What kills a final drive/differential?

[750IX]

Hi, how to treat a final drive to make it last "forever"? I got a 2013 Bmw 750IX so it has 2 differentials/final drives. Bmw claims it has "lifetime fluid" but i replaced the oil at 85 k miles and it was dark but ok. Will continue to replace oil together with transmission oil every 40k miles ish. I do no burning and stuff but other than that is there tings to do or not to do? My old E39 530D had 250k miles on it with the original diff.

What does really kill a final drive like the bearings and the gears? Heat? Old oil? Burning? Or just bad luck?

I use Ravenol DGL SAE 75W-85 GL-5 LS in both diffs.

 

[ABN_CBT_ENGR]

Well, everything has a design life but all things equal ( and considering the fluid is a wear item and excluding seals/gaskets)

I would expect bearings to reach "end of life" before gears ( assuming gears were properly made, set and lubricated throughout the life)

 

[MrHorspwer]

The biggest sure-fire killer for a final drive is heat. That is largely caused by an overloaded or overworked axle and most often associated with failures in trucks.

When seeing dark fluid, you have to consider if the axle has a clutch-type LSD. As the clutches wear, the spent clutch material is suspended in the fluid, which will certainly change the appearance. That doesn't mean the fluid has failed or will result in any damage.

Change the axle fluid in a vehicle with a 1st gen Haldex AWD system and it will look disgustingly black. Those systems use the difference between input and output shaft speeds (slippage) to turn a gerotor pump that pressurizes a clutch pack and begins to equalize speed ("locks" the axle). Thing is, if you were to change the fluid again in 5 miles, it would look nearly as bad as the first time. That's because the axle fluid that lubricates everything is also the fluid that is used to pressurize and engage the clutch pack. There is just a lot of spent clutch material in there. Many are surprised when they see how dark the old fluid coming out is versus the new fluid going in and think they just saved the life of their axle.

 

[Zaedock]

I would say it depends on the application.

In the case of a daily driver, I would say heat (from excessive loading) and water and foreign material contamination via seal degradation.
Otherwise, in a normally operated manner, final drives typically last a very long time, even without fluid changes.

When it comes to hobbies, I would say abuse. I have blown up 10 and 12 bolt axles at the drag strip. I had my trail Jeep off road and destroyed a rear Detroit Locker ( I swapped in a spare open carrier on the trail to keep going). Physical axle size and moving parts can be a weak link, so now I run a rear spool.

When I owned a shop, besides gear ratio changes & abuse repair for enthusiasts and typical seal maintenance, the only "killed" final drive I encountered was an older woman with a Highlander ('00+/-) where the rear axle seized on the highway. She was very lucky not to crash! Dissecting the axle revealed the axle was nearly bone-dry and the gears appeared to be friction welded together. It was impressive carnage. It appeared that a half shaft seal had been leaking for a long time.
She had little money, so we were able to locate a good used replacement and she also needed rear tires (flat spots) and we located used replacements with depth to match the front. The big man upstairs was definitely helping her that week.

 

[ABN_CBT_ENGR]

OK, so to convert to "engineer speak" (assign specific definitions to failure modes and mechanisms) for analysis...

Assuming a properly designed, built, installed and lubricated gear set ( That's the baseline for analysis to address the OP question and the zero point)

Would you agree that "loading" ( defined as loading in excess of design) is probably the most encountered failure mode? ( "heat" is not a failure mode technically since it is a result of something else)

My experience with industrial gearing shows this to be the most common but what about the car guys?

But this doesn't address the question of what component fails first

 

[Snagglefoot]

I witnessed a wheel and axle come out of a 1/2 ton pickup on a divided highway. It went flying across the median and across the two lanes of my side of the highway. The bearings went rolling into the ditch and started a grass fire, Evidently there was no fluid in the diff and the bearings welded to axle. The pickup was able to stop along the shoulder with plenty of sparks.

On a 1/2 ton Suburban I had, the bearings for the gear assembly started the growl when turning, especially noticeable on a cloverleaf. I had to get it rebuilt. I think this would be the most common failure.

 

[Gebo]

Heat and water.

 

[spasm3]

I witnessed a wheel and axle come out of a 1/2 ton pickup on a divided highway. It went flying across the median and across the two lanes of my side of the highway. The bearings went rolling into the ditch and started a grass fire, Evidently there was no fluid in the diff and the bearings welded to axle. The pickup was able to stop along the shoulder with plenty of sparks.

On a 1/2 ton Suburban I had, the bearings for the gear assembly started the growl when turning, especially noticeable on a cloverleaf. I had to get it rebuilt. I think this would be the most common failure.

Probably a c-clip rearend.

I'm going to say loss of oil. Leaking and running out of gear oil probably kills them.

 

[Zaedock]

OK, so to convert to "engineer speak" (assign specific definitions to failure modes and mechanisms) for analysis...

Assuming a properly designed, built, installed and lubricated gear set ( That's the baseline for analysis to address the OP question and the zero point)

Would you agree that "loading" ( defined as loading in excess of design) is probably the most encountered failure mode? ( "heat" is not a failure mode technically since it is a result of something else)

My experience with industrial gearing shows this to be the most common but what about the car guys?

But this doesn't address the question of what component fails first

I agree with your industrial experience regarding loading. As a car guy, I'm referring to exceeding the GVWR of the axle, which likely happens more often than people realize. A good example - my wife recently needed 1-2 cubic yards of topsoil. The owner of the center we go to was out and his son operated the largest front loader he had and gave us nearly 5 cubic yards and said don't worry about it! My wife was happy, but that's approx. 8+ thousand lbs (or more!) on my 7K rated flat bed all pulled by a half-ton Ford. I certainly exceeded the GVWR for the short ride home (it actually pulled fine though!). This season of home "upgrades" and honey-do lists will certainly affect my fluid (and trailer bearing) service intervals.

In my laymen's experience (as compared to an appropriate Engineer on the subject), I have found that the overloaded axle heat affects the lubricant and the overloaded bearings typically start to fail first. When the bearings exceed their pre-load tolerance, they allow the ring and pinion to prematurely wear.

 

[Zaedock]

Probably a c-clip rearend.

I'm going to say loss of oil. Leaking and running out of gear oil probably kills them.

I agree. The c-clip axle shaft is the inner bearing race.

 

[ABN_CBT_ENGR]

I agree with your industrial experience regarding loading. As a car guy, I'm referring to exceeding the GVWR of the axle, which likely happens more often than people realize.

In my laymen's experience (as compared to an appropriate Engineer on the subject), I have found that the overloaded axle heat affects the lubricant and the overloaded bearings typically start to fail first. When the bearings exceed their pre-load tolerance, they allow the ring and pinion to prematurely wear.

Sounds like a complete alignment of observations and experiences to me regarding the "prime mover' of the failure train.

One of the tricks of a proper failure analysis when identifying the true "root cause" is separating and ranking second, third etc. effects from the literal "match that lit the fuse" that results in people chasing symptoms rather than cures.

 

[Trav]

Going from my own experience, normal, stock street vehicles that do not see heavy loading, modified engines, plowing, etc fluid loss due to leaking and lack of maintenance to at least keep them full is the cause of most failures.
Most of the blown diffs I have seen over the years were low to very low on fluid.

Other causes are pretty much vehicle dependent, overloading on trucks, modified engines with too small a diff eg V8 Vega with stock diff was a guaranteed fail, blown out spider gears in plow trucks are all common. Actual fluid overheating is rare on street cars, rally cars and endurance cars typically use a circulating rear mounted diff cooler.
Basically the type of vehicle and how it is used can predict the type of drive line failures it may have.

 

[ABN_CBT_ENGR]

Going from my own experience, normal, stock street vehicles that do not see heavy loading, modified engines, plowing, etc fluid loss due to leaking and lack of maintenance to at least keep them full is the cause of most failures.
Most of the blown diffs I have seen over the years were low to very low on fluid.

Other causes are pretty much vehicle dependent, overloading on trucks, modified engines with too small a diff eg V8 Vega with stock diff was a guaranteed fail, blown out spider gears in plow trucks are all common. Actual fluid overheating is rare on street cars, rally cars and endurance cars typically use a circulating rear mounted diff cooler.
Basically the type of vehicle and how it is used can predict the type of drive line failures it may have.

OK, here is another failure case where loading is not a factor observed over groups of various vehicles.

Trav, you open the patient and do the surgery.

I take it lubrication is the “prime mover” of the failure (as the roll up) per your post

If you agree, then what “property” (s) (the drill down) of the lubricant enabled the failure.

Based on your professional observation, is it more likely than not….

The oil volume (simply didn’t have enough to stop the metal to metal contact and wear was the result) which is arguably not a failure/inadequacy of the lubricant proper.

A property(s) of the oil? ( if so, what properties and/or failures on the gears and bearings that led you to this conclusion?)

 

[40w8]

I have a 2014 BMW X1 s28i with only the rear differential. I read most articles about BMW's and read that the front drive train on Xmodels has a built in defect. The gears are designed straight instead of beveled. Not sure if it's on EVERY BMW.

As for rear diffs they are usually very strong. Many cars are used up and never have rear diff trouble. Mostly it would occur if the seals leaked, and it was run low. I drained my X1 rear diff at 3500 miles and the oil had no metal shavings like found on some cars. My RX8 Mazda had some metal mud at that mileage.

 

[Trav]

OK, here is another failure case where loading is not a factor observed over groups of various vehicles.

Trav, you open the patient and do the surgery.

I take it lubrication is the “prime mover” of the failure (as the roll up) per your post

If you agree, then what “property” (s) (the drill down) of the lubricant enabled the failure.

Based on your professional observation, is it more likely than not….

The oil volume (simply didn’t have enough to stop the metal to metal contact and wear was the result) which is arguably not a failure/inadequacy of the lubricant proper.

A property(s) of the oil? ( if so, what properties and/or failures on the gears and bearings that led you to this conclusion?)

I don't think it has much to do with the oil quality as it does with the volume of oil in the diff. The pinion is dependent on the ring gear not only bathing in fluid but also bringing the lube up the back of the cover/housing to it as well as the tubes that contain the bearings, as the level goes down it starves which leads to excess wear on the pinion and if it gets low enough the bearings.

Gale banks has a good video on this.

 

[Kestas]

In my microcosm I find bearings take a beating and fail from excessive debris denting. This is exacerbated by the fact that most manufacturers have gone to fill-for-life. Any break-in debris developed during use stays in the fluid, continually damaging the bearing until it fails out of warranty.

 

[ABN_CBT_ENGR]

In my microcosm I find bearings take a beating and fail from excessive debris denting. This is exacerbated by the fact that most manufacturers have gone to fill-for-life. Any break-in debris developed during use stays in the fluid, continually damaging the bearing until it fails out of warranty.

Yeah, ( not on cars per se) I have to second this even with scheduled changes and sampling PM's.

Here's a question for those who get into these and fix them to comment on relating to the OP. No right or wrong answers- just judgment and what the judgment is based on.

With loading removed ( looking for a second level primary failure mode since loading seems to be first) and lubrication nulled

In your OPINION ( based on your experience) did.... and what are you basing your opinion on

Is wear ( loosening of tolerance, damage etc.) of the bearing triggering gear wear?

Wear on the gear damaging the bearing?

Special case ( for off roaders)- can deflection of the axle tubes at the pumpkin change mesh angles and damage gears so they grind from then on?

 

[Snagglefoot]

Here are some additional ramblings. I understand the number of load cycles are important in the design of an automobile rear end. One complete revolution of the wheel imparts one complete load cycle on the ring gear and wheel bearings.

I once spoke with a professor of mechanical engineering who was of British background and was an old boy and a MGB fan. He told me during a discussion of MGB’s that for fun he had designed a differential for an MBG. He said that when it was finished, it had the dimensions of a differential of a Mac truck. He had to check his numbers because the size was ridiculously large. He then figured it out. He had designed it to never fail.

When he used more realistic numbers such as the number of load cycles in 100,000 miles of driving, the resulting size was that of an existing MGB rear end. So how many cycles were used for the BMW design? Meanwhile you can do things already mentioned like changing the lubricant on a regular basis and not driving the #$&@ out of it.

 

[Snagglefoot]

Here are some shots of a 14 bolt bolt rear end in a 3/4 ton Chev. Included is a shot of the magnetic located at the bottom of the pumpkin. I pushed my finger down in the center of the magnet and forced up a bunch of metal shards for lack of a better word. These shards shedded off the gears during 50,000 miles of driving with maybe 2,000 miles of towing a 7,000 lb trailer during hot conditions of summer.
Also shown is a photo of the cleaned up magnet. The rear end is still fine after 200,000 miles and 125 million loading cycles. Obviously it was designed to last. I used 75W90 synthetic from AC Delco and Motor Master (Canadian Tire). YRMV.

 

[ABN_CBT_ENGR]

magnetic located at the bottom of the pumpkin. I pushed my finger down in the center of the magnet and forced up a bunch of metal shards for lack of a better word. These shards shedded off the gears

Asking for a reason, are you sure the shards are hardened gear steel or could they possibly shavings from a casting or elsewhere?