1D freely propagating flame - N2 mass fraction behaviour

[Rafael]

Hello there!

I've been simulating a 1D premixed frelly propagating CH4 flame (CH4:0.05518667, O2:0.22014124, N2:0.7246721), with a skeptical mechanism (Yangpope1998) which does not hold any reactions with nitrogen. Therefore, N2 is indeed inert.

I've observed differences in N2 mass fraction depending on which diffusive transport model i'm using.

For UnityLewis approach i see no fluctuations in Y_N2. Meanwhile, if mixture-averaged is considered there are an oscillation in Y_N2 in the flame front.

It is not clear for me why does this oscillation happens? If have any physical meaning? And if affects or not the convergence of the flame structure and speed?

Any thoughts on it?

Thanks in advance, 

Rafael Coelho

[Ray Speth]

Hi Rafael,

I think this local perturbation to the N2 mass fraction is in fact real. One way to see that there must be some local variation in the mass fraction even for an inert species is by looking at the diffusive flux term in the mixture-averaged case. Here, the flux is proportional to the mole fraction gradient of N2. Since the mixture molecular weight is not a constant (at least, in the vicinity of the flame where you have various intermediates that reduce the local value), the mole fraction of N2 cannot be constant. This means that there must be a non-zero mole fraction gradient and a non-zero diffusive mass flux, leading in turn to a variation in the mass fraction. Of course, at the global scale you're correct that this must be balanced out, and you do see that the mass fraction eventually tends back toward the initial value.

This effect does not occur with the Unity-Le transport model because in that case the diffusive fluxes are defined in terms of the mass fractions.

Regards,

Ray