Hello,

I'm using multiple species model to simulate an internal combustion of the methane-air mixture (Options: volumetric Reaction, Diffusion Energy Source, Finite-Rate reaction model). I get a strange effect of the temperature rise in the core of the flame all the way up to maximum value specified in the Solution Limits panel (whichever value is entered). In all other aspects the solution looks plausible. I would quite appreciate any comments on the problem. Thanks,

Yulia.

I don't have experience with Fluent but generally speaking you may reduce the time step and/or reduce the ignition energy and see what happens.

first deleting diffusion energy source und trying to use finite-rate/eddy-dissiption model.

Thanks for the comments. I tried to cancel a diffusion energy source in the Species model, but this did not affect the solution. On the other hand, using of the finite rate/eddy dissipation reaction model eventually leads to extinction of the flame and stops the reaction altogether.

Hi,

is the flame turbulent or laminar? Is it premixed? With other words: are you forced to use finite rate?

If yes, then use the coupled solver. And try defining the diffusion coefficients e.g. by kinetic theory or maybe use the multicomponent diffusion approach (available since 5.1) . This seemed to help in my case

Volker

I assume you are solving turbulent premixed combustion case. (If not, you can use PDF approach or follow Volker's good advice) Again I assume that this case has a solution, I mean, that with your configuration and mixture parameters non-pulse combustion will occur. For premixed flames it is not always obvious. Combined finite rate and eddy dissipation model seems more physical for that case. To keep flame from extinguishing on early stage of solution you could try several things. First you may increase mixture temperature, then you can increase temperature at some boundaries, which are near the combustion area, and you can reduce excess air, making mixture more stoichiometric. After you start your calculations (patching "spark" I guess) keep pretty conservative energy under-relaxation coefficient. When you reach decent solution try slowly to return boundary conditions to their desired values.

Hope it helps.

Kostya

Hi!

Many thanks for all the comments and advice. As I mentioned in my original posting I'm using Multiple Species model (not Premixed or PDF), defining methane-air mixture consisting of 5 elements. The flow is turbulent (k-eps) and I'm supposed to arrive to steady mean flow. I use the segregated solver. The Cp of the mixture is constant and equals 1000. To tell you the truth, I had a big problem with ignition, so I end up by introducing artifitial thermal condition (radiator with specified heat flux) on a small annular surface near the fuel entrance (all other kinds of temperature rising conditions did not help). Then I got stable combustion. The surrounding temperature field (and actually, all other flow parameters) are quite plausible - on the contrary to situation described in recent posting by Devy, who reports very low temperatures around the flame.

Any new comments would be greatly appreciated.

Cheers,

Yulia.