Huge Shimano crank recall

[warthog1]

I liked the poking sh it at aluminium though, thrown back the same stuff that is said of carbon.

Except there’s nothing wrong with the aluminum, it’s the “glue” that’s failed.

Or you could say it's the manufacturing that failed.
It coincides with my experience of aluminium as a frame material. Cracks and fails.

[familyguy]

I liked the poking sh it at aluminium though, thrown back the same stuff that is said of carbon.

Except there’s nothing wrong with the aluminum, it’s the “glue” that’s failed.

Or you could say it's the manufacturing that failed.
It coincides with my experience of aluminium as a frame material. Cracks and fails.

Sometimes, yes. The point, though, is that the symptom here is cracking aluminium with the root cause appearing to be the failing adhesive bond. This overloads the aluminium piece which THEN fails as the system is broken apart.

I don't have anything that is impacted, luckily. But as above and stated again, if it's not a case of if but when, I think they might get demands for a replacement regardless of inspection.

I haven't checked but almost definitely my DA 9100 cranks are in the recall. I don't really like the inspection and if everything is ok, they don't replace them. Seems like any crank in the time period could crack and just because they haven't at the time of inspection doesn't mean that they are safe. I would much prefer them to replace all affected cranks.

[warthog1]

Except there’s nothing wrong with the aluminum, it’s the “glue” that’s failed.

Or you could say it's the manufacturing that failed.
It coincides with my experience of aluminium as a frame material. Cracks and fails.

Sometimes, yes. The point, though, is that the symptom here is cracking aluminium with the root cause appearing to be the failing adhesive bond. This overloads the aluminium piece which THEN fails as the system is broken apart.

If they didn't use aluminium would they be using that manufacturing process however?
I guess, yes it is the manufacturing process that is the root cause though.
Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

I like my shimano groupos but ultegra and dura ace cranks are a scratching for me and I expect many others.

[Thoglette]

It coincides with my experience of aluminium as a frame material. Cracks and fails.

Al (like most ductile materials) has a finite fatigue life. The way Al road bike frames are designed means that the frame will likely fail “early”, particularly if ridden hard by a big, strong rider.

Traditional Al components tend to be designed to not fail in their expected lifetime, after some bad experiences mid last century.

E.g. TooLongLegs is the only forum member I’ve seen to fatigue fail a traditional forged Al crank arm. That same crank arm in my service would a) take years to get the # of load cycles TLL does in a season and b) would likely take at least one order of magnitude of cycles to fail due to my much lower strength. So, I’ll take my entire life to fatigue my crank arms to failure whereas TLL can kill then in a season or two. His handle bars probably crack after a while, too!

Almost ever material material has a fatigue limit, it’s a question of what the designer does that determines when.

(Steel is weird, under a certain %ge load it doesn’t fatigue. Which is why we tend to make things like springs out of steel )

[warthog1]

It coincides with my experience of aluminium as a frame material. Cracks and fails.

Al (like most ductile materials) has a finite fatigue life. The way Al road bike frames are designed means that the frame will likely fail “early”, particularly if ridden hard by a big, strong rider.

Traditional Al components tend to be designed to not fail in their expected lifetime, after some bad experiences mid last century.

E.g. TooLongLegs is the only forum member I’ve seen to fatigue fail a traditional forged Al crank arm. That same crank arm in my service would a) take years to get the # of load cycles TLL does in a season and b) would likely take at least one order of magnitude of cycles to fail due to my much lower strength. So, I’ll take my entire life to fatigue my crank arms to failure whereas TLL can kill then in a season or two. His handle bars probably crack after a while, too!

Almost ever material material has a fatigue limit, it’s a question of what the designer does that determines when.

(Steel is weird, under a certain %ge load it doesn’t fatigue. Which is why we tend to make things like springs out of steel )

I remember TLL doing that too.
I still follow him on Strava over in France. He rides a lot less and more gravel, offroad. Some good photos at times.

[Duck!]

Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

Only the recalled models used the glued joint. 105 and 9 & 10-sp Ultegra and Dura-Ace are still hollow construction, but a different manufacturing process. Tiagra, Sora, Claris and all non-series models are solid arms. I haven't seen enough 12-sp. to see how they're made.

[warthog1]

Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

Only the recalled models used the glued joint. 105 and 9 & 10-sp Ultegra and Dura-Ace are still hollow construction, but a different manufacturing process. Tiagra, Sora, Claris and all non-series models are solid arms. I haven't seen enough 12-sp. to see how they're made.

I am going off info on the ww forum. There is a thread there on it and several posters have stated the newer versions are sill glued, it may be wrong however.

Edit had a quick look here; https://velo.outsideonline.com/road/roa ... ad-cranks/

The most obvious change on the driveside crankarm is the addition of a physical plug inside the chromoly steel spindle to keep water out. Although the shape of the crankarm is different from the Dura-Ace R9100 and Ultegra R8000 generations, the basic design appears to be otherwise unchanged.

The non-driveside crankarm sees a much bigger revision, at least for Dura-Ace. Whereas the two previous generations used a two-piece bonded construction like on the driveside, Dura-Ace 9200 reverts to a one-piece forged design (like Ultegra). Without a bond seam, there’s now no bond to fail — problem seemingly solved.

Changed non-driveside to one piece.
A plug to keep water out of the 2 piece driveside part.
Don't know how accurate that is.

[Nobody]

Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

Manufacturers generally don't take big risk with mid to lower end (volume) products as they say there will be too many coming back.

[warthog1]

Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

Manufacturers generally don't take big risk with mid to lower end (volume) products as they say there will be too many coming back.

Fair enough.
This thread made me look at the weights of 105 vs DA and Ultegra chainsets. It seems almost pointless.

https://www.cyclistshub.com/shimano-105 ... -dura-ace/

[Nobody]

Yeah I doubt anyone could really tell the difference in a blind test. So many other factors. People that are worried about 69g should really start analyzing what they eat. That should make far more difference over the long term.

[Duck!]

Seems counter-intuitive they are still using it, when the "lower spec" products that carry a small weight penalty but aren't glued, haven't failed.

Only the recalled models used the glued joint. 105 and 9 & 10-sp Ultegra and Dura-Ace are still hollow construction, but a different manufacturing process. Tiagra, Sora, Claris and all non-series models are solid arms. I haven't seen enough 12-sp. to see how they're made.

I am going off info on the ww forum. There is a thread there on it and several posters have stated the newer versions are sill glued, it may be wrong however.

Edit had a quick look here; https://velo.outsideonline.com/road/roa ... ad-cranks/

The most obvious change on the driveside crankarm is the addition of a physical plug inside the chromoly steel spindle to keep water out. Although the shape of the crankarm is different from the Dura-Ace R9100 and Ultegra R8000 generations, the basic design appears to be otherwise unchanged.

The non-driveside crankarm sees a much bigger revision, at least for Dura-Ace. Whereas the two previous generations used a two-piece bonded construction like on the driveside, Dura-Ace 9200 reverts to a one-piece forged design (like Ultegra). Without a bond seam, there’s now no bond to fail — problem seemingly solved.

Changed non-driveside to one piece.
A plug to keep water out of the 2 piece driveside part.
Don't know how accurate that is.

If true, seems like a dumb move, because it's not the left arms failing. From the pics and one in-the-flesh broken crank I've seen, the fractures propagate from the smaller bonding surfaces on the spider arms then spread down the crank. An impending crank failure should be felt by the rider, because as the bond separates the crank will bend well before it breaks entirely, and that will feel obviously weird when pedalling.

If the tokenistic but realistically useless "safety plate" in the left crank and its locating hole in the spindle, introduced on 5600 105 and used on every Shimano 2-piece crank since, were deleted, there would be no need for a sealing plug because the tension bolt would serve that purpose.

[Tim]

I ride 4 different bikes on a regular 4-6 days/week basis. 2 of them fitted with Ultegra 6800 cranks, one with a DA9000 crankset and the other with FSA carbon.
There's a combined mileage of over 100,000km's on them.
None have ever given me the tiniest amount of trouble.
I'm not in the slightest bit worried about any of them.
If one fails, so be it. I've had good service from them all and apparently there's a degree of warning before they give way.
That said, I'll be a little bit more observant at chain lube and service times.

[Lukeyboy]

Not a big surprise. Most of the failures I've seen has been from corrosion ie heavy sweaters or those that love industrial strength undiluted degreaser. Introduce some power and she'll crack.

[Retrobyte]

This thread made me look at the weights of 105 vs DA and Ultegra chainsets. It seems almost pointless.

https://www.cyclistshub.com/shimano-105 ... -dura-ace/

Yep - just 500gm difference. Less than the weight of the water in your bidon. I'd be perfectly happy with 105 on my next road bike, and use the saving on a wheelset upgrade.

[Duck!]

So only for cranksets above 175mm??

Also only TT/Tri spec 54/55T big rings.....

Seriously though, although on face value 4500-odd reported failures is a significant number, as a proportion of nearly 800,000 units just in North America, that is well under 1%. So yeah, if you have one of the affected cranks, get it checked out, but the probability of it having a problem is in reality very very low.

From my understanding is that is 4500 crank failures that were reported to the CPSC, so there are most likely a lot more failures in the field.

Both the 4500-odd reported failures and the total 760000 cranks stated in the articles are only North American numbers. On those figures, the failure rate is about 0.6%. Another article I read today mentioned the number of cranks affected globally could be as many as 2.8 million. What's the global number of failures? We don't know. But applying that 0.6% failure rate hypothetically suggests ≈16,800 broken cranks worldwide. Yes it's a big number on face value, but 0.6% is still a tiny proportion.

To those with cranks in the affected batches, get them checked and stay tuned in to any weird feel or noises, but the probability of breakage is very low.

[g-boaf]

This thread made me look at the weights of 105 vs DA and Ultegra chainsets. It seems almost pointless.

https://www.cyclistshub.com/shimano-105 ... -dura-ace/

Yep - just 500gm difference. Less than the weight of the water in your bidon. I'd be perfectly happy with 105 on my next road bike, and use the saving on a wheelset upgrade.

Although I couldn’t resist a quip to stir people on sram, I must say the FC-R9100P “QC” that I have gave me no troubles over the huge amount of distance I did with it, and the power meter battery was very long lasting on a charge.

That was on an extreme lightweight “hyperbike” however so weight savings mattered to get it down to normal correct weight. I didn’t go for the THM Carbon brakes or cranks however nor Meilenstein wheels, simple too expensive.

[blizzard]

Also only TT/Tri spec 54/55T big rings.....

Seriously though, although on face value 4500-odd reported failures is a significant number, as a proportion of nearly 800,000 units just in North America, that is well under 1%. So yeah, if you have one of the affected cranks, get it checked out, but the probability of it having a problem is in reality very very low.

From my understanding is that is 4500 crank failures that were reported to the CPSC, so there are most likely a lot more failures in the field.

Both the 4500-odd reported failures and the total 760000 cranks stated in the articles are only North American numbers. On those figures, the failure rate is about 0.6%. Another article I read today mentioned the number of cranks affected globally could be as many as 2.8 million. What's the global number of failures? We don't know. But applying that 0.6% failure rate hypothetically suggests ≈16,800 broken cranks worldwide. Yes it's a big number on face value, but 0.6% is still a tiny proportion.

To those with cranks in the affected batches, get them checked and stay tuned in to any weird feel or noises, but the probability of breakage is very low.

I suspect true failure rate is well above 0.6% as a lot of failures would never have been reported to the CPSC. Still probably only ~1% which is actually pretty high.

From the discourse online you would think that everyone who had a failure, face planted and lost all their teeth. Seems the vast majority of people notice a noise or spongey feeling as you say and find the crack before catastrophic in use failure.

[Dave_rh]

I liked the poking sh it at aluminium though, thrown back the same stuff that is said of carbon.

Except there’s nothing wrong with the aluminum, it’s the “glue” that’s failed.

The glue did not fail. The aluminimum spider/crank corroded due to a galvanic reaction with the steel axle. This reaction was accelerated by moisture acting as an electrolyte.

The only failure of the glue is that it does not stick to powdery aluminimum oxide.

[baabaa]

powdery aluminimum oxide

Hmmm sounds good - Do you know the Shimano Original Part number?

[jetglo]

Has anyone contacted Shitmano and know the procedure for the check?

I have one set here that has failed so that will be replaced.

The other four sets I have that fit into the serial/model/product code for the recall - are about to fail or will never fail is unknown to me.
How a visual inspection would determine this is another matter.
They should just all be replaced based on my experience on the failed set.

Time will tell on this one.

[P!N20]

Has anyone contacted Shitmano and know the procedure for the check?

USA only at this stage. Stay tuned.

[Snowie1]

Yeah I called Shimano Australia, they're not sure what's happening yet.

https://bike.shimano.com/en-AU/informat ... ogram.html

[Thoglette]

The glue did not fail. The aluminimum spider/crank corroded due to a galvanic reaction with the steel axle. This reaction was accelerated by moisture acting as an electrolyte.

The only failure of the glue is that it does not stick to powdery aluminimum oxide.

Got a source for that? The original bike radar piece suggests otherwise & one wonders how the surface corroded if the glue was still properly bonded.

I’ve certainly seen lots of al. corrosion (on boats and seaside facilities ) where SS fasteners are used. And the occasional seized seat post or quill stem (after a few decades of neglect )

[familyguy]

powdery aluminimum oxide

Hmmm sounds good - Do you know the Shimano Original Part number?

SH-PO19.

[P!N20]

E.g. TooLongLegs is the only forum member I’ve seen to fatigue fail a traditional forged Al crank arm.

I did this, does it count?