Shimano HollowTech crank arm failure (article)

[AngryScientist]

https://www.bikeradar.com/features/shimano-crank-failure/

Saw this posted at Vsalon and thought it worthy of discussion here as well.

I think the article is responsibly written, without jumping to baseless conclusions.

personally, i have ultegra cranks on my most abused gravel bike, and the finish does seem to wear especially fast. i do not ride in the wet all that much, but if i did regularly, i sure would inspect my cranks much more thoroughly/often and pull from service at any sign of seeping corrosion.

come to think of it, my rain bike uses a hollowtech crank too, but doesnt really see piles of miles, but that may be even worse? hmmm

interesting stuff.

https://images.immediate.co.uk/production/volatile/sites/21/2020/03/Shimano-cransket-main-8c77328-scaled.jpg?webp=true&quality=90&resize=620%2C413

[Mikej]

I analyze foundry sand castings, I will say this about Shimano- they have an incredible incredible amount of knowledge and experience in this area. I would go so far as to say Shimano is a foundry company that makes components. And yes, components can fail. That model crank - 6800, could be 7 years old.

[AngryScientist]

I analyze foundry sand castings, I will say this about Shimano- they have an incredible incredible amount of knowledge and experience in this area. I would go so far as to say Shimano is a foundry company that makes components. And yes, components can fail.

no doubt about that. they have been at it for a long time. everyone went to carbon for cranksets, but they are holding the line with hollow alu.

[Mikej]

no doubt about that. they have been at it for a long time. everyone went to carbon for cranksets, but they are holding the line with hollow alu.

Shimano did make a run of Dura Ace carbon cranks - maybe 7800?https://www.bikeradar.com/news/shimano-launches-carbon-dura-ace-crankset/

[Gummee]

In all the years these things have been out, I've seen exactly zero of them break. ...including the ones in my garage.

This seems to be a lot like BMW final drives. According to the guys that have had them go bad, they're a ticking time bomb. To those of us that have ridden them trouble free for many miles, it's much ado about not much

M

[unterhausen]

I think it was just bad glue.

[Mikej]

I think it was just bad glue.

Interesting comment. I worked with somebody that interned at Boeing. His job was to note metal adhesive failures caused by petroleum products. Tri flow would eat loc-tight. Very possible.

[benb]

These fail occasionally just like everything else. A couple people I know have broken cranks over the years.

Mostly really strong riders over 6’, 175mm or longer cranks, guys who are Cat 1/2/Pro who put down really huge watts, only own a small # of bikes and ride 10k or more miles a year so they’re putting tons of cycles into the same crank.

Basically the same riders who fatigue frames into cracking without crashing. One of the guys I know who broke a Shimano crank cracked 3 Ti frames at the BB and finally got into a dispute with the frame manufacturer because they stopped honoring the warranty.

Crank failure is a lot like rings wearing out causing skipping. Not that big of a deal if you’re seated, scary as hell if you’re out of the saddle.

[oldpotatoe]

no doubt about that. they have been at it for a long time. everyone went to carbon for cranksets, but they are holding the line with hollow alu.

Hollow aluminum that is then welded together. This 'failure' is not common but not unheard of either since they started welding these 2 halves together.

[Gphin]

My Ultegra crank broke on my Gravel bike last year after 2 seasons. Luckily my shop was able to get it warrantied. They don't make like they use to.:)

[zmalwo]

Ultegra 6800 drive side cranks were known to fail at the joint right above the T letter. This is why Shimano made the 2nd gen 11 speed cranks asymmetric with beefed-up upper area.

[zmalwo]

Plus who knows how many times this guy has crashed on the drive side to weaken the crank...

[robt57]

Plus who knows how many times this guy has crashed on the drive side to weaken the crank...


Or left it in hot cars often? like really hot cars or hot garage?? I use heat to un-bond lots of stuff.

Is the R8000 made with same construction design?

[ultraman6970]

I have to be honest, I thought shimano was producing a one piece hollow crank, not a 2 piece crank which produce a space between the two parts.

[cabriggs]

This happened to a friend of mine a month ago!

Coincidentally, three days later I came across this, which happens to include the same failure, among others.
https://cyclingmagazine.ca/sections/gear-reviews/bikesframes/justridingalong-finds-the-weirdest-mechanicals-on-the-internet/

[Velocipede]

This problem is more common than people realize. I've had 6 so far like this since the 6800 crank came out. And I'm just one person. While I love to say it's Customer Error, it's not. It's a design flaw. And there are ways to remove it but they clearly don't.

Also, I've seen this on a crankset(Dura Ace 9000) that was only raced on a triathlon bike. It's never been crashed or even laid on its side. It's never been in a hot car all day or anything like that. It's literally kept in a bike room, only raced 14 times a year but a 5'10", 145 pound guy who is all spin. He carries his bike in and out of the house and to and from the car to the race rack. It's a babied bike. All of their bikes are. I'm the only person who's worked on their bikes for the last 20+ years.

[CiclistiCliff]

Seen far too many, 6800; 9000; 8000; 9100.

[ryker]

Lots of these failures...

https://instagram.com/thanksshimano

[ultraman6970]

dont want to start something but clearly there is a problem and a common one.

THere's stats for this???

Campagnolo fails aswell but wonder how much more it fails, or not, compared with those shimano cranksets. Personally never had a campagnolo crankset to fail me, same with the old shimano squared tapper, neither suntour.

[robt57]

Smokes, i don't wanna ride my 11s Shimano chainsets now.

Is FC-5800 also among this issue??

[davidlee]

Everything eventually fails! :banana:

[mcteague]

dont want to start something but clearly there is a problem and a common one.

THere's stats for this???

Campagnolo fails aswell but wonder how much more it fails, or not, compared with those shimano cranksets. Personally never had a campagnolo crankset to fail me, same with the old shimano squared tapper, neither suntour.

I had a Chorus carbon crank where the pedal threads came loose. This also is not unheard of. It never came out and gave me a fair amount of warning that something was amiss. At first I thought the pedal was not holding the cleat tight enough. The only other crank failure I had was an old Avocet where it snapped at the pedal threads. Luckily it was on a hill and I was not going very fast so did not go down.

Tim

[kramnnim]

Smokes, i don't wanna ride my 11s Shimano chainsets now.

Is FC-5800 also among this issue??

No, I don't think they, have the same bonded clamshell design

[bikinchris]

I have seen several bikes with hollow Shimano crank cracks. Two of them on Dura Ace.

[AngryScientist]

at least with aftermarket rings, you can keep an eye on the glue job.

https://1.bp.blogspot.com/-7myEBQUit3Y/XZ9fvnSpVOI/AAAAAAAADVo/aWcqh77GOJoL1wgt-1ZA0jwZ7RJWynt-ACLcBGAsYHQ/s1200/IMG_4571.jpg

[fignon's barber]

My guess is that it seems likea lot of failures because there are so many out in the wild compared to other brands. That being said, what are the best alternative brands/model cranks if you used ultegra or dura ace groupsets?

[benb]

Is there actually any glue in these? That doesn't make a lot of sense.

If they're cast/welded (older ones didn't even have any welds) there is no way the heat of a car on a hot day is going to do anything.

The car would have to be on fire to affect aluminum, and who knows what else Shimano put in the alloys, it could be more or less heat resistant but we're still talking 5-10x what a car would get to.

My previous post I was actually referring to the older models.. I was talking about riders cracking the pedal end of the crack right off, not detaching it from the spider.

edit: Looked at the article.. I have the older version of this.. at what point did they start bonding these? I wonder how much weight it saved, it seems really hokey compared to the earlier design where the cranks are seamless. I bet these new ones are cheaper to make.

[unterhausen]

My guess is that it seems likea lot of failures because there are so many out in the wild compared to other brands.
I had a similar experience doing support on f-16s. When I was at the office, there was always a crisis with one jet or another. When I spent a couple of months in the field, nothing ever went wrong. If you go looking for problems with these cranks, you will find a lot of them, but most of them are doing just fine.


at least with aftermarket rings, you can keep an eye on the glue job. You can buy TA covers for aftermarket chainrings to hide the cracks if you want. Peter J. White carries them.

[robt57]

I guess a good question might be, do after market rings such as 30/46 set create extra stress or mitigate stress in term of these glued joints.

[kramnnim]

edit: Looked at the article.. I have the older version of this.. at what point did they start bonding these? I wonder how much weight it saved, it seems really hokey compared to the earlier design where the cranks are seamless. I bet these new ones are cheaper to make.

Started with 9000 and 6800.

[kramnnim]

They recalled Alivio cranks back 20-25 years ago. What needs to happen for these to be recalled? Deaths?

[kramnnim]

https://www.cpsc.gov/Recalls/1997/cpsc-shimano-announce-recall-of-bicycle-components

"Shimano has received more than 630 reports of cranks breaking in North America resulting in 22 injuries, including cuts and fractures."

How have these 6800s not been recalled?!

[benb]

I had a similar experience doing support on f-16s. When I was at the office, there was always a crisis with one jet or another. When I spent a couple of months in the field, nothing ever went wrong. If you go looking for problems with these cranks, you will find a lot of them, but most of them are doing just fine.



The difference with aviation is likely every part on an F-16 that could fatigue had engineering time to understand it's fatigue lifetime and it has a maintenance interval set to replace the part before it would fail like these cranks. The F-16 is probably right on the verge of the first planes that had this done to an extremely rigorous degree. My dad did his graduate research on this in the early 70s.. specifically with aviation as the application. Certainly the F-22 and F-35 have had this heavily taken into account. But those planes are also flow to a bleeding edge compared to civilian planes and there are always mistakes too.

I don't really think (but also don't really know) bike companies ever do this level of engineering... and Shimano is probably the pointy tip of the spear for bike companies doing legitimate engineering.

We don't get time/hour estimates on parts lifetime for anything on our bikes for the most part other than chains. We're mostly in the dark, probably cause the bike industry knows most of the bikes mostly sit in the garage.

This is one of the reasons it's always amazing when bikes cost more than motorcycles.. motorcycles are way closer to that level of engineering.

If we had things like "seatpost binder bolts must be replaced every time seat is adjusted" and "stem must be replaced every season, bolts must be replaced every time the stem is adjusted" and "This hollowtech II crankarm is rated for 5000 miles" I think consumers would not buy as many of the bleeding edge bikes. The motorcycles I owned had that level of caution/engineering in their maintenance procedures despite not being as bleeding edge as a lot of the road bikes being sold today.

[robt57]

We need ADs for bikes. AD=AirworthinessDirective.

In this case TDs, as in TesticleWortiness... never mind.

[Pastashop]

https://www.sheldonbrown.com/brandt/breaking-cranks.html

https://yarchive.net/bike/cranks.html

https://yarchive.net/bike/crank_break.html

Been there before.... diminishing performance returns on lightness & stiffness, growing risk to life and limb.

[benb]

We need ADs for bikes. AD=AirworthinessDirective.

In this case TDs, as in TesticleWortiness... never mind.

It's only partly a joke.

If a bike had an aircraft or even motorcycle level maintenance spec and warranty & liability hinged on following the maintenance schedule I think it would shift what bikes people bought.

If a $1000 aluminum bike weighing 20lbs ended up costing $X/hr in maintenance and a $10000 Carbon bike right at the UCI limit ended up costing $5-10X/hr in maintenance you might not see as many people jumping on the high end bike.

Maybe an F1 car is a better example... the bike industry has repeatedly compared road bikes to F1 cars. F1 cars have a very high maintenance cost/hr... the bicycles meanwhile ship out with "don't do anything till something stops working." The truth is somewhere in the middle but most of the time we don't do much proactive maintenance on our bikes. We clean & oil the chain when it starts to start making noise. We ride our cables till shifting gets screwy. We replace chains once shifting starts getting screwy. Most of us don't change tires till they the tread is 100% gone and the casing is showing. Headsets & wheel bearings never get looked at till they are shot & heavily degraded. Most frames never get taken apart and inspected for cracks, corrosion, rust, etc.. until the situation is very far gone.

I used to be really good about doing stuff periodically.. I'd overhaul my bikes at least 1x a season with pretty much complete disassembly & inspection and do cables/chains, etc.. 2x a season, tires & brake pads got changed long before they were worn out. Now I'm way more time crunched that was the first thing I left behind. Only thing I was always a bit remiss about was opening up hubs and such to check them.

[robt57]

Well, I was only 1/2 kidding really..

[choricoff]

This happened to my 6800 crankset in 2017 (different break than the photos in the Bikeradar article). It had around 15-18,000 miles and about 3 years of use when it happened. Luckily I noticed a creaking before it caused a crash or injury and luckily it wasn't on my primary bike when I noticed the failure.

I was really upset at the time, but Shimano was very quick about replacing the crankset (they replaced entire crank with chainrings even though I took the old chainrings off before I sent the broken one back to them). I wish I could find the correspondence I had with Shimano - it seemed like they were well aware of the issue and were eager to placate. Haven't had an issue with the replacement (yet).

https://i.imgur.com/eoevHg2l.jpg

[Velocipede]

This happened to my 6800 crankset in 2017 (different break than the photos in the Bikeradar article). It had around 15-18,000 miles and about 3 years of use when it happened. Luckily I noticed a creaking before it caused a crash or injury and luckily it wasn't on my primary bike when I noticed the failure.

I was really upset at the time, but Shimano was very quick about replacing the crankset (they replaced entire crank with chainrings even though I took the old chainrings off before I sent the broken one back to them). I wish I could find the correspondence I had with Shimano - it seemed like they were well aware of the issue and were eager to placate. Haven't had an issue with the replacement (yet).



Oh they are WELL aware of it. It bugs some of the employees I've spoken to. A couple of them just shake their heads or sigh heavily when it's spoken about. They are tired of being hit with this. Shimano isn't known for these kind of issues. That crank from the 90's and the stupid plastic brake spring caps. That's really it for them for the last 20+ years major warranty wise.

They hate the stigma of these. Why they are keeping them around? Honestly, I am betting its cost. These can't be that expensive to make and they've invested a ton into them. It's probably cheaper to keep the design around and do warranties on them than to redesign the whole crank. Maybe for a 13 speed or the new 12 that's coming out.

[oldpotatoe]

Oh they are WELL aware of it. It bugs some of the employees I've spoken to. A couple of them just shake their heads or sigh heavily when it's spoken about. They are tired of being hit with this. Shimano isn't known for these kind of issues. That crank from the 90's and the stupid plastic brake spring caps. That's really it for them for the last 20+ years major warranty wise.

They hate the stigma of these. Why they are keeping them around? Honestly, I am betting its cost. These can't be that expensive to make and they've invested a ton into them. It's probably cheaper to keep the design around and do warranties on them than to redesign the whole crank. Maybe for a 13 speed or the new 12 that's coming out.

Bing Bing Bing, we have a winner.

[benb]

Yah these parts are forged, probably not cheap to redo the tooling.

Except it seems like they’ve already redesigned them once unsuccessfully?

[Davist]

Except it seems like they’ve already redesigned them once unsuccessfully?

it would seem the 8k ultegra was a retooling (if not redesign). weird. In my corner of the world, I've seen lots more carbon cranks "eject" the pedal insert, but still only 2 total (one FSA one Specialized, I don't know who makes them).

I DO like the shimano design of keeping bearing tension separate from the clamp/pinch bolts.

[Velocipede]

it would seem the 8k ultegra was a retooling (if not redesign). weird. In my corner of the world, I've seen lots more carbon cranks "eject" the pedal insert, but still only 2 total (one FSA one Specialized, I don't know who makes them).

That's due to stupidity and trying to make a lighter part. There's an easy fix for it but no one is doing it.

[Blue Jays]

"...This happened to my 6800 crankset in 2017..."Good catch, @choricoff. Very glad you spotted the crack.

[benb]

Yah so I was wrong earlier in the thread.. I have a FC-6800 Ultegra crank on my Domane. (53/39 version)

Easy to get confused, cause I have 10-speed on that bike, but I bought a 6800 crank for it at the time because I was buying a Stages PM, and I wanted to future proof it. When I got the bike I put old wheels on it and used 10-speed because my old wheels had Shimano 10-speed hubs. The old wheels are gone now though. I mostly ride that bike in nice weather, so it's not seen that many miles and the components are lasting forever, but it'll get 11-speed at some point.

Guess I need to watch for this. My Space Horse has a 50/34 105 FC-5800 crankset on it as well. I wonder if those are also subject to this. That bike I beat on and it's got more miles on the crankset than the Domane does. Same situation on that bike, I bought a Stages PM so I wanted to future proof the bike for 11 speed. But that bike I still have the original 2013 Tiagra Hubs on.. LBS put new bearings them last winter. Those wheels are overbuilt and will last seemingly last forever. They got at least 15k miles before the hubs even needed service and they've never needed a truing even though I even occasionally ride that bike right down rocky singletrack.

[bicycletricycle]

I know adhesive engineers can give endless reasons and specifications proving that a glue joint is superior to a mechanical one.

but

Glue joints can be a real pain to do well and consistently in a production setting and I just try to avoid them in metal bicycle parts.

[bicycletricycle]

I replaced thousands of those plastic spring housings and cheap cranks.



Oh they are WELL aware of it. It bugs some of the employees I've spoken to. A couple of them just shake their heads or sigh heavily when it's spoken about. They are tired of being hit with this. Shimano isn't known for these kind of issues. That crank from the 90's and the stupid plastic brake spring caps. That's really it for them for the last 20+ years major warranty wise.

They hate the stigma of these. Why they are keeping them around? Honestly, I am betting its cost. These can't be that expensive to make and they've invested a ton into them. It's probably cheaper to keep the design around and do warranties on them than to redesign the whole crank. Maybe for a 13 speed or the new 12 that's coming out.

[Velocipede]

I replaced thousands of those plastic spring housings and cheap cranks.

Me too. Man they sucked. I still have spare spring housings in a drawer! I think I have a replacement crankset around as well.

[bicycletricycle]

the shimano guy gave us boxes of them, we replaced those things for years and years and years.

Me too. Man they sucked. I still have spare spring housings in a drawer! I think I have a replacement crankset around as well.

[A1A]

I just looked out of curiosity, I have an FC-CX70 crank on my Cyclocross bike, a Ridley X-Fire. Is this in the same "family" and subject to potential failure? Looks similar to the road cranks shown.

[robt57]

Did we decide if the FC-R7000 is bonded or still like the fc-5800?

Found a pic that suggests spider side like 5800 and not like bonded units.

[Velocipede]

Did we decide if the FC-R7000 is bonded or still like the fc-5800?

Found a pic that suggests spider side like 5800 and not like bonded units.

5800 and 7000 are not bonded. You can see it really well in the area of the chainrings. You can see the two pieces, their seam.

[robt57]

5800 and 7000 are not bonded. You can see it really well in the area of the chainrings. You can see the two pieces, their seam.

Yeah, thus my last sentence you quoted there. That was my conclusion from that pic.

Just ordered an R7000 to go along with my FC-5800 and GRX-810-2. Will be selling off two 6800 and one R8000. I put 30/46 rings on these anyway, 30 grams is not worth my knuts...

[jdh]

Yeah, thus my last sentence you quoted there. That was my conclusion from that pic.

Just ordered an R7000 to go along with my FC-5800 and GRX-810-2. Will be selling off two 6800 and one R8000. I put 30/46 rings on these anyway, 30 grams is not worth my knuts...

Are GRX-810-2 not bonded either?

[earlfoss]

the thanksshimano IG account is freaky!

[Tz779]

His job was to note metal adhesive failures caused by petroleum products. Tri flow would eat loc-tight. Very possible.

This could be handy if you cannot apply heat!