Mesh and Simulation

The way a structure is discretized strongly influences the accuracy of the results. The computed results converge towards the "continuous world” results when the mesh size gets smaller and smaller until the discrete lengths vanish or, being more exact, become differentials. So, generally speaking the structure and the fields are simulated best with a very fine mesh.

Mesh and Simulation Time

As far as the results are concerned, the finer the mesh, the better. However, a very fine mesh has a large number of mesh cells and with this a large number of unknowns to be solved. Every increase of unknowns extends the need of memory and simulation time.

For the increase of simulation time, there are two different mechanisms. One is quite obvious: the more unknowns that are present, the more operations must be computed. The other mechanism is caused by the properties of the solver. The solvers converge slower with smaller mesh step sizes than with larger ones.

.

See also

Mesh Generation Overview (Hexahedral), Mesh Generation Overview (Tetrahedral), Mesh View (Hexahedral), Mesh View (Tetrahedral), Mesh Properties (Hexahedral), Mesh Properties (Tetrahedral), Adaptive Mesh Refinement (Hex), Adaptive Mesh Refinement (Tet), Which Mesh to Use