Learning Forum

[joshua.wong1196]

Hello,

I am trying to simulate frost formation onto a vertical surface with a DEFINE_MASS_TRANSFER UDF. The temperature of frost (modeled as ice reaches 5000K (see picture). The ambient air is 295K and the cold surface temperature is 77K. My suspicion is that there is an issue with how the model treats the sensible/latent heat. Anyone mind offering any advice on how to address the problem? Thank you.

 

[Rob]

What models are you using? Have you got the sign correct in the heat transfer part of the UDF?

[joshua.wong1196]

Hi
I'm using the Eulerian multiphase model. I only included the mass transfer in the UDF. Based on the manual, it says that the DEFINE_MASS_TRANSFER function calculates the energy function well. From what I understand that means that fluent calculates the heat exchange automatically?

[Rob]

Yes, it should do. Have you made sure you got the phase change energy correctly assigned to the phase materials? Ie the sublimation energy. 

[joshua.wong1196]

Hi,

Could you elaborate on what you mean by that?  I have not changed anything related to phase change energy in the model since the mass transfer UDF automatically calculates the energy change due to mass transfer. Hence, there's no separate UDF for energy change. Right now I'm assuming that the energy change is calculated based on the thermodynamic properties of water vapour and frost stored in Fluent's material database. 

Inside the Eulerian model, I can't find any field that allows me to specify that the mass transfer is phase change nor am I able to find any field to assign the phase change energy.

[Amine Ben Hadj Ali]

Please add a screenshot of the panel. Aslo check if you are limiting your mass transfer and providing proper linearization.

[joshua.wong1196]

Hi, 

I was just looking through the model again and I realize that I have not assigned a method for heat transfer coefficient. I've then used the ranz-marshall method of specifying the heat transfer coefficient and the results I obtain are more sensible. Can I clarify if this is the solution to the problem or was this just lucky and I am missing something else?  

[Amine Ben Hadj Ali]

Yes heat transfer here is of course helpful.

[Amine Ben Hadj Ali]

And linearizing and source term limiting will be helpful too. 

[joshua.wong1196]

Thank you very much for your help. I will give that a try.

Just one last question related to modelling desublimation: I am now trying to simulate frost formation on a vertical surface. One of the problems I have now is that instead of initially forming on the surface, the mass transfer occurs in "air" and as such most of it is just falling off at the outlet. The closest I've tried to model this is by enabling surface adhesion. Admittedly, my current approach is rather simplified as well. Where should I start looking from (Fluent/manual) if I wanted to model it in such a way that frost is formed on the surface first?

 

 

 

 

[Ceng Jing Ru Shi]

Hello, I am also doing cold surface frosting simulation, mainly using Eular model and combination of UDF, I have encountered some problems, can learn from your UDF

[Ceng Jing Ru Shi]

Can I have a look at your UDF, I am also doing related work, common learning progress