Hi CFD friends!

I encoutered a kind of weird result in my pdf-combustion simulations (10 atm). Using the Discrete Ordinates Radiation Model with the WSGGM option, I got values of absorption coefficient up to 1.9! That seems to me apparently wrong, reminding the basic radiation law that the sum of absorption, transmission and reflection coefficients must be = 1. I point out that it's not a matter of grid dependency because it has been properly refined and values over 1 are spread all over the domain (and not within a confined region).

Anyone could give me an explanation: could these values make sense physically or are they definetely and anyway wrong? Are they somehow consistant with Fluent definition of absorption coefficient, which is dimensionally 1/m(lenght), whereas it should be adimensional?

Thanking you in advance.

Jamiro.

Hi,

I also have absorption coefficient>1 but i think that it is not wrong. The absorption coefficient is defined as 1/m so it is local definition. That means you have to integrate this parameter along a line. Of course, if the result of the integral is >1 then Fluent is physically wrong.

Hi Frederic,

thanks for your answer. But what do you mean for 'lines'? (Actually, using 'D.O.' model I can't integrate on 'rays': do you mean the characteristic cell size or the mean beam length, according to settings?)

Thanks again.

Jamiro

You are mixing up the absorption coeff. with the absorptivity. It applies: Absorptivity = emissivity (at the same wavelength); emissivity + transmisivity + reflectivity = 1, which means that for a black radiator with emissivity =1 there is no transmission or reflection.

The absorption coeff. gives the reduction of radiation passing through a medium (e.g. flue gas or even a solid like steel).