Mechanical properties and machine learning analysis of concrete incorporating waste glass as coarse aggregate

Abstract

The concrete industry has investigated the integration of waste glass (WG) as an alternative for traditional concrete materials in response to growing environmental concerns and the necessity to reduce waste while encouraging recycling. This study explores the use of Waste Glass Coarse Aggregate (WGCA) as a partial replacement for natural coarse aggregates in concrete, assessing its effects on mechanical performance and utilizing machine learning models for predictive analysis. The assessment involved concrete mixtures with replacement levels of 0%, 5%, 10%, and 15% WGCA, focusing on their compressive, tensile, and flexural strengths at both 7 and 28 days of curing. The experimental results demonstrate a gradual decrease in mechanical properties as WGCA content increases, showing a peak reduction of 10% in compressive strength, 13% in tensile strength, and 12% in flexural strength at a 15% replacement level. The findings correspond with suitable structural parameters, strengthening the potential of WGCA in the production of sustainable concrete. The application of predictive modeling through Multiple Linear Regression (MLR) and Support Vector Machine (SVM) revealed notable effectiveness, with MLR attaining a higher R2 score of 0.8193, exceeding SVM’s R2 of 0.7253 in terms of predictive accuracy. The investigation demonstrates the environmental advantages of WG reuse, serving a significant role in the principles of the circular economy by minimizing landfill waste, preserving natural resources, and lessening greenhouse gas emissions.