|
[Sponsors] |
July 9, 2021, 12:23 |
No convergence + AMG solve rejected
|
#1 |
New Member
Thanos Alexopoulos
Join Date: Jul 2021
Posts: 1
Rep Power: 0 |
Hello,
I am a new star-CCM+ user, and I am trying to perform a 2D supersonic (M=1.9) cavity flow, over a rectangular cavity (Length=100mm, Depth=20mm). Selected coupled flow, unsteady, turbulent and K-OMEGA models. My residuals keep being constant as I run the simulation, especially Continuity and Energy. Other residuals perform spikes, which I understand means instability (maybe Cr too high?). I am also getting errors such as ''Maximum Temperature limited to xx on yy faces on Body 1'' and ''AMG solve rejected''. Is the mesh quality the problem? What should I look for first to resolve the problem? I am doing my own studying too, but haven't got anywhere yet. |
|
July 10, 2021, 23:16 |
|
#2 |
Senior Member
|
It is not a good idea to start with coupled flow simulations directly unless segregated flow is not converging. Mesh resolution study can be done with segregated flow with less computational resources. Also first start with steady state. Note out what transient behaviour are missing and desired. Optimise your wall function criteria. See what results can be extracted from steady state. Then move on to next step.
Need more info like geometry pics, bc's, mesh parameters etc to comment anything related. |
|
July 20, 2021, 06:55 |
|
#3 |
New Member
Nikolaos Papafilippou
Join Date: Apr 2019
Posts: 25
Rep Power: 7 |
Your problem could also stem from your timestep and/or CFL number jumps.
Also, sometimes steady solutions cannot be achieved for certain domains. In addition, it is accepted that coupled flow works better with supersonic cases compared to segragated flow. However, I agree with ashokac7, if timestepping and changes in the mesh to get the desired CFL do not work, try to simplify your case in order to have a benchmark that you can build upon. |
|
August 14, 2021, 06:37 |
|
#4 |
Senior Member
Lefteris
Join Date: Oct 2011
Location: UK
Posts: 337
Rep Power: 15 |
If it will help you at all, you can enable the "temporary storage retained" or change the verbosity in your solvers (I don't remember which option, sorry ) to see which one is diverging.
__________________
Lefteris |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Fail to converge when solving with a fabricated solution | zizhou | FLUENT | 0 | March 22, 2021 06:33 |
Solver parameters to solve convergence issue | WilliamH | SU2 | 5 | February 5, 2021 06:19 |
Rising residual behaviour and convergence | tjim | SU2 | 0 | October 12, 2018 04:04 |
About the AMG Linear Sover Convergence Tolerance | LeeTao | STAR-CCM+ | 0 | March 21, 2016 23:01 |
AMG versus ICCG | msrinath80 | OpenFOAM Running, Solving & CFD | 2 | November 7, 2006 15:15 |