CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Melting of a pure metal - 1D (

Soheyl May 27, 2009 17:16

Melting of a pure metal - 1D

I want to model the 1D Stefan problem, for the melting of a pure metal (T_sol=T_liq).

1- What do you think is an appropriate mesh size, and time step size for this problem? (alpha, thermal diffusivity, is around 100) Also, do I need a discretization in the y-direction, although my problem is 1D? What is a logical size for the domain in y-direction?

2- For the material, even though my domain is initially all solid, I am using a fluid, because only in fluids, we can define melting/solidification parameters, such as T_sol, T_liq, and q_melt. Am I right?

3- My results always show a much slower process than the one expected from theory. Any suggestion why this happens?

4- Do I need to worry about the mushy zone constant?

Thank you

Ralf Schmidt May 28, 2009 02:58


I can answer question 1 partially:

As far as I know, Fluent can not model 1D! You have to take some assumption for the second axis. Maybe, you can have your domain 1 cell thick...


MASOUD June 11, 2009 21:08

melting process
Dear All,
I have an Electric Battery which gets warmer during the time with a specified heat generation rate. I have to reduce the battery temperature using a phase change material around the battery.

1.So I have to define boundary condition for the battery wall which separates battery from phase change material (phase change material is solid).
I want to know in the thermal tab for the wall boundary condition, which kind of thermal condition should I use? Convection? Temperature?...? None of the is reasonable!!!

2. Since I have a pure substance, what should I do for solidus and liquidus temperature? I have only a melting temperature.

I would sincerely appreciate if you could answer my question at your earliest possible.


All times are GMT -4. The time now is 07:54.