How does Order Analysis measure phase lag, particularly super-sync amplitudes?

[rladr]

In recent testing of a high speed gear box, an 8X amplitude was observed (on the radial displacement probes) as the gear passed through a certain speed range. An NI4472 pci card and vi's written in 7 Express, w/ Order Analysis tool kit were configured with radial displacement probes and keyphasor as input signals. The behavior of the phase lag measurement is significant. If the phase lag of the 8X amplitude rolled as the gear passed through a particular speed range then this could suggest one particular problem, but if the phase lag remained constant then this could indicate a different problem. So, someone asked: How does the system measure phase lag of supersynchronous amplitudes?

[Chen_Z]

See attached file for figure 1,2,3

The Phase measurement in the LabVIEW Order Analysis Toolkit uses the phase lag convention. Phase is defined as the angle difference measured from the peak of a vibration signal backward in time to the reference trigger point. This means the directions of numerically increasing angles are always set against the shaft rotation. The trigger point here is the tachometer pulses.

Figure 1 shows the relationship of the vibration signal and reference signal to zero degree phase. The shaft has a heavy spot and a keyway slot. When the keyway slot passes the tachometer, the tachometer detects a trigger pulse. The heavy spot causes the shaft to vibrate as the shaft rotates. When the heavy spot passes the proximity probe, the vibration reaches a peak. When the heavy spot passes the proximity probe and the keyway slot passes the tachometer simultaneously, the peak of the vibration does not lag or lead the reference trigger point. At this point the phase is zero degrees.

 The other part of the convention dictates that 90 degrees means that the peak of vibration lags 90 degrees behind the trigger point. Figure 2 1illustrates the 90 degree phase. When the vibration signal reaches the peak, rotate the shaft backward (counter the rotation direction) until the keyway slot passes the tachometer. The number of degrees you rotate is the phase lag, or the phase value in machinery vibration measurement. Figure 2 shows the relationship of the vibration signal and reference signal to the 90 degrees phase convention.

For the 8x phase, the measurement is conducted in the similar way. The difference is that the 8x order component go 8 times faster than the trigger pulses. Figure 3 shows the 90 degree phases ( tacho pulse in rising edge) of the 8x order components. The phase measurement is still conducted through calculating the delay between the peak in the 8x vibration components and the trigger pulses and comparing the delay to one cycle of the 8x components. In figure 3, the signals between the two green grid lines are used to output one phase results.

[Jack_A.]

I have attached a Word document which contains some of the supporting images that Chen mentioned.  You can also find most of this information in Chapter 5 of the OAT 2.1 User Manual under the Order Magnitude and Phase section. 

Hope this helps explain things.

-Jack

Message Edited by Jack A. on 09-12-2005 10:49 AM

[rladr]

Thank you both for the very informative and revealing replies. It would seem a correct conclusion that a signal occurring at eight (8) times rotative speed would have its phase lag relationship measured once every eight cycles of the super synchronous signal. Would you direct me to the VI's that make this measurement and where in the code it occurs?

My motivation here is two fold. One, I am trying to satisfy a customers question. Secondly, I am comparing National Instruments and Bently Nevada. Bently of course, is the established main stay of rotating machinery diagnostics. And, since you both appear very knowledgeable you may be familiar with Bently. Further, you are probably aware that NI and BN's business styles are about 180 degree's from one another, at least in my opinion. So, here I am trying to introduce NI into my company, an OEM of rotating machinery, and I am running into some skepticism. While I am not a degreed engineer, I have 15 years experience in the testing of rotating machinery. I know enough about Labview and diagnostic equipment to know a good thing when I see it. My mission is to learn, address concerns, and hopefully quiet the skeptics.

Thanks,

rladr

[preston_johnson]

Thanks for the comments, and for the details of your application. 

The VIs used in rotating machinery applications are in the Order Analysis Toolset. 

https://www.ni.com/en/shop/labview/select-edition.html

Here is the full manual

https://www.ni.com/docs/en-US/bundle/372879c/resource/372879c.pdf

Also, if you would like to purchase a system built with the order analysis toolset, there are a couple of turn key solutions using NI hardware and software available.  I would have you review the Nexjen VTS100 and the Vibdaq Vibdaq. 

Finally, NI can have one of our Field Engineers provide and on-site demonstration of the tools and related data acquisition hardware. 

We look forward to assisting you and your company. 

Sincerely,