I have twelve (12) gearboxes that have an input, two (2) intermediate and an output shaft with spur gears. I ordered one (1) sensor per bearing so I have eight (8) sensors per gearbox. In what direction would be the best direction (horizontal, vertical or axial) to mount the accelerometers to get the most useful data? On the intermediate shafts they can only be mounted in the vertical or axial. If anyone has any input from past experience, I would appreciate the advice.

Thanks,

Patrick

Patrick,

On parallel shaft reducers prefer vertically mounted at the six o'clock position and side exit cables. For one, the load zone is at the bottom. Two, the accels and cables are usually better protected under the bearing housing. Three, on parallel shaft reducers you usually can get a pretty direct transmission path from there. This can change depending on the contruction and type of gearing. The most important part is getting the most direct path from your accel through the bearing support to the bearing. If there are intermediate shafts with no bearing points exposed, locate the intermediate bearing supports and mount the accels directly over the attachment to the outer housing. I rarely take any readings on the housings that are not at attachment points to the intermediate bearing supports. Before you select your locations, take some measurements and compare.

If you need more help, send me a sketch of the drives and I will try.

Good Luck,

8-accelerometers on a gearbox is quite a lot, unless it is very large in size. In general, place accelerometer radial to shaft in-line with tooth-mesh location for radial bearings and use axial orientation for thrust bearings. A good compromise is to ensure all three principal directions are measured and then cover the most critical bearings/gears. Determine critical components based on failure/repair history, detailed vibration survey, or choose high speed shaft over others. A permanet speed sensor is necessary for diagnosing complex gear vibrations.

Why would a service company ask where to put accelerometers after deciding on amount and purchasing them?

What kind of gearboxes are these, Patrick? That has a great deal of bearing[no pun intended ] on the final location. I would not consider advising without knowing this. The arrangement on a Falk 2155Y3 would be nothing like that on a Lightnin 883Q250. [Got both! ]

Also, output & 2nd intermediate speeds need to be considered, a 100mv/g xdcr may not be appropriate at all locations[although IMI does make a 100mv/g LF xdcr]. Clearances at the high-speed shaft input and low-speed output may limit the position or style of a xdcr.

Dana

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

Walt, I appreciate the reply. To answer your question at the end, I was under the impression that a forum was for discussing ideas. I did not post my question as a service company but as an individual that enjoys vibration analysis. Yes, we have purchased the equipment and have already decided their location. I personally was interested in the opinions of the individuals that read this board.

Dana, I have attached a rough sketch of the gearbox. The unit has deep grove ball bearing and spur gears. The input RPM is 1200 and the output is 30. Tell me what you think?

Thanks,
Patrick

Well Patrick, The forum crapped out while I was replying and since it is past my bedtime and I've emailed you directly I won't try to reconstruct my post except to say I mostly agree with Danny Harvey about mounting position since these are using spur gears. Of course that depends on whether or not there is someplace to mount. Parallel shaft boxes are not well known for having a wide flange to mount sensors vertically below the shaft. If you are stuck with a tangental mount, axial might be better. Compare some measurements from the different directions.0

Later,
Dana

How can all the load zones be in the bottom of the bearings.

If the gear weight on the uploaded gears is greater than the vertical component of the gear reaction force, then the load zone will be in the bottom even for an uploaded gear. On this type of box the gears are pretty large and heavy so it is possible that all the load zones would be in the bottom. Possible but not likely.

Even so the other advantages of installing the transducer at 6:00 on each bearing out weighs missing the load zone in my opinion, and as a bonus you'll be at or near the load zone for at least all the down loaded gears.

There are three static forces from gears acting on the shaft bearings:

1) Torque reaction force that is perpendicular to line connecting mated gear centers
2) Gear separation force that is pushing gears apart in line with the mated gear centers
3) Gravity or mass force from weight of gear and shaft with a direction that depends on shaft axis orientation

Proximity probe gap measurements on journal bearings in a 40 MW Steam-Turbine reduction gear indicated that gravity was the smaller of the three forces in that situation.

I generally find that it is a waste of time trying to "Guess" where a bearing load zone is located. Ultrasound or shock-pulse measurements can provide a indication of load zone location better than low frequency vibration measurements. Is this paragraph controversial?

I'm sure this will be controversial, but I don't think it's as important where they are, as much as "they are". Most gearboxes "ring" like a bell. As long as you have one on, and it's the same accel, in the same place, with the same parameters, and the same collector, I think it will help a great deal.

Aside from that, do you have any failure history on these gearboxes? In other words, which gear or bearing normally goes out first?

In my experience, right angle sections and vertical thrust bearings are usually the culprits - but I'm sure it's different for others.

Steve,

You are probably correct about the possible, but not likely idea that the load zone is on the bottom. I was speaking from my direct experience with several really big drives that I have mounted accels on and given the configuration of the box the choices were directly above or below the bearings (as is the choice on most parallel shaft reducers). A comparison of readings showed the amplitudes to be measurably higher below, so it was assumed that this area was at least closer to the load zone than above.

Walt and Stan,

Sorry but I find no controversy in either of your statements.

Walt,

Would you mind taking the time to make a sketch illustrating the torque and gear separation reactions that you describe? I think I understand what you are saying, but I would like to make sure.

Stan,

I think that the right angle sections of gearboxes wear first because they are usually an even ratio, say 5:1, that gives an assembly phase. I have always been taught to avoid even ratios for that reason but it seems to be pretty common among gearbox designs. Any idea why?

In addition, I agree that it is most important that they are somewhere. For properly designed and applied gearboxes, trending the wear is most important and it should be pretty boring stuff.

BTW, this may be an application where CSI's Autostat program is very useful.

Good Luck

Here is an article I wrote a long time ago in a galaxy far far away, but it describes gear seperation and torque forces that result in what I called gear reaction force...

http://www.bently.com/articles/articlepdf/dec197.pdf

Steve,

Thanks for pulling up your fine article. The only force vector not shown on Figure 1 is from gravity. It would depend on gear orientation and weight of gear and shaft. Depending upon the magnitude of the other two forces, it may not be dominant. Almost forgot another force; helical gears produce an axial force in-line with shaft that is either offset by an opposite gear (double helical set or herringbone) or by thrust bearings.