Ruben,
I know the gears are not split on these printing rolls. I inspected the gears and counted the number of teeth so I would know my gearmesh frequencies. The gears are mounted on the opposite end from the spectrum I posted and the readings look much better on that end. I have asked for the bearing information but have not received it yet. Is there any chance the variable frequency drive could induce this type of vibration. I was told they had a problem with the vfd on this unit. They were having trouble getting the voltage balanced. I was told it was corrected to some degree but not totally. The motor is mounted on the end where my spectrum and waveform came from. At this point, I have not ruled out bearing or drive problems. But I am on hold until I get bearing info, oil sample results, and possibly take readings on the motor. I do appreciate the feed back from all who have written. Hopefully I will have some follow up info to post.
Thanks,
Ronnie

More often than not, a 1x and harmonics in Peakvues are indicitave of a rub. Possibly a seal,retainer, high spot. The waveform should be spiked(once per rev.), and the width of it indicative of the surface area being rubbed. If this is the case, you should be able to sync it with a strobe and determine exactly where the rub is.

Ronnie,
After looking at your spectrum, this is almost exactly what an inner race fracture spectrum looks like. The mounds will probably line up with BPFI faults, and the sidebands are turning speed.
If you have access to a contact ultrasonic probe, you can hear it. Demodulate it at about 100kHz.

Red
I am using RBMware 4.8 version. Where do I find the "Auto Correlate" function? And how do you use it.
Thanks
RLE

RLE
The 'Auto Correlate' function is on the graphics function bar on the right hand side of the screen when viewing a time waveform. The way it was explained to me is that the time waveform is taken and divided into 2. The 2 halves are overlayed on top of each other and any repeatable events are displayed as peaks in the resulting correlated waveform spaced apart by the frequency at which they occur.(Notice the correlated waveform is half the time of the original waveform). The amplitude of these is on a scale of 0 to 1 which is a measure of 'certainty' that there is a repeating event - 1 being 100% certain. I find this function very useful on peakvue time waveforms as a way of confirming a fault seen in the frequency spectrum. Fault frquencies can be overlayed as normal. Hope this helps

David

Ronnie,

Did you ever have a resolution to this issue? I have seen similar on a paper machine but it comes and goes.

Thanks,
Matt

Hi Matt,

I was stumbling through this section of the boards and have not read all the posts in this thread but what problems are you seeing on your paper machine?

Ronnie,

I am in agreement with those who had suggested inner race fault, likely a crack. Pay attention to the spacing between mounds (about 20x) in the PeakVue spectrum. It may be BPFI. Important thing is their sidebands. Hard to see but it looks they are spaced at 1x and modulation is strong. In combination with very impactive pattern in the TWF all this suggests cracked inner race.

How does the normal spectrum look like? Any signs of 1/2x?

I guess you know the answer by now. What is it?

This message has been edited. Last edited by: David_G,

Sorry Guys,
I have not been keep up with this thread. Bama Vib started a new thread on Sept 29 "Multiple Harmonics in Peakvue" asking me about this. I posted some follow ups to this with an explanation of the results. I am not sure why but I have not seen these most recent posts on this thread until I clicked on the link under Matts posting. Not sure why that is because I check the forum on a daily basis. Anyway, if anyone reads this, please reply on the other post by Bama Vib.

Thanks alot for the interest,
Ronnie

Ronnie, I am in the same boat you were in when you posted this. I have been "Tasked" with taking over a PdM Program. Right now all we are monitoring are our printing folders. Do you have any examples or data that might be useful to me . I would appreciate it.