blown 944

Just an FYI if you thread a bolt in it will make quite a clink every time you engage the clutch. I know

fortysixandtwo

I haven't taken one of these apart yet, but will a spring work instead of a solid spacer? It will allow movement, but keep the shaft from migrating out of the desired location.

TonyG

I'll take your word for it but it doesn't technically make sense. The axles shouldn't move laterally due to load since each ball/race is at an opposite angle adjacent ball/race. Thus the lateral loads are cancelled out.

TonyG

blown 944

Doesn't work that way. There must be some inertia that sends the axle outward. I've had a couple where the shaft would move all the way out and smack the stub shaft making a good racket

I don't think you could fasten a spring properly.

Van

Is it possible to retrofit a 996 axle? It's a bit of a different design - the stub axle in integrated on one side. The CV joints are also larger than 951 size.

blown 944

Van, I'm not familiar with them.you could change the inner flange easily but the outer would require some machining of the banana arm to fit a larger stub.

I am looking into some of the fwd Audi axles to see if the vanagon joints can be fit to the Audi shaft on the outside at the very least.

Some of the Audi axles are a torsion design and may hold up better for what I'm doing.

I am also looking into some BMW shafts and flanges.

TonyG

The axle moving laterally due to inertia is different than a torque applied to the axle causing the axle to move laterally.

My point is that the the torque of the engine doesn't cause the axle to move to one side or the other. This is impossible because of how the bearing races in the cv joints are designed. One race is at x angle and the next race is at the opposite of x angle, and so on... Thus any lateral (side to side) movement caused by torque applied to the axle are cancelled out. So there won't be any movement.

Now it is possible to assemble the cv joints incorrectly so that this (as well as binding) can occur. But if they are assembled correctly... the side to side torque is cancelled out.

Anyway... I think we should have axles that have cv joint cages designed so that they cannot move past the race. This will stop any excess lateral axle travel and will stop the axles from popping out of the cages (again... this is how the GKN Motorsport CV joints are designed.

TonyG

TonyG

eric951turbo

this is an excellent thread, thanks robstah for this info.

But I have one junior question : )

Is the right axle shaft assembly the same part as the left side on all cars... including my 951?

I got confused looking at this to buy one for my right side (it's clicking) in Canada :

http://www.autopartsway.ca/PartList..../pagenum1/tabS

If you click on the link, there is one made by FEQ that says Rear Axle Shaft Assembly only.

There are also the ones by EMPI, one for the right side, and one for the left side, with different part number.

There is also a 12$ difference between the FEQ and the EMPI

blown 944

Tony, I think the reason they move to one side (in my case the outside) is due to wear..

For everybody:

I did a bit of disassembly yesterday and suite a bit of measuring. I took a few photos as well.

I measured the vanagon axles and also an axle I borrowed from Kyle (mooreboost)

They both were machined down out toward the joint area.

I then took apart a gnk 951 axle and found it was not turned down. Iirc the replacement axles were .945 ish where I broke it and the 951 axles are 1.080 ish all the way down. See photos for exact measurements.

Before I remebered that I had a 951 axle hanging around, I assembled the borrowed axle (I'm assuming it may be a 944na axle, or some rfpkacement) I was able to install the good vanagon joints onto the shaft.

I then found the gnk axle that had been moving too much laterally in the past and making noise (reason it was replaced. In I was able to install one side of the joint . The other end of the shaft was too mushroomed.

So I'll VE looking for a different shaft to install the beefier vanagon joints onto. This may be the final solution as the vanagon joints do not have the cage machined out as far like the 944na? And the 951 gnk cages are..

See photos for some detail;

broken shaft


vanagon joint vs 944 and 951


assembled on 944 shaft


vanagon and 944 na measurement at breakage point


951 gnk at same location

m73m95

It wouldn't be that difficult to have a couple of axles machined. Make them out of chrome moly. Then for sure, the only thing you would have to worry about breaking is the joint.

blown 944

Mark, I've never broken a 951 shaft. They are 1.080 all the way down. I don't think they will break so really no need to re-invent the wheel here.


for absolute clarification..
My final solution to be able to get the strongest axle possible out of readily available parts is:

951 thicker shaft
Vanagon joints
delrin buttons threaded into the end of the shaft to eliminate excess lateral movement

the only real weak link I can see is the vanagon cages. They havn't been fully tested yet. Those cages could be machined out of a stronger less brittle material pretty easily.

m73m95

D'oh. I knew that

Glad the 951 shafts are the direct fit!

Anyway,

The delrin button sounds like a good idea. But I still can't work out in my head why the shaft moves so much.

When you launch the car, and squat the rear suspension, is it possible that when the rear axles straighten out, they become almost to long?

blown 944

I don't think so...actually no..

The lateral movement I have witnessed with the worn gnk that has the mushroomed end. I watched as Dave went forward and backward engagining and disengaging the clutch. The axle just went straight outward every time making a huge clunk. I think that after the cage becomes worn in that they naturally migrate to the running position. Once there is too much wear they go further than wanted. This is my theory..

If I had known now what I know, I would have flipped the axle to see if it would have then moved inward to confirm.

m73m95

I have a spare set of axles, so I'll have to check.... But,

What if you just flipped one of the joints, running it inside out? Maybe the way the grooves are machined for the *****/cage forces it out under load? Flipping a joint might make it stay centered, but still allow it to flex like its supposed to?

If its mushrooming steel, maybe delrin won't be strong enough?

odurandina