GregBBRD

????????

Come on.....you have looked at all those 3mm squares with cracks that are running laterally through a broken 28mm shaft and think that all happened when the shaft instantaneous sheared? If you think this, you need a new metalurgist. Those squares and lateral cracks have been there for a long time...the "exposed" exterior cracks even have rust, in sone cases. I have 28mm shafts, with cracks, that are not broken.

There's no "chicken before the egg or egg before the chicken, debate here. Cracks happen first!

I'm just completely confused about how you can be debating this stuff....

Perhaps I have an advantage in discussions like this, because of the literally thousands of these cars I've observed and worked on. When I say I've NEVER seen a 25mm shaft migrate before 80,000 miles....that's not from a tiny little sample. Hell, I tell people that are buying cars that are "pre 28mm shaft cars" with less than 60,000 miles not to even bother checking for flexplate preload, if they can verify the mileage. Complete waste of time and money....it's not going to have moved!

Let me simplify/review:

1. Every shaft made, regardless of who makes it out of whatever material is going to twist.

2. Original Porsche shafts do not twist enough, when new, to pull themselves out of the stock clamp, which is why we never see "shaft migration" until higher mileage.

3. Unless there is damage to the metallic structure of any shaft, it will never "twist" any farther than it does when it is first made. (Modulus of elasticity)

4. The fact that shafts pull out of the stock clamps at higher mileages means that they twist more....the shafts have micro cracks....as I have stated. These are easily seem, on high mileage 25mm shafts and 28mm shafts by Magnaflux. (Have you even ever taken a look at this....while you tell me it doesn't happen? Because I have....and there are always cracks.)

5. Increased clamping IS a patch....the real problem is the increased twisting of the shaft. Replace the shafts with a new shaft from 928 International, from you, or from me...and if the shaft is a good product, additional clamping is not required until that shaft also deteriorates.

6. 300M is the best modern material made, for this application. It will deteriorate slower and last longer than any other current material. 928 International has their shafts made from the original material....Mark paid to have it analyzed 20 years ago. It is "old technolgy", but should last as long as the original shafts....somewhere after 60,000 miles on a street car. (I have no idea of what material you are using.)

Jerry Feather

In looking at the pictures of the splines in question, both in this thread and those posted some time ago by Rob Edward in the 2013 thread, it is obvious that the overlap between the male and female splines is only about 50 to 60 percent. The splines as cut or rolled or otherwise formed in both components are formed much deeper than they need to be, all ending at the bottom of their grooves in a very tiny radius. One thing that is telling me is that although we are pretty well stuck with the splines as formed in the clamps, the ones that might be formed into the ends of the shafts in question might very well be formed not so deeply, ending in a broader either relative flat or even radiused bottom, and with the tops of the "lands" being much more narrow.

What I think this would do for us is to allow for the shaft to be made out of 26 mm bar stock or even larger and then turned to the largest diameter that will still fit into the clamps. Then, even if there is some reason to neck the shaft down next to the splines, which I highly doubt, the result would be a bit larger in diameter than otherwise. Too, even to the extent that there is no need to neck the shaft down next to the splines, the result will still be a significant amount of more material left there at what is still the weakest point(s).

Right now I am looking for a spline cutter with the correct profile to do just that with my shafts.

Another reason for my present interest in this approach to cutting the splines is that I am considering making my shaft(s) out of hollow bar, perhaps 1 1/2 inch with a half inch bore or maybe similar with a 3/8 inch bore. Either one of those is going to tend to emphasize the weak point just next to the splines, so I am trying to design in as much "meat" as I can there and still not have the shaft too beefy. Or I may just go to 1/14 or 1 1/8 inch solid bar and turn the ends down to the larger size I am suggesting in this post.

Just some of my present thoughts.

FredR

Good to see this thread is getting some good mechanical input. That the shaft twists is part of the design intent and that is why there is a flex plate to absorb the dynamic changes in length of the drive shaft. The notion that the drive shaft is over twisting and is likely over stressed at the base of the taper is the crux of the matter. Irrespective of anything else if the clamp cannot hold the shaft the system fails irrespective of what is happening metallurgically to the shaft itself. The notion that over twisting is pulling the shaft out of the clamp as the shaft shortens through over rotation may well be central to the issue but the notion that clamping the shaft securely may be a "bad thing" troubles me somewhat. When I pulled my original GTS motor apart after the PO experienced TBF, the rear thrust face of the crank was shot as was the rear most thrust bearing face and this could only have been caused by the elongated shaft [generated by clamp slippage] causing the requisite thrust. Irrespective of anything else, stopping that slippage is an absolute must to preserve the motor so any mod that achieves such must be appropriate. I can tolerate a snapped drive shaft but a wrecked motor would be the end of the 928 road for me.

Now if I understand what Greg is saying, if the clamp is maintained rigid then over-rotation of the weakened shaft will put more pressure on the forward thrust face and cause excessive wear [or failure] on that side. That being the case we should see increased end float on motors that have the flex plate clamp well and truly clamped so I will certainly be taking another look shortly at mine to see if I can detect signs of such. I will also see if I can organise some in-situ crack detection as I do my annual inspection in the next month or so.

I still cannot fathom out what Porsche were trying to achieve with that tapered design. First time I saw mine was after it snapped and the shaft was removed- I could not believe what I saw design wise when I saw the taper and the resulting mess where the splined end section had been.

Regarding failure of the shaft, inspection of the failed bits clearly showed a lot of axial surface cracking that led to delaminations and when the problem degenerates deep enough the shaft is left with what is in effect a central core that is way smaller than the shaft diameter and at some point the shear stress is too much for the metal left intact and then pure shear failure occurs across the inner core that had some integrity left which in my case was about 15mm [as I seem to remember] and that was not sufficient to sustain full of throttle torque that finished the poor thing off. This happened at 84k km [about 50k miles]. My current shaft has 150k km on it and the interesting question is how much more life is left in it? - not much by the sound of it!

So the interesting question is how many folks have 28mm shafts with 100k miles or more on them?

Rgds

Fred

Jerry Feather

This thread seems to have gone or might be going in a direction not intended at the outset, and that is in the direction of the old unresolved discussion about just what it is that causes the shaft to migrate in the front flex plate clamp. The sort of conclusion that seems to be accepted, at least in this thread, is that it is caused by the shortening of the shaft upon being twisted. However, my intuition tells me that the shafts are not likely to be twisting enough to make much if any change in the length due to shortening. I doubt that the shafts twist any more than about one eighth of a turn and likely not even that much. I don't know of any empirical data about that. Too, I doubt that in any case there could be enough shortening to first pull any flex out of the flexplate in the pull direction before then pulling the shaft out of the clamp.

So I went online and searched for some information about how to measure any shortening in a shaft due to twisting. What I found was some information that says:

"According to Roarks Formula for stress and strain a longitudinal stress and strain do exist, however it goes on to say that the longitudinal strain is a shortening, and that the longitudinal stress is a tension in the outer part and a BALANCING COMPRESSION in the inner part." (emphasis added by me).

Then a guy named Poynting "concluded by saying that a solid circular bar involving pure elastic torsion the longitudinal stress and strain can be a shortening, a lengthening, or zero, depending on the material."

Finally it says that a guy named Poynting "concluded by saying that for a solid circular bar involving pure elastic torsion the longitudinal stress and strain would not be large enough to have any engineering significance."

What that is telling me is that there is still no explanation about why the shaft and front clamp migrate apart under the force of the original clamp. The fact that Greg does not find any such migration of his shaft under the original clamp may be just a function of time, but I wonder if it too might be simply due to the different finish in the splines being cut rather than rolled. It also tells me that there is not likely to be any damage to the other side of the thrust bearing due to the theoretical shortening of the shaft(s), after the clamp is beefed up enough to stop the migration, whatever is causing it.

Constantine

Hi Jerry,

The main problem on the 928 automatic drive line which sets up the whole drive shaft pullout problem which can lead to engine TBF stems from the pinching of the drive shaft at the front flex plate. If this union can be re-designed to emulate the drive shaft and flywheel union of the 928 manual drive line, this whole problem will go away as evidenced by (almost) no TBF stories of 928 5-speeds versions.

We tried unsuccessfully to do so before we came up with the 928 Super Clamp. The main problem was the fix we initially devised did not have the drive shaft being pinned to the back of the crankshaft as in the 928/924/944/968 manual drive line designs. This is needed to locate and support the front of the drive shaft to stop the driveline harmonics from making this union noisy and which can lead to failure after prolonged use since the flywheel plate and front drive shaft splines would be damage from these harmonics.

We ran out of steam on this idea once we realized the amount of work needed to stand up these new parts and the overall costs of doing so. If owners believe our Super Clamp is pricey, then the price for these new parts would have caused them to faint. Maybe the time is different now, but I'm sure there wouldn't be too many takers still.

Jerry Feather

It would seem to me that the answer was to simply use the same or similar drive shaft as with the manual trans 928. That is, two pieces rather than just one. I suspect that the manual drive shaft is two pieces so that the short one can be hardened in order to resist the wear from moving a bit in the clutch disc splines and, as you suggest, from the driveline harmonics. The short shaft for the autos can also be hardened to resist this wear just as with the manual. Otherwise, what is the real difference between the splines in the clutch disc(s) and those in the flex plate clamp?

Constantine

Yes, but try this and you will see that a lot more is needed to get it done correctly in the automatic version. BTDT.

Jerry Feather

Well, that is kind of a non-answer, isn't it? However, as I seem to recall, this discussion went around here some time ago, then also without much of a conclusion. So the solution is to clamp the flexplate more firmly to the shaft! Now, what does that have to do with the present discussion about remaking and/or redesigning the shafts? And it certainly doesn't add anything to the old discussion about why the drive shaft migrates in the flexplate clamp, does it?

FredR

Jerry,

I think one has to look at this as a multi faceted problem. If the shaft is slipping in the clamp the reasons do not matter as much as stopping it from happening. Whether or not this helps the longevity of the shaft remains to be seen. As I have reported many times, when my shaft was slipping through the clamp a very noticeable characteristic vibration appeared at 3050 rpm- a dead giveaway as to what was happening. Porsche could not solve this problem but the Loctite did. Has this saved my current drive shaft? Well it has now covered 100k miles and still in one piece but of course it may snap next time I use it for all I know at the moment.

Whereas way too many shafts have snapped I still think the number that have gone pop is likely to be very small compared to the total number of units produced but without hard numbers it is difficult to form any conclusions. I put the randomness of the shaft slippage down to clamp tolerances and most likely other factors such as drivers with lead shoes. .

Constantine

Jerry,

There are a few different ideas of why the front flex plate loads up which have been discussed to death on this board. Shaft twists and gets shorter, rear clamp is not tight enough and allows movement at the front and currently the newest idea is the drive shafts get more "twisty" as they age.

What is known is that using a better clamp, Loctite in the old clamp or replacing the parts at the front of the 25mm drive shafts as Porsche originally designed the automatic drive line, all help to stop the drive shaft pullout at the front clamp which could lead to TBF.

Changing the drive line of the automatic to mimic the manual version would be the ultimate solution, but will be costly and not give anyone any better results than what is currently being used.

If you want to proceed into your own research into the "why", go for it!

Wisconsin Joe

Not trying to be argumentative, but how exactly is the drive shaft "pinned" to the back of the crank on the 928?

The long shaft (TT) is clamped to the short shaft (intermediate shaft), but the short shaft isn't really "pinned" to anything. It's a slip fit into the pilot bearing, and the clutch (bolted to the flywheel) slides back and forth on the splines. There's no direct attachment between the shaft and the crank that I know of.

I may be wrong, and welcome correction if I am.

GregBBRD

And, then there is the Elephant in the room that no one is considering.....

Torque tube shafts, as they age, twist more.The additional twist, causing the shaft to pull out of the stock clamp, absorbs engine torque (which is why shafts get so incredibly hot.) Changing the shaft to a shaft that twists less transfers more of the engine torque to the transmission.

Simply go to Harbor Freight and ponder the set of impact sockets with 6" extensions built into them (designed for tightening wheels). No one should have to hurt their head over this.....The more the shaft twists, the less toque that can be transmitted.

the flyin' scotsman

My 928 project this winter; replace TT and TC brgs in my '88 S4.

TT gets a new GB shaft and Constantine brgs with new OEM TC brgs from the man who knows his parts, Roger.

YMMV

FredR

Malcolm,

Have you got any symptoms that suggest intervention- strange vibrations at 3050 rpm for instance? Assuming you have the original shaft at 190k km that is a marker for longevity- wonder how many we have with more miles.

Just wondering whether Bill Ball had the 25mm or 28mm driveshaft in his 300k miler.

Rgds

Fred

hlee96

This is what I call, WIN, WIN, WIN!