Hello,

I am modeling a solidification problem. Initially a tank is filled with water, and there is a cold plate on the top to solidify the water (due to radiation). I can model this without too much problem, however from experiment I know the solid later created on top of water has a thickness of order 10 micron. VOF and Eulerian interface (water/ice) is not appropriate since I will have cells that are partially filled with water and ice, so no matter how small the mesh can get (before round off errors mess it up), the ice thickness is not going to be correct. So I think the only way to take care this problem is to be able to track the fluid surface using a Lagrangian tracking method, i.e. the mesh on the interface will be refined so that no cells cross the interface. Now here is the questions. Is it possible to do it in CFX? I have also thought about adding my own user function to do this, it will be complicated but not impossible.

Any comments would be much appreciated.

Kam

Hi,

What are you trying to look at with the simulation? Do you need CHT in the ice? Do you need to model the ice at all, can you use a moving top boundary with a mass sink?

Regards, Glenn Horrocks

Glenn, My ultimate goal is to control the thickness of the ice. At this time I am using a highly viscous fluid as ice (mu = 100 Pa.s). So essentially two fluids in the model but one (ice) doesn't move because of high viscosity. I find the bulk temperature in a cell and when it is below 0[C] I change the viscosity in that cell to 100 Pa.s (convert it to ice). I don't care about the temprature gradient inside the ice layer at this time. What I care is when I convert water to ice in a cell, I like that cell to be full of water instead of partially water/ice. Is it more clear?

What do you mean by "moving top boundary with a mass sink"? If I understand it correctly, the cells that I am going to add the mass sink will still be partially water/ice, so what would be different from my current approach?

Thanks for your time. Kam

Hi,

I was suggesting you only model the liquid part and define the top boundary as the ice level. The top boundary moves up and down according to the bottom level of the ice.

Whether this is appropriate or not will depend on exactly what you are trying to get out of the analysis.

Glenn Horrocks