Fuzzy modeling and particle swarm optimization of Al₂O₃/SiO₂ nanofluid

This study aims to model and optimize the performance of a shell and helically coiled tube heat exchanger application using Aluminum oxide (Al₂O₃) and Silicon dioxide (SiO₂) nanofluids. An adaptive network-based fuzzy inference system (ANFIS) in MATLAB was used to build the fuzzy logic model for density, viscosity, specific heat, and thermal conductivity properties of nanofluid. The accuracy of the model was evaluated by checking the mean square error (MSE) for training data, testing data, and all data. The particle swarm optimization (PSO) based on the effectiveness of heat exchanger and friction factor as objective functions were used to find the optimum nanofluid properties. The optimum properties can be achieved by minimizing the viscosity and specific heat and maximizing the thermal conductivity; while keeping the density at optimum value for this application. The optimal properties of nanofluids were found at 60 °C using a hybrid nanofluid consisting of 0.3 Al₂O₃ and 0.1 SiO₂. The optimum values of the density, viscosity, specific heat, and thermal conductivity were 985.77 kg/m³, 0.000471 Pa s, 4129.36 J/kg.°K and 0.71069 W/m. °K, respectively. The coupling between the fuzzy model and the PSO method was effective to find the optimum properties of nanofluids under operation.