Investigation on multi-objective optimization of thermal–hydraulic performance for enhanced heat transfer in jet impingement cooling

Exploring efficient, clean and energy-saving cooling solutions is one of the most important ways to decarbonize data centers (DCs). This study investigates a variable spacing multi-jet direct chip cooling device to increase the heat transfer coefficient and improve the temperature uniformity. Attaching the pin fins directly to the chip significantly reduces the thermal resistance. The study of different jet schemes shows that the variable spacing multi-jet cooling solution has better heat dissipation performance. Coolant flow rate, inlet temperature, pin fin design parameters, and jet hole spacings all have different effects on thermal resistance, pressure drop, temperature standard deviation, and Nusselt number. The dataset for surrogate model construction is obtained based on computational fluid dynamics and Latin hypercube sampling experimental designs. An artificial neural network model with structural and thermal parameters as inputs and thermal resistance, pressure drop and temperature uniformity as outputs is developed. The algorithm named constrained multi-objective optimization based on even search (CMOES) is used to implement the solution of the optimization model. The optimization results show that the optimal design reduces thermal resistance to 0.048 K/W, decreases pressure drop by 20.26 kPa, and lowers temperature standard deviation by 3.22 K compared to the initial single-hole jet design. The optimal design outperforms the initial design and the performance specifications of the existing studies in terms of both thermal and hydraulic performance, and has good prospects for engineering applications. Herein, the exploration of optimization for variable spacing multi-jet direct chip cooling will help chip to operate at higher performance.