Shape Optimization mesh file name error

Gramanzini · July 8, 2013, 15:45

I am trying to run tutorial 8, using parallel processing, I have successfully ran it with serial processing. I am getting a mesh file name error. The program is trying to find "mesh_opt_MDC_rot_NACA0012.SU2". Due to parallel processing (3 processors) the file names are numbered as mesh_opt_MDC_rot_NACA0012_1, mesh_opt_MDC_rot_NACA0012_2, mesh_opt_MDC_rot_NACA0012_3. Please refer to the attached error file.

Thank you

rktchip · July 25, 2013, 13:52

this seems like an old version of the code, can you please try the latest version?

economon · August 10, 2013, 14:34

Hi Joe,

Did using the latest version of the code fix the issue?

Cheers,
Tom

Gramanzini · August 10, 2013, 22:03

No I haven't i have been away for a few weeks. I have been using the SU2 2.0 for windows. Should I try to use the developers code instead. I am using the software for its adjoint and shape optimization capabilities. Thank you.

RcktMan77 · October 2, 2013, 21:02

I may be having the same issue under OS X. It appears a symbolic link is created for the FFD mesh file in the /DESIGNS/DSN_XXX directory that points to the original located up two parent directories. After the initial direct flow and adjoint solutions are completed, the gradient evaluation fails with:

Read the FFD information from mesh file.
0 Free Form Deformation (FFD) FFDBoxs.
24575 Free Form Deformation (FFD) nested levels.
There is no FFD box definition. Just in case, review the .su2 file
The input grid doesn't have the entire FFD information!
Press any key to exit...

My guess is that the code is unable to follow the symbolic link to the original and doesn't recognize the FFD definition.

economon · October 3, 2013, 03:02

Hi,

Just a note that in the recent versions of the code, the FFD specification has been adjusted in the mesh file. It is now specified using:

FFD_NBOX= 1
FFD_NLEVEL= 1
FFD_TAG= 0
FFD_LEVEL= 0
FFD_DEGREE_I= 5
FFD_DEGREE_J= 4
FFD_DEGREE_K= 1
FFD_PARENTS= 0
FFD_CHILDREN= 0
FFD_CORNER_POINTS= 8
-0.0403 0 -0.04836
0.8463 0 -0.04836
1.209 1.2896 -0.04836
0.6851 1.2896 -0.04836
-0.0403 0 0.04836
0.8463 0 0.04836
1.209 1.2896 0.04836
0.6851 1.2896 0.04836
FFD_CONTROL_POINTS= 0
FFD_SURFACE_POINTS= 0

You might try with version 2.0.8 of the code that is available on GitHub, along with the updated specification above (this example was taken from the Onera M6 mesh).

Hope this helps,
Tom

RcktMan77 · October 3, 2013, 14:45

Thanks for the update, Tom. Updating the FFD box description in the mesh file resolved the issue I was having.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[swak4Foam] GroovyBC the dynamic cousin of funkySetFields that lives on the suburb of the mesh	gschaider	OpenFOAM Community Contributions	300	October 29, 2014 19:00
pisoFoam compiling error with OF 1.7.1 on MAC OSX	Greg Givogue	OpenFOAM Programming & Development	3	March 4, 2011 18:18
checking the system setup and Qt version	vivek070176	OpenFOAM Installation	22	June 1, 2010 13:34
ParaView Compilation	jakaranda	OpenFOAM Installation	3	October 27, 2008 12:46
DecomposePar links against liblamso0 with OpenMPI	jens_klostermann	OpenFOAM Bugs	11	June 28, 2007 18:51

July 25, 2013, 13:52		#2
rktchip Member Trent Lukaczyk Join Date: Feb 2011 Location: Stanford, CA Posts: 75 Rep Power: 15	this seems like an old version of the code, can you please try the latest version?

August 10, 2013, 14:34		#3
economon Super Moderator Thomas D. Economon Join Date: Jan 2013 Location: Stanford, CA Posts: 271 Rep Power: 14	Hi Joe, Did using the latest version of the code fix the issue? Cheers, Tom

August 10, 2013, 22:03		#4
Gramanzini New Member Joe Gramanzini Join Date: Jul 2012 Posts: 3 Rep Power: 14	No I haven't i have been away for a few weeks. I have been using the SU2 2.0 for windows. Should I try to use the developers code instead. I am using the software for its adjoint and shape optimization capabilities. Thank you.

October 2, 2013, 21:02		#5
RcktMan77 Senior Member Zach Davis Join Date: Jan 2010 Location: Los Angeles, CA Posts: 101 Rep Power: 16	I may be having the same issue under OS X. It appears a symbolic link is created for the FFD mesh file in the /DESIGNS/DSN_XXX directory that points to the original located up two parent directories. After the initial direct flow and adjoint solutions are completed, the gradient evaluation fails with: Read the FFD information from mesh file. 0 Free Form Deformation (FFD) FFDBoxs. 24575 Free Form Deformation (FFD) nested levels. There is no FFD box definition. Just in case, review the .su2 file The input grid doesn't have the entire FFD information! Press any key to exit... My guess is that the code is unable to follow the symbolic link to the original and doesn't recognize the FFD definition.

October 3, 2013, 03:02		#6
economon Super Moderator Thomas D. Economon Join Date: Jan 2013 Location: Stanford, CA Posts: 271 Rep Power: 14	Hi, Just a note that in the recent versions of the code, the FFD specification has been adjusted in the mesh file. It is now specified using: FFD_NBOX= 1 FFD_NLEVEL= 1 FFD_TAG= 0 FFD_LEVEL= 0 FFD_DEGREE_I= 5 FFD_DEGREE_J= 4 FFD_DEGREE_K= 1 FFD_PARENTS= 0 FFD_CHILDREN= 0 FFD_CORNER_POINTS= 8 -0.0403 0 -0.04836 0.8463 0 -0.04836 1.209 1.2896 -0.04836 0.6851 1.2896 -0.04836 -0.0403 0 0.04836 0.8463 0 0.04836 1.209 1.2896 0.04836 0.6851 1.2896 0.04836 FFD_CONTROL_POINTS= 0 FFD_SURFACE_POINTS= 0 You might try with version 2.0.8 of the code that is available on GitHub, along with the updated specification above (this example was taken from the Onera M6 mesh). Hope this helps, Tom

October 3, 2013, 14:45		#7
RcktMan77 Senior Member Zach Davis Join Date: Jan 2010 Location: Los Angeles, CA Posts: 101 Rep Power: 16	Thanks for the update, Tom. Updating the FFD box description in the mesh file resolved the issue I was having.