Bill Ball

Thanks, Stan. The WSM makes note of carefully LIFTING the TC out of a vertical tranny. I'm wondering if we are going to do more harm than good trying to get the TC out and in when the tranny in the horizontal position. [LATE EDIT: In an S4 and newer, there is not enough room to remove the TC with the tranny in the car. The TC will come out half way, but no more before it hits the tunnel. Then you will be screwed as that will wreck the seal behind the TC. So, if you want to replace any seals behind the TC, you MUST drop the rear suspension and crossmember and then the tranny all the way out of the car.]

Bill Ball

All good questions, LO. I'm sure it's not cast but forged. Hard to say what the center looked like as it got distorted by all the post break rubbing. It would be interesting to do an autopsy, but I would not be know what I was looking at metallurgically speaking.

Hope to see you at the installation party.

69gaugeman

I have seen fractures like this before, Usually it stems from stress cracks at the root of the spline. I am not sure if they are heat treated or not ( I would think not ) but alignment issues will cause this kind of failure. (the splines would be 'flexing' each time it rotated) . I am not a spline expert but spend alot of time with ones who are. I also spent a couple of years in driveline repairs. Saw a few that looked just like this.

borland

I recently overhauled the transmission on my S4, so after seeing those photos, one thing that came to mind for me, was you could actually do a complete transmission overhaul without removing the transmission from the car. But that wouldn't be too practical or very easy.

On the converter seals. Don't go there. There are no converter shaft seal (wheel bearing type seal) or primary pump seal (flat gasket) leaks on this car's transmission.

WSM, also calls for positioning the transmission in the vertical position, while removing/installing the converter casing. That would allow the torque converter to rest on it's thrust bearing. One point that you probably already noticed, is that there are no alignment pins for the drive flange to torque converter bolts.

Since your not going to do this converter casing reassembly with it in the vertical position, you should tighten all six bolts bolts by hand first, then tighten them by torque wrench. That will require rotating the torque converter and should pull everthing back into alignment. You could check the installed torque converter by hand to see if there's any bearing slop, but I'll bet you don't find much if any.

Bill Ball

borland:

The TC seems solid. I see what you mean about the drive plate to TC alignment. I sensed no movement/sag when I removed the bolts, but we will do the hand snug , rotate, torque steps, as you mention.

I appreciate you point about the seals. I'd rather leave the TC alone.

SharkSkin

Matt, that's the idea, whatever is needed to make the same length extension as we used, with 3/4" extensions. Make sure we can adapt to the socket too, if all you have is 1/2".

Ernest, that second link is almost humerous. We would have probably twisted that thing 90° and/or snapped it.

BTW the seal I was thinking about is the red one shown 1/2 way down this SoCal928 thread.. I asked Keith about how he did his, hoping they are similar enough for the info to be relevant.

Also, I think it's pretty clear that the long splits were there, and one had sheared off, before Matt got the car. Stress fractures then later rust is the killer here IMHO.

Bill Ball

"Also, I think it's pretty clear that the long splits were there, and one had sheared off, before Matt got the car. Stress fractures then later rust is the killer here IMHO."

Dave, it certainly looks that way. It baffles me that 3 weeks earlier there was no evidence at all to the naked eye. You'd think there would have been some fine crack lines, rust, whatever - something. There was nothing there. It looked like my TT shaft. The strange thing also was that there was almost no preload (less than 2mm), and we believe this was the first time it was touched since this was a single owner dealer serviced car. The other fella from Australia reported the same thing - no signs two weeks prior to a break that looked identical to this.

Perhaps we were not paying close attention, but I don't think so. We stared very intently at the shaft, and particularly the exposed area where the break happened as I released the clamp so we could observe the preload release. So, to me, there does not appear to be a way to prevent this or detect warning signs of failure.

I wish we had taken pictures. This may not be real helpful, but here's one from Brian's TT a few weeks earlier. Matt's looked the same.
Had we seen the second image instead, I would have been real concerned....

GregBBRD

Really nice thread. Do you think it will go back together the same way when there is 4 more inches of torque tube shaft present on the new shaft?

Are you adding the shims and cup retainer with the circlip onto the early shaft to prevent TBF? It is really hard to get the circlip onto the end of the shaft and measure the required amount of shims needed to get the correct loading on the flexplate, but it can be done. Everyone that we've done has used almost the exact same amount of shims, if you need a starting place.

edco

Hey GregBBRD,

We measured the distance or the new TT and it looks like we’re ok on the overall length. Yes we are adding the shims and cup retainer with the circlip onto the early shaft. It would be great to have an idea of how many shims to get us in the ball park we’re going to do a complete write up on this once we finish, and post it to my web page.

Bill Ball

Hey Greg:

It has to go back together!

The tranny is tilted at such an angle that the front of the TT was well below the flywheel during removal, so shoving the new shaft into the rear coupling is simply a matter of positioning and some brute force, perhaps drawing it in with long bolts if needed. The front end of the shaft better clear the flywheel when we swing it back up. If not, we will just loosen the flywheel bolts so the flywheel can be tilted to give more clearance.

We don't have the shims/clip/bushing yet - they are coming from Germany. My understanding from the WSM is that the shims are set to hold the flexplate 0.3mm from the flywheel surface. So, with the flexplate loose, we will install too many shims so that the flexplate is held much further away from the flywheel, so the distance is measureable easily, figure how much excess clearance there is and slide the plate back and remove that amount of shims, then bolt the flexplate to the flywheel. I don't think you need to do all those X= A-B+0.3 measurements. Just measure the unbolted flexplate as I decribe or just use a 0.3mm feeler gauge before you bolt the flexplate down.

SharkSkin

Bill, I don't have metallurgy background beyond the most basic fundamentals, so I'm only speculating based on observation. I have no doubt that you couldn't see the fractures... there was probably just enough of a patina on the parts to obscure any evidence. Having thought about it for a while though, I have an idea for detecting these fractures that run lengthwise down the shaft.

Basically, just get a q-tip with some light machine oil on it(like 3-in-1), and paint a line around the bottom(visible) arc of the shaft with the oil. Watch carefully to see if you can see any evidence of it being carried along hairline fractures by capillary action. Wipe up excess oil, turn crank 1/2 turn and repeat.

That may or may not work, and maybe most TTs will show some fracturing -- I don't know. Still, even if some minor fracturing is found, I'm not sure there's any way to evaluate if/how long it will last, and the best answer may still be just to drive it till it breaks. Seems like we're in new territory here, just beginning to gather data.

GregBBRD

Bill:

Doing all the measurements with things apart like you have them will not be very easy......you are right on top of the program in doing it by trial and error. I can get you very close to start with.

Yes, the WSM does actually want the flexplate preloaded to the rear of the car with the shims in place by .3mm. Note that this is with the torque tube shaft pulled all the way forward so that the rear bolt bottoms out in the rear flexplate assembly and then that bolt gets torqued. Porsche definately did not the flexplate to ever get the chance to flex forward! I set up all of the ones I do at "zero"....or as close to that as possible. You will find that the "ears" on the front flexplate usually have a fair amount of runout and can be moved very easily. Make sure the front flexplate is as straight as possible....you can check it with a straight edge while it is out. Many flexplates get so distorted that one can never get them straight and need to be replaced. Mark Anderson still has a good stock of used, straight flexplates.....in case you need one.

Almost every late car that we have done this way has taken 7-900 234 117 02 shims (1mm) and 1-900 234 116 02 shim (.5mm).....plus or minus the small shim. I keep all the shims, collars, and snap rings in stock, if you guys get needy and tired of waiting for Germany to deliver......I don't know how you ordered them.

You need to trust me on this: You will not like putting that snap ring onto the end of the shaft more than once......it is not a "user friendly" job......and that is understated.

gb

Bill Ball

Greg:

OK, I can see the snap ring could be the issue due to restricted access.. I wasn't sure what the problem was, but you had previously commented this wasn't a fun procedure. We will check the flexplate for flatness.

Thanks,

GregBBRD

Bill:

Yes, when you get the circlip installed for the final time, you will officially be a "mechanical ninja".

gb

SharkSkin

I'll bring some stuff out of my bag o tricks that I use for reaching into impossible places...