[emilhelgren]

When modeling isothermal phase change ( T_solidus = T_liquidus) i expected to have a non-existent or at least very thin mushy zone, since there is no diffusion of latent heat*. When i observed a zone with a significant thickness (low_const.png) i thought about what might cause this. I came up with the explanation that because the solid phase is simulated as a fluid with a very big momentum sink**, there is still some little bit of convection of latent heat, since the velocity is not completely damped while
0 < Latent Heat content < Latent Heat of Fusion
And this would create the zone seen on the image. From this reasoning, increasing the mushy zone constant (which determines the steep-ness of the damping in the momentum sink term) would make the zone thinner, since the velocity is damped quicker in response to freezing (/ ramped up quicker in response to melting), but this is not what happened when i increased it from 10^5 to 10^8 (which is the limit of the recommended span), as can be seen on (high_const.png).

Does anybody have an idea what might cause this? Do you agree with the reasoning? Why is there a mushy zone still, and how to get rid of it(since there is no zone in the melting of water in reality)?


*As can be seen in the formulas used by Voller and Prakash in this paper:
https://www.sciencedirect.com/scienc...17931087901529
Which is the source provided in the ANSYS 18 theory guide for the case of T_solidus = T_liquidus in chapter 18.4 (Energy Equation)

**(A_mush * vel * (1 - LF)^2/(LF^3 + 0.00001))