Inverse design approach for predicting the optimal geometry of horizontal annular wavy perforated fins

Abstract

A three-dimensional inverse design problem for determining the optimal geometry of annular irregular wavy perforated fins is investigated in this study. The conjugate turbulent heat transfer problem is solved using the commercial CFD package CFD-ACE+, with the standard κ–ε turbulence model employed for the numerical simulations. The primary objective is to minimize the average shaft surface temperature, T b , of the heat sink through wavy fin geometry optimization using the Levenberg–Marquardt method (LMM) under a constant-volume constraint, without considering the pressure drop penalty. The design variables include the fin amplitude, wavelength, perforation radius, and perforation location. Four heat sink configurations are considered for comparison of the average shaft surface temperature: fin#1 (flat plate), fin#2 (regular wavy), fin#3 (irregular wavy), and fin#4 (irregular wavy with perforations). The numerical results show that, at an inlet velocity of 5.0 m/s, the fin#4 heat sink reduces the average shaft surface temperature, T b , by approximately 7.7%, 2.3%, and 1.6% compared with fin#1, fin#2, and fin#3, respectively. These improvements demonstrate the synergistic cooling enhancement achieved by combining an irregular wavy geometry with strategically placed perforations. To validate the numerical predictions, prototypes of the fin#1, fin#2, and fin#4 heat sinks were fabricated and tested in a wind tunnel. Shaft surface temperature distributions were measured using an infrared thermal camera. The experimental results show excellent agreement with the CFD-ACE + simulations using the standard κ–ε turbulence model, thereby confirming the accuracy of the numerical model and demonstrating that the fin#4 heat sink provides the best cooling performance among the configurations investigated, highlighting its strong potential for practical thermal management applications.