broke the 4472 on the first time to the track

[JSM]

Re: broke the 4472 on the first time to the track

My apologies to Mr Borg Warner for bashing his Tcase before I had all the facts.....

When I pulled the Tcase out this morning I found the output shaft of the tranny twisted into. Has anyone else ever broken an output shaft on the 4l80E conversion? The splines are twisted and it looks like it was just overpowered.


I ordered a new chain and a rebuild kit for the Tcase, I guess I will go ahead and do that while it is down. I was really hoping to make it to a dyno day this coming Saturday, but it isnt going to happen now

I would like to see a picture if you can. I have done a bit of solid modeling on the shaft, which although isn't perfect, didn't show it would do what you said it did.

If it ripped the splines off the shaft, then I would suspect a loose fit. If it twisted the shaft, then the heat treat must not have been great and it was soft.

Really if it breaks, it should break at the base of the shaft where it flares up to the larger diameter, at least my engineering guestimations show that. That will only happen over 1200 ft lbs of torque, again all in theory.

We all know theory and what really happens don't equal each other. Just curious though as your the first with an issue that I know of.

I'd like to see a picture of the end of the shaft also (transfer case side) if possible.

 

[l-eater]

Re: broke the 4472 on the first time to the track

Here is a pic of the broken piece. I will try to get a shot of the end of it sometime today.

 

['JustDreamin']

Re: broke the 4472 on the first time to the track

Here is a pic of the broken piece. I will try to get a shot of the end of it sometime today.

Ouch....That hurt.....That is alot of twist in the splines, though. I would have thought it should not have twisted as much, like maybe the heat treatment was off (not adequately hardened). But the end may provide more evidence (or it may be wiped clean because of spinning after it broke).

That will only happen over 1200 ft lbs of torque, again all in theory.

1200 ft lbs of torque isn't all that hard to get to. It seems like a big number, but it really isn't.

1st gear in an 80e is 2.48:1. To hit 1200 ft lbs of output torque requires 485 ft lbs of engine torque.

And that doesn't include any torque multiplication from the torque convertor (I've seen values in the 1.5 to 2.1 range). With a 1.5 multiplication TC (for example), that 485 ft lbs becomes 323 ft lbs. That not that hard to do, especially with a stout V8. I'd hope the limit is significantly higher.

But like I said above, the shaft has a bunch of twist in it, which suprises me. I would have thought it should have sheared off (at the reduced section) or failed in more of a shattering action. Maybe some additional pictures or some maybe hardness testing will provide useful information as to whether this is a design related failure (reaching the design limitations) or a manufacturing process error (bad heat treat (drawn to far back) or something).

'JustDreamin'

 

[JSM]

Re: broke the 4472 on the first time to the track

1200 was engine torque as I had done the reverse for the gear ratio already

Yes, it failed in not what I expected, looks like shaft was soft. The interesting thing is on a 2wd 4l80e output shaft there is a diameter that is very close to the smallest diameter once a shaft is converted to 27 splines. So a stock 4l80e shaft is not much if any stronger than one resplined.

I wonder if the heat treatment failed somehow, be curious to know what GM's Quality control is like for shafts.

 

['JustDreamin']

Re: broke the 4472 on the first time to the track

1200 was engine torque as I had done the reverse for the gear ratio already .

AHHH....That's good....Does that back-out include any torque convertor multiplication?

I wonder if the heat treatment failed somehow, be curious to know what GM's Quality control is like for shafts.

It would really depend upon how they heat treated it....Also would depend upon what they're trying to accomplish, that area might intentionally be soft (can't think of why they'd want to do that but....).If it's induction hardened, they selectively do portions of the shaft and leave others.....

I'm assuming re-heat treating the part probably isn't a possibility because of the potential for warpage (I wouldn't take the chance myself).

'JustDreamin'

 

[George Blake]

Re: broke the 4472 on the first time to the track

The shafts that are used for the syty's were chosen due to a deeper hardening of the particular application. This shaft just wasn't done like the rest. Basically it was too "soft". Normally they run from 58 rockwell down to as low as 40 rockwell. I really would like this shaft back, Berville. When you get it out of the trans please send the whole thing. I'll pay shipping on it. Your other one went out today. I'm betting this particular shaft just wasn't hardened correctly to begin with. There are people running down into the 8s and this is the first we've had do this. This stuff can happen when your racing but it's something that I cannot control. Basically when I have these machined in the future, I'm gonna have to have hardness testing done. This really sucks because this means I will have to pay for a shaft, have it machined then test it and it could be worthless and have to be thrown away. This shaft did not at all fail as ANY of the models had shown. Again leading to a poorly hardened piece. With over 30 of these out there now, it was just a matter of time I guess before a bad unit popped up.

 

['JustDreamin']

Re: broke the 4472 on the first time to the track

The shafts that are used for the syty's were chosen due to a deeper hardening of the particular application. This shaft just wasn't done like the rest. Basically it was too "soft". Normally they run from 58 rockwell down to as low as 40 rockwell. I really would like this shaft back, Berville. When you get it out of the trans please send the whole thing. I'll pay shipping on it. Your other one went out today. I'm betting this particular shaft just wasn't hardened correctly to begin with. There are people running down into the 8s and this is the first we've had do this. This stuff can happen when your racing but it's something that I cannot control. Basically when I have these machined in the future, I'm gonna have to have hardness testing done. This really sucks because this means I will have to pay for a shaft, have it machined then test it and it could be worthless and have to be thrown away. This shaft did not at all fail as ANY of the models had shown. Again leading to a poorly hardened piece. With over 30 of these out there now, it was just a matter of time I guess before a bad unit popped up.

I think checking the hardness may not be a bad idea, assuming there is an area on the part you can do it without causing problems....Hardness testing is semi-destructive....Leaves a little dimple....I wouldn't want that little dimple in an area where a seal runs or at the smallest diameter (where you'd have a stress concentration factor applied to an area with a stress concentration factor already)

I wonder if these shafts could be re-heat treated in such a manner as to minimize warpage. I would think conventional quench and temper could be done (which would harden it all the way through), as long as the part was oriented to minimize warpage (I would think vertical orientation as it goes into the quench tank). Might be interesting to experiment with a junk (soft) shaft and see what happens. Heat treat it, then do run-out testing, maybe? Might even be able to heat treat them before re-machining them, therefore not spending a bunch of time on a warped shaft.

I'm not sure there is another useful heat treating method for these shafts (I'm not a heat treat expert). But, I don't think Austemper would work, because the section thickness is too thick (and I know the process at my work would have the added drawback of not being able to control part orientation which would probably inducing warpage). I believe Induction requires fairly specialized equipment, and I'm not sure what the results would be like even if you could find somebody with the right hardware. Not sure about flame hardening either....

Just some thoughts.....

'JustDreamin'

 

[George Blake]

Re: broke the 4472 on the first time to the track

That may be an option. Turning it down. Re heat treat. Then cut splines. The only problem on the heat treat that if it is just done over the whole part, it will have the affect of annealing the rest of the structure (that means softening it for others reading this). I think Berville will have much better luck on the next shaft. He made 6 passes and the first one twisted. This one will do him good.

There are some other things on the horizon. It just costs a little more. Guys are complaining now that it costs too much. Of course they spend $18K on an engine and then balk on billet components for the drivetrain:roll: This is ongoing in developement. The 80e is the right direction. It's execution. I talked with another knowledgable gentleman last night and I think I have an idea that is a combination of factors from our conversation. I think it will bring it to the masses more as well.

One thing Jeff mentioned and I know most of you guys don't realize this but there is an O-ring on the 80e output shaft. The shaft is machined down for this ring. So great big 32 spline vs 27 spline doesn't mean anything because the ring is literally the size that the shaft gets turned down to before respline. So you could have big 32 spline engagement and then twist off at the groove. Again, opens it up to a billet piece down the road. It's coming. Someone WILL get into the 7's and the need will arise. There are some goodies for the transfer case being worked on as well. Borg Warner did one OUTSTANDING job on this case and it just needs a little help.

 

[Tydriver]

Re: broke the 4472 on the first time to the track

This thread is of particular interest to me because eventually I will be going 80e (if my truck ever runs again) :rant: . I am somewhat concerned about the particular output shaft I have since it was from the very first run of shafts ever conceived for our trucks. I believe it also was turned down by Moser (if I recall correctly, its been too long).

Anyhow when the concept of 'turned down & resplined' shafts was talked about in its infancy I wasnt completely opposed to the idea, but questioned why it was being done. I suspect its because it was easier to go this route than the route I proposed. I would have liked to see atleast some focus being placed on replacing the input drum assembly of the transfer case to match that of the spline count and diameter of the standard 80e shaft, and to see if it could be changed out with a different unit to accept the standard 80e output shaft for a 4x4 transmission. After all "bigger is better" and turning down a larger shaft to 700r4 specs seems sort of like shooting yourself in the foot since its been proven time and time again that the output shaft for the 700r4 can be snapped as well even if the rest of the transmission is holding together just fine. Essentially you are making the output shaft the "fuse" again in the drive train even in the 4L80e builds. I am not 100% certain this is even a viable option but if I had my choice I would rather go bigger than repline and go smaller. I realize this shaft snapped at a relatively LOW HP build (still nice numbers, just no where near the 'quoted' hp/tq numbers everyone has been throwing around), and its most likely NOT indicative of all the builds/shafts out there.

I dont have the contacts or even know where to start looking for a change for the input drum to increase it to match that of the STANDARD 4L80e shaft diameter/spline count. Transfer cases really aren't a mystery or anything but most are pretty bulletproof so they dont get alot of attention (like transmissions and engines do). I suspect it would have to be someone with some engineering contacts at New Process, NVG, Borg Warner, or possibly even GM. It would make more sense to me to alter the T-case input drum and at the same time replace the chain, these trucks aren't getting any newer and its probably about time the T-case got some attention anyhow.

I realize this post isnt going to be viewed positively by the guys making the conversion kits and building the transmissions because its going to cut into their market a bit. Its not intended to start a flame war or anything else, merely to offer a different perspective on a problem.

I am in full understanding of the cut down/respline/heat treat process. I understand the aspects behind the heat treat process as well (its quite common to do in aviation aluminum & steel) but I just dont think that a cut down 80e shaft thats designed to match 700r4 specs no matter how its made or machines can match the strength of an unmolested 4L80e shaft.

Anyone have any contacts or connections they'd like to exploit in pursuit of this idea? I dont even know for sure if it can be done, it may be outside the size limitations of the T-case altogether, I'd just like some hard data to review before completely blowing that option out of the water.

Again just my :2cents:

 

[George Blake]

Re: broke the 4472 on the first time to the track

This has been investigated already Todd. Like I said above, the 80e shafts have a groove cut in them that again makes a fuse. It's not any stronger than the 27 spline shafts.

I am working on the transfer case end and I think I have an option but it still wont' be "free" so to speak.
As far as "hurting" the market, there is no real syty market. Myself and others are doing tons of research for what? 10-12 trucks over a couple years? That's not a market. We are trying to help ourselves and the few that have these trucks. If you want to take that away, you can but we are working on it and cost unfortunately IS a factor. Like I wrote above. A billet 80e shaft would be HUGE dollars because of the large base. But if guys are gonna play, they have to pay. On the transfer case again, a VERY viable option has come but it won't be cheap either. I won't put up anything more publicly but it will bring this to the masses. So market? No. If I wanted to keep the market I wouldn't do a T-case design but I'm not trying to corner anything. Robert has done 8s with the identical shaft and again, the 27 spline 700r shaft fails at the HOLES. Not in the splined area. The holes are like serrations in paper that creates a weakness.

Does anyone else reading this understand what we're doing? We're not trying to corner this. We're trying to make the most economical setup because so few of us are doing it. If there were thousands it would be a no brainer how to do it but the REALITY is there are NOT thousands. We're gonna get this done though hopefully soon.

 

[JSM]

Re: broke the 4472 on the first time to the track

I looked at opening up the input shaft, but then you run into where you weaken it so to speak. I felt it was weaker and more expensive to open the input shaft up (splining internal holes is much more difficult). Now you could make the input shaft larger in that area, but then you would need a bigger bearing in the t-case, which means machining of the case, well you can see where this is going.

Billet shaft really would probably solve this for anybody.

If you guys get bored though, price just a bare chuck of 4140 or your choice of material. Figure 5" diameter x 8" long. let me know what you find and know that is just the billet, no machining started yet.

I really think this shaft was a fluke from GM quality as I have modeled the stresses on the shaft with 2 programs, 2 different times, go the same answers and they don't show what happened in this case.

 

[George Blake]

Re: broke the 4472 on the first time to the track

Please post up your results, Berville, after you get the new shaft in.

 

['JustDreamin']

Re: broke the 4472 on the first time to the track

If you guys get bored though, price just a bare chuck of 4140 or your choice of material. Figure 5" diameter x 8" long. let me know what you find and know that is just the billet, no machining started yet.

I checked with my local supplier (needed to call them anyway)

5" diameter 4140 hot rolled & annealed 8" long = $60.

But the machining time certainly would be the killer....Even in a fairly high hp CNC lathe, that's still got ALOT of material to take off.

'JustDreamin'

 

[#2875]

Re: broke the 4472 on the first time to the track

Jeff, have you considered friction welding? No one that i know of has ever broken a good friction weld at the weld... in an industrial application.

it would solve the problem of needing larger OD material and it would cut down greatly on the machining time.

-matt

 

[George Blake]

Re: broke the 4472 on the first time to the track

I already have a solution in the works and it doesn't require welding shafts together or machining hardened surfaces off. I've got everything in and I'm just waiting for a quote.

On the output shaft of the trans itself, you need to get in on a forged billet that's close to the shape you already need. I've checked on that also and it is still cost prohobitive. In an already non existant market, if you make a components that costs $1300 with machining and hardening on top of the purchase of the forged cores, what good is that? Gotta keep it realistic.

 

[#2875]

Re: broke the 4472 on the first time to the track

friction welding isn't welding like you and i can do in the basement. trust me it will NOT be the weak link. also, i HOPE you guys are machining these shafts in an annealed state and not after hardening...

another thought is maybe you could use a shaft that can be bought off the shelf, friction weld it to your 5" stock and finish machine in one operation in an NC lathe, no need for a second operation.

I don't know if there is anything available that you could use i just know machining and i'm throwing out ideas...

I already have a solution in the works and it doesn't require welding shafts together or machining hardened surfaces off. I've got everything in and I'm just waiting for a quote.

On the output shaft of the trans itself, you need to get in on a forged billet that's close to the shape you already need. I've checked on that also and it is still cost prohobitive. In an already non existant market, if you make a components that costs $1300 with machining and hardening on top of the purchase of the forged cores, what good is that? Gotta keep it realistic.

 

[George Blake]

Re: broke the 4472 on the first time to the track

I know what your talking about and the solution I have doesn't involve 27 spline grafts. They will be E4340 most likely and be a totally new component.

 

[l-eater]

Re: broke the 4472 on the first time to the track

I got the new shaft today....George has been great to deal with on this issue-thanks for the quick shipping on the new shaft. I am hoping to get the tranny put back together within a couple days and maybe be back in business soon,

Now, to find the next weak link........

 

[TT350]

Re: broke the 4472 on the first time to the track

This is little off topic but it is about T-Cases.

We all know there are 2 models a 1372 & 4472,
one has 27 spline output shaft and the other has
a 32 spline output shaft.

But there are different versions of both.

I have 5 T-Cases in the shop 2 have 27 spline output
shafts and 3 have 32 spline output shafts.

1 of the 27 spline cases has a large pilot hub,
this is the part that centers the factory T-Case adapter
to the T-Case, and 1 of the 32 spline cases has the large
hub as well.

The other 3 have a smaller hub.

This means that there are at least 4 different models.

Chris

 

[l-eater]

Re: broke the 4472 on the first time to the track

I noticed some of those differences in the Tcases when I was doing the research on what to rebuild/replace inside it. Apparently the AWD 4472 used in the Astro vans uses a unique chain and sprockets but I think the rest of it is pretty much the same thing.

The 2 cases I have are the 32 spline units. I just went through mine today and got it all ready to re-install.