Investigation on the two-phase flow and heat transfer behaviors in a new central uniform dispersion-type heat exchanger

Abstract

In order to improve the temperature uniformity of multiple chips, a new central uniform distributed heat exchanger (CUD) is designed. Firstly, adopting ethanol solution as coolant, the cooling ability of the heat exchanger is numerically investigated by using VOF gas-liquid two-phase flow model, and the two-phase flow experimental system is established to verify the correctness of the numerical results. The results show that the Tavg (average temperature of heat sources) error range between the experimental and numerical values is 1 K ~ 3 K, and the relative error of SDT (standard deviation of temperature) is 6% ~ 10%. This verifies the accuracy of the numerical simulation method. Secondly, the two-phase flow boiling heat transfer behavior are compared with that of single-phase flow with water as coolant. The results show that with the increase of inlet flow, the thermal resistance and average temperature of two-phase flow change are less than that of single-phase flow. This indicating that two-phase flow has better advantages in improving the temperature uniformity of multiple heat sources. Thirdly, the heat transfer characteristics of radiators with different Wi are studied. Compared with Case1, the boiling heat transfer coefficient of Case3 and Case4 can be increased by 12.3% and 18.3%, respectively. However, Case3 is a better choice by weighing the pressure drop. Finally, the prediction functions of the boiling heat transfer coefficient and pressure drop is established, and the multi-objective genetic algorithm is used to obtain the best parameters of this heat exchanger. The average heat transfer coefficient of the optimized heat exchanger can be increased by 7.56%, and the pressure drop can be reduced by 3.54% compared to the exchanger before optimization. Furthermore, the heat dissipation ability of the optimized heat exchanger design is studied numerically. The simulated results are in good agreement with the optimization results.