Analysis And Verification Of 2D Fluid-Thermal Field In Oil-Immersed Transformer Winding Based On Finite Element Method

Oil-immersed power transformers are widely distributed in different capacity and voltage level situations,which undertake the important task of power transmission and voltage conversion in power system.The abnormal temperature rising and hot-spot temperature of power transformer will affect its service life and stable operation.The key to calculate the temperature rising and hot-spot temperature of power transformer is to solve the fluid-thermal coupling problem.The numerical method to solve the fluid-thermal coupling problem of oil-immersed transformer must have high numerical accuracy,stability and convergence.In order to solve the fluid-thermal coupling problem of oil-immersed power transformer and accurately calculate the winding temperature rising,hot-spot temperature and its position,a hybrid finite element method based on dimensionless least-squares finite element method and upwind finite element method is adopted in this paper.In this method,the least-squares finite element method is used to calculate the flow field.The velocity-pressure coupling is treated as direct coupling.Then the upwind finite element method is used to calculate the temperature field.Finally,the sequential iteration method is used to calculate the flow-thermal coupling problem.In the calculation process,the change of physical parameters of transformer oil with temperature and the temperature effect of winding loss are considered.Through the verification of the lid-cavity model and the model of the local windings,it shows that this method can be used to solve the fluid-thermal coupling problem of the transformer windings.Due to the complex internal structure of transformer,the scale of grid and node are very large.In addition,the physical parameters of the transformer oil are quite different,which leads to the problem that the condition number of the stiffness matrix is too large,which seriously affects the convergence of the calculation.To improve the convergence,three preconditioned methods are used in this paper.One is the dimensionless method,which can transform the variables in the control equations into dimensionless ones,and the physical parameters can be transformed by introducing Reynolds number;the other is the two-side equilibration method,which can preprocess the rows(columns)of the stiffness matrix to have the same 1-norm to reduce the condition number;the third is the Jacobian preconditioned conjugate gradient method,due to the stiffness matrix of the least-squares finite element method is a symmetric positive definite sparse matrix,so the iterative method can be used to solve the equations,and then accelerate the convergence and reduce the calculation.The results show that the convergence of the hybrid finite element method is better than that of the Fluent software based on the finite volume method.To verify the effectiveness of the hybrid finite element method used in this paper,a product-level oil-immersed power transformer winding temperature rising and cooling test platform is built.The parameters such as winding loss,inlet flow and outlet pressure are obtained by the instrument,and then the temperatures of the measuring point are measured by the temperature measuring instrument.Combined with the actual model size and other parameters,the corresponding numerical model is established,and then the hybrid finite element method is used to simulate the flow field and temperature field of the model.Finally,the results are compared with the experimental results.The conclusion is that this method has high numerical accuracy.