[hnemati]

Dear all
I am using fluent to perform CFD simulations on combustion of CH4/air. I know that fluent itself has a two-step mechanism. Using this mechanism gives me unacceptable CO emissions compared to lab results.
Based on the drafts that I have attached to this paper, the authors suggested a new reaction mechanism, however, they reported other mechanisms that do not work properly. Does anybody have any hint on that? Or has any suggestion for a more accurate mechanism?
Thanks in advance
Hassan

[Andrea1984]

In order to get your CO prediction right it is important that your combustion model can account for non-equilibrium, since CO is a slow forming species (the same goes for NOx).

What are you using as a combustion model?

[hnemati]

Thank you Andrea
I am using species transport with Finite-rate/Eddy-dissipation model. I am using the same mechanism as reported in Table 7 of the attached work. What do you think about the mechanism?
http://www.vulkano-h2020.eu/wp-conte...combustion.pdf

Originally Posted by Andrea1984

In order to get your CO prediction right it is important that your combustion model can account for non-equilibrium, since CO is a slow forming species (the same goes for NOx).

What are you using as a combustion model?

[Andrea1984]

I don't think it is a good idea to use the Eddy Dissipation model with a multi-step mechanism, because all the steps will end up having the same reaction rate, which is not ideal.

I would try to use the same mechanism with the Eddy Dissipation Concept (I know, the name is similar, but the model is completely different from the Eddy Dissipation) and see if it makes any difference.

Another route would be to use a detailed mechanism (lookup the GRI3.0 for instance). The use of detailed mechanisms with the Eddy Dissipation Concept is quite computationally intensive, so I would look into a flamelet model capable of dealing with non-equilibrium if computational resources are an issue. The premixed and the partially-premixed combustion models in Fluent map the mechanism using a progress variable, so they should be be able to deal with non-equilibrium. In my experience though, predict CO production with such flamelet models can be a challenge.

Good luck!
Andrea

[hnemati]

Thanks for the time Andrea. I appreciate it.

Originally Posted by Andrea1984

I don't think it is a good idea to use the Eddy Dissipation model with a multi-step mechanism, because all the steps will end up having the same reaction rate, which is not ideal.

I would try to use the same mechanism with the Eddy Dissipation Concept (I know, the name is similar, but the model is completely different from the Eddy Dissipation) and see if it makes any difference.

Another route would be to use a detailed mechanism (lookup the GRI3.0 for instance). The use of detailed mechanisms with the Eddy Dissipation Concept is quite computationally intensive, so I would look into a flamelet model capable of dealing with non-equilibrium if computational resources are an issue. The premixed and the partially-premixed combustion models in Fluent map the mechanism using a progress variable, so they should be be able to deal with non-equilibrium. In my experience though, predict CO production with such flamelet models can be a challenge.

Good luck!
Andrea