Mesh is ignoring CAD altogether


#1

I am trying to mesh an aircraft and can’t figure out why the mesher is seemingly unaware of the geometry.

I am using hex dominant parametric meshing and the geometry is within the bounding box (except for a tiny sliver as recommended by the aerospace workshop). This seems like it should be a beginner question and I feel like there’s just a button I haven’t pressed, but I set this simulation up just like my other which ran fine.

Any help is much appreciated! My project is located here: https://www.simscale.com/workbench/?pid=7239633305600564175#tab_0-0

Thank you


#2

Hey @connorsousa1 and thanks for your question!

Did you make sure that your so called Material Point was inside the Bounding Box but not inside the aircraft? We have to make sure that we perform an external flow analysis! @dheiny explains in this post what the material point is for.

Let me know if that was the solution to your problem.

Good luck and happy SimScaling!

Jousef


#3

Yes @jousefm the material point matches those criteria =)


#4

The domain is a bit of an overshoot I would say. The geometry is watertight so that is no problem.

I would recommend having a look at this project and start from scratch. If you need any help let me know - will keep an eye on your progress in any case :+1:

Best,

Jousef


#5

@jousefm is correct, the domain is huge. The cells were so big they ‘missed’ your geometry. I reduced the bounding box to 1/4 the size and it picked the geometry up, but still looked severely under refined. So your options are to better refine the region surrounding the aircraft (either locally or globally) or reduce the domain size (or both).

Kind regards,
Darren


#6

@1318980 That solved the problem, thank you!

I will definitely be refining this mesh much better than it is now! I tried meshing quickly just to see where I was at as far as cell size.

As a follow up question, I’ve seen a rule-of-thumb domain sizing of 10L and 20L upstream and downstream respectively. The original domain was roughly that size. Did I misread/misunderstand that rule of thumb?

Thank you,
Connor


#7

@connorsousa1, I normally recommend 5L and 10L but that isn’t form a source it is just the numbers I remember from going on a course :smiley: The real answer is that you get used to approximating domain size based upon type of flow, for example, slow incompressible flow need less around it than a supersonic compressible flow (because of the huge pressure gradients).

There was nothing wrong technically with your original mesh just in our opinion was a lot bigger than it needed to be (and results may show differently) however if you make it that large you need to ensure that the area of interest is refined, I usually stick a large cylinder around the aircraft along the wing axis a few sizes around and refining it to get the accuracy. If your mesh had picked up the geometry it still wouldn’t have been accurate just because of the cell size in the near field.

best,
Darren