A Hybrid Finite Element and Adaptive Fuzzy Logic Approach for Optimizing Nano-Enhanced Heat Exchangers

Abstract

This work investigates approaches to enhance the thermal performance of U-tube heat exchangers by coupling them with phase change materials (PCMs) enhanced with Al2O3 nanoparticles. The focus is on investigating how nanoparticle addition and wavy tube configurations influence the melting behavior. A numerical finite element model was validated against a reference study, confirming its accuracy. The incorporation of Al2O3 nanoparticles (0%, 1%, and 2%) led to a substantial improvement, reducing melting times by up to 80%. Notably, the wavy tube system with 2% nanoparticle-enhanced PCM demonstrated superior performance. To circumvent the significant computational expense associated with finite element analysis, an Adaptive Network-Based Fuzzy Inference System (ANFIS) was employed as a surrogate model. With a high prediction accuracy (R² > 0.997) for melting dynamics, the ANFIS model was then employed for a detailed optimization analysis. The optimization identified an ideal nanoparticle concentration of 3.50%, which minimizes the melting time to 263.63 seconds under wavy tube conditions. The findings underscore the potential of a hybrid FEA-ANFIS methodology, suggesting that a synergistic combination of nanotechnology, advanced geometries, and artificial intelligence offers an effective approach to improving the design and operational efficiency of thermal energy storage units.