Some of our new electronic designs involve lithium ion batteries where we plan on surrounding each cell of the battery with a wax-like, phase change material (PCM). Due to safety requirements, we need to accomodate a catastrophic failure of the lithium ion cells, where the resulting heat during the cell failure is being absorbed by the PCM. The challenge in modeling the PCM is how to handle the all important phase change, which occurs at a well defined temperature, and the accompanying high latent heat of fusion. My approach has been to use a specific heat that is temperature dependent. Using the Solidworks' Engineering Database, I set the material's specific heat to its published value, except that for a range of 1 degree C, beginning at its melting point, I set the specific heat equal to the latent heat of fusion. Although the latent heat of fusion has the units of energy/mass, dividing that value by 1 degree C gives me the correct units for the material specific heat, energy/(mass*temperature) and I can use the same numerical value from the published latent heat of fusion. Before tackling the real design, I found that with a simple test case I need to have a time step equal to 0.001 sec to properly capture the material's latent heat of fusion. I'm now working on the real design, but have experienced very long run times (hours and hours), because of the very small time step needed.
Any ideas or your own experience in modeling phase change materials? Is using the specific heat as a function of temperature a good idea for mimicing the latent heat of fusion? Any help is greatly appreciated.