CPSC judgement on disk brakes and QR forks - Page 3

"David Damerell" <damerell@chiark.greenend.org.uk> wrote in message
news:CfD*gm+0p@news.chiark.greenend.org.uk...
> Spider <beelzebubba@hotmail.com> wrote:
> >How was the 0.6g decelleration figure arrived at? (This is one I've had a problem with from the
> >beginning - I don't know where to get a g-meter for my mtb. I would be willing to experiment if I
> >had one.)
>
> As was explained to you at the time, the maximum deceleration attainable without lifting the rear
> wheel on the flat is directly determined by the angle between the horizontal and a line drawn
> through the forward contact patch and the centre of gravity of the bike+rider.
> --
> David Damerell <damerell@chiark.greenend.org.uk> Kill the tomato!

Is that assuming that you don't shift your weight to the rear: sitting on the saddle? or is it
measured when your butt is behind the saddle, arms extended? Inquiring minds want to know...

Mike

In article <MPG.19b913a45ae348cd9896f6@netnews.comcast.net>, Super Slinky
<slinky@newmail.net> wrote:

> I was always willing to admit that James Annan had an interesting point, but it was the hype, the
> ego and the exaggeration that made me want to jump in and throw some obstacles in front of the
> band wagon, for whatever good it did.

No good at all, in fact attitudes like yours do harm instead.

It's very odd, unlike you and several others I saw no ego or hype or exaggeration in Annan's posts.
I initially disagreed with him- indeed, I was the first responder and told him his idea was all wet-
believing like many that all failures of the QR to hold the wheel in the dropout are the result of
user error. After all, that's what "everyone knows" is the case with QRs.

However, the situation with disk brake forks is different than the situation with other types of
wheels, in that the brake puts an ejection force on the wheel that is absent in all other designs.
That this force is asymmetrical adds another dimension to the problem, and that the force occurs in
suspension forks that have a degree of independence between the movement of the two legs further
adds to the problem.

> Less scare-mongering and arrogance might have actually resulted in a better reception by the world
> cycling community. Instead we get a guy trying to set himself up as the Ralph Nader of bicycles.
> How could any rational person not give a resounding Bronx cheer to that?

Oh bull****. This is a problem of your interpretation, not a problem with what Annan published here
and on his Web site. The rational response would have been to examine the analysis and the available
evidence, which in turn would have brought you to the conclusion that Annan has a valid point.
Further independent testing and examination has verified- in a preliminary fashion at least- both
the ejection force and the loosening of QRs.

> Fortunately, the bicycling community is smaller and more sophisticated than the auto industry and
> that turkey won't fly in this little world.

Only if you think that being hidebound and superstitious are the same as being "more sophisticated."
Because that's the situation with bicyclists- we have an extensive lore based on little more than
supersition and mistaken attribution. The turkey files just fine in this little world, but few will
open their eyes to see the fact.

In article <1fsukv40odj5jc2h07pbf283q5vsq25hma@4ax.com>, Doug Taylor
<taylordw@choiceonemail.com> wrote:

> I get it. It's a conspiracy! The government wants to hide this huge problem with disc brakes and
> quick releases, which is a dire threat to the health and freedom of all cyclists, so that Big
> Business (the bike industry? Hah) will thrive. But little Timmie McNamara knows the TRUTH. He
> heard it from James Annan, who crashed on his tandem, and Jobst Brandt, who knows everything. So
> it must be true. Forget the fact that if it were true, it would be in the interest of the CPSC and
> the cycling industry to nip it in the bud, BEFORE the tort lawyers blow the whole thing to holy
> hell. And that only a tiny minority of cyclists have been injured by this problem, IF it's a
> problem and IF the problem caused the injury. --dt

Ah, Dougie, you've been listening to too much Rush again. These paranoid theories of yours- and
Rush's as well- are treatable, after all. Consider looking into it. ;-)

In article <PrL3b.32217$cj1.11028@fed1read06>, "Mike S." <mikeshaw2@coxDOTnet> wrote:

> <jobst.brandt@stanfordalumni.org> wrote in message
> news:MhL3b.16451$dk4.541184@typhoon.sonic.net...
>
> > OK. How about a simple experiment, assuming you have access to a disk brake equipped bicycle.
> > Since retention ridges are only on the outside of the fork dropout, remove the QR skewer and
> > perform a static braking test by pushing the bicycle forward with the brake applied. You could
> > also do this at low speed without hazard. Then report whether this disengages the wheel or not.
> >
> > This is intended to demonstrate that there is a separating force, something that has been argued
> > here at length.
>
> Yeah, and I'll bet that if you did the same with cantis or V-brakes, it'd do the same thing. Since
> I haven't tried it, I don't know. Anyone willing to try?

Umm, it's been done. It's been written up in this newsgroup.

> I've got an F3000 out in the garage that I can experiment with in a bit. I'll let you know what
> happens.

Please do. Having checked this out personally, I can tell you what you will find. The wheel will
stay in the dropouts with caliper brakes, cantilever brakes and V brakes. The wheel will be pushed
out of the dropout with disk brakes.

That there is a ejection force created by disk brakes is not in dispute, although you seem to
continue to think that it is. This has been verified by informal testing by participants in this
newsgroup, and has been verified by independent testing within the bike industry. The reference for
this has been posted several times, and it is entertaining to note that Annan's critics continue to
be silent about this.

The alleged hype and ego are now on the side of the people trying to argue against observable
reality. At this point in the discussion, sheer obstructionism or lack of simple comprehension seem
to be the only explanations.

. Further independent testing and examination
> has verified- in a preliminary fashion at least- both the ejection force and the loosening of QRs.
>
When was the last time a QR came undone spontaneously? Snagged on trees, bushes, bashed on rocks,
yes. Loosen completely on its own? Hmmm... don't think so.

I just did the test that Jobst recommended and indeed, my disc equipped wheel migrated back and down
when there was no clamping pressure applied by the QR. However, once there was pressure applied by
the QR, this tendency was overcome.

The only way I can see my front wheel ejecting itself spontaneously from my forks is if the QR is
either installed incorrectly, breaks, or there's a massive side load that pulls the wheel partially
out of the dropout first. In the first instance, it is still user error. The second is equipment
failure which would be possible to be caught in time with proper maintenance. The third would
probably mean that you have to stop riding and fix your wheel before continuing, allowing you to
tighten an otherwise loose QR.

I can see how using Ti QRs would be asking for trouble when running discs. I can also see that
filing off the "lawyer lips" on a disc brake equipped bike is not a good thing either, if just for
the additional insurance "just in case."

I can also see why I wouldn't want my disc caliper on the front of my fork. That much delicate stuff
right in the way of trees, rocks, etc. What the "answer" is, I haven't a clue.

Questions that just popped into my mind: Exactly how much force does it take to eject the wheel from
a properly tightened QR/fork interface? How much force is a "typical" QR exerting? Exactly where is
the tendency of the wheel to come out of the bottom of the fork equal to the clamping force of the
QR? How much force is the disc brake exerting on the QR/fork interface? How deep would a dropout
have to be to keep a wheel in the fork? Maybe a "J" shaped dropout would work?

Mike

> > I've got an F3000 out in the garage that I can experiment with in a bit. I'll let you know what
> > happens.
>
> Please do. Having checked this out personally, I can tell you what you will find. The wheel will
> stay in the dropouts with caliper brakes, cantilever brakes and V brakes. The wheel will be pushed
> out of the dropout with disk brakes.
>
> That there is a ejection force created by disk brakes is not in dispute, although you seem to
> continue to think that it is. This has been verified by informal testing by participants in this
> newsgroup, and has been verified by independent testing within the bike industry. The reference
> for this has been posted several times, and it is entertaining to note that Annan's critics
> continue to be silent about this.
>
> The alleged hype and ego are now on the side of the people trying to argue against observable
> reality. At this point in the discussion, sheer obstructionism or lack of simple comprehension
> seem to be the only explanations.

I have never argued that there ISN'T a force ejecting the wheel. Just that there has to be something
else going on to make a wheel spontaneously jump out of a perfectly good dropout.

There I was, JRA, when BAM! my front wheel flew out of my dropout... Things like that don't just
happen. Something's gotta be wrong before it happens. What that is: user error, malfunction, who
knows. But to happen spontaneously? Ummm, yeah...

Mike

Mike Shaw writes:

> I just did the test that Jobst recommended and indeed, my disc equipped wheel migrated back and
> down when there was no clamping pressure applied by the QR. However, once there was pressure
> applied by the QR, this tendency was overcome.

How much pressure? Wellllll, if I brake harder, it will take even more pressure than on my bicycle
with conventional fork and rim brakes. How much? This variable setting defines the safe bicycle in
this respect. I am glad my safety has not been defined by how tight I make my QR. Besides, as has
been discussed here often in this forum, most front-wheel bearings fail from QR preload. The
over-center feel of a QR comes from stretch in the skewer and compression of the hollow axle. That
compression usually overloads the previously correctly adjusted bearings.

> The only way I can see my front wheel ejecting itself spontaneously from my forks is if the QR is
> either installed incorrectly, breaks, or there's a massive side load that pulls the wheel
> partially out of the dropout first.

That side load on the QR does not occur because a wheel will collapse from a load great enough to
extract the wheel with a freshly adjusted QR. Wheels are amazingly weak to side loads but because
the bicycle is a balanced single track vehicle (only light lateral loads) they do not suffer this
failure. This is true even during a broadslide. A wheel separating side load would throw the rider
to the ground... or occur as a result of crashing.

However, the massive load you mention is furnished by the disc brake. The separation force arises
from the ratio of tire rolling radius divided by brake disc radius minus (the axle load) one. Since
this ratio for common disc brakes, lies between 3 and 4, they can cause a force (on one side of the
axle) of as much as 500lbf. This can occur with hard braking on hard surface with the rear wheel at
lift-off for a combined rider and bicycle weight of 200lbs.

> In the first instance, it is still user error. The second is equipment failure which would be
> possible to be caught in time with proper maintenance. The third would probably mean that you
> have to stop riding and fix your wheel before continuing, allowing you to tighten an otherwise
> loose QR.

So in your opinion, is this reasonable condition for a bicycle sold to the public at large?

> I can see how using Ti QRs would be asking for trouble when running discs. I can also see that
> filing off the "lawyer lips" on a disc brake equipped bike is not a good thing either, if just for
> the additional insurance "just in case."

What difference should QR material make. Both are intended to secure the wheel. The way you say
that, it seems there is a thin line between a safe QR closure and an unsafe one.

> I can also see why I wouldn't want my disc caliper on the front of my fork. That much delicate
> stuff right in the way of trees, rocks, etc. What the "answer" is, I haven't a clue.

Oh you rascal. You ride under tough conditions exceeding imagination. You probably drive an SUV with
chromed jungle grills over the front with bumper bars to help you get to the real rough stuff to
ride bike. How well do you think your fork blades sustain running into rocks. This sounds like so
much Rambo. Where's Sylvester Stallone when we need him?

> Questions that just popped into my mind:

> Exactly how much force does it take to eject the wheel from a properly tightened QR/fork
> interface? How much force is a "typical" QR exerting? Exactly where is the tendency of the wheel
> to come out of the bottom of the fork equal to the clamping force of the QR? How much force is the
> disc brake exerting on the QR/fork interface? How deep would a dropout have to be to keep a wheel
> in the fork? Maybe a "J" shaped dropout would work?

You say that as if I and others have not given that any thought. I guess you just began reading
about this because the suggestions you offer have all been analyzed here. The different shaped
dropout has also been shown to not solve the problem.

Jobst Brandt jobst.brandt@stanfordalumni.org Palo Alto CA

Mike S. wrote:
> . Further independent testing and examination
>
>>has verified- in a preliminary fashion at least- both the ejection force and the loosening of QRs.
>>
>
> When was the last time a QR came undone spontaneously? Snagged on trees, bushes, bashed on rocks,
> yes. Loosen completely on its own? Hmmm... don't think so.

I received this by email recently, from a shop owner who was setting up a bike for a customer:

"Well, just in the parking lot, the disc will demonstrably begin to pull the wheel out of the drop
out. A few on-off cycles of braking and the nut can be seen backing off, if you keep it up, the QR
is loose enough to clear the lawyer tabs in just a minute of this action, mind you just in the
parking lot."

I guess that particular parking lot has more than its fair share of rocks and bushes to snag
the lever on.

Ben Cooper is another shop owner who has done similar tests and found a correctly fastened skewer
loosening repeatedly under very light use. If you massively overtighten the skewer, as many MTBers
have learnt by experience to do, you can reduce the likelihood of this occurring but there is no way
of knowing in advance how tight is tight enough, nor how much this overtensioning increases the
likelihood of snapping the skewer, another `operator error' that can have very serious consequences.

James

David Damerell <damerell@chiark.greenend.org.uk> wrote in message
news:<CfD*gm+0p@news.chiark.greenend.org.uk>...
> Spider <beelzebubba@hotmail.com> wrote:
> >How was the 0.6g decelleration figure arrived at? (This is one I've had a problem with from the
> >beginning - I don't know where to get a g-meter for my mtb. I would be willing to experiment if I
> >had one.)
>
> As was explained to you at the time

Why is this bit of condescension necessary? I just don't get why you folks think that this makes
your case more solid.

> the maximum deceleration attainable without lifting the rear wheel on the flat is directly
> determined by the angle between the horizontal and a line drawn through the forward contact patch
> and the centre of gravity of the bike+rider.

Yes, it's a nice theory. Too bad that's not the way it works in application.

1.) What is the coeffcient of friction of the two materials, acting together (travelling surface
and tire)?

2.) Since when does braking only happen on level ground? (This could *increase* the number, as the
CG may change.) In fact, braking most often happens on a slope. The number needs to be
recalculated.

3.) Most riders change their CG during braking, especially on a hill. An endo at speed is not
indicated for happy riding, so, the rider will shift his/her posterior back to change the weight
distribution. This also changes the calculation.

4.) On dirt, I have NEVER only used the front brake. The rear is always part of the equation. How
much is being used? I dont' know, and neither does Mr. Annan. This changes the calculation - by
some unknown amount.

Now, if someone could list the mu of knobby tire on dirt (sand, mud, etc) then we could see some
numbers that are not at the most extreme end of the spectrum, but something that a little closer to
what really happens.

Spider

"Mike S." <mikeshaw2@coxDOTnet> wrote in message news:<PrL3b.32217$cj1.11028@fed1read06>...
> <jobst.brandt@stanfordalumni.org> wrote in message
> news:MhL3b.16451$dk4.541184@typhoon.sonic.net...

> > This is intended to demonstrate that there is a separating force, something that has been argued
> > here at length.
>
> Yeah, and I'll bet that if you did the same with cantis or V-brakes, it'd do the same thing. Since
> I haven't tried it, I don't know. Anyone willing to try?
>
> I've got an F3000 out in the garage that I can experiment with in a bit. I'll let you know what
> happens.

From James' website, the free-body diagram shows clearly that there is an ejection force. It's quite
simple. The magnitude might not be accurate (due to an optimistic decel figure of 0.6g) but the
force exists. A lot of the questions you ask here have been answered previously, and especially on
James' web site.

Spider

"James Annan" <still_the_same_me@hotmail.com> wrote in message news:3F4FC7D3.80008@hotmail.com...
> Mike S. wrote:
> > . Further independent testing and examination
> >
> >>has verified- in a preliminary fashion at least- both the ejection force and the loosening
> >>of QRs.
> >>
> >
> > When was the last time a QR came undone spontaneously? Snagged on
trees,
> > bushes, bashed on rocks, yes. Loosen completely on its own? Hmmm...
don't
> > think so.
>
> I received this by email recently, from a shop owner who was setting up a bike for a customer:
>
> "Well, just in the parking lot, the disc will demonstrably begin to pull the wheel out of the drop
> out. A few on-off cycles of braking and the nut can be seen backing off, if you keep it up, the QR
> is loose enough to clear the lawyer tabs in just a minute of this action, mind you just in the
> parking lot."
>
> I guess that particular parking lot has more than its fair share of rocks and bushes to snag the
> lever on.
>
> Ben Cooper is another shop owner who has done similar tests and found a correctly fastened skewer
> loosening repeatedly under very light use. If you massively overtighten the skewer, as many MTBers
> have learnt by experience to do, you can reduce the likelihood of this occurring but there is no
> way of knowing in advance how tight is tight enough, nor how much this overtensioning increases
> the likelihood of snapping the skewer, another `operator error' that can have very serious
> consequences.
>
> James
>
Now, if I were to go to any of the shops here in San Diego, grab a bike off the rack, roll it out,
and test it, chances are I'll find that the QR probably isn't going to move. Just like people can
quote the Bible to say pretty well whatever they want, a few "tests" being quoted doesn't a
scientific paper make.

In the years that I've been riding discs, I have not one time, I say again, NOT ONE TIME, have had a
QR either come loose, or the wheel move in the dropouts. Am I extremely cautious? Nope. Am I riding
DH? No to that one too. Am I a competent to tighten a QR without supervision? I guess so.

I don't dispute that there's something going on, but it seems to me that the current setup is
adequate for the majority of riders. That there are a few failures out there doesn't surprise me.
Given a big enough statistical sample, there's going to be a failure in anything.

So, to answer Jobst here, yes I believe that the current setup is adequate. Is it perfect? God no.
Could the interface be better designed? Probably. Do I think it will be? Probably not any time soon.

So, all you that believe that there's an inherent problem in the design, what is the proposition for
fixing the situation? Remember that this design has to work with conventional wheels.

Mike

In article <sQN3b.32234$cj1.10026@fed1read06>, "Mike S." <mikeshaw2@coxDOTnet> wrote:

> . Further independent testing and examination
> > has verified- in a preliminary fashion at least- both the ejection force and the loosening
> > of QRs.
> >
> When was the last time a QR came undone spontaneously? Snagged on trees, bushes, bashed on rocks,
> yes. Loosen completely on its own? Hmmm... don't think so.

Again, like the rest of the Annan-bashers, you've chosen to avoid reality. Loosening QR with front
disk brakes has been demonstrated in an article published in the bike industry press in response
to Annan's Web site. I've posted an URL where you can read the article on three occasions in the
past 8 weeks.

> I just did the test that Jobst recommended and indeed, my disc equipped wheel migrated back and
> down when there was no clamping pressure applied by the QR. However, once there was pressure
> applied by the QR, this tendency was overcome.

Overcome in terms of visible movement for the time being, you mean.

> Questions that just popped into my mind: Exactly how much force does it take to eject the wheel
> from a properly tightened QR/fork interface? How much force is a "typical" QR exerting? Exactly
> where is the tendency of the wheel to come out of the bottom of the fork equal to the clamping
> force of the QR? How much force is the disc brake exerting on the QR/fork interface?

Those things have already been estimated, Mike, and discussed ad nauseum. Where ya been?

"Tim McNamara" <timmcn@bitstream.net> wrote in message
news:timmcn-A9E9D4.18353129082003@gemini.visi.com...
> In article <sQN3b.32234$cj1.10026@fed1read06>, "Mike S." <mikeshaw2@coxDOTnet> wrote:
>
> > . Further independent testing and examination
> > > has verified- in a preliminary fashion at least- both the ejection force and the loosening of
> > > QRs.
> > >
> > When was the last time a QR came undone spontaneously? Snagged on trees, bushes, bashed on
> > rocks, yes. Loosen completely on its own? Hmmm... don't think so.
>
> Again, like the rest of the Annan-bashers, you've chosen to avoid reality. Loosening QR with front
> disk brakes has been demonstrated in an article published in the bike industry press in response
> to Annan's Web site. I've posted an URL where you can read the article on three occasions in the
> past 8 weeks.
>
> > I just did the test that Jobst recommended and indeed, my disc equipped wheel migrated back and
> > down when there was no clamping pressure applied by the QR. However, once there was pressure
> > applied by the QR, this tendency was overcome.
>
> Overcome in terms of visible movement for the time being, you mean.
>
> > Questions that just popped into my mind: Exactly how much force does it take to eject the wheel
> > from a properly tightened QR/fork interface? How much force is a "typical" QR exerting? Exactly
> > where is the tendency of the wheel to come out of the bottom of the fork equal to the clamping
> > force of the QR? How much force is the disc brake exerting on the QR/fork interface?
>
> Those things have already been estimated, Mike, and discussed ad nauseum. Where ya been?

Watching "the disc brake/QR is faulty" crowd bash it out with the "there's no problem crowd" and
laughing at everyone. I fall firmly into the "there's no real problem" crowd till proven
conclusively that there's a significant occurrence of failures. This is especially true since I
actually mtn bike on a bike with discs, unlike some of the "its the design" crowd.

I just went out and examined my dropouts. You'd think with the prevalence of the problem we're
talking about, that there'd be scores where the axle/QR was moving around. Nothing other than the
normal imprint from attaching the wheel (in one spot) and there's no scoring on the lawyer lips from
the wheel partially ejected. I don't overtighten my QR, so please explain to me why there's a
problem again...

Mike

Doug Taylor <taylordw@choiceonemail.com> writes:

> Forget the fact that if it were true, it would be in the interest of the CPSC and the cycling
> industry to nip it in the bud, BEFORE the tort lawyers blow the whole thing to holy hell.

Like Ford nipped the Explorer/Pinto problems in the bud?

> And that only a tiny minority of cyclists have been injured by this problem, IF it's a problem and
> IF the problem caused the injury.

What percentage of said Ford owners had problems? A tiny minority, perhaps? How is this different?

Joe Riel

Tim McNamara <timmcn@bitstream.net> wrote in message
news:<timmcn-B51625.13413929082003@gemini.visi.com>...
> In article <PrL3b.32217$cj1.11028@fed1read06>, "Mike S." <mikeshaw2@coxDOTnet> wrote:
>
> > <jobst.brandt@stanfordalumni.org> wrote in message
> > news:MhL3b.16451$dk4.541184@typhoon.sonic.net...
> >
> > > OK. How about a simple experiment, assuming you have access to a disk brake equipped bicycle.
> > > Since retention ridges are only on the outside of the fork dropout, remove the QR skewer and
> > > perform a static braking test by pushing the bicycle forward with the brake applied. You could
> > > also do this at low speed without hazard. Then report whether this disengages the wheel or
> > > not.
> > >
> > > This is intended to demonstrate that there is a separating force, something that has been
> > > argued here at length.
> >
> > Yeah, and I'll bet that if you did the same with cantis or V-brakes, it'd do the same thing.
> > Since I haven't tried it, I don't know. Anyone willing to try?
>
> Umm, it's been done. It's been written up in this newsgroup.
>
> > I've got an F3000 out in the garage that I can experiment with in a bit. I'll let you know what
> > happens.
>
> Please do. Having checked this out personally, I can tell you what you will find. The wheel will
> stay in the dropouts with caliper brakes, cantilever brakes and V brakes. The wheel will be pushed
> out of the dropout with disk brakes.

Yes, it will. What is the magnitude of the risk of this design compromise? ATM, unknown. Too many
independent parameters to judge at this point.

> This has been verified by informal testing by participants in this newsgroup, and has been
> verified by independent testing within the bike industry.

If you mean by the German group, no, that is not yet in the realm of "verification." Even they say
that more experiments need to be done, as per the article in STW.

> The reference for this has been posted several times, and it is entertaining to note that Annan's
> critics continue to be silent about this.

Wrong again. I pointed out that it was merely preliminary data, and not conclusive.

> The alleged hype and ego are now on the side of the people trying to argue against observable
> reality.

Here's where you folks get into trouble. "Observable reality" is often at odds with what is really
going on. That is why optical illusions exist. Controlled study is the final arbiter, not claims in
a newsgroup, nor self-selected, unknown-condition samples to a website.

> At this point in the discussion, sheer obstructionism or lack of simple comprehension seem to be
> the only explanations.

Another condescending dismissal. It reminds me of folks who are pro-war with regards to Iraq. When
they hear someone has questions, or is hesitant, they level their gaze and say "So, why do you
support Saddam?" or "you're unpatriotic."

Solid hypotheses and theories will withstand pointed questions. In fact, pointed questions can serve
to bolster them. Those familiar with science understand this.

If you can't answer the questions, say "I don't know." If you know the answer, post it or the link.
If you find the questions tiresome, don't bother responding. You (and the others) do yourself a
DISservice by being defensive and obtuse.

Spider