Reconstructing Constant Velocity (CV) Joints

[Tut]

As I recall, my '87 has moon plates but no Schnor washers. I have new original-length bolts, as well as Schnorr washers. Is the thichness of the Schnorr washers going to cause a problem by "shortening" the bolt, thereby leaving fewer through-threads exposed?

[Wil Ferch]

Tut:

The answer to your question was posted by Grady somwehere within this post.....I touched upon it briefly in my note about thread engagement and protrusion beyond "flush"....

Wil

[Grady Clay]

Tut,

With your ’87 you should have the 108 mm CV joints and six M10 bolts. If so, you have the support plate 911.332.191.00. If you somehow have the 100 mm CVs with M8 screws and support plates, we need to find the part number – it isn’t in PET. An easy way to tell is by the Allen wrench for the CV screws. An 8 mm wrench is for the M10 screw, a 6 mm wrench is for the M8 screw.

To answer your question, you probably need to try it in situ and measure the penetration.




Assemble the CV joint to the transmission flange with clean steel-to-steel = 4760 ft-lbs. Get grease on the interface (the usual situation) = 950 ft-lbs. Normal full throttle in first gear with a 3.2 is 1191 ft-lbs (dynamic loads may be 2-3X that).

I find it hard to believe that Porsche didn’t consider this when changing from the big 108 mm to the 100 mm 923 part.

From Jim’s calculations it appears that the 100 mm CV joint is normally very close to the limit regardless of Schnorr washers and such. If there is grease on the joint surface, the connection will almost certainly slip in normal operation. It is next to impossible to attach the CV joint to the transmission axle flange and not get moly grease on the mating surfaces. This makes the case for installing an end cap on every CV joint that doesn’t have one.

It now makes sense why Porsche includes an end cap with the CV boot kits.
The 100 mm kit is 923 that I can’t find a P/N with the end cap.

What does everyone think of THIS “cap” that replaces the gasket on a 108 mm joint? Please be polite.

Installing the end cap isn’t a perfect solution because 911s with end caps still have CVs come loose.

Something we need to look at with the end cap is the effect of having “another piece of meat in the sandwich.” With the end cap, there is now a plated thin steel spacer between the CV joint and the flange. The instructions for installing an end cap include applying sealant between the CV joint and the end cap. It seems to me this is similar to greasing the interface. It might also make “settling” a greater issue. There has to be a better way to seal the grease inside the end cap and boot flange while insuring metal-to-metal contact. Suggestions?



Of course the real solution is to replace the 100 mm CVs with the 108 mm, six M10 screws, Schnorr washers & support plates and an end cap without sealant in the “sandwich.”

The 108 mm Turbo set-up
2 each Stub axles 930.332.232.04
2 each Lock nuts 900.910.093.02
2 each Axle assemblies, 108 mm CV 930.332.037.04
2 each Transmission axle flanges 915.332.209.12
24 each M10 socket head cap screws M10x50 900.067.123.01
12 each Support plates 911.332.191.00
24 each Schnorr washers M10 999.523.103.01

The 108 mm Carrera set-up
Stub axles not needed
2 each Lock nuts 900.910.093.02
2 each Axle assemblies, 108 mm CV with friction welded stub axles 911.332.024.14
2 each Transmission axle flanges 915.332.209.12
12 each M10 socket head cap screws M10x50 900.067.123.01
6 each Support plates 911.332.191.00
12 each Schnorr washers M10 999.523.103.01

Please don’t take these P/Ns as gospel; there are supersessions and these might not be the right parts for the conversion. Some research help would be appreciated.



Porsche should supply these parts kits to owners at a bargain price – one that would cover their unburdened manufacturing, shipping and distribution costs. They probably could even get GKN Lobro GmbH to pitch in.

They may be practicing the “let sleeping dogs lie” philosophy from prior incidents but one of these days someone is going to have a CV joint come loose – with disastrous results for everyone. A phineg of prevention is worth a €B of judgments. Classic 21st century CYA.

Best,
Grady

How do you spell the unit that is 1/100 Deutsche mark?

[Tut]

Thanks, Grady. Will do.

You had indicated that you'd like some Schnorr washers when I received them. How many?

[Wil Ferch]

However, your question was how much extra thickness is a sandwich of Schnorr washers and moon plates...Grady says 1mm + 2.5mm, respectively, or = 3.5 mm.

If you didn't have these pieces installed OEM...and you intend to add the washers and moon plates...then you'll likely need bolts 3.5 mm longer. Since bolts come in ( usually) 5 mm length increments, this may be a problem, but maybe it's not if you first check the backside penetration is still 1.5-2 threads.....

Your option then is to find these as Porsche parts that may be used elsewhere and come in 1 mm length increments...or having them pro-machined to length w/o damaging the shank or remaining threads.

Grady or Island....the only alternative is to keep the mating surfaces really clean. This has benefits regardless of the design safety factor applied.

Any advice here how to do that?... or is it simply a matter of being REAL careful and trying to be surgically clean?. Is this a case like the plumber's trick to stuff white bread into a copper pipe ( to hold back some water) so you can get applied torch heat to work on the solder- joint?. The bread later is "consumed" in the flowing water to no ill-effect. Do we need to find something like that ( in principle) applied here?

Wil

[randywebb]

re: clean thread surfaces
I tried to be careful.. then used Q-tips, and brake cleaner on more Q-tips to clean the femal threads. next, run the clean bolt in, pull it out and inspect. Repeat cleaning until done...

[Wil Ferch]

Randy:

Taking nothing away from what you said in terms of thread cleanliness...

I think...(in the context of torque load capability)....that we're talking about clean mating surfaces of the CV joint with the matching surface on the drive axle flanges.

Look at the substantially different torque transmitting values of the CV.... clean vs greased.

- Wil

[randywebb]

yeh - but just wipe that - cloth w/brake cleaner on it

seems like the roll pins should carry nearly all the torque (?) what am i missing w/that?

[76carrera3.2]

If clean mating surfaces are so vital why not a cross hatch or something similar on the surfaces so they would adhear to each other better when clamped together. Do smoothly machined surfaces offer better clamping and therefore higher torque or is there a better alternative.

[randywebb]

smoothly machined surfaces

- maybe - machinists blocks can be 'wrung'together and cohere so well that they will support heavy objects in tension

[Grady Clay]

Jim can probably be more accurate here but my take is:

The screws when torqued provide “clamping force” between the CV joint and a flange (stub axle or transmission).

It is the steel-to-steel interface that, with clamping force, provides the shear strength that allows the transmission of torque without the joint slipping.

The screws should only provide clamping force by their tension. They should not be part of the shear between the CV joint and flange.

If the interface slips, then the screws are in shear (and bending) and bad things happen. The screws are put in bending and shear while still in tension. If this is cyclic (repeated slip back & forth) the screw is likely to unscrew. When that happens the clamping force is reduced and the failure becomes progressive.

It is the interface that needs to be free from grease (difference of 4760 ft-lbs vs. 950 ft-lbs) for the joint to not slip. The more the mating surfaces are oil-free, flat, parallel and smooth, the better the clamping force-to-shear strength ratio. Sure, gauge block level of surface finish is ideal but impractical.

I think Porsche had it right with the 4-bolt & spiral pin arrangement with 108 mm CVs in ’69. Too bad the bean counters prevailed. I’ll bet had the engineers stood their ground, PAG would have made more money and customers would be happier.

Best,
Grady

[Jim Sims]

"4-bolt & spiral pin arrangement with 108 mm CVs in ’69."

One problem with these was getting them separated; probably more than one wheel hub or transmission output bearing met an early death from the "pounding" that was usually applied to get them apart. They badly needed a couple of tapped holes on the back side to use to jack them apart with screws - likely the bean counters again.

[Wil Ferch]

So.... for us 100mm guys with 8mm bolts....lets keep the mating surfaces clean...eh?.

Once done, then what ?.....Schnorr/moonplates added too? ( my 85 ...as noted....has demountable CV's on both ends of the halfshafts..with 100mm CV/ 8mm cap screws...no Schnorr/moons).

Not to derail....but why isn't a lock washer needed for "wheel" bolts and/or studs...it sees the same type of cyclical forces? Size? Contact area?

- Wil

[island911]

yep, dry . . that's been my soapbox for a while now.

since you bring up hole patterns and forces...

I saw that Grady posted three reasons for the moonsplates . . .thought I do doubt that those reasons hit any nail on the head - so to speak.

Anyone who has done much machine design knows about tolerance build-up that exists when matching multiple holes of matching parts . . .that is; the more holes you have to match-up, the harder (costly) it is to execute.

. . . if you had both the CV's and the flange match-drilled, you could have very precise bolts in very precise holes . . . such that, even if the fasteners are lightly torqued, the CV and flange can have little relative motion.

Of course that particular CV isn't too likely to fit on the opposite side tho -- thus the need for either tight ($) control over dimensions, or loosening of the tolerance.

The moonplates only have to hold 2 tight tolerances each (hole size, and the linear distance between.

that's my thought for the day.

[Matt_West]

Wonderful thread, but seems to be missing the piece of information that I need...

Yesterday, turning right out of a parking lot I heard a load bang followed by a rumbling in synch with the wheel speed. Drove real slow and pulled into the nearest parking lot.

Looked under the car and saw the loose outboard CV bolts. Removed the drive axle and found two bolts sheared. One spun easily out of the hole. The other is stuck. I made a drilling fixture but even so my bit walked and drilled the outer edge of the bolt and possibly some of the threads...

Minimum it seems I need to pull the thing that the six M8 bolts of the outboard CV screw into. From my reading it seems that I need to remove the large 32mm nut. Is this correct?

If so I will pick up the requisite 5'ft of galvanized pipe...

Thanks for the help.

[rcecale]

Matt,

You'll find all the info you need in this thread.

Randy

[Matt_West]

Thanks Randy,
Took off the axle nut, easy job with a five foot cheater bar. That provided great access to the sheared bolt.

I drilled a piece of bar stock and used two bolts to mount the flange to it. Clamped the bar stock in the drill press vice and used a center drill to start a hole in the bolt.

Switched to 5/32 drill and went all the way through the sheared bolt. Used an EZ out to remove the bolt. Chased the threads with a re-threader and am ready to reassemble tomorrow.

Will of course review this thread again to be certain to orientate washers correctly etc... I am also thinking about searching for that picture of the wired CV bolts.

Thanks again for the help.

[Tut]

Grady or Wil,

A while ago there was discussion about the bolts extending 1-2mm beyond the flange as being the ideal situationto to assure proper purchase, holding, call it what you will. Yesterday I crawled under my car. I have M10 50mm bolts. At the trans. end from the surface of the moon plate (no Schnorr washers present) to the end cap on the CV is 36-37mm. The flange is 15mm thick, including a "bump-out" to increase the thread depth. So the bolt is recessed 1-2mm into the bump-out. For the purposes of this discussion I'm assuming they are the original bolts, but we all know what happens when one assumes.

If I add the 1mm Schnorr washer that would mean that the new M10 50mm bolts that I have would be recessed 2-3mm, but would still have 12-13mm of thread depth in the flange. If I purcahse 55mm bolts I would have 2-3mm extending beyond the flange. The flange is awfully close to the sides of the case, and I'm concerned about the extended bolts coming in contact with the case. If my assumption regarding the originality of the bolts is correct, thereby indicating that Porsche didn't feel the need to have the bolts extend beyond the flange, and acknowledging that Porsche is not infallible, am I running a real risk (real risk vs. ideal situation) if the bolts do not extend beyond the flange given that there would still be 12-13mm of thread depth in the flange with 50mm bolts?

Thanks.

[Grady Clay]

Tut,

I would get the longer bolts and new Schnorr washers.

Before you disassemble anything, remove one bolt and check the clearance between the end of the new bolt and the case (without the Schnorr washer. Remember the axle flange has about 1-2 mm of axial free-play so you need to force it toward the transmission.

If there is contact you can add washers under the head of the bolt until you get the desired 1 ½ threads through the flange. This will let you calculate how much needs to be removed from the threaded end of the new bolts.

My local largest metric fastener supplier tells me I should figure on 25¢ per bolt to have them shortened in 100 quantities. In these lengths they are standard in 5 mm increments.
.
I think everyone agrees that the bolt should always extend through the flange.

These bolts have a thread pitch of 1.50 mm/turn. If your 50 mm bolts are recessed 1-2 mm and you add a 1 mm Schnorr washer, that should leave a 55 mm bolt protruding a correct 2-3 mm (1 ½ - 2 turns).

You said “…if the bolts do not extend beyond the flange given that there would still be 12-13mm of thread depth in the flange with 50mm bolts?
No, I just measured a flange and it is only 12 mm thick less some countersink on the CV joint side. I recall on of my earlier posts counting slightly less than six turns of engagement from the first thread to flush on the back side.


I’ll speculate you originally had 55 mm bolts and a PO replaced them with 50 mm. Upon installation they left off the Schnorr washers to get any thread engagement. When you have things apart; look for the tell-tail Schnorr markings in the plate washers under the heads of the bolts and on the back side of the plate washers (in case some were turned over).


Pay close attention to having the mating surfaces of the CV joint to flange clean and free of grease. Jim’s calculations above were an eye-opener for me.

Best,
Grady

[Tut]

Grady,

Thanks for the info. I'll order 55mm bolts from McMaster today.

By the way, I purchased a box of 1000 Schnorr washers directly fro Schnorr some time ago. You had indicated that you wanted some. How many sets do you want?

Regards,