Numerical Simulation Of Fluid-structure Coupling Of Riser Vortex Induced Vibration

With the rapid development of integrated circuit technology,the development trend of electronic chips tends to be highly integrated,miniaturized and multi-core,and the heat generated by the chip during work rises sharply;At the same time,the working conditions of the chip in actual work have been changing,the heat has been changing,and the working conditions of each core in the electronic chip are also different.If the same heat dissipation method is used for different cores,the temperature difference between the cores will cause temperature stress,resulting in reduced performance of the chip,reducing the service life of the chip,and even damaging the chip.Guided by the distributed heat dissipation of the array chip,this paper designs a semi-open microchannel with multiple entrances and single outlets to solve the heat dissipation problem of the array chip,and according to the simulated values of the microchannel in Fluent,the multi-objective genetic algorithm and the artificial neural network in machine learning are introduced to optimize the structure and construct a regression model respectively.It provides a solution to the distributed heat dissipation problem of array chips.It mainly includes the following three aspects:(1)The structural parameters of the staggered trapezoidal rib microchannels were optimized.Taking the front wedge angle and rib width coefficient of the trapezoidal rib in the microchannel as the independent variables,and the efficiency and pressure drop of heat transfer as the dependent variables of the study,the microchannel results of 16 groups of arbitrary combinations of front wedge angle and rib width coefficient were simulated by using Fluent simulation software,and the optimization model of microchannel structure parameters was constructed by using the NSGA-II multiobjective genetic algorithm.The results show that compared with the channel before optimization,the optimized microchannel structure has better heat transfer performance under the same pump power,the Nu number is increased by 30%,and the temperature distribution in the microchannel is more uniform.The comprehensive heat transfer performance is better.(2)Based on the optimized microchannel structural parameters,a semi-open multiinlet single-outlet staggered trapezoidal ridge microchannel was designed,the relationship between the inlet velocity difference and the distributed heat source temperature field was analyzed,4 working conditions were determined,a total of 5000 sets of data sets under different working conditions were collected through Fluent simulation,and a BP neural network optimized by genetic algorithm(GA-BPNN)was introduced,and a regression model between the heat flux density of distributed heat sources and different inlet fluid velocities under 4 working conditions was established.(3)The completed regression model is programmed in C++ and linked to the dynamic link library in Fluent to complete the fusion test simulation.The complex situation of sine wave variation and rectangular wave change of heat flux density of distributed heat source under four working conditions was tested.The simulation results show that the established regression model has good performance under four working conditions,and the inlet flow rate can control the temperature of different heat sources within 300K～302K under the condition of 0.021195 LPM.