AI-enhanced prediction and analysis of heat and mass transfer in NEPCM-filled wavy star-shaped cavities under magnetic, radiative, and chemical influences

Nano-Encapsulated Phase Change Material (NEPCM)-based thermal systems play a crucial role in enhancing heat and mass transfer performance in energy storage and chemical processing applications. This study investigates the thermal and solutal behavior of a wavy star-shaped cavity filled with NEPCM, influenced by magnetic fields , thermal radiation , and chemical reactions. The mathematical model incorporates the effects of double diffusion , including Soret and Dufour phenomena, along with buoyancy-driven convection. The research methodology integrates numerical simulations with an artificial intelligence (AI) framework, specifically utilizing the XGBoost regression model to predict the average Nusselt ( N u avg ) and Sherwood ( S h avg ) numbers. The results reveal that increasing the Soret number enhances particle dispersion , improving N u avg and S h avg by up to 40% and 45%, respectively. Thermal radiation improves uniformity in heat and mass transfer by approximately 30%, while magnetic fields enhance these processes by 20 %. Additionally, higher chemical reaction rates amplify heat and mass transfer by up to 60 %, highlighting the significance of reactive transport. The findings emphasize the critical role of initial conditions, with the configuration featuring high-temperature and high-concentration solid particles achieving superior performance. These insights contribute to optimizing NEPCM-based systems for thermal energy storage , advanced heat exchangers , and chemical process efficiency.