Reference density in buoyant multiphase-flow?
 

Hi everybody,

I am simulating a buoyant two-phase flow (air, water) through a packed column in CFX 5.7. Both phases are continuous. Which value has the reference density?

In the manual I read something about an average-density in the domain. One of my friends told me to set the density of the heavier fluid as reference density.

I hope that somebody can help me on this question.

Thanks in advance and best regards, Johannes

Re: Reference density in buoyant multiphase-flow?
 

Hi, Johannes

set the buoyancy reference density to that of the contiuous phase.

I think that in your case it is the air.

regards

Re: Reference density in buoyant multiphase-flow?
 

I my case both fluids are continuous-phases, as written above.

Regards, Johannes

Re: Reference density in buoyant multiphase-flow?
 

in your case, you can set the density of air as the buoyance reference. you can see the hydrostatic pressure distribution of water as it refer in the cfx manual BR Erica

Re: Reference density in buoyant multiphase-flow?
 

Ok, thanks for your help, Erica.

Johannes

Re: Reference density in buoyant multiphase-flow?
 

I read the cfx manual about the hydrostatic pressure gradient. There is written:

p_hydrostat = rho_ref * gravity * (r - r_ref)

If I use the density of air as the reference density I will not see any hydrostatic pressure gradient caused by the water column. I will just see the gradient caused by the air. Or am I wrong?

Regards, Johannes

Re: Reference density in buoyant multiphase-flow?
 

In your case, two phases are continuous one, you should set the light phase as reference one. It will cancel the buoyant force and hydrostatic pressure of the reference phase. In order to decrease the roundoff error,you should choose the lighter one as reference phase.

BR, Erica

Re: Reference density in buoyant multiphase-flow?
 

I have one more question: With air as reference phase, will I see any hydrostatic-pressure gradient caused by the water phase in the result-file? And if yes, in the pressure or in the absolut-pressure distribution?

Regards, Johannes

Re: Reference density in buoyant multiphase-flow?
 

Yes, you can it in the result file.It is pressure. You can see it in cfx-post.

BR, Erica

Re: Reference density in buoyant multiphase-flow?
 

Johannes,

Using the air density as the refernce density, you'll get body force in the air phase which is:

B = g * ( rho.air - rho.ref ) = 0

In the water phase, the body force will be:

B = g * ( rho.water - rho.ref )

Since rho.water is ~1000, and rho.ref = rho.air = ~1, this will give the hydrostatic force in the liquid phase. You should see the expected hydrostatic pressure. Note both phases share the same pressure.

I have had stability issues with this setting when bubbling a gas through a liquid tank and setting the liquid phase density as the reference density gave a more stable convergence (not sure why yet). In this case, the body force on the liquid (my continuous phase) is subtracted out, however it gives the correct flow solution. In this case, you can adjust the pressure in Post by adding the hydrostatic head back in to create a new variable and plot that instead of pressure.

Not sure what your continuous/continuous case will do.

Jeff