Utilization of Industrial Waste as a Sustainable Approach for Enhancing the Properties of Ternary Blended Concrete Prediction with Machine Learning

The construction sector is widely acknowledged as a major source of global carbon dioxide (CO₂) emissions, largely arising from the high energy consumption associated with cement manufacturing and the extensive exploitation of finite natural resources. Mitigating this environmental impact requires a comprehensive strategy that includes reducing material demand, promoting alternative and sustainable materials, and adopting efficient carbon management practices. Consequently, the incorporation of environmentally responsible and sustainable approaches in construction has become imperative, not only to lower the sector’s carbon footprint but also to support global efforts toward climate change mitigation. In this context, one of the notable developments in concrete technology in recent years is the emergence and increasing use of self-compacting concrete (SCC). SCC is characterized by its exceptional flowability and resistance to segregation, enabling it to spread and consolidate under its own weight without the need for external vibration [1]. This advanced concrete technology significantly enhances construction efficiency, delivers superior surface finish, and minimizes labor requirements as well as equipment usage. Owing to its ability to uniformly flow through intricate formwork and densely reinforced sections, SCC is particularly well suited for complex structural applications.

However, the workability of SCC can be adversely affected by the inclusion of fibers, which, while beneficial for enhancing mechanical properties and crack resistance, tend to reduce flowability and increase the risk of blocking [2]. Therefore, a balanced mix design is crucial to maintaining the rheological performance of SCC while achieving the desired structural benefits.

To achieve the required performance in the fresh state, SCC must satisfy a set of well-defined rheological and flow-related criteria. Key parameters such as slump flow, V-Funnel flow time, U-box height difference, and L-box blocking ratio are evaluated in accordance with the guidelines established by the European Federation of National Associations Representing Concrete (EFNARC). These standardized tests are essential for verifying the filling ability, passing ability, and resistance to segregation of SCC, thereby confirming its effectiveness for self-compacting applications.

Once hardened, SCC demonstrates mechanical strength and durability characteristics that are comparable to, and in many cases superior to, those of conventionally vibrated concrete. This improved performance is primarily due to the formation of a dense and uniform microstructure, which enhances load-bearing capacity and long-term durability [3]. In addition, the environmental impact of concrete can be significantly reduced by partially substituting Portland cement with pozzolanic industrial byproducts such as fly ash, silica fume, and ground granulated blast furnace slag (GGBS). The incorporation of these supplementary cementitious materials not only conserves natural resources and minimizes industrial waste but also improves durability-related properties of concrete [4, 5]. Such practices align with circular economy concepts and contribute effectively to lowering the carbon footprint associated with cement production.

By substituting a portion of the cement with a substantial quantity of waste material, concrete producers may realistically decrease their industry's adverse ecological impact and boost cost effectiveness. The consideration has been provided to the usage of the unused additives to boost mechanical properties of concrete. It is apparent that fly ash and Alccofine can be used in place of cement in concrete. To achieve zero net carbon emissions by 2050 is one of the objectives of Sustainable Development Goal 13 (climate action) [6]. In accordance with experimental research, adding more copper slag to novel materials strengthened concrete characteristics. At 50%, the industrial waste by product copper slag exhibited increased mechanical qualities for high-strength concrete (HSC) with suitable consistency and would be a viable alternative for sand [7]. Mineral admixtures are essential for overcoming cost constraints in manufacturing. The hardened concrete is dense, and solid and has constant engineering properties and sturdiness as that of conventional concrete [8]. Commuting the fine combination with copper slag associate degree industrial by-product helps with the increase of strength and sturdiness parameters. Inclusion of silica fume refines the pore structure resulting in stronger concrete. Silica fume has an exceedingly specific surface area and functions as a reactive pozzolona. Silica fume is a highly prospective mineral addition, particularly for ensuring high-strength to ultrahigh-strength concrete, with suitable particle dispersion ensured during mixing [9]. The utilization of Alccofine 1203 and silica fume will increase the strength (both in compression and flexure) to very good extent [10, 11]. The incorporation of Alccofine 1203 and silica fume in SCC positively impacts its engineering properties, making it a cost-effective and sustainable option. As the overuse of natural resources like river sand contributes to environmental concerns, this study examines the potential of copper slag as an alternative fine aggregate. Cement was partially replaced with 15% Alccofine and 20% silica fume, while copper slag was used in varying proportions to replace fine aggregate. The fresh and hardened properties of the resulting SCC mixes were evaluated. Results confirm that the use of industrial by-products enhances compressive strength and supports sustainable construction practices [12, 13]. The results derived from the experiments were compared with the predicted values obtained from Support Vector Machine (SVM) modelling [14-16]. The experimental data indicate that the optimal replacement level of fine aggregate with copper slag is 20%, leading to a 15.25% increase in strength [17-19]. Copper slag has attracted significant attention as a sustainable industrial by-product due to its potential for recycling and value-added applications. Adilov et al. [20] analyzed the challenges and opportunities associated with copper slag recycling, emphasizing regulatory barriers, variability in slag composition, and the need for integrated processing routes to enable large-scale sustainable utilization. Their study highlighted copper slag as a promising secondary resource within the circular economy framework. The environmental implications of copper slag reuse have been examined with a focus on heavy metal leaching and construction applications. Rathee and Misra [21] investigated the leaching behavior of copper slag and its feasibility as a replacement for natural fine aggregate in geopolymer concrete, demonstrating acceptable environmental performance and improved sustainability when proper mix design controls are applied. These findings support the safe incorporation of copper slag in cementitious systems. A comprehensive review by Klaffenbach et al. [22] provided a critical analysis of copper slag utilization pathways, including construction materials, metal recovery, and advanced processing techniques. The review emphasized the importance of holistic life-cycle assessment and techno-economic evaluation to ensure environmentally responsible and economically viable applications. Several studies have focused on the optimization of copper slag use in concrete and geotechnical applications. Turkane et al. [23] applied response surface methodology (RSM) to optimize concrete mixes incorporating copper slag as a fine sand replacement, reporting enhanced mechanical properties at optimal replacement levels. Similarly, Yeo et al. [24] demonstrated the effectiveness of RSM in optimizing concrete paving block mixtures, underscoring the robustness of statistical optimization techniques for sustainable material design. Beyond concrete, Zhang et al. [25] explored the use of copper slag combined with plant roots to reinforce soft soils using RSM, reporting significant improvements in soil strength and stability. This study highlighted the versatility of copper slag in geotechnical engineering applications.

This study aims to enhance concrete performance by incorporating industrial by-products—Alccofine, silica fume, and copper slag—into a ternary blended mix. The fine particle size and high reactivity of Alccofine and silica fume improve workability, cohesion, and compressive strength through enhanced calcium silicate hydrate (C-S-H) gel formation. Copper slag, used as a partial sand replacement, boosts strength due to its density and angular shape. These materials also enhance durability by reducing permeability and increasing resistance to abrasion and harsh environments. While beneficial overall, excessive copper slag can reduce workability, and Alccofine may slow early strength development. The experimental results were further validated by comparing them with the predicted values generated through SVM-based modelling, demonstrating close agreement between measured and estimated strengths [14]. The findings indicate that the optimal replacement level of fine aggregate with copper slag is 20%, at which the compressive strength exhibited an improvement of approximately 15.25% over the control mix [15]. The primary objective of this research is to promote the incorporation of industrial by-products into concrete production, thereby enhancing sustainability and reducing environmental burdens. Ternary blended concrete incorporating Alccofine, silica fume, and copper slag has shown superior mechanical and durability performance compared to conventional concrete mixes. The ultrafine particle size of Alccofine and silica fume significantly improves the workability, cohesiveness, and packing density of the cementitious matrix. Alccofine, owing to its high fineness and latent hydraulic activity, contributes to pore refinement and enhances compressive strength by accelerating the formation of additional C-S-H gel. Similarly, silica fume reacts pozzolanically with calcium hydroxide to generate secondary C-S-H gel, leading to substantial improvements in strength and long-term performance. Copper slag, when used as a partial substitute for natural sand, enhances mechanical characteristics due to its higher density, angularity, and superior particle interlocking. In terms of durability, Alccofine and silica fume substantially reduce permeability, thereby increasing resistance against chloride ingress, sulphate attack, and other aggressive environmental exposures. Copper slag further contributes by improving abrasion resistance, making the composite concrete particularly suitable for marine, coastal, and de-icing environments. Overall, the synergistic integration of Alccofine, silica fume, and copper slag in a ternary blend result in concrete with significantly enhanced strength, durability, and sustainability credentials. However, it may be noted that Alccofine can delay early-age strength development due to its slower hydration kinetics, and excessive copper slag content may adversely affect workability because of its angular and dense nature. Despite these limitations, ternary blended concrete presents a promising pathway toward high-performance and eco-efficient construction materials.

3.1 Workability

The rheological behavior of the formulated SCC mixes was evaluated through a detailed set of standardized workability tests, namely the L-Box, J-Ring, and V-Funnel Tests. All assessments were performed in compliance with the recommendations of the EFNARC. The outcomes of these tests are presented in Table 1, which summarizes the flowability and passing ability of the SCC mixtures. The data highlight the effect of incorporating different proportions of crushed sand as a partial substitute for natural river sand, thereby illustrating the role of fine aggregate type in governing the fresh-state performance of SCC.

Table 1. Workability test

Figures 2-4 present a graphical comparison of the rheological behavior of the SCC mixes as evaluated through the L-Box, J-Ring, and V-Funnel Tests, respectively. These visual representations clearly demonstrate a consistent trend: as the proportion of crushed sand replacing river sand increases, there is a notable enhancement in workability parameters across all test methods. This improvement in rheological performance is primarily attributed to the lower water absorption capacity of copper slag compared to river sand. Due to its reduced porosity and angular particle shape, copper slag retains less mixing water, thereby increasing the availability of free water within the concrete matrix. The presence of this additional free water effectively reduces inter-particle friction and enhances the lubricating effect among the aggregates, leading to improved flowability and self-compaction characteristics. In the L-Box Test, higher copper slag content corresponded with an increase in the H₂/H₁ ratio, indicating superior passing ability. Similarly, the J-Ring Test showed a reduced difference in flow diameter between the control and the copper slag-modified mixes, reflecting improved obstruction-free flow. The V-Funnel Test results also confirmed reduced flow times with increasing copper slag content, signifying better filling ability and lower viscosity. Furthermore, all measured values for the SCC mixes fall within the permissible limits specified by EFNARC, validating the adequacy of the mixtures in satisfying the fundamental criteria for SCC. These findings confirm that the inclusion of copper slag not only maintains but actively enhances the rheological properties of SCC, making it a viable and sustainable alternative to natural river sand in high-performance concrete applications.

Figure 2. L-Box

Figure 3. J-Ring

Figure 4. V-Funnel

3.2 Compressive strength test

Compressive strength evaluation was carried out on concrete cube specimens measuring 150 mm × 150 mm × 150 mm in accordance with IS 516:1959 (reaffirmed in 2018), using a standard compression testing machine. The test results, presented in Table 2 and illustrated in Figure 5, demonstrate a progressive increase in the compressive strength of SCC with a higher proportion of crushed sand, with the optimum performance observed at 30% replacement of natural river sand. This enhancement can be attributed to improved particle packing density and more efficient load transfer within the concrete matrix. Furthermore, the incorporation of Alccofine 1203 and GGBS contributed to additional strength development by refining the microstructure and promoting secondary hydration reactions.

Table 2. Compressive strength test

Figure 5. Strength in compression

Figure 6. Strength in compression

Strength through increased C-S-H gel formation, due to their high silica and calcium content. These materials refine the microstructure and improve the cement–aggregate bond, leading to better overall mechanical performance and sustainability.

The compressive strength test was performed on 150 mm concrete cubes as per IS 516:1959 (reaffirmed in 2018), using a compression testing machine. Results shown in Table 2 and Figure 6 indicate that compressive strength of SCC improved with increasing crushed sand content, peaking at 30% replacement of river sand. This improvement is linked to better particle packing and mix efficiency. Additionally, the inclusion of Alccofine 1203 and GGBS enhanced strength through increased C-S-H gel formation, due to their high silica and calcium content. These materials refine the microstructure and improve the cement–aggregate bond, leading to better overall mechanical performance and sustainability.

3.3 Split tensile strength test

The split tensile strength test was carried out in conformity with IS 516:1959 (reaffirmed in 2018) as shown in Figure 7. This test was performed on specimens of size 150mm diameter and 300 mm length. Table 3 and Figure 8 illustrate the outcome. Tensile strength of SCC has been increased at 30% replacement of copper slag as compared to other resulted output. Secondary cementitious material like Alccofine and GGBS enhance the properties.

Figure 7. Split tensile strength

Table 3. Split tensile strength test

Figure 8. Split tensile strength

3.4 Flexural strength test

The test was conducted in accordance with the IS 516:1959 standard (reaffirmed in 2018), as detailed in Figure 9. The specimens used were characterized by a cross-sectional dimension of 100 mm and a length of 500 mm. The output results, are presented in Table 4 and Figures 10 to 11. Notably, the bending strength of the SCC was enhanced by up to 30% with a 30% replacement of coarse aggregate, compared to other tested outputs.

Figure 9. Flexural strength test

Table 4. Flexural strength test

Figure 10. Output of MATLAB

Figure 11. Comparison of flexural strength

3.5 Prediction of strength values with machine learning

The prediction of strength values has been estimated using MATLAB through experimental values. Figure 10 present developed algorithm for compression strength prediction using MATLAB. Table 5 shows the obtained experimental and predicted compressive strength values. Strength versus the percentage substitution of aggregate is displayed in Figures 12 to 14. From these plots, it was seen that the experimental results are almost similar to the predicted results. The prediction accuracy rates of the six datasets in aspects of the coefficient of determination (R²), mean squared error (MSE), and root mean square error (RMSE), are found and the readings are demonstrated in Table 5. The R² values lies in between 0 to 1 in which more the value of R² greater the accuracy of the results obtained. The R² value of compressive strength was found to be 0.98 while for Splitting tensile strength the value of R² was 0.97 and for flexural strength the value of R² was 0.91 which shows that the prediction of SVM model was more accurate.

Table 5. Experimental and predicted compressive strength

From Figure 11 and Table 5 it is observed that the experimental and predicted values of model using MATLAB shows a close match. In this model compressive strength values are taken as data set for modelling with Radial Basis Function (RBF) algorithm was used. At 30% replacement getting optimistic strength values. SVM modelling is most reliable too.

3.6 Split tensile strength prediction using MATLAB

The code utilized for both split tensile and compressive strength calculations is the same; the only modifications are in the labels, titles, and legends. Table 6 shows the obtained experimental and predicted split tensile strength values.

Table 6. Experimental and predicted split tensile strength

It is observed that (Figure 13) the experimental and predicted values of model using MATLAB shows a close match. The SVM model was developed using RBF algorithm for prediction of tensile strength. The graphs showing relatively closer values with the experimental values. In evaluating the results MSE and R².

3.7 Flexural strength prediction using MATLAB

The code used for calculating split tensile and compressive strengths is identical, with the sole differences being in the labels, titles, and legends.

Figure 14. Prediction of flexural strength

From Figure 14 and Table 7 it is observed that the experimental and predicted values of model using MATLAB shows a close match. The values are identical with experimental values.

Table 7. Experimental and predicted flexural strength