|
[Sponsors] |
July 3, 2023, 11:56 |
Convergence and PIMPLE algorithm
|
#1 |
Senior Member
Join Date: Dec 2021
Posts: 251
Rep Power: 6 |
Hey everyone,
For the people familiar with Fluent, you can set a number of Max Iterations per timestep for transient simulations. I always assumed it allowed Fluent to behave similarly to the PIMPLE algorithm and the nOuterCorrectors. But in Fluent, you can check the convergence of the residuals absolutely or relatively, meaning compared to the very first timestep of the run, or compared the previous timestep (if my understanding is correct). The issue with setting the convergence check to Absolute is that you might end up with 0 iteration per timestep if your simulation reached a low absolute residual even though your flow is still evolving. The Relative setting should prevent that. My question is, when using residualControl in PIMPLE and nOuterCorrectors > 1, are the residuals checked absolutely (compared to the first timestep) or relatively (compared to the previous timestep)? I feel like it might make quite a difference for the tolerance we would set. For instance, Fluent default setting for a relative check is 0.05 versus 0.001 for an absolute check. Sorry if it is unclear, as I have not completely figured out the differences |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
How to use PIMPLE properly? | floquation | OpenFOAM Running, Solving & CFD | 27 | August 12, 2024 11:15 |
PIMPLE: No convergence criteria found (please help) | fawzi | OpenFOAM Running, Solving & CFD | 9 | February 8, 2024 18:51 |
PIMPLE – the value of the final under-relaxation factor | Zbynek | OpenFOAM | 9 | December 22, 2023 06:26 |
Help with PIMPLE algorithm in chtMultiregionFoam | Chris T | OpenFOAM Running, Solving & CFD | 0 | August 30, 2022 09:49 |
Additional PIMPLE convergence criteria | amanbearpig | OpenFOAM Programming & Development | 5 | August 14, 2019 16:35 |