11

I was toying about with a disk brake for the first time and managed to shear two bolts clean off while reassembling.

Curiously, both tore the head off, leaving the threadded shaft in the tapped hole in the hub. I was able to remove the shaft using bull-nosed pliers both times.

Looking closely at the remains, the head has torn off where it fastens to the shaft. Is this a weak point by design?

No, I was not using a torque wrench, because I don't own one. This is high on my list of wants though.

enter image description here

enter image description here

  • 3
    Always! use a torque wrench on these possibly live-endangering bolts and if possible new bolts as well! – Carel Feb 2 '17 at 11:32
  • 3
    Did you consider to measure the torque with a pocket scales at the exepense of another bolt? – Klaster_1 Feb 2 '17 at 11:33
  • 2
    Especially in this case you ought to look into corrosion though. Stress crack corrosion might have weakened the bolt leading to failure now. I find it a bit unusual that you as an experienced bike mechanic would overtighten so much to shear off the head of a good bolt. However post overtightening might have lead to cracks and the exposed position might then have speed up stress corrosion cracking. Has it been driven in winter (salt!)? Are those bolts out of stainless or mild steel? – gschenk Feb 2 '17 at 13:02
  • 3
    How can you not have a torque wrench? What do you use when you tighten crank bolts? – Daniel R Hicks Feb 2 '17 at 13:43
  • 2
    Torque wrenches are $25 usd. No reason not to have one. – RoboKaren Feb 2 '17 at 16:48
2

To directly address the question in the title:

There is no reason for a rotor to be designed to shear off, and no point in having the individual bolts be able to shear off more readily than the materials used would normally allow. One might argue that something might get caught in the holes of the disk, causing the wheel to lock, but there's always the danger of something getting caught in the spokes or a bearing locking up, or some such, and no provision is made to "protect" from these situations.

And it should be noted that the mode of failure of the bolts would not allow the disk to spin free, since even with the heads broken off the bolt bodies would still prevent the disk from rotating free. If anything, the disk would pop outward on one edge and make the brakes be more likely to seize.

The bolt failure here is not "by design" (unless someone is purposely sabotaging their parts to increase replacement parts sales).

14

Low-profile button-cap socket screws do have a design weakness. The button reduces the amount of material around the head and the recessed hex or Torx socket also does the same, exactly at the interface of the screw body and head. In this zoomed in portion of your larger photo, you can see just how little material connects the head to the rest of the screw:

Zoomed in portion of head

You have to be careful with following the torque recommendations with low-profile heads, especially if you're using a torx bit in a regular socket wrench. As a side note, this is a great guide for bike component torque specs. Rotor bolts shouldn't be overtorqued -- and need to be torqued to the same amount -- to prevent the rotor being warped.

Two things you can do:

  • Torque wrenches are cheap and you should have two: one low-range (<11 Nm) and one high range (20-6 0Nm). Beam-type torque wrenches are more accurate than uncalibrated, cheap clicker-type torque wrenches. In terms of the low range, many automotive torque wrenches don't read low enough for bike componentry, especially carbon frames. Brake rotors are supposed to be torqued only to 6.2 Nm for Avid, 2-4 Nm for Shimano. This is a cheap US$25 wrench that goes from 2 to 10 Nm (88 in/lb, you have to do the imperial correction manually). It's great for the smaller components on bicycles or for carbon frames. I used it for my own recent disc brake upgrade.

    Torque wrench

  • If you can find a Torx or hex screw where the threads don't go up all the way to the head, it's a bit stronger. But you have to be careful that this doesn't limit the depth of material that is being retained. Disc brake rotors I believe are 1.6~1.8mm thick when new so you should only have about 1.25~1.5 mm of unthreaded shank. This one is too long:

    Bolt

Note that you may not be at fault. The drunken lemur who initially assembled your hub and rotor might have applied too much torque at the factory, fatally weakening the screw. So when you redid the brakes, you were working with a weakened component.

sheared torx

As an addendum, even though button cap screw heads are inherently weak, the zoomed-in photo at the top and here seems a bit extreme in just how little material there is -- the torx socket seems recessed even further than normal. So don't rule out a manufacturing defect in the screws themselves... although one also shouldn't rule out that it may be a design feature and not a bug -- designed to prevent you from overtorquing the rotor bolts and warping the rotor.

  • 4
    What do you have against lemurs? They are actually quite an interesting animal! – Rider_X Feb 2 '17 at 20:29
  • 5
    Great animals but they can't assemble bikes for caca. – RoboKaren Feb 2 '17 at 20:38
  • 4
    can you quote a source? – Rider_X Feb 2 '17 at 20:39
  • 2
    Exhibit A and I rest my case: youtu.be/3iYc_3fWbUo – RoboKaren Feb 2 '17 at 20:53
  • 2
    @RoboKaren Case seems proven: the lemurs in that video didn't manage to assemble a single bicycle correctly! – David Richerby Feb 3 '17 at 13:54
8

Disc brake rotors with the six bolt design are attached with M5 fasteners. To shear off the head of a new M5 bolt requires a considerable torque. The design torque for medium resistance bolt is typically 5 Nm. It ought to take at least twice that torque before it fails so abruptly. A torque of 10 Nm is roughly equivalent to the weight force of 10 kg applied at a 10 cm lever. This would require quite considerable effort when using a common hand tool (eg a L-shaped X25 hexalobular-key).

Since this happened at used bolts I should rather suspect a different reason for failure. For example, Stress corrosion cracking (SCC) or corrosion fatigue (CF).

"[SCC] is the growth of crack formation in a corrosive environment. It can lead to unexpected sudden failure of normally ductile metals subjected to a tensile stress, especially at elevated temperature." [1]

"[CF] is fatigue in a corrosive environment. It is the mechanical degradation of a material under the joint action of corrosion and cyclic loading." [2]

Soft steels (ie ductile, austenitic steel) are often at risk to SCC when subject to chlorine. (Wooden truss roofs over swimmingpools are notorious for collapsing because of bolts failing due to SCC.)

While the environment in this case is not particularly corrosive, previous over-tightening of the bolts might considerably increase the susceptibility of the material to SCC. A corrosive source may be salt, which is very common on roads in high concentrations in cold climates.

Several other causes of fastener failure are discussed in an article by Hudgins et al [[3]]. It also shows and explains the surface where the fastener broke off. You might compare your fasteners with the pictures to identify typical structures associated with different failure modes. In all cases you ought to be able to see the ductile overload happened, ie the last bit holding to the head.

tl;dr A form of corrosion might have damaged the bolts such that they rupture easily. This is more likely to happen when tightening fasteners with too much torque.

trivia There's an old German proverb that fits here: "Erst fest dann ab!". An approximate translation is "First its tight, then broken".

[3]: A Hudgins, B James, FASM, Avanced materials & processes, August 2014, 18-22 http://www.asminternational.org/documents/10192/20564188/amp17208p18.pdf/5cddb014-2b4a-40a1-b0f6-07b58906754d

  • 4
    It should be noted that the fasteners, for some reason, are generally designed for a "Torx" wrench. With such a wrench one can apply much more torque to a fastener than with a standard hex "Allen" wrench. This could easily lead someone to believe that the fasteners must really be highly torqued, even though there's no need to tighten them any more than "normal" fasteners on a bike. – Daniel R Hicks Feb 2 '17 at 20:00
  • 1
    @DanielRHicks oh! I've overlooked it on the photo. Hexalobular fasteners are much easier to handle in machines and serial mass production and reduces tooling wear quite a bit. Bike shops usually prefer hexalobular as well. Textron's patents on Torx expired recently and hexalobular became an ISO standard. We may see these fasteners much more often, hopefully they can push those horrid Phillips finally out of the market. Cheers for pointing it out, I shall change it accordingly. – gschenk Feb 2 '17 at 20:16
  • 1
    +1. This is the technical explanation behind drunken lemurs. Lemurs may look weak but are actually quite strong for their size. – RoboKaren Feb 2 '17 at 20:40
  • 1
    Ps I've sheared off M5 bolts with my own monkey wrenching. 10nm doesn't feel like a lot of torque if you are using a socket wrench. – RoboKaren Feb 2 '17 at 20:55
  • 2
    I don't really buy the corrosion explanation. Looking at the pictures, there is no evidence of significant corrosion on any other components. – Daniel R Hicks Feb 4 '17 at 3:24
4

I own a road bike with disc brakes, have had them on and off countless times over the past few years, and have never had any type of problems with the bolts shearing.

The bolt heads should not shear off easily, unless you are applying a crazy excessive amount of torque. For M5 rotor bolts, Park recommends:

Shimano - 18-35 in-lb, 2.0-4.0 Nm

Avid - 55 in-lb, 6.2 Nm

Magura - 35 in-lb, 4.0 Nm

Hayes - 50 in-lb, 5.6 Nm


I also do not own a torque wrench either. I think I have a pretty good feel for the amount of torque I'm putting on small bolts, like rotor bolts, handlebar, seat post, etc. I torque the heck out of the rotor bolts, and have never come close to damaging one. I think you may have some worn/fatigued/defective bolts and I'd recommend picking up some new ones. You can often get them for free from your LBS if you ask nicely.

  • 1
    For reference: 55 in lb correspond to about 6 Nm. (ps: oughtn't it be lbf for poundforce, otherwise it wouldn't be torque but moment of inertia? no expert on non-SI though ...) – gschenk Feb 2 '17 at 20:22
  • 9
    I hate imperial. – RoboKaren Feb 2 '17 at 20:55
  • 2
    @RoboKaren Oi! Leave the Empire alone. Those are American customary units. (The liquid measures are different, as are tons.) – David Richerby Feb 3 '17 at 14:01
  • @RoboKaren the comments are for question clarification, not a place for you to complain. Everyone says that. It's arbitrary. – ebrohman Feb 3 '17 at 14:50
  • 3
    Imperial refers to the American Empire now. :) – RoboKaren Feb 3 '17 at 16:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.