CST2013: Adaptive Mesh Refinement (Transient)

Simulation: Solver Start Simulation Adaptive mesh refinement, Properties

The adaptive mesh refinement properties dialog box allows you to change the parameters used for the three-dimensional adaptive mesh refinement. Please note that the adaptive mesh refinement is only active when the corresponding button is checked in the transient solver control dialog box.

The adaptive mesh refinement is an automatic scheme to create a mesh better suited for the given problem. With the refined mesh a new calculation pass will be started. Here, the initial mesh is automatically generated by the expert system and is used for the first pass calculation. Afterwards, the adaptive mesh refinement improves the mesh until the stop criteria are satisfied.

As stop criteria, either the change in the S-parameters or the change of a user defined 0D result template from one pass to the next are available.

For the refinement of the meshes, two different strategies are available: The energy based mesh refinement stores the energy density distribution within the structure during a calculation. Based on these data, the mesh is refined afterwards in regions with high energy density. In contrast to this, the strategy based on the expert system successively changes the settings of the mesh expert system, so that finally an appropriate mesh is obtained that can be used afterwards for further parameter studies without activating the adaptive meshing again.

Number of passes frame

Minimum: This is the minimum number of passes that will be performed. The mesh adaptation will run for this minimum number of passes, even if the stop criteria are already met. The minimum setting is 2 passes.

Maximum: This setting determines the maximum number of passes to be performed for the mesh adaptation. The mesh adaptation will stop after this maximum number of passes, even if the stop criteria for the mesh adaptation have not been met. This setting is useful to limit the total calculation time to reasonable amounts. The minimum setting is 2 passes.

Adaptation stop criteria frame

Adapt to S-parameters: Use the change in the S-parameters as stop criterion. The change in the S-parameters has to be below a set threshold (Maximum delta) for a number of consecutive passes (Number of checks) for the stop criterion to be met.

The change in the S-parameters is determined as the maximal deviation of the absolute value of the complex difference of the S-parameters between two subsequent passes.

Keep in mind that a small shift in a pole location may result in a strong change of the S-parameters at the pole’s frequency and thus a large maximal deviation, especially in strongly resonant structures.

This feature is only available if S-parameters will be calculated.

Use full frequency range: If activated, the full frequency range is used for the calculation of the change in the S-parameters. If not activated, you can limit the frequency range by setting the values in the Fmin / Fmax edit fields.

Fmin / Fmax: You can limit the frequency range used for the calculation of the change in the S-parameters here.

Maximum delta: This sets the threshold for the change in the S-parameters.

Number of checks: This sets the number of consecutive passes for which the threshold for the change in the S-parameters has to be below the set threshold for the stop criterion to be met.

Adapt to 0D result template: Use the relative change in a user defined 0D result template as stop criterion. The relative change in the 0D result has to be below a set threshold (Maximum relative delta) for a number of consecutive passes (Number of checks) for the stop criterion to be met.

0D result template name: Here you can select the 0D result template to use as stop criterion. Only 0D result templates previously defined in the template based postprocessing dialog are available for selection.

Maximum relative delta: This sets the threshold for the relative change in the 0D result template.

Number of checks: This sets the number of consecutive passes for which the relative change in 0D result template has to be below the set threshold for the stop criterion to be met.

Refinement settings frame

Refinement strategy

Determine here the strategy used for the adaptive mesh refinement. Two different possibilities are available, either based on the modification of the mesh expert system or determined by regions of high field energy in the calculation domain.

Expert system based

Choosing this strategy, the mesh refinement is performed by successively changing the settings of the mesh expert system. This offers the possibility of running the mesh adaptation only once to generate an appropriate mesh line distribution for the current simulation model. This optimized mesh can be used afterwards for further calculation runs with modified parameters (e.g., parameter sweeps or optimization cycles) without activating the refinement process again.

Mesh increment: This parameter determines the changing of the mesh expert system by increasing the settings of Lines per wavelength and Lower mesh limit in the Mesh Properties dialog box.

Energy based

Choosing this strategy the mesh refinement is based on the given energy distribution of the electromagnetic fields inside the calculation domain.

Factor for mesh cell increase: This factor determines how many new cells are introduced between two subsequent passes of the mesh refinement. A setting of 0.5 means that 0.5 times more mesh cells are used for the next calculation than have been used for the previous one. In other words, the number of mesh cells increases about 50% from pass to pass.

Number of pulse widths to skip: The behavior of a narrow bandwidth structure is often investigated within a broader frequency band to keep the pulse lengths reasonably short. However, in these cases a big part of the energy fed into the structure is reflected almost immediately inside the device. Only a small part of the energy remains in the structure and is used to investigate the device’s behavior. However, because this small part is the actually important one, it may be misleading if this energy is summed up with the large (but reflected) part to determine the locations for the mesh refinement. In these cases, the mesh refinement would lead to strange refinements in the feeding lines without refining the critical parts inside the device. In such cases, it is often useful to start the energy calculation after the reflected energy has left the structure. Therefore, the number of pulse widths can be specified after which the investigation of the energy will be started. For narrow band structures, it is common to set this value up to 1-3. On the other hand, the value should remain at zero for wide band structures.

Weight for electric field energy: The electromagnetic field energy is stored during the calculation and the mesh refinement is based upon the energy density within the structure. However, sometimes a structure’s behavior is more critically coupled to changes of the electric field than to changes of the magnetic field. If this behavior is previously known, it is useful to increase the weight of the electric energy compared to the magnetic energy. By default, both electric and magnetic energies have the same weights.

Weight for magnetic field energy: As described above for the electric energy weight, it is sometimes useful to increase or decrease the weight for the magnetic field energy when it is previously known that the electric or the magnetic field energy is more critical to the device’s performance.

Refinement directions: If a button is checked, the mesh will be refined along the corresponding coordinate direction. This option is useful to avoid refining a mesh along coordinate directions in which the structure’s fields have no dependency (e.g., some quasi-2D structures).

OK

Accepts the changes and closes the dialog.

Defaults

Resets all dialog parameters to the their default settings.

Cancel

Closes this dialog box without performing any further action.

Help

Shows this help text.

See also

Transient Solver Properties