Shimano cranks failing. Time to do something about it?

I think in 2022 a company can’t ethically justify they make stuff that is deemed acceptable to break in less than 10 years. Yes poor people don’t buy new dura-ace component. However sooner or later those bikes end up in the second hand market at a much lower price. Chainring, chain and cassettes are wear items and can be considered consumables. Crankset aren’t and shouldn’t.

2 Likes

I totally see the environmental argument here, and I agree in many ways. That said, engineering something to last for that amount of time, seeing potentially daily use, not typically regularly maintained, is going to add weight or cost or probably both, and probably a lot.

You’ll need to do massive amounts more on QC, which will mean increasing machining/manufacturing costs, more manual checks, and a much higher reject rate.

Given the current concerns about cost already, I think a big further hike is going to be a hard sell, as is something a lot heavier.

I still think engineering something that has a solid 5 year minimum expected life span, is sensibly priced, reasonably lightweight, and can be recycled, should be the target.

1 Like

How long should I expect the crank in my toyota to last?

1 Like

Good thing then that no one is doing that then, I guess….#relief

Sarcasm asides, the fact it is the norm doesn’t mean we shouldn’t as consumers point a finger at them.

Wait a minute, whose cranks last five years? Shimano Hollowgram cranks last way more than 5 years for the massive, massive % of them. People would be furious if cranks only lasted five years. Even the old SRAM Red mechanical, king of too light for its own good, has a crank that will last indefinitely.

1 Like

Exactly. This point has been made over and over again. There is a huge difference between anecdotes and data. Even the CT article on the issue, with comments from multiple retailers, seems to confirm that this is a very small, but highly visible, issue.

2 Likes

#metoo - Fail on my Ultegra R8000 right crank after a month thinking it was my pedal feeling a bit strange (and being aware of this issue I had inspects the crank arm) my Shimano Gluetech crank finally failed.
Luckily (in the circumstances) it cracked going up hill rather than sprinting and was on a club ride so had support. Just about lasted till I got home but last 2k had to cycle left legged only.
Just 15,000 k on the crank clock.

The same thing has happened to two other guys in my club in Dublin :ireland: and a third experienced same issue on a rental bike abroad. Anecdotal / vocal minority it may be - but it seems to be more common than you’d expect.

Had to fork out and eye watering amount for a full chainset, as bike shop wouldn’t sell just crank arm…and hard to source fast online.
They are contacting Shimano re: replacement /covering cost which I hope comes through.




1 Like

I have the feeling it is not that anectdotal.

I mean yes in the grand scheme of things only a small amount fail that quickly, but I have the feeling that if you live in a coastal city and/or a place with high humidity levels the odds of having a failure is much much higher.

1 Like

Statistics > Feelings.

Shimano has the numbers and the data. We don’t. The fact that they have not had to issue a recall. (Mandatory at some point based on failure rate, but I forget the exact threshold) is a key indicator re: the pervasiveness of the issue.

It’s v low-tech part. For what you pay, expect it should last the lifetime of the bike. Wty is for mfg defects, etc. and is appropriately limited to 1 yr (or is it still 3yr for DA?). I have/had R6800, R7000, R8000. Statistically, if the fail mode is repeatable and high-consequence, you don’t need overwhelming counts to conclude there is a pb. There’s clearly a pb with the bonded plastic/alu cranks. That they won’t admit to it is an especially bad look. Take your lumps, lay blame where it is deserved and use that to fire the jackass folks who pushed this ‘great’ idea - pretty sure not any engineering mgr. That will do wonders for long-term reliability and reputation.

Aren’t recall only issued if safety is at risk?

Given the failure mode, there is very little risk in term of safety as the crank becomes noisy and noodly way before it is too late.

Raceface Next carbon crank were all failing repeatedly, some owners having it warrantied up to 3 times (most gave up and switch to other vrand way before) without Raceface issuing a recall. A crank just pose very little risk when failing, at worst you are hitting your foot on the ground which lead to a small, non life threatening injury.

Yeah, that simply isn’t true…people have crashed as a result of these failures.

You don’t really seem to be listening to what people have said, nor do you seem to understand the nature of the issue, so I am just moving on.

Yes I do but you just dismissed the point about the raceface carbon crank that had a high failure rate, any LBS can attest to that and were never recalled.

My point is that if say 0.001% of the crank worldwide, it is one data point. But it is irrelevant if 5% of those that are used in specific conditions (coastal, salt air + high humidity level) end up failing as well. In that case Shimano would have to revise its copy or issue a warning that those items must not be used in those conditions. Is 5% the case? I don’t know but it looks there is a pattern in the failure and humidity. The cyclingtips forum is very small, village sized, yet we see some direct reports about it from its members.

Shimano has the true numbers, sure. Would they communicate about it? Are you naive? This wouldn’t be the first time a company would hide the true reality under a carpet. Japanese companies in general have a terrible record in that regard.

The fact that Shimano altered the design without communicating about it shows that they are aware of the issue and treat it as worthy one to do something about it, regardless or my and your understanding.

1 Like

If this were really a pursuable issue, the personal injury class action vampires in the US world have done something by now.