search

smdogroup / caps2fun

Interface between CAPS and FUNtoFEM to automatically generate coupled structure and fluid meshes.
2 stars 2 forks source link

High stresses in the forward analysis. #3

Open bburke38 opened 2 years ago

bburke38 commented 2 years ago

Issue Type

Description

Producing high stresses in the trailing edge of the rib (due to the sharp corner present). @sean-engelstad tried chopping off the ends of the ribs in the .csm file, but then the structural mesh generation fails; tried unioning a wire object to close the edges on the surface of the OML.

Example

nacaWing

sean-engelstad commented 2 years ago

Was running a structural optimization with the diamondWing case where the stresses were pretty tame and changing with thickness, except the

membraneIssue

sean-engelstad commented 2 years ago

Ran the forward analysis for Sam's biconvex wing case and get 2/3 similar results and then one really bad result. Had stress about 28 times larger than the other one. samsWing_forward samsWing_forward2 samsWing_forward3

Also the temperatures went up a lot for the case with really high stress, 7x higher. The cl and cd values were really consistent and also very close to the actual result from test data, cl=0.192 and cd=0.029 from my data. The issue is the structure solution because we terrible structure mesh (no points in the internal structure except for rib and spar crossings). Need to send this to ESP/CAPS. nacaWing_points

bburke38 commented 2 years ago

What were the differences between the different cases for the biconvex wing? Did they have different flow conditions?

sean-engelstad commented 2 years ago

What were the differences between the different cases for the biconvex wing? Did they have different flow conditions?

The exact same inputs were used for each run case one after the other. So this should not happen. The cl and cd values were the same basically so the fun3d solution was basically the same. But the structural solution is not robust because we have terrible stringy elements in the internal structure. I'm contacing ESP/CAPS shortly to discuss improving the meshing capability.