Research On Topology Optimization Design Of Microchannel Heat Sink Based On Fluid Structure Thermal Coupling

With the advent of the global commercial 5G（5th generation mobile networks）era,the semiconductor field has ushered in another rapid development.The integration of 5G chips and modules and the density of parts have greatly increased,and the power consumption and heat density of 5G equipment have increased rapidly.Therefore,if we want to ensure the leading position of 5G,we must solve the problem of thermal management.Multiple inlet and outlet micro-channel heat sink can effectively reduce the average temperature of electronic equipment and improve temperature uniformity,but for the traditional structure design method,it cannot meet the high heat flow requirements of 5G chip module.Therefore,this paper adopts flexible and highly targeted topology optimization technology to design micro-channel heat sink,and conducts research and analysis of related fluid flow and heat transfer characteristics through numerical simulation,and further explores the performance potential of radiator through parameter optimization.The analysis and verification of experiments have been carried out,and the specific work results are as follows:First,through the combination of COMSOL Multiphysics 5.5 and MATLAB2017a,the design domain of multiple inlets and multiple outlets is topologically optimized,and the two objective functions of the lowest average temperature and the best temperature consistency are set,and passed The method of numerical fitting determines the best objective function,and finally obtains the two-dimensional flow channel configuration of the lowest average temperature model with a fluid channel volume fraction of 0.6 and the best temperature consistency model.The lowest average temperature model has a small number of channels.Larger solid fins have a slightly lower average temperature and the best temperature consistency model.The solid fins are small in size,have a large number of channels,and have better temperature uniformity.Secondly,through solidworks2017,the two flow channel configurations generated by topology optimization were converted into three-dimensional radiator models,and the lowest uniform temperature model and the optimal model of temperature consistency were performed with different pump powers through ANSYS Fluent 15.0 simulation software Characteristics research,and contrast with the traditional design of the annular radiator,respectively from the average Nusselt number,average pressure drop,temperature difference,thermal resistance and other aspects of specific research,the results show that:under the same pump power,topology optimization The designed radiator has low average temperature,small temperature difference,small pressure drop,and has better overall performance.Between the two topological optimization models,the temperature consistency is the best model,with the largest Nusselt number,the smallest temperature difference,and the smallest thermal resistance.The average temperature and average pressure drop are close to the lowest average temperature model,and the overall heat dissipation performance is the best.Thirdly,through the mathematical software MATLAB2017a,optimize the structural parameters of the radiator channel height W_ch and the bottom plate thickness W_b of the optimal temperature consistency model,use curve and surface fitting tools to fit the simulated data,and perform single-objective optimization through genetic algorithms.The final optimal structure parameters are radiator channel height W_ch=6.2mm and bottom plate thickness W_b=1.5mm.Compared with the simulation results,the optimization error is 3.39%.Under the same Reynolds number,the optimized radiator Nusselt number increased by157.24%.Finally,set up an experimental platform to conduct experimental research on the flow and heat transfer characteristics of the optimal temperature consistency model,record relevant data and analyze them.Compared with the simulation results,the maximum error of Nusselt number is 4.2%and the maximum error of average pressure drop is 7.75%,which is in good agreement.At the same time,the uncertainty analysis of the Nusselt number in the experiment shows that the maximum uncertainty of the Nusselt number is 4.37%.