16th International Conference on Composite Structures

ICCS 16
A. J. M. Ferreira (Editor)
FEUP, Porto, 2011

PRIMARY AND SECONDARY REINFORCEMENTS IN CORBELS

UNDER COMBINED ACTION OF VERTICAL AND HORIZONTAL
LOADINGS
Mehdi Rezaei1, S.A.Osman2, N.E. Shanmugam3
Department of Civil & Structural Engineering, Faculty of Engineering & Built Environment
National University of Malaysia (UKM)
1
[email protected], [email protected], [email protected]

Keywords: Corbel, Combination of Horizontal and Vertical Loading

Summary. This paper is concerned with the effects of primary and secondary reinforcements
in corbels under combined action of horizontal and vertical loadings. Strut-and-Tie models
and Cantilever Beam method recommended by PCI (Pre-stress/Precast Concrete Institute) are
considered. The two methods are assessed and validated. The results showed that Cantilever
Beam method is suitable if 20% or less of the vertical load is involved as the horizontal force.
It was also found that Strut-and-Tie method provides more primary and secondary
reinforcements compared to the Cantilever Beam method.

1
1 INTRODUCTION
Extensive experimental programs have been undertaken over the last decades to establish
the behavior of reinforced concrete non-flexural members, such as deep beams, nibs, corbels,
beam-column joints. Experimental studies have shown that non-flexural members display an
increased shear capacity relative to flexural members. Based on the results obtained from the
studies a number of empirical models were developed for design of corbels. The main
limitations of the empirical approaches are a limited range of shear span to depth ratios for
which design equations apply and, an inability to explain precisely the mechanics of corbels
behavior. Finite element studies incorporating non-linear materials models have been used in
combination with experimental programs to determine the mechanics of non-flexural
members including corbels. 16 life-size corbels were tested by Yong et al. [1994] to
investigate the effects of horizontal force, reinforcement ratio, and shear-span-to-depth ratio.
A specially designed test setup was used to induce a horizontal load equal to 20% of the
vertical load. The primary steel yielded before failure almost in all cases. The truss analogy
model was found to provide relatively accurate strength predictions compared to the
American Concrete Institutes procedure.

Six full-scale reinforced concrete corbel specimens were tested by Mohamed Almeer
[2004], to study the influence of steel and polypropylene fibers, headed bars, and horizontal
loading. The experimental values of ultimate load capacity were compared with predictions
using simplified and refined Strut-and-Tie models; the refined model was found to provide
better prediction. The results showed that the addition of horizontal force (20% of vertical) to
the test specimen resulted in ultimate load carrying capacity dropping by almost 20%.
Addition of this force increased the flexibility of the test specimens. Studies have been
carried out on high-strength reinforced concrete corbels by researchers [Stephan, 1997;
Mohamed et al., 1997; Fattuhi, 1994, 1995] with artificial substances such as steel fiber,
plastic mesh, and etc. Only vertical load was considered in these studies. Comparison
between the experimental and calculated strengths of corbels showed that the two values
obtained in each case are in satisfactory agreement. It is clear from what has been observed
that the studies on the effects of horizontal force on strength of reinforced concrete corbels are
inconclusive. One reason why the researchers avoided the study with the application of
horizontal and vertical forces simultaneously could be because of the practical problems that
may arise in realizing that type of loading in the tests. In order to fill the gap in the results for
the combined action of horizontal and vertical loading on corbels, the current study aims to
provide some information in this regard. In this paper corbels are designed with different
ratios of primary and secondary reinforcements and they are analyzed by two methods viz.
Strut-and-Tie model and Cantilever Beam model, recommended by PCI. Finite element
software package, LUSAS version14.1, was employed in the investigations.

2
2 FINITE ELEMENT ANALYSIS OF CORBELS

2.1 Finite Element Modeling and Its Accuracy

Finite element software package LUSAS was used to analyze corbels in this study. A
typical meshed corbel adopted in the analysis is shown in Fig. 1. It is necessary to establish
the accuracy of the finite element modeling. Therefore, seven corbels tested by Stephan
[1996] were considered in this study. All specimens were modeled and analyzed using
LUSAS and the results obtained are shown in Fig. 2 in which the experimental ultimate load
is plotted against the corresponding values obtained from the finite element analysis for all the
seven specimens. It is clear from the figure that finite element results are close to the
corresponding experimental values, maximum deviation being 15%. The difference could be
attributed to the assumptions used in the modeling since some of the material properties and
dimensions are not given in the paper. Since LUSAS was found to predict the results with
acceptable accuracy the software package was used for further analyses.

Fig. 1 A Half Meshed Corbel, Using LUSAS

Fig. 2 Comparison between Finite Element Results and Test Results

3
2.2 Details of The Corbels
Detailed dimensions of the corbels and the reinforcements are shown in Fig. 3. For all
corbels considered in the analyses the overall dimensions were kept the same. The corbel
width, 254 mm was the same as the width of the column supporting the corbels, the overall
corbel depth and its effective depth being 406 mm and 376 mm, respectively. The depth at
the free edge of the corbel was 203 mm. The cross section of the column supporting the
corbel was 305 mm 254 mm with total length taken as 965 mm for the purpose of analysis.
The column was reinforced with four 16-mm diameter bars located at the corners and tied by
9 mm diameter ties spaced at 216 mm center to center along the column length. The primary
and secondary reinforcements designed as per the code requirements are summarized in Table
1. Secondary reinforcement usually includes two or more steel bars. The total cross-sectional
area for secondary reinforcement is generally divided into a number of steel bars spread
equally. In this study, two secondary reinforcements were used. In all corbels, the materials
used were the same having properties such as compressive strength of concrete equal to 30
MPa, the yield stress of steel equal to 420 MPa, Youngs modulus of concrete and steel equal
to 26103 N/mm2 and 209103 N/mm2, respectively, and Poissons ratio of 0.2 and 0.3 for
concrete and steel, respectively.

There are two methods, Cantilever Beam method and Strut-and-Tie method, proposed in
the PCI recommendations to determine the primary and secondary reinforcements for corbels.
For the corbel dimensions chosen, area of reinforcements has been calculated by the two
methods for different combinations of vertical and horizontal loadings the details of which are
shown in Table 1. In the table, the corbel specimens with different combinations of horizontal
and vertical loadings are identified as H/V0.0, H/V0.1 etc. indicating the horizontal to vertical
loading ratios of 0.0 and 0.1 and so on. Both primary and secondary reinforcement areas
calculated by the two methods are presented in the table. In Strut-and-Tie method, however,
neither primary nor secondary reinforcement values are given for corbels H/V0.8, H/V0.9 and
H/V1.0. In these corbels, the quadratic expression derived for the determination of
reinforcement areas by truss analogy method results in larger column width, not covered in
the present study. In the finite element analyses of the corbels the top and the bottom ends of
the column were assumed clamped with the load applied incrementally on a bearing pad,
100mm 254mm 15mm, made of steel as shown in the figure. Taking advantage of
symmetry in geometry, loading and support conditions only a half of the specimen was
modeled for the analyses. Mesh size of 3030 mm and distribution of mesh similar to the
one shown in Fig. 1 were chosen based on convergence studies carried out to determine the
optimal mesh that gives a relatively accurate solution and one that takes low computational
time. Results are presented in the form of plots, horizontal to vertical load ratios against
reinforcement areas and, vertical and horizontal loads against the respective displacements.

4
 Fig. 3 Reinforcement Detailing (Dimensions in mm)
3 RESULTS AND DISCUSSION
Figs. 4 and 5 show the variation of primary and secondary reinforcements, respectively,
with horizontal to vertical load ratios. In each of the two figures, the variations as obtained by
the two methods, Cantilever Beam method and Strut-and-Tie method are compared. It can be
seen from Fig. 4 that Strut-and-Tie method provides larger cross sectional area for primary
reinforcements. The difference in the area of primary reinforcements gradually increases with
the load ratios. For example, the area of reinforcement predicted by the Cantilever Beam
method and Strut-and-Tie method for load ratio of 0.0 is 272 mm2 and 355 mm2, respectively
with a difference of 83 mm2 whilst the corresponding values for load ratio of 0.7 are 752 mm2
and 949 mm2 with a difference of 197 mm2. Variations in secondary reinforcements with load
ratio are totally different, as predicted by the two methods. It can be seen from Fig. 5 that the
Cantilever Beam method provides almost same area of secondary reinforcement irrespective
of the load ratios but on the other hand the area of secondary reinforcement provided by Strut-
and-Tie method increases linearly with load ratio. Strut-and-Tie method theoretically does
not provide any secondary reinforcement when horizontal force is not involved. A minimum
amount of horizontal stirrup reinforcement must, however, be provided to avoid diagonal
tension failure. Therefore, the ACI 318 recommends designing corbels for a minimum of 0.2
times vertical force applied in the horizontal direction. It is clear from Fig. 5 that Strut-and-
Tie method gives secondary reinforcement quite close to that by Cantilever Beam method for
a load ratio of around 0.2.

5
 Table 1: Primary and Secondary Reinforcements
Vertical Beam theory Truss Analogy
Horizontal
Designation Load
Load (kN) Aprim Asec Aprim Asec
(kN)
H/V0.0 200 0 272 136 355 0
H/V0.1 200 20 341 139 433 63
H/V0.2 200 40 409 141 513 126
H/V0.3 200 60 478 144 594 190
H/V0.4 200 80 546 146 677 254
H/V0.5 200 100 615 149 762 317
H/V0.6 200 120 683 151 850 381
H/V0.7 200 140 752 153 949 444
H/V0.8 200 160 821 156
H/V0.9 200 180 890 159
H/V1.0 200 200 958 161

Fig. 4 Cross Sectional Areas of Primary Reinforcement versus Ratios of Horizontal to Vertical Force

6
 Fig. 5 Cross Sectional Areas of Secondary Reinforcement versus Ratios of Horizontal to Vertical Force

It has been noted in the ultimate load analyses of the corbels that a crack initiates first at the
re-entrant corner and then propagates along the column-corbel interface. At the inner edge of
the bearing plate a second crack is formed propagating much faster than the first one. While
the first crack continues to propagate along the column face the second crack progresses
towards the junction of the column and the sloping face of the corbel. The second crack,
which becomes the primary or major crack, eventually runs between the inner edge of the
bearing plate and the column-corbel junction at the sloping face and results in the failure of
the corbel. Ten reinforced concrete corbels, listed in Table 1, designed by Cantilever Beam
method were analyzed by the finite element method and the results are presented in the form
of load-displacement plots as shown in Figs. 6 and 7. Vertical or horizontal displacements
plotted on the horizontal axis correspond to those measured under the load. Fig. 6 shows the
load-deflection curves for vertical loads and the corresponding plots for horizontal loads are
given in Fig. 7. In each of these figures curves are plotted showing the variation of deflection
with vertical load or horizontal load for different values of load ratios. It can be seen from the
figures that the load deflection relation is nonlinear even from the early stages of loading. The
horizontal and vertical load interaction is reflected in the two sets of curves. In the presence
of vertical load the gradual increase in horizontal load results in gradual drop in vertical load
capacity. In all cases the load tends to increase and stop further increase when the ultimate
load condition is reached. The results for ultimate loads are summarized in Table 2. The
results show that the Cantilever Beam method is appropriately able to predicate the behavior

7
of corbels without horizontal load, on the contrary it grossly overestimates the load-carrying
capacity of corbels when horizontal load is involved, even though, for the corbel involving
20% of the vertical force is still satisfactory. It is clear from the results that Cantilever Beam
method is overestimated when greater than 20% of vertical load is involved as horizontal load.
It should be noted that the horizontal to vertical force ratio is limited to 1.0 by the PCI
(Nu/Vu 1). Unlike vertical load-deflection curves, in horizontal load-deflection curves the
ductility enhances. Also, the rate change of the ductility of the horizontal load-deflection
curves is considerably greater than the vertical load-deflection curves as well as toughness of
the horizontal load-deflection curves.

Fig. 6 Vertical Load-Deflection Curves

Fig. 7 Horizontal Load-Deflection Curves

8
 Table 2 Results for ultimate loads
Ultimate Ultimate
Designation Vertical load Horizontal load
(kN) (kN)
H/V0.0 302 0
H/V0.1 215 22
H/V0.2 203 41
H/V0.3 251 75
H/V0.4 224 89
H/V0.5 197 99
H/V0.6 187 112
H/V0.7 161 113
H/V0.8 149 119
H/V0.9 135 121
H/V1.0 153 123

CONCLUSIONS
Finite element method has been employed to analyze corbels subjected to combined action of
vertical and horizontal loads. Two different methods, Cantilever Beam Method and Strut-
and-Tie method proposed by the PCI have been used to determine secondary and primary
reinforcements and the analyses carried out to determine the ultimate load carrying capacity.
From the results obtained following conclusions are drawn:

1. The amount of primary and secondary reinforcements obtained by using the Strut-and-Tie
method is more than the corresponding amount provided by the Cantilever Beam method.
2. As per the Strut-and-Tie method, the cross sectional area of secondary reinforcement is
considerably larger in the presence of horizontal load.
3. Cantilever Beam method is suitable when 20% or less of vertical load is involved.
4. As expected, the ultimate vertical load carrying capacity drops considerably in the
presence of horizontal load.

REFERENCES
[1] Yook-Kong Yong, and P. Blaguru, Behavior of Reinforced High-Strength- Concrete
Corbels, American Society of Civil Engineers (ASCE), Vol.120, No.4, pp. 1182-1201(1982).
[2] Nijad I. Fattuhi. Reinforced Corbel Made With High-Strength Concrete and Various
Secondary Reinforcement, ACI Structural Journal, Vol.91, No.3, pp.345-368 (1994).
[3] S.J. Foster, R.E. Powell, and H.S. Selim, Performance of High Strength Corbels, ACI
Structural Journal 93, pp. 555563 (1997).

9
 [4] Nijad I. Fattuhi ,Strength of FRC Corbels in Flexure, Journal of Structural Engineering,
Vol. 120, No. 2, , pp. 360-377 (1994).
[5] PCI Design Handbook (6th Edition), Precast and Prestressed Concrete Institute, Chapter 6,
46 pp (2004).
[6] Mohamed Almeer, Effects of Fibers and Headed Bars on the Response of Concrete
Corbels, MS Thesis, Department of Civil Engineering and Applied Mechanics, McGill
University (2004).
[7] Moahmed S., and Houssam T., Design of High Preformance Concrete Corbels, Journal of
Materials and Structures, Vol.31, November, pp 616-622 (1998).
[8] ACI Committee 318, Buildding Code Requiremtns for reinforced Concrete (ACI 318),
American Concrete Institute.

10