|
[Sponsors] |
January 26, 2013, 12:09 |
Large Eddy Simulation - Residuals
|
#1 |
Member
Anonymous
Join Date: Jan 2012
Location: Canada
Posts: 65
Rep Power: 14 |
Hi Foamers,
I have a question regarding the residuals I am getting for my initial testing on a LES problem. The mesh is very high quality (pure hexahedral), and the cells nearest the wall are designed such that y+ <1. In addition the timestep is sufficiently small so that the max Co number should be less than 0.5, but the log is indicating that it is much higher, but the solution is relatively stable. Here is a snippet of the log file: "Time = 0.000943 Courant Number mean: 0.0301863 max: 45.9277 DILUPBiCG: Solving for Ux, Initial residual = 0.0153451, Final residual = 1.39222e-06, No Iterations 3 DILUPBiCG: Solving for Uy, Initial residual = 0.0154658, Final residual = 1.3237e-06, No Iterations 3 DILUPBiCG: Solving for Uz, Initial residual = 0.0141484, Final residual = 1.36949e-06, No Iterations 3 GAMG: Solving for p, Initial residual = 0.521525, Final residual = 0.00047048, No Iterations 10 GAMG: Solving for p, Initial residual = 0.499248, Final residual = 0.000347157, No Iterations 10 time step continuity errors : sum local = 4.13254e-07, global = 9.54636e-08, cumulative = -7.48025e-05 GAMG: Solving for p, Initial residual = 0.495827, Final residual = 0.00037971, No Iterations 10 GAMG: Solving for p, Initial residual = 0.487864, Final residual = 0.000286617, No Iterations 10 time step continuity errors : sum local = 3.37066e-07, global = 7.99683e-08, cumulative = -7.47226e-05 GAMG: Solving for p, Initial residual = 0.484981, Final residual = 0.000339209, No Iterations 10 GAMG: Solving for p, Initial residual = 0.477989, Final residual = 0.000258817, No Iterations 10 time step continuity errors : sum local = 3.01685e-07, global = 7.23369e-08, cumulative = -7.46502e-05 ExecutionTime = 53404.1 s ClockTime = 54141 s Time = 0.000944" I am having a hard time understanding why this could be occurring. Here is my solution dict: " solvers { p { solver GAMG; preconditioner DIC; agglomerator faceAreaPair; tolerance 1e-05; relTol 0; nCellsInCoarsestLevel 25; mergeLevels 1; smoother GaussSeidel; nPreSweeps 0; nPostSweeps 2; nFinestSweeps 2; maxIter 10; } pFinal { solver GAMG; preconditioner DIC; agglomerator faceAreaPair; tolerance 1e-05; relTol 0; nCellsInCoarsestLevel 25; mergeLevels 1; smoother GaussSeidel; nPreSweeps 0; nPostSweeps 2; nFinestSweeps 2; maxIter 10; } U { solver PBiCG; preconditioner DILU; tolerance 1e-05; relTol 0; } k { solver PBiCG; preconditioner DILU; tolerance 1e-05; relTol 0; } B { solver PBiCG; preconditioner DILU; tolerance 1e-05; relTol 0; } nuTilda { solver PBiCG; preconditioner DILU; tolerance 1e-05; relTol 0; } } PISO { nCorrectors 3; nNonOrthogonalCorrectors 1; } relaxationFactors { fields { p 0.3; } equations { U 0.7; k 0.7; B 0.7; nuTilda 0.7; } } " All of my discretization schemes are 2nd order (Gauss Linear). Similar settings seem fine in FLUENT (my experience is mostly using commercial solvers). I know I must be missing something.... Anyone have any ideas? Any other information I should post? Thanks a lot! |
|
January 26, 2013, 17:50 |
Update
|
#2 |
Member
Anonymous
Join Date: Jan 2012
Location: Canada
Posts: 65
Rep Power: 14 |
I just realized I set the maxIter to 10, which probably isn't letting pressure converge properly. Can anyone comment on this?
BTW, when judging convergence should I monitor the initial or final residuals? Thanks a lot. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
2D Large Eddy Simulation with fluent? | Barry Cole | FLUENT | 4 | October 2, 2013 13:36 |
wall function in large eddy simulation | siamak1424 | FLUENT | 0 | November 22, 2009 09:51 |
Large Eddy Simulation of helium jets | Rajani | FLUENT | 0 | February 19, 2004 12:25 |
LARGE EDDY SIMULATION | Parshant Dhand | FLUENT | 0 | May 19, 2003 22:12 |
Large Eddy Simulation | Raul | CFX | 5 | October 6, 2002 09:29 |