Learning Forum

[Jack5864]

Hi there,nnI am building a model of ions and ion-exchange membranes in Fluent, using a UDS for the species, and hooking UDFs in order to impose constraints and adjust the transport equations (it gave me much more control than using the standard species transport model).n I am currently modelling the membranes as a porous zone with additional constraints (in the form of fixed value UDFs) and I need to impose an equilibrium relationship at the membrane-solution interface. I was wondering if there is a way such as using a step change or Heaviside function to impose this relationship in a UDF, or if there is a more direct way of imposing it. I've searched through the documentation and haven't been able to find anything.Without the equilibrium, the gradients at the boundary are very large, and fluent runs into trouble handling it.Thanks for your help.nJackn

[Rob]

For membranes we typically use a source+sink approach to move species from one side of the wall to the other. It's a question that keeps coming up so have a look in here. n

[Jack5864]

Thanks for your reply. I have looked at other membrane threads, and unfortunately, this approach does not work for my situation, since they're ion-exchange membranes, and there is a concentration equilibrium at the membrane-solution interface, for which I have equilibrium relations. I've tried implementing these using DEFFINE_ADJUST and DEFFINE_PROFILE macros, to specify a relationship on either side of the boundary, but like I said the steep gradients aren't handled well. Do you know of any way to implement something where I can have two concentration instances at a boundary, such as with vapour liquid equilibria? Or alternatively, make fluent ignore the transport equation at the boundary so I can specify the concentrations either side manually?n

[Rob]

I'm not sure I understand why you can't sink+source material. The concentration gradient on each side of the wall is based on the flow in that region. The concentration gradient on either side of the wall is down to the UDF: the solver shouldn't care as there's a wall in the way. n

[Jack5864]

Because unlike the other models discussed in this forum, I am modelling the membranes as a separate domain, not just a wall. If there is a way to do this using sink and source for a uds, then could you possibly elaborate or link to a discussion on it, since I haven't been able to find an approach which works for my situation?n

[Rob]

Why do you need to model the thickness of the membrane? If you need to transfer from the free stream to the membrane surface to the core to the surface & finally the other side then you may still want to use walls & a cell zone (possibly solid) in the gap. The same principle applies but you'll need to calculate and keep track of the bit in the middle and any accumulation/loss of the scalar. n

[Jack5864]

My flow channels are very thin, so the membranes make up a non-negligible volume (about 20% of the total volume), so the thickness must be modeled. Thank you, I'll give that a go. I am wondering, would a wall and a coupled boundary condition work? There isn't much documentation on how a coupled boundary works for a UDS, but on a forum, I read that the flux is matched on either side and the values are allowed to be different. Is this correct?n

[Rob]

Coupled is for energy, is heat passes through the material. Not sure about scalars as they tend to be attached to the flow so shouldn't pass through solids unless told to. n

[Jack5864]

Ok, the flux of my species scalars are in part driven by an electric field (perpendicular to the flow) as well as through diffusion and flow, so it should be able to go through the walls. Do you know if a coupled boundary would be appropriate? I'm not too sure how they work n