I want to model a chemically reacting flow. The flow is laminar and involves 6 species in 4 reactions. The rate expressions I have at hand do, however, not comply with the modified Arrhenius expression applied in Fluent although "my" rate expressions only differ by a multiplication of an additional term (1-B).

r = k*T^n*exp(-Ea/RT)*(1-B)

Does anyonw know if it is possible to simply make an adjustment in UDF of the rate calculated by Fluent?

I have considdered to make the adjustment via the macro DEFINE_VR_RATE...

Any help?

Is (1-B)=const ? In this case you could include it in k.

RoM

No, B is a function of species concentrations and thermodynamic properties, but I have all these data and properties at hand and can easily write this into a UDF...

In this case a DEFINE_VR_RATE udf would be the best option.

RoM

I have tried this using an interpreted as well as compiled version and both fails to run. I recieve an error during the first iteration. Do you know if a turbulence model must be invoked in order for the UDF to work. Even the example IU have found in the UDF manual seems to fail when the flow is laminar...

If not already doing so, try assigning values to both *rate and *rr_t (FLUENT will pick the limiting one as in the Eddy Viscosity model), i.e.

*rate = (whatever your formula is);

*rr_t = *rate;

Maybe this helps!

/Henrik

Have you tried the soultion procedure outlined in the fluent manual? Try to start with a cold flow solution with no reactions and then gradually add the reactions one by one. If the system is numerical stiff it will then start to diverge and you need to use the stiff chemistry solver.

RoM