[beyonder]

Dear All,

i have a 3D Navier Stokes solver with no turbulence model implemented but also I am not resolving the scales considering my Re is around 1000 and my grid is around 256X56X64.

So I want to find the wavenumber spectra from the 3d output of my simulation to see up to where I am resolving and after where it starts to accumulate energy. I found some stuff online but I thought here it is more reliable.
Do you know the simple commands in matlab to calculate this?

Thanks in advance

[FMDenaro]

Originally Posted by beyonder

Dear All,

i have a 3D Navier Stokes solver with no turbulence model implemented but also I am not resolving the scales considering my Re is around 1000 and my grid is around 256X56X64.

So I want to find the wavenumber spectra from the 3d output of my simulation to see up to where I am resolving and after where it starts to accumulate energy. I found some stuff online but I thought here it is more reliable.
Do you know the simple commands in matlab to calculate this?

Thanks in advance

Your question is not completed ... what about the type of flow? the Re is computed using what lenght and velocity?
you have the physical Kolmogorov lenght scale and 3 Nyquist cut frequencies pi/h_i for each direction.

[beyonder]

I have Fourier-Fourier-Chebyshev code. So two dimensions are periodic B.C. while the other one is imposed velocity, diricihlet.

I am solving non dimensionalized navier stokes, where in the code my only input is Re.

while my modes/cells are [256 X 56 X 64]
my domain is mapped onto [30D X 3D X 4D] where D represents the unit dimension. Because I have a turbine in the mid domain with a diameter of D, and represented via body forces (as in source term)

my Re is 1000. this must be the diameter based reynolds.
and delta t of 0.004.

also the inflow conditions are Gaussian perturbed. so in the end when the flow develops a bit you receive a video of such
https://www.youtube.com/watch?v=bFFb...ature=youtu.be