CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > SU2

Problem with RAE optimization tutorial

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   November 23, 2021, 13:40
Unhappy Problem with RAE optimization tutorial
  #1
New Member
 
Andrea Michelotti
Join Date: Mar 2021
Posts: 12
Rep Power: 3
AndreaMichelotti is on a distinguished road
Hello everyone,

I am trying to run the RAE optimization tutorial without the solution_adj_cd.dat and solution_adj_cmz.dat files because I have to optimize different configurations but the process ends after the DIRECT solution because those file are not present.

I tried to run each step of the optimization process individually but I reached the same problem. After running the DIRECT problem, I used the restart file as the solution one for the CONTINUOUS_ADJOINT problem but I received the same message:

Code:
Error in "void CSolver::Read_SU2_Restart_Binary(CGeometry*, const CConfig*, std::__cxx11::string)":
-------------------------------------------------------------------------
Unable to open SU2 restart file solution_adj_cd.dat
I saw that this is the only optimization tutorial which requires these files as in the others those files are generate within the shape_optimization procedure.

I can't figure out why I can't bypass those files or how to generate new ones because even with RESTART_SOL=NO and/or running the process step by step I face the same message.

Does anyone know what I am missing?

I attach my config files and the tutorial one.

Thank you in advance
Attached Files
File Type: zip RAE_optimization_cfg.zip (6.2 KB, 2 views)
AndreaMichelotti is offline   Reply With Quote

Old   November 23, 2021, 13:47
Default
  #2
pcg
Senior Member
 
Pedro Gomes
Join Date: Dec 2017
Posts: 419
Rep Power: 10
pcg is on a distinguished road
Can you try reproducing those files for your new configuration? i.e. run SU2_CFD directly with MATH_PROBLEM=CONTINUOUS_ADJOINT for your initial design.
pcg is offline   Reply With Quote

Old   November 23, 2021, 14:18
Default
  #3
New Member
 
Andrea Michelotti
Join Date: Mar 2021
Posts: 12
Rep Power: 3
AndreaMichelotti is on a distinguished road
Hi!

The file RAE2822_8_1_1_600_350.cfg is the one I am using for both inviscid and viscous flow with the appropriate changes.

I use shape_optimization.py to run the process.

In the inviscid case the optimization runs fine, for the viscous it doesn't because it requires the soluiton_adjoint files.

I also tried, just in case, to change the profile with the NACA0012 as in the previous tutorial but the RANS simulation gives the same message.

I copy here the .su2 mesh so it is possible to run the case.

https://drive.google.com/file/d/1qDj...ew?usp=sharing
AndreaMichelotti is offline   Reply With Quote

Old   November 26, 2021, 05:37
Default
  #4
New Member
 
Andrea Michelotti
Join Date: Mar 2021
Posts: 12
Rep Power: 3
AndreaMichelotti is on a distinguished road
I have tried to run again the viscous cfg file for both my case study and the tutorial with MATH_PROBLEM=DIRECT and after that MATH_PROBLEM=CONTINUOUS_ADJOINT with and without RESTART_SOL option and in every case I received the message:

Code:
Error in "void CSolver::Read_SU2_Restart_Binary(CGeometry*, const CConfig*, std::__cxx11::string)":
-------------------------------------------------------------------------
Unable to open SU2 restart file RAE2822_8_1_1_600_350_restart_adj_cd.dat
I have also tried to run the MATH_PROBLEM=CONTINUOUS_ADJOINT with SOLVER=EULER after the DIRECT solution of the RANS case and this time it has generated the restart_adj_cd.dat file without needing it (this always happens when running an inviscid optimization process).

After that, I have run the CONTINUOUS_ADJOINT problem for the RANS case and it seems to work because the restart_adj_cd.dat file now exists. Also, it seems to converge to a solution. I don't think it is a feasable one.

I found that the RESTART_SOL seems not to work because the program continues to search for the RESTART_ADJ_FILENAME_cd.dat even thought RESTART_SOL=NO.

I don't know what to do because I don't get why this is happening.

Another viscous optimization case study is the Unsteady NACA0012 one which doesn't require the RESTART_ADJ_FILENAME_cd.dat files as initial solution for the problem. I have tried to find discrepancies but I did not find anything.

I even tried to run the TestCases/optimization_rans/steady_oneram6 and I receive the same message.

No one has ever experienced this problem before?
AndreaMichelotti is offline   Reply With Quote

Old   December 23, 2021, 20:01
Default
  #5
New Member
 
zhou
Join Date: Mar 2019
Posts: 4
Rep Power: 5
qooin is on a distinguished road
I encountered the same problem. I tried 'cont_adj_rans' of TestCase, but the problem also occured. Is resrart file necessary for continuous adjoint based on RANS?

Last edited by qooin; December 24, 2021 at 00:39.
qooin is offline   Reply With Quote

Reply

Tags
adjoint, rae, shape_optimization, tutorial

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
CFD Fluent tutorial - Shell and tube heat exchanger - PROBLEM tom96 FLUENT 1 May 21, 2018 01:24
multiRegionHeater tutorial visualization problem Bufacchi OpenFOAM Running, Solving & CFD 3 September 18, 2015 10:20
Problem on Fluent Tutorial: Horizontal Film Boilig Feng FLUENT 2 April 13, 2013 06:34
newbie problem with cavity tutorial miki OpenFOAM Running, Solving & CFD 8 September 2, 2012 16:22
Help! Compiled UDF problem 4 Wave tank tutorial Shane FLUENT 1 September 3, 2010 03:32


All times are GMT -4. The time now is 19:13.