[Manuel CFD]

Dear all,

I would like to simulate a non-premixed combustion problem with two different fuel streams having different compositions and a single oxidizer stream using the flamelet model. I found out that two distinct flamelet solver packages have been developed in OpenFOAM: one by Alberto Cuoci and one by Hagen Müller. I would like to know whether one and/or both of them would be applicable for my problem or whether they are restricted to the case of a single fuel stream.

Thank you for your help.

Greetings,

Manuel

[dlahaye]

The limitation is not the implementation itself.

In case of two fuel streams, the combustion is staged. The fuel stream travel through the volume at different local stochiometric ratios. Capturing this properly, requires working with mixture fraction and mixture fraction ratios per fuel stream.

I am interested in discussing an implementation of the flamelet model that is entire based on function objects. This implementation should allow the required flexibility for two fuel streams.

[cryabroad]

The one by Alberto Cuoci is for RANS, if I remember correctly. You should look into the code and double check that.

But Domenico is right, the key is not the implementation itself.

[Manuel CFD]

Hi dlahaye and cryabroad,

thanks for your replies. Sorry for my late reply but in the meantime I started to work on another project and I put aside the question about the Flamelet model.

I understand that the implementation is not an issue. However, it is not clear to me yet whether the pre-processing phase, i.e. the generation of the flamelets, can be representative of the combustion of both fuel streams. From my understanding the flamelets are generated using boundary conditions describing the inlet conditions of a fuel stream and an oxidizer stream. But in my case I have two different fuel streams with different flowrates and different compositions (the first is methane + hydrogen + minor contributions, the second is CO2 + methane + hydrogen where CO2 is the most abundant). Shouldn't two different sets of flamelets be solved, each set being representative of one of the two fuel streams? Or my understanding is wrong and a single set of flamelets would be enough? In that case what boundary conditions should I use at fuel side to generate the flamelets?

Thanks for the help and the explanation.

Regards

Manuel

[dlahaye]

You will need two (2) sets of flamelets.

Boundary condition at the fuel side is the fuel mixture.

[Manuel CFD]

Dear dlahaye,

thank you for your clarification.

Regards

Manuel

[danhnam]

Hi Manuel

I think you need to use three-stream model for your problem.
You might want to refer to this paper.

https://web.stanford.edu/group/ihmeg...f/ihme2011.pdf

Danh Nam!

[dlahaye]

The Imhe2011 reference only has one (single) fuel inlet. It is different from a configuration having multiple fuel inlets.

[danhnam]

Hi Domenico Lahaye!

In my opinion, we could use three-stream model by applying the same idea of Imhe (2011) but for the fuel stream instead of for oxidizer stream. In that case, you need to define a new scalar variable, for instance, compositionFactor. This new scalar variable only affects the boundary condition. For example, you could define a nominal mole/mass fraction of fuel such that it can account for both 2 different fuel stream compositions. But the most important thing for this problem is that you need to generate a new 4D flamelet table instead of 3D table. In the flameletFoam solver, you also need to implement a new interpolation method as well as an additional transport equation for compositionFactor as mentioned above.

Danh Nam!

[Manuel CFD]

Hi Danh Nam,

thanks for your suggestion. Yes, introducing a new scalar variable seems to be the most direct approach to deal with this problem. However, I don't know whether I am going to work on such extension because I was asked to figure out whether the current version of the flameletFoam solver can deal with two fuel streams. Since this is not possible, we would have to consider a different option for this project. However, in case I will find some time to work on extending the current flameletFoam solver I will post an update here on the forum.

Thanks.

Best regards

Manuel