Double reduction extruder gear box:
HSP / Int. gear tooth count - 65 / 145
Int. / LS Bullgear count - 31 / 119
Been watching since around January 06 what looks like BPFO on the radial bearings of the low speed shaft. This shaft has 2 identical roller bearings, Rollway LP156UMR. # of rollers = 18. FTF = .427, BSF = 3.338, BPFO = 7.681, BPFI = 10.319. Motor is variable speed, output speed usually runs in the 80 - 90 rpm range. The peaks of interest look like dead ringers for the 4x and 8x BPFO with smaller harmonics dropping off. After digging in further after only thinking it was low speed shaft bearing defects, it seems that the intermediate shaft speed harmonics (shaft speed multiples of 8x) also line up as dead ringers. Oil analysis hasn't shown any unusual increase in wear metals. Wave forms aren't showing any impacting. Question is, got any tips to help determine exactly if it is shaft speed harmonics or BPFO?

If you can trigger the intermediate speed shaft, it seems like a good candidate for time synch averaging?

If the speed is constantly verying you will need a tachometer input.

If it is not varying during your readings, and depending upon the equipment you use, you can probably enter the input speed and have the software calculate your fault frequencies.

Also remember that assembly phase will be a concern in the input gear set so you could be looking at fractional harmonics of that gearmesh frequency.

If you can post your data and include the input speed, I might be able to help out more.

Shaft speeds in double and triple reduction and, etc. gearboxes are usually easy to identify or seperate from bearing faults and gearmesh frequencies.

Takes time and patience to decipher, but most of the time the challenge is worth it.

George: The intermediate shaft is not visible. I've never used the time synch averaging.

Danny: The speed is constant while taking all of the readings. Doesn't flucuate like a VFD air handler. I'll try and post some data later today.

Ralph: The problem I'm having is that the int. shaft speed harmonics and low speed shaft BPFO harmonics are less than 1 cpm difference. Even if the speed is higher or lower, they still follow each other that close.

Don,

I don't know what kind of equipment you are using but I think I would try to take as high of resolution reading as I could then check the higher multiples of both fundamental frequencies you suspect. If the defect has harmonics and you move out to the higher harmonics the separation gets greater. Why would you expect to see only the 4th and 8th order of outer race defect?
Some equipment has the capability to sync time average on a intermediate shaft using a reference on the other shaft by entering the gear tooth counts. On my 2130 it is called a pseudo-tach.

If there are modulation sidebands around the peaks in question, you might be able to tell the difference in shaft speed vs BPFO.

Ususlly BPFO does not show sidebands of shaft speed in the early stages.

If there are input speed sidebands around your mystery peaks, then I would guess it is shaft speed. If there are output speed sidebands around the peaks, I still would guess it being shaft speed since outer race has a tendency not to show running speed sidebands early. If it is FTF sidebands then I would say it is the outer race and possible other problems also. But if the amplitude of the peaks are quite small, it may be hard to even find sidebands at all.

What sidebands are showing around first and second gearmesh, intermediate speed on both or output on 2nd and input on 1st?.

Can you post a zoomed in area of the peaks with high resolution with sidebands marked?

Jim's 2130 sounds like it is the ticket to syncing but I see what you are up against with the frequencies being so close together, but wonder why the outer race would only be showing 4 and 8 times and not any lower frequencies and the shaft speed only having an 8X peak?

High resolution might be the answer, if you can get some 6400 lines or more.. All this "action" seems to be taking place under 100 Hz, so 1600 or 3200 lines would probably seperate every thing into "fine" sharp peaks and show any sidebands that might be there.

Of course you probably have tried this already.

It's just my opinion and I could be wrong.

Ralph, here's some data taken zoomed in with high res. The sidebands are output speed. Remember when I said shaft speed harmonics of the intermediate shaft match some BPFO harmonics? The 9x harmonic FTF of the bearings in question is right on the intermediate shaft turning speed. I've never seen it fall out and match up like this. But then again, Ive haven't seen everything in life either. I have attached some data off of the low speed and intermediate shafts. Also ran the GMF calcs and brg freq calcs.

Looks like I need some help with adding an attachment. ???

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

EXT_BIH.rtf (62 Kb, 16 downloads) BIH

Ohhh, so that's how!

EXT_IIH.rtf (61 Kb, 7 downloads) IIH

Here's the brg freq print out.

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

EXT_brgfreq.rtf (8 Kb, 8 downloads) BRG FREQ

And here's the gmf print out. Can you only just send one attachment at a time?

EXT_gearfreq.rtf (13 Kb, 12 downloads) GEAR FREQ

I ask a lot of questions, huh?

Here is another one.

Ralph, nothing wrong with a question. The 1st 2 sidebands around the 8x is gearbox input speed. On the 2nd question, those sidebands are 2x intermediate shaft speed.

quote:

Ralph, nothing wrong with a question. The 1st 2 sidebands around the 8x is gearbox input speed. On the 2nd question, those sidebands are 2x intermediate shaft speed.

Well, I would at this stage in the game, trust that the signal is coming from the intermediate shaft speed rather than from a bearing defect on the output shaft. But now, I guess one would have to try and figure out why there is an 8x and a 16 x from the intermediate shaft, modulated by output and input speed (which would be normal since both touch the intermediate). Or let her run.

What do the acceleration twf's look like?