Sound and Vibration

cancel
Showing results for 
Search instead for 
Did you mean: 

Why exist differences in calculated amplitudes using Order Waveform and Magnitude Plot?

Hi Guys, I try to analyse a vibration signal of a runup test with the LabVIEW Order Analysis Toolkit.

At first I determine the main orders and extract their waveforms using the "OAT Extract Most Significant Order Waveforms VI".

At second, I calculate the Magnitude Plot using the "OAT Order Magnitude and Phase VI".

Attached you find the plots in a pdf file. If you study the resultes of the two mentioned VIs, the question arises, why is the differenc between the amplutides of the 23E order waveform plot and 23E magnitude plot so large?

max. amplitudes (micro strain, 0-p):

23E Order Waveform = 360

23E Magnitude Plot = 280

The delta of 80 micro strain (0-p) is a huge amount for me.

Knows anybody what to do in this case? Thank you

0 Kudos
Message 1 of 5
(8,551 Views)

I don't know if this will help, but many years ago when I first started using the Sound and Vibration VI's, I found that some of the VI's produced power numbers, meaning, the values were squared. I was using an Octave Analysis VI (forget exactly which one), measuring acceleration in magnitude linear units (direct measurments in m/s^2).  I had to create a small VI to take each value, get the square root, and sum them all back up again to get the total power.

This was years ago, and I'm not sure if it is even relevent, or if this has changed.. Just something to look at. Make sure the units in both VI's are the same.

Jeff

0 Kudos
Message 2 of 5
(7,791 Views)

As Jeff mentioned the units are different on the waveform extraction versus the order tracking filters.

They were built for somewhat different purposes, but I am more than glad to discuss what you are doing with them and we can probably describe how to get the best of both worlds.

0 Kudos
Message 3 of 5
(7,791 Views)

Thank you for quick response, you're right

The units are different. If I multiply the magnitude plot by sqrt(2), it is closer to the order waveform plot.

OAT Extract Most Significant Order Waveforms VI [EU]

OAT Order Magnitude and Phase VI           [EU rms]

Now I have three questions:

  1. Why is the output of the OAT Order Magnitude and Phase VI only in RMS values possible?
  2. I'm not sure what to do. I have a signal containing information in terms of the most significant order and for sure noise too. What I want to know is the higest amplitude value of this order. Which order track is the best one for me? The peak amplitudes or the averaged amplitudes (RMS)?
  3. @MickyD: What are the different purposes? Could you pleas describe your answer including more details?

Regards Jens

0 Kudos
Message 4 of 5
(7,791 Views)

Hi Jens,

1. The output of the OAT Order Magnitude and Phase VI can only be delivered in RMS. For computational effeciency we are using a bandwidth limiting set of filtes that can very quickly return a magnitude RMS and phase value. Which usually tells you what you need to know for many condition monitoring situations as tracking the RMS or peak is usually equally valuable.

2. If you are just processing a signal, then I would use the OAT Extract Most Significant Order Waveforms VI. It gives you the most information and you could always calculate the RMS magnitude from that waveform. As a post processing tool that is the recommended one, it is just much more computationally intensive so often is not something that can be cone in a continuous monitoring use case, where smaller embedded processor may be used.

One other option would be to use the OAT Order Power Spectrum VI

http://zone.ni.com/reference/en-XX/help/372416A-01/lvoat/oat_order_pwr_spectrum/

If you need to monitor a peak value, it is less computationally intensive than the OAT Extract Most Significant Order Waveforms VI

3. Hopefully in answering the top 2 I have described their different purposes as well.

Let me know if I can provide more info.

Michael

0 Kudos
Message 5 of 5
(7,791 Views)