Adaptive Methods

Adaptive methods are based on error estimation. There are mainly two methods to improve the accuracy of the results of static studies:

The H-Method

The concept of the h-method is to use smaller elements in regions with high errors. After running the study and estimating errors, the software automatically refines the mesh where needed to improve results.

The P-Method

The concept of the p-method is to use more efficient elements in regions with high errors. After running analysis and estimating errors, the program increases the order of elements in regions with errors higher than a user-specified level and reruns the study. The p-method does not change the mesh. It changes the order of the polynomials used to approximate the displacement field. Using a unified polynomial order for all elements is not efficient. The software increases the order of the polynomial only where it is needed. This approach is called the selective adaptive p-method.

This option is supported for solid elements only. When this option is checked, the program may run the problem several times. After each loop, the program assesses the global and local errors and decides whether to make another run.

The program stops the loops when one of the following conditions is met:

  • The global criterion converges,
  • All local errors converge (i.e., for each element), or
  • The maximum number of loops is reached.

You can base the convergence check on total strain energy, Root Mean Square (RMS) of von Mises stresses, or RMS of resultant displacements.

In this release, the p-method works with solid elements only, shells are not supported.

After running a static problem using the p-adaptive method, you can generate convergence plots. For more information, refer to the Viewing Results chapter.

For an example on adaptive methods, refer to the Online Tutorial.