1 - information on gearbox vibration analysis 2 - gearbox vib problem 3

[electricpete]

1 - Can anyone recommend sites with good information on gearbox vibration analysis?

2 - I have a speed increaser gearbox driving a centrifugal pump which is showing increasing acceleration levels 8g's overall.

Time waveform shows impacting 20-40g's peak-to-zero at an interval corresponding to motor speed.

Looks to me like sidebands around gearmesh frequency are fairly low. I don't think I have the capability to get data at 2*GMF.

The cookbook analysis would be a chipped/worn gear tooth on the bull gear causing an impact once per revolution when it engages. Can anyone suggest any other possible causes?

 

[electricpete]

Any sites?
Thx in advance.

 

[electricpete]

I ask again, can anyone help recommend useful sites on this subject? (there may be a star in it for you! ;-) .

I have done some searching myself but mostly comes up with commercial sites with a lot of marketing and very little technical info.

Thx.

 

[nbucska]

Hi Pete: check this at <

http://www.uspto.gov>

NASA patent: using CEPSTRUM analysis for bearing defect...
(paraphrased !!! from memory )



<nbucska@pcperipherals.com>

 

[bentlytim]

Electricpete,
Hello. Sorry, do not have any good websites for free technical information. But here are a couple of resources:

Technical Associates of Charlotte – Wall chart is very good as a guideline. Courses are very good.
Vibration Consultants – Gear Analysis Handbook is supposed to be very good. I have not bought a copy yet, but I know several people who have. Courses can also be pretty good.

Although, with the identification of the impact at motor speed I would suspect that you are correct with a problem on the bull gear. If the change in the time waveform was gradual and is increasing I would suspect spalling or severe wear of the bull gear. If the change happened suddenly and then has stayed pretty consistent over time it could be a broken, chipped, or cracked tooth. I would also suspect wear or spalling if the noise floor of the spectrum is changing or increasing.

Typically individual (or even groups) of teeth will not show any significant change in the gear mesh. I would also like to suggest that you consider being able to go out to 3x of gear mesh for measurements. There have been documented cases that 1x and 2x did not identify gear alignment problems, but 3x did.

 

[chjura]

Electicpete,
I am just currently working on large gear/gearbox vibration monitoring procedure. I had also problems finding the appropriete standards and literature about it.

According to &quot;The Vibration Monitoring Handbook&quot; from Charles W. Reeves, externally mounted accelometers monitoring velocity or acceleration are best suited to detect the whole range of gearbox faults.

This book gives following satisfactory and maximum allowable levels for gearboxes at fans and pumps:

Satisfactory:
velocity: 10mm/s (RMS) 25mm/s (true peak)
acceleration: 70m/s2 (RMS) 250m/s2 (true peak)

Max. allowable:
velocity: 15mm/s (RMS) 40mm/s (true peak)
acceleration: 150m/s2 (RMS) 500m/s2 (true peak)

API 613 gives following maximum allowable casing vibraitonlevels (for horizontal offset gears, for vertical the allowable levels are twice those):

Velocity (10Hz-2.5kHz): 4.07mm/2 (unfiltered, peak)
2.5mm/s (filtered, peak)

Acceleration (2.5kHz-10kHz): 4.0g

It is also possible to monitor the external vibration of the gearbox using time domain averaging technique by Steward Hughes (this is also mentioned in &quot;The Vibration MOnitoring Handbook, but not so in detail). This technique is very usefull for detecting the gear faults.

Following standards recommand the necesary vibration pickups for the gears:

API 670, Appendix H-2 -Typical System Arrangemnet Plans for a Double Helical Gear,

DIN ISO 13373-1, Tabbelle A1, Grossgetriebe mit Gleitlagern (unfortunatelly available only in german)

Other standars that could be interesting are:
API 613, ISO 7919, ISO 10816,