Hello,

I would like to seek a very simple clarification. I read in the literature for modelling solid-liquid phase transitions, latent heat has been added in the energy equation as L(d.phi./dt) , L: latent heat , phi.: phase field parameter). A relevant change was also reported in free energy functional

Now doubt is, when numerically solving this problem, d.phi/dt will be effective or contribute only when at previous time iteration there was some other phase (i.e. say phi=1) and now there second phase (phi=0). I hope I am clear in understanding till here.

Now this will be true if you don't have any velocity field because as per my understanding, say if the gravity is present and we do the case of liquid-vapor transition, then it may happen that at a point where at previous time iteration there was some phase (say phi=1, liquid) now by virtue of gravity at that grid we can now have a gas phase (phi=0).

Now at this particular grid point, the phase parameter will change by virtue of motion of fluid and not phase-change. But in energy equation, the term L(d.phi/dt) will contribute which is physically not correct.

Can some one correct me if I understood something incorrectly? Do we need to add latent heat in case of liquid-vapour transition.

Wouldn't it evolve based on the dependence of free-energy which will have form such as :

F= bulk free energy + Latent heat*(1-T/T0) + k/2 * grad(phi)^2

here T0 is the saturation temperature.

So won't based on this function and thus temperature, the phase field parameter will evolve thereby indicating which node corresponds to which phase?

Looking for some help.

Merci

Deewakar