CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Multiple species and UDS source terms (

Ale September 4, 2002 06:05

Multiple species and UDS source terms
Hi to all,

I wanted to check the right way to write a source term in fluent, so I solved a simple problem of laminar non-premixed combustion, using the Multiple Species method and ticking off only the option Volumetric Reaction.

In the same simulation, I attempted to calculate the profiles of the species by UDF.

The source term I have written for methane is the following:

DEFINE_SOURCE(p_source_ch4, c, t, dS, eqn) {

double source_ch4;


return source_ch4; }

C_UDSI(c,t,0) is methane mass fraction; C_UDSI(c,t,1) is oxygen mass fraction. Other source terms for other species are similar. The kinetic expression is the one used by Fluent for the methane-air mixture.

The problem is that contours of UDS are different from the corresponding mass fraction contours calculated by Fluent.

I used for diffusivity the constant-dilute-appx (2.88e-5) and I set to the same value for UDS Diffusivity.

Has anyone idea of the reason of this discrepancy? Where's my mistake?

Thank you very much!


Greg Perkins September 8, 2002 21:54

Re: Multiple species and UDS source terms
If you're trying to replicate a chemical reaction you need to have sources for all the species participating in the reaction - some will be positive (products) and some negative (reactants).

Thus for ch4 + 2O2 -> co2 + 2h2o

you'll need sources for each of the four species to get the right mass fractions!

Ps: Its possible to solve this yourself - but for laminar systems where chemical kinetics generate a stiff matrix, you may find solutions diverge unless you are careful with the implementation. Fluent's added a stiff chemistry solver in version 6


Ale September 9, 2002 03:45

Re: Multiple species and UDS source terms
Thank you for reply.

I have written of course a source term for each of the four species; I reported only the one of methane for the sake of brevity.

Anyway, I noticed such discrepancy in the results of converged simulations (I didn't have problems with convergence).

I made these simulations just to check the right way to write a source term in fluent (I chose a laminar sistem just to avoid mixing effects).

Anyway, thank you very much for your response.


All times are GMT -4. The time now is 04:44.