[sharonyue]

Hi guys,

I am simulating a mixer3D case, totally mesh is about 200m. I expect MRFInterFoam could be faster than interDyMFoam, but via the log, interDyMFoam is faster.


Both the result is fine.So why?any comments would be appreciated.


MRFlog

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create mesh for time = 0

Reading field p_rgh

Reading field U

Reading/calculating face flux field phi

AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 1, 1.00723, 1.00001
AMI: Patch target weights min/max/average = 1, 1.00723, 1.00001
Reading transportProperties

Selecting incompressible transport model Newtonian
Selecting incompressible transport model Newtonian
Selecting turbulence model type RASModel
Selecting RAS turbulence model kEpsilon
bounding k, min: 0 max: 1 average: 1
bounding epsilon, min: 0 max: 20 average: 20
kEpsilonCoeffs
{
    Cmu             0.09;
    C1              1.44;
    C2              1.92;
    sigmaEps        1.3;
}


Reading g
Calculating field g.h

No finite volume options present

Creating MRF zone list from MRFProperties
    creating MRF zone: zone1

PIMPLE: Operating solver in PISO mode

time step continuity errors : sum local = 0, global = 0, cumulative = 0
DICPCG:  Solving for pcorr, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 0, global = 0, cumulative = 0
Courant Number mean: 0 max: 0

Starting time loop

Courant Number mean: 0 max: 0
Interface Courant Number mean: 0 max: 0
deltaT = 1.07991e-05
Time = 1.07991e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 1, Final residual = 0.0495223, No Iterations 19
time step continuity errors : sum local = 1.7138e-06, global = -1.29864e-08, cumulative = -1.29864e-08
DICPCG:  Solving for p_rgh, Initial residual = 0.122609, Final residual = 0.00580765, No Iterations 14
time step continuity errors : sum local = 4.70704e-07, global = -4.78148e-09, cumulative = -1.77679e-08
DICPCG:  Solving for p_rgh, Initial residual = 0.017567, Final residual = 9.90151e-08, No Iterations 161
time step continuity errors : sum local = 8.01748e-12, global = 5.02083e-14, cumulative = -1.77678e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 0.000142027, Final residual = 1.31604e-09, No Iterations 3
DILUPBiCG:  Solving for k, Initial residual = 1, Final residual = 1.12205e-09, No Iterations 7
ExecutionTime = 9.97 s  ClockTime = 10 s

Courant Number mean: 0.00216553 max: 0.53051
Interface Courant Number mean: 0 max: 0
deltaT = 1.0177e-05
Time = 2.09761e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.44695e-30  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.160646, Final residual = 0.0072516, No Iterations 26
time step continuity errors : sum local = 5.00609e-07, global = 9.94811e-09, cumulative = -7.8197e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.0921909, Final residual = 0.00435461, No Iterations 8
time step continuity errors : sum local = 2.83072e-07, global = 2.82049e-09, cumulative = -4.99921e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.0085199, Final residual = 9.89058e-08, No Iterations 170
time step continuity errors : sum local = 6.29855e-12, global = -1.17153e-13, cumulative = -4.99932e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 3.50951e-05, Final residual = 2.19052e-09, No Iterations 2
bounding epsilon, min: -0.961442 max: 12468.1 average: 319.429
DILUPBiCG:  Solving for k, Initial residual = 0.00849041, Final residual = 1.83622e-09, No Iterations 4
ExecutionTime = 18.62 s  ClockTime = 18 s

Courant Number mean: 0.00204717 max: 0.556823
Interface Courant Number mean: 0 max: 0
deltaT = 9.13693e-06
Time = 3.0113e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -7.2564e-31  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -7.53487e-31  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0608071, Final residual = 0.00295234, No Iterations 19
time step continuity errors : sum local = 2.65526e-07, global = -3.68452e-10, cumulative = -5.36777e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.0164258, Final residual = 0.000808253, No Iterations 13
time step continuity errors : sum local = 7.18554e-08, global = -1.17925e-09, cumulative = -6.54703e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00174533, Final residual = 9.50677e-08, No Iterations 153
time step continuity errors : sum local = 8.46815e-12, global = 6.33871e-14, cumulative = -6.54696e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 3.00088e-05, Final residual = 1.60449e-09, No Iterations 2
bounding epsilon, min: -3.14248 max: 11394.3 average: 319.679
DILUPBiCG:  Solving for k, Initial residual = 0.00602194, Final residual = 8.43019e-10, No Iterations 4
ExecutionTime = 26.5 s  ClockTime = 26 s

Courant Number mean: 0.00183615 max: 0.525867
Interface Courant Number mean: 0 max: 0
deltaT = 8.68739e-06
Time = 3.88004e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.04223e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.04223e-30  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0338219, Final residual = 0.00151052, No Iterations 19
time step continuity errors : sum local = 1.2876e-07, global = 1.12176e-10, cumulative = -6.43479e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00809159, Final residual = 0.000383803, No Iterations 13
time step continuity errors : sum local = 3.19975e-08, global = 2.05515e-10, cumulative = -6.22927e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.000821271, Final residual = 9.88127e-08, No Iterations 147
time step continuity errors : sum local = 8.20908e-12, global = -2.54011e-14, cumulative = -6.2293e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.74772e-05, Final residual = 2.21332e-09, No Iterations 2
bounding epsilon, min: -15.8509 max: 10622.5 average: 320.093
DILUPBiCG:  Solving for k, Initial residual = 0.0049088, Final residual = 8.57433e-10, No Iterations 4
ExecutionTime = 34.34 s  ClockTime = 34 s

Courant Number mean: 0.0017467 max: 0.513383
Interface Courant Number mean: 0 max: 0
deltaT = 8.46083e-06
Time = 4.72612e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.03781e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.6606e-30  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0229209, Final residual = 0.00111938, No Iterations 18
time step continuity errors : sum local = 8.28057e-08, global = 2.41502e-11, cumulative = -6.20515e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.0051387, Final residual = 0.000226864, No Iterations 14
time step continuity errors : sum local = 1.67717e-08, global = -6.4155e-12, cumulative = -6.21156e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.000532194, Final residual = 9.43961e-08, No Iterations 145
time step continuity errors : sum local = 6.98896e-12, global = 5.56208e-14, cumulative = -6.21151e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.59259e-05, Final residual = 4.43109e-09, No Iterations 3
bounding epsilon, min: -35.6881 max: 10019.6 average: 320.542
DILUPBiCG:  Solving for k, Initial residual = 0.00429238, Final residual = 6.20346e-09, No Iterations 5
ExecutionTime = 43.51 s  ClockTime = 43 s

Courant Number mean: 0.00170077 max: 0.510754
Interface Courant Number mean: 0 max: 0
deltaT = 8.28266e-06
Time = 5.55439e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.68908e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -4.38537e-30  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0144755, Final residual = 0.000649306, No Iterations 18
time step continuity errors : sum local = 4.65693e-08, global = 2.10894e-11, cumulative = -6.19042e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.0030582, Final residual = 0.000147816, No Iterations 13
time step continuity errors : sum local = 1.05051e-08, global = 9.01834e-12, cumulative = -6.1814e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.000317831, Final residual = 9.03361e-08, No Iterations 140
time step continuity errors : sum local = 6.41349e-12, global = 1.32232e-13, cumulative = -6.18127e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.46021e-05, Final residual = 1.10753e-09, No Iterations 2
DILUPBiCG:  Solving for k, Initial residual = 0.00385872, Final residual = 7.08998e-09, No Iterations 3
ExecutionTime = 52.81 s  ClockTime = 52 s

Courant Number mean: 0.00166523 max: 0.50848
Interface Courant Number mean: 0 max: 0
deltaT = 8.14356e-06
Time = 6.36875e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.73436e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.29233e-29  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0101018, Final residual = 0.000428819, No Iterations 18
time step continuity errors : sum local = 2.85356e-08, global = 2.1336e-12, cumulative = -6.17913e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00196009, Final residual = 9.6592e-05, No Iterations 13
time step continuity errors : sum local = 6.42842e-09, global = 8.08073e-12, cumulative = -6.17105e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00020677, Final residual = 9.61554e-08, No Iterations 138
time step continuity errors : sum local = 6.40504e-12, global = -1.40573e-13, cumulative = -6.17119e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.35861e-05, Final residual = 1.05688e-09, No Iterations 2
bounding epsilon, min: -1.13847 max: 9136.01 average: 321.43
DILUPBiCG:  Solving for k, Initial residual = 0.00354651, Final residual = 5.20685e-09, No Iterations 3
ExecutionTime = 60.61 s  ClockTime = 60 s

Courant Number mean: 0.0016372 max: 0.506738
Interface Courant Number mean: 0 max: 0
deltaT = 8.0348e-06
Time = 7.17223e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.28447e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -7.93829e-30  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.00678436, Final residual = 0.000295459, No Iterations 17
time step continuity errors : sum local = 1.95926e-08, global = -6.30779e-12, cumulative = -6.1775e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00124079, Final residual = 6.04477e-05, No Iterations 14
time step continuity errors : sum local = 3.98948e-09, global = -3.47633e-12, cumulative = -6.18098e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.000133578, Final residual = 9.99504e-08, No Iterations 127
time step continuity errors : sum local = 6.59323e-12, global = -2.21552e-14, cumulative = -6.181e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.27288e-05, Final residual = 1.00723e-09, No Iterations 2
DILUPBiCG:  Solving for k, Initial residual = 0.00330894, Final residual = 4.07864e-09, No Iterations 3
ExecutionTime = 68.02 s  ClockTime = 68 s

Courant Number mean: 0.00161543 max: 0.505443
Interface Courant Number mean: 0 max: 0
deltaT = 7.94783e-06
Time = 7.96701e-05

MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.27255e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -4.56795e-29  Max(alpha1) = 1
Phase volume in zone innerCylinder = 73689.4 ml 
DICPCG:  Solving for p_rgh, Initial residual = 0.0050451, Final residual = 0.000233662, No Iterations 16
time step continuity errors : sum local = 1.46823e-08, global = 4.15023e-12, cumulative = -6.17685e-09
DICPCG:  Solving for p_rgh, Initial residual = 0.00085883, Final residual = 3.82176e-05, No Iterations 16
time step continuity errors : sum local = 2.40204e-09, global = 6.85582e-12, cumulative = -6.16999e-09
DICPCG:  Solving for p_rgh, Initial residual = 9.63252e-05, Final residual = 9.71377e-08, No Iterations 126
time step continuity errors : sum local = 6.1099e-12, global = 9.08267e-14, cumulative = -6.1699e-09
DILUPBiCG:  Solving for epsilon, Initial residual = 2.20023e-05, Final residual = 9.74754e-10, No Iterations 2
DILUPBiCG:  Solving for k, Initial residual = 0.00312384, Final residual = 3.44313e-09, No Iterations 3
ExecutionTime = 75.58 s  ClockTime = 75 s

[sharonyue]

DyMlog

Create time

Create mesh for time = 0

Selecting dynamicFvMesh solidBodyMotionFvMesh
Selecting solid-body motion function rotatingMotion
Applying solid body motion to cellZone innerCylinder
Reading field p_rgh

Reading field U

Reading/calculating face flux field phi

AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 1, 1.00723, 1.00001
AMI: Patch target weights min/max/average = 1, 1.00723, 1.00001
Reading transportProperties

Selecting incompressible transport model Newtonian
Selecting incompressible transport model Newtonian
Selecting turbulence model type RASModel
Selecting RAS turbulence model kEpsilon
bounding k, min: 0 max: 1 average: 1
bounding epsilon, min: 0 max: 20 average: 20
kEpsilonCoeffs
{
    Cmu             0.09;
    C1              1.44;
    C2              1.92;
    sigmaEps        1.3;
}


Reading g
Calculating field g.h

No finite volume options present


PIMPLE: Operating solver in PISO mode

time step continuity errors : sum local = 0, global = 0, cumulative = 0
GAMGPCG:  Solving for pcorr, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 0, global = 0, cumulative = 0
Courant Number mean: 0 max: 0

Starting time loop

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0 max: 0
deltaT = 1.07991e-05
Time = 1.07991e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 1.07991e-05 transformation: ((0 0 0) (1 (0 -0.000169632 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999981, 1.01009, 1.00001
AMI: Patch target weights min/max/average = 0.999872, 1.00789, 1.00001
Execution time for mesh.update() = 0.57 s
time step continuity errors : sum local = 0, global = 0, cumulative = 0
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 5.85724e-05, No Iterations 6
time step continuity errors : sum local = 1.1423e-15, global = -2.09364e-16, cumulative = -2.09364e-16
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 1, Final residual = 0.0465405, No Iterations 2
time step continuity errors : sum local = 1.57011e-06, global = 1.87009e-08, cumulative = 1.87009e-08
GAMG:  Solving for p_rgh, Initial residual = 0.0789648, Final residual = 0.00246715, No Iterations 2
time step continuity errors : sum local = 2.88529e-07, global = 6.79494e-10, cumulative = 1.93804e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 0.0106098, Final residual = 4.05951e-09, No Iterations 9
time step continuity errors : sum local = 4.86709e-13, global = 1.19577e-14, cumulative = 1.93804e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 0.000141365, Final residual = 9.72214e-07, No Iterations 1
DILUPBiCG:  Solving for k, Initial residual = 1, Final residual = 9.04443e-07, No Iterations 4
ExecutionTime = 11.88 s  ClockTime = 12 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.00219444 max: 0.530883
deltaT = 1.01693e-05
Time = 2.09684e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 2.09684e-05 transformation: ((0 0 0) (1 (0 -0.000329371 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999903, 1.0128, 1.00002
AMI: Patch target weights min/max/average = 0.999753, 1.00851, 1.00002
Execution time for mesh.update() = 0.73 s
time step continuity errors : sum local = 4.58334e-13, global = 1.10632e-14, cumulative = 1.93804e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 1.85074e-05, No Iterations 6
time step continuity errors : sum local = 1.04091e-15, global = -3.97154e-17, cumulative = 1.93804e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = 0  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.40366e-31  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.40366e-31  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.111485, Final residual = 0.00298827, No Iterations 4
time step continuity errors : sum local = 1.87951e-07, global = -8.61162e-10, cumulative = 1.85192e-08
GAMG:  Solving for p_rgh, Initial residual = 0.0806247, Final residual = 0.0016671, No Iterations 2
time step continuity errors : sum local = 9.60968e-08, global = -1.21665e-09, cumulative = 1.73026e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 0.00729125, Final residual = 2.44918e-09, No Iterations 9
time step continuity errors : sum local = 1.22663e-13, global = 7.78887e-15, cumulative = 1.73026e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.26066e-05, Final residual = 6.98713e-08, No Iterations 1
bounding epsilon, min: -2.33489 max: 12270 average: 318.304
DILUPBiCG:  Solving for k, Initial residual = 0.00893094, Final residual = 1.52782e-06, No Iterations 2
ExecutionTime = 22.12 s  ClockTime = 22 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.00206649 max: 0.471772
deltaT = 1.07761e-05
Time = 3.17445e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 3.17445e-05 transformation: ((0 0 0) (1 (0 -0.000498641 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999807, 1.01565, 1.00003
AMI: Patch target weights min/max/average = 0.999626, 1.00917, 1.00003
Execution time for mesh.update() = 0.83 s
time step continuity errors : sum local = 1.30033e-13, global = 8.21103e-15, cumulative = 1.73026e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 2.08231e-05, No Iterations 5
time step continuity errors : sum local = 1.14651e-15, global = 1.625e-16, cumulative = 1.73026e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.12177e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -4.14319e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.00275e-30  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.00924303, Final residual = 0.000183372, No Iterations 2
time step continuity errors : sum local = 2.01927e-08, global = -3.82154e-11, cumulative = 1.72644e-08
GAMG:  Solving for p_rgh, Initial residual = 0.0010327, Final residual = 4.32465e-05, No Iterations 2
time step continuity errors : sum local = 4.73384e-09, global = -1.00248e-10, cumulative = 1.71641e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 0.000160943, Final residual = 6.14766e-09, No Iterations 6
time step continuity errors : sum local = 6.76895e-13, global = -2.3289e-14, cumulative = 1.71641e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.28376e-05, Final residual = 7.18948e-08, No Iterations 1
bounding epsilon, min: -25.3986 max: 11277.1 average: 318.745
DILUPBiCG:  Solving for k, Initial residual = 0.00691905, Final residual = 1.49221e-06, No Iterations 2
ExecutionTime = 30.31 s  ClockTime = 30 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.0021898 max: 0.476022
deltaT = 1.13173e-05
Time = 4.30618e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 4.30618e-05 transformation: ((0 0 0) (1 (0 -0.000676413 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999706, 1.01864, 1.00004
AMI: Patch target weights min/max/average = 0.999492, 1.00987, 1.00003
Execution time for mesh.update() = 0.77 s
time step continuity errors : sum local = 7.10885e-13, global = -2.4288e-14, cumulative = 1.71641e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 7.44623e-05, No Iterations 5
time step continuity errors : sum local = 1.1549e-15, global = -7.19047e-17, cumulative = 1.71641e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -7.97699e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.53358e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.10898e-30  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.00313902, Final residual = 4.86501e-05, No Iterations 2
time step continuity errors : sum local = 6.11297e-09, global = 2.00984e-11, cumulative = 1.71842e-08
GAMG:  Solving for p_rgh, Initial residual = 0.000415594, Final residual = 1.11061e-05, No Iterations 2
time step continuity errors : sum local = 1.39186e-09, global = 3.24225e-13, cumulative = 1.71845e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 5.9657e-05, Final residual = 7.36493e-09, No Iterations 5
time step continuity errors : sum local = 9.22573e-13, global = 5.22275e-14, cumulative = 1.71846e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.31122e-05, Final residual = 7.37862e-08, No Iterations 1
bounding epsilon, min: -17.1488 max: 10497.8 average: 319.131
DILUPBiCG:  Solving for k, Initial residual = 0.00581176, Final residual = 1.37789e-06, No Iterations 2
ExecutionTime = 39.07 s  ClockTime = 39 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.0022998 max: 0.481042
deltaT = 1.17617e-05
Time = 5.48235e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 5.48235e-05 transformation: ((0 0 0) (1 (0 -0.000861165 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999602, 1.02174, 1.00005
AMI: Patch target weights min/max/average = 0.999354, 1.0106, 1.00004
Execution time for mesh.update() = 0.81 s
time step continuity errors : sum local = 9.58819e-13, global = 5.42002e-14, cumulative = 1.71846e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 1.96255e-05, No Iterations 6
time step continuity errors : sum local = 1.14831e-15, global = 7.02381e-17, cumulative = 1.71846e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -5.85705e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.53586e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.25193e-29  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.00171074, Final residual = 2.69953e-05, No Iterations 2
time step continuity errors : sum local = 3.65318e-09, global = -1.63339e-11, cumulative = 1.71683e-08
GAMG:  Solving for p_rgh, Initial residual = 0.000266381, Final residual = 6.07096e-06, No Iterations 2
time step continuity errors : sum local = 8.20523e-10, global = 1.05994e-12, cumulative = 1.71693e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 3.85518e-05, Final residual = 3.00878e-09, No Iterations 5
time step continuity errors : sum local = 4.06444e-13, global = -2.41697e-14, cumulative = 1.71693e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.31698e-05, Final residual = 7.43642e-08, No Iterations 1
bounding epsilon, min: -14.0313 max: 9872.48 average: 319.451
DILUPBiCG:  Solving for k, Initial residual = 0.00509662, Final residual = 1.40194e-06, No Iterations 2
ExecutionTime = 47.4 s  ClockTime = 47 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.0023901 max: 0.484863
deltaT = 1.21285e-05
Time = 6.69519e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 6.69519e-05 transformation: ((0 0 0) (0.999999 (0 -0.00105168 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.999497, 1.02491, 1.00006
AMI: Patch target weights min/max/average = 0.999212, 1.01135, 1.00005
Execution time for mesh.update() = 0.76 s
time step continuity errors : sum local = 4.19134e-13, global = -2.48519e-14, cumulative = 1.71693e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 3.53546e-05, No Iterations 5
time step continuity errors : sum local = 1.04974e-15, global = 6.00047e-17, cumulative = 1.71693e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -6.68026e-30  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.23842e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -2.32055e-29  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.00112309, Final residual = 1.93869e-05, No Iterations 2
time step continuity errors : sum local = 2.77824e-09, global = 5.02977e-12, cumulative = 1.71743e-08
GAMG:  Solving for p_rgh, Initial residual = 0.000201329, Final residual = 4.32628e-06, No Iterations 2
time step continuity errors : sum local = 6.19553e-10, global = 4.85506e-12, cumulative = 1.71792e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 2.98488e-05, Final residual = 4.46162e-09, No Iterations 5
time step continuity errors : sum local = 6.38652e-13, global = -3.95783e-14, cumulative = 1.71791e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.3014e-05, Final residual = 7.2015e-08, No Iterations 1
bounding epsilon, min: -4.3537 max: 9380.48 average: 319.728
DILUPBiCG:  Solving for k, Initial residual = 0.00459607, Final residual = 1.00059e-06, No Iterations 2
ExecutionTime = 55.51 s  ClockTime = 55 s

Interface Courant Number mean: 0 max: 0
Courant Number mean: 0.00246466 max: 0.487793
deltaT = 1.24307e-05
Time = 7.93826e-05

solidBodyMotionFunctions::rotatingMotion::transformation(): Time = 7.93826e-05 transformation: ((0 0 0) (0.999999 (0 -0.00124694 0)))
AMI: Creating addressing and weights between 6660 source faces and 6660 target faces
AMI: Patch source weights min/max/average = 0.99939, 1.02815, 1.00006
AMI: Patch target weights min/max/average = 0.999066, 1.01213, 1.00005
Execution time for mesh.update() = 0.73 s
time step continuity errors : sum local = 6.54573e-13, global = -4.05042e-14, cumulative = 1.71791e-08
GAMGPCG:  Solving for pcorr, Initial residual = 1, Final residual = 3.58974e-05, No Iterations 6
time step continuity errors : sum local = 1.04694e-15, global = 5.07291e-17, cumulative = 1.71791e-08
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -5.06817e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -3.68502e-29  Max(alpha1) = 1
MULES: Solving for alpha1
Phase-1 volume fraction = 0.801613  Min(alpha1) = -1.12484e-28  Max(alpha1) = 1
GAMG:  Solving for p_rgh, Initial residual = 0.000809561, Final residual = 1.5265e-05, No Iterations 2
time step continuity errors : sum local = 2.29045e-09, global = 2.31451e-12, cumulative = 1.71814e-08
GAMG:  Solving for p_rgh, Initial residual = 0.000162648, Final residual = 3.58839e-06, No Iterations 2
time step continuity errors : sum local = 5.38243e-10, global = 4.18246e-12, cumulative = 1.71856e-08
GAMGPCG:  Solving for p_rgh, Initial residual = 2.49227e-05, Final residual = 3.39738e-09, No Iterations 5
time step continuity errors : sum local = 5.094e-13, global = -3.44719e-14, cumulative = 1.71856e-08
DILUPBiCG:  Solving for epsilon, Initial residual = 3.27583e-05, Final residual = 6.98992e-08, No Iterations 1
bounding epsilon, min: -3.47421 max: 8996.31 average: 319.978
DILUPBiCG:  Solving for k, Initial residual = 0.00422555, Final residual = 1.25871e-06, No Iterations 2
ExecutionTime = 64.44 s  ClockTime = 64 s

[wyldckat]

Greetings sharonyue,

I'm not very experienced on this topic, but I do have a few of questions/comments:

1. Which OpenFOAM version are you using?
2. Are you running both in parallel? If so, which decomposition method was used?
3. Why is the MRF version referring to AMI? Do the tutorials for MRFInterFoam also refer to AMI in their output?
4. Of the 200 million cells, how many cells are moving?
5. Are you certain that the same volume of fluid is filled on both cases?
6. Are there any cyclic patches, symmetry planes and so on?
7. This seems to be too early in the simulation to diagnose which one is fastest...

Best regards,
Bruno

[sharonyue]

Originally Posted by wyldckat

Greetings sharonyue,

I'm not very experienced on this topic, but I do have a few of questions/comments:

1. Which OpenFOAM version are you using?
2. Are you running both in parallel? If so, which decomposition method was used?
3. Why is the MRF version referring to AMI? Do the tutorials for MRFInterFoam also refer to AMI in their output?
4. Of the 200 million cells, how many cells are moving?
5. Are you certain that the same volume of fluid is filled on both cases?
6. Are there any cyclic patches, symmetry planes and so on?
7. This seems to be too early in the simulation to diagnose which one is fastest...

Best regards,
Bruno

Hi Bruno,

I have read lots of your post,Thats very helpful.Thanks for sharing your CFD experience.

Regarding to my case.
[*]Which OpenFOAM version are you using?

I am using 220
[*]Are you running both in parallel? If so, which decomposition method was used?

Yes, I am using simple method in my laptop just for a test.
[*]Why is the MRF version referring to AMI? Do the tutorials for MRFInterFoam also refer to AMI in their output?

Actually,in MRF, I dont need AMI,but if I want to run the same case in sliding mesh and MRF. I have to use two meshes. While in MRFInterFoam, I can use the mesh with AMI, so I dont need to make another mesh for dynamic solver.
[*]Of the 200 million cells, how many cells are moving?

In sliding mesh, There are 2127678 cells.Total cell is 2162754.
[*]Are you certain that the same volume of fluid is filled on both cases?

Yeah,of course,I use the same setFieldsDict.

[*]Are there any cyclic patches, symmetry planes and so on?

Nope, The majority of em are walls,impeller,Only one outlet.
[*]This seems to be too early in the simulation to diagnose which one is fastest...

Um...Perhaps I should try it another day for a long runing..?
In my laptop, it took about 8 hours calculating 0.08 seconds.

[wyldckat]

Hi sharonyue,

Originally Posted by sharonyue

[*]Why is the MRF version referring to AMI? Do the tutorials for MRFInterFoam also refer to AMI in their output?

Actually,in MRF, I dont need AMI,but if I want to run the same case in sliding mesh and MRF. I have to use two meshes. While in MRFInterFoam, I can use the mesh with AMI, so I dont need to make another mesh for dynamic solver.

That's somewhat of a dangerous assumption. I'm probably wrong, but given OpenFOAM's 2.2 "fvOptions" capabilities, you might be using both AMI dynamic meshing algorithms and MRF at the same time, without even knowing it! Or at the very least, half of the AMI algorithm might be in effect at the same time of the MRF, therefore leading to the slowdown you're seeing.

Originally Posted by sharonyue

[*]Of the 200 million cells, how many cells are moving?

In sliding mesh, There are 2127678 cells.Total cell is 2162754.

Did I understand you correctly? Over 98% of the cells are moving around? That's a lot of cells in motion, or at least in emulated motion...

Originally Posted by sharonyue

[*]This seems to be too early in the simulation to diagnose which one is fastest...

Um...Perhaps I should try it another day for a long runing..?
In my laptop, it took about 8 hours calculating 0.08 seconds.

But you showed only a minute's worth of output. I thought you only had run for about a minute
Roughly 1 or 2 hours each should be more than enough, I guess.
But first you better reconfigure the MRF case to not rely on AMI at all, if possible!

Best regards,
Bruno

[sharonyue]

Originally Posted by wyldckat

That's somewhat of a dangerous assumption. I'm probably wrong, but given OpenFOAM's 2.2 "fvOptions" capabilities, you might be using both AMI dynamic meshing algorithms and MRF at the same time, without even knowing it! Or at the very least, half of the AMI algorithm might be in effect at the same time of the MRF, therefore leading to the slowdown you're seeing.

Woo,This does make sense. Indeed MRF dont need AMI,

Did I understand you correctly? Over 98% of the cells are moving around? That's a lot of cells in motion, or at least in emulated motion...

Yeah,my impellers are a little complex and big.So in the stirred tank it take many cells.But surely I should try to cease the number of cells.

Is there anything different between this two situation?I regard the result is independent with the rotating cells number.

But you showed only a minute's worth of output. I thought you only had run for about a minute
Roughly 1 or 2 hours each should be more than enough, I guess.
But first you better reconfigure the MRF case to not rely on AMI at all, if possible!

I will try the MRF method without the AMI.
Thank you very much.This do me a lot favor~.

Regards.

[wyldckat]

Hi sharonyue,

Originally Posted by sharonyue

Is there anything different between this two situation?I regard the result is independent with the rotating cells number.

My very last theory is that MRF applied to so many cells is heavier than mesh motion, simply because the motion of the mesh affects a lot less cells mesh-wise, making the one with AMI easier/simpler to solve.

For example (all theory here, I have not checked the code!), here's an abstract concept:

• MRF applies a centrifugal force to the rotating area. Has to do 2 passes, 2 flux corrections and one or two relaxations. This is done over 98% of the arrays.
• AMI moves the mesh, displacing 10% of the cells. Does 1 or 2 passes for flux correction. The rest is all standard solver PISO, PIMPLE or SIMPLE.

Therefore, AMI would be lighter simply because it needs less operations to do the same thing.

If this theory is correct, then there is only one (or a few) remaining explanations as to why AMI was not used in the past: memory! AMI might require a lot more memory than MRF to run.
Or perhaps because the algorithm for AMI is a lot more complex... or harder to calibrate... overlapping cells/faces can also be a pain...

Best regards,
Bruno

[sharonyue]

Originally Posted by wyldckat

Hi sharonyue,


My very last theory is that MRF applied to so many cells is heavier than mesh motion, simply because the motion of the mesh affects a lot less cells mesh-wise, making the one with AMI easier/simpler to solve.

For example (all theory here, I have not checked the code!), here's an abstract concept:

• MRF applies a centrifugal force to the rotating area. Has to do 2 passes, 2 flux corrections and one or two relaxations. This is done over 98% of the arrays.
• AMI moves the mesh, displacing 10% of the cells. Does 1 or 2 passes for flux correction. The rest is all standard solver PISO, PIMPLE or SIMPLE.

Therefore, AMI would be lighter simply because it needs less operations to do the same thing.

If this theory is correct, then there is only one (or a few) remaining explanations as to why AMI was not used in the past: memory! AMI might require a lot more memory than MRF to run.
Or perhaps because the algorithm for AMI is a lot more complex... or harder to calibrate... overlapping cells/faces can also be a pain...

Best regards,
Bruno

That seems to be a guide to those who is doing work about rotating machines.yesterday I just make a new mesh.and there are 1280838 cells . and 841458 in the rotating part. and if try to cease the number of the rotating part,I am afarid its tough.

Anyway,this mesh is without AMI.I will try the same mesh with AMI and make a speed test another day.oh my laptop its a pity..

Thanks Bruno. You did a splendid job regarding CFD.

BTW, do you have a blog or something?
Regards.

[wyldckat]

Hi sharonyue,

Here's an interesting read about MRF: http://www.esi-cfd.com/faq/index.php...=19&artlang=en

Originally Posted by sharonyue

You did a splendid job regarding CFD.

My expertise are in installing OpenFOAM When it comes to CFD, I only know some stuff on the topic

Originally Posted by sharonyue

BTW, do you have a blog or something?

The first link in my first signature points to a post to the only blog I've got. Other than that, I've got twitter: @wyldckat

Best regards,
Bruno

[sharonyue]

Originally Posted by wyldckat

Hi sharonyue,

Here's an interesting read about MRF: http://www.esi-cfd.com/faq/index.php...=19&artlang=en


My expertise are in installing OpenFOAM When it comes to CFD, I only know some stuff on the topic


The first link in my first signature points to a post to the only blog I've got. Other than that, I've got twitter: @wyldckat

Best regards,
Bruno

Thanks Bruno,I will check it out deeply,Thats interesting!

Regards

[nash]

hi sharonyue,

this topic seems to be same as what i'm doing now. I have done the simulation for centrifugal fan 3D using simpleFoam with mrfZone and it works well.

I would like to try this simulation using pimpleDyMFoam which requires me to have AMI.
Have you done the one using AMI?? If yes how did you define the AMI? using MRF??

I have read threads regarding the AMI and also the propeller tutorial. In snappyHexMesh there are already faceZone and cellZone being created. Why there is a need to define that again using toposetDict.

I hope someone could help me to have a better understanding in continuing my project using pimpleDyMFoam.

Btw the simulation using simpleFoam MRF took me a day to be completed using 12 processors (including sHM: cells generated almost 8 000 000)

With the assumption from wyldckat that one using pimpleFoam could reduce the time, i would really like to try it and to confirm it..

[wyldckat]

Greetings nash,

Some of the details on this topic are available in this thread: http://www.cfd-online.com/Forums/ope...tml#post446517 namely start reading from post #184

Keep in mind that simpleFoam with MRF gives you a steady-state result. In theory, it should be faster than using pimpleDyMFoam, because this latter solver provides a transient solution with dynamic mesh, which should be several times slower to perform.

The comparable differences in wall clock time is only if you compare pimpleFoam+MRF vs pimpleDyMFoam+AMI, because both are transient solvers.
And keep in mind that MRF is by definition meant for steady-state (I learned this recently), so be very careful when using MRF in transient simulations.

Last but not least:

• MRF is for when all of the mesh is always static.
• AMI is for when some of the mesh is dynamic.
• "cyclicAMI" is a patch type for similar to "cyclic", but meant for when one or both surface mesh patches are moving; but it can still be used for static meshes and sometimes it's easier to use them with running the solver in parallel, than it is with "cyclic".

Best regards,
Bruno

[nash]

thanks Bruno for the explaination.

here is some information i get from the link posted by you:

Originally Posted by wyldckat

Here's an interesting read about MRF: http://www.esi-cfd.com/faq/index.php...=19&artlang=en

the flow of fan case is actually transient. However with the approach of the MRF, this complexity of transient case can be solved using steady state ( simpleFoam). Hence reducing the simulation time.

Lets say if i want to make the mesh move, i need to use AMI+transient solver which will lead to large simulation time as you mentioned above.

Maybe i can run the simulation using transient solver pimpleFoam and compare it with MRF-simpleFoam ( just to confirm the statement from the esi website)

Do you have any idea which solver i could also use to simulate this fan (rotor and stator)? Ideas from other foamers are also welcomed

thanks

[wyldckat]

Hi nash,

I was following nicely your post until this question:

Originally Posted by nash

Do you have any idea which solver i could also use to simulate this fan (rotor and stator)?

I'm not sure I understand your question. simpleFoam and pimple*Foam have already been stated. Anything beyond this is usually for compressible flow and include temperature/heat.

Without knowing more about the case in question, it's complicated to suggest anything else.

Best regards,
Bruno

[simt]

In the log files in the first post it can be seen that different solvers are used (DICPCG and GAMG / GAMGPCG). Make sure the solution control is the same for comparison to be valid.

Best regards