An issue about the generation of the mesh, in Hex-Dominant parmetrics

Project Support
1

I have been working on the mesh of a BWB with 4º AOA. To obtain a good resolution of the y+, I first try to obtain a y+=1, I create an average domain (X axis=11Lc, Y axis =4, Z=3). Then I created some boxes to apply region refinement. In the wake box ( the box in which inside is the BWB), I put a level similar to the final thickness of the boundary layer. The BL has 34 layers and the expansion ratio is obtained from the following calculator:
Fluid Mechanics 101.
I tried dozens of variations to obtain a proper BL with y+=1, but I never obtained the mesh, I always get the error that the machine has run out of memory, but in other cases, I obtained the meshes with higher resolution than the one used now.
I attached below the link, the delta X is 0.07149m:
https://www.simscale.com/workbench/?pid=3789687585094801634&mi=spec%3A750d9862-a124-4617-aa92-815e62bac512%2Cservice%3AMESHING%2Cstrategy%3A15&ps=algorithm%2Frefinements%2F1&sh=3
Thank you in advance.

2

Hi, I’d suggest rethinking the refinement strategy that you are using, because the current mesh settings are causing a huge amount of cells to be generated.

E.g. your bounding box is around 8.5 x 3.15 x 2.6 meters in size, with 121, 44, 33 base cells in X, Y, Z, respectively. A refinement level 5 would give you cells with around 0.002 meters edge length.

Taking exclusively the wake refinement, you have a box with 1.6 x 1 x 0.5 meters in dimensions, and a refinement level 5 being requested. If you hand calculate how many cells would take to fill this box, it would be around 75 million cells.

In short, this is way too much for a community license.

Cheers

3

Thank you so much. Would you recommend me the minimum dimension of the box to capture properly the lift and drag forces?

4

In this particular case the issue is more on controlling cell count, since you won’t be able to run a simulation with a gigantic mesh due to memory issues, related to your community license.

I’d shoot for 5-8M cells max.

Suggestion: calculate the size of level 0, 1, 2, 3, 4, 5 cells and use that information to calculate how many cells you’d spend in a certain region refinement. You likely won’t be able to use any level 5 region refinements due to the cells being too small at that level. Maybe level 5 surface refinements might be possible - it would require testing.

Cheers