Microwave-assisted beneficiation of recycled concrete aggregates

Construction and Building Materials 25 (2011) 3469–3479

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Microwave-assisted beneficiation of recycled concrete aggregates A. Akbarnezhad a,⇑, K.C.G. Ong a, M.H. Zhang a, C.T. Tam a, T.W.J. Foo b a b

National University of Singapore, 1 Engineering Drive 2, E1A 07-03, Singapore 117576, Singapore Ngee Ann Polytechnic, Blk 8 #03-00, 535 Clementi Road, Singapore 599489, Singapore

a r t i c l e

i n f o

Article history: Received 21 October 2010 Received in revised form 24 January 2011 Accepted 1 March 2011 Available online 25 March 2011 Keywords: Aggregate Concrete Recycling Beneficiation Microwaves Heating

a b s t r a c t The presence of mortar has been reported as the main factor causing the lower quality of recycled concrete aggregates (RCA) when compared to natural aggregates (NA). A novel microwave-assisted technique to increase the quality of RCA by partially removing the mortar adhering to RCA particles and breaking up the lumps of mortar present in RCA is introduced in this paper. The technique takes advantage of the difference in the electromagnetic properties of the adhering mortar and natural aggregates to generate high thermal stresses within the mortar, especially at the interface with the embedded natural aggregates, to cause delamination. The stresses generated also result in the breaking up of the lumps of mortar into smaller pieces. The results of an experimental study conducted to investigate the capability of the microwave-assisted RCA beneficiation technique and to compare its efficiency with other beneficiation methods proposed in available literature are presented. Moreover, the effects of incorporating various amounts of un-treated and microwave-treated coarse RCA on the mechanical properties of concrete are investigated. The temperature distribution and stresses developed in RCA when subjected to microwave heating during the beneficiation process are numerically calculated for a better understanding of the processes involved. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Recycled concrete aggregates (RCA) are produced by reducing the size of concrete debris through multiple crushing stages. Depending on the size, the crushed concrete particle may comprise one or more particles of natural coarse aggregate held together and surrounded fully or partially by a layer of mortar (Fig. 1a and b). In the present study, such RCA particles are referred to as Type I RCA. Moreover, as can be seen in Fig. 1c, after sieving into the respective aggregate size grades, recycled coarse aggregates which are essentially lumps of mortar with varying proportions of smaller size natural aggregates embedded are also present. In this paper, the RCA particles comprising entirely of mortar are referred to as Type II RCA. There is as-yet no practical method to separate the lumps of mortar from the RCA particles with the embedded natural coarse aggregates in a recycling plant. Therefore, batches of RCA usually comprise varying proportions of both types. The presence of mortar in RCA has been identified as the most important factor lowering the quality of recycled concrete aggregates [1]. The presence of cementitious mortar has been reported to result in a lower density, higher water absorption, lower Los Angeles abrasion resistance and higher soundness loss of RCA when compared to natural aggregates [2–5]. This has limited the ⇑ Corresponding author. Tel.: +65 65163786. E-mail address: [email protected] (A. Akbarnezhad). 0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2011.03.038

use of RCA in the structural concrete to only small replacement percentages of the natural aggregates (up to 30%) [6] because the differences in the properties of RCA and natural aggregates will in turn affect the fresh, mechanical and long term properties of the concretes made using RCA; e.g. lowering the density, workability, compressive strength, tensile strength, modulus of elasticity, frost resistance, chloride penetration, and increasing the creep and shrinkage. when compared to natural aggregate concrete [3,7–15]. The volume percentage of old mortar in RCA may vary between 20% and 70% [16,17], depending on the grain size, strength of the parent concrete and the crushing process used [16,18,19]. The undesirable effect of the mortar on the properties of RCA particles has been found to be proportional to its content [1]. A number of RCA beneficiation methods have been recently proposed to enhance the quality of RCA through reduction of the mortar present. In these methods, one or a combination of mechanical, thermal and chemical treatments is usually used to remove the mortar. However, these methods are either too time and energy consuming or do not adequately increase the quality of RCA produced [1]. A new microwave-assisted beneficiation method to reduce the mortar content of RCA is presented in this paper. This method takes advantage of the differences in electromagnetic properties, water absorption and the tensile strength of natural aggregates and mortar to break up and separate the mortar without damaging

3470

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

Fig. 1. Various types of RCA comprising (a) a granite particle surrounded by adhering mortar, (b) three granite particles held together and surrounded by mortar, (c) only mortar.

the natural aggregates. The process involved is twofold. First, microwave heating is used to generate high differential thermal stresses concentrated in the adhering mortar (AM), especially at its interface with the embedded natural aggregates (ENA), to cause delamination of the adhering mortar present in Type I RCA particles. Next, further microwave heating is used to break up the mortar delaminated from Type I particles as well as the mortar lumps present (Type II) to smaller pieces that can be easily collected through sieving. The capability of the microwave-assisted beneficiation method to increase the quality of RCA through reduction of the total mortar content is experimentally investigated and compared with the acid pre-soaking, conventional heating, mechanical rubbing and ‘‘heating and rubbing’’ methods as proposed in available literature. Furthermore, the efficacy of using the combined ‘‘microwave heating and mechanical rubbing’’ processes is examined. Numerical modeling of the microwave-RCA interaction is used to estimate the temperature distribution and the thermal stresses developed in an idealized RCA particle when exposed to microwaves for a better understanding of the phenomena involved. In addition the effects of incorporating various amounts of un-treated RCA and microwave-treated RCA (MRCA) on the compressive strength, modulus of elasticity and flexural strength of concrete are experimentally investigated.

eccentric-shaft rotor method, crushed concrete lumps are passed downward between an outer cylinder and an inner cylinder that rotates eccentrically at a high speed to separate the coarse aggregate from the mortar through grinding. In the mechanical grinding method, a drum is divided into small sections with partitions. The mortar portion of the RCA is removed by rubbing against the iron balls placed in each of the rotating partitioned sections of the drum. 2.3. Thermal–Mechanical beneficiation

2. Previously proposed RCA beneficiation methods

In this method a combination of the thermal stresses generated through conventional heating at temperatures from 300 °C to 500 °C and the mechanical stresses generated through rubbing is used to remove mortar from the RCA particles. In 1999, Shima et al. proposed a thermal–mechanical RCA treatment technique known as ‘‘heating and rubbing’’ [24]. In this technique, concrete debris are first heated at 300 °C in a vertical furnace to render the cement paste brittle due to dehydration. To remove the mortar, the heated concrete debris are fed into the rubbing equipment. In the equipment, the heated concrete is rubbed against steel balls and the mortar portion that is dislodged is discharged through the screening system provided [21]. Inventors of this method claimed that it can increase the quality of RCA to comply with the JCI (Japan Concrete Institute) standards for high quality recycled concrete aggregates [25].

2.1. Conventional heating (thermal beneficiation)

2.4. Acid Soaking beneficiation

In this method, RCA particles are heated at about 500 °C for a duration of about two hours. The thermal stresses generated through thermal expansion are used to fracture and thereby remove the mortar present [20]. Moreover, according to Shima et al. when concrete is heated at temperatures higher than 300 °C, mortar is made brittle due to dehydration; lowering its resistance against the thermal stresses developed [21]. It is believed that saturating the mortar before heating can increase the efficiency of this method because it can lead to pore pressure development which may result in the faster removal of mortar. It has also been reported that immersing the heated aggregates in cold water immediately after heating can lead to higher mortar removal yields through increasing the differential thermal stresses developed [1].

More recently, Tam et al. proposed a new method to remove the mortar by pre-soaking RCAs in 0.1 M acidic solutions for 24 h [26]. Three different acidic solutions (HCl, H2SO4, H3PO4) were considered in this study. They reported that the water absorption of RCA after treatment reduced, showing improvements in the range of 7.27–12.17%. A major drawback of this method is the increase in the chloride and sulfate content of the aggregates respectively after treatment with hydrochloric and sulfuric acids. The increase in the chloride and sulfate content of aggregates may cause durability problems.

2.2. Mechanical beneficiation In this technique, mechanical forces are used to grind and remove the mortar. Two techniques have been proposed in Japan; eccentric-shaft rotor [22] and mechanical grinding [23]. In the

2.5. Chemical–Mechanical beneficiation Abbas et al. proposed to use combined chemical degradation through exposure of RCA to sodium sulfate solution and mechanical stresses created through subjecting RCA to freeze-and-thaw action to separate mortar from RCA [27]. However, the main objective of this study was focused on quantifying the amount of mortar present for use in RCA classification. The technique ‘‘as is’’, is not suited for full scale RCA production.

3471

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

3. Microwave-assisted beneficiation A typical RCA particle is composed of natural aggregate (NA) and mortar. The type of coarse aggregate as well as the sand and cement used may vary widely from one source to another. In this study, due to the lack of reliable data on the electromagnetic properties of other types of material and because of the widespread use of granite coarse aggregate, natural river sand and Type I Portland cement in Singapore’s construction, only these were used to produce RCA tested in this study. However, the concepts discussed herein are general and are expected to be relevant for RCA originating from the concrete made with other types of coarse aggregates, sand and cement.

Attenuation Factor (neper / m)

100

Wet Mortar

80

Saturated Mortar Air-Dried Mortar

60

40

Coarse Aggregate

20

0 0

5

10

15

20

Microwave Frequency (GHz) 3.1. Working principle Fig. 2. The attenuation factors of coarse aggregate and cementitious mortar.

Natural aggregate and mortar are both dielectric materials. The extent and pattern of microwave heating of dielectric materials varies with the microwave frequency, microwave power and most importantly the electromagnetic (EM) properties of the material. A dielectric material can be characterized by two independent electromagnetic properties, the complex permittivity e and the complex (magnetic) permeability l [28]. Complex permittivity is defined as

e ¼ e0  ie00

ð1Þ

where e0 and e00 are the real and imaginary parts of complex permittivity, respectively. Dividing this by the permittivity of the free space, e0, the property becomes dimensionless and relative to the permittivity of the free space

e e0 e00 ¼ i e0 e0 e0

ð2Þ

or

er ¼ e0r  ie00r

ð3Þ

where e0 is the permittivity of the free space, er is the relative permittivity, e0r is the dielectric constant and e00r is the loss factor of the material. Dielectric constant is a measure of how much of the energy from an external electric field is stored in a material and the loss factor is a measure of how dissipative or lossy a material is to an external field [29]. The microwave power dissipated per unit volume of a dielectric material is directly proportional to its loss factor [29]:

PLðxÞ ¼

1 xe0 e00r jEj2 2

ð4Þ

where E is the electric field strength. As a result of the dissipation, microwave energy decays in the dielectric material. The power penetration depth is defined as the depth at which the transmitted power drops to 1/e of its value at the surface and is given by

dp ¼

1 2b

ð5Þ

Here, b is the attenuation factor which can be calculated using the basic electromagnetic properties of the material. As can be seen, in Eq. (5), the attenuation factor is inversely proportional to the dissipation depth and thus directly proportional to the amount of the microwave energy dissipation.

2p f b¼ c

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi e0r ð 1 þ tan2 d  1Þ 2

(1998), are compared in Fig. 2. The mortar specimens tested by Rhim et al. were cast with water/cement/sand mix ratio of 1:2.22:5.61 (by weight) using Type I Portland cement and natural river sand. Comparing the loss factor of air-dried mortar to coarse aggregates shows that for a similar moisture condition, the attenuation factor and thus the microwave energy absorption rate of mortar is higher than natural aggregates. Hence, if a RCA particle comprising one or more coarse aggregates held together and surrounded partially or fully by mortar is exposed to microwaves, mortar would be heated up much faster than the embedded natural aggregates and thus, significant differential thermal stresses may be developed within the mortar, especially at the mortar–aggregate interface. In microwave-assisted RCA beneficiation technique, such differential thermal stresses are harnessed to delaminate the adhering mortar from RCA particles (Type I particles). Moreover, further microwave heating of RCA is used to increase the stresses developed in the delaminated mortar and the mortar lumps (Type II particles) present from the start, gradually breaking them up into the constituent materials, sand, fines, cementious powder, etc. By fine tuning the process to avoid excessive heating, because of the slower heating rate and significantly higher tensile strength of natural aggregates compared to mortar, the respective stresses developed in the natural aggregates at the time of mortar delamination and pulverization are significantly lower than the aggregate’s tensile strength. Therefore, the RCA beneficiation process may be easily optimized to avoid damage to the integrity of natural aggregates, unless defects such as fissures were already present. The microwave-assisted technique can be easily incorporated into the conventional concrete recycling technique. The technology needed for procuring a workable microwave beneficiation system is readily available. Continuous microwave heating tunnels with conveyor belt are widely used in food and timber drying industries [29]. The conveyor belt of the microwave system may be easily connected to the conveyor system moving the crushed concrete pieces through the other conventional stages of concrete recycling. Nonetheless, one of the limitations of microwave-assisted beneficiation technique is that the amount of organic impurities such as asphalt in the recycled materials should be minimized before microwave processing to avoid occurrence of arching and fire ball phenomena in the microwave tunnel. This may pose additional limitations on the source of the concrete debris. 3.2. Effects of mortar water content

ð6Þ

Here, c is the speed of light, tan d the loss tangent of concrete and f the microwave frequency. The attenuation factors of natural coarse aggregates and cementitious mortar, as reported by Rhim et al.

The efficacy of microwave-assisted RCA beneficiation is directly proportional to the difference in the microwave energy absorption (heating rate) between mortar and aggregates which leads to the development of higher thermal stresses in the mortar rather than

3472

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

the aggregates. As can be seen in Fig. 2, the attenuation factor and therefore microwave energy dissipation (=energy absorption by material) in mortar increases significantly with an increase in the water content. Also, it is well known that mortar has considerably higher water absorption rate, enabling it to absorb water much faster, compared to natural aggregates. To accentuate this, immersing the RCA particles in water for a specified duration, e.g. 10 min is advocated. Moreover, it is well known that the w/c ratio increases at the interfacial transition zone (ITZ) and can be quite significant if bleeding is predominant; hence, the ITZ normally has a higher porosity and water absorption than the bulk cementitious mortar. Therefore, in the case of Type I RCA particles, as a result of higher water content, the ITZ is expected to heat up even faster when the RCA is exposed to microwaves. As a result, higher differential thermal stresses may develop at the ITZ, leading to faster delamination of the adhering mortar. 4. Experimental program This experimental study was conducted to investigate the capability of the microwave-assisted RCA beneficiation technique to reduce the mortar content of RCA and to compare its efficiency with that achievable using the acid pre-soaking beneficiation, conventional heating beneficiation, mechanical rubbing beneficiation and ‘‘conventional heating and rubbing’’ beneficiation methods as proposed in available literature. The efficacy of different RCA beneficiation methods was investigated by comparing the change in the water absorption, particle density and mortar content of the RCA particles after being subjected to the respective beneficiation process. Furthermore, the effects of incorporating the microwave-beneficiated RCA (MRCA) on the compressive strength, flexural strength and the modulus of elasticity of concrete were investigated and compared with concrete cast using un-treated RCA.

4.1. Material Experiments were carried out on the RCA produced by crushing the laboratorycast concrete specimens. The Mix proportions used for the parent (control) concrete are presented in Table 1. ASTM Type I normal Portland cement with a specific gravity of 3150 kg/m3 and fineness of 347 m2/kg was used for the concretes. Moreover, granite aggregate with a maximum size of 12 mm and natural sand were used for all the concretes. Bulk densities of the natural coarse and fine aggregates in oven dry (OD) condition were 2580 kg/m3 and 2540 kg/m3, respectively. Furthermore, the respective 24-h water absorption capacities of natural coarse and fine aggregates were approximately 0.6% and 0.5%. To produce RCA, 40 day old concrete specimens were crushed in a jaw crusher. According to available literature, the properties of RCA change with the particle size [1,16,17]. De Juan et al. (2009) showed that the amount of mortar present in RCA is inversely proportional to the size of particle. Therefore, in order to eliminate the effects of RCA size and thereby minimize variation of the particles properties within samples tested, the RCA particles belonging to the 8–12 mm size fraction were used throughout this experimental study. The choice of the 8–12 mm size fractions is mainly because a series of preliminary tests focusing on the properties of size fractions 4–8 mm, 8–12 mm, 12–16 mm and16–20 mm showed that the 8–12 mm fraction gave results that were close to the average of the 4–20 mm size fraction.

On average, 17% of the total particles in the samples tested (8–12 mm) were Type II RCA, comprising only mortar while the rest were comprised of one or more granite particles held together and partially or fully surrounded by mortar. Since, there is as-yet no practical method to separate the lumps of mortar from Type I RCA particles, a mixture of both types as produced were used in the present study. The 24-h water absorption, particle density (OD) and total mortar content of the RCA samples were, on average, 4.2%, 2370 kg/m3, and 47% (by mass), respectively.

4.2. Experimental methodology for RCA beneficiation 4.2.1. Microwave heating Commercially available microwave ovens operate at the intermediate power level and have a limited heating duration. On the other hand, in microwave heating, the degree of the heating and temperature rise is directly proportional to the volume of the material heated. Therefore, because of the limited power and heating duration, a very small amount of RCA may be heated in each batch using commercially available microwave ovens. To cope with these problems, in the present study, a pilot industrial microwave-assisted RCA beneficiation system was designed and installed in the Structural Engineering Laboratory of the Department of Civil Engineering, National University of Singapore. This system may operate continuously at a power level of up to 10 kW. As can be seen in Fig. 3, the microwave-assisted RCA beneficiation equipment comprises a 10 kW microwave generator unit and a RCA heating chamber with ventilation to remove the dust and water vapor generated during microwave heating. Moreover, this system comprises a series of waveguide components (inside the chamber) including an auto-tuner to minimize the reflection of microwave power, a direction coupler to measure the forward and reflected power and an isolator to protect the generator against possible reflected power. To investigate the efficacy of the microwave-assisted RCA beneficiation method, the system described was used to heat 2 kg (oven dried weight) RCA samples at the maximum power (10 kW) for 1 min. After heating, the samples were immediately cooled down by immersing in 25 °C water. To examine the effects of the RCA water content, RCA samples with two different initial moisture conditions were considered; (1) Air-dried (AD): 2 kg oven dried RCA samples were immersed in water for 24 h and were then kept at room temperature conditions for 21 days; (2) Saturated (SA): Air-dried RCA samples were immersed in water for 24 h.

4.2.2. Other RCA beneficiation techniques The methodology used in this study is described in the following sections.

4.2.2.1. Conventional heating beneficiation (thermal beneficiation). Saturated RCA samples (2 kg oven dry weight) were heated in a conventional furnace for 2 h. Two different heating temperatures of 300 °C and 500 °C were used. After heating, the RCA samples were immediately cooled down by immersing in a water tank filled with 25 °C water.

4.2.2.2. Mechanical rubbing beneficiation. Los Angles abrasion testing equipment with a load of 10 steel balls was used to rub the RCA samples (10 kg oven dried weight) against each other and the steel balls for 100 revolutions of the rotating drum.

4.2.2.3. Thermal–Mechanical beneficiation. RCA samples were heated at 500 °C as in the thermal beneficiation technique. Ten kilograms (oven dried weight) batches of heated RCA were then rubbed using the Los Angles abrasion testing equipment as in the mechanical rubbing beneficiation technique.

Table 1 The mix proportions of concrete cast with various amounts of coarse un-treated (RCA) and microwave treated (MRCA) recycled aggregates.

a

Specimen ID

RCA/MRCA replacement (%)

W/ C

Cement (kg/m3)

Mixing water (kg/m3)

Natural coarse agg. (kg/m3)

Natural sand (kg/m3)

RCA/MRCA (kg/m3)

Pre-mixing watera (kg/m3)

Super plasticizer (L/m3)

CC RAC20 RAC40 RAC60 RAC80 RAC100 MRAC20 MRAC40 MRAC60 MRAC80 MRAC100

0 20 40 60 80 100 20 40 60 80 100

0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45 0.45

375 375 375 375 375 375 375 375 375 375 375

167 167 167 167 167 167 167 167 167 167 167

1072 857.6 643.2 428.8 214.4 0 857.6 643.2 428.8 214.4 0

736 736 736 736 736 736 736 736 736 736 736

0 214.4 428.8 643.2 857.6 1072 214.4 428.8 643.2 857.6 1072

9.78 17.64 25.51 33.38 41.25 49.12 14.71 19.64 24.57 29.50 34.43

2 2 2 2 2 2 2 2 2 2 2

Calculated based on the 24 h water absorption of aggregates.

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

3473

Fig. 3. The pilot microwave heating system designed and fabricated for use in the experimental program, (a) Microwave generator unit (b) RCA beneficiation chamber.

4.2.2.4. Microwave heating and mechanical rubbing. To investigate the efficiency of combining the microwave heating and mechanical rubbing treatments, 10 kg batches of the saturated RCA were microwave heated as described in Section 4.2.1 and were then rubbed using Los Angles abrasion testing equipment as in the mechanical rubbing technique.

4.2.3.5. Acid soaking beneficiation. Two kilograms oven dried samples were placed into a plastic container which was then filled with the acid solution, diluted to the desired concentration. The samples were soaked for the specified duration and were then washed on a 4 mm sieve to remove the detached mortar and acid. Three sulfuric acid concentrations of 0.1, 0.5 or 1 M at two soaking durations of 1 day or 5 days were considered.

The mix proportions of the concrete cast are listed in Table 1. Similar cement, natural coarse and fine aggregate types and mix proportions as that used for the concrete crushed to produce RCA were used. The coarse and fine aggregates (including RCA and MRCA) were oven dried before casting. The fine and coarse aggregates were then mixed with the appropriate amount of water, calculated to compensate for the water absorption of natural and recycled aggregates, for 10 min before addition of the cement and mixing water. This is based on the fact that more than 90% of the aggregates 24 h water absorption takes place within the first 10 min of exposure to water. The water absorption versus time curve for the RCA used in the present study is shown in Fig. 5. All concrete specimens were moist cured at about 28 °C for 7 days and were then kept under ambient laboratory conditions (28 °C and relative humidity of about 80–85%) until the time of testing.

4.3. Measurement of the remaining mortar content

5. Numerical study There is as yet no standard method prescribed to determine the mortar content of RCA. The acid concentration (0.1 M) proposed in the acid pre-soaking method developed by Tam et al. [26] was reported to result in only 7.27–12.17% reduction in the water absorption of the RCA samples tested. However, considering the difference between the water absorption of natural aggregates (<1%) and a typical RCA (between 2% and 5%), such a small reduction in water absorption shows that there is still a considerable amount of mortar present in RCA. The preliminary results of the experiments conducted in the present study showed that in order to completely remove the mortar from the type of RCA tested, soaking in sulfuric acid concentration of at least 2 M for at least 5 days was required (Fig. 4). Hence, this procedure was adapted to measure the mortar content of the RCA batches treated by the various methods used in this study. After soaking in acid, coarse aggregates were collected by washing the samples on a 4 mm sieve. The weight loss of the coarse aggregates collected was used to calculate the mortar content.

Mc ¼

Wi  Wr  100 Wi

ð7Þ

where Mc is the mortar content, wi is the initial oven dried weight of the RCA sample and wr is the oven dried weight of the RCA samples retained on the 4 mm sieve after soaking in the acid. Subjecting samples of the natural aggregates to similar conditions confirmed that the testing procedure used leads to negligible weight loss of the natural aggregates themselves. Therefore, the measurements may be regarded as the mortar content in RCA.

4.4. Effects of RCA and MRCA on the mechanical properties of concrete To investigate the effects of RCA beneficiation on the properties of concrete made using the microwave-treated recycled aggregates (MRCA), six batches containing various amounts of coarse MRCA (0%, 20%, 40%, 60%, 80% and 100%, replacement by mass of coarse NA with MRCA) were cast. Moreover, for comparison purposes, similar batches of concrete were prepared using the un-treated RCA.

5.1. Model description To investigate the capability of the microwave-assisted RCA beneficiation technique when used on an industrial scale, rather than modeling the system configuration used in the experimental studies, a proposed deployable industrial system is numerically simulated. The microwave beneficiation system considered is schematically shown in Fig. 6. As can be seen, a typical industrial microwave heating system may comprise a heating tunnel with a conveyor belt that can be easily integrated as part of the processes of a typical concrete recycling plant. Other components of the system include the microwave generator unit, a horn (antenna) to spread the microwave power over a specific area, and a series of microwave waveguides to transfer the microwave power from the generator to the horn. In practice, stirrers are required to provide a more uniform heating pattern within the chamber and churning of the RCA particles during microwave heating. However, due to the lack of accurate methods to simulate their function, stirrers are not simulated in this study. This leads to non-uniform heating of RCA with heating primarily from the top exposed surface. Nevertheless, the simulation presented in this paper is useful to illustrate the capability of microwaves to heat the mortar and natural aggregate to different extents and to estimate the differential stresses developed. Similar to the system used in the experimental studies, the microwave generator simulated is assumed to operate at a fre-

3474

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

sions at its larger end as shown in Fig. 6. TE10 is the most common incident mode excited by industrial microwave heating systems, providing a sinusoidal microwave power distribution on the incident surface of the material heated.

Mortar removed / total adhering mortar (%)

100

5 days 3 days

80 60

5.2. Simulation methodology

1 day

40 20 0 0

1

2

3

4

5

Sulfuric Acid Molarity Fig. 4. Effects of the soaking duration and acid concentration on the amount of the mortar removed through acid soaking method.

Water absorption (%)

5

4

3

2

1

0 1

10

100

1000

Log [Time (min)] Fig. 5. Water absorption of recycled concrete aggregates up to 24 h of soaking.

quency of 2.45 GHz and 10 kW incident power. As an illustrative case, the temperature rise and thermal stresses in an idealized Type I RCA particle (Fig. 7) comprising a 10 mm granite core and 2 mm thick adhering mortar is considered to examine the capability of microwave heating to delaminate the adhering mortar. To investigate the effect of the RCA water content on the magnitude and pattern of the development of temperature rise and thermal stresses across the RCA particle, two different moisture content conditions are considered, saturated and air-dried. The RCAs are assumed to be directly exposed to a microwaves beam of TE10 mode through a standard WR340 waveguide (86.36 mm  43.18 mm) and a horn of 100 mm  100 mm dimen-

The numerical simulation of microwave-assisted RCA beneficiation generally involves estimating the microwave energy dissipation in the RCA through electromagnetic analysis, solving the heat transfer equation to predict the temperature rise and temperature gradient and the thermal-structural analysis to predict the thermal stresses developed. The problem of electromagnetic analysis on a macroscopic level involves the solving of Maxwell’s equations subject to the requisite boundary conditions. Maxwell’s equations are a set of equations, establishing the relationships between the fundamental electromagnetic quantities including the electric field intensity (E). Once, the electric field intensity throughout the RCA particle has been obtained by solving Maxwell’s equations, Eq. (1) may be used to estimate the microwave power dissipation at any specific point. The results obtained using Eq. (1) are used as the heat source function in the heat transfer equation to calculate the temperature rise and temperature gradient throughout the RCA particle. The temperature distribution function obtained is then used to calculate the thermal stresses developed in the RCA particle. In the present study, the ‘‘general equation solver’’ module of the commercially available Comsol Multiphysics finite element package is used to solve the coupled electromagnetic, thermal and structural problem described. A more comprehensive description of the microwave heating formulation and simulation methodology may be found in Refs. [30,31]. Properties of the mortar and the granite considered are listed in Table 2. 6. Results and discussion 6.1. Properties of the microwave-assisted beneficiated RCA (MRCA) The magnified surface of an individual Type I RCA particle before and after microwave heating is depicted in Fig. 8. Visual inspection showed that a considerable portion of the layer of adhering mortar was removed from the RCA particles after microwave heating. Removal of the adhering mortar was accompanied by moderately loud noises and started after the minimum heating duration of 10 s for the saturated samples and 30 s for the air dried samples. The amount of the steam generated was not considerable and was collected together with the dust generated using the attached vacuum system connected to the microwave heating

Fig. 6. The RCA beneficiation system considered for numerical simulation.

3475

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

Table 2 Electromagnetic, thermal and structural properties of granite and mortar used for simulation.

Dielectric constant Electric conductivity (mohs/m) Thermal conductivity (J/m2 s °C) Expansion coefficient (106/°C) Specific heat (J/kg °C) Modulus of elasticity (GPa)

Aggregate (granite)

Mortar

6 0.071 4.3 11 800 50

6.42 0.372 1.3 19 1600 25

6.2. Temperature rise and thermal stresses in microwave heated RCA (numerical results)

Fig. 7. The RCA particle considered in the numerical model.

chamber. Moreover, it was observed that most of the delaminated adhering mortar and the mortar lumps (Type II RCA) were broken up when microwave heating was continued for a duration of 1 min. In addition, inspection showed that the remaining adhering mortar and un-broken smaller mortar lumps were severely weakened so that they could be easily broken up by hand. The results of the experimental studies are listed in Table 3. The results presented are average values obtained by testing at least six samples. As can be seen, microwave heating alone resulted in almost 48% (47–24%) reduction in the mortar content when RCA samples were pre-saturated. Such a reduction in the mortar content led to almost 33% (4.2–2.8%) decrease in the water absorption as well as 3.8% (2370–2460 kg/m3) increase in the particle density of RCA. Results also showed that microwave heating of air-dried RCA particles led to, on average, 32% reduction in the mortar content, 19% reduction in the water absorption and 2.5% increase in the particle density of the RCA samples tested (Table 3). Comparison between the results obtained for the two different moisture conditions considered showed that an increase in the water content of RCA can significantly increase the heating rate and thereby the differential thermal stresses generated, leading to faster and more efficient removal of the mortar. In addition, results showed that using a mechanical rubbing stage after microwave heating can significantly improve the efficacy of the microwave-assisted RCA beneficiation by removing the remaining weakened adhering mortar as well as breaking up the microwave treated mortar lumps for sieving. An almost 85% (from 47% to 7%) reduction in the mortar content, 76% (4.2–1.1%) reduction in the water absorption and 7.6% (2370–2550 kg/m3) increase in the particle density of RCA were achieved using the combined microwave heating and mechanical rubbing, rendering the RCA properties significantly closer to that of natural aggregates. However, incorporating an additional rubbing stage can significantly increase the operation cost. The surface temperature of the RCA particles captured using an infrared camera after microwave heating is shown in Fig. 9. As can be seen, the maximum surface temperature of the RCA particles reached about 140 °C which is considerably lower than the 300– 500 °C required for removal of adhering mortar using conventional heating (thermal beneficiation) method. According to Homand and Houper, heating at high temperatures (>300 °C) can adversely affect the mechanical properties of granite. For example, the pressure resistance decreases by 16% at 400 °C and by 44% at 600 °C [32]. Hence, unlike the conventional heating beneficiation, microwave-assisted beneficiation does not seem to degrade the quality of NA.

The temperature distribution, temperature gradient, and thermal stresses developed in a Type I RCA particle (Fig. 7) for saturated and air-dried conditions are illustrated in Figs. 10 and 11, respectively. The tensile and compressive stresses are illustrated by positive and negative signs, respectively. As can be seen, as a result of the differences in the electromagnetic, thermal and mechanical properties of granite and the cementitious adhering mortar, the temperature gradient and thus thermal stresses increase significantly at the interfacial zone. Moreover, comparing the temperature rise and stresses developed for saturated and air-dried moisture conditions confirms that saturating the adhering cementitious mortar can significantly increase the efficiency and speed of the RCA beneficiation process. As can be seen in Fig. 10, for saturated moisture condition, after a considerably short microwave heating duration, the normal tensile stresses at the ITZ can easily exceed the typical bond strength between the adhering mortar and aggregate particle in normal concretes. Taking into account, the higher water content at the ITZ due to its higher porosity and water absorption (especially when bleeding is possible), in practice the temperature gradient at the ITZ may be even higher. Moreover, considering the significantly higher tensile strength of natural granite compared to the cementitious mortar, the stresses developed in the granite aggregate particles are not expected to cause any damage to its integrity unless fissures are present originally. The microwave beneficiation system modeled in this study uses a microwave horn to direct microwave power from the top of the microwave chamber towards the incident surface of the RCA particles. Hence, as can be seen in Figs. 10 and 11, only the mortar adhering to the top exposed surface of the RCA particles is substantially heated and stressed. In practice, stirrers should be used in the RCA beneficiation chamber to ensure a uniform microwave power distribution so that the RCA particles can be heated uniformly from all directions. Using a stirrer in microwave beneficiation systems may lead to a more uniform delamination of the mortar adhering to the RCA particles and thus higher quality RCA. 6.3. Comparison between the properties of MRCA and the RCA treated using other beneficiation techniques Table 3 compares the remaining mortar content, water absorption and particle density of RCA processed using the microwaveassisted RCA beneficiation technique to that obtained using other RCA beneficiation methods discussed. As can be seen, among the single-stage processing methods, microwave heating resulted in significantly higher improvements in the properties of RCA. Comparison between the results reported in Table 3 confirmed that as expected, there is an almost linear relationship between the mortar content and the water absorption and particle density of the RCA samples. As can be seen a decrease in the amount of the mortar led to an almost linear decrease in the water absorption

3476

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

Fig. 8. Surface of a RCA particle before and after microwave heating.

Table 3 Properties of RCA before and after treatment using various beneficiation techniques. Beneficiation process

Before beneficiation Single-stage processes Microwave heating Conventional heating Mechanical rubbing Acid soaking

Process duration (h)

24-h Water absorption (%)

Particle density (OD) (kg/ m3)

Mortar content (%) by mass

4.2

2370

47

0.02 0.02 2 2 0.1 24 120 24 120 24 120

2.8 3.4 4.1 3.8 3.5 4.1 4.1 3.9 3.4 3.5 1.6

2460 2430 2380 2390 2410 2380 2380 2390 2420 2410 2500

24 32 44 41 34 45 45 41 33 34 13

300 °C

2.1

3.3

2430

31

500 °C Pre-saturated RCA

2.1 0.12

2.1 1.1

2480 2550

21 7

Pre-saturated RCA Air-dried RCA 300 °C 500 °C 0.1 M sulfuric acid 0.5 M sulfuric acid 1 M sulfuric acid

Combined processes Conventional heating and mechanical rubbing Microwave heating and mechanical rubbing

Properties of RCA

Fig. 9. Surface temperature of RCA particles after 1 min microwave heating.

and increase in the particle density of RCA. Hence, in the following discussion, mortar content will be used as the basis for comparison between various RCA beneficiation techniques. As can be seen in Table 3, while a single run of microwave heating resulted in almost 48% reduction in the mortar content of RCA, conventional heating at 300 °C and 500 °C reduced the mortar con-

tent by only about 6.4% (from 47% to 44%) and 12.8% (from 47% to 41%), respectively. This is mainly because microwave heating leads to a faster and more concentrated heating of mortar, resulting in higher differential thermal stresses in mortar (especially at the ITZ), whereas conventional gradually heats up the entire RCA particle, granite and mortar, to the same temperature. In is noteworthy that compared to conventional heating, microwave heating is significantly more energy efficient because it volumetrically heats up only the RCA exposed to it, not the entire heating chamber together with its contents. The efficiency of microwave heating in the current state of art systems reaches as high as 80–90% [29]. The significantly shorter duration required for microwave heating (1 min) compared to that required for conventional heating (120 min) leads to significantly lower energy consumption of the microwave-assisted RCA beneficiation process compared to the conventional heating beneficiation. Moreover, due to the shorter heating duration and the lower temperatures reached in microwave-assisted processing compared to that of conventional heating, the quality of the original granite is more likely to remain unaffected after processing. A preliminary cost analysis conducted as a part of this study by considering the utility tariffs (electricity and water), labor cost and machinery maintenance costs in Singapore showed that this method may add about 15% to the final price of the RCA produced. However, considering

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

Fig. 10. (a) Temperature, (b) temperature gradient, (c) normal stress, and (d) tangential stress in a RCA particle comprising saturated adhering mortar and airdried granite core when subjected to microwaves of 2.45 GHz frequency and 10 kW power.

such an increase, the final price of RCA would be still lower than the price of natural coarse aggregate (in Singapore), while offering competitive qualities. The results presented in Table 3 also revealed that soaking the RCA particles in a 0.1 M sulfuric acid solution as proposed by Tam et al. [26] reduced the mortar content by only about 4%. Significantly higher improvements in the RCA properties (up to 72% reduction in the mortar content) were achieved using the higher acid concentrations (0.5 and 1 M) and longer soaking durations. However, such high acid concentrations are significantly more hazardous when used on an industrial scale. Moreover, the use of concentrated sulfuric acid may considerably increase the sulfate content of RCA after beneficiation which in turn would give rise to durability concerns. Moreover, acid soaking is too time consuming, taking at least 24 h overall. Hence, acid soaking at high acid concentrations may only be considered as an efficient testing method to measure the mortar content of small samples of RCA in the laboratory. Compared to conventional heating and acid soaking methods, slightly more promising improvements were achieved using the mechanical rubbing technique. The results showed that 100 revolutions of mechanical rubbing using the Los Angles abrasion testing machine filled with 10 steel balls reduced the mortar content of RCA by almost 28% (from 47% to 34%). Intuitively, the removal of mortar is likely to be proportional to the number of steel balls used, the volume of the material rubbed and number of rotations of the drum. A Tradeoff between the quality of the RCA, the processing time and the energy consumption is necessary. When compared to microwave heating, mechanical rubbing requires significantly longer processing time and energy to achieve a similar yield and quality and thus may not be an economical option in

3477

Fig. 11. (a) Temperature, (b) temperature gradient, (c) normal stress, and (d) tangential stress in an air-dried RCA particle subjected to microwaves of 2.45 GHz frequency and 10 kW power.

practice. Furthermore, mechanical rubbing is the noisiest technique among all the beneficiation methods discussed. In addition, results showed that combined ‘‘conventional heating and mechanical rubbing’’ technique resulted in almost 34% (from 47% to 31%) and 55% (from 47% to 21%) reduction in the mortar content of RCA when heating temperatures of 300 °C and 500 °C were used, respectively. These are more than four times better than that achieved using conventional heating alone, suggesting that there was weakening of the mortar after heating. However, the reduction in the mortar content achieved using the combined ‘‘conventional heating and rubbing’’ were about 30% less than that achieved using the combined ‘‘microwave heating and rubbing’’ technique, confirming that microwave heating is more efficient. 6.4. Properties of the concrete incorporating microwave-treated RCA and un-treated RCA 6.4.1. Compressive strength The 28 day compressive strength of the recycled aggregate concrete (RAC) incorporating various amounts of MRCA and that of the RAC incorporating various percentages of original un-treated RCA are compared in Fig. 12. As can be seen, compared to concrete with un-treated RCA, incorporating MRCA led to significantly smaller reduction in the compressive strength of concrete. Incorporating 100% coarse MRCA resulted in only 10% reduction in the compressive strength as compared to almost 30% reduction when a similar amount of un-treated RCA was used. Moreover, incorporating up to 40% MRCA seemed to have led to negligible reduction in the compressive strength of concrete. Similar to when un-treated RCA were used, the compressive strength seemed to decrease almost linearly with an increase in the amount of MRCA used. Besides the decrease in the mortar content of the RCA used, breaking up of the

3478

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479

Fig. 12. Effect of RCA/MRCA replacement (% by weight) on 28 day compressive strength of concrete.

Fig. 14. Effect of RCA/MRCA replacement (% by weight) on the modulus of elasticity of concrete.

The results obtained for the modulus of elasticity of the concretes incorporating various amounts of RCA and MRCA are presented in Fig. 14. As shown, the reduction in the modulus of elasticity of concrete made with MRCA almost resemble the reduction in compressive strength, being limited to only 10% when the coarse natural aggregates were completely replaced with MRCA. This may be compared with the up to 25% reduction in the modulus of elasticity when natural aggregates were replaced with un-treated RCA. 7. Conclusions

Fig. 13. Effect of RCA/MRCA replacement (% by weight) on 28 day flexural strength of concrete.

weakened RCA particles as a result of the stresses developed during microwave heating may be also be one cause of the significant improvements observed in the compressive strength of RAC cast using MRCA. 6.4.2. Flexural strength The results of the flexural strength tests conducted are shown in Fig. 13. As can be seen, for the concrete incorporating MRCA, the trend in flexural strength reduction is rather similar to that observed for compressive strength. However, the flexural strength seemed to be less affected compared to compressive strength and modulus of elasticity when untreated RCA was used. The 100% replacement of coarse NA with RCA led to only 15% reduction in modulus of rupture as compared to almost 30% reduction in compressive strength. This may be due to the fact that the flexural strength is highly dependent on the bond strength between the aggregate and matrix. The higher water absorption capacity of the adhering mortar in RCA may enhance the bond between the new mortar and RCA which in turn can partially compensate for the adverse effect of the weakness of old ITZ (between the adhering mortar and embedded NA in RCA) on the overall flexural strength. 6.4.3. Modulus of elasticity The modulus of elasticity of concrete varies significantly with the stiffness of the coarse aggregates. Mortar is usually weaker compared to natural aggregates, lowering the stiffness of RCA.

The feasibility of using microwave heating to improve the quality of RCA through reducing the amount of the mortar adhering to the RCA particles as well as the mortar present as entirely mortar particles was experimentally and numerically investigated in this paper. Results confirmed that microwave heating may be effectively used to partially remove the cementitious mortar through developing high temperature gradients and thus high thermal stresses within the mortar, especially at the interfacial zone with the natural aggregates. Furthermore results showed that saturating the RCA particles prior to exposure to microwaves can significantly increase the yield and quality of the RCA produced. Results showed that unlike conventional heating beneficiation, in the microwave-assisted beneficiation method, RCA particles are heated to a much lower temperature and for a significantly shorter duration which should eliminate concerns about the possible degradation of the aggregate particle during processing. Shorter processing time compared to acid soaking and combined chemical–mechanical beneficiation methods and less energy consumption in comparison with the conventional heating and mechanical rubbing methods are among the advantages of microwave-assisted beneficiation. Moreover, unlike the acid pre-soaking method, microwave-assisted beneficiation method does not seem to have potential durability concerns. In addition, results of this study showed that microwave-assisted treatment can significantly enhance the mechanical properties of concrete produced with RCA including the compressive and flexural strengths and modulus of elasticity. Incorporation of up to 40% of microwave-treated RCA seems to have negligible effect on the mechanical properties of concrete. References [1] De Juan MS, Gutierrez PA. Study on the influence of attached mortar content on the properties of recycled concrete aggregate. Construct Build Mater 2009;23(2):872–7. [2] Hansen TC, Boegh E. Elasticity and drying shrinkage of recycled aggregate concrete. ACI Mater 1985;82(5):648–52.

A. Akbarnezhad et al. / Construction and Building Materials 25 (2011) 3469–3479 [3] Ravindrajah RS, Loo YH, Tam CT. Strength evaluation of recycled aggregate concrete by in-situ tests. Mater Struct/Mater Construct 1988;21(124): 289–95. [4] Poon CS, Kou SC, Lam L. Use of recycled aggregates in molded concrete bricks and blocks. Construct Build Mater 2002;16(5):281–9. [5] Tabsh SW, Abdelfatah AS. Influence of recycled concrete aggregates on strength properties of concrete. Construct Build Mater 2009;23(2):1163–7. [6] Tam WY, Vivian Gao XF, Tam CM. Micro-structural analysis of recycled aggregate concrete produced from two-stage mixing approach. Cement Concr Res 2005;35(6):1195–203. [7] Tavakoli M, Soroushian P. Strengths of recycled aggregate concrete made using field-demolished concrete as aggregate. ACI Mater J 1996;93(2):182–90. [8] Kou S, Poon CS. Compressive strength, pore size distribution and chloride-ion penetration of recycled aggregate concrete incorporating class-F fly ash, 2006. J Wuhan Univ Technol, Mater Sci Ed 2006;21(4):130–6. [9] Rakshvir M, Barai SV. Studies on recycled aggregates-based concrete. Waste Manage Res 2006;24(3):225–33. [10] Won MC. Use of crushed concretes aggregate for pavement concrete. Research Section, Construction Division, Texas Department of Transportation, PhD Thesis; 1999. [11] Xiao JZ, Li JB, Zhang CH. On relationships between the mechanical properties of recycled aggregate concrete: an overview. Mater Struct/Mater Construct 2006;39(290):655–64. [12] Tam WY, Vivian, Tam CM. Diversifying two-stage mixing approach (TSMA) for recycled aggregate concrete: TSMAs and TSMAsc. Construct Build Mater 2008;22(10):2067–77. [13] Otsuki Nobuaki, Miyazato Shin-Ichi, Yodsudjai Wanchai. Influence of recycled aggregate on interfacial transition zone, strength, chloride penetration and carbonation of concrete. J Mater Civil Eng 2003;15(5):443–51. [14] Khatib JM. Properties of concrete incorporating fine recycled aggregate. Cement Concr Res 2005;35(4):763–9. [15] Fraaij ALA, Pietersen HS, de Vries J. Performance of concrete with recycled aggregate. In: International congress challenges of concrete construction; 5– 11 September, 2002 Dundee. p. 187–98. ISBN 07277 31777 [volume: sustainable concrete construction]. [16] Fleischer W, Ruby M. Recycled aggregates from old concrete highway pavements. In: Proceedings of international symposium-sustainable construction: use of recycled concrete aggregate, London; November, 1998. [17] Li X. Recycling and reuse of waste concrete in China. Part I: material behaviour of recycled aggregate concrete. Resour Conserv Recycl 2008;53(1–2):36–44. [18] Grubl P, Nealen A. Construction of an office building using concrete made from recycled demolition material. In: Proceedings of international symposiumsustainable construction: use of recycled concrete aggregate, London; 1998.

3479

[19] Haase R, Dahm J. Baustoffkreislauf am besonderen Beispiel von Beton in norddeutschen Raum (Material recycling with specific example of concrete recycling in North Germany. Zeitschrift Beton (Concrete Journal); 1998. [20] Barra M. Estudio de la durabilidad del hormigon de arido reciclado en su aplicacin´ como hormigon estructural (Study on recycled concrete durability for structural concrete), in Civil Engineering department, Universidad Politécnica de Catalua; 1996. [21] Shima Hirokazu, Tateyashiki Hisashi, Matsuhashi Ryuji, Yoshida Yoshikuni. An advanced concrete recycling technology and its applicability assessment through input-output analysis. J Adv Concr Technol 2005;3(1):53–67. [22] Yonezawa T, Kamiyama Y, Yanagibashi K, et al. A study on a technology for producing high quality recycled coarse aggregate. Zairyo/J Soc Mater Sci Jpn 2001;50(8):835–42. [23] Yoda K, Harada M, Sakuramoto F. Field application and advantage of concrete recycled in-situ recycling systems. London, E14 4JD, United Kingdom: Thomas Telford Services Ltd; 2003. [24] Shima H, Tateyashiki H, Nakato T, Okamoto M, Asano T, et al. New technology for recovering high quality aggregate from demolished concrete. In: Proceedings of fifth international symposium on East Asia recycling technology; 1999. [25] Japan Concrete Institute (JCI) Standards. Recycled aggregate for concrete-class H (JIS A 5021:2005); 2005. [26] Tam VWY, Tam CM, Le KN. Removal of cement mortar remains from recycled aggregate using pre-soaking approaches. Resour Conserv Recycl 2007;50(1): 82–101. [27] Abbas Abdelgadir, Fathifazl Gholamreza, Burkan Isgor O, et al. Proposed method for determining the residual mortar content of recycled concrete aggregates. J ASTM Int. 2008. p. JAI101087. [28] Rhim HC, Buyukozturk O. Electromagnetic properties of concrete at microwave frequency range. J ACI Mater 1998;95:262–71. [29] Metaxas AC, Meredith RJ, Institution of Electrical Engineers. Industrial microwave heating. IEE power engineering series. London: Peter Peregrinus on behalf of the Institution of Electrical Engineers; 1983. [30] Akbarnezhad A, Ong KCG. Microwave decontamination of concrete. Mag Concr Res 2010;62(12):879–85. [31] Goangseup Zi, Bazˇant ZP. Decontamination of radionuclides from concrete by microwave heating. II: computations. J Eng Mech 2003;129(7):785–92. [32] Homand-Etienne, Houper R. Thermally induced microcracking in granites: characterization and analysis. Int J Rock Mech Min Sci Geomech Abst 1989;26(2):125–34.