search

Exawind / hfm-fsi

Input files and data for HFM FSI Work
0 stars 0 forks source link

Pazy wing FSI validation problem #90

Open pscrozi opened 2 months ago

pscrozi commented 2 months ago

Kevin: still working on the Pazy wing. Ganesh gave a task to do verification tests on these. Had originally done the tests with the main OpenFAST, now going to redo on OpenFAST dev (just recompiled it, and going to be using that).

Nate: put together mesh of the Pazy wing (airfoil extrusion) to drop into the FSI framework. Making sure the low-fidelity model works on the beam side, and doing the aerodynamics.

pscrozi commented 2 months ago

Kevin: now checked vs beamdyne and matched tip deflection and torsion. Ran fully assembled matrices through modal analyses. Looking at ways to drive down a little more of the error (currently within a few percent. Taking model that Nate put together and using Eagle build and working through input file issues. Nate has been working on the input files for a different task.

Kevin is still working on input files for the full CFD version of this problem. Nate may be able to help, and may be able to run on Flight.

pscrozi commented 1 month ago

Kevin will take a break from this, but hopefully finish by the end of Q3. He will keep communication open, even though he will be on travel for much of the next month.

Neil could pick this up. Kevin and Neil will meet to hand off much of this to Neil.

pscrozi commented 4 weeks ago

Neil will now pick this up, reach out to Kevin and Nate.

pscrozi commented 3 weeks ago

Neil: having some build issues on Sandia machines, but Phil and Nate helped. I reproduced the prior results. Now running static case that Nate created, then will try the full FSI run.

Kevin: what are the current options for running ExaWind modules that are pre-built on any machines? Phil: no. We have the infrastructure to generate those modules, but no one has done it recently.

pscrozi commented 2 weeks ago

Neil: Regarding the Pazy wing, the steady simulation appears to be fine. I'm attempting to set up the FSI calculation.

Kevin: vacation and understaffing pulling him elsewhere.

Ganesh: Kevin may have a repo set up for this that we can look at. Can we get this from him? Eagle is closed. Could put it on Kestrel. Need mode shapes and beamdyne model.

Nate: I could also get the mesh and try to run this if wanted. Neil was looking for a place to run last week. Trouble finding a place to run these.

neilmatula commented 1 week ago

Been out sick for several days. Dealing with segfaults with hybrid solver on attaway, while amr-wind and nalu-wind (with an appropriate replacement for the overset BC) run fine separately. Same is true for nrel5mw case (which I'm using as a reference), so it's not an issue with the case setup. In case attaway is the problem, building on several other machines. If none work, will move to Kestrel.

Generating a new, smaller mesh to help with debugging.

pscrozi commented 1 week ago

Report looks good. Need to look at the logs and segfault messages. Please add these @neilmatula .

Might be a build issue on a platform we haven't been building a lot on. Might be best to provide Neil with a build for debugging. Attaway could be a tough platform to start out on. Kestrel would be better. Shouldn't have to build on kestrel --- someone else will have a module you can use. We do need to get builds going again on attaway, probably by Nate with Neil.