Melting Ice

[Epy]

I would observe the physical process and see what the approximate thickness is, if you can manage. If you are applying force on the copper block (either directly or by weight) in order to sink it into the ice, then the water layer will be much thinner on the bottom (in the path of motion) than the side layers, depending on the speed of the process. Guessing it's pretty fast if the copper block is getting that hot. If I were going to do the simulation I guess I'd do something very small on the bottom, like 1/32" and then maybe 1/8" on the sides. Make sure the mesh for this water layer is very fine; you'll probably want it to be at least 8-10 cells wide at the thinnest spot in order to get accurate results.

I'm not sure what you're doing with this process, but as far as having to shut off the heaters goes, I think you ought to use some dimmer switches to modulate the power input to the heaters. With the right duty cycle, they won't overheat and you can just leave the heaters on for the entire process. You may not be able to readily find the right kind of dimmer switch, so you could just make the circuit yourself with a resistor, variable capacitor, and SCR/TRIAC depending on if you're using AC or DC power.

Hope this helps.

[tetraeder]

Thank you. I will try it with the layer and i think that really could be the solution.

The experiment has been done to verify some small technical parts inside the copper block. The results are used for a melting probe. As a bachelor thesis i try to work out the melting process itself to apply this knowledge to this complex melting probe. I think this little testblock would be a good start.

[raback]

Hi

Just a quick note: as your problem is too low temperature rather than too high, the poor conductivity of water/ice will work for you in the right direction. Adding convection will improve the cooling and hence considering the extra effort you might try just to play with the conduction.

-Peter

[raback]

Hi

Just a quick note: as your problem is too low temperature rather than too high, the poor conductivity of water/ice will work for you in the right direction. Adding convection will improve the cooling and hence considering the extra effort you might try just to play with the conduction.

-Peter

[tetraeder]

Hi,

the idea with the layer was really good. But i have to do the simulation the melting process with the phase change (spatial 2), because its more realistc.

I started with a thin bar (length=1m) of ice, an initial temperature of -10°C and a temperature of 40°C at one side. I compared the results with an analytical approximation and it turned out, that the simulation gives the correct values!

I don't want to remesh. To emulate the melting process it might be possible to remove some heat per time from the boundary to get "new" unmelted ice. When i know the speed of the phase interface i could calculate the heat to remove. I hope you understand what i mean...


So my questions:
- is it possible to get the speed of the moving phase interface
- is there a way to remove heat from a boundary (maybe through convection)

[tetraeder]

Okay... i did a little sketch to explain my idea.