Problem with running SU2 with Euler solver on Windows

averis007 · October 30, 2024, 00:00

Hi All,
I'm facing a problem in running SU2 with Euler solver on Windows. Every thing is fine if I I'm using the RANS solver. But when i use the Euler solver, i keep getting the following error whenever SU2 is trying to write out the output files (Surface, Volume, Restart):

Error in "void COutput::SetVolumeOutputValue(const string&, long unsigned int, su2double)"
Cannot find output field with name SKIN_FRICTION-X

Don't understand why SU2 is trying to write out Skin friction when Euler solver is selected.
I have tried specifying the things I want in VOLUME_OUTPUT. i tried all different format options for Surface, Volume and Restart files and even tried skipping Restart files. But nothing seems to resolve the issue.

Versions tried: SU2-v8.0.0-win64 and SU2-v8.1.0-win64-mpi.
TIA.

bigfootedrockmidget · October 30, 2024, 02:51

Do you have a test case that reproduces this issue?

averis007 · October 30, 2024, 08:04

Quote:

Originally Posted by bigfootedrockmidget

Do you have a test case that reproduces this issue?

Please see the attached files

Also, there is another error on Windows, even with RANS solver. just comment out line 231 and uncomment line 232 after changing solver to RANS, so that the volume and surface output files are Tecplot Binary. There is an error related to tecZoneNodeMapWrite64 which doesn't happen for paraview format.

R.K · October 30, 2024, 10:10

I encountered this before as well. You need to set the wall boundary condition as MARKER_EULER. MARKER_ISOTHERMAL, _HEATFLUX and _HEATTRANSFER boundary conditions are for viscous simulations, so they work fine for RANS or NS solvers.

bigfootedrockmidget · October 30, 2024, 10:16

Thanks, R.K. , I remembered that there was an issue about this before but not what it was about. I think the long-term solution is to give an error message when the user wants to use viscous boundaries for a non-viscous solver.

averis007 · November 4, 2024, 03:26

Quote:

Originally Posted by R.K

I encountered this before as well. You need to set the wall boundary condition as MARKER_EULER. MARKER_ISOTHERMAL, _HEATFLUX and _HEATTRANSFER boundary conditions are for viscous simulations, so they work fine for RANS or NS solvers.

Thanks Rutvik. It worked.

averis007 · November 4, 2024, 03:27

Quote:

Originally Posted by bigfootedrockmidget

Thanks, R.K. , I remembered that there was an issue about this before but not what it was about. I think the long-term solution is to give an error message when the user wants to use viscous boundaries for a non-viscous solver.

Please look into the Tecplot format issues if possible.

R.K · November 4, 2024, 15:43

can you upload the files so i can recreate the error? Have you installed SU2 directly on windows or through a linux virtual box on windows?

it may be a compiler issue: https://bugzilla.redhat.com/show_bug.cgi?id=1879254

averis007 · November 5, 2024, 00:47

Quote:

Originally Posted by R.K

can you upload the files so i can recreate the error? Have you installed SU2 directly on windows or through a linux virtual box on windows?

it may be a compiler issue: https://bugzilla.redhat.com/show_bug.cgi?id=1879254

See this post below for the files and how to recreate.

Quote:

Originally Posted by averis007

Please see the attached files

Also, there is another error on Windows, even with RANS solver. just comment out line 231 and uncomment line 232 after changing solver to RANS, so that the volume and surface output files are Tecplot Binary. There is an error related to tecZoneNodeMapWrite64 which doesn't happen for paraview format.

I'm using the prebuiild Windows binaries (8.0.0 and 8.1.0-mpi versions). I don't necesaarily think there is any issue with the compilers, as the problem only arises if Tecplot format is chosen for output, paraview works fine. Perhaps a bug in the tecplot writer function.

R.K · November 5, 2024, 06:23

i think you deleted the files. i cant see them.

averis007 · November 5, 2024, 09:57

Uploaded again here

R.K · November 5, 2024, 11:12

I ran your case on version 8.0.1 and 8.1.0 (with all of the tecplot files as the output files) it worked fine, giving the tecplot output files. Can you post the verbatim of the error?

When you run any tutorial/testcase with an output for Tecplot, do you get the same error?

Did you follow the guide when installing? https://su2code.github.io/docs_v7/Build-SU2-Windows/

can you try running SU2 through WSL?

October 30, 2024, 00:00	Problem with running SU2 with Euler solver on Windows	#1
averis007 New Member Avijeet Join Date: Nov 2012 Posts: 24 Rep Power: 14	Hi All, I'm facing a problem in running SU2 with Euler solver on Windows. Every thing is fine if I I'm using the RANS solver. But when i use the Euler solver, i keep getting the following error whenever SU2 is trying to write out the output files (Surface, Volume, Restart): Error in "void COutput::SetVolumeOutputValue(const string&, long unsigned int, su2double)" Cannot find output field with name SKIN_FRICTION-X Don't understand why SU2 is trying to write out Skin friction when Euler solver is selected. I have tried specifying the things I want in VOLUME_OUTPUT. i tried all different format options for Surface, Volume and Restart files and even tried skipping Restart files. But nothing seems to resolve the issue. Versions tried: SU2-v8.0.0-win64 and SU2-v8.1.0-win64-mpi. TIA.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
problem regarding solver running two times	Vishal Garg	OpenFOAM Running, Solving & CFD	1	October 19, 2023 04:22
Fail to converge when solving with a fabricated solution	zizhou	FLUENT	0	March 22, 2021 07:33
Star cd es-ice solver error	ernarasimman	STAR-CD	2	September 12, 2014 01:01
SU2 running problem	andrea.ciarella	SU2	5	May 16, 2013 16:02
Divergence problem	Smaras	FLUENT	13	February 21, 2013 06:03

October 30, 2024, 02:51		#2
bigfootedrockmidget Senior Member bigfoot Join Date: Dec 2011 Location: Netherlands Posts: 676 Rep Power: 21	Do you have a test case that reproduces this issue?

October 30, 2024, 10:10		#4
R.K New Member Rutvik Khedkar Join Date: Mar 2024 Posts: 14 Rep Power: 2	I encountered this before as well. You need to set the wall boundary condition as MARKER_EULER. MARKER_ISOTHERMAL, _HEATFLUX and _HEATTRANSFER boundary conditions are for viscous simulations, so they work fine for RANS or NS solvers.

October 30, 2024, 10:16		#5
bigfootedrockmidget Senior Member bigfoot Join Date: Dec 2011 Location: Netherlands Posts: 676 Rep Power: 21	Thanks, R.K. , I remembered that there was an issue about this before but not what it was about. I think the long-term solution is to give an error message when the user wants to use viscous boundaries for a non-viscous solver.

November 4, 2024, 15:43		#8
R.K New Member Rutvik Khedkar Join Date: Mar 2024 Posts: 14 Rep Power: 2	can you upload the files so i can recreate the error? Have you installed SU2 directly on windows or through a linux virtual box on windows? it may be a compiler issue: https://bugzilla.redhat.com/show_bug.cgi?id=1879254

November 5, 2024, 06:23		#10
R.K New Member Rutvik Khedkar Join Date: Mar 2024 Posts: 14 Rep Power: 2	i think you deleted the files. i cant see them.

November 5, 2024, 09:57		#11
averis007 New Member Avijeet Join Date: Nov 2012 Posts: 24 Rep Power: 14	Uploaded again here

November 5, 2024, 11:12		#12
R.K New Member Rutvik Khedkar Join Date: Mar 2024 Posts: 14 Rep Power: 2	I ran your case on version 8.0.1 and 8.1.0 (with all of the tecplot files as the output files) it worked fine, giving the tecplot output files. Can you post the verbatim of the error? When you run any tutorial/testcase with an output for Tecplot, do you get the same error? Did you follow the guide when installing? https://su2code.github.io/docs_v7/Build-SU2-Windows/ can you try running SU2 through WSL?