The Effect of Recycled Waste Glass as a Coarse Aggregate on the Properties of Portland Cement Concrete and Geopolymer Concrete

: A signiﬁcant amount of waste glass is generated in Australia and around the world, which requires sustainable recycling. The use of recycled glass as aggregates in concrete is one of the many options for recycling. This study investigated the characteristics of ordinary Portland cement (OPC) and geopolymer concretes containing different proportions of recycled glass as a partial replacement of natural coarse aggregate. It was found that the 28-day compressive and tensile strengths of OPC concrete decreased up to 21%, and 7% and of geopolymer concrete decreased by 11–26% and 11–29% with the increase in the recycled glass coarse aggregate. The porosity, sorptivity and chloride permeability of OPC and geopolymer concrete increased and the drying shrinkage decreased due to the use of the recycled glass coarse aggregate. The microstructural analysis revealed the porous interfacial transition zone (ITZ) between the glass coarse aggregate and the paste/mortar matrix led to a decrease in the strength and an increase in the porosity, sorptivity and chloride permeability of the concrete due to the increase in the glass coarse aggregate. However, the mechanical and durability properties of OPC and geopolymer concrete containing 10 to 20% glass coarse aggregate were comparable to the corresponding properties of the control concrete sample containing a natural coarse aggregate.


Introduction
A significant amount of natural aggregates and ordinary Portland cement are used in the production of concrete.The excessive use of stone chips as coarse aggregates in the production of concrete leads to a shortage of natural sources.In this regard, recycled glass aggregates could be a potential alternative to using natural coarse aggregates.Approx.130 million tons of glass is produced annually worldwide, and approx.100 million tons of glass is discarded as waste [1].Only 21% of waste glasses are recycled and the remaining waste glasses are landfilled [1].Since glass powder contains a notable proportion of amorphous silica [2,3], it is suitable for use as a binder in concrete.Several studies were conducted on the use of glass powder as a binder in the production of concrete [4][5][6].Furthermore, the physical properties of crushed glass are suitable for use as fine [7,8] and coarse aggregates [9,10].
Coarse aggregate is the one of key components of concrete, which comprises a high volume of concrete products.Therefore, the utilization of crushed waste glass as a coarse aggregate in concrete can significantly increase its usage, and thus, help save the land needed for its storage.Furthermore, the use of a glass coarse aggregate will reduce the use of natural coarse aggregate, saving natural resources.Topçu and Canbaz [11] reported that glass aggregate has no significant effect on the workability of concrete.They found that the Buildings 2023, 13, 586 2 of 20 Schmidt hardness, as well as the compressive, tensile and flexural strengths, were slightly reduced with the increase in the glass aggregate.Terro [12] investigated the influence of a glass aggregate on the properties of concrete at ambient and high temperatures.The author reported that, at high temperatures of up to 700 • C, the compressive strength of concrete using a glass aggregate decreased up to 20%.Serpa et al. [13] evaluated the mechanical properties of concrete where glass was used as coarse, fine and a combination of coarse and fine aggregate.The authors found that concrete with a glass coarse aggregate provided a higher performance than those with a glass fine aggregate and a combined glass fine and coarse aggregate.However, in all the cases, the mechanical performance of concrete was reduced due to the use of a glass aggregate.In contrary, Sangha et al. [14] reported that the compressive and tensile strengths of concrete increased up to 60% due to the use of a glass aggregate as a replacement for the control flint aggregate.Omoding et al. [15] found that concrete using a 100% glass coarse aggregate provided a comparable abrasion resistance to concrete using 100% crushed limestone.De Castro and De Brito [16] reported that the durability of concrete did not change significantly due to the use of a glass aggregate.Kou and Poon [17] investigated the characteristics of self-compacting concrete using recycled glass as a fine and coarse aggregate.They found that, though the mechanical properties and drying shrinkage decreased due to the increase in the glass aggregate, the resistance to penetration of the chloride ion increased due to the increase in the glass aggregate.Chelik, A.I. et al. [18] also studied the mechanical properties of OPC concrete containing a crushed glass aggregate as a partial replacement for the coarse aggregate and also reported a decrease in the compressive and tensile strengths with an increase in the crushed glass aggregate.
Furthermore, the production of OPC requires a significant amount of energy and it emits large quantities of carbon dioxide (CO 2 ).Around 0.85 tons of CO 2 is released during the production of 1 ton of OPC.In this regard, geopolymer is a potential alternative binder for OPC.The substitution of OPC with geopolymer binder could reduce CO 2 emissions by 80% or more [19].Similar to the case of OPC concrete, glass powder and cullet could be used as a precursor and aggregate in geopolymer concrete, respectively.Gholampour et al. [20] found that the mechanical characteristics of geopolymer mortars comprising a glass fine aggregate are similar to those comprising a natural fine aggregate.In contrary, Lu and Poon [21] reported that the compressive strength of geopolymer mortar decreased due to the increase in the glass fine aggregate.Khan and Sarker [7] found that geopolymer mortars using a glass fine aggregate did not show any detrimental alkali-silica reaction.Saccani et al. [22] reported that the freeze-thaw resistance and the diffusion of sulphate ions in geopolymer mortar did not change due to the use of a glass fine aggregate.
From the previous studies, it is obvious that waste glass has a great potential for use as a partial or full replacement for natural coarse aggregate in OPC and geopolymer concrete.Though the influence of waste glass as a coarse aggregate on some basic mechanical characteristics of OPC concrete has been studied, limited studies were conducted on the durability of concrete containing waste glass as a coarse aggregate.Furthermore, though several studies were conducted on the use of glass as a fine aggregate in geopolymers, few studies were conducted on the use of glass as a coarse aggregate in a geopolymer system.Therefore, the present study investigated the effect of a recycled glass coarse aggregate on the OPC concrete and geopolymer concrete considering the mechanical and durability-related properties.The characteristics of geopolymers are highly dependent on the curing conditions, which needs to be investigated in detail.Hence, this study also evaluated the properties of geopolymer concrete using waste glass as a coarse aggregate at ambient and heat curing environments.The characteristics of the OPC and geopolymer concrete using a glass coarse aggregate were assessed by investigating the workability, compressive strength, tensile strength, modulus of elasticity, porosity, sorptivity, chloride penetration, alkali-silica reaction and drying shrinkage.A scanning electron microscopy (SEM) analysis was performed to investigate the interface between glass aggregate and binder matrix.

Materials and Mix Proportions
Ordinary Portland cement was used as a binder for the production of OPC concrete, Class F fly ash was used as a precursor to produce heat-cured geopolymer concrete and both Class F fly ash and blast furnace slag were used as precursors to produce ambientcured geopolymer concrete.The chemical compositions of OPC, slag and fly ash were determined using X-ray fluorescence (XRF), as shown in Table 1.
Table 1.Chemical compositions of OPC and fly ash [23].The crushed granite and waste glass were used as the coarse aggregate.The waste glasses were collected from a local supplier in Perth.The physical appearance of the crushed granite and glass aggregates are shown in Figure 1.It can be seen that the surface of glass aggregate is smoother than the crushed granite.However, the shape of the glass aggregate is more angular than the crushed granite.The natural sand with a fineness modulus of 1.95 was used as the fine aggregate to produce the OPC and geopolymer concretes.The natural sand and coarse granite aggregates were placed in saturated and surface dry (SSD) conditions while the crushed glass aggregates were in dry conditions before being mixed into the concrete.

Constituents
uildings 2023, 13, x FOR PEER REVIEW 3 of 21 alkali-silica reaction and drying shrinkage.A scanning electron microscopy (SEM) analysis was performed to investigate the interface between glass aggregate and binder matrix.

Materials and Mix Proportions
Ordinary Portland cement was used as a binder for the production of OPC concrete, Class F fly ash was used as a precursor to produce heat-cured geopolymer concrete and both Class F fly ash and blast furnace slag were used as precursors to produce ambientcured geopolymer concrete.The chemical compositions of OPC, slag and fly ash were determined using X-ray fluorescence (XRF), as shown in Table 1.
Table 1.Chemical compositions of OPC and fly ash [23].The crushed granite and waste glass were used as the coarse aggregate.The waste glasses were collected from a local supplier in Perth.The physical appearance of the crushed granite and glass aggregates are shown in Figure 1.It can be seen that the surface of glass aggregate is smoother than the crushed granite.However, the shape of the glass aggregate is more angular than the crushed granite.The natural sand with a fineness modulus of 1.95 was used as the fine aggregate to produce the OPC and geopolymer concretes.The natural sand and coarse granite aggregates were placed in saturated and surface dry (SSD) conditions while the crushed glass aggregates were in dry conditions before being mixed into the concrete.A mixture of 8 M of sodium hydroxide and sodium silicate (Na 2 O = 14.70%,SiO 2 = 29.40% and H 2 O = 55.90%) was used as an alkaline activator solution to produce the geopolymer concretes.The 8 M molarity of the sodium hydroxide was selected based on the authors' previous research and since a higher molarity increases the cost of the geopolymer.

Constituents
The mixture proportions of the OPC and geopolymer concretes are shown in Table 2, where the natural coarse aggregates (crushed granite) were replaced with a 0, 10, 20, 30 and 40% recycled glass aggregate.The water-to-binder ratio of OPC concrete was 0.38.The amount of the alkaline activator solution for the geopolymer concrete was 40% of the total precursor, and the ratio of the sodium silicate to sodium hydroxide solution was 2.5.A Master Rheobuild 1000 NT superplasticizer was added to the OPC concrete to achieve the target slump in the range of 100-125 mm.No superplasticizer was used in the geopolymer concrete.The mix proportions of the heat-cured and ambient-cured geopolymer concretes were the same except that a 50% slag was used for a partial replacement of the Class F fly ash in the ambient-cured concrete since the presence of CaO helps set and harden the geopolymer concrete under ambient temperatures.

Sample Preparation and Testing
To create the OPC concrete, the saturated surface dry (SSD) fine and coarse granite aggregates and crushed glass aggregates were first dry mixed for 2 min.After that, the OPC was added to the concrete and mixing was performed again for approx. 2 min.The water and superplasticizer were then slowly added and mixing was continued until a uniform concrete was obtained.To prepare the geopolymer concrete, the fine and coarse aggregates and fly ash were dry mixed similar to the OPC concrete, and then the alkaline solution was added to obtain a uniform geopolymer concrete.The fresh mixtures were used for the slump test as a measure of the workability.The concrete mixture was then placed in the molds and compacted.The OPC concrete samples were demolded after one day of casting and then the samples were cured in lime water at 23 • C for 28 days.On the other hand, heat curing and ambient curing were applied to geopolymer concrete samples.For heat curing, after casting, the molds with the geopolymer sample were put in the oven at 70 • C for 24 h.After that, the hardened samples were demolded and stored at room temperature for further testing.For ambient curing, the geopolymer samples with the mold were stored at a controlled temperature of 20 ± 3 • C and relative humidity of 65 ± 5%.The samples were demolded after 24 h of casting and left under the same controls until testing.The typical mixing, casting, compaction and slump testing of the concrete in this study are illustrated in Figure 2.
As a measure of the workability, the slump test was conducted according to AS 1012.3.1 [24].The compressive strength and modulus of elasticity of the hardened samples were determined using 100 mm × 200 mm cylinder samples in accordance with AS 1012.9 [25] and ASTM C469 [26], respectively.The splitting tensile strength test was performed using 150 mm × 300 mm cylinder samples following AS 1012. 10 [27].The compressive strength of all the mixes was measured at 7, 28, 56 and 90 days, while all the other properties were measured at 28 days.The test setup for the compression, indirect tension and elastic modulus of concrete is shown in Figure 3.
the mold were stored at a controlled temperature of 20 ± 3 °C and relative humidity of 65 ± 5%.The samples were demolded after 24 h of casting and left under the same controls until testing.The typical mixing, casting, compaction and slump testing of the concrete in this study are illustrated in Figure 2.  As a measure of the workability, the slump test was conducted according to AS 1012.3.1 [24].The compressive strength and modulus of elasticity of the hardened samples were determined using 100 mm × 200 mm cylinder samples in accordance with AS 1012.9 [25] and ASTM C469 [26], respectively.The splitting tensile strength test was performed using 150 mm × 300 mm cylinder samples following AS 1012.10 [27].The compressive strength of all the mixes was measured at 7, 28, 56 and 90 days, while all the other properties were measured at 28 days.The test setup for the compression, indirect tension and elastic modulus of concrete is shown in Figure 3.The porosity of the concrete sample was determined using the volume of permeable voids (VPV) test following ASTM C642 [28].The sorptivity test was performed to determine the capillary suction of the concrete in accordance with ASTM C1585 [29].The rapid chloride permeability test (RCPT) was carried out following ASTM C1202 [30].The 50 mm thick discs were cut from 100 mm × 200 mm cylinder samples to conduct the VPV, sorptivity and RCPT test.Roughly, 75 mm × 75 mm × 285 mm concrete prism samples were used to determine the drying shrinkage and alkali-silica reaction (ASR) test as per AS 1012.13 [31] and ASTM C1293 [32], respectively.Figure 4 shows the typical test setup used to perform the RCPT and sorptivity tests.A scanning electron microscopy (SEM) analysis was performed to investigate the microstructure of the aggregates-matrix interface of the concrete specimens.The SEM analysis was conducted using the Tescan Mira3 microscope.The porosity of the concrete sample was determined using the volume of permeable voids (VPV) test following ASTM C642 [28].The sorptivity test was performed to determine the capillary suction of the concrete in accordance with ASTM C1585 [29].The rapid chloride permeability test (RCPT) was carried out following ASTM C1202 [30].The 50 mm thick discs were cut from 100 mm × 200 mm cylinder samples to conduct the VPV, sorptivity and RCPT test.Roughly, 75 mm × 75 mm × 285 mm concrete prism samples were used to determine the drying shrinkage and alkali-silica reaction (ASR) test as per AS 1012.13 [31] and ASTM C1293 [32], respectively.Figure 4 shows the typical test setup used to perform the RCPT and sorptivity tests.A scanning electron microscopy (SEM) analysis was performed to investigate the microstructure of the aggregates-matrix interface of the concrete specimens.The SEM analysis was conducted using the Tescan Mira3 microscope.

Results
The slump values of the OPC concretes were within the designed range as shown in in Table 2, and slightly decreased due to an increase in the glass aggregate contents.However, the slump vales of the geopolymer concrete increased due to an increase in the glass aggregate, as shown in Table 2.This increase in the slump was ascribed to the lower water absorption and smoother surface of the glass aggregate than the natural aggregate.

Compressive Strength
Figure 5a shows the compressive strength of the OPC concrete containing different proportions of the recycled glass coarse aggregate.It can be seen that the compressive strength of the OPC concrete decreased due to an increase in the recycled glass aggregate.For instance, the 28-day compressive strength of the recycled glass aggregate concrete decreased by 0.21, 3.00, 10.28 and 20.99% compared to the control concrete (without the recycled glass aggregate) due to the use of a 10, 20, 30 and 40% recycled glass aggregate,  The porosity of the concrete sample was determined using the volume of permeable voids (VPV) test following ASTM C642 [28].The sorptivity test was performed to determine the capillary suction of the concrete in accordance with ASTM C1585 [29].The rapid chloride permeability test (RCPT) was carried out following ASTM C1202 [30].The 50 mm thick discs were cut from 100 mm × 200 mm cylinder samples to conduct the VPV, sorptivity and RCPT test.Roughly, 75 mm × 75 mm × 285 mm concrete prism samples were used to determine the drying shrinkage and alkali-silica reaction (ASR) test as per AS 1012.13 [31] and ASTM C1293 [32], respectively.Figure 4 shows the typical test setup used to perform the RCPT and sorptivity tests.A scanning electron microscopy (SEM) analysis was performed to investigate the microstructure of the aggregates-matrix interface of the concrete specimens.The SEM analysis was conducted using the Tescan Mira3 microscope.

Results
The slump values of the OPC concretes were within the designed range as shown in in Table 2, and slightly decreased due to an increase in the glass aggregate contents.However, the slump vales of the geopolymer concrete increased due to an increase in the glass aggregate, as shown in Table 2.This increase in the slump was ascribed to the lower water absorption and smoother surface of the glass aggregate than the natural aggregate.

Compressive Strength
Figure 5a shows the compressive strength of the OPC concrete containing different proportions of the recycled glass coarse aggregate.It can be seen that the compressive strength of the OPC concrete decreased due to an increase in the recycled glass aggregate.For instance, the 28-day compressive strength of the recycled glass aggregate concrete decreased by 0.21, 3.00, 10.28 and 20.99% compared to the control concrete (without the recycled glass aggregate) due to the use of a 10, 20, 30 and 40% recycled glass aggregate,

Results
The slump values of the OPC concretes were within the designed range as shown in in Table 2, and slightly decreased due to an increase in the glass aggregate contents.However, the slump vales of the geopolymer concrete increased due to an increase in the glass aggregate, as shown in Table 2.This increase in the slump was ascribed to the lower water absorption and smoother surface of the glass aggregate than the natural aggregate.

Compressive Strength
Figure 5a shows the compressive strength of the OPC concrete containing different proportions of the recycled glass coarse aggregate.It can be seen that the compressive strength of the OPC concrete decreased due to an increase in the recycled glass aggregate.For instance, the 28-day compressive strength of the recycled glass aggregate concrete decreased by 0.21, 3.00, 10.28 and 20.99% compared to the control concrete (without the recycled glass aggregate) due to the use of a 10, 20, 30 and 40% recycled glass aggregate, respectively.A similar trend of the decreased compressive strength due to the use of a recycled glass aggregate in OPC concrete was reported in previous studies [11][12][13].Omoding et al. [15] found that the use of a 12.5, 25, 50 and 100% glass aggregate reduced the compressive strength by 4, 16, 20 and 27%, respectively.
respectively.A similar trend of the decreased compressive strength due to the use of a recycled glass aggregate in OPC concrete was reported in previous studies [11][12][13].Omoding et al. [15] found that the use of a 12.5, 25, 50 and 100% glass aggregate reduced the compressive strength by 4, 16, 20 and 27%, respectively.The compressive strength of the heat-cured geopolymer concrete containing different percentages of a recycled glass coarse aggregate is shown in Figure 5b.It can be seen that the compressive strength of the heat-cured geopolymer concrete decreased slightly with an increase in the recycled glass aggregate, similar to the OPC concrete.At 28 days, the heat-cured geopolymer concrete mixtures with a 0, 10, 20 and 30% recycled glass aggregate provided the compressive strengths of 65.1, 58.1, 55.5 and 48.3 MPa, respectively.Furthermore, the strength of the heat-cured geopolymer concrete was higher than that of the corresponding OPC concrete.For instance, at 28 days, the compressive strength of the heat-cured geopolymer concrete increased by 39.40, 24.68, 22.52 and 15.27% compared to the OPC concrete due to the use of a 0, 10, 20 and 30% recycled glass aggregate, respectively.A higher degree of geopolymerization occurred at higher temperatures [33] and, consequently, a higher strength was identified at higher temperatures.It should be noted that most of the strength of the heat-cured geopolymer was gained in the early age due to high temperature, and the strength after 28 days was not monitored in the heat-cured geopolymer concrete.
Figure 5c shows the compressive strength of the ambient-cured geopolymer concrete containing a recycled glass coarse aggregate.It can be seen that the compressive strength of the ambient-cured geopolymer concrete decreased due to the use of a recycled glass aggregate similar to the OPC and heat-cured geopolymer concretes.It can be seen that, at the early age of 7 days, the compressive strength of the ambient-cured geopolymer was very low compared to the corresponding OPC and heat-cured geopolymer concretes.For instance, at 7 days, the concrete mix with a 10% recycled glass aggregate (RGA) provided the compressive strength of 11.8 MPa at ambient curing conditions, which was 40.2 and 59.7 MPa for the OPC and heat-cured geopolymer concretes, respectively.On the other hand, at 28 days and afterward, the compressive strength of the ambient-cured geopolymer concrete significantly increased compared to its early age strength.For instance, for a 10% RGA, the compressive strength of the ambient-cured geopolymer concrete increased from 11.8 MPa at 7 days to 32.1, 37.8 and 41.2 MPa at 28, 56 and 90 days, respectively.The increase in the compressive strength after 7 days was ascribed to the delayed development of the reaction product under ambient conditions [34].
Overall, the compressive strength of the OPC and geopolymer concretes was reduced due to the use of a recycled glass aggregate.The size, shape and surface texture of the aggregate had a significant effect on the strength of the concrete [35].The decrease in the compressive strength due to the use of a recycled glass aggregate was ascribed to the smooth surface of the glass aggregate, which produced a poor bond with the binder matrix.The poor geometry of the glass aggregate might have hampered the homogeneous distribution of the aggregate [11].In consequence, the compressive strength was reduced with an increase in the recycled glass aggregate.However, the compressive strength due to the use of a 10 to 20% glass aggregate was comparable to the compressive strength of the control (without the recycled glass aggregate) concrete.

Splitting Tensile Strength
Figure 6 presents the 28-day splitting tensile strength of the OPC and geopolymer concretes containing different proportions of a recycled glass coarse aggregate.It can be seen that for both the OPC and geopolymer concretes, the splitting tensile strength was reduced due to the use of a recycled glass aggregate.The splitting tensile strength of the OPC concrete was 3.88, 3.86, 3.61, 3.67 and 3.51 MPa for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.A similar trend can be observed for the geopolymer concrete, where the splitting tensile strength decreased due to an increase in the recycled glass aggregate.The use of a 10, 20 and 30% recycled glass aggregates as a substitution for the natural coarse aggregate decreased the splitting tensile strength of the heat-cured geopolymer concrete by 10.91, 21.07 and 29.12%, respectively.Similar observations of a decreased splitting tensile strength due to the use of a recycled glass aggregate were found in previous studies [11,13,15].Omoding et al. [15] reported that the use of a 12.5, 25, 50 and 100% glass aggregate reduced the splitting tensile strength by 19, 1, 10 and 18%, respectively.Similar to the case of the compressive strength, the decrease in the splitting tensile strength due to the use of a recycled glass aggregate was ascribed to the poor geometry of the glass aggregate and an inadequate bond between the smooth glass surface and the binder matrix.Furthermore, it can be seen that the heat-cured geopolymer concrete provided a higher splitting tensile strength than the OPC and ambient-cured geopolymer concretes.This was ascribed to a higher degree of geopolymerization in the heat-cured geopolymer, as described in Section 3.1.
ometry of the glass aggregate and an inadequate bond between the smooth glass surface and the binder matrix.Furthermore, it can be seen that the heat-cured geopolymer concrete provided a higher splitting tensile strength than the OPC and ambient-cured geopolymer concretes.This was ascribed to a higher degree of geopolymerization in the heatcured geopolymer, as described in Section 3.1.
A comparison of the measured 28-day compressive and tensile strengths of the OPC concrete, containing different percentage replacements of a crushed recycled glass aggregates with the proposed prediction equation by Celik et al. [18], found that the proposed equation overestimated the compressive and tensile strengths of the concretes containing 20% or more crushed recycled glass aggregates.However, at the 10% replacement level, the prediction equation worked well with the measured strengths.

Modulus of Elasticity
The effect of the recycled glass aggregate on the modulus of elasticity of the OPC and geopolymer concretes is shown in Figure 7.It can be seen that the modulus of elasticity of the OPC concrete varied between 34.2 and 38.6 GPa.There was no certain trend due to the increase in the glass aggregate.The use of a 10% and 30% glass aggregate increased the modulus of elasticity of the OPC concrete by 2.39% and 1.33%, respectively.On the other hand, the modulus of elasticity was reduced by 6.10% and 9.28% due to the use of a 20% and 40% glass aggregate, respectively.A similar behavior was found in the previous studies [13,15], where no definite trend of the modulus of elasticity of the OPC concrete due to an increase in the glass aggregate was observed.Serpa et al. [13] reported that the modulus of elasticity of concrete increased by 5% and 3% due to the use of a 5% and 20% glass aggregate, whereas the modulus of elasticity was reduced by 1% due to the use of a 10% glass aggregate.It can also be seen in Figure 4 that the modulus of elasticity of the geopolymer concrete decreased due to the use of a recycled glass aggregate.For instance, the modulus of elasticity of the heat-cured geopolymer concrete decreased by 8.64, 14.75 and 20.68% due to the use of a 10, 20 and 30% glass aggregate, respectively.Moreover, the heat-cured geopolymer concrete provided a higher modulus of elasticity than the OPC and ambient-cured geopolymer concretes for all the mixtures.A comparison of the measured 28-day compressive and tensile strengths of the OPC concrete, containing different percentage replacements of a crushed recycled glass aggregates with the proposed prediction equation by Celik et al. [18], found that the proposed equation overestimated the compressive and tensile strengths of the concretes containing 20% or more crushed recycled glass aggregates.However, at the 10% replacement level, the prediction equation worked well with the measured strengths.

Modulus of Elasticity
The effect of the recycled glass aggregate on the modulus of elasticity of the OPC and geopolymer concretes is shown in Figure 7.It can be seen that the modulus of elasticity of the OPC concrete varied between 34.2 and 38.6 GPa.There was no certain trend due to the increase in the glass aggregate.The use of a 10% and 30% glass aggregate increased the modulus of elasticity of the OPC concrete by 2.39% and 1.33%, respectively.On the other hand, the modulus of elasticity was reduced by 6.10% and 9.28% due to the use of a 20% and 40% glass aggregate, respectively.A similar behavior was found in the previous studies [13,15], where no definite trend of the modulus of elasticity of the OPC concrete due to an increase in the glass aggregate was observed.Serpa et al. [13] reported that the modulus of elasticity of concrete increased by 5% and 3% due to the use of a 5% and 20% glass aggregate, whereas the modulus of elasticity was reduced by 1% due to the use of a 10% glass aggregate.It can also be seen in Figure 4 that the modulus of elasticity of the geopolymer concrete decreased due to the use of a recycled glass aggregate.For instance, the modulus of elasticity of the heat-cured geopolymer concrete decreased by 8.64, 14.75 and 20.68% due to the use of a 10, 20 and 30% glass aggregate, respectively.Moreover, the heat-cured geopolymer concrete provided a higher modulus of elasticity than the OPC and ambient-cured geopolymer concretes for all the mixtures.

Porosity
The porosity of the concrete samples was determined by measuring the volume of permeable voids (VPV).Figure 8 shows the influence of the recycled glass aggregate on the volume of permeable voids of the OPC concrete and the heat-cured and ambient-cured geopolymer concretes.It can be observed that the porosity of the OPC and geopolymer concretes slightly increased due to an increase in the glass aggregate content.The VPV values of the OPC concrete were 11.47, 11.76, 12.11, 12.19 and 12.33% for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.A similar trend can be noticed for the heat-cured geopolymer concrete, where the VPV values of 8.04, 8.05, 8.47 and 8.61% were found for using a 0, 10, 20 and 30% recycled glass aggregate, respectively.The VPV values of the ambient-cured geopolymer concrete were 8.24 and 8.30% for using a 0 and 10% recycled glass aggregate, respectively.Ouldkhaoua et al. [36] and Khan and Sarker [37] also reported that the use of a glass fine aggregate as a substitution for natural sand increased the porosity of the OPC and geopolymer concretes.The increase in the porosity due to an increase in the recycled glass aggregate was ascribed to the increase in the voids in the concrete due to the poor bond and angular shape of the glass aggregate.

Porosity
The porosity of the concrete samples was determined by measuring the volume of permeable voids (VPV).Figure 8 shows the influence of the recycled glass aggregate on the volume of permeable voids of the OPC concrete and the heat-cured and ambient-cured geopolymer concretes.It can be observed that the porosity of the OPC and geopolymer concretes slightly increased due to an increase in the glass aggregate content.The VPV values of the OPC concrete were 11.47, 11.76, 12.11, 12.19 and 12.33% for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.A similar trend can be noticed for the heat-cured geopolymer concrete, where the VPV values of 8.04, 8.05, 8.47 and 8.61% were found for using a 0, 10, 20 and 30% recycled glass aggregate, respectively.The VPV values of the ambient-cured geopolymer concrete were 8.24 and 8.30% for using a 0 and 10% recycled glass aggregate, respectively.Ouldkhaoua et al. [36] and Khan and Sarker [37] also reported that the use of a glass fine aggregate as a substitution for natural sand increased the porosity of the OPC and geopolymer concretes.The increase in the porosity due to an increase in the recycled glass aggregate was ascribed to the increase in the voids in the concrete due to the poor bond and angular shape of the glass aggregate.

Porosity
The porosity of the concrete samples was determined by measuring the volume of permeable voids (VPV).Figure 8 shows the influence of the recycled glass aggregate on the volume of permeable voids of the OPC concrete and the heat-cured and ambient-cured geopolymer concretes.It can be observed that the porosity of the OPC and geopolymer concretes slightly increased due to an increase in the glass aggregate content.The VPV values of the OPC concrete were 11.47, 11.76, 12.11, 12.19 and 12.33% for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.A similar trend can be noticed for the heat-cured geopolymer concrete, where the VPV values of 8.04, 8.05, 8.47 and 8.61% were found for using a 0, 10, 20 and 30% recycled glass aggregate, respectively.The VPV values of the ambient-cured geopolymer concrete were 8.24 and 8.30% for using a 0 and 10% recycled glass aggregate, respectively.Ouldkhaoua et al. [36] and Khan and Sarker [37] also reported that the use of a glass fine aggregate as a substitution for natural sand increased the porosity of the OPC and geopolymer concretes.The increase in the porosity due to an increase in the recycled glass aggregate was ascribed to the increase in the voids in the concrete due to the poor bond and angular shape of the glass aggregate.Moreover, it was noticed that, for the same recycled glass aggregate, the geopolymer concrete provided a lower porosity compared to the OPC concrete.For instance, for a 10% recycled glass aggregate, the VPV of the OPC concrete was 11.76% and the VPV of the heat-cured and ambient-cured geopolymer concretes were 8.05 and 8.30%, respectively.The alkaline aluminosilicate gel of the geopolymer concrete might produce a compact structure, and thus, reduce the porosity of the concrete compared to the OPC concrete [34].The durability characteristics as per the VPV of concrete by VicRoads [38] are also shown in Figure 8.According to VicRoads, concrete with VPV values between 11% and 13% is categorized as good and that less than 11% is categorized as excellent concrete.Therefore, the geopolymer concrete could be classified as excellent concrete and the OPC concrete could be classified as good concrete in terms of the volume of permeable voids.

Sorptivity
The influence of the recycled glass aggregate on the sorptivity coefficient of the OPC and geopolymer concretes is shown in Figure 9.It can be noticed that the sorptivity coefficient of the ambient and heat-cured geopolymer concretes increased due to the increase in the glass aggregate.The sorptivity coefficients of the heat-cured geopolymer concrete were 3.70 × 10 −3 , 5.57 × 10 −3 , 8.73 × 10 −3 and 10.50 × 10 −3 mm/s 1/2 for using a 0, 10, 20 and 30% recycled glass aggregate, respectively.In the case of the ambient-cured geopolymer concrete, the sorptivity coefficients were 13.10 × 10 −3 and 15.53 × 10 −3 mm/s 1/2 for using a 0 and 10% recycled glass aggregate, respectively.On the other hand, it can be seen that the sorptivity coefficient of the OPC concrete containing a natural coarse aggregate (without a glass aggregate) was higher than that of the OPC concrete containing 10 to 30% glass coarse aggregates.For instance, the sorptivity coefficients of the OPC concrete decreased from 15.80 × 10 −3 mm/s 1/2 for a 0% glass aggregate (100% natural coarse aggregate) to 13.65 × 10 −3 , 14.36 × 10 −3 and 14.65 × 10 −3 mm/s 1/2 for a 10, 20 and 30% glass aggregate, respectively.The lower sorptivity of glass coarse aggregate concrete compared to natural coarse aggregate concrete was ascribed to the lower water absorption capacity of the glass aggregate compared to the natural coarse aggregate [37].However, the sorptivity coefficient of the OPC concrete also increased due to the increase in the glass aggregate, similar to the geopolymer concrete.For instance, the sorptivity coefficients of the OPC concrete increased from 13.65 × 10 −3 mm/s 1/2 for a 10% glass aggregate to 14.36 × 10 −3 , 14.65 × 10 −3 and 18.55 × 10 −3 mm/s 1/2 for a 20, 30 and 40% glass aggregate, respectively.The increase in the sorptivity due to the use of the recycled glass aggregate was ascribed to the increase in the porosity due to the poor bond and angular shape of the glass aggregate, as described in the previous section.However, the sorptivity coefficients of all the samples were well below the allowable limit (27.11 × 10 −3 mm/s 1/2 ), as recommended by Cement Concrete & Aggregates Australia [39].Furthermore, for the same recycled glass aggregate, the heat-cured geopolymer concrete provided a low sorptivity compared to the OPC and ambient-cured geopolymer concretes.A higher degree of geopolymerization for the heatcured geopolymer provided a compact structure which led to a decrease in the sorptivity and an increase in the compressive strength, as described in Section 3.1.

Chloride Ion Penetration
Figure 10 presents the RCPT results of the OPC and geopolymer concretes containing different proportions of a recycled glass coarse aggregate.The classification of the concretes per the charge passed by the ASTM C1202 [30] are also shown in Figure 10.It can be seen that the total charge passed through the OPC concrete increased due to the increase in the recycled glass aggregate.The amount of charge that passed through the OPC concrete was 359, 388, 389, 464 and 538 C for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.It was also observed that the total charge that passed through the heat-cured geopolymer concrete varied between 573 C and 1047 C, and the total charge increased due to the increase in the recycled glass aggregate, similar to the OPC and ambient-cured geopolymer concretes.The amount of charge that passed through the ambient-cured geopolymer concrete was 859.50 and 787.50 C for using a 0 and 10% recycled glass aggregate, respectively.The increase in the chloride permeability due to the increase in the glass aggregate was ascribed to the increase in the porosity of the concrete, as described in Section 3.4.However, the chloride permeability of the heat-cured geopolymer sample containing a 30% glass aggregate was classified as 'low' and all the other concrete samples were classified as 'very low' in accordance with the ASTM C1202 [30].

Chloride Ion Penetration
Figure 10 presents the RCPT results of the OPC and geopolymer concretes containing different proportions of a recycled glass coarse aggregate.The classification of the con cretes per the charge passed by the ASTM C1202 [30] are also shown in Figure 10.It can be seen that the total charge passed through the OPC concrete increased due to the in crease in the recycled glass aggregate.The amount of charge that passed through the OPC concrete was 359, 388, 389, 464 and 538 C for using a 0, 10, 20, 30 and 40% recycled glas aggregate, respectively.It was also observed that the total charge that passed through th heat-cured geopolymer concrete varied between 573 C and 1047 C, and the total charg increased due to the increase in the recycled glass aggregate, similar to the OPC and am bient-cured geopolymer concretes.The amount of charge that passed through the ambi ent-cured geopolymer concrete was 859.50 and 787.50 C for using a 0 and 10% recycled glass aggregate, respectively.The increase in the chloride permeability due to the increas in the glass aggregate was ascribed to the increase in the porosity of the concrete, as de scribed in Section 3.4.However, the chloride permeability of the heat-cured geopolyme sample containing a 30% glass aggregate was classified as 'low' and all the other concret samples were classified as 'very low' in accordance with the ASTM C1202 [30] Low Moderate High

Chloride Ion Penetration
Figure 10 presents the RCPT results of the OPC and geopolymer concretes containing different proportions of a recycled glass coarse aggregate.The classification of the concretes per the charge passed by the ASTM C1202 [30] are also shown in Figure 10.It can be seen that the total charge passed through the OPC concrete increased due to the increase in the recycled glass aggregate.The amount of charge that passed through the OPC concrete was 359, 388, 389, 464 and 538 C for using a 0, 10, 20, 30 and 40% recycled glass aggregate, respectively.It was also observed that the total charge that passed through the heat-cured geopolymer concrete varied between 573 C and 1047 C, and the total charge increased due to the increase in the recycled glass aggregate, similar to the OPC and ambient-cured geopolymer concretes.The amount of charge that passed through the ambient-cured geopolymer concrete was 859.50 and 787.50 C for using a 0 and 10% recycled glass aggregate, respectively.The increase in the chloride permeability due to the increase in the glass aggregate was ascribed to the increase in the porosity of the concrete, as described in Section 3.4.However, the chloride permeability of the heat-cured geopolymer sample containing a 30% glass aggregate was classified as 'low' and all the other concrete samples were classified as 'very low' in accordance with the ASTM C1202 [30]  Furthermore, it was observed that, for the same recycled glass aggregate, the total charge that passed through geopolymer concrete was higher than the OPC concrete.For instance, for a 10% recycled glass aggregate, the charge that passed through the OPC concrete was 388 C and through heat-cured and ambient-cured geopolymer concretes were 675 C and 816 C, respectively.It was also noticed that, though the geopolymer concrete provided a lower porosity than the OPC concrete (Figure 5), the chloride permeability of the geopolymer concrete was higher than the OPC concrete.This might be due to the higher pore solution conductivity of the geopolymer concrete than the OPC concrete [37, 40,41].

Drying Shrinkage
Figure 11 presents the drying shrinkage of the OPC concrete, heat-cured geopolymer concrete and ambient-cured geopolymer concrete containing different percentages of a recycled glass coarse aggregate.In general, at a certain age, the drying shrinkage of both the OPC and geopolymer concrete decreased due to the use of a recycled glass aggregate.For instance, at 28 days, the drying shrinkage of the OPC concrete decreased from a 360 microstrain for a 0% glass aggregate to a 285, 273, 254 and 283 microstrain for a 10, 20, 30 and 40% glass aggregate, respectively.Ling and Poon [42] reported that the use of a large glass aggregate reduced the drying shrinkage of the OPC concrete.A similar decreasing trend of the drying shrinkage due to the use of a glass fine aggregate as a replacement for natural sand in the OPC concrete was reported in previous studies [43,44].The decrease in the drying shrinkage due to the use of a recycled glass aggregate was ascribed to the low water absorption of the glass aggregate [42,45].(c) Ambient-cured geopolymer concrete  Moreover, at a particular age, the heat-cured geopolymer concrete provided a very low drying shrinkage compared to the corresponding OPC and ambient-cured geopolymer concretes.For instance, for a 10% glass aggregate, the 28-day drying shrinkage of the heat-cured geopolymer concrete was a 85 microstrain while the OPC and ambient-cured geopolymer concretes were a 285 and 187 microstrain, respectively.The high curing temperature reduced the evaporable water and produced a compact matrix.In consequence, the drying shrinkage of the heat-cured geopolymer decreased [1,46].

Alkali-Silica Reactivity
The alkali-silica reactivity (ASR) is an important characteristic which should be investigated prior to using glass as an aggregate material.Figure 12 shows the change in the length after the ASR test for the OPC and heat-cured geopolymer concretes containing a 20% recycled glass aggregate.It can be seen that the expansion of both concrete samples was very low.In the case of the OPC concrete, the maximum expansion was 0.007% after 6 months of the ASR test.Yuksel et al. [47] found that the expansion of the control concrete microbar without a glass aggregate was around 0.002% at 30 days.On the other hand, the expansion of the geopolymer sample was very low compared to the OPC concrete sample.Additionally, the geopolymer concrete sample showed shrinkage after 28 days.A similar observation of shrinkage during the ASR test of mortar using glass fine aggregate was found by Xie et al. [48].Xie et al. [48] reported that the the cause for this type of shrinkage related to the binder materials instead of the glass aggregates.However, after 6 months of the ASR test, no visible expansion cracks were found in the OPC and geopolymer concrete samples containing a 20% glass coarse aggregate, as shown in Figure 13.

Microstructural Analysis
The microstructures of the OPC concrete and heat-cured geopolymer concrete containing a 10% recycled glass aggregate were investigated at 28 days.Figure 14 shows the SEM images of the microstructure of the OPC concrete and the heat-cured geopolymer concrete at different magnifications, respectively.It can be seen that, in both cases (the OPC concrete and the geopolymer concrete), the natural coarse aggregate (stone chips) and the paste produced stronger bond than the recycled glass aggregate and paste.A large gap can be seen at the interfacial transition zone (ITZ) between the RGA and the paste matrix.This poor bond at the ITZ between the RGA and the paste matrix was ascribed to the smooth surface of the recycled glass aggregate.Similar observations were reported in previous studies, where glass was used as a replacement for a natural coarse aggregate [15] and a natural fine aggregate [49].The properties of the concrete were highly dependent on the surrounding ITZ of aggregates [50].The weak ITZ between the RGA and the paste matrix led to a decrease in the mechanical strength (Sections 3.1 and 3.2) and an increase in the porosity (Section 3.4) of the concrete due to the use of a glass aggregate.It can also be seen that, due to the sharp edge of the glass aggregate, some microcracks in the matrix were caused by the glass aggregate, which further propagated and increased the porosity of the concrete.

Alkali-Silica Reactivity
The alkali-silica reactivity (ASR) is an important characteristic which should be investigated prior to using glass as an aggregate material.Figure 12 shows the change in the length after the ASR test for the OPC and heat-cured geopolymer concretes containing a 20% recycled glass aggregate.It can be seen that the expansion of both concrete samples was very low.In the case of the OPC concrete, the maximum expansion was 0.007% after 6 months of the ASR test.Yuksel et al. [47] found that the expansion of the control concrete microbar without a glass aggregate was around 0.002% at 30 days.On the other hand, the expansion of the geopolymer sample was very low compared to the OPC concrete sample.Additionally, the geopolymer concrete sample showed shrinkage after 28 days.A similar observation of shrinkage during the ASR test of mortar using glass fine aggregate was found by Xie et al. [48].Xie et al. [48] reported that the the cause for this type of shrinkage related to the binder materials instead of the glass aggregates.However, after 6 months of the ASR test, no visible expansion cracks were found in the OPC and geopolymer concrete samples containing a 20% glass coarse aggregate, as shown in Figure 13.

Alkali-Silica Reactivity
The alkali-silica reactivity (ASR) is an important characteristic which should be investigated prior to using glass as an aggregate material.Figure 12 shows the change in the length after the ASR test for the OPC and heat-cured geopolymer concretes containing a 20% recycled glass aggregate.It can be seen that the expansion of both concrete samples was very low.In the case of the OPC concrete, the maximum expansion was 0.007% after 6 months of the ASR test.Yuksel et al. [47] found that the expansion of the control concrete microbar without a glass aggregate was around 0.002% at 30 days.On the other hand, the expansion of the geopolymer sample was very low compared to the OPC concrete sample.Additionally, the geopolymer concrete sample showed shrinkage after 28 days.A similar observation of shrinkage during the ASR test of mortar using glass fine aggregate was found by Xie et al. [48].Xie et al. [48] reported that the the cause for this type of shrinkage related to the binder materials instead of the glass aggregates.However, after 6 months of the ASR test, no visible expansion cracks were found in the OPC and geopolymer concrete samples containing a 20% glass coarse aggregate, as shown in Figure 13.

Conclusions
The effect of using recycled waste glass as a coarse aggregate on the properties of ordinary Portland cement concrete and geopolymer concretes was investigated.On the basis of the obtained results, the following conclusions were drawn.
-In general, the compressive and tensile strength of the OPC and geopolymer concretes decreased due to the increase in the glass aggregate.The OPC concrete mixtures with a 0, 10, 20, 30 and 40% recycled glass coarse aggregate provided a 28-day compressive strength of 46.7, 46.6, 45.3, 41.9 and 36.9MPa, respectively.The decrease in the strength due to the use of a glass coarse aggregate was ascribed to the smooth surface of the glass aggregate that produced a poor bond with the binder matrix.

Conclusions
The effect of using recycled waste glass as a coarse aggregate on the properties of ordinary Portland cement concrete and geopolymer concretes was investigated.On the basis of the obtained results, the following conclusions were drawn.

-
In general, the compressive and tensile strength of the OPC and geopolymer concretes decreased due to the increase in the glass aggregate.The OPC concrete mixtures with a 0, 10, 20, 30 and 40% recycled glass coarse aggregate provided a 28-day compressive strength of 46.7, 46.6, 45.3, 41.9 and 36.9MPa, respectively.The decrease in the strength due to the use of a glass coarse aggregate was ascribed to the smooth surface of the glass aggregate that produced a poor bond with the binder matrix.
(a) Concrete mixing (b) Mixed concrete (c) Concrete in molds (d) Concrete under compaction in vibrating table (e) Compacted concrete after finishing (f) Slump testing

Figure 2 .
Figure 2. Illustration of concrete sample testing in this study.

Figure 2 .
Figure 2. Illustration of concrete sample testing in this study.

Figure 3 .
Figure 3. Test setup to measure the mechanical properties of concrete (Left: compression test, middle: split cylinder test and right: Elastic modulus test).

Figure 4 .
Figure 4. Test setup of the rapid chloride permeability test (a) and sorptivity test (b).

Figure 3 .
Figure 3. Test setup to measure the mechanical properties of concrete (Left: compression test, middle: split cylinder test and right: Elastic modulus test).

Figure 3 .
Figure 3. Test setup to measure the mechanical properties of concrete (Left: compression test, middle: split cylinder test and right: Elastic modulus test).

Figure 4 .
Figure 4. Test setup of the rapid chloride permeability test (a) and sorptivity test (b).

Figure 4 .
Figure 4. Test setup of the rapid chloride permeability test (a) and sorptivity test (b).

Figure 6 .
Figure 6.Splitting tensile strength of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 6 .
Figure 6.Splitting tensile strength of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 7 .
Figure 7. Modulus of elasticity of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 8 .
Figure 8. Volume of permeable voids of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 7 .
Figure 7. Modulus of elasticity of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 8 .
Figure 8. Volume of permeable voids of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 8 .
Figure 8. Volume of permeable voids of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 11 .
Figure 11.Drying shrinkage of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.Figure 11.Drying shrinkage of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 11 .
Figure 11.Drying shrinkage of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.Figure 11.Drying shrinkage of the OPC, heat-cured and ambient-cured geopolymer concretes containing recycled glass coarse aggregates.

Figure 12 .Figure 13 .
Figure 12.Length of change after the ASR test for the OPC and heat-cured geopolymer concretes containing 20% recycled glass aggregates.

Figure 12 .
Figure 12.Length of change after the ASR test for the OPC and heat-cured geopolymer concretes containing 20% recycled glass aggregates.

Figure 12 .Figure 13 .
Figure 12.Length of change after the ASR test for the OPC and heat-cured geopolymer concretes containing 20% recycled glass aggregates.

Table 2 .
Mix proportions of the OPC and geopolymer concretes (kg/m 3 ) and slump values.