|
[Sponsors] |
Whats the procedure for choosing tolerances in fvSolution? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
June 15, 2013, 12:08 |
Whats the procedure for choosing tolerances in fvSolution?
|
#1 |
Senior Member
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,208
Rep Power: 26 |
in sonicFoam/ras/prism the fvSolution is so:
Code:
/*--------------------------------*- C++ -*----------------------------------*\ | ========= | | | \\ / F ield | OpenFOAM: The Open Source CFD Toolbox | | \\ / O peration | Version: 2.2.0 | | \\ / A nd | Web: www.OpenFOAM.org | | \\/ M anipulation | | \*---------------------------------------------------------------------------*/ FoamFile { version 2.0; format ascii; class dictionary; location "system"; object fvSolution; } // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // solvers { "rho.*" { solver diagonal; } "p.*" { solver PBiCG; preconditioner DILU; tolerance 1e-08; relTol 0; } "(U|e|R).*" { $p; tolerance 1e-05; } "(k|epsilon).*" { $p; tolerance 1e-08; } } PIMPLE { nOuterCorrectors 2; nCorrectors 1; nNonOrthogonalCorrectors 0; } and what's the general way to have a good selection of tolerances in fvSolution?
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King. To Be or Not To Be,Thats the Question! The Only Stupid Question Is the One that Goes Unasked. |
|
June 15, 2013, 13:55 |
|
#2 |
Retired Super Moderator
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 10,974
Blog Entries: 45
Rep Power: 128 |
Hi Ehsan,
Oddly enough, I think I know some parts to this question, although I'm in no way an expert on this topic, so I ask anyone who has more experience on this to fill in the gaps. First of all, the OpenFOAM User Guide explains what the parameters "tolerance" and "relTolerance" are, in subsection "4.5.1.1 Solution tolerances": http://www.openfoam.org/docs/user/fv...-1190004.5.1.1 Secondly, since it's about "tolerance" you're asking about, as the User Guide indicates, refers to the target final residual for the equation solution, which should be the default option to be used for transient simulations. As for values, it all depends on the level of accuracy your looking for. But in this case, I suspect it's because the original author doesn't want errors from the turbulence fields to introduce unwanted cumulative errors into the solution, since "U" and "p" are affected by the turbulence fields. Most likely, it's also due to this: Code:
div(phi,k) Gauss upwind; div(phi,epsilon) Gauss upwind; While for "U", tolerance of 1e-5 is "good enough". AFAIK, "U" and "p" should be at least below 1e-4, so 1e-5 is a good target value. And since (I think) they are solved with 2nd degree div schemes, lower values would be a time waster, unless you were dealing with micro-fluids or something like that (it's just a guess here!). Best regards, Bruno
__________________
|
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
SimpleFoam convergence problems | brahim | OpenFOAM Running, Solving & CFD | 20 | June 9, 2015 10:09 |
how to do the fvSolution configuration - SIMPLE Foam - complex cylinder geometry | despaired student | OpenFOAM Running, Solving & CFD | 10 | July 5, 2012 04:36 |
interFoam solution tolerances | mgdenno | OpenFOAM | 4 | September 13, 2011 13:58 |
[ICEM] Procedure : how to mesh correctly two semicircles in an air flow ? | eeesprit | ANSYS Meshing & Geometry | 2 | April 10, 2011 16:14 |
Calculation of Averages in lesInterFoam Erroneous Procedure | ngj | OpenFOAM Bugs | 6 | April 15, 2008 14:21 |