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

Why did the residuals increase after flattened out as I continued the iteration?

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

Like Tree1Likes
  • 1 Post By LuckyTran

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   May 13, 2018, 05:41
Default Why did the residuals increase after flattened out as I continued the iteration?
  #1
Member
 
Join Date: Aug 2011
Posts: 33
Rep Power: 15
pchoopanya is on a distinguished road
Dear all,

Please take a look at my residuals plot of a 3D flow past a row of compressor blade cascade, using Spalart-Allmaras turbulence model.

As you can see in the plot the residuals dived satisfactorily, then flattened out, then they rose up again, and again...

Luckily, I also activated point monitors of pressure and velocity, and lift and drag coefficients, and they all had stabilised, unlike the residuals.

There was no modification to the settings at all during the iteration process. Why did the residuals rise again after being seemingly flattened out/converged?

The residuals plot does not look great in terms of their value, but other point monitors and force coefficients had converged, can I conclude that my simulation has already converged? Or should I continue the iteration with new under-relaxation factors to get as low as possible residual values?

Judging the point monitors, I am happy with my solution but I am just being curious at what made the residuals rise as I kept iterating without making any change to the settings?
Attached Images
File Type: jpg 20180513_145937.jpg (57.8 KB, 63 views)
pchoopanya is offline   Reply With Quote

Old   May 20, 2018, 04:31
Default
  #2
Senior Member
 
Lucky
Join Date: Apr 2011
Location: Orlando, FL USA
Posts: 5,760
Rep Power: 66
LuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura about
You'll never know why the residuals jumped around unless you follow what the solution was doing at that time; and this depends also on what the initial guess was.

Flattened residuals is a hint that the solution is no longer changing, which it sounds like you've verified using monitors. If your solution has converged then it has converged.

URF's are for stabilizing the solution. If it ain't diverging leave it alone. If your monitors converged then your solution is converged. Remember, the values of residuals measure imbalances & does not tell you anything about convergence. If you want different values of converged residuals, then start tinkering with your settings & mesh to try to improve the quality of the solution.
ari003 likes this.
LuckyTran is offline   Reply With Quote

Old   April 24, 2024, 22:34
Default
  #3
New Member
 
Ertugrul
Join Date: Apr 2024
Posts: 4
Rep Power: 2
Aeroma is on a distinguished road
What if the residuals had a spike and then started a diverging trend? I am running a supersonic ramjet engine simulation. Residuals were stable for almost 2500 iterations and at that point the temperature plot is still increasing and the mass flow rate lines had still not met. Then it has a drastic increase and a drop followed by a diverging trend while the temperature still increased while thrust were flattened out and mass flow rates at different sections (inlet and outlet) were still coming closer and had not met each other. I thought the mass flow rates at the inlet and before nozzle must be good indicators for convergence since it is a ramjet engine.
Attached Images
File Type: jpg IMG_0830.jpg (180.9 KB, 8 views)
File Type: jpg IMG_0831.jpg (191.9 KB, 5 views)
Aeroma 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
[snappyHexMesh] SnappyHexMesh running killed! Mark JIN OpenFOAM Meshing & Mesh Conversion 7 June 14, 2022 02:37
how to increase "Newton Pressure Iteration Limit" kus CFX 9 April 21, 2013 02:54
Accuracy and convergence pawan1989 FLUENT 5 April 15, 2010 18:17
Why do residuals increase jrg FLUENT 3 October 4, 2007 18:43
Convergence - scaled vs unscaled residuals HS FLUENT 1 November 7, 2005 06:45


All times are GMT -4. The time now is 11:45.