Power Spectral Density

amir.teimourian · March 4, 2015, 10:10

Hi everybody!

I am working on vortex shedding from bluff bodies and I have done some FFT analysis on my data (velocity data) and ploted the power spectral density against frequency. And I know the peak in power spectrum will correspond to the shedding frequency. I have done this for different sets of experiment. But my question is: does the change in magnitude of peaks from one experiment to the other experiment mean anything?

Thanks a lot in advanced.

agd · March 4, 2015, 14:52

Your question is somewhat open-ended. You refer to a number of experiments you performed to generate this velocity data. Does this mean that you ran your solutions at different values of freestream velocity? If so then that could account for the amplitude differences you are seeing. On the other hand, if you ran different time steps (or effectively different sampling rates) for a single set of freestream conditions, then your amplitude variations could be the result of straddle loss in the FFT. There are several things that could account for the variation you are seeing.

amir.teimourian · March 5, 2015, 03:17

thanks @adg

actually by different experiment, I mean different gap between two bluff bodies to investigate the effects of gap between two bluff bodies on the vortex shedding and St number. so by my question I meant is there any physical meaning of these changes of peak magnitude from one experiment to another? let call this case 1.

another case is as Saha (Saha, A. K. (2007). Far-wake characteristics of two-dimensional flow past a normal flat plate. Physics of Fluids (1994-present), 19(12), 128110.) investigate the far wake.
he perform FFT on transverse velocity at different X behind the bluff body for the same Re number. Although the Freq was changing the magnitude of the peaks was changing as well!

now my question: is there any physical interpretation or meaning for this changes in peak for each case? or those have no physical meaning?

thanks again for your reply
good day

FMDenaro · March 5, 2015, 04:26

Quote:

Originally Posted by amir.teimourian

thanks @adg

actually by different experiment, I mean different gap between two bluff bodies to investigate the effects of gap between two bluff bodies on the vortex shedding and St number. so by my question I meant is there any physical meaning of these changes of peak magnitude from one experiment to another? let call this case 1.

another case is as Saha (Saha, A. K. (2007). Far-wake characteristics of two-dimensional flow past a normal flat plate. Physics of Fluids (1994-present), 19(12), 128110.) investigate the far wake.
he perform FFT on transverse velocity at different X behind the bluff body for the same Re number. Although the Freq was changing the magnitude of the peaks was changing as well!

now my question: is there any physical interpretation or meaning for this changes in peak for each case? or those have no physical meaning?

thanks again for your reply
good day

Just think about the kinetic energy associated to the eddy having that wavenumber... it tells you that vortex of that size is less or more containing energy

agd · March 5, 2015, 11:53

As FMDenaro points out, the peak amplitudes have physical meaning. But something else to be careful of is that the FFT is not just a black box. If you are comparing experiments then you need to make sure that the transforms were taken so that the frequency resolution is comparable between the experiments. The amplitudes indicate that energy is in the flow at that frequency, but the amount of energy can be badly underestimated by an FFT if the frequency resolution is not carefully considered. Do some research on straddle loss and FFTs. Your question is a lot bigger than you might think.

March 4, 2015, 10:10	Power Spectral Density	#1
amir.teimourian New Member Amir Teimourian Join Date: Mar 2015 Posts: 2 Rep Power: 0	Hi everybody! I am working on vortex shedding from bluff bodies and I have done some FFT analysis on my data (velocity data) and ploted the power spectral density against frequency. And I know the peak in power spectrum will correspond to the shedding frequency. I have done this for different sets of experiment. But my question is: does the change in magnitude of peaks from one experiment to the other experiment mean anything? Thanks a lot in advanced.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Power Spectral Density	Madhusanka	CFX	0	March 2, 2011 14:30
power spectral density	vetnav	Main CFD Forum	8	September 3, 2010 15:28
power spectral density	cfd_newbie	FLUENT	1	December 10, 2007 01:53
REAL GAS UDF	brian	FLUENT	6	September 11, 2006 09:23
Warning 097-	AB	Siemens	6	November 15, 2004 05:41

March 4, 2015, 14:52		#2
agd Senior Member Join Date: Jul 2009 Posts: 358 Rep Power: 19	Your question is somewhat open-ended. You refer to a number of experiments you performed to generate this velocity data. Does this mean that you ran your solutions at different values of freestream velocity? If so then that could account for the amplitude differences you are seeing. On the other hand, if you ran different time steps (or effectively different sampling rates) for a single set of freestream conditions, then your amplitude variations could be the result of straddle loss in the FFT. There are several things that could account for the variation you are seeing.

March 5, 2015, 03:17		#3
amir.teimourian New Member Amir Teimourian Join Date: Mar 2015 Posts: 2 Rep Power: 0	thanks @adg actually by different experiment, I mean different gap between two bluff bodies to investigate the effects of gap between two bluff bodies on the vortex shedding and St number. so by my question I meant is there any physical meaning of these changes of peak magnitude from one experiment to another? let call this case 1. another case is as Saha (Saha, A. K. (2007). Far-wake characteristics of two-dimensional flow past a normal flat plate. Physics of Fluids (1994-present), 19(12), 128110.) investigate the far wake. he perform FFT on transverse velocity at different X behind the bluff body for the same Re number. Although the Freq was changing the magnitude of the peaks was changing as well! now my question: is there any physical interpretation or meaning for this changes in peak for each case? or those have no physical meaning? thanks again for your reply good day

March 5, 2015, 11:53		#5
agd Senior Member Join Date: Jul 2009 Posts: 358 Rep Power: 19	As FMDenaro points out, the peak amplitudes have physical meaning. But something else to be careful of is that the FFT is not just a black box. If you are comparing experiments then you need to make sure that the transforms were taken so that the frequency resolution is comparable between the experiments. The amplitudes indicate that energy is in the flow at that frequency, but the amount of energy can be badly underestimated by an FFT if the frequency resolution is not carefully considered. Do some research on straddle loss and FFTs. Your question is a lot bigger than you might think.