[aditya_sin]

Hi all,

I am not able to run twoPhaseEulerFoam with turbulence. However without turbulence it is working fine. I am running a bubble column case with freeboard at the top of the water column.

Can anyone help?

[raagh77]

Hi..

even I am trying the same case with free surface. I was able to run with turbulence being "ON" with some changes
. including relaxation factors.
. increasing number of correctors PISO to 8 in fvSolution

The simulation was quite stable but time accurate predictions was not that good..(I mean I got the a bit higher time period what the experiments says)

and also this might (Alberto suggestions) help you
http://www.cfd-online.com/Forums/ope...ditions-2.html

[aditya_sin]

Hi Raghavendra,

Thanks for your suggestions.

I tried increasing the PISO correctors to 10 and reducing the timestep by two order of magnitude but still the simulation is crashing. I also tried underrelaxing but that does not work too.

Can you tell me what were the turbulence boundary conditions you were using?

Also I would appreciate any other suggestions you may have.

[raagh77]

what is your inlet velocity?
based on the inlet velocity calculate k and epsilon ..boundary conditions are epsilon and k



dimensions [0 2 -3 0 0 0 0];

internalField uniform 1e-06;

boundaryField
{
wall
{
type zeroGradient;
}
frontwall
{
type zeroGradient;
}
backwall
{
type zeroGradient;
}
sidewalls
{
type zeroGradient;
}
outlet
{
type zeroGradient;
}
inlet
{
type fixedValue;
value uniform 1e-6;
}
}


and for K

dimensions [0 2 -2 0 0 0 0];

internalField uniform 1e-06;

boundaryField
{
wall
{
type zeroGradient;
}
frontwall
{
type zeroGradient;
}
backwall
{
type zeroGradient;
}
sidewalls
{
type zeroGradient;
}
outlet
{
type zeroGradient;
}
inlet
{
type fixedValue;
value uniform 1e-06;
}
}

[raagh77]

As suggested by my supervisors I tried running the simulation with turbulence being on
from the flow field developed by laminar flow..
Probably this might work.

In my case the simulation was quite stable and results were also quite satisfactory only for velocity profiles (as it is in the experiments)
but the oscillation frequency of the plume developed didn't match.

[irinsun]

in the case turbulence on why shouldn't we use
kqRwallfunction and epsilonWallFunction as the boundary condition on the wall?