Hi, I was wondering if I can get some opinions of this subject.

A scheduled overhaul was made on the gearbox of a 10 inch rubber extruder. The second gear train was replaced, in other words, the pinion on the second shaft and the bull gear on the third shaft. The following picture shows the measurements points.


Before starting the equipment, a contacting test was made to see the mesh of the teeth. A problem was found on one gear, as shown in the next image. The decision was to run the gear unit even with this abnormality.


At the startup, the gearbox showed no problem with no load. But when we started applying load on the gearbox, a knocking was heard. The next vibration plots show the impacts. The frequency corresponds to the 3rd shaft rpm. The spectrum analysis shows the 2nd gearmesh with sideband spaced at the 3rd shaft running speed. All the data pinpoints to the 3rd shaft bull gear. Knowing this, we think it is possible that the discontinuity of the charge transfer could be the cause of this vibration.




This seems hard to determine when only analyzing amplitude though… Any thoughts or recommendations?

Thanks

(Sorry about the big pics, this is new to me)

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

Pictures unavailable - just a box reading "Image Expired" "The FreeImageHosting.com"

John,
That's strange, I am seeing the diagram, the picture of the gear and the spectrums and waveforms. Wonder why you aren't.

Ronnie

Actually, I found a different host after reading John's post(pics were down)... Changed it a few hours ago. Thanks for the warning.

Any opinions?

Looks pretty bad to me wigwag. Did you collect data at different points for the first and second spectrums that you have posted. The first spectrum and waveform is dated Jan 4, 06 and the amplitude is fairly low, but the data from the second spectrum is dated two days later and things look different. Why are you looking at your data in acceleration in the second plot that you have posted? With this fmax, velocity will be a better indicator of condition. I see that it was collected separate from route data due to the Analyze Spectrum tag at the top right of the spectrum. Did you mean to collect the data in acceleration. The .0222 Gs rms overall that you are seeing in this plot corresponds to approximately 1.5 in/sec in velocity at the indicated speed. This really seems unexeceptable to me.

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

Pat,

You may have already resolved this problem, but anyway, here is my opinion.

This knocking definitely does not "sound" good. The sharp impactive pattern rather then amplitudes is more indicative of a problem (assuming impacting is coming from inside of the gear box, not from the outside). One tooth is clearly either chipped off or cracked, less likely worn out. There is also a possibility of a bearing IR being cracked. To see real amplitude I suggest taking PeakVue reading.

You also have looseness in bearings indicated by families of sidebands. It appears (although not readily clear from the first spectrum) that you have another set of GMF (7430 CPM) with sidebands also showing symptoms of looseness and possibly misalignment. (I am not clear what this wide blue line on a tooth in the picture represents.)

I'd be interested to see you findings.

We didn't find the problem yet. The only abnormality we found is the contact of the teeth shown in the upper picture. As I already mentionned, there is no broken tooth, this gear is 3 weeks old. We were only able to identify that the knocking is coming from the 3rd shaft. I was just wondering if there are some inspection that can be made from the outside. We think that the manufacturer who made the gear did a bad job.

I took measurements on all the others shaft but the knocking is visible only at point G and H(see picture above). Also, you can hear and feel the knock when you are beside the gear unit.

We look forward to replace the whole unit at the July shutdown. It's not possible to open the unit again, it's a critical equipement for the plant and there is no backup.

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

Have you computed the torsional resonances? If so, are there any potential forcing functions?

Torsional testing?

If I understand right you changed a set of gear? . Gears are supposed to be matched. If not you will some wearing and possibly knocking.
Just a thought.

I think you've got all your answers already in the 1/rev impact, and the picture of the 1 tooth with the funny contact pattern.

It'll either last until July, or it won't. If you really can't take it out before then, make sure your new parts are quickly available should the gearbox decide it's own fate.

Good luck.

Wigwag,
I'm going to have to side with tthew on this ... get your parts in while your have a chance.

If your managers want more proof before buying parts, draw some oil samples and have them sent to an independent analysis lab.

Take Care,

we changed the set of gears, which means the gears on the 2 and 3 shaft.

we already orderes the new gearbox for july. Until then, we are taking measurements every week and oil analysis every week. As a problem never comes alone, after start up, we found 1500ppm of water in the oil. We found that the oil cooler failed

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