Proceedings of the fib Symposium 2019

Concrete - Innovations in Materials, Design and Structures

SHEAR ENHANCEMENT IN RC BEAMS WITH CONCOMITANT

LOADS NEAR AND FAR FROM SUPPORTS

Marcus V. F. Pastore1 and Robert L. Vollum1

1
Department of Civil and Environmental Engineering, Imperial College London, London, United Kingdom
Corresponding author email: [email protected]

Abstract
Shear enhancement occurs in reinforced concrete beams when loads are applied within a distance of
around twice the beam effective depth (d) from supports. Eurocode 2 (EC2) and fib Model Code 2010
(MC 2010) reduce the component of design shear force arising from loads applied within 2d of supports
on the basis that part of this load is directly transferred to the supports through arching action. On the
other hand, the superseded UK code BS8110 increases the shear resistance provided by concrete, relying
on the hypothesis that shear enhancement is related to the angle of the failure plane. These two
approaches can give very different predictions of shear resistance for beams loaded with multiple
concentrated loads within the same shear span. The paper describes a test programme in which 12 beams
were tested to investigate the influence of loading arrangement on the shear resistance. The beams were
notionally, geometrically identical and divided into three groups: first without links, second with 8-mm
links at 200 mm centres and third with 8-mm links at 300mm centres. The beams were loaded with
either one or two concentrated loads positioned in the critical shear span within 2d of the support.
Another concentrated load was positioned at 3d from the same support where shear enhancement is
minimal. The paper compares the shear strengths of the tested beams having shear reinforcement with
the predictions of EC2, BS8110 and MC2010 as well as nonlinear finite-element analysis (NLFEA). For
beams with shear reinforcement, the EC2 method of shear enhancement performs very poorly with
predictions overwhelmingly on the safe side. The predictions of MC2010 and BS8110 are significantly
better than those of EC2 for the tested beams with shear reinforcement. Tests indicate that the shear
resistance of beams with shear reinforcement is related to the angle of the failure plane as assumed by
BS8110.

Keywords: Shear enhancement, reinforced concrete, short span beams, multiple loads.

1. Introduction
This paper is concerned with shear enhancement in reinforced concrete beams. The vast majority of
laboratory tests investigating shear enhancement have been carried out on beams with single
concentrated loads applied within twice the beam effective depth (d) of supports (e.g. Brown and
Bayrak, 2007 and Sagaseta and Vollum, 2010). This paper describes an experimental programme in
which 12 beams were tested to determine the shear resistance of beams loaded with one/two point load(s)
positioned within 2d of the nearest support and another load applied near mid-span. The tests were
designed to investigate shear enhancement in beams without and with shear reinforcement. This paper
presents detailed results for four of the beams with shear reinforcement. The strength of the tested beams
is compared with the design provisions of Eurocode 2 (EC2) (BSI, 2004), fib Model Code 2010 (fib,
2013) (MC2010) and the superseded UK code BS8110 (BSI, 1997) as well as nonlinear finite element
analysis. EC2 and MC2010 account for shear enhancement by reducing the component of design shear
force due to loads applied within 2d of supports by a multiple  = av/2d. Conversely, BS8110 increases
the shear resistance provided by concrete within 2d of supports by the multiple 1/. The paper considers
the merits of these two approaches for beams with shear reinforcement.

1763
2. Code provisions for shear enhancement
2.1. Eurocode 2
EC2 allows the component of shear force due to loads applied at distance ai from the support centreline,
with clear shear span avi  2d , to be reduced by the multiple:

avi
i  2d
(1)

Where avi  2d is the clear shear span measured to concentrated load Pi and d is the beam effective
depth. For av ≤ 0.5d, the value av = 0.5d should be used.
The shear resistance corresponding to clear shear span avi is given by:


VRd ,i  max VRd ,c ( EC 2), nav Aswi f yd
i
 (2)

in which nav = number of sets of links with cross-sectional area Aswi within the central ¾ of the
i
clear shear span avi and fyd = fyk/s is the design reinforcement yield strength (where fyk is the
characteristic yield strength and s is the partial factor for reinforcement which EC2 takes as 1.15 for
design). EC2 takes the shear resistance without shear reinforcement (VRd,c) as:
1
 100 Asl 3  200 
VRd ,c ( EC 2)  0.18bd  f ck   1   /c (3)
 bwd   d 

Where bw is the minimum beam width in the tensile area, Asl is the area of flexural tension
reinforcement, fck is the characteristic concrete compressive cylinder strength and c is the partial factor
for concrete which equals 1.5 for design.
In the absence of shear enhancement EC2 calculates the shear resistance of beams with shear
reinforcement as follows:
VRd ,s  0.9 Asw f yd d cot  / s  VRd ,max (4)

Where
VRd ,max  0.9bwd1 f cd /  cot   tan   (5)

 f 
in which f cd  f ck  c , 1  0.6 1  ck  and 1  cot   2.5 .
 250 

2.2. fib Model Code 2010 (MC2010) shear enhancement

As noted by Vollum and Fang (2015), equation (2) can give illogical results for beams with shear
reinforcement and multiple point loads within 2d of supports. The illogicality arises because in the case
of a beam loaded at av2, the application of a notional point load at av1 < av2 can significantly reduce shear
resistance due to fewer links contributing to shear resistance within av1. fib Model Code 2010 (fib, 2013)
avoids this problem by reducing the design shear force as in EC2 but calculating the shear resistance
VRd,i corresponding to shear span avi in its normal way. In the case of point loads applied with av < d, the
design shear force is calculated in MC2010 with  = 0.5 as if the load was applied at av = d. Design
methods in MC2010 depend on the adopted Level of Approximation (LoA). This paper presents results
calculated with LoA II and LoA III but only LoA II is described in detail. In LoA II, the shear resistance
of beams with shear reinforcement is given by:
A
VRd ,s ( MC 2010)  sw zf yd cot   VRd ,max (6)
s

1764
 f
VRd ,max  kc ck bw z sin  cos (7)
c

min  20  10000 x (8)

kc  k fc (9)

1
k   0.65 (10)
1.2  551
1
 30  3
 fc    1 (11)
 f ck 

1   x   x  0.002 cot 2  (12)

1  M Ed 
x    VEd  (13)
2 Es Asl  z 
Where Es is the Young’s modulus of the longitudinal reinforcement, Asl is the area of flexural tension
reinforcement, z is the lever arm which may be taken as 0.9d for shear and MEd is the design bending
moment taken at d from the support or concentrated load but not closer than d from the face of the
support. LoA III, which is more accurate, also includes a concrete contribution to shear resistance.

2.3. BS8110 shear enhancement

BS8110 relates the enhanced shear resistance to the angle of the failure plane which is assumed to run
from the outside edge of the support to the inside edge of the load. The shear resistance provided by
concrete is assumed to increase in proportion to 1/. For beams with shear reinforcement, the shear
strength within 2d of the support is given by:
V d
VRdi  Rd ,c  na Aswi f yd (14)
i avi vi

Where VRdi is the shear resistance for failure along an inclined plane with horizontal projection avi
and VRd,c is the shear resistance of the beam without shear reinforcement. To allow direct comparison
of the shear enhancement methods in EC2 and BS8110, VRdc is calculated with EC2 in this paper when
applying the BS8110 shear enhancement method.

2.4. Stress field analysis

EC2 also permits shear enhancement to be modelled using the use of stress field analysis as described
by Sigrist et al. (1995). In this approach, which EC2 does not explain in any detail, part of the load is
transferred to the support by direct strut action with the remainder being transferred to the links, if
present, through fan shaped stress fields. As shown by Sagaseta and Vollum (2010), as well as Fang and
Vollum (2015), this approach gives reasonable strength predictions for short span beams loaded with
concentrated loads. The novelty of the current work is that it considers shear enhancement in beams
loaded both within and outside 2d of supports experimentally and with NLFEA. Though relevant, stress
field analysis is not considered further in this paper.

3. Experimental programme
3.1. Introduction
The 12 tested beams were divided into three series with: 1) no vertical links in the main shear span, 2)
8 mm links spaced at 200 mm centres and 3) 8 mm links spaced at 300 mm centres in the shear span of

1765
interest. All 12 beams measured 4350 mm long × 250 mm (wide) × 500 mm (deep). Top longitudinal
reinforcement was two 16 mm diameter straight bars. Loading and bearing plates were fabricated from
steel and measured 250 mm (wide) ×100 mm (long) ×30 mm (thick). Each series of beams was cast and
tested separately. The span between centrelines of supports was 3350 mm. This paper describes the
results of the second series of tests in which beams were reinforced with 8 mm diameter links at 200
mm centres. The bottom longitudinal reinforcement consisted of four 25 mm diameter bars detailed as
shown in Figure 1. Additional links were provided in the end regions of the beams for anchorage of
flexural reinforcement.

3.2. Material properties

The beams in Series 2 were cast from ready mix concrete specified to have strength of class C25/30
and limestone aggregate with maximum size of 20 mm. During the testing period, the average
compressive cylinder strength of the air cured specimens varied between 29.3 and 32.0 MPa. The
strength of water cured specimens varied between 27.7 and 30.1 MPa. The average concrete tensile
strength of the water cured specimens varied between 2.3 MPa on the first day of testing and 2.5 MPa
on the last day of testing. The corresponding strengths of the air cured specimens were 2.0 and 2.1 MPa
respectively. Table 1 summarises the age of loading, estimated concrete strengths on day of testing and
loading arrangement for each beam. The estimated concrete compressive strengths are averages of the
air and water cured strengths. Table 2 summarises the properties of the tensile reinforcement.

a)

b)

c) d)

Figure 1. Geometry and typical loading arrangements of the beams: a) Series 2, b) loading arrangement for
beam 2P-1.1d/1.9d-50:50-425s, c) cross-section details, f) labelling of specimens.

1766
 Table 1. Description of the beams of series 2.

Age at test fc ft Loading

Beam
(days) (MPa) arrangement

P-1.5d-425s 28 28.5 2.2 Fig 1b - P1=100%, P2=0%

2P-1.1d/1.9d-50:50-425s 31 28.9 2.2 Fig 1d - P1= P2=50%

2P-1.5d/3d-40:60-425s 32 29.0 2.2 Fig 1b - P1=40%, P2=60%

P-3d-425s 36 29.4 2.2 Fig 1b - P1=0%, P2=100%

2P-1.5d/3d-60:40-425s (re-test) 51 31.1 2.3 Fig 1b - P1=60%, P2=40%

Table 2. Steel reinforcement properties.

Diameter E fy fu
(mm) (GPa) (MPa) (MPa)
8 200100 510 695
16 194900 550 645
25 200600 540 650

3.3. Test results and discussions

Beams P-1.5d-425s, with a single point load, and beam 2P-1.1d/1.9d-50:50-425s, with two point
loads, were loaded such that the centroid of the resultant concentrated load was at 1.5d in both beams.
Both beams failed in shear but beam P-1.5d-425s failed at a maximum shear force of 387 kN compared
with 632 kN in beam 2P-1.1d/1.9d-50:50-425s. This shows that it can be overly conservative to replace
multiple point loads within 2d of the support by their resultant when assessing shear resistance. Both
beams failed due to the formation of a critical shear crack which ran between the outside edge of the
support and the inside edge of the adjacent loading plate. However, the kinematics and shape of the
critical shear cracks were rather different. The critical shear crack in beam P-1.5d-425s with only one
load at 1.5d (Fig.3a) initiated at the inner edge of the support and extended towards the inner edge of
the loading plate. Near failure the crack joined up with a secondary crack which initiated at the inner
corner of the loading plate. Conversely, in beam 2P-1.1d/1.9d-50:50-425s the failure crack initiated,
near failure, at the outer corner of the loading plate and extended towards the support subsequent to
secondary cracks forming in the span. The authors attribute the greater shear resistance of beam 2P-
1.1d/1.9d-50:50-425s, compared with P-1.5d-425s, to the steeper angle of its failure plane. This
assumption is in agreement with Mihaylov at al. (2013). Beam 2P-1.5d/3d-40:60-425s failed in flexure
due to concrete crushing at the outer load subsequent to reinforcement yield. The beam was repaired
and retested as beam 2P-1.5d/3d-60:40-425s with the loads in the same positon but with loading ratio
60:40. Beam 2P-1.5d/3d-60:40-425s also failed in flexure.

4. Comparison of code strength predictions and NLFEA

The shear capacity of the beams was calculated with i) EC2, ii) MC2010 LoA II, iii) MC2010 LoAIII,
iv) BS8110 and v) NLFEA using ATENA (Cervenka et al., 2018). The shear enhancement methods
were applied as described in Section 2. The MC2010 shear resistance depends on the strain in the
flexural reinforcement at the critical section which was taken at d from the face of the support for all
beams except P-3d-425s where it was taken at d from the inside edge of the loading plate. MC2010
does not state whether or not the flexural reinforcement strain used to calculate shear resistance should
be based on the actual moment or that calculated with the reduced shear force. In this paper, the strain

1767
is calculated in terms of the actual shear force on the basis that the shear resistance depends on the actual
strain. This assumption reduces the calculated shear resistances by 5-10% compared with basing the
strain on the reduced design shear force. All links within the shear span were considered effective since
strain gauge readings show the reinforcement to have either yielded or been close to yielding. The
material strengths used in the analyses are given in Tables 1 and 2. Material partial factors were taken
as 1.0. In ATENA, concrete was modelled with nonlinear cementitious 2, which is a combined fracture-
plastic model (Cervenka and Papanikolaou, 2008). The fixed crack model was used with a shear factor
of 3000 for beams with links. The concrete elastic modulus and fracture energy are calculated in terms
of the measured concrete compressive strength using MC2010. Default values were used for the other
material parameters. Concrete was meshed using 50×50×50 mm3 8-node hexahedral elements with
linear interpolation. The arc-length solution procedure was used with initial load step set at 1% of the
measured failure load.

a) b)

c) d)

e) f)

Figure 3. Details of Series 2. Final crack patterns of: a) P-1.5d-425s, b) 2P-1.1d/1.9d-50:50-425s, c) 2P-
1.5d/3d-40:60-425s, d) concrete crushed in the flexural zone of 2P-1.5d/3d-40:60, e) P-3d-425s, f)
strengthening of failure zone with grout.

Table 3 shows the measured and calculated maximum shear resistances. EC2 gives a good estimate
of the shear resistance of beam P-3d-425s with a single point load at 3d but very poor estimates for the
other beams since the enhanced shear resistance is less than the unenhanced resistance. The BS8110
predictions are the best of the code predictions as found by Vollum and Fang (2015) for a large database

1768
of test specimens. The MC2010 LoA III predictions are also reasonable but more variable than the
BS8110 ones. The ATENA predictions are most accurate. A parametric study was carried out to
investigate the effect of varying the load proportions between the two loading points of beams with point
loads positioned at 1.5d and 3.0d. The geometry and reinforcement details used in the parametric study
were the same as used for test Series 2 and 3. The concrete compressive strength was taken as 30 MPa.
Test Series 3 was similar to test Series 2 but the link spacing was increased to 300 mm and six 25-mm
bottom bars were provided, in two layers, to ensure shear failure. For each load ratio, the shear resistance
was calculated according to the design codes and NLFEA. The results of the parametric study are
presented in Figure 4a (links at 200 mm centres) and 4b (links at 300 mm centres) which also include
test results from Series 2 and 3 respectively. Figure 4 shows that both the test results and ATENA
predictions show the shear resistance to be fairly uniform for load ratios between 40% and 100% in
which the % refers to the proportion of the total load applied at 1.5d. The test results and NLFEA crack
patterns show that the transition in behaviour is associated with a change in angle of the failure plane as
found for specimens P-1.5d-425s, and 2P-1.1d/1.9d-50:50-425s (see Section 3.3). The critical shear
crack extended from the support to the load at 1.5d for load ratios above 40%. Below a load ratio of
40%, the shear force in the outer span was critical.

Table 3. Measured and predicted ultimate shear force.

Maximum shear Vtest/Vcalc
force
Beam Type of Flexural
failure Test MC2010 MC2010 ATENA
failure EC2 BS8110
(kN) LoAII LoAIII (NLFEA)
(kN)
2P-1.1d/1.9d-
Shear 632 705 3.05 2.13 1.81 1.48 1.04
50:50-425s
P-1.5d-425s Shear 387 686 1.56 1.25 1.07 1.19 0.83
2P-1.5d/3d-
Flexure 535 548 2.59 1.99 1.65 1.61 1.14
60:40-425s
2P-1.5d/3d-
Flexure 492 455 2.62 1.97 1.63 1.50 1.05
40:60-425s
P-3d-425s Shear 297 324 1.13 1.49 1.21 1.13 0.93

Figure 4. Parametric study of varying the load proportions between loading points for series 2 (left) and 3
(right).

1769
5. Conclusions
This paper describes an experimental programme which was carried out to investigate the effect of
loading arrangement and link ratio on shear enhancement in simply supported RC beams. The paper
presents detailed results for four of the 12 tested beams. The strengths of the tested beams were evaluated
using EC2, MC2010 and BS8110 as well as NLFEA. Of the code methods, BS8110 gave the best
strength predictions and EC2 the worst. BS8110 assumes that shear resistance is related to the angle of
the failure plane. This assumption is consistent with the observed behaviour of the tested beams with
shear reinforcement but not for beams without shear reinforcement. For example, the shear resistance,
and failure mode, of beams in Series 3, with links at 300 mm centres and loads applied at 1.5d and 3.0d,
was almost the same for load ratios at 1.5d of 40% and 100% of the total. The EC2 shear enhancement
rule of only accounting for the strength provided by the concrete or the links produced extremely
conservative predictions with the enhanced strength being less than the unenhanced strength.

Acknowledgements
The first author wishes to acknowledge the financial support of the National Council for Scientific and
Technological Development (CNPq), through the Science Without Borders Programme, funded by the
Brazilian Ministry of Science and Technology.

References
Brown MD and Bayrak O (2007) Investigation of deep beams with various load, configurations. ACI Structural
Journal 104(6): 11–20.
BSI (1997). BS 8110. Part 1: Structural use of concrete: code of practice for design and construction. London, UK:
BSI.
BSI (2004). European Standard EN-1992-1-1: 2004. Eurocode 2: Design of concrete structures. Part 1, general
rules and rules for buildings. London, UK: BSI.
Cervenka, J. and Papanikolaou, V. K. (2008). Three dimensional combined fracture-plastic material model for
concrete. International Journal of Plasticity, 24, 2192-2220.
Cervenka, V., Jendele, L. & Cervenka, J. (2018). ATENA Program Documentation. Prague: Cervenka Consulting
s.r.o.
fib (2013). fib Model Code for Concrete Structures 2010. Germany: Ernst &amp; Sohn.
Mihaylov BI, Bentz EC, Collins MP (2013) Two-Parameter Kinematic Theory for Shear Behavior of Deep Beams,
ACI Structural Journal, V. 110, No. 3, 447-455
Sagaseta J and Vollum RL (2010) Shear design of short-span beams. Magazine of Concrete Research 62(4): 267–
282, http://dx.doi.org/10.1680/macr.2010.62.4.267.
Sigrist V, Alvarez M, Kaufmann W (1995). Shear and flexure in structural concrete beams, Comité Euro-
International du Béton (CEB), Bulletin d’Information No. 223, “Ultimate Limit State Design Models”, A
state-of -art report.
Vollum, R. L. and Fang, L. (2014). Shear enhancement in RC beams with multiple point loads. Engineering
Structures, 80, 389-405.
Vollum, R. L. and Fang, L. (2015). Shear enhancement near supports in RC beams. Magazine of Concrete
Research, 67, 443-458.

1770