FFT of RMS velocity component
 

Dear all

I hope someone can point me in the right direction here, I think I am missing something. I want to do an FFT of the fluctuating velocity component of an LES simulation in order to see if I have captured kolmorogov's -5/3 relation in the inertial region. I believe that I should be able to use the RMS velocity terms that fluent can output under unsteady statistics, however when a put points in the areas I am concerned about and attempt to output the integral monitor values at those points I am left with column of zeros. I appear to be able to generate an output using mass or area average values but that to my mind would destroy the fluctuating components I am concerned with.

Any thoughts or shoves in the right direction would be most welcome

Many thanks

Two points:

1) You are right, most of the monitor functions return averaged values. What you are looking for, according to my memory, is the facet average, which should be something like a "nearest cell center value". Still, i would check the manual.

2) You can't use the RMS from the Fluent statistics to do what you want to do. You can either work directly with the instantaneous velocity or with a custom field function where you put the instantaneous velocity minus the average one. These are the only options. The fluent statistics are like RANS fields, they are not time dependent and do not have any spectral content. Would you expect the RMS of any signal to be just a number or a signal itself?

Hi Paolo

Thank you for your response, I will have a hunt around the manual, I have been trying to avoid using an averaged value entirely up to now as I felt it would eliminate any useful components.
I did wonder whether that was the case, I stumbled across something somewhere that that was fluents way out outputting the fluctuating component. Irritatingly I cant find it again now. I have had pope's book arrive in the post today so I am hoping that will shed some more light on the issue to.


Many thanks