ACI STRUCTURAL JOURNAL TECHNICAL PAPER

Title no. 105-S63

Strength Evaluation of Interior Slab-Column Connections

by Ying Tian, James O. Jirsa, and Oguzhan Bayrak

Test data for interior flat-plate slab-column connections subjected

to concentric gravity load and combined gravity and lateral loads
were collected. The connection strength was evaluated using ACI
Code design equations. From test data, equations for connection
gravity load capacity as a function of concrete strength, slab
reinforcement ratio and yield strength, and the ratio of column size
to slab effective depth were developed. A model for strength of
connections subjected to lateral load using a beam analogy
concept that includes resistance from flexure, shear, and torsion at
the critical section is proposed.

Keywords: flat plate; lateral strength; slab-column connection; two-way

shear strength.

INTRODUCTION
Considerable experimental research has been carried out Fig. 1—Typical behavior of slab-column connections
on slab-column connections of reinforced concrete flat-plate subjected to concentric gravity load.1
structures subjected to concentric gravity load. Characteristic
behavior of slab-column connections under gravity loads can
be illustrated by Fig. 1, taken from a study by Criswell.1 where b is the circumference of the column, d is the average
Generally, as the tensile reinforcement ratio ρ decreases, the effective depth of the slab, c is the column size, and fc′ is the
behavior becomes more ductile, however, with reduced concrete compressive strength.
connection strength. It is noteworthy that, in a typical two-way slab system, top
Punching failure that results in an instant loss of gravity reinforcing bars are curtailed outside the negative bending
load capacity was observed in most reported isolated moment region and may have a nonuniform spacing that
connection specimens. However, punching failure may not results in varied flexural capacity along a yield line. Thus, Vflex
take place prior to general flexural yielding. Moe2 defined may be difficult to determine using the yield-line theory. In
shear failure as one that is mainly caused by transverse addition, for connections controlled by shear, the study by
forces and takes place before the flexural capacity Vflex is Marzouk and Hussein5 indicated that the influence of concrete
reached. Vflex was defined as the gravity shear at the formation strength is overestimated when shear strength is correlated
of a yield-line mechanism. If the ratio of ultimate load Vult to with the square root of concrete compressive strength.
Vflex, φ, was less than unity, the failure was classified as a Only a limited number of tests have been carried out on
shear failure; otherwise, it was classified as a flexural failure slab-column connections transferring both gravity and
with punching a secondary phenomenon. seismic lateral loads. Equivalent beam width method and
Behavior of slab-column connections cannot be compared beam analogy are two approaches for evaluating connection
with that of beams without transverse reinforcement. For lateral strength. The beam analogy concept is more attractive
connections controlled by shear, the diagonal tension cracks because it describes all actions including shear, flexure, and
were usually generated in the early loading stage but torsion acting on the critical section of a connection. It is
remained stable. For connections controlled by flexure, difficult, however, to define the lateral resistance contributed
inclined cracking resulted in significant connection deflection by shear and torsion when the ultimate unbalanced moment
but concrete crushing was not observed in the compression is reached.
zone in a typical test. The research reported herein focuses on strength evaluation
In addition to slab concrete strength and flexural reinforcement of interior connections with square columns. Relevant
ratio ρ and yield strength fy, the ratio of column size to slab tests2,3,5-38 satisfying the following conditions were
effective depth (c/d) and boundary conditions also influence collected to form the basis of the study: 1) connections were
the connection gravity load-carrying capacity.3,4 Using test constructed with normalweight concrete; 2) slab flexural
data for footings and slabs that failed in shear (Vult < Vflex), reinforcement strength as measured by ρfy was less than 8 MPa
Moe2 defined the connection shear strength as (1200 psi); 3) no shear reinforcement was used; and 4) slab
thickness was at least 75 mm (3 in.).
15 ⎛ 1 – 0.075 ---⎞
c
⎝ d⎠ ACI Structural Journal, V. 105, No. 6, November-December 2008.
V n = ( bd f c′ ) ------------------------------------- (in customary units) (1) MS No. S-2007-042 received January 24, 2007, and reviewed under Institute
bd f c′ publication policies. Copyright © 2008, American Concrete Institute. All rights reserved,
1 + 5.25 --------------- - including the making of copies unless permission is obtained from the copyright proprietors.
V flex Pertinent discussion including author’s closure, if any, will be published in the September-
October 2009 ACI Structural Journal if the discussion is received by May 1, 2009.

692 ACI Structural Journal/November-December 2008

ACI member Ying Tian is an Assistant Professor of civil and environmental engineering
at the University of Nevada, Las Vegas, NV. He received his PhD from the University
of Texas at Austin, Austin, TX. He is a member of ACI Committee 369, Seismic Repair
and Rehabilitation, and Joint ACI-ASCE Committees 421, Design of Reinforced Concrete
Slabs, and 445, Shear and Torsion. His research interests include experimental
investigation and multi-scale modeling of reinforced concrete members and systems.

ACI Honorary Member James O. Jirsa holds the Janet S. Cockrell Centennial Chair
in Engineering at the University of Texas at Austin. He is a Past President of ACI and
a Past Chair of the Technical Activities Committee. He is a member of ACI Committee
318, Structural Concrete Building Code, and the Concrete Research Council.

ACI member Oguzhan Bayrak is an Associate Professor of civil, architectual, and

environmental engineering and a Fellow of the Clyde E. Lee Professorship at the University
of Texas at Austin. He is a member of ACI Committees 341, Earthquake-Resistant
Concrete Bridges, E803, Faculty Network Coordinating Committee, and Joint ACI-
ASCE Committees 441, Design of Reinforced Concrete Slabs, and 445, Shear and Torsion.

RESEARCH SIGNIFICANCE
The effectiveness of using code design equations to evaluate
gravity load capacity and lateral resistance of slab-column
connections was examined. According to analyses of available
test data, the effects of several variables on connection behavior
were individually studied and an equation was formulated
for calculating connection shear strength. An algorithm using
beam analogy concepts was developed for evaluating lateral
resistance of lightly-reinforced slab-column connections.

CONNECTION SHEAR STRENGTH Fig. 2—Range of parameters in database (84 tests).

Two-way shear design of ACI Code
ACI 318-0539 defines the two-way shear strength Vc of a
typical interior slab-column connection with a square column as

Vc = vc Ac = 0.33 f c′ Ac (in SI units) (2)

Vc = vc Ac = 4 f c′ Ac (in customary units) (3)

where vc = 0.33 f c′ (MPa) (vc = 4 f c′ [psi]) is the nominal

shear strength and Ac = 4d(c + d) is the critical section area.
Joint ACI-ASCE Committee 32640 simplified Eq. (1) and
derived Eq. (3) to represent a conservative estimate of tests
that failed in shear (Vult < Vflex). Thus, it is evident that
Moe’s2 definition of connection failure modes was inherent
in ACI Code provisions for two-way shear design of a slab-
column connection. Because Eq. (2) and (3) were not
intended for preventing connection failure dominated by
flexure, these equations should not be used to estimate
gravity load-carrying capacity of connections with relatively
low slab tensile reinforcement ratio ρ and yield strength fy.
As an illustration, Eq. (2) was applied to 84 specimens with
uniformly distributed slab flexural reinforcement. The range Fig. 3—Comparison of measured and calculated shear
of major parameters for these specimens including concrete strength based on Eq. (2).
strength fc′ , tensile reinforcement index ρfy, slab thickness h,
and c/d is shown in Fig. 2. The measured strength is plotted
Zsutty41,42 for predicting beam shear strength. The effects of
in Fig. 3 against the strength estimated based on Eq. (2). All
slab size and boundary condition on shear strength were not
specimens, for which Eq. (2) overestimates the shear
considered due to lack of test data for connections supported
capacity, have a low value of tensile reinforcement index ρfy
on square columns. The connection shear strength was
equal to or less than 4.1 MPa (595 psi) (for Grade 60 steel, ρ
assumed to take the following general form
≤ 1%).

Evaluation of shear strength based on test data Vn = q · ( fct)α · (ρfy)β · Ac · y (c/d) (4)
The manner in which slab-column connections fail in
punching under concentric gravity load is still not under- where fct is the concrete tensile strength; Ac = 4d(c + 2s) is
stood well. An empirical solution was developed in this the area of a critical section located at a distance s from the
study using an approach similar to that adopted by column faces; and y is a dimensionless function of c/d.

ACI Structural Journal/November-December 2008 693

Table 1—Determination of α and β 4. The average deviation for all tests used in the analysis
No. of was calculated as follows
No. of groups
Variable tests (m) Data range e ni
m

∑ ∑ eij
fc ′ 13.7 to 70.0 MPa 0.093 0.066 0.048
24 9
(1990 to 10,200 psi) (α =1) (α = 2/3) (α = 1/2)
1.58 to 7.93 MPa 0.16 0.081 0.065 e = = 1j = 1 -
i---------------------- (9)
ρfy 39 15 m
(230 to 1150 psi) (β = 0) (β = 1/3) (β = 1/2)
∑ ni
i=1

Assuming fct is related to f c′ , Eq. (4) is rewritten as 5. The average deviation e was minimized by adjusting the
value of α or β and repeating Steps 2 to 4.
Vn = k · ( fc′ )α/2 · (ρfy)β · Ac · y (c/d) (5) The number of tests, groups, data range of variables, and
results of the evaluation are given in Table 1. It was found
that the contribution of concrete strength and slab flexural
The parameters k, α, β, s, and the expression of y were reinforcement is best described by assuming the punching
determined from analyses of the test results. Test data of strength to be proportional to (fc′ )0.25 and (ρfy)0.5 (α = β = 1/2).
specimens with top bars evenly distributed in two orthogonal The significant contribution of flexural reinforcement is not
directions of the slab were used to generate Eq. (5). If the surprising because tests indicated that:
concrete strength was reported as cube compressive strength • Response of lightly-reinforced slab-column connections
fcu, value was transformed to cylinder compressive strength is dominated by flexure;
according to • Flexural reinforcement restrains the propagation of
inclined cracking that takes place when loads as low as
fc′ = 0.80fcu (6) 50% of the ultimate load are reached2,26;
• As flexural strength increases, concrete strain tangential to
Effect of concrete and flexural reinforcement strength— the column face at the compressive surface of slab also
To examine the effect of a single variable on connection increases.5 The resulting restraint from the surrounding
shear capacity while eliminating the influence of others, the slab enhances the effect of a triaxial state of stress for
following procedures were adopted: concrete in the connection region; and
1. An intermediate term Y was defined as Y = ( fc′ )0.5α and • Once an inclined crack develops, tensile longitudinal
Y = (ρfy)β when the effect of concrete strength and slab flexural reinforcement provides a doweling force at the crack
reinforcement was evaluated, respectively. The specimens and may provide a significant fraction of the shear
with nearly identical properties (difference limited to 5%), strength.26,43
except for the variable contained in the term Y to be examined, Critical section location—The procedures described
were placed into the same group. For example, when previously were similarly applied to 10 groups of test data
investigating the influence of concrete strength, the specimens (49 specimens) to determine an appropriate critical section
with similar flexural reinforcement ratio and strength, location. In each group, the specimens had nearly identical
column size, and slab effective depth but with different values of effective depth d and c/d. The following ratio was
concrete strength formed a data group. Each group consisted used in the investigation
of at least two specimens. Using this approach, m groups
were generated, as indicated in Table 1. V ult V ult
2. An initial value was chosen for the parameter (α or β) Y = ---------------------------------
1 1
- = -------------------------------------------------------
1 1
- (10)
--- --- --- ---
and the ratio of measured strength Vij to Yij was calculated 4 2 4 2
for each specimen. The subscript i denotes the group (i = 1, ( f c′ ) ( ρf y ) A c ( f c′ ) ( ρf y ) [ 4d ( c + s ) ]
2, …, m) and j denotes the specimen in each group ( j = 1,
2, …, ni). Clearly, this ratio should be similar for tests in the It was found that, if s ≤ 2d, the deviation e was insensitive
same group if the initial value of α or β closely reflects the to s. Therefore, s = 0.5d, the same value as used in ACI 318-05,
contribution of the investigated variable (concrete strength was adopted.
or flexural reinforcement). The average ratio for each group, Effect of c/d—A review of the data indicted a pronounced
as defined by the following equation, was then calculated. tendency toward decreased connection strength with increased
c/d. To illustrate this situation, the following ratio as a function
ni of c/d was evaluated for specimens with a slab thickness of
V
∑ -----
Y ij
ij
- approximately 150 mm (6 in.). The results are plotted in Fig. 4.

X i = j---------------
=1 (7) V ult
ni Z = -------------------------------------------------------- (11)
1 1
--- ---
4 2
3. For each specimen, the deviation of the prediction from ( f c′ ) ( ρf y ) [ 4d ( c + d ) ]
the average of the corresponding group was defined as
The c/d is likely related to the lateral restraint imposed on
V ij concrete under a triaxial state of stress at the base of the
------ – X i inclined crack. Based on a regression analysis, as shown in
Y ij
e ij = -------------------- (8) Fig. 4, the effect of c/d on the connection strength can be
Xi expressed by taking function y in Eq. (5) as

694 ACI Structural Journal/November-December 2008

 y ⎛ ---⎞ =
c d
--- (12)
⎝ d⎠ c

Formulation of connection shear strength—Based on data

analyses described previously, the shear capacity of a slab-
column connection was derived as

1
---
2
V n = 0.65ξA c ( ρf y f c′ ) (in SI units) (13)

1
---
2
V n = 2.3ξA c ( ρf y f c′ ) (in customary units) (14)

where

Ac = 4(c + d)d (15)

d
ξ = --- (16)
c
Fig. 4—Effect of c/d on connection shear strength.
If 95% of the test results are expected to exceed the
strength defined by Eq. (13) or (14), a reduction factor of 0.83
should be applied to the right-hand side of these equations to
derive a characteristic strength.
Comparison of calculated and measured strength—The
comparison of shear capacity estimated using Eq. (13) with
the measured strength (for the same database used in Fig. 3)
is shown in Fig. 5. Computed values are now closer to
measured values as the scatter is reduced.
Reinforcement concentration—Equations (13) and (14) were
established from tests with uniformly spaced reinforcement.
Banded flexural reinforcement at the column is used for two
reasons: 1) to increase slab stiffness; and 2) to allow the
connection to carry unbalanced moments imposed by lateral
deformation as well as by gravity loads. Therefore, the area
where banded reinforcement is effective for shear strength,
that is, the area where the reinforcement ratio ρ is evaluated,
must be defined. For this purpose, the strength of 11 specimens
with concentrated flexural reinforcement was evaluated
according to Eq. (13) or (14) using reinforcement ratios
defined over various slab widths centered on the column. It
was found that the reinforcement ratio defined in a width
c + 12d, nearly the entire region of column strip, provided
the best estimate of the test results.
Fig. 5—Comparison of measured and calculated shear
CONNECTION LATERAL STRENGTH strength based on proposed formulation (Eq. (13)).
ACI 318-05 provisions
For slab-column connections carrying both gravity shear
and unbalanced moment, the design approach recommended 2 ( V c – V g )J c 5 ( Vc – Vg )
M n, v = ------------------------------
- = ------------------------- (17)
in ACI 318-0539 is shown in Fig. 6. It is assumed that a γ v ( c + d )A c ( c + d )A c
portion (γv = 40%) of the unbalanced moment M is trans-
ferred by eccentric shear stress, whereas the remaining
where
portion (γf = 60%) of M is resisted by flexure provided by the
reinforcement situated within a width c + 3h at the column.
Following the aforementioned assumptions, the unbalanced 2 3 1 3
J c = --- d ( c + d ) + --- ( c + d )d (18)
moment at shear failure Mn,v and that at flexural failure Mn,f 3 6
can be determined from Eq. (17) and (19), respectively. The
minimum of Mn,v and Mn,f provides an estimation of the M fn, 1 + M fn, 2 M fn, 1 + M fn, 2
connection lateral resistance, Mn,ACI, as well as the corre- M n, f = -------------------------------
- = -------------------------------
- (19)
sponding failure mode. γf 0.6

ACI Structural Journal/November-December 2008 695

 must be made for modeling the connection lateral strength
Mn. The formulations for various actions and the associated
assumptions are presented as follows.
Unbalanced moment resisted by flexure at bending faces—
The flexural strength at the back face of critical section Mn,1 and
that at the front face Mn,2 can be determined as44

(c + d) A s,1 f y, 1⎞
M n, 1 = A s,1 f y, 1 d 1 ---------------- ⎛ 1 – 0.59 -----------------
- (22)
s1 ⎝ s 1 d1 f c′ ⎠

Fig. 6—ACI 318-05 design approach for connections

(c + d) A s, 2 f y, 2 ⎞
transferring unbalanced moment and gravity shear. M n, 2 = A s, 2 f y, 2 d 2 ---------------- ⎛ 1 – 0.59 -----------------
- (23)
s2 ⎝ s 2 d 2 fc′ ⎠

where As,1, fy,1, d1, and s1 denote the bar area, yield strength,
effective depth, and average spacing of top reinforcement
placed in the lateral loading direction within a width c + d
centered on the column, respectively; As,2, fy,2, d2, and s2 are
defined similarly but for the bottom reinforcement.
For connections with bottom reinforcement terminated at
the column, fy,2 may not be developed. In this case, Mn,2 is
limited to the flexural cracking moment evaluated in c + d.
In addition, neither steel strain hardening nor shear-flexure
interaction was considered.
Unbalanced moment resisted by shear at bending faces—
Fig. 7—Internal forces acting on critical section. It was assumed that Vn,1 and Vn,2, the maximum shear values
that can be developed at the front and back faces, are limited to
where Mfn,1 and Mfn,2 are the maximum values that can be
Vn,1 = 0.25Vn (24)
developed for Mf,1 and Mf,2.

Model based on beam analogy Vn,2 = 0.25Vn′ (25)

Tests26 indicated that the slab inelastic flexural deformation
was not concentrated at the slab-column interface but at the where Vn is the shear capacity of a connection subjected only to
location of the inclined cracks where the punching failure gravity load; Vn′ , however, is the shear capacity if an upward
under lateral deformation was triggered. In addition, a critical vertical load acts on the slab and causes tension in the slab
section located 0.5d away from the column faces has been bottom surface at the column. Vn is determined from Eq. (13).
suggested for Eq. (13). Thus, for convenience, the beam The variables in Eq. (13), however, should be associated with
analogy procedures described in this study assumed a critical the bottom reinforcement properties to calculate Vn′ .
section location for both flexure and shear as 0.5d from the The connection rotational resistance from shear Vn,1 and
column faces. Vn,2 is calculated as a moment about the critical section
Figure 7 shows the internal forces acting on the critical centroid as
section that must balance the external moment M and the
gravity shear transferred from slab to column Vg. Among the
M vn, 1 = V n, 1 ⎛ ------------⎞
c+d
four critical section faces, two are referred to as bending faces, (26)
⎝ 2 ⎠
including the front and back faces with respect to the lateral
loading direction. The other two faces are termed as side faces.
The vector notation for the internal forces is given as
M vn, 2 = V n, 2 ⎛ ------------⎞
c+d
follows: M1 and V1 are flexural moment and shear at the (27)
⎝ 2 ⎠
back face subjected to negative bending (top steel in
tension); M2 and V2 are flexural moment and shear at the
front face; T and VT are torsion and shear at the side faces. Unbalanced moment resisted by torsion at side faces—It
The aforementioned actions must satisfy the following was assumed that, at the side faces, the existence of shear
equilibrium equations always reduces the torsional capacity. The interaction
between shear and torsion was assumed to be a function of
c+d VT /Vn,1 and Tn/T0, where Tn and T0 are the torsional strength
M = M 1 + M 2 + 2T + ( V 1 + V 2 ) ------------ (20) with and without the presence of VT. T0 was investigated first
2 according to available test data, and then an interaction
relationship between shear and torsion was developed.
Vg = V1 – V2 + 2VT (21) Using the test setup shown in Fig. 8, Kanoh and
Yoshizaki27 conducted a series of tests to determine T0 for a
Because slab-column connections are statically indeterminate slab-column connection without transferring gravity shear.
and internal forces at the critical section, especially shear and In the tests, the slab was connected to one face of the column
torsion, cannot be measured in a test, a series of assumptions and pin-supported at two slab edges. Torsion was induced by

696 ACI Structural Journal/November-December 2008

 Table 2—Torsional strength T0
vT
fc′ , fy , T0 , kN-m --------
-
Reference Label MPa (psi) ρ, % MPa (ksi) (k-in.) f c′
(1) (2) (3) (4) (5) (6) (7)
* 25.7 (3730) 0.97 377 (54.7) 7.62 (67.4) 1.44
T2
27 T3* 25.7 (3730) 1.93 377 (54.7) 10.1 (89.6) 1.92
T4* 25.7 (3730) 0.55 377 (54.7) 7.66 (67.8) 1.45
H5 22.8 (3310) 0.97 361 (52.3) 9.43 (83.5) 2.14
28
H6 22.8 (3310) 1.93 361 (52.3) 9.77 (86.5) 1.98
S1 45.8 (6640) 0.65 323 (46.8) 7.54 (66.7) 2.03
29 S2 35.1 (5090) 0.98 330 (47.9) 7.30 (64.6) 2.26
S3 33.9 (4920) 1.31 335 (48.6) 7.21 (63.8) 2.12
*
Using test scheme shown in Fig. 8.

Fig. 8—Test scheme for determining torsion capacity T0.27 and for the sake of simplicity, T0 was defined as a function
of critical section geometry and concrete strength. It was
assumed that at T0 the concrete within a width c + d and
between the centroids of top and bottom reinforcement
layers at the side faces of the critical section can develop full
plastic torsional capacity defined as

T0 = vT (c + d)d02 (29)

where vT is the nominal torsional shear stress that was assumed

to be proportional to f c′ and d0 is the distance between the
Fig. 9—Test setups for lateral loading without applying center of top and bottom reinforcement layers (Fig. 10).
gravity load.28,29 Column (7) in Table 2 provides the ratio of vT to f c′ for
each test. Based on the results for Specimens H5, H6, S1, S2,
and S3 with continuous slab at the column, vT was defined as
applying a monotonic lateral load on the column. Although vT = 2.1 f c′ (MPa) (vT = 25 f c′ [psi]). It is noted that this
the physical conditions for torsion of a typical slab-column strength value is approximately six times larger than the
connection could not be fully duplicated, such a test scheme shear stress at cracking for a beam (with a section of c + d by
provided an approach for investigating pure torsional d) subjected to pure torsion. The significantly enhanced
resistance. It was observed that T0 and the column rotation strength can be attributed partly to the presence of slab flexural
at which general yielding occurred were largely unaffected reinforcement and partly to the restraint of the concrete due
by the reinforcement ratio and slab width. In addition, the to slab continuity.
column ultimate rotation was as high as 0.02 to 0.04 radians, A review of previous tests indicated that 90% of the
indicating a relatively ductile failure mode. specimens that satisfied the following conditions achieved
T0 can also be implicitly estimated using tests of slab- general yielding prior to the connection failure: 1) top
column connections transferring only unbalanced moment. reinforcement ratio in c + 12d region was less than 1.5%; and
Such tests have been carried out by Kanoh and Yoshizaki28 2) gravity shear ratio Vg /Vc was no larger than 0.40, where
and Morrison et al.29 using a test setup shown in Fig. 9(a) Vc is the code-defined two-way shear strength of a connection.
and (b), respectively. The tests, different from those illustrated It was therefore assumed that the full capacities of M1, M2,
in Fig. 8, were conducted on conventional isolated connections V1, and V2 were developed in these specimens at the maximum
with continuous slab surrounding the column. In addition, a unbalanced moment. Such specimens are listed in Table 3
cyclic lateral load was applied to simulate the seismic effects. and were used to formulate an appropriate shear-torsion
Because no gravity load was applied and general yielding interaction equation because shear VT and torsion Tn can be
was observed in the tests, it can be assumed that all internal determined from equilibrium according to Eq. (20) and (21).
actions at the critical section developed and maintained their Various shear-torsion interaction relationships were
full capacity. Thus, T0 can be determined from equilibrium as examined. It was found that the linear function as expressed
in Eq. (30) resulted in reasonable agreement between the
c+d calculated connection lateral resistance Mn and the measured
T 0 = 0.5 M ult – M n, 1 – M n, 2 – ( V n, 1 + V n, 2 ) ------------ (28)
2 Mult. The ratios of Mn/Mult are given in Column (8) of Table 3.

where Mult is the measured maximum unbalanced moment. Tn VT

Properties of the aforementioned tests and the calculated ----- + ---------- = 1 (30)
T 0 V n, 1
values of T0 based on Eq. (28) are given in Table 2.
Comparing the values in Columns (4) and (6) for each series
of tests does not reveal a clear trend regarding the effect of Based on Eq. (21) and (30), the torsional strength Tn in the
reinforcement on torsional resistance. For lack of sufficient presence of gravity shear Vg for a slab-column connection
test data to quantify the influence of slab flexural reinforcement was determined as

ACI Structural Journal/November-December 2008 697

Table 3—Calculated versus measured connection lateral strength
Suggested model ACI 318-05 approach

Mn,1 + Mn,2, Mvn,1 + Mvn,2, 2Tn , Mn, Mult, M Mu,f , Mu,v, Mn,ACI, M n, ACI
--------n- ---------------
Reference Label kN-m (k-in.) kN-m (k-in.) kN-m (k-in.) kN-m (k-in.) kN-m (k-in.) M ult kN-m (k-in.) kN-m (k-in.) kN-m (k-in.) M ult
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
S1 44.4 (393) 39.9 (353) 43.0 (381) 127 (1127) 145 (1280) 0.88 175 (1545) 88.0 (779) 88.0 (779) 0.61
31
S4 44.2 (391) 39.2 (347) 34.2 (303) 118 (1041) 125 (1110) 0.94 173 (1532) 75.4 (667) 75.4 (667) 0.60
H5 12.1 (107) 9.1 (81) 19.9 (177) 41.2 (364) 42.5 (376) 0.97 39.7 (351) 33.4 (299) 33.4 (299) 0.80
H6 21.9 (194) 11.5 (102) 19.9 (177) 53.4 (473) 53.1 (470) 1.01 65.4 (579) 33.4 (299) 33.4 (299) 0.63
28
H9 11.5 (102) 8.8 (7.8) 6.0 (53) 26.3 (232) 32.4 (287) 0.81 37.7 (334) 23.1 (204) 23.1 (204) 0.71
H10 16.1 (143) 9.9 (88) 10.7 (94) 36.7 (325) 35.4 (313) 1.04 50.9 (450) 22.5 (199) 22.5 (199) 0.64
1 8.8 (78) 9.0 (79) 11.3 (100) 29.1 (258) 30.5 (270) 0.95 24.5 (217) 28.1 (249) 24.5 (217) 0.80
31 2 9.4 (83) 9.5 (84) 14.1 (124) 33.0 (292) 37.7 (334) 0.87 25.9 (229) 32.5 (288) 25.9 (229) 0.69
3C 8.0 (70) 8.6 (76) 11.6 (103) 28.2 (249) 35.8 (317) 0.79 22.0 (195) 29.8 (264) 22.0 (195) 0.61
SM0.5 21.1 (187) 30.3 (269) 54.7 (484) 106 (939) 100 (888) 1.06 62.0 (549) 115 (1017) 62.0 (549) 0.62
32
SM1.0* 40.5 (359) 41.9 (371) 73.4 (649) 156 (1379) 127 (1128) 1.22 119 (1055) 107 (949) 107 (949) 0.84
S1 6.0 (53) 9.2 (81) 20.4 (180) 35.6 (315) 35.0 (310) 1.02 14.7 (130) 60.7 (537) 14.7 (130) 0.42
S2 9.0 (80) 10.6 (94) 17.8 (158) 37.5 (332) 39.0 (345) 0.96 21.9 (194) 53.1 (470) 21.9 (194) 0.56
29 S3 11.9 (105) 12.3 (109) 17.5 (155) 41.8 (370) 42.0 (372) 0.99 29.0 (257) 52.2 (462) 29.0 (257) 0.69
S4 8.7 (77) 10.5 (93) 13.7 (122) 32.9 (291) 36.0 (319) 0.91 21.1 (187) 48.9 (433) 21.1 (187) 0.59
S5 9.2 (81) 10.8 (96) 9.4 (83) 29.4 (260) 38.0 (336) 0.77 22.4 (198) 44.6 (395) 22.4 (198) 0.59
1 22.3 (197) 19.9 (176) 24.0 (212) 66.1 (585) 64.0 (566) 1.03 57.4 (508) 63.2 (559) 57.4 (508) 0.90
33
3 22.2 (197) 19.6 (173) 46.1 (408) 87.9 (778) 97.9 (866) 0.90 57.2 (506) 76.5 (677) 57.2 (506) 0.58
A 23.9 (211) 15.7 (139) 23.3 (206) 62.9 (556) 66.2 (586) 0.95 53.3 (472) 57.6 (510) 53.3 (472) 0.81
34
B* 24.8 (219) 15.8 (140) 6.7 (59) 47.3 (418) 41.4 (366) 1.14 55.5 (491) 43.1 (381) 43.1 (381) 1.04
35 SC0 23.4 (207) 17.1 (151) 18.3 (162) 58.8 (520) 61.7 (546) 0.95 67.8 (600) 58.0 (513) 58.0 (513) 0.94
DNY_1 13.3 (118) 11.2 (99) 24.8 (219) 49.3 (436) 47.2 (418) 1.04 35.1 (310) 53.8 (476) 35.1 (310) 0.74
DNY_2 12.7 (112) 10.4 (92) 12.1 (107) 35.1 (311) 33.4 (296) 1.05 33.5 (296) 38.1 (337) 33.5 (296) 1.00
36
DNY_3 12.6 (112) 11.4 (101) 22.8 (202) 46.8 (414) 48.4 (428) 0.97 33.2 (294) 46.9 (415) 33.2 (294) 0.69
DNY_4 12.1 (107) 9.6 (85) 20.5 (181) 42.2 (373) 44.1 (390) 0.96 31.9 (282) 38.9 (344) 31.9 (282) 0.72
37 IC 17.5 (155) 14.7 (130) 24.8 (219) 56.9 (504) 51.2 (453) 1.11 54.4 (481) 61.5 (544) 54.4 (453) 1.06
38 C02 19.0 (168) 18.0 (160) 7.3 (65) 44.3 (392) 44.4 (393) 1.00 77.0 (681) 44.0 (389) 44.0 (389) 0.99
26 L0.5 30.9 (273) 39.6 (351) 62.0 (548) 133 (1173) 128 (1135) 1.03 78.6 (696) 158 (1400) 78.6 (696) 0.61
Average 0.98 Average 0.73
Standard deviation 0.10 Standard deviation 0.16
*
No general yielding existed.

The procedures recommended in ACI 318-05 were

applied to the test specimens listed in Table 3. The ratios of
Mn,ACI /Mult given in Column (12) indicate that using the
Code procedures may significantly underestimate the lateral
strength of a slab-column connection.

CONCLUSIONS
Analyses of available test data of interior slab-column
connections subjected to concentric gravity loading indicated
that: 1) the capacity of slab-column connections estimated
using ACI 318-05 two-way shear design equations may vary
considerably from the test results; 2) the contribution of
Fig. 10—Effective area of calculating torsion capacity T0. concrete strength is best described by assuming connection
punching capacity to be proportional to (fc′ )0.25 instead of
V g – V n, 1 + V n, 2⎞ f c′ ; and 3) strength of slab tensile reinforcement and c/d
T n = ⎛ 1 – -------------------------------------
- (31) significantly affect the connection strength. Based on regression
⎝ 2V n, 1 ⎠
analyses, an equation was developed to estimate gravity
load-carrying capacity of interior slab-column connections.
Columns (3), (4), and (5) of Table 3 provide connection The ACI 318-05 design equations underestimate the
lateral resistance from flexure, shear, and torsion, respectively. unbalanced moment-carrying capacity of most connections
According to these data, the fraction of the unbalanced tested. The equations are conservative for design but may be
moment resisted by flexure ranges from 17 to 52%, by shear too conservative for use in evaluation of existing flat-plate
from 22 to 41%, and by torsion from 14 to 57%. structures, thereby making structural retrofit more costly and

698 ACI Structural Journal/November-December 2008

less feasible. To use the beam analogy approach to predict 5. Marzouk, H., and Hussein, A., “Experimental Investigation on the
lateral resistance of interior slab-column connections: 1) the Behavior of High-Strength Concrete Slabs,” ACI Structural Journal, V. 88,
No. 6, Nov.-Dec. 1991, pp. 701-713.
unbalanced moment resisted by shear at the critical sections 6. Corley, W. G., and Hawkins, N. M., “Shearhead Reinforcement for
was related to the shear capacity that considered the effects Slabs,” ACI JOURNAL, Proceedings V. 65, No. 10, Oct. 1968, pp. 811-824.
of slab flexural reinforcement and c/d; and 2) lateral resistance 7. Swamy, R. N., and Ali, S. A. R., “Punching Shear Behavior of
from torsion with and without the presence of gravity load Reinforced Slab-Column Connections Made with Steel Fiber Concrete,”
ACI JOURNAL, Proceedings V. 79, No. 5, Sept.-Oct. 1982, pp. 392-406.
was defined based on the test data. The proposed equations offer 8. Mokhtar, A.; Ghali, A.; and Dilger, W., “Stud Shear Reinforcement
an alternate approach to evaluate the lateral strength of lightly- for Flat Concrete Plates,” ACI JOURNAL, Proceedings V. 82, No. 5, Sept.-
reinforced interior slab-column connections. Oct. 1985, pp. 676-683.
9. Alexander, S. D. B., and Simmonds, S. H., “Tests of Column-Flat
Plate Connections,” ACI Structural Journal, V. 89, No. 5, Sept.-Oct. 1992,
ACKNOWLEDGMENTS pp. 495-502.
The authors gratefully acknowledge the funding for this research
10. Theodorakopoulos, D. D., and Swamy, N., “Contribution of Steel
provided by the National Science Foundation under Grant No. 0301632 and
Fibers to the Strength Characteristics of Lightweight Concrete Slab-
0301395. Any opinions, findings, and conclusions or recommendations
Column Connections Failing in Punching Shear,” ACI Structural Journal,
expressed in this paper are those of the authors and do not necessarily reflect
V. 90, No. 4, July-Aug. 1993, pp. 342-355.
the views of the sponsor.
11. Gardner, N. J., and Shao, X., “Punching Shear of Continuous Flat
Reinforced Concrete Slabs,” ACI Structural Journal, V. 93, No. 2, Mar.-
NOTATION Apr. 1996, pp. 218-228.
Ac = critical section area 12. Marzouk, H.; Emam, M.; and Hilal, M. S., “Effect of High-Strength
As,1 = area of top reinforcement Concrete Columns on the Behavior of Slab-Column Connections,” ACI
As,2 = area of bottom reinforcement Structural Journal, V. 93, No. 5, Sept.-Oct. 1996, pp. 545-554.
c = size of square column 13. Marzouk, H., and Jiang, D., “Experimental Investigation on Shear
d = slab average effective depth Enhancement Types for High-Strength Concrete Plates,” ACI Structural
d0 = distance between center of top and bottom reinforcement layers Journal, V. 94, No. 1, Jan.-Feb. 1997, pp. 49-58.
d1 = slab effective depth (top reinforcement in tension) 14. Marzouk, H.; Emam, M.; and Hilal, M. S., “Effect of High-Strength
d2 = slab effective depth (bottom reinforcement in tension) Concrete Slab on the Behavior of Slab-Column Connections,” ACI
fc′ = concrete cylinder compressive strength Structural Journal, V. 95, No. 3, May-June 1998, pp. 227-237.
fy, fy,1 = yield strength of slab top reinforcement 15. Broms, C. E., “Elimination of Flat Plate Punching Failure Mode,”
fy,2 = yield strength of slab bottom reinforcement ACI Structural Journal, V. 97, No. 1, Jan.-Feb. 2000, pp. 94-101.
h = slab thickness 16. McHarg, P. J.; Cook, W. D.; Mitchell, D.; and Yoon, Y., “Benefits of
M = unbalanced moment transferred from column to slab Concentrated Slab Reinforcement and Steel Fibers on Performance of
Mfn,1 = flexural capacity for negative bending evaluated in c + 3h Slab-Column Connections,” ACI Structural Journal, V. 97, No. 2, Mar.-
Mfn,2 = flexural capacity for positive bending evaluated in c + 3h Apr. 2000, pp. 225-234.
Mn = nominal unbalanced moment-carrying capacity 17. Osman, M.; Marzouk, H.; and Helmy, S., “Behavior of High-Strength
Mn,1 = flexural capacity for negative bending evaluated in c + d Lightweight Concrete Slabs under Punching Loads,” ACI Structural Journal,
Mn,2 = flexural capacity for positive bending evaluated in c + d V. 97, No. 3, May-June 2000, pp. 492-498.
Mn,ACI = estimated unbalanced moment-carrying capacity based on ACI 18. Ebead, U., and Marzouk, H., “Strengthening of Two-Way Slabs
design approach Using Steel Plates,” ACI Structural Journal, V. 99, No. 1, Jan.-Feb. 2002,
Mn,f = unbalanced moment at flexural failure pp. 23-31.
Mn,v = unbalanced moment at shear failure 19. Harajli, M. H., and Soudki, K. A., “Shear Strengthening of Interior
Mult = measured unbalanced moment-carrying capacity Slab-Column Connections Using Carbon Fiber-Reinforced Polymer
Mvn,1 = unbalanced moment resisted by Vn,1 Sheets,” Journal of Composites for Construction, ASCE, V. 7, No. 2, May
Mvn,2 = unbalanced moment resisted by Vn,2 2003, pp. 145-153.
s1 = average spacing of slab top reinforcement placed in c + d 20. Pilakoutas, K., and Li, X., “Alternative Shear Reinforcement for
s2 = average spacing of slab bottom reinforcement placed in c + d Reinforced Concrete Flat Slabs,” Journal of Structural Engineering,
T0 = torsional capacity without interaction with shear ASCE, V. 129, No. 9, Sept. 2003, pp. 1164-1172.
Tn = torsional capacity with interaction with shear 21. Binici, B., and Bayrak, O., “Punching Shear Strengthening of
Vc = nominal two-way shear strength (ACI 318-05) Reinforced Concrete Flat Plates Using Carbon Fiber Reinforced Polymers,”
Vflex = gravity shear at formation of yield-line mechanism Journal of Structural Engineering, ASCE, Sept. 2003, pp. 1173-1182.
Vg = gravity shear transferred from slab to column 22. Ospina, C. E.; Alexander, S. D. B.; and Cheng, J. J. R., “Punching
Vn = nominal connection shear capacity of Two-Way Concrete Slabs with Fiber-Reinforced Polymer Reinforcing
Vn′ = nominal connection shear capacity when slab bottom surface in Bars and Grids,” ACI Structural Journal, V. 100, No. 5, Sept.-Oct. 2003,
tension pp. 589-598.
Vn,1 = maximum shear developed at back face of critical section 23. Ebead, U., and Marzouk, H., “Fiber-Reinforced Polymer Strengthening
Vn,2 = maximum shear developed at front face of critical section of Two-Way Slabs,” ACI Structural Journal, V. 101, No. 5, Sept.-Oct. 2004,
VT = shear force at torsional face of critical section pp. 650-659.
Vult = measured connection gravity load capacity 24. Adetifa, B., and Polak, M. A., “Retrofit of Slab Column Interior
Connections Using Shear Bolts,” ACI Structural Journal, V. 102, No. 2,
ρ = top reinforcement ratio evaluated in c + 12d at column
Mar.-Apr. 2005, pp. 268-274.
γf = fraction of unbalanced moment transferred by flexure
25. Chen, C., and Li, C., “Punching Shear Strength of Reinforced Concrete
γv = fraction of unbalanced moment transferred by shear
Slabs Strengthened with Glass Fiber-Reinforced Polymer Laminates,” ACI
ξ = parameter reflecting effect of c/d on punching strength
Structural Journal, V. 102, No. 4, July-Aug. 2005, pp. 535-542.
26. Tian, Y.; Jirsa, J. O.; Bayrak, O.; Widinato; and Argudo, J. F.,
REFERENCES “Behavior of Slab-Column Connections of Existing Flat-Plate Structures,”
1. Criswell, M. E., “Static and Dynamic Response of Reinforced ACI Structural Journal, V. 105, No. 5, Sept.-Oct. 2008, pp. 561-569.
Concrete Slab-Column Connections,” Shear in Reinforced Concrete, SP-42, 27. Kanoh, Y., and Yoshizaki, S., “Strength of Slab-Column Connections
V. 2, American Concrete Institute, Farmington Hills, MI, 1974, pp. 721-746. Transferring Shear and Moment,” ACI JOURNAL, Proceedings V. 76, No. 3,
2. Moe, J., “Shearing Strength of Reinforced Concrete Slabs and Footings Mar. 1979, pp. 461-478.
under Concentrated Loads,” Development Department Bulletin No. D47, 28. Kanoh, Y., and Yoshizaki, S., “Experiments on Slab-Column and
Portland Cement Association, Skokie, IL, Apr. 1961, 130 pp. Slab-Wall Connections of Flat Plate Structures,” Concrete Journal, Japan
3. Elstner, R. C., and Hognestad, E., “Shearing Strength of Reinforced Concrete Institute, V. 13, June 1975, pp. 7-19. (in Japanese)
Concrete Slabs,” ACI JOURNAL, Proceedings V. 53, No. 1, July 1956, pp. 29-58. 29. Morrison, D. G.; Hirasawa, A. M.; and Sozen, M. A., “Lateral-Load
4. Regan, P. E., and Braestrup, M. W., “Punching Shear in Reinforced Tests of R/C Slab-Column Connections,” Journal Structural Engineering,
Concrete: A State of the Art Report,” Bulletin D’Information, No. 168, ASCE, V. 109, No. 11, Nov. 1983, pp. 2698-2714.
Comité Euro-International du Béton, Lausanne, 1985, 232 pp. 30. Hawkins, N. M.; Mitchell, D.; and Sheu, M. S., “Cyclic Behavior of

ACI Structural Journal/November-December 2008 699

Six Reinforced Concrete Slab-Column Specimens Transferring Moment 37. Roberson, I. N.; Kawai, T.; Lee, J.; and Enomoto, B., “Cyclic Testing
and Shear,” Progressive Report 1973-1974, NSF Project GI-38717, Section II, of Slab-Column Connections with Shear Reinforcement,” ACI Structural
Department of Civil Engineering, University of Washington, Seattle, WA, Journal, V. 99, No. 5, Sept.-Oct. 2002, pp. 605-613.
1974, 50 pp. 38. Stark, A.; Binici, B.; and Bayrak, O., “Seismic Upgrade of Reinforced
31. Islam, S., and Park, R., “Tests on Slab-Column Connections with Concrete Slab-Column Connections Using Carbon Fiber-Reinforced
Shear and Unbalanced Flexure,” Proceedings, ASCE, V. 102, No. ST3, Polymers,” ACI Structural Journal, V. 102, No. 2, Mar.-Apr. 2005, pp. 324-333.
Mar. 1976, pp. 549-568. 39. ACI Committee 318, “Building Code Requirements for Structural
32. Ghali, A.; Elmasri, M. Z.; and Dilger, W., “Punching of Flat Plates Concrete (ACI 318-05) and Commentary (318R-05),” American Concrete
under Static and Dynamic Forces,” ACI JOURNAL, Proceedings V. 73, No. 10, Institute, Farmington Hills, MI, 2005, 430 pp.
Oct. 1976, pp. 566-572. 40. Joint ACI-ASCE Committee 326, “Shear and Diagonal Tension,”
33. Pan, A., and Moehle, J. P., “An Experimental Study of Slab- ACI JOURNAL, Proceedings V. 59, No. 3, Mar. 1962, pp. 353-396.
Column Connections,” ACI Structural Journal, V. 89, No. 6, Nov.-Dec.
1992, pp. 626-638. 41. Zsutty, T. C., “Beam Shear Strength Prediction by Analysis of Existing
34. Robertson, I. N., and Durrani, A. J., “Gravity Load Effect on Seismic Data,” ACI JOURNAL, Proceedings V. 65, No. 11, Nov. 1968, pp. 943-951.
Behavior of Interior Slab-Column Connections,” ACI Structural Journal, 42. Zsutty, T., “Shear Strength Prediction for Separate Categories of
V. 89, No. 1, Jan.-Feb. 1992, pp. 37-45. Simple Beam Tests,” ACI JOURNAL, Proceedings V. 68, No. 2, Feb. 1971,
35. Wey, E. H., and Durrani, A. J., “Seismic Response of Interior Slab- pp. 138-143.
Column Connections with Shear Capitals,” ACI Structural Journal, V. 89, 43. Kinnunen, S., and Nylander, H., “Punching of Concrete Slabs
No. 6, Nov.-Dec. 1992, pp. 682- 691. without Shear Reinforcement,” Meddelande Nr. 38, Institute for
36. Durrani, A. J.; Du, Y.; and Luo, Y. H., “Seismic Resistance of Byggnadsstatik, Kungliga Tekniska Hogskolan, Stockholm, Sweden, 1960.
Nonductile Slab-Column Connections in Existing Flat-Slab Buildings,” 44. Park, R., and Gamble, W. L., Reinforced Concrete Slabs, second
ACI Structural Journal, V. 92, No. 4, July-Aug. 1995, pp. 479-487. edition, John Wiley & Sons, 2000, 736 pp.

700 ACI Structural Journal/November-December 2008