Sie sind auf Seite 1von 12

Journal of Cleaner Production 197 (2018) 656e667

Contents lists available at ScienceDirect

Journal of Cleaner Production


journal homepage: www.elsevier.com/locate/jclepro

Effects of the addition of silica fume and rubber particles on the


compressive behaviour of recycled aggregate concrete with steel fibres
Jianhe Xie, Chi Fang, Zhongyu Lu*, 1, Zijian Li, Lijuan Li
School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou 510006, Guangdong, China

a r t i c l e i n f o a b s t r a c t

Article history: The use of recycled aggregates made from waste concrete and scrap rubber in structural concrete is a
Received 28 November 2017 sustainable solution to dealing with solid waste. This technology reduces the serious impact on
Received in revised form ecological environments caused by a shortage of natural mineral resources. The aim of the present study
20 May 2018
is to investigate the coupling effects of incorporating silica fume (SF) and rubber particles on the
Accepted 22 June 2018
compressive performance of rubberized steel-fibre recycled aggregate concrete (RSRAC). The SF and
Available online 26 June 2018
rubber contents were the main test parameters. The compressive strength, elasticity modulus, energy
dissipation capacity, and failure mechanism of RSRAC were analysed based on a series of axial
Keywords:
Recycled aggregate concrete (RAC)
compression tests, and the carbon emissions of RSRAC were estimated. The interfaces between the
Silica fume (SF) recycled coarse aggregate (RCA), rubber particles, steel fibre, and cement paste in RSRAC without SF are
Rubber particle generally weak; however, the addition of SF enhances these interfacial bonds, resulting in an improve-
Steel fibre ment in the compressive strength of RSRAC. Such strength increases with the amount of SF. Based on a
Compressive behaviour synthetical consideration of the compressive properties and carbon emissions, RSRAC with 100% recycled
Carbon emission coarse aggregate, 10% SF, and 5% rubber is a more environmentally friendly alternative to normal con-
crete for use in the compression member of concrete structures.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction fabrication using RA than that using NA (Henry et al., 2011; Shi
et al., 2016). In addition to the C&D waste, the “black pollution”
Due to the rapid urbanization in developing countries, consid- in China due to waste rubber is also increasing sharply. At present,
erable construction and demolition (C&D) waste has been gener- the scrap tyre production in China is over 10 million tons per year,
ated, and the amount of C&D waste is increasing annually. which is the highest rate worldwide (Tai et al., 2017). Determining
According to current statistics, 15 billion tons of construction waste how to address the increase in waste rubber has become a serious
were produced in China in 2015 (Tai et al., 2017). One of the pro- environmental problem. Since rubber does not decompose easily,
posed solutions is to use the C&D waste as concrete aggregate. The the traditional disposal methods may lead to secondary pollution
use of recycled aggregate (RA) in concrete would help to reduce the (Gheni et al., 2017). Currently, scrap rubber commonly is broken
negative environmental impacts of C&D waste. It should be noted into fine particles or powder and used as fine aggregate in concrete
that the ever-increasing construction demands may be met (Thomas and Chandra Gupta, 2016), and the rubber can improve
without depleting natural resources by replacing natural aggregate some of the mechanical properties of the concrete (Rashad, 2016).
(NA) with RA (Kurad et al., 2017). Furthermore, CO2 emissions and In general, the management of solid waste with respect to waste
raw material consumption are lower during structural member recycling can be improved in China. As mentioned above, trans-
forming (C&D) waste and scrap rubber into aggregates for use in
concrete production is a promising technology that promotes the
recycling of solid waste, consequently reducing not only environ-
* Corresponding author. mental pollution but also the consumption of raw materials (Xie
E-mail addresses: jhxie@gdut.edu.cn (J. Xie), fangchi1993@hotmail.com
(C. Fang), luzy@gdut.edu.cn (Z. Lu), 2111709010@mail2.gdut.edu.cn (Z. Li), lilj@
et al., 2015).
gdut.edu.cn (L. Li). However, because RA has many unfavourable properties, such as
1
Postal address: School of Civil and Transportation Engineering, Guangdong a high porosity, high water absorption, and low strength, concrete
University of Technology, No. 100 Waihuan Xi Road, Guangzhou Higher Education containing RA exhibits inferior mechanical properties compared to
Mega Center, Panyu District, Guangzhou 510006, China.

https://doi.org/10.1016/j.jclepro.2018.06.237
0959-6526/© 2018 Elsevier Ltd. All rights reserved.
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 657

effectively improve the compressive performance of the rubberized


List of acronyms concrete.
Moreover, the addition of fibre into brittle concrete is an
ITZ interfacial transition zone increasingly used technology in civil engineering. Many studies
LVDTs linear variable displacement transducers reported that the bridging and pulling effects of fibre can also
NA natural aggregate greatly improve the brittleness, impact resistance, tensile and
NAC natural aggregate concrete flexural strength, and energy dissipation of NAC (Al-Masoodi et al.,
NCA natural coarse aggregate 2016; Nazarimofrad et al., 2008). Similarly, steel fibre improves
RA recycled aggregate both the cracking resistance of RAC (Guo et al., 2014) and its frac-
RAC recycled aggregate concrete turing process (Carneiro et al., 2014). Consequently, Katzer and
RCA recycled coarse aggregate Domski (2013) noted that RAC with fibre has a considerable po-
RSRAC rubberized steel-fibre recycled aggregate concrete tential for use in secondary structural elements. The performance
SF silica fume of fibre in concrete depends on the aspect ratio, fibre type and
optimal fibre content (Won et al., 2012; Yoo et al., 2013). Regarding
the optimum content of steel fibre in concrete, Wang and Wang
(2013) and Lau and Anson (2006) concluded that the volume ra-
those of concrete containing NA (Chakradhara Rao et al., 2011; tio of the steel fibre ranged from 1% to 1.5% is feasible. Furthermore,
Evangelista and de Brito, 2014; Sivakumar et al., 2014; Tam et al., Nguyen et al. (2010) reported that the combination of rubber and
2005). The compressive strength, elastic modulus and flexural steel fibre has a positive synergistic influence on cement-based
strength of recycled aggregate concrete (RAC) gradually decrease as mortars: the steel fibre increases the post-crack strength; the
the content of RA increases (Chakradhara Rao et al., 2011; rubber increases the strain capacity.
Etxeberria et al., 2007; Guo et al., 2014; Henry et al., 2011;
Sivakumar et al., 2014; Xiao et al., 2005). Etxeberria et al. (2007) 2. Research significance
and Chakradhara Rao et al. (2011) reported that the replacement
of NA with RA by up to 25% does not significantly change the me- Studies have showed that it is promising to enhance the
chanical properties of the concrete. However, replacing 50% or rubberized RAC performance by adding optimal silica fume or steel
more of the coarse and fine NA with RA can significantly reduce the fibre; however, previous studies have focused on the independent
compressive strength of the concrete (Xiao et al., 2012). For the role of rubber, steel fibre or silica fume on the mechanical perfor-
rubberized concrete, several studies reported that its impact mance of concrete, very limited information is available for the
behaviour can be improved significantly due to the addition of coupling effect of rubber and steel fibre on RAC. Actually, rubber-
rubber (Al-Tayeb et al., 2012; Liu et al., 2012; Najim and Hall, 2012; ized steel-fibre recycled aggregate concrete (RSRAC) is considered a
Taha et al., 2008; Vadivel et al., 2014), as well as the fatigue per- promising type of environmentally friendly concrete (Guo et al.,
formance (Liu et al., 2013) and the reeze-thaw resistance 2014): (1) recycled aggregates of waste concrete and rubber are
(Richardson et al., 2012). Thomas et al. (2016) also reported that the mainly included because of their environmental significance, (2)
rubberized concrete is highly resistant to the aggressive environ- rubber particles are also used to reduce the overall density, and
ments and can be implemented in the areas where there are improve the freeze-thaw resistance and dynamic properties of
chances of acid attack. However, extensive experimental results concrete, and (3) steel fibre is used to improve the concrete per-
indicated that the compressive strength of concrete reduces with formance. However, experimental results reported by Guo et al.
the increase of rubber content (Antil et al., 2014; Gesoglu et al., (2014) and Xie et al. (2015) indicate that a simple combination of
2014; Holmes et al., 2014; Li et al., 2014; Valadares et al., 2012). rubber and steel fibre without any mineral additives cannot achieve
Guo et al. (2014) and Xie et al. (2015) conducted a series of tests to a significant positive synergistic effect on RAC. To effectively utilize
study the effects of different rubber contents on the compressive this new type of concrete in structural applications, it is essential to
properties of rubberized RAC without mineral additions; their re- consider the use of RSRAC with silica fume.
sults indicated that a concrete mix fabricated completely with The object of this research study is to investigate the coupling
recycled coarse aggregate requires mineral additives to achieve an effects of incorporating silica fume and rubber particles on the
adequate compressive strength. compressive performance of RSRAC, as well as the carbon emis-
In order to enhance the mechanical properties of RAC and sions of RSRAC. To maximize the recycling of C&D waste, recycled
rubberized concrete, ground granulated blast furnace slag (Patra coarse aggregate (RCA) was added to the concrete mixture to
and Mukharjee, 2017), fly ash (Corinaldesi and Moriconi, 2009) completely replace the natural coarse aggregate (NCA). A series of
and silica fume (SF) (Pedro et al., 2017; Shi et al., 2016) are cylindrical specimens were cast and tested under compressive
frequently used as mineral additives in the concrete. Corinaldesi loading by considering different amount of silica fume and rubber.
and Moriconi (2009) reported that SF has a more significant in- The compressive strength, stressestrain curve, toughness, elastic
fluence on improving the compressive behaviour of RAC than that modulus and the failure mechanism of RSRAC were investigated,
of either fly ash or ground granulated blast furnace slag. SF is a and the carbon emissions of RSRAC were estimated. Based on the
waste material produced in the smelting industry. When SF is synthetical consideration of compressive properties and environ-
mixed with concrete, the pozzolanic and micro-filler effects of SF mental effects, the optimal combination of SF and rubber content
can enhance the bonding of interfacial transition zone (ITZ) be- was evaluated.
tween the aggregates and paste, subsequently changing the me-
chanical properties of the concrete (Dilbas et al., 2014; Kou et al., 3. Experiment set-up
2011). Interestingly, Cakir (2014) reported that the incorporation
of SF might cause greater improvement for the mechanical prop- 3.1. Raw materials
erties of RAC than that of natural aggregate concrete (NAC). Addi-
tionally, Güneyisi et al. (2004) and Onuaguluchi and Panesar (2014) 3.1.1. Aggregates
added SF to rubberized concrete and found that the SF can In this study, the NCA was granite. Most of the RCA came from
compensate for the negative influence of the rubber particles and crushed waste concrete, which was derived from several civil
658 J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667

buildings, and had a compressive strength of 20e30 MPa. A small


amount of RCA produced from crushed bricks was also included. To
improve the purity and quality of the RA, the impurities such as
wood and glass were manually removed. After the RCAs were
repeatedly flushed with water, they were dried by sunshine before
use. Both the NCA and RCA had continuous grading from 5 to
20 mm. The basic properties of the NCA and RCA are listed in
Table 1. In this study, rubber particles and sand were included in the
fine aggregates. The coarse river sand had a fineness modulus of
2.52 and a specific gravity of 2.69. The rubber aggregates with a
density of 1.05 were obtained by shredding waste tyres, and their
Fig. 1. Photographs of the test materials: (a) RCA, (b) Rubber and steel fibre.
diameters ranged from 0.85 mm to 1.40 mm. The rubber particles,
steel fibre, RCA and SF are shown in Fig. 1, and the size grading of
the aggregates is presented in Fig. 2.

3.1.2. Cement, silica fume and steel fibre


Ordinary Portland cement with a strength of 42.5 MPa was used
in this study, according to the Chinese standard (2007). Silica fume
complying with the Chinese standard (2017) was used in the tested
concrete mixtures. Fig. 3 shows the XRD spectra of SF, and Table 2
list the properties of SF. Based on the consideration of cost, the steel
fibres with a low aspect ratio were used into concrete mix in this
study. Table 3 presents the properties of steel fibres. Fig. 2. Size grading of aggregates: (a) Coarse aggregate, (b) Fine aggregate.

3.1.3. Superplasticizer
A polycarboxylate-based high performance superplasticizer was
used to improve the initially low workability of the mixtures, which
was provided by the manufacturer Guangdong Jiangmen strong
building materials technology Co., Ltd, China. The water reduction
rate of a superplasticizer with a reddish-brown liquid is 20%e45%.

3.2. Mixtures

Nineteen groups of concrete mixtures containing the afore-


mentioned materials were produced in the laboratory, as listed in
Table 4. The SF content and rubber content were used as the main
variables. The SF contents tested were 0%, 5% and 10% by quantity
substitution of the cement. Moreover, the rubber contents tested
were 0%, 5%, 10%, 15% and 20% by volume substitution of the sand.
Additional water was added into the mixtures to achieve the
designed water/cement ratio due to the relatively higher water
absorption of the RCA compared with the NCA. Based on the slump
tests that were performed in accordance with the Chinese standard
(2016), the superplasticizer introduced in the mixtures was set as Fig. 3. XRD spectra of silica fume.
3% of the cement in weight. A total of 57 cylindrical specimens with
150 mm (diameter) by 300 mm (height) were prepared in this
study, and three specimens were made from each of the mixtures. 4000 kN. To measure the stressestrain relationship of the concrete,
The specimens were sealed with plastic film to keep wet condition two linear variable displacement transducers (LVDTs) were
in indoor temperature for 24 h. After demolding, they were cured employed to measure the axial compressive deformation value.
by water for a total of 14 days with twice a day and then stay in Two strain gauges were bonded at the cylinder mid-height to get
room temperature condition. the hoop strains. Centring and preloading were carried out before
loading. Axial loads controlled by the displacement were applied to
3.3. Test procedure the specimens at a speed of 0.18 mm/min. The majority of the tests
were stopped as the compressive load decreased to 10% of the peak
Ninety days after casting, all the specimens were tested under stress of the specimen. The measurements of displacement, strain
compressive loading. The compression test was conducted by the and load data were collected by a Tokyo Sokki Kenkyujo Co., Ltd.
Italian MATEST material testing machine with a capacity of high-speed acquisition system.

Table 1
The basic properties of coarse aggregates.

Type Particle size/mm Apparent density/(kg/m3) Bulk density/(kg/m3) Water absorption/% Crushing index/%

NCA 5e20 2652 1350 0.9 8.33


RCA 5e20 2435 1200 2.1 11.67
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 659

Table 2
The SF properties.

Dry mattercontent/% Loss on ignition/% Water-soluble sulphates/% Silica content/% Chlorides content/% FeCaO content/%

99.3 1.9 0.45 94.61 0.01 <0.1

Table 3
The basic properties of steel fibres.

Shape feature Length/mm Aspect ratio Tensile strength/MPa Density/g/cm3 Elongation/%

Shear-wave 32 45 600 7.82 5e20

Table 4
Test mixture.

Group Specimen Mix proportion (kg/m3)

Water Cement Sand Natural coarse aggregate Recycled coarse aggregate Silica fume Additional water Steel fibre Rubber Superplasticizer

NC NC 172 358.3 785.5 1041.2 e e e e e 1.075


NC-f NC-S0f 172 358.3 785.5 1041.2 e e e 78 e 1.075
NC-S5f 172 340.4 785.5 1041.2 e 17.9 e 78 e 1.075
NC-S10f 172 322.5 785.5 1041.2 e 35.8 e 78 e 1.075
RC-R0 RC-R0S0 172 358.3 785.5 e 964.9 e 19.30 78 e 1.075
RC-R0S5 172 340.4 785.5 e 964.9 17.9 19.30 78 e 1.075
RC-R0S10 172 322.5 785.5 e 964.9 35.8 19.30 78 e 1.075
RC-R5 RC-R5S0 172 358.3 746.2 e 964.9 e 19.30 78 15.7 1.075
RC-R5S5 172 340.4 746.2 e 964.9 17.9 19.30 78 15.7 1.075
RC-R5S10 172 322.5 746.2 e 964.9 35.8 19.30 78 15.7 1.075
RC-R10 RC-R10S0 172 358.3 707.0 e 964.9 e 19.30 78 31.3 1.075
RC-R10S5 172 340.4 707.0 e 964.9 17.9 19.30 78 31.3 1.075
RC-R10S10 172 322.5 707.0 e 964.9 35.8 19.30 78 31.3 1.075
RC-R15 RC-R15S0 172 358.3 667.7 e 964.9 e 19.30 78 47.0 1.075
RC-R15S5 172 340.4 667.7 e 964.9 17.9 19.30 78 47.0 1.075
RC-R15S10 172 322.5 667.7 e 964.9 35.8 19.30 78 47.0 1.075
RC-R20 RC-R20S0 172 358.3 628.4 e 964.9 e 19.30 78 62.7 1.075
RC-R20S5 172 340.4 628.4 e 964.9 17.9 19.30 78 62.7 1.075
RC-R20S10 172 322.5 628.4 e 964.9 35.8 19.30 78 62.7 1.075

Note: NC denotes natural aggregate concrete; RC denotes recycled aggregate concrete; f denotes steel fibre; S0, S5, S10 denote the SF substitution ratios of 0%, 5%, and 10%,
respectivelyt; and R0, R5, R10, R15 and R20 denote the rubber volume substitution ratios of 0%, 5%, 10%, 15% and 20%, respectively.

4. Experimental results and discussions through experimental testing, Carneiro et al. (2014) found that the
addition of fine steel fibres can improve the compressive strength
4.1. Compressive strength of concrete. In general, the addition of steel fibre has little effect on
the compressive strength of concrete, which depends on the fibre
The compressive strength of tested specimens is listed in type, fibre content and the fibre aspect ratio (Afroughsabet et al.,
Table 5. The listed value is the average of the test results from the 2016).
three specimens in a group. Table 5 shows the compressive The compressive strength reduced by 9.3% when the NCA was
strength of the natural concrete had a slight reduction due to the 100% replaced with RCA, as shown in Table 5. It is well known that
addition of steel fibre. This change is because the compressive the concrete strength relies on the strength of the aggregates and
strength of the concrete is significantly affected by the interfacial cement matrix, and their ITZ (Dilbas et al., 2014). Apart from the
bonding of the steel fibre and cement matrix. Commonly, the bond inferior properties of the recycled aggregates themselves, the
properties between thick steel fibres and cement matrix are weak, decrease in the RAC strength is commonly attributed to the weak
leading to a decrease in the compressive strength of the concrete. bonding of ITZ between recycled aggregates and cement matrix
Similar results were reported by Chen and Pan (2013). However, (Cakir, 2014).
The influence of the rubber content on the RSRAC strength is
illuminated in Fig. 4. Moreover, some experimental results of pre-
Table 5 vious studies of rubberized NAC (Al-Tayeb et al., 2012; Khatib and
Compressive strength test results. Bayomy, 1999; Topçu, 1995; Valadares et al., 2012) and rubberized
Group Rubber content Compressive strength RAC (Xie et al., 2015), in which the rubber replaced the sand, are
also presented in Fig. 4. The ordinate of Fig. 4 refers to the ratio of
0% SF 5% SF 10% SF
the compressive strength of concerned specimen to that of the non-
NC 0% 45.65 e e
rubber specimen in same group. It can be seen from Fig. 4 that a
NC-f 0% 40.70 44.42 58.84
RC-R0 0% 36.12 40.41 47.86 gradual decrease in the compressive strength can be found with the
RC-R5 5% 33.89 34.19 42.83 increase of rubber content in the results of both this study and the
RC-R10 10% 28.44 29.70 33.59 previous studies. There are mainly two reasons for the decrease in
RC-R15 15% 24.73 26.96 31.68 the rubberized concrete strength. First, the low elastic modulus and
a
RC-R20 20% 25.95 30.96
strength of the rubber particles predominantly decrease the
a
Data not obtained.
660 J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667

120%
Xie et al. (2015) (Cube)

Compressive strength ratio with respect


Xie et al. (2015) (Cylinder)
100% Al-Tayeb et al. (2012) (Cube)
Topcu (1995) (Cube)
Valadares (2012) (Cube)

to non-rubber specimen
Khatib (1999) (Cylinder)
80% Present data (No SF)
Present data (5% SF)
Present data (10% SF)

60%

40%

20%

0%
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
Rubber content
Fig. 4. Relative compressive strength vs a function of the rubber content.

strength of the concrete (Güneyisi et al., 2004). Second, the rubber 1.5
particles are highly hydrophobic, increasing the porosity of the
concrete matrix and consequently reducing its strength (Aslani,
Relative compressive strength

1.4
2016; Topcu and Bilir, 2009). However, there is disagreement NC-f
with the previously obtained results for investigations of rubber- RC-R0
ized NAC. As observed in Fig. 4, when the rubber content was less 1.3 RC-R5
than 10%, the RSRAC without SF underwent a smaller decrease in RC-R10
RC-R15
the compressive strength than that of the NAC, indicating that the
1.2
RSRAC is a more favourable concrete than traditional concrete.
Similar results were reported by Xie et al. (2015). A possible reason
for this phenomenon is that RAC is more porous than NAC, and a 1.1
small quantity of fine rubber could fill the pores in the RAC, sub-
sequently reducing the negative effect on the strength due to the
addition of rubber. Additionally, it can be found that the RSRAC 1.0
containing SF had a similar trend as that of the NAC, verifying that
the addition of SF could decrease the RAC porosity. However, Fig. 4 0.9
shows that the decrease rate in the compressive strength of the 0% SF 5% SF 10% SF
RSRAC without SF increases with the rubber amount, but the
rubberized NAC or RSRAC with SF exhibits the opposite behaviour
SF content
as the rubber amount increases. These observations indicate that Fig. 5. Influence of SF content on the RSRAC strength.
the addition of rubber can significantly decrease the compressive
strength of RAC; however, this negative effect can be reduced by
adding SF. compressive strength of the specimens in Groups RC-R5, RC-R10
To demonstrate the influence of SF content on the compressive and RC-R15, respectively. With an increase in SF content, the
strength of the RSRAC, Fig. 5 presents the relationship between the compressive strength of the RSRAC significantly increased.
SF content and the relative compressive strength, which is defined Compared with the results of incorporating 5% SF, the incorporation
as the ratio of the compressive strength of concerned specimen to of 10% SF caused an increase of 25.3%, 13.1%, 17.5% and 19.3% for the
that of the non-SF specimen in same group. As shown in Fig. 5, the specimens in Groups RC-R5, RC-R10, RC-R15 and RC-R20, respec-
addition of SF can increase the compressive strength of NAC and tively. Moreover, the compressive strength of Specimen RC-R5S10
RAC with steel fibre, and this improvement increases with (42.83 MPa) is greater than that of Specimen NC-S0f (40.7 MPa),
increasing SF content. The replacement of 5% of the cement with SF as shown in Table 5. This result indicates that if only the
increased the compressive strength of the NAC and RAC without compressive strength is considered, the RSRAC with 10% SF content,
rubber by 9.1% and 11.9%, respectively. When the SF content 100% RCA content and 5% rubber content is more suitable for use
increased to 10%, the compressive strength of the NAC and RAC than the steel-fibre NAC because both the filler effect and the
without rubber increased by 44.6% and 32.5%, respectively. pozzolanic effect of SF enhance the mechanical behaviour of the
For the RAC with rubber, the incorporation of 5% SF had little concrete, especially its compressive strength (Gonza lez-Fonteboa
effect on the concrete strength. As illustrated in Fig. 5, the incor- and Martínez-Abella, 2008). Although the RAC with rubber is
poration of 5% SF caused an increase of 1.0%, 4.4% and 9.0% for the more porous than the NAC, when the RSRAC is prepared with SF,
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 661

part of the SF penetrates into the pores, consequently improving


the ITZ bonding between the cement matrix and aggregates.
Furthermore, hydration process fills the cracks which originally
present in the RCA, resulting in an improvement in the mechanical
properties of the RSRAC.

4.2. Stressestrain curves

Stressestrain curves of the specimens were obtained from the


compression tests. The axial strains of the concrete in compression
were measured by using LVDTs, and the hoop strains were recorded
by the strain gauge placed in the middle of the specimen. For each
mix, three cylindrical specimens were tested, and the stressestrain
curve shown in Figs. 6 and 7 represents the average of the three
specimens.
Fig. 6 shows the stressestrain curves of the specimens with
different rubber contents. It can be seen from Fig. 6(a) that when
the content of SF is 0%, for the NAC, adding steel fibre resulted in a
Fig. 7. Stresseaxial strain curves of the mixtures with different SF contents: (a) SF
definite change in the curve shape. The stressestrain curve had a content: 0%, (b) SF content: 5%, (c) SF content: 10%.
slight change in the ascending stage, and the peak stress decreased,
but the descending stage clearly became flatter. These changes are
mainly because the steel fibre did not have a dominant effect in the the shape of the stressestrain curves. With an increase in SF con-
ascending stage, while the weak bonding between the steel fibre tent, incorporating 10% SF resulted in a considerable change in the
and paste resulted in the decrease in the peak stress; as the con- curve shape. In general, with the incorporation of SF, the
crete strain increased, in the descending stage, the steel fibre fully descending stage of both the NAC and the RSRAC became steeper,
exhibited its high-strength characteristics and increased the resis- indicating that SF may increase its brittleness, similar to the results
tance to cracking (Lau and Anson, 2006). For the steel-fibre NAC, of the previously reported studies for normal concrete (Tasdemir
Fig. 6(a) shows that the addition of SF resulted in a visible alteration et al., 1996).
in the curve shape, and the curve in the descending stage became The strains at the peak stresses (i.e., peak strain) from the
steeper as the SF content increased. However, for the RSRAC, as stressestrain curves are summarized in Table 6. For the NAC
shown in Fig. 6(b)-(f), the addition of 5% SF generally did not change specimens, the incorporation of steel fibre resulted in higher peak
strains, as shown in Table 6; this result was attributed to the
cracking resistance that the steel fibre provides (Guo et al., 2014).
Furthermore, the substitution of NCA by RCA resulted in higher
peak strains in the concrete. For the NAC, incorporating 5% SF
slightly decreased the strain at the peak stress, but its peak strain
did not change significantly with a further increase in SF content, as
observed in Table 6. For the RSRAC, with an increase in SF content,
the peak strain first decreased (the incorporation of 5% SF) and then
increased (the incorporation of 10% SF). These results indicate that
RAC has better ductility than NAC. This difference is mainly because
the RCA has a lower elastic modulus than that of the NCA and
because the incorporation of RCA increases the porosity of the
concrete (Cakir, 2014).
The stressestrain curves of the concretes with different SF
contents are shown in Fig. 7. Fig. 7(a) shows that a full replacement
of NCA by RCA led to a greater peak strain and a flatter post-peak
branch, indicating that the RAC exhibits not only a greater
ductility but also a less strength and stiffness than those of the NAC.
The addition of rubber resulted in a significant change in the curve
shape, not only reducing the peak stress but also resulting in a

Table 6
Peak strain test results.

Specimen Rubber content SF content

0% 5% 10%

NC 0% 0.00243 e e
NC-f 0% 0.00263 0.00231 0.00223
RC-R0 0% 0.00278 0.00234 0.00255
RC-R5 5% 0.00209 0.00206 0.00239
RC-R10 10% 0.00219 0.00185 0.00246
RC-R15 15% 0.00229 0.00187 0.00221
Fig. 6. Stresseaxial strain curves of the mixes with different rubber contents: (a) NAC a
RC-R20 20% 0.00167 0.00184
specimens, (b) Rubber content: 0%, (c) Rubber content: 5%, (d) Rubber content: 10%, (e)
a
Rubber content: 15%, (f) Rubber content: 20%. Data not obtained.
662 J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667

decrease in the peak strain, as demonstrated in Fig. 7 and Table 7. thereby making the concrete with rubber more deformable. As
With the increasing of the rubber content, the peak stress of the shown in Fig. 8, the increase of the rubber content from 5% to 20%
RSRAC clearly showed monotonic decrease, but the change in the resulted in a slight reduction in the elastic modulus from approx-
rubber content had little effect on the peak strain, which first imately 10%e25%. This reduction in the elastic modulus was less
decreased and subsequently increased slightly. Fig. 7(a) and (b) also than the decrease in the compressive strength.
show that the incorporation of rubber decreased the peak stress of Fig. 9 shows the relative elastic modulus as a function of the SF
the RSRAC but had a small effect on the slope of the descending addition into the concrete. The ordinate of Fig. 9 refers to the ratio
stage of the stressestrain curve for rubber contents up to 10%. With of the elastic modulus of concerned specimen to that of the spec-
the further increase of the rubber content, the post-peak branch of imen without SF in same group. The compressive strength of the
the stressestrain curve flattened. However, for the RSRAC with 10% NAC increased steadily as the SF content in the NAC increased from
SF, the incorporation of rubber content up to 20% caused the 0% to 10%. However, as shown in Fig. 9, the elastic modulus of the
opposite trend for the post-peak branch, i.e., it became steeper, as NAC slightly decreased when the SF content was 5% of the NAC and
shown in Fig. 7(c). Thus, the combination of a large amount of both then increased when the SF content exceeded 5%. These results
rubber and SF is not benefit for the compressive properties of the indicate that a large content of SF will increase the elastic modulus
RSRAC. of NAC, but a small content of SF has an insignificant effect on the
elastic modulus of NAC. Similar results were reported by Dilbas
4.3. Modulus of elasticity et al. (2014).
The elastic modulus of the RAC specimens, as shown in Table 7,
The test results of the elastic modulus of the specimens are were lower than that of the NAC specimens with equal SF contents.
presented in Table 7. The secant elastic modulus of the concrete was It is well known that the unfavourable properties of RCA, such as
obtained within one-third of the peak stress. cracks and higher porosity in the adhered low-rigidity mortar,
To present the effect of the addition of rubber, the relative decrease the elasticity modulus of RAC (Vadivel et al., 2014).
modulus of elasticity is illustrated in Fig. 8, which is defined as the However, the use of SF did not appear to have a significant impact
ratio of the elastic modulus of RSRAC to that of the non-rubber on the elastic modulus of RAC with or without rubber. As shown in
specimen in same group. Fig. 8 shows that the elastic modulus of Fig. 9, the change of elastic modulus of Groups RC-R0, RC-R5, and
all the RAC specimens decreased with the increasing of rubber RC-R10 are similar, with values between approximately 7% and
content. That is, the concrete with rubber was less rigid than the 13%, although their compressive strengths are significant different.
other concretes; these results are in accordance with the previous This result indicates that the stiffness of the RAC was not signifi-
studies (Naito et al., 2013). This effect is mainly because the elastic cantly influenced by the presence of mineral additions.
modulus of the rubber is considerably lower than that of the sand, A possible reason for the abovementioned difference in the SF
effect on the NAC and the RAC is because the presence and number
of voids in concrete substantially influences its stiffness. In high-
Table 7
porosity concrete, the stiffness is relatively low during compac-
Elastic modulus test results.
tion under compressive loading. After compaction, the deformation
Group Rubber content Elasticity modulus (GPa) of concrete generally depends on the high-stiffness phases,
0% SF 5% SF 10% SF including aggregates, ITZs and cement paste (Fonseca et al., 2011).
NC 0% 46.63 Thus, since RAC with 100% RA is more porous than NAC, the addi-
NC-f 0% 43.28 42.26 50.56 tion of SF cannot significantly reduce the negative effect of high
RC-R0 0% 37.23 42.05 39.23 porosity on the stiffness of the RAC, especially for the RAC with
RC-R5 5% 35.97 37.73 35.92 rubber, whereas the elastic modulus of the NAC with a lower
RC-R10 10% 35.17 36.82 33.70
porosity is influenced by the strengthening of the ITZs and cement
RC-R15 15% 31.44 35.46 31.29
RC-R20 20% a
31.35 30.17 paste caused by the presence of SF. In this context, the modulus of
a elasticity depends substantially on the kind of aggregate in the
Data not obtained.

Fig. 8. Influence of the rubber content on the elastic modulus of RSRAC. Fig. 9. Influence of the SF content on the elastic modulus.
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 663

concrete, as reported in Etxeberria et al. (2007) and Corinaldesi and seen from Table 8 and Fig. 10 that the incorporation of rubber
Moriconi (2009). generally decreased the toughness of the RAC. With the increasing
of rubber content, the RSRAC toughness varied, which first
4.4. Energy absorption capacity decreased and then increased (or flattened) and ultimately
decreased. When the RSRAC did not include SF or had 5% SF, the
Commonly, the energy absorption capacity of concrete is char- specific toughness of the RSRAC had a similar trend as that of the
acterized by its compressive toughness, which can be defined as the toughness with the increase in rubber content. However, for the
area under the compressive stressestrain curve (Nataraja et al., RSRAC with 10% SF, the specific toughness initially increased and
1999; Taerwe, 1993). Moreover, the specific toughness defined as then decreased. Thus, a proper rubber content can improve the
the ratio of the stressestrain curve area to the compressive strength specific toughness of the RSRAC with SF. This improvement is
was also discussed in this section. In this study, following Poon et al. mainly due to the role of rubber in absorbing energy. When crack
(2004), the toughness is calculated by the enveloped area of the tip touched the rubber in concrete, rubber can absorb the part of
stress-strain curve within 1.5% strain. The test results of compres- the energy causing crack further propagation, like a damper
sive toughness are listed in Table 8. (Turatsinze et al., 2006). However, rubberized concrete displays a
Table 8 shows that the incorporation of 1% steel fibre increased reduced overall density, which indicates the presence of internal air
the NAC toughness by 9% and the specific toughness by 22%. It is voids (Richardson et al., 2016). Therefore, excessive rubber content
well accepted that, during the compression process of concrete, the would negatively affect the energy dissipation capacity of the
steel fibre effectively decreases the stress concentration at the tip of concrete. This result is consistent with the findings of Khaloo et al.
the crack and helps to dissipate the energy (Guo et al., 2014). The (2008) and Turatsinze and Garros (2008). Moreover, either the
difference in toughness between Specimen NC-S0f and RC-R0S0 toughness or the specific toughness of the RSRAC underwent a
was 13.9%. However, the specific toughness of Specimen NC-S0f turning point with the increase in the rubber content. It can be
was close to that of Specimen RC-R0S0, attributable to the higher found from Table 8 and Fig. 10 that the RSRAC specimens with
compressive strength of Specimen NC-S0f. Therefore, RAC has a about 15% rubbers exhibited the excellent toughness and specific
high energy absorption capacity, which mainly is due to its porosity toughness, representing the appropriate rubber content for the
(Xiao et al., 2012). energy absorption capacity of the RSRAC.
The influences of the rubber content on the compressive To compare the influence of the SF content on the energy ab-
toughness of the RSRAC are shown in Fig. 10. In this figure, the sorption capacity, the relative toughness of the specimens with the
ordinate refers to the ratio of the toughness of concerned specimen increasing of SF is shown in Fig. 11, which is defined as the ratio of
to that of the non-rubber specimen with same SF content. It can be the toughness of concerned specimen to that of the specimen
without rubber in same group. For the NAC with steel fibre, as
shown in Table 8 and Fig. 11, the addition of SF only slightly
Table 8 decreased the toughness but significantly decreased the specific
Compressive toughness test results. toughness. Furthermore, with the increase in SF content, the spe-
Group Rubber content Toughness Specific toughness (%) cific toughness of NAC gradually decreased, verifying that the
(MPa  102) addition of SF increased the brittleness of the NAC. As demon-
0% SF 5% SF 10% SF 0% SF 5% SF 10% SF strated in Table 8, the RAC without rubber exhibited changes in
toughness with the increase in SF content similar to those of the
NC 0% 0.388 0.85
NC-f 0% 0.423 0.385 0.400 1.04 0.87 0.68 NAC.
RC-R0 0% 0.364 0.331 0.384 1.01 0.82 0.80 Interestingly, when the RAC included rubber, its toughness
RC-R5 5% 0.262 0.249 0.355 0.77 0.73 0.83 increased with the SF content except for the specimens with 20%
RC-R10 10% 0.196 0.199 0.286 0.69 0.67 0.85 rubber content, as shown in Fig. 11 and Table 8. When the rubber
RC-R15 15% 0.221 0.238 0.278 0.89 0.88 0.88
RC-R20 20% a
0.220 0.167 a
0.85 0.54
content was 5% and 10%, the addition of 10% SF significantly in-
a
creases the specific toughness. However, as the rubber content
Data not obtained.

Fig. 10. Influence of the rubber content on the concrete toughness. Fig. 11. Influence of the SF content on the RSRAC toughness.
664 J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667

exceeded 15%, the positive effect of the SF on the specific toughness


of the RSRAC decreased, indicating that the negative effect of a large
content of rubber cannot mitigated by the addition of SF. Moreover,
compared with the addition of 5% SF, the replacement of 10% SF
provides additional benefits for the energy absorption capacity of
the RSRAC with less than 15% rubber. Thus, if only considering the
specific toughness, incorporating both 10% SF and 15% rubber is
appropriate for RSRAC.

4.5. Failure mechanism


Fig. 13. Appearance of the local failure: (a) RSRAC without SF, (b) RSRAC with SF.
The failure modes of the cylindrical specimens are shown in
Fig. 12. Fig. 12(a)-(c) show that the addition of SF led to significant
changes in the cracks of both the NAC and the RAC. With the in- reasons for this observation are presented here. First, rubber par-
crease in the SF content, the number of cracks decreased, the width ticles are organic materials with hydrophobic surfaces; in contrast,
of the major macro-cracks increased, and the propagation direction the cement paste is an inorganic material with ionic compounds,
of the major macro-cracks became more vertical. Comparing and the surface of the cement paste has a strong water absorption
Fig. 12(a) and (c), it can be clearly demonstrated that major macro- ability (Richardson et al., 2012). The different material surface
cracks formed along the lengths of the concrete cylinders with 10% properties do not combine effectively. Second, the elastic modulus
SF, whereas multiple small longitudinal cracks can be observed on of the rubber particles is low, as aforementioned; these particles act
the concrete without SF. The phenomenon may be mainly attrib- as dampers and can decrease the stress concentrations at the crack
uted to the formation of stronger ITZs due to the pozzolanic and tips, resulting in a delay of the coalescence of the cracks (Turatsinze
filler effects of SF (Dilbas et al., 2014). Fig. 13(a) and (b) present the et al., 2006; Xie et al., 2015). Therefore, the interfacial bond be-
local failure of the RSRAC with SF and without SF, respectively. tween the rubber and cement paste is weak (due to the first
Fig. 13(a) shows that the failure surface of the RSRAC without SF reason); when the cracks reach a rubber particle, the propagation
was clearly loose, and the bonds between the aggregate, steel fibre direction of the crack possibly changes to turn along the interface,
and paste look very weak. When SF was added into the mixture, the since some of the fracture energy is absorbed by the rubber particle
paste of the RSRAC became denser, improving the bond between (due to the second reason). However, as aforementioned, the
the components in the concrete, as shown in Fig. 13(b). Thus, the addition of SF can improve the interfacial bond strength between
failure modes of the specimens with SF demonstrate that although the rubber and cement paste, allowing the cracks to propagate
the compressive strength of the RSRAC with SF could be improved, through rubber particles. Therefore, the addition of rubber can
the RSRAC became more brittle, similar to the NAC. change the distribution of cracks in concrete.
When the content of rubber was 15%, as shown in Fig. 12, the Based on the observations of the compressive loading process
crack widths in the concrete were small, the crack paths were more and failure mode, the crack propagation paths of the RSRAC with SF
twisted, and more cracks emerged on the surface of the concrete. and without SF are described in Fig. 14. As mentioned above, the
Therefore, the incorporation of rubber can affect the propagation of interfaces between the RCA, rubber, thick steel fibre and cement
cracks in concrete under compressive loading. Two possible paste are often weak in RSRAC. The strength of the RCA itself might

Fig. 12. Failure modes: (a) Specimens without SF, (b) Specimens with 5% SF, (c) Specimens with 10% SF.
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 665

used in this study.

X
n X
Eu ¼ Mi mij GWPj (1)
i j

where Eu is the total amount of greenhouse gas (GHG) (CO2, CH4,


HFC, PFC, and SF6) emissions produced in each unit process, Mi is
the consumption of the i-th energy or resource in each unit process,
mij is the emissions factor of the j-th GHG of the i-th energy or
resource, and GWPj is the global warming potential (GWP) of the j-
th GHG, which for CO2 is 1.
At present, carbon emissions make up the largest portion of
greenhouse gas emissions in concrete construction engineering.
Therefore, CO2 emissions are focused upon in this study, including
the three unit processes of RSRAC production, namely, the pro-
duction of raw materials, mixing, and curing.

5.2. CO2 emissions from the production of raw materials

The CO2 emissions from the production of raw materials are an


important cause of carbon emissions from concrete. Table 9 lists the
CO2 emissions factors of the raw materials used in RSRAC. It should
be noted that the CO2 emissions of silica fume were neglected
Fig. 14. Schematic description of the crack propagation in RSRAC: (a) RSRAC without
SF, (b) RSRAC with SF. because this material is a waste by-product arising from the use of
an industrial electric furnace (Turner and Collins, 2013). As shown
in Table 9, cement accounts for over 80% of the total amount of CO2
be higher than the interface strength. Based on the energetic emissions through the production of raw materials. Thus, among
viewpoint, a crack usually propagates along the path with mini- the raw materials of RSRAC, cement contributes the most carbon
mum energy consumption (Dilbas et al., 2014). Therefore, the emissions, which is similar with normal concrete.
propagation of internal cracks under compression is commonly not
through the RCA but along the interface between the RCA and the 5.3. CO2 emissions from concrete mixing and curing
paste, bypassing the steel fibre and rubber, which bear a certain
degree of resistance to cracking, as shown in Fig. 14(a). As a result, Commonly, CO2 emissions during the mixing of concrete are
interface fractures easily occur, resulting in the failure of RSRAC attributed to the energy consumption of the equipment used,
without SF. For the RSRAC with SF, the cracks commonly propagate mainly the concrete mixer and vibratory tamper. The emissions
through the RCA and rubber particles due to the increase in the factor of the construction equipment can be estimated based on the
interfacial bond strength but still bypass the steel fibre, as shown in local one-machine-team energy consumption. According to the
Fig. 14(b). Because of the high strength of steel fibre, many steel Chinese specification (2013), the CO2 emissions factors of a concrete
fibres are pulled out and bent during failure of the RSRAC. More- mixer and vibratory tamper are 8.4 and 3.9 in China, respectively.
over, as indicated in Fig. 14, the development of cracks is hindered In addition, the curing process of concrete through water
by the combined effect of the fibre and rubber, causing the cracks to maintenance introduces carbon emissions, which are also attrib-
propagate along a tortuous path. uted to the energy consumption. The CO2 emissions factor of tap
water is 0.91 kg/m3, which is considered acceptable, as reported by
5. Carbon emissions of RSRAC production Jiang et al. (2014). In accordance with the curing method presented
in this study, when RSRAC is watered for a total of 14 days twice a
With the development of construction industrialization, the day (comsuming 200 ml water each time for one specimen), the
emissions of greenhouse gases, which have been given priority over CO2 emissions derived from the curing process of RSRAC can be
carbon dioxide, caused by cement-based materials have been calculated as 0.9 kg/m3.
receiving greater attention. Thus, in addition to the mechanical
performance of RSRAC, as a new cement-based concrete, it is also 5.4. Total CO2 emissions
imperative to estimate the carbon emissions of RSRAC production.
The total CO2 emissions of RSRAC are shown in Table 10, as based
5.1. Calculation model on Eq. (1). It can be seen from Table 10 that the addition of rubber
and steel fibre increases the CO2 emissions of RSRAC, with the
Following Li and Chen (2017), a model for estimating the carbon former having a smaller effect on the CO2 emissions than the latter.
emissions of the unit processes applied to concrete production was However, the incorporation of silica fume offsets the increase in

Table 9
Emissions factors of raw materials of RSRAC.

Materials Cement (Luo Sand (Luo et al., NCA (Luo et al., RCA (Luo et al., Water (Jiang Superplasticizer (Jiang Rubber (Gustavsson Steel fibre (Li and
et al., 2016) 2016) 2016) 2016) et al., 2014) et al., 2014) et al., 2010) Chen, 2017)

Emissions 685.74 kg/t 3.19 kg/m3 4.00 kg/m3 2.60 kg/m3 0.91 kg/m3 28.49 kg/t 133.34 kg/t 220.60 kg/t
factors
666 J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667

Table 10 the RCA and rubber particles due to the increase in the inter-
CO2 emissions of RSRAC. facial bond strength but still bypassed the steel fibre.
Group Rubber content CO2 emissions (kg/m3) 6. The incorporation of silica fume and RCA offset the rise of the
0% SF 5% SF 10% SF
CO2 emission caused by the addition of rubber and steel fibre.
Based on the synthetical consideration of compressive proper-
NC 0% 257.9
ties and carbon emissions, RSRAC with 10% SF content and 5%
NC-f 0% 275.1 262.5 250.5
RC-R0 0% 274.4 261.9 249.9 rubber content is a more environmentally friendly alternative to
RC-R5 5% 276.9 264.4 252.4 normal concrete for use in the compression member of concrete
RC-R10 10% 278.8 266.3 254.3 structures.
RC-R15 15% 280.7 268.1 256.2
RC-R20 20% 282.6 270.1 258.1
Acknowledgments

The authors gratefully acknowledge the financial support pro-


CO2 emissions caused by the addition of rubber and steel fibre. As vided by National Natural Science Foundation of China (No.
demonstrated in Table 10, the CO2 emissions of RSRAC significantly 11602060), the Science and Technology Planning Project of
decrease with the increase in amount of silica fume, indicating that Guangdong Province (No. 2017B020238006), and the Science and
silica fume not only improves the mechanical performance of Technology Project Foundation of Guangzhou (Nos. 201704030057
RSRAC, it also reduces the carbon emissions of concrete to a certain and 201707010364)).
extent (Turner and Collins, 2013). Additionally, when recycled ag-
gregates completely replace natural aggregates, a slight decrease in
carbon emissions from RSRAC can be found, as indicated in References
Table 10, although the carbon emissions of RCA itself are signifi-
Afroughsabet, V., Biolzi, L., Ozbakkaloglu, T., 2016. High-performance fiber-
cantly less than those of NCA, as shown in Table 9. This is mainly reinforced concrete: a review. J. Mater. Sci. 51 (14), 6517e6551.
due to the fact that the carbon emissions of cement are dominant, Al-Masoodi, A.H.H., Kawan, A., Kasmuri, M., Hamid, R., Khan, M.N.N., 2016. Static
and are significantly greater than those of other raw materials. and dynamic properties of concrete with different types and shapes of fibrous
reinforcement. Construct. Build. Mater. 104, 247e262.
Compared with NAC using steel fibre, RSRAC with 10% silica fume Al-Tayeb, M.M., Abu Bakar, B.H., Akil, H.M., Ismail, H., 2012. Effect of partial re-
and 5% rubber can not only save natural resources and reduces the placements of sand and cement by waste rubber on the fracture characteristics
land used required by concrete waste, it also reduces CO2 emissions of concrete. Polym. Plast. Technol. Eng. 51 (6), 583e589.
Antil, Y., Verma, E.V., Singh, E.B., 2014. Rubberized concrete made with crumb
by over 8%. Thus, RSRAC can bring about considerable environ- rubber. Int. J. Sci. Res. 3 (5), 1481e1483.
mental benefits. Aslani, F., 2016. Mechanical properties of waste tire rubber concrete. J. Mater. Civ.
Eng. 28 (3), 4015152.
Cakir, O., 2014. Experimental analysis of properties of recycled coarse aggregate
(RCA) concrete with mineral additives. Construct. Build. Mater. 68, 17e25.
6. Conclusions Carneiro, J.A., Lima, P.R.L., Leite, M.B., Toledo, R.D., 2014. Compressive stress-strain
behavior of steel fiber reinforced-recycled aggregate concrete. Cement Concr.
Compos. 46, 65e72.
The use of scrap rubber and waste concrete as aggregates to Chakradhara Rao, M., Bhattacharyya, S.K., Barai, S.V., 2011. Behaviour of recycled
produce environment-friendly concrete is a promising technology. aggregate concrete under drop weight impact load. Construct. Build. Mater. 25
The aim of this research was to investigate the coupling effects of (1), 69e80.
Chen, B., Pan, H.Y., 2013. Studies on the effect of micro silica powder on some
adding SF and rubber particles on the compressive behaviour of properties of high-strength steel fiber concrete. Concrete 70, 18e20.
RSRAC. Moreover, the carbon emission of RSRAC production was China specification, 2013. Pricing Specification of Bill of Quantities for Construction
estimated. Based on the above results, the following conclusions Project (GB50500). China Plans Publisher, Beijing, China.
Chinese standard, 2007. Common Portland Cement (GB175). China Architecture and
can be drawn:
Building Press, Beijing, China.
Chinese standard, 2016. Standard for Test Method of Performance on Ordinary
1. The RSRAC with silica fume shows good compressive behaviour Fresh Concrete (GB/T 50080). China Architecture&Building Press, Beijing, China.
Chinese standard, 2017. Mineral admixtures for High Strength and High Perfor-
due to the superior coupling effect of silica fume, steel fibres and
mance Concrete (GB/T18736). China Architecture and Building Press, Beijing,
rubber particles. China.
2. The interfaces between the aggregate, cement matrix and steel Corinaldesi, V., Moriconi, G., 2009. Influence of mineral additions on the perfor-
fibre with a low aspect ratio in the RSRAC without SF are mance of 100% recycled aggregate concrete. Construct. Build. Mater. 23 (8),
2869e2876.
generally weak. The incorporation of SF can improve the Dilbas, H., Simsek, M., Cakir, O., 2014. An investigation on mechanical and physical
bonding properties, resulting in an increase of the compressive properties of recycled aggregate concrete (RAC) with and without silica fume.
strength of RSRAC. Moreover, the compressive strength in- Construct. Build. Mater. 61, 50e59.
Etxeberria, M., Vazquez, E., Marí, A., Barra, M., 2007. Influence of amount of recycled
creases with the SF content. coarse aggregates and production process on properties of recycled aggregate
3. The compressive strength of RSRAC markedly decreases with concrete. Cement Concr. Res. 37 (5), 735e742.
the increase of rubber content, but this negative effect can be Evangelista, L., de Brito, J., 2014. Concrete with fine recycled aggregates: a review.
Eur. J. Environ. Civ. Eng. 18 (2), 129e172.
significantly reduced by the incorporation of SF when the rub- Fonseca, N., de Brito, J., Evangelista, L., 2011. The influence of curing conditions on
ber content is less than 20%. the mechanical performance of concrete made with recycled concrete waste.
4. With the addition of rubber into the RSRAC, the decrease in the Cement Concr. Compos. 33 (6), 637e643.
Gesoglu, M., Guneyisi, E., Khoshnaw, G., Ipek, S., 2014. Investigating properties of
elastic modulus was smaller than that in the compressive pervious concretes containing waste tire rubbers. Construct. Build. Mater. 63,
strength. A large content of SF, replacing up to 10%, considerably 206e213.
increased the elasticity modulus of the NAC, but slightly Gheni, A.A., Abdelkarim, O.I., Abdulazeez, M.M., ElGawady, M.A., 2017. Texture and
design of green chip seal using recycled crumb rubber aggregate. J. Clean. Prod.
increased the stiffness of the RSRAC. This difference may be
166, 1084e1101.
attributable to the high porosity of the RSRAC. Gonz alez-Fonteboa, B., Martínez-Abella, F., 2008. Concretes with aggregates from
5. The propagation of the internal cracks in the RSRAC without SF demolition waste and silica fume. Materials and mechanical properties. Build.
under compression commonly occurred along the interface Environ. 43 (4), 429e437.
Güneyisi, E., Gesog €
lu, M., Ozturan, T., 2004. Properties of rubberized concretes
between the RCA and paste, bypassing the steel fibre. With the containing silica fume. Cement Concr. Res. 34 (12), 2309e2317.
incorporation of SF, the cracks commonly propagated through Guo, Y.C., Zhang, J.H., Chen, G.M., Xie, Z.H., 2014. Compressive behaviour of concrete
J. Xie et al. / Journal of Cleaner Production 197 (2018) 656e667 667

structures incorporating recycled concrete aggregates, rubber crumb and Environ. 5 (1), 46e82.
reinforced with steel fibre, subjected to elevated temperatures. J. Clean. Prod. Richardson, A., Coventry, K., Edmondson, V., Dias, E., 2016. Crumb rubber used in
72, 193e203. concrete to provide freezeethaw protection (optimal particle size). J. Clean.
Gustavsson, L., Joelsson, A., Sathre, R., 2010. Life cycle primary energy use and Prod. 112, 599e606.
carbon emission of an eight-storey wood-framed apartment building. Energy Richardson, A.E., Coventry, K.A., Ward, G., 2012. Freeze/thaw protection of concrete
Build. 42 (2), 230e242. with optimum rubber crumb content. J. Clean. Prod. 23 (1), 96e103.
Henry, M., Pardo, G., Nishimura, T., Kato, Y., 2011. Balancing durability and envi- Shi, C.J., Li, Y.K., Zhang, J.K., Li, W.G., Chong, L.L., Xie, Z.B., 2016. Performance
ronmental impact in concrete combining low-grade recycled aggregates and enhancement of recycled concrete aggregate - a review. J. Clean. Prod. 112,
mineral admixtures. Resour. Conserv. Recycl. 55 (11), 1060e1069. 466e472.
Holmes, N., Dunne, K., O'Donnell, J., 2014. Longitudinal shear resistance of com- Sivakumar, N., Muthukumar, S., Sivakumar, V., Gowtham, D., Muthuraj, V., 2014.
posite slabs containing crumb rubber in concrete toppings. Construct. Build. Experimental studies on high strength concrete by using recycled coarse
Mater. 55, 365e378. aggregate. Int. J. Eng. Sci. 4 (01), 27e36.
Jiang, J.J., Ye, B., Ma, X.M., 2014. The construction of Shenzhen‫׳‬s carbon emission Taerwe, L.R., 1993. Influence of steel fibers on strain-softening of high-strength
trading scheme. Energy Pol. 75, 17e21. concrete. Mater. J. 89 (1), 54e60.
Katzer, J., Domski, J., 2013. Optimization of fibre reinforcement for waste aggregate Taha, M.M.R., El-Dieb, A.S., El-Wahab, M.A.A., Abdel-Hameed, M.E., 2008. Me-
cement composite. Construct. Build. Mater. 38, 790e795. chanical, fracture, and microstructural investigations of rubber concrete.
Khaloo, A.R., Dehestani, M., Rahmatabadi, P., 2008. Mechanical properties of con- J. Mater. Civ. Eng. 20 (10), 640e649.
crete containing a high volume of tireerubber particles. Waste Manag. 28 (12), Tai, J., An, M., Wang, X.D., 2017. Annual Research Report on the Development of
2472e2482. Urban Environmental and Sanitation Industry in China: 2015-2016. Shanghai
Khatib, Z.K., Bayomy, F.M., 1999. Rubberized Portland cement concrete. J. Mater. Civ. Jiao Tong University press, Shanghai.
Eng. 11 (3), 206e213. Tam, V.W.Y., Tam, C.M., Gao, X.F., 2005. Microstructural analysis of recycled
Kou, S.C., Poon, C.S., Agrela, F., 2011. Comparisons of natural and recycled aggregate aggregate concrete produced from two-stage mixing approach. Cement Concr.
concretes prepared with the addition of different mineral admixtures. Cement Res. 35 (6), 1195e1203.
Concr. Compos. 33 (8), 788e795. Tasdemir, C., Tasdemir, M.A., Lydon, F.D., Barr, B.I.G., 1996. Effects of silica fume and
Kurad, R., Silvestre, J.D., de Brito, J., Ahmed, H., 2017. Effect of incorporation of high aggregate size on the brittleness of concrete. Cement Concr. Res. 26 (1), 63e68.
volume of recycled concrete aggregates and fly ash on the strength and global Thomas, B.S., Chandra Gupta, R., 2016. Properties of high strength concrete con-
warming potential of concrete. J. Clean. Prod. 166, 485e502. taining scrap tire rubber. J. Clean. Prod. 113, 86e92.
Lau, A., Anson, M., 2006. Effect of high temperatures on high performance steel Thomas, B.S., Gupta, R.C., Panicker, V.J., 2016. Recycling of waste tire rubber as
fibre reinforced concrete. Cement Concr. Res. 36 (9), 1698e1707. aggregate in concrete: durability-related performance. J. Clean. Prod. 112,
Li, L., Chen, K., 2017. Quantitative assessment of carbon dioxide emissions in con- 504e513.
struction projects: a case study in Shenzhen. J. Clean. Prod. 141, 394e408. Topçu, I.B., 1995. The properties of rubberized concretes. Cement Concr. Res. 25 (2),
Li, L.J., Ruan, S.H., Zeng, L., 2014. Mechanical properties and constitutive equations 304e310.
of concrete containing a low volume of tire rubber particles. Construct. Build. Topcu, I.B., Bilir, T., 2009. Experimental investigation of some fresh and hardened
Mater. 70, 291e308. properties of rubberized self-compacting concrete. Mater. Des. 30 (8),
Liu, F., Chen, G.X., Li, L.J., Guo, Y.C., 2012. Study of impact performance of rubber 3056e3065.
reinforced concrete. Construct. Build. Mater. 36, 604e616. Turatsinze, A., Garros, M., 2008. On the modulus of elasticity and strain capacity of
Liu, F., Zheng, W.H., Li, L.J., Feng, W.X., Ning, G.F., 2013. Mechanical and fatigue Self-Compacting Concrete incorporating rubber aggregates. Resour. Conserv.
performance of rubber concrete. Construct. Build. Mater. 47, 711e719. Recycl. 52 (10), 1209e1215.
Luo, Z., Yang, L., Liu, J., 2016. Embodied carbon emissions of office building: a case Turatsinze, A., Granju, J.L., Bonnet, S., 2006. Positive synergy between steel-fibres
study of China's 78 office buildings. Build. Environ. 95, 365e371. and rubber aggregates: effect on the resistance of cement-based mortars to
Naito, C., States, J., Jackson, C., Bewick, B., 2013. Assessment of crumb rubber con- shrinkage cracking. Cement Concr. Res. 36 (9), 1692e1697.
crete for flexural structural members. J. Mater. Civ. Eng. 26 (10), 04014075. Turner, L.K., Collins, F.G., 2013. Carbon dioxide equivalent (CO2-e) emissions: a
Najim, K.B., Hall, M.R., 2012. Mechanical and dynamic properties of self-compacting comparison between geopolymer and OPC cement concrete. Construct. Build.
crumb rubber modified concrete. Construct. Build. Mater. 27 (1), 521e530. Mater. 43, 125e130.
Nataraja, M.C., Dhang, N., Gupta, A.P., 1999. Stressestrain curves for steel-fiber Vadivel, T.S., Thenmozhi, R., Doddurani, M., 2014. Experimental behaviour of waste
reinforced concrete under compression. Cement Concr. Compos. 21 (5e6), tyre rubber aggregate concrete under impact loading. Iran. J. Sci. Technol. 38
383e390. (C1), 251e259.
Nazarimofrad, E., Shaikh, F., Nili, M., 2008. Effects of steel fibre and silica fume on Valadares, F., Bravo, M., de Brito, J., 2012. Concrete with used tire rubber aggregates:
impact behaviour of recycled aggregate concrete. J. Sustain. Cem. Mater. 1 (6), mechanical performance. ACI Mater. J. 109 (3), 283e292.
54e68. Wang, H.T., Wang, L.C., 2013. Experimental study on static and dynamic mechanical
Nguyen, T.H., Toumi, A., Turatsinze, A., 2010. Mechanical properties of steel fibre properties of steel fiber reinforced lightweight aggregate concrete. Construct.
reinforced and rubberised cement-based mortars. Mater. Des. 31 (1), 641e647. Build. Mater. 38, 1146e1151.
Onuaguluchi, O., Panesar, D.K., 2014. Hardened properties of concrete mixtures Won, J.P., Hong, B.T., Choi, T.J., Lee, S.J., Kang, J.W., 2012. Flexural behaviour of
containing pre-coated crumb rubber and silica fume. J. Clean. Prod. 82, 125e131. amorphous micro-steel fibre-reinforced cement composites. Compos. Struct. 94
Patra, R.K., Mukharjee, B.B., 2017. Influence of incorporation of granulated blast (4), 1443e1449.
furnace slag as replacement of fine aggregate on properties of concrete. J. Clean. Xiao, J., Li, J., Zhang, C., 2005. Mechanical properties of recycled aggregate concrete
Prod. 165, 468e476. under uniaxial loading. Cement Concr. Res. 35 (6), 1187e1194.
Pedro, D., de Brito, J., Evangelista, L., 2017. Mechanical characterization of high Xiao, J., Li, W., Fan, Y., Huang, X., 2012. An overview of study on recycled aggregate
performance concrete prepared with recycled aggregates and silica fume from concrete in China (1996e2011). Construct. Build. Mater. 31 (1), 364e383.
precast industry. J. Clean. Prod. 164, 939e949. Xie, J.H., Guo, Y.C., Liu, L.S., Xie, Z.H., 2015. Compressive and flexural behaviours of a
Poon, C.S., Shui, Z.H., Lam, L., 2004. Compressive behavior of fiber reinforced high- new steel-fibre-reinforced recycled aggregate concrete with crumb rubber.
performance concretesubjected to elevated temperatures. Cement Concr. Res. Construct. Build. Mater. 79, 263e272.
34, 2215e2222. Yoo, D.Y., Lee, J.H., Yoon, Y.S., 2013. Effect of fiber content on mechanical and frac-
Rashad, A.M., 2016. A comprehensive overview about recycling rubber as fine ture properties of ultra high performance fiber reinforced cementitious com-
aggregate replacement in traditional cementitious materials. Int. J. Sust. Built. posites. Compos. Struct. 106, 742e753.