Hi all,

Which method is the more commonly used for turbulent methane/air combustion.

PrePDF or Eddy Dissipation and why is one used more than the other?

Cheers,

Mark

Well, they are both basically pretty coarse models actually. The main problem with ED model is that it not made for equilibrium reactions and that it is meant to be used with no more than only one or two reaction steps. Very simple therefore, and not handy if you want to predict NOx where you need more intermediate species. It also doesn't take effect of turbulence on reaction rate into account, which could affect your accuracy. The good thing about the ED model is that it is cheap (CPU wise), and you can tweak the coefficients (A, B and reaction coefficients) to match experimental data. Despite its shortcomings I personally like the method, also because it is not so vague as the PDF method.

BUT, the PDF method allows for many species and takes turbulence-chemistry interaction into account (although I still can't figure out how a pure mathematical thing like a beta-pdf can describe the time or spatial variation of fuel-oxidizer ratio but that's another thing). It is more CPU intensive, but most people that need to look closely at flames seem to prefer this method. One problem: the PDF assumption is that all reactions are fast and in equilibrium, which means that you can also forget predicting CO formation with this method.

The new ISAT method in Fluent 6.1 could be a major improvement over the existing methods although from the description it seems to me that it is not altogether straightforward to use it.

Hope this helped a little...

Cheers back.

Thanks for your reply Erwin,

I am using prepdf 4.1 for natural gas to input into fluent. I initially conduct an adiabatic and then a non-adiabatic model. My peak temperatures are the same for both models. Surely this is wrong?

I set the species and their compositions, including products. I left the products composition blank, I assumed this is what is done. I may just be reading the wrong results. The fluent manual isn't very clear on reading the results.

Any ideas where I might be going wrong?

Cheers, Mark.

What you look at are the adiabatic results (the adiabatic flame temperature), since PrePDF cannot know where your heat is going... that will only later be a factor inside the fluent claculation.

Thanks Erwin,

So the non-adiabatic model just tells the model to prepare for the heat transfer when its being solved in Fluent, but can't account for it in the prepdf solution itself because it doesn't have the required information?

I don't fully understand the graphs and readings given by the prepdf solution. Do you know of any tutorials online, the fluent tutorial are useless.

Mark.

No, I don't know anything online. I try to read books or PhD theses on combustion whenever I can get my hands on them. If you find a good source online let me know...