Structures Congress 2011 © ASCE 2011 152

Experimental Study of Shear Capacity of Reinforced Concrete Slabs

Eva Lantsoght1, Cor van der Veen1 and Joost Walraven1

1
Concrete Structures, Delft University of Technology, Stevin II-North 2.06,
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

Stevinweg 1, 2628 CN Delft, The Netherlands; PH 0031(0)152787449; FAX

0031(0)152787438; email: [email protected]

ABSTRACT

When assessing the capacity of existing reinforced concrete slab bridges

under the increased traffic loads prescribed in the current codes, shear may become
the critical failure mode. To better evaluate the shear capacity of reinforced concrete
slab bridges, a series of experiments is carried out on continuous one-way slabs
loaded close to the support. Eight continuous slabs of 5m x 2,5m x 0,3m are tested.
The loading position is taken at different a/d ratios. Six slabs with a standard
concrete mixture and two slabs with a higher strength concrete are tested. The
influence of the loading history, the shear span to depth ratio and the concrete
compressive strength is discussed. Conclusions on the influence of these parameters
on the one-way shear capacity of reinforced concrete slabs are drawn.

INTRODUCTION

Over the past decades traffic loads have increased significantly. Therefore,
the capacity of existing reinforced concrete bridges, designed for lower traffic loads,
is currently under discussion. In particular the shear capacity of existing slab bridges
is sometimes found to be insufficient, if checked with the current codes.
The governing design methods present a lower bound solution and usually are
a conservative prediction of the shear capacity. Generally structural members are
designed in such a way that flexural failure will occur before shear failure. Regarding
existing structures, residual bearing capacity is expected from factors such as the
increase of the concrete strength due to decades of further hydration of the cement
and possibly a higher concrete strength under long-term loading. The shear capacity
of deck slabs loaded close to the supports might also be higher than expected. Most
publications about shear in reinforced concrete concentrate either on one-way shear
in beams or on punching shear (two-way shear) in slabs. One-way slabs typically are
designed using beam shear formulas. For loads close to the supports and very wide
slabs a certain effective width beff is assumed. This effective width represents the
width of the support which carries the load. The shear stress v is taken as uniformly
distributed over the effective width. In practice this load spreading for wheel loads is
assumed to occur under 45° from the load towards the support, Figure 1. A minimum
value of 2d for loads in the middle of the width and of d at the edge and corner of the
slab is used in practice. Even though the design values of the shear capacity are

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 153

found to be very low and sometimes insufficient, previous small-scale testing

(Regan, 1982) has shown a considerable additional shear capacity for slabs under
concentrated loads close to the supports.
To investigate this phenomenon an extensive experimental program is conducted
at Delft University of Technology, The Netherlands. The variables in this test
program are:
• the shear span to depth ratio (a/d),
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

• the reinforcement layout,

• the concrete compressive strength,
• the width of the slab,
• the position of the load along the width,
• the size of the loading plate,
• the type of support (simple or continuous),
• the type of reinforcement (deformed bars versus plain bars),
• the support layout (line support or bearing) and
• the loading history.
More background to the test program can be found in (Lantsoght et al., 2010b). The
influence of the size of the wheel load, the amount of transverse flexural
reinforcement, the type of the support (simple or continuous) and a comparison with
code provisions and with Regan’s method (Regan, 1982) for the first four tested
slabs is given in (Lantsoght et al., 2010a).

Figure 1: Effective width assuming 45° load spreading, top view showing
concentrated load and line support.

EXPERIMENTS

Test setup. Slabs of 5m x 2,5m x 0,3m representing bridge decks on scale 1:2 are
tested. A sketch of the top view and side view of the test setup is presented in Figure
2. The load can be moved along the width and span of the slab and can be placed
both close to support 1 (sup 1 in Figure 2) and support 2 (sup 2 in Figure 2). Loading
plates of 200mm x 200mm and 300mm x 300mm are used to simulate a wheel load.
The supports are line supports and consist of a steel beam of 300mm wide, a layer of
plywood eccentric from the center line of the steel, away from the span, and a layer
of felt of 100mm wide. Support 1 represents a simple support. Prestressing bars are
used to restrain the rotation at support 2, simulating a continuous support. The
prestressing force is applied before the start of every test. When the load increases
during testing (ΔP), the rotation over support 2 increases (Δθ). The prestressing bars

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 154

counteract the increase in rotation, which leads to an increase in the force in the
prestressing bars (ΔFps). The mechanics of the increasing force in the prestressing
bar during testing is shown in Figure 3. In reality, the settlement of the support (felt
and plywood) and the elongation of the prestressing bars result in Δθ ≠ 0º.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

Figure 2. Sketch of test setup, (a) top view, (b) side view, dimensions in mm.

Figure 3. Increase in force in prestressing bar during testing.

Specimens. An overview of the properties of the eight tested slabs can be found in
Table 1, in which:
fc’ the cube compressive strength of the concrete at the age of testing the slab,
fct the splitting tensile strength of the concrete at the age of testing the slab,
ρl the amount of flexural reinforcement in the span direction,
ρt the amount of flexural reinforcement in the transverse direction,
a/d the shear span to depth ratio,

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 155

a the distance between the center of the load and the center of the support,
d the effective depth,
M/S the position of the load along the width; middle (M) or side (S),
Aload the size of the square loading plate.

Table 1. Properties of slabs S1 to S8.

Slab fc’ fct ρl ρt a/d M/S Aload
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

(MPa) (MPa) (%) (%) (mm x mm)

S1 35,8 3,1 0,996 0,132 2,26 M 200 x 200
S2 34,5 2,9 0,996 0,132 2,26 M 300 x 300
S3 51,6 4,1 0,996 0,258 2,26 M 300 x 300
S4 51,7 4,2 0,996 0,182 2,26 S 300 x 300
S5 48,2 3,8 0,996 0,258 1,51 M 300 x 300
S6 50,6 3,9 0,996 0,258 1,51 S 300 x 300
S7 82,1 6,2 0,996 0,258 2,26 S 300 x 300
S8 77,0 6,0 0,996 0,258 2,26 M 300 x 300

On every slab six tests were carried out, except for S3 (five tests), S5 (five tests) and
S8 (three tests). M denotes initial loading at the middle of the slab width. S denotes
initial loading at the side at 438mm from the free edge, Figure 2. In case of M-
loading, subsequent S-loading is carried out on a fully cracked and locally failed slab
which indicates the residual capacity for S-loading. In case of S-loading, subsequent
M-loading is carried out on a fully cracked and locally failed slab which indicates the
residual capacity for M-loading.
The maximum aggregate size was 16mm. Slabs S1 to S6 contained blast
furnace cement and S7 and S8 contained Portland cement, blast furnace cement and
fly ash. During casting, two slabs were cast simultaneously. As a result, there are
always two slabs (for example S1 and S2) which were made out of the same concrete
mixture and thus have the same concrete properties. Two types of concrete were
used: normal strength concrete with a target value for fc’ of 43MPa (C28/35) for
slabs S1 – S6 and high strength concrete with a target value for fc’ of 73MPa
(C53/65) for slabs S7 and S8.
The layout of the reinforcement of S1 and S2 is shown in Figure 4(a) and
Figure 4(b). The layout of S3 and S5 – S8 is shown in Figure 4(c) and the layout of
S4 is shown in Figure 4(d). Reinforcement S500 with an average yield stress of
580MPa and ultimate stress of 710MPa is used. The amount of flexural
reinforcement ρl was determined as necessary to resist a bending moment created by
a load of 2MN (maximum capacity of the jack) at position M along the width and
600mm along the span (a/d = 2,26). The amount of transverse flexural reinforcement
ρt is traditionally taken as 20% of ρl. In this series of tests 13,3% of ρl is used in S1
and S2 and 25,9% of ρl is used in S3, S5, S6, S7 and S8. In S4 the amount of
transverse flexural reinforcement is only doubled in the vicinity of the supports,
Figure 4(d).
The influence of the support is important for a/d-values of 2,5 and less (Kani,
1964). Slabs S1 – S4, S7 and S8 were loaded at 600mm from the center of the

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 156

support, resulting in an a/d-ratio of 2,26. Slabs S5 and S6 were loaded at 400mm

from the center of the support, resulting in an a/d-ratio of 1,51.
The different sizes of the loading plate were chosen to simulate different sizes
of tires. The 200mm x 200mm load is a 1:2 scale representation of the 400mm x
400mm axle load used in load model 2 of EN1991-2:2002.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

Figure 4. Reinforcement layout: (a) plan view of slabs S1 and S2, (b) cross-
section of slabs S1 and S2, (c) cross-section of slabs S3, S5-S8, (d) cross-section
of slab S4.

Test results. The values of the peak loads obtained for S5 to S8 are summarized in
Table 2, in which the following abbreviations are used:
br the distance between the load and the free edge taken along the width,
SS the load is placed close to the simple support,
CS the load is placed close to the continuous support,
Pu the peak load,
s the test gives a shear capacity, obtained on a virgin specimen,
r the test gives a residual shear capacity, obtained on a predamaged specimen.

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 157

The results of slabs S1 up to S4 can be found in Lantsoght et al. (2010a).

Table 2. Results of tests of S5 to S8.

Name a/d br (mm) SS/CS Pu (kN) s/r
S5T1 1,51 1250 CS 1804 s
S5T2 1,51 438 CS 1395 r
S5T4 1,51 1250 SS 1755 s
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

S5T5 1,51 438 SS 1295 r

S5T6 1,51 438 SS 1286 r
S6T1 1,51 438 CS 1446 s
S6T2 1,51 438 CS 1423 s
S6T3 1,51 1250 CS 1897 r
S6T4 1,51 438 SS 1366 s
S6T5 1,51 438 SS 1347 s
S6T6 1,51 1250 SS 1384 r
S7T1 2,26 438 SS 1121 s
S7T2 2,26 438 CS 1172 s
S7T3 2,26 438 CS 1136 s
S7T4 2,26 1250 CS 1128 r
S7T5 2,26 438 SS 1063 s
S7T6 2,26 1250 SS 1011 r
S8T1 2,26 1250 SS 1481 s
S8T2 2,26 1250 CS 1356 s
S8T5 2,26 438 SS 868 r

Figure 5 shows how the peak load is determined from the measured load-deflection
diagram during testing. Typical brittle shear failures (Figure 5(a)) are observed as
well as shear failures with a limited amount of post-peak ductile behavior (Figure
5(b)).

Figure 5: Determination of the peak load (Pu) for (a) S5T4, brittle failure and (b)
S8T2, with post-peak ductility.

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 158

Figure 6 shows the loading sequence and the meaning of s and r. First, test S5T1 is
carried out with the load close to the continuous support and results in a virgin shear
capacity denoted by “s”. This test causes local failure and heavy damage to the slab
at the side of the continuous support. An example of damage caused by testing the
shear capacity is shown in Figure 7. Next, S5T2 is carried out on the damaged slab.
The result of this experiment is influenced by the cracking and failure caused by
S5T1 and therefore gives a residual shear capacity, denoted r. Bridges which might
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

have been in service for several decades, do not have locally failed areas like the
specimens tested for residual capacity. As a result, the measured residual shear
capacity is a conservative lower bound of the capacity of existing bridges that are
fully cracked in bending and might have been overloaded in the past.

Figure 6: Explanation of s and r.

Figure 7. Cracking caused by S2T1, bottom face.

RESULTS AND DISCUSSION

Loading history. The residual shear capacity gives a lower bound estimate of the
shear capacity of bridge slabs which are fully cracked in bending and have been in
service for several decades. Therefore experiments were carried out slabs locally

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 159

failed due to previous testing, with results denoted by “r” in Table 2, loading
sequence shown in Figure 6. To study the influence of predamaging, the results of
tests on S3 and S4, on S5 and S6 and on S7 and S8 are compared. For these pairs of
slabs, all parameters were held constant, only the load history was altered. It was
observed that the width of cracks due to previous testing increased during testing for
residual capacity. In only one case the test gave a 5% higher residual capacity, as
shown in Table 2 for S5T1 and S6T3 (tests in the middle of the width at the
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

continuous support). All other comparisons gave lower residual capacities, as

expected. The results per pair of slabs are given in Table 3, in which:
Pr the capacity obtained in an experiment for residual shear capacity, denoted r
in Table 2,
Ps the capacity obtained in a shear capacity experiment on an undamaged slab,
denoted s in Table 2.
The overall average is a residual capacity of 84% of the undamaged shear strength.
This result is surprisingly high, as it was not expected that slabs which had been
tested up to their ultimate capacity and showed cracks of sometimes a few
centimeters wide (Figure 7) would be able to resist again considerable levels of
loading.
Table 3. Residual capacity.
Slabs Pr/Ps
S3 – S4 0,80
S5 – S6 0,95
S7 – S8 0,78

Shear span to depth ratio. The shear span to depth ratio is known as an important
parameter influencing the shear capacity of reinforced concrete beams. This
influence is traditionally shown in Kani’s valley (Kani, 1964). The influence of the
shear span to depth ratio on reinforced concrete slabs subjected to point loads is less
clear. A 45º load spreading as shown in Figure 1 leads to a decreasing effective width
for a decreasing distance to the support. The theoretical ultimate shear stress vu
remains the same, but the smaller effective width would lead to a smaller maximum
shear load Vu = vubeffd. Therefore, the shear span to depth ratio (a/d) was studied in
this series of experiments.
The slabs S3 and S5 are similar with regard to their properties. They are both
loaded at the middle (M), but for different values of a/d (2,26 and 1,51 respectively).
Also, the slabs S4 and S6 are similar. They are both loaded at the side (S) but for
different values of a/d (2,26 and 1,51 respectively). Only test results obtained on
virgin specimens, denoted by s in Table 2, are considered. Figure 8 shows the results
as a function of a/d. It is clear that in all cases the shear capacity increases with
decreasing value of a/d.
In EN1992-1-1:2005 the shear span to depth ratio is incorporated for small
shear spans. §6.2.2 (6) prescribes: “For members with loads applied on the upper
side within a distance 0,5d ≤ av ≤ 2d from the edge of a support (or center of bearing
where flexible bearings are used), the contribution of this load to the shear force VEd
may be multiplied by β = av/2d. … This is only valid provided that the longitudinal
reinforcement is fully anchored at the support. For av ≤ 0,5d the value av = 0,5d

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 160

should be used.” For slabs S3 and S4 β = 1 is obtained and for S5 and S6 β = 0,66. A
51% increase in shear capacity is expected using this procedure. Comparing the test
results shows an average increase in capacity of 26% with a coefficient of variation
of 6%. The observed increase in shear capacity for slabs is thus less than expected on
the basis of the factor β from the EN 1992-1-1:2005. This factor was derived based
on results from shear tests on beams.
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

Figure 8. Influence of distance of load to support on shear capacity for normal

strength concrete.

A possible reason for the influence of the shear span to depth ratio is related to the
formation of a direct compression strut between the load and the support. When the
load is applied close to the support, an internal compression strut can develop
between the load and the support. Hence, another mechanism of shear transfer is
activated: the strut-and-tie action, in which the beam transfers the force between the
load and the support through a direct compression strut.
When applying this concept to the shear capacity of slabs under concentrated
loads, a lower increase in capacity for decreasing a/d could be explained. The width
of the element influences the possibility of developing a direct compression strut
between the load and the support. A three-dimensional strut-and-tie model can now
develop. In beams, the only possible load path is directly from the load to the support
at small a/d ratios. In slabs, different load paths can develop. As a result of these
different load paths, the average value of a/d will be larger, Figure 9.

Figure 9: Larger average a/d ratio for slabs as compared to beams.

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 161

Concrete compressive strength. Traditionally, the concrete tensile strength is seen

as the most important factor influencing the shear strength of beams. For design
purposes, the tensile strength is calculated based on the concrete compressive
strength with a square root (for example, ACI 318-08) or cube root (for example EN
1992-1-1:2005) relationship. Design methods for the punching shear resistance of
concrete slabs also involve the concrete compressive strength.
To experimentally study the influence of the concrete compressive strength
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

on the one-way shear capacity of slabs, the results of S2, S3, S4, S7 and S8 are
compared. S2, S3 and S4 are produced with normal strength concrete, while S7 and
S8 are produced with high strength concrete. The results are shown in Figure 10. All
tests were carried out at a/d = 2,26. S2 contained less transverse flexural
reinforcement (ρt in Table 1), but it has been previously shown that the influence of ρt
on the ultimate load is small (Lantsoght et al., 2010a). Only the results of the
undamaged slabs (denoted s in Table 2) are used. No increase in capacity with
increasing compressive strength is observed. An increase in scatter is observed for
the test results with high strength concrete. The measured average increase in shear
capacity for S3 and S4 versus S7 and S8 is 0,46% for an increase in concrete
compressive strength of 54%. The expected increase in capacity is 24% for a square
root relationship between the concrete compressive strength and the tensile strength
and 15% for a cube root relationship. The experimental data do not reflect this
observation. The splitting tensile strength increased even with 47%, as shown in
Table 1. Comparing S2, S3/S4 and S7/S8 leads to the similar conclusions.

1600

1400

1200

1000
Pu (kN)

800

600
S3
400 S4
S7
200 S8
S2
0
0 10 20 30 40 50 60 70 80 90
fc' (MPa)
Figure 10. Influence of the concrete compressive strength (cube strength), for
a/d = 2,26, for loading at the middle (upper series) and loading at the side (lower
series).

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 162

The concrete compressive strength influences the shear capacity of beams because it
directly influences the concrete tensile strength, the dowel capacity limited to the
tensile strength of the concrete cover supporting the dowel, the aggregate interlock
capacity and the strength of the compression zone. However, for high strength
concrete the aggregate interlock capacity is reduced. The aggregate itself is then the
weakest element through which the shear crack passes. The result is a very smooth
crack surface. It was observed in S7 and S8 that the shear crack indeed cut through
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

the aggregates. (Sherwood et al., 2007) suggest that 70% of the shear capacity is
carried by aggregate interlock. The reduced aggregate interlock capacity of the high
strength concrete might explain why no increase in shear capacity is observed for an
increasing concrete strength. However, the relation between high strength concrete
and their aggregate interlock capacity is still a point of discussion. The same
observation has been made by (Angelakos et al., 2001). They reported: “high
cylinder strengths do not necessarily result in high values of failure shear stress”.
Another possible explanation is that the mechanism of shear transfer works in
a different way for slabs under point loads than for beams in shear. Due to the width
of the element, redistribution of forces can occur. Therefore, a three-dimensional
strut-and-tie model should be developed in order to explain the observations.

CONCLUSIONS

To evaluate the capacity of existing reinforced concrete slab bridges a series of large-
scale experiments on reinforced concrete slabs have been carried out. Test results
concerning the loading history, the shear span to depth ratio and the concrete
compressive strength are discussed. The results and conclusions are summarized in
the following:
1. Heavily cracked and locally failed slabs still have about 84% of the shear
capacity of an undamaged slab when loaded again in the vicinity of an already failed
region.
2. The shear capacity clearly increases for a decrease in distance to the support, but
the increase is smaller than expected based on the Eurocode formula for beam shear.
3. No increase in capacity with increasing concrete compressive strength was
observed. The smaller aggregate interlock capacity in high strength concrete might
explain this.
The experimental results show that the mechanism of shear transfer in slabs under
wheel loads is different from the mechanism in beams. Redistribution of forces along
the width might result in a three-dimensional system of forces which needs to be
evaluated through a three-dimensional strut-and-tie model.
Future work includes:
• experiments on slabs with a different width,
• experiments on slabs reinforced with plain bars, and
• experiments on slabs supported by elastomeric bearings.

Structures Congress 2011

 Structures Congress 2011 © ASCE 2011 163

ACKNOWLEDGEMENTS

The funding for this research provided by the Dutch Ministry of Transport, Public
Works and Water Management (Rijkswaterstaat) is gratefully acknowledged.

REFERENCES
Downloaded from ascelibrary.org by UNIVERSITE LAVAL on 06/06/14. Copyright ASCE. For personal use only; all rights reserved.

ACI Committee 318, (2008). Building Code Requirements for Structural Concrete
(ACI 318-08) and Commentary, American Concrete Institute, Farmington
Hills, MI, USA.
Angelakos, D., Bentz, E. C., and Collins, M. P. (2001). "Effect of concrete strength
and minimum stirrups on shear strength of large members." ACI Structural
Journal, 98(3), 290-300.
CEN, (2002). Eurocode 1 – Actions on Structures - Part 2: Traffic loads on bridges,
EN 1991-2, Comité Européen de Normalisation, Brussels, Belgium.
CEN, (2005). Eurocode 2 – Design of Concrete Structures: Part 1-1 General Rules
and Rules for Buildings, EN 1992-1-1, Comité Européen de Normalisation,
Brussels, Belgium.
Kani, G. N. J. (1964). "The Riddle of Shear Failure and Its Solution." ACI Journal
Proceedings, 61(4), 441-467.
Lantsoght, E. O. L., van der Veen, C., and Walraven, J. (2010a). "Experimental study
of reinforced concrete bridge decks under concentrated loads near to
supports." Proceedings of the 13th International Conference and Exhibition
on Structural Faults and Repair, Edinburgh, UK.
Lantsoght, E. O. L., van der Veen, C., and Walraven, J. (2010b). "Shear tests of
reinforced concrete slabs with concentrated loads near to supports."
Proceedings of the 8th fib international PhD Symposium in Civil
Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark, 81-
86.
Regan, P. E. (1982). Shear Resistance of Concrete Slabs at Concentrated Loads
close to Supports. Polytechnic of Central London.
Sherwood, E. G., Bentz, E. C., Collins, M. P., (2007). "Effect of aggregate size on
beam-shear strength of thick slabs." ACI Structural Journal, 104(2), 180-190.

Structures Congress 2011