Hybrid Approach of Radial Basis Function and Finite Element Method for Electromagnetic Problems

2015 ◽  
Vol 51 (3) ◽  
pp. 1-4 ◽  
Author(s):  
Yang Zou ◽  
Gang Lei ◽  
Keran Shao ◽  
Youguang Guo ◽  
Joe Zhu ◽  
...  
2021 ◽  
Vol 13 (01) ◽  
pp. 2150007
Author(s):  
Shahram Hosseini ◽  
Gholamhossein Rahimi

This paper investigates the nonlinear bending analysis of a hyperelastic plate via neo-Hookean strain energy function. The first-order shear deformation plate theory (FSDPT) is used for the formulation of the field variables. Also, the nonlinear Lagrangian strains are considered via the right Cauchy–Green tensor. The governing equations and nonlinear boundary conditions are derived using Euler–Lagrange relations. The meshless collocation method based on radial basis function is used to discretize the governing equations of the hyperelastic plate. Square and circular plates are studied to evaluate the accuracy of the meshless collocation method based on thin-plate spline (TPS) and multiquadric (MQ) and logarithmic thin-plate spline (LTPS) radial basis function. Also, the results of the meshless method are compared to those of the finite element method. In some cases, the meshless method is more efficient than the finite element method due to no meshing. The linear and nonlinear natural boundary conditions are directly imposed on the stiffness matrix and are compared to each other. The maximum differences between linear and nonlinear natural boundary conditions are 1.43%. The von-Mises stress using meshless collocation method based on TPS basis function is compared to those of the finite element method.


2013 ◽  
Vol 2013 ◽  
pp. 1-13
Author(s):  
Xia Xiaozhou ◽  
Zhang Qing ◽  
Wang Hong ◽  
Jiang Qun

In the frame of the extended finite element method, the exponent disconnected function is introduced to reflect the discontinuous characteristic of crack and the crack tip enrichment function which is made of triangular basis function, and the linear polar radius function is adopted to describe the displacement field distribution of elastoplastic crack tip. Where, the linear polar radius function form is chosen to decrease the singularity characteristic induced by the plastic yield zone of crack tip, and the triangle basis function form is adopted to describe the displacement distribution character with the polar angle of crack tip. Based on the displacement model containing the above enrichment displacement function, the increment iterative form of elastoplastic extended finite element method is deduced by virtual work principle. For nonuniform hardening material such as concrete, in order to avoid the nonsymmetry characteristic of stiffness matrix induced by the non-associate flowing of plastic strain, the plastic flowing rule containing cross item based on the least energy dissipation principle is adopted. Finally, some numerical examples show that the elastoplastic X-FEM constructed in this paper is of validity.


2013 ◽  
Vol 2013 ◽  
pp. 1-9
Author(s):  
Wei Liu ◽  
Feifan Wang ◽  
Xiawei Yang ◽  
Wenya Li

This paper addresses the upset prediction problem of friction welded joints. Based on finite element simulations of inertia friction welding (IFW), a radial basis function (RBF) neural network was developed initially to predict the final upset for a number of welding parameters. The predicted joint upset by the RBF neural network was compared to validated finite element simulations, producing an error of less than 8.16% which is reasonable. Furthermore, the effects of initial rotational speed and axial pressure on the upset were investigated in relation to energy conversion with the RBF neural network. The developed RBF neural network was also applied to linear friction welding (LFW) and continuous drive friction welding (CDFW). The correlation coefficients of RBF prediction for LFW and CDFW were 0.963 and 0.998, respectively, which further suggest that an RBF neural network is an effective method for upset prediction of friction welded joints.


2010 ◽  
Vol 670 ◽  
pp. 311-317
Author(s):  
T. Arudchelvam ◽  
D. Rodger ◽  
S.R.H. Hoole

An enhanced multi-grid method eliminating the error correction process of the conventional multi-grid method is presented for solving Poissonian problems and tested on two simple two-dimensional magnetostatic field problems. The finite element method (FEM) was used to solve for the vector potential in a sequence of grids. The gains in computation time are shown to be immense compared to the standard multi-grid methods, especially as the matrix system grows in size. These gains are very useful in solving electromagnetic problems using the finite element method.


Sign in / Sign up

Export Citation Format

Share Document