[cheng1988sjtu]

Hi All,

I'm running a twoPhaseEulerFoam case using 1DV ( just 1 grid in x and z direction, but in y direction I'm using 200 grids) and find something interesting:
The mass of particle phase is not conserved basically. After Initialization, the volume averaged fraction is 0.198, after 400s' running, the volume averaged fraction becomes about 0.244, apparently, the mass of dispersed phase is not conserved.

In the above case, I'm using OF-1.7.1, I've turned on the PackingLimter, and the alphaEx=0.62, I've set g0=0, and turned the kinetic theory on.

I guess my question is what do you think the problem is? Could somebody explain something about the packingLimiter?

Thanks a lot!

Zhen

[kaifu]

Originally Posted by cheng1988sjtu

Hi All,

I'm running a twoPhaseEulerFoam case using 1DV ( just 1 grid in x and z direction, but in y direction I'm using 200 grids) and find something interesting:
The mass of particle phase is not conserved basically. After Initialization, the volume averaged fraction is 0.198, after 400s' running, the volume averaged fraction becomes about 0.244, apparently, the mass of dispersed phase is not conserved.

Of course the void fraction is not conserved unless it goes into steady state.

Originally Posted by cheng1988sjtu

In the above case, I'm using OF-1.7.1, I've turned on the PackingLimter, and the alphaEx=0.62, I've set g0=0, and turned the kinetic theory on.

I guess my question is what do you think the problem is? Could somebody explain something about the packingLimiter?

Physically, the PackingLimter means nothing and we should not have PackingLimter. The value alpha should only updated by its transport eqn. Technically, since there might exist a serious problem when alpha goes too high, we may want to keep alpha roughly below a value (but cannot promise alpha<alphaMax strictly). Future model together with mass transport eqn should be developed in order to keep/promise a reasonable alpha.

[cheng1988sjtu]

Thanks, at first I thought it was due to the artificial redistribution of alpha in packingLimiter.H, however, investigation of the code just proved me wrong, there is no problem with the packingLimiter.H

Maybe that's because I'm just using 1 grid in x and z direction. since when I change it to 2D, there seems to be no such problem.

Zhen

Originally Posted by kaifu

Of course the void fraction is not conserved unless it goes into steady state.


Physically, the PackingLimter means nothing and we should not have PackingLimter. The value alpha should only updated by its transport eqn. Technically, since there might exist a serious problem when alpha goes too high, we may want to keep alpha roughly below a value (but cannot promise alpha<alphaMax strictly). Future model together with mass transport eqn should be developed in order to keep/promise a reasonable alpha.