Structures 20 (2019) 214–225

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Bending and shear behaviour of two layer beams with one layer of rubber T
recycled aggregate concrete in tension
⁎
R.B. Ataria, Y.C. Wang
University of Manchester, UK

A R T I C LE I N FO A B S T R A C T

Keywords: This paper presents the results of an experimental, numerical and analytical study of bending and shear beha-
Recycled aggregate viour of simply supported reinforced concrete beams with two layers of different grades of concrete. The top
Crumb rubber layer (1/3rd) concrete, mainly in compression, is higher grade, and the bottom (2/3rd) layer, in tension, is lower
Rubber recycled aggregate concrete grade using rubber recycled aggregate concrete. A total of 8 tests were conducted. The results confirm that the
Two layer reinforced concrete beam
two layer beam achieves the same bending resistance as the control reinforced concrete beam made entirely of
Bending resistance
Shear resistance
the higher grade concrete.
In beams without shear reinforcement, the two layer beam attained lower shear resistance than the control
beam. This was attributed to the unzipping effect: once the lower strength concrete has failed in shear, it loses its
shear resistance and transfers the shear force from the failed lower strength concrete to the control concrete.
Further FE simulations reveal that the top layer higher grade concrete plays no role in influencing the beam's
shear resistance. Therefore, when calculating shear resistance of concrete of the two layer beam, the lower
concrete grade should be used. However, since the shear force in reinforced concrete beams is mainly resisted by
shear links, the lower shear resistance of rubber recycled aggregate concrete has very minor implication on beam
shear resistance.
The proposed construction facilitates wider adoption of recycled concrete without having to undergo ex-
pensive processes to achieve the same mechanical properties of concrete made of fresh aggregates.

1. Introduction stockpiled in landfills.

Even where the rate of recycling solid wastes is high, applications in
The issue associated with disposal of solid waste from the con- the construction industry are limited to non-structural members such as
struction industry is alarming, especially in developed countries. For fences, road curbs and road base filling materials [6] etc. One main use
instance, it is estimated that about 180 million tons of construction and of recycled solid wastes is as recycled aggregates in concrete. There
demolition waste (CDW) are generated every year in the European have been numerous research studies on mechanical properties of re-
Union EU [1]. Of this volume of CDW generated, only 28% is being cycled aggregate concrete and one of the reasons for a lack of use of
reused for other purposes while 72% is conveyed to landfills or in- recycled materials (recycled aggregates, crumb rubber) in structural
cinerated. Only countries with limited land like the Netherlands and members is the difficulty of using recycled solid wastes to produce
Japan recycle solid wastes to a high level of about 90% and 98% re- concrete having the same properties as concrete made from fresh ma-
spectively [2,3]. Discarded tyres from vehicles are another source of terials. For instance poor workability at the wet stage is observed
solid waste causing serious environmental problems all over the world. [7–10], so are poor mechanical properties (compressive and tensile
About one billion scrap tyres are generated every year around the strength) due to low quality of recycled aggregates [11,12]. Weak in-
globe, out of which 280 million is estimated in Europe alone [4,5]. terfacial zone between the recycled aggregate and the surrounding
About 60% of these tyres are being recovered by means of road and concrete [13–16] is also reported. Incorporating rubber particles in
landfill construction, reuse, recycling, energy recovery and asphalt concrete mix is a possibility [17–30]. However, the aforementioned
rubber modified concrete. The cheapest way of discarding these waste research studies again reveal lower strength and lower elastic modulus
tyres would be to burn them but this causes pollution due to the amount when rubber particles from motor vehicles are used in concrete.
of smoke generated and as such is prohibited in most countries, or The other reason why concrete made of solid waste (recycled

⁎
Corresponding author.
E-mail address: yong.wang@manchester.ac.uk (Y.C. Wang).

https://doi.org/10.1016/j.istruc.2019.03.014
Received 9 October 2018; Received in revised form 19 March 2019; Accepted 26 March 2019
2352-0124/ © 2019 Published by Elsevier Ltd on behalf of Institution of Structural Engineers.
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Table 1
Summary of research studies on structural behaviour of reinforced concrete beams using recycled concrete (RC) with or without crumb rubber.
Bending tests

Author Beam size (mm) Reinf ratio Compressive strength MPa Bending capacity KNm Deflection at yield (mm)

W XD XL RC CC RC beam CC beam RC beam CC beam

[6] a* 150 × 300 × 3000 1.46 42.5 43.7 202 202.5 22.2 22.5
[31] a* 300 × 460 × 3000 0.47 30.5 37.2 119.3 123.4 9.1 7.1
[32] a* 150 × 200 × 2800 1.06 30.4 30.6 12.9 13.5 8.5 8.9
[33] b* 250 × 250 × 2440 2.03 48.35 65.61 90.7 100.68 11.6 11.5

Shear tests

Author Beam size(mm) a/d Compressive strength Mpa Ultimate shear capacity KN Maximum crack width (mm)

W XD XL RC CC RC beam CC beam RC beam CC beam

[34]b* 250 × 2501500 1.98 48.4 65.6 97.99 125.35 5.5 8

[35] a* 300 × 4600 × 4300 3 30 37.2 114.8 121.2 – –
[36] a* 200 × 400 × 2200 3.25 22.56 24.66 115.54 142.1 5.03 5.27
a* [37] 200 × 350 × 3050 3.3 39.65 40.2 90.64 88.86 5.57 4.75

a*: Recycled concrete (RC) without crumb rubber; b*: RC with crumb rubber; a/d-shear span to depth ratio.

Fig. 1. Experimental arrangement.

Fig. 2. Dimensions of two layer beam for bending tests.

215
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 3. Dimensions of two layer shear test beams.

Table 2
Mix composition of control concrete and rubber recycled aggregate concrete with 10% of rubber particles of recycled aggregate weight (RRAC10) per m3.
Specimen designation Cement (kg) Water (kg) w/c* ratio Natural aggregates (kg) Recycled aggregates (kg) Fine sand (kg) Crumb rubber (kg) SP (kg)

free Added

CC 550 220 – 0.4 939 – 626 – –

RRAC10 SP 550 220 39.5 0.4 – 845.1 626 93.9 5.5

*CC-Control concrete; *RRAC10SP-Rubber recycled aggregate concrete with 10% of recycled aggregate weight as crumb rubber, *SP-Superplasticizer.

Table 3
Water absorption of natural and recycled aggregates.
Type Apparent particle density Particle density on oven dry bases Particle density on saturated and oven dry bases Water absorption (%)

Natural aggregate 2.69 2.62 2.65 1.05

Recycled aggregates 2.50 2.27 2.36 4.21

concrete beams, they should achieve the same mechanical properties.

Therefore, research studies on reinforced concrete beams relevant to
using recycled concrete have focused on beams made entirely of the
same concrete. Table 1 is a summary of such investigations. Some of the
research studies in Table 1 were on reinforced concrete beams using
recycled concrete (RC) with or without crumb rubber with lower
strength than the control beam using fresh control concrete (CC), while
others used RC with similar strength as control concrete by improving
RC performance by either increasing cement content or adding fly ash
in recycled aggregate concrete, or by introducing steel fibres in crumb
rubber concrete [6,32–36,38–42].
From the highlighted results [31,33] of Table 1, it is evident that the
bending performances (load carrying capacity, deflection) and shear
capacities [34,35] of beams using recycled concrete with lower me-
chanical properties than control concrete beams were worse than those
of the control beams. As expected, the beams using RC with very similar
properties as CC achieved almost the same performances as beams
Fig. 4. Grading of natural aggregates, recycled aggregates and fine sand.
using CC [6,32,37]. Improving mechanical properties of recycled con-
crete is feasible and performances of structures made of such concrete
aggregates, crumb rubber) is limited to non-structural applications is would be similar to those using fresh concrete, but the additional
the mind-set that if concrete made of solid wastes is to replace concrete processes necessary to improve mechanical properties of RC to achieve
made of fresh materials in structural applications such as reinforced the same properties of fresh concrete are expensive and

216
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 5. (a) Recycled aggregates (b) natural aggregates (c) crumb rubber from worn out tyres.

Fig. 6. (a) Pull out bond strength and (b) normalized strength of the recycled concrete compared to control concrete.

Table 4 maintain composite behaviour of the two layers of concrete [44].

Properties of fresh and hardened concrete for bending and shear tests. Clearly, these additional processes add to the cost of construction.
Mix type Slump mm 28 days cylinder Unit wt Tensile UPV In practical design and construction, the durability (porosity, creep
strength Mpa (kg) strength Mpa (km/s) and shrinkage, freeze and thaw, chloride penetration, water absorption)
of the recycled concrete should be carefully considered [10,16,45,46]
Bending test
to ensure that the tensile reinforcement in the recycled concrete is
CC 165 40.1 2.39 4.11 4.19
RRAC10SP 170 30.3 2.19 2.82 3.61
prevented from corrosion. Incorporating fly ash as partial replacement
Shear test of cement in recycled concrete can enhance its durability performance
CC 165 41.4 – 4.15 – to achieve that of virgin concrete [45]. Furthermore, Richardson et al.
RRAC10SP 175 27.0 – 2.6 – [10] have shown that the use of air entrainment and polypropylene
fibres in recycled concrete can achieve the same freeze and thaw dur-
RAC-Recycled concrete without crumb rubber; RRAC5 and RRAC15-Recycled
ability compared to control concrete. Yadav and Pathak [46] also at-
concrete with 5 and 15% of crumb rubber.
tained acceptable durability performance of recycled concrete by re-
moving the attached mortar on the recycled aggregates used. Therefore,
environmentally damaging.
this paper will only focus on mechanical performance of the structure.
However, if one recognises that the mechanical properties of con-
crete in tension are ignored in bending design, it is possible to use
concrete made of recycled aggregates and crumb rubber from solid 2. Research significance and objectives
wastes in the tension region of structural members such as reinforced
concrete beams. This simple technology has been demonstrated pre- Application of recycled materials in structural members is limited
viously in prestressed concrete bridges [43] and reinforced concrete because of the difficulty of using recycled solid wastes to produce
beams [44] involving normal concrete in tension and high strength concrete having the same properties as concrete made from fresh ma-
concrete in compression. Lapko et al. [44] unveiled similar bending terials. The aim of this research is to enable concrete using recycled
performance between the high strength performing concrete aggregates and crumb rubber from tyres to be used in mainstream re-
(115.4 MPa) beam and the two layer beam with normal concrete inforced concrete beams. The proposed construction method is to place
(25.35 MPa) in the bottom layer and high performing concrete in the concrete made from recycled aggregates and crumb rubber in the ten-
top layer (115.4 MPa). However, because the two layers of concrete sion zone of reinforced concrete beam. Thus, the proposed beam has
were cast with long time difference, their interface had low shear re- two layers of concrete, concrete made of fresh materials in the com-
sistance so interfacial shear failure was a serious concern. Interfacial pression region and concrete using recycled aggregates and crumb
failure could be avoided by using vertical shear links or by an addi- rubber in the tension region. The method of constructing reinforced
tional layer of liquid adhesive agent at the interface in order to concrete beams proposed in this paper has the potential to make sub-
stantial contribution to reducing the demand on fresh materials in

217
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 7. Interfacial test (a) Test arrangement (b) Failure mode.

Table 5 concrete.
Properties of shear and longitudinal reinforcement for all test beams. The objectives of this paper are; to demonstrate the viability of this
Beam type 16 mm rebar 8 mm rebar Mpa 6 mm rebar 5 mm rebar
simple idea of using two layer concrete in beams (with recycled con-
Mpa Mpa Mpa crete in tension and control concrete in compression), to investigate
means of eliminating interfacial shear failure without expensive means,
Control beam for Y = 551 Y = 547.02 _ Y = 280 and to recommend design calculation methods for bending and shear
bending tests ULT = 643 ULT = 666.4 ULT = 310
resistances of such beams.
Two layer beam Y = 524.1 Y = 547.02 Y = 383 Y = 280
for bending ULT = 616 ULT = 666.4 ULT = 466 ULT = 310
tests
Control and two Y = 551 Y = 547.02 _ Y = 280 3. Experiments
layer beams ULT = 643 ULT = 666.4 ULT = 310
for shear 3.1. Specimen design
tests

*Y- Yield strength; *ULT- Ultimate strength. A total of eight simply supported reinforced concrete beams were
tested, consisting of six tests for assessment of bending resistance
(bending tests) and two tests for assessment of shear resistance (shear
tests). Fig. 1 shows the overall four-point bending arrangement.

Fig. 8. Failure modes of beams in bending.

218
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

longitudinal reinforcement ratio of 2.64% at the bottom.

3.2. Concrete mix and properties

The target grade for the control mix (using fresh materials) was
40Mpa compressive cylinder strength. Table 2 compares material
compositions of the control concrete (designated as CC) and the rubber
recycled aggregate concrete with 10% replacement of recycled ag-
gregates by crumb rubbers (RRAC10). The recycled concrete composi-
tion for this study was selected based on the results of the different
mixes made to meet workability and mechanical properties for struc-
tural applications. Additional water was added to the recycled concrete
to compensate the excessive water absorbed by the recycled aggregates
in Table 3. The added water was carefully measured based on the
amount of water absorbed (4.21%) by the recycled aggregates in order
to maintain same free water cement ratio with control mix. From the
series of mix designs made, this research also used superplasticizer of
1% of cement weight to meet the workability requirement of the pro-
posed recycled concrete due to the added crumb rubber. The size of
Fig. 9. Comparison of experimental load-longitudinal reinforcement bar strain
curves of bending test tests.
both the recycled and natural aggregate is 10 mm from the sieve ana-
lysis conducted in Fig. 4.
The crumb rubbers were supplied by SRC products Ltd., Stockport in
the United Kingdom. They were free from wire strings with an average
length of 8 mm in Fig. 5(c). They were not pre-treated before in-
corporating into the concrete mix. The recycled aggregates were sup-
plied from the commercial market (Offerton sand and gravel, Man-
chester) within UK comprising of other debris (bricks, glass, and tiles)
in Fig. 5(a). The 10% of rubber particles of the recycled aggregate
weight was carefully selected in order to maintain the basic mechanical
properties of the recycled concrete for structural application and the
bond strength existing between rebar and recycled concrete in tension.
Pull-out tests were carried out on different recycled concrete mixes to
justify the selected mix for this research in order to meet up bonding
requirements. Recycled concrete with 5% and 15% crumb rubber of
recycled aggregate weight pull out bond strength compared to the
control concrete is shown in Fig. 6. The pull out strength of the recycled
concrete without crumb rubber is 16% higher than the control. This is
attributed to the fact that the crushed recycled aggregates have rough
edges that enables better grip to the ribbed bar compared to the un-
crushed natural aggregates with smooth edges. However the bond
strength of the recycled concrete is reduced with increase in the content
of crumb rubber though the normalized bond strength in Fig. 6(b) still
indicate acceptable bond strength of up to 15% of crumb rubber in
concrete. Therefore in order to maintain acceptable bond strength of
Fig. 10. Load deflection curves of bending test beams.
the recycled concrete and rebar, this study limit the incorporation
crumb rubber to 10% of the recycled aggregate weight as in Table 2.
Fig. 2 shows the overall dimensions of bending test beams. The Table 4 compare the wet and hardened concrete properties between
bending test beams were designed to fail under flexure bending ac- control concrete and rubber recycled aggregate concrete for bending
cording to EC2 [47], therefore, they were provided with vertical shear and shear test respectively.
links to prevent vertical shear failure. The cross section of all beams is As mentioned previously, an issue with two layer beams is inter-
150 mm width and overall depth of 150 mm. In the two layer beams facial shear failure. For the proposed two layer beam, the time interval
(for both bending and shear tests), the top 50 mm of concrete was made between casting the two layers of concrete can be made short so long as
of fresh materials and the bottom 100 mm of concrete was of recycled the bottom layer of recycled concrete is set so that the top layer of
aggregate concrete, as shown in Fig. 2. According to calculations using control concrete does not compromise its quality by flowing into the
EC2 [47], the plastic neutral axis is just above the interface of the two bottom layer. The short time delay between casting the two layer of
different types of concrete, so the recycled concrete is entirely in the concrete can ensure that the interface achieves high bond strength so
tension zone. that interfacial shear failure is prevented. To demonstrate this, inter-
A longitudinal reinforcement ratio of 2.03% was used for all beams facial shear bond test was carried out according to BSEN12615 [48], on
in bending: consisting of two 16 mm diameter ribbed bars at the bottom specimens with an inclined interface as shown in Fig. 7(a). The two
and two 8 mm diameter rebar bars at the top. parts of the specimens were cast with a 4 h delay. Fig. 7(b) indicates
Fig. 3 shows the shear test beam dimensions. The shear test beams that the specimen failed in a monolithic fashion, with compressive
had no shear reinforcement in the shear span length. The shear span failure rather than interfacial shear failure.
depth ratio was 3.4. Only four shear links were provided as hanger bars Table 5 lists details of longitudinal and shear reinforcement prop-
located at the support and loading positions. The shear test beams erties from tensile test used for all the beams.
where designed to preclude flexural failure, which was ensured by a

219
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 11. Failure modes of shear test beams.

Fig. 14. Results of mesh sensitivity analysis of reinforced concrete beam.

(132 mm) ratio was 3 in the bending tests and 3.4 in the shear tests
(450 mm/132 mm). The beams had an unsupported length of 150 mm
and the supports were made of cylindrical steel rollers of 50 mm dia-
meter. Two actuators applied loads to create a four point bending test.
Fig. 12. Load deflection curves for the control and two layer shear test beams. Two linear strain gauges (FLA-6-11) with a gauge length and a gauge
factor of 6 and 2.13 respectively were attached to the bottom long-
3.3. Test set up and instrumentation itudinal bars at the centre of the beam to obtain reinforcement strain
data. The deflection of the beam was measured at the bottom of the
As shown in Fig. 2, all beams were simply supported with a span of centre span by means of a linear potentiometer (LP). The test was
1.2 m between the supports. The shear span (400 mm) to depth displacement controlled at a rate of 0.02 mm per second until failure.

Fig. 13. Stress crack opening width curve for cracked concrete [49].

220
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 15. Comparison of experimental and numerical simulation results for load-
longitudinal tensile steel strain curves for all beams.
Fig. 17. Comparison of experimental and numerical load deflection plots for
control and two layer beams in bending.
All beams were tested at least 28 days after casting.

3.4. Results and discussions 3.4.1.2. Load-deflection curves. Figs. 9 and 10 compares the yield load
of the bottom steel rebar and load deflection curves for bending test
3.4.1. Bending tests beams respectively. They show almost identical elastic behaviour. The
3.4.1.1. Crack pattern and failure modes for bending test. Fig. 8 shows the ultimate load carrying capacities of the two layer beams is slightly
bending test beams after failure. In all cases, the failure mode of control lower than those of the control beams. This is attributed to the lower
beams was flexural bending, as indicated by the vertical cracks in the yield strength of the 16 mm diameter longitudinal rebar used for the
pure bending region shown in Fig. 8. There was no interfacial shear two layer beams than those for the control beams in Table 5. The load-
failure, which is in agreement with findings of interfacial bond test deflection curves in Fig. 9 indicate that the two types of beam have very
(Fig. 7). The tensile reinforcement reached yield, as shown in Fig. 9 for similar stiffness at a serviceability limit state load of 60% of the
the control and two layer beams. ultimate load carrying capacity. The numerical results reveal that

Fig. 16. Numerical simulation results of crack pattern and failure mode of all beams.

221
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 20. Comparison of numerical results of load-deflection at mid-span of two

layer beams of different recycled concrete grades with control beam.

Fig. 18. Comparison of experimental and numerical load-deflection plots, for

control and two layer beam in shear.

Fig. 19. Comparison of numerical results of strain in longitudinal reinforce-

ment of two layer beams of different recycled concrete grades with control
beam. Fig. 21. Influence of ratio of top layer thickness to bottom layer thickness on
shear resistance of two layer beam.

with identical materials, the bending resistances of two layer beams are
the same as those of control beams (Figs. 15 and 16). 4. Finite element modelling using Abaqus

The general finite element software ABAQUS was used to carry out
3.4.2. Shear test beams numerical simulations to help explaining the structural behaviour of the
The crack patterns and failure modes of the control beam and two two layer beam and to extend the scope of the limited experiments, in
layer beams were the same. However, the failure modes of both beams particular, to provide guidance on how to evaluate shear resistance of
were shear as intended, as clearly indicated by the wide diagonal cracks two layer beams.
in the shear span shown in Fig. 11(a) and (b). Shear failure mode of To validate the ABAQUS model, all the bending and shear tests were
both beams is further confirmed by brittle response of both beams, as simulated. In the ABAQUS model, concrete was modelled with C3D8R
shown by their load-deflection curves in Fig. 12, after reaching peak SOLID elements comprising of eight nodes with three degrees of
loads. freedom at every nodal point. This type of element can be used to model
linear and nonlinear behaviour involving contacts, plasticity, cracking,
crushing and large deformations. For both longitudinal and shear

222
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Table 7
Different methods of calculating shear strength of beams without stirrups.
Proposed model Ultimate shear strength

EC2 1 3⎤ 1
Vc = ⎡0.18K (100ρfck ) b∗d
⎣ ⎦

(
K = Size effect factor K = 1 + 200
d ) ≤ 2.0
[51] 1
3
(
υc = 2.1746 fck ∗ ρ ∗ ) [ Mpa ] for ≥ 2.5
d
a
a
d
1
3
= (2.5 ) ∗ 2.1746 (f ∗ ρ ∗ ) [ Mpa] for
d d a
υc ck ≤ 2.5
a a d
[52]
0.8313 ρ ⎛ ρ
⎞
υc = ∗ ⎜ fck + 249 5 ⎟ [Mpa]
d ⎜ a ⎟
1 +⎛
⎝ 25
∗ d0 ⎞
⎠ ⎝ ( d) ⎠

defining the stress crack opening curve as specified in CEB-FIP [49] and
shown in Fig. 13.
The post peak behaviour of stress-crack opening is assumed to be
bilinear and fracture energy is required to define the model parameters.
The fracture energy of concrete (defined as the energy required to
propagate a tensile crack per unit area of concrete) was determined
based on provisions made in CEP-FIP [50] and given in Eq. (3).
0.7
f
Fig. 22. Influence of top layer control mix on shear resistance of two layer GF = GFO ⎛⎜ cm ⎞⎟
beam. f
⎝ cmo ⎠ (3)

where fcm is the compressive strength of concrete and fcmo taken as

reinforcements, beam elements were used and are embedded in solid
10Mpa. The GFO is the base value of fracture energy which depends on
concrete elements. The beam elements have two nodes with three de-
the maximum size of aggregates (14 mm) in the concrete mix. Linear
grees of freedom at each node. The interface between the two concrete
interpolation ranging from 8 mm to 16 mm as provided in the [50] code
layers is modelled as a perfect contact. The perfect contact is assumed
was adopted to predict the base fracture energy in this study as speci-
due to the fact the interface strength in Figs. 10(b) and 12(b) of the
fications were not made the code in this regard.
experimental results was sufficient to transfer forces from the bottom
The steel reinforcement bars were modelled as elastic perfectly
concrete to the top layer without failure.
plastic material embedded in concrete. The steel reinforcement bars
were modelled as elastic perfectly plastic material embedded in con-
4.1. Material properties and constitutive models crete.

Concrete damage plasticity (CDP) model was used in defining the

behaviour of concrete in the inelastic range. The main failure criteria of 4.2. Finite element discretization
concrete damage plasticity model are tensile cracking and compressive
crushing of concrete. Mesh sensitivity test was carried out to determine the most appro-
The uniaxial compressive stress strain curves of both the control priate mesh density for the simulation model. Fig. 14 compares the
concrete and rubber recycled aggregate concrete were based on the mesh sensitivity study results. Accordingly, a mesh size of 15 mm was
authors' material test results. They were then converted into true stress- chosen. Mesh sizes 10 to 20 mm give the same bending performance
plastic strain curves using Eqs. (2) and (3). based on load deflection plots.
σtrue = σengineering ∗ (1 + εengineering ) (1)
4.3. Comparison between Abaqus numerical simulation results with
σtrue
εInplastic = In (1 + εengineering ) − experimental results
E (2)

where σtrue and εInplastic are the true stress and plastic strain respectively. 4.3.1. Tensile strain in longitudinal reinforcement
σengineering is the engineering stress, εengineering is the engineering strain, E Fig. 15 compares numerical simulation and experimental results for
is the modulus of elasticity. In(1 + εengineering) represents the total strain tensile strain of the bottom reinforcement at the mid-span of beams.
and σtrue is the elastic strain (recoverable part of the strain). There is good correlation for both the control and two layer beams. The
E
After cracking, concrete retains a small amount of tensile strength reinforcement for all beams fully reached its yield strain and stayed on
normal to the direction of crack. This phenomenon is called tensile the yield plateau to enable the beam to reach full bending resistance
stiffening. In this study, tensile stiffening of concrete is accounted for by prior to failure.

Table 6
Comparison of bending resistance for all beams to EC2.
Beam type Concrete strength MPa Reinforcement yield strength Mpa Test bending resistance kN-m EC2 result/test or simulation result

Control beam 40.0 D16 = 551 D8 = 547.02 27.4 0.977

Two layer beam 40.0/30.0 D16 = 551 D8 = 547.02 27.3 0.980
Two layer beam 40.0/22.0 D16 = 551 D8 = 547.02 27.3 0.980

223
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

Fig. 23. Comparison of ratios of test/simulation result to calculation result of shear strength of two layer beams using different calculation method.

4.3.2. Crack pattern and failure mode It suggests that the shear resistances of these beams are almost iden-
Fig. 16 shows the simulated crack patterns of the control and two tical. Also the influence of the top layer control mix was investigated to
layer beams. They correspond well with the test observations in Fig. 7. aid understanding of shear failure mechanism of two layer beams.
Vertical cracks in the bending region fully developed in control and two Fig. 22 show that the beam made entirely of recycled concrete has
layer beams beam Fig. 16(a) and (b) respectively, indicating flexural identical shear resistance as the two layer beams. This was attributed to
bending failure mode. Fig. 16(c) and (d) indicates shear failure of the the unzipping effect: once the lower strength (recycled concrete) con-
control and two layer beams respectively with diagonal cracks running crete has failed in shear, it losses its shear resistance and transfers the
from the support to the point of applied load. The result correlates well shear force from the failed lower strength concrete to the control con-
the test results in Fig. 10. crete.

4.3.3. Load deflection curves at mid-span 5. Assessment of design methods

Figs. 17 and 18 compare experimental and numerical simulation
results of load-deflection curves for both the control and two layer 5.1. Bending resistance
beams for bending and shear test respectively. The simulation results
include some post-peak results to ensure that the numerical models The experimental and numerical simulation results have confirmed
reached full capacities of the beams. Agreements between the experi- that two layer beams reach the same bending resistance as control
ment and simulation results are very good. beams. Table 6 compares design calculation results using EC2 with all
The results in Figs. 17 and 18 confirm that the concrete damage the test and simulation results of bending resistance.
plasticity model in ABAQUS can be used to model two layer concrete As expected, Table 6 reveals that the two layer beams attain the
beam behaviour. same bending resistance as the control beam. EC2 calculation results
are accurate and on the safe side. Therefore, this research concludes
4.4. Numerical parametric study results that the EC 2 method of calculating reinforced concrete beam bending
resistance can be used, without modification, for the two layer beams.
The validated ABAQUS model was used to investigate the effects of
changing a number of design parameters, such as grade of recycled 5.2. Shear resistance
concrete, depth of recycled concrete layer.
According to the shear resistance mechanism discussed earlier, due
4.4.1. Effects of recycled concrete grade on bending performance of two to unzipping effect, the shear resistance of a plain two layer concrete
layer beam beam is the same regardless of depth and grade of the top layer (con-
In order to re affirm the hypothesis that concrete grade in the ten- trol) concrete. Hence, there is no additional contribution of the control
sion zone plays no role in determining bending resistance, two different concrete in the top layer on shear strength of two layer beams.
grades of recycled concrete (22Mpa, 30Mpa) were used while main- Therefore, the shear resistance of a two layer beam can be calculated in
taining the control mix of 40Mpa at the top. Fig. 19 shows that the the same way as a beam made entirely of recycled concrete without
longitudinal reinforcement in all beams reaches its yield strain and considering the higher strength of control concrete in the top layer. To
maintains a long plateau. Fig. 20 confirms that all beams achieve the verify this, the shear resistances for all the test and simulation beams
same bending resistance. were checked against EC2 and prediction methods of Zsutty [51] and
Bazant and Kim [52]. The calculation methods are summarized in
4.4.2. Plain concrete shear resistance mechanism: effects of top layer/ Table 7.
bottom layer concrete depth ratio and top layer control mix on shear Fig. 23 summarizes the calculation results, the vertical axis showing
capacity of two layer beams ratio of test/numerical simulation result to analytical calculation result.
From the two shear test validated model, the ratio of control con- In all cases, all different analytical methods conservatively calculate the
crete depth to recycled concrete depth was changed (ratio = 0.5, 1, 2) shear capacity of two layer beams by using the strength of the recycled
to develop understanding of shear resistance mechanism of two layer concrete. EC2 results conservatively predict the shear capacity of the
beams. Fig. 21 compares the simulation results of load-deflection curve. two layer beams than the other two methods based on the recycled

224
R.B. Ataria and Y.C. Wang Structures 20 (2019) 214–225

concrete strength. aggregates with different amounts of old adhered mortars. Mater. Des.
2014;58:19–29.
[15] Kwan WH, Ramli M, Kam KJ, Sulieman MZ. Influence of the amount of recycled
6. Conclusions coarse aggregate in concrete design and durability properties. Construct Build
Mater 2011;26(1):565–73.
This paper has presented the results of an experimental programme [16] Manzi S, Mazzotti C, Bignozzi MC. Short and long-term behavior of structural
concrete with recycled concrete aggregate. Cem Concr Compos 2013;37:312–8.
and associated numerical simulations to investigate the effects of re- [17] Aiello MA, Leuzzi F. Waste tyre rubberized concrete: properties at fresh and har-
placing concrete in the tension region of simply supported reinforced dened state. Waste Manag 2010;30:1696–704.
concrete beams with weaker concrete made of rubber and recycled [18] Atahan AO, Yücel AÖ. Crumb rubber in concrete: static and dynamic evaluation.
Construct Build Mater 2012;36:617–22.
aggregate. The aim of the research was to investigate whether it would [19] Eldin NN, Senouci AB. Measurement and prediction of the strength of rubberized
be possible to use rubber recycled aggregate concrete in primary concrete. Cem. Concr. Compos. 1994;16(1994):287–98.
loadbearing members without having to improve mechanical properties [20] Ganjian E, Khorami M, Maghsoudi AA. Scrap-tyre-rubber replacement for aggregate
and filler in concrete. Construct Build Mater 2009;23:1828–36.
of recycled concrete. The following conclusions may be drawn:
[21] Herna'ndez-Olivares, F., Barluenga, G., Bollati, M. & Witoszek, B. 2002. Static and
It is feasible to use rubber recycled aggregate concrete without re- dynamic behaviour of recycled tyre rubber-filled concrete. Cem Concr Res, 32,
ducing beam's bending resistance. The bending resistance of two layer 1587–1596.
beam was found to be same as the control beam provided other con- [22] Holmes N, Browne A, Montague C. Acoustic properties of concrete panels with
crumb rubber as a fine aggregate replacement. Construct Build Mater
ditions are the same. The grade of recycled concrete in the tension zone 2014;73:195–204.
has no influence. Consequently, the EC 2 method of calculating beam [23] Khaloo AR, Dehestani M, Rahmatabadi P. Mechanical properties of concrete con-
bending resistance can be used, without modification, to the proposed taining a high volume of tire-rubber particles. Waste Manag 2008;28:2472–82.
[24] Long G, Feys D, Khayat KH, Yahia A. Efficiency of waste tire rubber aggregate on
two layer beam construction. Using weak, recycled aggregate concrete the rheological properties and compressive strength of cementitious materials. J
with rubber crumb in the tension zone slightly reduces the beam stiff- Sustain Cem Based Mater 2014;3:201–11.
ness. [25] Moustafa A, Elgawady MA. Mechanical properties of high strength concrete with
scrap tire rubber. Construct Build Mater 2015;93:249–56.
Under vertical shear failure mode, the proposed two layer beam [26] Najim KB, Hall MR. Crumb rubber aggregate coatings/pre-treatments and their
experiences unzipping effect: once the lower strength concrete has effects on interfacial bonding, air entrapment and fracture toughness in self-com-
failed in shear, it loses its shear resistance and transfers the shear force pacting rubberised concrete (SCRC). Mater Struct 2013;46:2029–43.
[27] Top~U LB. The properties of rubberized concretes. Cem Concr Res
from the failed lower strength concrete to the control concrete. 1995;25(2):304–10.
Consequently, when calculating shear resistance of plain concrete of the [28] Toutanji HA. The use of rubber Tire particles in concrete to replace mineral ag-
proposed two layer beam, the EC2 calculation method can still be used gregates. Cement & Concrete Composites 1995;18:135–9.
[29] Wang L, Huang YH. Study on rubber particles modified concrete. Appl Mech Mater
but the recycled concrete strength in the bottom layer should be used in
2013;477-478:953–8.
calculations of beam shear resistance. [30] Zheng L, Sharon Huo X, Yuan Y. Experimental investigation on dynamic properties
Horizontal interfacial shear failure can be minimised by casting the of rubberized concrete. Construct Build Mater 2008;22:939–47.
two layers of concrete at short intervals (4 h or less) so that the interface [31] Arezoumandi M, Smith A, Volz JS, Khayat KH. An experimental study on flexural
strength of reinforced concrete beams with 100% recycled concrete aggregate. Eng
could achieve sufficient bond strength. Struct 2015;88:154–62.
[32] Sato R, Maruyama I, Sogabe T & M S. 2007. Flexural behaviour of reinforced re-
Acknowledgments cycled concrete beams. J Adv Concr Technol, 5, No. 1, 43–61.
[33] Ismail MK, Hassan AAA, Hussein AA. Structural behaviour of reinforced concrete
beams containing crumb rubber and steel fibres. Mag Concr Res 2017;69:939–53.
The authors wish to express their gratitude and sincere appreciation [34] Ismail MK, Hassan AAA. Shear behaviour of large-scale rubberized concrete beams
to Mr. Paul Nedwell and Mr. John Mason for their support in the reinforced with steel fibres. Construct Build Mater 2017;140:43–57.
[35] Arezoumandi M, Smith A, Volz JS, Khayat KH. An experimental study on shear
Laboratory tests. strength of reinforced concrete beams with 100% recycled concrete aggregate.
Construct Build Mater 2014;53:612–20.
References [36] Choi HB, Yi CK, Cho HH, Kang KI. Experimental study on the shear strength of
recycled aggregate concrete beams. Mag Concr Res 2010;62:103–14.
[37] González-Fonteboa. Shear strength of recycled concrete beams. Construct Build
[1] Fisher C, Werge M. EU as a recycling society. ETC./SCP Working Paper 2/2009. Mater 2007;21:887–93.
2009. [38] González-Fonteboa B, Martínez-Abella F, Martínez-Lage I, Eiras-López J. Structural
[2] Sakai K. Environmental design for concrete structures. J Adv Concrete Technol shear behaviour of recycled concrete with silica fume. Construct Build Mater
2005;3:17–28. 2009;23:3406–10.
[3] Symonds A. COWI, PRC BOUWCENTRUM. Construction and demolition waste [39] Ismail MK, Hassan AAA. An experimental study on flexural behaviour of large-scale
management practices and their economic impact. Report of the project group to concrete beams incorporating crumb rubber and steel fibres. Eng Struct
the European Commission. 1999. 2017;145:97–108.
[4] ETRA European Tyre Recycling Association (ETRA) [40] Malešev M, Radonjanin V, Marinković S. Recycled concrete as aggregate for
[5] UTWG, U. T. W. G. tyre recycling. Department of Trade and Industry, London (UK). structural concrete production. Sustainability 2010;2:1204–25.
[6] Ignjatović IS, Marinković SB, Mišković ZM, Savić AR. Flexural behavior of re- [41] Mendis ASM, Al-Deen S, Ashraf M. Flexural shear behaviour of reinforced crumbed
inforced recycled aggregate concrete beams under short-term loading. Mater Struct rubber concrete beam. Construct Build Mater 2018;166:779–91.
2012;46:1045–59. [42] Xiao J, Li W, Fan Y, Huang X. An overview of study on recycled aggregate concrete
[7] Barbudo A, De Brito J, Evangelista L, Bravo M, Agrela F. Influence of water-reducing in China (1996–2011). Construct Build Mater 2012;31:364–83.
admixtures on the mechanical performance of recycled concrete. J Clean Prod [43] Iskhakov I, Ribakov Y. Two-layer concrete bridge beams as composite elements.
2013;59:93–8. Struct Concr 2013;14:271–7.
[8] Dilbas H, Şimşek M, Çakır Ö. An investigation on mechanical and physical prop- [44] Lapko A, Sadowska-Buraczewska B, Tomaszewicz A. Experimental and numerical
erties of recycled aggregate concrete (RAC) with and without silica fume. Construct analysis of flexural composite beams with partial use of high strength/high per-
Build Mater 2014;61:50–9. formance concrete. J Civ Eng Manag 2005;11:115–20.
[9] Mefteh H, Kebaili O, Oucief H, Berredjem L, Arabi N. Influence of moisture con- [45] Kou SC, Poon CS. Enhancing the durability properties of concrete prepared with
ditioning of recycled aggregates on the properties of fresh and hardened concrete. J coarse recycled aggregate. Construct Build Mater 2012;35:69–76.
Clean Prod 2013;54:282–8. [46] Yadav, S. & Pathak, S. 2018. Durability aspects of recycled aggregate concrete: an
[10] Richardson A, Coventry K, Bacon J. Freeze/thaw durability of concrete with re- experimental study. Int J Civ Environ Eng, Vol:12, No:3, 2018.
cycled demolition aggregate compared to virgin aggregate concrete. J Clean Prod [47] BS-EN1992-1-1:2004 Eurocode2: design of concrete structures.
2011;19:272–7. [48] BSEN12615 1999. Products and systems for the protection and repair of concrete
[11] Bolouri Bazaz J, Khayati M. Properties and performance of concrete made with structures.
recycled low-quality crushed brick. J Mater Civ Eng 2012;24:330–8. [49] CEB-FIP 1990. Model Code for concrete structures.
[12] Crentsil KKS-, Brown T, Taylor AH. Performance of concrete made with commer- [50] CEP-FIP. Model code 1990. London: Thomas Telford services Ltd.; 1990. [437pp].
cially produced coarse recycled concrete aggregates. Cem Concr Res [51] Zsutty TC. Shear strength predictions for separate categories of simple beam tests.
2001;31:707–12. ACI J Proc 1971;68(2):138–43.
[13] De Juan MS, Gutiérrez PA. Study on the influence of attached mortar content on the [52] Bazant Z, Kim J. Size effect in shear failure of longitudinally reinforced beams. Proc
properties of recycled concrete aggregate. Construct Build Mater 2009;23:872–7. ACI J 1984;81(5):456–68.
[14] Duan ZH, Poon CS. Properties of recycled aggregate concrete made with recycled

225