Effective case on gears, how much is too much??

[MTPTurbo]

We are trying to determine an appropriate case depth "window" for some of our carburized 8620 parts (8dp, 25 degree PA) and are reluctant to widen it much beyond our current specs. Our Heat treater's metallurgist would like us to provide a minimum and consider anything in excess of that amount good. We have a considerable amount of test data that demonstrates what happens if we have too little case, but I have found little more than a few references regarding the negative effects of having too much. Have any of you encountered tooth chipping due to excessive effective case depth? Are any of you aware of papers/texts which provide more detailed insights regarding the matter. I have reviewed Dudley's texts, as well as numerous AGMA and ASM books, standards etc... any insight would be greatly appreciated

 

[CESSNA1]

MIPTURBO: If you are successful with what you are doing I see no reason to change it. Try another heat treater.

Regards
Dave

 

[MintJulep]

I've never actually read this book, but it's very thick.

http://www.normas.com/SAE/pages/AE-15.html

 

[tbuelna]

MTPTurbo,

There are many variables involved in the outcome of the finished case on your gear. Prior to finish grinding, the carburized case can be anywhere from .030 to .070 inch thick, depending on the exact carburizing process you employ. Depending on how you quench and temper the part, the properties of the core area can also vary greatly.

With a highly loaded 8dp, 8620 alloy gear, a finished (after grinding) case depth of .025 to .035, a case hardness of Rc58-62, and a core hardness of Rc 38 min, should be quite adequate.

The only problem you can encounter with excessive case depth is that with small teeth and/or sharp corners, you can "through harden" the tooth or corner, causing it to be very brittle. With 8dp teeth, you should not have to worry about this problem. Just remember to radius your gear teeth edges appropriately, or copper mask where necessary.

If you process the part properly, a deeper case depth will not produce "chipping". If you're experiencing chipping (or more correctly: spalling), it's most likely because your contact stresses are too high for your design.

The only other reason that I can think of for your "chipping" is the loss of temper due to grinding burn. You can check for this with a Nital etch process.

Good Luck.
Terry

 

[rob768]

I think case depth should depend on the amount of subsurface herzian stress you expect. Higher stress (from more heavy loaded gears) requires a deeper case.
Perhaps the book from Lester W. Alban (Systematic Analysis of Gear Failures)sheds some more light on the subject

 

[tbuelna]

rob768,

I agree, a deeper case will allow a greater contact stress limit. Of course, the strength of the underlying core material supporting the case also has a significant effect on the allowable contact stress (ie. brinelling). There are also limits (both metalurgical and economical) to how thick a case can be produced. And the exact carburizing, quench and temper process employed determines those properties. The typical case hardening processes for 8620 alloy steel are-

a) carburize and direct quench, followed by tempering, or
b) carburize, cool, reheat, single quench and temper, or
c) carburize, cool, reheat, double quench and temper

These various processes can produce a core hardness that varies from about 120 ksi to 180 ksi, and a case that varies from about .020 to .070 inch thickness.

Also remember that a case hardened gear undergoes significant distortion during the quench, so it must be finished by grinding. The amount of distortion is dependent upon the geometry of the part and whether a quench fixture is employed. The more the part distorts, the more material that must be removed during the finish grind and the thinner the finished case.

The actual level of contact stress that a gear experiences in service can vary greatly, depending upon the accuracy of the manufactured gears, the accuracy of the mounted gears and any deflection of the gears under load and/or thermal displacement. For example, a spur gear mesh has a nice line of contact, in theory. But that quickly becomes a point contact condition (edge loading) with just a couple thousanths of an inch of misalignment. And your contact stresses go up by several factors. That's why the standard formulas you find in Dudley or Buckingham are very conservative.

As for MTPTurbo's original query, you will never really end up with "too thick of a case" on your parts, due to inherent limitations with most commercial carburizing processes. It's typical practice with carburized gears to specify a finished case depth range (after grinding) and hardness and to where it applies (flanks, tips, root fillets, etc.), as well as a minimum core hardness. It's then up to your gear vendor to optimize his processing to meet your finished part requirements. Remember, he's in business to make money, so he won't leave your parts in the carburizing furnace any longer than necessary to produce the requisite case depth (time is money).

Good Luck.
Terry

 

[MTPTurbo]

Thanks for taking the time to offer your thoughts. We have thoroughly evaluated these designs using UTS gear design software to determine the case depths needed to support the design load, however doing so suggests a "window" which is rather narrow (typically only 0.005")and difficult for our heat treater to hit with any level of consistency given the inherent errors associated with micro hardness testing, material variations etc... In light of this we were considering extending the upper end of the envelope by an additional .003-.005" to give them a larger target. If we go any thinner we know they will fail in severe applications. The question is, what will happen if we go heavier? The texts and specifications I have reviewed are rather vague, but they imply that too much case will make the teeth prone to crack/chip/shatter/spall...etc.. Our Heat Treater's head metallurgist who claims to have an extensive background with gear applications claims that "you can't have too much case". We have been in the business for years and don't recall any failures which appeared to be the result of "overcasing". I am just wondering if any of you have observed this phenomenon.

 

[diamondjim]

I would spec .010 minimum.
I would use 8640 rather
than the 8620. You probably
are having subsurface shear
stresses in excess of the
allowable. Are you checking
the stresses at 50Rc transition
depth and .020 below this point?
I feel that the .005 is too
little of a case. With the 25
degree pressure angle you may
want to spec out a simple tip
relief before hardening as well.
You might want to consider
induction hardening these parts.
You did not mention the number
of teeth, so it is hard to know
the value of the tooth thickness
at the tips.

 

[Frederick]

I suggest that you look into vacuum carburizing these parts to give tighter control of the case depth. Vacuum carburizing will give a more even case as the carbon is infused in a controlled manner after the gear is up to temperature. In normal gas carburizing, the case depth can vary significantly as the tip of the tooth heats up more rapidly than the root. The infusion of carbon is temperatre dependent, the higher the carbon the faster the infusion. In vacuum carburizing, the part is heated, then the carbon is introduced, so the case becomes very even.

 

[tbuelna]

MTPTurbo,

You did not specify what case depth you are trying to achieve on your finished part. As I mentioned in my previous post, there are practical limits to how deep a case you can produce in 8620 with carburizing, so your metallurgist is correct: It is highly unlikely you are achieving "too much case depth".

Also, as I mentioned, there can be a significant difference between the case thickness before and after finish grinding. Depending upon the amount of distortion produced during quench. For example, are you using a quench fixture?

If your analysis shows you only need a case depth of about .015 inch or less, I would recommend switching to a nitriding process. The processing temperatures are much lower (about 950degF) and there is no quench required, so distortions are kept to a minimum. It is also much easier to control the case depth with nitriding. And to finish the gear profile, all that is usually required is a light hone to remove the "white layer".

Finally, as a good reference, take a look at MIL-S-6090 for standard practices employed in the aircraft industry for carburizing or nitriding.

Good Luck.
Terry

 

[MTPTurbo]

More info on one of the parts in question:

We have a gear that is 8 DP, w/ 30 teeth the face width is in excess of 2 inches. We currently specify an effective case depth of 0.035"-0.040" (thus providing the window/target of 0.005" I spoke of in my previous posts). Our heat treater would prefer to see us simply set a 0.035" minimum and permit any thing in excess of that to be cinsidered a good part. These are very heavily loaded planetary gears which are used in applications reaching intermittent torques in excess of 300,000 in-lbs. We may consider tip relief but would prefer to minimize the number of operations on the gears if at all possible.

DiamondJim, in light of the range of case depths we are looking at, do you still recommend looking at 8640? What are the particular advantages that you are aware of?

Also, in summary what I am trying to figure out is this: have any of you seen gear failures which clearly resulted from an "excessive" effective case depth?

Thanks again to all for your efforts....

 

[TVP]

MTPTurbo,

Yes, in certain instances, excessive case depth can cause failures. tbuelna provided one example earlier. Internal stress rupture at the case/core interface can occur under conditions where the case depth is excessive, due to the high residual tensile stresses that develop. Discussion of this phenomenon starts on page 159 of Alban's book. I suggest you develop a reasonable window for the heat treater to attain. You should not change alloys due to this issue.

 

[diamondjim]

I would think that the core hardness
would be greater with the extra carbon.
What core hardness are you now specifying?
Couldn't you live with .030 to .045 depth
of 50Rc case? That seems like a lot with
an 8 dp gear. Is this after grind?
I assume these are ground. What
quality level are these?
What depth is the protuberance hob cutting
the dedendum of the pinions?

I would be concerned about the tips
being too brittle and too sharp.
It looks like the tip width is about
.034 so you are hardening across the tips.
Can the tips be deburred before
hardening if these are not ground or
deburred after grind? Since you are
having double tooth contact when the
tips are loaded, are the contact
stresses still that high at the tip?

I would think that the case would cause
a compressive stress and not a tensile
stress at the case transition gradient.

Is the carrier driving the pinions or
vice versa?

 

[tbuelna]

diamondjim,

With 8620, you should be able to easily achieve a case depth of .040 inch with a case hardness in excess of Rc55. The underlying core hardness will be a function of the quench and temper procedure used.

I don't really see where an alloy of 8640 (twice the carbon content of 8620) would provide any benefit. If 8620 alloy steel is not sufficient for this application, then your next logical choice would be 9310 alloy steel (AMS 6260). 9310 can be case hardened by carburizing, but it produces a much tougher core due to it's lower carbon content. Most high performance (carburized) gears for aircraft, or automotive (rear ends) applications use 9310.

As for avoiding "through hardening" of tips, you can always apply a copper mask prior to rough hobbing and carburizing.

 

[Carburize]

AGMA 2001 gives a recommended maximum case depth of
0.4/NDP
or
0.56 x the width of the top land
See equation 43 section 16.1

 

[TVP]

Terry,

The 3% Ni has something to do with the increased toughness of 9310 as well.

 

[MTPTurbo]

I would think that the core hardness would be greater with the extra carbon. What core hardness are you now specifying?
Couldn't you live with .030 to .045 depth of 50Rc case?
-we don't specify

That seems like a lot with an 8 dp gear. Is this after grind? I assume these are ground.
-they are not currently ground, our analysis shows that a Class 7 gear will give us sufficient life

What quality level are these?
- AGMA 7

Can the tips be deburred before hardening if these are not ground or deburred after grind?
-We would prefer to avoid it if possible, but that is something I may look into..


Since you are having double tooth contact when the
tips are loaded, are the contact stresses still that high at the tip?
-they aren't excessive under normal conditions, my understanding is that the real problem is when misalignment occurs

I would think that the case would cause
a compressive stress and not a tensile
stress at the case transition gradient.
-agreed

Is the carrier driving the pinions or
vice versa?
- this is a speed reducer, the carrier is attached to the output shaft and the ring is stationary

 

[diamondjim]

The sharp corners are the only
thing that bothers me. You are
already hardening across the
tips so greater depth may not
be a problem. AGMA 7 after
hardening is decent. I assume you
are getting shrinkage with that
depth of hardening. So you probably
have an AGMA 9 before hardening?
Do you already have broken tips?
If not, go for it. What is the liability
if you have a failure? Can it be
easily exchaned or replaced?

 

[MTPTurbo]

We have yet to see any tip breakage failures, but are only recently getting into 25 degree PA gears (most of our older products use 20 degree PA's). As far as liability, we might lose some high volume accounts, should field failures occur. My guess would be that broken tips themselves might not produce critical failures, but the "chips" may damage other surfaces, thereby accelerating pitting failures or perhaps inducing bending failures should the pieces bind between teeth (if they are very large..) Pre HT I would say we are in the realm of AGMA 8-9, our Heat treater has done a nice job of minimizing distortion, but has struggled to obtaining sufficient effective case depth. It has been a real challenge...

Thanks again for all of the advice. I suppose no news is good news... (this thread isn't flooded with replies from those who have seen failures which were the result of tooth chipping) I will continue to dig...