0 500 1000 1500 2000
Microstrain
Microstrain Microstrain
Microstrain

(b)
250 B1L-Bt2-e1-p1 (Critical section) 250 B1L-Bt2-e1-p1 (Mid-length section)
10 1 8 7 6 5 4
200 3 200
11 9
(kN)

kN)
LLooaadd (kN)

150 150
LoLoaadd ((kN)

2
100 Yield strain 100 Yield strain
3(11) 8
50 50
1(9) 2(10) 4 5 6
7
0 0
0 1000 2000 3000 4000 5000 6000 -1000 0 1000 2000 3000 4000
Microstrain
Microstrain Microstrain
Microstrain

(c)
350 B1L-Bt3-e1-p1 (Critical section) 350 B1L-Bt3-e1-p1 (Mid-length section)
11 2 7
300 3 300 8 6 5 4
1
250 10 250
kN)

9
LoLoaadd ((kN)

kN)
LoLaodad((kN)

200 200
150 Yield strain 150 Yield
100 3(11) 100 strain 8
50 1(9) 2(10) 50
4 5 6
7
0 0
0 2000 4000 6000 -1000 0 1000 2000 3000 4000
Microstrain
Microstrain Microstrain
Microstrain

(d)

Fig. 8. Load versus microstrain curves of test specimens: (a) A1L-Bt2-e1-p1 (b) A1S-Bt2-e1-p1 (c) B1L-Bt2-e1-p1 and (d) BL-Bt3-e1-p1.
B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 9

deformation of the end fixture and the test machine, and the stiffness of Johnson-Cook damage model [43] was used to quantify the relationship
the spring elements was calibrated using the initial stiffness of the test between the fracture strain and stress triaxiality in FE models of angles,
response curves in the linear stage. A ‘dynamic/explicit’ analysis and the following equation was adopted.
governed by a displacement-based control algorithm was used to simu-
−cη
late the pseudo-static loading procedure, and the displacement was ap- ε f ¼ a þ be ; ð3Þ
plied to the other gusset plate as shown in Fig. 9. It is worth noting that
the bolts were snug-tightened in the test preparation, and thus the bolt where εf= fracture strain, and η= stress triaxiality. The material param-
pretension force was insignificant. eters (a, b, and c) were calibrated using an iterative procedure. For S690
An isotropic elastoplastic model governed by the von Mises yield cri- steel, a = 0.038, b = 0.600, and c = 2.500. For S275 steel, a = 0.063, b =
terion was used in the material modules. Engineering stress–strain 0.938, and c = 2.800. The rationale and limitations of using an iterative
curves extracted from the coupon results were converted to true procedure in fracture simulation have been discussed in recent studies
stress–strain responses as the data input. In particular, the engineering [44–48].
stress–strain responses were divided into two stages: response curves
before necking and after necking [42]. The correlation between the 4.2. FE analysis results and further discussions
true stress–strain curves and the engineering curves before necking
can be determined using the following equations: The corresponding test-to-FE ratio determined by the ratio of the ul-
  timate load from the tests (Ptest) to FE prediction (PFE) is summarised in
σ true ¼ σ eng 1 þ εeng ð1Þ Table 1. In summary, the test-to-FE ratio (i.e. the ultimate load from test
  σ true to FE prediction, Ptest/PFE) ranged from 0.89 to 0.99 for the HSS angles,
εptrue ¼ ln 1 þ εeng − ; ð2Þ and a mean value of 0.93 with a coefficient of variation (CoV) of 0.042
E
was obtained. For the NS angle specimens, the Ptest/PFE ratio varied
where σtrue = true stress, σeng = engineering stress, εptrue = true plastic from 0.90 to 1.10. The mean value and corresponding CoV were 1.01
strain, εeng = engineering strain, and E = Young's modulus. Note that and 0.071, respectively. In general, the FE predictions were in satisfac-
the above equations were not valid when the materials progressed to tory agreement with the test results. FE models of selected specimens
the necking stage due to evident non-uniform strain. Thus, a trial-and- (i.e. A1L-Bt2-e1-p1, A1L-Bt3-e1-p1, B1S-Bt2-e1-p1, B1S-Bt2-e1-p2,
error procedure was adopted to determine the true stress versus strain B1S-Bt3-e1-p1, and B1L-Bt3-e1-p1) covering varied parameters (i.e.
responses in the post-necking stage based on the coupon test data. The steel grades, bolt arrangements, and connection legs) were examined
iteration was initiated by adopting the presumption that the true stress- in detail to interpret the behaviour and failure mechanism of the angles.
true strain response is linear after necking [42]. The “Damage for Ductile A numerical study with a broadened spectrum of parameters was sub-
Metals” module in ABAQUS was used to simulate the behaviour deteri- sequently conducted to analyse the effect of the outstanding leg length
oration and fracture of the tension coupons. A threshold value quantify- and the connected leg length.
ing the cracking initiation and a damage evolution index were used to The applied load versus elongation responses of the selected speci-
reproduce the fracture of the coupons. The ‘fracture strain’ quantifies mens were compared with the FE results, as illustrated in Fig. 10. In gen-
the maximum strain at which a cracking of the material is triggered, eral, the overall trend of the test responses was captured by the FE
whilst the ‘damage evolution index’ quantifies the deterioration rate analysis. The damage-initiation processes simulated by the FE models
of the material. A linear method was used to trigger the damage evolu- are also shown in Fig. 10 and are correspondingly indicated in the re-
tion of the material from the inception of cracking to complete stiffness sponse curves. As shown in the figure, it can be reconfirmed that the
loss. To account for the complex stress state in the angle specimens, the block shear failure of the angles was activated by the necking of the ma-
terial near the bolt hole, followed by a rupture of the net tension plane.
According to the FE predictions of the comparison group A1L-Bt2-e1-
p1/A1L-Bt3-e1-p1 and B1S-Bt2-e1-p1/B1S-Bt2-e1-p2, the beneficial ef-
fect of increasing the connection length, which is produced by increas-
ing the bolt number or the pitch, can be confirmed, as the block shear
resistance increases with the pitch and bolt number. In addition, the
FE predictions of group B1S-Bt3-e1-p1/B1L-Bt3-e1-p1 echoed the test
observations that the effect of the connection leg on the block shear re-
sistance was not significant, as the difference in block shear resistance
predicted by the FE models was not pronounced. The typical failure
modes of specimens compared well with those of the FE results, as
shown in Fig. 11.
The stress distributions of model B1S-Bt2-e1-p1 at the ultimate load
were extracted and are shown in Fig. 12. They offer an in-depth
understanding of the stress state in the vicinity of the block of the HSS
angle. The tensile stress distributions over the critical section are
demonstrated in the figure. In general, a non-uniform tensile stress dis-
tribution at the critical section occurred in the outstanding leg; how-
ever, the net tension plane had already developed stresses above the
yield stress of the material, upon reaching the ultimate load. This obser-
vation echoed the strain distribution results in Section 3.3 and findings
reported in the literature [31–33]. The shear stresses along the gross
shear plane and the active shear plane [31–33,49] are plotted in
Fig. 12. The shear stress along the active shear plane at the ultimate
load level shows that shear yielding occurred along the plane. In con-
trast, the shear stress along the gross shear plane was not uniform,
and on average, was lower than that of the active shear plane. These
Fig. 9. Overview of finite element (FE) model. findings were also in line with those drawn from the block shear tests
10 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

Fig. 10. Comparison between test results and FE predictions.

on cold-reduced steel sheets conducted by Teh and Clements [31] and slightly enhanced for both HSS angles. A similar observation was ob-
Teh and Yazici [32]. tained for the NS angles. Thus, the effect of the outstanding leg length
As mentioned previously, the available test data may not be suffi- on the block shear resistance of the tension angles was insignificant,
cient to fully reveal the influence of the outstanding leg and connected reconfirming the test observations and the research findings from pre-
leg length on the behaviour and block shear resistance of tension angles. vious works on NS tension angles [1,15]. Similar observations were ob-
Therefore, an FE analysis including ten (10) FE models of HSS and NS tained in cases with varied connected leg lengths, and the effect of the
tension angles with varied sections and connection legs was performed. connected leg length on the block shear resistance of the tension angles
The information on these FE models is provided in Table 3. Note that the was also insignificant. These phenomena imply that increasing the out-
connection detail was kept identical (i.e. edge distance = 28 mm, pitch standing leg length or the connection leg length does not appreciably af-
distance = 50, bolt number = 2, and end distance = 30 mm) to provide fect the stress distribution around the shear block at ultimate, although
a comparable basis. With the outstanding leg length increased from 90 the connection resistance is slightly enhanced owing to a correspond-
mm to 120 mm and 150 mm, the block shear resistance was only ingly increased cross-section area.
B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 11

Fig. 11. Comparison of failure modes between test results and FE predictions.

5. Design comments plane'. In fact, this split block shear failure mode is similar to those ob-
served from the test specimens of this study. The area of the active
The effectiveness of the design equations in AISC 360–2016 [34], shear plane was computed as the average area of the gross shear
Eurocode 3 [35], and CSA S16–14 [36] for quantifying the block shear ca- plane and the net shear plane. In addition, a modification factor was fur-
pacity of the tension angle specimens was examined. In addition, design ther introduced to the equation accounting for the non-uniform tensile
equations proposed by Topkaya [15], Teh and Yazici [32], and Teh and stress distribution over the net tension plane. More recently, Teh and
Deierlein [33] were revisited. The design equations mentioned above Deierlein [33] developed an updated design equation recognising that
are summarised in Table 4. The design method in AISC 360–2016 [34] the shear yielding in the active shear plane was generally accompanied
assumes that block shear failure is dominated by the fracture of the by a full strain hardening upon ultimate load [33,39], and hence the
net tension plane combined with the shear yielding of the gross shear value of 0.6fy used in Teh and Yazici's Eq. [32] was replaced by 0.6fu in
plane or shear fracture of the net shear plane (Fig. 1). Comparatively, Teh and Deierlein’ Eq. [33].
Eurocode 3 [35] assumes that the tensile rupture of the net tension The test-to-predicted ratios obtained from the design equations
plane combined with the shear yielding of the net shear plane (Fig. 1) mentioned above are summarised in Table 5. Note that the resistance
governs the block shear failure. For the design equation in CSA S16–14 factors and partial factors in the design equations were taken as unity
[36], the contribution of the shear plane is quantified by the product for determining the resistance, and the measured material properties
of the area of the gross shear plane and an equivalent shear stress (i.e. (i.e. yield stress and tensile strength) from coupon tests and measured
the average of the yield stress and the tensile strength of the material). dimensions of the specimens were used. Only specimens failed by
Topkaya [15] presented a single design equation developed based on an block shear were examined. It can be seen that AISC 360–2016 [34]
FE database of tension members. In the equation, it was postulated that and Eurocode 3 [35] both gave conservative predictions for S690 and
the block shear resistance of a tension member was achieved upon the S275 steel angles. Comparatively, CSA S16–14 [36] and the design
rupture of the net tension plane, and concurrently the gross shear plane method proposed by Topkaya [15] produced un-conservative estimates
developed a percentage of yield stress based on a regression analysis. for the block shear capacity of S690 and S275 steel angles. The design
Alternatively, Teh and Yazici [32] proposed a design equation for equation proposed by Teh and Yazici [32] resulted in an average test-
predicting the split block shear (Fig. 2) resistance of cold-reduced sheets to-predicted ratio of 1.00 with a CoV of 0.068 for the S690 steel angles,
in which the block shear failure was assumed to be governed by the rup- and an average test-to-predicted ratio of 1.26 with a CoV of 0.073 was
ture of the net tension plane and shear yielding of an ‘active shear achieved for the S275 steel angles. The equation proposed by Teh and
12 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

Fig. 12. Typical stress distribution in the vicinity of the block.

Deierlein [33] led to an average test-to-predicted ratio of 0.91 with a shear plane strength contribution, provided a good estimate of the
CoV of 0.070 for the S690 steel angles, and an average test-to- block shear capacity of the S690 test specimens. However, the predic-
predicted ratio of 0.92 with a CoV of 0.063 was obtained for the S275 tions by the Teh and Deierlein Eq. [33], which uses 0.6fu to evaluate
counterparts. the shear plane strength contribution, are slightly on the un-
Based on the limited test data above, it appears that the reduced duc- conservative side as the fu/fy ratio of S690 steel is small (i.e. 1.12 from
tility (εu approximately equal to 0.06) of S690 steel does not have a sig- the coupon tests). In contrast, due to a much larger fu/fy ratio of S275
nificant effect on the block shear capacity of the tension angle steel (i.e. 1.52 from coupon tests), Teh and Yazici's Eq. [32] excluding
specimens. The specimens were able to develop shear yielding in the ac- the strain hardening effect of the active shear plane [33,39] led to con-
tive shear plane prior to the tensile fracture of the tension plane. As servative estimates of the block shear resistance of the S275 steel angles.
mentioned, Teh and Yazici's Eq. [32], which uses 0.6fy in assessing the Accordingly, based on the limited test data, it is preliminarily
B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068 13

Table 3 mechanism. The test findings show that the failure was activated by a
Further FE analysis results. rupture of the net tension plane, and a pronounced shear deformation
Model Dimension (mm) Steel grade PFE (kN) of the shear area was observed. A total tear-out of the shear block was
690-S1 90 × 65 × 6 S690 227
observed after the specimens were loaded to complete failure.
690-S2 120 × 65 × 6 S690 229 The primary test parameters included the steel grades, bolt arrange-
690-S3 150 × 65 × 6 S690 231 ment (i.e. pitch distance, bolt number, and side edge distance), and con-
690-L1 90 × 65 × 6 S690 224 nection legs. It was confirmed from the test results that in general the
690-L2 120 × 65 × 6 S690 238
behaviour and failure pattern of the HSS tension angles resembled
690-L3 150 × 65 × 6 S690 241
275-S1 90 × 65 × 6 S275 153 those of the normal steel (NS) angles. The block shear resistance of
275-S2 120 × 65 × 6 S275 155 the tension angles increased with pitch distance, bolt number, and
275-L1 90 × 65 × 6 S275 145 edge distances. In addition, it was seen that increasing the outstanding
275-L2 120 × 65 × 6 S275 152 leg length and varying the connected legs did not produce an evident ef-
fect on the block shear capacities of either the HSS or NS tension angles.
To gain a more comprehensive understanding of the behaviour and
block shear resistance of tension angles, an FE model validated by the
Table 4
Design equations for block shear resistance. test results was developed. The FE analysis of the test specimens pro-
vided a further interpretation of the test responses. To examine the in-
Codes or Design equations
fluence of connection legs on the block shear capacities of tension
reference
angles, an FE analysis was conducted with varied outstanding leg
AISC 360–2016 PAISC = Ubsfu Ant + 0.60fuAnv ≤ UbsfuAnt + 0.60fyAgv lengths and connected leg lengths. It was found that the block shear re-
[34] where Ubs = 1.0 for tension angles with single bolted lines,
sistance of the tension angles was not appreciably influenced by the
fy = yield stress, fu = tensile strength, Ant = net tension area,
Anv = net shear area and Agv = gross shear area. outstanding leg lengths and connected leg lengths.
!
Eurocode 3 Part 1 The test results were compared with the design predictions by using
1.8 [35] P EC3 ¼ f u Ant þ pffiffiffi f y Anv
3 methods documented in major design provisions and those reported in
CSA S16–14 [36] PCAN = UtfuAnt + 0.6Agv(fy + fu)/2 the literature. It was shown that the design equations in AISC 360–2016
where Ut = 0.6 for steel angles connected by one leg. For steel [34] and Eurocode 3 [35] generally produced conservative estimates of
grades fyN 460 MPa, (fy + fu)/2 should be replaced by fy the block shear capacities for both the S275 and S690 steel angle speci-
!
Topkaya [15] f
P Topkaya ¼ f u Ant þ 0:2 þ 0:35 u f y Agv mens. The method in CSA S16–14 [36] and a design equation proposed
fy
by Topkaya [15] overestimated the block shear resistance of the S690
Teh and Yazici PTY = (0.9 + 0.05d/e)fu Ant + 0.6fyAav
[32] where d = bolt shank diameter, e = edge distance; and Aav =
steel angles. The design equation proposed by Teh and Yazici [32] pro-
(Agv + Anv)/2. duced satisfactory predictions of the block shear capacity of S690 steel
Teh and PTD = fu Ant + 0.6fuAev tension angles, while conservative estimates were obtained in the
Deierlein [33] where Aev = (Agv + Anv)/2. cases of S275 steel angles. Based on the test results, it appears that the
reduced ductility (εu approximately equal to 0.06) of S690 steel does
not have a pronounced effect on the block shear capacity of the tension
angles. Accordingly, it is preliminarily recommended to use Teh and
Table 5
Performance of design equations.
Yazici's [32] equation to evaluate the block shear strength of bolted
HSS angles; however, more test data are required to confirm the prelim-
Specimen Ptest/ Ptest/ Ptest/ Ptest/ Ptest/PTopkaya Ptest/ inary recommendation.
PAISC PEC3 PCAN PTY PTD

A1S-Bt2-e1-p1 1.05 1.36 0.90 1.13 0.82 0.84


A1S-Bt3-e1-p1 1.18 1.63 0.91 1.26 0.87 0.90 Declaration of Competing Interest
A1L-Bt2-e1-p1 1.19 1.55 1.02 1.28 0.93 0.95
A1L-Bt3-e1-p1 1.27 1.72 0.99 1.35 0.94 0.97
Mean 1.17 1.57 0.96 1.26 0.89 0.92
There is no financial/personal interest or belief that could affect our
CoV 0.078 0.098 0.062 0.073 0.063 0.063 objectivity. There are no potential conflicts of interest either.
B1S-Bt2-e1-p1 1.21 1.33 0.97 1.06 0.88 0.96
B1S-Bt2-e1-p2 1.07 1.18 0.90 0.96 0.82 0.87 Acknowledgement
B1S-Bt3-e1-p1 1.11 1.23 0.83 0.94 0.77 0.85
B1L-Bt2-e1-p1 1.13 1.24 0.92 1.00 0.83 0.91
B1L-Bt3-e1-p1 1.10 1.23 0.99 0.94 0.76 0.85 The work described in this paper was fully supported by a grant from
B2E-Bt2-e1-p1 1.10 1.20 0.82 0.97 0.80 0.88 the Research Grants Council of the Hong Kong Special Administrative
B3L-Bt2-e1-p1 1.28 1.41 0.89 1.12 0.93 1.02 Region, China (Project No. PolyU 152650/16E). The constructive com-
Mean 1.14 1.26 0.90 1.00 0.83 0.91 ments and contribution from Dr. Lip H Teh of the University of Wollon-
CoV 0.065 0.064 0.071 0.068 0.073 0.070
gong for improving the manuscript are highly appreciated. The
contributions by Mr. Qun He and Mr. Ping Zhang to this work are grate-
fully acknowledged.
recommended that Teh and Yazici's [32] equation be used to evaluate
the block shear strength of bolted HSS angles. More test data are re- References
quired to confirm the preliminary recommendation. [1] J.G. Orbison, M.E. Wagner, W.P. Fritz, Tension plane behavior in single-row bolted
connections subject to block shear, J. Constr. Steel Res. 49 (3) (1999) 225–239.
6. Conclusions [2] J.M. Ricles, J.A. Yura, Strength of double-row bolted-web connections, J. Struct. Eng.
109 (1) (1983) 126–142.
[3] C. Topkaya, Finite element modeling of block shear failure in coped steel beams, J.
A full-scale test programme including 13 specimens was conducted Constr. Steel Res. 63 (4) (2007) 544–553.
to examine the behaviour and block shear resistance of high-strength [4] M.C.H. Yam, Y.C. Zhong, A.C.C. Lam, V.P. Iu, An investigation of the block shear
strength of coped beams with a welded clip angle connection—part I: experimental
steel (HSS) single-line bolted angles with a nominal yield stress of 690
study, J. Constr. Steel Res. 63 (1) (2007) 96–115.
MPa and normal steel (NS) angles with a nominal yield stress of 275 [5] C. Fang, A.C.C. Lam, M.C.H. Yam, K.S. Seak, Block shear strength of coped beams with
MPa. Out of the 13 specimens, 11 failed by a typical block shear single-sided bolted connection, J. Constr. Steel Res. 86 (2013) 153–166.
14 B. Jiang et al. / Journal of Constructional Steel Research 170 (2020) 106068

[6] A.C.C. Lam, C. Fang, M.C.H. Yam, W. Wang, V.P. Iu, Block shear strength and design of [27] X.H. Tian, M.Z. Su, M. Lian, F. Wang, S. Li, Seismic behavior of K-shaped eccentrically
coped beams with double bolt-line connections, Eng. Struct. 100 (2015) 293–307. braced frames with high-strength steel: shaking table testing and FEM analysis, J.
[7] M.C.H. Yam, C. Fang, A.C.C. Lam, J.J.R. Cheng, Local failures of coped steel beams—a Constr. Steel Res. 143 (2018) 250–263.
state-of-the-art review, J. Constr. Steel Res. 102 (2014) 217–232. [28] F. Wang, M.Z. Su, M. Hong, Y.R. Guo, S.H. Li, Cyclic behaviour of Y-shaped eccentri-
[8] S.G. Hardash, R. Bjorhovde, New design criteria for gusset plates in tension, Engl. J. cally braced frames fabricated with high-strength steel composite, J. Constr. Steel
22 (2) (1985) 77–94. Res. 120 (2016) 176–187.
[9] C. Topkaya, Block shear failure of gusset plates with welded connections, Eng. Struct. [29] B.B. He, B. Hu, H.W. Yen, G.J. Cheng, Z.K. Wang, H.W. Luo, M.X. Huang, High disloca-
29 (1) (2007) 11–20. tion density–induced large ductility in deformed and partitioned steels, Science. 357
[10] P.L. Rosenstrauch, M. Sanayei, B.R. Brenner, Capacity analysis of gusset plate connec- (6355) (2017) 1029–1032.
tions using the Whitmore, block shear, global section shear, and finite element [30] J.M. Gross, J.G. Orbison, R.D. Ziemian, Block shear tests in high-strength steel angles,
methods, Eng. Struct. 48 (2013) 543–557. Engl. J. 32 (3) (1995) 117–122.
[11] M.K.S. Madugula, S. Mohan, Angles in eccentric tension, J. Struct. Eng. 114 (10) [31] L.H. Teh, D.D.A. Clements, Block shear capacity of bolted connections in cold-re-
(1988) 2387–2396. duced steel sheets, J. Struct. Eng. 138 (4) (2011) 459–467.
[12] H.I. Epstein, An experimental study of block shear failure of angles in tension, Engl. J. [32] L.H. Teh, V. Yazici, Unconventional block shear failures of bolted connections in cold-
29 (2) (1992) 75–84. reduced steel sheets, Eng. Struct. 56 (2013) 567–571.
[13] T.J. Cunningham, J.G. Orbison, R.D. Ziemian, Assessment of American block shear [33] L.H. Teh, G.G. Deierlein, Effective shear plane model for tearout and block shear fail-
load capacity predictions, J. Constr. Steel Res. 35 (3) (1995) 323–338. ure of bolted connections, Engl. J. 54 (2017) 181–194.
[14] G.L. Kulak, G.Y. Grondin, Block shear failure in steel members–a review of design [34] American Institute of Steel Construction, Specification for Structural steel buildings,
practice, Engl. J. 38 (4) (2001) 199–203. in ANSI/AISC360-16, 2016 , Chicago, America.
[15] C. Topkaya, A finite element parametric study on block shear failure of steel tension [35] European Committee for Standardization, Design of steel structures, Part 1.8: Design
members, J. Constr. Steel Res. 60 (11) (2004) 1615–1635. of joints, in EN 1993-1-8, 2005 , Brussels, Belgium.
[16] M.J. Samimi, J.A. Marnani, S.M.S. Otaghsaraie, S.R.S. Otaghsaraie, Block shear exper- [36] Canadian Standards Association, Design of steel structures, In CSA S16–14, 2014.
imental and numerical studies on hot rolled channel and gusset plate with stag- [37] ASTM, A370, Standard Test Methods and Definitions for Mechanical Testing of Steel
gered bolted connection, Thin-Walled Struct. 108 (2016) 153–162. Products, American Society for Testing and Material, Philadelphia, PA, 2002.
[17] Y.B. Wang, G.Q. Li, W. Cui, S.W. Chen, F.F. Sun, Experimental investigation and [38] L.H. Teh, M.E. Uz, Effect of loading direction on the bearing capacity of cold-reduced
modeling of cyclic behavior of high strength steel, J. Constr. Steel Res. 104 (2015) steel sheets, J. Struct. Eng. 140 (2014), 06014005.
37–48. [39] L.H. Teh, M.E. Uz, Block shear failure planes of bolted connections-direct experimen-
[18] G. Shi, M. Wang, Y. Bai, F. Wang, Y.J. Shi, Y.Q. Wang, Experimental and modeling tal verifications, J. Constr. Steel Res. 111 (2015) 70–74.
study of high-strength structural steel under cyclic loading, Eng. Struct. 37 (2012) [40] M.D. Elliott, L.H. Teh, A. Ahmed, Behaviour and strength of bolted connections failing
1–13. in shear, J. Constr. Steel Res. 153 (2019) 320–329.
[19] L.T. Hai, F.F. Sun, C. Zhao, G.Q. Li, Y.B. Wang, Experimental cyclic behavior and con- [41] ABAQUS Analysis User's Manual. ABAQUS Standard, Version 6.12; 2012.
stitutive modeling of high strength structural steels, Constr. Build. Mater. 189 [42] L.J. Jia, H. Kuwamura, Ductile fracture simulation of structural steels under mono-
(2018) 1264–1285. tonic tension, J. Struct. Eng. 140 (5) (2013), 04013115.
[20] H.C. Ho, X. Liu, K.F. Chung, A.Y. Elghazouli, M. Xiao, Hysteretic behaviour of high [43] G.R. Johnson, W.H. Cook, Fracture characteristics of three metals subjected to vari-
strength S690 steel materials under low cycle high strain tests, Eng. Struct. 165 ous strains, strain rates, temperatures and pressures, Eng. Fract. Mech. 21 (1985)
(2018) 222–236. 31–48.
[21] K. Ke, Y.H. Xiong, M.C.H. Yam, A.C.C. Lam, K.F. Chung, Shear lag effect on ultimate [44] L.J. Jia, H. Kuwamura, Ductile fracture model for structural steel under cyclic large
tensile capacity of high strength steel angles, J. Constr. Steel Res. 145 (2018) strain loading, J. Constr. Steel Res. 106 (2015) 110–121.
300–314. [45] F.F. Liao, W. Wang, Y.Y. Chen, Ductile fracture prediction for welded steel connec-
[22] A.D.C. Sousa, A. Nussbaumer, Multiaxial ultra low cycle fatigue in welded high tions under monotonic loading based on micromechanical fracture criteria, Eng.
strength steel structural components, J. Constr. Steel Res. 153 (2019) 473–482. Struct. 94 (2015) 16–28.
[23] Y.B. Wang, G.Q. Li, W. Cui, S.W. Chen, Seismic behavior of high strength steel welded [46] L. Li, W. Wang, Y.Y. Chen, Y. Lu, Effect of beam web bolt arrangement on catenary
beam-column members, J. Constr. Steel Res. 102 (2014) 245–255. behaviour of moment connections, J. Constr. Steel Res. 104 (2015) 22–36.
[24] R. Pucinotti, N. Tondini, G. Zanon, O.S. Bursi, Tests and model calibration of high- [47] F. Sadek, J.A. Main, H.S. Lew, Y.H. Bao, Testing and analysis of steel and concrete
strength steel tubular beam-to-column and column-base composite joints for mo- beam-column assemblies under a column removal scenario, J. Struct. Eng. 137 (9)
ment-resisting structures, Earthq. Eng. Struct. Dyn. 44 (9) (2015) 1471–1493. (2011) 881–892.
[25] Y.Y. Chen, K. Ke, Seismic performance of high-strength-steel frame equipped with [48] B. Yang, K.H. Tan, Numerical analyses of steel beam-column joints subjected to cat-
sacrificial beams of non-compact sections in energy dissipation bays, Thin-Walled enary action, J. Constr. Steel Res. 70 (70) (2012) 1–11.
Struct. 139 (2019) 169–185. [49] B.V. Mai, C.H. Pham, G.J. Hancock, G.D. Nguyen, Block shear strength and behaviour
[26] K. Ke, M.C.H. Yam, L. Deng, Q.Y. Zhao, A modified DEB procedure for estimating seis- of cold-reduced G450 steel bolted connections using DIC, J. Constr. Steel Res. 157
mic demands of multi-mode-sensitive damage-control HSSF-EDBs, J. Constr. Steel (2019) 151–160.
Res. 150 (2018) 329–345.

You might also like