Snapped spindle

[classic33]

As I showed several times earlier in the thread: 20 grammes extra material in the spindles wall (thickness, towards the inside ofcourse), would have increased its resistence to torsion forces with 29%.
The next question then is, which % of bicycle users would bother that 20 grammes extra?

The capital difference between HollowTech 1 and HollowTech 2 is that 1's crankset has 3 pieces and 2's crankset has 2 pieces - 1 being the left crank, 2 being a made-as-one-piece spindle+right crank.
So, there is no "Hollowtech 2 spindle" - it's one piece with spider / right crank, since your shown experience with bicycles, you surely know this, yet you ask me here to limit my talk to a part... of a part?

And, even in the hypothetical case HollowTech 2 had been 3 pieces as 1 and Octalink, there is a clear relevance: Weight Whining leading to safety risk.
There's the golden rule to regularly (let) inspect your bicycle (and any mechanical apparatus), but cranks, spiders and axles aren't wearing parts, who'd expect these to break?
That's as ridiculous as having to regularly inspect your car inside outside, including the cushioning of its seats, in order to discover a trend towards accident soon enough.

See Shimano's settlement text (with Consumer Safety department of US State), also mentioned on their website:
https://www.shimanocranksetsettlement.com/

The inspection procedure for dealers:
https://si.shimano.com/en/pdfs/dm/RAFC012/DM-RAFC012-01-ENG.pdf
Note there the many points to check, even also the other piece - the left crank, proving that failures also occurred there, and ALSO on various places alike around the pincher bolts.
But nothing about the integrated spindle.
But then we read (example this topic) about spindle breaks too, and clearly ALSO due to initiation of material defects, resulting in movement, resulting in fatigue.
The very same scenario of all other locations on the crankset.

You implicitly state here crank and other failures as irrelevant - another story than the spindle section of the part, but ALL data indicates the very same story.
Imagine the TopicStarter had let his crankset inspected. If the inspecting person hadn't bother to check OTHER places than those listed in Shimano's manual, it would have been flagged as OK, and a next day he could have been dead due to spindle sections failure and bad luck.

Conventional wisdom says increasing the outer diameter has a bigger difference in the strength of a spindle, not decreasing the inner.

 

[EckyH]

As I showed several times earlier in the thread: 20 grammes extra material in the spindles wall (thickness, towards the inside ofcourse), would have increased its resistence to torsion forces with 29%.

... which tells abolutely nothing about the long life fatigue strength.

As I cited before from a website you referred in this thread: long life fatigue strength isn't "add a safety margin, done".
That's basic mechanical engineering knowledge that even I have learned.
To give a example: good spokes are thinner in the middle part than near the thread and the head. That's not primarily for weight saving but for long life fatigue strength.

in the Church of the Light Religion

Repetition doesn't make a bad joke better. (Lord Mc Rashley).

E.

 

[Webbo2]

... which tells abolutely nothing about the long life fatigue strength.

As I cited before from a website you referred in a thread: long life fatigue strength isn't "add a safety margin, done".
That's basic mechanical engineering knowledge that even I have learned.
To give a example: good spokes are thinner in the middle part than near the thread and the head. That's not primarily for weight saving but for long life fatigue strength.


Repetition doesn't make a bad joke better. (Lord Mc Rashley).

E.

You are trying get a square taper fundamentalist to listen to reason. I think you may have to be patient.😱

 

[classic33]

You are trying get a square taper fundamentalist to listen to reason. I think you may have to be patient.😱

You are forgetting he might be a cotter pin fundamentalist.

 

[EckyH]

You are trying get a square taper fundamentalist to listen to reason.

It's more that I cannot let these obviously wrong assertions go unchallenged.
Hopefully other people - if they make the mistake to read this thread - will come to the conclusion that even if something snaps or breaks in more or less rare cases (relatively to the amount of pieces used out there) it could be a solitary case due to different circumstances: production mistake by the parts manufacturer, assembly mistake at the bike manufacturer or something else. Every case of a failure might be different.
Principally construction or production mistakes usually lead to recalls by the manufacturer - and yes, there are differences how manufacturers handle recalls.

And for the record: square taper cranks wear out at the square taper every time they are taken off and put on again.
Nevertheless I ride both square taper cranks and Hollowtech cranks.

E.

 

[EckyH]

You are forgetting he might be a cotter pin fundamentalist.

Decades ago one of the "under 19" riders in my cycling club sheared off a cotter pin. Fortunately it was at relatively low speed because he was at the start of a time trial on a cycling track.

E.

 

[silva]

The number and %age in use of H2 chainsets which have failed because their spindle failed is negligible: we read about it rarely. The mean number of rotations between spindle failures is bound to be in the tens of millions - decades in typical rates of use.
I made this MTBF point in Post #3 in this thread.
Any time it happens it is remarkable but that does not mean the design is unsatisfactory nor that any action is needed: it's happenstance.

The same has been declared about other sections of it, and the other part of the HT2 assembly, until when people started a blog with pictures of fails, until so many fail pictures that keeping repeating those declarations ceased to be an option, being publicly forced to admit it, forced to do something about it, and the data given in the legal case that some started in the US, proved that Shimano was aware of the design fail during their entire denial decade.

You stick the word "remarkable" on this topics break of the spindle section - there has been a time that the fails elsewhere in the crankset were also described like that. In the end, as was proven, and also logical - everything happens for a reason, including a break, why would the spindle section be an exception?
It's simple: this specific spindle section should never break. NEVER. Nog a single occasion, for the simple reason that it's not a axle of a wheel that supports a weight and could thus be overloaded, in which the blame is on the one that overloads, not on the axle / it's producer.
While human leg power is well known.
Thus, if a cranksets axle breaks, it's due to its design, most likely due to the Light Religion to cheat in races. A minimalistic design, as less as possible material, resulting in parts/sections operating near their mechanical limits.
I also said here nothing new.

MTBF has nothing to do with this - it's just a crankset lol, just two parts lol, not some complex machinery.

 

[silva]

...
And for the record: square taper cranks wear out at the square taper every time they are taken off and put on again.
Nevertheless I ride both square taper cranks and Hollowtech cranks.

E.

That was exactly the invention reason of the "taper" in "square taper": allowing to compensate for wear. Nothing needs to be replaced, the crank just shifts abit further on the axle, and as long as it doesn't bottom out, No Problem in the Champaign Room Sir!

 

[Webbo2]

That was exactly the invention reason of the "taper" in "square taper": allowing to compensate for wear. Nothing needs to be replaced, the crank just shifts abit further on the axle, and as long as it doesn't bottom out, No Problem in the Champaign Room Sir!

Please explain why light in races is cheating and it would be helpful if you didn’t make the answer several pages long.

 

[classic33]

The same has been declared about other sections of it, and the other part of the HT2 assembly, until when people started a blog with pictures of fails, until so many fail pictures that keeping repeating those declarations ceased to be an option, being publicly forced to admit it, forced to do something about it, and the data given in the legal case that some started in the US, proved that Shimano was aware of the design fail during their entire denial decade.

You stick the word "remarkable" on this topics break of the spindle section - there has been a time that the fails elsewhere in the crankset were also described like that. In the end, as was proven, and also logical - everything happens for a reason, including a break, why would the spindle section be an exception?
It's simple: this specific spindle section should never break. NEVER. Nog a single occasion, for the simple reason that it's not a axle of a wheel that supports a weight and could thus be overloaded, in which the blame is on the one that overloads, not on the axle / it's producer.
While human leg power is well known.
Thus, if a cranksets axle breaks, it's due to its design, most likely due to the Light Religion to cheat in races. A minimalistic design, as less as possible material, resulting in parts/sections operating near their mechanical limits.
I also said here nothing new.

MTBF has nothing to do with this - it's just a crankset lol, just two parts lol, not some complex machinery.

I twisted a Sturmey Archer five-speed hub axle through 90°. Sending a letter to ask how much a replacement would cost, if not covered by their guarantee, got the response that it was more than likely some other part that had failed. Thereby stopping the hub from operating correctly.
I rang the number on the letter, got put through to someone who said what I was describing was impossible, but he realised that I was talking about the piece in my hand at the time. Was there any difference between the axle in their three-speed hubs?
I was asked to return it for them to examine and test. They supplied a replacement axle, free of charge, along with a reply that said what I'd done shouldn't have been possible. They were tested to beyond what was expected to be put through them, before being released for assembly.

Failures happen, and I have no reason to disbelieve what I was told.