Strategy: Identifying and Resolving Potential Trouble Spots in a Model
To help identify potential trouble spots or modeling errors before you invest a great deal of time in a model, we recommend that you obtain a first pass of the results within a relatively short period of time.
• Define and run an analysis using the quick check convergence option. With this option, the engine runs the analysis at a polynomial order of 3. You can then check stress and deformation results for structural analyses, or temperature gradient and flux results for thermal analyses. These results can reveal problems in the model, such as cracks or missing loads and constraints.
• Then, you might want to consider running an analysis with reduced tolerance or reduced maximum polynomial order values. For example, if the model is large and it would require several hours to achieve 10% convergence, consider first setting convergence to 20% and the maximum p-order to 6.
After the run is complete, review a fringe plot of the p-level reached by each edge. For edges that reach the maximum p-order, you may want to make the following changes:
◦ Spread any loads or constraints over a larger area.
◦ Seed your model with datum points near the high p-order zones to adjust the size of elements so the elements in the high p-order zones are smaller.
◦ Add more elements.