Hey guys, I've got a Hoffman blower being turned by a Falk gearbox by way of an electric motor.

Turning speed for the motor is variable, but for this collection it was 1139 rpm. Through the gearbox, which only has two shafts and one gearset, the output is 694 rpm.

I've got a twice per revolution impact in the time waveform at the inboard input shaft gearbox point and at the inboard output shaft gearbox point that matches twice the speed of the output shaft (~1399 or so). So basically, it seems that there maybe a tooth broken on either side of the low speed gear and it is impacting at twice the run speed of the low speed shaft since this doesn't match any other known faults. Is this a reasonable assumption? Is does seem somewhat unlikely that two teeth would break on either side of a gear. What's the chances right?? Anyway, what do you guys think??

I am attaching a Word document with some waveforms and spectrums.

This_is_the_time_waveform_from_the_inboard_input_gearbox_shaft_horizontal_point.doc (90 Kb, 78 downloads)

Bwze,

For a Falk gearbox with about 2:1 ratio, I would expect the tooth mesh frequency to be about 27,000 cpm (about 24xSS). This comes close to the high peak in spectrum and not your TM markers. Either I misinterpreted your data or there is a mistake identifying TM.

Is the "Hoffman" blower a rotory positive displacement blower with two lobes on each rotor? If yes, then the 2x output shaft speed is from blower and not a gear fault. If not, then?? I doubt that there is a gear tooth fault. If it is easy to shut machine down, then open gearbox inspection cover and look for gear tooth faults. Otherwise, make sure you know exactly what is in the machine before making a diagnosis.

Walt
w_f_strong [at] msn [dot] com

Walt, tooth count on the input pinion is 64 and on the output gear it's 105. I've put the spectrum in this post. At an input speed of 1139, this places gear mesh at ~73,000 cpm which is what is shown below. If you look a little closer, you can see the first GMF lines up perfectly. There is also some energy present at 2 x GMF which may be from this impacting. All the energy at the lower end of the spectrum is spaced at about 695 cpm which as I stated in the Word document may be sidebands around the gear natural frequency. The peak list for the spectrum below is showing 92% of the energy is nonsychronous and since nothing in that part of the spectrum is lining up with a bearing fault, I can only assume this be the case.

BWZE,

OK, so you confirmed the tooth count, but I still doubt that 2xLS is from broken teeth on low speed shaft. You have verified that most of the vibration is nonsynchronous to high speed shaft, but what about being synchrounous with low speed shaft? What about my comment regarding the blower? Is a bearing fault on your suspect list?

Walt

From the peak list for the inboard bearing on the output shaft, 98% of the energy is non-synchronous. I was under the impression, however, that impacts from cracked teeth don't tranfer well to the frequency spectrum.
Yes, it is a positive displacement two lobe blower. I don't see this impact pattern at the blower at all though.
I have added the bearings from the blower to the gearbox points where I'm seeing these impacts, but they don't line up at all.

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

Bwze,

I have done a lot of vibration work on these blower trains, including torsional vibration measurements. A normal vibration is 2x blower speed in the gearbox from the torque reaction, so it modulates the TM frequency (creates sidebands). The amplitude is sensitive to blower load (pressure), so only compare data for the same load. Does machine have Falk Grid-Flex couplings? If yes, then be sure couplings are in good conditon and full of appropriate grease. This will help to dampen torsional vibrations. I'll bet a beer you don't have ANY cracked gear teeth!

Walt

I am in agreement with Walt. Also, as you have mentioned Bwze, it is very unlikely to have two cracked teeth spaced exactly at 180 deg. The two impacts per rev are more likely due to impact as result of backlash occurring because of load variation, 2 times per rev. Apparently, this peak at around 13,942CPM is a resonance frequency of one of the gears. This could be verified separately in order to confirm the above guess.

One more suggestion. Switch units in the TWF to velocity. It may give you a better picture as to what is happening in the gear box..

Take a PeakVue measurement as well. Time waveform impacts from broken teeth show very well there. Sounds like Walt is right on though.

OK, thanks for the input guys. This is what I was looking for. Yes Walt, it is a Falk grid-flex coupling. I'll just keep watching it and let maintenance know that they need to check the coupling for lubrication when the blower comes down next time.

With PeakVue isn’t it possible to auto correlate the waveform and swap it over to the polar plot view and actually be able to see a tooth defect. I’ve seen the perfect representation of gear teeth as long as the rpm is exact and would imagine that if there were a defective tooth it would show up as some form of anomaly on the defective tooth.
I might be out to lunch with my thinking but seems it would work. Works for showing a BPFO.
Can anyone verify this?

Wally K