Shell element benchmark testing is the structured evaluation of shell finite elements across basic element checks, convergence measures, shell geometries, asymptotic behavior classes, and mesh patterns.
How It Works
The source argues that shell element performance should not be judged only from a displacement at one point. A useful test set should include zero-energy mode checks, membrane and bending patch tests, isotropy checks for triangular elements, convergence curves, and global error measures such as S-norm.
The test problems should also vary the shell's Gaussian curvature, thickness, layer behavior, asymptotic behavior, and element mesh shape. This exposes whether the element is only tuned for a narrow problem class or is robust enough for design analysis.
Benchmark Dimensions
Basic element checks: zero-energy modes, patch tests, and element isotropy.
Geometry: positive, zero, and negative Gaussian curvature.
Asymptotic behavior: bending-dominated, membrane-dominated, and mixed-dominated shell problems.
Error view: global error norms and field distributions, not only one output point.
Mesh sensitivity: element distortion, anisotropic meshes, and orientation effects.