Overset meshes are challenging. Be prepared to deal with overset mesh errors during your simulation. To minimize the chance of wasting a lot of time later in your simulation workflow, there a few tips to keep in mind and avoid fixing time consuming errors. Here we go!
- Check you mesh. Meshing is by far the most time consuming step in any CFD simulation. But if we’re talking Overset, we’re dealing with a different beast. It’s just not important but critical to ensure a high quality mesh at the overset interface to ensure an adequate interpolation of the results but especially at boundaries that cross or exit the background domain. Contact is the most typical case.
- A handy way to improve you mesh near the contact area is by using a volumetric control of the size and shape of your overlaping interface. In this way you can define the size of your cells in both regions and ensure that they will have about the same size for a proper interpolation.
- Make sure that you activate zero mesh gap technology and that you keep a minimum number of good quality cells in the contact area. You can control how many layers you want to deactivate with the zero gap layers under physics values.
- Before firing up your final simulation, check first the translation of your overset simulation during the entire trajectory. You can do this by freezing all the fluid solvers and running the transient simulation. It can take some time depending on the timestep but trust me, it will never take as much time as having to deal with an overset error later. It might be the case that your simulation runs well for a long part of the process and suddenly diverges because of an overset error.
- If you are confronted with these kind of errors is very likely that you’ll have to remesh. A new challenge appears as Star-CCM+ remesh the parts but translate them to their original position.
- There are two alternatives here: To remesh AND translate the region or to Translate the parts AND remesh the region. I prefer the second option since this is the option that affects the solution the less (because it remesh the parts in-place). In any case it is better not to have to remesh during solution.
- It is very useful to keep track of the overset region’s position. You can do this with a point derived part and with a position report using that point as input. In this way you’ll know the location of your overset region during the whole simulation. Should you remesh, then you’ll know where you’ll have to position your part or region again.
Happy oversetting your meshes!