Heat Treating OE Gears a waste? 4

[apex944]

I've got a guy with a big turbo on a small car that used to make 100hp. Now he's pushing 350 and breaking gears all the time (imagine!)

Anyway he wont pay for proper racing straight cut gears and shafts so he wants to try heat treating.

Is there anyway that a heat treatment could be this successful on gears made for 1/3 the HP this engine now makes?
And if so, what type of treatment are we talking about?

Thanks in advance.

- Bill

 

[tbuelna]

Re-tempering a case hardened gear to produce a softer core would make the gear more prone to a subsurface fatigue failure, where the transition between the case and core occurs. It would appear as a case spalling failure (ie. big chunks of the case material popping out of the surface of the gear).

The production gear guys really know their stuff, so I think it's safe to assume that they have gotten the best performance possible out of the particular material that those gears are made from.

 

[izzmus]

I have had extremely good luck with the Shockproof gear oil in a semi daily driver vehicle that saw regular competition abuse, similar to autocross but off-road and with more runs per event and a lot more shifting. Previously on all manner of different gear oils, I would killing the bearings in as little as one outing. The current transmission went in a year ago and shows no signs of failure despite roughly ten events and 20,000 street miles.

My transmissions had been experiencing bearing failure, not gear failure, but bearing failure leads to gear failure in a hurry. Synchronizer action seemed to be about the same or slightly improved, despite Redline's warnings on the subject.

After a month of good luck with the transmission, I changed the differential oil to Shockproof as well. The breakaway torque (clutch type differential) went from 40lb to under 10lb in a very short time, even though Redline's copy claimed that it was safe for limited slip units. The gears themselves are still in good shape, though.

 

[NickE]

izzmus - RallyX? - I also have excellent performance with shockproof. In fact even at -20F, although I do get bad bad noises till she warms up. (I mix it down with ~25-50% 75-90NS)

(5000+ competition miles, 135,000+ total)

Although my turbo is small.

 

[Tmoose]

http://www.strangeengineering.net/catalog/pdf/002.pdf

http://www.strangeengineering.net/catalog/pages/023.jpg

Tip 4, partway down the page.

http://www.hotrod.com/techarticles/chevy_12_bolt_differential_install/

 

[apex944]

Transmission gears, not R&P

 

[tbuelna]

Obviously the geniuses at Hot Rod magazine don't have a clue as to the difference between case hardness and core hardness in a case hardened gear.

 

[IceStationZebra]

They just print whatever Strange Engineering tells them. :{
ISZ

 

[swall]

Ironically, these kind of tech articles in Hot Rod and Car Craft got me interested in metallurgical engineering in the first place. Once I got my degree, I could see how some of their tech content misconstrued things or left out important details.

 

[NickE]

Also like most not in our field they state this:

"Softer construction allows for a certain amount of tooth "bending,"

Nope softer means weaker, and possibly more resistant to shock, but definately not more flexible.

Nick
I love materials science!

 

[pontiacjack]

Okay... so how would a metallurgical engineer explain the good performance of 'Pro' ring & pinion gears (from U.S.Gear, Richmond, etc.) in extremely highly loaded applications (dragsters, etc.) without mentioning deformation shy of the yield stress value? I know firsthand that their surface hardness is less than typical OEM gears. Is their performance strictly attributable to the properties of the 9310 alloy? If the manufacturer had treated the 9310 up to higher surface hardness, would they be even 'stronger' yet?
Thanks for all input.

 

[tbuelna]

pontiacjack,

Since you, or the gear vendors you seem to be endorsing, have not substantiated your claims with any hard statistical or analytical data, I would simply attribute these claims to "old wives tales" (very common in drag racing circles). That is, their claims are anecdotal.

Gears are case hardened not to prevent "wear", but to improve their surface fatigue life. A properly designed, installed and lubricated gear mesh operates in a hydrodynamic lubrication regime, which means you never have metal-to-metal contact. The peak hertzian stresses, at the location of the EHD oil film supporting the contact load, can be quite high. For carburized 9310 gears, the allowable hertzian stress can easily exceed 180 ksi, even for long life production gears. According to classical gear theory, a gear with a "softer" case would perform worse than a gear with a "harder" case, simply because the allowable magnitude and number of load cycles would be lower.

There are also many factors that affect gear life besides metallurgy. The primary factor being contact geometry under load. Since no drag racers (including those at Strange Engineering) have the technical expertise or resources to properly evaluate the contact pattern of their gear mesh under load (using strain gauges), they will never know for sure how their gears are responding. Like most shade tree mechanics, they simply coat the gears with a dye, install them, and perform a roll thru without operating loads. All this ends up telling you in the end is that you have a satisfactory contact pattern in an unloaded condition, which is relatively meaningless. Even looking at the contact pattern of a gear with a few runs on it is misleading because the contact pattern is an aggregate of numerous load/speed conditions.

Hypoid gear geometry and metallurgy are complex subjects. If you would like to understand heat treatment of 9310 steel, check this reference:

http://www.latrobesteel.com/engineering/Sheets/Bearing_Gear/9310.pdf

Good luck.

 

[firewire67t]

there are only a finite number of cars that one can take from 100whp to 350whp in one's garage. Any vehicle with this much aftermarket support is bound to have a wealth of aftermarket transmission upgrades. Your customer needs to invest in his vehicle is such a way as to keep all aspects of it's performance equal. Shelling out to Skip-Barber will certainly help too!

 

[pontiacjack]

tbuelna- I'm not seeing any answers to my questions, in your 'answer' to me.
I see you refuting real-world experience as "anecdotal". Yeah, they are anecdotes- but thousands of them running every day. None of the top classes in drag racing (and a few other motorsports) could be at their present performance levels without 'Pro' ring & pinion gears. How many anecdotes constitute "hard statistical data"? "Wives tale"? My post was in invitation to a qualified engineer to dispel any wives tales.
I didn't "endorse" any vendor- I don't see how you inferred that from my question. I'm merely trying to better understand the metallurgy. Anybody got a straight answer, as to whether 'Pro' gears could withstand even more torque if the manufacturer were to create them with more surface hardness?
Thanks.

 

[patprimmer]

tbuelna

Top level drag race parts have life spans measured in seconds and fatigue life measured in a few thousand cycles.

People with experience might look at a machinist blue contact patch and aim for a predetermined set misalignment. They may have enough experience to know that it will distort under load to give a good contact patch, or at least acceptable life for the application.

Many racers change to these parts when their OEM std parts fail, so there is good real world experience to back up improved strength in the parameters that count most in the application.

I have not seen data on the grade of steel, heat treatments or hardness of core or surface, but I have seen evidence that the Richmond gear "Pro" series do improve life under very high load conditions.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[tbuelna]

pontiacjack,

Sorry if I offended, my comments were directed at gear vendors making claims without any scientific merit, not at your question.

The short answer to your question is no, at least for very limited life drag racing gear sets.

Your question regarding torque capacity versus case hardness is somewhat ambiguous though. To give a definitve (or "straight") answer you would need to pose a very specific question, since there are numerous variable that determine gear performance.

Gears fail in basically three ways: scuffing, bending or pitting. Scuffing is more of a lubrication and geometry related issue, so case hardness is not of relative importance.

Pitting is a surface fatigue issue, so case hardness does have some importance. But a drag racing gear set does not see enough load cycles (a few thousand maybe?) to fail the gear in surface fatigue.

Tooth bending may be a concern if the loads are very high, but the highest stresses in bending will occur at the root fillet area, not on the flank of the gear. To maximize the bending strength of a carburized gear, this area is usually left unground, with the full case intact. But minor variations in case hardness at the fillet will have very little effect on allowable tooth bending loads. More important to tooth bending strength is the operating "pressure angle". In general, higher pressure angle will produce a stronger tooth.

If you will look at the reference noted in my previous post, you will see that a high quality, vacuum melt 9310 steel alloy, carburized, quenched and tempered at 300degF will have a case hardness of Rc 60 to 62, and a core strength of 187ksi UTS with elongation approaching 16%. These are very impressive numbers and are the reason 9310 is the gold standard for carburized gears. You'll note that Strange uses a lower grade carburizing steel alloy (8620) for their street gears.

These "as heat treated" numbers for 9310 are only half of the story though. The carburized case does not have a uniform hardness through it's thickness. It gets "softer" the deeper you go. And all of the quenching the gear goes through during heat treatment produces distortion. These distortions must be eliminated by a finish grinding process. Since the effective carburized case is only about .060 inch to start with, finish grinding can remove a substantial portion of it. In fact, the remmaining case may not be anywhere close to original Rc60 hardness level after grinding. With high quality aerospace gears, it is a requirement that a maximum of about only .008 inch of material is allowed to be removed during finish grinding to clean up the gear surface. If it requires more, the gear is scrapped. I don't think Strange adheres to such a rigid standard, since they are a commercial endeavor. So I would expect that a commercial grade gear produced in low quantities(like a Strange R&P) would have a relatively wide range of case properties from tooth to tooth and gear to gear.

As for the loads a gear tooth experiences, I don't know if "shock" is a good description to use. Any gear in the drive line of a recip piston engine powered vehicle (especially one with a slipping clutch and spinning wheels) sees an instantaneous torque of varying magnitude. Under normal conditions, the gear teeth should never experience a "shock" or impact load, since there is always at least one set or more of gear teeth in contact. This gear characteristic is referred to as "contact ratio". In fact, a high contact ratio is the main reason for using a hypoid type gear geometry. Unfortunately, increasing the pressure angle, as noted above to increase the tooth bending strength, reduces the contact ratio. So a balance must be struck.

Sorry for the long answer, but I hope you will gain something from my explanation of carburized gears. If I can explain something further, please let me know. I promise to try and not be so obnoxious.

 

[pontiacjack]

tbuelna- Thanks for your considerable effort to educate a non-metallurgist (me). Frankly, my pet peeves include the wives' tales that drive many of the decisions in motorsports.
I guess I should rummage through my scrap piles for broken 'Pro' gear pieces to hardness-test. I do remember that I was able to dress tooth corners with a file, which I would not have been able to do unless they were something less than 60 Rockwell-C. And, not having any "insider" at U.S.Gear or Richmond, I don't foresee ever being privy to such information.
You said a mouthful regarding the complexity of hypoid gears. As much as I utilize geometric relationships, I can't imagine trying to develop expressions to represent the surfaces of hypoid teeth! In fact, I don't even know how many parameters define a set- is it just five?
1) Effective diameter of pinion
2) Effective diameter of ring
3) Centerline-offset
4) Tooth pitch
5) Gear 'widths' (tooth lengths)
Or is it possible to have non-90-degree sets also?
Thanks for all responses (any day I learn a LITTLE something I consider a good day!)

 

[NickE]

I hope I didn't start something bad.

While bending strenght will definately increasse with hardness, bending stiffness will not. -- Thats what I was trying to say.

Thanks tbuelna -- your posts have been informative.

 

[patprimmer]

I have found this thread robust and informative and therefore of considerable value.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[axledude]

pontiacjack,

There is reason to believe your softer gears would perform better in an extremely short run situation.

If the failure mode is impact/overload, and the gear is at its limit for deflection in a given housing, it is probably being tip-loaded on the heel of the gear. In this case, IF the gear is Rc 50 or so it will "give" rather than a brittle fracture of the tooth. This kind of gear would work only if the team was changing gearsets every weekend. If it is lasting longer, than they aren't seeing the kind of loads that make it necessary. In this case, the standard processing, with an appropriate shot peen would actually be better.

What you're really looking at here is the area under the curve in a stress/strain test. While the yield/ultimate are lower at Rc 50, the overall toughness may be better, which for a very specialized application, can show an improvement.

I agree with Tbuelna that much of racer information is anecdotal and misguided. Furthermore, a proper engineering analysis of most applications using these softer gears would probably find them to be unnecessary. However, as you point out, the top classes would not be using them if they were failing. I believe for a weekend of 1/4 mile runs, these may offer an advantage.

As far as your question on parameters defining a set, there are 43 main parameters, and then many more machine settings that go into a given development. This is where a change can be made that would help an aftermarket gear by making tradeoffs that aren't acceptable in mass production (for instance, dramatic changes in pressure angle, spiral angle, contact ratio) due to noise or other concerns.

Hope this helps!

 

[pontiacjack]

axledude- Thanks for the response. 43 parameters? Holy cow- I know even less than I thought I did about hypoid gears!
My understanding is that lower vehicle floorpans was one of the main advantages of switching to hypoid rear gears- as late as 1949 for Ford. If this pinion offset weren't needed, would the more-efficient bevel gears serve us just as well as hypoids in modern vehicles?