基于拓扑优化的风冷热沉研究

Research on Air-cooled Heat Sink Based on Topology Optimization

随着电子芯片的集成度越来越高,相应的热管理问题也愈加严峻,对热沉进行有效设计则成为一种很好的解决办法。采用变密度的拓扑优化方法对强制风冷热沉进行优化设计,建立由导热基座层和热流设计层组成的伪三维优化模型,并考虑更有实际意义的对流边界和利用RAMP法对热导率和对流传热系数进行插值;以压降最小化为优化目标,以空气吸收热Q最小值为2W为热约束,分析三组基座温度Tb(314K、330K、350K)下的优化结果。利用拉伸三维热沉的数值模拟验证伪三维优化模型的物理有效性。结果表明,基座温度Tb升高过程中,翅片四周光滑度有所提升但表面积在缩小,占设计区域的比例分别为18%、16%、14%;目标函数φ逐渐降低,分别为0.073、0.036、0.025;设计区域内速度最大值逐渐变小但速度和温度分布更加均匀。三维热沉模拟结果中沿Z轴的速度和温度变化量很小,从而验证伪三维优化模型的物理有效性。

As the integration of electronic chips becomes higher and higher,the corresponding thermal management problems become more and more serious,and the effective design of heat sinks has become a good solution.The variable density topology optimization method is used to optimize the design of the forced air-cooled heat sink,and a pseudo three-dimensional optimization model consisting of a thermally conductive base layer and a heat flow design layer is established;more practical convection boundaries are considered,and the RAMP method to interpolate the thermal conductivity and convection heat transfer coefficient is used.Taking the minimization of pressure drop as the optimization goal,and taking the minimum value of air absorption heat Q as 2 W as the thermal constraint,the optimization results under three sets of base temperature Tb(314 K,330 K,350 K)are analyzed.Finally,the numerical simulation of the tensile three-dimensional heat sink is used to verify the physical validity of the pseudo three-dimensional optimization model.The results show that during the process of increasing the base temperature Tb,and the smoothness of the fins is improved but the surface area is shrinking,accounting for 18%,16%and 14%of the design area;the objective functionφgradually decreases,respectively 0.073,0.036,0.025;the maximum velocity in the design area gradually decreases,but the velocity and temperature distribution are more uniform.The velocity and temperature changes along the Z axis in the 3 D heat sink simulation results are very small,thus verifying the physical validity of the pseudo 3 D optimization model.