• Develop an intelligent optimization scheme based on Bayesian Optimization algorithm. • Optimum design is performed by integrating the structure modeling, meshing, and evaluation. • Bayesian Optimization algorithm demonstrates a great advantage over evolutionary algorithms in efficiency. • The comprehensive thermal performance factor of the heat sink with the optimal structure increases by 17.6%. • The enhancement is explained by the reduction of the synergy angle between the velocity and the pressure gradient. The optimal structure of a sinusoidal wavy plate fin heat sink with crosscut (SWHS-WC) is determined by an efficient intelligent optimization method based on Bayesian Optimization (BO) algorithm. The comprehensive thermal performance factor (TPF) is the objective function, which is associated with the heat transfer and the pressure drop. The parametric modeling, meshing, and numerical calculation are integrated to optimize three structural parameters, including the fin amplitude, period, and phase shift angle of the heat sink, in an iterative and automated way. The result shows that the TPF of the SWHS-WC with the optimized structure is increased by 17.6%, due to the significant reduction in the pressure drop penalty, and the corresponding synergy angle between the pressure gradient and the velocity decreases by about 6.2°. Compared to the evolutionary algorithms, the BO algorithm demonstrates a great advantage in calculation efficiency, which can be used to find the global optimum design at a much smaller computational cost.
Read full abstract