Mathematical Modeling Of Thermal Conductivity Of Nanofluids And Optimization Of Microchannel Structure Parameters

In the fields of energy and power,micro-electronic technology,military nuclear power and other engineering fields,there are many heat transfer problems that need to be solved urgently.With the development of micro-nanoscience and technology,more and more efficient heat transfer fluids and structures,such as nanofluids and micro-channel structures,are being used to rapidly remove excess heat from heat transfer devices during convective heat transfer.However,the heat transfer performance of nanofluids and microchannels is affected by many factors.Therefore,it is of great importance to improve the modeling method of thermal conductivity and micro-channel structure of hybrid nanofluids.In this paper,a combination of experiment,simulation and theory is used.Based on the experimental data of the thermal conductivity of the hybrid nanofluid and the simulated value of the microchannel structure,the Artificial Neural Network(ANN)in Machine Learning(ML)is introduced.Multi-Objective Evolutionary Algorithm(MOEA)builds mathematical models and optimizes them separately,providing a theoretical basis for micro-nano enhanced heat transfer research.It mainly includes the following three aspects:(1)A two-step method was used to prepare a Cu-Al₂O₃/water-glycol mixed nanofluid with a mass fraction of 1%,and the factors affecting its stability and thermal conductivity were studied.The important factors affecting the stability of the hybrid nanofluids were analyzed by precipitation and electron transmission electron microscopy(TEM).The results show that when the mass ratio of water to ethylene glycol in the base fluid is 50:50 and the ultrasonic time is 30 minutes,the stability of the hybrid nanofluids is the best;meanwhile,its thermal conductivity depends on the nanoparticle volume fraction,mixing ratio and temperature Increase and increase.(2)Artificial neural network is used to mathematically model and optimize the hybrid nanofluids.Taking mixing ratio and temperature as research objects,BP neural network(BPNN),genetic algorithm optimized BP neural network(GA-BPNN),radial basis neural network(RBFNN),and mind evolutionary algorithmoptimized BP neural network(MEA-BP)Predict the change of thermal conductivity of Cu-Al₂O₃/water-glycol nanofluid and compare it with multiple linear regression(MLR)model.The study found that the determination coefficient(R²)and root mean square error(RMSE)of the five models were 0.0035,0.9983;0.0034,0.9995;0.0048,0.9974;0.0031,0.9997;0.0107,0.9984.Among them,the prediction accuracy of the MEA-BPNN model is the highest,indicating that this algorithm can effectively realize the data-driven modeling of the thermal conductivity of the hybrid nanofluids.(3)The structural parameters of the circular cavity and internal rib microchannels were used as independent variables,and the heat transfer efficiency and pressure drop were used as the dependent variables.The Fluent software was used to simulate 30groups of microchannels with arbitrary combinations of cavity and internal rib heights,and a mathematical model of micro-channel structure optimization was constructed with multi-objective genetic algorithm(MOEA).The research results show that,compared with the reference channel,the optimized microchannel structure(e₁=0.0368mm,e₂=0.0193mm)has a more uniform temperature field distribution and better overall heat transfer performance(enhanced heat transfer factor?=1.23).