Learning Forum

[prebenjs]

I am having a multiphase simulations where i want one phase to convert to the other (i.e. be consumed and produce the other phase) at a surface. The reaction is from gas to liquid. How do I approach this? Using mass transfer or reactions? Have anybody any experience in this?

[Rob]

Like boiling?

[prebenjs]

Quite similar to condensation. However, the rate at which the reaction happens cannot be determined by the condensation model. I though about using constant rate, and mass transfer mechanism, but I don't know how to "connect" this to be a surface reaction ? The best option would be to a chemical reaction there or to have a custom source term that makes this reaction happen when gas is present at the surface (but I can't find any literature on writing this as an UDF or implementing it another way). Any advice ?

[prebenjs]

Do anybody have any thoughts around this?

[Rob]

If the resulting material can move the wall films might make more sense? Otherwise you need a reaction scheme that has a surface species as a product.

[prebenjs]

When you're talking about a reaction scheme and surface species as a product, wouldn't this need to be related to heterogenenous reactions and implementation through UDF ? Since it's from gas to liquid. I thought about just implementing a site species but I couldn't get it to work. Furthermore, would the site species only acts as a precipitation of solids, or is it possible to get this to work in my case as well ?

[Rob]

If the liquid is moving then you need either multiphase or a wall film. If it's not moving then a surface reaction ought to work. The EWF has a source term option, so whilst it may not be easy you can write your own mass transfer model. Otherwise you'll need to resolve the phase that's created and the cell count may become excessive.

[prebenjs]

When you say multiphase what do you mean? I'm using VOF and that is a multiphase. Do you mean multifluid VOF ?
Yes but the only problem here is that I believe that i need to implement this as an UDF due to gas-liquid conversion ? Is this correct or could i use surface reactions in the GUI provided by Fluent?
The last point you make, is this related to creating a mass transfer mechanism through UDF ? And that the sources which are assigned to the cells where this reaction happens will become very mesh dependent and will not be feasible ?

[Rob]

With VOF you need to resolve the film thickness: can you afford the cell count to do this? Multifluid VOF is more forgiving, but will still require a fine mesh.
You can use surface reactions, or a multiphase model: it depends on what happens after the material is created. The EWF approach works on the wall facet but will use the fluid data from the near wall cell.

[prebenjs]

Thanks for the knowledge rob! I have just some clarification and questions:
When you say that surface reactions might be used, you explicitly refer to heterogeneous reactions defined in UDF ? Or is it possible to in the GUI ?
Okay, so after the mass transfer at the surface the water flows due to a pressure gradient. Thus I do not really not if a wall film would be the correct alternative ? Furthermore, this would implicate setup by the addition of particle diameter and drag that needs to be defined.
When you say I need to resolve the film thickness and that multifluid is more forgiving, what do you mean by this ? the thickness of the produced liquid would probably be adjacent cell(s) to the surface, that contains gas. which will build/coalescene a water area and flow in other areas (but will not be continously drained).

[Rob]

As I don't know the rates etc in your model I can't comment, the standard solver may work, or you may need UDFs.
If the film is "thin" then the wall film models are applicable. If the film becomes thick we can convert to the VOF model (care must be taken with this as you may then need to add mass to VOF from your model).
Re resolving the thin film, read the background theory on the VOF model, my comments about mesh resolution will then become clearer.