Personal tools

Error Estimation and Adaptive Meshing

Creating the perfect mesh for EM simulation problems is often a daunting task for novice and experienced users alike. 

FEKO's automated meshing assists the user by taking into account the frequency, solution method, media properties, curvature and electrical size of a model and creating a fine, standard or coarse mesh accordingly. This allows for rapid meshing, while custom meshing is available for the manual definition of mesh parameters such as the size of mesh elements and the adjustment of advanced mesh properties e.g. the mesh growth rate.

jerusalem_cross_auto_mesh_coarse.png  jerusalem_cross_auto_mesh_standard.png  jerusalem_cross_auto_mesh_fine.png

Automatic meshing of a Jerusalem Cross FSS element (coarse, standard and fine).

Mesh refinement rules can be set in problematic areas that the user is aware of. However, the right balance must be sought between meshing finely in critical areas of the model and using as few (course) mesh elements as possible in other areas of the model to save on simulation time and memory requirements, while ensuring accurate answers.

FEKO's error estimation routine is an a posteriori error indicator that will help users refine mesh parameters by estimating the quality of a mesh. Areas with high error values can easily be refined before further simulation runs by using adaptive mesh refinement rules.

jerusalem_cross_error_estimates.png  jerusalem_cross_adaptive_mesh_refinement_rules.png  jer_cross_mesh_adapted.png 

Error estimates calculated for a Jerusalem Cross FSS element, indicators showing the associated adaptive point refinement rules and the resulting mesh.



Surface meshes:  Large errors are likely to occur near small gaps in complex models.



Volume meshes:  Largest errors occur near the septa in a waveguide model.