|
[Sponsors] |
September 28, 2019, 16:26 |
Speed up convergence with PimpleFOAM
|
#1 |
New Member
Milad Mozayyani
Join Date: Jun 2019
Posts: 4
Rep Power: 7 |
Hey all, so I have been running a case on pimple foam for around a week now and am looking for some pointers to try and speed up convergence. I'll post logs but it is running stably at a maxCo of 150, resulting in time steps of around 0.0001. Each time step converges in about 18 or so iterations but this takes about 5 minutes between each timestep. I am trying to capture at least 0.5s of flow in order to make sure any long period oscillations are captured. It seems like velocity converges pretty quickly because it basically never iterates inside any of the loops, but pressure takes a while. Posted below are the last few iterations of a timestep and my fvSchemes and fvSolution files.
Thanks! Code:
PIMPLE: iteration 12 smoothSolver: Solving for Ux, Initial residual = 4.74636e-09, Final residual = 4.74636e-09, No Iterations 0 smoothSolver: Solving for Uy, Initial residual = 4.83958e-07, Final residual = 4.83958e-07, No Iterations 0 smoothSolver: Solving for Uz, Initial residual = 9.97602e-08, Final residual = 9.97602e-08, No Iterations 0 GAMG: Solving for p, Initial residual = 0.000129754, Final residual = 9.73145e-06, No Iterations 2 time step continuity errors : sum local = 2.95278e-11, global = -1.77276e-14, cumulative = 1.72506e-09 GAMG: Solving for p, Initial residual = 8.87815e-05, Final residual = 8.26856e-07, No Iterations 4 time step continuity errors : sum local = 2.50891e-12, global = -7.31017e-15, cumulative = 1.72505e-09 PIMPLE: iteration 13 smoothSolver: Solving for Ux, Initial residual = 3.32e-09, Final residual = 3.32e-09, No Iterations 0 smoothSolver: Solving for Uy, Initial residual = 3.33212e-07, Final residual = 3.33212e-07, No Iterations 0 smoothSolver: Solving for Uz, Initial residual = 6.44759e-08, Final residual = 6.44759e-08, No Iterations 0 GAMG: Solving for p, Initial residual = 9.44212e-05, Final residual = 7.96073e-06, No Iterations 2 time step continuity errors : sum local = 2.41549e-11, global = -1.46594e-14, cumulative = 1.72503e-09 GAMG: Solving for p, Initial residual = 6.52984e-05, Final residual = 7.41643e-07, No Iterations 4 time step continuity errors : sum local = 2.25034e-12, global = -1.1951e-14, cumulative = 1.72502e-09 PIMPLE: iteration 14 smoothSolver: Solving for Ux, Initial residual = 2.41505e-09, Final residual = 2.41505e-09, No Iterations 0 smoothSolver: Solving for Uy, Initial residual = 2.40896e-07, Final residual = 2.40896e-07, No Iterations 0 smoothSolver: Solving for Uz, Initial residual = 4.39377e-08, Final residual = 4.39377e-08, No Iterations 0 GAMG: Solving for p, Initial residual = 7.02713e-05, Final residual = 6.38299e-06, No Iterations 2 time step continuity errors : sum local = 1.93676e-11, global = -1.82954e-14, cumulative = 1.725e-09 GAMG: Solving for p, Initial residual = 4.89942e-05, Final residual = 5.92647e-07, No Iterations 4 time step continuity errors : sum local = 1.79824e-12, global = -1.32138e-14, cumulative = 1.72499e-09 PIMPLE: iteration 15 smoothSolver: Solving for Ux, Initial residual = 1.81951e-09, Final residual = 1.81951e-09, No Iterations 0 smoothSolver: Solving for Uy, Initial residual = 1.81386e-07, Final residual = 1.81386e-07, No Iterations 0 smoothSolver: Solving for Uz, Initial residual = 3.10416e-08, Final residual = 3.10416e-08, No Iterations 0 GAMG: Solving for p, Initial residual = 5.38205e-05, Final residual = 4.89759e-06, No Iterations 2 time step continuity errors : sum local = 1.48604e-11, global = -1.76769e-14, cumulative = 1.72497e-09 GAMG: Solving for p, Initial residual = 3.76034e-05, Final residual = 9.16247e-07, No Iterations 3 time step continuity errors : sum local = 2.78011e-12, global = 1.46448e-15, cumulative = 1.72497e-09 smoothSolver: Solving for omega, Initial residual = 2.15262e-06, Final residual = 2.15262e-06, No Iterations 0 smoothSolver: Solving for k, Initial residual = 2.05128e-05, Final residual = 2.95276e-07, No Iterations 1 bounding k, min: 4.28226e-16 max: 750.312 average: 13.4291 PIMPLE: converged in 15 iterations ExecutionTime = 673218 s ClockTime = 677344 s Code:
FoamFile { version 2.0; format ascii; class dictionary; object fvSolution; } // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // solvers { p { solver GAMG; tolerance 1e-06; relTol 0.1; smoother GaussSeidel; } Phi { $p; } pFinal { $p; smoother DICGaussSeidel; tolerance 1e-06; relTol 0; } "(U|k|omega)" { solver smoothSolver; smoother symGaussSeidel; tolerance 1e-05; relTol 0.1; } "(U|k|omega)Final" { $U; tolerance 1e-05; relTol 0; } } SIMPLE { nNonOrthogonalCorrectors 2; consistent yes; } PIMPLE { nOuterCorrectors 100; nCorrectors 2; residualControl { U { tolerance 1e-4; relTol 0; } p { tolerance 5e-5; relTol 0; } } } potentialFlow { nNonOrthogonalCorrectors 10; } relaxationFactors { fields { p 0.4; pFinal 0.4; } equations { "U|k|omega" 0.6; "(U|k|omega)Final" 0.6; } } Code:
FoamFile { version 2.0; format ascii; class dictionary; object fvSchemes; } // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // ddtSchemes { default backward; } gradSchemes { default cellLimited leastSquares 1; grad(U) cellLimited leastSquares 1; } divSchemes { default none; div(phi,U) Gauss upwind; div(phi,k) Gauss upwind; div(phi,omega) Gauss upwind; div((nuEff*dev2(T(grad(U))))) Gauss linear; } laplacianSchemes { default Gauss linear limited 0.7; } interpolationSchemes { default linear; } snGradSchemes { default limited 0.7; } wallDist { method meshWave; } |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
pimpleFoam convergence | kandelabr | OpenFOAM Running, Solving & CFD | 4 | February 8, 2018 13:38 |
Convergence problem with pimpleFoam with OpenFOAM-v1706 | kaszt | OpenFOAM Running, Solving & CFD | 1 | September 5, 2017 21:00 |
Bad convergence for flow separation in T-junction | MelroseBing | CFX | 2 | May 17, 2016 01:59 |
Speed of Convergence faster with smaller time step | Torque_Converter | CFX | 1 | May 15, 2012 20:22 |
Problem with the convergence of the speed profiles | Inma Cabanillas | FLUENT | 0 | February 3, 2009 07:03 |