Enhancing thermal efficiency in twisted tri-lobe double pipe heat exchangers via integrated CFD and AI approaches

Industrial energy efficiency relies on precise heat exchanger design. Engineers aim to enhance heat transfer while minimizing pressure drop. Twisted tube Double Pipe Heat Exchangers (DPHEs) excel, inducing secondary flows that disrupt boundary layers and enhance fluid mixing. This investigation aims to provide insights into the relationships among specific factors and the thermal performance of a DPHE with twisted tri-lobe tubes, utilizing water as the working fluid. This is accomplished by employing an approach that integrates computational fluid dynamics with an artificial neural network, coupled with a single-objective genetic algorithm. Data inquiry was facilitated using response surface methodology, considering five design variables including the Reynolds numbers of the outer and inner twisted tubes, outer and inner twist ratios, and twisting direction. The simulation employed the turbulent k-ω shear-stress transport model, and the governing equations were solved using the finite volume method. The results underscored the efficacy of the developed multi-approach algorithm in maximizing the Performance Evaluation Criterion (PEC) both within the specified domain (10000 ≤ Re ≤ 20000 & 5≤twist ratio≤20) and beyond the domain (5000 ≤ Re ≤ 50000 & 2.5≤twist ratio≤25) of design variables. In the specified domain, a PEC of 1.16 was achieved, demonstrating the algorithm's effectiveness. Furthermore, the proposed method, extrapolated to the data space, yielded a noteworthy improvement in PEC of 16.35 %. In both domains, the optimized set of design variables for maximizing the PEC comprised a minimum outer Reynolds number, maximum inner Reynolds number, maximum outer twist ratio, and minimum inner twist ratio.