Professional Documents
Culture Documents
Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct
a r t i c l e i n f o a b s t r a c t
Article history: Six one-third scale reinforced concrete (RC) beam-column substructures, each with different design
Received 20 June 2011 detail, span length and span aspect ratio, were designed and tested to investigate the dynamic load redis-
Revised 23 April 2012 tribution performance of RC substructures following predefined initial damage. The initial damage was
Accepted 24 April 2012
caused by the sudden removal of the corner support. The tested variables include: the longitudinal
Available online 31 May 2012
and transverse reinforcement ratio in the beams, columns and joints, design span length and span aspect
ratio. Extensive instrumentation was installed in predetermined locations on the tested specimens prior
Keywords:
to conducting the tests, recording data such as acceleration, velocity and displacement distribution. The
Dynamic performance
RC beam-column substructures
histories of the bending moments, as well as the vertical and horizontal reaction forces at the fixed sup-
Loss of a corner column ports were also noted. The test results indicated that the column removal apparatus was effective and
Progressive collapse confirmed that the design span length significantly affected the capability of the frames to resist progres-
sive collapse. Furthermore, it was demonstrated that seismically detailed specimens could mount a more
robust performance in resisting progressive collapse.
Ó 2012 Elsevier Ltd. All rights reserved.
0141-0296/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.engstruct.2012.04.016
Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167 155
Nomenclature
[13] and Yi et al. [16] indicated that the upper and lower floors Six one-third scale substructures of varying detailing and
operated in tandem as a unit, given that the dimensions and rein- dimensions were designed and tested at NTU, Singapore to inves-
forcement details of both members were identical. Thus, the tigate the dynamic performance of three-dimensional (3D) beam-
behavior of multi-storey frames could be simplified to that of a sin- column substructures following the removal of a ground corner
gle-storey structure under the proper boundary conditions. support. It should be emphasized that the results presented in this
paper are limited to experimental data based on an extensive
instrumentation plan. Moreover, the slab and in-fill wall effects
on the progressive collapse performance of the RC frame was not
included in this study although Kai and Li [17] has indicated that
RC slab could increase the resistant capacity up by 63% and Tsai
and Huang [18] has concluded that the in-fill walls could signifi-
cantly reduce the inelastic displacement.
1. Chain block
7 R2 R1
10 7 8
6
D
12
13
DF3, and DF4 were subjected to full service loads but wind and
snow loads were ignored. The load combination is 1.2DL + 0.5LL,
(V)
where DL is the dead load and LL is the live load. Specimens DF1
Steel Box and DF2 were subjected to a reduced load combination of
0.9 (1.2DL + 0.5LL). The calculation process of the design axial
Z
force in the corner support of each specimen is given in Appendix
A.
Corner Stub Fig. 5 illustrates the typical reinforcement layout of Specimens
0 DF1 and DF2. The concrete cover of the beam and column was
X 10 and 20 mm respectively. For DF2, the transverse reinforcements
were provided in the joint region using hoop stirrups with 135°
Fig. 4. The detailing of steel assembly.
bends. For the remaining specimens, non-seismic detailing was
provided while transverse reinforcements were provided using
adjacent columns to simulate the fixed boundary conditions as hoop stirrups with 90° bends. In addition, no transverse reinforce-
provided by the surrounding structural elements. In Component ment was installed in the joint region. It should be noted that dou-
2, a column removal apparatus comprising a specially designed bly continuous longitudinal rebar were installed in the beam as
steel column, a pin support and a load cell (Item 6 in Fig. 2, details scaled specimens were tested in the current study. The anchorage
shown in Fig. 3) was designed to simulate the sudden removal of a failure of the bottom longitudinal reinforcement at the beam end
Table 1
Specimen properties (unit: mm).
Specimen ID Elements Longitudinal rebar Transverse reinforcement Load case Measured axial force (kN)
Beam-T Beam-L Beam-T Beam-L Joint Beam-T Beam-L
Control specimen DF1 (F3) Type aa Type aa 4-T10 4-T10 None R6@180 R6@180 0.9 DoD[2] 16.9
Seismically detailed specimen DF2 (F2) Type aa Type aa 4-T13 4-T13 R6@55 R6@60 R6@60 0.9 DoD[2] 16.9
High service load specimen DF3 (F3) Type aa Type aa 4-T10 4-T10 None R6@180 R6@180 1.0 DoD[2] 18.7
Modified detailed specimen DF4 (F4) Type aa Type aa 4-T10 4-T10 None R6@80 R6@80 1.0 DoD[2] 18.8
Long span specimen DF5 (F5) Type bb Type bb 4-T10 4-T10 None R6@160 R6@160 0.8 DoD[2] 23.3
Unequal span specimen DF6 (F6) Type aa Type bb 4-T10 4-T10 None R6@180 R6@160 1.0 DoD[2] 23.2
Beam-L = Longitudinal beam Beam-T = Transverse beam; 0.9 DoD [2] = 0.9(1.2DL + 0.5LL)
a
Clear span = 2175 mm, cross-section = 180 100.
b
Clear span = 2775 mm, cross-section = 240 100.
Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167 157
Table 2
Dynamic strain meter channel label for each specimen.
R6@180 R10@55 CH No. DF1 and DF2 DF3, DF4 and DF5 DF6
4T10 4T16 1 Reaction force D1 D1
Detail A-A Detail B-B 2 D1 BSTL1 BSTL1
3 D2 BSBL1 BSBL1
D4&A4
4 D3 BSTL8 BSTL8
5 D4 BSBL8 VR1-T
R6@60 6 BSTL1 VR1-T⁄ VR2-T
Bar anchorage 7 BSBL1 VR2-T⁄ HR1-T
D3&A3
T10=Deformed bar of 10 mm diameter Note: VR1-T⁄ and VR2-T⁄ were used to measure the vertical reaction of the trans-
T13=Deformed bar of 13 mm diameter verse support.
Bar anchorage T16=Deformed bar of 16 mm diameter HR1-T⁄ was utilized to measure the horizontal reaction in the transverse fixed
detail 1
support.
D1, D2, D3, and D4 were the LVDTs distributed along the transverse beam (see
D1&A1
DF1 and DF2. For the remaining specimens, only one LVDT with asymmetrical failure modes in the longitudinal and transverse
300 mm travel was installed to measure the response of vertical beams of DF1 were observed. Severe diagonal shear cracks and
displacement in the corner column. Two compression/tension load slight concrete spalling occurred in the transverse beam end near
cells (Items 8 and 10 in Fig. 2) were installed vertically to measure to the fixed support (BENF) while narrow diagonal shear cracks oc-
the vertical reaction and to determine the moments on the fixed curred in the longitudinal BENF. Hairline flexural cracks were ob-
supports. One compression/tension load cell (Item 9 in Fig. 2) served in the BENC while symmetric hairline shear cracks were
was mounted horizontally to measure the horizontal reaction force observed in the corner joint. This indicated that the direction of
at the fixed support. In total, 12 strain gauges were mounted on the the bending moment in the BENC, which was initially negative
reinforcement at strategic locations before casting. However, only (tension at the top) under gravity loading, was changed after re-
the strain gauges listed in Table 2 were monitored during the dy- moval of the corner support.
namic test. The locations of the accelerometers, LVDTs and strain For DF2, only some flexural cracks were observed in the BENFs
gauges are illustrated in Fig. 5. (refer to Fig. 7b). Similar to DF1, the direction of the bending mo-
ment was changed in the BENC. However, no cracks occurred in
the corner joint of DF2. It should be emphasized that DF1 and
3. Test observations and results
DF2 were subjected to a reduced service load of 0.9(1.2DL + 0.5LL)
before removal of the corner support.
A total of six 3D beam-column substructures, each of different
DF3 had similar dimensions and reinforcement details as DF1.
design detailing and span length, were constructed and tested to
However, a full service load 1.0(1.2DL + 0.5LL) was applied. As pre-
evaluate the dynamic redistribution performance of the RC frame
sented in Fig. 7c, severe flexural and shear cracks were observed in
when subjected to the loss of a ground corner column. The test re-
the BENFs. Slight concrete crushing occurred in the bottom of the
sults of the six specimens are presented below.
BENF. More flexural cracks developed in the BENCs and more se-
vere diagonal shear cracks were observed in the corner joint of
3.1. Evaluation of the efficiency of the designed column removal DF3 compared to DF1. Although more severe damage took place
apparatus in DF3 due to the higher service load, both specimens managed
to survive the tests. DF4 had a higher transverse reinforcement ra-
The column removal apparatus designed in the current study tio in the potential plastic hinge zone of the beam compared to
involved impacting down by a heavy hammer to simulate the sud- DF3. In general, the crack patterns of DF4 were similar to those
den removal of the column by extreme loading. As the duration of of DF3. The crack width in the BENFs and corner joint was slightly
corner support removal will significantly affect the accuracy of the less than that in DF3. DF5 had a longer design span length than that
dynamic response of the substructures, the effectiveness of the de- of DF3. The dimensions and reinforcement details are given in
signed apparatus must be ensured. The history of the axial force in Table 1. DF5 was subjected to a reduced service load of
the corner column was monitored by a load cell (Item 4 in Fig. 3). 0.8(1.2DL + 0.5LL) before removal of the corner support. As dis-
As illustrated in Fig. 6, initially, the corner column of DF1 and DF2 played in Fig. 7e, DF5 suffered severe damage after removal of
was subjected to an axial force of 16.9 kN (negative represents the corner support. Extremely wide shear cracks and flexural
compressive force). The axial force started to release (compressive cracks were observed in the BENFs while significant spalling was
force begin to reduce) at 0.2 s. The column axial forces in DF1 and observed in the corner joint. The damage in the BENCs was negli-
DF2 were totally released by 0.2035 s and 0.2030 s, respectively. gible relative to that in the BENFs. It should be pointed out that
Thus, the duration of the force release was 0.0035 s and 0.0030 s the collapse of DF5 was stopped by the pin support of the column
for DF1 and DF2, respectively. This provided proof that the de- removal apparatus when the maximum displacement exceeded
signed apparatus satisfied the requirement of the guideline DoD the allowed displacement of about 360 mm. It can be predicted
[2], which requires the duration of the column removal to be less that in the absence of the pin support, DF5 would have collapsed.
than one-tenth of the natural period of the vertical motion of the DF6 had unequal spans in the longitudinal and transverse direc-
structure after losing the column (the measured nature period of tions. The dimensions and reinforcement details are given in Table
the vibration is about 0.15 s for DF1 and DF2). 1. DF6 was subjected to a full service load of 1.0(1.2DL + 0.5LL) be-
fore removal of the corner support. As shown in Fig. 7f, DF6 col-
3.2. Crack pattern and failure mode lapsed following the removal of the corner support. However,
asymmetrical damages were observed in the longitudinal and
The crack pattern and failure mode of each specimen after the transverse beams. Similar to DF5, extreme damage was concen-
dynamic tests are illustrated in Fig. 7. As presented in Fig. 7a, trated in the transverse BENF. However, the damages occurring
in the longitudinal BENF of DF6 were much milder. As illustrated
0
DF1 in Fig. 7f, severe cracks were observed in the corner joint; these
Axial force in the corner column (kN)
DF2 cracks were not symmetrical due to the beams being of unequal
-3
spans.
-6
3.3. Global displacements
-9
In order to estimate the global displacement responses of the
tested specimens, several LVDTs were placed on the specimens.
-12
Fig. 5 graphically depicts the locations of the LVDTs. One LVDT with
300 mm travel was placed in the center of the corner joint for all
-15
specimens. Three additional LVDTs were placed at the quarter,
-16.90 kN
middle and three-quarter spans of the transverse beams of DF1
-18
0.190 0.195 0.200 0.205 0.210 0.215 0.220 0.225 0.230
and DF2 to monitor the displacement distribution along the beams.
Fig. 8 shows the vertical displacement history of the corner joint of
Time (s)
each specimen. Negative values represent downward displace-
Fig. 6. Recorded reaction force history in the corner column during test. ments. The column was removed at time 0.2 s for all specimens.
Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167 159
(a) (d)
Specimen DF2 Specimen DF5
(b) (e)
Specimen DF3 Specimen DF6
(c) (f)
Fig. 7. Cracking patterns and failure mode of each specimen.
For DF1, the first peak vertical displacement was 83.9 mm at addition, vertical motions of both specimens were stopped by the
0.38 s. The downward displacement was reduced to 79.3 mm pin support of the apparatus and the measured distance from the
by 0.48 s but subsequently rose back to 82.6 mm at 0.55 s. The bottom face of the corner column to the pin support was about
oscillation was almost eliminated by 1.5 s and the substructure 360 mm.
had a permanent downward displacement of 81.3 mm, which
was 96.9% of the first peak displacement. The first peak displace- 3.4. Displacement distribution
ments measured in DF2, DF3 and DF4 were 29.0, 146.6 and
99.4 mm, respectively. It should be noted that the peak displace- Fig. 9 illustrates the vertical displacement distribution along the
ments of DF5 and DF6 exceeded 300 mm, which was the mea- transverse beam of DF1. As can be seen from the figure, the maxi-
surement capacity of the LVDT placed in the corner joint. In mum displacements at D1, D2, D3 and D4 of DF1 were 83.9,
73.4, 46.8 and 20.0 mm, respectively. This indicated that the
distribution of the deformation was non-linear and the deforma-
50 tion of the beam occurred in a manner different from that in a can-
tilever beam due to the partial rotational constraint applied to the
Vertical Displacement (mm)
0
corner joint. As mentioned in the above section, the column was
-50
removed at a time of 0.2 s. However, the movement at the third-
quarter of the beam span only commenced at 0.235 s. From this
-100 it can be reasoned that the dynamic force took at least 0.035 s to
propagate to that point. DF2 had a similar displacement distribu-
-150 DF1
DF2 tion response as DF1; the maximum displacements at D1, D2, D3,
DF3 and D4 of DF2 were 29.0, 26.0, 17.0 and 4.9 mm, respec-
-200 DF4
DF5 tively. A similar initial time lag of 0.037 s at D4 was also observed.
DF6
-250
0.0 0.5 1.0 1.5
3.5. Acceleration response
Time (s)
Fig. 8. Recorded history of the vertical displacement of the corner joint of each In progressive collapse analysis, it is prudent to understand that
specimen. it is a dynamic event, and that the sudden removal of an element in
160 Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167
-10 DF1-D3 vibration along an inclined line. It indicated that large noise existed
DF1-D4
-20 in A1 of DF2. DF3 and DF4 displayed similar behavior to the first
-30 two specimens. However, no reliable data was recorded for DF5
-40 and DF6 as their collapse was stopped by the pin support of the
-50 apparatus, bringing about large random noises in the acceleration
-60 readings.
-70
-80 3.6. Integral velocity response
-90
0.0 0.5 1.0 1.5 No measured velocity response could be presented as no veloc-
Time (s) ity sensors were installed in the specimens. However, by assuming
a zero initial condition, it is possible to acquire the velocity re-
Fig. 9. Recorded vertical displacement distribution of DF1.
sponse by directly conducting an integral transform based on the
measured acceleration data:
a structure causes an immediate geometric change, resulting in the
Z t
2.0 2.0
0.0 0.0
-1.0 A1 -1.0 A2
-2.0 -2.0
-3.0 -3.0
-4.0 -4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
2.0 2.0
0.0 0.0
-1.0 -1.0 A4
A3
-2.0 -2.0
-3.0 -3.0
-4.0 -4.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Time (s) Time (s)
10 10
10 10
0
-10
(c) 0 (d)
-10
-20 -20
-30 DF1-Measured-D3 -30 DF1-Measured-D4
DF1-Integral-D3 DF1-Integral-D4
-40 -40
-50 -50
-60 -60
-70 -70
-80 -80
-90 -90
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Time (s) Time (s)
Fig. 11. Comparison the measured displacement history with the corresponding integral displacement history of DF1.
acceleration records. Even a small offset in acceleration can produce and V4 in DF1 were 0.61, 0.60, 0.32 and 0.14 m/s, respec-
significant drifts in velocities and displacements. Another major tively. Similar to the displacement and acceleration response dis-
source of unrealistic drift in velocities and displacements may come tribution along the transverse beam, the velocity response
from the accumulation of the random noise in accelerations result- distribution was also non-linear. The velocities of V1, V2, V3, and
ing from the integrations. Thus, for the graph tail, polynomial curve V4 were reduced to zero at 0.425 s, 0.406 s, 0.437 s and 0.439 s,
fitting was performed and the graph was subsequently corrected to respectively. The velocity responses of the remaining specimens
the zero-velocity baseline as shown in Fig. 12a. were similar to DF1. It should be noted that the acceleration, veloc-
In order to evaluate the accuracy of this integral method to ob- ity and displacement response of the full-scale specimens may be
tain the velocity response, the displacement responses obtained by up to three times of the reported figures as only one-third scale
integral transform from the recorded acceleration data were com- specimens were tested.
pared with the measured displacement histories. As presented in
Fig. 11, the integral displacement matched well with the measured 3.7. Vertical reaction force response
displacement, although the integral displacement had a slight time
lag. Thus, the accuracy of the proposed integral method and the Fig. 13 illustrates the recorded history of the total vertical reac-
reliability of the velocity response obtained from integral trans- tion forces of DF3 and DF4. As displayed in Fig. 13, the initial total
form can be ascertained. Fig. 12 illustrates the velocity responses vertical reaction force measured from the longitudinal and trans-
at prescribed locations of DF1, the peak velocities of V1, V2, V3 verse fixed supports of DF3 was 20.4 kN. This vertical reaction
0.4 0.4
Intergal V1 after
(a) (b)
Vertical velocity (m/s)
0.2
0.0 0.0
-0.6 -0.6
-0.8 -0.8
-1.0 -1.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
0.4 0.4
(c) (d)
Vertical velocity (m/s)
0.2
Vertical velocity (m/s)
0.2
0.0 0.0
-0.2 Integral V3 -0.2 Integral V4
-0.4 -0.4
-0.6 -0.6
-0.8 -0.8
-1.0 -1.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Time (s) Time (s)
0.0 enhancement, the dynamic increase factor for the released column
DF3
axial forces in DF3 and DF4 were 1.29 and 1.37 respectively. Fur-
DF4
-10.0
ther discussions for the dynamic effects are shown below.
Total Reaction Force (kN)
Initial Values
-20.4 kN for DF3
-20.0 -20.7 kN for DF4 3.8. Horizontal reaction force response
4.6 kN for DF4 2.45 kN m, respectively. The bending moment in the longitudinal
0.0
and transverse fixed supports suddenly increased to 20.8 and
20.1 kN m, respectively after removing the corner support. After
free vibration, the permanent moment was 16.6 and 16.3 kN m in
-10.0
the longitudinal and transverse fixed supports, respectively. It can
be seen that the bending moment suddenly increased by 682.0%
and 720.5% respectively in the longitudinal and transverse fixed
-20.0
supports after the removal of the corner support. For DF4, the initial
bending moments at the longitudinal and transverse supports were
2.79 kN m and 2.51 kN m, respectively. The bending moments in
-30.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 the longitudinal and transverse fixed supports suddenly rose to
Time (s) 20.1 kN m and 19.5 kN m, respectively after removing the corner
support. After free vibration, the permanent moments recorded in
Fig. 14. Recoded history of the horizontal reaction force in the transverse beam of the longitudinal and transverse fixed supports were 14.0 kN m
DF3 and DF4. and 14.8 kN m, respectively. It can be seen that the bending
25 25
DF3 DF4
Bending Moment (kN.m)
20 20
15 15
10 10
Initial Values Initial Values
2.45 kN.m in Trans. Beam 2.51 kN.m in Trans. Beam
5 2.66 kN.m in Long. Beam 5 2.79 kN.m in Long. Beam
Longitudinal Beam Longitudinal Beam
0 Transverse beam 0 Transverse Beam
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Time (s) Time (s)
(a) (b)
Fig. 15. Calculated histories of the bending moments in the fixed supports of DF3 and DF4.
Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167 163
Strain (με)
1200
In general, the trends of the strain gauge results in each speci-
1000
men were similar and thus, only the results of DF1 are explained
800 Initial Values
in detail in this paper. As shown in Fig. 16, the strain in the top 430 με in BST4 of DF1
of the BENC (BSTL1) was 360le before the removal of the corner 600
390 με in BST4 of DF2
support. However, this tensile strain decreased and ultimately 400 BSTL1 BSTL8
compressive strain in BSBL1 suddenly decreased to zero and ulti- Time (s)
mately attained the maximum tensile strain of 2488le at a time
of 0.42 s. Both results verified that the direction of bending mo- Fig. 17. Strain gauge recording of the transverse reinforcement in DF1 and DF2.
ment in the BENC was changed after the removal of the corner sup-
port. It can also be seen that the strain reading significantly
of DF1. Comparing the peak vertical displacements of these two
decreased after 0.5 s. The permanent strain in BSTL1 and BSBL1
specimens (Fig. 8), it can be seen that the peak vertical displace-
at the end of vibration was 723 and 1731le, respectively. The ini-
ment of DF2 was only 34.6% that of DF1. Similar trends were ob-
tial strain readings of BSTL8 and BSBL8 were 578 and 205le,
served for the acceleration and velocity results.
respectively. After removing the corner support, the readings of
BSTL8 and BSBL8 increased to 4487 and 1817le at the times of
4.2. The effect of different service load conditions
0.41 and 0.42 s, respectively. This demonstrated that the bending
moment increased considerably in the BENF (consistent with the
As listed in Table 1, DF3 was subjected to a full service load
bending moment results). For other specimens, similar behavior
1.0(1.2DL + 0.5LL) while DF1 was subjected to a reduced service
was observed. Fig. 17 illustrates the strain gauge reading in the
load 0.9(1.2DL + 0.5 LL) before the removal of the corner support.
transverse reinforcements of DF1 and DF2. As presented in the fig-
Comparing the crack patterns of DF3 with DF1 (refer to Fig. 7),
ure, the strain in BST4 was initially 430le and it increased to
more severe cracks were observed in the corner joint and longitu-
1807le by 0.42 s. The permanent tensile strain in BST4 at the
dinal BENF in DF3. However, the shear cracks occurring in the
end of the vibration was 1425le. It should be emphasized that
transverse BENF of DF1 were more severe than those of DF3. This
the yield strain of steel rebar R6 was 2650le. Thus, the transverse
was an unexpected phenomenon and was possibly due to inherent
rebar did not yield during the test. However, as shown in the final
drawbacks existing in DF1, such like a honeycomb structure devel-
crack patterns of DF1 (refer to Fig. 7a), severe diagonal cracks oc-
oped in the BENF of DF1 due to insufficient vibration during cast-
curred in the BENF. Similar behavior was observed in DF2. How-
ing. It can be seen in Fig. 8 that DF3 increased the peak vertical
ever, the peak strain of BST4 was only 1451le, much less than
displacement by 74.7%.
the BST4 in DF1, possibly due to the higher transverse reinforce-
ment ratio in the beam potential plastic hinge zone.
4.3. The effect of modified detailing
4. Discussion of test results DF4 had a much higher transverse reinforcement ratio in the
potential plastic hinge zone of beams compared to that of DF3.
4.1. The effect of seismic detailing Similar to DF3, DF4 was subjected to a full service load
1.0(1.2DL + 0.5LL) before removing the corner support. Comparing
The non-seismically detailed specimen DF1 had severe flexural the crack pattern of DF4 to that of DF3 (refer to Fig. 7), it can be
and shear cracks in the BENFs while only limited flexural cracks in seen that milder cracks were observed in the BENFs and corner
the BENFs were observed in the seismically detailed specimen DF2. joint in DF4 although, in general, similar crack patterns were ob-
Moreover, no cracks were observed in the corner joint of DF2 while served in both specimens. As seen in Fig. 8, DF4 sustained a de-
several hairline diagonal cracks were observed in the corner joint crease in the peak vertical displacement by 32.2%.
BSTL8
2000
360 mm. In fact, without the pin support of the column removal
1000
BSBL1 apparatus or should there be sufficient space to allow the gravity
0 BSTL1 weights to drop freely, the specimen would be expected to fail
completely.
-1000
-2000
BSBL8 4.5. The effect of span aspect ratio
-3000
0.0 0.5 1.0 1.5 2.0 2.5 3.0
DF6 has different span lengths in the longitudinal and trans-
Time (s)
verse directions, and it was subjected to the reduced service load
Fig. 16. Strain gauge recording of the flexural reinforcement in DF1. of 0.9(1.2DL + 0.5LL). It can be seen from Fig. 7f that extensive
164 Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167
3: Steel column
2
4: Comp/tension load cell
5: Steel assembly
3
6: LVDT with 300 mm travel
7: RC substructure 5 4
8 and 9: LVDTs 6
Longitudinal beam
9 8
10 and 11: Comp/tension load cell
7
11
Transverse beam 10
Table A1
Co-relationship between the prototype frames with corresponding test models in the current study (unit: mm).
Test Prototype Beam (mm) Prototype Beam Rebar Model Beam (mm) Model Beam Rebar
Trans. Beam Long. Beam Trans. Beam Long. Beam Beam-T Beam-L Trans. Beam Long. Beam
Top Bottom Top Bottom
DF1 (F3) 540 300 540 300 2T20 + T32 2T20 + T32 2T20 + T32 2T20 + T32 180 100 180 100 4T10 4T10
DF2 (F2) 540 300 540 300 3T32 3T32 3T32 3T32 180 100 180 100 4T13 4T13
DF3 (F3) 540 300 540 300 2T20 + T32 2T20 + T32 2T20 + T32 2T20 + T32 180 100 180 100 4T10 4T10
DF4 (F4) 540 300 540 300 2T20 + T32 2T20 + T32 2T20 + T32 2T20 + T32 180 100 180 100 4T10 4T10
DF5 (F5) 720 300 720 300 2T25 + T32 2T25 + T32 2T25 + T32 2T25 + T32 240 100 240 100 4T10 4T10
DF6 (F6) 540 300 720 300 2T20 + T32 2T20 + T32 2T25 + T32 2T25 + T32 180 100 240 100 4T10 4T10
66:8 kN
¼ 16:7 kN ðA:18Þ
Fig. A.3. Equivalent uniform line pressure on beam of prototype slab. 4
Q. Kai, B. Li / Engineering Structures 42 (2012) 154–167 167
A.3. For DF5 [6] Tsai MH, Lin BH. Investigation of progressive collapse resistance and inelastic
response for an earthquake-resistant RC building subjected to column failure.
Eng Struct 2008;30(12):3619–28.
The design pressure [7] Kwasniewski L. Nonlinear dynamic simulations of progressive collapse for a
multistory building. Eng Struct 2010;32(5):1223–35.
qm ¼ 0:8 ð1:2DL þ 0:5LÞ ¼ 10:3 kN=m2 ðA:19Þ [8] Izzuddin BA, Vlassis AG, Elghazouli AY, Nethercot DA. Progressive collapse of
multi-storey buildings due to sudden column loss—Part I: simplified
assessment framework. Eng Struct 2008;30(5):1308–18.
The total gravity load of the slab: [9] Vlassis AG, Izzuddin BA, Elghazouli AY, Nethercot DA. Progressive collapse of
multi-storey buildings due to sudden column loss—Part II: application. Eng
10:3 3:0 3:0 ¼ 92:9 kN ðA:20Þ Struct 2008;30(5):1424–38.
[10] Mohamed OA. Assessment of progressive collapse potential in corner panels of
The design axial force is: reinforced concrete buildings. Eng Struct 2009;31(3):749–57.
[11] Fu F. 3-D nonlinear dynamic progressive collapse analysis of multi-storey steel
92:9 kN composite frame buildings—Parametric study. Eng Struct
¼ 23:2 kN ðA:21Þ 2010;32(12):3974–80.
4 [12] Sasani M, Bazan M, Sagirolu S. Experimental and analytical progressive
collapse evaluation of an actual reinforced concrete structure. ACI Struct J
2007;104(6):731–9.
A.4. For DF6 [13] Sasani M. Response of a reinforced concrete infilled-frame structure to
removal of two adjacent columns. Eng Struct 2008;30(9):2478–91.
[14] Sasani M, Kropelnicki J. Progressive collapse analysis of an RC structure. Struct
qm ¼ 1:2DL þ 0:5L ¼ 12:9 kN=m2 ðA:22Þ Des Tall Special Build 2008;17:757–71.
[15] Sasani M, Sagiroglu S. Gravity load redistribution and progressive collapse
The total gravity load of the slab: resistance of 20 story reinforced concrete structure following loss of interior
column. ACI Struct J 2010;107(6):636–44.
12:9 3:0 2:4 ¼ 92:9 kN ðA:23Þ [16] Yi WJ, He QF, Xiao Y, Kunnath SK. Experimental study on progressive collapse-
resistant behavior of reinforced concrete frame structures. ACI Struct J
The design axial force is: 2008;105(4):433–9.
[17] Kai Q, Li B. Slab effects on the response of reinforced concrete substructures
92:9 kN after the loss of a corner column. ACI Struct J, in press.
¼ 23:2 kN ðA:24Þ [18] Tsai MH, Huang TC. Numerical investigation on the progressive collapse
4
resistance of an RC building with brick infills under column loss. World Acad
Sci Eng Technol 2011;58:946–53.
[19] CP 65. Structural use of concrete, Part 1 Code of practice for design and
References construction. Singapore Standard; 1999.
[20] ACI Committee 318. Building code requirements for structural concrete (ACI
[1] ASCE/SEI 7. Minimum design loads for buildings and other structures. Reston 318-08) and commentary (318R-08). American Concrete Institute, Farmington
(VA): Structural Engineering Institute, American Society of Civil Engineers; Hills, MI; 2008. p. 456.
2010. [21] Kai Q. Experimental and analytical study of reinforced concrete substructures
[2] DOD. Design of building to resist progressive collapse. Unified Facility Criteria, subjected to a loss of ground corner column scenario. PhD dissertation,
UFC 4-023-03. Washington (DC): US Department of Defense; 2009. Nanyang Technological University, Singapore; 2011. p. 300.
[3] GSA. Progressive collapse analysis and design guidelines for new federal office [22] Kai Q, Li B. Performance of three-dimensional reinforced concrete beam-
buildings and major modernization projects. Washington (DC): US General column substructures under loss of a corner column scenario. J Struct Eng,
Service Administration; 2003. submitted for publication.
[4] Marjanishvili SM. Progressive analysis procedure for progressive collapse. J [23] Marchand K, Mckay A. Stevens DJ. Development and application of linear and
Perform Construct Facil ASCE 2004;18(2):79–85. non-linear static approaches in UFC 4-023-03 structures 2009: Don’t Mess
[5] Marjanishvili SM, Agnew E. Comparison of various procedures for progressive with Structural Engineers@2009. ASCE; 2009. p. 1729–38.
collapse analysis. J Perform Construct Facil 2006;20(4):365–74.