PSA: Loose Rear Axleshaft to Diff Housing Bolts

[froop]

Story

I haven't heard of this issue before but I'd just like to share my experience with everyone so that hopefully others can learn from my issue and can do some preventative maintenance if required.

I recently had an entire M3 rear end installed into my 135i which included the axleshafts and diff housing. Everything had been fine up until maybe around 1 month ago. I started noticing an intermittent metallic clicking/tapping noise which was clearly rotational in origin. It wasn't too loud and was heard mostly when coasting and when driving alongside objects which reflected the sound. I put the car up on a lift and tried to turn the wheel however everything seemed fine. I noticed that there was a nail in the wheel so I attributed it to that. As I didn't have any time or tools at the time I just left the nail in to pull out and repair another time when I was doing some other work on the car.

Fast forward a couple weeks and I started to notice that there was a bit of clunking when going from neutral to 1st gear. I wasn't sure whether this was normal and that I just hadn't noticed it before but it was also intermittent and didn't seem very severe so I just put it off for a bit however I started to notice that the metallic clicking/tapping noise was now a constant thing and was now also extremely loud. Loud to the point where I could hear it from inside the car even with the windows closed and everyone on the street as well as probably people in their cars with their windows open could probably hear it very clearly. I got a lot of funny looks.

As I had to install a stud kit I removed all 4 wheels, installed the studs and pulled out 3 foreign metallic objects in 3 different tyres. I figured that this would solve the issue however I was clearly wrong as as soon as I went to drive it, the exact same noise was still there.

This afternoon I took it down to the mechanics to have a look at it, got it up on the hoist again and this time when spinning the rear wheels the noise was clearly reproduced. Had a look underneath the car and noticed that most, if not all of the rear axleshaft to diff housing bolts had loosened and/or backed out. The noise was coming from one of the washers that was hanging loose as one of the bolts had actually fallen out completely and as the axleshaft rotated, it was dinging on the exhaust pipe.

We fixed it by re-installing and retightening all of the bolts with threadlocker to prevent them from backing out again. Another fix would be to just replace all of the bolts with new bolts. According to the mechanic, this was a common issue on our cars which had previously had suspension or diff work done which required the diff housing to be detatched from the axleshafts. Apparently the bolts stretch and therefore will not hold torque properly

I am lucky that I caught it when it did as who knows what would have happened had it gotten worse and completely failed whilst I was driving. The mechanic did mention that another car which had the same problem actually got to the point where the loose bolt or washer was hitting on brake lines but had luckily not destroyed them completely or he could have gotten complete brake failure.

Symptoms
Clunking going in/out of gear
Odd rotational noises from the rear of the vehicle

Issue
Loose and/or missing rear axleshaft to differential housing bolts.

Causes
Re-used bolts after removing and re-installing the rear differential and/or axleshafts

Remedy
Threadlocker
New Bolts

[Nugget]

I had the same issue, I only picked it up as I was changing my exhaust a few days after my diff was swapped. Most of the bolts were only finger tight.
Are you going to put in new bolts?

[froop]

Originally Posted by Nugget

I had the same issue, I only picked it up as I was changing my exhaust a few days after my diff was swapped. Most of the bolts were only finger tight.
Are you going to put in new bolts?

With thread locker it should be fine however I'm definitely going to be checking them again after this weekend's motorkhana event and maybe after another month or two to be certain they're not backing out.

But probably next time I order something from ECS or where ever I'll be ordering a new set of bolts just in case.

[Nugget]

I have checked mine a few times since and they have stayed tight, but might do the same and install new ones.
Do you know what the part number is?

[Tab135]

I'm pretty sure the aluminum bolts need to be replaced each time

[ozimmer]

Do you know what material the bolts are made from?
I can understand if they are an aluminium alloy that they may stretch, but am surprised if steel bolts in this application would be stressed beyond their elastic limit. BTW, I recall reading that many of the Al-alloy engine bolts are single use only.

[dcaron9999]

Wow!
Froop, thanks for sharing this ...

Im putting in a Quaife diff shortly, and I will definitely make sure fresh axle bolts + thread lock are used. Having said that, you would think an experienced mechanic doing this job on a regular basis would know about this preventive measure though, no?

Im actually going to put my mechanic to the test, and ask him if this is something that is part of his diff swap checklist ...

These are the part numbers on REALOEM:
Description: Torx screw with ribs.
Bolt spec: M10X43 ZNS3.
Part # is 33207572716.
12 bolts are needed.

I think the ZNS3 refers to a coating process, for anti-corrosive properties.

[128Convertibleguy]

I'm guessing they're "torque to yield" bolts. Which could be made out of anything. They have definite advantages, but are one time use only. A fact which is simply ignored by many mechanics.

"Torque to yield fasteners are mounting hardware which is torqued beyond the state of elasticity and therefore undergoes plastic transformation, causing it to become permanently elongated. Torquing a fastener to yield results in a high preloading of the fastener which, depending on the load frequency and amplitude, can significantly increase the fatigue life of the fastener. When the applied load doesn't surpass the clamping force of the fastener the strain of the fastener will be lower than when the preloading would be smaller than the applied load. It is therefore beneficial in high-frequency high-load situations with a higher risk of fatigue related failure, like a bolted down cylinder head, to use torque to yield bolts.

Compared to normally tightened hardware, a smaller sized TTY bolt/screw may be used while still maintaining the same clamping force. A drawback with TTY hardware is that it normally has to be replaced when loosened, for example when the cylinder head is removed."

Who's to blame here? A manufacturer using state of the art techniques, or a mechanic who says "that's the way I've always done it."?

Threadlocker is a bandaid, not a fix. The used bolt will never develop the same clamping force as a new one. The fix is to RTFM, and new bolts, torqued carefully to spec. Good luck finding a mechanic who always works like that. "Calibrated arms" are the rule. Mechanics actually take pride in those, even though they shouldn't. A guy using his calibrated arm on a NASCAR team, or in an aerospace factory, will be gone fast.

[MZWIE]

I'm quite certain that those bolts are not re-usable.

[fe1rx]

Originally Posted by e34m5

I'm quite certain that those bolts are not re-usable.

Agreed. Bentley's 3-series Service Manual indicates that that the torque specs for the drive shaft to differential bolts:

a) they must be replaced

b) they must be torqued to a specified value and then tightened a further 45° or 90° depending on application.

Both of the above are consistent with a torque-to-yield fastener, which is to say that correct installation permanently stretches the bolt.

[MZWIE]

Some times those types of bolts are called "stretch bolts" in fact. They are typical in use to bolt the head to a block.

[CGMike]

You should always check suspension/driveline bolts after a few days of doing work on them to make sure everything is still tight.

[128Convertibleguy]

Originally Posted by CGMike

You should always check suspension/driveline bolts after a few days of doing work on them to make sure everything is still tight.

How would you check a torque to yield bolt in any meaningful way? As stated above, typically you torque to a specified value plus some specified rotation more. Messing with that will likely make things worse, not better. All you could find would be a seriously undertightened bolt.

For highly stressed fasteners, particularly torque to yield, the job simply has to be done correctly the first time. New fastener specced correctly (probably means OEM), female threads clean but not "chased" with a cutter. Everything good enough so that you can spin the fastener in by hand, and that torque builds smoothly. Lubricated or antiseized or threadlocked (or not) exactly as specified in the manual, no ad hoc "bright" ideas.

Do that right, and you can do no more.

[Solo_M_Tech]

Originally Posted by 128Convertibleguy

How would you check a torque to yield bolt in any meaningful way? As stated above, typically you torque to a specified value plus some specified rotation more. Messing with that will likely make things worse, not better. All you could find would be a seriously undertightened bolt.

For highly stressed fasteners, particularly torque to yield, the job simply has to be done correctly the first time. New fastener specced correctly (probably means OEM), female threads clean but not "chased" with a cutter. Everything good enough so that you can spin the fastener in by hand, and that torque builds smoothly. Lubricated or antiseized or threadlocked (or not) exactly as specified in the manual, no ad hoc "bright" ideas.

Do that right, and you can do no more.

The right way to test a torque to yield bolt is to measure its length with an outside micrometer.

[CGMike]

I would say with torque to yield (if thats what these bolts are) would be checked by verifying the torque again. I know on other cars the torque to yield bolts are usually "100 ft/lbs plus 60 degrees of rotation." So on something critical, after a few days, I would go back and check to ensure it was still torqued at 100 ft/lbs. I might even set my torque wrench to 95 ft/lbs to make sure I dont move the bolt as I check the torque, but thats being anal. If the bolt moves, it is no longer tight. If it doesnt move, its torqued to at least what the manufacturer wants. There is the small posibility that it isnt exactly torqued to what its suppossed to because of the "yield" part but at least I know its torqued to and not completely loose.

[RimasRS]

I saw that problem on M3 forums

[ShocknAwe]

Found a few of these loose on my rear end. Thanks for the part number above, does anyone know the precise torque spec on these and torx socket size?

[ShocknAwe]

Know what the socket size and torque spec is on these bolts?

Originally Posted by dcaron9999

Wow!
Froop, thanks for sharing this ...

Im putting in a Quaife diff shortly, and I will definitely make sure fresh axle bolts + thread lock are used. Having said that, you would think an experienced mechanic doing this job on a regular basis would know about this preventive measure though, no?

Im actually going to put my mechanic to the test, and ask him if this is something that is part of his diff swap checklist ...

These are the part numbers on REALOEM:
Description: Torx screw with ribs.
Bolt spec: M10X43 ZNS3.
Part # is 33207572716.
12 bolts are needed.

I think the ZNS3 refers to a coating process, for anti-corrosive properties.

[ShocknAwe]

Update: according to internet searches it's between 48-55 ft lbs. I give it a go at 50 and call it a day.