Adaptive mesh refinement
This function allows you to automatically adjust the FE mesh to the requirements of the calculation. The program performs a mesh refinement in several steps. In each step, the FE mesh is recreated based on an error comparison of the results in the previous calculation step. The error indicator for the surface element results is based on the method according to Zienkiewicz-Zhu.
Note
An adaptive mesh refinement eliminates the need to define refinement areas manually. The program does this automatically, reliably, and flexibly.
After selecting the check box, the parameter fields can be accessed. The Maximum number of refinement iterations controls how many steps are carried out during the mesh refinement. The larger the number, the finer the FE mesh becomes in the areas of significant result changes - which can also have an unfavorable effect on the extreme values of singularities. The default setting is therefore only one refinement iteration, which is usually sufficient. The Average error tolerance represents the error indicator for the surface elements described above.
To avoid an FE mesh that is too rough, it is possible to specify a Minimum length of elements.
The error comparison is carried out for a linear structural analysis. To do this, use the list to define a load case or load combination whose results are the basis for adjusting the FE mesh. This FE mesh is then used for the calculation of all other load cases and combinations.
Note
The following FAQs provide more information about adaptive mesh refinement:
https://www.dlubal.com/en/support-and-learning/support/faq/002794
https://www.dlubal.com/en/support-and-learning/support/faq/002922