Hard Cutting/Skyving of case hardened gears

[gearcutter]

I found this on another forum and am interested in this forums views as we are seeing more of this type of finishing process. I've posted the original question along with my reply.

Ivan D.s question:
From Time to time we see unexplained tooth surface failure on large cyclo-palloid spiral bevel gears that we use on marine transmissions. The gears are case hardened and hard cut. The hard cutting is common for finishing of small hypoid and spiral bevel gears for automotive drive axles. But on large gears, hard cutting looks to be resulting in subsurface damage because of the very high cutting forces. Some of the large hard cut gears fail due to tooth surface damage much sooner than they should. We know that hard turning is not recommended for critical parts because an alternative grinding can be used. But how to avoid damage from hard cutting on large spiral bevel gears? We deal with AGMA class Q12 spiral bevel gears 1000mm – 2150mm in diameter. Would somebody know an alternative tooth finishing process that would not damage the hardened case as much as hard cutting.

Ron V.s reply:
I don't know about the cyclo-palloid system but I do know that the Gleason system can be finish ground.
The only other thing I can think of would be to lap them after heat treatment. I've had this done on spiral gears up to 800mm diameter with good results............I'm not sure about gears as large as you are quoting.
The reduction in resistance to surface fatigue, due to the hard cutting process, that you are seeing is a cause for concern as the process seems to be picking up in popularity.
From your results; do you think that these gear sets should be de-rated?


Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[gearcutter]

I've conducted several failure investigations where the primary failure mode occured as a direct result of this type of finishing process.

The photos are from one of those investigations. They're quite large so give them time to load.
The pitch of these rolling mill pinions was 18 module.
Notice the micro pitting (frosty/dull grey) developing on the high points of the generation flats left behind by the skyving hob. The micro pitting has then progressed to macro pitting.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[CoryPad]

You've mentioned the other traditional options, grinding and lapping.

 

[TVP]

I would not call this "unexplained tooth surface failure". The rough surfaces, etc. produced by hard cutting will lead to premature failures, it is as simple as that. Some subsequent processing must be used to improve the gear surfaces, such as grinding, lapping, or polishing.

 

[CoryPad]

TVP's mention of polishing shook my memory loose on a newer surface finishing technique - ISF, or isotropic superfinishing, which is a chemical polishing techique.

http://www.remchem.com/page3-49/GearingandBearings

 

[tbuelna]

gearcutter,

Your pictures seem to show spalling damage on both the active profile and tips. In addition, there seems to also be some brinnelling, which would likely indicate debris passing through the mesh.

Spalling of the tips is usually the result of case fractures caused by through hardening of corners that had inadequate relief prior to case hardening.

Some of the pictures also seem to show evidence of material transfer and fretting. Fretting would be the result of mesh contact caused by scoring. The oil film Lambda value was not sufficient to support the loads, and the contact condition became boundary (ie. metal-to-metal).

Hard shaping of case hardened gears is a legitimate process if done correctly. But as with any other finishing process (like grinding), the high forces involved can generate surface heat which will easily de-temper the case structure if not done correctly. That is why hardened and finish-ground surfaces undergo Nital etch inspections.

Honing or ISF processes don't remove much material, so there is not usually enough heat generated to produce case de-tempering.

Regards,
Terry

 

[mfgenggear]

it appears to me as the case harden surface where reharden or tempered. It would be benificial to Temper(Nital)Etch,in addition Mag & demaged for indications.

 

[mfgenggear]

GEARCUTTER

May I ask what was the case depth?
& what was the the stock removal.

Just for my information.

 

[gearcutter]

mfgenggear

Effective case depth after machining = 1.80mm
Stock removed = 0.30mm - 0.40mm

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[gearcutter]

After having attended one of Ray Drago’s Gearbox CSI courses recently; I am re-assessing my diagnosis for this gear set and have also, once again, realised how much there is to learn in this field.
There is evidence of misalignment causing heavy bearing towards the left hand end of these pinions.
Due to the localised contact condition this creates; the ability of the teeth to handle both bending and surface loads, in this area, has been significantly reduced.
In this case; excess loading has caused significant tooth deflection resulting in evidence of tooth tip interference or ‘hard line contact’ below the dedendum (this can be easily seen in photos 2 & 3). Along with the misalignment, this causes excessive surface loads at the non involute contact points of which the result is sub surface fracturing which then leads to pitting. As a result of the sliding contact conditions gear teeth experience as well as hydraulic propagation; the pitting then progresses to spalling which is evident in the photos. Spalling in gears is a type of macro pitting but is unique to contact conditions that experience high amounts of sliding; note the classic fanned out appearance on photos 2 & 3. Spalling in gears is different to spalling seen on bearings where the generally the contact conditions are rolling in nature.
The photos clearly show how cracks are travelling, both surface and sub surface, from the dedendum area to the tips of the teeth causing material to be ejected along the way.
Also notice the dull grey or frosty appearance that can be seen on the surface of the active profiles. This is known as micro pitting or grey staining. The micro pitting is concentrated along axial lines traversing the tooth width. These lines are high points left over from the generation process. The gaps between each line are known as generation flats and are left behind by the skiving hob. There is evidence that some of the surface degradation is as a result of micro pitting progressing to macro pitting.


To “tbuelna”; Terry, you were correct in diagnosing the large fan shapped craters as spalling. Thanks for the response.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

http://www.aussieweb.com.au/email.aspx?id=1194181

 

[unclesyd]

I think you should look at superfinishing for your gears. I have heard of some outstanding results with a couple of gear sets.

http://www.remchem.com/articles1-9/TheEffectofSuperfinishingonGearMicropittingPartI

http://www.cryogenicsuperfinish.com/pdf/AAMPaper 9.27.07.pdf

Do a search on Google on super finishing gears.

 

[tbuelna]

gearcutter,

It's too bad you couldn't have spent a couple of months listening to Ray Drago, instead of a couple of hours during a course. When it comes to high performance gears, his accumulated knowledge is priceless. It's too bad that there are very few young engineers coming along that he can pass that knowledge on to. It's the type of knowledge that you can't get from a textbook.

But back to those gear photos, all of the failures appear to be spalls close to the root or tip. Spalls are almost always due to excessive contact stress, and not due to bending stress. A bending stress failure would typically result in a crack that propagates from the tooth root area and goes through the tooth (or possibly through the rim if the rim section is too thin).

My guess is that those root and tip spalls might be due to inadequate tip reliefs and index errors causing excessive dynamic contact loads as multiple teeth try to load share when they enter and exit the mesh. The original post stated that the gears were hard shaped and had a quality requirement of class 12. I don't know much about hard shaping, but I can't imagine that you would achieve a class 12 gear with any shaping operation.

Regards,
Terry