Optimization of mechanical properties and composition of M-sand and pet particle added concrete using hybrid deep neural network-horse herd optimization algorithm

Abstract

• Natural sand extraction, Granite & Basalt wastes are a problem, hence five Novel concrete design mixes are made. • Concrete performance and Durability tests are conducted and compared with the Predicted results of familiar algorithms. • Results show that the Design Mix P7.5 Containing 7.5% PET particles is the best Mix among all the Mixes made. • While Hybrid DNN-HHO proved to be the best algorithm among the other algorithms used in the investigation. Concrete is one of the broadly utilized materials due to the unavoidable nature of construction. The requirement for improvement in mechanical properties and environmental hazards with the usage of natural sand requires an alternative choice in raw materials. This research work is carried out to investigate the compressive and tensile strength properties of concrete prepared by adding manufactured sand (M-sand) sand and polyethylene terephthalate (PET) particles. Pullout test, Water Absorption (WA), Electrical Resistivity (ER), Rapid Chloride Permeability Test (RCPT), X-Ray Diffraction (XRD) and Scanning Electron Microscopy (SEM) are also analyzed in addition to compressive and tensile strength. A suitable proportion of materials is studied from the preliminary investigation and produced five different concrete specimens for this study. New proportions of concrete mixes with cost-effective materials such as PET and M sand are contributed to enhancing the strength and tensile properties. Cylindrical-shaped concrete specimens are created with final mix proportions and their properties are studied for six different days. The results show that the final specimen has acquired a maximum compressive strength of 45.3 MPa, a maximum tensile strength of 3.77 MPa, a maximum flexural strength of 4.72 MPa, and a minimum water absorption of 8.087 were all obtained by 7.5P specimen after 90 days of synthesis. The X-Ray Diffraction (XRD) and Scanning Electron Microscopy (SEM) analysis results taken for 7.5P specimen after 90 days indicate the presence of PET provide more strength and tensile properties. The properties of the proposed concrete composition is validated and compared with predicted values obtained from Deep Neural Network-Horse herd Optimization Algorithm (DNN-HHO) to find accurate results. The proposed concrete indicates an enhanced result in mechanical properties along with the reduction in environmental issues and can be utilized in the construction field of small applications.