CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Siemens (
-   -   Simpler Chemical Reaction Problem (

Sam February 2, 2006 17:13

Simpler Chemical Reaction Problem
I have the following problem. Main gas stream is Air with a 30 PPM (parts per million by volume) of SO3 at the inlet. I'm injection NH3 downstream of the inlet. The chemical reaction is NH3 + SO3 + H2O --> NH4HSO4.

The reaction is temperature dependent but instantaneous. I don't know the reaction constants. Is there are way to combine the intesections of the scalars of SO3 and NH3 and H2O to form a new scale of NH4HSO4? Or, how do I model this chemical reaction. I'm new to this.

Thanks for the help.


Guillaume February 3, 2006 10:36

Re: Simpler Chemical Reaction Problem
It seems you want to use a `mixed-is-burned' model, don't you?

You have to define three passive scalars with the same transport equation than NH3, SO3 et H2O respectively. It gives you the concentration of these reactants if there were no reaction. You can thus determine which one would be in default and deduce the effective concentration of NH3, SO3, H2O and NH4HSO4 on a cell-by-cell basis.

In Star-CD, use the sorsca.f exemple coding to impose this effective concentration. The 'chemico-thermal' enthalpy option should give you the temperature from the concentration field.

What do you mean by `the reaction is temperature dependent but instantaneous'?

Sam February 3, 2006 10:43

Re: Simpler Chemical Reaction Problem
"Mixed is burned" is correct. For our purposes the reaction is instantaneous. By temperature dependent I mean that the reaction will not occur below a certain temperature. I'd like to see the three passive scalars enter the domain, the reacted scaler where the temperature is above a certain T_act, and the remining reactants scalars.

Will Star automatically create the 4th scalar when I set up the reaction? I'm looking at the methodology volume and I believe I want to use the EBU model.

I'm not familiar with the sorsca.f coding. Is there anywhere that I can learn more about this? Does the sorcsa.f coding eliminate the need to model the chemical reaction?

Guillaume February 3, 2006 14:48

Re: Simpler Chemical Reaction Problem
The reason I spoke about sorsca.f is that you have to implement the combustion model by yourself. EBU is not 'mixed-is-burned'.

You have to define all the scalars (reactants, products and passive scalars) + the background medium properties.

sorsca.f is described (a bit) as any other user routines in Chap. 18 of the user's guide (`USER PROGRAMMING'). When this routine is written down, you will have sample coding.

Otherwise, you may run the `EQUILIBRIUM CHEMISTRY MODEL' with `MULTIPLE FUELS' but you will have few control as it is an in-build feature. On the other hand, it does almost everything for you. You still have to define reactants and background fluid. Furthermore, the flow has to be turbulent, is it the case?

Sam February 3, 2006 15:01

Re: Simpler Chemical Reaction Problem
How do I implement the Equilibrium Chemisry Model? Is it in the GUI?

Guillaume February 6, 2006 10:24

Re: Simpler Chemical Reaction Problem
It is in the GUI. Check the user's guide.

Sam February 6, 2006 14:28

Re: Simpler Chemical Reaction Problem
Okay. I have implemented the reaction. Again the original equation is NH3+SO3+H2O = NH4HSO4. When I set up the reaction a N2 scalar is setup. I set up the inlet as 95% N2 and 5% H2O and 30 PPM SO3. However, when I plot the N2 scalar it is not constant in the entire domain. It is reacting as well. Is there any reason for this. I only want the scalars in the reaction to show any change in concentration.

Guillaume February 7, 2006 10:28

Re: Simpler Chemical Reaction Problem
If star predicts a reaction, there should be a reaction.

If you do not want it, create a virtual N2, VN2, and assign to him the same properties as N2.

All times are GMT -4. The time now is 17:25.