Robust mesh generation for electromagnetic models with geometric defects through node alignment and mesh boolean operations.

This paper presents a mesh fault-tolerant repair algorithm tailored to address common geometric issues inherent in electromagnetic models. During the actual production design phase, electromagnetic models frequently display a variety of geometric imperfections due to factors such as design errors and model complexity. These imperfections encompass surface leakage, interpenetrating assemblies, contact, excess material, and minute gaps among multiple components. To tackle these geometric challenges and their associated mesh requirements in electromagnetic models, this study devises a sophisticated mesh fault-tolerant repairing algorithm process. Initially, the algorithm employs an enhanced bidirectional alignment technique to conduct global alignment operations on surface meshes. This method synchronously aligns curves and surfaces across different levels and permits the customization of tolerances for mesh alignment and node fusion in distinct regions, thereby ensuring mesh compatibility between adjacent planes and assemblies within the electromagnetic model. Following this, a watertight repair algorithm is utilized to rectify mesh leaks, effectively resolving issues such as holes and degenerate triangles. Lastly, a mesh Boolean algorithm grounded in large integer representation is applied to address problems related to interpenetrating assemblies and contact in geometric defects, ultimately generating compatible, watertight meshes. The mesh alignment process not only enhances watertight repair and mesh Boolean operations but also contributes to the overall quality of the mesh. Through this comprehensive approach, the method generates higher-quality meshes for electromagnetic models with geometric imperfections while preserving the original model's geometric characteristics to a significant extent. Additionally, the method exhibits robustness, with repaired models achieving a high success rate in tetrahedral filling. In comparison with several open-source mesh generation tools and a set of commercial mesh generation tools, this algorithm demonstrates superior mesh generation capabilities.