has anybody solved non-linear convection-diffusion (convection dominated case) equations using Fluent? We have a non-standard problem and would like to slove it using UDFs, has anybody attempted this before? There is little difficulty in setting up the equations but boundary conditions for UDF based equations seem to conflict with standard |Fluent terms, can you help with this?

I've solved these type of problems using USER DEFINED SCALARS (UDS). My experience has been mainly using UDS to model scalars without convection - actually I model solid phases that are generally stationary.

I have modelled them with convection using a user routine for the convective flux, but I found coding this very tricky. While I have got it to work, I still don't completely understand what FLUENT is doing. I think part of my difficulty was I used my own convective flux and continuity for the second phase.

If you are talking about modelling scalar transport which is convected with the existing flow field, then I think it will be much easier, since you wont have to code the convective fluxes.

While Fluent provides standard bc for the scalars, its possible to completely override these by applying your own flux formulation at the boundary. So you can do whatever you want - it just takes a quite a while to work it out, though the documentation has improved a lot with Fluent 6.

Greg

about this problem, I do not know your boundary condition. However, I think you can modify the standard boundary condition offered by Fluent. I have succeeded in doing it by UDF.