Solid Structure Damage Analysis with Segmented Quasi-Bond Finite Elements

The patent segments the structure into finite element regions and quasi-bond regions based on crack risk assessment. Only quasi-bonds in high-risk regions are traversed and analyzed, while continuous regions use traditional finite element methods. This segmentation reduces the number of quasi-bonds that need to be processed from all quasi-bonds to only those in potential crack regions, significantly improving calculation efficiency while maintaining simulation accuracy where it matters most.

The patent applies different calculation methods to different regions: quasi-bond method with full traversal for regions prone to cracking, and traditional finite element method for continuous regions. This local differentiation ensures high simulation accuracy in critical areas while avoiding unnecessary computational overhead in non-critical areas, resolving the contradiction between accuracy and efficiency.

The patent extracts and removes the non-local interaction calculation from the overall system by identifying and eliminating redundant calculations. By using a standardized stiffness matrix assembly process and avoiding repeated non-local interaction computations for all particles, the calculation cost is significantly reduced while preserving the essential micro-crack propagation simulation capability in critical regions.

The patent applies the computationally intensive quasi-bond method with non-local interaction only partially - specifically in regions where crack propagation is likely to occur - rather than excessively applying it to the entire structure. This partial application maintains simulation accuracy for micro-crack propagation where needed while avoiding unnecessary calculation costs in regions where such detailed analysis is not required.