Gear Measuring - Determining when a gear has had too much wear

[jfiorin]

Hello all,

I am trying to find out when a gear has faced too much wear. In other words, I want to be able to determine whether or not a new gear is necessary for an assembly. I would assume there is a standards chart or something that would let me know what the standard backlash is for certain gear sizes, or if some gears have a standard tooth thickness, but I have been unable to find any kind of document/link online. I have been researching gears, finding out about their nomenclature and how to measure them using a Vernier Gear Tooth Caliper, without much luck.

In summary, is there any way I can learn more about gears and their standards, if they have them? Or knowing when a gear is shot?

Thanks for the help!

 

[Jboggs]

I think the standards you are looking for probably vary from industry to industry. For example, I'm confident there are standards set by SAE or by individual manufacturers for acceptable backlash, wear, etc. on the gears in an automotive transmission. They would be based on the loading, expected life, and other factors. On the other hand, gears in a hand cranked winch off the shelf at Grainger see a totally different world, and therefore require different standards. You didn't mention your application. Have you checked with AGMA (

http://www.agma.org)?

They might be able to help.

 

[tbuelna]

jfiorin-

A properly designed gear set should not experience "wear" in the classical sense. While a new gear set may produce a small amount of debris during an initial run-in period from shearing of microscopic surface asperity tips, once the gears have been fully run-in they should experience no further abrasive wear under normal operating conditions.

Properly designed gear sets are usually limited by tooth bending or surface contact fatigue life. By the time your gear set has accumulated enough abrasive tooth wear to produce any measurable change in tooth thickness or backlash, they are definitely "shot". But simply replacing the gears will not likely resolve the root cause of the problem. If your gear set is experiencing abrasive tooth wear, you need to address the problem by modifying the design of the gears and/or lube system. As Jboggs noted, the AGMA documents provide an excellent reference for designing gear systems.

Good luck.
Terry

 

[BrianPetersen]

The standards for acceptable gear condition will be very different for a 10,000 rpm motorcycle transmission, than they will be for an 18th-century flour grinding mill. It will help if the original poster gives us some clue about the application.

 

[jfiorin]

Thanks everybody! It seems I was not specific enough. What my company does is repair and replace existing assemblies, and one of our specialties is gearboxes. Essentially we get a junky gearbox, fix it up, and send it back. If we knew more about gears and when we don't need to buy a new one (thus making the cost of the job lower), it would be better for our customer AND we'd learn new, more efficient techniques.

@Jboggs- I have checked with the AGMA, and is one of the reasons I posted in the first place. There's a broad spectrum of standards they offer, but I didn't want to make a choice and end up wasting money on a set of standards that don't pertain with what I need.

@tbuelna- you can see from what I said about the company above, that is precisely what we do. We see what causes a gearbox's wear, then we fix it. What I need to know is when a gear is still usable, or if we need an entire new one.

@Brian Petersen- Our applications vary greatly, but a basic RPM range we work with is 1100-1700 (we are certainly not limited to that). As for strengths/ratios, we typically work with steel and paper mills, which also tend to vary (hence the medium speed on the RPMs). We've worked with gears with high enough ratios that the output hardly even moves when the input is rotated.

I hope all this helps. Thanks again!

 

[BrianPetersen]

In something like that, any gear that is not visually in perfect condition on the tooth mating faces, needs to be replaced. Perhaps on lower-speed portions of the gearbox, you might get away with minor chips, but the problem is that the lower-speed portions of a gearbox are also the higher-torque portions, and those surface imperfections can become origins of fatigue cracks.

You presumably wouldn't have the gearbox in for repair unless something broke. Broken stuff inside a gearbox tends to send metal shrapnel everywhere inside, which can end up jamming itself between gear teeth and getting into bearings and wrecking all sorts of parts that weren't directly involved in the original failure. If maintenance is as per usual - keep going until you absolutely cannot go any further - I'm sure you've seen the results.

I'm working on a motorcycle transmission in my own shop right now, in which an engagement dog broke off and ended up in the oil pan without damaging anything else, and there was no outward indication of any fault with it because the other two continued to carry the load! Those are the exceptions. I've had the same type of failure result in the bike having to come in on the recovery truck (at a track day).

 

[jfiorin]

@Brian Peterson
That's a valid point. For some gearboxes, however, things like backlash and tooth length aren't as critical as other gearboxes correct? So we could trim down the teeth on the gear to remove blemishes, while shortening the teeth to make the teeth stronger (or at least the torque on the teeth less). Obviously we would not be able to remove much, but for minor abrasions that could save money. Would that be possible?

You lucked out there, we've seen what one missing tooth can do, I can tell you!

 

[rob768]

If you don't want a regular visual inspection, you may consider vibration measurements. You will see an increase in the harmonics of the gear mesh with increased wear. It is still difficult to set a rejection limit for that, but you might combine the trend with regular visual inspections.
Other methods are oil samples.
No matter what the condition: if it stable under the applied operating conditions, it is good.
"better" is only relative.

 

[geesamand]

"Thanks everybody! It seems I was not specific enough. What my company does is repair and replace existing assemblies, and one of our specialties is gearboxes. Essentially we get a junky gearbox, fix it up, and send it back. If we knew more about gears and when we don't need to buy a new one (thus making the cost of the job lower), it would be better for our customer AND we'd learn new, more efficient techniques."

My company does this too. Our basic approach is: new or like new gets full warranty, slight wear gets much shorter warranty, worn stuff gets no warranty. Customer gets to choose.

The real answer always depends. What is the failure mode of the gearing? Is it surface failure? Are your customers sensitive to noise and vibration, or do they only care that there are enough teeth intact to keep the output shaft turning?

David

 

[rob768]

in that case you might need a more accurate method. We sometimes make replicas of a surface which we take home for roughness or geometrical measurements. We use a polymer, and this gives a very accurate "negative" of a a surface (in your case: a single tooth).
A substance to do this with is Technovit 3040, but there are more, and they all do the job