Sustainable Lightweight Self-Compacting Concrete Incorporating Recycled Waste Glass: Influence of Particle Size and Replacement Levels on Performance

2.2.1. Portland Cement

In the current study, Ordinary Portland Cement known by Al Mass was applied. This cement meets the Iraq specification IQS NO.5-2019. The cement is produced by (AL-Mass Company) in the north of Iraq (Kurdistan region). Table 1 shows the chemical analysis and physical test.

2.2.2. Fine Aggregate

All LWSCC mixes design by using natural fine aggregate obtained from the Al-Ekhaider region. The physical properties of sand were specific gravity 2.64, water absorption 2% and sulphate content 0.344. The grading curve of the applied sand are given in Figure 1, according to Iraqi Specifications No. 45/1984, the used sand is in zone II.

Figure 1:

2.2.3. Recycled Waste Glass

The waste glass is collected from landfill juice bottles and waste windows glass, then washing it by water tap to remove impurities such as dust, labels, and juice contaminants. The cleaned glass is left to air-dry under ambient conditions to remove any excess moisture. The dried glass was crushed in the loss Angeles apparatus then sieved into two distinct fractions based on particle size. The first fraction, referred to as coarse glass, consisted of particles passing the 4.75 mm sieve and retained on the 1.18 mm sieve. The second fraction, referred to as fine glass, consisted of particles passing the 1.18 mm sieve as shown in the Figure 2. both types were stored in containers to prevent moisture absorption prior to mixing. The prepared glass materials were used as partial replacements for natural sand at specified replacement ratios (10, 20, 30 and, 40) % in the concrete mixture. The chemical analysis of the waste glass used in this study were illustrated in Table 2.

Figure 2:

2.2.4. Supplementary Cementitious Materials (Fly Ash and Limestone Powder)

Class F fly ash and limestone powder (LP) were utilized as supplementary cementitious materials in this study to improve the performance of LWSCC. The fly ash conforms with ASTM C 618 requirements.

Limestone powder, locally known as Al-Ghubra, was used as a filler material to increase the cohesiveness and improve particle packing of the mixture, Its high calcium oxide (CaO). Table 3 show the chemical analysis of fly ash type F and limestone powder.

2.2.5. Super-plasticizer

Superplasticizer (Sp) (high-range water-reducing admixture), commercially known as Master Glenium^® 54, was used to achieve the required workability. It is based on modified polycarboxylate ether and complies with ASTM C494 Type F. The chemical admixture properties are white-like to straw coloured liquid form, relative density 1.07 and PH 5–8.

2.2.6. Attapulgite Aggregate

Attapulgite lightweight aggregate as shown in the Figure 3 was employed as coarse aggregate in this study. Attapulgite study from the Al-Najaf region the aggregate was obtained. Attapulgite aggregate particle size distribution within the range of 5–14 mm to conforms to BS EN 882:1992, it has a dry unit weight of 866 kg/m³, water absorption of 26%, and specific gravity of 1.54, the grading of LWA in Figure 4 below:

Figure 3:

Figure 4:

To reduce absorption water of attapulgite aggregate, the aggregate surface was treated with cement impregnation according to (Yew et al., 2023) cement paste is made by using ordinary Portland cement known by Al Mas and tap water. Dry the attapulgite aggregate in before treated it. Prepare the grout by using cement and water with w\c=0.6 and mix it by using the blade mixer. Soaking LWA aggregate in the grout and mix it for 5 minutes. then, leave it for a period until the coating is evenly distributed, empty it into a basket (mesh) to eliminate of excess grout. After that, spread it on a horizontal, non-absorbent surface until the initial setting occurs. Sieving aggregate by using mesh to remove any Excess cement from its surface. Then cured it for Figure 6: Treated Attapulgite by Cement Paste three days to ensure sealed the pores and getting additional strength. Figure 5 shows the results of treated attapulgite by cement paste.

Figure 5:

6.2.1. Equilibrium Density

The equilibrium density results clearly show a systematic reduction with increasing glass replacement for both GC and GF mixtures compared to the reference mixture (1646 kg/m³), confirming the sensitivity of LWSCC density to aggregate type and particle size distribution. Table 6 and Figure 16 show the results of Equilibrium density for LWSCC of GC and GF mixtures.

Figure 16:

From the above results, LWSCC contained coarser glass (GC mixtures), the equilibrium density decreased from 1637 kg/m³ (10%) to 1620 kg/m³ (40%), corresponding to a total reduction of approximately 1.6%,0.55%, 1.21%, and 1.58% at 10%, 20%, 30%, and 40%, respectively. This reduction indicates that the coarser glass fraction maintained relatively better packing efficiency and particle interlock within the cementitious matrix.

LWSCC contained finer glass (GF mixtures) showed decreasing in the density from 1633 kg/m³ at (10%) to 1594 kg/m³ at (40%), the reduction was approximately 0.79%, 1.34%, 1.64%, and 3.16% at 10%, 20%, 30%, and 40%, respectively. The rate of reduction in GF mixtures was higher than that of GC mixtures. This behaviour can be attributed to the fine glass particles has a higher specific surface area of, which increases required water demand and helps of particle agglomeration, leading to reducing of packing efficiency and the creation of micro-voids within the matrix. These results stay within the acceptable limits of structural lightweight concrete according to ACI 318-11 (Standard, 2011), which specifies that the equilibrium density of structural lightweight concrete normally ranges between 1440 and 1840 kg/m³.

6.2.2. Compressive Strength

Compressive strength is an important mechanical property of concrete that denotes concrete ability to resist applied loads, and it is widely used to evaluate its quality and performance. In LWSCC, compressive strength influenced by many factors such as aggregate type, particle size, and the replacement materials like waste glass. Table 7 and Figure 17 illustrated the results of compressive strength at 7, 28 and 90 days.

The compressive strength results show a significant influence of glass particle size and replacement ratios on the mechanical properties of LWSCC. For GC mixtures, a continuous reduction in compressive strength was observed with increasing replacement levels at all curing ages. At 7 days, the strength decreased by approximately 2.8%, 4.9%, 7.8%, and 10.4% for 10%, 20%, 30%, and 40% replacement, respectively, compared to the control mix. A similar trend was observed at 28 days, where the reduction about 2%, 4.5,6.4% and 8.6% at 10%, 20%, 30%, and 40% replacement, while at 90 days, the decrease in compressive strength at10%, 20%, 30%, and 40% was approximately 2.8%, 4.5%, 6.2% and 8.9%, respectively. This behavior is mainly attributed to the smooth surface of glass particles and coarse glass behaves mainly as an inert aggregate resulting in a weaker interfacial transition zone (ITZ), which leads to poor bonding and the formation of internal micro cracks. These findings are consistent with previous studies (Tong et al., 2024) (Qaidi et al., 2022) (Dadouch et al., 2024), which reported that increasing the proportion of glass aggregate, particularly with larger particle sizes, reduces compressive strength due to weak adhesion and unfavorable particle morphology.

For GF mixture displayed an improvement in compressive strength. At 7 days, the strength increased by approximately 2.5% and 3.3% at 10% and 20% replacement, respectively, before slightly decreasing at higher ratios 1.2%,1.3% at 30% and 40%. At 28 days, the improvement became more pronounced, with increases of about 4.7%, 7.3%, and 4.1% at 10%, 20%, and 30% replacement, respectively, while the strength remained nearly equal to the control mix at 40% replacement. Similarly, at 90 days, slight improvements of about 1.9% and 1.7% were observed at 10% and 20%, followed by reductions of approximately 1.0% and 3.1% at 30% and 40%, respectively. The enhancement at 10% and 20% levels can be attributed to the particle packing, void filling, and refinement of the pore structure. Although finely ground waste glass may exhibit pozzolanic activity because of its amorphous silica, which raise the formation of additional C–S–H gel and result in a denser microstructure. This behavior agrees with recent studies (Poudel et al., 2025) (Ishaq & Mohammed, 2026)]. Though, at higher replacement levels (30% and 40%), the reduction in strength for GF mixtures due to the increase in specific surface area, which required a raise water demand and leads to particle agglomeration and higher porosity, Similar results were obtained by (Kassed & Ali, 2021).

Figure 17:

For both GC, GF mixtures at glass replacement percentage 30% and 40%, attained compressive strength values within the acceptable limits of structural lightweight concrete according to ACI 318-11. However, at these percentages the observed reduction is not desirable, it causes a decline in mechanical efficiency, so it can still be acceptable for specific structural applications where reduced self-weight is useful.

6.2.3. Splitting Tensile strength

The splitting tensile strength results show obvious dependency on both glass particle size and replacement percentage fraction, exhibiting a behavior closely aligned with the compressive strength trends. Table 8 and Figure 18 display experimental results of splitting strength test.

For LWSCC mixtures incorporating coarse glass (GC), experimental results show a progressive reduction in tensile strength with increasing replacement levels at all specified ages. At 7 days, the reduction ranged from approximately 1.3%,3.1%, 10.3% and 16.5% at 10%,20%, 30% and 40% replacement ratios, respectively, while at 28 days, the decrease was approximately 3.2%, 5.3%,7.9% and 10.9% at the same replacement at age 7days. At the same replacement levels, a similar decreasing trend was also marked at 90 days, with reductions 6.2%,7.3%,9.1% and 10.4%. This deterioration can be primarily attributed to the inherently smooth and non-porous surface of glass particles, which significantly weakens the interfacial transition zone (ITZ), thus reducing the efficiency of stress transfer and of micro cracks were initiated under tensile loading.

Figure 18:

For LWSCC mixtures incorporating with fine glass (GF), at 7 days, increases of approximately 1.3%, 5.4% at 10% and 20% relative to the control mixture, whereas GF30 and GF40 showed reductions of approximately 1.3% and 6.3%. This improvement enhanced more pronounced at 28 days, where the tensile strength increased by approximately 3.5%, 5.9%, 2.1% for GF10, GF20, and GF30, respectively, while GF40 exhibited a reduction of approximately 1.5%. The greatest performance was attained at 20% replacement, elsewhere which a gradual decline was observed. At 90 days, for GF10, GF20, GF30, and GF40 are increased +2.4%, +3.8%, +0.8%, and −0.8%, respectively, This improvement can be attributed to the double function of fine glass particles as both micro-fillers and pozzolanic materials, which contribute to pore refinement, improved packing density, and the formation of additional C–S–H gel, ultimately leading to a denser and more cohesive microstructure. Conversely, at higher replacement levels (greater than 30%), the tensile strength begins to decline. This reduction is due to the specific surface area were increased, which leads to higher water requirement, facilitates particle agglomeration, and leads to increased porosity and weak internal cohesion between concrete consistence. Similar observations have been reported in recent studies, where excessive glass content resulted in a dilution effect and increased void formation, ultimately compromising mechanical performance. Such findings are greatly agreement with previous experimental studies (Fayed et al., 2024) (Oan & Elrefaei, 2022).

6.2.4. Flexural Strength

The flexural strength results of LWESCC mixtures containing coarse glass (GC) showed a consistent deterioration with increasing replacement ratio at all curing ages, indicating that coarse glass adversely affected the bending resistance of the matrix. Table 9 and Figure 19 show the results of flexural strength tests.

Figure 19:

At 7 days, the flexural strength decreased by 0.5%, 3.2%, 7.0%, and 8.9% at 10%, 20%, 30%, and 40% replacement, respectively, compared with the reference mix. A similar downward trend was observed at 28 days, where the reductions reached 4.2%, 5.6%, 9.1%, and 10.7%, respectively, while at 90 days the corresponding decreases were 2.2%, 3.9%, 6.7%, and 7.8%. The progressive reduction can be attributed to the smooth and non-porous surface of coarse glass particles, which weakens the interfacial transition zone (ITZ) and reduces the efficiency of stress transfer under flexural loading. Also, the angular and brittle nature of coarse glass may help microcrack initiation and propagation, thereby reducing the capacity of the concrete to resist tensile stresses generated by bending. This behavior is consistent with previous findings (Saha et al., 2024) (Qaidi et al., 2022).

LWSCC mixtures containing fine glass (GF) exhibited a more favorable response, particularly at low replacement levels, confirming that glass particle size plays a decisive role in flexural behavior. Over 7 days, the flexural strength increased by 7.0% and 8.1% at 10% and 20% replacement, respectively, before declining by 1.1% and 4.3% at 30% and 40%. At 28 days, the enhancement became more pronounced, with increases of 6.5% and 9.5% at 10% and 20%, respectively, whereas slight and moderate reductions of 0.7% and 4.4% were recorded at 30% and 40%. Likewise, at 90 days, the strength improved by 2.0% and 4.8% at 10% and 20%, respectively, but then decreased by 1.5% and 5.0% at 30% and 40%. The best performance was attained at 20% replacement, especially at 28 days, where the highest improvement 9.5% was recorded. This development is primarily attributed to the filler effect of fine glass particles and their pozzolanic influence, which increases packing density, reduces internal voids, produces a more compacted microstructure, and reinforce the ITZ. Though, once the replacement level exceeded the optimum threshold, the positive effect gradually decreased that causes reduction in flexural strength. This behavior can be mainly combined with dilution effects, increased surface area, localized particle agglomeration, and the formation of higher porosity inside the matrix. This trend agrees well with the literature of (Jahami et al., 2022) (ElNemr & Shaltout, 2025).

6.2.5. Modulus of Elasticity

The modulus of elasticity results of LWASCC mixtures demonstrated a clear dependency on both glass particle size and replacement ratio, with distinct trends observed for (GC) and (GF) across curing ages (28 and 90 days). Table 10 and Figure 20 show the results of modulus of elasticity test.

For GC mixtures, a continuous reduction in elastic modulus was recorded with increasing replacement levels. At 28 days, the modulus of elasticity decreased from 16.74 GPa (reference mix) to 16.22, 15.87, 15.42, and 15.16 GPa at 10%, 20%, 30%, and 40% replacement, corresponding to reductions of approximately 3.1%, 5.2%, 7.9%, and 9.4%, respectively. A similar trend was observed at 90 days, where the modulus declined from 17.96 GPa to 17.35, 17.18, 16.76, and 16.46 GPa, representing reductions of about 3.4%, 4.3%, 6.7%, and 8.4%, respectively. This reduction is mainly attributed to the smooth surface texture and weak interfacial transition zone (ITZ) associated with coarse glass particles, which reduces stiffness and stress transfer efficiency within the cement matrix. Additionally, the angular shape and lower bonding characteristics of glass aggregates contribute to increased microcracking and reduced rigidity, as reported in previous studies (Ahmed & Rana, 2023). In contrast, GF mixtures exhibited an initial improvement in modulus of elasticity at lower replacement levels, followed by a reduction at higher contents. At 28 days, the modulus increased to 17.62 and 17.84 GPa at 10% and 20% replacement, corresponding to increases of approximately 5.3% and 6.6%, respectively, before decreasing to 17.31 GPa (3.4% increase) and 16.90 GPa (1.0% increase) at 30% and 40%. Similarly, at 90 days, the modulus increased to 17.82 and 18.32 GPa (−0.8% decrease and +2.0% increase relative to control), then decreased to 17.87 (−0.5%) and 17.46 GPa (−2.8%) at higher replacement levels. The enhancement at low replacement levels can be attributed to particle packing, void filling, and refinement of the pore structure. Although finely ground waste glass may exhibit pozzolanic activity because of its amorphous silica content that improve matrix densification, thereby increasing stiffness. This behavior is well supported by recent findings (Chen et al., 2022) indicating that finely ground glass enhances microstructure and elastic properties through secondary hydration reactions [38]. Moreover, the improvement in mechanical performance at moderate replacement levels (≈20%) has been widely reported due to pore refinement and improved packing density. However, beyond the optimum replacement level, the modulus decreases due to dilution effects, insufficient calcium hydroxide for continued pozzolanic reaction, and increased porosity, which weakens the internal structure. Furthermore, increasing glass content leads to disruption in aggregate interlocking and increased micro voids, resulting in reduced stiffness. Similar observations were reported in previous studies (Gholampour et al., 2025).

Figure 20:

6.2.6. Impact Resistance by Drop-weight Method

Impact resistance is considered one of the key properties of LWSCC. The results of the impact resistance test for the control mix G0 and other mixes (GC and GF) at (10–40) % replacement ratios at the age of 7, 28 and 90 days are presented in Table 11 and illustrated in Figure 21. The performance was evaluated based on the number of blows required to initiate the first crack (N1) and to reach failure (N2).

For LWSCC mixtures incorporating coarse glass (GA), a general reduction in impact resistance was observed with increasing replacement ratio. At 10% replacement, a slight improvement or comparable performance was recorded; however, beyond 20%, both N1 and Nf decreased progressively, indicating reduced energy absorption capacity and increased brittleness of the concrete matrix. This behavior can be attributed to weaker ITZ and smooth surface of coarse glass particles, which limit crack bridging and promote early crack propagation. This behavior can be attributed to weaker ITZ and smooth surface of coarse glass particles, which limit effective stress transfer and promote early crack propagation, particularly at higher replacement levels. Regardless of the replacement ratio, impact resistance generally increased with treatment age from 7 to 90 days, as reflected by increase in both first-crack (N1) and failure (Nf) blow counts. This age-dependent improvement can be attributed to constant hydration and progressive densification of cementitious matrix, which enhanced matrix strength and resistance to crack initiation and propagation. Nevertheless, the relative performance of mixtures with higher coarse-glass contents remained lower, indicating that prolonged curing could improve the overall impact resistance but could not fully overcome the adverse effect of the weak glass–matrix interface. Similar findings have been reported in previous studies, where higher glass content led to reductions in tensile-related properties due to poor bonding and increased microcracking (Jassam et al., 2022).

In contrast, LESCC mixtures incorporating fine glass (GF mixtures) exhibited a noticeable improvement in impact resistance at moderate replacement levels (10–20%), where both N1 and Nf values increased related to reference mix. This enhancement is mainly attributed to filler effect and improved particle packing, which densify the microstructure and enhance crack resistance. Moreover, pozzolanic activity of fine glass particles contributes to strengthening the cement matrix, resulting in improved energy absorption under repeated impact loading. However, at greater replacement levels (30–40%), a decline in impact resistance was saw, which can be linked to particle agglomeration, increased porosity, and decrease matrix cohesion. The beneficial effect of curing age was also evident in the GF mixtures, with both N1 and Nf generally increasing from 7 to 28 and 90 days. The improvement at later ages was particularly associated with the continued development of cementitious matrix and the contribution of fine glass to particle packing and possible pozzolanic reactions. Consequently, combined effects of matrix densification and fine-glass incorporation enhanced resistance to crack initiation and failure, particularly at moderate replacement levels. These results are in good agreement with recent studies indicating that optimal glass content ranges between 10–20%, beyond which mechanical and durability properties tend to deteriorate (Ahmed & Rana, 2023) (Ubeid et al., 2020).

Figure 21:

6.3.1. Water Absorption

The results of the water absorption resistance test for the control mix G0 and other mixes (GC and GF) at (10–40) % replacement ratios at the age of 7, 28 and 90 days are presented in Table 12 and illustrated in Figure 22.

Figure 22:

For all mixtures, the water absorption values decreased with increasing curing age, which can be attributed to the continued hydration process and the resulting refinement of the pore structure. However, the incorporation of coarse glass generally increased water absorption compared with the control mixture. At 7 days, the absorption of the control mixture was 9.26%, whereas the GC10, GC20, GC30, and GC40 mixtures recorded 9.48%, 9.82%, 9.39%, and 9.77%, respectively. These values represent increases of about 2.4%, 6.0%, 1.4%, and 5.5% relative to the control mixture. Although the results fluctuated with increasing GC content at this early curing age, all GC mixtures exhibited higher absorption than the control. At 28 days, the water absorption increased progressively from 8.29% for the control mixture to 8.41%, 8.86%, 9.11%, and 9.56% for GC10, GC20, GC30, and GC40, corresponding to increases about 1.4%, 6.9%, 9.9%, and 15.3%, respectively. A similar progressive trend was observed at 90 days, when the control mixture recorded an absorption of 7.32%, compared with 7.47%, 7.68%, 7.94%, and 8.22% for the respective GC mixtures. These results correspond to increases of approximately 2.0%, 4.9%, 8.5%, and 12.3%. The increase in water absorption associated with coarse-glass incorporation, particularly at the higher replacement ratios, may be attributed to the smooth and relatively impermeable surface of the glass particles, which can weaken the bond with the surrounding cementitious matrix. Consequently, a less compact interfacial transition zone (ITZ), containing micro-voids and interconnected capillary pores, may develop and facilitate water penetration. Nevertheless, further microstructural observations and statistical analysis are required to confirm this interpretation. Furthermore, the increase in water absorption is strongly correlated with the observed reduction in compressive strength at higher replacement levels, confirming that the deterioration in microstructure adversely affects both mechanical and durability properties. Similar trends have been reported in previous studies. (Seddik Meddah, 2019) (Kuri et al., 2023).

In contrast, mixtures incorporating fine glass (GF) generally exhibited lower water absorption than the control mixture. At 7 days, the water absorption decreased from 9.26% for the control mixture to 8.65%, 8.37%, 8.89%, and 8.64% for GF10, GF20, GF30, and GF40, respectively. These values correspond to reductions of approximately 6.6%, 9.6%, 4.0%, and 6.7%. At 28 days, the control mixture recorded a water absorption of 8.29%, whereas the corresponding values for GF10, GF20, GF30, and GF40 were 7.92%, 7.80%, 8.11%, and 8.16%, representing reductions of approximately 4.5%, 5.9%, 2.2%, and 1.6%, respectively. Similarly, at 90 days, water absorption decreased from 7.32% for the control mixture to 7.12%, 7.04%, 7.26%, and 7.21% for GF10, GF20, and GF30, corresponding to reductions of approximately 2.7%, 3.8%, 0.8%, and 1.2% respectively.

The results indicate that a 20% fine-glass replacement ratio was the best level for reducing water absorption at all curing ages. The improved performance at low and moderate fine glass contents may be attributed to the filler effect and potential pozzolanic activity of the fine glass particles, which refine the pore structure and promote the formation of additional C–S–H gel, thereby producing a denser cementitious matrix. At higher replacement ratios, however, the beneficial effect became less pronounced, possibly because of particle agglomeration and the dilution of the cementitious material. These effects may limit the availability of calcium hydroxide for pozzolanic reactions and increase the volume of accessible pores. These findings are consistent with those reported by (Chen et al., 2022) (Roy, 2025).

6.3.2. Sorptivity

When the surface of hardened concrete is submerged in water or exposed to moisture, the concrete pores absorb water by action of capillary suction and the rate at which water is absorbed is called sorptivity. Durability, deterioration under wet-dry cycle and service lifetime of concrete can be related with its rate of sorptivity. The results of capillary water absorption at 28 days of curing are presented in Table 13 and Figure 23.

The sorptivity results demonstrate a clear dependency on both glass particle size and replacement level across all curing ages (7, 28, and 90 days). For mixtures incorporating coarse glass GC, a continuous increase in sorptivity was observed with increasing replacement ratios. At 7 days, sorptivity increased from 0.068 mm/s^0.5 in the control mix to 0.073, 0.079, 0.083, and 0.087 mm/s^0.5 at 10%, 20%, 30%, and 40% replacement ratios, corresponding to increases of approximately 7.4%, 16.2%, 22.1%, and 27.9%, respectively. A similar trend was noted at 28 days, where sorptivity increased by 9.6%, 19.2%, 40.4%, and 50% for the same replacement levels. On 90 days, the increase ranged from 15.2% to 39.1%. This behavior can be attributed to the weak interfacial transition zone (ITZ) and increased porosity associated with coarse glass particles, which reduces packing density and facilitates capillary water ingress, previous studies confirmed that (Kirane et al., 2025) (Jahami et al., 2025).

In other hand, mixtures containing fine glass (GF) exhibited a reduction in sorptivity at 10% and 20% replacement ratios, followed by a slight increase at30% and 40% replacement ratios. At 7 days, sorptivity decreased from 0.068 mm/s^0.5 to 0.064 and 0.061 mm/s^0.5 at 10% and 20% replacement, representing reductions of 5.9% and 10.3%, respectively, before increasing to 0.068 and 0.071 mm/s^0.5 at 30% and 40%. A similar trend was observed at 28 days, where reductions reached 7.7% and 11.5% at 10% and 20%, respectively. At 90 days, the sorptivity decreased by up to 17.4% at 20% replacement, indicating improved durability. This reduction may be associated with the filler effect and pozzolanic reactivity of fine glass particles, which refine the pore structure and enhance matrix densification. Particle packing, void filling, and refinement of the pore structure. Although finely ground waste glass may exhibit pozzolanic activity because of its amorphous silica content, this activity can reduce capillary pores, leading to lower water absorption rates. This finding agrees with previous studies (Roy, 2025) (Raman et al., 2025).

Figure 23:

6.3.3. Alkali-silica Reaction (ASR)

The results ASR presented in Table 14 and Figure 24

Figure 24:

The ASR expansion results indicate that the control mixture (G0) exhibited a high expansion value of 0.24%, exceeding the ASTM C1260 limit and confirming the presence of deleterious alkali–silica reaction. The incorporation of glass as a partial replacement of sand significantly influenced the expansion behavior depending on both the replacement ratio and particle size. When replacing fine aggregate by coarse glass (GC), the expansion remained unchanged at 10% replacement (0.23%), indicating no improvement, while at 20%, 30%, and 40% it decreased to 0.20%, 0.17%, and 0.14%, compared to the control mixture G0. However, although this reduction occurred, the expansion values of GC mixtures remained relatively high, indicating that coarse glass may act as a reactive aggregate. This may be attributed to the presence of internal microcracks within larger glass particles, which facilitate ASR gel formation and expansion, this finding agreement with (Rajabipour et al., 2010). When using fine glass as a fine aggregate replacement showed a more significant reduction in expansion, decreasing from 0.21% at 10% replacement to 0.18%, 0.15%, and 0.13% at 20%, 30%, and 40%, relative to the control. This pronounced reduction indicates that fine glass is more effective in mitigating ASR expansion, particularly at higher replacement levels, where the expansion approached the non-deleterious range. The improved performance of GF mixtures is mainly attributed to its finer particle size to produce dense CSH gel by pozzolanic reaction (pozzolanic activity), reduces calcium hydroxide content, and improves the microstructure, thus controlling ASR gel formation. These findings are in good agreement with previous investigations (Gökşen et al., 2023) (Huang et al., 2021) (Kioko, 2021), which confirmed that the ASR expansion is controlled by glass particles size, where coarse glass acts as a reactive aggregate while fine glass acts as a pozzolanic material able of mitigating expansion, especially at higher replacement levels. This mechanism provides a possible literature-based explanation; however, it was not directly confirmed in the present study.