shaaark89

made an interesting and concerning observation today on my 89 s4 auto wilst preparing it for supercharger install.
background: car has 69k, tbf fix was done 12/00 by myself and earl gillstrom using loctite on the drive shaft splines and both front and rear pinch bolts. end play was .008". marked the driveshaft with paint and i've peered through the hole in the bellhousing cover from time to time to see no movement of the driveshaft in the pinch collar. engine and trans have not been removed since this work was done.
today i pulled the bellhousing cover off and the driveshaft still hadn't moved in the pinch collar but the flexplate was obviously flexed forward quite significantly. released both the front and rear pinch bolts with slight movement, but the loctite on the splines prevented any real movement of the driveshaft and subsequent release of presssure on the flexplate.
spoke with earl and we decided to heat the collar which combined with some heavy prying seemed to releive most of the pressure.
measured my end play and it was still at .008" (car was driven maybe 10k since the fix was done) so doesn't appear that any damage was done.
however, this seems to suggest that there is more going on further back that may contribute to the pressure on the flexplate (ie torque converter ballooning or some lengthening of the driveshaft due to stress or torsion?)
i will continue to check this as time goes to see if it occurs again, but in the meantime anyone monitoring their tbf fix should consider taking the bell housing off and really looking at things.

Garth S

Tom,
When you broke the loktite 290 bond, did you measure the additional release distance? If so, was it comparable to the distance when first done and 'paint' marked? Also, the rear pinch bolt runs in a slot in the prop shaft such that releasing it's torque shouldn't alter the shaft's effective length( or so I thought)?
I applied this release/loktite 290 method to my '88 ~6 months ago, with no discernable movement since: but in light of your interresting post...
For what it's worth, the measurements on this 'new to me' S4 @ 52KKm (31Kmiles) were 0.122" flex plate release and 0.003 max. crank end play.
The first figure strikes me as being close to the max. reported movement - in a car with very few miles; therefore one could speculate that the movement is not related to milage - your current experience indicates time is not necessarily a (linear) factor re. the ageing of these cars. It seems possible that this movement can/could have occured very early in the lives of auto 928's and holds in the 0.100-0.120" range until released - usually, many years later.
The TBF speculation continues....

Erik - Denmark

Tom,
Very interesting news that the shaft is moving forwards/increase length even it is locked with Loctite after Earls method
'The 1000$ question' is still what happen?

If you are sure the shaft has not moved in the splines (for and rear) there is only one possibility left - The shaft grow in lenght as you said due to stress or torsion
I mean the balloning effect that only a rumor and in facht impossible, because if that should happen, then:
1. The rear flexplate (mounted on the converter) should be bend backwards
2. The converter shaft should move forwards and that is not possible due to the rear convershaft bearing, this is before 87 model locked (for forwards moving) by a spring-ring and after 87 model of an reces in the housing (See shop manual page 37-126 and 37-130)
That's well known, that some newer models break the trans shaft, see:
http://forums.rennlist.com/upload/transshaft.jpg , the question is, what happen to the shaft before it breaks - Is it then gradually growing in lenght?

Jay Wellwood

Thanks for the update Tom. Good info.

shaaark89

garth,
unfortunately, the 290 bond is very hard to break and the pressure was relieved only by VERY heavy prying after applying heat with a propane torch for 10+ minutes. even then there was not an abrupt, discernable release, only less obvious flex in the plate. before release, the flexplate was flush against the inner hole in the flywheel, after release there was an obvious gap.
i agree that the movement likely occurs early and reaches a max, then stays there.
as far as the rear pinchbolt goes, yes it does run in a groove in the shaft, but there is a slight amount of play in that groove which would allow movement both fore and aft. the fact that releasing the rear bolt didn't fully relieve the flexplate pressure suggets that there was more going on.

erik,
interesting comment about the torque converter. if in fact the ballooning effect is not possible, then we are left with driveshaft lengthening, which we have postulated might occur as a precursor to breakage.

maybe we can get some input from steve cattaneo about the converter

John Struthers

???
Tom & Erik,
I see, literally, what Erik is saying but I'm having a hard time coming to grips with the TT/driveshaft getting a woody. And I'd need something demonstrable to show where the shaft alone 'grew' enough to to push the crank past allowable thrust tolerances in order to do engine damage.
I have no doubt in my mind that due to torsion/friction ...what-have-you... that there is some play -push- attributed to the driveshaft ALONE. However, I'm going to need some convincing that the TC converter is not the direct line cause of MOST of the fwd thrust.
Steve C. feel free to leap in here with both feet as it is possible that I'm totally wrong.
Someone posted that Mercedes had significant knowledge of TC balloning -not necessarily on the 928 -and that said balloning could cause serious travel of the driveshaft/TT in a forward direction. There was further comment that - while not in every case - once the TC housing/body balloned it usually did not revert to original shape upon cool down or torsional release. I believe there was a hint -not fact- that if you could make a thrust adjustment to within tolerance after balloning that was O.K. as once the initial balloning occured there would not be a significant additional balloning event.
I know, for a fact, is this a problem Steve?- that the mount ears on Pattycakes TC drive flange are bowed FORWARD from center as she sits this would indicate a rearward movement, or truly, FORCED adjustment from the front clamp and one that I would not think the forward clamp/single bolt could sustain unless severely over-torqued.
Also in light of the more or less ridged mounting of the transmission, axles,
and all the suspension/link mounts back there my first suspicion on travel would be in the fwd direction, second would be what's causing the 'significant travel'?
I don't think any of the 'slip spline connections' are perfect. So there has to be some give/play that whatever is expanding has to overcome before moving forward while at the same time pushing rearward as well
(against the relatively rigid and unyeilding transmission).
This, to me, is indicating TT/driveshaft growth numbers in lengths that don't seem possible. If all those aluminum and steel pieces from the front of the motor to the drive axles start expanding due to friction and torsional loads -a thousandth here a thousandth there- I might buy into something other than TC, but it has to bumped against total freeplay and not just allowable to spec. play.
In Erik's defense:
There are a number of things that may or may not influence driveline travel.
Friction/expansion TT shaft, TC (balloning), gear pac expansion due to heat or thrust/torsional loads from the drive axles, aluminum transmission case expansion, expansion of the aluminum engine block due to heat, torsional loads due to detonation, or even overtightening of the T belt -a reach-, and perhaps even chassis flexing.
There is also a possibility that auto shifting both up and particularly down at well above recommended road speeds are a cause.
Further defence of Erik's arguement against balloning is construction of the TC itself -you ever really look at one of these puppies?-
the metal ain't mild steel -anymore- it rings like hardened steel, welded completely around the outer middle joint faces, strengthening/balancing weld bumps on the 'flats' on both sides, and if you look in the valley on both sides-in toward the drive necks there is an additional row of weld bumps for balance/strength. On top of this look at the metals color, it appears to have been stress relieved and either oil or air quenched for even more strength.
Bottom line, I don't know! And I don't have the time/tools or a spare 928 automatic to do an engineering study.
My suspicion is -and not just because I'm lazy- that there is enough background info on this subject that we follow Tom's lead:
Inspect on a bi-annual basis and adjust if necessary.
I need to do this, along with at least half or better of the 928 automatic community.
If anyone doesn't have a will made out;
1, You need a will.
2. Could you leave your Manuals to me? Please?
Postscript:
Glitch on the computer, again. When I correct spelling, etc....words disappear, and sentence/fragments are hopping around . To lazy to start over. Reply is kind of broken in one spot from my end. Nope, not correctable with insert ect. like on Word...
You guy's own 928's so you can figure this out. Sorry!

Garth S

Readily agree that semi-annual inspection of the flex plate/pinch bolt is appropiate, especially in light of additional 'unknowns'.
The definition of this inspection , which initially only evaluated movement of the locking coupling of the flex plate along the shaft splines, must now include a look for additional forward movement of the flex plate(with no further coupling/ spline movement) as per Tom's observations.
Easily enough done - unless further 'release' is required:especially for those of us who have already done the locktite 290 fix..... I'll keep an extra cylinder of propane around, just in case.
Thanks for the 'head's up'

Steve Cattaneo

It was I who first suggested torque converter ballooning as a possible cause of thrust bearing failure to the 928 community.




That torque tube shaft shows signs of twisted spines, a failure like that will shorten the shaft thus reducing the load on the crank shaft... Also a sudden failure like that will not place a forward load on the crank long enough to cause thrust bearing failure. It would take months, if not years for an engine failure to happen.


It is a fact that torque converter ballooning is in part and in some cases directly responsible for thrust bearing failures in engines.

It is a fact that when torque converter ballooning happens in a Mercedes Benz TC it physically increases the length of the torque converter. MB issues a tech bulletin on this.



Up to 1986, the torque converter bell housings have two circlips and the 1987 and up cars have a lip in the bell housing to support the bearings. It would not take much for an 11.34’’ diameter pressurized hydraulic rotating ram mass (TC) to exert a forward pressure of only thousands of an inch... That is all it would take to lock the crank against the thrust bearing.



The TC is a possibility and what makes it more of a possibility is that the torque tube drive system; rear and front flex plate do not flex so any forward movement is transmitted directly to the crank.




The question is why do most thrust bearing failures happen with automatic transmissions with torque converters?

shaaark89

steve,
is the tc ballooning a progressive thing or does it balloon to a certain stage and then stop?

Steve Cattaneo

Tom,


The TC is charge with fluid once the engine is started. It is under the most loads when moving the car from a dead stop that is when torque multiplication happens at a ratio of 2:1 until 2400rpms.


I think I found the simple solution this problem we can do a group buy. A part would have to be made and tested. No profit involved only for the benefit of the 928 community. I have a busy shop so with work and putting up a new building my time is very limited. I would need help are you interested.

John Struthers

Steve,
What do you need in the way of help?

Still curious about the TC. That kind of construction being susceptable to 'balloning', but then I've seen bent and snapped rods due to the uncompressiblity of fluids.
I was wondering about the flex plate comment...
I can 'feel' the mount ears thru the inspection hole and after rotation all ears seem to be bowed forward - ears forward, center back- .
This might be a tactile mistake on my part.
I was thinking the same thing about the load/force being transmitted thru the center/splined bearing area almost exclusively and that it would be sudden rather than gradual.
Is that your take on this topic? Or am I simplifying?
I'll give what assistance and cash that I can towards a cure...for totally selfish reasons, of course.
When, not if, Murph and Company address the 4.5L engine crowds NEED for a SC kit I think the auto owners are going to see more of this problem rather than less.

Erik - Denmark

Steve,
Maybe you are an expert in transmission and converters (I am not) but here, I mean, you are you are on a 'wrong track', and I risk to give my comments:
1. I mean the ballooning effect for a 928 is a myth
2. I agree - Was only shown as an example to show that the shaft can break, this picture is from one of the 'Danish GTS vans'
3. Did you see this yourself in a 928? How did the rear flex-plate look? How was the condition of the converter shaft bearings?
4. OK, but the Mercedes has no converter-shaft - As I remember the converter is placed direct on the flywheel
5. Here we are not talking of thousand of inch (0.025 mm) we are talking of several mm's movement of the trans shaft (Remember the two flex plates)
Has you seen the circlip/lip damaged as you said? I think the rear bearing will collapse before the circlip/lip
6. I do not understand what you mean - Please explain again
7. Because the manual cars has the clutch where the shaft can move in clutch dish plate
Regards from Erik in Denmark

T_MaX

Steve, I think it's great that you should offer your experience and whatever else to help solve this problem.

I'm in!

If it helps, there is a guy close to me (Brownsville, Tx) that I have worked with on custom converters before. He primarily does rebuilds for sales to Trans-shops, but will do custom work if he knows you.

Time and money are tight right now, but I will do what I can.

Steve Cattaneo

HI Erik,
I am only saying Torque converter ballooning is a possibility.



On the last 5 928 transmission that I rebuild all had bad, worn TC bearings. The rear bearing next to the TC had the most damage. The inner race had a lot of lateral movement. The snap ring only supports the outer race of the bearing. Contrary to your belief the rear flex plate flange can move forward in the bell housing pushing on the shaft. When I have more time I will post pictures.

What facts do you have to support your shaft flex theory? BTW manual transmissions engine do suffer from TB failure. Not as often as automatics.


I think the fix to this problem is to replace the front pinch bolt style flange with a free floating York style flange. This would allow the torque tube drive shaft to float in the front. So when the toque tube shaft migrates forward there would be no forward load on the crank shaft.


http://www.mastertechtrans.com/post/TT 002.jpg[/IMG]



Questions ?????????????????????????????????????????/

1 Can a machinist make a custom solid flange from the old one? Or does he need a cad drawing?
2 Can anyone recommend a machinist that does excellent custom work?
3 If he needs a cad drawing can any one he draw one?

4 I have been told that billet steel can only handle 600HP. Does anyone know of stronger steel or other metals for this application that can handle more HP?

For Jim or mark from 928 international.
5 Do all models of automatic cars have the same torque tube drive shaft length, diameter, and the same front spline count? If not which ones are the most common?


6 Do all models of automatic cars have the same front flange? If not which ones are the most common?


John, Max, thanks for the offer. The help I need is leg work and phone calls.

For those of you thinking of SC your cars this is a must along with a custom billet drive shaft.
I am not interested in any profit from this venture. If it works we can do a group buy. The parts would have to be tested before any buys are made.

Maybe Mark Anderson can use his track car as a test mule



Any comments are appreciated.

Tony

Constantine in FL has been running his own version of "coupling" for some time now on his track car...nearly 2 yrs is my guess. I think he actually tried a free floating design but with no luck??

Im not sure if he reads the list or not but perhaps he will chime in with his findings. With your knowledge of the 928 tranny Steve and the work Constantine has behind him already perhaps you guys can put your heads together?




added: If i recall the new corvettes use a layered rubber and metal sandwich....rubber...metal..rubber..metal..etc etc... for the coupling. Its is strong enough for the torque yet has just enough compression to absorb for and aft movement. I think investigating the corvette set up will lend some ideas for the cause. Anyone have any links..leads?