Jump to content

Airfoil Analysis - issues with mesh


---
 Share

Recommended Posts

Fairly often, I will encounter issues with airfoil analysis where the software does not recognize my actual mesh section as an outer airfoil profile.  In many cases, this is because some of the scanned and polygonized mesh may contain stray polygons in an area where there is a trailing edge slot (see attached from the Direct Help for "Camber Line" - red arrows represent surface normals of stray polygons). 

Is it possible to include parameters in the camber line creation dialog that will filter out these stray polygon normals to allow the airfoil analysis to be more robust with respect to difficult scan areas?

image.png

Link to comment
Share on other sites

Hi Ray,

Thanks for the feedback. 

A bigger picture of the process would help, to understand why those stray polygons appear on first place.
Regarding the camber line: As you mention, the camber line creation command checks if the section is suitable for the underlying algorithm. I recorded your suggestion in a software ticket, but I am afraid that without test data no further action is possible.
Please contact support@gom.com for this (Confidentiality is guaranteed)

Regards,

Sergio

 

Link to comment
Share on other sites

Hi again,

After a short internal discussion, couple of ideas that could help:

- To prevent the noise on the measurements: please have a look at the "Scan Point Computation" entry on the direct help
- To ignore this noise when creating the section: try to adapt the selection on the mesh to avoid selecting the stray polygons -> getting in the end a clean section. One possibility to achieve this parametrically (without manual interaction on each case) is to create a patch compound on the CAD and applying the measuring principle "Fitting element" (which allows some control over the selection). The sections can then be created on those elements.

Hope this helps.
Anyway we are of course interested in improving the robustness of the camber line computation and as mentioned test data is very welcomed.

Cheers,
Sergio

Link to comment
Share on other sites

 Share

×
×
  • Create New...