PREDICTION OF CONCRETE STRENGTH WITH MODIFIED PLASTIC WASTE AGGREGATE AS PARTIAL REPLACEMENT FOR COARSE AGGREGATE

Abstract

The disposal of plastic waste presents significant environmental challenges, including degradation of landfills and water bodies, greenhouse gas emissions, and soil contamination. Utilizing plastic waste in concrete production offers a solution to illegal dumping and reduces the reliance on mined aggregates, promoting sustainable construction practices. Polyethylene Terephthalate (PET), commonly found in plastic bottles and food containers, is a readily available source of plastic waste. This study investigates the effects of treating PET waste with calcium hypochlorite solution (Ca(ClO)2) before incorporating it into concrete as a partial replacement for coarse aggregate. Various compressive strength, ultrasonic pulse velocity (UPV), and density tests were conducted for three replacement percentages: 15 %, 30 %, and 45 % of conventional coarse aggregate with modified plastic aggregates (MPA). The findings show that chemically treated plastic aggregates maintained fresh density while reducing slump value at 30% and 45% replacement levels, even with the addition of polycarboxylate acid (superplasticizer), possibly due to surface roughness and irregular shapes of the MPA. However, concrete with 30% MPA achieved a 28-day compressive strength, UPV, and density of 23.13 N/mm², 3643 m/s, and 1996 kg/m³, respectively, which conforms with BS EN 206-1 (2013) standards for the minimum requirement of structural lightweight concrete. Additionally, three machine learning models which include Artificial Neural Network (ANN), K-Nearest Neighbor (KNN) and Random Forest (RF) were developed to predict water absorption and sorptivity. Pre-processing, statistical methods and data visualization techniques were employed for data understanding. Experimental results were used to generate a dataset, and the models ...