Implementation Procedures of Parallel Preconditioning with Sparse Matrix Based on FEM

Guo Liang Ji; Yang De Feng; Wen Kai Cui; Liang Gang Lu

doi:10.4028/www.scientific.net/AMM.166-169.3166

Paper Titles

Experimental Verification of a Full-Size Buckling Restrained Brace
p.3147

Thermal Stress Analysis of a Certain Concrete Wind Cover
p.3151

Correct Wavelet-Based Multilevel Numerical Method of Local Solution of Boundary Problems of Structural Analysis
p.3155

Tests of Buckling-Restrained Braces Using Low-Yield Point Steel as Core Material
p.3159

Implementation Procedures of Parallel Preconditioning with Sparse Matrix Based on FEM
p.3166

Experimental Research on Mechanical Properties of Fundamental Frame Units in Multi-Ribbed Composite Wallboard
p.3174

Study on Bending-Torsional Coupling Vibration of Intermediate Axis
p.3180

Mechanical Behavior of Concrete-Filled Double Steel Tubular Columns under Axial Compression Load
p.3184

Study on the Influence of M Value to Internal Force and Deformation for Transverse Bent of High-Piled Wharf
p.3189

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 166-169Implementation Procedures of Parallel...

Implementation Procedures of Parallel Preconditioning with Sparse Matrix Based on FEM

Abstract:

A technique to assemble global stiffness matrix stored in sparse storage format and two parallel solvers for sparse linear systems based on FEM are presented. The assembly method uses a data structure named associated node at intermediate stages to finally arrive at the Compressed Sparse Row (CSR) format. The associated nodes record the information about the connection of nodes in the mesh. The technique can reduce large memory because it only stores the nonzero elements of the global stiffness matrix. This method is simple and effective. The solvers are Restarted GMRES iterative solvers with Jacobi and sparse appropriate inverse (SPAI) preconditioning, respectively. Some numerical experiments show that the both preconditioners can improve the convergence of the iterative method, and SPAI is more powerful than Jacobi in the sence of reducing the number of iterations and parallel efficiency. Both of the two solvers can be used to solve large sparse linear system.