[doubledang]

Hi Guys,

I modified a case from "mixerVessel2D (air+oil+water+mercury four phases)" tutorial under reactingmultiphaseEulerFoam, I dropped mercury while keeping oil+water+air as a test (three phase in a 3D stirred tank, MFR is used to model the impeller rotation).

The pureIsothermalPhaseModel was used to turn off the energy equation (induced temperature variations, especially for air).

I used OpenFOAM 4.0+reactingMultiphaseEulerFoam+k-e turbulence model for each phase.

I think the mesh is okay (see attached log.checkmesh):

My situation is that the simulation is not running, crashed (See the attached log.reactingMultiphaseEulerFoam) with some symptoms:

1. the time step become smaller and smaller (1e-4 to 1e-8 s);
2. minus volume fraction for air:
at t= 0.1 s: air volume fraction, min, max = 0.000552653 -0.00194183 0.053265
3. bounding epsilon+ bounding k:
smoothSolver: Solving for epsilon.water, Initial residual = 1, Final residual = 2.74972e-10, No Iterations 3
bounding epsilon.water, min: -3.59429e+18 max: 3.72478e+21 average: 3.66396e+16
smoothSolver: Solving for k.water, Initial residual = 1, Final residual = 9.07657e-13, No Iterations 1
bounding k.water, min: -4.3176e+08 max: 1.16267e+11 average: 1.68291e+06
smoothSolver: Solving for epsilon.oil, Initial residual = 1, Final residual = 6.19214e-10, No Iterations 5
bounding epsilon.oil, min: -8.1113e+24 max: 5.66417e+23 average: -7.27513e+19
smoothSolver: Solving for k.oil, Initial residual = 1, Final residual = 2.23996e-06, No Iterations 10
bounding k.oil, min: -4.54615e+21 max: 5.69049e+16 average: -3.93674e+16

I tired to change boundary conditions, drag, some settings in fvSchemes, and reduce tolerance for p_rgh, etc, but it turns out in vain (all attempts crashed with the same symptoms as listed above).

In the attached case, the initial volume fraction of air is 0.001, which is way a bit low, however, I also tested air volume fraction equal to 0.1, the simulation crashed in the same way.


By the way, I also noticed that the it seems the initial volume fraction of air could not be set to "0", if the initial volume fraction of air is set to "0", the reactingMultiPhaseEulerFoam solver simply not start even.

I have a little feeling that the problem might be possibly originated from the boundary conditions (but I also tried hard to make changes to the boundary conditions).

Could you guys take a look, and see if my boundary settings are correct?

Besides, do you guys have some suggestions as regard to the settings in fvSchemes and fvSolution files?

Any suggestions and comments are warmly welcome!

Thanks a million!

Best regards,

[ehsan_am86]

Dear Dang,

I have faced the exact problem which you mentioned above.

Did you manage to solve the problem?

Originally Posted by doubledang

Hi Guys,

I modified a case from "mixerVessel2D (air+oil+water+mercury four phases)" tutorial under reactingmultiphaseEulerFoam, I dropped mercury while keeping oil+water+air as a test (three phase in a 3D stirred tank, MFR is used to model the impeller rotation).

The pureIsothermalPhaseModel was used to turn off the energy equation (induced temperature variations, especially for air).

I used OpenFOAM 4.0+reactingMultiphaseEulerFoam+k-e turbulence model for each phase.

I think the mesh is okay (see attached log.checkmesh):

My situation is that the simulation is not running, crashed (See the attached log.reactingMultiphaseEulerFoam) with some symptoms:

1. the time step become smaller and smaller (1e-4 to 1e-8 s);
2. minus volume fraction for air:
at t= 0.1 s: air volume fraction, min, max = 0.000552653 -0.00194183 0.053265
3. bounding epsilon+ bounding k:
smoothSolver: Solving for epsilon.water, Initial residual = 1, Final residual = 2.74972e-10, No Iterations 3
bounding epsilon.water, min: -3.59429e+18 max: 3.72478e+21 average: 3.66396e+16
smoothSolver: Solving for k.water, Initial residual = 1, Final residual = 9.07657e-13, No Iterations 1
bounding k.water, min: -4.3176e+08 max: 1.16267e+11 average: 1.68291e+06
smoothSolver: Solving for epsilon.oil, Initial residual = 1, Final residual = 6.19214e-10, No Iterations 5
bounding epsilon.oil, min: -8.1113e+24 max: 5.66417e+23 average: -7.27513e+19
smoothSolver: Solving for k.oil, Initial residual = 1, Final residual = 2.23996e-06, No Iterations 10
bounding k.oil, min: -4.54615e+21 max: 5.69049e+16 average: -3.93674e+16

I tired to change boundary conditions, drag, some settings in fvSchemes, and reduce tolerance for p_rgh, etc, but it turns out in vain (all attempts crashed with the same symptoms as listed above).

In the attached case, the initial volume fraction of air is 0.001, which is way a bit low, however, I also tested air volume fraction equal to 0.1, the simulation crashed in the same way.


By the way, I also noticed that the it seems the initial volume fraction of air could not be set to "0", if the initial volume fraction of air is set to "0", the reactingMultiPhaseEulerFoam solver simply not start even.

I have a little feeling that the problem might be possibly originated from the boundary conditions (but I also tried hard to make changes to the boundary conditions).

Could you guys take a look, and see if my boundary settings are correct?

Besides, do you guys have some suggestions as regard to the settings in fvSchemes and fvSolution files?

Any suggestions and comments are warmly welcome!

Thanks a million!

Best regards,