CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   variable transport from one domain to other (http://www.cfd-online.com/Forums/cfx/71119-variable-transport-one-domain-other.html)

cosine December 16, 2009 05:07

variable transport from one domain to other
 
Hello,
I have a problem with two domains (imagine one tube in a bigger one). I add a variable to calculate mass transfer to the first domain (inner tube). i want this variable to penetrate to other domain and to be calculated there.
What boundary condition is appropriate? This penetration is shear stress depended.
Thanks a lot

ghorrocks December 16, 2009 17:41

I have no idea what you are trying to do. Is the inner tube porous? Where do you want to calculate the mass transfer?

cosine December 16, 2009 18:27

Thanks for your answer!
Let me explain better my problem. I have two flows, one in the inner tube which is fluid and one in the outer which is porous domain. Interaction exists only at the common interface where penetration of a molecule exists.
Inner tube:
I define inlet boundary condition at one edge and outlet at the other. Also i add a variable to calculate the diffusion of a molecule concentration. Here is the problem. I want the common "interface" of the tube to be permeable and the molecule to be diffused to the porous domain. Also this diffusion is shear stress depended. So the mass transfer need to be calculated in both domains.
I have tried to solve it using two configurations but i think that the molecule is not diffused to the porous domain.
Also i have tried using an interface but then i think that i cannot use a shear stress depended diffusion.
I hope that you could understand now, otherwise i could post some photos
Thanks a lot again

ghorrocks December 17, 2009 06:02

I have no experience in this sort of modelling, but can't you define your concentration as a mass fraction with a different diffusivity defined in the fluid and porous regions? And hopefully you can put some sort of resistance at the interface and with some luck make it a function of local wall shear stress.

cosine December 28, 2009 16:23

Thanks for your answer.
I will try shortly.

kingjewel1 December 29, 2009 19:13

Diffusion can be modelled algebraically via the use of the CEL for the kinematic transport. I don't quite know how what you're looking for but how about using: kinematic diffusivity = ave(Shear Strain)@small pipe, or some such?

cosine December 29, 2009 19:39

Thanks,
I want to calculate the concentration in the porous domain, while this arrives from the fluid domain. Could i connect the results of a calculations between two different analysis systems?
I thought, i could use "flux in" at the porous domain using an expression which is depend on the wall shear stress as calculated in the fluid domain.
I hope you could understand now.:)

kingjewel1 December 30, 2009 06:58

So in fact it is the porosity of the domain (sort of osmosis) based on the stress you are interested in not specifically the kinematic molecular diffusion. I still think this is possible using CEL through porous loss though I haven't tried it.

cosine December 30, 2009 07:04

Yes, exactly. The interface is a biological membrane with specified properties which someone could model using Kedem-Katchalsky equations. The problem is that i cannot use the wall shear stress of one domain to the second one. When i did it, actually i use the wall shear stress of the porous domain.
I use the CEL to describe all necessary equations, but i don't know how to define one calculated variable from the fluid domain in the porous domain.


All times are GMT -4. The time now is 07:57.