Xie, ChangguiSun, XinSalahshour, SoheilEmami, N.Alkhalifah, TamimBouallegue, Belgacem2026-01-152026-01-1520260735-19331879-017810.1016/j.icheatmasstransfer.2025.1102082-s2.0-105024411951https://doi.org/10.1016/j.icheatmasstransfer.2025.110208https://hdl.handle.net/20.500.14517/8697Nanoparticles (NPs) can improve the thermo-physical properties of fluids and increase the effectiveness of heat transfer systems. In this way, achieving optimal properties of nanofluids (NFs) is an important subject. The present work aims to model and optimize the thermo-physical properties of dynamic viscosity (DV) and thermal conductivity (TC) of CuO/Cyclohexane + Diethylamine (DEA) as a non-polar nanofluid with binary base fluids. The input parameters include the temperature and the solid volume fraction (SVF) of NF. Based on available experimental data, the molar weight ratio of the NPs ranges from 0.01 % to 0.06 % with temperatures varying from 298 K to 318 K. The NF is modeled by two trained two-layer feedforward artificial neural networks (ANNs) for the prediction of DV and TC at a specified range of temperature and SVF. The average and maximum relative errors for test datasets are 0.4872 and 0.9106 for DV and 0.4279 and 0.7338 for TC prediction networks, respectively. Through the ANNs' sensitivity analysis, the importance of the SVF rather than the temperature on DV and TC was revealed. Based on the proposed model, a multi-objective optimization problem was formulated to maximize TC and minimize DV simultaneously, and solved using the multi-objective particle swarm optimization (MOPSO) method. Finally, the optimal values of the objective functions and the corresponding input parameters were plotted along with the Pareto optimal points.eninfo:eu-repo/semantics/closedAccessNanofluidDynamic Viscosity (DV)Thermal Conductivity (TC)Artificial Neural NetworkMulti-Objective OptimizationArtificial IntelligenceThermal Conductivity ( TC )Dynamic Viscosity ( DV )Article