文献类型：期刊文献

英文题名：Multi-objective optimization of manifold microchannel liquid cold plate combining data-driven and genetic algorithm approach

作者：Bie, Yuzhen[1];Shao, Xiang[1];Wang, Ning[2];Zhang, Tengteng[3];Xu, Hongtao[1]

机构：[1]Univ Shanghai Sci & Technol, Sch Energy & Power Engn, Shanghai 200093, Peoples R China;[2]East China Univ Sci & Technol, Sch Mech & Power Engn, Shanghai 200237, Peoples R China;[3]Suzhou Thermal Win Energy Technol Co Ltd, Suzhou 205300, Peoples R China

年份：2026

卷号：178

期号：P5

外文期刊名：INTERNATIONAL COMMUNICATIONS IN HEAT AND MASS TRANSFER

收录：;EI(收录号：20262721049681);Scopus(收录号：2-s2.0-105043767009);WOS:【SCI-EXPANDED(收录号:WOS:001821224100001)】；

语种：英文

外文关键词：Manifold microchannel cold plate; Thermal management; Multi-objective optimization; Hydraulic loss; Data-driven

摘要：Developing efficient and energy-saving thermal management technology is essential for maintaining stable operation of next-generation electronic devices and improving energy efficiency. In this study, a novel symmetric manifold microchannel liquid cold plate (SMMLCP) is proposed for high-power chip thermal management. Its thermal and hydraulic performance is numerically investigated and compared with parallel microchannel liquid cold plate (PMLCP) and normal manifold microchannel liquid cold plate (NMMLCP). The Sobol global sensitivity analysis method based on Gaussian process regression (GPR) surrogate model is implemented to reveal the contribution of the multi-parameters and their interactions. Subsequently, a data-driven GPR surrogate model coupled with NSGA-II algorithm is proposed for the multi-objective optimization of the SMMLCP. The results indicate that severe velocity maldistribution is observed at the microchannel entrance in the NMMLCP due to the appearance of vortex, whereas the SMMLCP design effectively eliminates this phenomenon, achieving the lowest pressure drop of 15.4 kPa, chip max temperature of 79.5 degrees C, and the best overall performance with the highest COP of 4941. The microchannel width exhibits the primary impact on chip max temperature and thermal resistance, whereas the inlet mass flow rate performs the most evident impact on pumping power. Under the condition of 100 W cm-2 heat flux, the SMMLCP decreases chip max temperature, pumping power, and thermal resistance by 2.4%, 34.8%, and 21.2%, respectively, and improves COP by 52% through achieving more uniform velocity and temperature distributions. This model framework provides an efficient optimization strategy and design guidance for novel manifold microchannel liquid cold plates.