That is the kind of inspirational disquisition I like to hear
Disappointed (gearbox replacements)
- Thread starter Sandrae
- Start date
- Replies 130
- Views 34K
ericy
Well-Known Member
I wish the youtube video had English captions, but right at the end it was pretty obvious what had failed. Not at all clear *why* it failed.
I have always been curious as to what the insides of the reduction gear assembly looked like. Now I know. I didn't expect to see tapered roller bearings
In comments to this video the author mentions bad quality of the roller bearing as the cause. AFAIU the faulty bearing makes the gearbox vibrate and this leads to damage of the splined connection between the motor and the gearbox. In case of prolonged intense vibration the electric motor’s output shaft bearings can as well be damaged. This is why it’s becoming necessary to change the motor together with the gearbox
Genevamech
Active Member
That bearing was certainly the main problem in that case, but bad bearings don't necessarily explain the various attempts at repair by rotating the motor or lubricating the splines that I read about. It's possible those "remedies" were just the engineer's attempts to see what sticks I suppose.
A missing roller would absolutely cause a rhythmic thunk as the inner race dropped down into the gap with each revolution of the roller cage. Must've been hammered to sh*t!
As for why tapered roller bearings are used, they are very common on drive differentials as far as I'm aware. They're typically specified whenever axial load is expected, and used in pairs if the load could be in either direction. I don't know how much axial load can be expected on a front-wheel-drive differential (don't the driveshafts already have splines to change length as the suspension travels?). Rollers also have a larger contact area than balls do, so you'll see roller bearings where there are higher radial loads... but again there should be little radial load on that shaft; it's not supporting the weight of the car or anything!
Stranger still is they used ball bearings for the rest of the gearbox. The helix cut gears produce some side thrust but I guess they figure the ball bearings can handle that (angular contact type perhaps?).
A missing roller would absolutely cause a rhythmic thunk as the inner race dropped down into the gap with each revolution of the roller cage. Must've been hammered to sh*t!
As for why tapered roller bearings are used, they are very common on drive differentials as far as I'm aware. They're typically specified whenever axial load is expected, and used in pairs if the load could be in either direction. I don't know how much axial load can be expected on a front-wheel-drive differential (don't the driveshafts already have splines to change length as the suspension travels?). Rollers also have a larger contact area than balls do, so you'll see roller bearings where there are higher radial loads... but again there should be little radial load on that shaft; it's not supporting the weight of the car or anything!
Stranger still is they used ball bearings for the rest of the gearbox. The helix cut gears produce some side thrust but I guess they figure the ball bearings can handle that (angular contact type perhaps?).
ericy
Well-Known Member
It looks like there was all sorts of metallic debris that seemed to have gotten mixed in with the grease. So was the roller missing all along, or did it disintegrate and wear down until nothing was left? But what's special about that one roller that caused it to disintegrate? I saw signs of metallic debris elsewhere in the video, but I didn't spot much of anything else that had huge amounts of wear on it. But we didn't really get a good look at all of the mating surfaces.
Earlier in the video, they focused on the half-shafts, and seemed to be pointing something out - possible some wear? Later - when the gearbox was still in the car, he was pushing on that shaft with a wrench, and noting slop where there should be none.
Then there is the question of whether this was a one-off, or whether this is the same the failure that other cars have? We normally never get to see the things when the gearbox is replaced - AFAIK, they go back to Hyundai for whatever reason.
I guess the reason the tapered bearings surprised me is that I saw the teardown of the Bolt reduction gear, and GM didn't bother with anything other than normal ball bearings. As you say - there ought not be much axial load due to the spline, and these things ought not be supporting all that much weight (not the weight of the full car, that's for sure).
ericy
Well-Known Member
Here is a further thought. If you opt to use tapered roller bearings, they would need to be carefully shimmed when the reduction gear is being assembled. And then what happens when the temperature changes and you have differing rates of expansion and contraction for the dissimilar metals? The case *looks* like cast aluminum - the shaft must be some steel alloy. But as temperature decreases, aluminum contracts faster than most grades of steel. I mention this because I remember someone here mentioning that the ticking noise seemed temperature dependent.
Actually looks like the axial bearing for the differential gear is the cause of the issue, not the motor spline shaft bearing as indicated by this diagram:
https://insideevsforum.com/communit...g-vibration-type-sound.7262/page-8#post-96897
Could be a valid point on the dissimilar metals expansion/ contraction ratios.
Which would also indicate that the "turn motor shaft 180 degrees " as a first step to repair diagnosis, would in this case prove didily squat, except to prove that it is not the gear near the motor shaft that is the issue- still needs a new gearbox, just not necessarily a new motor as well, unless as mentioned the noise has gone on long enough to cause other issues such as lubrication in the area around the motor shaft bearing.
https://insideevsforum.com/communit...g-vibration-type-sound.7262/page-8#post-96897
Could be a valid point on the dissimilar metals expansion/ contraction ratios.
Which would also indicate that the "turn motor shaft 180 degrees " as a first step to repair diagnosis, would in this case prove didily squat, except to prove that it is not the gear near the motor shaft that is the issue- still needs a new gearbox, just not necessarily a new motor as well, unless as mentioned the noise has gone on long enough to cause other issues such as lubrication in the area around the motor shaft bearing.
Last edited:
Genevamech
Active Member
I suspect it was missing the entire time. There doesn't seem to be much damage to the other rollers and the cage looks to be in very good shape; not something I'd expect if there was jagged chunks of hardened steel tumbling around in there. I also can't imagine it wore away to absolutely nothing, though it's possible it wore so much that it fell out upon disassembly... still, though, I'd expect more collateral damage.
By the way there isn't supposed to be any grease in that gearbox; just gear oil (that goopy brown liquid). I have no idea what that grey paste is but I bet it's not supposed to be there! Looks like maybe a magnet caught all the debris?
Yeah it looked like he was pointing out how the driveshaft was worn, from rubbing from the whole thing being kattywompus. If you go back to about 4:08 where he's poking it with the wrench, you can see the bottom edge of the casting is egg-shaped! That slop is due to the failed bearing. You don't have to be a mechanic to know that ain't how it supposed to be! I bet it rattled on rough roads, too.
Here's hoping it's just a one-off, or at worst a small number of cases from a bad batch of bearings...
Also, you make an excellent point about the thermal expansion, though aluminum differential housings are not exactly rare - I bet you'll find them on most FWD vehicles. I don't know exactly how they account for it but I have to assume thermal expansion is a "solved problem(tm)."
That noise at 4:08 brings back unpleasant memories on my 2019 model - identical to the intermittent sound I was having, and also a direct correlation with temperature changes. Funny it occurred very intermittently while in forward at low speed, but most all the time when reversing out of my driveway in the morning, when the metals were cold along with the ambient temperature.
KiwiME
Well-Known Member
Yep, tapered rollers on the final drive seen out of place in this design, especially since helical spur gears are used rather than hypoid which require tight axial retention. Certainly the axial loading due to the helical-cut gears is at a maximum on this particular shaft where the torque is highest but larger ball bearings would be the correct choice as they don't require a precise preload, not easy to set prior to the cases being closed up. Plus the aluminium cases are not very rigid axially and a suitable preload of perhaps several 10s of kg might not be maintained over the entire life. Tapered rollers must run with enough axial preload to keep the rollers geometrically in place when under maximum radial load. Mentioning the Chevy Bolt, they also got the primary shaft bearing arrangement correct as well, all following good design practices.
I find it difficult to believe that the missing roller had completely disappeared yet didn't significantly score the races or damage either the roller cage or adjacent rollers. It seems to have left the cage in such good condition that I have to wonder if it dropped out during factory assembly as the play was evident before he disassembled the gearbox. The silver-coloured debris he pointed out at the inner raceway looked to me like aluminium filings, as did the debris in the sump. The rollers and raceways are hardened steel and won't go anywhere without a massive fight and leaving obvious damage.
Another puzzle is what appeared to be damage on the motor shaft nose and if the missing roller was the cause. It seems a bit far-fetched to me that there is a connection but it's possible if the driver ignored the resulting noise. On the other hand, from a statistics perspective, it's unlikely that two independent defects would appear at the same time.
I find it difficult to believe that the missing roller had completely disappeared yet didn't significantly score the races or damage either the roller cage or adjacent rollers. It seems to have left the cage in such good condition that I have to wonder if it dropped out during factory assembly as the play was evident before he disassembled the gearbox. The silver-coloured debris he pointed out at the inner raceway looked to me like aluminium filings, as did the debris in the sump. The rollers and raceways are hardened steel and won't go anywhere without a massive fight and leaving obvious damage.
Another puzzle is what appeared to be damage on the motor shaft nose and if the missing roller was the cause. It seems a bit far-fetched to me that there is a connection but it's possible if the driver ignored the resulting noise. On the other hand, from a statistics perspective, it's unlikely that two independent defects would appear at the same time.
ericy
Well-Known Member
I am kind of thinking out loud here. Part of me has a hard time believing that they could assemble the thing with a missing roller and not notice. Did they fall out in the factory during assembly? Were the things missing even before they reached the factory?
But I agree that it is also true that if there were a roller there which failed in some manner (seized, for example), the collateral damage would be considerable - more than we see in the video. Unless just one of the rollers hadn't been properly heat treated, but that would still cause a lot of damage.
As I look at the video again, that "grease" as I put it starts to look more like iron/steel fragments in the vicinity of a magnet - possibly with some gear oil still mixed in. And there is some of this stuff in both halves of the case, but that would imply two magnets, which doesn't make sense.
But I agree that it is also true that if there were a roller there which failed in some manner (seized, for example), the collateral damage would be considerable - more than we see in the video. Unless just one of the rollers hadn't been properly heat treated, but that would still cause a lot of damage.
As I look at the video again, that "grease" as I put it starts to look more like iron/steel fragments in the vicinity of a magnet - possibly with some gear oil still mixed in. And there is some of this stuff in both halves of the case, but that would imply two magnets, which doesn't make sense.
KiwiME
Well-Known Member
It does seem that the debris is clearly centered around one tapered roller bearing so I'll agree that it seems to be a single roller failure even though I've never seen or heard of one that didn't damage all the other rollers and raceways. I'm a little surprised that we didn't know about this video earlier.
I didn't notice a sump magnet nor is there evidence one is included in the drain plug, but perhaps I shouldn't be surprised. The resting location of the debris looks due to that being a quiet place in the oil turbulence due to the small dam.
10,000 (kms assumed) according to the reply to the second comment.
I'm not seeing any damage elsewhere in the gearbox but all bearings would have to have been replaced to do the job right. Gears are fairly robust.
Note that a single missing roller on the differential shaft would theoretically produce a clunk at half wheel speed and most if not all reports have the noise placed at motor speed. However it could clunk at several points in the rotation due to the instantaneous compounding of radial forces on the bearing.
One owner in Iceland did have a catastrophic gearbox failure about the same time as this video was made. The FB posts are the last 2 images.
Perhaps this defect could be detectable by checking axle shaft radial play or by an oil analysis.
I didn't notice a sump magnet nor is there evidence one is included in the drain plug, but perhaps I shouldn't be surprised. The resting location of the debris looks due to that being a quiet place in the oil turbulence due to the small dam.
10,000 (kms assumed) according to the reply to the second comment.
I'm not seeing any damage elsewhere in the gearbox but all bearings would have to have been replaced to do the job right. Gears are fairly robust.
Note that a single missing roller on the differential shaft would theoretically produce a clunk at half wheel speed and most if not all reports have the noise placed at motor speed. However it could clunk at several points in the rotation due to the instantaneous compounding of radial forces on the bearing.
One owner in Iceland did have a catastrophic gearbox failure about the same time as this video was made. The FB posts are the last 2 images.
Perhaps this defect could be detectable by checking axle shaft radial play or by an oil analysis.
Attachments
Last edited:
Catch the problem early enough and avoid further damage to the motor shaft as well. Sounds like a more appropriate approach and an excellent point made
ericy
Well-Known Member
Here is a hypothetical. What if you had a progressive failure of the roller that started with a flat spot forming on the roller? The ticking noise from a flat spot wouldn't be at shaft speed - it would be at some multiple of the shaft speed.
One question I have is why not everyone is having this issue. I have a 2019 model, 35000 km and I live in really cold weather (Montreal). So following the information in the thread, I should have the issue. Just trying to understand if it's a matter of time for everyone to get to that point.
ericy
Well-Known Member
We really don't know, and Hyundai isn't saying much. We don't often get a look inside of the thing after it has failed, and it isn't even clear whether this failure is the same as any of the others.
R P
Well-Known Member
I don't get how this above gearbox damage is related to the tick, tick, ticking issue. Mine came on at about 18K kms. They first replaced the gearbox, but the ticking resumed within 400 kms. So a new gearbox did not fix it. Then they replaced the motor (701) and the ticking is gone (now over 6K kms ago). If the ticking is due to gearbox damage, should it not have been replaced again as well?
I still think it has something to do with clamping of the gearbox and motor. When they first replaced my gearbox, the tech showed me some metal shavings that appeared like it could be due to some fretting of the mating surfaces. The new motor came with new bolts that would allow a higher torquing of the mating surfaces. And maybe the reason some cars don't have this problem at all is that the initial factory assembly had a truer mating and stronger torque.
Of course, just speculation on my part, and doubt that Hyundai will ever tell us exactly what the problem was.
I still think it has something to do with clamping of the gearbox and motor. When they first replaced my gearbox, the tech showed me some metal shavings that appeared like it could be due to some fretting of the mating surfaces. The new motor came with new bolts that would allow a higher torquing of the mating surfaces. And maybe the reason some cars don't have this problem at all is that the initial factory assembly had a truer mating and stronger torque.
Of course, just speculation on my part, and doubt that Hyundai will ever tell us exactly what the problem was.
Genevamech
Active Member
My impression is that Hyundai/Kia is not capable of finding the gearbox design or having problems to produce it with the required torque/speed ratio. Enclosing an extract from Hyundai datasheet with manufactured reduction gear. Notice that model G1F24 used in Kona and Niro (and probably 64KW Soul) has lower torque than G1F30 used in 27KW Soul, while the speed of 12'000 rpm is higher than 10'500 rpm. This doesn't correspond with increased torque of the traction motor (285 Nm for 27KW Soul compared with 395 Nm for Kona and Niro). This could explain why the gearbox is getting damaged in some cases, maybe it's related to the aggressive driving style.
https://www.hyundai-transys.com/resources/upload/ko/product/Powertrain.pdf page 24
https://www.hyundai-transys.com/resources/upload/ko/product/Powertrain.pdf page 24
