Optimizing thermophysical properties of nanofluids using response surface methodology and particle swarm optimization in a non-dominated sorting genetic algorithm

Abstract

Abstract The purpose of this study is to optimize the thermal conductivity and viscosity of the Al2O3/water, CuO/water, SiO2/water, and ZnO/water nanofluids. Both thermophysical properties are modeled using the experimental data via Response Surface Methodology (RSM) and Artificial Neural Network (ANN). The thermal conductivities of the Al2O3/water and CuO/water nanofluids demonstrate maximum increment at all the temperatures and volume fractions. However, the viscosity variations of various nanofluids have no noticeable difference. The models of the ZnO/water and CuO/water nanofluids indicate the highest accuracy among the proposed models of relative viscosity and relative thermal conductivity, respectively. The deviation values of the RSM model are greater than those of the ANN model for predicting the relative viscosity, and the minimum error of the ANN for prediction of this output is related to the ZnO/water nanofluid. The results show that the most appropriate models for predicting the relative thermal conductivity and relative viscosity are the RSM model and ANN model, respectively. The multi-objective optimization based on RSM and Multi-Objective Particle Swarm Optimization (MOPSO) is performed by the Non-dominated Sorting Genetic Algorithm (NSGA-II), and the optimal points for both thermophysical properties are presented. Based on the results, the highest temperature provides simultaneous optimization of both thermophysical properties.