STUDY ON THE HAINGING NODES IN THE FINITE-COVER-BASED MESHFREE METHOD

For the meshfree methods based on the Galerkin method, all nodes are distributed in the computational domain or on domain boundaries. This study explores the feasibility of using hanging nodes for meshfree methods as the nodal arrangement is independent of background cells. A unified and uniform strategy of the nodal arrangement is proposed, along with a corresponding scheme for the distribution of background cells. Numerical examples are employed to discuss the following influential factors: the necessity of using hanging nodes, the number of hanging nodes, the shape and size of nodal influential domain, and the background integral scheme. Results show that the proposed method can greatly improve the accuracy of meshfree methods, especially in the calculation of stresses near domain boundaries.