[pefi]

Hello everybody,

i want to simulate the melting of metal particles with the discrete phase model. Does anyone have an idea how to implement such a problem with UDF?

I figured out two options but none of them seems to be praticable:
- Treat the metal particle as an inert particle (Inert heating law) and add somehow fusion heat when melting temperature is reached. Problem: How to add fusion heat? to which fluent parameter? temperatureupdate makes no sense here...

- implementing a routine such as evaporating. problem: dpm evaporating is based on mass update (bc mass loss of the particle). metal melting has no mass loss: melted mass sticks to the particle

any ideas?

[pat-cfd]

I did this by following your first option (i.e. using inert particles). The way I modeled melting is by setting the heat capacity really high over a small temperature range starting at the melting temperature. This way there will be only a small temperature change for a relatively large amount of heat transfer (i.e. latent heat transfer), which is close to what happens during melting.

The way to determine how high the heat capacity should be is to plot a graph of heat capacity vs. temperature and ensure the area under the high heat capacity section of the plot is equal to the latent heat of melting. That way the particle has to absorb that amount of latent heat energy in order to overcome this small temperature change.

Regarding the small temperature gap, I initially had it be really small (0.1°C). The issue I found was that the particles seemed to "skip over" this. I then tried larger ones (i.e. 100°C), which worked much better. I think the reason for this "skipping" is that I had very high temperature gradients and particle velocities in my model, so it might not happen to you. Another thing to note is if you're using unsteady particle tracking, which is what I used, then decreasing the particle time step size will help prevent particles from "skipping" over the high heat capacity region.

Hope this helps