Hi,

I'm modelling heat transfer from a non-premixed turbulent flame in a rotary kiln using the two-step eddy dissipation model. I have converged a second order reaction solution for the model and everything looks good.

Now I want to add the discrete transfer radiation heat transfer model to the solution using default values for the DO model.

I can't get a solution when the radiation model is used. In fact it seems to extingush the flame in the kiln.

I have tried to change the urf for radiation as low as 0.7 but this doesn't work either!!

Can anyone offer advice, I've been trying to solve this for way to long!!

Thanks

Mark

Hi Mark,

and what about the radiative flame properties? you don't mention setting them, so I'd recommend cell-based wsggm for absorption coefficient zero scattering coeff. Hope it helps!

Hi George,

Sorry yea, I set absorption coeff to wsggm-domain-based, as I'm really only interested in the overall heat transfer, and I set zero scattering for gas cmobustion, and default values for reflection and scatter phase.

Sould I change the absorption coeff to wsggm cell based?

Thanks

Mark

No, that will lead to an error in your heat balance. Leave it as is, domain based DO is correct (or a manual input of the mean beam length).

The WSGGM should lead to absorption coefficients in the range of 0.2 to 0.4. Correct?

If the flame extinguishes, your radiative flux may be too high. This means the wall temperatures could be too low. Did you model the insulation properly?

Hi Erwin,

Thanks for the reply. I'm not exactly sure what you mean by, "model insulation properly"?

The insulation is set as a solid and not participating in radiation, as there is an wall between the combustion space and the insulation itself. The material properties are correctly applied.

I have checked the absorption Coeff and there is definitely something wrong there - the total range is from 0.2 (just downstream of the burner) to 0.95(in the flue). Another point is, as the gases flow from the burner they flow through a restriction in the geometry. This is the point where the absorption Coeff. starts to change. There is a front in the flow where the values change from 0.3 approx. to 0.65. It is also at this point that flame seems to extingush!!

This seems to be causing the trouble, how can I lower the absorption Coeff. at this point?

Any ideas, getting desparate at this stage!!!!!!!!

Thanks

Mark

What I meant by 'model insulation' was 1) it should be there and 2) it should have the right thermal cond. value. But it sounds like you did that.

You can always manually input the absorption coefficient, but it makes more sense to find out why it increases so much at this particular point. What is the velocity? Also check what reaction products you get here, what is the flue gas composition.

The other value that determines the absorption coefficient is the pressure; maybe the local stagnation pressure is unrealistically high at that point.