Hello, I have written a code for 3D simulation of a turbulent flow in an internal combustion engine and I am about to incorporate in it the combustion of hydrogen.

Does anybody know if the reduced chemical models (5-7 chemical equations instead of 19 which is the "accurate") give quantititative correct results?

Can anybody suggest me any good articles for premixed or partially premixed combustion involving hydrogen and how this is implemented numerically?

Thank you in advance.

I mean how the source terms of the species concentration change, and which terms are treated implicitly and which explicitly. This kind of questions I have and asking for an article.

Thank you again.

Hi Georg, hydrogen combustion in ICE - pretty nice topic!

There is quite a number of possible approaches for modelling combustion of hydrogen in engines and it depends very much on your conditions (turbulent, HCCI whatever?). If send me e-mail I can send you some material.

Why do you want to use reduced kinetic mechanism? Is there a special reason? Under turbulent conditions it may be a difficult approach (however not impossible). Reducing mechanism is an art. Even more as the complete mechanism for hydrogen under engine conditions is still topic of scientific research. May be one of the automatic reduce approaches is of interest for you (ILDM or in-situ adaptive tabulation).

Regards, J.

PS: I am not often here.

Thank you for your response. I will gather all my questions and soon I will send you an e-mail with some remarks-questions.

Best regards, George Kosmadakis

I want to simulate turbulent combustion using the k-epsilon turbulence model in a spark-ignition engine. I wanted to use a reduced chemical mechanism, so that the computational time will be reduced and less equations will be solved, without the results to be bad. But I am confused, because all the papers that I see are about chemical kinetics only and I think that in spark-ignition engines the air with hydrogen are partially premixed, therefore chemical equilibrium mechanisms should be also incorporated. I would appreciate if you send me some material and maybe your thoughts about the above.

Thank you in advance.

Best regards George

You should not use equilibrium because combustion is not instantaneous in an ICE. Only kinetics is important.

Thank you for your response. I think equilibrium is used for the basic reactions and for the creation of nitric oxide chemical kinetics. Or not?

NOx chemistry is even slower than combustion reactions. It cannot be obtained by equilibirum. Only by kinetics.

What do you mean by `basic reactions'?

The reactions except from the ones NOx is involved. Do you have to recommend me any paper, where all these things are involved?

Thank you in advance

Nothing in mind but you should do a search like "kiva + hydrogen" on scholar.google.com.

What if I add to the source term of each Y species equation the term (from equilibrium or kinetics) that has the largest time scale (is a more slow reaction). By this way all reactions will be governed from both equilibrium and kinetics according to the slowest one. Is this approach logical and correct? Any arguments?

Thank you in advance

You cannot put the source term of A+B->... to C+D->... simply because the reaction time is larger.

OK, but let's say we have a reaction A+B->E+F, and we try to find its reaction rate. The reaction rate for the specific reaction would be the smaller one that comes from equilibrium and kinetics. R=min(Requil,Rkin). Or you mean that there are other reactions used for equilibrium and others for kinetics?

I really appreciate your help-advice.

Best regards, George

By definition, there is no reaction (in term of progress variable source/sink) in equilibirum. Equilibrium considers that the kinetics is infinite. Technically, the mixture is always in the burned state (minimum of the local Gibbs function).

And then how would you suggest me to continue? For hydrogen as a fuel in a spark-ignition engine, which is the most correct way to deal with chemistry, equilibrium or kinetics, or some combination of these two?

Thank you in advance

If the CFD code you use has it or permits it to be implemented (which is usually the case), I would first begin with the PaSR (Partially Stirred Reactor Model). Although it has been developed for diesel engines by Chalmers U., it is versatile enough (and also simple to implement) to simulate spark ignited engines. The PaSR approach is certainly were to begin, to my opinion.

This model is quite simple? Because I had in mind to go for the chemical kinetics approach and maybe for the whole reactions (19 in number). Which one do you think could give more reliable results?

Thank you for your response

The PaSR model is for kinetics, not for equilibirum