摘要：In our previous work, a new structure of a microchannel with triangular cavity and rib was proposed to enhance heat transfer. To obtain excellent thermal performance, a multi-objective optimization using Response Surface Methodology (RSM), Non-dominated Sorting Genetic Algorithm (NSGA-II), and k -mean clustering was performed to obtain a Pareto front. Thirty sets of numerical data were used to optimize design variables (cavity height, rib height, and Reynold number) of objective functions (thermal resistance R th and pumping power PP ). Results showed that compared heat transfer amount in the process, convective heat transfer is the dominant part than heat conduction and heat loss. Four representative solutions obtained by k -means clustering divided the Pareto front into five regions, and the moderate R th and PP resulted into the best heat transfer characteristics. Temperature difference between fluid and heating wall decreased from 26 K, obtained for the rectangular microchannel, to 17 K, obtained for the optimized one. The best thermal performance of Case 4 ( e 1 = 0.0572 mm, e 2 = 0.0224 mm) was obtained for a thermal enhanced factor of 1.2305 due to good synergy between the velocity field and temperature gradient. Therefore, rational structural designs efficiently improve heat removal ability in limited heat exchange area.
