Computational modelling and optimization of mechanical properties of red brick dust‐epoxy composites using machine learning approach
Abhilash Purohit, Sekaran Sathees Kumar, S. Padmanabhan, Priyabrat Pradhan, Gaurav Gupta, Pravat Ranjan Pati
Abstract
Utilizing construction waste as reinforcement in polymer composites is becoming more popular these days. In the current research, composites with varying amounts of red brick dust in an epoxy matrix are developed. The mechanical characteristics of these composites have been assessed. As the filler content increases, a progressive rise in impact strength and microhardness values is noted. However, as compared to neat epoxy, epoxy‐red brick dust composites show a slight reduction in tensile and flexural strength. A maximum hardness of 30.82 HV 2 and impact strength of 25.7 kJ/m 2 are obtained by adding 30 wt. ‐ % of red brick dust to the epoxy matrix; nevertheless, the tensile and flexural strengths are determined to be 23.37 MPa and 10.37 MPa, respectively. The impact strength of neat polymer increased by 43.57 % with an addition 30 wt. ‐ % filler. Simple linear regression is used to compare the results of mechanical properties. The machine learning simple linear regression model's projected outcomes closely match the mechanical values. Mechanical behaviors of the composites are also investigated with the help of Digimat software and compared with the investigational results. It has been noted that the experimental results are in excellent agreement with the numerically predicted values.