Research And Application Of Three-dimensional Inverse Design For Axial Fan

How to effectively control the internal flow of the rotor and obtain high-efficiency turbomachinery according to the flow and loss mechanism obtained by flow field analysis is a difficult point in the design process.Since the definite solution condition of the inverse design method is the aerodynamic parameters of the flow field,the relationship between the design input and the characteristics of the flow field is established,so that the designer can directly control the internal flow characteristics of the rotor by adjusting the aerodynamic parameters for more targeted design.Therefore,the inverse design method has received more and more attention,but its development still lags behind the direct design method.Prominent problems that need to be further developed for the inverse design method include:the mesh update caused by blade deformation,which affects the calculation accuracy and efficiency of the inverse design;the efficiency of solving the flow field control equation needs to be further improved;the definite solution conditions directly determine the effect of the inverse design method.And the given measures of the reasonable definite solution conditions have always been the difficulties of the inverse design method.Based on the calculation of the three-dimensional viscous flow field,this paper carries out the research on the inverse design of the axial flow fan,focusing on the mesh update caused by the blade,the efficiency of the numerical solution of the flow field,and the definite solution conditions of the inverse problem.The main work and research results are as follows:(1)Research on adaptive generation and deformation strategy of grid.Aiming at the problem of automatic mesh generation in the calculation process of the inverse problem method,the paper presents a method of using partial differential equations to complete geometric adaptive topological structure division and generating fine meshes through transfinite interpolation.The grid needs to be updated repeatedly.A mesh deformation strategy combining radial basis function and transfinite interpolation method is given.The rapid deformation is achieved by moving the mesh nodes,which reduces the timeconsuming of mesh generation.The return experiment of blade profile verifies the improvement of the calculation accuracy and efficiency of the inverse problem method based on the grid update strategy.(2)Research on accelerated convergence measures for numerical calculation of flow field in inverse problem method.Aiming at the nonlinear characteristics of the flow field control equations in the inverse problem design,the Jacobian free Newton Krylov(JFNK)method is introduced as the solver of the incompressible Navier Stokes equations to accelerate the convergence of the flow field.The effect of the improved preconditioning operator,preconditioning operator velocity relaxation factor,correction term relaxation factor,relative linear tolerance and Krylov subspace maximum iteration steps in the JFNK method on the efficiency of the flow field solution are studied,and the efficiency of the flow field solution is further improved.The influence of the stiffness factor and the camber line update frequency in the airfoil reconstruction relation on the calculation efficiency of the entire inverse problem is analyzed.The calculation efficiency of the JFNK and SIMPLE algorithms as the flow field solver is compared,and the effect of the preconditioning JFNK method in accelerating the convergence is verified.(3)Research on the definite conditions for the design of the inverse problem for the axial fan based on the blade loading distribution.Taking a typical axial fan with a small hub ratio as an object,the conditions for the inverse problem design of the axial fan are explored.The design effects and influence of loading distribution in different span and flow directions are analyzed in detail from the perspectives of the external characteristics of the fan,internal flow characteristics and loss distribution.The results show that for the loading span distribution,an appropriate increase in the tip loading factor can effectively improve the efficiency of the fan,and the stacking at the leading edge can increase the usable design range of the high tip loading factor.Regarding the loading flow distribution,the matching relationship between the leading edge loading and the maximum loading and the influence of the flow direction loading position on the tip leakage flow determine the flow loss and fan efficiency.An inverse design scheme was optimized and determined by combining various influence laws,and the effect of the method was verified by the experimental results.(4)Inverse problem design based on vortex dynamics diagnosis.The vorticity dynamics diagnosis technology-boundary vorticity flux developed in recent years is introduced into inverse problem design as the basis for modifying the loading distribution.According to the flow field diagnosis,the dynamic root of the adverse flow in the rotor is found,and the loading is adjusted through the inverse problem design,and the revised rotor model is obtained.Combined with the design example of the axial flow fan,the clear correspondence between the boundary vorticity flux and the loading distribution,loss,and boundary layer is demonstrated.The complex and adverse flow shown by the boundary vorticity flux diagnosis is reflected in the increase of loss and the change of the boundary layer.The flow structure of the rotor obtained by inverse design with modifying the loading distribution has been significantly improved,and the improvement effect of the aerodynamic performance of the improved fan is verified through the comparison of the aerodynamic performance test.