Snapped spindle

Page may contain affiliate links. Please see terms for details.

silva

Über Member
Location
Belgium
How can there be discussion when I speak from what I actually do from experience and works without fuss? What you found on the Internet is generic, and not written for a specific model or even make of Hollowtech crankset. You are simply wrong if you claim I need more than Allen keys to remove my HT2 cranks. Even the YouTube video your links eventually reach does not suggest hammer and chisel to remove the BB. It says nothing about the BB.

This reminds me of a teenager who has found a book on driving, never having driven, explaining to a police advanced driving instructor what he is doing wrong.
How can there be discussion if you react on what you said, not me?
I did NOT claim that YOU need more than Allen keys, I, and ParkTools, said it can needed to have more, which I witnessed in a bike shop, which is mentioned in the ParkTools Howto. See again here: you edited, thus, you are discussing actually with yourself here.
Also, my previous bike had Hollow Technology 2, and its bearings, despite larger, failed, which was a first time on a bike ever.
Also, my current (avatar) bike has had an Octalink period, as a test, and its left crank kept on loosening, indicating a design problem. I Looked It Up On The Internet (something you seem to flag as a BAD THING) and indeed, a design problem.
So your implicit claim that I only know things "from the internet", for what it's worth (nothing) is a lie, a rather careless one, since I posted alot experience with my bikes during the years I was active here.
So (and again) why your General with capital G negative attitude?
You don't have a reason for it and nobody is served anything.
I said that the Hollow Technology Light Religion serves the people that wanna win 1 sec on 1 km (not by better physical condition but by cheating), and/or, want to keep up with a rest, and that's it.
For all others (which is most), its cost/benefit ratio is awful.

Unlike you, I reacted to what is said, not what I selected/edited from it.
You tried to restart this so-called "discussion" twice.
Repeating doesn't make more true or more false.
It('s as simple as that, and you resorted many pages ago already to such forum behaviour.
That's the choice you make, not me, for me, it's just boring page scrolling, which is why I pointed you out your two restart attempts.
But as I see with your post here - to no avail.
 

fossyant

Ride It Like You Stole It!
Location
South Manchester
Many millions of people are perfectly happy with Hollowtech 2 and the competitors versions. You do get failures, that's life, but it's very small if you exclude the Ultegra/Dura Ace issue on the crank arms.

You've had a failure. Yeh the bearings don't last as long, but they are cheap and easy to replace. But, your opinion alone doesn't mean they are crap, nor is it some religion of people wanting things faster etc etc.

I've had no issues with hollowtech and SRAM and FSA's version. The FSA and SRAM have had a much harder life - both fitted to the bikes that get covered in mud. The SRAM equipped bike get's ridden in horrendous conditions (MTB) and that does suffer bearings wear from ingress of mud and dirt. Costs me about £10 a side to replace the bearings (and they aren't SRAM) - I use Enduro and it's a Praxis BB shell.

I've had no axel's failing, and 99.9% of others haven't. The OP has gone out and replaced it with the same part.

I'll agree, Octalink wasn't particularly great on the axel/crank interface, but if you fitted it correctly and to the correct torque, you'd be OK - ride it with loose or badly fitted cranks, then you are stuffed - that will happen with square taper - you'll knacker the crank first.

Everyone is entitled to a view, and we/you should respect other's opinions, so please be aware 99.9% of people using these cranks don't thing they are crap. In your opinion they are crap, in the majority of opinions they are reliable.
 

silva

Über Member
Location
Belgium
It has been suggested by someone here, can't remember who, that Shimano should simply increase the diameter of the spindle to make it stronger. This is an interesting suggestion that requires some nerdy analysis.

Let's call the spindle/axle/middlethingy a torque tube for now. A torque tube is something that resists turning or, transmits a turning force. torque tubes are well understood in engineering. They follow some laws:
1) given a certain amounty of material - say 100 grams, a large hollow tube of 100 grams is strongere (resists torque more)n than a solid rod of the same 100grams

2) The bigger the tube diameter, the stronger the tube

3) Material added to the inside of the tube (increased wall thickness) only increases strength linearly. In other words, double the wall thickness only doubles the strength.
4) Increasing the tube diameter increases strength to the cube (power of 3) of the diameter.

Simply put, bigger and thinner is stronger than smaller and thicker.

And herin lies the genius of Shiman's design. Shimano specified the maximum diameter that will fit inside a standard Engilsh BB shell, with reasonable clearance. This means that the design was backward compatible, lighter and simpler.
Shimano moved the space-eating bearings to the outside of the standard BB, leaving the maximum diameter for the tube.
Shimano specc'd the tube diameter to satisfy 99% of our needs whilst still keeping the weight weenies happy. Remember, at the time of the invention, the weight weenies were farting around with hollow axle and titanium axle BSA BBs in order to make them lighter. These things were super unreliable, including Shimano's own hapless Octalink 1 and II

So yes, Shiman can increase the wall diameter to the inside of the tube but the gains will be marginal. Instead, they went with CroMo tubing that keeps most of us happy.

The above is a separate issue from the failed Octalink, ISIS and GXP. It trumps them all in every single way. Also, keep the Hollowtech II glued cranks out of the equasion. That's another folly on its own and again your fault - you buy bike parts as if yoiu're buying cocaine - by the gram.

The opposition, i.e. GXP is super crap. Its left side is only 19mm (IIRC) and thus, following the bigger and thinner law, pretty stupid. They had to make the 19mm end very thick and very heavy, yet it still flexes and muches that bearing.

Forgive Shimano for Octalink, embrace Hollowtech I

Now, if only Shimano changed its nomenclature and didn't confuse us with the same name for hollow cranks and hollow axles, that would have been nice.
All I know, didn't check if others here did, is that I asked earlier in this thread if the inner diameter of the tube could be decreased, as to have a thicker and thus more torsion/break resisting wall, since torsion, as you also stated, is a rather clear (the 45° angle) cause of the break.

Because, for reason you also state here, the outer diameter would be a drastic change that wipes all compatibility with existing frames.
As I looked up on the internet (that some flag as a bad thing), in a techie doc, a formula was shown, that contained inner and outer diameters, to deliver the torsion resistence, which is rather logical, since any more material in the wall, can only increase the torsion resistence, not in a linear fashion ofcourse, BUT, the smaller the diameter, the more linear (= more effect) with it, and of course vice versa, as you also stated here the other way: the larger the diameter, the less the impact of the wall thickness on the torsion resistence.

The original question thus is, is a smaller inner diameter possible? It would ofcourse add again weight to the spindle (as I illustrated in earlier post here with a proposal of a 110 instead of 130 grammes overall (whole of crankset) "gain" relative to square taper), so for the Light Religion a partly but maybe acceptable setback, to trade for Safety First plus (chance of break under leg power becoming nil).
Can such smaller inner diameter void something else of the HT2 design?
There is a "bolt" that pushes (shifts) the left crank over the splines into place, that's all, that latter, the actual mechanical mounting is the clamp forcefully closed with the 2 sideways ("pinch") bolts.
That bolt has to be given thread, so "inside" the hollow, there must be some part with a threaded hole, and the outer diameter of that part might void a decreased (to get a thicker tube wall) diameter of the spindle/tube.
Also the other, drive side. The tube must be connected to the spider somehow, with something internally.

If that is the case, then a thicker inner wall, even if Shimano had be willing, is not technically possible, out of question, rendering HollowTech 2 to a compatibility design at the end of what was possible.
An interesting question would be why they changed the 3 parts of HT2 version 1, to 2 parts in version 2. Was that solely to further reduce weight / making mount easier, or was there a problem with version 1, that they solved by permanent fixing of drive side to spindle.
 

silva

Über Member
Location
Belgium
Many millions of people are perfectly happy with Hollowtech 2 and the competitors versions. You do get failures, that's life, but it's very small if you exclude the Ultegra/Dura Ace issue on the crank arms.

You've had a failure. Yeh the bearings don't last as long, but they are cheap and easy to replace. But, your opinion alone doesn't mean they are crap, nor is it some religion of people wanting things faster etc etc.

I've had no issues with hollowtech and SRAM and FSA's version. The FSA and SRAM have had a much harder life - both fitted to the bikes that get covered in mud. The SRAM equipped bike get's ridden in horrendous conditions (MTB) and that does suffer bearings wear from ingress of mud and dirt. Costs me about £10 a side to replace the bearings (and they aren't SRAM) - I use Enduro and it's a Praxis BB shell.

I've had no axel's failing, and 99.9% of others haven't. The OP has gone out and replaced it with the same part.

I'll agree, Octalink wasn't particularly great on the axel/crank interface, but if you fitted it correctly and to the correct torque, you'd be OK - ride it with loose or badly fitted cranks, then you are stuffed - that will happen with square taper - you'll knacker the crank first.

Everyone is entitled to a view, and we/you should respect other's opinions, so please be aware 99.9% of people using these cranks don't thing they are crap. In your opinion they are crap, in the majority of opinions they are reliable.
Many Many Many people were and still are perfectly happy with Square Taper, especially those who went to the Hollow Light Church, returned with less in their wallet, to then be struck with a Lightning from the Hollow Light God, including me and the one of this topics creation.

You state here that I should respect other's opinion, as like I criticized someones choice, while the opposite is true, my experience and my findings were treated with no respect.

About your 99.9% Shiny Happy People from Rolling Eyes Movement, also mentioned earlier here in another fashion, 99.9(feel free to add 9's) % people do NOT die while riding the road, yet, the remainder that DOES die, IS used for rules, for the 100%.

Your last sentence then, it wasn't my "opinion", it was my "experience".
The only plus I experienced, also already said here and elsewhere on the forum, from Hollow Technology, was the chainring centering of Hollow Technology version Hollow Technology 2: on the bike used before my avatar bike, and used while the latter was on repair, the tension variation of the chain was near to nil (remember, case singlespeed). But its bearings failed after 5 years, and the Octalink version of Hollow Technology, failed for me due to that loosening crank, that inflicted me near accidents upon (re)discovery, and dozens if not hundred of stops along the road to check the -*--bolt.
There's only one word for such: CRAP, and I used it, Mister fossyant.

What OTHER people do, is THEIR trade-off choice, since THEIR wallet end THEIR risk.
MY trade-off Square Taper choice makes my bicycle gain 130 gram, and a square taper fail due to design as such I never had, the cranks of my Stronglight crankset broke out there pedal eye, just because there was too less alu around them, which very likely delivered the "light" in "Stronglight".
Also never had that before, so you can understand MY attitude to "light" wherever.

What's 130 gram, when the "Tarra" (that is, no luggage other than default) weight of my fixed gear avatar bicyle is E-bike grade 30 kilo, and I sometimes carry a loudspeaker of 20 kilo on my back, plus, 1 of 20 kilo on my rear rack, and, in the pocket of my T-shirt, an apple of 130 gram?
It would be ridiculous, for MY use of a bike, to join the Light Religion, let alone after the ugly experiences I had with it, in this usage.
 

Drago

Legendary Member
I'll agree, Octalink wasn't particularly great on the axel/crank interface, but if you fitted it correctly and to the correct torque, you'd be OK - ride it with loose or badly fitted cranks, then you are stuffed - that will happen with square taper - you'll knacker the crank first.
Octalink MTB, AKA V2, with the deeper splines was fine but, as you suggest, few bother to torque it correctly and that was the cause of most problems.
 

Dogtrousers

Lefty tighty. Get it righty.
There's only one solution. We need to get together and invent a new standard that
we can all agree on.

1769772026456.png


https://xkcd.com/927/
 
Last edited:
OP
OP
R

roubaixtuesday

self serving virtue signaller
No stopping him.

Why would anyone want to? Shares in popcorn companies are booming.
 
Location
Loch side.
All I know, didn't check if others here did, is that I asked earlier in this thread if the inner diameter of the tube could be decreased, as to have a thicker and thus more torsion/break resisting wall, since torsion, as you also stated, is a rather clear (the 45° angle) cause of the break.

Yes, the inner diameter can be decreased, but that would be the wrong thing to do. It will produce a minimal increase on strength and a maximum increase in weight. It really is a silly thing to do. I believe that Shimano has reached an excellent compromise between weight and strength whilst maintaining compatibility.
Because, for reason you also state here, the outer diameter would be a drastic change that wipes all compatibility with existing frames.
As I looked up on the internet (that some flag as a bad thing), in a techie doc, a formula was shown, that contained inner and outer diameters, to deliver the torsion resistence, which is rather logical, since any more material in the wall, can only increase the torsion resistence, not in a linear fashion ofcourse,
No, it is absolutely linear. Whereas increasing the diameter deliver exponential benefits to the power of 3.

The original question thus is, is a smaller inner diameter possible? It would ofcourse add again weight to the spindle (as I illustrated in earlier post here with a proposal of a 110 instead of 130 grammes overall (whole of crankset) "gain" relative to square taper), so for the Light Religion a partly but maybe acceptable setback, to trade for Safety First plus (chance of break under leg power becoming nil).
We've gone of this and asking it again won't change the answer. Weight is a very important indicator of component success and therefore company success. Also, the safety concern is minimal. Read the poem One Hoss Shay to understand what's going on here. Once you've made the spindle unbreakable, the pedal axle is next in line and so on. Strong enough is strong enough.

Can such smaller inner diameter void something else of the HT2 design?
There is a "bolt" that pushes (shifts) the left crank over the splines into place, that's all, that latter, the actual mechanical mounting is the clamp forcefully closed with the 2 sideways ("pinch") bolts.
That bolt has to be given thread, so "inside" the hollow, there must be some part with a threaded hole, and the outer diameter of that part might void a decreased (to get a thicker tube wall) diameter of the spindle/tube.
Also the other, drive side. The tube must be connected to the spider somehow, with something internally.
I think you're taking about the preload cap. Yes it has thread on the inside of the crank, but so what? That's not a weak point.

Yes the other side is connected to the spider, making it a one-piece? So what. A one-piece means one less assembly error and cheaper manufacture and elegance. It is fine, leave it alone.
If that is the case, then a thicker inner wall, even if Shimano had be willing, is not technically possible, out of question, rendering HollowTech 2 to a compatibility design at the end of what was possible.
An interesting question would be why they changed the 3 parts of HT2 version 1, to 2 parts in version 2. Was that solely to further reduce weight / making mount easier, or was there a problem with version 1, that they solved by permanent fixing of drive side to spindle.
I honestly don't understand the question. But I suspect it is moot. We've addressed all concerns and questions.
 

Keezx

Regular
Location
The Netherlands
@Yellow Saddle
It has been proven in several Dutch (Silva is from Belgium though) fora that's absolutely useless* to bring in something technical sound arguments against Silva, the more you argue, the more he bites himself in his fixed beliefs.
I won't do that anymore, I must have tried that in 300 posts in the recent past.

*There might be 1 exception, I think i'm responsable for the fact that he uses a 47 front chainring now...instead ot a 48x16 which turned out to be very unfavourable for the chain and chainring life...

BTW the Dutch discssions were actually much fun, how could any human being have so much trouble with, what is basically a transport bike..
The local bike shops must have had nightmares about him.....;)
 
Last edited:

fossyant

Ride It Like You Stole It!
Location
South Manchester
@Yellow Saddle
It has been proven in several Dutch (Silva is from Belgium though) fora that's absolutely useless* to bring in something technical sound arguments against Silva, the more you argue, the more he bites himself in his fixed beliefs.
I won't do that anymore, I must have tried that in +100 posts in the recent past.

*There might be 1 exception, I think i'm responsable for the fact that he uses a 47 front chainring now...instead ot a 48x16 which turned out to be very unfavourable for the chain and chainring life...

BTW the Dutch discssions were actually much fun, how could any human being have so much trouble with, what is basically a transport bike..
The local bike shops must have had nightmares about him.....;)

Ah, thanks. Yes I've seen his bike is transport and it breaks a lot. I suspect it's the user not looking after it, but if he's also prolific back home on forums........

I'll get my mate to log in that's an ex pro and lives out there these days. Does loads and loads of miles on hollowtech cranks.
 

silva

Über Member
Location
Belgium
Octalink MTB, AKA V2, with the deeper splines was fine but, as you suggest, few bother to torque it correctly and that was the cause of most problems.
Consider my Octalink V1 story:
- dealer mounted it - not in my view - I assume correct torque.
- 3 months later, bolt suddenly came loose.
- back to same dealer, he losened it, put grease on it, remounted it with a huge allen wrench with thick plastic grip, I assume correct torque.
- 3 weeks or so later (have to look it up to be sure) bolt again suddenly came loose.
- I tensioned it with a default allen key.
- Ride, check, if loose, retension, and so on.
- bolt didn't came loose anymore.
- 6 months later, after a downpour, a 10 cm pool water the entire road width.
- I slowly cycled through it.
- 1 km further I felt my left crank moving abit, and of course, knowing the past, I immediately realized what was happening, so stopped and therefore avoided possible disaster.
- I went again a couple days through aboves procedure, and again success, the bolt stopped loosening.

What does this experience indicate, and for me even prove: that the mount slips due to water acting as a "lubricant" over the splines and/or bolt thread, rendering the default, normally sufficient, press of the fit insufficient.

Now, above was Octalink 1, let's go to Octalink 2, with 9>10 and longer splines, question how would the outcome of the water lubrication differ?
One could compare it with slipping with shoes over a wet floor, how much in outcome would a shoe size 50 differ to a size 35? I'd rather say that longer splines slip even easier, and, if not the same, case longer threaded bolt length of V2 versus V1, same story. One would be given more time to react to a loosening bolt than in case V1, a bigger chance to notice the cranks sideways movement before it's too late, that is, spline damage and ultimately crank pushed off, the potential disaster scenario.
 

silva

Über Member
Location
Belgium
Yes, the inner diameter can be decreased, but that would be the wrong thing to do. It will produce a minimal increase on strength and a maximum increase in weight. It really is a silly thing to do. I believe that Shimano has reached an excellent compromise between weight and strength whilst maintaining compatibility.

No, it is absolutely linear. Whereas increasing the diameter deliver exponential benefits to the power of 3.


We've gone of this and asking it again won't change the answer. Weight is a very important indicator of component success and therefore company success. Also, the safety concern is minimal. Read the poem One Hoss Shay to understand what's going on here. Once you've made the spindle unbreakable, the pedal axle is next in line and so on. Strong enough is strong enough.


I think you're taking about the preload cap. Yes it has thread on the inside of the crank, but so what? That's not a weak point.

Yes the other side is connected to the spider, making it a one-piece? So what. A one-piece means one less assembly error and cheaper manufacture and elegance. It is fine, leave it alone.

I honestly don't understand the question. But I suspect it is moot. We've addressed all concerns and questions.
https://amesweb.info/Torsion/torsion-of-shaft-calculator.aspx
Filled in, for what it matters, just some rough bicycle alike values.

Torque [T] 500
Rotation speed [ω] 60 rpm
Shaft outer radius [c2] 28 mm
Shaft inner radius [c1] 26 mm
Shaft length [L] 100
Modulus of rigidity [G] 300 GPa
= 2 mm wall thickness
Results
Parameter Value
Maximum shear stress [τmax] 56.524
Angle of twist [θ] 0.001
Power requirement [P] 3141.593
Polar moment of inertia [J] 247683.165

Torque [T] 500
Rotation speed [ω] 60 rpm
Shaft outer radius [c2] 28 mm
Shaft inner radius [c1] 25 mm
Shaft length [L] 100
Modulus of rigidity [G] 300 GPa
= 3 mm wall thickness
Maximum shear stress [τmax] 39.783
Angle of twist [θ] 0
Power requirement [P] 3141.593
Polar moment of inertia [J] 351907.072

Maximum shear stress difference is 56.524-39.783 = 16.741
That's 30% less (what you call "minimal") stress for a 50% thicker (what you call "maximal") wall.

Formulas are more accurate than minimal, maximum, and Believe.
 
Last edited:
Top Bottom