Creating high-quality solid meshes is critical to simulation accuracy; this article gives examples of effective meshing techniques. Successful meshing requires watertight surface inputs and knowing the appropriate meshing algorithm to apply.
FEA meshes require a surface mesh input. Discretizing complex implicit geometry may cause irregularities. However, we can take steps to promote quality results.
- Mesh from Implicit Body is the default option for creating a mesh body and highly capable of capturing features.
- Mesh from CAD Body discretizes CAD geometry when implicit nTop geometry is not required. If the imported CAD part has errors or holes, this may cause issues with the resulting mesh.
- Remesh Surface gives more control over the size, shape, and uniformity of the original mesh. Use this block to generate FEA quality meshes for downstream analysis.
- Simplify Mesh by Amount decreases the polygon count of an object by a given percentage, also a way to shrink the file size.
- Simplify Mesh by Threshold decreases the polygon count of an object by a given threshold, also a way to shrink the file size.
Using a minimum feature size improves the robustness of the Remesh Surface block. The example file uses 5% of the edge length as a starting point.
Voxel translations boost the quality of the surface mesh but introduce tolerance to the geometry:
- Apply Voxel Grid from Implicit Body along with Mesh from Voxel Grid to create a watertight surface body. Note that this function eliminates sharp edges.
For creating a solid mesh, two separate blocks employ distinct meshing algorithms:
- Robust Tetrahedral Mesh is a robust solution and exceptionally tolerant of low-quality inputs and ensures FEA quality elements. While this technique is remarkably robust, speed is a tradeoff.
- Volume Mesh is the preferred method for simple geometry, as it is speedy but less tolerant of defects. Volume Mesh may also work after taking steps to repair input errors.
Finally, meshing operations may demand significant memory resources, for optimal performance we recommend 32 GB RAM.