Too much EP additive leading to accelerated wear?

[geesamand]

I have an industrial enclosed gearbox that is showing accelerated wear of all gear meshes, particularly the spiral bevel set on the output shaft. We see both micropitting and pitting on the high speed mesh. No evidence of water in this case. Contact patterns reach all corners of the teeth, indicating the crowned areas are completely worn away. The bearings don't look bad, just light rolling pattern marks.

We are concerned that an automotive gear lube was used instead of AGMA-type EP gear oil. Viscosity was matched well but I'm told that the EP chemistry of the automotive oil is much more aggressive and could lead to early pitting and accelerate the wear rate. We have a sample of the reclaimed oil and it seems quite low in viscosity, so we're having it tested as a double-check.

This gearbox design has an excellent history and in this case very good service factor (>2.0). Carburized and ground steel gearing. No evidence of shock load and the driver is an electric motor. So we're really left with the oil formulation as our root cause.

Has anyone else had similar experience with overactive EP additive packages?

David

 

[gearcutter]

No, I haven't seen this before..........have certainly read about it though and can say that it's a very slow process which leaves tell-tale signs.
Any chance you could post some good quality photos?

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

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

 

[geesamand]

I can look for permission to do so, but the gears have classic micro-pitting and pitting over the whole face. Compared to normal pitting wear I cannot tell the difference. They look like they could be 20 years old and run to failure.

I spoke with the lubricant manufacturer and we're taking steps to ensure that the lubricant we collected from the returned gearbox matches their VOA analysis information. The reclaimed lubricant was markedly clean considering the level of damage inside the box. Either they were changing oil on very short intervals, aggressively filtering it to stave off failure, or...something else happened.

David

 

[gearcutter]

but the gears have classic micro-pitting and pitting over the whole face. Compared to normal pitting wear I cannot tell the difference. They look like they could be 20 years old and run to failure.

From what I understand; micropitting and pitting (macro pitting) are "normal pitting wear" damage modes.

From your posts; you seem to be talking about two different failure modes.

Micropitting is a hertzian fatigue damage mode. In fact, using more chemically aggressive EP additives could help reduce hertzian fatigue problems.

Chemical attack from oil additives is usually associated with polishing wear.

This is why pictures can be so important as, at this point, we can only reply based on your version of the story.

Some things to consider when attempting to reduce micropitting, I've found number 10 to be the most important for helping to reduce future problems with any gearbox:

1. Grind/hone/ polish gear teeth.
2. Avoid shot-peened flanks.
3. Make hardest gear as smooth as possible.
4. Make pinion 2 HRC point harder than gear.
5. Use oil with high micropitting resistance (EP additives).
6. Keep lube cool, clean & dry.
7. Use high viscosity lube.
8. Use high speeds.
9. Coat teeth with phosphate, Cu, or AG.
10. Run-in with special lube and controlled loads.

Source; Robert Errichello - GEARTECH




Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

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

 

[gearcutter]

Here's an example of polishing wear.
Remember that this image shows a damage mode that can be as a result of chemical attack.
Something to keep in mind; if oil additives are in fact attacking the gear flanks, the oil would not look clean. By-products of the chemical reaction cause the oil to change colour.

Polishing Wear

Image by Robert Errichello - GEARTECH.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

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

 

[geesamand]

EP additives are generally considered to resolve the issue of scuffing / scoring. Automotive gear lubes are designed to provide very high levels of anti-scoring EP additives, up to 6.5% by weight. Automttive applications do not consider micropitting as a concern, but in industrial applications micropitting is a failure model. Industrial gear oils have a less aggressive EP package (<1% by weight) but improved anti-micropitting performance. For example, Mobilgear SHC and SHC XMP, in which XMP has less aggressive EP additives to maximize anti-micropitting performance, and both are less aggressive than automotive lubes.

Sources:

http://www.engineersedge.com/gears/gears_lubrication_general.htm

http://www.machinerylubrication.com/Read/417/micropitting-gear-teeth

I have since yesterday learned that our service factor may be 1.2, not 3.0. Obviously this seems to be the major issue, as any deviation from ideal operating conditions will cause a failure. Poor choice of oil may be one of them. I also note that the references about too much EP indicate this is prevalent in high load applications.

Here are pictures of the gear surfaces:

http://files.engineering.com/getfil...-3ef3-46fa-b87c-3ff7a4710977&file=gearing.zip

Obviously the gearbox was run to failure, so there are several levels of failure modes.

Ron, thanks for your list of considerations. Here is where this example stands:
1) Teeth are ground on the helical, lapped on bevel. AGMA Q14 for helical.
2) No shot peening anywhere.
3-4) Both gear and pinion of helical are equal hardness, carburized and ground to 32Ra min.
5) Oil was high EP (anti-scuffing) meeting GL-5. Unknown anti-micropitting, as micropitting is not a recognized risk in automotive applications.
6) No information
7) Lube was on the high end of viscosity for this application. However, the lube used was a multi-weight and our viscosity recommendations are based on straight-weight industrial oils.
8) Speeds were relatively high - motor was 1800rpm so maximum for this application.
9) No coatings. Our company has not used coatings to my knowledge.
10) Our gear designer advises me that run-in is not really applicable to carburized/ground gearing of high accuracy.

Thanks for your discussion.

 

[gearcutter]

geesamand - As yet; there is not enough evidence available to back the claim that additives can be a cause of micropitting or the promotion of it.

As Errichello points out in one of the links that you provided - "Experiments show widely varying, and sometimes conflicting, results for the influence of additives on micropitting. For example, some tests show antiscuff additives containing sulfur and phosphorus (S-P) promote micropitting, whereas other tests seem to prove S-P additives impart resistance to micropitting."

Anyway, based on the images that you've provided; I seriously doubt that overly aggressive oil additives have caused any one of the various failure modes shown.

It would be interesting to know what the result of your investigation finds. Please let us know, as the forum rarely gets to hear the outcome of these investigations.

Ron Volmershausen
Brunkerville Engineering
Newcastle Australia

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

 

[mfgenggear]

Geesamand

Quote:
Ron, thanks for your list of considerations. Here is where this example stands:
1) Teeth are ground on the helical, lapped on bevel. AGMA Q14 for helical.
2) No shot peening anywhere.
3-4) Both gear and pinion of helical are equal hardness, carburized and ground to 32Ra min.
5) Oil was high EP (anti-scuffing) meeting GL-5. Unknown anti-micropitting, as micropitting is not a recognized risk in automotive applications.
6) No information
7) Lube was on the high end of viscosity for this application. However, the lube used was a multi-weight and our viscosity recommendations are based on straight-weight industrial oils.
8) Speeds were relatively high - motor was 1800rpm so maximum for this application.
9) No coatings. Our company has not used coatings to my knowledge.
10) Our gear designer advises me that run-in is not really applicable to carburized/ground gearing of high accuracy. Unquote:

#10) as stated by you gear designer is incorrect. there must be a break in. and coatings that Gearcutter advised help with that process. it is used for the initial break in only. or excessive damage occurs. many precision aircraft gears we fabricated are coated with copper & silver for break in.

for as drastic as it was described. there was excessive wear on those teeth. the first mistake in my opinion is to jump to conclusions before analyzing all data.
for example for gear teeth to be on the edge definitely caused the final failure.

Mfgenggear

 

[mfgenggear]

Geesamand

first verify the bevels center distance and
the patterns are correct.

secondly it would be beneficial to pull parts from stock if there is any and do a destructive test for met lab properties. case depth & case hardness. any austenitic grain visible. did to transform correctly to martensitic.

the helical involute needs to be verified and to make sure it is acceptable.
are the M.O.W. correct, are the leads correct.

are the parts all dimensionally correct.

and the list goes on.

Mfgenggear

 

[tbuelna]

geesamand-

Thanks for the pictures, they are helpful. However, I don't see any evidence of the "classic micro-pitting" you described in an earlier post. If we exclude what appears to be indentation type surface damage that occurred as a result of debris passing thru the mesh contact, the surface damage mostly seems to be adhesive type failures like scuffing, along with a few locations of macro-pitting/spalling.

Adhesive type surface damage like scuffing is caused by insufficient oil film thickness at the contact to prevent mechanical adhesion between the mating surface asperities. There are a number of factors that can influence scuffing, such as lubricant viscosity, lubricant temperature, flash temperature rise in the oil film due to contact sliding, roughness of the mating surfaces, or lube oil additives (like EP additives) that inhibit mechanical adhesion between the contacting surfaces.

Micro-pitting and macro-pitting are both forms of surface damage resulting from Hertzian contact fatigue. Macro-pitting damage is usually localized and much deeper, typically the result of a shear fracture initiated at some subsurface flaw that propagates to the nearest surface and produces a spall. Micro-pitting damage is usually widely distributed and very shallow, and is typically indicative of insufficient strength in the material surface for the pressures produced in the hydrodynamic oil film. EP additives are only really helpful for limited conditions where modest scuffing might occur, such as start-up. Normally, gears/bearings using oil film contacts are designed with extremely generous margins when it comes to scuffing since the scuffing limit can be very difficult to accurately establish due to the large number of variables involved.

Lastly, you mentioned that your helical gear set was AGMA Q14, which is very high quality. But simply having high precision gears is of no help unless the rest of the gear drive is designed properly. For example, you also noted that the flank surface roughness on these AGMA Q14 gears was Ra32, which is not very good for a ground gear. Obviously you would need to finish grind a carburized gear to get Q14 precision, but I would expect to see a finish ground surface roughness of Ra8-10 rather than Ra32. This difference in surface roughness on the tooth flanks can have a huge effect on scuffing limit.

Hope that helps.
Terry

 

[geesamand]

Hi Terry,
I think the challenge I'm faced with is that the damage is so severe that the initial failure modes have been wiped clean of the teeth. Even the pinion shows step wear that corresponds with both edges of the gear teeth and that can only happen when the several thousandths of material is worn away. If micropitting was present, it's gone now with the material removal that's taken place.

Generally speaking, for industrial gearing, we see micropitting as the long-term failure mode. Does it not lead to more severe wear forms? (Because if it was harmless it wouldn't be something we talk about.)

Our grinding finish is better than 32Ra. I do not have an as-ground number for it. The gearing was actually ground to Q14, in excess of our drawing requirements. In any case this is a proven design with good history and this kind of failure is extremely unusual. Our quality for these components is very good, and because this failure involved two sets of gearing made from completely independent supply channels, we checked surface hardness, have had no other issues with this design, suggests that further quality checks are unjustified.

Mfgenggear,
The addition of coatings is a whole new thing for our industry and worthy of a whole other discussion. Your point about the mixed results re: micropitting resistance and S-P EP additives is noted.

I will try to update later. We are primarily focused on a combination of low service factor, if higher loads are present, lubricant viscosity / temperature, possibility of an incorrect lubricant being used, etc.

David

 

[mfgenggear]

Geesamand

WE had the same issues we some gears. it turned out as Terry & Gearcutter have advised.
A) it needed a high pitch line modified involute.
B) for high torque, & high RPM the micro finish had to be 8 Micro or better.
part required super finish, gear grinding and honing was not good enough.
the high micro finish was required to prevent micro pitting. do the the high
Gear Cutter Quote "Micropitting is a hertzian fatigue damage mode." Unquote.

which is also to my experience is a factor.

In our case all the met lab, and dimentional attributes were correct.
even thou lubricant is well known to be the # 1 cause of failure.
is was not the factor in our case.

oh yes backlash is a big factor. not enough backlash with high torque components and there is severe rapid damage. even if it high precision gears.
I guaranty some of our gears where Q14 or Q15 of the preveious version of AGMA. and it did not help.

Mfgenggear

Mfgeneggear

 

[tbuelna]

geesamand-

When you stated these gears were carburized and ground to AGMA Q14 quality it indicates this is not a run-of-the-mill industrial application. It is very costly to produce gears to that level of quality, and the cost would only be acceptable for an application where high gear mesh PLVs are involved and/or where low levels of drivetrain noise/vibration are required.

Your comment about micro-pitting being the most common failure mode experienced in high-performance gear drives is basically correct, and this is by design. There are basically three primary failure modes for oil lubed gears- tooth bending fatigue, surface contact fatigue, and scuffing/scoring. Tooth bending fatigue and scuffing/scoring failures are difficult to detect, occur fairly rapidly, and are usually catastrophic in nature. The design philosophy used for high-performance or critical application gear drives is to design them with very conservative margins so that they never experience a tooth bending fatigue or severe scuffing/scoring type failure. On the other hand, micro-pitting type failures tend to progress very slowly and are very easy to detect with lube oil magnetic chip detectors long before they present a serious problem. With properly designed, manufactured and maintained gear drives they will normally be taken out of service and overhauled when onset of micro-pitting is detected. Thus this is the most common type of failure mode you would expect to see.

One question I have is that if your gear drive application justified the high cost of carburized/ground AGMA Q14 gears, then why didn't it have a simple magnetic chip detector in the lube oil system? A lube oil magnetic chip detector would have caught a micro-pitting failure many hours before it became a serious problem.

Interesting discussion.
Terry

 

[geesamand]

We specify AGMA 10 gear quality but the Hofler Rapid and Helix generating grinders that do the grinding produce no less than Q13. I understand it's all due to the Hofler on-board GMM that feeds back corrections in between grinding passes - by the final grind pass it's nearly perfect. I've seen the newer Gleason grinders overachieve as well.

Perhaps these grinders are not economical for high volume production, but for our small quantities we pay the same as before and get incredibly consistent quality in all of our helicals. I wish our housing accuracy and bevel gearing could be at that level of quality for similar cost. (I'm hopeful that milling small-run bevels may get there someday)

This failed drive was identified with vibration spectrum analysis weeks before it tanked. I suspect that those involved were not permitted to stop their production and just wanted to keep it running as long as possible.

David

 

[tbuelna]

We specify AGMA 10 gear quality but the Hofler Rapid and Helix generating grinders that do the grinding produce no less than Q13.

God bless the gear grinder OEMs. Ask for AGMA 10 and get AGMA 13. Machines with built-in inspection probes are indeed fantastic at compensating for tiny profile errors and grinding wheel wear. But as you noted, all of that probing adds time to the finish grind process.

Back to the issue of micro-pitting, here's what AGMA has to say about it:
"Factors that influence micropitting are gear tooth geometry, surface roughness, lubricant viscosity, coefficient of friction, load, tangential speed, oil temperature and lubricant additives. Common methods suggested for reducing the probability of micropitting include:
-- reduce surface roughness;
-- increase film thickness;
-- use higher viscosity oil;
-- reduce coefficient of friction;
-- run at higher speeds if possible;
-- reduce oil temperature;
-- use additives with demonstrated micropitting resistance;
-- protect gear teeth during run--in with suitable coatings, such asmanganese phosphate, copper or silver plating.
CAUTION: Silver or copper plating of carburized gear elements will cause hydrogen embrittlement, which could result in a reduction in bending strength and fatigue life. Thermal treatment shortly after plating may reduce this effect.
Surface roughness strongly influences the tendency to micropit. Gears finished to amirrorlike finish have been reported to eliminate micropitting.
Gear teeth have maximum micropitting resistance when the teeth of the high speed member are harder than the mating teeth and are as smooth as possible.
Currently there is no standard test for determining micropitting resistance of lubricants. However, FVA Information Sheet 54/IV describes a test that uses the FZG C--GF type gears to rank micropitting performance of oils. At present, the influence of lubricant additives is unresolved. Therefore, the micropitting resistance of a lubricant should be determined by field testing on actual gears or by laboratory tests."

The first six factors listed are all related to lamda ratio and dynamic oil film thickness at the contact area. You mentioned an oil was used that did not have the proper viscosity characteristics, but you still could have problems even when using an oil with the correct viscosity. Conditions such as high oil input temps, high gear flank surface temps, excessive mesh contact sliding, or contact frictions due to excessive surface roughness all contribute to increased oil film flash temperatures. Flash temperature rise reduces oil viscosity within the dynamic fluid film, and the reduced oil viscosity then results in a thinner dynamic oil film. The process is somewhat self-perpetuating.

Lastly, regarding this comment from one of your earlier posts: "Our gear designer advises me that run-in is not really applicable to carburized/ground gearing of high accuracy." This is not true with gears operating with oil film contacts. Even gears manufactured with a very smooth surface will usually see a slight improvement in their surface finish after an initial run-in period. High-performance gearboxes are often given a "green run" followed by a cleaning prior to delivery. The green run usually generates a fair bit of fine metallic debris in the oil which needs to be removed. When performing an analysis for gear scuffing/scoring it is also common to assume a gear flank surface roughness value existing after a run-in period.

Good luck with diagnosing your gearbox problem.
Terry