Carlton Reid on QR safety

[[email protected]]

James Thomson writes:

>> I don't see where that proves the thread on these bolts is rolled.
>> The ones I have for inspection do not have the classic bolt end,
>> have a thread whose major diameter is exactly the same as the
>> bearing surface of the shaft

> [snip]

>> The problem with that is the thread and shaft are the same diameter
>> within less than 0.01mm, something that is not readily repeatable
>> with a rolled thread.

> I've just measured the threads and shafts of the three caliper
> brakes I have lying around. The figures are each the mean of five
> readings, shaft first, then thread, then the diameter of the shaft
> adjacent to the root of the thread:

> Shimano 6208: 5.90mm , 5.83mm , 5.28mm

> SunTour Superbe Pro: 5.96mm , 5.76mm, 5.38mm

> and here's the interesting one:

> Shimano 1050: 5.80mm, 5.90mm, 5.37mm

> In other words, the diameter of the thread of the 1050 is greater
> than the maximum diameter of the plain section of the shaft.

I have only standard reach Campagnolo Record brake bolts from earlier
times at hand, and they measure 5.87mm on both thread and shaft on the
brake caliper side and 5.90mm on thread and shaft on the bicycle frame
side. These bolts have lathe centers on both sawed off ends.

From what has been offered on this subject I suspect that these brake
bolts could have rolled threads. Their thick shiny chrome finish
hiding any fine telltale features that might make a cut thread
obvious. Screw machines can be made to leave transitions that do not
reveal thier final cut, however, the lathe centers seem to imply a
grind finish (in the days when these were made).

Jobst Brandt

 

[Jay Beattie]

"Tim McNamara" <[email protected]> wrote in message
news:[email protected]...
> "Marvin" <[email protected]> writes:
>
> > But you haven't actually *given* a magnitude for this force.
All
> > you've done is show a relationship to another force (which
you can't
> > measure either). If you can put a number to any of these
forces,
> > that would indeed be a step forward.
>
> Those numbers have already been given, they are available on
James
> Annan's Web page. Numbers from Cannondale have also been
given:
>
>
http://www.ne.jp/asahi/julesandjames/home/disk_and_quick_release/
>
> Generally the forces have been given in Newtons, however, which
is
> confusing to the non-engineers and non-physicists. It's an
> appropriate unit to use, though, since this problem is largely
about
> Newton's law (for every action...). ;-) Let me try using more
common
> units of measurement. I am sure that my math and my
assumptions will
> be adequately scrutinized by both sides of the discussion.
>
<big snip>

> I hope that helps make the picture clearer, Marvin. If you use
disk
> brakes, use a good skewer such as a Shimano MTB as jim beam has
> pointed out in the past. Also as jim pointed out, use hubs
with
> serrated locknuts to "bite" into the inner faces of the
dropouts. And
> clamp those babies down tight. I'd carry a second skewer, BTW,
> because we've had reports of skewers breaking. Use your rear
brake a
> lot since it doesn't have this problem- it pushes the axle into
the
> dropout instead of trying to push it out, because the rear
caliper is
> mounted in front of the axle. Stop every so often to check on
the
> tightness of the skewer. And when the problem is fixed by
eliminating
> the ejection force or eliminating the open dropout, upgrade!

Other confounding factors include the design of the drop out and
the location of the disc caliper. I have an Avid mechanical disc
which has a caliper mounted at a fairly high angle in
relationship to the drop out. The drop out also has downward
facing, "U" shaped opening rather than the usual swept-back
opening on a traditional road dropout. In combination with thick
lawyer's lips, I think the chance of QR failure and wheel
ejection is rather remote. I still check my QRs, though because
I always have (except before that one race . . . ) -- Jay
Beattie.

 

[Marvin]

jim beam wrote:
> Marvin wrote:
> > jim beam wrote:
> >
> >>James Annan wrote:
> >>
> >>>jim beam wrote:
> >>>
> >>>
> >>>
> >>>>"interference between materials"??? "dead end"? so how come /i/ am
> >>>>the poor dolt that had to point out that axle faces are serrated and
> >>>>that subsequent indentation significantly increases retention force?
> >>>>i can't see you bothering to point out such trivial details tim.
> >>>>
> >>>
> >>>All I need to do is point to the pictures that jim beam so kindly
> >>>provided:
> >>>
> >>>http://home.comcast.net/~carlfogel/download/nishiki_before.jpeg
> >>>http://home.comcast.net/~carlfogel/download/nishiki_indent.jpeg
> >>>
> >>>Of course, a particularly dull-witted person might not wonder why it is
> >>>that someone (who?) bothered to scrape the paint off the dropout of an
> >>>old fork that happened to be lying around, nor what the picture might
> >>>have looked like had the fork end been left normally painted. But I
> >>>think most people reading this thread will have got the message clearly
> >>>enough.
> >>
> >>eh? that's spectacular b.s. that is a fork i happened to have laying
> >>about. the paint was already off. the snide implication that i somehow
> >>"prepared" it to alter the result is as bogus as it is desperate. it
> >>also shows incredible ignorance if you think a simple paint layer offers
> >>any significant resistance to steel indentors.

So how come the paint was off? More to the point, how come the fork
didn't have any indents before, if forks indent even through paint?

I'd hold the above photo up as a shining example of a fork that's been
used repeatedly, then had the paint scraped off to reveal pristine
metal.

> > Yes, it does. I did the experiments, I reported the results. Google
> > for them if you can't remember.
> >
> >
> >>i can see b.s. may be the only way you could have any come-back on the
> >>evidence annan, but resorting to such a pathetic diversion is below even
> >>you. integrity and credibility go hand in hand, in case you never
> >>learned it before. [why did you leave scotland again?]
> >>
> >>
> >>>Of course, for those who are still unconvinced, there is also Marvin's
> >>>comment posted earlier:
> >>>
> >>>"Every new bike I assembled today, I checked before and after on the
> >>>indentations. All of them embossed the paint to a fairly minor degree
> >>>with a single clamping, repeated clampings on one test subject made
> >>>them a little more obvious. None of them were to the same level as jim
> >>>beam's example above."
> >>>
> >>>
> >>>James
> >>
> >>straw clutch city. that was from a single thumb-pressure application.
> >>what now - suggest i have some kind of digital deformity and super-human
> >>strength so the result can be "ignored"? do the math on the indentation
> >>force. no, wait, /google/ for the math on the indentation force - it's
> >>been done for you. but i warn you, you'll be straying into retention
> >>force territory annan - you probably won't want to go there.
> >
> >
> > Oh, pull your head out of your ****, Beam. As I said at the time
> > (google for the post if you want), that was repeated on several
> > different bikes, one of which I checked repeatedly to see if this
> > embossing got more obvious over time. I don't think I reported it, but
> > I took one skewer and ramped it as tight as my padded hands could cope
> > with (far tighter than any recommendation) - still didn't emboss
> > through the virgin paint to the metal. These aren't exactly super
> > thick tough paint jobs we're talking about here, either.
>
> with respect, marvin, you seem to be implying that somehow the paint
> only indents and the metal is unscathed.

Yes, that's exactly what I'm suggesting. I'm suggesting that the metal
compresses the much softer paint in the same way as you'd expect two
hard surfaces to do to a soft one, and that neither metal surface is
affected. The serrations emboss the paint, and not the metal
underneath.

Try an experiment: take one flat piece of metal, one flat piece of
wood, and clamp them both in a vice (with jagged edges). See if the
vice indents the metal through the wood. Same principle.

And before you ask, I've checked the depth of the skewer serrations
(0.5mm) and the depth of average paint (0.7-1.0mm, measured from paint
flakes off various bikes).

> maybe you should remove the
> paint with solvent /after/ you've done your indent test.
> then you'll
> see that the metal does indeed deform under the paint.
> > So my actual, experimental evidence suggests that on new bikes you
> > don't have a metal to metal interface no matter what you do to the QR.
> > This substantially lowers the required pullout force.
>
> that's like saying that because you have an oil film separating the
> teeth when meshing in a gear box, that torque is not transmitted! i
> don't think you understand what you're looking at.

> > Now if you've got a larger sample size, I'd love to hear it.
>
> google. i've posted my own, thanks.

I don't see you having repeated any experiments ten or more times on a
variety of different bikes. Do tell me what to google for to find
them.

> > Until
> > then I think I'm one of the only people who's even attempted actual,
> > unbiased experiments on this topic - ironically enough, something I
> > recall Beam complaining loudly about the lack of. Funny how he
> > complains even louder when things don't go his way.
>
> eh? that makes no sense. your misinterpretation of results is not my
> problem.

 

[Ian Blake]

On Tue, 14 Feb 2006 06:18:09 -0800, jim beam <[email protected]> wrote:

>Ian Blake wrote:
>> On Mon, 13 Feb 2006 20:28:17 -0800, jim beam <[email protected]>
>> wrote:
>>
>>>2. caliper ahead of the fork has been abandoned by engineers that
>>>fortunately know more about materials than you do! your comprehension
>>>of fatigue is abysmal jobst.
>>
>>
>> So, there is never enough force from braking to move the wheel held by
>> friction alone in the dropouts but if the caliper is mounted ahead the
>> same forces will destroy the bike? You can not have your cake and eat
>> it.
>
>the cake is called "fatigue". check it out some time.

Lets see a new and increased fatigue level for forward mounted brakes. Lets
see, What can you mean? Perhaps you mean the caliper. It make no difference
to the caliper whether it is pulled left or right. Perhaps you mean the bolts
attaching the caliper to the fork. Again these are pulled the same wherever
the brake is attached. So that means you expect the fork to fail.

In my case I do not expect so. Argos[1] built me a Reynolds 525 steel fork
with a reinforced leg where the front facing brake is attached. I am not
expecting a failure anytime soon. The original fork with no reinforcement
failed where the rear facing brake was attached. So I have an example of a
possible fatigue failure of a fork with rear facing disc brake. Now produce an
example with a front facing brake. I can see no reason why a properly built
fork will fail from fatigue whichever way it is attached.

[1] Argos is a very good frame builder in Bristol UK. Their expertise is
mostly in road going bikes but a few customers request mountain bikes.

 

[Tim McNamara]

jim beam <[email protected]> writes:

> Tim McNamara wrote:
>> jim beam <[email protected]> writes:
>>
>>>James Annan wrote:
>>>
>>>>Ed Pirrero wrote:
>>>>
>>>>
>>>>>James Annan wrote:
>>>>>
>>>>How would you know what a "real scientist" would think? Have you
>>>>ever met one? James
>>>
>>>he is one annan. he's all over the web if you know who you're
>>>looking for.
>>
>> He's all over Usenet, anyway, but that's only proof of an Internet
>> connection. I find nothing by him that is scientific or scholarly,
>> certainly not in this newsgroup. Is he just another sock puppet of
>> yours, perhaps?
>
> the chickens taking up too much of the intellectual resources tim?
> "ed" is not his real name. there, that was hard to figure out,
> wasn't it.

Then we only have your word and his, and neither of you is credible.

 

[Tim McNamara]

Luke <[email protected]> writes:

> In article <[email protected]>, Tim McNamara
> <[email protected]> wrote:
>
>> "Marvin" <[email protected]> writes:
>>
>> > But you haven't actually *given* a magnitude for this force. All
>> > you've done is show a relationship to another force (which you
>> > can't measure either). If you can put a number to any of these
>> > forces, that would indeed be a step forward.
>>
>> Those numbers have already been given, they are available on James
>> Annan's Web page. Numbers from Cannondale have also been given:
>>
> <snip>
>
> A clear and objective assessment of the hazard. Thank you.

You're welcome. As I said, I am sure the numbers will be scrutinized
and errors will be pointed out. But at least it was a starting point
for (hopefully) more reasoned discussion.

 

[Tim McNamara]

Mike Causer <[email protected]> writes:

> On Mon, 13 Feb 2006 18:38:37 -0600, Tim McNamara wrote:
>
>
>> every action...). ;-) Let me try using more common units of
>> measurement. I am sure that my math and my assumptions will be
>> adequately scrutinized by both sides of the discussion.
>
> Excellent summary Tim.

Thanks.

 

[Tim McNamara]

"Jay Beattie" <[email protected]> writes:

> "Tim McNamara" <[email protected]> wrote...
>> "Marvin" <[email protected]> writes:
>>
>> > But you haven't actually *given* a magnitude for this force. All
>> > you've done is show a relationship to another force (which you
>> > can't measure either). If you can put a number to any of these
>> > forces, that would indeed be a step forward.
>>
>> Those numbers have already been given, they are available on James
>> Annan's Web page. Numbers from Cannondale have also been given:
>>
> http://www.ne.jp/asahi/julesandjames/home/disk_and_quick_release/
>>
>> Generally the forces have been given in Newtons, however, which is
>> confusing to the non-engineers and non-physicists. It's an
>> appropriate unit to use, though, since this problem is largely
>> about Newton's law (for every action...). ;-) Let me try using
>> more common units of measurement. I am sure that my math and my
>> assumptions will be adequately scrutinized by both sides of the
>> discussion.
>>
> <big snip>
>
>> I hope that helps make the picture clearer, Marvin. If you use
>> disk brakes, use a good skewer such as a Shimano MTB as jim beam
>> has pointed out in the past. Also as jim pointed out, use hubs
>> with serrated locknuts to "bite" into the inner faces of the
>> dropouts. And clamp those babies down tight. I'd carry a second
>> skewer, BTW, because we've had reports of skewers breaking. Use
>> your rear brake a lot since it doesn't have this problem- it pushes
>> the axle into the dropout instead of trying to push it out, because
>> the rear caliper is mounted in front of the axle. Stop every so
>> often to check on the tightness of the skewer. And when the
>> problem is fixed by eliminating the ejection force or eliminating
>> the open dropout, upgrade!
>
> Other confounding factors include the design of the drop out and
> the location of the disc caliper. I have an Avid mechanical disc
> which has a caliper mounted at a fairly high angle in
> relationship to the drop out. The drop out also has downward
> facing, "U" shaped opening rather than the usual swept-back
> opening on a traditional road dropout.

Yes, one undeclared assumption- for the sake of simplicity- was that
the direction of the ejection force was parallel to the exit of the
dropout. In the real world, there are variations in the location of
the caliper and the exit direction of the dropout.

 

[Johnny Sunset aka Tom Sherman]

Boyle M. Owl wrote:
> ...
> Following your [Jobst Brandt's] example.
>
> I would not even pretend to be an expert in physics, but I would not
> expect a physicist who has not been in my field to claim to be an expert
> in machining or toolmaking, either.

When did Jobst Brandt metamorphose from a mechanical engineer to a
physicist?

--
Tom Sherman

 

[Johnny Sunset aka Tom Sherman]

Boyle M. Owl wrote:
> ...
> I, myself, don't see the big deal about QR issue. If you're
> experiencing pullout, replace the hollow axels with solid ones and nuts....
^^^^^^

What does changing figure skating maneuvers have to do with the
discussion on the potential for disc brakes to cause front wheel
ejection?

My bicycle and trike with disc brakes have single-side mount axles
(sic), so wheel ejection is not a significant concern.

--
Tom Sherman

 

[Benjamin Lewis]

Tim McNamara wrote:

> "Ed Pirrero" <[email protected]> writes:
>> Logical fallacy - calling for the proof of a negative.
>
> Good, at least you have some basic grasp of logic. I was starting to
> wonder.

Not really. "Proof of a negative" is quite possible. Or perhaps our sock
puppet has proof that you can't prove a negative?

--
Benjamin Lewis

Now is the time for all good men to come to.
-- Walt Kelly

 

[Benjamin Lewis]

Tim McNamara wrote:

> jim beam <[email protected]> writes:
>> 1. ejection incidence is effectively zero and cannot be distinguished
>> from operator error.
>
> That does not mean that there are no incidents. Since we all have
> been brought up on the notion that all wheel ejections are the fault
> of operator error, it is likely that riders will attribute any wheel
> ejections to operator error.

Yes, and also end-overs caused when the axle slips but the wheel is
retained by lawyer lips. Of course, because I can't *prove* that this has
ever occurred in practice, it must not have.

--
Benjamin Lewis

Now is the time for all good men to come to.
-- Walt Kelly

 

[Benjamin Lewis]

Boyle M. Owl wrote:

> I, myself, don't see the big deal about QR issue. If you're experiencing
> pullout, replace the hollow axels with solid ones and nuts. Problem
> solved. However, I can sit here, close my eyes, and be confounded at how
> wheel pullout can happen if the skewers are properly tightened in the
> first place. The braking force has to overcome the clamping force of the
> properly tightened skewer _and_ make the skewer jump over the lawyer
> lips. For that to happen, you have to make the skewer _stretch_ by making
> it exceed its yeild point.

As Jobst pointed out earlier, axle slipping can cause end-over even if the
wheel is retained by lawyer lips, due to the sudden non-linearity in
braking response when the disk cants sideways in the caliper.

--
Benjamin Lewis

Now is the time for all good men to come to.
-- Walt Kelly

 

[Ed Pirrero]

Tim McNamara wrote:
> jim beam <[email protected]> writes:
>
> > Tim McNamara wrote:
> >> jim beam <[email protected]> writes:
> >>
> >>>James Annan wrote:
> >>>
> >>>>Ed Pirrero wrote:
> >>>>
> >>>>
> >>>>>James Annan wrote:
> >>>>>
> >>>>How would you know what a "real scientist" would think? Have you
> >>>>ever met one? James
> >>>
> >>>he is one annan. he's all over the web if you know who you're
> >>>looking for.
> >>
> >> He's all over Usenet, anyway, but that's only proof of an Internet
> >> connection. I find nothing by him that is scientific or scholarly,
> >> certainly not in this newsgroup. Is he just another sock puppet of
> >> yours, perhaps?
> >
> > the chickens taking up too much of the intellectual resources tim?
> > "ed" is not his real name. there, that was hard to figure out,
> > wasn't it.
>
> Then we only have your word and his, and neither of you is credible.

You making a judgement on scientific credibility is laughable. Here's
a clue - the social "sciences" really aren't much about science. But
hey, if it makes *you* feel more important, that's fine by me.

Hell, I think they even call economics "science" nowadays. ROTFL.

E.P.

 

[James Annan]

Ed Pirrero wrote:
> James Annan wrote:
> >
> > I'm just asking you to clarify what you mean by a "third party", and
> > what you might consider "verification".
>
> What difference does my opinion in this matter make?

Since you are the one asking for "verification" by a "third party" it
seems worthwhile for me to check what you mean, so you don't invent
spurious grounds for rejection.

If you can't specify what you mean by either of these terms, it seems
like a pretty meaningless request. If you won't specify what you mean,
it seems like an obviously dishonest one.

James

 

[Ed Pirrero]

[email protected] wrote:
> Ed Pirrero wrote:
> >
> >
> > There is a force. Whether the force is sufficient to cause anything
> > but consternation in engineering purists *has not yet been proven*.
>
> Rather, "has not been proven" to _your_ satisfaction (or jim beam's).

When there are actual experiments (as opposed to drawings and
calculations) I will consider those numbers to be the final word.

> There will always
> be people so welded to their own view that no amount of evidence and
> logic will suffice.

A lovely strawman. Knock it down!


> > > There is a known mechanism for loosening
> > > threaded fasteners, and there is no reason to presume that that
> > > mechanism cannot apply to bicycles.
> >
> > Another presumption. It *might*. But *does it*? We don't know,
> > because nobody has actually done any controlled testing.
>
> Our of curiosity, Ed, why do you think that well known mechanism for
> loosening of industrial threaded fasteners would _not_ apply to
> fasteners with far less locking power and far greater transverse loads?

Who says they don't?

Oh, that's right, another strawman.

But hey, *maybe*, since all the other fasteners on a bike seem to hold
without the two anti-loosening strategies used commonly in QRs, it
seems incredibly odd that MTBs don't routinely fall to pieces.

Or maybe, just maybe, not everything is known about the system. It
could be that James is 100% correct. I don't discount that possiblity,
because I understand how hypotheses work. But it sure would be nice to
see SOMETHING other than supposition and conjecture.

> Can you identify any specific mechanical reason, or is it just another
> "you can't prove it" thing?

The onus is not on me to disprove James' hypothesis. You know that
just as well as I. Your query is yet another rhetorical device in a
long line of folks too frightened to gather any sort of data
whatsoever.

Where's the testing? So far, only Cannondale has bothered. And their
test seems to not support James' hypothesis. "It's not a fair test!"
James moans. OK, that's fine - but in the absence of your own data
collection, that's all there is. And I consider data over hypothesis
every day.

E.P.

 

[Ed Pirrero]

James Annan wrote:
> Ed Pirrero wrote:
> > James Annan wrote:
> > >
> > > I'm just asking you to clarify what you mean by a "third party", and
> > > what you might consider "verification".
> >
> > What difference does my opinion in this matter make?
>
> Since you are the one asking for "verification" by a "third party" it
> seems worthwhile for me to check what you mean, so you don't invent
> spurious grounds for rejection.

Nonsense. If you had any sort of confirmation, you'd have presented it
ages ago. Even the most questionable of third parties get quoted by
you immediately when they say anything that resembles support of your
hypothesis.

Does anyone other than yourself, besides some lawyer, claim on the
record that the accidents *you claim* were caused by disk brakes were
actually caused that way?

No? Then all your fishing for rhetorical advantage comes to nothing.
It's easy to reject third-party confirmation when none actually exists.

You got something, or not? If not, then knock it off. If so, let's
see it. Your weaselling is tiresome.

E.P.

 

[Ed Pirrero]

James Annan wrote:
> Ed Pirrero wrote:
>
> >
> > "Many" is not quantitative. Nor is it broken down by type or quality
> > of data.
> >
> > Some guy said his QR slipped. OK.
>
> Do you think this sugggests a problem or is worthy of any
> investigation?

It may suggest user error, or it may suggest a deeper issue. It's hard
to tell without "further investigation." Drawing a diagram and doing
some simplified calculation is not the furthest that such investigation
can be taken.

Unless you are James Annan, I guess.

E.P.

 

[James Annan]

Ed Pirrero wrote:
> [email protected] wrote:
> > Ed Pirrero wrote:
> > >
> > >
> > > There is a force. Whether the force is sufficient to cause anything
> > > but consternation in engineering purists *has not yet been proven*.
> >
> > Rather, "has not been proven" to _your_ satisfaction (or jim beam's).
>
> When there are actual experiments (as opposed to drawings and
> calculations) I will consider those numbers to be the final word.

But:

Ben Cooper's experiments don't count
Velotech.de's experiments don't count.
Dave Gray and Brant Richards' own personal experiences don't count
Journalists describing repeated skewer loosening on a new,
manufacturer-supplied test bike don't count

What actually would count?

James

 

[James Annan]

Ed Pirrero wrote:
> James Annan wrote:
> > Ed Pirrero wrote:
> >
> > >
> > > "Many" is not quantitative. Nor is it broken down by type or quality
> > > of data.
> > >
> > > Some guy said his QR slipped. OK.
> >
> > Do you think this sugggests a problem or is worthy of any
> > investigation?
>
> It may suggest user error, or it may suggest a deeper issue. It's hard
> to tell without "further investigation."

Do you think the manufacturer, when informed of the problem, should
undertake this "further investigation"?

James