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

net/publication/276251742

Shear enhancement near supports in RC beams

Article  in  Magazine of Concrete Research · May 2015

DOI: 10.1680/macr.14.00309

CITATIONS READS

7 4,214

2 authors, including:

Robert L. Vollum
Imperial College London
89 PUBLICATIONS   703 CITATIONS   

SEE PROFILE

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

MCTI / CNPQ / Universal 14/2014 - Desenvolvimento de Armadura Inovadora para o Combate à Punção em Lajes Lisas de Concreto View project

Shear Enhancement in Deep Beams View project

All content following this page was uploaded by Robert L. Vollum on 25 June 2018.

The user has requested enhancement of the downloaded file.

Magazine of Concrete Research Magazine of Concrete Research
http://dx.doi.org/10.1680/macr.14.00309
Shear enhancement near supports in RC Paper 1400309
beams Received 22/09/2014; revised 24/11/2014; accepted 27/11/2014
Vollum and Fang
ICE Publishing: All rights reserved

Shear enhancement near

supports in RC beams
Robert L. Vollum Libin Fang
Reader in Concrete Structures, Department of Civil and Environmental Former PhD student, Department of Civil and Environmental Engineering,
Engineering, Imperial College London, London, UK Imperial College London, London, UK

Shear strength is enhanced within the short shear span of reinforced concrete beams that are loaded on their upper
side within a distance of around 2–2 .5d of supports (where d is the beam effective depth). Eurocode 2 and fib Model
Code 2010 (MC2010) account for this by reducing the design shear force, unlike the previous UK code BS 8110, which
increases the shear resistance provided by concrete. Eurocode 2 and MC2010 also allow shear enhancement to be
modelled using strut-and-tie models. Very few test data are available to assess the comparative merits of these
approaches for the design of beams with multiple point loads within 2d of supports. Consequently, 12 beams were
tested to investigate the influence of loading arrangement on shear resistance. Comparisons are made between the
strengths of the tested beams and the predictions of BS 8110, Eurocode 2, MC2010, strut-and-tie modelling and non-
linear finite-element analysis. Significantly, the BS 8110 approach of enhancing shear resistance is found to give better
strength predictions than the load reduction methods of Eurocode 2 and MC2010. Accuracy of the non-linear finite-
element analysis and strut-and-tie model is broadly comparable, but the former requires calibration, unlike the latter.

Notation lb length of support

As cross-sectional area of flexural tension reinforcement lt length of loaded area
Asl area of flexural tension reinforcement Mfd moment at the control section for shear
Asw cross-sectional area of each set of stirrups nave number of stirrups with cross-sectional area Asw within
av clear shear span between adjacent edges of load and central three-quarters of clear shear span avs
support s stirrup spacing
bw effective web width, which equals b for rectangular Td longitudinal component of force resisted by strut III
section T i9 longitudinal component of force resisted by strut VII
c distance from bottom of beam to centroid of bottom Tsi yield capacity of stirrups in tie Tsi
flexural reinforcement VRd design shear resistance
d beam effective depth VRdc design shear resistance provided by the concrete alone
dg maximum aggregate size Vcalc calculated shear resistance
Es elastic modulus of reinforcement Vtest measured shear resistance
ei distance to line of action of stirrup force Tsi from wb width of strut III at its bottom end
centreline of support wt width of strut III at its top end
fc measured concrete cylinder strength Æ inclination of strut VII to the horizontal
fcd design concrete uniaxial compressive strength  proportion of force Td + T i9 transferred to bottom node
fck characteristic concrete cylinder strength by strut III
fcnt flexural compressive stress used in strut-and-tie model ªc partial factor for concrete
fcsb concrete strength at bottom end of direct strut ªs material factor of safety for reinforcement
fcst concrete strength at top end of direct strut x longitudinal strain at mid-depth of member
fy yield strength of reinforcement fc coefficient used in calculation of maximum shear
fyd design yield strength of reinforcement resistance
h beam depth Ł inclination of strut III to the horizontal
kc coefficient used by MC2010 in calculation of maximum k P2 /P1, where P1 and P2 are defined in Figure 3
possible shear resistance º proportion of shear force carried by strut III
kV coefficient used my MC2010 in calculation shear
resistance provided by concrete Introduction
k coefficient used in calculation of maximum shear The present research was prompted by differences between the
resistance approaches adopted in BS 8110 (BSI, 1997), Eurocode 2 (BSI,

1
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

2004) and fib Model Code 2010 (MC2010) (fib, 2013) for
modelling shear enhancement, which is assumed to occur when
beams are loaded on their upper side within 2d of supports. BS
8110 increases the shear resistance provided by concrete by the
3000
multiple 2d/av (where av is the clear shear span), unlike Eurocode
(a)
2 and MC2010 which reduce the component of the design shear
160
force due to loads applied on the upper side of the beam within
2d of supports by the multiple av =2d: All three codes also limit

437·5 (462·5)
the maximum possible shear resistance. MC2010 and Eurocode 2
also allow the strut-and-tie method (STM) to be used to model 500
shear enhancement near supports, bringing into question which
method to use. Many beams have been tested with single 50 (25)
concentrated loads within 2d of supports, but there is an almost 2H25
25
complete lack of data on beams with two or more concentrated (b)
loads within 2d of supports (Brown and Bayrak, 2007). To
address this omission, the authors tested 12 simply supported Figure 1. Reinforcement arrangement for series 1 beams:
beams with up to two point loads positioned within 2d of each (a) elevation; (b) cross-section
support. The tests investigated the influences on shear resistance
of reinforcement, concrete cover, bearing plate dimensions and
loading arrangement. This paper briefly describes the beam tests shear reinforcement were used, S1 and S2, which are depicted in
and a STM that is suitable for the analysis of beams with up to Figures 2(b) and 2(c), respectively. Figure 3, which should be read
two point loads adjacent to supports. Full details of the beam in conjunction with Table 1, gives details of the loading arrange-
tests and STM can be found elsewhere (Fang, 2014; Vollum and ments. In all the tests, loads were applied with a single actuator
Fang, 2014). The main contribution of this paper is the compara- that reacted against an internal reaction frame. All the loading
tive evaluation of STM, non-linear finite-element analysis and support plates were 50 mm thick. Figure 3(d) shows the
(NLFEA) and the sectional shear enhancement methods of BS method of load application for loading arrangement C with four
8110, Eurocode 2 and MC2010. This is done through parametric point loads. The shaded rollers in Figure 3(d) were fixed by
studies and analysis of data from the tested beams and 57 other welding the roller to the lower bearing plate. The test specimen
short-span beams with shear reinforcement. The parametric stud- was supported on bearings which allowed horizontal translation as
ies investigate the influence of loading arrangements not con- well as rotation. In loading arrangement B with two point loads,
sidered in the tests. NLFEA is used to assess the influence of the solid steel beams shown in Figure 3(d) were omitted and the
varying the loading arrangement on the shear strength of beams test specimen was loaded through the concrete-filled rectangular
with the same geometry and reinforcement arrangements as those hollow section (RHS). No instabilities were encountered for any
tested by the authors. The predictions of the STM and codified of the loading arrangements as failure was approached. Further-
sectional design methods are subsequently compared with the more, the reactions measured at the right-hand end of the beam
NLFEA predictions. Suggestions are also made for improving the showed that the loads were applied as intended. The stirrup
shear enhancement provisions of Eurocode 2. positions in Figure 2(c) were chosen to clarify the definition of
the ties Ts1 and Ts2 in the STM of Figure 4. In each set of beams,
Description of beam tests failure was expected to occur on the side of the narrowest support,
The beams were cast in two groups of six, which are denoted which was on the right-hand side as seen in Figure 3.
series 1 and 2. The beam geometry and reinforcement arrange-
ments are shown in Figures 1 and 2 for series 1 and 2, The beams were cast from ready-mix concrete in which the coarse
respectively. The first six beams were 500 mm deep 3 160 mm aggregate was marine dredged gravel with a maximum aggregate
wide and were reinforced with two 25 mm diameter high-tensile size of 10 mm. Push-off tests were not carried out as the aggregate
reinforcement bars as shown in Figure 1. The cover to the flexural was similar to that tested by Sagaseta and Vollum (2011a, 2011b).
reinforcement was either 25 mm or 50 mm. The cover was varied Twelve concrete cylinders were cast with each batch of beams, of
because STM predicts the strength to increase with cover, whereas which six were cured in air alongside the beams and six in water
the sectional design methods of BS 8110, Eurocode 2 and at 208C. Three water- and three air-cured cylinders were tested at
MC2010 predict the strength to decrease owing to the reduction the same time as the first and last beams of each series. Table 2
in effective depth d. The second six beams, which measured shows the estimated water-cured concrete strengths at the times of
505 mm deep 3 165 mm wide, were grouped into three pairs testing the beams, which were derived assuming a linear increase
which are depicted A, S1 and S2, of which pair A was in strength between the times of testing of the first and last beams
unreinforced in shear. Unlike the first set, four 8 mm diameter in each group. The strength of the air-cured cylinders was on
stirrups were provided at the ends of the beams to improve the average around 5% less than that of the water-cured cylinders,
anchorage of the flexural reinforcement. Two arrangements of which were used in the strength assessments of the beams to give

2
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

100 100 100 100 2200 100 100 100 100

(a)

100 100 100 100 312·5 200 200 200 187·5 187·5 200 200 200 312·5 100 100 100 100
(b)

100 100 100 100 312·5 100 100 200 100 100 187·5 187·5 100 100 200 100 100 312·5 100 100 100 100
(c)
165

48
2H16

505
H8

50
4H25
37·5

25
(d)

Figure 2. Reinforcement arrangement for series 2 beams:

(a) A beams; (b) S1 beams; (c) S2 beams; (d) cross-section

an upper bound to the predicted strengths. Table 3 gives the concrete surface along the reinforcement. Strains were also meas-
reinforcement 0 .2% offset yield strengths. ured with electrical resistance strain gauges in the flexural and
shear reinforcement of beams A-1, S1-1 and S2-1. The gauges
Reinforcement strains were estimated in all of the beams from were positioned in the shear span adjacent to the narrowest
displacements measured between Demec points mounted on the support, which STM predicts to be critical. In reality, beams S1-1

3
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

av P

lh

l1 l2 l3 l0 l⬘3 l⬘2 l⬘1

(a)

av P P

lh

l1 l2 l3 l0 lc l0 l⬘3 l⬘2 l⬘1

(b)

P1 P2 P2 P1
av2
av1

Failure plane P1
lh
Failure plane P2

l1 l2 l3 l0 l4 l0 lc l0 l⬘4 l0 l⬘3 l⬘2 l⬘1

(c)

Load through spherical seating

Load cell
Concrete-filled RHS

Solid steel beam

Clear span av1 Load cell

Clear span av2

(d)

Figure 3. Loading arrangements: (a) A; (b) B; (c) C; (d) loading rig

for C

4
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Beams Load Rebar l0 l1 l91 l2 l92 l3 l93 l4 l94 lc lh

arrangement arrangementc

B1-25a A 1a 200 500 550 200 150 700 700 — — — 500

B1-50a A 1a 200 500 550 200 150 700 700 — — — 500
B2-25a B 1a 100 700 700 200 150 325 325 — — 400 500
B2-50a B 1a 100 700 700 200 150 325 325 — — 400 500
B3-25a C 1a 100 450 500 200 150 325 325 275 275 100 500
B3-50a C 1a 100 450 500 200 150 325 325 275 275 100 500
A-2b B 2a 100 450 500 200 100 700 750 — — 100 505
S1-2b B 2b 100 450 500 200 100 700 750 — — 100 505
S2-2b B 2c 100 450 500 200 100 700 750 — — 100 505
A-1b C 2a 100 450 500 200 100 325 375 275 275 100 505
S1-1b C 2b 100 450 500 200 100 325 375 275 275 100 505
S2-1b C 2c 100 450 500 200 100 325 375 275 275 100 505
a
Series 1.
b
Series 2.
c
Figure number in which reinforcement is shown.

Table 1. Description of loading arrangements (to be read in

conjunction with Figure 3)

a2 a2
a1 P1 W P2 P1 W
a1 P2

Strut IV CIV Strut IV

Strut V CIV
Strut I α CIII
Strut II CIII
II
tV

CVI
II

CVI
tV

α
ru

α V
ut
St

Ts1 II
ru

Ts2
Y h tI Str Y
St

h III ru Ts1 T
ut V I St θ tV
I s2
Str θ rut tru
St 2c 2cβ S
2c 2cβ
0·5lb e1
e1 X
0·5lb X e2
e2
(b)
a2
a1 P1 W P2

Strut I CII
II
Strut
σIII
I
2c 2cβ h u t II Y
T Str
σ1 ⭐ 0·85vfcd 2c
Ts1 P1 ⫹ P2 ⫺Ts1
λlb 0·5lb X
lb

(a)
(c)

Figure 4. STM for beams with two point loads within 2d of a

support: (a) STMa for P2 > Ts1 + Ts2; (b) STMb for P2 , Ts1 + Ts2;
(c) STMc with no shear reinforcement (adapted from Vollum and
Fang, 2014)

5
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Beam fc: MPa No. loads d: mm ÓP: kN Ptest /Pcalc

Test Flexa Eurocode 2 BS 8110 MC2010c

Equation 4

Lb Rb Lb Rb Lb Rb

B1-25 45 .7 1 462 .5 368 558 1 .61 1 .61 1 .61 1 .61 1 .75 1 .73
B1-50 45 .7 1 437 .5 352 510 1 .67 1 .67 1 .67 1 .67 1 .78 1 .77
B2-25 45 .7 2 462 .5 977 1001 1 .98 1 .98 1 .98 1 .98 3 .54 3 .50
B2-50 45 .7 2 437 .5 929 942 2 .04 2 .04 2 .04 2 .04 3 .45 3 .42
A-2 34 .6 2 442 .5 349 890 1 .53 1 .64 1 .53 1 .64 1 .54 1 .61
S1-2 36 .0 2 442 .5 601 890 1 .09 1 .17 1 .19 1 .27 1 .04 1 .08
S2-2 36 .7 2 442 .5 820 890 1 .00 1 .07 1 .27 1 .36 1 .05 1 .08 d
B3-25 45 .7 4 462 .5 480 726 1 .53 1 .53 1 .05 1 .05 2 .02 2 .00
B3-50 45 .7 4 437 .5 580 684 2 .01 2 .01 1 .37 1 .37 2 .56 2 .53
A-1 33 .1 4 442 .5 823 1235 2 .68 2 .94 1 .83 1 .96 3 .19 3 .24
S1-1 33 .7 4 442 .5 1000 1235 1 .76 1 .95 1 .28 1 .47 1 .53 1 .55
S2-1 35 .2 4 442 .5 1179 1235 1 .38 1 .53 1 .26 1 .45 1 .33 1 .34d
a
Flexural failure load calculated with fy.
b
Calculated for right (R)/left (L) shear span as defined in Figure 3 (bold type denotes critical shear span in test).
c
Level 2 for beams without shear reinforcement and level 3 for beams with shear reinforcement.
d
Calculated with average stirrup spacing of 120 mm.

Table 2. Measured and predicted failure loads

Eurocode 2:
Type Es: GPa fy: MPa

H8 200 540 1a: 1 P1 þ 2 P2 < V Rdav1 ¼ max(V Rd,c , nav1 Asw f yd )

H16 200 540
H25 200 560

Table 3. Reinforcement properties 1b: 2 P2 < V Rdav2 ¼ max(V Rd,c , nav2 Asw f yd )

and S2-2 unexpectedly failed on the side of the wider support, MC2010:
which has implications for design when using STM. The failure
loads of all the beams are listed in Table 2, which also gives the 2: 1 P1 þ 2 P2 < V Rd
calculated flexural failure loads and the shear span within which
failure occurred.
av1 av2
Shear enhancement in Eurocode 2, MC2010 3:
1 ¼ , 2 ¼
2d 2d
and BS 8110
MC2010 and Eurocode 2 reduce the component of the design
shear force due to loads applied within 2d, unlike BS 8110 which where V Rd is the design shear resistance, V Rd,c is the design shear
increases the shear resistance provided by concrete. However, resistance provided by the concrete alone and navi is the number
MC2010 and BS 8110 add the design shear resistances provided of stirrups with cross-sectional area Asw within the central three-
by the concrete and shear reinforcement, unlike Eurocode 2 quarters of the clear shear span avi, which is defined in Figure
which takes the shear resistance as the greater of the two. BS 3(c). Eurocode 2 limits  to a minimum of 0 .25, whereas
8110 and Eurocode 2 assume stirrups to be effective if positioned MC2010 limits it to a minimum of 0 .5.
within the central three-quarters of the clear shear span. The three
design methods are compared below for beams symmetrically Equation 1 gives rise to illogical results for load case C of Figure
loaded with two point loads positioned within 2d of each support, 3(c) if nav1 Asw f yd . V Rd,c , since the application of an infinitesi-
as shown in Figure 3(c). mally small load at av1 reduces the design shear resistance from

6
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

nav2 Asw f yd to nav1 Asw f yd : Other ambiguities in Equation 1 have where Asl is the area of flexural tension reinforcement, bw is the
been reported by Jackson et al. (2007). The illogicality of width of the cross-section, d is the beam effective depth, V Rd,s is
Equation 1 arises when V Rdav1 , V Rdav2 and can be avoided by the design shear resistance with shear reinforcement, z ¼ 0 .9d
interpreting Eurocode 2 as follows and f ck is the characteristic compressive concrete cylinder
strength. 1 ¼ 0.6[1  ( f ck =250)] f cd and 1 < cotŁ < 2.5: The
4: 1 P1 =V Rdav1 þ 2 P2 =V Rdav2 < 1 reinforcement ratio Asl /bd is limited to a maximum of 0 .02 in
Equation 6. ªc is a partial factor for concrete which has a design
value of 1 .5 but is taken as 1 .0 in the strength assessments of this
in which V Rdav1 and V Rdav2 are defined in Equation 1. paper. Eurocode 2 does not define the value of cotŁ to be used in
Equation 8 when av , 2d. This omission is rectified in the
According to BS 8110, failure occurs along either plane P1 or P2 background document to the UK National Annex to Eurocode 2
in Figure 3(c) for which (BSI, 2010), which takes cotŁ at av /h but not less than 1 .0.

MC2010 gives four levels of design approximation for shear, of

5a: 1 (P1 þ P2 ) < V Rdav1 ¼ (V Rd,c þ 0.5nav1 Asw f yd )
which levels 2 and 3 are intended for detailed design and
assessment of existing structures.

MC2010 level 2: V Rd ¼ max(V Rd,c , V Rd,s ) where

5b: 2 P2 < V Rdav2 ¼ (V Rd,c þ 0.5nav2 Asw f yd )
pffiffiffiffiffiffiffi
f ck
V Rd,c (MC2010) ¼ k v zbw
BS 8110 takes the failure load as the least of the enhanced shear 9: ªc
resistances V Rdav1 =1 and V Rdav2 =2 , which avoids the illogicality
that arises with Equation 1. Consideration of Equations 5(a) and
5(b) shows that for uniformly distributed stirrups at spacing s, the 0 .4 1300
enhanced shear reinforcement contribution is V Rd,s ¼ kv ¼
10: 1 þ 1500x 1000 þ k dg z
0.75(Asw =s)df ywd if nav ¼ 0.75av =s: The corresponding enhanced
shear reinforcement contribution from Equation 1(b) is
V Rd,s ¼ 1.5(Asw =s)df ywd , which implies Eurocode 2 gives no
shear enhancement when the shear resistance is governed by 32
k dg ¼ > 0.75
shear reinforcement. Equation 2 of MC2010 avoids the problems 11: 16 þ d g
of Equation 1 by making the shear resistance V Rd independent
of av.
   
Calculation of shear resistance with Eurocode 1 M Ed V Ed (d þ 0.5lb )
x ¼ þ V Ed > þ1
2 and MC2010 12: 2Es As z 2Es As z
Eurocode 2 and MC2010 use the following equations to calculate
the design shear resistance V Rd of reinforced concrete beams with
av /d . 2. where dg is the maximum aggregate size, lb is the length of the
support, z ¼ 0 .9d and MEd is the moment at the control section
Eurocode 2: V Rd ¼ max(V Rd,c , V Rd,s ) where for shear, which is located at a distance d from the edge of
supports or concentrated loads. The former control section
V Rd,c (Eurocode 2) ¼ 0.18bw d governs for the beams considered in this paper as av < 2d.
 1=3 rffiffiffiffiffiffiffiffi!,
100Asl 200 Asw
3 f ck 1þ ªc 13:
V Rd,s ¼ zf yd cotŁ < V Rd,max
6: bd d s

f ck
V Rd,s ¼ 0.9Asw f yd dcotŁ=s < V Rd,max V Rdmax ¼ k c bw z sin Ł cos Ł
7: 14: ªc

8: V Rd,max ¼ 0.9bw d1 f cd =(cotŁ þ tan Ł) 15: Łmin ¼ 20o þ 1000x

7
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

 1=3
1 30 the measured to calculated failure loads (Ptest /Pcalc) are presented
k c ¼ k  fc ¼ . in Table 2 for each shear span, with the shear span in which
1 2 þ 551 f ck
failure occurred highlighted in bold. Table 2 shows that all the
 1=3
1 30 predictions are safe but the relative accuracy of each method
where . < 0.65 and <1
16: 1 2 þ 551 f ck depends on the loading arrangement and presence or otherwise of
shear reinforcement.

Strut-and-tie modelling
17: 1 ¼ x þ (x þ 0.002)cot2 Ł Strut-and-tie modelling is a powerful technique for the design
and analysis of D regions like beam–column joints (Vollum and
Newman, 1999) and deep beams for which it is a practical
MC2010 level 3: V Rd ¼ V Rd,c þ V Rd,s where all the terms are as alternative to NLFEA-based design procedures such as that of
defined above except Amini Najafian et al. (2013). Both MC2010 and Eurocode 2
allow STM to be used for modelling shear enhancement near
 
0 .4 V Ed supports. Both codes give design concrete strengths for struts,
kv ¼ 1 >0
18: 1 þ 1500x V Rd,max (Łmin ) with and without transverse tension, and nodes which are
summarised in Table 4. Figure 4 shows the geometry of the
proposed STM with and without shear reinforcement. Only
essential details are given here as the development of the STM is
Evaluation of shear enhancement methods of described in detail elsewhere (Fang, 2014; Vollum and Fang,
MC2010, BS 8110 and Eurocode 2 2014). The stirrup forces in Figures 4(a) and 4(b) are subdivided
The shear enhancement methods of MC2010, BS 8110 and into components Ts1 and Ts2 as shown and CIII, CIV and CVI are
Eurocode 2 were used, with ªc ¼ ªs ¼ 1 .0, to predict the strengths the horizontal components of force resisted by struts III, IV and
of the authors’ beams. The Eurocode 2 strengths were calculated VI. The STM geometry depends on the magnitude of P2 relative
with Equation 4 as Equation 1 incorrectly predicts the strengths to Ts1 + Ts2 as shown in Figure 4. STMa in Figure 4(a) ceases to
of beams with shear reinforcement to be less under loading be applicable when Ts1 + Ts2 . P2 because the vertical component
arrangement C with four point loads than loading arrangement B of force in strut VI is no longer sufficient to balance Ts1 + Ts2. In
with two. Equation 4 avoids this illogicality, but incorrectly gives this case, STMb of Figure 4(b) should be used. Figure 4(c) shows
the same shear enhancement for load cases B and C as the geometry of the STM when stirrups are not present. Failure is
1 =nav1 ¼ 2 =nav2 , whereas in reality the enhancement was great- assumed to occur in flexure, shear or bearing provided the
er for the tested beams under loading combination C than B. In flexural reinforcement is sufficiently anchored. Shear failure is
order to compare the respective merits of the shear enhancement assumed to occur due to combined yielding of the stirrups and
methods in BS 8110 and Eurocode 2, the BS 8110 strengths were crushing of strut III at its bottom end. As strut VII is fan shaped,
calculated using Equation 5 with VRd,c from Equation 6 of it is assumed to be adequate provided that the bearing stress at
Eurocode 2. The maximum possible shear resistance V Rdmax was the bottom node does not exceed the Eurocode 2 limit of 0 .85vfcd
calculated in accordance with PD 6687-1 (BSI, 2010), but was for C–C–T nodes, in which case bearing failure is critical. The
not critical for any of the tested beams. All the stirrups were bearing stress is not checked at the back of the bottom node as
assumed to yield on the basis of strain measurements. Ratios of this is not required by Eurocode 2.

C–C–C node C–C–T node Strut III C–C–T node interface Strut III C–C–C node
interface
       
f ck f ck f ck f ck f f ck f ck f ck
STM-Eurocode 2 1 0.85 1  0.60 1  ck 1
250 ªc 250 ªc 250 ªc 250 ªc
STM-MCFTa 0.85 f ck 0.85 f ck  f ck =(0.8 þ 1701 ) < 0.85 f ck b 0.85 f ck
 1=3  1=3  1=3  1=3
30 f ck 30 f ck 30 f ck c 30 f ck
STM-MC2010 1.0 0.75 0.55 1.0
f ck ªc f ck ªc f ck ªc f ck ªc
a
 is a capacity reduction factor with a recommended design value of 0 .65.
b
1 ¼ (L þ 0.002)cot2 Ł,where Ł is defined in Figure 4 and L is the strain in the flexural reinforcement corresponding to T ¼ T9i þ T d :
c
Applicable for strut angles Ł , 658.

Table 4. Design concrete strengths used in STM

8
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Derivation of design equations for STM X

cotŁ ¼ ¼
The design equations are derived from considerations of geo- Y
metry and equilibrium following the procedure adopted by a1 þ W  0.5lb (1  º)
Sagaseta and Vollum (2010) for beams with single concentrated
24: h  fT s1 cotÆ þ [0.5ºT s1 =(1  º)]cotŁg=bf cnt  c
loads within 2d of supports. Only essential details are given here
as the model is described in more detail elsewhere (Vollum and
Fang, 2014). The design shear resistance is defined in terms of
the stirrup force Ts1 as follows in which the dimensions e1, X, Y and W are defined in Figure 4.
W is the horizontal distance between P1 and the intersection of
the centreline of strut III with the line of action of CIII. If
T s1 P2 > Ts1 + Ts2, consideration of geometry (see Figure 4(a)) and
V Rd ¼ P1 þ P2 ¼ P1 (1 þ k) ¼
19: (1  º) equilibrium leads to

fT s1 [º  1=(1 þ k)]  T s2 (1  º)g

where ºVRd is the vertical component of force in strut III at  
0.5lt (1  º)(T s1 þ T s2 )(1 þ k)
failure, k ¼ P2 =P1 and Ts1 is the yield capacity of the effective a2  a1 
kT s1
stirrups within av1.
þT s2 (1  º)(e2  a1 )
W¼
Consideration of horizontal equilibrium at the bottom node leads 25: ºT s1
to the following relationships

in which the dimensions e1 and e2 define the positions of the

  centrelines of the stirrup forces Ts1 and Ts2 as shown in Figure
T d ¼ T ¼ T i9 ¼ T s1 cotÆ
20: 1 1 4(a).

If P2 , Ts1 + Ts2, consideration of Figure 4(b) leads to

º [º(1 þ k)  1](e2  a1 )
T d ¼ ºV Rd cotŁ ¼ T s1 cotŁ W¼
21: 1º 26: º(1 þ k)

At failure, the horizontal component of force in strut III is given by

in which T ¼ T d þ T i9 is the force in the flexural reinforcement at
the bottom node, Td and T i9 are the horizontal components of
27: T d ¼ bf csb (ºlb sin Ł þ 2c cos Ł) cos Ł
force in struts III and VII, respectively, and Æ,  and Ł are
defined in Figure 4.
where fcsb is the concrete strength at the bottom end of strut III.
Equating Equations 20 and 21 and rearranging leads to the
following equation for cotÆ Equating Equations 20 and 27 for Td and rearranging gives

 
º 1 Y þ Y cot2 Ł  ºlb
cotÆ ¼ cotŁ 28:
¼ <1
22:  1º 2ccotŁ

in which
The angles Æ and Ł can be calculated from geometry in terms of
Ts1, Ts2, , º and the assumed flexural compressive stress fcnt as
follows º T s1
Y ¼
29: (1  º) bf csb

e1 þ 0.5ºlb
cotÆ ¼
23: h  0.5T s1 cotÆ=bf cnt  c(1 þ ) In the case of beams without shear reinforcement (see Figure
4(c)), strut VII vanishes, º ¼  ¼ 1 and

9
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

a þ ka2
cotŁ ¼ 1  ment was modelled with embedded elements in which reinforce-
0.5T d ment strains are calculated from the displacements of the
(1 þ k) h  c 
30: bf cnt concrete element assuming perfect bond (TNO, 2007). The
reinforcement was conservatively assumed to be a von Mises
perfectly elasto-plastic material with the yield strengths fy shown
where Td is given by Equation 27. in Table 3. Following a mesh sensitivity study (Fang, 2014), ten
elements of equal depth were used through the beam depth in
Design concrete strengths both the 2D and 3D analyses. The elements were approximately
The design strength at the bottom end of strut III was calculated 50 mm square in elevation but their exact width was varied to
in accordance with MC2010, Eurocode 2 and the modified give a whole number of equal width elements between the edges
compression field theory (MCFT) of Collins et al. (2008). The of the load and support plates. Three equal width elements were
corresponding design concrete strengths are given in Table 4. The used through the beam thickness in the 3D analyses. It was found
upper limit of 0 .85fck in the MCFT is applicable to the ends of necessary to locally increase the concrete strength adjacent to the
struts not crossed by tension ties. The flexural compressive stress loading plates in the 2D analyses of beams without flexural
fcnt was taken as the relevant C–C–C node strength from Table 4. compression reinforcement to prevent premature failure. The
All the STM strength assessments were carried out with strength was locally increased by a multiple of three, above which
ªc ¼ ªs ¼  ¼ 1 .0. further enhancement had no effect on the calculated failure load
(Fang, 2014).
Solution procedure for STM
The STM geometry, and hence failure load, can be calculated Table 5 gives values of Ptest /Pcalc obtained using 2D and 3D
once º, cotŁ and fcsb are known. The stress fcsb is predefined in NLFEA. Both analyses were reasonably accurate, with little
STM–Eurocode 2 and STM–MC2010, but depends on the flexural advantage being gained from 3D NLFEA which was much more
reinforcement strain and cotŁ in STM–MCFT. The unknowns º time consuming than 2D NLFEA. Figures 5 and 6 show typical
and cotŁ can be found from Equations 22–24 and 28 using a non- comparisons between measured and predicted reinforcement
linear equation solver like the generalised reduced gradient strains. The agreement is reasonably good and shows that the
(GRG2) non-linear solver in Microsoft Excel (Microsoft, 2014). NLFEA was able to simulate aspects of the observed structural
The first step is to evaluate the right-hand sides of Equations 22– response. For example, Figure 6 shows that the positions of the
24 and 28 in terms of assumed values for ºest, cotŁest and fcsb est. peak measured and predicted stirrup strains overlap. Fang (2014)
The GRG2 solver can then be used to find the values of ºest, also showed that the peak predicted strains occurred at or near
cotŁest and fcsb est at which the assumed and calculated (e.g. right- the intersection of the stirrups with the critical shear crack as
hand sides of Equations 22–24 and 28) values converge. observed. The measured strains in Figure 6 show that all the
stirrups yielded in beam S1-2, which was typical.
Non-linear finite-element analysis
The response of the tested beams was simulated with NLFEA Evaluation of STM and comparison with other
(Fang, 2014). Both two-dimensional (2D) and three-dimensional methods
(3D) analyses were carried out with Diana (TNO, 2007) using its Failure loads were calculated for all the tested beams using the
fixed crack total strain model. The concrete compressive behav- STM. The failure loads of beams S1-2 and S2-2 with two point
iour was modelled with the parabolic stress–strain relationship of loads and stirrups were calculated assuming Ts1 ¼ Ts2 with
Feenstra (1993), which relates the compressive stress after failure k ¼ P2 =P1 ¼ 1000: The distance a1 to P1 was taken as 475 mm,
to the compressive fracture energy Gc – taken as 100 times the as in tests S1-1 and S2-1, but the resulting failure loads are
tensile fracture energy Gf. The maximum concrete compressive insensitive to this assumption. All the stirrups were assumed to
stress was related to the principal tensile strain in accordance be effective. Both shear spans were analysed even though the
with the recommendations of Vecchio and Collins (1993). The shear span with the narrowest bearing plate is predicted to be
effect of confinement was modelled in the 3D NLFEA using the critical in all cases. In fact, half of the 12 beams failed in the
approach of Selby and Vecchio (1997). Tension softening was shear span supported by the widest bearing plate. This is
modelled with the Hordijk (1991) model, which relates the tensile significant since it suggests that the strategy of increasing shear
stress after cracking to Gf, which was calculated using Model strength by increasing bearing plate width could lead to unsafe
Code 1990 (CEB, 1993). Poisson’s ratio was taken as zero, as designs. The results are given in Table 5, which shows that the
proposed by Vecchio and Collins (1993) for the MCFT. The STM–MCFT predictions are generally better than those of
pffiffiffiffiffiffiffi
concrete tensile strength was taken as 0 .33 f ck as adopted by STM–Eurocode 2 and STM–MC2010, which are almost identi-
Sagaseta and Vollum (2010). The shear retention factor was cal. The STM–Eurocode 2 and STM–MC2010 predictions are
determined by sensitivity studies on the tested beams and was unsafe for all the beams without shear reinforcement except the
taken as 0 .25 in the 2D analyses and 0 .07 in the 3D analyses. B2 beams with av1 /d , 0 .7 and beam A-1. Table 5 also gives
The concrete elastic modulus was calculated in terms of the the strength predictions obtained with 2D and 3D NLFEA. The
concrete compressive strength with Eurocode 2. The reinforce- STM–MCFT predictions are seen to compare favourably with

10
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Beam P test : kN Ptest /Pcalc

STM-Eurocode 2 STM-MCFT STM-MC2010 NLFEA

La Ra La Ra La Ra 2D 3D

B1-25 368 0 .74 0 .82 1 .16 1 .22 0 .75 0 .84 0 .99 1 .04
B1-50 352 0 .61 0 .65 1 .06 1 .08 0 .62 0 .66 0 .98 1 .06
B2-25 977 1 .09 1 .25 1 .15 1 .20 1 .11 1 .28 1 .04 1 .28
B2-50 929 0 .92 1 .02 1 .07 1 .09 0 .94 1 .04 1 .05 1 .20
A-2 349 0 .72 0 .89 1 .06 1 .26 0 .70 0 .87 0 .94 1 .01
S1-2 601 0 .89 1 .04 1 .03 1 .14 0 .86 1 .03 0 .96 1 .02
S2-2 820 1 .07 1 .23 1 .17 1 .28 1 .04 1 .22 1 .25 1 .18
B3-25 480 0 .73 0 .83 0 .95 1 .00 0 .75 0 .84 0 .79 1 .00
B3-50 580 0 .78 0 .84 1 .12 1 .13 0 .79 0 .85 1 .06 0 .99
A-1 823 1 .37 1 .76 1 .63 1 .98 1 .32 1 .71 1 .28 1 .33
S1-1 1000 1 .22 1 .55 1 .22 1 .44 1 .20 1 .52 1 .32 1 .39
S2-1 1179 1 .31 1 .64 1 .27 1 .47 1 .29 1 .62 1 .43 1 .35
a
Calculated for right (R)/left (L) shear span as defined in Figure 3 (bold type denotes critical shear span in test)

Table 5. Measured and predicted STM failure loads

2·0
the NLFEA predictions, which depend significantly on user-
1·6 defined input parameters of which the shear retention factor has
the greatest effect. Table 6 presents a statistical analysis of the
accuracy of the calculated strengths of the tested beams for
εs: ⫻103 με

1·2
STM, NLFEA and codified sectional methods. BS 8110 and
MC2010 give more consistent values of Ptest /Pcalc than Eurocode
0·8
Test 2 for beams with shear reinforcement. This is partly because
both methods, unlike Eurocode 2, predict greater strengths for
0·4 NLFEA the beams with four point loads than the corresponding beams
with two point loads.
0
0 125 250 375 500 625 750 875
Parametric studies
Distance from centre of beam: mm
A series of parametric studies was carried out on beams with four
Figure 5. Comparison between predicted and experimental tensile point loads to investigate the influence of varying the loading
strains along the flexural reinforcement in beam S1-1 at 645 kN ratio k ¼ P2/P1 on shear strength. The aim of the study was to
investigate loading arrangements not considered in the tests. The

Stirrup 4 Stirrup 3 Stirrup 2 Stirrup 1

500 Beam S1-2
2·69 0·05 0·00 P ⫽ 550 kN
400

300 Demec
0·23 2·04 5·56 ⫺0·03 NLFEA 2D
200 NLFEA 3D
0·07 2·09
100 0·89 3·78

0
Figure 6. Comparison between measured and predicted stirrup
strains in right-hand shear span of beam S1-2 at P ¼ 550 kN
(Pu ¼ 601 kN) (note: shaded numbers denote Demec strains
3103)

11
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Method Sectional STM- NLFEA

Beams Statistic Eurocode 2 BS 8110 MC2010 Eurocode 2 MCFT MC2010 2D 3D

All Mean 1 .76 1 .57 2 .07 1 .13 1 .27 1 .12 1 .09 1 .16
COV 0 .28 0 .19 0 .43 0 .32 0 .21 0 .31 0 .17 0 .13
With stirrups Mean 1 .43 1 .39 1 .26 1 .36 1 .33 1 .35 1 .24 1 .23
COV 0 .28 0 .07 0 .18 0 .21 0 .12 0 .20 0 .16 0 .14

Table 6. Statistical analysis of Vtest /Vcalc for tested beams

analyses were carried out on beams with the same geometry, 400
reinforcement and loading positions as beams B3-25 with 350 STM–MCFT
fck ¼ 45 .7 MPa (see Figure 1), S1-1 with fck ¼ 36 MPa (see STM–Eurocode 2
300
V ⫽ P1 ⫹ P2: kN

STM–MC2010
Figure 2(b)) and S2-1 with fck ¼ 36 MPa (see Figure 2(c)). The 250 NLFEA 2D
B3-25
loading positions are given in Table 1, which should be read in 200 NLFEA 3D
Test
conjunction with Figure 3. 150 B1-25
100
Failure loads were initially obtained with 2D and 3D NLFEA 50
using the same finite-element meshes, material parameters and
0
solution procedure as used for the tested beams. The resulting 0 1 2 3 4
shear strengths are compared with those given by STM in Figure P1/P2
7, along with relevant test data from this programme. Figure 7 (a)
shows that of the STM, STM1–MCFT gives the best estimates 600
of the measured beam strengths as well as simulating the
500 S1-1 STM–MCFT
observed increase in strength with P1/P2 most accurately. The STM–Eurocode 2
V ⫽ P1 ⫹ P2: kN

NLFEA predictions are comparable but less consistent than the 400 S1-2 STM–MC2010
STM–MCFT predictions which seem better. The 3D NLFEA is NLFEA 2D
300 NLFEA 3D
slightly more conservative than the 2D NLFEA, which over-
Test
estimates the strength of beam B3-25. More conservative 2D 200
NLFEA strength predictions could be achieved by reducing the
100
shear retention factor  from its adopted value of 0 .25 but this
was not done as  was chosen to give a reasonable fit to all the 0
test results. 0 1 2 3 4
P1/P2
Figure 8 compares the code predictions with STM–MCFT, which (b)
is the best of the STMs, and relevant test data. The BS 8110 700
predictions are seen to be closest to the STM–MCFT predictions. 600 S2-1 STM–MCFT
The BS 8110 predictions become independent of P2/P1 above a STM–Eurocode 2
V ⫽ P1 ⫹ P2: kN

500
threshold value of P2/P1 because failure plane P1 in Figure 3(c) S2-2 STM–MC2010
400 NLFEA 2D
becomes critical. Failure loads were also calculated with Equation
NLFEA 3D
4 using VRdav1 and VRdav2 from Equations 5(a) and 5(b) of BS 300
Test
8110. Figure 8 shows that the resulting predictions (BS 8110 200
Equation 4) are acceptable but more conservative than given by
100
BS 8110. The MC2010 shear enhancement method works reason-
ably well for the beams with shear reinforcement (level 3) but is 0
0 1 2 3 4
overly conservative for beams without shear reinforcement (level
P1/P2
2). Three Eurocode 2 predictions are given for beams with shear
(c)
reinforcement which are denoted Eurocode 2 Equation 1, Euro-
code 2 Equation 2 and Eurocode 2 Equation 4 in Figure 8. The Figure 7. Influence of load ratio P1 /P2 on shear strengths
predictions correspond to the shear enhancement methods of calculated with STM for: (a) beam B3-25; (b) beam S1-1;
equations 1 and 2 of MC2010 with VRd ¼ max(V Rd,c from (c) beam S1-2
Equation 6, V Rd,s from Equation 7) and Equation 4 with VRdav1

12
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

350
Comparison with other test results
300 The analyses in this paper show that the Eurocode 2 method of
BS 8110
B3-25 shear enhancement requires improvement for beams with shear
250 Eurocode 2
V ⫽ P1 ⫹ P2: kN

B1-25 MC2010 reinforcement. There is a lack of data to properly assess these

200 methods for beams with multiple point loads within 2d of supports.
STM–MCFT
150 Test Therefore, potential improvements to the Eurocode 2 method of
shear enhancement were assessed with data from 57 beams
100
(Lehwalter (1988) (one beam); Regan (1971) (six beams); Tan et
50 al. (1995) (one beam); Tan et al. (1997) (two beams); Vollum and
0 Tay (2001) (two beams); Clark (1951) (27 beams); Kong and
0 1 2 3 4 Rangan (1998) (two beams); Sagaseta and Vollum (2010) (six
P1/P2 beams); Birrcher et al. (2013) (eight beams) and the authors’
(a) beams S1-2 and S2-2) with av/d between 1 and 2, shear reinforce-
600 BS 8110 ment and single point loads within 2d of the supports. All the
BS 8110 beams had just vertical stirrups except those of Birrcher et al.
500 S1-1 Equation 4
(2013), which also had horizontal web reinforcement. Further
Eurocode 2
V ⫽ P1 ⫹ P2: kN

400 Equation 1
details of all these beams except those of Birrcher et al. (2013) and
S1-2 Eurocode 2 this programme can be found in Table 3 of Sagaseta and Vollum
300 Equation 2 (2010). The enhanced shear strengths Vcalc of these beams were
Eurocode 2 calculated using the following five methods, with V Rd,c from
200 Equation 4 Eurocode 2, as well as STM–MCFT and MC2010.
MC2010
100 STM–MCFT
Test BS 8110:
0
0 1 2 3 4
P1/P2 V calc ¼ (V Rd,c þ 0.5nav Asw f yd )=
(b)
700 BS 8110
Eurocode 2 Equation 1(b):
BS 8110
600 S2-1 Equation 4
500 Eurocode 2 V calc ¼ max(V Rd,c , nav Asw f yd )=
V ⫽ P1 ⫹ P2: kN

Equation 1
S2-2
400 Eurocode 2
Equation 2
Eurocode 2
Eurocode 2 Equation 2:
300
Equation 4
200 MC2010 V calc ¼ max (V Rd,c , V Rd,s from Equation 7)=
STM–MCFT
100
Test
0 Eurocode 2 M1:
0 1 2 3 4
P1/P2
(c) V calc ¼ (V Rd,c þ nav Asw f yd )=

Figure 8. Influence of load ratio P1 /P2 on shear strengths

calculated with STM–MCFT and code based methods for: Eurocode 2 M2:
(a) beam B3-25; (b) beam S1-1; (c) beam S1-2
V calc ¼ max (V Rd,c =, V Rd,s from Equation 7)

and VRdav2 from Equations 1(a) and 1(b), respectively. The spike The STM failure loads were calculated with Ts1 ¼ Ts2, P1
in the Eurocode 2 Equation 1 results of Figures 8(b) and 8(c) positioned at the centre of the clear shear span and
arises because Equation (1a) governs if P1 /P2 . 0 but Equation k ¼ P2 =P1 ¼ 1000: The resulting failure loads are similar to
(1b) if P1 /P2 ¼ 0. Eurocode 2 Equation 2 gives better predictions those given by the STM of Sagaseta and Vollum (2010) for
than Eurocode 2 Equation 1, but they are relatively high com- beams with one point load within 2d of supports. Eurocode 2
pared with the MC2010 and STM1–MCFT predictions. Eurocode Equation 2 uses equation 2 of MC2010 but VRd is calculated with
2 Equation 4 predicts shear resistance to be independent of P1 /P2 Eurocode 2. Eurocode 2 M1 is given in the Eurocode 2
since 1 =nav1 ¼ 2 =nav2 for beams S1-1 and S1-2. commentary (European Concrete Platform ASBL, 2008) but was

13
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

3·0 Mean 1·23 Mean 1·41

3·0
5% 0·91 5% 0·94
2·5 STM-MCFT 2·5 BS 8110

2·0 R2 ⫽ 0·02 Linear (BS 8110)

2·0

Vtest / Vcalc
R2 ⫽ 0·39
Vtest / Vcalc

1·5 1·5

1·0 1·0

0·5 0·5

0 0
0 5 10 15 0 5 10 15
100Asw f y /(fcbs) 100Asw fy /(fcbs)
(a) (b)
3·0 Mean 1·73 3·0 Mean 1·08
5% 0·87 5% 0·40
2·5 Eurocode 2 Equation 1(b) 2·5 EC2 Equation 2
Linear (Eurocode 2
Linear (EC2 Equation 2)
2·0 Equation 1(b)) 2·0
R2 ⫽ 0·41
R2 ⫽ 0·56
Vtest / Vcalc
Vtest / Vcalc

1·5 1·5

1·0 1·0

0·5 0·5

0 0
0 5 10 15 0 5 10 15
100Asw fy /(fcbs) 100Asw fy /(fcbs)
(c) 3·0 (d)
3·0 Mean 1·02 Mean 1·38
5% 0·60 5% 0·51
2·5 Eurocode 2 M2
2·5 Eurocode 2 M1
Linear (Eurocode 2 M2)
Linear (Eurocode 2 M1) 2·0
2·0
Vtest / Vcalc

R2 ⫽ 0·65
Vtest / Vcalc

2
R ⫽ 0·53 1·5
1·5

1·0 1·0

0·5 0·5

0 0
0 5 10 15 0 5 10 15
100Asw fy /(fcbs) 100Asw fy /(fcbs)
(e) (f)
3·0
Mean 1·15
5% 0·66
2·5
MC2010 Level 3
2·0 Linear (MC2010 Level 3)
R2 ⫽ 0·59
Vtest/Vcalc

1·5

1·0

0·5

0
0 5 10 15
100Aswfy /(fcbs)
(g)
Figure 9. Comparison between measured and calculated shear
strengths for: (a) STM–MCFT; (b) BS8110; (c) Eurocode 2
Equation 1(b); (d) Eurocode 2 Equation 2; (e) Eurocode 2 M1;
(f) Eurocode 2 M2; (g) MC2010 level 3

14
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

omitted from the final version of the code. Like Equations 1 and width. The Eurocode 2 method of shear enhancement is adequate
5 of Eurocode 2 and BS 8110, respectively, Eurocode 2 M2 for beams without shear reinforcement but not for beams with
applies no enhancement to the shear resistance provided by the shear reinforcement, where it fails to predict the influence of
shear reinforcement, which is calculated using the standard loading arrangement. A STM is presented which is shown to give
Eurocode 2 method for beams. Vtest/Vcalc is plotted against the reasonable predictions of shear resistance of the tested beams,
shear reinforcement ratio 100Asw f y =(bsf c ) for each design meth- with the best predictions being obtained if the direct strut strength
od in Figure 9, which also shows the trend between the two. The is calculated in accordance with the recommendations of the
correlation between Vtest/Vcalc and 100Asw f y =(bsf c ) is reflected in MCFT. The STM–MCFT and NLFEA predictions are shown to
the R2 value, which is 1 for perfect correlation. The figures also be of comparable accuracy but the STM has the advantage of not
show that, as well as being most accurate, the STM–MCFT requiring calibration of input parameters. The BS 8110 method
predictions are the only ones that are independent of of modelling shear enhancement was found to be significantly
100Asw f y =bsf c : BS 8110 performs best of the sectional methods better for beams with shear reinforcement than the methods of
and significantly better than the current Eurocode 2 method Eurocode 2 and MC2010, with the advantages particularly
(Equation 1(b)), which is overly conservative at low reinforce- evident for beams with multiple point loads. Consequently, it is
ment ratios. Eurocode 2 Equation 2, Eurocode 2 M1, Eurocode 2 suggested that Equation 5 of BS 8110 is used to calculate the
M2 and MC2010 all overestimate shear resistance at high shear shear resistance V Rdav in future revisions of Eurocode 2. Multiple
reinforcement ratios. This is even the case for Eurocode 2 M2 in point loads are most easily handled using the BS 8110 approach
which no enhancement is applied if the shear resistance is of enhancing shear resistance. However, Equation 4 also gives
governed by shear reinforcement. With the exception of Eurocode reasonable results as shown in Figure 8.
2 M1, the overestimate arises in part because these methods fail
to acknowledge that shear reinforcement is only effective if Acknowledgements
placed within the central three-quarters of the clear shear span av. The authors would like to acknowledge the financial support of
The overestimate is also consistent with the previous finding of The Concrete Centre, UK and the technical assistance of the staff
Sagaseta and Vollum (2011b) that Eurocode 2 overestimates the of the Structures Laboratory at Imperial College London.
shear strength of beams with rectangular cross-sections and high
shear reinforcement ratios. Table 7 presents a statistical analysis REFERENCES
which confirms that the STM–MCFT and BS 8110 methods are Amini Najafian H, Vollum RL and Fang L (2013) Comparative
most accurate. assessment of finite element and strut and tie based design
methods for deep beams. Magazine of Concrete Research
Conclusions 65(16): 970–986, http://dx.doi.org/10.1680/macr.13.00006.
Twelve beams were tested to investigate the effects of loading Birrcher DB, Tuchscherer RG, Huizinga M and Bayrak O (2013)
arrangement and shear reinforcement ratio on the relative accura- Minimum web reinforcement in deep beams. ACI Structural
cies of the shear enhancement methods in MC2010, BS 8110 and Journal 110(2): 297–306.
Eurocode 2. Different support widths were adopted at each end of Brown MD and Bayrak O (2007) Investigation of deep beams
the beams to determine whether failure occurred on the side of with various load, configurations. ACI Structural Journal
the narrower support as predicted by STM. Notably, half the 104(6): 11–20.
beams failed on the side of the wider support, which casts doubt BSI (1997) BS 8110. Part 1: Structural use of concrete: code of
on the realism of the underlying assumptions of STM and the practice for design and construction. BSI, London, UK.
strategy of increasing shear resistance by increasing support BSI (2004) European Standard EN-1992-1-1: 2004. Eurocode 2:

Methoda STM MCFT BS 8110 Eurocode 2 Eurocode 2 Eurocode 2 Eurocode 2 MC2010 MC2010
Equation 1(b) Equation 2 M1 M2 level 2 level 3

Mean 1 .23 1 .41 1 .73 1 .08 1 .02 1 .38 1 .34 1 .15

COV 0 .16 0 .20 0 .30 0 .39 0 .25 0 .39 0 .38 0 .26
Characteristicb 0 .91 0 .94 0 .87 0 .40 0 .60 0 .51 0 .51 0 .66
a
Short-span beams used were 47 beams from Table 3 of Sagaseta and Vollum (2010), beams S1-2 and S2-2 from this program and eight beams
from Birrcher et al. (2013).
b
5% lower characteristic strength.

Table 7. Statistical analysis of Vtest /Vcalc for 57 short-span beams

with shear reinforcement and 1 , av /d < 2

15
Magazine of Concrete Research Shear enhancement near supports in RC
beams
Vollum and Fang

Design of concrete structures. Part 1, general rules and rules Sagaseta J and Vollum RL (2010) Shear design of short-span
for buildings. BSI, London, UK. beams. Magazine of Concrete Research 62(4): 267–282,
BSI (2010) PD 6687-1. Background paper to the National Annexes http://dx.doi.org/10.1680/macr.2010.62.4.267.
to BS EN 1992-1 and BS EN 1992-3. BSI, London, UK. Sagaseta J and Vollum RL (2011a) Influence of aggregate fracture
CEB (1993) CEB-FIP Model Code 1990. Thomas Telford, on shear transfer through cracks in reinforced concrete.
London, UK. Magazine of Concrete Research 63(2): 119–137, http://
Clark AP (1951) Diagonal tension in reinforced concrete beams. dx.doi.org/10.1680/macr.9.00191.
Journal of the American Concrete Institute 23(2): 145–156. Sagaseta J and Vollum RL (2011b) Influence of beam cross-
Collins MP, Bentz EC, Sherwood EG and Xie L (2008) An section, loading arrangement and aggregate type on shear
adequate theory for the shear strength of reinforced concrete strength. Magazine of Concrete Research 63(2): 139–155,
structures. Magazine of Concrete Research 60(9): 635–650, http://dx.doi.org/10.1680/macr.9.00192.
http://dx.doi.org/10.1680/macr.2008.60.9.635. Selby R and Vecchio F (1997) A constitutive model for analysis of
European Concrete Platform ASBL (2008) Commentary Eurocode reinforced concrete solids. Canadian Journal of Civil
2. European Concrete Platform, Brussels, Belgium. Engineering 24(3): 460–470.
Fang L (2014) Shear Enhancement in Reinforced Concrete Tan KH, Kong FK, Teng S and Guan S (1995)
Beams. PhD thesis, Imperial College London, London, UK. High-strength concrete deep beams with effective span
Feenstra PH (1993) Computational Aspects of Biaxial Stress in and shear span variations. ACI Structural Journal 92(4): 395–
Plain and Reinforced Concrete. PhD, Delft University of 405.
Technology, Delft, the Netherlands. Tan KH, Kong FK, Teng S and Weng LW (1997) Effect of web
fib (Fédération Internationale du Béton) (2013) fib Model Code reinforcement on high-strength concrete deep beams. ACI
for Concrete Structures 2010. Ernst & Sohn, Germany. Structural Journal 94(5): 572–582.
Hordijk DA (1991) Local Approach to Fatigue of Concrete. PhD TNO (2007) Finite Element Analysis Users Manual Release 9.2.
thesis, Delft University of Technology, Delft, the Netherlands. TNO Diana, Delft, the Netherlands.
Jackson P, Salim S, Shave J and Denton S (2007) Shear Vecchio FJ and Collins MP (1993) Compression response of
enhancement at short shear spans in EN 1992. The Structural cracked reinforced concrete. Journal of Structural
Engineer 85(23/24): 37–42. Engineering ASCE 119(12): 3590–610.
Kong YL and Rangan BV (1998) Shear strength of high- Vollum RL and Fang L (2014) Shear enhancement in RC beams
performance concrete beams. ACI Structural Journal 94(6): with multiple point loads. Engineering Structures 80: 389–
677–688. 405.
Lehwalter N (1988) The Bearing Capacity of Concrete Vollum RL and Newman JB (1999) Towards the design of
Compression Struts in Strut and Tie Models with Examples of eccentric beam-column joints. Magazine of Concrete
Deep Beams. PhD thesis, Technical University of Darmstadt, Research 51(6): 397–407, http://dx.doi.org/10.1680/
Germany. macr.1999.51.6.397.
Microsoft (2014) Article 82890. See http:// Vollum RL and Tay UL (2001) Strut and tie modelling of shear
support.microsoft.com/kb/82890 (accessed 29/06/2014). failure in short-span beams. Proceedings of the Concrete
Regan PE (1971) Shear in Reinforced Concrete – An Communication Conference, UMIST, Manchester. British
Experimental Study. Construction Industry and Information Cement Association and the Concrete Society, Camberley,
Association (Ciria), London, UK. Surrey, UK, pp. 193–199.

WHAT DO YOU THINK?

To discuss this paper, please submit up to 500 words to
the editor at journals@ice.org.uk. Your contribution will
be forwarded to the author(s) for a reply and, if
considered appropriate by the editorial panel, will be
published as a discussion in a future issue of the journal.

16
View publication stats