Arak University Colloid & Nanoscience Journal 2980-9215 4 1 2026 01 01 Modeling and optimization of thermal conductivity and viscosity of water-based hybrid nanofluids containing graphene oxide combined with silicon dioxide (Go-SiO2, 50:50) using MLP & GA 733628 EN Hamid Reza Sabouni Department of Mechanical Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran Ali Mokhtarian Department of Mechanical Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran Mojtaba Rahimi Department of Petroleum Engineering, Kho.C., Islamic Azad University, Khomeinishahr, Iran; Stone Research Center, Kho.C., Islamic Azad University, Khomeinishahr, Iran 0000-0002-2567-0508 Journal Article 2025 11 25 In this study, architecture and training of two artificial neural networks (ANNs) were designed to predict the thermal conductivity (TC) and viscosity properties of nanofluids GO-SiO2 (50:50, Graphene Oxide and Silicon Dioxide)-based nanofluids. The nanofluid samples investigated comprised various weight fractions (0.1–1%) of a 50:50 mass-ratio mixture of graphene oxide (GO) and silicon dioxide (SiO₂) nanoparticles dispersed in the base fluid; their viscosity and thermal conductivity were measured at temperatures ranging from 30 to 60°C. For each ANN developed to predict either nanofluid viscosity or TC, the regression plots corresponding to the training, validation, and test datasets provided a clear demonstration of the network’s optimal performance. Specifically, the mean and maximum relative errors obtained for the test dataset were 0.3425% and 0.8359%, respectively, for viscosity prediction, and 0.2465% and 0.4069%, respectively, for TC prediction. Furthermore, following a sensitivity analysis of the networks, we found that the weight fraction exerts a more pronounced influence on nanofluid viscosity and TC than temperature. Subsequently, a genetic algorithm (GA) applied to the trained ANN models identified the optimal conditions as a weight fraction in the range of 0.1–1% and a temperature of 60°C.