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

net/publication/236591221

Shear capacity of slabs and slab strips loaded close to the support

Article · January 2012

CITATIONS READS

8 2,896

3 authors, including:

Eva Lantsoght Cor van der Veen

Universidad San Francisco de Quito (USFQ) Delft University of Technology
108 PUBLICATIONS   362 CITATIONS    91 PUBLICATIONS   387 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Assessment of concrete bridges View project

Minor Bend and Break | CiTG | TU Delft View project

All content following this page was uploaded by Eva Lantsoght on 27 May 2014.

The user has requested enhancement of the downloaded file.

 SHEAR CAPACITY OF SLABS AND SLAB STRIPS LOADED CLOSE TO THE SUPPORT
Eva O. L. Lantsoght, Cor van der Veen and Joost C. Walraven

Synopsis: In reinforced concrete one-way slabs, two limit states related to shear need to be checked: beam shear
over an effective width at the support and punching shear on a perimeter around the load. Current code
provisions are based on shear tests on heavily reinforced slender beams under point loads. The question remains
if these procedures are valid for wide beams and slabs under point loads close to the support. To evaluate the
shear capacity of reinforced concrete slabs and the associated effective width, a series of experiments is carried
out on eight continuous one-way slabs and twelve continuous slab strips loaded close to the simple and
continuous supports. Test results are compared to current code provisions and methods to calculate the shear
capacity from the literature. The influence of the shear span to depth ratio, the size of the loading plate and the
overall width of the specimen are discussed. From these results follows that the behavior in shear of slabs and
beams is not identical. The effective slab width, used for calculating the beam shear capacity, is recommended
to be based on load spreading under 45° from the far side of the loading plate towards the support.

Keywords: effective width; experiments; one-way slabs; punching; shear.

1
Eva O. L. Lantsoght is a Ph.D. candidate at Delft University of Technology, Delft, The Netherlands. She
received the Degree of Engineer from Vrije University Brussel, Brussels, Belgium; and M.S. from the Georgia
Institute of Technology, Atlanta, Ga.

Cor van der Veen is an associate professor at Delft University of Technology, Delft, The Netherlands. He
received his M.Sc. and Ph.D. from Delft University of Technology. He is a member of various National
Committees. His research interests include (very) high strength (steel fiber) concrete, concrete bridges and
computational mechanics.

Joost C. Walraven is a full professor at Delft University of Technology, Delft, The Netherlands. He received
his M.Sc. and Ph.D. from Delft University of Technology, The Netherlands. He was convener of the Project
Team for Eurocode 2 “Concrete Structures”. He is the chairman of fib Special Activity Group 5 “New Model
Code for Concrete Structures”.

INTRODUCTION
In the Netherlands, 60% of all bridges have been built before 1975. These existing structures have
deteriorated over time, while traffic loads and volumes are continuously increasing. After the Laval bridge
collapse in Quebec (Wood 2008) and the observation of shear cracks in the webs of a prestressed box-girder
bridge, concerns regarding the safety of the existing bridges rose. The Dutch Ministry of Transport, Public
Works and Water Management (Rijkswaterstaat) decided to reexamine the existing structures. For an optimal
assessment of these structures, a better estimate of the real bearing capacity (all failure modes) is needed.
Calculations of design offices based on the current code provisions showed that the loading exceeds a multiple
of the shear capacity, while during site checks no signs of distress were found (Walraven 2010). For solid slab
bridges, the loading case in which concentrated loads are placed close to the supports was found to be governing
for the shear capacity. Due to the larger member width and the transverse reinforcement of slabs as compared to
beams for which the one-way shear criteria were developed, a larger shear capacity could be expected for slabs.
In this paper, the shear capacity of one-way slabs under concentrated loads close to the support is discussed.
Shear in reinforced concrete one-way slabs loaded with a concentrated load near the support is typically
checked in two ways: by calculating the beam shear capacity over a certain effective width of the support and by
checking the punching shear capacity on a perimeter around the load. The beam shear (one-way shear) capacity
formulas have been derived from experiments on beams under a concentrated load. These heavily reinforced
slender beams generally had a total width smaller than their depth so that the web width shows up in the
expression for the shear capacity. However, for slabs with point loads near to the support, the beam shear
capacity should be calculated taking into account a defined effective width beff. This effective width represents
the width of the support which carries the load. The punching shear (two-way shear) capacity in code formulas
was basically developed for two-way slabs. Most empirical methods for punching shear have been derived from
tests on slab areas around a column.
Recent research concerning shear in slabs has mainly focused on one-way slabs under line loads (Sherwood
et al. 2006; Lubell 2006; Sherwood 2008). It was experimentally shown that one-way slabs under line loads
behave like beams and that beam shear provisions led to good estimates of the shear capacity. However, test
data regarding the shear capacity of one-way slabs under concentrated loads representing wheel loads are scarce.
Even less data (36 reported tests) are available regarding the shear strength of one-way slabs loaded close to the
support (shear span to depth ratio ≤ 2.5), Table 1, in which the following symbols and abbreviations are used:
b specimen width,
d effective depth,
a center to center distance between load and support,
bload width of the loading plate,
lload length of the loading plate,
ρl flexural reinforcement ratio,
fc,cyl cylinder concrete compressive strength: for references which include the cube compressive strength,
the cylinder compressive strength is calculated as 85% of the cube compressive strength,
Vu maximum load reached during test,
FM reported failure mode,
SS loading near to simple support,
CS loading near to continuous support,
C combined failure mode of punching and wide beam shear,
P punching failure,
WB wide beam shear failure.
The data in Table 1 show that the majority of the cited experiments have been carried out on small-scale
specimens. No results for slabs with d > 160mm (6.3 in.) are available.

2
 Table 1 — Overview of test data regarding slabs in shear under concentrated loads close to the support
(a/d ≤ 2.5)
Reference Nr. b d a/d bload × lload ρl fc,cyl Vu FM
(m) (mm) (mm × mm) (%) (MPa) (kN)
Regan 1982 2SS 1.2 83.5 2.16 100 × 100 0.6 23.0 130 P
2CS 180 P
3SS 1.68 30.1 195 P
3CS 250 WB
4SS 1.44 35.1 230 P
5SS 2.16 200 × 100 30.3 190 P
7SS 1.68 36.7 200 P
7CS 2.16 230 P
Furuuchi et al. 1998 A-10-10 0.5 160 1.75 100 × 50 2.23 26.1 294 C
A-10-20 20.2 294 WB
A-10-30 23.8 333 WB
A-20-10 200 × 50 19.6 340 -
A-30-10 300 × 50 23.8 450 -
B-10-10 0.65 100 × 50 2.29 29.4 368 -
C-10-10 0.5 160 1.25 100 × 50 2.23 34.6 480 WB
C-20-10 200 × 50 32.1 525 WB
C-30-10 300 × 50 31.5 626 WB
C-50-10 500 × 50 34.9 811 WB
C-10-20 100 × 50 36.4 483 -
C-10-30 100 × 50 30.7 520 -
D-10-10 2.25 100 × 50 35.2 294 -
Graf 1933 1243 a1 2 115 1.30 100 × 150 0.65 19.1 314 WB
1243 a2 2.17 235 C
1243 b1 0.65 355 P
1243 b2 1.52 206 WB
1244 a1 104 1.92 1.14 13.3 275 WB
1244 a2 2.40 196 WB
1244 b1 1.68 157 WB
1244 b2 2.16 147 WB
1245 a1 2.4 106 1.89 1.52 23.6 333 C
1245 a2 2.36 257 WB
1245 b1 1.65 196 C
1245 b2 2.12 206 C
Richart & Kluge 1939 2-2 6.1 140 1.64 150 (disc) 0.91 29.1 369 C
Leonhardt & Walther 1962 P12 0.5 142 2.46 80 × 80 0.95 12.6 101† WB
Ekeberg et al. 1982 2nd fl nr. 3 5 108 2.18 100 × 100 0.52 17.8 465 -
Note: 1m = 3.3ft, 1mm = 0.04 in, 1MPa = 0.145 ksi, 1kN = 0.225 kip.
-: Photographs or a description of the failure mode were not provided.
†: self weight is reported to be included in the value of the peak load.

Regan (1982) suggested that slabs under point loads close to the support have a higher shear capacity than
expected based on beam shear formulas. To investigate this claim and quantify the real bearing capacity of solid
slab bridges, an extensive experimental program is carried out at Delft University of Technology, The
Netherlands. The research is limited to non-prestressed slabs. The variables in the test program are:
- the shear span to depth ratio (a/d),
- the reinforcement layout,
- the concrete compressive strength,
- the position of the load along the width,
- the size of the loading plate,
- the type of support (simple or continuous),
- the overall width of the specimen,
- the type of reinforcement (ribbed or plain bars),
- the support conditions (line support or bearings).

3
 RESEARCH SIGNIFICANCE
Only a small number of results from shear tests on one-way slabs subjected to concentrated loads close to
the support is available. The experimental data presented in this paper extend the test data and knowledge on
slabs under concentrated loads failing in shear. These data allow for a comparison with the test data of beams
failing in shear. Indications are found that shear behavior in one-way slabs under concentrated loads is different
from that in beams. The outcome of this research will be used to assess existing slab bridges in the Netherlands.

ONE-WAY SHEAR PROVISIONS

The one-way shear capacity of slabs is calculated using the beam shear formulas over a certain effective
width (beff). The method of load spreading, resulting in the effective width, depends on local practice. In most
cases, such as in Dutch practice, load spreading is assumed under a 45° angle from the center of the load
towards the support, Figure 1(a). The lower limit for the effective width is taken in Dutch practice as 2d for
loads in the middle of the width and d for loads at the edge and corner of the slab. In French practice, load
spreading is assumed under a 45° angle from the far corners of the loading plate towards the support, Figure
1(b).

support support
beff beff

load load
(a) (b)
Figure 1 — Effective width (a) assuming 45º load spreading from the center of the load; (b) assuming 45º
load spreading from the far corners of the load; top view showing concentrated load and support.

According to EN1992-1-1:2005 §6.2.2 (1) the maximum shear force for a section without stirrups is
calculated as follows (SI units, k1 = 0.15):
VRd , c = CRd , c k (100 ρ l fck ) + k1σ cp  bw d ≥ (vmin + k1σ cp )bw d
1/ 3
(1)

200
k = 1+ ≤ 2.0 (2)
d
with
fck the characteristic concrete cylinder compressive strength in MPa (1MPa = 0.145ksi),
d the effective depth in mm (1 mm = 0.04in.),
σcp the average normal concrete stress over the cross section, positive in compression,
bw the smallest width of the cross section in the tensile area, not to exceed beff.
Eq. 1 applies for members with or without prestressing.
The values of CRd,c and vmin depend on the National Annex. It is recommended to take CRd,c = 0.18/γc where γc is
the partial factor for concrete (γc=1.5 in general) and vmin:
vmin = 0.035k 3 / 2 f ck1/ 2 (3)
In the French National Annex (Chauvel et al. 2007) a different approach is used for vmin. For slabs benefiting
from transverse redistribution under the load case considered vmin is defined as:
ν min = 0.34 f ck 1/ 2 (4)
and for beams and slabs, other than those described by eq. 4, the expression becomes
vmin = 0.053k 3 / 2 f ck1/ 2 (5)
Loads applied within a distance 0.5d ≤ av ≤ 2d from the edge of a support can be reduced by β = av/2d, where av
is the clear shear span from the face of the load to the face of the support.
According to ACI 318-08 the inclined cracking load is described by formula (11-5) from §11.2.2.1. The
sectional shear force is not allowed to exceed the inclined cracking load for elements without web reinforcement.
For normal weight concrete (λ = 1) and with notations altered to facilitate comparison with the previously used
SI notations, the expression becomes:
 V d
Vc =  0.16 f ck + 17 ρl ACI  bw d ≤ 0.29 f ck bw d (6)
 M ACI 
with
fck the characteristic concrete cylinder compressive strength in MPa (1MPa = 0.145ksi),
ρl the flexural reinforcement ratio,

4
d the effective depth,
bw the width of the web, not to exceed beff.
VACI factored shear force at a section,
MACI factored moment at a section.
Regan’s method (1982) was developed especially for slabs under point loads close to the support (av ≤ 2d;
where av is the clear shear span from the face of the load to the face of the support) based on the punching
provisions from the British Code of Practice CP110 (Technical Committee CSB/39 1982) and the results of
small-scale experiments (reported in Table 1). The method is based on the definition of a critical perimeter
around the concentrated load, Figure 2. For loads close to the free edge the method is extended based on the
principles used when determining the punching perimeter, Figure 2(c, d). As shown in Figure 2 the support is
subdivided into different parts of the perimeter u1 and u2. Likewise, the contributions to the capacity of u1 and u2
(influenced by the nearby support) are accounted for differently. The resistance of the part of the perimeter
parallel to the support u2 (Figure 2) is defined as:
 2d  fcu
PR 2 =   ξ s vc u2 d < u2 d (7)
 av  γm
500
ξs = 4 (8)
d
0.27
vc = 3
100 ρ f cu (9)
γm
in which
fcu the cube concrete compressive strength, in MPa (1 MPa = 0.145 ksi),
γm the partial safety factor for materials.
In eq. (8), d needs to be taken in mm (1mm = 0.04in). The resistance of the remainder (Σu = u1) of the perimeter
(Figure 2) is defined as:
PR1 = ∑ ξ s vc ud (10)
For the sections parallel to the support, the longitudinal properties are to be used (dl and ρl of the longitudinal
reinforcement), similar to a calculation for beam shear. For the sections perpendicular to the support, the
transverse properties are to be used (dt and ρt of the transverse flexural reinforcement), as indicated in Figure 2.
The contributions of PR2 from eq. (7) and PR1 from eq. (10) are then summed to give the total resistance against
shear failure. At a continuous support, the total shear resistance is multiplied with a factor depending on M1
(larger moment at the end of the shear span) and M2 (smaller moment), with M1 and M2 as absolute values.
M1 + M 2
α= (11)
M1
av av
1.5dl lload 1.5dl 1.5dl lload 1.5dl
1.5dt

1.5dt
support

support

0.5bload
b

load load
load

subdivision
of perimeter
1.5dt

br

u1
(a)
(c) u2
av lload 1.5d
l av lload 1.5d properties
l
dl ρl
1.5dt b

1.5dt

dt ρt
support

support

0.5b

load
load

load
load
1.5dt

br

(b)
(d)
Figure 2 — Subdivision of perimeter and slab properties to be used for parts of the perimeter: for a load
in the middle of the width of the slab (a) for 2dl > av >1.5dl, (b) for av < 1.5dl and for a load close to the edge
of the slab (c) for 2dl > av >1.5dl, (d) for av < 1.5dl, based on (Regan 1982).

5
 RESEARCH CARRIED OUT AT DELFT UNIVERSITY OF TECHNOLOGY
Test setup
A top view of the test setup with a slab is presented in Figure 3. The line supports (sup 1 and sup 2 in
Figure 3) are composed of a steel beam (HEM 300) of 300mm (11.81 in.) wide, a layer of plywood and a layer
of felt of 100mm (3.94 in.) wide. Experiments are carried out close to the simple support (sup 1 in Figure 3) and
close to the continuous support (sup 2 in Figure 3). The rotation at support 2 is restrained by vertical
prestressing bars which are fixed to the strong floor of the laboratory. This restraint results in a moment over
support 2 which hence behaves as a continuous support. The prestressing force is applied on the bars before the
start of every test, offsetting the self weight of the slab. During the course of the experiment, some rotation
could occur over support 2 due to the deformation of the felt and plywood and the elongation of the prestressing
bars.

simple continuous
support support

300 mm 300 mm
11.81in 11.81in

sup 1 sup 2

prestressing bars

2500mm
sup 1 load sup 2

8.2ft
1250mm (M)
4.10ft

load
438mm (S)
1.44ft

sup 1 sup 2

300mm 3600mm 600mm 500mm

11.81in 11.81ft 23.62in 19.69in
Figure 3 — Sketch of test setup, top view.

Specimens
The main properties of the eight slabs and twelve slab strips are given in Table 2, in which the following
symbols are used:
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,
a the center-to-center distance between the load and the support,
M loading at the middle of the slab width,
S loading at the side at 438 mm (1.44 ft.) from the free edge, Figure 3.
All slabs and slab strips had a thickness of 300 mm (11.81 in) and an effective depth d of 265mm (10.43 in).
The numbering for the slabs starts with “S”, while for the beams (“B”) the numbering is subdivided
according to the size: S, M, L, X. The slabs were either loaded at the middle of the slab width (position M) at the
simple and continuous support, resulting in two tests per slab, or consecutively at the east and west side
(position S) at the simple and continuous support, resulting in four tests per slab.
Ribbed reinforcing bars with a diameter of 10mm (0.4 in, cfr. #3 bars) (measured yield strength fy = 635
MPa = 92.1 ksi and measured ultimate strength fu = 710 MPa = 102 ksi) and 20mm (0.8in, cfr. #6 bars) (fy = 601
MPa = 87.2 ksi and fu = 647 MPa = 93.8 ksi) were used. The flexural reinforcement was designed to resist a
moment caused by a load of 2 MN (450 kip, maximum capacity of the jack) at position M (Figure 3) along the
width and 600 mm (23.62 in) along the span (a/d = 2.26). In practice, the amount of transverse flexural
reinforcement for slabs is taken as 20% of the longitudinal flexural reinforcement. In the slabs discussed, 13.3%
of the longitudinal flexural reinforcement was used in S1 and S2 and 25.9% in S3, S5, S6, S8, S9 and the slab

6
strips. In S4 the amount of transverse flexural reinforcement was only doubled as compared to S1 and S2 in the
vicinity of the supports. Figure 4 shows elevation, cross-section and detailing of the reinforcement in S1 and S2.

(a)

(b)

(c)
Figure 4 — Reinforcement layout of slabs: (a) plan view of S1 and S2, (b) section of S1 and S2, (c) Detail
of top reinforcement. All slabs had similar longitudinal reinforcement. Units: mm (1 mm = 0.04 in.)

Two types of concrete have been used: normal strength concrete with a target cylinder strength fc,cyl of
43MPa (6.2 ksi) for slabs S1 – S6 and high strength concrete with a target strength fc,cyl of 73MPa (10.6 ksi) for
slabs S8, S9 and the slab strips. Glacial river aggregate with a maximum aggregate size of 16 mm (0.63 in.) was
used.
Two center-to-center distances between load and support (600mm = 23.6in, a/d = 2.26 for S1 – S4, S8 – S9,
BS1 – BX1, BS3 – BX3 and 400mm = 15.8in, a/d = 1.51 for S5, S6, S9, BS2 – BX2) were used to study the
influence of direct load transfer to the support, which becomes significant for a/d ≤ 2.5 (where a is the center-to-
center distance between the load and the support) and less (Kani 1964).
The 200 mm × 200 mm (7.9 in. × 7.9 in.) load is a 1:2 scale representation of the 400 mm × 400 mm (15.8
in × 15.8 in) axle load used in load model 2 of EN1991-2:2002.

Table 2 — Properties of slabs S1 – S6, S8 – S9 and slab strips BS1 – BX3.

Slab b f c’ fct ρl ρt a/d M/S bload × lload cast date test date
nr. m (MPa) (MPa) (%) (%) (mm × mm) (dd-mm-yy) (dd-mm-yy)
S1 2.5 35.8 3.1 0.996 0.132 2.26 M 200 × 200 08-10-09 05-11-09
S2 2.5 34.5 2.9 0.996 0.132 2.26 M 300 × 300 08-10-09 03-12-09
S3 2.5 51.6 4.1 0.996 0.258 2.26 M 300 × 300 20-11-09 22-01-10
S4 2.5 51.7 4.2 0.996 0.182 2.26 S 300 × 300 20-11-09 04-02-10
S5 2.5 48.2 3.8 0.996 0.258 1.51 M 300 × 300 02-02-10 05-03-10
S6 2.5 50.6 3.9 0.996 0.258 1.51 S 300 × 300 02-02-10 15-03-10
S8 2.5 77.0 6.0 0.996 0.258 2.26 M 300 × 300 23-02-10 12-04-10
S9 2.5 81.7 5.8 0.996 0.258 1.51 M 200 × 200 10-03-10 26-05-10
BS1 0.5 81.5 6.1 0.996 0.258 2.26 M 300 × 300 23-02-10 19-04-10
BM1 1.0 81.5 6.1 0.996 0.258 2.26 M 300 × 300 23-02-10 26-04-10
BL1 1.5 81.5 6.1 0.996 0.258 2.26 M 300 × 300 23-02-10 31-08-10
BS2 0.5 88.6 5.9 0.996 0.258 1.51 M 200 × 200 10-03-10 14-09-10
BM2 1.0 88.6 5.9 0.996 0.258 1.51 M 200 × 200 10-03-10 14-09-10
BL2 1.5 94.8 5.9 0.996 0.258 1.51 M 200 × 200 10-03-10 06-09-10
BS3 0.5 91.0 6.2 0.996 0.258 2.26 M 200 × 200 22-03-10 20-09-10

7
BM3 1.0 91.0 6.2 0.996 0.258 2.26 M 200 × 200 22-03-10 20-09-10
BL3 1.5 81.4 6.2 0.996 0.258 2.26 M 200 × 200 22-03-10 09-09-10
BX1 2.0 81.4 6.0 0.996 0.258 2.26 M 300 × 300 25-11-10 11-01-11
BX2 2.0 70.4 5.8 0.996 0.258 1.51 M 200 × 200 9-12-10 17-01-11
BX3 2.0 78.8 6.0 0.996 0.258 2.26 M 200 × 200 25-11-10 04-01-11
Note: 1m = 3.3ft, 1 MPa = 145 psi, 1 mm = 0.04 in.

RESULTS AND DISCUSSION

Test results
The experimental results of the slabs and slab strips discussed are summarized in Table 3, in which the
following symbols are used:
br the distance between the center of the loading plate and the free edge of the slab along the width,
SS/CS the position of the load: close to the simple support (SS) or the continuous support (CS),
Pu the measured ultimate load during the experiment,
Mu,p the moment at failure over the continuous support caused by the force as measured in the prestressing
bars.
The number of the specimen is followed by the number of the test carried out on this specimen (eg. S6T2: 2nd
test on 6th slab). The failure modes are denoted as: WB (wide beam shear failure with inclined cracks observed
at the bottom face), P (punching shear failure with partial development of a punching cone on the bottom face),
B (beam shear failure with a shear crack at the side face).
Figure 5 shows how the peak load is determined based on the measured load-deflection relation. Typical brittle
shear failures (Figure 5(a)) are observed as well as shear failures with a limited amount of post-peak ductility
(Figure 5(b)). The damage caused by S4T1 and S4T2 extended towards the continuous support, such that only
two tests could be carried out on S4 (loading position S).

Table 3 —Experimental results for slabs and slab strips.

Test a/d br (mm) SS/CS Pu (kN) FM Mu,p (kNm)
S1T1 2.26 1250 SS 954 WB 98
S1T2 2.26 1250 CS 1023 WB 83
S2T1 2.26 1250 SS 1374 WB + P 168
S2T4 2.26 1250 CS 1421 WB 198
S3T1 2.26 1250 SS 1371 WB 151
S3T4 2.26 1250 CS 1337 WB + B 172
S4T1 2.26 438 SS 1160 WB + B 122
S4T2 2.26 438 SS 1110 WB + B 112
S5T1 1.51 1250 CS 1804 WB + B 141
S5T4 1.51 1250 SS 1755 WB + B 168
S6T1 1.51 438 CS 1446 WB + B 110
S6T2 1.51 438 CS 1423 WB + B 128
S6T4 1.51 438 SS 1366 WB + B 117
S6T5 1.51 438 SS 1347 WB + B 147
S8T1 2.26 1250 SS 1481 WB + B 140
S8T2 2.26 1250 CS 1356 WB + B 167
S9T1 1.51 1250 SS 1523 WB + P 105
S9T4 1.51 1250 CS 1842 WB + P 153
BS1T1 2.26 250 SS 290 B 22
BS1T2 2.26 250 CS 623 B 127
BS2T1 1.51 250 SS 633 B 60
BS2T2 1.51 250 CS 976 B 160
BS3T1 2.26 250 SS 356 B 34
BS3T2 2.26 250 CS 449 B 64
BM1T1 2.26 500 CS 923 WB + B 96
BM1T2 2.26 500 SS 720 WB + B 76
BM2T1 1.51 500 SS 1212 WB + B 100

8
 BM2T2 1.51 500 CS 1458 WB + B 157
BM3T1 2.26 500 SS 735 WB + B 66
BM3T2 2.26 500 CS 895 WB + B 110
BL1T1 2.26 750 SS 1034 WB + B 129
BL1T2 2.26 750 CS 1252 WB + B 192
BL2T1 1.51 750 SS 1494 WB + B 127
BL2T2 1.51 750 CS 1708 WB + B 166
BL3T1 2.26 750 SS 1114 WB + B 145
BL3T2 2.26 750 CS 1153 WB + B 187
BX1T1 2.26 1000 SS 1331 WB + P 195
BX1T2 2.26 1000 CS 1596 WB + B + P 201
BX2T1 1.51 1000 SS 1429 WB + B + P 130
BX2T2 1.51 1000 CS 1434 WB + P 100
BX3T1 2.26 1000 SS 1141 WB + P 147
BX3T2 2.26 1000 CS 1193 WB + B 126
Note: 1 mm = 0.04 in., 1 kN = 0.225 kip, 1 kNm = 8.851 kip in.

Pu
Pu 1400
1800

1600 1200

1400
1000
1200
800
F (kN)

1000
F (kN)

800 600

600
400
400
200
200

0 0
(a) 0 5 10 15 20 25 30 (b) 0 5 10 s (mm) 15 20 25
s (mm)
Figure 5 — Experimental determination of the peak load (Pu) based on the load-displacement diagram
for (a) S5T4, brittle failure and (b) S8T2, failure mode with post-peak ductility. Note: 1 mm = 0.04in., 1
kN = 0.225 kip.

Shear span to depth ratio (a/d)

From shear tests on reinforced concrete beams it is known that the shear span to depth ratio (a/d) is an
important parameter influencing the shear capacity and was identified as such in early research on shear in
beams (Talbot 1909; Richart 1927; Clark 1951). Later, it was shown to affect the ratio of maximum moment to
theoretical flexural failure moment MCR/MFL resulting in the so-called shear valley (Kani 1964). For small a/d
distances (a/d ≤ 2.5) direct transfer of the load from its point of application towards the support leads to an
increasing shear capacity for a decreasing distance between the load and the support. A compression strut or fan
of struts between the load and the support develops and the capacity of this strut increases as its inclination
becomes steeper. As previously mentioned, direct load transfer is incorporated into EN 1992-1-1:2005 §6.2.2
(6) for loads applied within a clear distance 0.5d ≤ av ≤ 2d from the edge of a support. The contribution of a
load applied near to the support to the overall shear force VEd can be reduced by multiplication with β = av/2d.
For av ≤ 0.5d a constant β = 0.25 is used. This value is determined from beam shear tests (Regan 1998).
The results from the literature from Table 1 show an influence of a/d on the ultimate load as shown in
Table 4. Table 4 shows the factor β=av/2d calculated based on the results reported in the literature, the reported
ultimate load Pu and the expected ultimate load Pu,exp,β. This expected value was determined based on the
measured Pu on the previous row and the ratio of factors β from the previous and current row in Table 4. The
increase is consequently smaller than the increase which would be expected using the factor β from EN 1992-1-
1:2005. It should be noted that the results from (Graf 1933) of series “a” were omitted from this comparison, as
these tests were executed close to the support over which a moment existed due to applying half of the load on
the small cantilevering end and the full load in the span with a/d as denoted in Table 1.

9
 Table 4 — Increase in capacity for a decreasing distance to the support as reported in literature
compared to factor β from EN 1992-1-1:2005.
Reference Nr a/d β Pu Pu,exp,β
(kN) (kN)
Regan 1982 2SS 2.16 0.72 130
3SS 1.68 0.48 195 195
4SS 1.44 0.24 230 390
Furuuchi et al. 1998 D-10-10 2.25 0.97 294
A-10-10 1.75 0.72 294 396
C-10-10 1.25 0.40 480 529
Graf 1933 1244b2 2.16 0.36 147
1244b1 1.68 0.25 157 212
1245b2 2.12 0.36 206
1245b1 1.65 0.25 196 297
Note: 1kN = 0.225kip.

Applying the concept of an effective width as shown in Figure 1 to a decreasing distance between load and
support results in a smaller associated effective width for a smaller a/d distance. When the maximum shear
stress vu, according to the code, remains the same, a smaller theoretical shear capacity Vu = vubeffd would be
expected. As a result of the additional dimension in slabs as compared to beams, it can be expected that direct
load transfer of a point load towards a line support will take place differently in beams than in slabs. While in a
beam a well-defined compressive strut develops, in a slab a three-dimensional strut-and-tie system is expected
to develop. When considering the plan view of a slab under a concentrated load, Figure 6, the different load
paths between the load and the support, resulting from this three-dimensional force equilibrium, can be
visualized. These different load paths result in a larger average a/d ratio for a slab than for a beam in which the
load is placed at the same position. In slabs under point loads near to the support, a lower increase in shear
capacity can thus be expected for a decreasing a/d than in beams.

support
a/d = 1.2
a/d = √2 a/d = 1
a/d = 1.2
a/d = √2

load
Figure 6 — Larger average a/d ratio for slabs as compared to beams.

To study the influence of the distance between the load and the support (a/d) and its relation to element type
(slab or slab strip/beam) S3, S4, BS3, BM3, BL3 and BX3 were loaded at a/d = 2.26 and S5, S6, BS2, BM2,
BL2 and BX2 at a/d = 1.51. The slabs were loaded with a load of 300mm × 300mm (11.8in. × 11.8in.) and the
slab strips with a load of 200mm × 200mm (7.9in. × 7.9in.). The slabs were normal strength concrete and the
slab strips were high strength concrete. The experimental observations are summarized in Table 5, showing the
measured average increase for a/d decreasing from 2.26 to 1.51. The results in Table 5 show a clear increase in
shear capacity with decreasing distance to the support as well as a clear influence of the overall member width b
on the quantity of this increase. These results can be compared to the expected increase based on the factor β
from EN 1992-1-1:2005. For slabs S3 and S4 β = 0.755 is found and for S5 and S6 β = 0.377. Using this
procedure, the capacity of the slabs would have been doubled while in the experiments a 26% increase was
observed. For the slabs strips BS3, BM3, BL3 and BX3 β =0.849 is found and for BS2, BM2, BL2 and BX2 β =
0.472. The expected increase in capacity is 80%. Comparison of these expected increases with the observed
increases shows the difference in behavior between beams/slab strips with mainly two-dimensional load-
carrying behavior and slabs with mainly three-dimensional load-carrying behavior. The result of the load
spreading as shown in Figure 6 is reflected in the experimental results. The factor β from EN 1992-1-1:2005 is
based on test results from small (effective depths between 250mm = 9.8 in. and 400mm = 15.7in.), heavily
reinforced (ρl between 1% and 3%) rectangular beams (Regan 1998) with mostly cube compressive strengths
between 20 MPa (2.9 ksi) and 40MPa (5.8 ksi). For slabs, more appropriate methods taking into account the
resulting load paths and relating the capacity to the effective width in shear should be developed. Such a
procedure would be according to the philosophy of ACI 318-08, which recommends the use of nonlinear
analysis or strut-and-tie models for members with concentrated loads within a distance twice the member depth

10
from the support. These observations are also reflected by the results from the literature presented in Table 4
which show generally smaller increases than expected based on the factor β from EN 1992-1-1:2005.

Table 5 —Measured increase in capacity for a/d decreasing from 2.26 to 1.51.
Specimens b Average
(m) increase
BS2 – BS3 0.5 98%
BM2 – BM3 1.0 64%
BL2 – BL3 1.5 41%
BX2 – BX3 2.0 23%
S3/S4 – S5/S6 2.5 26%
Note: 1m = 3.3ft.

Size of the loading plate

The size of the tire contact area is not a constant. For design purposes however, the size of the tire contact
area used for road bridges is prescribed in the code provisions as a fixed value. The tire contact area as
prescribed in EN 1991-2:2003 is 400mm × 400mm (15.7 in. × 15.7in.) in load model 1 which covers most of the
effects of trucks and cars and 600mm × 350mm (23.6in. × 13.8in.) in load model 2, a single axle load which
covers the dynamic effects of the normal traffic on short structural members. In AASHTO LRFD 2007 the tire
contact area is 510mm × 250mm (20in × 10in) for design truck and tandem. For other design vehicles, the tire
contact area should be determined by the engineer.
As shown in Figure 1(b), the size of the contact area influences the effective width based on load spreading
from the far side of the loading plate over which theoretically a uniform shear stress vu is assumed. Contrarily,
for the load spreading path from the center of the load towards the support, as shown in Figure 1(a), the size of
the loading plate is not assumed to influence the effective width and the resulting shear capacity of the slab.
It was noted previously (Sherwood et al. 2006), based on limited test data (Sherwood et al. 2006; Serna-Ros
et al. 2002), that load and support points narrower than the specimen width may have a small detrimental effect
on the shear capacity of a one-way spanning member. The reference point for this conclusion was a one-way
slab under line load over the full specimen width, behaving as a beam, and may not be representative for a slab
under a concentrated load. Test results in Table 1 (Furuuchi et al. 1998; Regan 1982) show an increasing shear
capacity for an increasing width of the loading plate. A comparison is shown in Table 6. The increase in loaded
area as compared to the loaded area of the experiment on the previous row and the increase in ultimate capacity
as compared to the ultimate capacity of the experiment on the previous row are shown. It is remarkable that the
increase in ultimate capacity becomes larger for the largest tested loading plates, while the increase in size of
these loading plates is percentage-wise smaller than for the smaller tested loading plates. For smaller a/d
distances, smaller increases in capacity are reported than for loads applied further away from the support. In the
experiments compared in Table 6, the length was kept constant, leading to an increasing degree of
rectangularity. This rectangularity might have resulted in the slabs behaving like a wide beam, not allowing for
transverse redistribution of forces and three-dimensional force flow.

Table 6 — Increase in capacity for an increasing size of the loading plate as reported in literature.
Reference Nr a/d load increase Pu increase
(mm × mm) load size (kN) Pu
Furuuchi et al. A-10-10 1.75 100 × 50 294
1998 A-20-10 200 × 50 100% 340 16%
A-30-10 300 × 50 50% 450 32%
C-10-10 1.25 100 × 50 480
C-20-10 200 × 50 100% 525 9%
C-30-10 300 × 50 50% 626 19%
C-50-10 500 × 50 60% 811 30%
Regan 2SS 2.16 100 × 100 130
1982 5SS 200 × 100 100% 190 46%
3SS 1.68 100 × 100 195
7SS 200 × 100 100% 200 3%
Note: 1mm = 0.4 in., 1kN = 0.225kip.

To study the influence of the size of a square loading plate on the shear capacity of one-way slabs, the
results of S1 and S2 can be compared, as well as BS1 and BS3, BM1 and BM3, BL1 and BL3 and BX1 and
BX3. The slabs were normal strength concrete and the slab strips were high strength concrete. The results of the

11
comparison of the experimental data are shown in Table 7. The measured average increase in capacity for an
increase in size of loading plate from 200mm × 200mm (7.9in. × 7.9in.) to 300mm × 300mm (11.8in. × 11.8in.)
is displayed.

Table 7 —Measured increase in capacity for an increase in the size of the square loading plate from
200mm × 200mm (7.9in. × 7.9in.) to 300mm × 300mm (11.8in. × 11.8in.).
Specimens b (m) Average increase
BS1 – BS3 0.5 10.1%
BM1 – BM3 1.0 0.5%
BL1 – BL3 1.5 0.7%
BX1 – BX3 2.0 25.2%
S1 – S2 2.5 41.5%
Note: 1m = 3.3ft.

The results in Table 7 show an increasing influence of the loading plate size as the overall width of the
specimen increases. Observation of the cracking patterns at the bottom face of the specimen showed large
similarities between the patterns in the BX-series and the S-series. The specimens of smaller width (BS and BM
series) showed a different cracking pattern at the bottom face, consisting mainly of straight cracks running from
one free edge to the other free edge, Figure 7(a). In the wider specimens, a more grid-like pattern was visible,
Figure 7(b). These observations correspond with the concept of transverse load redistribution in slabs. For
members of smaller width, transverse load redistribution does not take place and the load is carried directly from
its point of application to the support. In this case, the size of the loading plate should not influence the capacity
of the member. However, for a wider member, load spreading in the transverse direction becomes more
important. For a larger loading plate, a larger surface is available out of which the load can be carried through
the previously mentioned three-dimensional struts towards the support.

(a) (b)
Figure 7 — Difference in cracking pattern between beam and slab: (a) cracking pattern at bottom face
after BS2T1, (b) cracking pattern at bottom face after S9T1. The dashed lines denote the location of the
loading plate. Thicker lines in (b) denote areas of punching damage.

Overall width and effective width

If the concept of an effective width can be applied to concrete slabs loaded in shear, then the shear capacity
should cease to increase as the width is increased after reaching the effective width. According to this concept,
increasing widths will lead to the same capacity, as only the effective width of the support can carry the load.
Previous research (Regan and Rezai-Jorabi 1988) observed increasing maximum shear capacities for
increasing widths up to a certain value (1m, 3.3ft) for a/d = 5.42 after which the maximum shear capacity
remained around the same value. From the results in Table 1 only A-10-10 and B-10-10 (Furuuchi et al. 1998)
can be compared. Load spreading based on Figure 1(a) results in an effective width of 510mm (20.1in.), limited
to 500mm (19.7in.) for A-10-10 by its physical width. The value of a/d is 1.75. Using the principles from
Figure 1(b) results in an effective width of 660mm (26in.), limited to 500mm (19.7in.) for A-10-10 and to
650mm (25.6in.) for B-10-10 by their physical width. As a result, the expected increase in capacity based on
load spreading as shown in Figure 1(a) is 2% and based on load spreading as shown in Figure 1(b) is 30%. The
measured capacities in Table 1 had an increase in capacity of 25%. Thus, load spreading as shown in Figure
1(b) agrees best with these experimental data and the observed increase in capacity is almost linear with the
increase in width.
Another aspect related to the overall width is the transverse redistribution capacity. According to this
concept, larger widths lead to larger capacities. In wide elements disturbances caused by local weaknesses will
be smoothed out.

12
 To study the influence of the overall width on the shear capacity, the results of S8 and S9 are compared to
the results of the series of slab strips (BS1 – BX3). All considered specimens are composed of high strength
concrete. The results of the comparison of the experimental data are shown in Figure 8. These results show that
the concept of using an effective width for slabs is indeed necessary as the shear capacity does not increase
linearly for larger widths. In Figure 8 the shear capacity increases linearly for the results of BS, BM and BL.
Comparing the test results of BL, BX and S8/S9 shows that the increase in shear capacity with increasing width
for these slab widths is rather small, indicating that the threshold value for effective width for this load
configuration is achieved.

2000

1800

1600

1400

1200
Pu (kN)

1000

800

600 BS
BM
400
BL
S
200
BX
0
0 500 1000 1500 2000 2500

b (mm)

Figure 8 — Influence of overall width on shear capacity. Test results for BS, BM, BL, BX, S8 and S9 are
shown. Conversion: 1 mm = 0.04 in., 1 kN = 0.225 kip.

The results for the calculated effective width based on the experimental results are given in Table 8 and
compared to the calculated widths based on the load spreading methods from Figure 1(a) and Figure 1(b). The
series of results are divided into the following categories:

A loading plate 300mm × 300mm (11.8in. × 11.8in.), simple support, a/d = 2.26,
B loading plate 300mm × 300mm (11.8in. × 11.8in.), continuous support, a/d = 2.26,
C loading plate 200mm × 200mm (7.9in. × 7.9in.), simple support, a/d = 1.51,
D loading plate 200mm × 200mm (7.9in. × 7.9in.), continuous support, a/d = 1.51,
E loading plate 200mm × 200mm (7.9in. × 7.9in.), simple support, a/d = 2.26,
F loading plate 200mm × 200mm (7.9in. × 7.9in.), continuous support, a/d = 2.26.
The symbols in Table 8 are
bmeas effective width as calculated from the series of experiments with different widths;
beff1 effective width calculated as shown in Figure 1(a),
beff2 effective width calculated as shown in Figure 1(b).

Table 8 — Effective width as calculated from the experimental results.

Serie bmeas (m) beff1 (m) beff2 (m) bmeas/beff1 bmeas/beff2
A 2.12 1.10 1.70 1.92 1.24
B 1.81 1.10 1.70 1.65 1.07
C 1.25 0.70 1.10 1.78 1.26
D 1.11 0.70 1.10 1.58 1.12
E 1.63 1.10 1.50 1.48 1.09
F 1.33 1.10 1.50 1.21 0.89
Note: 1m = 3.3ft.

The results from Table 8 show a difference between loading at the simple and continuous support. Consistently,
lower effective widths are found at the continuous support as compared to the simple support. The moment

13
distribution in the shear span is thus observed to influence the load spreading mechanism which takes place in
the slab. The results from Table 8 also show a different effective width depending on the size of the loading
plate. As previously discussed, load spreading from the center of the load towards the support would not imply
an influence of the size of the loading plate on the effective width or the overall shear capacity. The results of
this series of experiments show the influence of the size of the loading plate on the effective width: a larger
loading plate leads to a larger effective width and thus a wider mechanism of load spreading. This observation is
also reflected in the concept of load spreading as shown in Figure 1(b). Moreover, the results from Table 8
show that the effective width becomes smaller as the a/d ratio decreases. This experimental result corresponds
with both load spreading mechanisms shown in Figure 1. Table 8 also contains the values for the effective
width as calculated with the load spreading methods from Figure 1(a) and Figure 1(b). Load spreading from
the center of the load towards the support (Figure 1(a)) underestimates the effective width. Better results are
obtained through the method of load spreading from the far sides of the loading plate (Figure 1(b)).

Comparison to code methods

All test results are compared to the previously discussed code methods: EN 1992-1-1:2005 with
recommended values, EN 1992-1-1: 2005 with values from the French National Annex and load spreading
according to French practice (Figure 1(b)), Regan’s formula (Regan 1982) and ACI 318-08. The comparison is
based on the measured mean material properties and all safety factors are omitted. Calculations according to EN
1992-1-1:2005 and ACI 318-08 for the punching shear capacity were carried out but punching shear was never
found to be the critical failure mode. For comparison with the method from the French National Annex, all slabs
and the BL and BX series were considered as slabs benefitting from transverse redistribution, while the BS and
BM series were not categorized as such. This a posteriori distinction is based on the cracking pattern, as shown
in Figure 7. Elements showing a two-way cracking pattern on the bottom face were categorized as slabs.
Beam shear calculations according to EN 1992-1-1:2005 and ACI 318-08 were combined with the load
spreading method as shown in Figure 1(a). To study the influence of the assumed load spreading mechanism
used to determine the effective width, the values for EN 1992-1-1:2005 were also calculated with the effective
width as shown in Figure 1(b). The previously mentioned lower bounds from Dutch practice of 2d for loading
in the middle of the width and of d for loading close to the edge haven been taken into account. The critical
perimeter as used in Regan’s method is taken with 4 sides for loading in the middle of the width for the slabs
(Figure 2(a) and Figure 2(b)), with 3 sides for loading near to the edge of the width for slabs (Figure 2(c) and
Figure 2(d)) and with 2 sides for the smallest slab strips. The distribution of moments based on the measured
Mu,p from Table 3 was used for calculating the factor α (eq. 11).
In Figure 9 the comparison between the test results Pu and the calculated values Pcalc according to the codes
are shown. As denoted in Table 2, S1, S2, S3, S5, S8, S9 and the slab strips were tested in the middle of the
width, resulting in two tests per specimen in Figure 9. On S4 two tests were carried out close to the simple
support, causing damage extending up to the vicinity of the continuous support. On S6 four tests were carried
out close to the respective edges. The results as shown in Figure 9 show that Regan’s method and EN 1992-1-
1:2005 with load spreading from the far side of the support plate as shown in Figure 1(b) estimate best the shear
capacity of slabs under concentrated loads close to the support. The method from the French National Annex
with load spreading from French practice as shown in Figure 1(b) overestimates the shear capacity. This
method overestimates the shear capacity because significantly higher stresses (127% larger if k = 2) are allowed
for slabs benefitting from the effects of transverse redistribution, because the effect of the factor β is large due to
the short shear spans and because the effective width is larger as a result of load spreading from the far side of
the loading plate. Comparing the test data to the shear capacities as calculated with ACI 318-08 shows that this
method is not suitable for slabs under concentrated loads close to the support. This observation is not surprising,
as the ACI 318-08 method predicts an inclined cracking load and not an ultimate load in shear. Due to the short
shear spans, the load was observed to increase after the occurrence of inclined cracking. Moreover, the ACI
318-08 method was developed for slender beams. For elements with short shear spans loaded in shear ACI 318-
08 recommends the use of non-linear methods or strut-and-tie models. This is confirmed by the ACI 318-08
values calculated for the shear capacity of all tests on S5, S6, S9, BM2, BL2 and BX2 being a multiple lower
than the experimental values. For slabs under concentrated loads, the strut-and-tie model recommended by ACI
318-08 should be a three-dimensional strut-and-tie model.

14
 8
EC2, beff1
EC2, beff2
7 Regan
ACI
6 France

5
Pu/Pcalc

0
S1T1
S1T2
S2T1
S2T4
S3T1
S3T4
S4T1
S4T2
S5T1
S5T4
S6T1
S6T2
S6T4
S6T5
S8T1
S8T2
S9T1
S9T4
BS1T1
BS1T2

BS2T1
BS2T2

BS3T1
BS3T2
BM1T2
BM1T1

BM2T1
BM2T2

BM3T1
BM3T2

BX1T1
BX1T2
BX2T1
BX2T2
BX3T1
BX3T2
BL1T1
BL1T2

BL2T1
BL2T2

BL3T1
BL3T2
Figure 9 — Comparison of test results to values calculated according to EN 1992-1-1:2005 with load
spreading as shown in Figure 1(a) and Figure 1(b), French National Annex, ACI 318-08 and Regan’s
method.

In Table 9, the average (AVG), standard deviation (STD) and coefficient of variation (COV) are given for
the comparison between the experimental data and the values calculated from EN 1992-1-1:2005 with load
spreading beff1 as shown in Figure 1(a), with load spreading beff2 as shown in Figure 1(b), with the method from
the French National Annex and with Regan’s method. The first row of Table 9 shows the comparison of all
experimental data with the calculated values. The consecutive rows show the comparison (2nd row) only for
slabs S1 – S6, S8 and S9, (3rd row) only for the slab strips BS1 – BX3 and (4th row) only for the slab strips BS1
– BM3, which only showed a one-dimensional cracking pattern at the bottom. On the 5th row the results for
loading at the simple support for slabs S1 – S6, S8 and S9 are shown and on the 6th row for loading at the
continuous support. The results show that currently, Regan’s method leads to the best estimates for the shear
capacity of slabs under concentrated loads close to the support. The largest coefficient of variation with Regan’s
method is found when comparing the test results from BS1 – BM3 to the calculated results. Only two edges of
the perimeter of Regan’s method (Figure 2) were used (parallel to the support) as the edges perpendicular to the
support were too close to the free edge and did not contribute to the load carrying capacity of the member. As
Regan’s method was developed for slabs, it is not surprising that this method leads to better results for slabs
than for slab strips with a smaller width. For slabs, the method from the French National Annex leads to a
reasonable coefficient of variation (14%), but the average value of Pu/Pcalc (0.84) shows that this method
overestimates the capacity of slabs benefitting from transverse redistribution. The shear capacity based on vmin
(eq. 4) from the French National Annex was larger than VRd,c from eq. 1 for all slabs and slab strips from the BL
and BX series. The slab strips from the BS and BM series were calculated using the expressions for beams for
which VRd,c was larger than the shear capacity based on vmin from eq. 5 in all cases. Using this procedure resulted
in the very large coefficient of variation as observed in Table 9.

Table 9 — Statistical properties obtained from comparing S1 – S6, S8, S9 and the slab strips with values
calculated according to EN1992-1-1:2005 with load spreading according to Figure 1(a) and Figure 1(b),
French National Annex, and with Regan’s method.
Test data Pu/PEC2,beff1 Pu/PEC2,beff2 Pu/ PFR Pu/PRegan
AVG STD COV AVG STD COV AVG STD COV AVG STD COV
All 2.15 0.46 21% 1.56 0.23 15% 1.02 0.41 40% 1.06 0.14 13%
Slabs 2.43 0.30 12% 1.59 0.19 12% 0.84 0.12 14% 1.11 0.15 13%
Strips 1.94 0.45 23% 1.53 0.26 17% 1.15 0.50 43% 1.02 0.12 12%
BS1 – BM3 1.67 0.42 25% 1.57 0.35 22% 1.57 0.35 22% 1.01 0.16 15%
Slabs, SS 2.38 0.26 11% 1.58 0.17 11% 0.84 0.11 13% 1.18 0.12 10%
Slabs, CS 2.50 0.35 14% 1.60 0.22 14% 0.85 0.13 16% 1.02 0.14 14%

15
 Regan’s method explicitly takes the moment distribution at the continuous support into account (Eq. 11).
The results from Table 9 however show that for all considered methods the values for the average Pu/Pmethod,
standard deviation and coefficient of variation are different for loading at the simple and continuous support.
The methods based on EN 1992-1-1:2005 do not differ between loading at the simple or continuous support, and
result in higher average Pu/Pmethod values at the continuous support than at the simple support. Regan’s method
seems to overestimate the influence of the moment distribution at the continuous support, and leads to lower
average Pu/Pmethod values at the continuous support than at the simple support.
Comparing the columns for the calculation of test results over calculated result Pu/Pcalc for EN 1992-1-
1:2005 with load spreading as shown in Figure 1(a), beff1, and Figure 1(b), beff2 shows that using beff2 agrees
better with the experimental results. The average value for Pu/Pcalc becomes smaller and more uniform: compare
the range of 1.7 – 2.4 for beff1 to the range of 1.5 – 1.6 for beff2. The standard deviation becomes smaller, as well
as the coefficient of variation. These statistical parameters confirm that load spreading from the far side is to be
preferred for determining the effective width.
The results also show that EN 1992-1-1:2005 underestimates the shear capacity of slabs. EN 1992-1-1:2005
with load spreading from the far side of the loading plate (Figure 1(b)) underestimates the shear capacity by a
factor 1.5. It can thus be proposed to change the factor β = av/2d to β = av/3d. The results of the comparison
between the experimental data Pu and the values calculated according to EN 1992-1-1:2005 with the proposed
method PEC2,prop are shown in Table 10. The proposed method gives results which are as good as the results
obtained with Regan’s method. However, two improvements need to be made to this method. First, the average
value of test result to calculated result should be similar for different values of a/d. Table 10 shows that this
requirement is not yet fulfilled: the average tends to decrease as the distance between the load and the support
decreases. Second, the factor for direct load transfer to the support can only be applied for a/d ≤ 2.5. The curve
for β = av/3d only reaches the value β = 1 for av = 3d. Limiting this range to av ≤ 2d leads to discontinuity in the
results. Therefore, a method to correctly account for the influence of the shear span to depth ratio on the shear
capacity of slabs still needs to be developed based on the force flow in slabs.

Table 10 — Statistical properties obtained from comparing S1 – S6, S8, S9 and the slab strips with
values calculated according to EN1992-1-1:2005 with load spreading according to Figure 1(b) and β =
av/3d.
Test data Pu/PEC2,prop
AVG STD COV
All 1.04 0.16 15%
Slabs 1.06 0.13 12%
Strips 1.02 0.18 17%
BS1 – BM3 1.05 0.24 23%
slabs, a/d = 2.26 1.11 0.11 9.5%
slabs, a/d = 1.51 0.99 0.12 13%

Regan’s method uses a punching perimeter in which the side of the perimeter close to the support is
calculated enhanced with a factor 2d/av accounting for the vicinity of the support. This factor is also used in EN
1992-1-1:2005: β. Regan’s method combines elements from a punching shear calculation (perimeter around the
load) with elements from regular beam shear calculations (such as the enhancement factor), showing that the
mechanism developing from loading through a concentrated load on a slab is a combination of one-way and
two-way action. The experimental observations regarding the cracking patterns on the bottom face also show a
clear distinction between the cracking pattern in the slabs and the slabs strips with a small width, Figure 7. The
shear calculation methods from EN 1992-1-1:2005 and ACI 318-08 were based on test results from small,
heavily reinforced, slender beams with a width-to-depth ratio smaller than 1. These methods underestimate the
capacity of slabs under concentrated loads, showing an additional capacity in slabs as a result from the
additional dimension of the width over which load can be distributed. It can be concluded that one-way slabs
under concentrated loads close to the support do not behave in the same way as beams in shear, but an additional
capacity and behavior resembling elements of two-way shear failure are observed.
For evaluating existing solid slab bridges, it is currently advised to use EN 1992-1-1:2005 with an effective
width based on load spreading from the far side of the support, beff2. As can be seen in Table 9, the average
Pu/PEC2,beff2 is larger than 1.50. The contribution of the concentrated loads to the overall shear force can therefore
be reduced by at least 25%, which has been recommended as a reduction factor for assessing existing bridges
based on this research. Furthermore, the contribution of the concentrated loads has to be reduced by β=av/2d as
proposed in EN 1992-1-1:2005 to take direct load transfer into account.

16
 CONCLUSIONS
To evaluate the shear capacity of existing slab bridges, it is necessary to assess the shear resistance of slabs
under concentrated loads close to the support and to estimate the effective width in shear. A survey of test data
on shear in slabs under concentrated loads with short shear spans showed the need for a large-scale experimental
program. Test data are reported and discussed with an emphasis on the influence of the width and loading
configuration.
Based on the test results and the results from the literature, it can be concluded that the influence of the
distance to the support is found not to be as large as in beams. A possible explanation for this is the three-
dimensional load spreading in slabs under concentrated loads, resulting in a multitude of possible compression
struts to carry the load from its point of application towards the support.
An increase in shear carrying capacity is observed for an increase in size of the loading plate for both the
experimental results and the results from the literature. The experimental results also show that the influence on
the capacity becomes larger as the member width increases. The results from the literature show that the
influence on the capacity decreases as the distance to the support becomes smaller. These observations could as
well be explained by the transverse load redistribution in slabs.
The influence of the width is observed in the results from the literature as well as in the experimental
results. The concept of an effective width which provides a threshold value for increasing capacities with
increasing widths was experimentally observed. The effective width was found to be influenced by the moment
distribution in the shear span, the size of the loading plate and the location of the load. Future work will relate
the effective width to the equilibrium of forces and compatibility.
The experimental results are compared to the values calculated according to EN 1992-1-1:2005 (with two
different methods of load spreading), the method from the French National Annex, ACI 318-08 and Regan’s
method. The best estimators are EN 1992-1-1:2005 with load spreading from the far side of the loading plate
and β = av/3d and Regan’s method, which is developed for slabs under concentrated loads close to the support.
Load spreading under 45º from the far side of the loading plate is recommended to be used for design in
combination with the beam shear formulas. Due to the extra dimension of the width in slabs, slabs under
concentrated loads are observed to behave differently from beams.
The results of this research can be used for evaluating existing solid slab bridges. It is advised to check the
shear capacity by using EN 1992-1-1:2005. The contribution of concentrated loads to the sectional shear force
can be reduced by combining a 25% reduction in combination with an effective width based on load spreading
from the far side of the loading plate and taking β = av/2d into account.

ACKNOWLEDGMENTS
The authors wish to express their gratitude and sincere appreciation to the Dutch Ministry of Transport,
Public Works and Water Management (Rijkswaterstaat) for financing this research work.

REFERENCES
AASHTO, 2007, “AASHTO LRFD Bridge Design Specifications,” American Association of State
Highway and Transportation Officials, Washington D.C., 4086 pp.
ACI Committee 318, 2008, “Building Code Requirements for Structural Concrete (ACI 318-08) and
Commentary,” American Concrete Institute, Farmington Hills, MI, 465 pp.
CEN, 2003, “Eurocode 1 – Actions on Structures - Part 2: Traffic loads on bridges, EN 1991-2,” Comité
Européen de Normalisation, Brussels, Belgium, 168 pp.
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, 229 pp.
Chauvel, D., Thonier, H., Coin, A. and Ile, N., 2007, “Shear Resistance of slabs not provided with shear
reinforcement,” CEN/TC 250/SC 02 N 726, France, 32 pp.
Clark, A. P., 1951, "Diagonal Tension in Reinforced Concrete Beams," ACI Journal Proceedings, V. 48,
No. 10, Oct., pp. 145-156.
Ekeberg, P. K., Sjursen, A., and Thorenfeldt, E., 1982, "Load-carrying capacity of continuous concrete
slabs with concentrated loads," Nordisk betong, V. 4, No. 2, pp. 153-156.
Furuuchi, H., Takahashi, Y., Ueda, T., and Kakuta, Y., 1998, "Effective width for shear failure of RC deep
slabs," Transactions of the Japan concrete institute, V. 20, pp. 209-216.
Graf, O., 1933, "Versuche über die Widerstandsfähigkeit von Eisenbetonplatten unter konzentrierter Last
nahe einem Auflager," Deutscher Ausschuss für Eisenbeton, V. 73, pp. 10-16.
Kani, G. N. J., 1964, "The Riddle of Shear Failure and Its Solution," ACI Journal Proceedings, V. 61, No. 4,
Apr., pp. 441-467.
Leonhardt, F., and Walther, R., 1962, "Beitrage zur Behandlung der Schubprobleme in Stahlbetonbau - 2.
Fortsetzung des Kapitels II - Versuchsberichte," Beton- und Stahlbetonbau, V. 57, No. 3, pp. 54-64.

17
 Lubell, A.S., 2006, “Shear in wide reinforced concrete members,” Ph.D. dissertation, University of Toronto,
Toronto, ON, Canada, 455 pp.
Regan, P. E., 1982, “Shear Resistance of Concrete Slabs at Concentrated Loads close to Supports,”
Structures Research Group, Polytechnic of Central London, U.K., 24 pp.
Regan, P. E., and Rezai-Jorabi, H, 1988, "Shear Resistance of One-Way Slabs under Concentrated Loads,"
ACI Structural Journal, V. 85, No. 2, Mar.-Apr., pp. 150-157.
Regan, P. E., 1998, "Enhancement of shear resistance in short shear spans of reinforced concrete - an
evaluation of UK recommendations and particularly of BD44/95." University of Westminster, London, U.K., 16
pp.
Richart, F. E., 1927, "An investigation of web stresses in reinforced concrete beams," University of Illinois
Engineering Experiment Station Bulletin No. 166, V. 24, No. 43, University of Illinois, Urbana, 120 pp.
Richart, F. E. and Kluge, R. W., 1939, “Tests of reinforced concrete slabs subjected to concentrated loads: a
report of an investigation,” University of Illinois Engineering Experiment Station Bulletin No. 314, V. 36, No.
85, University of Illinois, Urbana, 86 pp.
Serna-Ros, P., Fernandez-Prada, M. A., Miguel-Sosa, P. and Debb, O. A. R., 2002, “Influence of stirrup
distribution and support width on the shear strength of reinforced concrete wide beams,” Magazine of concrete
Research, V. 54, No. 3, June, pp. 181-191.
Sherwood, E. G., Lubell, A. S., Bentz, E. C. and Collins, M.P., 2006, “One-way Shear Strength of Thick
Slabs and Wide Beams,” ACI Structural Journal, V. 103, No. 6, Nov.-Dec., pp.794-802.
Sherwood, E. G., 2008, “One-way shear behaviour of large, lightly-reinforced concrete beams and slabs,”
Ph.D. dissertation, University of Toronto, Toronto, ON, Canada, 547 pp.
Talbot, A. N., 1909, “Tests of reinforced concrete beams: resistance to web stresses,” University of Illinois
Engineering Experiment Station Bulletin No. 29, V. 6, No. 15, University of Illinois, Urbana, 90 pp.
Technical Committee CSB/39, 1982, “The structural use of concrete (Draft revision of CP110),” British
Standards Institution, London, U.K., Feb., 118 pp.
Walraven, J. C., 2010, “Residual shear bearing capacity of existing bridges,” Shear and punching shear in
RC and FRC elements – Proceedings of a workshop held on 15-16 October 2010 in Salò, Lake Garda, Italy, fib
Bulletin 57, Oct., pp. 129-138.
Wood, J. G. M., 2008, "Implications of the collapse of the de la Concorde overpass." The structural
engineer, V. 86, No. 1, Jan., pp. 16-18.

18

View publication stats