[hitchenxue]

Hi everyone, I am now using Fluent to model the Melting process and Considering the Buoyancy（natural convection）. A problem comes to me in the PCM properties definition.

Usually, the thermal properties are given as two, one for the solid phase of PCM and the other for fluid phase.

If I use the Boussinesq Model in FLUENT and can set the Operating Temperature ,Thermal Expansion Coefficient, and Select boussinesq in the drop-down list for Density, here however only can enter a constant value for density. I think this one is the desity of fluid phase of PCM.

How to consider the solid phase's density? Thanks all.

[LuckyTran]

The boussinesq assumption is a constant density field. If you want variable density then don't use boussinesq.


Buoyancy doesn't affect the pure solid region. The buoyancy is the pure fluid region is as if there was not any solidifcation/melting occurring. The mushy zone is treated as a porous medium where the density is already weighted based on the liquid fraction. But this doesn't affect the Boussinesq buoyancy force.

[hitchenxue]

Hi, Lucky Tran.
Thanks for your reply.
I have seen some papers, the two densities were given and the authors said they used the boussinesq assumption. This confused me.
If choose the boussinesq assumption, only one constan density and the Thermal Expansion Coefficient can be input.
Someone gives me advice that using the piecewise-linear definition（Density as a Profile Function of Temperature） to set the two destinies instead of boussinesq assumption. is it right ?
Thanks again.

[LuckyTran]

There is density for the flow part of the problem (i.e. rho*u everywhere) and density in the gravity force term (the rho*g*z) part. The point of Boussinesq is to assume the rho is constant in the rho*u part and simplify the rho*g*z and make it a simple function of temperature using the thermal expansion coefficient. The benefit of doing it this way, is you can use an incompressible solver and that simplifies a lot of things.

Specifying a temperature dependent density is much better than a Boussinesq approximation. The disadvantage is now there is stronger coupling between momentum and energy equation and makes it a bit trickier to solve.

With today's computational capabilities, there's few reasons to use Boussinesq except for learning purposes.

[nguyentbtuyen]

Originally Posted by hitchenxue

Hi everyone, I am now using Fluent to model the Melting process and Considering the Buoyancy（natural convection）. A problem comes to me in the PCM properties definition.

Usually, the thermal properties are given as two, one for the solid phase of PCM and the other for fluid phase.

If I use the Boussinesq Model in FLUENT and can set the Operating Temperature ,Thermal Expansion Coefficient, and Select boussinesq in the drop-down list for Density, here however only can enter a constant value for density. I think this one is the desity of fluid phase of PCM.

How to consider the solid phase's density? Thanks all.

Hi Xue Chen,
I wonder if you have solved your problem yet, I hope yes.

Please advise some clues as I am facing the same problem.
Thank you.