Iso-parametric finite elements have been around for a while now, but are rarely used in CFD. The difficulty of creating high quality curved elements in the boundary-layers is one big reason for this. This is changing, as we’ll see subsequently, but there are other areas where these elements pose challenges. See this article for a discussion on why visualizing results is harder to do.

This github page is a good way to wrap up, for now, our discussion on grid adaptation which featured Pointwise, SU2 and Tecplot. For background on SU2’s approaches see this 2016 presentation. The impact of 10 adaptive iterations, 20 minutes on a laptop is eye-catching. If you have any interesting results that use our solver-independent approach, do let us know!

The remarkable performance of a Bombardier engineer’s grid-generation using Pointwise led to a powerful new feature: automatic synchronization of spacing-constraints. The first release of this feature, detailed in this video, is now even stronger as described here. Whether you’re using structured or unstructured grids, learning the nuances of this feature is well worth the effort.