AUTHOR=Özkılıç Yasin Onuralp , Bahrami Alireza , Güzel Yusuf , Soğancı Ali Sinan , Karalar Memduh , Althaqafi Essam , Çelik Ali İhsan , Zeybek Özer , Jagadesh P. TITLE=Waste ceramic powder for sustainable concrete production as supplementary cementitious material JOURNAL=Frontiers in Materials VOLUME=11 YEAR=2025 URL=https://www.frontiersin.org/journals/materials/articles/10.3389/fmats.2024.1450824 DOI=10.3389/fmats.2024.1450824 ISSN=2296-8016 ABSTRACT=

A detailed examination was carried out by substituting waste ceramic powder (WCP) for specific ratios of cement in concrete. To achieve this, five different WCP percentages (10%, 20%, 30%, 40%, and 50%) were used in manufacturing of concrete. First, the workability and slump values in the fresh state of concrete were determined by performing a slump test. Subsequently, several tests, including compressive strength (CS), splitting tensile strength (STS), and flexural strength (FS), were conducted on the specimens to assess the effectiveness of concrete fabricated using WCP. Variations in the strength were determined in terms of the various amounts of WCP. The findings demonstrated that by including WCP at levels of 10%, 20%, 30%, 40%, and 50%, there were corresponding reductions in CS of 5.8%, 21.8%, 47.1%, 63.2%, and 73.6%, respectively. The decreases in STS were 6.3%, 13.8%, 35.2%, 49.7%, and 65.4%, respectively, when a concrete STS value of 1.59 MPa was considered. Similarly, when the WCP content increased, FS was reduced by 15.3%, 21.4%, 31.6%, 44.9%, and 54.1%, respectively. This is very significant because it represents one of the key issues in calculating the optimal quantity of WCP in relation to both the strength and the amount of WCP utilized. Furthermore, taking into account our experimental research and previous studies on concrete produced utilizing WCP, straightforward equations were provided for practical use to predict CS, STS, and FS. In addition, scanning electron microscopy was done to validate the findings obtained from the experimental part of the study. The artificial neural network modeling technique was adopted to estimate the concrete properties with average coefficients of determination (R2) as 0.945 (CS), 0.901 (STS), and 0.856 (FS) with K-fold cross-validation.