Application of multilayered porous media for heat transfer optimization in double pipe heat exchangers using neural network and NSGA II

This study investigates the use of multi-layered porous media (MLPM) to enhance thermal energy transfer within a counterflow double-pipe heat exchanger (DPHE). We conducted computational fluid dynamics (CFD) simulations on DPHEs featuring five distinct MLPM configurations, analyzed under both fully filled and partially filled conditions, alongside a conventional DPHE. The impact of various parameters such as porous layer arrangements, thickness, and flow Reynolds numbers on pressure drop, logarithmic mean temperature difference (LMTD), and performance evaluation criterion (PEC) was assessed. The results demonstrate that the PEC improves by approximately three times with the optimal MLPM configuration, where the porous layers decrease from the interfacial wall outward under fully filled conditions. To further optimize performance, we trained a neural network using 507 simulations to establish a continuous correlation between input variables and output results. A multi-objective optimization was then implemented using the non-dominated sorting genetic algorithm II (NSGA-II) to identify the optimal operating conditions. The Pareto front of the optimal points was established, allowing designers to select specific points based on their operational requirements. This research provides valuable insights into the potential of MLPM to significantly enhance the thermohydraulic performance of DPHEs and establishes a framework for future optimization studies.