Tips & Tricks for FEA Modeling of Rubber and Elastomers - Part 2
Last week, in Part 1 of this post, I discussed three tips to consider when solving a detailed stress analysis of an O-ring or seal, including: material testing, material law selection, and testing of the selected material law . This week, I'll discuss four more tips for enhancing the accuracy and convergence of your simulation!
4) Element Formulation: Selecting the best element formulation is another key modeling decision, especially when analyzing parts in compression where element locking is of concern. Mike Bak’s blog “Dealing with Incompressibility” provides excellent guidance to follow for these types of analyses.
5) Mesh: With hyperelastic materials, the deformed element shapes are more important than the initial element shapes. Plan for the deformation by skewing the undeformed mesh to distort into more of a regular pattern. Lower order elements are more stable, and often an all-triangular lower order mesh is recommended to overcome excessive element distortion. A finer mesh might not always be better, since small elements in the areas of peak strain often become more distorted than with a coarser mesh and thus are more susceptible to element failure. Consider rezoning if excess element distortion cannot be avoided in the undeformed mesh. Figure 3 above illustrates an axisymmetric seal simulation where automated progressive rezoning is used to capture the significant seal distortion and provide the accurate refined mesh in the final stages of the simulation.
6) Load control: Use a displacement controlled solution if possible, and force a large number of sub-steps since cumulative unbalanced residuals are a common cause of non-convergence. By using small load steps, one can minimize these errors and accurately track the large deflections and material nonlinearities.
7) Adjust the volume compatibility constraint: This is typically an analysis control parameter that allows for some error in volumetric compatibility. In ANSYS, the default for the Vtolparameter is 1e-5. Adjusting this parameter on the ANSYS SOLC command to a value as low as 1e-2 can sometimes be the key in overcoming convergence issues when large compressive strains are encountered.
What challenges have you overcome when modeling rubber materials? I would love to hear other tips and tricks for solving these complex analyses.
https://caeai.com/blog/tips-tricks-fea-modeling-rubber-and-elastomers-part-2