The effects of recycled clay brick content on the engineering properties, weathering durability, and resilient modulus of recycled concrete aggregate

Accepted Manuscript The Effects of Recycled Clay Brick Content on the Engineering Properties, Weathering Durability, and Resilient Modulus of Recycled Concrete Aggregate Mababa Diagne, James M. Tinjum, Kongrat Nokkaew PII: DOI: Reference:

S2214-3912(14)00051-8 http://dx.doi.org/10.1016/j.trgeo.2014.12.003 TRGEO 36

To appear in: Received Date: Revised Date: Accepted Date:

13 August 2014 23 December 2014 27 December 2014

Please cite this article as: M. Diagne, J.M. Tinjum, K. Nokkaew, The Effects of Recycled Clay Brick Content on the Engineering Properties, Weathering Durability, and Resilient Modulus of Recycled Concrete Aggregate, (2015), doi: http://dx.doi.org/10.1016/j.trgeo.2014.12.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The Effects of Recycled Clay Brick Content on the Engineering Properties, Weathering Durability, and Resilient Modulus of Recycled Concrete Aggregate

Mababa Diagne1,2, James M Tinjum2, Kongrat Nokkaew2

1. Corresponding Author Institut des Sciences de la Terre ( ) Faculté des Sciences et Techniques Université Cheikh Anta DIOP, Dakar, SENEGAL BP 5396 Dakar-Fann, SENEGAL E-mail: [email protected] Phone: +221 33 825 25 30, Fax: +221 33 824 63 18 2. Department of Civil and Environmental Engineering, University of Wisconsin-Madison 2214 Engineering Hall 1415 Engineering Drive Madison, Wisconsin 53706-1607 USA Phone: +1 (608) 262-0785, Fax: +1 (608) 262-5199 E-mail: [email protected] E-mail: [email protected]

Abstract This paper presents a laboratory investigation of Recycled Clay Bricks (RCB) from demolished building when mixed with Recycled Concrete Aggregates (RCA) as an unbound base course in road construction. The hydraulic properties (saturated and unsaturated) of 100% RCB, 30%RCB, 15%RCB, 5% RCB and 100% RCA were determined at the optimum water content and 95% of dry density using rigid-wall hydraulic conductivity test and multi-step outflow respectively. The drying path of Soil-Water Characteristic Curves (SWCCs) was measured using hanging column tests. They are used for the estimation of unsaturated hydraulic conductivities and pores size distribution (pores diameter occupied by free water). Results show that the drainage of water is faster when the percentage of RCB increases due to larger pores. The effects of weathering (wet and dry cycles W-D, freeze and thaw cycles F-T) and abrasion (Micro-Deval and Los Angeles tests) were performed to see the stiffness changes with the number of cycles and the effects of abrasion on particle degradation of the specimens. The Summary Resilient Moduli (SRM) were calculated using the Power function model developed by Moosazedh and Witczak and the National Cooperative Highway Research Program (NCHRP) model under 1-28A procedure. The trend shows that the SRM and the constrained modulus decrease when the percentage of brick and the number of F-T cycles increase. Micro-Deval (MDE) and Los Angeles (LA) coefficients increase with the percentage of RCB on the specimens because they have low particle density and great porosity compared to natural aggregates. The W-D cycle results show that the durability of the material is affected by the number of W-D cycles and an increase in the percentage of fine particle is observed. The optimum moisture content of the specimens combined with the variation of temperature may influence the characteristics of the unbound road base pavement made out of recycled aggregates by the action of wet-dry or freeze-thaw cycling. The durability of a pavement can be influenced by repeated freeze-thaw cycles, wet-dry cycles, or a combination of both. During the thawing period, the melting of ice on the specimens lead to an unsaturated material, inducing an important bearing capacity loss.

Keywords: Recycled Clay Bricks, Recycled Concrete Aggregates, Durability, Resilient Modulus, Constrained Modulus, Base Course

The Effects of Recycled Clay Brick Content on the Engineering Properties, Weathering Durability, and Resilient Modulus of Recycled Concrete Aggregate

Mababa Diagne1,2, James M Tinjum2, Kongrat Nokkaew2

3. Institut des Sciences de la Terre (Corresponding Author) Faculté des Sciences et Techniques Université Cheikh Anta DIOP, Dakar, SENEGAL BP 5396 Dakar-Fann, SENEGAL E-mail: [email protected] Phone: +221 33 825 25 30, Fax: +221 33 824 63 18

4. Department of Civil and Environmental Engineering, University of Wisconsin-Madison 2214 Engineering Hall 1415 Engineering Drive Madison, Wisconsin 53706-1607 USA Phone: +1 (608) 262-0785, Fax: +1 (608) 262-5199 E-mail: [email protected]

E-mail: [email protected]

1. Introduction The recycling of demolition wastes is increasingly important because products coming from the demolition of buildings, roads, and other structures can be an alternative material for utilization in concrete production (Poon. 2007), in paving blocks (Poon and Chan 2006), and in road construction (Figueroa 1987). For the latter, these materials are often used if blended with recycled concrete aggregates (RCA) for pavement sub-base and base layer for sidewalks and low traffic roads. Many studies have been conducted to evaluate the suitability of using recycled materials like recycled clay brick (Arulrajah, 2011), recycled concrete aggregates, reclaimed asphalt pavement (Rahardjo et al. 2102, Rahman et al. 2014), and recycled crushed glass (Ali and Arulrajah, 2012) in pavement applications. Disfani et al. (2012) provided excellent evidence for using natural aggregates with recycled glass in a wide range of road construction including fill material in structural and non-structural applications and road pavement. Rahman et al. (2014) stated that recycled concrete aggregate and crushed brick were found to have the physical, geotechnical and chemical properties recommended by the Australian environmental protection authorities for backfill applications, whereas reclaimed asphalt pavement material did not meet some specified requirements. In practice, the use of RCA in road construction is more documented than the use of recycled clay bricks (RCB). Brick aggregate is different from RCA because of differences in the raw materials used and the manufacturing process of clay bricks. Arulrajah et al. (2011) showed that blend with other recycled aggregates could further enhance the properties of the crushed brick and improve performance in pavement applications. For Cameron et al. (2012), a high substitution of aggregate in unbound sub-base by 25% and 50% of recycled clay masonry reduced the maximum dry density and increased the optimum moisture content compared to RCA. RCB aggregates have a low particle density compared to natural aggregates. This low density is associated with a great porosity. Concrete made out of RCB aggregates has a lower density, a higher water absorption, a higher soundness mass loss, and a higher content of foreign material compared to crushed aggregates. The higher water absorption content of RCB combined with the seasonal variation of moisture and temperature may influence the characteristics of the unbound road base pavement made out of RCB or of material containing RCB by the actions of wet and dry (W-D) or freeze and thaw (F-T) cycling and repeated loading (represented by action of standard wheels loading on the road). During the thawing period, the melting of snow and

pavement ice causes materials saturation, which causes in turn an important bearing capacity loss. The structural integrity of an unbound granular base should be maintained throughout annual environmental cycling. Even if unsaturated, freezing of unbound granular material causes a net volume gain and an increase in water content, which affects pavement performance during the spring (Bilodeau 2009, Bilodeau et al. 2011). The purpose of this research paper is to evaluate hydraulic properties of recycled materials by measuring the soil-water characteristic curves (SWCC), the effects of weathering (W-D, F-T cycles) and the degradation on RCB aggregates when mixed with RCA as an unbound base course in road construction. The saturated hydraulic conductivity was conducted using rigid-wall compaction-mold permeameter (ATSM D5856), and the unsaturated hydraulic conductivity was performed using multi-step outflow (MSO). The constrained (D) and resilient (Mr) moduli were determined for each specimen at the end of 0, 5, 10 and 20 cycles of freeze and thaw. The relations between physical properties, type of material and unsaturated hydraulic parameters are developed. 2

Background

Some studies have dealt with the use of RCB mixed with RCA as an unbound road base. Cameron et al. (2012) reported the use of 20% RCB with 80% RCA by mass for the construction of an unbound granular pavement which respected the Australian technical specifications. Poon and Chan (2006) reported the possibility of using RCA and RCB aggregates in unbound sub-base materials because the use of crushed brick only did not meet all the technical specifications, such as the Los Angeles (LA) coefficient. This high value of LA found on RCB can be linked to the origin of the material (manufacturing process). The replacement of 20% of RCA by RCB increased the LA value by 6 (Cameron et al. 2012). The use of RCB as sub-base does not appear to pose a crucial problem due to less restrictive mechanical characteristics. The allowable amount (% by dry mass) of RCB in RCA for road construction is variable. Grab et al. (2012) reported that some European specifications allow a maximum of 30% of recycled clay masonry blended with RCA. In South Australia, the Department for Transport, Energy and Infrastructure (DTEI, 2001) allows 20%. In the USA, according to the Greenbook Specification for Construction, the amount of RCB cannot exceed 3% by weight (Greenbook, 2009). In the Wisconsin Department of Transportation (WisDOT) base course specifications, bricks can be used as “reprocessed material” in a combination of 80% of crushed concrete and asphaltic

pavement or surfacing and 20% of crushed stone or gravel, concrete block, brick, cinder, or slag. The percentage of brick weight in the mixture (20%) cannot exceed 5%. Percentage differences of RCB necessary for the road layers are due to the lack of standardized guidelines or performance-based specifications for RCB (Arulrajah et al. 2011). For Arulrajah et al. (2014) the compacted construction and demolition materials (recycled concrete aggregate, crushed brick, reclaimed asphalt pavement, waste excavation rock, fine recycled glass, and medium recycled glass have the potential to be used in pavement base/subbase applications as they have the required minimum effective friction angles. Moisture and temperature are the two environmental parameters that can significantly affect the resilient modulus of unbound materials. The water flow relations for unsaturated materials (base course) can be described by the soil-water characteristic curve (SWCC) and the fitting parameters of the Fredlund and Xing equation are part of the Mechanistic-Empirical Pavement Design Guide (MEPDG). The measurements of hydraulic conductivity in unsaturated soils are difficult to obtain, partly due to its variability in the field. This measurement is time-consuming. That is why the models of van Genuchten (1980) and Fredlund and Xing (1994), reviewed by Leong and Rahardjo (1997), have been developed for calculating the unsaturated hydraulic conductivity from more easily measured SWCC, using the saturated hydraulic conductivity as an initial value. The durability of a pavement can be influenced by repeated F-T, W-D cycles, or by a combination of both (Khoury et al. 2005). Previous studies reveal no standard laboratory procedure to evaluate the effect of W-D action on unbound materials. W-D cycles were usually performed on stabilized base courses with cement or flash. Khoury and Zaman (2007) showed that up to eight W-D cycles a reduction of Mr was observed. In this research paper, Mr was not tested due to the bad quality of the specimens at the end of W-D cycles. When the compacted unbound road base material is saturated and that water begins to freeze, ice crystals start to form and expand within the capillary voids. If the water continues to freeze, hydraulic pressure develops by the expansion of water in the transformation into ice. When exposed to negative temperatures, unbound granular material and soil may heave due to the interstitial water freezing and to the formation of segregated ice. The freezing process may be accompanied by the formation of ice lenses that create zones of greatly reduced strength in the pavement when thawing occurs (NCHRP, 2004). For the NCHRP (2004), the recommended ratio between Mr just

after thawing to Mr of unfrozen material for coarse material following these specifications: P4 < 50%, P200 < 6%, and PI < 12% (which is the case of our materials, AASTHOO: A-1-a) - termed reduction factor RF- should be 0.85 for the Design Guide approach. Arulrajah et al. (2011) found that the resilient modulus of recycled crushed bricks at 90% of OMC and at different stress stages (by varying deviator stress and confining pressure) varies between 125 to 300 kPa. Gabr and Cameron, 2012 showed that resilient moduli of both RCA and Virgin Aggregates products were significantly greater than the minimum requirement of 300 MPa when tested in accordance with the South Australian Department of Transportation. Cameron et al. (2012) reported that the resilient modulus decreased when the crushed masonry content increased. For these authors, it is also influenced by the matric suction. 3

Materials and Methods

3.1 Materials RCB and RCA used in this research paper were from a building demolition on University Avenue in Madison and Peters Concrete Company in Green Bay (Wisconsin State) respectively. The RCB was crushed to meet the requirements of the Wisconsin Department of Transportation (WisDOT) for dense graded base, and the maximum size was 19 mm. Figure 1 gives the grain size distribution according to ASTMD 6836 of all materials used in this research paper that fit the WisDOT size gradation requirements. Hydraulic conductivity tests were performed and SWCC was generated on RCA, RCB and RCB/RCA mixes (0, 5, 15 and 30%RCB) according to ASTM D5856. Freeze and Thaw (ASTM D6035 Modified), Wet and Dry (ASTM D559 Modified), Micro-Deval (ASTM D6928), Los Angeles Abrasion (ASTM C131) and Long Term Soaking tests were also performed to study the effects of weathering and durability on the degradation of the material. Constrained and Resilient moduli tests were performed on the specimens after 0, 5, 10 and 20 freeze and thaw cycles. The engineering properties of each mixture, RCA and RCB, are shown in Table I. 3.2 Methods 3.2.1 Resilient and Constrained Moduli

This study uses the Power function model proposed by Moosazedh and Witczak (1981) in Equation 1 and the NCHRP model (2004) in Equation 2. The resilient modulus is described as the ratio of applied deviator stress to recoverable or “resilient” strain.

M r = k 1 (θ ) k 2

(1)

Mr is resilient modulus, θ is bulk stress (θ = σ1 + σ2 + σ3), k1 and k2 are empirical fitting parameters and depend on the kind of material to be tested, σ1, σ2 and σ3 are the principal stresses applied on the specimen.

 θ − 3 k6 M r = k 1 pa   pa

  

k2

 τ oct   + k 7   pa 

k3

(2)

pa is the atmospheric pressure (101.35 kPa), k1, k2, k3, k6, and k7 are constants. τoct. is the octahedral shear stress: τ oct . = ((σ 1 − σ 2 ) 2 + (σ 1 − σ 3 ) 2 + (σ 2 − σ 3 ) 2 )1/ 2 = (

2 3) × σ d

and σd = σ1 −σ3

is the deviator stress. For base course, the Summary Resilient Modulus (SRM) corresponds to the Mr at a bulk stress of 208 kPa and τoct is the octahedral shear stress= 48.6 kPa, (NCHRP 1-28a). The Mr is performed on specimens subject to F-T cycles. Unbound granular materials, when exposed to freezing and thawing, may generally experience volume changes, loss in shear strength and sometimes alterations in their hydraulic conductivities. The weakening and deteriorating effects of freezethaw cycles on the compacted soil are confirmed in a study by Li et al. (2012). If water continues to freeze, hydraulic pressure will be developed by the expansion of water, which becomes ice. For the F-T cycles tests, specimens are compacted in plastic molds (PVC) of 6-in diameter and 12-in high at the optimum moisture content and 95% of their maximum dry unit weight. After compaction, samples were kept in their plastic molds, plastic-wrapped to conserve the optimum water content and placed in a freezer for one day (24h) at the temperature of -10 °F (-23 °C). After that the specimens are held at a room temperature (+21 °C for 23 hours) according to ASTM D560. The plastic molds were sealed carefully to prevent exposure to moisture during F-T cycling. After the designated number of F-T cycle (0, 5, 10 and 20), the specimens were extruded frozen and the constrained modulus performed. The constrained modulus, D, is calculated using Equation 3,where Vp is the compression wave velocity obtained by dividing the length of the

specimens (30.5 cm) by the corresponding travel time of wave through the specimen and ρ is the bulk density of the specimens.

D = ρV p

2

(3)

After determining the constrained modulus, the specimens are thawed inside the resilient modulus cell in order to perform the resilient modulus test. 3.2.2 Wet and Dry cycles The wet and dry cycles (W-D) for compacted soil-cement mixtures (ASTM D559) were modified to perform the same test on compacted unbound base material since there is no standard for that. The procedure used in previous works in UW-Madison was verified in advance for complete saturation and drying time of the specimens. The influence of W-D cycles is determined by the level of degradation of the specimens (grain size distribution and Micro-Deval tests). W-D cycles constitute environmental aggressions that can impact negatively on the behavior of the unbound base road. The moisture content in the pores of the base or sub-base pavement layer is a parameter that can influence the stiffness change with the number of W-D cycles. The properties of recycled aggregates greatly influence their performance as unbound granular pavement base layer. Complete saturation and drying of specimens occurred in 1 hour and 24 hours respectively, on the basis of a calibration using 100% RCA. Then, a path of 1hour saturation and 24-hour drying was applied to all samples subject to W-D and was considered as one cycle. Before soaking, samples were completely wrapped in a geotextile membrane to prevent the particle leakage and sealed between two aluminum plates by four (4) screws bolts to allow water flow. 0, 5 and 10 W-D cycles were tested and at the end of each of them, grain size distribution and Micro-Deval tests were performed. The variations of percentage of fines of each specimen were calculated. Long-term soaking tests for Micro-Deval were also performed. The samples (100% RCA, 5, 15, 30 and 100% of RCB) were washed properly and dried to eliminate the fine particles on the aggregates. 1,500 ± 5 g of each blends were placed in a small plastic bag entirely covered with small holes that allowed water flowing through the samples and then soaked in a container full of water for 104 days. 3.3.3 Soil Water Characteristic Curves - SWCC

The SWCCs, in coarse material, were determined by using hanging column test following the drying path according to ASTM D6836. The cell size is 7.6 cm in height and 38.5 cm in inner diameter. The suctions were measured using a ceramic pressure plate with air entry value of 100 kPa saturated with de-aired water for 48 hours. The specimens were compacted in three layers to reach 95% of the maximum dry unit weight by hand in order not to break the ceramic plate. The suctions were applied by putting out of level two tanks filled with water and the difference of level in the manometer, which gave the suction applied to the specimens, was measured. After compaction, the soils presented unsaturated conditions due to the presence, at the same time, of both air and water in the pores. Then, the specimens were saturated completely. When modeling unsaturated moisture flow beneath a road pavement, the hydraulic conductivity of the base course and subgrade materials, as a function of water content, must be known. This function can be estimated on the basis of the SWCC. It expresses the relation between volumetric water content (θ) or gravimetric water content (ω) and soil suction (ψ) in unsaturated soils. SWCC can determine the engineering behavior of unsaturated soil since it permits to have soil functions such as hydraulic conductivity and volume change. Matric suctions were determined using a hanging column test and a large scale (for details see Nokkaew et al. 2012) was used to reduce the scale effect of sample size on hydraulic properties. Two numerical models have been proposed to describe the SWCCs, which relate matric suction to volumetric water content. These two models were developed by Fredlund and Xing (1994) and van Genuchten (1980) in Equations (4) and (5) respectively.

θ = C (ψ ) ×



θs

[ln (e + (ψ / a ) )] bf

f

 θ − θr θ = θ r +  s n  1 + (αψ )

  

cf

where C (ψ ) = 1 −



ln (1 +ψ ψ r )   ln(1 + 106 /ψ r ) 

(4)

m

(5)

where θ is the water content corresponding to matric suction ψ; θs is the saturated volumetric water content; θr is the residual water content, af, α, bf, cf, m, n, ψr are mathematical fitting parameters, n is related to the pore size distribution and m gives an indication of the asymmetry of the curve( m = 1 − n −1 ). C(ψ) is an adjusting function used to force the volumetric water content to zero at 1 GPa.

3.3.4 Unsaturated hydraulic conductivity Measurements of unsaturated hydraulic conductivity were conducted using Multi-Step Outflow method (MSO) and interpreted according to equation 6 which is Gardner’s analytical solution (Gardner, 1956).  V − Vt ln  ∞  V∞

  π 2t   8   = ln  2  − Dθ  2  π    4L 

(6)

The unsaturated hydraulic conductivity kΘ was calculated using Equation 7. The water diffusivity Dθ is calculated from the slope of the linear relationship in Equation 6 (Breitmeyer and Benson, 2011).  ∆θ k Θ = γ w Dθ   ∆ψ

  

(7)

V∞ the cumulative outflow volume for a suction (ψ) in m3,Vt the cumulative outflow volume in m3 at elapsed time (s) from the initial application of ψ in kPa; L the thickness of the specimen in m, Dθ water diffusivity on the specimen in m2.s-1, ∆θ the variation in equilibrium of water content for a suction increment ∆ψ and γω, the unit weight of water. Based on the shape of the SWCC and the saturated hydraulic conductivity, it is possible to predict the unsaturated hydraulic conductivity function (Equation 8 and 9) using the van Genuchten fitting parameters defined in Equation 5 with the van Genuchten-Mualem model (van Genuchten 1980; Mualem 1976). Equations 8 and 9 give the same results.

(1 − (αψ ) × (1 + (αψ ) ) ) = (1 + (αψ ) )

n −m 2

n−1

kΘ

n m/2

(

× k sat .

)

2

k Θ = Θl 1 − (1 − Θ1 / m ) m × k sat.

(8)

(9)

Θ is the effective saturation, given by Equation 10.

[

Θ = 1 1 + (αψ ) n

]

m

(10)

n is the pore size distribution parameter, l is a parameter linked to the tortuosity of soil or pore connectivity term, ksat is the saturated hydraulic conductivity and kθ is the unsaturated hydraulic conductivity for the considered matric suction.

4

Results and discussion

4.1 Soil Water Characteristic Curves

Soil behavior under completely dry or completely saturated conditions is better understood than under unsaturated conditions because of matric suction caused by water surface tension on the curved pore air/pore water interface (Health et al. 2004). The suction at which water content starts to decrease significantly on the specimen is defined as the air entry value (ψa).

For higher values of suction the volumetric water content tends to be residual (θr). The air entry values range from 0.55 kPa (100% RCA) to 0.15 kPa (100% RCB). They decrease when the percentage of RCB in the blends increases due to its porosity. This trend is confirmed by the pore size distribution, which showed a smallest dominant pore size when the percentage of RCA increases. These results are similar to the results of Rahardjo et al. 2013. They found that the air entry values decrease when the materials are porous. The measured values of saturated permeability were 3.00, 5.88, 5.07, 4.51 and 1.48×10-5 m.s-1 respectively for 100%RCA, 30% RCB + 70%RCA, 15% RCB + 85%RCA, 5% RCB + 95%RCA and 100% RCB. The high sensitivity of RCB to water content changes, reflects the large change in suction when the percentage of RCB increases in the mixture. For this reason, Coronado et al. (2011) suggested to control the drainage conditions to ensure that water content will never exceed the project design value.

The slope desorption of all the specimens, which is linked to the fitting parameters, af and bf, has almost the same trend. Water drainage is faster when the percentage of RCB increases due to higher porosity and lower density. Specimen desorption is linked to saturated hydraulic conductivity (ksat.), the faster the desorption is, the higher ksat. is. The data obtained from the SWCC were fitted to Fredlund and Xing (1994), and van Genutchen (1980) models as shown in Equations (3) and (4). Figures 2.a and 2.b showed the best fitting curves compared to the experimental data. The fitting parameters of both models are summarized in Table 2. Good fitting curves were obtained.

4.2 Unsaturated hydraulic conductivity

For all the specimens, the graphics giving the unsaturated hydraulic conductivity kunsat as a function of the suction show that this parameter decreases faster for the lower suctions due to the rapid drainage of water from the specimens. For higher suction, kunsatdecreases slowly and tends to be minimal when the residual volumetric water content (θr) was reached. At this time, the free water in the pores is minimal. The predicted unsaturated hydraulic conductivity calculated with van Genuchten-Mualem model (Equation 8 or 9) gave good fitting compared to the measured values (Equation 7). Figure 3 gives good correlations between the unsaturated hydraulic conductivity measured and predicted for all the specimens tested with an R2 of 0.99. This model is good to estimate the unsaturated hydraulic conductivity of coarse materials like base course materials. The van Genuchten-Mualem SWCC model combined to unsaturated hydraulic conductivity kΘ shows that the pore size distribution parameter (n) increases with the percentage of RCB aggregates (Figure 4). This means the percentage of pores in the materials used increases when the percentage of RCB increases. RCB is more porous than RCA. Figure 4 shows how the size of the pores occupied by free water increases with the percentage of bricks (Figure 4.a). This volumetric percentage of pore is calculated using the effective degree of saturation (Equation 10) and the equivalent pore diameter D, if the pores are considered as cylindrical flow channels, is calculated using Kelvin’s equation (Equation 11).

D = 4Ts cosα /ψ

(11)

α = contact angle of the meniscus which is supposed to be 1, Ts = surface tension of water

(72.75×10-6 kN∕m at 20°C). The results show RCA has the smallest pore size, its pore size distribution parameter is 1.70 and RCB has the biggest one, which is equal to 2.21 (Figure 4.b). According to Noh et al. (2012), the pore size distribution parameter cannot be less than 1. 4.3 Wet and Dry Cycles results

Micro-Deval testing was performed on the materials to evaluate the effects of W-D cycles on abrasion resistance. Figure 5 shows that the MDE, which looks at the effects of moisture on the abrasion resistance of aggregates, increases with the number of W-D cycles but also with the percentage of RCB on the specimen. Results obtained are compared to Wisconsin limestone as a reference at 0 cycle. The results are shown in Figure 5.

The standard in Micro-Deval loss (ASTM D6928) for 19.0 mm maximum size aggregate found, for abrasion losses between 5% and 20%, the single operator coefficient of variation is 3.4%. No significant change in particle size distribution was observed. During the wetting path, fine particles passed through the geotextile membrane to a water container, this phenomenon prevents the real variation of fines after W-D cycles for the unbound materials, and the performing of Mr. The variation of the percentage of fines (∆f) after 5 cycles of W-D ranged from 1.30% to 3.66% for RCA and RCB. For 10 cycles, it is reduced from 0.82% to 1.71%. This reduction of ∆f is due to the long soaking period of the specimens, which allows the leaching of fine particles through water. But it is observed in all cases that the percentage of fines increases at the end of each cycle. The loss in the percentage of limestone at 0 cycle is equal to 21%; this value is almost the same as RCB at 5 cycles of W-D. Azam and Cameron (2012) found that the replacement of 20% of the RCA with recycled clay masonry increased the Los Angeles value by about 6. According to the MDE and to Los Angeles tests, a reduction of the durability of the specimen with an increase in W-D cycles was observed. 4.4 Resilient and Constrained Moduli results 4.4.1 Constrained Modulus

Seismic modulus testing is a non-destructive method monitoring the change of constrained modulus recycled material during each F-T cycle. Previous studies were found on cementstabilized materials. It can measure the mechanical property (constrained modulus) changes of the specimens during weathering exposure. The constrained modulus for 0 cycle ranged from 0.77 GPa for the 100% RCB to 1.82 GPa for 100% RCA. Its reduction after 5, 10, and 20 cycles ranged from 0% to 29%, 12% to 33%, and x to y respectively. The results are compiled in Figure 6, and they show a general reduction trend when the percentage of RCB increases. This trend can be attributed to the density of the specimens but also to the effect of weathering which reduces the durability of the unbound material. The mechanical behavior (durability) has a close correlation with the number of freeze-thaw cycles. Rosa (2006) reported that repetitive F-T cycling degrades the strength of soils through generating cracks with ice lens melting and inducing more water into the voids. Freeze and thaw cycles have a significantly weakening effect on the compacted granular materials. Equation 12 can be used to describe approximately the

relationship between the constrained modulus, the percentage of RCB, the duration of a cycle, and the number of F-T cycles.

Dx = D( x −i ) exp

 k ×T ×% RCB × N    1000  

(12)

x = 0, 5, 10 and 20 cycles of F-T, i equal to 0, 5, and 10 (x – i ≠ 15), k is calculated from the slope of the linear relationship giving the constrained modulus as a function of the percentage of RCB for each number of cycle applied, T is the time of one cycle, N is the number of cycles. The results obtained are compiled in Figure 6. The results show that the constrained modulus decreases when the number of F-T cycles increases but also when the percentage of RCB increases. It is important to take into account the effects of weathering (action of frizzing and water content) on the behavior of the pavement. Freeze and Thaw cycles lead to the generation of fine particles, increasing therefore the voids in the medium. 4.4.2 Resilient Modulus

The internal SRM was evaluated after the determination of the constrained modulus D. The same specimen was tested both for D and SRM. For more details on Resilient Modulus test procedure refer to Ba, 2013. According to the Mechanistic- Empirical Pavement Design Guide, the resilient modulus (Mr) is an important input parameter for calculating the response of pavements under traffic loading. The main advantage of this model is its consideration of the stress state (i.e., change of normal and shear stress) of the material during testing. For this study, the results show that the internal SRM decreases when the number of F-T cycles increases (Figure 7). The results found with the external SRM without 30% RCB show that the SRM tends to decrease when the number of F-T cycles and the percentage of RCB increase and they also show that the resilient modulus increases when moisture content decreases. Gabr and Cameron, 2012 obtained the same trend. The SRM based on internal LVDT measurements of deformation were found to

be higher than those based on external LVDT measurements of deformation for all the specimens. For all the specimens the internal SRM ranges between 374 and 106 MPa and the external SRM between 142 and 73 MPa. The results obtained in this study show that RCB only cannot be used as base course materials since its resilient modulus at 0 cycle is below the required value (300 MPa). RCB

needs to be mixed with RCA for base course. Bozyurt et al. (2012) found the same results. The

reduction of Mr with the number of F-T cycles can be linked to the action of F-T (weathering conditions), which breaks the granular material and reduces the durability of the specimen over time. Figure 7 also shows that the internal SRM decreases when the optimum water content increases. 5. Conclusion

This study was conducted to understand the properties of base course made out of recycled materials. It consisted in evaluating the hydraulic properties (saturated and unsaturated conductivities), the effects of weathering (Wet and Dry cycles, Freeze and Thaw cycles) and degradation (Los Angeles, Micro-Deval, Constrained and Resilient Modulus of recycled clay brick (0, 5, 15 and 30%) when mixed with recycled concrete aggregate (RCA) as an unbound base course in road construction. Results show that addition of RCB in the base course increases hydraulic conductivity. They also reveal that the weathering conditions (freeze and thaw, wet and dry) play an important role in the degradation of the specimen. From 0 to 20 cycles of F-T, the constrained modulus decreases to the point of reaching 53, 63, 63, and 70% from 100% RCA to 100% RCB. The resilient modulus decreases when the percentage of RCB increases. This fact can be linked to hydraulic conductivity and weathering conditions. 6. Acknowledgements

This research project was supported by the Office of Scholarship Program (OSP) of the Islamic Development Bank under its Merit Scholarship Program for High Technology (MSP) 2011, (Number: 66/SN/D33).

References Ali M.M.Y, Arulrajah A. Potential Use of Recycled Crushed Concrete-Recycled Crushed Glass Blends in Pavement Subbase Applications. GeoCongress. 2012; 3662-3671. Arulrajah A, Piratheepan J, Aatheesan T, Bo M. Geotechnical Properties of Recycled Crushed Brick in Pavement Applications. J. Mater. Civ. Eng. 2011; 10:1444-1452. Arulrajah A, Disfani MM, Horpibulsuk S, Suksiripattanapong C, Prongmanee N. Physical properties and shear strength responses of recycled construction and demolition materials in unbound pavement base/subbase applications. Constr. and Build. Mat. 2014; 58:245257. ASTM C131. Resistance to degradation of small size coarse aggregate by abrasion and impact in the Los Angeles machine; 2006. ASTM D559. Wetting and drying compacted soil-cement mixtures; 1996. ASTM D560. Freezing and Thawing Compacted Soil-Cement Mixtures; 2003. ASTM D5856, Measurement of hydraulic conductivity of porous material using a rigid-wall, compaction-mold permeameter; 2008. ASTM D6035. Standard test method for determining the effect of freeze-thaw on hydraulic conductivity of compacted or undisturbed soil specimens using a flexible wall permeameter; 2002.

ASTM D6928. Resistance of coarse aggregate to degradation by abrasion in the microDeval apparatus; 2010. ASTMD6836. Determination of the soil water characteristic curve for desorption using hanging column, pressure extractor, chilled mirror hygrometer, or centrifuge; 2008. Ba M, Nokkaew K, Fall M, Tinjum JM. Effect of Matric Suction on Resilient Modulus of Compacted Aggregate Base Courses. Geotech. Geol. Eng DOI 10.1007/s10706-013-9674-y; 2013. Bilodeau J-P, Dore G, Schwarz C. Effect of seasonal frost conditions on the permanent strain behavior of compacted unbound granular materials used as base course. Int. J. of Pavement Eng 2011; 5:507-518. Bilodeau J-P. Optimisation de la granulométrie des matériaux granulaires de fondation des chaussées, Dissertation PhD, Université Laval. Québec ; 2009. Bozyurt O, Tinjum JM, Son Y-H, Edil TB, Benson CH. Resilient Modulus of Recycled Asphalt Pavement and Recycled Concrete Aggregate. GeoCongress 2012; 3901-3910. Breitmeyer RJ, Benson CH. Measurement of unsaturated hydraulic properties of municipal solid waste. Geo-Frontiers 2011; 1433-1442.

Cameron DA, Azam AH, Rahman MM. Recycled clay masonry and recycled concrete aggregate blends in pavement. GeoCongress Oakland, CA 2012; 1532-1541. Department for Transport, Energy and Infrastructure (DTEI) TSA 2000/02428. Standard specification for supply and delivery of pavement materials, Government of South Australia, Revision 1, Adelaide, Australia; 2001. Coronado O, Caicedo B, Taibi S, Correia A.G, Fleureau J.M. A macro geomechanical approach to rank non-standard unbound granular materials for pavements. Engineering Geology. 2011, 119:64-73. Disfani MM, Arulrajah A, Bo MW, Sivakugan N. Environmental risks of using recycled crushed glass in road applications. Journal of Cleaner Production. 2012; 20:170-179. Figueroa JL, Zhou L, Chang WF. Use of phosphate mining waste in secondary road construction. Transp. Res. Rec. 1987; 1:59-64. Fredlund DG, Xing A. Equation for the soil-water characteristic curve. Can. Geotech. J. 1994; 4:521-532. Gardner WR. Calculation of capillary conductivity from pressure plate outflow data. Soil Sci. Soc. of Am. J. 1956; 3:317-320. Grab AG, Cameron DA. Properties of recycled concrete aggregate for unbound pavement construction. J. Materials in Civil Engineering 2012; 6:754-764. GREENBOOK, Construction Materials, Section 200-Rock Materials, In Greenbook standard specifications for public works construction; (2009). Heath A, Pestana J, Harvey J, Bejerano M. Normalizing behavior of unsaturated granular pavement materials. J. Geotech. Geoenviron. Eng. 2004; 9:896-904. Khoury N, Zaman M, Laguros J. Behavior of stabilized aggregate bases subjected to cyclic loading and wet-dry cycles. Adv. in Pavement Eng doi: 10.1061/40776 (155)5 2005; 1-10. Khoury N, Zaman MM. Durability of stabilized base courses subjected to wet-dry cycles. Int. J. of Pavement Eng. 2007; 4:265-276. Leong EC, Rahardjo H. Review of Soil-Water Characteristic Curve Equations. J. Geotech. Geoenviron, Eng 1997; 123:1106-1117. Li G, Ma W, Zhao S, Mao Y, Mu Y. Effect of freeze-thaw cycles on mechanical behavior of compacted fine-grained soil. Cold Reg. Eng doi: 10.1061/9780784412473.008 2012; 72-81. Moosazedh J, Witczak M. Prediction of Subgrade Moduli for Soil that Exhibits Nonlinear Behavior. J. of Transp. Res. Board 1981; 810:10-17. Mualem Y. A New Model for Predicting the Hydraulic Conductivity of Unsaturated Porous Media. Water Resour. Res. 1976; 3:513-522.

NCHRP. Laboratory determination of resilient modulus for flexible pavement design, NCHRP Research Results Digest; (2004). Noh JH, Lee SR, Park H. Prediction of Cryo-SWCC during freezing basedon pore-size distribution. Int. J. Geomech 2012; 12:428-438. Nokkaew K, Tinjum JM, Benson CH. Hydraulic properties of recycled asphalt pavement and recycled concrete aggregate. GeoCongress 2012; 1476-1485. Poon CS, Chan D. Feasible use of recycled concrete aggregates and crushed clay brick as unbound road sub-base. Constr. and Build. Mat. 2006; 20:578-585. Poon CS, Chan D. Paving blocks made with recycled concrete aggregate and crushed clay brick. Constr. and Build. Mat. 2006; 20:569-577. Rahardjo H, Satyanaga A, Leong E-C, Wang J-Y. Unsaturated properties of recycled concrete aggregate and reclaimed asphalt pavement. Engineering Geology. 2013; 161:44–54. Rahman M.A, Imteaz M, Arulrajah A, Disfani M.M. Suitability of recycled construction and demolition aggregates as alternative pipe backfilling materials, Journal of Cleaner Production. 2014; 66:75-84. Rosa MG. Effect of freeze and thaw cycling on soils stabilized using fly ash. MS Thesis, University of Wisconsin-Madison, Madison, WI; 2006. Van Genuchten M. A Closed-Form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. of Am. J. 1980; 5:892-898.

Figure captions

Figure 1. Grain size distribution of the materials used.

Figure 2. Soil-Water Characteristic Curves, a) Measured WCC data fitted to Fredlund and Xing (1994) model, b) Measured WCC data fitted to van Genuchten (1980) model.

Figure 3. Correlations between unsaturated hydraulic conductivity kunsat measured and predicted.

Figure 4. Variation of volumetric percentage of pores as a function pores diameter occupied by free water (a) and Pore size distribution parameter as a function of percentage of RCB (b).

Figure 5. Variation of the Micro-Deval coefficient of the aggregates at different Wet and Dry Cycles.

Figure 6. Variation of the Constrained Modulus at different Freeze and Thaw Cycles.

Figure 7. Variation of the Resilient Modulus at different Freeze and Thaw Cycles.

List of Figures

Figure 1

a)

b) Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

List of Tables Table 1 Engineering Properties of Aggregates Used.

MATERIALS

D max (mm)

Cu

Cc

Gs

Mortar (%)

Water Abs. (%)

LA (%)

ωopt. (%)

γd max (kN/m3 )

k× × 10-5 (m.s-1)

Grave l (%)

Sand (%)

100% RCB

19.1

25.83

2.69

2.25

21.50

5.12

36.8

8.6

18.7

14.8

56.0

38.2

5% RCB + 95% RCA

19.1

34.09

2.93

2.39

1.08∗

4.76

30.1

6.4

20.8

4.51

56.0

39.6

15% RCB + 85% RCA

19.1

48.00

3.95

2.36

3.23∗

4.86

31.9

6.8

20.6

5.07

52.3

42.2

30% RCB + 70% RCA

19.1

30.91

2.17

2.33

6.45∗

4.95

33.7

6.8

20.4

5.88

53.0

43.9

100% RCA

19.1

41.67

1.25

2.41

-

4.64

29.9

6.1

20.9

3.00

51.0

47.2

LA: Los Angeles percent loss, γd max: modified Proctor maximum dry density, ωopt: modified Proctor optimum water content, Gs: Specific Gravity, USCS: Unified Soil Classification System, AASHTO: American Association of State Highway and Transportation Officials, k: Saturated Hydraulic Conductivity, ∗ Theoretical mortar content calculated from the initial mortar content of the raw crushed bricks.

F

Table 2 SWCCs fitting parameters in Fredlund and Xing (1994) and Genuchten (1980) models. Materials ψa(kPa) θs(m3/m3) bf cf af(kPa) ψr(kPa) θr(m3/m3) θs α (kPa-1) n m

100% RCA

5% RCB

15% RCB

30% RCB

0.55 0.35 0.25 0.15 Fredlund and Xing (1994) fitting parameters 0.23 0.232 0.232 0.24 2.4 2.2 2.17 2.5 0.58 0.55 0.54 0.6 1.12 0.75 0.57 0.27 90 7000 20 20 Genuchten (1980) fitting parameters 0.0365 0.0568 0.0499 0.0438 0.230 0.232 0.232 0.240 0.77 1.30 1.76 3.35 1.71 1.75 1.76 1.93 0.41 0.43 0.41 0.48

100% RCB

0.15 0.24 2.7 0.63 0.25 130 0.0450 0.240 3.29 2.21 0.55