[Frenk_T]

Hello Everybody,

I am using multiphaseEulerFoam (OF 2.2.2) to simulate a gas-liquid flow through a vertical pipe. My simulation works fine but there are very sudden velocity spikes without any obvious cause (and consequently high Courant number and small time step). However the simulation is not diverging afterwards and these spikes happen at different moment for different runs. I am looking at convergence to a quasi steady state: these velocity spikes seems to be just transient events which damp down, but they do slow the process of course...

I already tried to reduce maxCo (now I have: maxAlphaCo = 0.5, nAlphaSubCycles = 5, maxCo = 1, maxDeltaT = 0.0005), different numerical schemes and linear solver. The case is running fine with interFoam but gives unsatisfactory results, while multiphaseEulerFoam (with interface compression) predicts good agreement in terms of pressure drop and liquid holdup (if the spikes are not considered in their calculation).

Courant Number mean: 0.03354866023 max: 0.5000822782
Interface Courant Number mean: 0.03291234457 max: 0.3301922196
deltaT = 3.279629714e-05
Time = 4.057664

smoothSolver:  Solving for omega, Initial residual = 0.0004233724736, Final residual = 6.483884521e-09, No Iterations 6
smoothSolver:  Solving for k, Initial residual = 0.0008657965129, Final residual = 8.034028879e-09, No Iterations 9
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935806296 -1.206282135e-06 1.000053919
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064353529 -5.78019834e-12 1.000795038
Phase-sum volume fraction, min, max = 1.000015982 0.9682864229 1.047777234
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935810693 -9.551659934e-07 1.000053951
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064349129 -5.779948852e-12 1.000795241
Phase-sum volume fraction, min, max = 1.000015982 0.9682452577 1.047778703
MULES: Solving for alphawater
water volume fraction, min, max = 0.193581509 -7.133944747e-07 1.000053984
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064344729 -5.779697411e-12 1.000795445
Phase-sum volume fraction, min, max = 1.000015982 0.9682044639 1.04778016
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935819488 -4.803904539e-07 1.000054016
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064340329 -5.779445978e-12 1.00079565
Phase-sum volume fraction, min, max = 1.000015982 0.9681640235 1.047781602
alphawater phase volume in zone AB = 995.5642595 ml
alphaair phase volume in zone AB = 4147.374824 ml
alphawater phase volume in zone pipe30D = 574.3235992 ml
alphaair phase volume in zone pipe30D = 2511.472611 ml
GAMG:  Solving for p, Initial residual = 0.006124217476, Final residual = 3.692759527e-05, No Iterations 4
GAMG:  Solving for p, Initial residual = 0.0002267828647, Final residual = 2.12083789e-06, No Iterations 11
time step continuity errors : sum local = 1.436653067e-08, global = -1.385048028e-11, cumulative = -2.031026982e-05
GAMG:  Solving for p, Initial residual = 0.0003193666751, Final residual = 3.041920407e-06, No Iterations 9
GAMG:  Solving for p, Initial residual = 4.551235006e-05, Final residual = 4.080258665e-07, No Iterations 11
time step continuity errors : sum local = 2.764300915e-09, global = -1.142671761e-10, cumulative = -2.031038409e-05
GAMG:  Solving for p, Initial residual = 0.0004827085861, Final residual = 4.492423075e-06, No Iterations 36
GAMG:  Solving for p, Initial residual = 5.059787016e-05, Final residual = 9.289342781e-07, No Iterations 16
time step continuity errors : sum local = 6.295698326e-09, global = 6.69707674e-12, cumulative = -2.031037739e-05
ExecutionTime = 349381.87 s

Courant Number mean: 0.07088353367 max: 19054.01219
Interface Courant Number mean: 0.03293067193 max: 13.63270585
deltaT = 8.606139431e-10
--> FOAM Warning :
    From function Time::operator++()
    in file db/Time/Time.C at line 1039
    Increased the timePrecision from 7 to 10 to distinguish between timeNames at time 4.057663995
Time = 4.057663995

smoothSolver:  Solving for omega, Initial residual = 1.190823172e-08, Final residual = 3.336175019e-14, No Iterations 1
smoothSolver:  Solving for k, Initial residual = 2.334359121e-08, Final residual = 1.532873766e-13, No Iterations 1
--> FOAM Warning :
    From function Time::operator++()
    in file db/Time/Time.C at line 1039
    Increased the timePrecision from 10 to 11 to distinguish between timeNames at time 4.057663994
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935819488 -3.638118437e-07 1.000054016
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064340329 -1.805508425e-12 1.00079565
Phase-sum volume fraction, min, max = 1.000015982 0.9681640234 1.047781603
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935819488 -2.473435533e-07 1.000054016
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064340328 -1.51373208e-16 1.00079565
Phase-sum volume fraction, min, max = 1.000015982 0.9681640232 1.047781603
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935819488 -1.309847101e-07 1.000054016
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064340328 -1.510822602e-16 1.00079565
Phase-sum volume fraction, min, max = 1.000015982 0.9681640231 1.047781603
MULES: Solving for alphawater
water volume fraction, min, max = 0.1935819488 -1.473444722e-08 1.000054015
MULES: Solving for alphaair
air volume fraction, min, max = 0.8064340328 -1.510822596e-16 1.000795649
Phase-sum volume fraction, min, max = 1.000015982 0.9681640229 1.047781603
alphawater phase volume in zone AB = 995.5642597 ml
alphaair phase volume in zone AB = 4147.374824 ml
alphawater phase volume in zone pipe30D = 574.323599 ml
alphaair phase volume in zone pipe30D = 2511.472611 ml
GAMG:  Solving for p, Initial residual = 0.9990622067, Final residual = 0.009030592757, No Iterations 7
GAMG:  Solving for p, Initial residual = 0.01058062846, Final residual = 9.957216274e-05, No Iterations 31
time step continuity errors : sum local = 9.470303504e-13, global = 7.257069631e-17, cumulative = -2.031037739e-05
GAMG:  Solving for p, Initial residual = 0.07303887886, Final residual = 0.000665584782, No Iterations 14
GAMG:  Solving for p, Initial residual = 0.001270285605, Final residual = 1.229286986e-05, No Iterations 37
time step continuity errors : sum local = 1.023783463e-13, global = -4.891270899e-16, cumulative = -2.031037739e-05
GAMG:  Solving for p, Initial residual = 0.0004048321254, Final residual = 3.79214576e-06, No Iterations 14
GAMG:  Solving for p, Initial residual = 0.0001298401762, Final residual = 9.118084172e-07, No Iterations 15
time step continuity errors : sum local = 7.590509253e-15, global = -3.318214066e-16, cumulative = -2.031037739e-05
ExecutionTime = 349386.51 s

Courant Number mean: 1.057986249e-06 max: 0.08949085779
Interface Courant Number mean: 8.6273426e-07 max: 8.337475345e-05
deltaT = 1.032736731e-09
Time = 4.0576639961

Attached are an image of the mesh (inlet: annular flow), checkMesh output, initial residual, fvSchemes and fvSolution.

Any help is appreciated.

Thanks,

Francesco

linear.png

Inlet.jpg

fvSolution.txt

fvSchemes.txt

checkMesh.txt

[mohsen_ba65]

Hi Francesco,

I am using this solver too.
I have kind of the same issue. I get those spikes too.
After trying many things I ended up increasing the nCorrectors to 5.
Also I am using a very small time step, 2.5e-5, and the Co is about 0.1
If I double the time step or for Co number larger than 0.2 the simulation will be unstable.
Another problem that I have is that the simulations blow up after some many time steps and if I restart it form the latest time it continues without any problem.

Did you find any solution?

Thanks,
Mohsen

[Frenk_T]

Hi Mohsen,

In my case I reduced the maximum time step to 5e-6 which gives a maximum Co below 0.05 during the run. In this way the simulation was stable, however the increased computational effort did not allow me to study the problem further.
Are you sure that if you restart the simulation, it continues till you reach a steady state and it's not blowing up later?
If you are looking for a steady state I would give a try also to the PIMPLE algorithm, setting nOuterCorrectors higher than 1.

Kind regards,
Francesco

[mohsen_ba65]

Hi Frank,

Maybe I need to reword what I typed.
If I restart a simulation it continues for a few minutes (let's say 30 min) and then blows up. Then I can start from the time that it blow up and again it continues for a few minutes (another 30 mins) and again blows up!

I was not able to find any sources of problem. At the time that simulation blows up, yo can find some cells, nodes, with very high velocity.

I tried using PIMPLE and also using very low relaxation factors but it didn't help.

[mohsen_ba65]

I finally could solve the problem by increasing the number of non orthogonal corrector.
Set the nNonOrthogonalCorrectors to a number greater than 1. It should help alot.

Cheers

[chpjz0391]

Hi Francesco,
I think you can find the answer from this post. http://www.cfd-online.com/Forums/blo...hm-part-i.html
Regards,
Ping