基于多目标遗传算法的串联通道水冷散热器优化设计

Optimization of Serpentine Channel Heat Sink Based on Multi-objective Genetic Algorithm

为最大限度的提高水冷散热器的散热效率,需要对外形尺寸固定的水冷散热器进行优化设计。将串联通道散热器按通道数离散成N个并联的热阻单元,各个热阻单元之间用焓变热阻相连,并以此为基础建立散热器的热阻网路模型及压力损失模型。以热阻和压力损失模型为目标函数,在层流状态下,采用多目标遗传算法对散热器的结构参数,如通道数、通道宽度和高度以及入口水流速度进行优化,通过多次迭代得到Pareto优化解。采用计算流体力学仿真方法验证散热器的热阻模型和压力损失模型的准确性,并从优化解集中选择部分解进行比较分析。这种方法能够同时降低串联通道散热器的热阻和压力损失,为串联通道水冷散热器的优化设计提供理论依据,在工程上也具有很好的指导意义。

In order to improve the efficiency of serpentine channel heat sink, optimization design must be done for it on the condition that outline dimensions of the heat sink（length, width and height） are chosen. The thermal resistance network model and pressure loss model are proposed based the discrete thermal resistance unit, which are parallel of N channels. Thermal resistance and pressure loss model are selected as objective functions, and structural parameters of heat sink are optimized in laminar flow range based on multi-objective genetic algorithm with fixed length, width and height. The variables include the number of channels, channel width and height and the velocity at the inlet. The Pareto optimal solutions are obtained after much iteration. Computational fluid dynamics simulation is performed to validate the correctness of thermal resistance model and pressure drop model. This method not only can provide some theoretical base for the optimization design of serpentine channel water-cooling heat sink, but also can be used in the engineering.