This study aims to address the persistent challenges of high labor costs and limited robustness in conventional manual mesh generation for CAE, and to achieve the “process invisibility” of mesh generation as prescribed by the NASA CFD Vision 2030 Study.
The national numerical wind tunnel mesh generation software NNW-GridStar implements a systematic framework. For CAD linkage issues, a virtual geometry-based defect repair framework resolves dirty geometry through gap filling, patch merging, and topology reconstruction. For mesh quality optimization, surface remeshing and anisotropic optimization algorithms are integrated. For mesh generation, conformal geometry theory and AI methodologies are leveraged to explore automatic structured surface mesh generation. A comprehensive approach enables high-quality layer-preserved boundary layer mesh generation, while a building-block approach automates spatial topology construction. A parameterized configuration framework supports domain-specific customization.
The application of these technologies significantly reduces mesh generation time while enhancing mesh quality consistency. Comparative experiments with commercial software validate the effectiveness and feasibility of the proposed methodologies.
This work presents a systematic technology framework specifically designed to overcome critical challenges in automated mesh generation. Its novelty lies in the integration of virtual geometry repair, advanced surface mesh optimization techniques, the exploration of AI-driven structured surface meshing, layer-preserved boundary layer generation, automated spatial topology building, and parameterized customization, collectively advancing towards the goal of invisible mesh generation for CAE.
