ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 102-S78

Dynamic Responses of Flat Plate Systems with

Shear Reinforcement
by Thomas H.-K. Kang and John W. Wallace

Use of flat plate floor systems consisting of a conventional or post- intensities considerably less than those expected in the
tensioned reinforced concrete slab-column system incorporating design basis event (for example, 10% in a 50 year event).
shear reinforcement within the slab-column connection region has
A code change has been approved for ACI 318-057 to
become increasingly popular as lateral-force resisting systems in
intermediate seismic zones and gravity systems in high seismic clarify the intent of the code. The approved code change
zones. Relatively sparse experimental data, however, are available assesses the need for shear reinforcement at slab-column
to assess the dynamic responses of such systems, particularly for connections based on the inter-story lateral drift and the
post-tensioned systems. To address these gaps, shaketable tests of gravity shear stress on the slab critical section, unless
a conventional reinforced-concrete slab-column frame (RC calculations are made to show that the connection is
specimen) and a post-tensioned slab-column frame (PT specimen) capable of sustaining the drift associated with the design
were conducted. An overview of the test program and the findings displacement without punching. Analysis of the slab-
is presented. Lateral drift ratios of approximately 3 and 4% were
column frame requires that appropriate models be used and
achieved during testing for the RC and PT system frames, respectively,
with relatively little loss of lateral load capacity. Also, although slab- that the drift associated with punching failures be documented
column punching failures occurred for both specimens, the degree of for a range of conditions.
damage observed was less extensive than damage observed for prior An effective slab width model is commonly used to model
quasi-static tests of specimens without shear reinforcement. the lateral-load stiffness of the slab-column frame.1 In the
effective slab width method, the three-dimensional system is
Keywords: concrete; flat plate; post-tensioning; reinforced concrete; modeled as a two-dimensional frame using an effective slab
shear reinforcement; stud. width l2 and a conventional column, where the -factor is
derived using elastic plate theory to result in an equivalent
INTRODUCTION slab width with uniform rotation across the effective slab
Building systems constructed to resist seismic forces are width that yields the same rotation stiffness as the original
often designed with a lateral-force resisting system to resist system with nonuniform rotation.8 The -value for the effective
the seismic forces and a nonparticipating system (or gravity- slab width model depends primarily on the column and slab
force resisting system) to support vertical loads. Flat plate aspect ratios, whereas the influence of cracking on stiffness
systems have become very popular gravity systems; there- is accounted for by using an additional factor, commonly
fore, their behavior, when subjected to lateral forces or referred to as a -factor. 2
deformations, has been studied extensively (for example, According to ACI 318-02 6 requirements, design shear Vu
Moehle and Diebold; 1 Hwang; 2 Moehle; 3 and Robertson et and unbalanced moment Mu,unb from a gravity load analysis
al.4). In general, studies of flat plate systems have focused on of a slab-column frame are assumed to be transferred to the
developing and refining models that describe the stiffness column by direct shear and eccentric shear acting on a
of the slab-column frame and the strength of the slab-
critical section within the slab adjacent to the column
column connection.
perimeter, and by slab flexural yielding within a slab transfer
Use of flat plate systems to resist earthquake lateral loads width of c2 + 3h centered on the column. The maximum
is not permitted in Zones 3 and 4 (Uniform Building Code5); shear stress on the slab critical section is a sum of the direct
however, no specific requirements are stated in ACI 318-02,6 shear stress from gravity load and the maximum shear stress
Section 21.11, Frame Members Not Proportioned to Resist from the eccentric shear stress model. If the connection
Forces Induced by Earthquake Motions for slab-column strength is insufficient to transfer the unbalanced moment
systems. Therefore, the systems are designed for gravity loads
anywhere on the slab critical section, then a punching shear
and for deformation compatibility with the lateral-force
failure is assumed to occur. Common approaches used to
resisting system to ensure that the system can maintain
increase the shear strength of the slab-column connection are
support for vertical loads. ACI 318-02,6 Chapter 7, requires
to provide either a drop panel or slab shear reinforcement.
continuous bottom reinforcement (integrity reinforcement)
through the column to prevent progressive collapse; most Currently, use of slab shear reinforcement is a popular
engineers assume that providing the integrity reinforcement option, as it obviates the need to provide a drop panel, and
satisfies the deformation compatibility requirement a priori, thus tends to reduce costs associated with labor and formwork.
versus designing the slab-column connection to avoid a
punching failure when the slab-column frame is checked for ACI Structural Journal, V. 102, No. 5, September-October 2005.
deformation compatibility. Given that the punching shear MS No. 04-325 received October 8, 2004, and reviewed under Institute publication
policies. Copyright 2005, American Concrete Institute. All rights reserved, including the
capacity may be nearly exhausted under gravity loads, making of copies unless permission is obtained from the copyright proprietors. Pertinent
discussion including authors closure, if any, will be published in the July-August
however, punching failures may occur for ground motion 2006 ACI Structural Journal if the discussion is received by March 1, 2006.

ACI Structural Journal/September-October 2005 763

 tensioned slab-column frame (PT) flat plate system for
ACI member Thomas H.-K. Kang is a lecturer and postdoctoral researcher at the
University of California-Los Angeles (UCLA), Los Angeles, Calif. He received his BS common construction in intermediate and high seismic regions
from Seoul National University, Korea, in 1998 and his PhD from UCLA in 2004. He in the U.S., respectively, were constructed and subjected to
is a member of Joint ACI-ASCE Committee 352, Joints and Connections in Monolithic
Concrete Structures. His research interests include the behavior and modeling of
uniaxial shaking. Due to dimensional limitations of the
reinforced and post-tensioned concrete systems subjected to dynamic loading. University of California-Berkeley shaketable, a scale
factor of approximately 1/3 was used for both specimens. The
John W. Wallace, FACI, is an associate professor of civil engineering at UCLA. He is
a member of ACI Committee 318-H, Seismic Provisions; 369, Seismic Repair and
specimens included six columns, with two full spans in the east-
Rehabilitation; 374, Performance-Based Seismic Design of Concrete Buildings; and west (E-W) direction, and one full span and two half-spans in the
Joint ACI-ASCE Committee 352, Joints and Connections in Monolithic Concrete north-south (N-S) direction (Fig. 1). This geometry was selected
Structures. His research interests include response and design of buildings and
bridges to earthquake actions, laboratory and field testing of structural components to provide symmetry and stability, as well as appropriate
and systems, and structural health monitoring and use of sensor networks. boundary conditions for uniaxial shaketable tests.
Column flexural strengths for the specimens were selected
Although the use of post-tensioned flat plate systems with to ensure that yielding and damage would be concentrated in
shear reinforcement has become very common in both the slab-column connection region versus within the column.
intermediate and high seismic regions in the U.S., a majority Resulting column cross sections for both the RC and PT
of existing experimental work has focused on conventional specimens were 203 x 203 mm reinforced with eight 12.7 mm
reinforced concrete slab-column connections with or diameter (No. 4) longitudinal bars with a nominal yield
without shear reinforcement. Tests have been conducted stress of 414 MPa (60 ksi). The connection regions for both
on conventionally reinforced slabs with shear reinforcement, specimens were designed such that punching failures were
such as the tests by Megally9 and Robertson et al.,4 which expected to occur prior to yielding of slab flexural (column
included specimens with shear (stud-rail) reinforcement strip) reinforcement in the absence of slab shear reinforcement.
subjected to quasi-static, reversed cyclic loading. In most of With the addition of slab shear reinforcement, yielding of
these tests, the specimens consisted of isolated, interior or slab reinforcement within the column strip was anticipated;
exterior connections without drop panels or shear capitals. however, punching failures were expected to occur prior to
Test results for post-tensioned systems subjected to yielding of slab reinforcement across the full width of the
earthquake-type loadings are limited to tests of seven slab (Table 1). Design of the test specimens is summarized
isolated connections conducted by Martinez-Cruzado, in the following paragraphs with more detailed information
Qaisrani, and Moehle.10 No dynamic tests of post-tensioned provided by Kang.11
systems have been conducted.
RC specimen
RESEARCH SIGNIFICANCE Typical slab span-depth ratios for RC flat plate construction
Given that use of flat plate systems, particularly post- are roughly 20 to 30 (25 to 30 in the U.S., 20 to 25 in Japan);
tensioned systems with shear reinforcement at the slab-column
therefore, a bay width of 2.06 m (6 ft. 9 in.) and a slab thickness
connections, is popular and test results for reversed cyclic
of 89 mm (3.5 in.) were selected, resulting in a span-depth
and dynamic loading are scarce, a research study was
ratio of 23.1 for the test specimen. Gravity load consisted
undertaken to conduct shaketable tests and analytical
of the slab self-weight (2.2 kPa [45 lb/ft2]) and lead
studies. An overview of the experimental research program
weights affixed to the slab to provide an additional 5.7 kPa
and general results is presented. The test results provide
(120 lb/ft2) to represent the self-weight of the prototype
valuable data on the lateral-load behavior of flat plate
floor slab of 267 mm thickness (6.5 kPa [135 lb/ft2]), as
systems as well as information needed to verify analytical
models and to assess the extent of damage for dynamic well as an additional gravity load of 1.4 kPa (30 lb/ft2). The
loading. These issues are particularly important given the resulting gravity shear ratios on the slab critical sections
changes approved for ACI 318-057 for nonparticipating adjacent to the 203 x 203 mm columns for the interior and
slab-column frames. exterior connections were 0.25Vc and 0.20Vc, respectively,
which are fairly representative of reinforced concrete flat
plate construction in the U.S. and Japan.
EXPERIMENTAL PROGRAM
Two flat plate slab-column frames, designed to represent a The slab reinforcement consisted of 9.5 mm-diameter
reinforced concrete slab-column frame (RC) and a post- (No. 3) deformed bars with a yield stress of 458 MPa (66.4 ksi)
having 180-degree standard hooks at the slab edge.
Concrete clear cover was 9.5 mm for both top and bottom
slab reinforcement. Sufficient slab reinforcement was
placed within c2 + 3h (Fig. 2), where c2 is the column
dimension perpendicular to the direction of the applied loads
and h is the slab thickness, to resist the fraction of the
unbalanced moment transferred by flexure f t Mu,unb as well
as satisfy the requirements in ACI 318-02,6 Section 21.12.6.3.
The ratio of top to bottom reinforcement for unbalanced
moment transfer within the transfer width of c2 + 3h was
selected as approximately two as recommended in ACI 318-02,6
Section R.13.5.3.3, and slab minimum reinforcement was
provided to satisfy ACI 318-02,6 Section 7.12.2.1. In
addition, ACI 318-026 requirements of Section 21.12.6
were satisfied, as 71% of top bars were placed within the
Fig. 1Test specimen on shaketable. column strip and all bottom bars were continuous. Two

764 ACI Structural Journal/September-October 2005

continuous bottom bars were placed within the column cage in section located at d/2 from the column face (column critical
accordance with ACI 318-02,6 Section 13.3.8.5. section in Fig. 3). Outside the shear-reinforced region
Shear reinforcement, in the form of stud-rails, was (outer critical section in Fig. 3), shear resistance is
selected because available test results have shown them to be provided by concrete alone.
effective (Ghali and Hammill;12 Megally;9 and Kang and
Wallace13) and they are commonly used. Design shear PT specimen
strength of the shear reinforced connections was based on The configuration of the PT specimen was similar to that
ACI 318-026 and ACI 421.1R-9914 requirements. Stud-rails used for the RC specimen (Fig. 4), except a slab thickness of
were provided to resist combined direct gravity shear stress 76 mm (3 in.) and bay widths of 2.84 m (9 ft. 4 in.) in the
vu,direct and eccentric shear stress vu,unb due to the unbalanced direction of testing were selected (Fig. 4). The resulting
moment transferred by eccentric shear on the slab critical span-to-depth ratio of 37.3 is slightly less than typical values

Fig. 2Reinforced-concrete slab-column frame specimen.

Table 1Shear demand and capacity for column and outer critical sections
Reinforced concrete Reinforced concrete Reinforced concrete Reinforced concrete Post-tensioned Post-tensioned
slab-column frame- slab-column frame- slab-column frame- slab-column frame- slab-column slab-column
interior, Min v exterior, Max v interior, Min v exterior, Max v frame-interior frame-exterior
vu,direct 0.434 0.434 0.345 0.345 0.641 0.555
1* 1.710/0.848 2.475/1.613 0.738/0.041 2.772/3.468 2.613./1.331 3.508/2.972
vu
2* NA 0.614/0.234 NA 0.855/0.317 0.889/0.221 1.103/0.531
vu
1 2.392/1.531 3.571/2.703 1.007/0.131 4.096/5.288 2.896/1.613 3.896/3.358
vu
vu 2 NA 0.841/0.469 NA 1.207/0.600 0.972/0.303 1.200/0.614
1
vc 1.482 1.482 1.482 1.482 1.648 1.648
2
vc NA 0.745 NA 0.745 1.110
1.110
vn1 2.227 2.227 2.227 2.227 2.965 2.965
* _
Based on Mu,unb = M+y,cs (+ M y, cs); (sum of yield moments within column strip).
_
Based on Mu,unb = M+v,fs(+ M v, fs); (sum of yield moments across full span).
1/2
Upper limit of vn = 2/3(fc) (based on ACI 421.1R-99).
1On critical section at d/2 from column face (column critical section).
2On critical section at d/2 from outermost peripheral line of studs (outer critical section).

Notes: All units are in kPa; vu = vu,direct + vu, unb = (Vu/Ac) (v Mu,unbc/Jc); vc1 = 1/3(f c )1/2, vc2 = 1/6(f c )1/2, vn = 1/2(f c )1/2, where v and f c are in MPa; NA = not available.
Min v = 0.25 (reinforced-concrete slab-column frame-interior), 0 (reinforced-concrete slab-column frame-exterior), NA (reinforced-concrete slab-column frame, exterior, outer critical section)
(based on ACI 318-02).
Max v = 0.40 (reinforced-concrete slab-column frame-interior), 0.384 (reinforced-concrete slab-column frame, exterior, column critical section), 0.305 (reinforced-concrete
slab-column frame, exterior, outer critical section).
v = 0.40 (post-tensioned slab-column frame-interior), 0.386 (post-tensioned slab-column frame, exterior, column critical section), 0.313 (post-tensioned slab-column frame,
exterior, outer critical section).

ACI Structural Journal/September-October 2005 765

used for PT slab-column frame construction on the west Tendons were placed according to ACI 318-026 and
coast of the U.S. (40 to 45). Lead weights were affixed to the ACI 423.3R-9616 provisions and recommendations,
76 mm-thick floor slabs to provide 3.6 kPa (76 lb/ft2) of respectively, as shown in Fig. 4. Six banded tendons, with
gravity load, to result in a total slab self-weight of 5.5 kPa two passing through the column cage, were used along each
(114 lb/ft2), which corresponds closely to the self-weight of column line in the E-W direction (direction of testing), and
the prototype 229 mm (9 in.) thick post-tensioned slab. The seven (two passing through the column cage) distributed
resulting gravity stress ratios on the interior and exterior tendons between each column line were used in the N-S
connections are approximately 0.33Vc (0.39Vc) and 0.25Vc direction. Minimum clear cover was 9.5 mm over both the
(0.30Vc), which are less than the limit of 0.4Vc for top and bottom of tendons or bonded reinforcement (No. 2,
intermediate slab-column frames of ACI 318-02,6 Section db = 6.35 mm), except at interior connections where top clear
21.12, but low for typical nonparticipating slab-column cover of 7.9 and 12.7 mm were designated for tendons and
frames used in the U.S., where ratios as high as 0.50 to bonded (deformed) reinforcement, respectively, due to space
0.60Vc are sometimes used. limitations (Fig. 4). An effective tension force of 44 kN
Slabs were post-tensioned with 7.9 mm (5/16 in.) diameter, (9.9 kips) per strand was applied based on flexural strength
seven-wire strand with a yield strength of 1627 MPa (236 ksi). requirements, resulting in an average compressive stress of
The post-tensioning strands were greased and wrapped directly 1.39 MPa (202 psi) in the banded (testing) direction (E-W)
in an extruded plastic coating per PTI recommendations15 and and 1.37 MPa (199 psi) in the other direction (N-S). Given
post-tensioned to provide a design average compressive the effective tendon force, the post-tensioned slab balanced
concrete stress of fpc = 1.38 MPa (200 psi). The diameter of the a gravity load of 4.3 kPa (90 lb/ft2), or 79% of dead load
unbonded extruded tendons was kept to 11.1 mm (7/16 in.), (including lead-weights).17
which was desirable given the slab thickness of 76 mm (3 in.). Bonded top reinforcement was provided at connection
regions as required by ACI 318-02,6 Section 18.9.3.3 (Fig. 4).
Although not required, an equal quantity of bottom reinforce-
ment was provided to control cracking along the bottom of
the slab anticipated due to applied dynamic (cyclic) load
history as well as to improve the hysteretic energy dissipation
capacity of the system.

Shear reinforcement
Shear reinforcement for the RC and PT specimens
consisted of stud-rails (Fig. 2 and 4) satisfying requirements
of ACI 318-026 and ACI 421.1R-99.14 The size and spacing
of the stud-rails were selected to require flexural yielding of
Fig. 3Stud-rails and critical sections at connection bonded bars prior to punching failure on a critical section d/2
regions. from the column face (Table 1). Based on ACI 421.1R-99,14

Fig. 4Post-tensioned slab-column frame specimen.

766 ACI Structural Journal/September-October 2005

the first studs were located at 19 mm (0.75 in.) (= 0.27d = slab-column punching failure assessment; and 5) flexural
0.40d) away from the column faces, and studs were spaced at transfer width and slab effective width evaluation.
0.5d = 35 mm (1.38 in.) on center for both the RC and PT
specimens. The stud-rails extended 2.5h (note that the Free vibration tests and shaketable tests
minimum length of 4h will be required in ACI 318-057 provi- Free vibration tests were conducted prior to initial testing
sions) from the column face, to ensure that the punching and after each shaketable test to assess changes in stiffness
failure would occur within the shear-reinforced region (at the and damping of the specimens (Table 4). The damped
slab critical section d/2 from the column face). Two stud- fundamental period Tn was determined by averaging the
rails were placed perpendicular to each column face 187 mm time between response peaks for several cycles and dividing
(7.35 in.) apart for the RC specimen, whereas three stud-rails by the number of cycles for 17 and 15 channels for the RC
were placed perpendicular to each column face for the PT and PT frames, respectively. Damping ratios were
specimen (due to the thinner slab). determined using the logarithmic decrement method for 4 to

Materials
Table 2Material properties
Design compressive concrete strength was 27.6 MPa
(4000 psi) with a maximum aggregate size of 9.5 mm (3/8 in.). Esec,
MPa fc , kPa o
Table 2 summarizes material properties for the concrete
used in the test structure. A representative concrete stress- Concrete 19,305 27,668 0.0024
strain relation for a 152 x 305 mm (6 x 12 in.) cylinder is db, mm As, mm2 Es, MPa fy, MPa y fs*, MPa fu, MPa
presented in Fig. 5(a). Material properties obtained for the
No. 2 bars 4.1 13.1 207,717 481 0.00432 373 684
deformed reinforcing bars, stud-rail shaft, and the seven-
wire strand are shown in Table 2 and the associated stress-strain No. 3 bars 6.3 31.3 239,384 458 0.00395 401 652
relations are depicted in Fig. 5(b). More detailed material No. 4 bars 9.1 64.3 204,756 491 0.00242 698
information is available elsewhere.18 Studs 4.0 12.8 213,867 635 0.00489 743
Strands 8.0 37.5 200,898 1625 0.00979 1834
*f = stress
Instrumentation s at 0.002 strains.

During testing, information was collected from 193 channels Yield stress at 0.2% offset.
for each specimen. A detailed summary of the instrumentation
used is provided by Kang.11 The instrumentation layout was
selected to allow the determination of story shear and over-
turning moments, story displacements, reinforcing bar and
stud-rail strains, and slab and column moments and
curvatures, with a particular emphasis on evaluating the
unbalanced moment transferred at slab-column connections
during the dynamic tests.

Testing and ground (table) motions

A ground motion with fairly long duration and without
significant near-fault effects, the CHY087W record from the
September 21, 1999, earthquake in Taiwan was selected. The
original ground motion was time-compressed by 1/3
corresponding to the dimension scale factor of 1/3 (Harris
and Sabnis19). The peak values of the spectral pseudo-
acceleration response spectrum of the modified ground
motion were located between 0.2 to 0.4 s, where the
fundamental periods of the specimens were located. The
specimens were subjected to several runs of uniaxial shaking
with the intensity of the motions increasing with each test to
study specimen behavior for: 1) low-level excitation; 2)
service-level excitation to approximately 2/3 of the yield
displacement; 3) moderate-level excitation to produce
limited yielding within the structure; and 4) damage-level
excitation. For the PT specimen, a fifth test was run with
very intense motions given that relatively little damage was
observed after Run 4 (Table 3).

EXPERIMENTAL RESULTS AND DISCUSSION

Observations and experimental results obtained from
evaluation of the data collected from the free vibration tests
and the shaketable tests are presented and discussed in the
following sections. The information presented includes: 1)
general findings and observed damage; 2) periods and Fig. 5Stress-strain relations of materials: (a) one concrete
damping ratios; 3) base shear versus drift ratio relations; 4) cylinder; and (b) steel bars and seven-wire strands.

ACI Structural Journal/September-October 2005 767

6 cycles of response. Fourier amplitude spectra (FAS) of the PT specimens. To identify these changes, response histories
response histories obtained during the free vibration tests of absolute story accelerations for Run 4 were divided into
also were used to identify first and second mode periods Tp1 two time windows and FAS analyses for each window were
and Tp2. As anticipated, periods and damping ratios increased as compared. The fundamental period before and after 12.70 s
the intensity of the table motions were increased, with the most was 0.42 and 0.53 s, respectively, for the RC specimen,
significant increases noted for Run 4 for the RC specimen. whereas the fundamental period before and after 11.10 s was
Table 5 summarizes the responses of the test specimens to 0.43 and 0.54 s for the PT specimen. As will be discussed in
all shaketable test runs. First-mode and second-mode periods more detail later, these period increases were caused by
were identified from the FAS of the absolute story punching failures at the slab-column connections, as well as
displacement histories. Significant changes in the funda- yielding at the base of the columns. Punching was noted for
mental period occurred during Run 4 for both the RC and some connections between 12.68 s and 12.72 s for the RC

Table 3Testing sequence and table motion characteristics

Reinforced concrete slab-column frame specimen Post-tensioned slab-column frame specimen
Test ID PHA, g IA, cm/s D, s Tp, s Tm , s Test ID PHA, g IA, cm/s D, s Tp , s Tm , s
Free vibration Tests 1 and 2 (FV 1, 2) Free vibration Tests 1 and 2 (FV 1, 2)
Run 1 0.109 12.7 7.9 0.290 0.307 Run 1 0.095 10.2 8.8 0.318 0.320
Run 2-1 0.287 89.2 11.3 0.438 0.295 Run 2 0.276 77.7 8.9 0.346 0.363
Free vibration Tests 3 and 4 (FV 3, 4) Free vibration Tests 3 and 4 (FV 3, 4)
Run 2-2 0.256 94.2 11.3 0.439 0.286 Run 3 0.413 202.4 11.0 0.407 0.408
Free vibration Tests 5 and 6 (FV 5, 6) Free vibration Tests 5 and 6 (FV 5, 6)
Run 3 0.407 223.5 10.9 0.438 0.319 Run 4 1.144 1730.0 9.9 0.558 0.518
Free vibration Tests 7 and 8 (FV 7, 8) Free vibration Tests 7 and 8 (FV 7, 8)
Run 4 1.254 1775.0 9.9 0.439 0.366 Run 5 1.275 2221.0 9.8 0.585 0.552
Free vibration Tests 9 and 10 (FV 9, 10) Free vibration Tests 9 and 10 (FV 9, 10)
Note: PHA = Peak horizontal acceleration; IA = Arias intensity (Arias 1970); D = time interval between points at which 5 and
75% of Arias intensity has been recorded; Tp = predominant period; and Tm = mean period.

Table 4Free vibration tests

Tn , s Tn, % Tpl, s Tp1 , % Tp2, s Tp2, % , %
Reinforced-concrete slab-column frame specimen
FV 1, 2 0.233 0.240 0.057 0.0119
FV 3, 4 0.285 22 0.308 28 0.071 25 0.0323 171
FV 5, 6 0.293 3 0.319 4 0.076 7 0.0311 4
FV 7, 8 0.318 9 0.334 5 0.075 1 0.0394 27
FV 9, 10 0.481 51 0.566 69 0.113 51 0.0608 54
Post-tensioned slab-column frame specimen
FV 1, 2 0.276 0.286 0.071 0.0163
FV 3, 4 0.305 11 0.315 10 0.074 4 0.0213 31
FV 5, 6 0.333 9 0.336 7 0.076 3 0.0339 59
FV 7, 8 0.432 30 0.444 32 0.096 26 0.0419 24
FV 9, 10 0.465 8 0.502 13 0.105 9 0.0503 20

Table 5Responses of shaketable tests

Peak base Peak first story Peak first story Peak base First-mode Second-mode
t t
acceleration, g acceleration, a1 , g acceleration, a2 , g shear, kN Tp1, s Tp2, s
Reinforced-concrete slab-column frame specimen
Run 1 0.109 0.241 0.328 82.7 (0.25W) 0.271 0.0776
Run 2-1 0.287 0.420 0.644 149.3 (0.46W) 0.324 0.0843
Run 2-2 0.256 0.556 0.681 182.5 (0.56W) 0.346 0.0870
Run 3 0.407 0.767 1.055 281.2 (0.87W) 0.404 0.1333
Run 4 1.254 1.065 1.752 433.8 (1.33W) 0.456
Post-tensioned slab-column frame specimen
Run 1 0.095 0.222 0.326 96.6 (0.27W) 0.319 0.0758
Run 2 0.276 0.460 0.600 175.2 (0.48W) 0.346 0.0807
Run 3 0.413 0.714 0.854 175.2 (0.77W) 0.404 0.0880
Run 4 1.144 0.968 1.469 409.3 (1.13W) 0.544
Run 5 1.275 0.805 1.327 352.9 (0.97W) 0.600

768 ACI Structural Journal/September-October 2005

specimen and between 11.08 and 11.11 s for the PT spec- Based on measured responses, the RC specimen was
imen, whereas the columns began to yield at the base at 12.64 subjected to peak drift levels of approximately 3% with only
and 10.86 s for the RC and PT specimens, respectively. modest strength deterioration. The measured base shear of
the RC specimen began to degrade after t = 12.68 s, when the
Observed damage top drift ratio reached 2.5% (Fig. 9(a)). Strength deterioration
Significant cracking was observed both on the top and was observed for the PT specimen after the top drift ratio
bottom of the slab adjacent to columns of the RC specimen. reached 2.78% at t = 11.08 s (Fig. 9(b)). During the first
On the top surface of the slab, approximately d/2 (35 mm) significant excursion into the nonlinear range, a slight
from the column faces, where d is the effective slab depth, decrease in the base shear of the RC specimen was observed.
extensive cracking was observed and radial extension of In contrast, more substantial deterioration in base shear was
cracks from the critical sections occurred (Fig. 6(a)); however, observed for the PT specimen.
in general, cracks were not observed outside the shear- Although substantial loss of stiffness is apparent for both
reinforced region for the RC specimen. The most significant specimens, results for the RC specimen reveal significant
slab damage was observed at the exterior connections of the pinching of the hysteresis loops relative to the PT specimen.
RC specimen where substantial torsional cracking and The hysteresis loops for the lateral load versus top
concrete spalling occurred during Run 4 (Fig. 6(b)). displacement relations for the PT specimen are narrower
than the corresponding loops for the RC specimen. For the
For the PT specimen, less severe cracking was observed PT specimen, after yielding of the nominal quantity of
within the connection regions; however, cracking was bonded slab reinforcement and at the column bases, the PT
observed outside of the region reinforced with the banded specimen acts as an essentially elastic system with reduced
post-tensioning at the east edge (Fig. 6(c)). Yield lines were stiffness due to the presence of the unbonded post-tensioning
observed to extend the full width of the slab at the interior reinforcement. This behavior is apparent in both Runs 4 and 5.
connections and on the west edge of the PT specimen as
well, indicating that the post-tensioning was effective in
developing the full nominal moment strength of the slab.
Overall, the damage observed at the RC connections was
more significant and more widely distributed than the
damage observed for the PT specimen.
Significant relative rotation between the slabs and the
columns was observed during Run 4 for both the RC and PT
specimens, indicating that moment transfer strength at some
connections had been lost, that is, effectively a hinge had
formed and the connection was free to rotate. Inspection of
the connections (using a hammer to tap the concrete) after
Run 4 also indicated that connection damage had occurred.
The limited damage at punching failure is in sharp contrast
to that observed for tests of reinforced concrete flat plate
tests without shear reinforcement2 (for example, Fig. 6(d)).
The test results indicate that use of slab shear reinforcement
substantially reduces the extent of damage and, in particular,
prevents the dropping of the slab observed in reinforced
concrete connections where shear reinforcement is not Fig. 6(a) to (d): Observed damage at slab-column
provided. More detailed assessment of slab-column connection region.
punching failures and the drift that resulted in connection
failure is provided by Kang.11

Base shear versus drift ratio

Displacement transducers were used to monitor the
absolute displacements of each floor slab level and at the top
of the column footing. These measurements, corrected to
remove the lateral displacement due to footing rotation, were
used to obtain relative displacements between the second
floor (top) and the column base, as well as inter-story
displacements.11 Base shear forces were measured using
triaxial load cells as well as derived from absolute floor
acceleration records multiplied by the floor masses (Fig. 7).
In addition, interstory displacements obtained from the
external displacement transducers compare favorably with
the corresponding displacements obtained from the transducers
mounted diagonally between floor levels (Fig. 8), providing
confidence that the measured responses are reliable. Average
responses obtained for the two frames (Frames N and S in
Fig. 1) are plotted in Fig. 9(a) and (b) for Run 2-2 and 4 (RC),
and Run 2, 4, and 5 (PT), respectively. Fig. 7Comparison of base shear.

ACI Structural Journal/September-October 2005 769

Slab curvature versus column curvature punching failures and the relationship between the gravity
When punching failure occurs, the slab loses its ability to shear ratio and drift for the RC and PT specimens, as well as
transfer moment from the slab to the column; therefore, one from tests conducted by other researchers, is presented by
approach to assessing if connection punching has occurred Kang.11 Results presented by Robertson et al.4 based on
for shaketable tests is to plot the relationship between slab quasistatic tests of slab-column specimens with shear
curvature and column curvature. As slab curvature increases, reinforcement suggest that providing shear reinforcement
the column curvature should either increase or remain roughly doubles the drift capacity at which punching failures
approximately constant if the slab has yielded, unless slab are observed for isolated interior slab-column connections.
moment capacity drops (that is, punching occurs) or the Results for this study, such as the results reported in the
column yields. Column yielding is not expected because the preceding paragraph for connection RC-FL2NW (punching
columns of the two specimens were designed to be stronger at 3.12% second-story drift at punching), indicate that
than the slab; therefore, for this study, a drop in column punching failures occurred at significantly lower drift ratios
curvature when significant slab curvatures exist signals that than expected based on the quasistatic testing (punching
the strength deterioration is occurring at the slab-column capacity of approximately 7% based on Fig. 18 of Robertson
connection (that is, a punching failure has occurred). et al.4). This discrepancy could be attributed to the difference
A representative relationship for slab curvature versus in the specimens (isolated slab-column specimen versus a
column curvature is presented in Fig. 10, for additional multiple-span specimen); however, this trend is not apparent
results, refer to Kang.11 Calculated yield curvatures for the in available (although limited) test results. Although more
slab and column also are indicated on these figures. As detailed studies of the reasons for this discrepancy are
shown in Fig. 10, for positive curvatures, column curvatures needed, preliminary observations indicate that the reduced
begin to drop for higher slab curvatures, indicating that the drift capacity might be due to loss of interface shear capacity
moment transfer capacity of the slab-column connection is at the column-slab interface, which was not observed in the
degrading, which is consistent with punching failure. Therefore, quasi-static tests.
the results indicate that a punching failure occurred for
connection RC-FL2NW (refer to Fig. 1) of the RC specimen Effective flexural transfer width
at approximately t = 12.72 s for Run 4. The ratio of column Slab reinforcement strains were measured to assess if the
curvature to slab curvature (K1) at t = 12.72 s is 0.057, is test results were consistent with the ACI 318-026 defined
reduced to K2 = 0.022 at t = 30.35 s, for a relatively modest
increase in slab curvature. The smaller value of K2 relative
to K1 is a result of a punching failure. For negative bending,
column curvature remains relatively constant at 0.0001/cm
for a large range of slab curvatures (0.00006 to 0.00024/cm),
indicating that the slab has yielded, but no degradation in
the strength of the slab-column connection has occurred.
Degradation of connection moment capacity, however,
appears to occur after t = 14.38 s where K3 = 0.054,
versus K4 = 0.023 at t = 30.06 s. Similar results are noted
for other connections.11
A detailed presentation of the results for each connection
is beyond the scope of this paper, a detailed assessment of

Fig. 8Comparison of first story displacement relative

to footing. Fig. 9Base shear versus top drift.

770 ACI Structural Journal/September-October 2005

transfer width for interior and exterior connections. Rein- account for footing rotation. The slab-column joints were
forcement strains for top and bottom bars at the first story assumed to be rigid, as studies indicate that use of this
exterior connection (RC-FL1NW) and interior connection assumption tends to produce better correlation with test
(RC-FL1NC-w) are plotted in Fig. 11. Results indicate that results.22 Gravity loads tributary to the column strip were
the tensile strains in the top bars at the column centerline included in the model as a uniform load on the effective slab-
significantly exceeded the yield strain and that bars placed beam. Additional point loads were applied directly to the
within approximately c2 + 3h reached the yield (Fig. 11(a) columns to account for the tributary gravity load from the
and (c)). Bars located outside of c2 + 3h, however, reached slab middle strips.
maximum tensile strains of only 50 to 70% of the yield Appropriate values of and were determined by varying
strain. Crack patterns at exterior connections for the RC modeling parameters and comparing analytical and
specimen indicate that the moment transfer width was limited experimental responses. Values for were determined using
due to torsional cracking (Fig. 6(b)), which is consistent with the free vibration test results prior to Run 1 (that is, prior to
observations from prior studies.20 Slab reinforcement outside of conducting shaketable tests), whereas values for were
the torsional yield lines are not sufficiently anchored and thus do determined using results for Run 2. Predominant periods
not develop significant tensile strains (Fig. 11(a)). Results determined from the analytical models for Run 1 for the RC
obtained for average slab curvatures from displacement specimen were 0.233 and 0.240 s for -values of 0.75 and
gauges mounted above and below the slab produce similar 0.5, respectively, indicating that the fundamental period of
conclusions for both interior and exterior connections of the the model was relatively insensitive to changes in the -value
RC specimen.11 (Table 6). Analytical results for the PT specimen indicated
For the exterior connections of the PT specimen, torsional that the fundamental period was more sensitive to changes in
cracks formed outside the region reinforced with banded the -value, as the predominant period increased from
tendons at the east edge (Fig. 6(c)) and a flexural crack 0.271 to 0.315 s for -values of 0.75 and 0.5, respectively.
developed across the entire slab width at the west edge; Fundamental periods of the models were matched with
therefore, the banded post-tensioning tendons provided fundamental periods determined from the experimental
sufficient membrane forces to move the torsional yield line
outside the banded tendon region, or to suppress the torsional
yield line completely.

Lateral load stiffnesseffective slab width model

Analytical models for each specimen were created to
allow comparison of responses obtained with analytical
models with responses obtained during the tests. Responses
for Runs 1 and 2 (or 2-2) for each specimen were compared
to determine appropriate - and -factors for the effective
slab width model, because responses for these runs were
effectively within the elastic range. The specimens were
modeled as a plane frame using a finite element program
designed for earthquake engineering simulation.21 A fiber
model, using as-tested material properties, was used for the
columns and rotational springs (k = 565 kN-m/rad) were
included in the model at the base of column footings to

Fig. 10Curvature diagram (RC-FL2NW). Fig. 11Slab strain gauge data.

ACI Structural Journal/September-October 2005 771

results to obtain -values of approximately 0.75 and 0.70 for Response history analyses of the RC and PT models with
the RC and PT specimens, respectively. Table 6 presents the same - and -values noted in the previous paragraph
comparisons of the -factors obtained for the RC and PT were conducted and computed responses are compared to
specimens with various recommended values (Pecknold;8 experimental results in Fig. 12. Analytical results are plotted
Allen and Darvall;22 and Federal Emergency Management for two values of damping (1 and 3%), to assess sensitivity.
Agency 23), which are used to model the uncracked lateral- Portions of the response histories compare favorably for the
load stiffness of slab-column frames. The results obtained in RC specimen (Fig. 12(a)), although the peak values are not
this study are quite close to the values recommended in well represented, whereas results for the PT specimen
Federal Emergency Management Agency.23 compare quite well for 1% damping (Fig. 12(b)). The greater
Responses of the test structures to Run 2-2 (RC) or Run 2 discrepancy for the RC specimen may be due to the influence
(PT) were used to assess the reduction of slab stiffness due of footing rotations, which were measured directly for the PT
to cracking by comparing the relative stiffness values for specimen whereas footing rotations were derived for the RC
these runs with results for the free vibration tests prior to the specimen using results from the PT specimen test.18
shaketable testing. Given the -factors of 0.75 and 0.70 for
the RC and PT specimens determined from the free vibration SUMMARY AND CONCLUSIONS
testing, respectively, -values of approximately 1/3 and 2/3 Two, approximately 1/3-scale, RC and PT frames were
resulted in good agreement between experimentally and subjected to gravity loads and increasing intensity of
analytically determined periods for Runs 2-2 (RC) and 2 uniaxial base acceleration histories. Based on the observed
(PT) (Table 6). These -values are appropriate to account for damage at slab-column connections, it was concluded that
the stiffness reduction due to concrete cracking for the analysis connection punching had occurred during the test program,
of slab-column frames within the elastic range. despite the somewhat limited connection damage observed
compared with tests conducted under slowly varying loads
on specimens of similar scale. The limited damage at
punching failure is in sharp contrast to that observed for prior
quasi-static tests of reinforced concrete flat plate without
shear reinforcement. The damage observed at the RC
connections was more significant and more widely distributed
than those observed for the PT specimen. The test results
indicate that use of slab shear reinforcement substantially
reduces the extent of damage, and in particular, prevents the
dropping of the slab observed in reinforced concrete
connections where shear reinforcement is not provided. This
may be an important factor in post-earthquake repairs.
An evaluation of measured responses indicates that the
moment transfer at the slab-column connections degraded
during the tests; however, mean lateral drift ratios of
approximately 3 and 4% were achieved for the RC and PT
specimens, respectively, with relatively little loss of lateral
load capacity for the system. The ability of the systems to
reach these relatively high deformation levels was impacted
by several factors. First, a capacity design approach was used
for the columns to force yielding into the slab as well as
preclude column yielding or shear failure. Second, shear
Fig. 12Top relative displacement histories. (stud-rails) reinforcement was used to enhance post-punching

Table 6Effective slab width factors and

Tp1, s* Tp1, % Tp1 , s* Tp1, %
high = 1.00 0.222 5 high = 1/2 0.267 10
Reinforced- concrete avg = 0.75 0.233 0 avg = 1/3 0.297 0
slab-column frame specimen
low = 0.50 0.240 +3 low = 1/6 0.343 +16
high = 1.00 0.261 8 high = 1 0.289 4
Post-tensioned slab-column avg = 0.70 avg = 2/3
frame specimen 0.285 0 0.302 0
low = 0.50 0.315 +11 low = 1/3 0.333 10
Reinforced-concrete slab-column frame specimen Post-tensioned slab-column frame specimen
Measured- 0.75 0.70
Theoretical Theoretical/Measured Theoretical Theoretical/Measured
Pecknold7 0.78 1.04 0.71 1.01
19 0.57 0.76 0.57 0.81
Allen and Darvall
20 0.74 0.99 0.61 0.87
FEMA-274
*Tp1 = Analytical results obtained using a finite element program.21

772 ACI Structural Journal/September-October 2005

deformation capacity, and lastly, the ratio of top to bottom under Vertical and Lateral Loads, PhD thesis, Department of Civil and
reinforcement for unbalanced moment transfer within the Environmental Engineering, University of California-Berkeley, Berkeley,
Calif., 1989, 270 pp.
transfer width of c2 + 3h was maintained as approximately 3. Moehle, J. P., Seismic Design Considerations for Flat-Plate Construc-
two. A capacity design approach is uncommon in the U.S., tion, Mete A. Sozen Symposium, SP-162, J. K. Wight and M. E. Kreger,
and the use of bottom reinforcement within c2 + 3h for eds., American Concrete Institute, Farmington Hills, Mich., 1996, 460 pp.
gravity systems within high seismic zones is not common. 4. Robertson, I. N.; Kawai, T.; Lee, J.; and Enomoto, B., Cyclic Testing
of Slab-Column Connections with Shear Reinforcement, ACI Structural
Punching failures were verified by plotting slab curvature Journal, V. 99, No. 5, Sept.-Oct. 2002, pp. 605-613.
versus column curvature (or unbalanced moment) relations 5. International Conference of Building Officials, Uniform Building
for a representative connection. Punching failures were Code, Whittier, Calif., 1997, 492 pp.
noted where column curvature dropped as slab curvatures 6. ACI Committee 318, Building Code Requirements for Structural
increased. Given the gravity shear ratios on the connections, Concrete (ACI 318-02) and Commentary (318R-02), American Concrete
Institute, Farmington Hills, Mich., 2002, 443 pp.
preliminary results for the drift ratios at which punching failures 7. ACI Committee 318, Building Code Requirements for Structural
were observed are substantially less than those suggested based Concrete (ACI 318-05) and Commentary (318R-05), American Concrete
on tests of similar scale specimens subjected to quasistatic, Institute, Farmington Hills, Mich., 2005, 430 pp.
reversed cyclic loading, possibly due to the loss of interface 8. Pecknold, D. A., Slab Effective Width for Equivalent Frame Analysis,
ACI Structural Journal, V. 72, No. 4, Apr. 1975, pp. 135-137.
shear capacity at the slab-column interface. 9. Megally, S. H., Punching Shear Resistance of Concrete Slabs to Gravity
Based on measured reinforcing bar strain data and slab and Earthquake Forces, PhD thesis, Department of Civil Engineering,
curvature data, yielding of slab flexural reinforcement was University of Calgary, Calgary, Alberta, Canada, 1998, 468 pp.
found to be limited to the effective flexural transfer width of 10. Martinez-Cruzado, J. A.; Qaisrani, A.-N.; and Moehle, J. P.,
c2 + 3h for both the interior and exterior connections, and Post-Tensioned Flat-Plate Slab-Column Connections Subjected to Earth-
quake Loading, Proceedings of 5th National Conference on Earthquake
yield lines formed at approximately 45 degrees for the exterior Engineering, Earthquake Engineering Research Center, University of
connections of the RC specimen. These findings are generally California-Berkeley, Berkeley, Calif., 1994, pp. 139-148.
consistent with existing ACI 318-026 provisions. 11. Kang, T. H.-K., Shake Table Tests and Analytical Studies of
Analytical studies were conducted to assess appropriate Reinforced and Post-Tensioned Concrete Flat Plate Frames, PhD thesis,
Department of Civil and Environmental Engineering, University of California
modification factors to use for effective slab width () and at Los Angeles, Los Angeles, Calif., 2004, 299 pp.
cracking () for lateral load analysis of slab-column frames. 12. Ghali, A., and Hammill, N., Effectiveness of Shear Reinforcement
Using an effective slab with model and a column fiber model, in Slabs, Concrete International, V. 14, No. 1, Jan. 1992, pp. 60-65.
-values of 0.75 and 0.70, and -values of 1/3 and 2/3, resulted 13. Kang, T. H.-K., and Wallace, J. W., Shake Table Tests of Reinforced
in generally good correspondence between experimental and Concrete Flat Plate Frames and Post-Tensioned Flat Plate Frames,
Proceedings of 13th World Conference on Earthquake Engineering,
analytical results for periods and top displacements in the elastic Vancouver, British Columbia, Canada, 2004.
range, and are similar to the values suggested in Federal 14. Joint ACI-ASCE Committee 421, Shear Reinforcement for Slabs
Emergency Management Agency.23 (ACI 421.1R-99), American Concrete Institute, Farmington Hills, Mich.,
1999, 15 pp.
15. PTI Committee for Unbonded Tendons, Field Procedures Manual for
ACKNOWLEDGMENTS Unbonded Single Tendons, Post-Tensioning Institute, Phoenix, Ariz., 2000.
The work presented in this paper was sponsored by Phase IV of the joint 16. Joint ACI-ASCE Committee 423, Recommendations for Concrete
CUREE-Kajima Research Program. The authors would like to acknowledge the Members Prestressed with Unbonded Tendons (ACI 423.3R-96), American
active participation of Kajima Corp. researchers K. Igarashi and N. Suzuki Concrete Institute, Farmington Hills, Mich., 1996, 19 pp.
throughout the project, as well as D. Cylde, W. Neighbour, S. Takhirov, 17. Nilson, A. H.; Darwin, D.; and Dolan, C. W., Design of Concrete
and C. Machado at the Earthquake Engineering Research Center at the Structures, McGraw Hill, Singapore, 2004, 896 pp.
University of California-Berkeley, Berkeley, Calif., for their assistance. The 18. Kang, T. H.-K.; Rha, C.; Wallace, J. W.; Igarashi, K.; and Suzuki, N.,
views expressed are those of authors and do not necessarily represent those Seismic Performance Assessment of Flat Plate Floor Systems, CUREE-
of the sponsor. Kajima Joint Research ProgramPhase IV, Report No. CKIV-02, 2003.
212 pp.
NOTATION 19. Harris, A. H. N., and Sabnis, G. M., Structural Modeling and
Ac = area of slab critical section Experimental Techniques, CRC Press, New York, 1999, 808 pp.
c = distance from centroid of critical section to perimeter of critical section 20. Joint ACI-ASCE Committee 352, Recommendations for Design of
Jc = polar moment of inertia of critical section Slab-Column Connections in Monolithic Reinforced Concrete Structures (ACI
vc = nominal concrete shear stress 352.1R-89), American Concrete Institute, Farmington Hills, Mich., 22 pp.
vn = nominal shear stress 21. OpenSees Development Team, OpenSees: Open System for
vu = combined direct gravity shear stress and eccentric shear stress Earthquake Engineering Simulations, Version 1.5, Berkeley, Calif., 2000.
22. Allen, F. H., and Darvall, P., Lateral Load Equivalent Frame, ACI
JOURNAL, Proceedings V. 74, No. 7, July 1977, pp. 294-299.
REFERENCES 23. Federal Emergency Management Agency, NEHRP Commentary on
1. Moehle, J. P., and Diebold, J. W., Experimental Study of the Seismic the Guidelines for the Seismic Rehabilitation of Buildings, Report
Response of a Two-Story Flat-Plate Structure, Report No. UCB/EERC-84/08, No. FEMA-274, Washington DC, 1997, 440 pp.
Earthquake Engineering Research Center, University of California-Berkeley, 24. Arias, A., A Measure of Earthquake Intensity, Seismic Design for
Berkeley, Calif., 1984, 244 pp. Nuclear Power Plant, R. J. Hansen, ed., MIT Press, Cambridge, Mass.,
2. Hwang, S. J., An Experimental Study of Flat-Plate Structures 1970, pp. 438-483.

ACI Structural Journal/September-October 2005 773