Researchers have developed optimized concrete mixes using Ground Granulated Blast Furnace Slag (GGBS), Eggshell Powder (ESP), and Waste Glass Powder (WGP) to improve compressive and tensile strength.
Study: Machine learning–driven optimization of compressive and tensile strength in concrete with GGBS, eggshell powder, and waste glass powder. Image Credit: Ermak Oksana/Shutterstock.com
Published in Scientific Reports, the research explores how these supplementary materials can improve concrete's mechanical properties while promoting more sustainable construction practices.
Addressing Carbon Emissions in Construction
The construction industry is a major contributor to global carbon dioxide (CO2) emissions. For example, Ordinary Portland Cement (OPC) production alone accounts for roughly 8 % of the total.
As cities expand and infrastructure demands grow, the construction sector faces mounting pressure to adopt lower-carbon alternatives.
That’s where supplementary cementitious materials (SCMs) come in. By using GGBS, WGP, and ESP - byproducts of industrial and agricultural processes - this study taps into materials that not only reduce emissions but also improve concrete performance.
GGBS enhances durability and long-term strength through its latent hydraulic properties. WGP promotes the formation of calcium silicate hydrate (C-S-H), which contributes to strength development. ESP, when used in moderation, improves particle packing and early-age strength due to its fine particle size, although excessive use may dilute the binder and reduce performance.
Together, these SCMs provide a pathway to reduce cement usage while maintaining or even improving mechanical properties.
Experimental Design and Methodology
To assess how these materials affect concrete performance, researchers prepared 64 unique mix designs. Each mix varied the content of GGBS (0–30 %), ESP (0–12 %), and WGP (0–15 %). The binder-to-aggregate ratio was kept constant at 1:1.5:3, and all mixes used a water-to-binder ratio of 0.50, aligning with ASTM C192 standards.
Testing included both workability (via slump tests) and mechanical strength evaluations, specifically compressive strength (ASTM C39) and split tensile strength (ASTM C496) after 28 days of curing.
Beyond physical testing, the study integrated machine learning (ML) models to predict strength outcomes based on variables such as water-to-binder ratio, cement content, and curing time. This helped streamline the optimization process and reduced the need for excessive experimental trials.
Key Findings from the Research
The results showed that using SCMs in combination had clear effects on performance. Compressive strength across the mixes ranged from 15.5 MPa to 24.7 MPa, while split tensile strength ranged from 1.8 MPa to 2.77 MPa.
GGBS was particularly effective at 10–20 % replacement, consistently improving strength by encouraging C-S-H formation. The best-performing mix, designated M39, combined 20 % GGBS, 4 % ESP, and 10 % WGP. This formulation achieved the highest compressive and tensile strengths: 24.7 MPa and 2.77 MPa, respectively.
When it came to workability, mixes with GGBS and finely ground WGP generally saw improved slump values, indicating better flow. In contrast, higher ESP contents tended to reduce workability due to its fine particles increasing water demand.
The ML models proved reliable in predicting outcomes. The Gradient Boosting Regressor (GBR) achieved an R2 of 0.937 for compressive strength and 0.906 for tensile strength. SHAP (Shapley Additive Explanations) analysis revealed that ESP was the most influential factor across predictions. GGBS had a variable impact depending on the mix, while WGP showed minimal direct influence.
Practical Applications and Sustainability Impacts
The findings offer tangible insights for real-world application. The optimized concrete mixes are suitable for structural elements like beams, slabs, and columns - providing strong performance while reducing environmental impact.
By partially replacing cement with SCMs, the industry can significantly reduce its carbon footprint and promote circular economy practices. Repurposing waste materials not only lowers emissions but also reduces raw material costs, making sustainable concrete a more practical option across a range of projects.
Conclusion and Future Directions
Overall, this study highlights the potential of GGBS, ESP, and WGP as effective cement alternatives in concrete. Properly balanced blends can deliver strong mechanical performance while reducing environmental harm. The integration of machine learning adds another layer of efficiency, allowing for a more targeted mix design with fewer physical tests.
That said, the study focused solely on mechanical properties after 28 days. It did not assess long-term durability or microstructural behavior, which are important factors for widespread adoption. Future research should explore how these mixes perform over extended periods and under various environmental conditions. Techniques like scanning electron microscopy (SEM) and X-ray diffraction (XRD) could also provide deeper insight into how these materials interact at the microscopic level.
Journal Reference
Alnadish, A.M., &. et al. (2025). Machine learning–driven optimization of compressive and tensile strength in concrete with GGBS, eggshell powder, and waste glass powder. Sci Rep 15, 40499. DOI: 10.1038/s41598-025-24438-1. https://www.nature.com/articles/s41598-025-24438-1
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.