Promoting the sustainable construction: A scientometric review on the utilization of waste glass in concrete

4.1.1 Reduction in landfills

One of the main waste sources in the globe is glass. Glass is a common substance in the world’s largest landfills and in the oceans, where it affects not just people but also marine life. Egypt produces over 3.45 million tons of glass waste per year, with 84% of it ending up in landfills [69]. Glass has a tremendous potential for use in concrete as a partial replacement for cement because of its pozzolanic qualities.

4.1.2 Reduction in environmental pollution

The large portion of the waste produced worldwide is made of glass. Basically, burning and melting glass to shape it into various jars and bottles will produce more pollution. According to Shekhawat and Aggarwal [73], environmental legislation is driving the increased availability supplementary aggregates for concrete made of WG. This is due to the impact of regulations governing containers, end-of-life vehicles, and waste electrical goods. When concrete contains GP in place of some of the cement, glass acts as pozzolan. In addition to increasing glass utilization, using glass in concrete will reduce pollution by preventing glass burning, which is good for the environment.

4.1.3 Improvement in basic properties of concrete

The WGP has a particle size that is smaller than cement. In comparison to regular cement concrete, smaller particles will help the concrete have less porosity, which will help the concrete have fewer cracks. Concrete’s mechanical strength will increase with the addition of GP with smaller particles. Khatib et al. [74] reported an increase in the compressive strength of concrete with a maximum observed enhancement of up to 10.0% while utilizing GP in place of cement. Concrete’s workability is also influenced by particle size; smaller particles will be more workable than typical cement concrete.

4.1.4 Resistance against sulfate attack

Concrete is significantly affected by sulfate attack. The sulfate attack can result in spalling, expansion, delayed ettringite formation, and cracking. According to Rashad [75], long-term chloride permeability performance was increased by the durability of concrete containing WG. However, concerns were raised regarding ASR. Over the course of the concrete’s existence, noxious chemical components like sulfides, sulfates, and alkalis provide a higher danger of ASR. It was noted that a suitable pozzolan can effectively mitigate ASR and consume lime, thereby significantly reducing efflorescence. GP can help concrete be more sulfate resistant than regular concrete by being used in the concrete-making process. The glass will withstand the sulfate attack because it is less permeable than cement and does not react with salt.

4.1.5 Resistance against ASR

When the mixture contains silica material that is reactive, an ASR occurs. ASR damages the structure and generates significant cracks. GP does not react with silica; hence it can help to keep the silica material from reacting with other substances. In a study by Du and Tan [76], a unique behavior of concrete was observed upon the addition of GP. When 60% of the cement was replaced with GP, the researcher discovered that the resistance to the ASR and water absorption dramatically increased.

4.1.6 Resistance against FT effect

When water is present in small voids in the concrete, a cycle of freezing and thawing causes the concrete to crack. Concrete cracks as a result of freezing and thawing’s expansion and contraction of the material. GP is very fine particle, which will absorb less water than cement. Concrete will be less vulnerable to the effects of freezing and thawing because of the smaller size of GP will lead to less voids. Furthermore, the study found that adding more WGP to concrete increases the FT resilience of plain concrete [77].

4.1.7 Improvement in hydration process

Concrete’s mechanical properties are crucial. The majority of researchers investigate a variety of approaches to enhance the concrete’s mechanical properties. Because it is smaller and absorbs less water than cement, the improved strength of the concrete will be aided by the use of GP in its manufacturing. As a result, there will be enough water available for the reaction to fully hydrate and develop its maximum strength. Based on hydration studies, investigations were made on the WGP’s pozzolanic reactivity through experimental analysis at different cement replacement levels (0, 15, 30, 45, and 60% by weight). The results demonstrated that up to 30% cement substitution, due to the pozzolanic reaction between glass particles and cement hydration products, the compressive strength of concrete was unchanged after 28 days [78].

4.2.1 Glass color

Glass is available in a wide range of forms and originates from numerous sources. Glass comes in a variety of colors as well. It can often be very difficult to remove its color without the help of chemicals. Different chemicals are used to remove the color of glass, but they are also costly and dangerous [81]. The properties of glass are also changed by its color. It is quite hard to use colored glass in concrete since it also reacts with some materials when a reaction occurs. Jiang et al. [82] have discussed how the color of the glass affects pozzolanic and ASR reactivity [83], durability of alkali-activated materials (AAMs) with GP [84], performance of glass fume in ultra-high-performance concrete (UHPC) [85], and use of GP in phase change materials and sintering products [86].

4.2.2 Glass sources

Many different household and other products are made from glass. Glass is not obtained from a single source in landfills or other locations. The majority of WGP is sourced from landfills as the non-recyclable fraction, because it is difficult to achieve the standards for quality in remanufacturing glass. Consequently, usually, this portion is abandoned and dumped in landfills [87]. Glass is gathered from a variety of places. When glass from various sources is used in the preparation of concrete, the concrete’s strength will be impacted. In certain areas, the concrete’s strength will be higher, while in others, it will be lower, which is risky for important structures [88].

4.2.3 Difficulty to work with glass

Glass is a very sharp material that easily cuts through anything. When glass is used in concrete, it is done so in the form of powder. Due to the high percentage of amorphous silica (50–70%) in large particle size glass aggregates, the ASR poses a serious concern [89] and sodium (8–15%) [90] in soda-lime glass [91]. Glass is crushed down to a fine powder to use it in concrete. When someone comes into contact with the smaller glass, it can hurt both humans and animals since it is so sharp. The GP behaves sharply like crystal, and working with it can be challenging [92].

4.5.1 Heat of hydration

The increase in initial temperature, which occurs when cement reacts with water, is known as the “heat of hydration.” The cementitious characteristics of the material, curing temperature, water-to-cement ratio, and material fineness all have an impact on the heat of hydration. The mechanical properties of normal concrete with WG varies at high temperatures. Despite this, little is known about how reactive glass (RG) affects the high-temperature performance of geopolymer UHPC, especially in light of the glass’s heat stability and lack of a defined melting point [117]. When GP is finer, it will absorb more water, changing the concrete’s water-to-cement ratio as a result. In this condition, concrete will have enough water for hydration because cement absorbs more water than GP. Due to GP’s chemical makeup and resistance to sulfates, making concrete with it will require less heat of hydration [118]. Thermal cracking in the concrete will be less likely to take place with less heat of hydration, and it will not affect the volume.

4.5.2 Volume stability

The volume of new concrete varies when it is settling up initially. The volume of the concrete increases during mixing, when more heat is produced, and it will shrink after 24 h due to heat loss. Early age shrinkage and drying shrinkage will increase the number of cracks and damage the concrete. In order to mitigate drying shrinkage, WG cullet was utilized as a substitute for natural aggregates [119]. There will be less volume change in concrete because GP does not produce as much heat as cement. By adding more GP to concrete, early age shrinkage and drying shrinkage cracks will be reduced.

4.5.3 Water–cement ratio

The term “water–cement ratio” refers to the mass ratio of cement and water used to build concrete. As the water–cement ratio is raised, the strength of the concrete will drop. However, as the water–cement ratio rises, the concrete becomes more workable. GP does not absorb much water; therefore, making concrete with less water results in stronger concrete. Due to its angular shape and decreased water–cement ratio, GP slightly decreases workability. Meena et al. [120] have written a research paper about partial replacement of cement. GP content in M30 concrete was 0, 2.5, and 7%. The water–cement ratio was 0.43 on each occasion. The highest compressive strength was 43.1 N·mm⁻².

4.5.4 Pozzolanic reaction of WGP

Due to the material’s inclusion of silica and alumina, a pozzolanic reaction occurs when moisture is added to a material that already has certain cementitious properties. Due to their pozzolanic behavior, concrete can be made partially without cement by using a variety of ingredients. According to the study, if the glass particle size is less than 38 µm, it exhibits favorable pozzolanic behavior, whereas between 38 and 75 µm, its pozzolanic activity decreases or becomes negligible. Furthermore, it was observed that finer glass particles result in higher concrete strength, particularly at later stages [121]. Fly ash (FA), SF, and granulate blast furnace slag are a few of these materials, all of which have demonstrated great performance when employed as cementitious materials. High levels of silica in GP indicate that it shows pozzolanic behavior. A high silica content will facilitate the initiation of the pozzolanic reaction when mixing GP. According to Parghi and Alam [122], due to the potential for the pozzolanic reaction to increase the density of the microstructure, the hardened mixture’s water absorption was observed to be decreased and its density increased with the addition of GP. Smaller GP will give better results because of the increase in pozzolanic reaction as the size of the GP decrease. The pozzolanic reactivity of GP is also affected by color. Due to its chemical composition and glass structure, the green color has a higher pozzolanic reaction. According to Li et al. [123], because of its pozzolanic feature and low thermal conductivity, GP was said to have had a key influence in improving the strength development and lowering thermal conductivity. This was possible because blocks physio-mechanical characteristics could be easily met.

4.6.1 Setting time

Concrete’s initial and final settings both take a certain amount of time to complete. The qualities of the concrete are still fresh during the initial setting phase, and they harden during the final setting time. The initial setting time takes around 45 min to mold the concrete in shape, and the final setting time takes about 10 h for the concrete to harden. Similar findings have been reported by Samarakoon et al. [124] where the addition of GP as a precursor in place of up to 30% FA led to an increase in the initial and final setting times. The water–cement ratio will increase when GP is added to concrete as a partial replacement for cement since glass needs less water to hydrate while other components in the concrete mix need more water. The initial setting time of concrete will be impacted by this. Both the initial and final setting times of concrete are increased by GP. Guo et al. [125] conducted a study, in which they concluded that when GP is used in the concrete, the reaction rate experienced reduction in polycondensation rate due to which the formation of activation products is slower. According to Lu et al. [126], the increase in setting time that resulted from the addition of GP, which increased the water content in the fresh AAM, was caused by the physical characteristics of the GP (Figure 23).

Figure 23

Influence of GP on setting time [127].

4.6.2 Workability

For usage in concrete, glass is ground into a fine powder. The fundamentally angular shape of the GP creates a link with other concrete components. GP makes concrete less workable because of its angular shape, but because it did not absorb as much water, there is more water in the mixture, making the concrete more workable. Si et al. [128] conducted a study, in which he evaluated the mini-slump test’s ability to assess the workability of AAMs created by substituting GP for metakaolin (MK) by 20%. As one property increases and one property drops, the workability of the concrete practically remains the same. Therefore, adding GP to concrete will not significantly change the workability of concrete. These results agree with that in the study by Samarakoon et al. [124] where using GP as the precursor instead of up to 30% FA led to an improvement in the workability of the AAMs of up to 80%. According to the study’s findings, fresh concrete became more workable with a higher GP content, as can be seen in Figure 24.

Figure 24

Influence of GP on slump flow [127].

As shown in Figure 25 and reported by Liao et al. [129], GP has been used up to 40% in the concrete flow of concrete. The flow has increased by 11, 22, 28, and 33%, respectively, when 10, 20, 30, and 40% GP has been used in concrete. These studies have concluded that when GP is used in concrete, the workability of the concrete improves as the GP quantity increases. In another study, Jiang et al. [130] utilized GP as a replacement for FA in AAMs by up to 30%. The study’s findings also point to an improvement in the workability of fresh AAMs with more GP. When compared to AAM made with only FA as the precursor, the flow increased by 3, 11, and 14%, respectively, when GP was used as 10, 20, and 30% replacement of the FA.

Figure 25

Influence of GP on workability [127].

4.7.1 Compressive strength

The force used to compress an object against it is known as its compressive strength. Because of the inclusion of coarse aggregate and the adherence of the constituent materials, concrete has a good compressive strength. Table 9 demonstrates the results of concrete samples using different amounts of GP in terms of compressive strength. The samples are labeled as “Sample,” and the table displays the compressive strength values in N·mm⁻² for each GP percentage. The results reveal the compressive strength values achieved when using 0, 5, 10, 15, 20, and 40% GP content. Notably, the data show that as the GP percentage increases, the compressive strength generally follows an increasing trend, with a few exceptions. The highest compressive strength is observed at 55.1 N·mm⁻² for a 15% GP content, while the lowest value is 24.22 N·mm⁻² for a 0% GP content. It is worth mentioning that some entries lack compressive strength values, indicated by dashes in the table. This table provides valuable insights into the effect of GP on the compressive strength of concrete, allowing researchers and practitioners to assess the performance of concrete mixes incorporating various amounts of GP. According to Sethi et al. [133], WGP can be added to the geopolymer matrix together with other source materials to strengthen the bonding between aggregates and binder in the transition zone. GP also helps in filling small voids present in the concrete. As the amount of GP in the sample grows, so does its compressive strength. In another research, Chen et al. [104] conducted the compressive test on the FA-type geopolymer binder by substituting WGP at various concentrations (10–40%). According to Liao et al. [129], increasing the WGP content by 30% boosts the compressive strength of the ground granulated blast-furnace slag (GGBS)-based geopolymer binder.

As shown in Figure 26, the graph shows how the amount of GP used in concrete and its effect on compressive strength are related to one another. The x-axis represents the various levels of GP replacement, ranging from 0 to 40%, the concrete’s compressive strength, measured in N·mm⁻², is shown on the y-axis. The graph clearly shows that adding GP to the concrete mixture results in a general boost in compressive strength. Compressive strength consistently and noticeably improves as the fraction of GP replacement rises from 0 to 15%. This indicates that incorporating GP in concrete positively influences its ability to withstand compressive forces. However, beyond the 15% threshold, the graph shows a decline in compressive strength with higher percentages of GP replacement. This decline suggests that there may be limitations or adverse effects associated with using higher amounts of GP in the concrete mix. It can be inferred from the graph’s tendencies that 15% of GP is the ideal amount to use in order to achieve the highest compressive strength. At this level, the concrete exhibits significant improvements in strength without experiencing a decline in performance. Utilizing a 15% GP replacement in concrete is thus recommended to maximize the compressive strength and ensure optimal performance.

Figure 26

Effect of GP on compressive strength.

4.7.2 Split tensile strength

Concrete’s split tensile strength is crucial for the concrete. The split tensile force is applied on the cylinder to split the cylinder into two parts. Split tensile force extends the size of the cracks in the structure. Basically, the concrete is weak in tensile strength and good in compressive strength. Sethi et al. [133] have shown an improvement in the split tensile characteristics of the geopolymer matrix when WGP was employed as a ternary source material as opposed to reference concrete made with FA and GGBS. Table 10 illustrates the concrete samples’ split tensile strength when GP was added in various amounts. The table includes different samples labeled as “Sample” and presents the split tensile strength values in N·mm⁻² for each GP percentage. The data demonstrate the split tensile strength values achieved at 0, 5, 10, 15, 20, and 40% GP content. The highest split tensile strength value recorded is 5 N·mm⁻² at a 15% GP content, while the lowest value is 2.32 N·mm⁻² for the 0% GP content. Some entries in the table are marked with dashes, indicating missing values. This table provides valuable insights into how the inclusion of GP influences the split tensile strength of concrete, aiding researchers and practitioners in assessing the performance of concrete mixes incorporating different amounts of GP. Hemanth et al. [139] have done a study from which they concluded that adding different WGP replacement percentages had an impact on the matrix of the FA-type geopolymer. The experimental investigations’ findings indicated that an improvement in split tensile strength of 3.3 MPa is produced by a 10% increase in WGP.

According to Figure 27, it is evident that adding GP to concrete results in a rise in split tensile strength. The graph shows the effect of the proportion of GP used in the concrete mix on the split tensile strength, which ranges from 0 to 40%. The graph shows that, compared to conventional concrete without GP, split tensile strength consistently and significantly improves up to a 15% replacement of GP. This indicates that the inclusion of GP positively affects the concrete’s ability to resist tensile forces. The graph, however, reveals a decrease in split tensile strength when the proportion of GP replacement rises above the 15% limit point. This decline suggests that there may be limitations or adverse effects associated with higher percentages of GP in the concrete mix. Comparing the GP-containing samples to the normal concrete, it is notable that the GP-containing samples generally exhibit higher split tensile strength. This finding suggests that adding GP to concrete can improve the concrete’s tensile strength, which is a good result for using GP in concrete applications. Overall, the data from Figure 27 demonstrate that incorporating GP in concrete can result in increased split tensile strength. It has been discovered that 15% of GP is the ideal amount for generating the highest split tensile strength. However, further investigations and considerations may be required to understand the potential limitations and effects associated with higher percentages of GP in concrete.

Figure 27

Effect of GP on split tensile strength.

4.7.3 Flexure strength

The flexure strength of the concrete is obtained by applying load on the beam sample. The tensile strength of unreinforced concrete is measured by the flexure strength of the concrete. The flexure strength is the beam resistance against bending. Sethi et al. [133] conducted a study in which they concluded that the flexural strength of a geopolymer matrix, made with ternary blends of WGP and GGBS as partial replacements for FA, demonstrated significant improvements. Table 11 displays the flexure strength of concrete samples containing varying percentages of GP. The table includes different samples labeled as “Sample” and presents the flexure strength values in N·mm⁻² for each GP content. The data represent the flexure strength achieved at 0, 5, 10, 15, 20, and 40% GP content. The highest flexure strength recorded is 5.76 N·mm⁻² at a 10% GP content, while the lowest value is 4.03 N·mm⁻² at the 0% GP content. Some entries in the table contain dashes, indicating missing values. This table provides insights into the influence of GP content on the flexure strength of concrete, offering researchers and practitioners valuable information for evaluating concrete performance when incorporating different levels of GP. Chen et al. [144], studied the impact of different WGP concentrations (10%–30%) on the flexural strength of FA-type geopolymers. According to them, adding more than 10% WGP negatively affected the mixture’s flexural strength.

Figure 28 shows that the flexural strength of concrete increases with the addition of GP. The graph shows how the amount of GP, which can range from 0 to 40% in the concrete mix, affects the flexural strength. The graph shows that, compared to conventional concrete without GP, there is a continuous and significant gain in flexural strength up to a 10% GP replacement. This indicates that the inclusion of GP positively affects the concrete’s ability to withstand bending or flexural stresses. However, beyond the 10% threshold, the graph shows a decline in flexural strength as the percentage of GP replacement increases. Similar to what was seen in earlier data for split tensile strength and compressive strength, this decline raises the possibility that increasing percentages of GP in the concrete mix may have restrictions or unfavorable impacts. Upon comparing the GP-containing samples to the normal concrete, it is evident that the GP-containing samples generally exhibit higher flexural strength. This finding indicates that incorporating GP in concrete is beneficial for enhancing the concrete’s ability to withstand bending forces, which is a positive outcome for the use of GP in concrete applications. Overall, the data from Figure 28 demonstrate that incorporating GP in concrete can result in increased flexural strength. The optimal percentage of GP for achieving the highest flexural strength is observed to be around 10%. However, it is important to consider the potential limitations and effects associated with higher percentages of GP in concrete to ensure optimal performance and durability.

Figure 28

Effect of GP on flexure strength.

4.7.4 Durability

Millions of tons of glass are being dumped in landfills, and the amount is rising extremely fast. Because glass has extremely slow rate of degradation and the millions of years it takes to decompose, disposal is one of the primary issues with glass. Glass can be recycled in concrete, which makes it a great option for use in concrete. The concrete industry has shown its ability to use glass in concrete, which will start to decrease the amount of waste going to landfills. By adding glass to concrete, the construction will become less dependent on natural aggregate and move closer to being sustainable. There are certain issues with replacing aggregate with GP, since the concrete’s durability will be reduced by the ASR. According to Maraghechi et al. [147], with 30% GP present, the compressive strength of up to 55 and 78 MPa, respectively, could be achieved after 3 and 28 days, although no overt ASR susceptibility was found. These results relate to mechanical and durability aspects. GP shows pozzolanic behavior, making it suitable for use as a cementitious material and a partial replacement for cement to some extent. Concrete having up to 15% GP has a substantially positive effect on compressive, split tensile, and flexure strength because of the pozzolanic characteristics of the GP and its ability to fill voids. According to various studies, in terms of durability, adding GP as an additional cementitious component works well.

Table 12 summarizes a number of studies looking into the impact of WG incorporation on concrete’s durability performance. The references cited in the table investigate various percentages of cement replacements with WG and analyze a number of effects, including oxygen permeability, porosity, electrical resistivity, chloride diffusivity, drying shrinkage, compressive strength, flexural strength, split tensile strength, modulus of elasticity, transport properties, chemical attack, ASR, FT, water absorption, chloride penetration, and sulfate attack. The main findings suggest that incorporating WG into concrete can lead to improved durability characteristics, including reduced permeability, enhanced resistance to chemical attack, ASR, FT cycles, and reduced water absorption. GP displays pozzolanic properties, especially with smaller particle sizes (<75 µm), which prevent the development of ASR gel and enhance durability and strength performance. Particle size, shape, color, and type are just a few variables that affect how much WG is to be replaced at any one time. Overall, the studies indicate that WG can contribute positively to the durability of concrete, forming denser microstructures, enhancing hydration, reducing permeability, and increasing strength.

The water absorption tendencies of ultra-high performance self-compacting concrete (UHPSCC) that incorporates different proportions of GP and lime powder (LP) replacements for cement are shown in Figure 29, and we see a strong tendency. As the content of GP and LP increases within the mix, a noticeable reduction in water absorption becomes evident when compared to the control mix. The conventional sample exhibits a water absorption rate of 1.151%, while the UHPC 4 mix (30% GP) and UHPC 8 mix (30% LP) remarkably show substantially lower water absorption percentages of 1.05 and 0.98%, respectively. This marks a significant decrease of 8.77 and 14.8% for UHPSCC 4 and 8 mixes when contrasted with the control mix [151]. The increased adhesion of the aggregate paste matrix and the density gain brought on by the addition of glass and lime powder are credited with this decrease in water absorption. It is important to highlight that the nano-particle size of the lime powder contributes to its role as a filler, effectively densifying the concrete and, in turn, reducing water absorption. In contrast, Kanellopoulos et al. [152] reported opposing results concerning recycled LP’s impact on water absorption. However, LP’s potential as a replacement filler in cement remains promising and calls for additional investigation.

Figure 29

Impact of water absorption on the durability of GP-based concrete after 90 days [151].

The assessment of sorptivity, a key indicator of concrete’s long-term durability, involves a technique where UHPSCC samples are partially submerged in bottles filled with distilled water. This evaluation revolves around capillary suction, influenced by the presence of capillary pores and their continuity within the concrete [151]. Sorptivity results for all UHPSCC combinations are shown in Figure 30, with values ranging from 6.5 to 4.1 g·cm⁻²·s^−0.5. Notably, when compared to the control, all UHPC combinations had much reduced sorptivity values. This observation is consistent with the findings of Dou et al. [153], underscoring the impact of nano glass on UHPC durability. The lowest sorptivity values among the mixtures are displayed by UHPSCC 4 (30% GP) and UHPSCC 9 (20% GP + 20% LP), which have sorptivity values of 4.1 and 4.2 g·cm⁻² s^−0.5, respectively. These numbers indicate a considerable decline of 36.9 and 35.3% from the control mix. The significant pozzolanic reaction caused by the presence of GP and SF, which effectively shrinks pores, is blamed for the decrease in sorptivity. Additionally, pore plugging, which increases the concrete’s resistance to water penetration, is greatly aided by the tiny LP particles, which act as fillers and nanoparticles.

Figure 30

Sorptivity’s influence on durability with GP and LP replacement ratios at 90 days [151].

4.8.1 EIS

An effective and non-destructive method for examining the electrical characteristics of materials, particularly cementitious composites made of glass, is electrical impedance spectroscopy (EIS). In the context of studying the microstructure of these composites, EIS provides valuable insights into their internal electrochemical behavior and can offer a deeper understanding of their durability and performance [164]. During EIS analysis, an alternating current is applied to the material at different frequencies, and the resulting impedance (resistance and reactance) is measured [165]. This impedance response is then plotted on a Nyquist plot [166] or a Bode plot [167], providing a clear representation of the material’s electrical behavior. In the case of glass-based cementitious composites, EIS can offer valuable information on several aspects of the microstructure. For instance, it can help assess the presence and extent of porosity, as different pore structures and sizes can affect the electrical conductivity of the material. A denser microstructure with fewer pores is expected to exhibit lower electrical conductivity compared to a more porous one [160]. Furthermore, EIS can provide insights into the electrochemical reactions occurring within the composite. It can be used to research the interactions between glass particles, cementitious matrix, and other elements as well as the hydration products’ production and characteristics, such as the calcium silicate hydrate (C–S–H) gel [168]. The durability and overall performance of the glass-based cementitious composites are greatly influenced by these parameters. Additionally, EIS is a helpful tool for tracking the long-term performance and stability of these composites in practical applications since it may assist detect any changes in electrical behavior over time. Overall, EIS is a valuable technique for characterizing the microstructure of glass-based cementitious composites. Its non-destructive nature and ability to probe the internal electrochemical behavior make it an essential tool for researchers and engineers seeking to optimize the microstructural design and enhance the performance and durability of these sustainable construction materials [110].

4.8.2 Pore size distribution

Pore size distribution is a critical aspect of the microstructure analysis for various materials, including glass-based cementitious composites. It refers to the characterization of pores in terms of their size [169], shape [104], and distribution within the material. Understanding the pore size distribution is essential because it directly influences several key properties and performance aspects of the composites. In the context of glass-based cementitious composites, pore size distribution plays a vital role in determining their mechanical strength [163], durability [169], permeability, and other important characteristics. A well-optimized pore size distribution can lead to enhanced properties, while an undesirable distribution may result in reduced performance. The mechanical properties of the composite, such as compressive strength and flexural strength, are often improved by a denser microstructure with a smaller average pore size. Smaller pores provide more efficient load transfer mechanisms, resulting in increased overall strength and resistance to mechanical forces. Additionally, the relationship between pore size distribution and material permeability is simple to understand [157]. A lower porosity and a narrow range of pore sizes can lead to reduced permeability, making the composite less susceptible to water ingress and chemical attack. This improved durability is especially crucial in structures exposed to harsh environmental conditions. Analyzing the pore size distribution can also help in understanding the potential for pore connectivity within the material. If larger pores are prevalent and interconnected, it may create pathways for the ingress of harmful agents, such as chlorides and sulfates, leading to possible deterioration of the composite over time. Researchers and engineers use various techniques to determine the pore size distribution in glass-based cementitious composites. Mercury intrusion porosimetry [159], gas adsorption methods [170], and image analysis [171] using scanning electron microscopy are some common techniques employed to characterize the pore structure. In summary, pore size distribution is a critical microstructural characteristic in glass-based cementitious composites. A crucial consideration in the design and development of high-performance and sustainable building materials is that the distribution of pores may considerably enhance the material’s mechanical strength, durability, and permeability features.

Figure 31 presents the pore size distribution of various components, including GP, lime stone (LS), SF, and MK. Notably, the particle size of SF is the smallest among these materials, followed by LS and MK, which have similar particle sizes. In comparison, GP exhibits a larger pore size distribution, a characteristic attributed to its particle size. This variation in pore size can significantly influence concrete properties and durability. The benefit of GP’s larger particle size lies in its potential to enhance interparticle packing, leading to improved overall density. Reduced pore connectivity in concrete mixtures also highlights the intriguing observation that the pore size of GP’s performance is comparable to that of OPC, highlighting its potential as an additional cementitious ingredient.

Figure 31

Comparison of particle size distribution with GP [172].

4.8.3 XRD analysis

An effective approach for examining the crystallographic structure and mineralogical makeup of materials, particularly glass-based cementitious composites, is the XRD analysis [173,161]. XRD is a non-destructive method that provides valuable information about the arrangement of atoms within the material and helps identify the phases present in the sample. In the context of studying the microstructure of glass-based cementitious composites, XRD analysis is particularly useful for understanding the cement hydration products and the interactions between cementitious materials and WG particles. Multiple crystalline phases and amorphous components may be identified in the composites using the diffraction patterns produced from the XRD examination [110]. XRD can detect various mineral phases in the cementitious matrix, including C–S–H [174], calcium hydroxide (CH) [175], and calcium aluminate hydrates (C–A–H) [176]. Additionally, it can detect the crystalline phases of WG, such as silica (SiO₂) [168] or other glass-forming oxides, which can undergo reactions with cementitious materials. XRD analysis is also useful in evaluating the degree of pozzolanic reaction in the composite. The appearance of new phases or shifts in diffraction peaks can indicate the formation of pozzolanic compounds, such as C–S–H gel or calcium–aluminum–silicate–hydrates (C–A–S–H), resulting from the interaction between the WG particles and cement hydration products. Moreover, XRD can provide information about the amorphous content in the material, which can significantly impact its performance. Amorphous components, such as glassy phases, may have an impact on the composite’s mechanical and durability qualities by adding to its pozzolanic activity. Researchers often utilize XRD analysis to assess the changes in the microstructure of glass-based cementitious composites over time, such as during the curing process or in response to environmental exposure. This helps in comprehending the material’s stability and long-term behavior in practical applications. In summary, XRD analysis is a useful method for describing the microstructure of cementitious composites made of glass. It provides critical information about the crystallographic structure, mineralogical composition, and pozzolanic reactions within the material [76]. By understanding the XRD patterns, researchers and engineers can optimize the microstructural design, tailor the composite’s properties, and develop high-performance and sustainable construction materials.

Figure 32 presents a comparative analysis of the properties of SF (a) and GP (b) through XRD analysis. The XRD spectra for both materials, designated as spectrum 8, spectrum 9, spectrum 23, and spectrum 24, provide insights into the composition of oxygen, calcium, carbon, silicon, aluminum, and magnesium. Specifically, spectra 8 and 9 correspond to SF, while spectra 23 and 24 correspond to GP. In spectrum 8, the composition percentages are 42.5% oxygen (O), 36.3% calcium (Ca), 9.7% carbon (C), 9.1% silicon (Si), 1.4% aluminum (Al), and 1% magnesium (Mg). Spectrum 9 exhibits composition percentages of 51.5% O, 22.5% Ca, 14.7% C, 8.1% Si, 1.4% Al, 1.1% Mg, and 0.6% carbon (C). Spectrum 23 showcases composition percentages of 45.4% O, 28.9% Ca, 9.6% C, 3.6% Si, 2.3% Al, and 1.4% Mg. In contrast, spectrum 24 illustrates composition percentages of 50.4% O, 21.1% Ca, 15.4% C, 11.3% Si, 1% Al, and 0.7% Mg. The data from this analysis highlight the elemental makeup of SF and GP, shedding light on their chemical composition and potential benefits for concrete performance.

Figure 32

X-ray diffraction analysis of (a) SF and (b) GP. The graphs characterize the chemical composition and structural properties of SF and GP [177].

4.8.4 RCP

RCP values, or rapid chloride permeability values, are a measure of the resistance of concrete, including glass-based cementitious composites, to the penetration of chloride ions [160]. Chloride ions can be particularly harmful to concrete structures as they can initiate and accelerate corrosion of embedded reinforcement, leading to structural deterioration and reduced service life. In chloride-rich settings, such as coastal locations or places exposed to deicing salts, RCP testing is a frequently used technique to evaluate the resilience and long-term performance of concrete [178]. The test involves subjecting the concrete specimen to an electrical potential, which drives chloride ions into the material [179]. By monitoring the rate of chloride ion penetration over a specified duration, the RCP value can be determined. In the context of glass-based cementitious composites, RCP values can provide valuable information about the material’s permeability to chloride ions [160]. A composite is less vulnerable to chloride-induced corrosion if it has a lower RCP value, which implies a stronger resistance to chloride penetration. This is particularly important in marine or harsh environments where chloride ingress can be a major concern for the durability of concrete structures. The incorporation of WG in the cementitious composite can influence the RCP values due to the effects on the microstructure. As glass particles contribute to a denser and less porous microstructure, the pathways for chloride ions to migrate through the composite are reduced, resulting in lower permeability and lower RCP values. Lower RCP values indicate that glass-based cementitious composites are less prone to chloride-induced corrosion, making them suitable for use in structures exposed to chloride-rich environments. By extending the service life of concrete buildings and decreasing maintenance expenses, this improved resistance to chloride penetration can increase sustainability. In conclusion, RCP values play a crucial role in assessing the durability of glass-based cementitious composites by evaluating their resistance to chloride ion penetration. The incorporation of WG can contribute to lower RCP values due to the positive effects on the material’s microstructure, making these composites promising candidates for sustainable and long-lasting concrete applications in challenging environments.

The RCP values of three distinct kinds of concrete – plain concrete, GP concrete, and FA concrete – are shown in Figure 33. These measurements were collected after 28, 56, and 90 days with varied curing times. Notably, plain concrete exhibits the highest RCP value among the three types, indicating its relatively lower permeability to chloride ions. The inclusion of GP in concrete brings forth a notable observation from these data. Despite plain concrete showcasing the highest RCP value, the RCP values of GP concrete demonstrate its potential to achieve chloride permeability levels comparable to or even better than that of FA concrete. This demonstrates how well GP works to reduce chloride ion penetration inside the concrete matrix, improving the concrete’s durability and resistance to chloride-induced corrosion. This suggests that GP, as an additional cementitious material, can efficiently contribute to the improvement of concrete’s resistance to chloride ingress, making it an important component for enhancing the long-term durability and performance of concrete structures, especially in corrosive environments.

Figure 33

Comparison of RCP values among plain concrete, GP-based concrete, and FA-based concrete [160].

4.8.5 NSSMCs

NSSMC is a parameter used to characterize the transport properties of ions [180], particularly chloride ions, within concrete materials, including glass-based cementitious composites. It is a valuable measure to assess the material’s resistance to chloride ingress and its overall durability performance in chloride-rich environments. In the NSSMC test, a concrete specimen is electrically potentiated, and the rate at which chloride ions diffuse through the material over time is observed. This test differs from the RCP test in that it examines the transport behavior of ions under non-steady-state conditions, representing more realistic conditions that occur in real-world scenarios. The NSSMC values provide insights into the effectiveness of the composite’s microstructure in mitigating chloride ion penetration. Lower NSSMC values indicate a higher resistance to chloride migration, signifying a denser and less porous microstructure that restricts the movement of chloride ions. The durability and service life of concrete buildings can be greatly impacted by chloride-induced corrosion, which can be prevented by this feature. The incorporation of WG in the cementitious composite can influence NSSMC values due to its impact on the microstructure [181]. Glass particles can enhance the material’s overall density, reduce porosity, and create a more tortuous path for chloride ions, leading to lower NSSMC values and improved chloride resistance. Researchers and engineers can better understand the transport characteristics of glass-based cementitious composites and how well they might function in chloride-contaminated settings by analyzing NSSMC values. Lower NSSMC values are indicative of a material’s enhanced durability and resistance to chloride ingress, making these composites suitable for use in infrastructure exposed to aggressive chloride-rich conditions. In conclusion, NSSMC is a crucial parameter for assessing the chloride resistance and durability performance of glass-based cementitious composites [160]. The integration of WG can positively influence NSSMC values by enhancing the microstructure and reducing chloride ion penetration, making these composites promising candidates for sustainable and resilient construction materials in challenging environments.

Three different forms of concrete – plain concrete, GP concrete, and FA concrete – have different NSSMC values, which are shown in Figure 34. These measurements were taken at various curing intervals of 28, 56, and 90 days. Notably, the NSSMC values reflect the rate at which chloride ions migrate through the concrete matrix, offering insights into the materials’ resistance to chloride penetration. Examining the information in this image reveals a huge advantage of using GP in concrete. Favorably, 20% GP concrete had the highest NSSMC value of all the concrete combinations, demonstrating that it has a higher level of resistance to chloride ion migration than other mixtures like plain concrete and 20% FA concrete. This shows that adding GP to concrete enhances the material’s capacity to obstruct the flow of chloride ions, increasing its longevity and lowering the likelihood of chloride-induced corrosion within concrete buildings. GP concrete has excellent performance in terms of NSSMC values, indicating that it has the potential to be a useful additive for extending the lifespan and service life of concrete in chloride-exposed situations, resulting in more robust and durable concrete buildings.

Figure 34

Comparison of NSSMC values among plain concrete, GP-based concrete, and FA-based concrete [160].

4.8.6 Enhanced microstructure

Enhanced microstructure refers to the improved arrangement and distribution of components within a material, such as glass-based cementitious composites, which leads to superior properties and performance compared to conventional materials. The concept of enhancing the microstructure involves optimizing the interaction and bonding between various constituents, resulting in a more refined and tailored composite with desirable characteristics [182]. Enhancing the microstructure of glass-based cementitious composites often involves utilizing WG in place of some of the cement or aggregates. The cementitious matrix may benefit from the inclusion of finely ground glass particles to achieve a more uniform distribution of elements and a denser, less porous structure. When WG is added, pozzolanic reactions may occur. In these reactions, glass particles combine with cement hydration products and water to generate new binding substances such as C–S–H gel. The interfacial transition zone between the glass particles and the cementitious matrix is strengthened and made more robust as a result of these processes, which improves the composite’s load-transfer mechanisms and overall mechanical performance [183]. Furthermore, the inclusion of WG in the microstructure can aid to boost packing density and minimize pore diameters, which improves resistance to the entry of hazardous substances like chlorides and sulfates and lowers permeability. This enhanced durability makes glass-based cementitious composites more suitable for applications in aggressive environments or structures exposed to harsh weather conditions. The optimized microstructure also influences the long-term performance of glass-based cementitious composites. It can reduce the potential for early-age cracking and enhance the material’s resistance to thermal stresses and environmental factors, ensuring the structural integrity and service life of the composite over time [184]. By carefully changing the amount and properties of WG and other cementitious materials, researchers and engineers hope to build glass-based cementitious composites with improved microstructure. By understanding the relationships between the composition, microstructure, and properties, they can develop high-performance and sustainable construction materials tailored to specific applications. In summary, enhancing the microstructure of glass-based cementitious composites involves improving the arrangement and interaction of components to create a denser, more durable, and better-performing material. The incorporation of WG plays a vital role in achieving these improvements, making glass-based cementitious composites promising candidates for sustainable construction solutions with enhanced properties and longer service life.

Figure 35a illustrates the effects of replacing sand with untreated GP, where the presence of silica initiates a reaction with hydrated cement, causing microcracks of approximately 3 µm width to develop. This reaction creates silica gel, which is highly attracted to water and causes swelling and pressure at the glass-cementitious matrix contact. The structure develops cracks as a result of this pressure. The problem is made worse by the rough surface roughness of the GP, which prevents effective pore-filling and reduces the strength of the concrete matrix. Additionally, the addition of GP causes a delay in the synthesis of C–S–H, potentially as a result of a slowed cement hydration process. The link between cement paste and glass particles becomes much weaker when considerable ettringite production starts to show after adding GP. On the other hand, Figure 35b shows that using GP that has been silane-treated improves the concrete’s resistance to fracture propagation. The increased adhesion between glass particles and cement paste made possible by the presence of silane is responsible for the decreased frequency of detected fractures. Through the formation of silanol groups, the smoother surface of glass particles treated with silane improves pore-filling and boosts bonding with all constituents of concrete. The internal toughness of the concrete is greatly increased as a result of this treatment, which creates a denser structure rich in C–S–H, particularly where the cementitious matrix and glass meet. Additionally, reducing ettringite forms, which increases the overall strength of the concrete, is another benefit of using GP that has been silane-treated. While using silane-treated GP increases adhesion, pore-filling, and strength, resulting in a denser structure rich in C–S–H and improved crack resistance, using untreated GP instead of sand causes the development of microcracks because of chemical reactions and reduced pore-filling.

Figure 35

Microstructural examination of (a) untreated GP concrete and (b) silane-treated GP concrete [185].

4.8.7 Interactions and binding mechanisms

The term “Interactions and Binding Mechanisms” in the context of glass-based cementitious composites refers to the intricate chemical and physical interactions that take place between the various components of the composite [186]. This includes the interactions between cementitious materials, WG particles, and other supplementary materials used in the mixture. Understanding the interactions and binding mechanisms is crucial as they directly influence the microstructure and overall performance of the composite material. These interactions can occur at both the macroscopic and microscopic levels, affecting properties such as mechanical strength, durability, and long-term stability. One of the key interactions is between the cementitious matrix and WG particles. Glass can experience pozzolanic interactions with the calcium hydroxide (CH) found in the cement when it is included in the cementitious composite. As a result, new binding substances are created, such as C–S–H gel, which adds to the material’s overall toughness and durability. Additionally, WG particles can act as nucleation sites for the cement hydration products, enhancing the development of the microstructure and improving the overall packing density of the composite [187]. The interactions between the glass particles and the cementitious matrix strengthen the bond and improve the material’s capacity for load transmission. The binding mechanisms within the composite can also influence the resistance to various forms of degradation, such as chemical attack and ASR [188]. Proper interactions between the glass particles and cementitious materials can lead to the formation of stable and less reactive compounds, reducing the potential for harmful chemical reactions and enhancing the material’s long-term performance. Moreover, the interactions between supplementary materials, like pozzolans or mineral admixtures, and the cementitious matrix can further contribute to the optimization of the microstructure. These substances can make it easier for cement hydration products and glass particles to stick together, which will increase performance and allow for more effective usage of cementitious substances. To achieve desirable interactions and binding mechanisms, researchers and engineers carefully control the composition, proportion, and characteristics of the components in the composite. By understanding the complex interactions between the constituents, they can design glass-based cementitious composites with enhanced properties and tailor them for specific applications, such as sustainable construction, infrastructure development, and building materials. In conclusion, the study of “Interactions and Binding Mechanisms” in glass-based cementitious composites is essential for understanding the chemical and physical relationships between the various components. These interactions play a key role in the creation of high-performance and environmentally friendly building materials because they have a considerable impact on the microstructure, mechanical characteristics, and durability of the composite.

4.8.8 Porosity and permeability characteristics

Porosity and permeability characteristics are vital parameters in the study of glass-based cementitious composites, as they directly influence the material’s durability, resistance to environmental factors, and overall performance [173]. Porosity refers to the volume fraction of voids or pores within the composite, while permeability refers to the ease with which fluids or gases can penetrate the material. Reducing porosity is essential for improving the mechanical characteristics of glass-based cementitious composites, such as compressive strength and tensile strength [189]. A lower porosity implies a more compact and denser microstructure, which enhances load transfer mechanisms and results in improved overall mechanical performance. The incorporation of WG particles can contribute to a reduction in porosity by promoting a more efficient packing of cementitious materials and filling up voids within the composite. On the other hand, permeability is a crucial factor in determining how well the composite resists the entry of hazardous elements like chloride ions or dangerous chemicals. Lower permeability indicates a reduced ability for these agents to penetrate the material, thereby enhancing its durability and resistance to degradation. By creating a denser microstructure with smaller pore sizes, WG particles can significantly lower the permeability of the cementitious composite. In addition to the macroscopic porosity, the pore size distribution within the composite is also essential [163]. A well-optimized pore size distribution can further enhance the material’s durability and mechanical properties. Smaller and more uniform pores contribute to improved resistance against chloride ingress and other forms of chemical attack, ensuring a longer service life for concrete structures. Researchers and engineers use various techniques, such as mercury intrusion porosimetry and image analysis, to evaluate the porosity and permeability characteristics of glass-based cementitious composites. These studies aid in the creation of high-performance composites for use in sustainable building applications and offer insightful information on the microstructure of the material. In summary, porosity and permeability characteristics are crucial factors in determining the performance and durability of glass-based cementitious composites. The incorporation of WG particles can lead to a reduction in porosity, creating a denser microstructure that enhances mechanical properties. Additionally, the reduced permeability contributes to improved resistance against harmful agents, making glass-based cementitious composites promising materials for sustainable and long-lasting infrastructure development.

4.8.9 Microstructure optimization for sustainable construction

Microstructure optimization for sustainable construction is a crucial approach aimed at creating high-performance materials with enhanced properties while minimizing their impact on the environment [85]. The aim of this optimization method is to increase the mechanical strength, durability, and eco-friendliness of the material, such as glass-based cementitious composites, by fine-tuning the arrangement, distribution, and interactions of components inside the material. In the context of glass-based cementitious composites, microstructure optimization involves several key aspects. The use of WG, a sustainable and environmentally beneficial element that can improve the composite’s overall performance, is one crucial component. Additional binding compounds are created as a result of pozzolanic interactions between the WG particles and the cement hydration products. As a result, the interfacial transition zone between the glass and cementitious matrix becomes denser and more robust, eventually enhancing the load transmission mechanisms. Another essential consideration in microstructure optimization is the reduction in porosity within the composite [190]. By promoting a more efficient packing of cementitious materials and filling up voids within the matrix, a lower porosity can be achieved. This leads to enhanced mechanical strength and resistance to degradation, making the composite more robust in various applications. Moreover, controlling the permeability of the composite is crucial to prevent the penetration of harmful substances, such as chloride ions or aggressive chemicals. A denser microstructure with smaller pore sizes can significantly lower the permeability, ensuring the composite’s long-term durability and service life. To further enhance microstructure optimization, supplementary materials like pozzolans or mineral admixtures can be added. These materials improve the interaction between the glass particles and cementitious matrix, leading to a more efficient use of cementitious materials and improved overall performance. Comprehensive performance testing is a fundamental part of microstructure optimization [159]. Through thorough evaluations of mechanical properties, durability characteristics, and resistance to environmental factors, engineers and researchers ensure that the composite meets the required performance standards for sustainable construction. By optimizing the microstructure of glass-based cementitious composites, we can create materials that offer superior mechanical strength, enhanced durability, and reduced environmental impact. By increasing the service life of infrastructure, lowering maintenance requirements, and encouraging the effective use of waste materials in construction, these composites have the potential to make a substantial contribution to sustainable building methods.

4.9.1 Linear regression (LR) model

LR model applied to GP in concrete predicts how changes in GP percentage influence concrete properties like strength and durability. It establishes a linear relationship between these variables using experimental data. This model provides insights into the optimal GP range for desired concrete performance, aiding in sustainable mix design. However, it assumes a linear correlation and should be used alongside other methods for a comprehensive understanding of GP’s effects on concrete.

Eq. (1) represents the thorough examination of the variables affecting the compressive strength of typical concrete mixes containing WG dust. This complex inquiry was carried out using optimization strategies, including the use of least squares and sum of error squares procedures. Utilizing the Solver tool in an Excel software speed up the procedure. Each of the parameters in the equation is important: “w/b” stands for the water-to-binder ratio, “C.C.” represents the cement content, “G.P.” stands for glass powder content, “F.A.” stands for fine aggregate content, “C.A.” stands for coarse aggregate content, and “T” stands for time or a related time-dependent variable. The equation effectively captures the interplay of these parameters and their collective impact on the compressive strength of concrete, offering valuable insights for the optimization of WG-incorporated concrete mixtures. This optimization was conducted in a designated cell termed the “objective cell,” bounded by values from other relevant equations. The connection between anticipated and actual compressive strengths of regular concrete mixes incorporating WG particles is shown graphically in Figure 36, both for training and testing datasets. Statistical methods were used to do a thorough evaluation of the model, which resulted in R ² values of 0.918 for training data and 0.923 for testing data. While the mean absolute error (MAE) was computed as 2.41 MPa (training) and 2.58 MPa (testing), the root mean square error (RMSE) was 3.07 MPa (training) and 3.28 MPa (testing). The structural index (SI) value indicated good efficiency within the range of 0.1–0.2, and the objective (OBJ) value was equal to 6.039. Furthermore, the model’s application is uncomplicated; however, it does exhibit a limitation, displaying slightly reduced predictive capability for compressive strength with higher error percentages compared to alternative models. Both the training and testing datasets in the study have an error margin of about ±25%. The compressive strength of low-strength normal concrete mixes tended to be somewhat overestimated by the model under development, whereas samples of high-strength normal concrete were found to have compressive strengths that were slightly underestimated.

Figure 36

LR model [203].

4.9.2 Nonlinear regression (NLR) model

NLR model applied to GP in concrete predicts concrete properties considering the complex, nonlinear relationship between GP percentage and outcomes. Unlike linear models, it captures intricate effects and offers accurate insights into real-world behavior. This model is particularly useful for materials like GP, where effects on concrete may not follow a linear trend. However, it requires careful parameter estimation and validation to prevent overfitting. It is a valuable tool for optimizing concrete mix designs, providing deeper understanding of how GP levels impact properties and aiding sustainable construction practices.

For both the training and testing datasets, connections between predicted compressive strength and experimentally acquired compressive strength data from typical concrete combinations are highlighted. The suggested equation for the NLR model combining several variable factors is shown in Figure 37. Eq. (2), generated through thorough analysis, produced outstanding results using 110 data points for training and 43 for testing, with an amazing R ² of 0.921 for training data and 0.929 for testing data. While the MAE data showed accuracy with 2.34 MPa (training) and 2.43 MPa (testing), the RMSE values showed precision with 3.022 MPa (training) and 3.159 MPa (testing). Notably, the equation’s parameters hold distinct importance: “w/b” represents the water-to-binder ratio, “C.C.” symbolizes the cement content, “G.P.” represents the GP content, “F.A.” represents the fine aggregate content, “C.A.” denotes the coarse aggregate content, and “T” represents time or a relevant time-dependent variable. The resultant OBJ value of 5.1 underlined the model’s efficacy, while the SI value, which fell between 0.1 and 0.2, underlined successful performance. Importantly, the assessed datasets showed error margins of –15 and +20% for both testing and validation datasets, and a ±20% error line for training data. Analogous to the LR model, the NLR model exhibited a slight underestimation for high-strength normal concrete mixes, accompanied by a moderate overestimation for low-strength normal concrete samples. The equation’s multifaceted parameters collectively contribute to a comprehensive understanding of their impact on predicting the compressive strength of concrete infused with WGPs.

Figure 37

NLR model [203].

4.9.3 ANN

ANN mirrors the human brain’s neural structure and are pivotal in deciphering intricate material behaviors within glass-based cementitious composites. By amassing extensive experimental data encompassing glass content, curing conditions, and mechanical and durability traits, researchers train the ANN to grasp the nuanced links between input factors and material properties. Comprising interconnected neurons organized in layers, the ANN processes input data through mathematical transformations, learning from patterns and refining connection weights to boost predictive accuracy. Trained ANNs exhibit potent predictive capabilities, projecting the mechanical strength, durability, permeability, and more of novel glass-based compositions. They unveil complex interactions between factors, elucidating how composition adjustments impact performance. This nonlinear aptitude enables ANNs to capture intricate relationships often evading traditional linear models. ANNs continually enhance predictions as they learn from fresh data, ensuring ongoing refinement. Nonetheless, their efficacy hinges on substantial data and careful parameter tuning to avert overfitting. Armed with adaptability and computational prowess, ANN stands as a potent asset for predicting and optimizing material properties in glass-based cementitious composites, significantly advancing sustainable and high-performance construction materials.

In order to maximize the effectiveness of the ANN, several combinations of neurons, momentum rates, learning rates, and iterations were investigated. The results showed that the configuration of a single hidden layer with 9 neurons, 0.2 momentum, 0.1 learning rate, and 50,000 iterations produced the best precise predictions for the compressive strength of regular concrete mixes incorporating WG granules. The ANN model was calibrated using training datasets and subsequently validated against independent testing datasets to forecast compressive strength values based on appropriate input parameters. Comparing projected and experimentally determined compressive strength values for regular concrete mixes containing WG particles was part of the assessment, which took into account both training and testing datasets. Depicted in Figure 38 is a graphical representation that showcases the outcome of data analysis. Specifically, this analysis pertains to the accuracy of predictions made by the ANN model in comparison to other models. The data points in the graph are indicative of the extent to which the predicted values align with the actual values. Importantly, for both the training and testing datasets, it is noted that the error line, which reflects the difference between projected and actual values, is considerably less than ±25%. According to this observation, the forecasts of the ANN model are noticeably closer to the actual values than the predictions of the other produced models. It means that, practically speaking, the ANN model is more accurate in predicting the compressive strength of regular concrete mixes that contain WG particles. Following this study, more assessments are made in order to fully gauge the effectiveness of the ANN model. These evaluations are guided by specific metrics that are widely utilized in predictive modeling. The MAE, which measures the average magnitude of absolute errors, the RMSE, which measures the average magnitude of prediction errors, and the coefficient of determination (R ²), which measures the percentage of the variance in the dependent variable that can be explained by the independent variables are some examples of these metrics. In essence, the graphical representation in Figure 38 provides an initial visual insight into the relative accuracy of the ANN model’s predictions, emphasizing its favorable performance in comparison to other models. Subsequent quantitative assessments through metrics such as RMSE, MAE, and R ² offer a more detailed and precise understanding of the model’s predictive capabilities and its overall fit to the observed data.

Figure 38

ANN model [203].

The efficacy of the ANN model in predicting the attributes of GP-based concrete is intricately analyzed in Table 13. The table systematically explores various configurations of hidden layers and neurons within those layers, detailing the resulting performance metrics. The first column indicates how many hidden layers there are in the ANN architecture, and the next two columns show how many neurons are located on the left and right sides of each hidden layer. The quantitative metrics of the model’s prediction precision are shown in the following columns. The coefficient of determination, denoted by the “R ²” column, measures how well the model’s forecasts match the data. Higher R ² values signify a stronger correlation between predictions and actual values. The “MAE” column represents the MAE, which quantifies the average magnitude of prediction errors, as opposed to this, the “RMSE” column displays the root mean square error, which calculates the average magnitude of prediction mistakes and gives larger errors more weight. As we analyze the table, we observe trends in the performance metrics based on the different configurations. For instance, the R ² values range from approximately 0.9696 to 0.9892, indicating a relatively high degree of correlation between predictions and actual values across various settings. Similarly, the MAE and RMSE values provide insights into the average error magnitudes, with lower values indicating better predictive accuracy. In summary, Table 13 serves as a comprehensive guide to understanding how different combinations of hidden layers and neurons impact the ANN model’s ability to predict properties of GP-based concrete. Researchers can use this table to select optimal configurations that yield the highest levels of accuracy and precision in predictions, enhancing the model’s reliability and utility in practical applications.

4.9.4 SVMs

SVMs are robust machine learning models used to predict material behaviors in glass-based cementitious composites. Employing a classification algorithm, SVM determines an optimal hyperplane for segregating data points into distinct classes. In the realm of these composites, SVM proves valuable for forecasting whether a composite meets specific performance criteria based on its composition and microstructure. Researchers gather experimental data on diverse glass-based cementitious composites, encompassing attributes like glass content, cement type, curing conditions, and mechanical measurements. SVM endeavors to identify a hyperplane that maximizes separation between classes, ensuring robustness and precision. Trained SVMs classify new composites based on input features, estimating factors such as load-bearing strength and durability against environmental effects. SVM’s adaptability accommodates various classification tasks, enabling exploration of different performance standards. Its capability to manage multiple features facilitates analysis of intricate variable-property relationships. Notably, SVM accuracy hinges on quality training data, necessitating continuous data collection and model refinement. With the ability to inform material design and enhance sustainable construction, SVMs and similar machine learning models are instrumental in propelling innovation and bolstering construction material efficiency and resilience.

The compressive strength of concrete cubes has been predicted using the SVM approach. Several statistical indicators, including the coefficient of correlation (CC) and RMSE, were used to evaluate the efficacy of this strategy. The coefficient of correlation is shown in Eq. (3) as CC. In a dataset of “n” paired observations, this statistical measure evaluates the magnitude and direction of the linear connection between two variables, denoted as “x” and “y,” inside the dataset. The equation involves calculating the covariance between the deviations of the paired observations from their respective means, and then normalizing it by the product of the standard deviations of the two variables. With values ranging from −1 to +1, CC assesses how closely the different permutations of “x” and “y” align. Greater values of “x” correlate with greater values of “y,” and vice versa. A positive number denotes a positive connection. Conversely, a negative value indicates an inverse correlation, while a value close to zero suggests a weak or no linear correlation between the variables.

RMSE, sometimes known as Eq. (4), is a statistical concept. It serves as a gauge for the average size of the discrepancies between observed (“x”) and projected (“y”) values. This metric provides insight into the overall accuracy of a predictive model or method. The equation involves squaring the differences between observed and predicted values, averaging these squared differences, and then taking the square root of the result. RMSE gives a sense of how well a predictive model’s outcomes align with actual observations. Lower values of RMSE indicate a smaller average prediction error, reflecting a higher level of accuracy in the predictions. In the context of assessing predictive models for concrete properties, like compressive strength, RMSE provides a quantitative measure of the model’s performance, enabling researchers to gauge how closely their predictions match the real-world data.

In this case, x _i and y _i stand for the experimental and anticipated compressive strengths of the WGP concrete, respectively. n denotes the total number of data points included in the model, and x ̅ i and y ̅ i stands for the mean experimental and projected compressive strength, respectively.

4.9.5 Comparison of models

The comparison of different prediction models for glass-based cementitious composites highlights their diverse approaches for forecasting material properties. These models play a pivotal role in optimizing composite design, production, and application, ultimately contributing to sustainable construction practices. Machine learning models, statistical analyses, and computational simulations are commonly utilized to develop prediction models. The effectiveness of three well-known machine learning models, SVMs, RF, and ANN, in predicting the strength of RAC including WG particles was specifically examined in this study.

Linear regression model offers insights into the linear relationship between GP percentage and concrete properties. It employs optimization techniques to determine parameter influence on compressive strength, optimizing error squares via Solver in Excel. The relationship between expected and actual compressive strengths for typical concrete mixes with WG particles is shown graphically in Figure 36. R ², RMSE, and MAE are evaluation measures that show how well the model performs. For both training and testing data, the LR model displays an error line of around ±25%, slightly underestimating high-strength concrete and overestimating low-strength concrete. NLR model captures the complex, nonlinear relationship between GP content and concrete properties. Figure 37 showcases the model’s equation and its accuracy evaluation. The analysis reveals an error line within ±20% for training data and error margins of –15 to +20% for testing and validation datasets. This model provides accurate insights into real-world behavior. ANNs simulate the human brain’s structure to predict material behaviors. Trained on extensive experimental data, ANN uncovers intricate links between input parameters and material properties. ANN’s architecture, training process, and predictive power are detailed. Figure 38 illustrates a comparative analysis of ANN’s predictive accuracy against other models. The error line is notably smaller than ±25%, indicating higher predictive accuracy. Table 13 further dissects ANN’s predictive performance by exploring various configurations of hidden layers and neurons. The table provides a comprehensive view of R ², MAE, and RMSE values for different setups, aiding researchers in selecting optimal configurations for accurate predictions. SVMs employ a classification algorithm to predict material behaviors. SVM determines an optimal hyperplane for classifying data points. Statistical indicators like the CC and RMSE are used in SVM’s application for the prediction of compressive strength. Based on experimental and predicted results for WGP concrete compressive strength, these measures determine the prediction accuracy.

The ANN exhibits superior dependability in comparison to the other proposed models, according to the computed R ², MAE, and RMSE values shown in Figures 39–42. Specifically, the R ² value in the ANN model surpassed that of the regression model by 6% and exceeded the NLR by 5.7%. Compared to the regression model and the NLR, the RMSE value in the ANN model was significantly lower at 91.8 and 88.8%, respectively. Similar to how the MAE value in the ANN model significantly improved, being 87.2% lower than the NLR and 92.8% lower than the regression model. The ANN model’s objective (OBJ) value was 95.8 and 89.6%, respectively, less than those of the LR and NLR models. Additionally, as shown in Figure 43, the SI value of the ANN model was within the desirable range of 0–0.1, suggesting outstanding performance.

Figure 39

R ² value for three proposed models [203].

Figure 40

RMSE value for proposed models [203].

Figure 41

MAE value for proposed models [203].

Figure 42

OBJ value for proposed models [203].

Figure 43

SI value for proposed models [203].

In summary, these models offer distinct approaches for predicting material properties in glass-based cementitious composites, catering to different complexities and nuances of the relationship between input parameters and material behavior. Researchers can leverage these models to make informed decisions during composite design, ensuring that the performance meets specific criteria while advancing sustainable construction practices.

4.10.1 Applications

Glass-based cementitious composites offer a wide range of applications in sustainable construction, showcasing their versatility and potential to revolutionize the industry. One of the main uses is the creation of high-performance concrete, where these composites improve the mechanical characteristics, such as compressive and flexural strength, which makes them perfect for load-bearing components in crucial infrastructure [204]. Additionally, their use as sustainable construction materials stands out, as they help alleviate the burden of glass waste disposal and reduce reliance on traditional cement, contributing to a greener environment. The enhanced durability of glass-based cementitious composites makes them suitable for aggressive environments like marine structures, where they resist chemical attack and ASR more effectively [108]. Moreover, incorporating these composites in the manufacturing of lightweight and insulating panels improves building energy efficiency and reduces carbon emissions. Their compatibility with prefabricated elements streamlines construction processes, reducing time and costs. As research and technology continue to advance, new applications are continually being explored, solidifying the position of glass-based cementitious composites as an innovative and sustainable solution in modern construction.

Table 14 provides an overview of the applications of glass-based cementitious composites in sustainable construction. Each application is described, along with its associated advantages and real-world projects where these composites have been used. High-performance concrete utilizes these composites to enhance mechanical properties in load-bearing structures like high-rise buildings and bridges, improving structural integrity. They are also employed as sustainable construction materials, utilizing WG to promote resource conservation and reduce environmental impact in projects such as green buildings and eco-friendly developments. Glass-based cementitious composites exhibit enhanced durability, making them ideal for use in coastal structures and wastewater facilities, where they resist chemical attack and environmental factors, prolonging the service life of these structures. Incorporating these composites in lightweight and insulating panels improves energy efficiency and reduces carbon emissions in buildings, while their application in prefabricated elements streamlines construction processes, leading to faster and cost-effective construction, particularly in modular construction projects. Overall, Table 13 showcases the versatility and potential of glass-based cementitious composites to address various construction needs, contributing to greener and more resilient infrastructure. The real-world projects mentioned demonstrate the practicality and success of using these composites in actual construction scenarios, emphasizing their growing importance in sustainable and high-performance building practices.

4.10.2 Challenges

The adoption of glass-based cementitious composites in sustainable construction presents several challenges that need careful consideration and innovative solutions [210]. It is crucial to ensure the performance and durability of these composites over the long term, which necessitates careful study and testing to determine how they respond to a range of loading situations and environmental factors. Standardization and the development of codes specific to these materials are crucial to ensure consistent quality and safety standards. Handling and processing WG during manufacturing can be complex, necessitating the development of efficient and cost-effective production processes. Maintaining quality control is essential to prevent variations in material properties [211]. The initial cost of glass-based cementitious composites may pose a barrier to their adoption, making it imperative to demonstrate their long-term benefits and cost-effectiveness. Integrating these composites into existing design practices may require adjustments, and promoting public perception and acceptance of new materials is essential. Additionally, establishing a robust recycling and circular economy infrastructure is vital to ensure a consistent supply of WG. Continued research and development efforts will further enhance their performance and sustainability. The integration of glass-based cementitious composites into mainstream construction will be speeded up by addressing these issues through cooperation between researchers, industry experts, and policymakers, leading to the development of future infrastructure that is more environmentally friendly and robust [212].

Table 15 provides a comprehensive overview of the challenges associated with the adoption of glass-based cementitious composites in sustainable construction. Every challenge is explained, stressing the most pressing issues that must be resolved to enable the broad application of these cutting-edge materials. As it influences the belief in the material’s dependability over time, the first difficulty highlights the significance of assuring the long-term durability and performance of glass-based cementitious composites. Rigorous testing and real-world monitoring are essential to assess its behavior under various environmental conditions and loading scenarios. Another significant challenge lies in establishing standardized testing methods and codes specific to these composites, ensuring consistent quality and safety standards. Collaborating with regulatory authorities will aid in developing and implementing these standards effectively. The complexities involved in handling WG during the manufacturing process pose challenges to production efficiency. Developing specialized equipment and processes will streamline the handling and processing of WG. Maintaining consistent quality during production is vital, as material properties may vary. Robust quality control measures are necessary to ensure that the final product meets desired performance standards. Additionally, the higher initial cost of glass-based cementitious composites compared to conventional materials may deter adoption, particularly in budget-constrained projects. Convincing stakeholders of the long-term cost benefits of these composites becomes crucial. Integrating these composites into existing design practices requires collaboration between designers and engineers to fully capitalize on their unique properties. Moreover, public perception and acceptance of new materials incorporating WG can significantly impact their market demand and adoption. Educating stakeholders about the benefits and sustainability aspects of these composites can help overcome resistance to change. Ensuring a consistent supply of WG for production is crucial for the continuous use of these composites, promoting recycling infrastructure development will help maintain a reliable source of WG. Continual research and development are essential to improve the performance and sustainability of glass-based cementitious composites, unlocking their full potential. Finally, transitioning the construction industry to adopt new materials and practices requires significant investment and coordination among stakeholders. Infrastructure development and training programs are vital for successful integration. The potential of glass-based cementitious composites for greener, more resilient, and sustainable infrastructure for the future will be unlocked by addressing these issues through cooperative efforts among researchers, industry experts, and policymakers.