The ABC’s of Limiting User-to-User Mesh Variations
Posted by Rob Kaczmarek on August 01, 2013
We’ve all experienced it… The old “Our group in ‘pick your location’ is getting very different results from ours.” When it comes to CFD there have been numerous advances in the types of meshing technologies used: tetrahedral, hexahedral, polyhedral, chimera, etc., but little has been done until recently to limit the user-to-user variations from the mesh. We’ve assembled the top three ways to limit your user-to-user variations.
Automate, Automate, Automate
Meshing automation has arrived in various manifestations and it looks to be the best solution to the issue, if done right. This new innovative feature can take the human error out of any simulation, ensuring that case after case is consistent. While the “automated meshing” of a few years ago was clunky, mostly scripted, and not truly automated at all, the present day automated meshing is stable and validated. In some cases meshing automation has opened the door for grid refinement studies proving the technology is fully mature and one of the best ways to limit user-to-user meshing variations.
Assembling a best practice around your project in which you determine which meshing technique best fits the problem will help to reduce the effects of meshing variations. This may seem like a no-brainer, but you would be surprised at how often this occurs. Differences in opinion on meshing can cause real problems when trying to derive consistency in CFD results.
Specifying the desired deliverables upfront can help to eliminate these user-to-user variations by setting the focus of the simulations ahead of time. This will ensure that all cases run will be focused on the same desired deliverables. All too often one group is focusing a simulation on the wrong phenomenon and then comparing results.