This might be an interesting case for discussion. There is unusual vibration, but it is not necessarily failing in the near future. I have an idea what mechanism might cause this pattern, but I may not be able to get the answer for quite awhile. It won't get rebuilt unless this really increases, or we get another issue. It is quite a few man hours to rebuild this particular piece of equipment.

2x_TS_PV_Centri-Screen.pdf (226 KB, 64 downloads) circular plot 2x TS PeakVue

Most spectacular fail I have seen in one of these if it´s the thing I think, was a arm rotating inside that broke from fatigue due to a perfect resonance tuned to a existing impulse. It took some 10 years of operation and it may be that time again about now if nothing was changed. Would not be certain to say that it would be indicated like this but at least I saw it happen once. It made a total mess of the internals so the knock test was on the spare parts. Olov

One possibility is belt load combined with asymmetric stiffness of screen shaft...possibly created by the keyway. When keyway is in-line with keyway or 180 from there, shaft stiffness is minimum. Rotate shaft 90 degrees and shaft stiffness is minimum. The stiftness reaches minimum twice per revolution and maximum twice per revolution... it has a 2x component.

There may be other causes of asymmetric shaft stiffness as well. Shaft crack was one mentioned in another thread when I posted 2x on belt driven machine.

The screen I assume is something like a loop that rotates between two rollers? In that case, a single irregularity on the screen could be a cause. The screen passes each roller once per revolution... the irregularity on the screen would be in contact with a roller twice per revolution.

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

Added picture of a similar system. I have not seen the inside of one of these yet, but I think it is a sort of centrifugal sorting screen. I'll try to get some drawings next week.

Oli,

Do yo know the insides of this beast? Any thing that could cause 2 impacts each revolution?


Pete,

I was thinking someting similar. I believe this could be a loose key/keyway. I have seen this same pattern in the past on a few different pieces of equipment. I may be able to inspect this on the next down day in a few weeks. Not sure if I can see they key or get a picture. We'll see.

Yes, that´s exactly the thing, it´s like 10 years ago so memory is not that detailed there is a screen basket inside and some rotating arms from centre and out my memory say it was 3 arms but I would not swear on it. I try to find out. Olov

Some of these centrifuges have two blades inside that need adjustment every so often. Medium make effect and depending on product and type of medium may have an affect from its very nature that gives an uneven thickness. The setup of bearing clearance is critical and some may have excessive axial float causing 2X.

Now, is medium product or is it a catalyst or filter? The cake may have variances that come and go during operation, so this 2x phenomenon will keep changing w/time.

This is a way it could look, not the one I saw but sure messy inside and yes it seems to be versions at least for testing with 2 wings:
http://www.uilo.ubc.ca/partner...iles/pulpscreen.html

/Olov

US3713536Screen1.pdf (316 KB, 17 downloads) screen patent

The picture looks like what we called "Selectifier Screens" at the papermill I retired from years ago. These had 2 blade "paddles". Never saw much 2x on them with much amplitude at the fundamental position, They would sort of act like pumps with "cavatation" noise and usually in the "higher" frequency area of the spectrum.
I see your Peakvue is a 500 Hz HP Filter, which might be bringing the "2x" into the spectrum from sidebands impacts of bladepass located in a "noise" area above 500 Hz. Have you a data set covering this area, say 1000 Hz Fmax in acceleration showing the 500 to 1000 Hz area in non Peakvue.

As pointed out: this unit has multiple applications.

Many of these the tub will have a tendency to crawl up if it isn't restricted properly. I've seen some where .4" clearance was along the shaft and dependence on gravity was implied but not always the case.

Historically, if an old unit; is there a file? Has sheave diameters been changed?

I would suggest to get back to the fundamentals of data processing and data with integrity.

The peakvue time waveform has two very clean, distinct impacts 180 degerees apart. Besides the loose key, I like the 2 paddles impacting something theory (if it has 2 paddles). I'll try to get more information and let you know.