Characterization of sustainable concrete made from wastewater bottle caps using a machine learning and RSM-CCD: towards performance and optimization

Abstract. The properties of concrete, a widely used building material across the globe, have changed due to technological breakthroughs. Cement, sand, coarse aggregate, and water are four components build concrete. Technological improvements increase human comfort, yet environment is also harmed. Therefore, recycling reuse vital environmental engineers because they help reduce amount plastic bottle garbage disposed as solid waste. In this study, water-cement ratios 0.5, 0.55, 0.6 in lieu concrete various percentages, including 0, 6, 12% aggregate replaced by caps, analyze behavior concrete's compressive strength experimentally. Based on experimental results, models based artificial neural networks—Levenberg Marquardt Response Surface Methodology—Central Composite Design were developed forecast final made part from bottles. results demonstrate that for accurately predicting ANN-LM model yields best result, R2=0.98, which close 1 R2 = 0.85 RSM-CCD, respectively.