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

net/publication/265244525

Punching Shear of Thick Plates with and without Shear Reinforcement

Article in ACI Structural Journal · September 2011

CITATIONS READS

30 1,387

3 authors:

Emad Rizk Hesham Marzouk

Memorial University of Newfoundland Ryerson University
13 PUBLICATIONS 137 CITATIONS 171 PUBLICATIONS 3,016 CITATIONS

SEE PROFILE SEE PROFILE

A. Hussein
Memorial University of Newfoundland
72 PUBLICATIONS 1,011 CITATIONS

SEE PROFILE

All content following this page was uploaded by Hesham Marzouk on 19 February 2015.

The user has requested enhancement of the downloaded file.

 ACI STRUCTURAL JOURNAL TECHNICAL PAPER
Title no. 108-S55

Punching Shear of Thick Plates with and without

Shear Reinforcement
by E. Rizk, H. Marzouk, and A. Hussein

oc m
Thick concrete plates are currently used for offshore and nuclear
containment concrete walls. In this research, five thick concrete
slabs with a total thickness of 300 to 400 mm (12 to 16 in.) were
tested under concentric punching loading. Four specimens had no
.fa
As a consequence, there is a need for a rational model that
correctly describes and accounts for size effect. This paper
presents an experimental investigation. Five thick plates
with different flexural reinforcement ratios were tested

glo
shear reinforcement, whereas the remaining one included T-headed
shear reinforcement consisting of vertical bars mechanically to examine the accuracy of available design equations. A
anchored at the top and bottom by welded anchor plates. The simplified practical punching shear equation that accounts
main focus of this research was to investigate the influence of the for the size effect factor is proposed. The proposed equation
size effect on the punching shear strength of thick high-strength is verified using experimental test results and is compared
concrete plates. All tests without shear reinforcement exhibited with the predictions of different design codes.

.ht b
brittle shear failures. The addition of T-headed shear reinforcement
with a shear reinforcement ratio of approximately 0.68% by volume
changed the failure mode to ductile flexural failure. The test results
Previous research
revealed that increasing the total thickness from 350 to 400 mm (14 to Although extensive research has been done on the punching
16 in.) resulted in increased punching capacity and at the same time shear strength of slabs, to date there is still no generally
resulted in a small increase in ductility characteristics. An equation applicable rational theory. There is also great discrepancy
based on fracture mechanics principles is recommended to account between different design codes. Some of these codes, such
for the size effect factor. The proposed equation is verified using

ksa
as ACI 318-08,1 do not even account for basic and proven
the test results and is compared with the predictions of different factors affecting the shear capacity of concrete members,
design codes.
such as reinforcement ratio and size effect. The analysis of
Keywords: punching; shear reinforcement; size effect; thick plates. databases available from the literature indicates a lack of
experimental data regarding member size.3 The testing of
INTRODUCTION thick slabs represents a great challenge to many researchers

tria -
North American design codes such as ACI 318-081 and due to expensive test setups compared to limited research
CSA A23.3-042 provide practical equations to design flat budgets and limited lab spaces.
plates subjected to punching loads. The design requirements Richart4 presented the results of a number of reinforced
provided by the ACI code1 do not account for the size effect concrete footing tests. The researcher reported that high
factor. At the same time, the size effect factor provided by tensile stresses in the flexural reinforcement led to extensive
the Canadian standard2 is empirical and does not apply to cracking in the footings. This cracking reduced the ability

iod
slabs having an effective depth less than 300 mm (12 in.). of the section to resist shear, resulting in the footings failing
The main reason for disregarding the size effect is the lack of at lower shearing stresses than expected. The maximum
enough experimental tests, especially for thick high-strength shearing stresses, computed at the conventional critical
concrete slabs. North American design codes also do not have section, varied considerably with the effective depth of the
provisions for minimum shear reinforcement requirements
footing, being larger for the thinner footings.
for thick slabs over 300 mm (12 in.) in thickness; these
Based on the results of a finite element parametric study
provisions could be required to prevent the brittle shear
failure behavior of thick plates due to size effect. One more on 81 slabs by Marzouk et al.,5 the authors recommended a
nontraditional expression for shear strength for high-strength

.w m
reason to provide shear reinforcement requirements is to
allow the use of a slab with a large amount of the flexural concrete application as follows
reinforcement ratio. Increasing the flexural reinforcement
−0.4
ratio increases the punching capacity but strongly decreases C l 
the deformation capacity of the slab. The focus of this research vu = 0.88 ft   3 r  ch  (1)
 d  h
is to experimentally investigate the influence of size effect on
the punching shear strength of thick high-strength concrete

ww
plates to better understand the punching mechanism of where ft is the direct tensile strength of concrete; C is the
thick plates. column side length; d is the slab depth; lch is the characteristic

RESEARCH SIGNIFICANCE
Rational models and design formulae for punching shear ACI Structural Journal, V. 108, No. 5, September-October 2011.
are based on the results of experimental tests performed MS No. S-2009-398.R3 received July 29, 2010, and reviewed under Institute
publication policies. Copyright © 2011, American Concrete Institute. All rights
mostly on thin slabs. Design codes, however, are also used reserved, including the making of copies unless permission is obtained from the
in the design of thick plates and footings. The few available copyright proprietors. Pertinent discussion including author’s closure, if any, will be
published in the July-August 2012 ACI Structural Journal if the discussion is received
tests performed on thick slabs exhibit a notable size effect. by March 1, 2012.

ACI Structural Journal/September-October 2011 581

E. Rizk is a PhD Candidate and Research Assistant at Memorial University of
than 35 degrees for the more slender footings (a/d = 2.0).
Newfoundland, St. John’s, Newfoundland, Canada, and is an Assistant Lecturer This failure angle seems to be steeper than that for a flat
at Menoufia University, Shebin El-Kom, Egypt. He received his BSc and MSc from slab. Furthermore, the a/d significantly affected the punching
Menoufia University in 1999 and 2005, respectively. His research interests include shear capacity.
cracking of offshore structures and shear strength of two-way slabs. Birkle and Dilger11 studied the influence of slab thickness
ACI member H. Marzouk is the Chair of the Civil Engineering Department at on punching shear strength. A total of nine slab-column
Ryerson University, Toronto, ON, Canada. He received his BSc from Cairo University, assemblies were tested. It was concluded that without shear
Giza, Egypt, and his MSc and PhD from the University of Saskatchewan, SK, Canada. reinforcement, a slab of 230 mm (9.1 in.) thickness may not
He is a member of ACI Committees 209, Creep and Shrinkage in Concrete, and have a sufficient factor of safety if designed in accordance
213, Lightweight Aggregate and Concrete. His research interests include structural
and material properties of high-strength concrete, lightweight high strength, creep,
with ACI 318-08.1
fracture mechanics, and finite element analysis. One of the effective solutions to the problem of size effect
for thick plates is to include provisions for shear reinforcement
ACI member A. Hussein is an Associate Professor of civil engineering at Memorial (for example, shear studs). The extensive investigations that
University of Newfoundland. He received his BSc from Ain-Shams University, Cairo,
Egypt, in 1984, and his MEng and PhD from Memorial University of Newfoundland in
were conducted by different researchers12-14 on full-size slab-
1990 and 1998, respectively. His research interests include the use of fiber-reinforced column connections verified that stud-type reinforcement can
polymer in concrete, the mechanical and structural behavior of high-strength concrete, substantially increase the strength and prevent brittle failure
and nonlinear finite element analysis of concrete structures. of thick plates. Berner et al.15 developed T-headed bars as
shear reinforcement for thick flat plates. The advantage of
this type is the ability to replace conventional stirrups and
length, which is described later in the paper; and h is the hence facilitate concrete placement and vibration.
slab thickness. Different codes and design guidelines allow the use of
The punching shear behavior of interior slab-column shear reinforcement for thick plates. Nonetheless, there
connections was investigated by Kevin.6 The response of is no accepted code-based formula to provide minimum
six two-way slab specimens, which were designed such shear reinforcement requirements for thick plates. The only
that they would fail in punching shear, was examined. The exception is for the Eurocode 2,16 which requires that where
overall thickness h of the slabs varied from 135 to 550 mm shear reinforcement is required in reinforced concrete slabs,
(5.3 to 21.6 in.). The results showed a strong size effect with the minimum area of shear reinforcement must be equal to
deeper members having smaller shear stresses at failure than that needed for the same cross section that is to be designed
shallow ones. It was also concluded that there is a significant as a beam.
size effect for effective depths greater than approximately
200 mm (8 in.). Experimental Investigation
Regan and Braestrup7 studied the effect of the shear The main variables in the experimental investigation were
span-depth ratio (a/d) on the shear strength. Although the the reinforcement ratio, concrete strength, and slab effective
data in this area are limited, it could be concluded that the depth. A total of five thick concrete plates were tested. One
shear strength rises quite sharply when the a/d is less than normal-strength concrete (NSC) plate and four high-strength
approximately 1.5 but is relatively constant for larger values concrete (HSC) plates were selected, as detailed in Table 1. In
of the ratio. For very short shear spans a, the support location numbering the specimens, an extra “S” was used to represent
significantly interferes with the failure surface. According to the size effect parameter and was added to the specimens’
Hallgren,8 slender slabs are those slabs with a/d of more than numbers. This was done to distinguish between specimens in
3 to 4. This ratio could be used to distinguish between thick this research and previous specimens tested and published by
and thin slabs. the authors.17 In this research, the thicknesses ranged from
Experimental investigations on the punching behavior 300 to 400 mm (12 to 16 in.). The specimens were designed
of reinforced concrete footing were conducted recently by to examine the effect of depth on the structural behavior of
Hegger et al.9,10 A total of 22 reinforced concrete footings thick concrete plates. The slabs had an a/d of 3.33 to 4.8. The
were tested to investigate the punching shear behavior of details of a typical test specimen are shown in Fig. 1.
footings. The test parameters were a/d, concrete strength, and All the specimens were designed to fail under punching
punching shear reinforcement. The experimental investigation failure except Specimen HSS1, which was designed to fail
indicated that for the footings without shear reinforcement, under flexure. Specimen HSS2 included T-headed shear
the inclination of the failure shear crack seems to be mainly stud reinforcement and was designed to examine the effect
influenced by a/d and not by the concrete strength. The of shear reinforcement on the structural behavior of thick
observed inclinations of the failure crack were approximately concrete plates. The shear reinforcement consisted of vertical
45 degrees for the compact footings (a/d = 1.25) and less bars with a diameter of 15 mm (No. 5) and a specified yield

Table 1—Details of test specimens

Compressive Bar spacing, Concrete cover Slab thickness, Flexural reinforcement Shear reinforcement
Slab no.* strength fc′, MPa Bar size, mm mm Cc, mm mm Depth, mm ratio r, % ratio rz , %
HSS1 76 25 368 70 350 267.5 0.50 —
HSS2 79 25 195 70 300 217.5 1.42 0.68
HSS3 65 35 289 70 350 262.5 1.42 —
NSS1 40 35 217 70 400 312.5 1.58 —
HSS4 60 35 217 70 400 312.5 1.58 —
*
NS is normal-strength slabs; HS is high-strength slabs.
Notes: 1 MPa = 145 psi; 1 mm = 0.0394 in.

582 ACI Structural Journal/September-October 2011

strength of 400 MPa (58 ksi). The bars were anchored at the
top and bottom by welded anchor plates. The layout of the
shear reinforcement is shown in Fig. 2.

Materials
Type 10 SF cement blended with silica fume was used for
all the mixtures. Local fine aggregate that had a composition
similar to that of the coarse aggregate was used. Crushed
sandstone fine aggregate and crushed sandstone with a
coarse aggregate of 19 mm (0.75 in.) maximum nominal
size were used. Two concrete mixtures with a compressive
strength of 35 and 70 MPa (5076 and 10,152 psi) after
28 days were used. The water-cement ratio (w/c) was 0.55 and
0.29 for the NSC and HSC mixtures, respectively. A non-
chloride water-reducing agent of polycarboxlate base and a
retarder of organic base conforming to ASTM C494 Types C
and D was adopted. Three 150 x 300 mm (6 x 12 in.) cylinders
Fig. 1—Dimensions and reinforcement details of Specimen HSS2
were cast from each batch and used to determine the concrete
with T-headed shear reinforcement. (Note: Dimensions in mm;
compressive strength. Reinforcing bars consisting of Grade 400
1 mm = 0.0394 in.)
steel—conforming to CSA standards with an actual tested
yield strength of 460 MPa (67 ksi)—and a yield strain of
approximately 2250 µε were used.

Test specimens
The test slabs had a side dimension of 2650 mm (106 in.)
in both directions. The test specimens were simply supported
along all four edges with the corners free to lift. A concentric
load was applied on the slabs through a 400 x 400 mm (16 x
16 in.) column stub. The specimen represents the region of
negative bending moment around an interior column and the
simply supported edges simulate the lines of contra-flexure.
Reinforcement ratios of 0.50, 1.42, and 1.58% were selected
for flexural reinforcement. The compression reinforcement
ratios were selected to satisfy ACI 318-081 for the minimum
reinforcement ratio to control shrinkage. The T-headed shear
reinforcement consisted of a 15 mm (No. 5) bar as a stem
and two 30 x 70 mm (1.2 x 2.8 in.) steel plates individually
welded to both ends as anchor plates. A total of 40 T-headed
studs were placed in Specimen HSS2, as shown in Fig. 2.
The rectangular plate anchors had an area that was at
least 10 times the area of the stem. 12 The stud spacing
s was chosen to be 0.5d as the lower of two spacings
recommended by Joint ACI-ASCE Committee 421.18 The
shear studs were extended to approximately 2d from the Fig. 2—Arrangement of T-headed shear reinforcement in
column faces as recommended by Marzouk and Jiang. 14 Specimen HSS2 (h = 300 mm [12 in.]). (Note: Dimensions
The distance between the first row of studs and the in mm; 1 mm = 0.0394 in.)
column face s o was taken as 0.4d to avoid shear failure
between the column and the first row of shear studs.
walls and to ensure that the test setup would act as a rigid
Test setup self-supporting unit. A picture of the test setup is shown in
A new test setup was designed and fabricated in Fig. 3(a). A hydraulic jack was used to apply a concentric
the structural laboratory of Memorial University of load on the column stub in a horizontal position. The jack
Newfoundland (MUN). The main function of this setup is was a hydraulic jack cylinder with a maximum capacity of
to apply direct transverse load through a hydraulic jack. 3110 kN (700 kips) and a maximum displacement of
The test setup consists of four retaining walls; two of them 300 mm (12 in.).
were used for supporting steel beams. The beams act as a
support for the test slab. The steel beams were anchored Test procedure
to the retaining walls. The walls were anchored to the 1 m The test slabs were placed in the frame in a vertical position.
(3.28 ft) thick structural floor. The third and fourth retaining The slabs were simply supported along all four edges with
walls were used to support the hydraulic jack that applied the corners free to lift. The test specimens were instrumented
the load directly on the column stub. The retaining wall units to measure the applied load, central deflection, strains in
were restrained at the top and lower edges by self-supporting concrete, and rein­forcement. The load was applied at a
closed rigid steel frames. The function of the frames was to selected load increment of 44.0 kN (10 kips). The load-
minimize the lateral displacement of the supporting retaining deflection curves were obtained using the linear variable

ACI Structural Journal/September-October 2011 583

 (a) (b)

(c) (d)

(e) (f)

Fig. 3—Test setup: (a) photograph; (b) arrangement of LVDTs; (c) locations of steel strain gauges; (d) locations of concrete
strain gauges; (e) crack displacement transducer (CDT); and (f) typical specimen indicates size effect challenge. (Note:
Dimensions in mm; 1 mm = 0.0394 in.)

differential transformer’s (LVDT’s) measurements at four locations on the compression faces of the concrete slabs. The
predetermined locations on the tension surface, as shown in concrete strains were measured with electrical 50 mm (2 in.)
Fig. 3(b). The slabs were carefully inspected at the end of strain gauges, as shown in Fig. 3(d).
each load step. Steel strains at five locations were monitored, The cracks were marked manually and the maximum
as shown in Fig. 3(c). Concrete strains were recorded at eight visible crack width was measured using CDTs. The gauges

584 ACI Structural Journal/September-October 2011

were mounted at the concrete surface cracks to measure the slab. In all the test slabs, the initial observed cracks were
crack opening displacement, as shown in Fig. 3(e). The crack first formed tangentially under the edge of the column stub,
displacement transducer is a waterproof-enabled gauge followed by radial cracking extending from the column edge
with a length of 64 mm (2.6 in.). The range of the gauge is toward the edge of the slab. As the load increased, the tension
between ±2 to ±5 mm (±0.08 to ±0.2 in.). The accuracy of
reinforcement yielded, which resulted in a significant in­
the measurements improved as the cracks started to widen.
The testing of thick specimens posed some challenge, as crease in the crack width and deflection. It was noted that
shown in Fig. 3(f). The data from the transducers were the ratio of the yield load to the cracking load increased with
measured using a high-speed data acquisition system and increasing the reinforcement ratio.
were stored on a personal computer. All tests were terminated
after punching occurred and the load dropped considerably. Load-deflection characteristics
The applied load versus the de­flection at the center of the
Test results slab for the different specimens is shown in Fig. 4. The yield
The first crack in each specimen was visually inspected displacement is calculated at the first yield of reinforcement
and the corresponding load was recorded. The yield steel bars. The first yielding of the bottom reinforcement is
strain was assumed to occur at 2000 με, which produced a
indicated by a circle on each curve. Table 2 presents the
stress in the steel reinforcing bar equal to 400 MPa (58 ksi).
The yield strain was measured first at the center of the measured deflection at first crack, at first yield of tension
steel, at ultimate load, and at post-ultimate load. It should be
noted that the load-deflection curves can be used in classifying
failure type.17 Post-ultimate loading capacity refers to
slab capacity at ultimate deflection. Also, post-ultimate
deflection is commonly defined as the deflection in the post-
peak load range and when approximately 20% of the peak
load is lost (that is, deflection when the load is 80% of the
peak value). For slabs failing in punching, there could be a
complete loss of load-carrying capacity for some slabs that
failed in punching. Hence, a descending portion of the load-
deflection curve would not be obtained in that case. As a
result, it may not be possible to apply the common definition
of post-ultimate deflection.
The load-deflection curve of Specimen HSS1 indicated
that it failed in flexure. The slab reached the state of steadily
increasing deflection at constant load. Thus, it displayed
ductile behavior characterized by a continuously increasing
capacity with increasing deflection after overall yield of the
flexural reinforcement. This is a normal charac­teristic for a
lightly reinforced concrete specimen experiencing flexural
failure. Specimen HSS2 failed in flexure failure, as indicated
by its load-deflection curve. Specimens HSS3, NSS1, and
HSS4 failed in punching; this is characterized by a sudden
drop in the load-deflection curve.
The inclusion of T-headed shear reinforcement
(Specimen HSS2) improved the ultimate load capacity
as compared to that of Specimen HSS3 without shear
Fig. 4—Load-deflection characteristics at center span reinforcement. The provision of shear enhancement
of test slabs: (a) Specimens HSS1, HSS2, and HSS3; and also provided a post-ultimate behavior after the ultimate
(b) Specimens HSS1 and HSS4. (Note: Dimensions in loading ca­pacity was reached and eliminated the so-called
mm; 1 mm = 0.0394 in.; 1 kN = 0.2248 kips.) punching shear failure. On the other hand, the punching

Table 2—Deflection characteristics of test slabs

Concrete Flexural Yield load Ultimate Ultimate load
steel strength reinforcement ratio First crack First crack Yield load deflection Dy, load Pu, deflection DPu, Post-ultimate Post-ultimate load
Slab no. fc′, MPa r, % load, kN deflection, mm Py, kN mm kN mm load PDu, kN deflection Du, mm
HSS1 76 0.50 312 5.9 790 7.3 1722 22.5 360 32.0
HSS2 79 1.42 320 5.7 1219 18.9 2172 35.5 1067 43.5
HSS3 65 1.42 258 7.0 1381 17.3 2090 24.1 617 35.1
NSS1 40 1.58 276 4.1 2094 11.5 2234 13.1 549 24.6
HSS4 60 1.58 317 4.4 2081 9.4 2513 13.1 507 25.7
Notes: 1 MPa = 145 psi; 1 mm = 0.0394 in.; 1 kN = 0.2248 kips.

ACI Structural Journal/September-October 2011 585

Table 3—Observed ductility and energy absorption failure occurred and no post-ultimate loading capacity
was observed for Specimen HSS3.
Concrete Flexural
strength fc′, reinforcement Ductility, Energy absorption
Slab no. MPa ratio r, % Du/Dy capacity, kN.mm × 103 Ductility and energy absorption characteristics
HSS1 76 0.50 4.38 33.44 Ductility is a term that reflects the deformation capacity
HSS2 79 1.42 2.30 50.00 of a structural member before failure. Ductility U is defined
HSS3 65 1.42 2.03 35.19 as the ratio of the post-ultimate deflection ΔU to the deflection
NSS1 40 1.58 2.14 29.98 at first yield Δy. The energy-absorption capacity is defined
HSS4 60 1.58 2.73 39.53
as the area under the load-deflection curve. The ductility at
Notes: 1 MPa = 145 psi; 1 kN.mm = 0.00886 kip.in.
failure and the energy-absorption capacity of all test slabs,
as de­fined previously, are given in Table 3. The test results
indicated that as the steel reinforcement increased, the
ductil­ity decreased. For example, increasing the reinforce­ment
ratio from 0.50 to 1.42% de­creased the ductility by more than
50%. The test results revealed that as the depth of the slab
increased, the ductility decreased. HSC Specimen HSS3 had
almost the same reinforcement ratio and concrete strength
compared to Specimen HSS4. The results showed that the
ductility and energy absorption capacity of Specimens HSS3
and HSS4 were almost the same; this could indicate that after
a certain depth limit (d = 260 mm [10.4 in.]), increasing the
effective depth resulted in increasing the punching capacity
but at the same time did not result in a significant increase
in ductility and energy absorption. In addition, by increasing
the slab effective depth, the structural behavior became more
brittle. This is known as the size effect.
The ductility of Specimen HSS2 was almost the same as
the ductility of Specimen HSS4. At the same time, the energy
Fig. 5—Load-concrete strain for typical test Specimen HSS2. absorption capacity of Specimen HSS4 was approximately
(Note: Dimensions in kN; 1 kN = 0.2248 kips.) 80% of Specimen HSS2. This reflected the enhanced structural
behavior of Specimen HSS2 by using shear reinforcement.
The brittle structural behavior of Specimen HSS4 can be
transformed to ductile structural behavior by using shear
reinforcement. Adding shear reinforcement ensures using the
full benefit due to increasing the slab effective depth. The
test results indicated that increasing the slab thickness
from 350 to 400 mm (14 to 16 in.) resulted in increased
punching capacity and at the same time resulted in only
a 25% increase in ductility characteristics. The possible
explanation for the slight increase in the ductility ratio
for Specimen HSS4 compared to Specimen HSS3 is the
increase in brittleness of Specimen HSS4; this is due to the
size effect factor.
Flexural reinforcement alone cannot provide adequate
ductility of slab-column connections, especially when
deformations are large (for example, during seismic events).
Adding shear reinforcement to the slabs at the column area
can substantially increase the punching shear capacity and
ductility, which was shown by several researchers (Dilger
and Ghali13 and Megally and Ghali19).

Concrete strains
For all the test slabs, measurements were made to
determine the distribution of the concrete strain along a
radius of the slab. Figure 5 shows the load-concrete strain
for a typical test specimen. None of the concrete strains in
the tangential or the radial directions reached a limiting
value of 3000 με for any of the test slabs, except for
Specimen HSS2. The concrete strains for Specimen HSS2
reached a value of almost 3000 με at a distance equal to
Fig. 6—Load-steel strain for test slabs: (a) Specimens HSS1, 100 mm (4 in.) from the column face. The load-concrete
HSS2, and HSS3; and (b) Specimens NSS1 and HSS4. (Note: strain curves for Specimens HSS1, HSS3, NSS1, and HSS4
Dimensions in kN; 1 kN = 0.2248 kips.) were linear until the first cracking load.

586 ACI Structural Journal/September-October 2011

Fig. 7—Crack pattern of test Specimen HSS1 (radial Fig. 8—Crack pattern of test Specimen HSS4 (orthogonal
crack pattern.) crack pattern.)

Steel strains tangential crack; orthogonal cracking was the most dominant
Measurements were made to determine the strain crack pattern (Fig. 8).
distribution along a radius for all test slabs. The measured For Specimen HSS2, inclined shear cracks were observed
steel strains are shown in Fig. 6. For all slabs, the tension beyond a distance 2d from the column face. From the test
reinforcement yielded before punching occurred. The degree observations, flexural cracking occurred first, and it advanced
to which yielding spread in the tension steel varied as the roughly from the column outlines towards the slab edges
reinforce­ment ratio changed. For high reinforcement levels, parallel to flexural reinforcement. Subsequently, tangential
the yielding of the tension reinforcement occurred at higher cracks developed around the vicinity of the column outline.
applied loads and was localized around the column stub. For In terms of inclined cracking, the T-headed shear studs were
the lightly reinforced Specimen HSS1, yielding initiated at the not intersected by the inclined shear cracks. The final failure
column stub and gradually progressed throughout the whole progressed with most of the reinforcement pulling out of the
tension reinforcement. Moreover, Specimen HSS1 reached outer part of the slab. Quite large areas enveloped by outmost
the state of steady steel strains at a constant load, which is tangential cracks were observed on the tension face of the
a normal behavior of slabs failing in flexure. The highest slab. In that case, the punching shear failure was eliminated,
strain and consequently initial yielding occurred below the and the failure transformed into a ductile flexure one.
stub column. Specimen HSS2 experienced flexure failure,
with the slab failing at 100% of the flexure strength. For Crack spacing
both Specimens HSS1 and HSS2, which used full flexural Numerous cracks were developed on the tension face of
capacity, the yielding of flexural reinforcement occurred at the slab at the time of failure. It was found that, for all the
almost 50% of the failure load. Specimens HSS3, NSS1, specimens, the first crack formed along the reinforcing bar
and HSS4 experienced punching failure. The yielding of the passing through the slab center or close to the slab center. The
flexural reinforcement occurred at almost 75 to 85% of the second crack formed along the similar reinforcing bar in the
ultimate load. In general, the slope of the load-strain curve other direction. CDTs were mounted on the concrete surface
was almost the same for HSC and NSC slabs that failed of the first, second, and third visible cracks to measure the
in shear. Regarding the T-headed shear studs, none of the crack-opening displacement. The corresponding load at
shear studs reached yielding. The contribution of the shear which each crack occurred was recorded. The cracks formed
reinforcement did not take place before the slab reached at this stage had no effect on the characteristics of the crack
60% of the ultimate load. The slope of the load-strain curve pattern and primarily depended on the concrete strength.
for shear studs was small and gradual. One parameter—namely, the bar spacing—was examined to
investigate its effect on crack spacing and crack width.
Cracking and failure characteristics All slabs, except Specimen HSS1, exhibited an orthogonal
For the slab failing in flexure (Specimen HSS1), the crack crack pattern that formed along the direction of the
pattern observed prior to punching consisted of one tangential reinforcement. The orthogonal cracks were a function of the
crack, roughly at the column outline, followed by radial bar spacing as it was noticed for Specimens HSS2, HSS3,
cracking that extended from the col­umn. Flexure yield lines NSS1, and HSS4. Once the bar spacing was increased,
were well developed (Fig. 7). This failure can be classified the average orthogonal crack spacing was increased. For
as flexure failure. For the slabs failing by punching, the crack Specimens HSS1, HSS3, and HSS4, the average crack
pattern observed prior to punching consisted of almost no spacing was less than the bar spacing (Table 4). In

ACI Structural Journal/September-October 2011 587

Table 4—Measured crack spacing and crack width
Slab no. Concrete cover Cc, mm Slab thickness h, mm Bar spacing s, mm fc′, MPa Average crack spacing Sm, mm Maximum crack width wk, mm
HSS1 70 350 368 76 221 0.581
HSS2 70 300 217 79 228 0.876
HSS3 70 350 289 65 264 0.435
NSS1 70 400 217 40 250 0.439
HSS4 70 400 217 60 210 0.469
Notes: 1 MPa = 145 psi; 1 mm = 0.0394 in.

thick concrete cover. At sections between successive cracks,

some tensile stress is retained in the concrete around steel
bars due to the action of bond, contributing to the bending
stiffness of the member; this is reflected by a reduction in
tensile strain in the reinforcement. This is called the tension-
stiffening effect. The crack width can be calculated by
multiplying the crack spacing by the average steel strain after
reducing the crack width due to tension stiffening. The steel
strain can be determined at any loading by determining the
neutral axis and assuming linear strain distribution.

Effectiveness of shear reinforcement

The shear reinforcement has little effect before the
occurrence of the inclined shear cracks inside the slab. After
the development of the inclined shear cracks, however, shear
Fig. 9—Steel strain versus crack width for typical test reinforcement transfers most of the forces across the shear
Specimen HSS2. (Note: Dimensions in mm; 1 mm = crack and delays the further widening of the shear crack,
0.0394 in.) thus increasing the punching shear capacity and ductility of
the slab. To achieve this, the reinforcement needs to be well
Specimen HSS1, the crack pattern was radial, which is a anchored and have enough ductility to allow the mobilization
normal behavior of a slab having a small reinforcement of many legs of the reinforcement. In the case of Specimen
ratio and exhibiting large deflections. Most of the slabs HSS2 with T-headed shear reinforcement, none of the shear
that failed in punching exhibited a large radius of punching studs were intersected by inclined shear cracks. The existence
cone on the tension face of the slab. of T-headed shear reinforcement forced the shear crack to
develop outside the shear reinforcement zone, resulting
Crack width in increasing the punching perimeter, and this resulted in
The crack width was measured at each load stage. Figure 9 increasing the ultimate load capacity. At the same time, the
shows the crack opening displacement versus the steel strain existence of T-headed shear reinforcement did not result in
for a typical test slab. The crack width increased as the applied increasing the ductility of Specimen HSS2 because the shear
load increased; however, this increase was not very smooth, as studs were not intersected by inclined shear cracks. A possible
concrete is not a homogenous material. It was noticed that the reason for the low ductility of Specimen HSS2 is the wide
crack width versus the steel strain can be represented by one (large) coverage area of T-headed shear stud reinforcement,
straight line up to a value approximately equal to 2000 µε of the coverage zone extended to more than 2d from the column
steel strain (yield strain), except for Specimen HSS3. In most of face, as shown in Fig. 2, and this did not give the chance for
the slabs, the crack width versus the steel strain curve tended to flexure reinforcement to be fully mobilized after the occurrence
behave nonlinearly after the steel strain reached the yield strain. of the shear crack. As a result, the failure mode was less
In Specimen HSS1, the crack width continued to increase after ductile, as indicated by the load-deflection curve (Fig. 4(b)).
the steel strain reached the yield point, which is an expected The structural behavior of Specimen HSS2 could be enhanced
behavior for a slab that failed in flexure. by either increasing the flexure reinforcement development
All measurements reported in Table 4 were taken at a steel length or decreasing the shear reinforcement ratio to force the
stress level of 267 MPa (38.7 ksi), which corresponded to shear crack to develop inside the shear reinforcement zone.
approximately 0.67fy. The results showed that the maximum
crack width can be influenced by as much as 50% when the bar Code requirements for two-way shear
spacing was increased from 217 to 368 mm (8.5 to 14.5 in.). Modern European codes of practice treat punching shear
This means that for the same concrete cover, increasing the in terms of shear stresses calculated at control perimeters
bar spacing by approximately 70% resulted in increasing the located at relatively large distances from the column
crack width by approximately 50%. The test slabs included or loaded area. In the CEB-FIP 1990 model code,20 the
two specimens, Specimens NSS1 and HSS4, reinforced with distance is 2d. In BS 8110-97,21 it is 1.5d, but the perimeter
a high reinforcement ratio of 1.58%. The two specimens had has square corners as compared with CEB-FIP20 rounded
the same bar spacing of 217 mm (8.5 in.). The test results corners. In North American codes such as ACI 318-081 and
revealed that crack control (crack width) can still be achieved CSA A23.3-04,2 the control punching shear perimeter is
by limiting the spacing of the reinforcing steel despite using a only 0.5d away from the loaded area. ACI 318-081 does not

588 ACI Structural Journal/September-October 2011

include the influence of either the flexural reinforcement or The punching resistance according to BS 8110-9721 is
the size effect on the limiting shear stress.
The test results are compared to the predictions of 1/ 3
 f ′ 0.78 
ACI 318-08, 1 CSA A23.3-04,2 Eurocode 2,16 CEB-FIP,20 = 0.79  100ρ c (400 d )1/ 4 ud (SI units) (6a)
VBS 
and BS 8110-97.21 ACI 318-081 requires that the ultimate  25 
shear resistance for slabs Vc is the smallest of
1/ 3
 f ′ 0.78 
 2 VBS = 4.08  100ρ c  (16 d )1/ 4 ud (U.S. units) (6b)
Vc = 0.17  1 +  λ fc ′ bo d (SI units)  25 
 β
(2)
 4
or Vc =  2 +  λ fc ′ bo d (U.S. units) where 4 400 d is a size effect factor; u is the length of a
 β
square perimeter at 1.5d from the loaded area for both
circular and square loaded areas; and r is the flexural
where β is the ratio of long side to short side of the column, reinforcement ratio.
concentrated load, or reaction area;
Size effect
α d  For design engineers, the size effect is a useful concept
Vc = 0.083  s + 2 λ fc ′ bo d (SI units)
 bo  that is based on fracture mechanics. The size of the fracture
(3)
process zone is represented by a material property called
α d 
or Vc =  s + 2 λ fc ′ bo d (U.S. units) the characteristic length lch. It relates the fracture properties
 bo 
of the concrete, such as the modulus of elasticity Ec,
the fracture energy Gf, and the tensile strength ft and is
where αs is 40 for interior columns, 30 for edge columns, 20 for calculated as
corner columns; and
Ec G f
lch = (7)
ft 2
Vc = 0.33λ fc ′ bo d (SI units)
(4)
or Vc = 4λ fc ′ bo d (U.S. units) where Gf is defined as the amount of energy required to
cause one unit area of a crack; hence, it can be obtained
where bo is the perimeter of critical section for shear in slabs as the area under the load-crack width opening curve. A
and footings; d is the slab effective depth; fc′ is the cylinder higher value of lch reflects that the material is less brittle
compressive strength of concrete; and l is the modification and a smaller value means that the material is more brittle.
factor reflecting the reduced mechanical properties of In an earlier investigation by Marzouk and Chen,22 the
lightweight concrete. CSA A23.3-042 accounts for the slab fracture energy Gf was determined experimentally for HSC
effective depth. If the effective depth d used in two-way
to be 160 N/m (11 lb/ft) compared to 110 N/m (7.5 lb/ft)
shear calculations exceeds 300 mm (12 in.), the value of
for NSC. The character­istic length lch was estimated to have
punching shear resistance Vc shall be multiplied by 1300/
an average value of 500 and 250 mm (20 and 10 in.) for
(1000 + d). Fracture mechanics concepts suggest that the size
NSC and HSC, respectively.
effect factor is not related to the member thickness only but
must be related to the concrete strength as well.
Proposed change for size effect
In CEB-FIP,20 the punching shear resistance VCEB is assumed
This section presents a proposal to modify the ACI 318-081
to be proportional to (fck)1/3, where fck is the characteristic
equation by the inclusion of the brittleness ratio h/lch to
compressive strength of concrete. The influences of
account for the size effect factor as proposed earlier in
reinforcement and slab depth are also considered in this design
Eq. (1) by Marzouk et al.5 However, it should be noted
code. The punching resistance according to CEB-FIP20 is
that any size effect factor that is derived using linear elastic
 200  fracture mechanics (LEFM) could be applied. It is also
VCEB = 0.18 3 100ρfck  1 + u1 d (SI units) (5a)
 d  proposed that punching shear strength is proportional to the
reinforcement ratio to the power of 0.33, as given by most
 8 European codes.16,20,21
VCEB = 0.93 3 100ρfck  1 + u1 d (U.S. units) (5b)
 d  Nominal shear strength in the proposed equation for
punching has been evaluated using the control perimeter,
( )
where 1 + 200 d is a size effect factor; u1 is the length concrete strength, reinforcement ratio, and brittleness ratio
of the control perimeter at 2d from the column; and r is the h/lch. In this case, it is proposed that the punching shear
flexural reinforcement ratio. strength is proportional to the ratio of lch/h to the power

ACI Structural Journal/September-October 2011 589

Table 5—Comparison of code predictions with test results
Pcode /Pu
Nominal shear stress
Slab no. Ultimate load Pu, kN v = Pu/bod√fc′ BS 8110-97 CEB-FIP ACI 318-081 Eurocode 216 CSA A23.3-042 Eq. (8) Pu/Pflex
21 20

HSS1 1722 0.28 0.85 0.89 1.19 0.89 1.37 0.90 1.25
HSS2 2172 0.46 0.96 1.08 0.91 1.06 1.15 0.93 0.99
HSS3 2090 0.37 0.88 0.93 0.89 0.93 1.02 0.94 0.64
NSS1 2234 0.40 0.92 0.98 0.83 0.98 0.95 0.92 0.46
HSS4 2513 0.36 0.94 1.00 0.91 1.00 1.03 0.96 0.50
Average 0.91 0.98 0.95 0.97 1.10 0.93
Standard deviation 0.04 0.07 0.14 0.07 0.17 0.02
Coefficient of variation 0.05 0.07 0.15 0.07 0.15 0.02
Notes: 1 MPa = 145 psi; 1 kN = 0.2248 kips.

of 0.33. Based on the previous assumptions, the following ratio and were designed to fail under punching shear. This
equation is recommended confirms that the size effect factor cannot be taken as a
constant number related to the member depth but it must
0.33
l  also be related to the mechanical properties of concrete.
Vc − eq = 0.33 fc ′  ch 
 h
(100ρ)0.33 bo d (SI units) (8a) 3. The enhanced structural behavior of Specimen HSS2
(300 mm [12 in.]) with T-headed studs reflects the benefit
0.33 of using shear reinforcement. The structural behavior of
l 
Vc − eq = 4 fc ′  ch 
 h
(100ρ)0.33 bo d (U.S. units) (8b) Specimen HSS4 (400 mm [16.0 in.]) could be enhanced by
adding shear reinforcement. Adding shear reinforcement
will ensure using the full benefit due to the increasing slab
where bo is the critical shear control perimeter calculated at a effective depth. The test results revealed a reduced ductility
distance equal to 0.5d from the support face. and energy absorption capacity despite increasing the
effective depth. The addition of shear reinforcement with
Test results versus code predictions a reinforcement ratio of approximately 0.68% by volume
The ultimate recorded test loads versus the code predictions changed the punching failure mode to a ductile flexure failure.
are given in Table 5, together with the estimated values by 4. A size effect factor is recommended based on the
the proposed equation. The nominal shear stress presented thickness of the slab and fracture mechanics material
in Table 5 is defined as v = Pu/bod√fc′. The nominal shear property represented by the brittleness factor known as the
stress at failure of Specimen HSS4 is lower than the characteristic length lch.
nominal shear stress at failure of Specimen NSS1. Both
slabs have the same thickness and the same reinforcement
ACKNOWLEDGMENTS
ratio and were designed to fail under punching shear. The authors are grateful to the Natural Sciences and Engineering Research
This confirms that the size effect factor cannot be taken Council of Canada (NSERC) for providing the funds for the project.
as a constant number related to the member depth but it Sincere thanks are due to M. Curtis, S. Organ, D. Pike, and the technical
must be related to the concrete strength as well. It should staff of the Structural Engineering Laboratory of Memorial University of
be noted that the proposed equation does not cover the Newfoundland for their assistance during the preparation of the specimens
effect of shear enhancement for T-headed reinforcement. and during testing. Sincere thanks are extended to Capital Ready Mix Ltd.,
Newfoundland, Canada, for providing the concrete for this project.
It is clear from Table 5 that ACI 318-081 underesti­mated
the punching shear capacity of all specimens with the
exception of Specimen HSS1. For example, in the NOTATION
bo = perimeter of critical section for shear in slabs and footings
case of Specimen NSS1, ACI 318-08 1 underestimated C = side length of square column
the punching capacity by 17%, whereas in the case of d = slab effective depth
Specimen HSS1, which has the minimum amount of fc′ = uniaxial compressive strength of concrete (cylinder strength)
flexural reinforcement ratio, ACI 318-081 overestimated fck = characteristic compressive strength of concrete in MPa
the punching capacity by 19%. ACI 318-081 is applied ft = uniaxial tensile strength of concrete
with the omission of the capacity reduction factor. The fy = yield stress of reinforcement
Gf = fracture energy
most accurate ultimate load predictions for the size effect h = slab thickness
were given by the CEB-FIP20 equation. lch = characteristic length
s = spacing between peripheral lines of shear reinforcement
CONCLUSIONS u = length of control perimeter at distance equal to 1.5d from loaded area
1. North American design codes do not provide guidance u1 = length of control perimeter at distance equal to 2d from loaded area
for thick plates over 250 mm (10 in.) and do not account for Vc = nominal shear strength provided by concrete
the size effect factor. as = constant used to compute Vc in slabs and footings
DU = post-ultimate deflection
2. The nominal shear stress at failure of Specimen HSS4 Dy = deflection at first yield
(400 mm [16.0 in.]) is lower than the nominal shear stress l = modification factor reflecting reduced mechanical properties of
at the failure of Specimen NSS1 (400 mm [16 in.]). Both lightweight concrete
slabs have the same thickness and the same reinforcement r = flexural reinforcement ratio

590 ACI Structural Journal/September-October 2011

 REFERENCES 12. Elgabry, A., and Ghali, A., “Design of Stud-Shear Reinforcement for
1. ACI Committee 318, “Building Code Requirements for Structural Slabs,” ACI Structural Journal, V. 87, No. 3, May-June 1990, pp. 350-361.
Concrete (ACI 318-08) and Commentary,” American Concrete Institute, 13. Dilger, W., and Ghali, A., “Shear Reinforcement for Concrete Slabs,”
Farmington Hills, MI, 2008, 473 pp. Journal of the Structural Division, ASCE, V. 107, No. 12, 1981, pp. 2403-2420.
2. CSA A23.3-04, “Design of Concrete Structures for Buildings,” 14. Marzouk, H., and Jiang, D., “Experimental Investigation on Shear
Canadian Standards Association, Rexdale, ON, Canada, 2004, 258 pp. Enhancement Types for High-Strength Concrete Plates,” ACI Structural
3. Comité Euro-International du Béton-Fédération de la Précontrainte Journal, V. 93, No. 1, Jan.-Feb. 1997, pp. 49-58.
(CEB-FIP), “Punching of Structural Concrete Slabs,” Bulletin No. 12, 15. Berner, D.; Gerwick, B.; and Hoff, G., “T-Headed Stirrup Bars,”
Lausanne, Switzerland, 2001, 307 pp.
Concrete International, V. 13, No. 5, May 1991, pp. 49-53.
4. Richart, F., “Reinforced Concrete Wall and Column Footings,” ACI
16. BS EN 1992-1-2, “Eurocode 2: Design of Concrete Structures—Part
JOURNAL, Proceedings V. 45, No. 2 and 3, 1948, pp. 97-127 and 237-260.
5. Marzouk, H.; Emam, M.; and Hilal, M., “Sensitivity of Shear Strength 1-1: General Rules and Rules for Buildings,” Brussels, Belgium, 2004, 230 pp.
to Fracture Energy of High-Strength Concrete Slabs,” Canadian Journal of 17. Marzouk, H., and Hussein, A., “Experimental Investigation on the
Civil Engineering, V. 25, No. 1, 1998, pp. 40-50. Behavior of High-Strength Concrete Slabs,” ACI Structural Journal, V. 88,
6. Kevin, K., “Influence of Size on Punching Shear Strength of Concrete No. 6, Nov.-Dec. 1991, pp. 701-713.
Slabs,” master’s thesis, McGill University, Montreal, QC, Canada, 2000, 92 pp. 18. Joint ACI-ASCE Committee 421, “Shear Reinforcement for Slabs
7. Regan, P., and Braestrup, M., “Punching Shear in Reinforced Concrete: (ACI 421.1R-08),” American Concrete Institute, Farmington Hills, MI, 2008, 15 pp.
A State of Art Report,” Bulletin d’information No. 168, Comité Euro- 19. Megally, S., and Ghali, A., “Seismic Behavior of Edge Slab-Column
International du Beton, Lausanne, Switzerland, 1985, 232 pp. Connections with Stud Shear Reinforcement,” ACI Structural Journal,
8. Hallgren, M., “Punching Shear Capacity of Reinforced High Strength V. 97, No. 1, Jan.-Feb. 2000, pp. 53-60.
Concrete Slabs,” doctoral thesis, Department of Structural Engineering, 20. Comité Euro-International du Béton-Fédération de la Précontrainte
Royal Institute of Technology, Stockholm, Sweden, 1996, 206 pp. (CEB-FIP), Model Code 1990, Bulletin d’information No. 203-305,
9. Hegger, J.; Sherif, A.; and Ricker, M., “Experimental Investigations
Lausanne, Switzerland, 1990, 462 pp.
on Punching Behavior of Reinforced Concrete Footings,” ACI Structural
Journal, V. 104, No. 3, May-June 2006, pp. 604-613. 21. BS 8110, “Structural Use of Concrete, Part 1: Code of Practice for
10. Hegger, J.; Ricker, M.; and Sherif, A., “Punching Strength of Design and Construction,” British Standards Institution, London, UK,
Reinforced Concrete Footings,” ACI Structural Journal, V. 106, No. 5, 1997, 117 pp.
Sept.-Oct. 2009, pp. 706-716. 22. Marzouk, H., and Chen, Z., “Fracture Energy and Tension Properties
11. Birkle, G., and Dilger, W., “Influence of Slab Thickness on Punching Shear of High-Strength Concrete,” Journal of Materials in Civil Engineering,
Strength,” ACI Structural Journal, V. 105, No. 2, Mar.-Apr. 2008, pp. 180-188. ASCE, V. 7, No. 2, 1995, pp. 108-116.

ACI Structural Journal/September-October 2011 591

View publication stats