lable at ScienceDirect

Journal of Cleaner Production 72 (2014) 193e203

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Compressive behaviour of concrete structures incorporating recycledconcrete aggregates, rubber crumb and reinforced with steel fibre,subjected to elevated temperatures

Yong-chang Guo a,*, Jian-hong Zhang a, Guang-ming Chen a, Zhi-hong Xie b

a School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou, Chinab School of Civil Engineering and Transportation, South China University of Technology, Guangzhou, China

a r t i c l e i n f o

Article history:Received 16 September 2013Received in revised form12 February 2014Accepted 15 February 2014Available online 1 March 2014

Keywords:Rubber crumbSteel fibreRecycled concrete aggregateCompressive propertiesHigh temperature

* Corresponding author. Tel.: þ86 20 39322538; faE-mail address: guoyc@gdut.edu.cn (Y.-c. Guo).

http://dx.doi.org/10.1016/j.jclepro.2014.02.0360959-6526/� 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

In this paper, effects of elevated temperatures on the compressive behaviour of rubber crumb and steelfibre reinforced recycled aggregate concrete (RSRAC) are presented. RSRAC is a new concrete materialproposed by the authors. In the RSRAC, steel fibre is used to improve the performances of concrete beforeexposure (e.g. ductility, cracking) and after exposure (explosive spalling) to evaluated temperature, andthe inclusion of rubber particles is mainly for the consideration of environment protection and reducingthe risk of spalling after exposure to high temperatures. A series of concrete mixes were prepared withOrdinary Portland Cement (OPC), recycled concrete coarse aggregates (RCA) or natural coarse aggregates(NCA), 1% steel fibre (by volume) and rubber particles with different fine aggregate (sand) replacementratios. The compressive properties, including compressive strength, Young’s modulus (stiffness), stressestrain curves and energy absorption capacity (toughness) of the different concrete mixes subjected toelevated temperatures (25 �C, 200 �C, 400 �C and 600 �C), were obtained in accordance to ASTM stan-dards. The results of weight loss and failure modes were recorded and presented in this study. The resultsshowed that both the compressive strength and stiffness of concrete mixes decreased after exposure toelevated temperature, with higher replacement of fine aggregate by rubber leading to lower compressivestrength and stiffness magnitude. Nevertheless, rubber crumbs significantly enhanced the energy ab-sorption capacity and explosive spalling resistance.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Waste concrete, often referred to recycled concrete aggregate(RCA), has been reused as a replacement of the natural aggregatefor new concrete mainly for the consideration of environmentalbenefit and effective utilization of resources. Various authors havestudied the properties of concrete preparedwith RCA. However, theincorporation of RCA leads to a significant loss of fluidity of themixture (Mefteh et al., 2013) caused by the attachedmortar contentof the RCA. This reduction certainly can be compensated by water-reducing admixtures (Barbudo et al., 2013). It is also known that theuse of recycled aggregates in concrete decreases its strength andYoung’s modulus compared to those of natural aggregate concrete(Miguel and de Brito, 2012). Poon et al. (2002) reported that thereplacement of coarse and fine natural aggregates by RCA (Recycled

x: þ86 20 39322511.

Concrete Aggregate) at higher levels (e.g. 50% or above) signifi-cantly reduced the compressive strength; while an air-driedaggregate that contained not more than 50% of RCA was optimalfor producing the Recycled Aggregate Concrete (RAC) with normalstrength (less than 60MPa) (Poon et al., 2004a,b). It is worth notingthat various methods have been attempted to compensate for thelower quality (e.g. lower strength, less durability) of concreteproducts with recycled aggregates and good results have beenachieved. Kou and Poon (2009) pointed out that the properties(mainly the compressive strength and tensile splitting strength) ofthe self-compacting concretes made from river sand and crushedfine recycled concrete aggregates (with 0, 25%, 50%, 75% and 100%replacement rates) showed only slight differencewith the inclusionof fly ash, demonstrating the feasibility of utilizing fine and coarserecycled concrete aggregates together with fly ash for self-compacting concretes. It has also been shown that the negativeeffect of RCA on durability properties of mixes can be mitigated byincorporating a certain amount of mineral admixtures, such as flyash and volcanic ash (Kou and Poon, 2012). These research results

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203194

have clearly promoted the promising use of RCA in construction. Todate, RAC has been successfully applied in pavements and buildingstructures in China, as shown by Li et al. (2009).

Steel fibre reinforced concrete (SFRC) was recognised to improvethe brittleness and lower tensile capacity of plain concrete. Thestudies showed steel fibres inside concrete matrix can increase thetoughness and cracking resistance of concrete mainly due to thebridging/tying effects of steel fibres on surrounding concrete, buthave little effect on the compressive behaviour of concrete probablybecause of the reduction/loss of the above effects in concrete undercompression (Atis and Karahan, 2009; Olivito and Zuccarello, 2010).Yang et al. (2006) and Gao et al. (2007) showed that when recycledaggregate concrete is reinforced with a certain amount of steel fi-bres, its compressive performance is similar or slightly lower thanthe natural aggregate concrete reinforced with equivalent amountof steel fibres, but significantly higher than ordinary plain concrete.This suggests that steel fibre reinforced recycled aggregate concretemay be used to replace ordinary concrete in the construction ofstructural members. Furthermore, steel fibres have been exten-sively used to improve the ductility of concrete. It has been foundthat steel fibres can reduce spalling and cracking and improve theresidual strength of concrete after exposure to elevated tempera-tures (Peng et al., 2006; Poon et al., 2004a,b). In particular, Poonet al. (2004a,b) showed that the energy dissipation capacity(toughness) of SFRC subjected to high temperatures can be almosttwo times that of plain concrete. Existing research also indicatedthat when steel fibre content is higher than 1.5% by volume of theconcrete, the increase of steel fibre content results in littleimprovement or even reduction of the above performances ofconcrete (e.g. residual strength, toughness) (Lau and Anson, 2006).As a result, many of the current studies of steel fibre reinforcedconcrete used around 1.0% steel fibres.

The fast development of automotive industry after the SecondWorld War has led to the rapid accumulation of waste tire rubber.Waste tire rubber is extremely difficult to degrade in landfilltreatment. As a result, the treatment of waste tire rubber hasrecently become a world-known environmental problem. Existingstudies showed that concrete performances can be significantlyimproved by including recycled rubber crumbs obtained fromwaste tires into the basic concrete composition (Hernández-Olivares and Barluenga, 2004; Lau and Anson, 2006; Hernández-Olivares et al., 2002; Son et al., 2011; Khaloo et al., 2008)Hernández-Olivares et al. (2002) showed that a small volumetricfraction of crushed tire rubber crumbs are of great contribution tothe dynamic behaviour of concrete under low-frequency dynamicactions. Mustafa Maher Al-Tayeb et al. (2013) received the similarconclusion that the use of hybrid rubberized concrete beam im-proves flexural impact performance of the beam during dynamicloading compared to static loading. Moreover, the addition ofrubber improved the toughness and deformation ability of thenormal concrete. Son et al. (2011) found that the rubber crumbsmay greatly improve the deformation capacity of the concretealthough the compressive strength of concrete may be slightlyreduced. Khaloo et al. (2008) indicated that the brittleness ofconcrete can be significantly decreased with increasing rubbercontent, with the crack width and crack propagation velocity in therubberized concrete (i.e. concrete with rubber content) beingobviously lower than those of plain concrete. Li et al. (2009) alsoobtained the similar conclusions in their experimental study onhigh strength concrete filled by recycled rubber. Furthermore, it hasbeen found that rubber crumbs can effectively reduce the risk ofexplosive spalling and strength loss rate of concrete after exposureto elevated temperatures (Hernández-Olivares and Barluenga,2004). This is because rubber crumbs, if burnt after exposure toevaluated temperatures, can release space for the escaping of water

vapour in concrete and thus protect the concrete body fromexplosive spalling (Li et al., 2011). Apparently, the inclusion ofrubber in concrete composition not only reduces the risk ofexplosive spalling and strength loss rate for concrete subjected toelevated temperatures, but also has a significant environmentadvantage as mentioned above.

Recently, it has been found that rubber content had no adverseimpact on the bridging and tying effects of steel fibres on sur-rounding concrete and the positive synergy between steel fibresand rubber particles has the advantage of enhancing the resistanceto shrinkage cracking (Turatsinze et al., 2006) and improving thefracture behaviours even subjected to elevated temperature (Guoet al., 2014).

Against the above background, rubberized steel fibre reinforcedrecycled aggregate concrete (RSRAC) was proposed by the authors(China invention patent No.: ZL. 201010019345.3). This new type ofmaterial has been coined based on the following considerations: 1)the steel fibre is used to improve the performances of concrete bothbefore exposure (e.g. toughness, ductility, cracking) and afterexposure (explosive spalling) to evaluated temperatures, 2) theinclusion of rubber particles is mainly for the consideration ofenvironmental protection and reducing the risk of spalling afterexposure to high temperatures, and 3) the beneficial interactionexists between steel fibre and rubber as mentioned above. Theenhanced ductility and resistance to crack of RSRAC make it suit-able in structures subjected to dynamic load, such as the pavementof road and bridge, while its improved resistance to explosivespalling makes it useful in fire-resistant structures. Several series oftests have been conducted in the authors’ research group toinvestigate the different behaviours of the proposed RSRAC. Thispaper presents the study on the effects of crumb rubber content onthe compressive behaviours (residual strength, Young’s modulus,stressestrain relationship and energy dissipation ability) of RSRACsubjected to elevated temperatures. From test results presented inthis paper, a preliminary understanding of the compressive failuremechanism of RSRAC after exposure to elevated temperatures canbe achieved. This study thus provides a basis for the furtherresearch on RSRAC and its potential applications.

2. Experimental details

A total of 6 groups of specimens, each consisting of 12 standardcylinders with dimensions of 150 mm � 300 mm (diameter andheight), were designed and prepared in this research. In thefollowing context, basic properties of the constituent materialsused, their mix proportioning, specimen preparation proceduresand loading scheme will be explained.

2.1. Materials

The cementitious material used in this study was ordinaryPortland cement with a strength of 42.5 MPa according to Chinesestandard GB175-2007. Fine aggregates were naturally sourcedmedium-coarse rive sand with a specific gravity of 2.69, a finenessmodulus of 2.52 and water absorption rate of 0.8%. Natural coarseaggregates were obtained from limestone and had a maximumparticle diameter of 12.5 mm. Recycled concrete coarse aggregatesused in the present study (with aggregate size ranging from 4.75 to12.5 mm) were made from crushed waste concrete. The water ab-sorption rates of the natural coarse aggregate and the recycledcoarse aggregate are 0.76% and 3.82% respectively, and the specificgravity of them is separately 2.65 and 2.43. Crumb rubber used inthis study was obtained from waste tires through the process ofcrushing, cleaning and screening; the rubber has an average par-ticle diameter corresponding to 14e20 sieve size (i.e. 0.85e

Fig. 1. Three materials used in preparing the concrete mixes.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203 195

1.40 mm according to ASTM-E11-09e1 (ASTM, 2009)), a specificgravity of 1.05, and a melting temperature of 170 �C. The steel fibresused were shear-wave type steel fibres with a length of 32 mm, anaspect ratio of 45 and a tensile strength of 600 MPa. This type ofsteel fibres, which were made from ordinary steel, with a meltingtemperature of 1538 �C and a density of 7.82 g/cm3, are loose inform at delivery as shown in Fig. 1. In addition, a commerciallyavailable naphthalene-based super-plasticizer with a solid contentof 30% and a water-reducing rate of 20% was used as admixture toachieve the requiredworkability of the concrete mixes. The amountof plasticizer was determined as 1.0% by weight of cement based onslump tests according to BS 1881: Part 102 (BS, 1983). Appearancesof recycled concrete aggregates, crumb rubber and steel fibres areshown in Fig. 1.

2.2. Mix proportioning

In this study, a total of 6 concrete mixes, each with a differentmix proportion, were designed and prepared to study the effect ofcrumb rubber content on the axial compressive behaviour of RSRACmixes subjected to elevated temperatures (25 �C, 200 �C, 400 �Cand 600 �C). The coarse aggregates of the first mix were the naturalaggregates; while in the remaining 5 mixes, recycled concrete ag-gregates were used to fully replace the natural coarse aggregates byvolume, which is referred to as recycled concrete with or withoutrubber crumb in the following context. In the five recycled concretemixes, the crumb rubber content was followed by 0%, 4%, 8%, 12%and 16% (by volume) of sands. The 6 concrete mixes used the samewater-to-cement ratio of 0.35 and contained 1.0% steel fibres by theconcrete volume as mentioned previously. Furthermore, additional3.82% water (by the weight of RCA) was added to the 5 rubberizedconcrete mixes to cater for the higher water absorption of RCA.Details of the mix proportions are summarized in Table 1.

2.3. Specimen preparation

The concrete mixes mentioned above were prepared in a con-crete mixer. For each of the concrete mixes, 12 standard cylinders of150 mm in diameter and 300 mm in height were cast using plasticmolds. Concrete cylinders of the same mix were mixed and castedin the same batch to ensure the uniformity. The procedure of pre-paring the concrete mixture is as follows, with a reference to ASTMC 192 (ASTM, 2006). Coarse aggregates and steel fibres were firstadded to the mixer followed by approximately one third of waterrequired, then the mixer was started and the mixing continued for1.5 min until sands, crumb rubber and cement were added to therotating mixer gradually, after which the mixing continued foranother 1.5 min. The rest water mixed with super-plasticizers wasadded to the mixer, mixing continued for 2 min. Fresh mixes weremeasured for workability by concrete slump test according to ASTMC 143. After being casted, concrete specimens were covered withplastic membrane sheets and kept in the laboratory at the roomtemperature for 24 h. Then the specimens were removed from themolds and cured in still water at 23 �C for 28 days, and thenconditioned in an environmental chamber at a temperature of 25 �Cand a relative humidity of 75% for another 60 days before beingheated to the prescribed temperatures.

2.4. Test method

Among 12 cylindrical specimens in each mix, three were testedimmediately after the conditioning without being heated (at theroom temperature of 25 �C), the remaining 9 specimens weredivided into 3 groups and subjected to 3 temperature exposureconditions (200 �C, 400 �C and 600 �C) in an electrical furnace. In

Table 1Mix proportions.

Mix Mix proportions (unit weight: kg/m3)

W/C W OPC S NCA RCA AW SF R WRA

NC-R0 0.35 170 485 645 1052 e e 78 e 4.85RC-R0 0.35 170 485 645 e 954 37 78 e 4.85RC-R4 0.35 170 485 625 e 954 37 78 7.9 4.85RC-R8 0.35 170 485 605 e 954 37 78 15.7 4.85RC-R12 0.35 170 485 585 e 954 37 78 23.6 4.85RC-R16 0.35 170 485 565 e 954 37 78 31.5 4.85

Note: NC ¼ natural concrete, RC ¼ recycled concrete, R0, R4, R8, R12 and R16 forvolume substitution ratio of rubber is 0%, 4%, 8%, 12% and16%, W/C ¼ water/cementratio (mass), W ¼ water, OPC ¼ ordinary Portland cement, S ¼ sand, NCA ¼ naturalcoarse aggregate, RCA ¼ recycled concrete aggregate, AW ¼ additional water,SF ¼ steel fibre, R ¼ crumb rubber, WRA ¼ naphthalene-based high-range water-reducing admixture.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203196

the furnace, the specimens were heated at a constant rate of 8 �C/min, from the room temperature to the prescribed temperatures.The theoretical heating procedure described by the temperatureetime curves shown in Fig. 2 were used in heating the test specimensin the present study (Peng et al., 2006). The target temperature wasmaintained for 2 h before electric heating was turned off and thenthe specimens were naturally cooled down to the room tempera-ture (see Fig. 2). During the heating period, water vapour wasallowed to escape freely.

Compression strength tests were carried out on the cylindricalspecimens with reference to ASTM C 39 (ASTM, 2003) using aMATEST compression machine with a 4000 kN capacity. An axialload was applied at a constant displacement rate of 0.18 mm/min.The axial shortening of the cylinders under compression wasmeasured using 2 linear variable displacement transducers (LVDTs)set between two height levels with a vertical distance of 120 mm attwo opposite locations in the mid-height region of the cylinder, asshown in Fig. 3(a). The hoop strains of the concrete cylinders weremeasured using strain gauges. For each concrete cylinder, 2 straingauges with a gauge length of 80 mmwere bonded at two oppositepoints on the mid-height of the cylinder, as shown in Fig. 3(b).Before testing, the upper and lower surfaces of each cylinderspecimen were levelled with gypsum with a compressive strengthof 800 MPa, so as to eliminate the eccentricity of loading.

3. Results and discussions

The effects of elevated temperature and crumb rubber contenton the compressive properties, including residual strength, Young’smodulus (stiffness), stressestrain curves and energy absorptioncapacity (ductility) were measured and analysed during the

Fig. 2. Temperatureetime curves in heating the test specimens.

compression strength tests. Meanwhile, visual inspection and massloss tests of each concrete mix after exposure to elevated temper-ature were also carried out for the compression strength tests.

3.1. Visual inspection of concrete specimens

3.1.1. Colour and appearance changesAn apparent change in the colour of the cylinder specimens can

be identified by visual inspections after exposure to elevatedtemperatures. At room temperature, the concrete was light grey,which was turned to dark red at 200 �C, yellowish gray at 400 �Cand gray white at 600 �C. The colour change of the specimens isassociated with the chemical and physical changes experienced bythe concretematerials after exposure to high temperatures (Li et al.,2011).

3.1.2. Crack and spalling behaviourFor concrete specimens without rubber content (NC-R0 and RC-

R0), an increasing number of micro-cracks were observed on thesurfaces of the specimens with the increasing of target tempera-ture, regardless whether the recycled concrete aggregates wereused or not. Existing research already showed that for normalconcrete (without rubber) subjected to elevated temperatures,stresses within the concretematerials (cement paste) caused by thewater vaporation, thermal expansion, drying shrinkage and inter-action of them account for the cracking of concrete (Son et al.,2011).

Fig. 3. Test setup and location of LVDTs and hoop strain gauges in a specimen.

Fig. 4. Concrete samples weight loss percentages vs. treatment temperature and theircontent of the rubber.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203 197

For crumb rubber reinforced concrete specimens (RC-R4, RC-R8,RC-R12 and RC-R16), the situations were different: at the targettemperature of 400 �C, only several micro-cracks were detected onthe surfaces of concrete cylinders with crumb rubber volume of 4%and 8% (RC-R4 and RC-R8), and nearly no cracks appeared on thesurfaces of specimens with larger crumb rubber volume (RC-R12and RC-R16); at the target temperature of 600 �C, micro-cracksappeared on the surfaces of all the specimens but the number ofmicro-cracks decreased with the increasing volume of crumbrubber. Obviously, crumb rubber helps to alleviate the initiationand development of cracks in concrete under the elevated tem-peratures. It is mainly due to the fact that rubber is melted underthe temperature of around 170 �C, providing space for the evapo-rated water in concrete to escape from the concrete, thus signifi-cantly reducing the pore pressure caused by the water vapour, oneof the main reasons leading to the cracking of concrete underhigher temperature (Netinger et al., 2011).

In addition, there was no evidence of explosive spalling for allthe concrete specimens during the process of heating even thoughexplosive spalling under high temperature has been widelyobserved in concrete especially for high strength concrete (Penget al., 2006; Li et al., 2011). The absence of spalling in presentstudy might be a result of steel fibres which increases the concreteresistance to the concrete spalling under elevated temperatures(Lau and Anson, 2006). The rate of heating may also affect theoccurrence of the concrete structures spalling, but the assessmentof this dependence for the concrete samples that have compositionlike the mixtures used in the present study, when subjected toelevated temperatures, needs further research.

3.2. Weight loss

The weight loss ratios of different concrete mixes after exposureto elevated temperatures are shown in Fig. 4. It should be noted thatthe test result corresponding to each point in Fig. 4 was obtainedfrom the average of the test results of three cylinders in a group.The same way was used in obtaining the test results in thefollowing context if not otherwise stated.

It is obvious from Fig. 4(a) that for all the concrete mixes, weightloss increases with the increase of target temperature. Fig. 4(a) alsoshows that the higher the target temperatures, the lower thegradient of the concrete weight loss. In particular, the averageweight loss ratios are 6.5%, 8.6% and 9.9% at the temperatures of200 �C, 400 �C and 600 �C, respectively. Thus it can be said thatmost of theweight lost occurs during the temperature range of 25e200 �C. This is mainly because evaporation of water, one of themaincauses leading to the weight loss of concrete specimens during theheating process, occurs between 25 and 200 �C. It should be notedthat concrete includes capillary water, physically absorbed water(Gel water) and chemically bound water in calcium silicate hydrate(CeSeH) and calcium hydroxide (Ca (OH)2) (Savva et al., 2005),among which capillary water and physically absorbed water takeup a large proportion of cement paste weight and can be driven outof concrete by evaporationwhen the ambient temperature is 200 �Cor above (Zhang, 2011). On the other hand, chemically bound wateris the part of cement hydrate compounds and often called non-evaporable water for it can not be released from cement pasteuntil the chemical decomposition of the CeSeH occurs at a highertemperature. It should be noted that the weight loss ratio of RC-R0mix is considerably larger than that of NC-R0 mix; this is partiallydue to the fact that concretewith RCA containsmorewater than thenormal concrete (Table 1) due to higher water absorption.Furthermore, crumb rubber melts at the temperature of around170 �C, which contributes to the weight lost of the concrete withrubber. The weight lost above 200 �C is mainly due to the

decomposition of CeSeH at about 400 �C and Ca (OH)2 at about600 �C (Savva et al., 2005; Janotka and Nürnbergerová, 2005).

Fig. 4(b) shows effect of rubber content on weight lost underdifferent elevated temperatures. From the figure it can be seen thatrubber content has less effect on the weight loss under highertemperatures, especially for the temperature above 200 �C. Apossible explanation is that at higher temperatures, the contribu-tion of the melting of rubber to the total weight loss is significantlyless than the contribution of water evaporation and decompositionof concrete materials.

3.3. Failure modes

The failure modes of different concrete mixes after thecompression tests were shown in Fig. 5. It should be noted that each

Fig. 5. Failure modes of concrete mixes exposed to elevated temperatures.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203198

of the pictures shown in Fig. 5 was obtained randomly from one ofthe three cylinders in a batch after the compression test. Fig. 5(a)clearly shows that a major macro crack crossed the height of con-crete cylinders without any crumb rubber, while on the surfaces ofthe specimens with crumb rubber, only multiple thinner crackswere observed. Such a phenomenon is attributed to the lowYoung’s modulus of crumb rubber which not only enhances thecapacity of deformation before cracking but also prevents furtherpropagation and coalescence of micro-cracks by decreasing thestress concentration like a damper (Turatsinze et al., 2006). Ingeneral, it can be concluded that the concrete containing crumbrubber have significantly higher ductility than the concretewithoutrubber at room temperature (Khaloo et al., 2008; Turatsinze and

Garros, 2008). The failure modes of the specimens after exposureto the elevated temperatures were apart shown in Fig. 5(b)e(d). Itcan be seen that except for specimen RC-R16, the increase of rubbercontent generally decreases thewidth of the critical crack and leadsto a more distributed crack pattern. This is mainly because theevaporation of moisture/water in concrete after exposure to hightemperatures can cause a severe damage to the bond betweencement paste and aggregates due to the so-called vapour pressuremechanism (Peng et al., 2006), as a result, micro-cracks initiatealong the cement pasteeaggregate interfaces which may furthercoalesce into macro cracks during the compression test; never-theless, for the rubberized concrete, rubber is melted and inner-connected pores are formed at the temperature of 170 �C or

Fig. 6. Stresseaxial strain curves of mix exposed to elevated temperatures.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203 199

above, providing space for the vapour to escape, consequently, thedamage from inner vapour pressure can be significantly decreased.As a result, the damage to the concrete structure prior to thecompression test should be very small, so the micro-cracks inrubberized concrete might be caused mainly by the compressiontesting. It is noted that wide/deep macro cracks also appear inspecimen RC-R16 especially after exposure to higher temperature(600 �C). This is probably because in this specimen, the damageeffects of the large space/void left by the melted rubber on thecylinder structure have become more pronounced than its benefi-cial effect of reducing water vapour pressure, as a result, thespecimen becomes more susceptible to cracking under compres-sion. Spalling phenomena were not observed in any of the speci-mens during the heating process probably due to the beneficialeffect of steel fibre as mentioned previously.

3.4. Stressestrain cures

Fig. 6 shows the complete stressestrain curves of the unheatedspecimens and the specimens after exposure to elevated temper-atures, which were obtained from the compression tests on cylin-ders. It should be noted that both axial strains and hoop strainswere measured during the tests; the hoop strain was measuredwith strain gauges with a gauge length of 80 mm and the axialstrain of concrete in compression was obtained from the mid-height region of the cylinders using LVDTs as mentioned above.The strain values obtained via the LVDTs are generally smaller thanthose obtained from the full height shortening of the cylindersbecause of the end effects (Poon et al., 2004a,b).

It can be seen from Fig. 6 that the addition of crumb rubberresulted in a significant change in the shape of the stressestraincurves. For both the unheated and the heated specimens, the peakstresses (i.e. compressive strengths) and the initial slop of thestressestrain the curves (i.e. initial stiffness) were generallydecreased with the increase of the exposure temperatures. Theyalso decreased with the increase of rubber content except afterexposure to the temperature of 600 �C. Furthermore, more flat-tened descending paths appeared in the stressestrain curves forconcrete mixes with larger rubber content. The rubber contentappears to have a very slight effect on the magnitude of strain atpeak stress, but the latter is apparently increased with the increaseof exposure temperature (see Table 3). The main reason is that, forthe concrete with RCA, the degradation of the stiffness (Young’smoduli) is much quicker than that of compressive strength (peakstress) with the increase of exposure temperature, as will be furtherdiscussed next.

3.5. Compressive strength

The residual compressive strengths of each concrete mix,including both unheated specimens and specimens after exposureto elevated temperatures, were shown in Table 2. The effects oftemperature on compressive strength of concretes without or withcrumb rubber are shown in Fig. 7. It should be noted that Fig. 7(a)and (b) shows the results of strength values and relativestrengths with reference to the compressive strengths of the cor-responding unheated concrete mixes. It can be seen from Table 2and Fig. 7 that, for the unheated concrete specimens (25 �C), afull replacement of NCA by RCA only results in a decrease of 9.04%in the compressive strength, while increasing the rubber contentfrom 4% to 16% by volume of sand leads to the decrease ofcompressive strength from 4.57% to 30.21% (with respect to thecompressive strength of RC-R0), with the largest strength lossoccurring between the rubber content of 4% and 8%. A further in-crease of rubber content beyond 8% leads to a very slight decrease

of the compressive strength for both the unheated and heatedspecimens (Table 2). An average of 84.30% of the compressivestrength of the unheated concrete was retained after exposure to200 �C, which was further reduced to 49.66% and 24.77% after

Table 2The results of compressive strength and Young’s modulus of concretes.

Mix Volume of crumbrubber

Slump constant(mm)

Compressive strength (MPa) Young’s modulus (GPa)

25 �C 200 �C 400 �C 600 �C 25 �C 200 �C 400 �C 600 �C

NC-R0 0% 132 56.52 45.66 24.71 16.28 34.91 18.38 6.19 1.35RC-R0 0% 125 51.41 43.55 28.64 10.98 26.58 13.82 5.42 1.03RC-R4 4% 125 49.06 40.44 26.21 12.79 25.20 13.66 4.51 1.29RC-R8 8% 122 39.41 34.54 19.70 10.21 21.25 12.82 3.98 1.11RC-R12 12% 123 37.61 32.15 17.61 8.30 20.88 11.60 3.60 0.93RC-R16 16% 122 35.88 31.18 17.17 8.28 19.15 11.21 3.32 0.91

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203200

exposure to 400 �C and 600 �C respectively. A comparison of thecurves in Fig. 7(a) or (b) shows that the inclusion of rubbergenerally reduces the rate of concrete strength loss and the trend ismore obvious for the elevated temperatures between 400 and600 �C. This is mainly because the rubber, after melted at 170 �C,leaves space for water vapour to escape and helps to release thepore pressure and thus reduces its damage on the concrete struc-ture (Li et al., 2011). As a result, after exposure to the temperature of600 �C, the rubberized concrete retained a residual strength similar

Fig. 7. Effect of temperature on compressive strength.

to (e.g. RC-R8) or even higher (e.g. RC-R4) than that of the concretewithout rubber (RC- R0) (Table 2). It should be noted that the in-crease of rubber content from 4% to 8% or above just results in aslight decrease in the rate of concrete strength loss after exposureto elevated temperature, but leads to a significant concrete strengthreduction for the unheated specimens as mentioned above,implying that to achieve a balanced compressive strength for bothunheated specimens and heated specimens, an appropriate amountof rubber content should be included. It should also be noted thatthe concrete mixes suffered the highest loss in compressivestrength in the temperature range of 200e400 �C. This might beattributed to that calcium silicate hydrate (CeSeH), the mainsource of concrete strength, usually decomposes at about 400 �C(Janotka and Nürnbergerová, 2005). NC-R0 mix suffered a quickerloss in compressive strength after exposure to elevated tempera-tures, especially in the temperature range of 200e400 �C. This canbe attributed to the relatively dense microstructures of concretewith natural concrete aggregate, which results in a quicker accu-mulation of high internal pressure during heating, as comparedwith concrete mixes with RCA (with and without crumb rubber)(Behnood and Ziari, 2008).

The strength degradation mechanism of the rubberized con-crete after exposure to elevated temperatures and how the inclu-sion of crumb rubber alleviates the strength degradation should befurther studied by examining the microstructure of concrete in thefuture, as in Li et al. (2011).

3.6. Young’s modulus (stiffness)

Table 2 lists the Young’s moduli of the concrete mixes afterexposure to different temperatures (including 25 �C), while Figs. 8and 9 separately show the effects of temperature and rubber con-tent on the Young’s modulus. It should be noted that each of theYoung’s moduli shown in Table 2 and in Figs. 8 and 9 was taken asthe secant modulus of the corresponding stressestrain curve at onethird of the peak stress following the method of Poon et al.(2004a,b). In Figs. 8(b) and 9(b), the moduli were given as therelative values with reference to the Young’s moduli of the corre-sponding unheated concrete mix and RC-R0 mix respectively. Theresults presented in Fig. 8 and Table 2 show that the degradation ofYoung’s moduli is much quicker than that of compressive strength.Only 55.55% of the Young’s moduli of the unheated concretes(25 �C) were retained, on average, after exposure to 200 �C, whichwas further decreased to 18.22% and 4.55% after exposure to 400 �Cand 600 �C respectively. The quick degradation of concrete Young’smoduli implies that elevated temperature has a significant damageon the stiffness of the concrete cylinder. This trend appears to beindependent of the rubber content of the concrete mixes. This canbe explained as follows. On the one hand, the void left by themelted rubber or the low stiffness of rubber (before melting) maycause additional damage to the concrete cylinders (Effect I) inaddition to the degradation of stiffness caused by the elevatedtemperatures; one the other hand, the void left by the melted

Fig. 8. Effect of temperatures on Young’s modulus.

Fig. 9. Effect of crumb rubber content on Young’s modulus.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203 201

rubber after exposure to elevated temperature can effectivelyrelease thewater vapour, help to reduce the damage/cracks existingin concrete cylinder before compression test and thus prevent theloss of the stiffness (Effect II). When Effect I is dominant, the stiff-ness of concrete decreases with the increases of rubber content,which is the situation for unheated concrete (25 �C) and the con-cretes after exposure to the temperatures of 200 �C and 400 �C(Fig. 9). It should be noted that after exposure to 600 �C, the elasticmodulus of concrete is firstly increased and then decreased withthe increase of rubber content, implying that under a specificexposure temperature (e.g. 600 �C) (Fig. 9), the above two effects(Effect I and Effect II) of rubber may interact with each other andthe weight of their influences on the stiffness depends on therubber content.

3.7. Energy absorption capacity (toughness)

While many existing studies have been carried out on thecompressive behaviours of concrete both unheated (Najim and Hall,2012; Güneyisi et al., 2004) and after exposure high temperatures(Peng et al., 2006; Behnood and Ziari, 2008), few studies have

documented the compressive toughness of rubberized concreteafter exposure to elevated temperatures. To fill the gap in theexisting studies, in this study, the energy absorption capacity(compressive toughness) of RSRAC mixes were measured andanalysed in terms of the so-called specific toughness (Poon et al.,2004a,b), defined as the ratio of the area under the stressestraincurve (i.e. toughness) of each concrete mix to its correspondingcompressive strength. Effects of exposure temperature and rubbercontent on the specific toughness are apart shown in Fig. 10(a) and(b), with the values of the specific toughness being listed in Table 3.It should be noted that each value of the specific toughness listed inTable 3 is the average value of three specimens in a group asexplained previously. It is well-recognised that the total area underthe stressestrain curve should be calculated to evaluate thetoughness; nevertheless, in the present study, the stressestraincurves were recorded up to a strain value of around 1.0% and 1.5%for unheated specimens and heated specimens respectivelybecause the stressestrain curves beyond the above values becameunstable andwere not collected due to the limitation of the test set-up. As a result, the toughness was evaluated based on a stress cri-terion as follows: the energy absorbed was calculated at the point

Fig. 10. Specific toughness of mix unheated and exposed to elevated temperatures.

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203202

where a stress reduction of 20% from the peak stress is achieved (i.e.at 0.80fc0 of the descending branch). Obviously, the closer to using100% reduction rate, the better; nevertheless, the above reductionrate was chosen to be applicable for most of the specimenscompared, as shown in Fig. 6. For the group of the specimensexposed to 600 �C, the descending branch of the stressestraincurves terminated before they decrease to 0.8fc0, which is a damagecritical point of concrete structure, due to the larger deformation ofspecimens subjected to a high temperature and the limited dataacquisition capacity of the testmachine. As a result, the group of thespecimens after exposure to 600 �C were not included in thetoughness evaluation.

Table 3Effects of temperatures on deformability and ductility of concrete mixes subjected to ele

Mix Volume of crumbrubber

Strain at peak stress (%)

25 �C 200 �C 400 �C 600 �C

NC-R0 0% 0.30 0.39 1.01 1.65RC-R0 0% 0.34 0.45 0.78 1.77RC-R4 4% 0.33 0.43 0.97 1.74RC-R8 8% 0.28 0.36 1.02 1.56RC-R12 12% 0.27 0.44 0.89 1.32RC-R16 16% 0.25 0.38 0.70 1.21

Concrete mixes after exposure to temperatures of 200 �C and400 �C retained, on average, 1.49 and 2.12 times of their specifictoughness separately, when compared with the specific toughnessof the unheated ones. It can be seen that increasing temperature(�400 �C) results in an apparent reduction of concrete strength, butgreatly improved its energy absorption capacity (see Fig. 10(a) andTable 3). It can be seen from Table 3 and Fig. 10(b) that a fullreplacement of NCA by RCA leads to a significant decline in thespecific toughness especially for the concrete subjected to a highertemperature, showing that mixes with RCA are more brittle. It isalso noted that while the rubber content is increased from 4% to16%, the specific toughness is first increased and then decreasedwith the increase of rubber content, with RC-R12 at 25 �C and200 �C and RC-R8 at 400 �C having the highest specific toughness,indicating that an appropriate amount of rubber content increasesthe energy absorption capacity of the concrete but too much rubbercontentmay has a negative effect on the energy absorption capacityof the concrete. As a result, it can be concluded that to effectivelyimprove the energy absorption capacity of concrete mixes (bothunheated and heated specimens), an appropriate amount of rubbershould be used.

From the results of the compressive strength and energy ab-sorption properties discussed above, it can be concluded that, ifconcrete strength is a major concern (e.g. in building structures),the optimal rubber content should be less than 4% (�4%), whichinduces a very limited strength decrease (Fig. 7 and Table 2) but asignificant energy absorption capacity enhancement (Fig. 10 andTable 3), in addition to other advantages obtained by using rubber(e.g. enhanced resistance to explosive spalling, environment pro-tection). However, if the energy absorption capacity of concrete is ofinterest (e.g. in road pavement), the advisable rubber contentshould be 8e12%, which leads to an apparent increase in the energyabsorption capacity despite some strength decrease after exposureto the elevated temperatures explored in this study.

4. Conclusions

In this paper, the effects of elevated temperatures on thecompressive properties of RSRAC concrete mixes, a new materialfirstly proposed by the authors, have been explored in detail basedon the results of axial compression tests on standard cylinders withreference to ASTM standards. The following conclusions can bedrawn from the test results, analyses and discussions presented inthis paper:

� Elevated temperature has a significant effect on the compressivebehaviour of RSRAC: after exposure to 200 �C, the concretemixes retained 84.30% of their compressive strengths, onaverage, which was further reduced to 49.66% and 24.77% afterexposure to 400 �C and 600 �C separately. While the loss of thestiffness of concrete mixes was much quicker than the loss ofcompressive strength after exposure to the elevated

vated temperatures.

Area under stressestrain curve(MPa * 10�2)

Specific toughness (%)

25 �C 200 �C 400 �C 25 �C 200 �C 400 �C

23.98 27.10 22.66 0.42 0.59 0.7921.32 23.20 19.18 0.41 0.53 0.6718.77 24.96 19.43 0.40 0.62 0.7417.02 23.13 22.21 0.43 0.67 1.1321.55 24.49 19.61 0.57 0.76 1.1112.42 17.98 15.68 0.35 0.58 0.91

Y.-c. Guo / Journal of Cleaner Production 72 (2014) 193e203 203

temperatures, and the loss of specific toughness was apparentlyslower than the loss of compressive strength. The strains at thepeak stresses of the concretes after exposure to 200 �C, 400 �Cand 600 �C were, respectively, on average, about 1.4, 3.1 and 5.3times that of the unheated concretes, mainly resulting from thesevere degradation of the stiffness.

� NC-R0 mix suffered a quicker loss in compressive strength afterexposure to elevated temperatures, especially after exposure tothe temperature of 400 �C, probably for their relatively densemicrostructures.

� For the unheated concrete mixes, a full replacement of NA byRCA only resulted in a decrease of 9.04% in the compressivestrength; when the rubber content was increased from 4% to16%, the strength reduction was increased from 4.57% to 30.21%and the decrease of stiffness from 5.19% to 27.95%, both withreference to RC-R0 mix, while specific toughness was firstlyincreased and then decreased.

� A certain amount of crumb rubber is effective in reducing thedegradation rate of compressive strength after exposure toelevated temperatures, but too much rubber results in a slightdecrease in the rate of concrete strength loss and a significantconcrete strength loss for the unheated specimens. After expo-sure to the same elevated temperature, the specific toughness ofconcrete mixes was first increase and then decreased with theincrease of rubber content, with RC-R12 at 25 �C and 200 �C andRC-R8 at 400 �C having the highest specific toughness valuescompared with the other mixes. Hence, to improve the energyabsorption properties of concrete after exposure to high tem-perature, an optimal amount of rubber content should beincluded.

� Base on the limited test result presented in this study, if strengthis a major concern, the optimal rubber content should be lessthan 4% (�4%), which leads to very slight strength decrease butbrings other advantages, e.g. enhanced toughness, enhancedresistance to explosive spalling and environmental protection; ifenergy absorption capacity is of interest, the advisable rubbercontent should be 8e12%, which contributes to an apparentincrease in the energy absorption capacity despite somestrength decrease of concrete after exposure to the elevatedtemperatures explored in this study.

� Effects of elevated temperatures on the damage mechanism ofthe RSRAC mixes and how the inclusion of the crumb rubbercontributes to alleviate the damage should be further studied byexamining the change in the microstructures of concrete in thefuture. Effects of steel fibre on the compressive behaviours ofRSRAC also need further study, which is the aim of an ongoingproject of the authors’ research group.

Acknowledgements

The authors gratefully acknowledge the financial support pro-vided by the National Natural Science Foundation (Project Nos.51278132, 11372076), and Science and Technology Planning Projectof Guangdong Province (2011B010400024), Technology PlanningProject of Huangpu District (201356) and Foundation of GuangdongProvincial Department of Transport (Project Nos. 2013-02-017,2013-04-006).

References

Al-Tayeb, Mustafa Maher, Bakar, B.H. Abu, Ismail, Hanafi, Akil, Hazizan Md, 2013.Effect of partial replacement of sand by recycled fine crumb rubber on theperformance of hybrid rubberized-normal concrete under impact load: exper-iment and simulation. J. Clean. Prod. 59 (12), 284e289.

ASTM, 2003. ASTM C 39/C 39M. Standard Test Method for Compressive Strength ofCylindrical Concrete Specimens. American Society for Testing Material.

ASTM, 2006. ASTM C 192/C 192M. Standard Practice for Making and Curing Con-crete Test Specimens in the Laboratory. American Society for Testing Material.

ASTM, 2009. ASTM E11-09e1. Standard Specification for Woven Wire Test SieveCloth and Test Sieves. Subcommittee, E29.01.

Atis, C.D., Karahan, O., 2009. Properties of steel fibre reinforced fly ash concrete.Constr. Build. Mater. 23 (1), 392e399.

Barbudo, A., de Brito, J., Evangelista, L., Bravo, M., Agrela, F., 2013. Influence of water-reducing mixtures on the mechanical performance of recycled concrete.J. Clean. Prod. 59, 93e98.

Behnood, A., Ziari, H., 2008. Effects of silica fumes addition and water to cementratio on the properties of high-strength concrete after exposure to high tem-peratures. Cem. Concr. Compos. 30 (2), 106e112.

BS, 1983. BS 1881-102. Testing Concrete, Method for Determination of Slump.British Standards Institution, London, UK.

Gao, D.Y., Lou, Z.H., Wang, Z.Q., 2007. Experimental study on the compressivestrength of steel fibre reinforced recycled aggregate concrete. J. ZhengzhouUniv. (Eng. Sci.) 28, 5e10 (in Chinese).

Güneyisi, E., Geso�glu, M., Özturan, T., 2004. Properties of rubberized concretescontaining silica fume. Cem. Concr. Res. 34 (12), 2309e2317.

Guo, Y.C., Zhang, J.H., Chen, G., Chen, M.G., Xie, Z.H., 2014. Fracture behaviors of anew steel fiber reinforced recycled aggregate concrete with crumb rubber.Constr. Build. Mater. 53, 32e39.

Hernández-Olivares, F., Barluenga, G., 2004. Fire performance of recycled rubber-filled high-strength concrete. Cem. Concr. Res. 34 (1), 109e117.

Hernández-Olivares, F., Barluenga, G., Bollati, M., 2002. Static and dynamic behav-iour of recycled tire rubber-filled concrete. Cem. Concr. Res. 32 (10), 1587e1596.

Janotka, I., Nürnbergerová, T., 2005. Effect of temperature on structural quality ofthe cement paste and high-strength concrete with silica fumes. Nucl. Eng. Des.235 (17e19), 2019e2032.

Khaloo, A.R., Dehestani, M., Rahmatabadi, P., 2008. Mechanical properties of concretecontainingahighvolumeof tireerubberparticles.WasteManag.28(12),2472e2482.

Kou, S.C., Poon, C.S., 2009. Properties of self-compacting concrete preparedwith coarseand fine recycled concrete aggregates. Cem. Concr. Compos. 31 (9), 622e627.

Kou, S.C., Poon, C.S., 2012. Enhancing the durability properties of concrete preparedwith coarse recycled aggregate. Constr. Build. Mater. 35 (0), 69e76.

Lau, A., Anson, M., 2006. Effect of high temperatures on high performance steelfibre reinforced concrete. Cem. Concr. Res. 36 (9), 1698e1707.

Li, L.J., Chen, Z.Z., Xie, W.F., Liu, F., 2009. Experimental study of recycled rubber filledhigh strength concrete. Mag. Concr. Res. 61 (7), 549e556.

Li, L.J., Xie, W.F., Liu, F., Guo, Y.C., Deng, J., 2011. Fire performance of high-strength con-crete reinforced with recycled rubber particles. Mag. Concr. Res. 63 (3), 187e195.

Mefteh, H., Kebaïli, O., Oucief, H., Berredjem, L., Arabi, N., 2013. Influence of mois-ture conditioning of recycled aggregates on the properties of fresh and hard-ened concrete. J. Clean. Prod. 54, 282e288.

Bravo, Miguel, de Brito, Jorge, 2012. Concrete made with used tyre aggregate:durability-related performance. J. Clean. Prod. 25, 42e50.

Najim, K.B., Hall, M.R., 2012. Mechanical and dynamic properties of self-compactingcrumb rubber modified concrete. Constr. Build. Mater. 27 (1), 521e530.

Netinger, I., Kesegic, I., Guljas, I., 2011. The effect of high temperatures on the me-chanical properties of concrete made with different types of aggregates. FireSaf. J. 46 (7), 425e430.

Olivito, R.S., Zuccarello, F.A., 2010. An experimental study on the tensile strength ofsteel fibre reinforced concrete. Compos. Part B: Eng. 41 (3), 246e255.

Peng, G.F., Yang, W.W., Zhao, J., Liu, Y.F., Bian, S.H., Zhao, L.H., 2006. Explosive spallingand residual mechanical properties of fibre-toughened high-performance con-crete subjected to high temperatures. Cem. Concr. Res. 36 (4), 723e727.

Poon, C.S., Kou, S.C., Lam, L., 2002. Use of recycled aggregates in molded concretebricks and blocks. Constr. Build. Mater. 16 (5), 281e289.

Poon, C.S., Shui, Z.H., Lam, Z.H., Kou, S.C., 2004a. Influence of moisture states ofnatural and recycled aggregates on the slump and compressive strength ofconcrete. Cem. Concr. Res. 34 (1), 31e36.

Poon, C.S., Shui, Z.H., Lam, L., 2004b. Compressive behavior of fibre reinforced high-performance concrete subjected to elevated temperatures. Cem. Concr. Res. 34(12), 2215e2222.

Savva, A., Manita, P., Sideris, K.K., 2005. Influence of elevated temperatures on themechanical properties of blended cement concretes prepared with limestoneand siliceous aggregates. Cem. Concr. Compos. 27 (2), 239e248.

Son, K.S., Hajirasouliha, I., Pilakoutas, K., 2011. Strength and deformability of waste tirerubber-filled reinforced concrete columns. Constr. Build. Mater. 25 (1), 218e226.

Turatsinze, A., Garros, M., 2008. On the Young’s modulus and strain capacity of self-compacting concrete incorporating rubber aggregates. Resour. Conserv. Recycl.52 (10), 1209e1215.

Turatsinze, A., Granju, J.L., Bonnet, S., 2006. Positive synergy between steel-fibresand rubber aggregates: effect on the resistance of cement-based mortars toshrinkage cracking. Cem. Concr. Res. 36 (9), 1692e1697.

Yang, R.N., Yin, J.R., Xiao, H.M., Liu, Q.F., 2006. Experimental study of steel fibrereinforced regenerated concrete’s mechanical properties. Concrete 01, 27e30(in Chinese).

Zhang, Binsheng, 2011. Effects of moisture evaporation (weight loss) on fractureproperties of high performance concrete subjected to high temperatures. FireSaf. J. 46 (8), 543e549.