POD Analysis Of Flow And Heat Transfer Performance Of Flat Tube Fin Heat Exchanger

Traditional numerical simulation methods have shortcomings of large amount of calculation,longer calculation time and high requirement for computer capability,so traditional numerical simulation methods can not meet development requirement of modern industry.The Proper Orthogonal Decomposition(POD)reduced-order model is an efficient numerical method and can shorten calculation time.The method of combining body-fitted coordinates and POD is used to calculate reduced-order heat transfer units of selected flat tube fin heat exchanger and improve calculation speed under the premise of ensuring calculation accuracy.The method can improve the efficiency of numerical design for heat exchangers with complex structures.This method can meet the actual needs of engineering in more fields.The Proper Orthogonal Decomposition(POD)reduced-order model of flat tube fin heat exchanger is established.And its calculation results are compared with calculation results of the finite volume method(FVM).The main research contents are as follows.(1)In simulation computation of using FVM to calculate heat transfer units in flat tube fin heat exchanger,modified parameters are fin spacing,transverse tube spacing and air side Reynolds number.Conditions of single-parameter variables,two-parameter variables and three-parameter variables are calculated under boundary conditions of uniform wall temperature and uniform heat flux.According to simulation results,distribution characteristics of temperature fields and velocity fields are analyzed.(2)Samples needed for POD are obtained from calculation results of FVM.And samples include simulation conditions of single-parameter variables,two-parameter variables and three-parameter variables under boundary conditions of uniform wall temperature and uniform heat flux respectively.So samples have six groups.The snapshot method is used to calculate sample variable conditions with the proper orthogonal decomposition and obtain corresponding basis functions.The linear interpolation method,Newton interpolation method and Lagrangian interpolation method are adopted to solve spectral coefficient for non-sample single-parameter variable conditions.Basis functions and corresponding spectral coefficients are linearly superposed to obtain reconstructed fields.The linear interpolation method is adopted to calculate spectral coefficients for two-parameter variable conditions and three-parameter variable conditions.And reconstructed fields are obtained.(3)Temperature fields and velocity fields calculated by POD and FVM under the same conditions are compared.And accuracy and rate of the POD interpolation method are analyzed.Simulation computation results show that when multivariate problems are calculated by the POD interpolation method,the relative deviation increases with the number of variables between the POD and FVM.The maximum error of parameter variables of reconstructed velocity fields is 1.91% of three-parameter variable of reconstructed velocity fields.The maximum error of parameter variables of reconstructed temperature fields is 0.563% of three-parameter variable of reconstructed velocity fields.The calculation accuracy of the uniform heat flux boundary conditions is higher than that of the uniform wall temperature boundary conditions.Numerical calculation rate of traditional FVM method is accelerated to 3093.4 times under the premise of ensuring calculation accuracy by the way of adopting the POD method.So the POD linear interpolation method not only has high calculation accuracy but also get fast calculation speed when used to calculate temperature fields and velocity fields.In addition,the POD interpolation method has little difference in the calculation accuracy under the conditions of equal heat flow boundary and the same wall temperature.The POD technology is applied to the trial of complex flow and heat transfer.Acceleration characteristics is magnified in solving parameters of velocity fields and temperature fields of heat exchangers with complex structures.The results give theoretical reference for optimal design of flat tube fin heat exchangers.And the research work has important engineering guiding significance.