Fast Fourier Transform and CFD

Fun_Doge · January 29, 2017, 05:14

Hi,

Since people here have an understanding of CFD and are good at explaining stuff to us newbies, I have to ask something that may be silly.

I was doing a FFT of the Nusselt number over time (for a RANS simulation) and got a graph where there is one peak, and the signal decreases periodically. The thing I don't understand is why you want to do a FFT of the Nu number in the first place. Is the first peak related to the first VK vortex? I am self-taught and just gave this a try by following a tutorial to learn plotting in CFD. The theory was not explained.

FMDenaro · January 29, 2017, 05:21

RANS is a steady formulation, maybe you are talking about URANS?
The Fourier analysis of a signal is a very classical tool to analyze turbulence, it produces the energy spectra and you can detect the wavelenght of characteristic flow structures.
In your case you have a periodic time-window?
Could you post your results?

Fun_Doge · January 31, 2017, 04:55

You are correct.

FMDenaro · January 31, 2017, 05:00

try a log-log scale, the peak should correspond to the frequency of max spectral density...however, I do not remember analysis of spectra of the Nu number....

Fun_Doge · February 1, 2017, 06:25

Thank you, I will try it!

January 29, 2017, 05:14	Fast Fourier Transform and CFD	#1
Fun_Doge New Member Sarah M Join Date: Oct 2016 Posts: 8 Rep Power: 10	Hi, Since people here have an understanding of CFD and are good at explaining stuff to us newbies, I have to ask something that may be silly. I was doing a FFT of the Nusselt number over time (for a RANS simulation) and got a graph where there is one peak, and the signal decreases periodically. The thing I don't understand is why you want to do a FFT of the Nu number in the first place. Is the first peak related to the first VK vortex? I am self-taught and just gave this a try by following a tutorial to learn plotting in CFD. The theory was not explained. Last edited by Fun_Doge; January 29, 2017 at 05:16. Reason: forgot a word

January 31, 2017, 04:55	plot	#3
Fun_Doge New Member Sarah M Join Date: Oct 2016 Posts: 8 Rep Power: 10	You are correct. Last edited by Fun_Doge; February 1, 2017 at 06:25. Reason: wrong file

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Cannot achieve negative wall shear stress values in CFD Post (ANSYS)	SAtkins	CFX	3	August 3, 2016 15:27
new to CFD, changing angle of attack/incidence?	bjk87	STAR-CCM+	9	October 2, 2015 04:32
CFD meshing projects for MS	sandipan_b2002	Main CFD Forum	1	August 7, 2012 05:01
[ICEM] Import a Nastran file into ICEM CFD	norger	ANSYS Meshing & Geometry	0	May 14, 2012 08:37
EXTRUDE option in ICEM CFD Geometry	Mohammad Faridul Alam	CFX	4	October 25, 2007 18:47

January 29, 2017, 05:21		#2
FMDenaro Senior Member Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,893 Rep Power: 73	RANS is a steady formulation, maybe you are talking about URANS? The Fourier analysis of a signal is a very classical tool to analyze turbulence, it produces the energy spectra and you can detect the wavelenght of characteristic flow structures. In your case you have a periodic time-window? Could you post your results?

January 31, 2017, 05:00		#4
FMDenaro Senior Member Filippo Maria Denaro Join Date: Jul 2010 Posts: 6,893 Rep Power: 73	try a log-log scale, the peak should correspond to the frequency of max spectral density...however, I do not remember analysis of spectra of the Nu number....

February 1, 2017, 06:25		#5
Fun_Doge New Member Sarah M Join Date: Oct 2016 Posts: 8 Rep Power: 10	Thank you, I will try it!