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

Direct simulation converging but not the Adjoints

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

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   September 23, 2016, 14:54
Default Direct simulation converging but not the Adjoints
  #1
New Member
 
ANAND AMRIT
Join Date: May 2016
Posts: 1
Rep Power: 0
amrit38 is on a distinguished road
Hello everyone,

I am running a CRM wing case and is using EULER equation. My DIRECT simulation converges pretty well.

However when it does ADJOINT simulation (continuous adjoints), it says there are some non physical points in the solution and the solution never converges, rather the residuals go very high. I have set MGLEVEL= 0, still its the same problem

I have attached the figures plotting the residuals for both direct and adjoint simulations. Also the configuration file is attached.

adjoint_simu.JPG

direct_simu.JPG

CRM.txt

Thanks
Anand
amrit38 is offline   Reply With Quote

Old   November 17, 2016, 17:59
Default
  #2
hlk
Senior Member
 
Heather Kline
Join Date: Jun 2013
Posts: 309
Rep Power: 10
hlk is on a distinguished road
Quote:
Originally Posted by amrit38 View Post
Hello everyone,

I am running a CRM wing case and is using EULER equation. My DIRECT simulation converges pretty well.

However when it does ADJOINT simulation (continuous adjoints), it says there are some non physical points in the solution and the solution never converges, rather the residuals go very high. I have set MGLEVEL= 0, still its the same problem

I have attached the figures plotting the residuals for both direct and adjoint simulations. Also the configuration file is attached.

Attachment 50693

Attachment 50694

Attachment 50695

Thanks
Anand
It is a common problem to have more difficulty in converging the adjoint, and here are some tips which usually help:
1. Greater convergence on the direct solution. The closer the residuals are to machine zero, the better.
2. Use the CFL_REDUCTION_ADJFLOW parameter to reduce the CFL value compared to what is used for the direct problem.
3. In order to turn of MG for the adjoint solution but still use it for the direct solution, use MG_ADJFLOW = NO
4. If the issue is that the mesh isn't great, try using RELAXATION_FACTOR_ADJFLOW (set to something slightly less than 1.0).
5. Try using a 1st-order method, at least to start up the solution (CONV_NUM_METHOD_ADJFLOW / SPATIAL_ORDER_ADJFLOW)
hlk is offline   Reply With Quote

Reply

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
TimeVaryingMappedFixedValue for Direct Numerical Simulation inlet johndeas OpenFOAM 5 May 21, 2014 07:11
Incompressible simulation brugiere_olivier SU2 2 April 15, 2014 10:12
Simulation and Optimisation of centrifugal fan 3D to 2D eRzBeNgEl STAR-CCM+ 0 January 31, 2013 13:21
Dynamic mesh simulation not converging mrestrepo30 FLUENT 0 March 8, 2010 14:15
3-D Contaminant Dispersal Simulation Apple L S Chan Main CFD Forum 1 December 23, 1998 10:06


All times are GMT -4. The time now is 12:00.