Documente Academic
Documente Profesional
Documente Cultură
HIGH-STRENGTH-CONCRETE CORBELS
By Yook-Kong YongI and P. Balaguru,2 Members, ASCE
Downloaded from ascelibrary.org by National Institute of Technology Warangal on 09/20/18. Copyright ASCE. For personal use only; all rights reserved.
INTRODUCTION
While extensive research has been done on the various aspects of high-
strength concrete, the use of high-strength concrete in reinforced corbels
has received very little attention from researchers (Yong et al. 1985). Most
of the existing research on the behavior of corbels deals only with n o r m a l
strength concrete (Franz and Niedenhoff 1964; H a g b e r g 1983; H e r m a n s e n
and Cowan 1974; Kriz and Raths 1965; M a t t o c k et al. 1976; Mast 1968).
Most code provisions also are based on the results obtained using concretes
with strengths less than 40 MPa. This p a p e r presents some results of an
investigation on the behavior of high-strength corbels. The main objective
of the experimental investigation was to study the behavior of h i g h - s t r e n g t h -
concrete corbels subjected to both horizontal and vertical loads.
The researchers who conducted experimental studies using normal-strength
concrete concluded that the shear strength is a function of: (1) S h e a r - s p a n -
t o - d e p t h ratio; (2) reinforcement ratio; (3) concrete strength; and (4) the
ratio of the horizontal and vertical components of the applied loads (Kriz
and Raths, 1965; Mattock et al. 1976). A minimum amount of horizontal
stirrup reinforcement must be p r o v i d e d to avoid diagonal tension failure.
It was also r e p o r t e d that in most cases the primary reinforcement steel
yielded before failure. Based on these results, s h e a r - s p a n - t o - d e p t h (a/d)
and reinforcement ratios were chosen as the primary variables. Increase in
concrete strength can be expected to increase the shear strength of the
corbel. However, since high-strength concrete is typically more brittle than
normal-strength concrete, it should be ascertained that higher strength can
be used without leading to catastrophic brittle failure. Yielding of steel
reinforcement prior to failure and extensive cracking typically lead to ductile
1Assoc. Prof., Dept. of Civ. and Envir. Engrg., Rutgers-the State University of
New Jersey, P.O. Box 909, Piscataway, NJ 08855-0909.
2prof., Dept. of Civ. and Envir. Engrg., Rutgers-the State University of New
Jersey, Piscataway, NJ.
Note. Discussion open until September 1, 1994. To extend the closing date one
month, a written request must be filed with the ASCE Manager of Journals. The
manuscript for this paper was submitted for review and possible publication on
August 14, 1992. This paper is part of the Journal of Structural Engineering, Vol.
120, No. 4, April, 1994. 9 ISSN 0733-9445/94/0004-1182/$2.00 + $.25 per
page. Paper No. 4611.
1182
EXPERIMENTAL PROGRAM
The primary independent variables were concrete strength, amount of
reinforcement, s h e a r - s p a n - t o - d e p t h (a/d) ratio, and the presence of hor-
izontal load. The response variables were failure mode, rotational capacity,
strain variations in the reinforcement, and the strength capacity. A total of
16 life-size corbels were tested. The pertinent details are presented in Table
1. The targeted range for concrete strength was 50-90 MPa. The actual
variation was 39.4-87.1 MPa. An attempt was made to use higher-strength
concrete for corbels containing more reinforcement. However, strength var-
iations could not be controlled well; therefore, the series D corbels with
higher reinforcement ratios had lower-strength concrete compared with the
series C corbels (Table 1).
The area of primary reinforcement was varied from 0 to 1,923 mm 2.
The corbels were designed using AC1 318-89 ("Building" 1989). Series B,
C, and D had reinforcement at or slightly below the minimum specified
in the code. Corbels E3, F2, and G3 had 50% more steel than the ACI
code-recommended maximum limit. The higher level of reinforcement was
1183
~-305 -~
305
I i/~.... 3o~
L/
I017
{a} l- ~'13 (b) (ALL DIMENSIONS IN mm)
FIG. 1. Corbel Dimensions: (a) Series A - F (for A - D , a = 140 mm; E, 89 mm; and
F, 178 mm); and (b) Series G
1184
t ~ 4 - #5 BARS
305
+ #3.._/// ~ U # 3 TIES @ 216 mm c/~
2o32~ ] -#3
-[5o
254 =~
1185
INSTRUMENTATION
1186
~
LOCATION
OF
STRAIN CAGES
2
11
i
Q
LVDT3
LVDTI
SUPPORT SUPPORT
FIG. 3(a). Location of Strain Gages and LVDTs plus Load Setup for Series B - D
OAOES 5 AND 6
\ _e /
, , ,, .
FIG. 3(b). Location of Strain Gages and LVDTs for Series E-G
Fewer gages were placed on the stirrups because the strains there were
typically low.
For series B - D , three L V D T s were used to measure the downward de-
flection and rotation of cantilevers at the column face, and their locations
are shown in Fig. 3(a). Since these deformations were very small, this mea-
1187
TEST PROCEDURE
~ 7 0 0 k ~
HYD. JACK
STEEL
PLATE
----Z__-.-LOADING
WEDGE PLATE
~SUPPORTING BEAM
o ............ o.;
FIG. 4. Details of Test Setup
1188
9 ' 9 9
Downloaded from ascelibrary.org by National Institute of Technology Warangal on 09/20/18. Copyright ASCE. For personal use only; all rights reserved.
BEARING PAD
p. .."
...'.,
SUPPORT
. .,..b~.
t.~
0.2 was chosen to satisfy the ACI 318-89 ("Building" 1989) requirement
that the corbels be designed for a minimum of 0.2 vertical load applied in
the horizontal direction. A Teflon bearing pad was placed at the top wedge,
and the bottom wedge was covered with a stainless-steel plate to minimize
friction. The Teflon pads used were similar to those used for roller supports
in bridge construction and manufactured to comply with near-zero-friction
specification. The load was transferred to the corbel using a channel welded
to the primary steel (Fig. 5). Since the wedges at the two supports had the
same geometry, the specimens remained in equilibrium.
The loads were applied in increments of 45 kN using a hydraulic loading
system. The strains and deformations were recorded for each increment
using a computer-controlled Hewlett-Packard recording system. For each
corbel, the companion 150 by 300 mm cylinders were tested on the same
day to obtain the compressive strength of the concrete.
TEST RESULTS
General Behavior and Crack Patterns
The unreinforced corbels, series A, failed right after initiation of the first
crack at the reentrant corner of the horizontal surface of corbel and the
vertical face of column9 The crack propagated rapidly along the column-
corbel interface, resulting in a sudden brittle failure.
Series B - D specimens shared similar failure patterns. All six specimens
in these series had a s h e a r - s p a n - t o - d e p t h ratio of 0.39 and were subjected
to only a vertical load. First, a crack initiated at the reentrant corner, and
while this crack was propagating along the column-corbel interface, a second
crack formed at the inner edge of the bearing plate, as shown in Fig. 6. The
second crack propagated much faster than the first crack. While the first
crack continued to propagate along the column face, the second crack pro-
gressed toward the junction of the column and the sloping face of the corbel.
The second crack, which now became the primary or major crack, eventually
ran between the inner edge of the bearing plate and the column-corbel
junction at the sloping face, and was responsible for the failure of the corbel.
1189
FAILURE
PLANE~ I
INITIAL
FLEXURE
CRACK
FIG. 6. Failure Pattern: Series B - D
Failure was defined as the point at which the load could no longer be
increased. Some minor secondary cracks also formed before failure (Fig.
6). The failure was stable and ductile. The failure pattern for series B - D
could be classified as beam-shear failure, which is a mode of failure initiated
by a flexural-type crack, and failure is caused by diagonal-shear-type cracks
(Mattock et al. 1976).
The failure and crack patterns of series E - G corbels were somewhat
similar to series B - D in spite of the variation in the a/d ratio from 0.25 to
0.75 and the presence of the horizontal load. The first crack to appear
during the loading sequence was a flexural crack similar to that of a can-
tilevered beam. After approximately one-quarter to one-third of the ulti-
mate load was applied, a second crack started at the bearing plate, and
propagated toward the junction of the column and the sloping face of the
corbel. This crack caused the ultimate failure of the corbel. The failure was
ductile, and the mode of failure (beam-shear failure) was the same for all
the specimens.
The major differences between series B - D , and E - G were in the area
of secondary cracks and the spalling of the concrete between the cracks.
For series E - G , the crack that started at the reentrant corner propagated
into the column slightly, as shown in Fig. 7(a)-7(c), respectively. As the
a/d ratio increased, the second crack, which initiated at the bearing plate,
followed a more curved path to the top of the corbel, as shown in Fig. 7(c).
FAILURECRACK
BEAM- SHEAR | I
~
FAILURE / / /
CRA~ \
/
INITIAL
1191
~
o W " .o'"
250 m" " .u'"
m" s'"
o
~' I I ~
0.2 0,4 0.6
the results are shown for only one specimen. The following observations
can be made regarding the (primary) steel strain variations.
The curves had either two or three segments. The first segment, which
was the steepest, terminated at the occurrence of the first crack. After the
initiation of the first crack, the strains increased rapidly for a small load
increment. This is quickly followed by the initiation of the second (primary)
crack, at which time, the load-strain response became relatively linear until
approximately 90% of the ultimate load. For specimen G3, which failed
before the yielding of steel, the curve had only these two segments. For all
the other specimens that failed after yielding of the primary steel, the load-
strain curves had a third segment which was relatively horizontal.
Corbels E3, F2, and G3 had 50% more reinforcement over the limit
specified in ACI 318-89. Only in the case of G3 did the reinforcement not
yield. Even in this case, the failure was not sudden. Considerable spalling
and cracking occurred before the failure.
The a/d ratio does not seem to influence the yielding of the reinforcement.
80(
Downloaded from ascelibrary.org by National Institute of Technology Warangal on 09/20/18. Copyright ASCE. For personal use only; all rights reserved.
~a
E2
o~ 60( -- CORBEL
o~
400
200
800
-- CORBEL, Pl P2
60O
o~ 400
200
widths in Fig. 12 since these readings primarily represent the width of the
crack openings, as explained earlier. The locations of the LVDTs are shown
in Fig. 3(b). The LVDT curves follow a pattern similar to that of the primary
steel curves. The first segment is very steep, followed by a kink, and rapid
increase of crack width coincident with the formation of the primary crack.
Shear Capacity
The shear strength data is presented in Table 4. The unreinforced corbels
A1 and A2 had shear strength of 206.8 and 166.8 kN, respectively. The
corresponding compressive strengths of concrete were 87.1 and 70.7 MPa.
The 20% decrease in shear capacity of A2 compared with A1 was about
the same as the decrease in the compressive strength of 19%. The first crack
load of the reinforced corbels, which varied from 250 to 320 kN, was always
1193
ILl"
o SO0
250
//,, 0,08
STEEL STRAIN, %
0.16
800
..~ 600
400
200
higher than the shear capacity of the unreinforced corbels. The increase in
Young's modulus of the uncracked corbel due to the presence of the re-
inforcement could account for this increase.
If the reinforcement and a/d ratio are the same, corbels with higher-
strength concrete recorded higher shear capacity (corbels B1, D2, E2, and
G2 versus B2, D1, E l , and G1). The increase in shear strength, however,
was not proportional to the increase in the compressive strength of concrete.
The influence of primary reinforcement on shear strength could not be
seen clearly because of the variation in compressive strength of concrete.
There is a general trend indicating increase in strength with an increase in
reinforcement.
Corbels with an a/d ratio of 0.75 had much lower shear strengths than
the other corbels. This is expected because the influence of bending moment
1194
CORBEL F2
Downloaded from ascelibrary.org by National Institute of Technology Warangal on 09/20/18. Copyright ASCE. For personal use only; all rights reserved.
500
400
300 G2
200
lOO
[ I I ;
2 4 6 8 10
CRACKWIDTH, mm
is more substantial at this a/d ratio compared with ratios of 0.25, 0.39, and
0.5.
Shear Maximum
friction Flexure shear a Truss Test/ACI b Test/truss
Corbel Test [(1)] [(2)] capacity analogy method analogy
(1) (2) (3) (4) (5) (6) (7) (8)
A1 206.8 -- -- --
A2 166.8 . . . . .
B1 778.2 423.7 415.0 415.0 502.6 1.88 1.55
B2 667.2 423.7 414.8 414.8 502.6 1.61 1.33
C1 796.2 501.4 553.0 501,4 631.2 1.59 1.26
C2 836.2 501.4 551.3 501.4 618.2 1.67 1.35
D1 700.6 552.0 726.8 552.0 711.7 1,27 0.98
D2 800.6 552.0 737.4 552.0 756.2 1.45 1.06
E1 711.6 564.5 769.0 564.5 631.6 1.26 1.13
E2 800.6 591.2 817.9 591.2 631.6 1.35 1.27
E3 1,078.6 1,173.5 1,600.2 1,173.5 796.2 0.92 1.35
F1 911.8 737.9 666.8 666.8 435.9 1.37 2.09
F2 845.1 1,000.1 880.1 880.1 355.8 0.96 2.37
G1 333.6 871.2 569.0 569.0 289.1 0.59 1.15
G2 411.4 866.8 573.4 573.4 293.6 0.72 1.40
G3 556.0 1,173.5 764.5 764.5 266.9 0.73 2.08
"The shear capacity based on maximum permissible shear stress is not used for com-
parison for reasons explained in the text. The shear capacity based on a limit of 5.5bwd
is 498.1 kN (112 kips). If this limit is used, ACI 318-89 overestimates shear capacity for
corbels G1 and G2.
bAC[ values equal maximum shear capacity (column 5).
1196
;HEAR FAILUR~
1- ~RIMARY STEEL
AT INTERFACE YIELD OR BEAM
SHEAR FAILURE~
.4
~f
=N
/
FIG. 13. Four Basic Failure Modes ("Building" 1989)
at the interface of the column and the corbel. The unreinforced corbels A1
and A2 failed in this mode. The second mode of failure is assumed to occur
by the yielding of primary reinforcement (or tension tie). All 14 reinforced
corbels can be assumed to have failed in this mode of failure and not in the
third mode of failure, which occurs by crushing the concrete strut. This is
because there was no crushing of the concrete, even though there was
spalling in some corbels. The specimens were designed to prevent bearing
failure, which is also shown.
The analytical procedure specified in the code involves the computation
of the shear capacity based on shear failure and failure by yielding of steel.
The corbels have to be designed to prevent bearing failure. The equations
needed for the computation of shear capacity, V, are as follows ("Building"
1989).
For shear failure
V . = A,,ffyp~ (1)
1197
TENSION TIE IV
d
SION
9 Either the steel yields or the concrete crushes during failure. The
elastic strains of both concrete and steel are negligible compared
with the strains generated during the crushing of concrete or yielding
of steel.
9 Strength of concrete determined using cylinders represents the strength
of the compression strut in the corbel. The tensile-strength capacity
of concrete is negligible compared with the tensile strength of steel
reinforcement.
9 The reinforcements are assumed to be properly anchored.
The analysis indicates that all the corbels would fail by yielding of the
steel. Since high-strength concrete was used for all corbels, the shear strengths
computed based on strut failure were two to five times greater than those
computed for failure due to the yielding of steel. The failure loads and the
ratios of experimental to predicted loads using this truss analogy model are
shown in Table 4.
From Table 4, it can be seen that the truss analogy method provides a
better prediction than the ACI 318-89 procedure. The prediction was un-
conservative in one instance and the error was 2%. If the corbels containing
1199
CONCLUSIONS
Following are the experimental observations and comments on the anal-
ysis of experimental results.
All 14 corbels containing reinforcement failed in a beam-shear failure,
and the failure mode is stable.
The primary steel yielded before failure in all but one case. Even when
the failure occurred before the yielding of primary steel, the failure was not
sudden. Considerable spalling occurred before the failure.
The strain in the primary steel increases linearly after the initiation of
the flexure-type crack in the reentrant corner. This stable linear relationship
between steel strain versus load indicates that the behavior of high-strength
corbels is stable in the postcracking range.
Increase in compressive strength leads to increased corbel capacity, but
does not result in brittle failure. Increase in reinforcement ratio also in-
creases the strength capacity. Within the reinforcement ratios tested, the
increase in reinforcement did not increase the ductility at failure.
None of the specimens failed in a shear-friction mode of failure, even
though the shear strength capacities calculated for this type of failure was
much lower than the experimental strengths.
The behavior of high-strength-concrete corbels is similar to that of nor-
mal-strength-concrete corbels. The increase in concrete strength does not
adversely affect the ductility of corbel.
The American Concrete Institute code ACI 318-89 ("Building" 1989)
provision that the "average shear stress at the interface should be less than
5.5 MPa (800 psi)" limits the full use of the available shear strength of the
high-strength concrete. The high-strength-reinforced concrete corbels with-
stood a higher average stress without sustaining a brittle failure. In addition,
for high-strength-concrete corbels with a shear-span-to-depth (a/d) ratio
of 0.75, ACI 318-89 procedure leads to an unconservative prediction of
strengths. These provisions need changes for the use of high-strength con-
crete in corbels. In general, the experimental results, while limited in the
number of specimens and number of parameters considered, suggests that
the truss analogy method (Hagberg 1983) provided a better prediction of
shear strengths as compared to the ACI 318-89 procedure.
ACKNOWLEDGMENTS
The contributions of Douglas McCloskey and Richard Petrino for the
experimental program are gratefully acknowledged.
APPENDIX. REFERENCES
"Building code requirements for reinforced concrete. Special provisions for brackets
and corbels." (1989). ACI 318-89, American Concrete Institute, Detroit, Mich.,
318/318R-163-318/318R-166.
Franz, G., and Niedenhoff, H. (1964). "Reinforcement for brackets and short deep
beams." C & CA Library Translation No. 114, Cement and Concrete Association,
London, England.
Hagberg, T. (1983). "Design of concrete brackets: On the application of the truss
analogy." ACI J., 80(1), 3-12.
1200
Mattock, A. H., Chen, K. C., and Soonswang, K. (1976). "The behavior of rein-
forced concrete corbels." J. Prestressed Concr. Inst., 21(3), 18-42.
Mast, R. F. (1968). "Auxiliary reinforcement in concrete connections." J. Struct.
Div., ASCE, 94(6), 1485-1504.
Standards for concrete and mineral aggregates. (1989). Vol. 04.02, ASTM, Phila-
delphia, Pa.
Yong, Y. K., McCloskey, D. H., and Nawy, E. G. (1985). "Reinforced corbels of
high-strength concrete." A CI SP87-11, American Concrete Institute, Detroit, Mich.,
197-212.
1201