Multi-Objective Optimization of Internally Finned Rectangular Heat Exchangers Using Surrogate Modeling, MOPSO, and TOPSIS

This research focuses on the thermo-hydraulic optimization of internally finned rectangular heat exchangers, wherein thermal enhancement is mainly constrained by significant hydraulic losses. A surrogate-assisted multi-objective optimization framework is developed by integrating high-fidelity computational fluid dynamics (CFD), regression-based surrogate modeling, multi-objective particle swarm optimization (MOPSO), and TOPSIS decision-making. Seven configurations, including a finless baseline and cylindrical and rectangular fins arranged in in-line and staggered patterns, are systematically investigated under identical fin surface area constraints and laminar flow conditions (0.5–2.5 L·min⁻¹). A comprehensive dataset is generated and then used to develop polynomial and log-polynomial surrogate models that closely reproduce the numerical results, with high predictive accuracy (R² > 0.98, RMSE ≈ 1%). The optimization process concurrently maximizes the heat transfer rate, overall heat transfer coefficient, and efficiency, while minimizing the friction factor and pressure drop. The results indicate that staggered cylindrical fins offer the highest heat transfer rate (up to +46.6%), whereas in-line rectangular fins provide a better thermo-hydraulic performance. Pareto analysis reveals distinct regimes and the thermal–hydraulic trade-off, enabling optimal design selection.