[cfdonline2mohsen]

Hello all Foamers

I usually choose the time step in LES based on the CFL number to be less than unity (or 0.5) in order to have a stable run.
Now, I want to calculate the smallest resolved time scale in my LES run to check whether the selected time step is suitable or not.
Any one can help me how to do that?

[hxaxtma]

see here:

Kolmogorov Microscales

https://en.wikipedia.org/wiki/Kolmogorov_microscales

[klilla]

Hi,

the smallest resolved time scale present in your simulation only can be approximated based on the numerical scheme you are using. A rule of thumb is that for a 1st order scheme you need 25 nodes/wavelength, for a 2nd order scheme appr. 16 nodes/wavelength (for standard schemes).

This means:
your scheme gives you NPW - number of points per wavelength
your mesh size is DX

Then the resolved wavelength can be calculated as: k = 2*PI/(NPW*DX)
From wavelength you can calculate the resolved frequency: f = k*c/(2*PI)
where c is the problem's propagation speed: incompressible simulation - flow speed, compressible simulation: sound speed.

Again, this is only a theoretical approximation.

If you really want to know exactly what scales you resolve and what not, plot the energy spectrum in some points and that gives you directly the cut-off frequency.

Best,
Lilla

[hxaxtma]

I think I missunderstood the question :-) Thanks killa for your answer