Hi all ,

Does anybody have any experience regarding the implementation of surface reactions in Fluent? The problem I have is that I'm looking at a 23 step reaction mechanism, each step of which involves surface species as both reactants and products. According to the Fluent manual, the standard code does not allow surface species to be reactants. Has anybody found a way around this, perhaps using user-specified subroutines, and how easy/difficult is this?

Thanks for your time.

Richard Carroni

Richard

I'm currently attempting to develop something similar for pack-bed combustion and gasification. In this case I have a hetrogeneous reaction model from which I calculate source term UDFs for the gas phase species and energy equations.

I'm still trying to get it all working at this stage.

I think you'd have to do something like this using UDFs because of Fluent's restrictions as you mentioned.

The complexity of the UDFs will probably depend on whether you need to calculate the solid consumption over time or just a steady state mass transfer at the surface. In the former case, its probably easiest to compute a recession velocity of the surface using the later technique and use this to compute the new position of the surface at the next time-step (thus avoiding a fully-time dependent formulation). This info can then be used to generate conditions for a new case file, which could be automatically generated using scripts in Gambit & Fluent.

Hope this helps

Regards

Greg

Greg

Thanks for your response, at least I now know the problem is solveable. This begs for an answer to the next question: how long does it take to write all these UDF's, and how much is the problem compounded if one has never actually written UDF's before? As a matter of interest, how long did it take you to get to the current stage?

Cheers,

Richard

I think writing basic UDFs is quite easy if you're familiar with C and can take 1-2 days. Access to the Fluent UDF manual is a must for this.

However, when it comes to more complicated UDFs where you need to access variables/functions that aren't explained in the UDF manual it takes much longer and gets much harder. Of course, testing that your UDFs actually perform and validating the results also takes longer with more complex routines.

I've being developing my code for several months. I'm trying to include the effects of a solid phase throughout the domain which can react so I am actually solving additional transport equations (solid phase continuity & energy) using the user defined scalar capability as well.

Hopefully you can get a basic reaction module up and running a lot sooner.

Send me an email if you like once you get started.

Regards

Greg