Shrinkage performance of Crumb Rubber Concrete (CRC) prepared by water-soaking treatment method for rigid pavements

Cement & Concrete Composites 62 (2015) 106–116

Contents lists available at ScienceDirect

Cement & Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp

Shrinkage performance of Crumb Rubber Concrete (CRC) prepared by water-soaking treatment method for rigid pavements Iman Mohammadi ⇑, Hadi Khabbaz Centre for Built Infrastructure Research (CBIR), School of Civil and Environmental Engineering, University of Technology, Sydney, Australia

a r t i c l e

i n f o

Article history: Received 7 May 2014 Received in revised form 25 October 2014 Accepted 1 February 2015 Available online 23 June 2015 Keywords: Crumb Rubber Concrete (CRC) Rubber treatment methods Water-soaking Plastic and drying shrinkage Bleeding of rubberised concrete Shrinkage of rubberised concrete Rigid pavements

a b s t r a c t This investigation deals with the shrinkage properties of rubberised concrete pavement. Arrays of concrete samples were prepared with different water–cement ratios and rubber content. The experimental results revealed that the introduction of rubber into concrete mixes results in the control of shrinkage cracks if the optimised content of rubber is selected. Accordingly, the optimised rubber content was determined based on the mix characteristics, mechanical properties and the results of plastic and drying shrinkage tests. The mechanical strength, toughness, bleeding, plastic shrinkage and drying shrinkage tests were conducted in this experimental program. Analysing the results revealed that the most promising performance results were achieved for samples prepared with the rubber contents of 20% and 25% of fine aggregates, and water–cement ratios of 0.45 and 0.40, respectively. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction and background In many countries around the world, the adverse environmental impacts of stockpiling waste tyres have led to investigate alternative options for disposal of waste tyres. The disposal of waste tyres has been found to be an environmental concern due to waste tyres resisting degradation [1]. Waste tyres occupy large landfill spaces that contain nesting insects and rats. Stockpiles of tyres destined for landfill are also known to be flammable. One option to reduce this environmental concern is for the construction industry to consume a high amount of recycled tyres accumulated in stockpiles. In the pavement industry, trialling the use of crumb rubber has been initiated with asphalt mixes. However, some difficulties, such as the high viscosity of the rubberised bitumen and the requirement of higher temperature for production of rubberised asphalt have been found to be barriers which limit its application [2]. Consequently, the introduction of rubber to concrete pavements was the basis of this investigation. It is a well-established fact that introduction of rubber to concrete can cause a decline in concrete strength [3], however rubberised concrete has been found to be suitable for rigid pavement applications, where a lower range of strengths are required in design [4,5].

A variety of rubber treatment methods have been applied to counteract the negative impact of rubber on concrete strength. One of the treatment methods, recently introduced to enhance the mechanical properties of crumb rubber concrete (CRC), is the water-soaking method [6]. Although the strength of rubberised concrete by applying the water-soaking method has been notably improved [6], detailed investigations regarding other properties of the produced rubberised concrete is still required. For the past several decades, the design of concrete has typically been based on the concrete strength since it can easily be measured for quality control and quality assurance purposes. Although the strength requirements are frequently met, it is clear that the durability performance may not always be satisfied. The main aim of this research is to enhance the know-how of shrinkage properties of rubberised concrete pavements prepared by application of the water-soaking method. This will be conducted through systematic laboratory tests, which can illustrate the possible advantages and limitations of adding different volumes of crumb rubber into concrete pavements.

2. Literature review 2.1. Shrinkage of rubberised concrete

⇑ Corresponding author at: Cement Concrete & Aggregates Australia (CCAA), PO Box 124, Mascot, NSW 1460, Australia. E-mail address: [email protected] (I. Mohammadi). http://dx.doi.org/10.1016/j.cemconcomp.2015.02.010 0958-9465/Ó 2015 Elsevier Ltd. All rights reserved.

Literature indicates when the concrete elements are well designed, cast and cured there should not be any noticeable

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

cracking, and hence most shrinkage cracks can be eliminated. However, the surfaces of concrete pavements are highly exposed to the ambient conditions, which need more attention compared to other types of concrete. The previous investigations have revealed that shrinkage cracks on large exposed surfaces of concrete pavements may happen upon a combination of various factors, such as concrete early-age tensile strength, deformation capacity, and the ability of concrete to resist against fracture. The aims of the previous studies were to increase the ductility capacity of concrete to improve the concrete cracking resistance. Rubber particles have low elastic modulus in the range of 0.01–0.1 GPa, and are highly flexible, which can undergo large deformation. Introducing rubber into concrete has two opposite effects. The negative impact is the reduction of tensile strength, whereas the positive effect is associated with the enhancement of concrete ductility. The degree of positive and negative effects depends on the properties and content of the introduced rubber. Different approaches can be found in the literature, regarding the reason for shrinkage cracking occurring. These approaches can be classified in three main categories, which are based on the concrete tensile strength, the concrete ultimate tensile strain capacity and, the concrete fracture capacity. The first approach discusses the low tensile strength of concrete, especially during the concrete, plastic state. Shrinkage cracks develop on the surface of concrete, after its surface becomes set and hardened. At the same time, concrete is not strong enough yet to resist against the tensile stress resulting from the restrained shrinkage. In other words, cracking happens, when the tensile stress rtsh induced by the restrained shrinkage strain esh exceeds the tensile strength ft of concrete, especially at early-ages even the first hours after the casting [7–10].

rtsh  A ¼ ðE  esh Þ  A P f t

ð1Þ

where rtsh is the tensile stress resulted from shrinkage in MPa, esh is the shrinkage strain in microstrain, E is the concrete elastic modulus in GPa, ft is concrete tensile strength in kN and A is the cross sectional area of the element subjected to the tensile stress in mm2. As can be seen from Eq. (1), shrinkage cracking is not only a function of concrete constituent properties, but also highly dependent on the concrete element size and geometry [11]. Two concrete elements with the same constituents and ambient conditions, but with different geometries, may substantially undergo different shrinkage cracking. The second approach includes the shrinkage cracking definition, based on the ultimate strain or the elongation capacity, which is considered as an important parameter, indicating the crack-resistance behaviour of concrete [12]. In this approach, shrinkage cracks may happen at any time, when the generated shrinkage tensile strain exceeds the ultimate tensile strain capacity of concrete structures [13]. Literature reported that concrete samples with the same tensile strength of 5 MPa have different strain capacities [14]. Therefore, two concrete samples with a same tensile strength can perform substantially different regarding shrinkage, due to different ultimate tensile strain capacities. Early age plastic shrinkage is a concern because, during the early hours after casting, strain capacity of concrete pavement has its minimum value [15,16]. Investigation performed by Byfors [16] revealed that concrete has the lowest tensile strain capacity in the early hours after it is set but has not gained enough tensile strain capacity. Finally, the third approach is the fracture based approach, which attempts to provide an explanation for shrinkage cracking characteristics by investigating fracture-based failure of concrete [11,17,18]. This approach might be better aligned for studying of shrinkage cracking especially in brittle materials, because the combination of sensitivity to crack propagation and high tensile stress at the stage of failure is considered in this approach. On the basis of

107

concrete fracture property, it can be indicated that when the shrinkage energy exceeds the fracture energy of concrete, cracking occurs on the surface of concrete [17,18]. There is no standard test available for measuring toughness of rubberised concrete. If the toughness test is conducted based on energy measuring methods, an optimum content of rubber may be determined by maximising the area under the load–deflection curve. It was reported that an increase in toughness was achieved by adding rubber to concrete [19]. In another study the load–displacement curve of tensile tests were plotted for both plain concrete and concrete containing 20% of rubber by volume of fine aggregates. Experimental results revealed that the normal concrete had a faster failure compared to the rubberised concrete [20]. However, a slow failure does not mean a high toughness of concrete. The reviewed test results available from the literature imply that compared to plain concrete, rubberised concrete had higher energy absorption capability during failure [19,21], also had limited residual capacity after the failure [22]. The low-stiffness rubber particles play three roles to limit the shrinkage cracks by reducing the internal restraint, lowering of the elastic modulus and trapping cracks from propagation in concrete. Concrete aggregates are solid particles resisting against the shrinkage. The restraint offered by aggregates results in shrinkage cracking of concrete [23]. Replacing a portion of aggregates with a highly elastic rubber reduces the internal restraint of concrete and lowers the concrete elastic modulus. It can be seen from Eq. (1) that lower elastic modulus may lead to lower thermal and shrinkage stress. In addition, Crumb rubber particles in the concrete matrix, trap cracks and do not let them pass through. Hence, it can be considered as a momentum reducing effect of rubber particles on crack propagation [12], which may provide resistance against shrinkage cracking. The shrinkage crack property of rubberised concrete is investigated in this research. In a study conducted by Raghavan and Huynh [24], rubberised mortar properties containing different portions of rubber particles in average length of 5–10 mm were assessed by evaluating the resistance of rubberised mortar against the plastic shrinkage cracking [24]. It was observed that all specimens were cracked within the first 3 h of placement. Moreover, it was indicated that modifying mortar with rubber had the key effect on allowing multiple cracks to occur over the width of the specimen, when it was compared with an intense single crack on the top surface of mortar prepared without rubber. Moreover, the total crack area in the case of rubber-filled mortar, found to be decreased with an increase in the rubber content [24]. It is noted that the crack resistance of the mortar contained crumb rubber with the size of 1.5 mm was improved substantially by adding crumb rubber particles [17]. Kang and Jiang [17] reported that adding rubber led to a reduction in both the tensile strength and the generated shrinkage stress, but the degree of reduction was found to be different for various content of rubber. It was found that when rubber fraction was less than 20% in volume, the cracking time was retarded and shrinkage properties were improved [17]. Another study investigated the development of the free drying shrinkage of rubberised concrete over 150 days [25]. That study indicated that by increasing the rubber content the drying shrinkage increased [25]. However, the higher free drying shrinkage will not result in higher shrinkage cracking if the strain capacity of concrete is enhanced. The free drying shrinkage test cannot provide clear understanding of shrinkage cracking behaviour of rubberised concrete without the study of restrained plastic shrinkage and fracture behaviour of concrete. In a recently conducted research by Bravo and de Brito [26] drying shrinkage of different rubberised concrete mixes with different water–cement ratios between 0.43 and 0.48, were investigated.

108

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

Different sizes of recycled rubber, prepared by different processes were examined. The results of that study showed that there is a significant positive correlation between rubber content and drying shrinkage. In addition, it was observed that increasing the content of fine size crumb rubber resulted in higher drying shrinkage [26]. Similar results were reported by Sukontasukkul and Tiamlom [27] for drying shrinkage of rubberised concrete samples and by Uygunog˘lu and Topçu [28] regarding rubberised mortar drying shrinkage. 2.2. Measuring plastic and drying shrinkage Autogenous shrinkage, carbonation shrinkage, plastic and drying shrinkage are the most important types of concrete shrinkage [29,30]. In Australia, water cement ratio and cement content of pavement mixes are usually set higher than 0.40 and less than 400 kg/m3, respectively. Thus, the plastic and drying shrinkage are selected as the main types of shrinkage that concrete pavements undergo. The shrinkage strain is significantly affected by the restraint applied to a concrete member and its geometry [11]. Environmental conditions, such as temperature, relative humidity, and the ambient wind velocity can also affect the magnitude of shrinkage [10,13]. In order to have a clear understanding of shrinkage performance of rubberised pavement concrete, measuring the plastic and drying shrinkage should be conducted in a controlled condition, which is addressed in shrinkage test specifications. Shrinkage takes place over a long period, but the effect of early age shrinkage is highly important in the total magnitude of shrinkage. Recent research in Australia, sponsored by Cement Concrete & Aggregates Australia (CCAA), showed that early-age shrinkage can develop strains in concrete with similar magnitudes to those resulting from the drying shrinkage [31]. Two different approaches can be applied for evaluating concrete plastic shrinkage. The first one is conducted by measuring the free plastic shrinkage strain directly. Measuring the magnitude of the plastic shrinkage strain is not a simple task, compared to the drying shrinkage. This problem arises from the fact that concrete in the plastic stage is semifluid. Therefore, pins such as ones used for measuring drying shrinkage cannot be set on samples. Moreover, the magnitude of plastic shrinkage varies throughout depth of samples. Therefore, it is more practical to measure restrained plastic cracking instead of plastic shrinkage. The second approach, which is adopted in this research, measures the cracking of the restrained elements. In order to have a realistic simulation of restrained plastic shrinkage cracking, different types of tests have been conducted

by researchers. These involve performing tests on restrained slabs with stress raisers at the middle [32,33], restrained ring tests [17,34], modified beam tests [35], and slabs placed on subbase layer [36]. The average width of cracks is calculated by applying a manual method of measuring the cracks on the surface of samples. However, accurate measurement of plastic cracks is a complex task and requires a high level of interpretation if it is performed manually. Applying the ‘‘Image Analysis’’ of the cracked areas is reported by literature to be a more suitable and accurate option rather than measuring crack properties manually [37–39]. In this research, an advanced program was developed for plastic shrinkage crack analysis. In this study both plastic and drying shrinkage tests were conducted. The shrinkage model introduced in the Australian concrete standard, AS3600, does not suggest a method for measuring the plastic shrinkage of concrete. Therefore, it was decided to apply the Standard ASTM C 1579 to assess plastic shrinkage. In Australia, the drying shrinkage is measured specifically based on the Australian standards, either by conducting AS1012.13 shrinkage test or using the AS3600 numerical model [31]. Any amount of concrete shrinkage should be of concern since higher shrinkage means there are greater risks of cracking and deterioration [13]. The concrete pavement specification prepared for the state of New South Wales [40] controls shrinkage of rigid pavements by limiting the measured 21-day free drying shrinkage strain. Moreover, the experimental results of drying shrinkage for rubberised concrete were compared to the results of the numerical model provided by the Australian standard AS3600. 3. Experimental program 3.1. Mixing constituents Characteristics of shrinkage limited (SL) cement utilised in this study complied with the requirements of the Australian Standard AS3972. The selected Type SL cement is designed and commonly used by the construction industry for pavement applications. The fine and coarse aggregates used to accomplish this research were sourced from Dunmore, Australia. 10 mm and 20 mm crushed Latite gravels were employed as coarse aggregates. The available resource of sand was 50/50 blended sand. All types of aggregates complied with the grading requirements of the Australian Standard AS2758.1, Austroads Specification [41] and NSW RTA Specification [40] as shown in Table 1. Applied crumb rubber had a nominal grading size less than 4.75 mm and was prepared by grinding tire in an ambient grinding procedure.

Table 1 Particle size distribution for coarse and fine aggregates. Sieve size (mm)

a b

Crumb

Fine aggregates

Coarse aggregates

Rubber

0.075–4.75 mm

10 mm (nominal size)

Combined aggregates 20 mm (nominal size)

Passing (%)

Limitsa (%)

Passing (%)

Limitsa (%)

Passing (%)

Limitsa (%)

Passing (%)

Limitsb (%)

Passing (%)

26.50 19.00 13.20 09.50 04.75 02.36 01.18 0.600 0.300 0.150 0.075

– – – – 100 60 35 5 0 – –

– – – 100 90–100 60–100 30–100 15–80 5–40 0–25 0–20

– – – 100 98 81 65 55 36 8 4

– – 100 85–100 0–20 0–5 0–2 – – – –

– – 100 87 11 3 2 – – – –

100 85–100 – 0–20 0–5 – 0–2 – – – –

100 95 51 14 4 3 2 – – – –

100 95–100 75–90 55–75 38–48 30–42 22–34 16–27 5–12 0–3 0–2

100 98 80 62 40 32 25 20 12 3 1

Absorption (%) Density (kg/m3)

0.89 1150

1.2 2650

Australian Standard AS 2758.1. Austrods Standard [41] and RTA specification [40].

1.8 2700

1.6 2710

109

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

The mix design of concrete samples selected based on the typical mix design proportioning addressed in Section 3.1 of the Austroads Standard [41]. The target slump for all mix series was selected 60 ± 10 mm, in accordance with the Austroads Standard [41] and NSW RTA Specification [40]. The selected slump range is lower than the typical slump values selected for structure applications and is more preferable for concrete pavement applications. The selected target slump was achieved by the use of WR Sika Plastiment10. The type of water used for this research was potable water conditioned to 23 ± 2 °C used for all mix series. 3.2. Testing methodology and test standards List of the conducted tests with the required criteria is presented in Table 2. This was performed to assess any improvement or drawback in shrinkage performance due to the effect of adding

crumb rubber into concrete. The fresh, hardened and the early-age concrete mechanical strengths were assessed, which involved the possible negative impact of rubber on the concrete strength. The prepared rubberised concrete should comply with the requirement of the Australian specifications for rigid pavement construction. In order to quantify the brittleness of concrete pavements, two series of tests were performed, based on Standard ASTM C1609 and the modified ASTM C1609 tests. The modified ASTM C1609 method was introduced and applied for fibre reinforced concrete in a recently conducted investigation [42]. The procedure of this test is exactly the same as ASTM C1609 test method except for the loading rate, which is set 10 times lower than the standard ASTM C1609 test. The plastic shrinkage tests were conducted in accordance with the Standard ASTM C1579. Concrete samples were casted into the prepared moulds, then screeded and finished with a trowel

Table 2 Selected tests carried out to optimised rubber content in concrete pavement.

a

Test name

Concrete status

Standard No.

Test intervals

Requirement for pavement

Specifications

Slump Air content Mass per volume Bleeding Compressive strength Flexural strength Plastic Shrinkage Drying Shrinkage Flexural Fracture Flexural Fracture

Fresh Fresh Fresh Fresh Hardened Hardened Fresh Hardened Hardened Hardened

AS1012.3.1 AS1012.4.2 AS1012.5 AS1012.6 AS1012.9 AS1012.11 ASTM 1579 AS1012.13 ASTM 1609 ASTM 1609⁄a

Batching day Batching day Batching day Batching day 1, 3, 7 and 28 days 3, 7 and 28 days Batching day Up to 56 days 28 days 28 days

60 ± 10 mm <6.0 % 2100–2800 kg/m3 <3.0 % 32 MPa at 28 days 16 MPa at 7 days 4.5 MPa at 28 days Checked for any improvement <450 microstrains at 21 days Checked for any improvement Checked for any improvement

Austroads [41] Austroads [41] AS1379 RTA R83[40] Austroads [41] AS1379 Austroads [41] – RTA R83[40] – –

ASTM 1609* is conducted with the same procedure as ASTM 1609 but with a loading rate 10 times lower.

(a) 355×variable mm

(b) 20×20 mm

(c) 20×20 mm

(d) 2×2 mm

(e)

Fig. 1. Process of quantifying plastic shrinkage cracks on the surface of samples (a) cracked surface of concrete slabs, (b) 20  20 mm taken from the cracks, (c) boundary of cracks were defined in the AutoCAD software, (d) grids generated inside each cracks, (e) the interface of the programmed VBA plug-in running on an AutoCAD software platform.

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

in accordance with the Standard ASTM C1579. A special chamber was prepared for curing of samples for a period of 24 h after casting. Samples were put into the chamber after finishing within a time interval less than 30 min after start of concrete mixing. The prepared chamber provided the ambient temperature of 36 °C ± 3 °C, relative humidity of 30 ± 10%, wind velocity of 5 ± 1 m/s, and evaporation rate of over 1 kg/m2/h for test samples. Two specimens with a volume of approximately 18 l prepared from each batch of concrete, and then samples were put into the conditioning chamber. Thereafter, specimens were taken out of the chamber after 24 h. The procedure of crack analysis is shown in Fig. 1(a) and (b). A series of images in the size of 20  20 mm were taken from the cracks on the surface of samples. These images were taken from the same height of 100 mm and by the accuracy of eight megapixels. Subsequently, pictures were imported into AutoCAD software, then they were scaled and analysed based on the procedure presented in Fig. 1(c) and (d). This task conducted by setting the boundary of cracks in software, and then a set of grids with spacing of 100 lm created over each crack (Fig. 1(d)). The created grids were set in a perpendicular direction to the paths that cracks were propagated. Unlike the horizontal grid, applied by Qi [29] for carrying out the image processing of shrinkage cracks, the procedure applied in this study provided grids, which were perpendicular to the cracks path. Applying this method significantly improved the accuracy of results especially for measuring the crack widths. The applied method provided essential data regarding crack analysis, which involved crack widths, lengths and area for each concrete sample. In order to provide a systematic method for quantifying the shrinkage cracking of concrete, a piece of sophisticated code, written in VBA programming language. The prepared VBA code was run in AutoCAD software as shown in Fig. 1(e). Applying the advanced image analysis technique, a reliable and consistent assessment was conducted for measuring the plastic shrinkage cracks.

Volume (%)

00 10 20 25 30 40 0 10 20 30 40 746 746 746 746 746 746 726 726 726 726 726 [WC ratio]/[rubber content] CR, 0.40/10CR refers to a mix with WC of ‘‘0.40’’ and 10% crumb rubber. Volumes in respect to the fine aggregate volume. a

b

10 mm (kg/m )

416 416 416 416 416 416 404 404 404 404 404 677 609 541 508 474 406 658 593 527 461 395 100 90 80 75 70 60 100 90 80 70 60

Volume (%)

370 370 370 370 370 370 370 370 370 370 370 0.40 0.40 0.40 0.40 0.40 0.40 0.45 0.45 0.45 0.45 0.45 0.40/00CR 0.40/10CRa 0.40/20CR 0.40/25CR 0.40/30CR 0.40/40CR 0.45/00CR 0.45/10CR 0.45/20CR 0.45/30CR 0.45/40CR

Cement (kg/m3) WC Mix ID

Table 3 Mix series designations for unit volume of mix.

Fine aggregates

3

Weight (kg/m )

3

Coarse aggregates

3

148 148 148 148 148 148 166 166 166 166 166

Water (kg/m3)

Rubber

b

20 mm (kg/m )

00 30 59 74 89 118 00 29 58 86 115

3

Weight (kg/m )

3.196 3.990 3.945 4.052 4.351 4.443 1.151 1.225 1.436 1.554 2.072

WR (L/m3)

2356 2318 2280 2261 2242 2204 2325 2288 2251 2214 2177

Total (kg/m3)

110

3.3. Mixing procedure and identification of mix series This research applied the innovative water-soaking method introduced by Mohammadi et al. [6] in a recently conducted investigation. Mohammadi et al. [6] introduced different contents of rubber (up to 70%) to a series of concrete mixes, and found that the maximum applicable content should be limited to 40% to provide a homogenous distribution of rubber particles in the mix and avoid accumulation of rubber on the top layer of samples. Moreover, the water cement ratios should be selected in the range of 0.40–0.45. Considering the outcomes of that investigation, eleven sets of concrete were prepared as demonstrated in Table 3. Previous studies have shown that the cement paste in the plastic stage undergoes a volumetric contraction. The magnitude of the contraction was in the order of 1% of the absolute volume of dry cement content of the mix [30]. Thus, the cement content for all mix series was kept constant (370 kg/m3) to avoid any inconsistency in evaluation of shrinkage performance. Moreover, WC ratios for each array of concrete samples were kept constant, in order to keep the volumes of cement paste equal in each array of mix samples. The mixing procedure was the same as procedure specified in the Standard AS 1012.2. 4. Results and discussion 4.1. Fresh and hardened mechanical properties As can be inferred from Table 4, the fresh properties of concrete were affected by introducing rubber. The pattern, which concrete

111

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116 Table 4 The results of fresh and hardened tests. Fresh propertiesa

Mix properties

a b c

Hardened properties 3

Mix ID

WC

CR (%)

Slump (mm)

AC (%)

MPV (kg/m )

0.40/00CR 0.40/10CR 0.40/20CR 0.40/25CR 0.40/30CR 0.40/40CR 0.45/00CR 0.45/10CR 0.45/20CR 0.45/30CR 0.45/40CR

0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5 0.5 0.5 0.5

0 10 20 25 30 40 0 10 20 30 40

50 60 60 60 65 65 55 65 55 55 65

1.90% 2.40% 3.20% 3.70% 4.50% 6.10% 1.50% 2.10% 3.00% 4.30% 5.60%

2442 2387 2319 2290 2266 2213 2426 2370 2314 2253 2142

Compressive strength (MPa)b

Flexural strength (MPa)c

fcm,1

fcm,3

fcm,7

fcm,28

fctm,3

fctm,7

fctm,28

16.0 ± 0.4 14.3 ± 0.3 12.4 ± 0.3 11.1 ± 0.1 9.5 ± 0.2 7.4 ± 0.2 17.4 ± 0.9 14.7 ± 0.6 11.1 ± 0.5 9.9 ± 0.4 7.9 ± 0.4

33.1 ± 1.6 30.2 ± 1.4 25.8 ± 1.3 23.3 ± 1.1 20.1 ± 1.0 16.7 ± 0.8 29.2 ± 1.4 25.7 ± 1.3 21.0 ± 1.1 17.5 ± 0.7 12.6 ± 0.3

50.5 ± 3.6 44.8 ± 2.6 32.8 ± 2.2 29.2 ± 1.7 27.8 ± 1.5 20.8 ± 1.5 45.3 ± 3.2 35.4 ± 2.5 27.6 ± 1.5 21.8 ± 1.1 16.5 ± 0.5

63.0 ± 3.3 54.1 ± 0.8 44.3 ± 1.2 35.9 ± 1.4 30.9 ± 1.3 22.9 ± 0.6 55.6 ± 2.0 45.8 ± 2.0 34.9 ± 1.3 23.8 ± 1.0 18.2 ± 0.4

5.0 ± 0.1 4.6 ± 0.9 4.1 ± 0.1 3.9 ± 0.1 3.8 ± 0.1 3.5 ± 0.1 4.0 ± 0.2 3.9 ± 0.2 3.7 ± 0.0 3.3 ± 0.1 3.0 ± 0.1

6.1 ± 0.1 5.4 ± 0.3 4.8 ± 0.2 4.3 ± 0.2 4.3 ± 0.2 3.9 ± 0.1 5.4 ± 0.2 4.7 ± 0.3 4.2 ± 0.1 3.6 ± 0.2 3.0 ± 0.1

6.9 ± 0.2 6.2 ± 0.3 5.6 ± 0.2 5.4 ± 0.2 5.2 ± 0.1 4.6 ± 0.1 6.0 ± 0.2 5.4 ± 0.1 5.0 ± 0.2 4.1 ± 0.2 3.6 ± 0.2

Where AC is air content and MPV is mass per volume. Where fcm,t is the mean of compressive strength ± standard deviation of the results at the time of t days. Where fctm,t is the mean of flexural strength ± standard deviation of the results at the time of t days.

properties were changed, was found to be the same for both arrays of concrete samples prepared with different WC ratios. According to the requirement of the Australian standard for pavement concrete the slump number should be placed in the range of 60 ± 10 mm. In order to keep the slump (workability) of mixes at the same level, the volume of the added WR was adjusted for different rubber contents. Moreover, it was found that the air content was subjected to inclining, when the rubber content of mix series increased. In addition, mass per unit volume was declined by replacing aggregates with lighter rubber particles. The results of hardened property tests are presented in Table 4. For all testing ages, including the early-ages, it was found that the introduction of rubber adversely affected the concrete strength. However if rubber content is limited to a maximum of 20% or 25%, strength properties of prepared rubberised concrete can comply with the minimum requirements of the Australian Pavement specification shown in Table 2. In addition, other fresh and hardened properties of concrete prepared with 20% and 25% complied with the requirements of Australian Standards demonstrated in Table 2. Overall, results of fresh and hardened property tests were supported by the previous investigations in the area of rubberised concrete [3,4]. The effects of rubber on bleeding properties of concrete were also evaluated in this research. The results presented in Fig. 2(a) show a reduction of bleed water by the increase of rubber content for both of the mix arrays with WC ratios of 0.40 and 0.45. Considering only the volume of bleed water is not a proper way to compare mix arrays with different WC ratios and water contents, because those properties substantially affects the bleeding

of rubberised concrete. The Australian Standard AS1012.6 provides a relationship for quantifying bleeding in a standard and consistent way. According to the AS1012.6 the bleeding percent can be calculated for mix series, using the Eq. (2).

Bleeding ¼

V1  M S  V 2  10

ð2Þ

where V1 is the volume of bleed water in mL, M is total batch mass of concrete in kg, V2 is the total volume of unbound water in mix in L, and S is mass of test specimen in kg. In Australia 1.5–3% bleeding water is expected for concrete pavement. On the one hand, a low bleeding rate results in a low paste volume on the surface of concrete pavement, and causes difficulty in finishing. On the other hand, a high pavement with high bleeding rate results in a pavement with low surface strength and abrasion resistance. According to the results demonstrated in Fig. 2(b), rubber reduced the bleeding percentage of rubberised concrete. It was observed that introducing 30% or more rubber into concrete mixes resulted in bleeding lower than 1.5%. This indicated that there may be a difficulty with the surface finishing of the rubberised concrete prepared with 30% rubber content or more. In addition, it was revealed that introducing rubberised mix series prepared with 40% rubber content had 20% lower bleeding compared to the control samples (Fig. 2(b)). This effect might be the result of substituting a portion of aggregates with rubber particles, which are considered as impermeable constituents. The impermeable property of rubber might prevent migration of the free water to the top surface of concrete. In addition, it was reported by Singh [43] that using an air entraining agent results in higher air content

Fig. 2. (a) Quantity of bleed water and (b) calculated bleeding percent index.

112

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

and lowers the concrete bleeding. On the basis of this fact, the effect rubber on the increase of concrete air content (Table 4) can reduce bleeding. The higher air content, results in more internal voids within the concrete matrix, which trap free water from migration to the surface of concrete. It is important to take into account that both high and low bleeding rates negatively affect concrete pavements. For conventional concrete pavements, a low bleeding rate might lead to a fast drying out of the top surface of concrete, which results in plastic shrinkage cracks. On the contrary, limiting concrete bleeding to a certain maximum level of 3% increases its wear resistance. Therefore, considering concrete bleeding properties is not enough and other effects of rubber on shrinkage properties of concrete should also be taken into account before drawing any conclusions regarding the positive or negative effect of introducing rubber to concrete pavements. 4.2. Fracture properties Fracture properties of concrete samples were assessed according to ASTM C1609 test method and based on the formula shown in Eqs. (3) and (4).

RD150

¼

150  T D150

ð3Þ

2

f1  b  d

where RD150 is the toughness, T D150 is the area under the load vs. deflection curve up to 2 mm deflection, f1 is the peak strength in MPa presented in Eq. (4), b and d are sample width and depth in millimetres

f1 ¼

Pp  L

ð4Þ

2

bd

where f1 is the peak strength in MPa, Pp is the peak load in kN, b and d are sample width and depth in millimetres.

According to the result of the performed tests demonstrated in Table 5, adding rubber enhanced toughness of concrete by increasing the area under the load–deflection plotted chart. It was observed that samples with WC ratios of 0.40 and 0.45, which did not contain rubber, had very low toughness. However, the enhancement was found insignificant, mostly because of low residual load capacity after the first peak. The most significant observed enhancements was the large deflection of samples under the load. In addition, rubberised samples demonstrated a high resistance against a sudden splitting failure also quick propagation of cracks through the concrete matrix structure. The test results attained for the modified ASTM C1609 method did not show any significant difference compared to the first series of tests. The absence of any significant difference in results proved that lowering the loading rate of the standard test is not beneficial for rubberised concrete. In contrast, it made the test significantly time-consuming and difficult to be performed.

4.3. Plastic and drying shrinkage properties The plastic and drying tests were conducted in accordance with the Standards AS1012.13 and ASTM C1579, respectively. The ASTM C1579 test was found to be the most suitable, considering the restraining friction of subbase that is applied to concrete pavements. As can be seen from Fig. 3, there was an optimum point for rubber content, where the plastic shrinkage cracking was minimal. According to results presented in Fig. 4, adding rubber up to 20–25% provided a significant continuous improving effect on the resistance of concrete against plastic shrinkage. It was observed that up to 20–25% rubber content, the average crack length, width and area were reduced for both of the concrete arrays with WC ratios of 0.40 and 0.45. On the contrary, it was found that introducing extra rubber than the 20–25% content, cancelled out the

Table 5 Toughness test results conducted based on Standard ASTM C1609. Mix ID

dP = d1a

Dimension

dL/600 = 0.50 mmb

dL/600 = 0.50 mmc

Toughnessd

b (mm)

d (mm)

d (mm)

PP (kN)

fP = f1 (MPa)

PD 600 (kN)

f 600 (MPa)

PD 150 (kN)

f 150 (MPa)

TD 150 (J)

RD 150 (%)

101.8 100.3 101.1

99.6 99.8 99.5

0.17 0.12 0.14

22.34 22.93 23.92

6.64 6.89 7.17

0.00 0.00 0.00

0.00 0.00 0.00

0.00 0.00 0.00

0.00 0.00 0.00

1.93 1.42 1.71

4.32 3.10 3.57

Average: 0.40/00CR 0.40/25CR 97.2 0.40/25CR 101.1 0.40/25CR 100.5

102.1 100.7 101.1

0.18 0.19 0.21

16.59 17.10 17.87

4.91 5.00 5.22

0.94 0.97 0.97

0.28 0.28 0.28

0.57 0.59 0.60

0.17 0.17 0.18

2.84 2.90 3.20

8.56 8.48 8.95

Average: 0.40/25CR 0.40/40CR 99.7 0.40/40CR 100.3 0.40/40CR 100.7

97.8 98 98.1

0.27 0.26 0.28

14.79 15.08 15.38

4.65 4.70 4.76

3.88 3.93 4.06

1.22 1.22 1.26

0.31 0.33 0.33

0.10 0.10 0.10

5.47 5.66 5.92

18.49 18.77 19.25

Average: 0.40/40CR 0.45/00CR 100.4 0.45/00CR 99.8 0.45/00CR 101.4

102.4 99.9 100.1

0.28 0.29 0.29

21.47 22.43 22.75

6.12 6.75 6.72

0.00 0.00 0.00

0.00 0.00 0.00

0.00 0.00 0.00

0.00 0.00 0.00

3.20 3.29 3.34

7.45 7.34 7.34

Average: 0.45/00CR 0.45/20CR 99.9 0.45/20CR 101.4 0.45/20CR 99.5

101.8 98.4 99.7

0.25 0.29 0.28

15.86 16.43 16.60

4.60 5.02 5.04

3.94 3.95 4.15

1.14 1.21 1.26

0.49 0.50 0.50

0.14 0.15 0.15

5.88 5.12 5.35

18.54 15.58 16.11

Average: 0.45/20CR 0.45/40CR 98 0.45/40CR 99.6 0.45/40CR 100.4

101.4 99.4 99.1

0.24 0.3 0.33

12.21 12.74 13.10

3.64 3.88 3.99

0.59 0.62 0.63

0.18 0.19 0.19

2.15 2.17 2.20

0.64 0.66 0.67

3.66 4.01 5.30

14.99 15.74 20.22

0.40/00CR 0.40/00CR 0.40/00CR

D

D

3.66

8.66

18.84

7.38

16.74

Average: 0.45/40CR a b c d

RD 150 (%)

dP = net deflection at peak, d1 = net deflection at first peak, PP = peak load, fP = f1 = peak strength. D dL/600 = 0.5 mm net deflection, P D 600 = residual load at deflection of 0.5 mm, f 600 = residual strength at deflection of 0.5 mm. D dL/150 = 2 mm net deflection, P D 150 = residual load at deflection of 2 mm, f 150 = residual strength at deflection of 2 mm. Toughness calculated based on Eqs. (3) and (4).

16.98

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

113

Fig. 3. Different WC ratios and rubber content vs. (a) total crack length (b) average crack width (c) maximum crack width, (d) total crack area.

  AverageCrack Width of Modified Concrete ½mm CRR ¼ 1  AverageCrack Width of Reference Concrete ½mm  100 ð5Þ

Fig. 4. Total crack area and compressive strength vs. rubber content.

achieved improving effects and by continuing the addition of rubber, shrinkage cracking properties of concrete negatively impacted and became even worse than the concrete without rubber. Test results indicated that there was a correlation between the minimum required strength of the rubberised concrete (32 MPa) and the lowest plastic shrinkage cracking of concrete. It can be noted that concrete samples, which had very high or very low strengths did not show satisfactory resistance against plastic shrinkage cracks. Furthermore, a previous study discussed and recommended recording the time that the first crack occurs during the plastic shrinkage test [10]. Subsequently, this analysis was accomplished and the results are presented in Fig. 5(a). The plotted results indicate that the time of first crack occurrence was significantly longer for the optimum content of rubber. ASTM C1579 test method introduced the crack reducing ratio (CRR), which is calculated according to Eq. (5). This index is applied to quantify the amount of improvement in plastic shrinkage properties of modified concrete compared to the unmodified control concrete samples.

Results demonstrated in Fig. 5(b) revealed that rubberised concrete samples with the optimised rubber content showed the best shrinkage performance with high CRR value. In contrast, by introducing more rubber into concrete, the CRR decreased and turned to a negative value, which means it did not improve the resistance of concrete against shrinkage cracking. It is a well-established fact that concrete samples with high WC ratios have higher drying shrinkage because they contain more unbound free water in mix [13]. This idea was found valid for the rubberised concrete as shown in Fig. 6(a). Based on the results demonstrated in Fig. 6(a) it was confirmed that adding rubber into concrete mix series will result in the increase drying shrinkage for both the mix series with different WC ratios. The highest drying shrinkage was observed for mix 0.45/40CR which was prepared with the highest rubber content and WC ratio. It was observed that by introducing 40% rubber content the drying shrinkage values at different ages were increased 5% to 15%. The achieved results were supported by previous studies [26,27], however, the 15% increase in drying shrinkage was lower than the reported values in the previously published papers. For instance there was 90% higher drying shrinkage reported by Sukontasukkul & Tiamlom [27] for rubberised concrete prepared with 30% rubber. Fig. 6(b) presents the change of ‘‘mass loss’’ parameter for concrete arrays over 56 days of drying condition. This parameter can be calculated using Eq. (6).

Mass loss ¼ 1 

Massat age Massat age of

of T

ð6Þ

7th day

where Massat age of T is the samples mass at drying age of T in gram, Massat age of 7th day is the surfaced dried (SSD) weight of samples at 7th day after casting in gram. It can be seen that ‘‘mass loss’’ index results showed in Fig. 6(b) followed the same trend as drying shrinkage presented in Fig. 6(a) over the 56 days of testing period. Moreover, the results

114

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

Fig. 5. (a) Time the first crack observed on concrete slabs and (b) the calculated crack reducing ratio (CRR).

Fig. 6. (a) The drying shrinkage and (b) the mass loss under the standard drying condition.

shrinkage strain of concrete ecs should be calculated as the sum of the chemical (autogenous) shrinkage strain ecse and the drying shrinkage strain ecsd.

ecs ¼ ecse þ ecsd

ð7Þ

where ecs is design shrinkage, ecse autogenous shrinkage is and ecsd is drying shrinkage strain. The autogenous shrinkage strain shall be taken as

ecse ¼ ecse  ð1  e0:1t Þ 0

¼ ð0:06f c  1Þ  ð1  e0:1t Þ  50  106

Fig. 7. The measured drying shrinkage at 21 days.

demonstrated in Fig. 7 indicated that except for samples prepared with 40% rubber content, all mixes complied with the requirement of RTA Specification [40] which limits the drying shrinkage to 450 microstrains at 21-day drying shrinkage. According to Australian Standard AS3600, the drying shrinkage strain ecs can be determined for conventional concrete by conducting measurement after eight weeks of drying conditioning, in accordance with AS1012.13 test method or by calculation in accordance with Clause 3.1.7.2 of AS3600 code. In addition to AS1012.13 test method, this study investigated the applicability of cluse 3.1.7.2 of the Australian Standard AS3600. The Australian Standard AS3600 introduces drying shrinkage model to predict the free drying shrinkage respond of rubberised concrete instead of measuring concrete shrinkage strain directly, using AS1012.13 test method. Australian Standard AS3600 denoted that the design

ð8Þ

where ‘‘t’’ is time in days after setting of concrete, ecse is final auto0 genous shrinkage and f c is the 28-day characteristic compressive strength of concrete in MPa. At any time ‘‘t’’ after the commencement of drying, the drying shrinkage strain shall be taken as

ecsd ¼

ð0:8 þ 1:2e0:005th Þ  t 0:8 0  k4  ð1  0:008  f c Þ  ecsd:b t0:8 þ 0:15th

ð9Þ

where t is the time after the commencement of drying in days, k4 is selected 0.6 based on the ambient condition that drying shrinkage 0 occurring, f c is the 28-day characteristic compressive strength of concrete in MPa, th is hypothetical thickness of a member in mm and ecsd:b is the final basic drying shrinkage strain, which is 800  106 for Sydney.

th ¼

2Ag ue

ð10Þ

where Ag gross cross-sectional area of a member in mm2 and ue exposed perimeter of a member cross-section plus half the perimeter of any closed voids contained therein in millimetres.

115

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116 Table 6 Experimental drying shrinkage results up to 56 day vs. the prediction of AS3600 model. Mix ID

WC

Rubber content (%)

7-day results

14-day results

AS1012.13 (ls)

AS3600 (ls)

AS1012.13 ls)

AS3600 (ls)

AS1012.13 (ls)

AS3600 (ls)

AS1012.13 (ls)

AS3600 (ls)

0.40/00CR 0.40/25CR 0.40/40CR 0.45/00CR 0.45/20CR 0.45/40CR

0.4 0.4 0.4 0.45 0.45 0.45

0 25 40 0 20 40

223 246 258 245 258 255

305 ± 92 326 ± 98 337 ± 101 310 ± 93 327 ± 98 341 ± 102

332 370 385 393 411 427

384 ± 115 419 ± 126 436 ± 131 392 ± 118 420 ± 126 443 ± 133

373 407 423 464 480 506

453 ± 136 502 ± 151 526 ± 158 464 ± 139 504 ± 151 535 ± 160

496 514 538 554 558 591

502 ± 151 565 ± 170 596 ± 179 516 ± 155 567 ± 170 607 ± 182

Consideration shall be given to the fact that ecs provided by Clause 3.1.7.2 of the code AS3600 has a range of ±30% error. The calculated drying shrinkage results, based on AS3600 model, compared with the real test results, measured according to drying shrinkage test AS1012.13, as shown in Table 6. It can be seen that introducing rubber did not reduce the validity of the model introduced by AS3600 and the experimental drying shrinkage results for rubberised concrete were aligned with the numbers calculated from AS3600 drying shrinkage model. 5. Conclusions In this research, the effects of using recycled crumb rubber on shrinkage properties of concrete were quantified based on 11 sets of concrete samples. The following conclusions can be drawn from the results: (a) Analysis of results for the rubberised concrete strength revealed that the concrete strength decreased by increasing the rubber content. Therefore, to comply with the Australian pavement specification, the maximum rubber contents should be limited to 20% and 25% for water to cement ratios of 0.45 and 0.40, respectively. (b) It was observed that adding more rubber into the mix series, decreases concrete bleeding index significantly. In addition, test results indicated that introducing 30% or more rubber into concrete results in difficulty with finishing of pavement surface and should be avoided. (c) Although, adding rubber reduces the maximum load that samples can resist in fracture tests, the total area under the load–deflection curve increases slightly by increasing the rubber content. However, the observed enhancement in the toughness index was not found significant. (d) Plastic shrinkage test results revealed that the average crack widths, lengths and areas reduced significantly by adding 20–25% rubber into the mix series. It was observed that by introducing the optimised content of rubber to concrete mix series, the crack reducing ratio (CRR) index was improved notably. Moreover, the time of the first crack occurring was delayed significantly. However, adding extra rubber to mix series eliminated all these improvements and showed an inverse impact on the plastic shrinkage properties of the rubberised concrete slabs. (e) It was found that adding rubber into concrete resulted in higher free drying shrinkage strain. In addition, drying shrinkage test results revealed that the AS3600 numerical model for prediction of the concrete design shrinkage remains valid to be applied for the rubberised concrete.

References [1] Atech Group, A National Approach to Waste Tyres, The Australian Commonwealth Department of Environment, 2001, pp. 1–180.

28-day results

56-day results

[2] I. Mohammadi, H. Khabbaz, Challenges associated with optimisation of blending, mixing and compaction temperature for asphalt mixture modified with Crumb Rubber Modifier (CRM), Appl. Mech. Mater. 256–259 (2012) 1837–1844. [3] F. Valadares, M. Bravo, J. de Brito, Concrete with used tire rubber aggregates: mechanical performance, ACI Mater. J. 109 (2012) 283–292. [4] A. John, A.J. Kardos, Beneficial Use of Crumb Rubber in Concrete Mixtures, University of Colorado Denver (2011). [5] A.C. Ho, A. Turatsinze, A. Abou-Chakra, D.C. Vu, Rubberised concrete for the design of pavement on soil, Mater. Eng. Innovat. 3 (2012) 101–116. [6] I. Mohammadi, H. Khabbaz, K. Vessalas, In-depth assessment of Crumb Rubber Concrete (CRC) prepared by water-soaking treatment method for rigid pavements, Constr. Build. Mater. 71 (2014) 456–471. [7] M. Collepardi, A. Borsoi, S. Collepardi, J.J. Ogoumah Olagot, R. Troli, Effects of shrinkage reducing admixture in shrinkage compensating concrete under nonwet curing conditions, Cement Concr. Compos. 27 (2005) 704–708. [8] E. Boghossian, L.D. Wegner, Use of flax fibres to reduce plastic shrinkage cracking in concrete, Cement Concr. Compos. 30 (2008) 929–937. [9] J. Newman, B.S. Choo, Advanced Concrete Technology, ButterworthHeinemann, 2003. [10] a. Sivakumar, M. Santhanam, A quantitative study on the plastic shrinkage cracking in high strength hybrid fibre reinforced concrete, Cement Concr. Compos. 29 (2007) 575–581. [11] W. Jason Weiss, Wei Yang, Surendra P. Shah, Influence of specimen size/geometry on shrinkage cracking of rings, J. Eng. Mech. 126 (2000) 93–101. [12] M. Khorrami, A. Vafai, A. Khalilitabas, C.S. Desai, M. Ardakani, Experimental investigation on mechanical characteristics and environmental effects on rubber concrete, Int. J. Concr. Struct. Mater. 4 (2010) 17–23. [13] E. Holt, M. Leivo, Cracking risks associated with early age shrinkage, Cement Concr. Compos. 26 (2004) 521–530. [14] S. Tongaroonsri, S. Tangtermsirikul, Influence of mixture condition and moisture on tensile strain capacity of concrete, ScienceAsia 34 (2008) 59–68. [15] M. Emborg, Thermal Stresses in Concrete Structures at Early Ages, Lulea University of Technology (1989). [16] J. Byfors, Plain concrete at early ages, Swed. Cement Concr. Res. Inst. (1980). [17] J. Kang, Y. Jiang, Improvement of cracking-resistance and flexural behavior of cement-based materials by addition of rubber particles, J. Wuhan Univ. Technol.-Mater. Sci. Ed. 23 (2008) 579–583. [18] K.V. Subramaniam, R. Gromotka, S.P. Shah, K. Obla, R. Hill, Influence of ultrafine fly ash on the early age response and the shrinkage cracking potential of concrete, J. Mater. Civ. Eng. 17 (2005) 45–53. [19] I.B. Topcu, The properties of rubberized concretes, Cem. Concr. Res. 25 (1995) 304–310. [20] M.M. Balaha, A.A.M. Badawy, M. Hashish, Effect of using ground waste tire rubber as fine aggregate on the behaviour of concrete mixes, Ind. J. Eng. Mater. Sci. 14 (2007) 427–435. [21] N. Segre, I. Joekes, Use of tire rubber particles as addition to cement paste, Cem. Concr. Res. 30 (2000) 1421–1425. [22] N. Eldin, A. Senouci, Rubber-tire particles as concrete aggregate, J. Mater. Civ. Eng. 5 (1994) 478–496. [23] A. Radocea, A Study on the Mechanism of Plastic Shrinkage of Cement-based Materials, Chalmers University of Technology (1992). [24] D. Raghavan, H. Huynh, Workability, mechanical properties, and chemical stability of a recycled tyre rubber-filled cementitious composite, J. Mater. Sci. 3 (1998) 1745–1752. [25] K. Jingfu, H. Chuncui, Z. Zhenli, Strength and shrinkage behaviors of rollercompacted concrete with rubber additives, Mater. Struct. 42 (2008) 1117– 1124. [26] M. Bravo, J. de Brito, Concrete made with used tyre aggregate: durabilityrelated performance, J. Cleaner Prod. 25 (2012) 42–50. [27] P. Sukontasukkul, K. Tiamlom, Expansion under water and drying shrinkage of rubberized concrete mixed with crumb rubber with different size, Constr. Build. Mater. 29 (2012) 520–526. _ Topçu, The role of scrap rubber particles on the drying [28] T. Uygunog˘lu, I.B. shrinkage and mechanical properties of self-consolidating mortars, Constr. Build. Mater. 24 (2010) 1141–1150. [29] C. Qi, Quantitative Assessment of Plastic Shrinkage Cracking and Its Impact on the Corrosion of Steel Reinforcement, Purdue University (2003). [30] A. Neville, J.J. Brooks, Concrete Technology, 2010.

116

I. Mohammadi, H. Khabbaz / Cement & Concrete Composites 62 (2015) 106–116

[31] Australian DataSheet, Early-Age Shrinkage of Concrete, Cement Concrete & Aggregate Australia, 2005. [32] P. Kraai, A proposed test to determine the cracking potential due to drying shrinkage of concrete, Concr. Construct. 30 (1985) 775–778. [33] W. Weiss, W. Yang, S. Shah, Factors influencing durability and early-age cracking in high-strength concrete structures, ACI 182 (2001) 387–409. [34] K. Wang, S. Shah, P. Phuaksuk, Plastic shrinkage cracking in concrete materials – influence of fly ash and fibers, ACI Mater. J. 11 (2001) 458–464. [35] J, Mora, R. Gettu, C. Olazabal, M. Martin, Effect of the incorporation of fibers on the plastic shrinkage of concrete, Fiber-Reinforced Concretes (FRC), in: RILEM Symposium on Fibre-reinforced Concretes, Lyon, France, 2000, p. 705–714. [36] N. Banthia, C. Yan, S. Mindess, Restrained shrinkage cracking in fiber reinforced concrete: a novel test technique, Cem. Concr. Res. 26 (1996) 9–14. [37] J.-L. Chermant, Why automatic image analysis? An introduction to this issue, Cement Concr. Compos. 23 (2001) 127–131.

[38] S. Mindess, S. Diamond, A preliminary SEM study of crack propagation in mortar, Cem. Concr. Res. 10 (1980) 509–519. [39] S. Diamond, D. Bonen, A re-evaluation of hardened cement paste microstructure based on backscatter SEM investigations, in: Materials Research Society Symposium, Materials Research Society, 1995. [40] RTA-R83, Joint Concrete Base Specification R83, 2010. [41] Moss J. Austroad, K. Sharp, D. Dash, J. Hodgkinson, D. Boon et al., Guide to Pavement Technology: Part 4C – Austroads, 2009. [42] S.T. Islam, Study of Some Parameters Affecting the Measured Flexural Toughness of Fiber Reinforced Concrete, The University of British Columbia (2012). [43] B. Singh, Bleeding in concrete, Int. J. Civil Eng. Technol. (IJCIET) 4 (2013) 247– 249.