non-watertight meshes from wing and fuselage intersections?

[James Haley]

I'm working on generating CFD meshes for practice with using in CBAERO. 

While the end goal is to finish up an X-24 model, I have switched to a simple X-1 Model that experiences the same mesh issues.

It seems that the curvature of the fuselage where the wing and vertical tail intersect are too complicated for the mesh to generate reliably, I have tried hundreds of mesh parameters to no avail.

Last night I switched the critical fuselage sections to rounded rectangles, (near perpendicular intersections) and it meshed fine (water tight), however making the X-1 or X-24 more boxy and rectangular can't be the best solution. 

Also I was unable to mesh at all ( vsp crashes ) with NACA 6-series foils, I had to change to 4 - series and it began to work fine. 

I could only find one thread that similar issues were mentioned for but it was for vsp 2.0 , I am currently using 3.0.4.

Any help would be appreciated, 

- James Haley

  

[Rob McDonald]

James,

The best documentation we have for the meshing is the presentation I gave at the last OpenVSP Workshop.

http://www.openvsp.org/wiki/lib/exe/fetch.php?media=workshopv3:vsp_file_types_meshing_mcdonald.pdf

In that file, I present a general meshing strategy that works for me (Slide 26).  I recommend you start with the 'most global' settings and then work towards the fine tuning.

Usually, when CFDMesh has problems it isn't really the meshing that is the problem, but instead the intersection of the underlying surfaces.  A wing intersecting a curved fuselage shouldn't pose a problem.  Instead, problems are caused by surfaces meeting with nearly exact tangency, or corners meeting in awkward ways that force bad triangles.

Your meshes go from extremely fine resolution to extremely coarse resolution -- I doubt you are going to want that coarse of resolution in your eventual model.  I generally start by using just the 'max edge length' parameter to get an overall mesh that is in the ballpark.

The mesh visualization allows you to view just the bad triangles.  That can help you figure out what the problem is.

If you can't figure anything out, and you can share your models, you can send them my way and I'll see if I have any better luck.

Rob

--
You received this message because you are subscribed to the Google Groups "OpenVSP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openvsp+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[andrew...@nasa.gov]

Hi James;

Your problem is most likely not mesh parameters, but the model itself. There is a natural tendency for humans to produce models that the computer can't distinguish one point from another. If you read http://www.openvsp.org/wiki/lib/exe/fetch.php?media=workshopv3:open_vehicle_sketch_pad_aircraft_modeling_strategies-hahn.pptx.pdf and watch the accompanying video http://mediasite01.ceng.calpoly.edu/Mediasite/Catalog/pages/catalog.aspx?catalogId=2770f727-7f81-4203-a847-ae1d98395d53, then you will be able to make small changes in your model that should make meshing far more robust.

Good luck,

Andy