Discussion Forum

Hello,

I have a question about the setup similar to the melting tin model. I have block with two layers, one on top of the other. The upper layer is granite, the lower layer is ice. On the upper surface I have a prescribed heat flux, in the middle I have a set temperature at the melting point of ice and at the lower surface I have a set temperature a few degrees below freezing. The model is setup exactly the same as the melting tin model and does work. Using the assumption that any phase change from ice to water allows the water to leave the system. However, I realized while watching the results, that by using the set middle temperature and the weak constraints which give me the Lagrangian which I use to calculate the mesh velocity, the ice cannot cool. So while I am able to send heat to the boundary and melt the ice, I cannot remove heat across the boundary and cool the ice.

I am wondering what other methods are possible in COMSOL to perform the same task, yet allow for removal of heat from below the melt interface. I have been reading through the forum but don't see anything similar to this problem. I was thinking maybe I'd have to calculate the heat flux at each node along the boundary, for each timestep and then feed that into the model. This, however, is so far beyond me (I'm still a bit of a newbie).

I'm just curious if anyone out there might have any suggestions as to how to do this. Any help would be appreciated.

Also, I should apologize for sending a private message to a forum guru. I realized after sending it that the whole point of the open forum is so that everyone can see the discussions and get help where they can find it. Hopefully this question will be answered and help others as well.

Cheers,

J