Hi, I'm actually an undergrade student doing my final year project in modeling seawater between membrane channels using computational fluid dynamics. The deadline for myy project submisssion is close and i'm still stuck with some problems. I'm quite new to CFD and have very few knowledge in UDF.

1)Since the membranes are permeable walls, I set the wall as velocity inlet in the boundary condition of Fluent and specify the Y-Velocity component to be equal to 0.00015 m/s and the X-velocity component to be 0 m/s. But unfortunately when do the simulation, the velocity profile at the membrane is slanted instead of being perpendicular to the wall.

2)In setting the transport species for seawater, do i need to set the mixture as only on species i.e NaCl or do i need to specify two species i.e NaCl and H2O? It is actuallly a species transport without reactions.

I have been searching for all topics related to my project but most of the previous messages have been left unanswered. I hope that my post will help those who are doing similar simulation to come forward and discuss about it. Hoping to hear from you very soon and thanking you all beforehand.

This not a very clear description of the problem, so I will do my best to make suggestions? You say that you have permeable walls that you have set as velocity inlets. Dose this mean that there is another inlet elsewhere in the simulation? If this other inlet is perpendicular to the permeable wall inlets, the resultant magnitude might be causing the slanted profile you speak of. Also, if you are using a velocity inlet to represent a wall you will not be using any wall function. see 10.8 in the manual. Also, what is the permeable wall letting through, is it the sea water or NaCl ions? If it is ions selective, you do not need to worry too much about having a velocity inlet because there will not be a large mass source at the permeable wall.

As for species modelling, if there is an even distribution of NaCl in the inlet, and there is not reaction (mass sink or source) for the species, I do not think you need to model it, as the concentration will be the same through the domain.

Hope some of this helps. Good luck.

Hi Andrew.. Thanx alot for your help.. really appreciated.. I think i forgot to mention before that actually there is an hoirizontal inlet flow(laminar, steady, incompressible) moving in between the two parallel membranes.

1)The feed is a diluted aqueous solution with constant density, viscosity and containing a solute with constant diffusivity. The permeat(solution) should be perpendicular to the membranes. I checked the data for the velocity at the membrane wall and it obeyed the boundaryy conditions i specified before(U=0 m/s ; V=0.00015 m/s.

2) Hence a the solute concentration will increase at the membrane sides compared to the middle of the channel. I was thinking of modeling the concentration profile in the channel using transport species. As u said, I got a uniform concentration through out the domain. How do i proceed to model the conceentration profile?

Hope to hear from u very soon.. Thanx alot again for yr time..

I have done simialr stuff in the past. I am guessing that the membrane is changing the concentration of NaCl in the inlet solution. For this reason I wouldn't specify the membrane as a veloicty inlet. If the increased concentration has little effect on the fluid properties (vicosity and density) I would consider, as I did, using a User Defined Scalar to model the concentraiton. This can be found in the manual.

Ok Thanx alot.. then how to proceed to model the permeate flux from the membrane?.. I dont think porous media is a good solution to it since the membrane is permeable and not porous... Thanx

Hi Andrew, After reading through the manual i manage to set up the contours of the mass flow rate and the concentration in the channel. But however I'm still having some doubts..

> Do i need to set the source terms in the fluid panel even if got only one species with no reaction?

> I will be grateful if u could tell me which code to use to input the equation which relates the concentration at the wall surface with respect to the length of the channel. I'm quite new to udf and I'm kinda blur now..

Thank you before hand

The use of a forum presents certain problems for answering questions, this is really the lack of imidiate feedback you get. If you are going to ask questions they have to be specific, and bear in mind that I don't know exactly what your problem is that you are trying to simulate. In answer to your first question, do I need a source term in the fluid panel, if by this you mean a species source term, if there is no volumetric reaction you do not need a source term, unless there is some additional species source that you haven't mentioned. The code to use to input the equation which relates the concentration can be any number of things. If you mean the surface reaction model this can be found in manual at 13.2. You say that your project deadline is soon, and if you are unfamiliar with UDF's it probably isn't worth starting to learn. If you are going to learn to use UDF's and you wish to write a code for a surface reaction model, this can be found in the UDF manual in 4.3.9. My general adivice would be to work out what you are trying to simulate and the equations you want to solve before commencing any computational work.