Differential output flange not staying in place

[PbFut]

Has anyone else had issue with the output flange not staying in place? Mine randomly pops out of position. I can’t figure out the cause. Bad enough, this weekend it pulled so far out the spline disengaged and I lose all drive. Ruined my weekend as I did not notice the cause until I got home on the trailer.
I have two ideas to correct. Cut the grove for the spring clip keeper a few thou deeper making it harder for the clip slide off. This is dangerous. If too deep the clip will not come out of the grove under pressure and the flange won’t be removable. The other idea is turn down a spacer similar in idea to a hub centric wheel spacer and add where the CV mates to the Flange. I have about 14 mm of room on each side that would then create less pull on the flange by the CVs on the half shaft..
It seems to happen when I am on track in warmup. If the drive train is not loaded in a turn, the half shaft pulls out the flange. Or at least that seems to be happening but I am not 100% sure this is truly when, how or why it pulls out.
Anyone else have this issue on or off track?

[gmushial]

Two thoughts: is it always happening on one side vs the other? Might that clip have softened and offer less resistance to being pulled out? The other thought being - is it possible that some grime or maybe a chip has gotten in the groove keeping the clip from seating properly? Also, any understanding on why the half-shaft might be pulling excessively on the hub? Seems that with your tight suspension there should be even less needed travel than on a street car - a trailing arm bushing going away; or maybe one of the tangs on the subframe has unwelded itself? But for me: it'd try to answer two questions - is the clip really seating properly, ie, when you seat and unseat the output flange, do you get a nice clean click/snap in and out - if not, why not; and two why is the half-shaft pulling outward hard enough to unseat the flange?
good luck - greg

[PbFut]

Seems to be either side but the side with new halfshaft will do it more. Groves are spotless. Replaced clips. Seemed better but has returned. Agree on suspension. I am wondering if it's happening in the pits. Thinking about it, seems it may happen when I lift car on one side to swap tires. I noticed it happened after a tire swap in the past but until now did not put the two actions together. Wonder if it's pulling the clip lose then working out over time. Still, why is it pulling lose to begin with just by lifting the car. Alignment is nothing crazy. About 2.5- camber, zero to a tick toe out. It's all just a guess.

[Rabbi Doom]

why is the half-shaft pulling outward hard enough to unseat the flange?

This. Especially on a car lower than stock ride height where there is even less play in the half shaft.

Maybe the spline or socket are worn from repeated thrashings that it's forcing the flange out of place. If there's a little bit of play in the spline putting lots of force on it could pop it out and once it starts happening it's just going to get worse because each time it comes out it's going to deform it more. I've seen it happen on heavy equipment but never on car.

[gmushial]

Dan - are the output flanges special, or just run of the mill E30 flanges? If nothing special, what about getting a pair from the junkyard, or maybe Jonathan will have a set around, and swap then in and see if that doesn't change anything [per Rabbi Doom's comments] - if such changes the incidence of pulling out, then you have at least part of an answer. The other thing I'd try is to tie a wire from the back side of the wheel bearings (one on each side - but away from interfering with the half-shafts and generally out of harm's way) to the diff case - either set it up tight so that if the trailing arms are flexing/moving outward (providing the pull on the flanges) the wire gets broken; or set up so that it's taut, but can be pulled, so that when you get back to the pits that it's a witness of the maximum outward displacement of the arms, ie, if you go out with the wire taut and come back with it sagging, that'll clearly tell you that the arms aren't anchored (or in the other case, if it's broken, that'll tell you the same, but not give a measure. [in the breaking wire case I'd use something pretty thin - maybe 20ga; in the other case I'd use something with more "memory" - maybe 14ga.] W/re alignment - I don't think that's it, ie, as you say, it's not crazy. I'd look at: where is the pull on the flanges coming from, especially if you've swapped clips and the grooves are spotless. Best of luck - greg

[PbFut]

With the two replies, I tend to agree and lean toward the flanges just being a little worn. Splines look good except the end of one that popped so far out it lost drive. I am thinking I'll buy a new one and new clips. Install and see if the force to pull out increases. I recall originally I needed a pry bar. Now I can pop the clip with my hands. Could be the lip of the grove has worn allowing the clip to slide off easier, but seems unlikely. The spline shaft is very hard metal. This will be far easier than turning a spacer.

[Gofast]

You're not spinning the wheels at full droop, are you?
IIRC, the CV joints aren't spec'd for the angle created by full suspension droop, so they don't like to turn down there.

[gmushial]

With the two replies, I tend to agree and lean toward the flanges just being a little worn. Splines look good except the end of one that popped so far out it lost drive. I am thinking I'll buy a new one and new clips. Install and see if the force to pull out increases. I recall originally I needed a pry bar. Now I can pop the clip with my hands. Could be the lip of the grove has worn allowing the clip to slide off easier, but seems unlikely. The spline shaft is very hard metal. This will be far easier than turning a spacer.

But this again raises the question: where is all the force coming from to pull the hubs out, and/or, why are the half-shafts moving that far outward so that they can run out of the spline - that's a lot of movement. I'm assuming (I hope safely) that you're running a limited travel suspension: short hard spring, and short shaft shocks (or coil-overs to the same effect) - 3 or 4 inches of wheel travel, ie, even when you jack a corner, the wheels aren't going to be dropping down that far, ie, they're not going to be pulling on the spline enough to pull the hub... so my question is: are you sure that under full load, that the trailing arms aren't flexing or being allowed to move so far outward to pull the splines free? How much clearance do you have btwn the outside of the rear tires and the fender? How about stuffing some Styrofoam in there to fill most of that space (but not enough to touch the tire, or where you'd expect it to travel during normal conditions), and see if you don't come back from a few hot laps with it missing, or badly marked by the tire, ie, evidence that the wheel is moving outward enough to in fact pull the half-shaft and splines free? (Or, maybe you already have unexplained wheel rub: fender to sidewall??) good luck - greg

[PbFut]

But this again raises the question: where is all the force coming from to pull the hubs out, and/or, why are the half-shafts moving that far outward so that they can run out of the spline - that's a lot of movement.good luck - greg

Agree. That question has been bugging me as well. I think I might fashion up a go-pro and stick down there and see if anything looks odd. Biggest problem is finding a place to run the car. Taking to a track to test this simple thing is a major effort. Wife works at local cop shop but I have agreed to never take advantage of this so a blast in the area is out of the question. If I get popped I'm hosed on two fronts. I'll have to find something. Maybe a big parking area. Car has so much power, is so quick and loud, very hard to test outside of a race track. Weekend looks free. I'll find some place to mess with it.

[rf900rkw]

I used to use an uncontrolled county air strip.

[Blacklane]

Could you have the wrong half-shaft installed? The /M version has a wider rear and maybe the new half-shaft you installed was for the non-/M version and is too short.

[rf900rkw]

M shafts are shorter and larger diameter than the standard 6cyl.

[spidertri]

Are the CV joints binding? Could it be temperature/grease related? I know some track/race guys have rebuilt their axles with redline CV2 grease. It is supposed to be better for high performance applications.

Also found this thread, there was mention of not making the CV boots air tight. I guess if the joint heats up and it is air tight it could turn into a balloon and not let things move as they should.

http://forums.bimmerforums.com/forum...out-of-my-diff

[mcoupemindy]

Don't have time to read every comment, so sorry if this was already said ...

First step is to replace the retaining clips in the pressure assembly and use a set of newer output flanges. I've had a few customers that had a similar problem and that solved it. If you need output stubs or the clips, let me know - I have plenty around.

The actual engagement of the stub to the pressure assembly is roughly 1", so it's not much pull until it become disengaged.

[PbFut]

Thanks Jonathan
Email sent

[LAWLence]

Dan,
Someone linked to my thread from last year earlier in this thread. I had the same issue. Is it happening on the left side only by any chance?

[PbFut]

No, seems both sides will do it. It's good to know this is not isolated and from some other issue, I hope. In this last situation, it was the right side. I am thinking Jonathan is correct. The grove on the flange may be very slightly rounded by a few mils and allowing the clip to slide off a little more easily. Just enough to allow the flange to pull free from time to time. I noticed that I can pull off with my hands. It seems that years ago, I had to use a tool to apply enough pressure to pop off. That would also indicate the groove has changed. Jonathan offered to send me some used replacements to try. I've taken him up on the offer and will report back results.

[Randy Forbes]

Looks like you've already gotten all the intelligent responses, so let me throw this one out there...

Is the vent hole in the differential cover plugged? If the stub-axles aren't popping out due to full suspension droop/bounce, perhaps it's a build-up of internal pressure that forces the flange out until the seal is broken? Do you usually find both, or just one (1) side has dislodged?

- - - Updated - - -

Seems to be either side but the side with new halfshaft will do it more. Groves are spotless. Replaced clips. Seemed better but has returned. Agree on suspension. I am wondering if it's happening in the pits. Thinking about it, seems it may happen when I lift car on one side to swap tires. I noticed it happened after a tire swap in the past but until now did not put the two actions together. Wonder if it's pulling the clip lose then working out over time. Still, why is it pulling lose to begin with just by lifting the car. Alignment is nothing crazy. About 2.5- camber, zero to a tick toe out. It's all just a guess.

Makes a good argument for lifting the rear end via the differential housing...

[mcoupemindy]

Is the vent hole in the differential cover plugged? If the stub-axles aren't popping out due to full suspension droop/bounce, perhaps it's a build-up of internal pressure that forces the flange out until the seal is broken? Do you usually find both, or just one (1) side has dislodged?

If it's clogged, you're going to know it. Oil everywhere. I've only seen a handful that were truly clogged, and they are a disastrous mess. On one car, I pulled the drain plug before the fill, and had 2 quarts of hot gear oil in the mouth. It was a reminder to always open the fill first, even when the whole diff is coming out.

Parts are on the way to PbFut, so I'm sure he will let us know the outcome.

[Randy Forbes]

Good to know; oily mess = check the vent hole!

As a result of a co-worker some decades ago, I ALWAYS loosen a fill plug before removing the drain plug. He drained the differential fluid in his S-10, only to discover the fill plug was seized solid to the cast housing. Fortunately, it has a removable rear cover, and I was able to weld half a 3/4" pipe-coupling at his "marked elevation" so he could get oil back in the thing.

Whether you're working on a differential or a transmission, ALWAYS make sure you can get the filler plug out BEFORE you pull the drain plug.

[gmushial]

Looks like you've already gotten all the intelligent responses, so let me throw this one out there...

Is the vent hole in the differential cover plugged? If the stub-axles aren't popping out due to full suspension droop/bounce, perhaps it's a build-up of internal pressure that forces the flange out until the seal is broken? Do you usually find both, or just one (1) side has dislodged?

- - - Updated - - -

Makes a good argument for lifting the rear end via the differential housing...

I hugely like it and appreciate it when people think outside of the box - very refreshing to see such: the sign of an active mind :-) ... but with a thermal coefficient of expansion for oil being something like 0.04%/degF, I wonder if it's even possible to get enough of a temperature rise to generate enough pressure to push the flanges out.

[rf900rkw]

Ah, but most of the volume in the diff is air, and the expansion rate of air is considerably higher. Many an F150 has burned to the ground due to a plugged transmission vent. AIr pressure rise inside the unit due to the plugged vent causes the dipstick tube to become the vent. Alas, the tube end is below the dynamic fluid level, so expansion pushes fluid out the tube onto the exhaust and cats.

Poof.

[gmushial]

Interesting image... ;-) but I see the situations as being a bit different, or at least the amount of force required/involved is more than a bit different [and yes, air expansion is approx. 200ppm/F at temps we're interested in] - in the case of the F150 the pressure only has to move an ounce or two of fluid to start the weenie roast... in this case we're talking about generating enough force to overcome the detent rings, which is going to be in the many of pounds range (even if they're compromised as it sounds)... but on the otherhand, and having been around here for long enough: the two of you have been right on so many situations that seemed entirely unobvious from the get-go... I think I'll sit back and see if this isn't another learning experience ;-)

[mcoupemindy]

Interesting image... ;-) but I see the situations as being a bit different, or at least the amount of force required/involved is more than a bit different [and yes, air expansion is approx. 200ppm/F at temps we're interested in] - in the case of the F150 the pressure only has to move an ounce or two of fluid to start the weenie roast... in this case we're talking about generating enough force to overcome the detent rings, which is going to be in the many of pounds range (even if they're compromised as it sounds)... but on the otherhand, and having been around here for long enough: the two of you have been right on so many situations that seemed entirely unobvious from the get-go... I think I'll sit back and see if this isn't another learning experience ;-)

Sealing surface is ~1.75" in diameter or 2.40 in^2. With oiled, aged seals and a worn c-clip, I would estimate less than 10# force required to pop the flange. 4-5psi is definitely plausible if a vent is clogged.

Again, it would be evidenced by an oily mess also, so I don't think that's the OP's problem - just wanted to give you information that it is theoretically possible to pop them via pressure if a vent is clogged.

[gmushial]

Not going to doubt it... but wouldn't 5psi require a 1500F increase in temp? (assuming 200ppm/F, which actually gets smaller at higher temps)... will be interesting to see what Dan finally finds.