[mariam.sara]

Dear all,
I am attempting to simulate a double diffusion (Heat trasnfer+Mass transfer) problem in a square enclosure in which there is a high concentration at the left wall (a thin film where the evaporation occurs) and low concentration at the right wall (cold wall where the condensation occurs) as illustrated in picture. May somebody guides me if I it is possible to solve this problem in CFX and which model I can use at the simulations.

Mariam

[ghorrocks]

Yes, you can do this in CFX.

The heat transfer is modelled using the normal heat transfer methods in CFX.

The mass transfer can be modelled by a few approaches:
* If the mass transfer does not affect material properties then you can use an additional variable approach.
* If the mass transfer affects material properties then use a multicomponent mixture approach.

In both of these mass transfer approaches you can apply similar boundary conditions to the heat model - dirichlet, neumann etc.

[mariam.sara]

Thanks for the valuable answer. May you tell me where I can find the multicomponent mixture approach? Is there any tutorial assist me at the problem setup?

[Svetlana]

https://www.sciencedirect.com/scienc...29549317304478 for example ?


https://www.youtube.com/watch?v=BfdbkNTkElo

[mariam.sara]

Many thanks guys for the answer. Can I make the flow as Boussinesq? i.e. I want to model Boussinesq flow with different fluid concentrations at the cavity side walls? knowing that the cavity is filled with humid air?

Mariam

[ghorrocks]

Boussinesq was a wide ranging researcher. Which Boussinesq approximation are you talking about? Buoyancy linearly varies with temperature, surface waves or eddy viscosity in turbulent flows? (https://en.wikipedia.org/wiki/Boussinesq_approximation)

[mariam.sara]

Originally Posted by ghorrocks

Boussinesq was a wide ranging researcher. Which Boussinesq approximation are you talking about? Buoyancy linearly varies with temperature, surface waves or eddy viscosity in turbulent flows? (https://en.wikipedia.org/wiki/Boussinesq_approximation)

The buoyancy caused by temperature difference only.

[ghorrocks]

OK, in that case it is easy. Use a thermal energy model with a material which has a constant density but a thermal expansivity defined. That is all you need for that buoyancy model.

[mariam.sara]

Originally Posted by ghorrocks

OK, in that case it is easy. Use a thermal energy model with a material which has a constant density but a thermal expansivity defined. That is all you need for that buoyancy model.

OK what about mass transfer among the boundaries (as my illustration on my first post) can I couple it with the Boyonancy effect?

[ghorrocks]

As I said in my previous post, you can do the mass transfer by either an additional variable or multicomponent mixture depending on how the mass transfer affects material properties.

[Opaque]

Originally Posted by mariam.sara

OK what about mass transfer among the boundaries (as my illustration on my first post) can I couple it with the Boyonancy effect?

You can do it, but you need to use the multicomponent approach. That is,

define a variable composition mixture material
assign the material as a fluid to your domain
activate the buoyancy model, set gravity
activate the model for each of the mixture components
set appropriate boundary conditions

[mariam.sara]

I attempting to follow the steps you mentioned but I face a problem in setting the mass fraction of H2O vapor at the side boundaries of my geometry because all boundaries in it are walls? and in wall BC I think adding mass fraction is unavailable?

Originally Posted by Opaque

You can do it, but you need to use the multicomponent approach. That is,

define a variable composition mixture material
assign the material as a fluid to your domain
activate the buoyancy model, set gravity
activate the model for each of the mixture components
set appropriate boundary conditions

[Svetlana]

Can you specify it in the initial conditions? Then it will be conserved during the simulation.

[mariam.sara]

But if I defined it at the initial conditions it will take it to the whole domain? I want a specified value at the right and left walls?

Originally Posted by Светлана

Can you specify it in the initial conditions? Then it will be conserved during the simulation.

[Svetlana]

Initial conditions are for the whole domain, not only on the boundaries.

[mariam.sara]

But I do not want set it to the whole domain? I want to specify it just at the boundaries? Each BC has a different concentration.

Originally Posted by Светлана

Initial conditions are for the whole domain, not only on the boundaries.

[Svetlana]

Sorry, I did not realize that each BC has a different concentration.

Enable 'Species transport' in your models. Then your wall boundaries will have a 'Species' tab, where instead of 'zero diffusive flux' you can specify the mass fraction of the species. (Please note that someone may need to check this is correct, I have not modeled this kind of physics before.)

[mariam.sara]

Sorry I did not know how to enable the "Species transport" I went through default domain> fluid models it not exist? May you send me illustrated picture?

Originally Posted by Светлана

Sorry, I did not realize that each BC has a different concentration.

Enable 'Species transport' in your models. Then your wall boundaries will have a 'Species' tab, where instead of 'zero diffusive flux' you can specify the mass fraction of the species. (Please note that someone may need to check this is correct, I have not modeled this kind of physics before.)