Index
Home
About
Blog
Newsgroups: rec.bicycles.tech
From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: Crank Tester
Date: Mon, 18 Dec 1995 18:28:39 GMT
Eric W. Bryant writes:
> I'm talking about breaking the pedal out, but this is certainly not due to
> pedaling forces. I know one rider up here who has sent back 2 pairs of LX
> and a pair of XT Shimano cranks due to breakage. Seeing that this person
> spends more time in the air than on the ground and his bike is about a
> year old, it's not due to fatigue from pedaling. I'm very sure that the
> pedal eye just happens to be the weak spot in this case.
I think you are unclear on the subject of failures. If it was a
forced rupture it would have broken the first time he made such a
jump; either that or he would have bent the crank. Therefore, it was
a fatigue failure because there was no deformation, showing that the
material, as a whole, did not approach yield. Cranks are ductile and I
have inspected enough of them after being run over by cars to know
that they bend appreciably when taken to yield. They do this without
breaking.
> Whether or not I land with my knees locked, large forces are going to be
> transmitted. You're confusing my argument with others, as I never mentioned
> locking my knees (I see this as a good way to destroy some cartilage). I just
> feel that the knee can't move fast enough to absorb 100% of the shock, and
> therefore some (if not most) of the landing is going to be absorbed by the
> bike.
It's not the knee that does the main shock absorption but the ankle.
The knee is important to your body but the ankle is the absorber that
takes out the peak force that might damage a crank. After that, its
all gravy for the crank.
> BTW, I ride a 15.5" mountain bike and often lower my seat in technical
> terrain. I have plenty of clearance to land one-legged. My BMX has a 10"
> frame - definitely enough room there.
I understand, but that still doesn't allow huge forces. It will
accelerate pedal eye breakout no doubt. Aluminum has some
undesirable fatigue characteristics that lead to more failures than
might be expected from the stress levels (with respect to the yield
stress).
> It might also interest you that a discussion is taking place on the
> trials mailing list about broken cranks on Crestone trials bikes.
> So, just what kind of fatigue is causing these cranks to break? I
> don't recall a lot of sprinting, much less pedaling, occurring in
> your typical trials section. Any of the pedaling that does occur is
> often done at the 3 and 9 o'clock positions, as the cranks often
> ground at the 6 o'clock position.
I haven't had the opportunity of inspecting such failures. I know of
no trials riders in my area and I don't know whose cranks these riders
are using. I am definitely not a fan of any of the NC machined cranks
offered these days. I have been breaking cranks regularly for many
years and about six or eight years ago switched to the old Dura Ace
cranks that have held well for years. All my Campagnolo Record cranks
failed in a year or less.
Jobst Brandt <jbrandt@hpl.hp.com>
Newsgroups: rec.bicycles.tech
From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: Crank Tester
Date: Tue, 19 Dec 1995 17:01:21 GMT
Lines: 55
Drew W Saunders writes:
> Let me see if I follow this:
> Current pedal design: (left part is pedal body, right part is
> threaded part of axle)
> ____
> |___
> ___|
> ____|
> The way it ought to be:
> ____
> \___
> ___|
> ____/
> I'm guessing the beveled part would be only a few mm's long?
It would be just as big as the current spindle diameter (at 45
degrees) is at the contact face of mainline pedals today.
> This doesn't look to be a particularly difficult retrofit. Just
> make the cranks accordingly and provide several pairs of the
> threaded rings with each pair of cranks sold. Any idea why
> nobody's tried it?
Because no one has recognized pedal eye failures as having any
relation to this interface. I'm not even sure how to model this in a
structural analysis program because the contact face is hard to define.
> How difficult would it be to cut existing cranks and then add the
> ring, or would that likely damage the cranks?
It would require machining the rings and countersinking the crank. Most
people who look at this concept reject it because the fear the loss of
two thread pitches will weaken the joint. In fact those thread
pitches are doing nothing currently. Only when a crank is new, do
these threads bear any load but that goes away by fretting damage.
> I have cracked cranks at the pedal eye, so I have some interest in
> significantly reducing the possibility of this happening again, as
> it can be a rather unpleasant experience. I'm guessing that most
> folks just don't put cranks under that kind of stress, so this
> would be a hard sell requiring way too much explaining on the part
> of the shop staff or manufacturer.
It is a marginal situation as is nearly everything on a bicycle but it
is a serious one, and one that is easily improved. When a left crank
fails adjacent to moving traffic it is most likely fatal because the
crank breaks primarily when standing and when standing at failure it
always causes the rider to fall to that side... under the wheels of
moving vehicles.
Bar stems and seat posts break also, yet the most fragile and lightest
ones receive the most attention in the press and cost the most.
Jobst Brandt <jbrandt@hpl.hp.com>
Newsgroups: rec.bicycles.tech
From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: stripped crank threads at pedal?
Date: Fri, 8 Mar 1996 00:51:43 GMT
Mark Hickey writes:
> You couldn't have possibly damaged the threads "forcing the pedals
> off", unless they had corroded to the crank (in which case it was
> inevitable anyway). You probably hosed up the threads by
> cross-threading them when you put them *on*.
I disagree with this assessment. I have removed pedals by heating the
aluminum crank to sizzle when touched with the wet finger with no
damage to the threads after pedal wrenches had previously been damaged
by removal attempts. I am certain that light corrosion in the
interface of a tight pedal is enough to make the pedal unremovable
(without heat) without damaging the crank.
> Your best option at this point is to make sure they're good and
> tight, and that they are properly aligned with the crank (tighten
> down flush) and never, ever take them off again. If they make any
> funny noises or show the least sign of looseness, retire the crank
> (and possibly the pedals, though the much tougher steel threads
> shouldn't be hurt by the aluminum crank).
Either you can get them tight without stripping the thread or you
can't. It doesn't take much 9/16-20 thread to get an amazingly strong
hold.
> If you do find it necessary to take the pedals off, you may be able
> to chase out the threads with a tap, or perhaps even a screw of the
> same size. Of course, you'll only be able to use a standard tap or
> screw on the *right* crank since the left one is reverse threaded.
Pedal taps are available at any reasonable bicycle shop. Besides,
there are Helicoil inserts for cranks that some shops carry from the
old days when the throw-away way of business wasn't fully established.
> You should start the tap or bolt from the back of the crank where
> the threads are untouched by all the abuse.
That's not a bad idea but I hope the threads are in better shape than that.
Jobst Brandt <jbrandt@hpl.hp.com>
Newsgroups: rec.bicycles.tech
From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: rim cracks - advice needed
Date: Tue, 9 Apr 1996 15:40:47 GMT
Alex D Rodriguez writes:
> Does this mean that if I weigh less than 170 lbs I should not be
> worried about breaking cranks? Will my Dura-Ace cranks have an
> infinite life span for me? I know cranks are most likely to break
> where the pedal threads in, but where is a BB most likely to break?
> Will there be visible cracks before the part fails?
As I said, people over 170 should be aware that cranks fail, not that
people below that or any other specific weight are immune to such
failures. Cranks develop cracks most often at the web between the
right crank and the spider finger trailing the crank in rotation.
These cracks often cease growing after progressing a few millimeters
while others advance to crank failure. Cracks at the pedal threads
seem always to lead to failure. Both of these areas are easily
inspected for cracks if the crank is wiped clean.
Crank spindle failure, in contrast, cannot easily be detected because
the break almost always occurs inside the press fit of the left crank
for the reasons explained in the FAQ. The crack develops in the
middle of the flat about two millimeters inside the press fit and the
crack does not open because the spindle is hardened steel and does not
flex much before failure.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: crank problem
Date: 1 Dec 1996 03:29:17 GMT
Ed Chait writes:
> Why would greasing the pedal threads and remounting them have any
> effect on the cyclic "click" I was experiencing? I can see how if
> the pedal bearings were not adjusted correctly, that could produce
> the problem, but I don't understand how grease on the pedal threads
> could make a difference. The pedals were mounted nice and tight.
If you inspect a pedal crank that has been ridden for a while, you
will find that the shoulder of the steel pedal has gouged the face
of the crank by fretting motions. These are microscopic movements
of the pedal in its threads that occur because there is no radial
retention of the pedal except by friction. If this is a dry joint,
the aluminum oxide can make clicking noises. Titanium does this even
more so. Lubricant does not prevent this action but it silences it.
> I read the crank arm mounting FAQ, which was written by Jobst and I
> had some trouble understanding some of the points made. It seemed
> as if the gist of the article was that if the cranks are properly
> mounted and torqued initially, there should be no need for
> subsequent adjustments.
That is correct. The reason you question this is that the most common
advice is to tighten the crank repeatedly because the subsequent
looseness of the screw is misinterpreted. If you tighten repeatedly,
you will split the crank and be admonished not to use grease, when in
fact repeated tightening caused the failure, not the grease.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: crank problem
Date: 2 Dec 1996 17:04:34 GMT
Ed Chait writes:
>> If you inspect a pedal crank that has been ridden for a while, you
>> will find that the shoulder of the steel pedal has gouged the face
>> of the crank by fretting motions. These are microscopic movements
>> of the pedal in its threads that occur because there is no radial
>> retention of the pedal except by friction. If this is a dry joint,
>> the aluminum oxide can make clicking noises. Titanium does this even
>> more so. Lubricant does not prevent this action but it silences it.
> Thanks for the clarification Jobst. I appreciate your help. I
> checked the cranks on my bikes and I see the gouging that you are
> describing. I'm going to remove, grease, and remount all my pedals.
> Just out of curiosity; would placing a hard nylon washer between the
> pedal and crank arm prevent the gouging?
No. The forces at this interface are such that even steel cranks
become damaged from the fretting action. I have tried to get any
manufacturer to make a 45 degree taper joint for years now without
success. This could even be a retrofit simply by making a split steel
ring tapered washer for a tapered crank face. New pedals would have
tapered faces to engage the crank.
>> That is correct. The reason you question this is that the most
>> common advice is to tighten the crank repeatedly because the
>> subsequent looseness of the screw is misinterpreted. If you
>> tighten repeatedly, you will split the crank and be admonished not
>> to use grease, when in fact repeated tightening caused the failure,
>> not the grease.
> Then what would be the correct procedure to follow if someone
> notices that there is any looseness in the cranks? Are they ruined
> if any degree of looseness exists?
Yes, because they worked themselves loose and that process destroyed
the fit. No manner of tightening can again extrude the shape of the
bore to what it should be.
> If the cranks are removed at the first sign of looseness and
> reinstalled with grease, is there a chance that they would still be
> serviceable?
Too late! You can try it and think you've got it but it will work loose
again with time and each time it gets worse.
> Most of what I have read in regards to crank maintenance has advised
> that normal use will produce some "seating in" of the cranks on the
> spindle which will cause looseness, especially right after the
> cranks have been first mounted, and that the rider should re-tighten
> them after a certain number of miles. From our discussion here, I
> would gather that this in incorrect advice and that properly mounted
> crank arms should never loosen.
I think if you re-read the item about crank installation, this aspect
is explained clearly. If it doesn't make sense to you, it is probably
because it is bad news for your cranks and you trying to find a way
around buying new ones.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Frame lifespan
Date: 4 Sep 1997 21:19:28 GMT
Charles Maurer writes:
>>> I have ridden bikes plenty long, and in that time have had more
>>> than two dozen Campagnolo Record cranks break....
>> Jobst, you can't stop here. How on earth have you broken two dozen
>> cranks?
I ride a lot and I weigh about 180 lbs. I ride mostly in the
mountains around here in the SF Peninsula cost range, the Sierra and
the Alps. See:
http://www-math.science.unitn.it/Bike/Countries/Switzerland/
These cranks broke off mainly at the pedal eye but also at the spider
on right cranks and neck of the crank on both sides. Many of these
were discovered before separation but cracked about half way through.
Crank spindles are more dangerous because they cannot be inspected
without disassembly, something that you shouldn't do any too often.
The early Campagnolo aluminum cranks failed within two thousand miles
and I switched back to cottered steel cranks in the 1960's In the
1970's I switched back to Campagnolo because steel crank spindles,
cups, and cotters were no longer available. I was not amused with the
durability of aluminum cranks, then and now.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: Don't disassemble BBs (was Re: Frame lifespan)
Date: 08 Sep 1997
Newsgroups: rec.bicycles.tech
Carl Kuck writes:
>> Crank spindles are more dangerous because they cannot be inspected
>> without disassembly, something that you shouldn't do any too often.
> Why is this? I would think that any greased part on a bike should
> be pulled, cleaned, and re-greased on a regular basis. Are there
> hidden fatigue costs associated with dis/reassembly?
> Tell us more!!!
The crank is pressed on with fairly high stress and invariably
scrapes off a bit of metal with every installation. Because the crank
is completely without lubrication after use, regardless of whether it
lubricated for installation, the extraction is always a metal smearing
process in the crank. This may not be much but repeated at monthly
maintenance would probably add up to a crank that pressed on farther
that one would like.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: greased cranks slip farther up the tapers
Date: 18 Sep 1997 18:50:55 GMT
Damon Rinards writes:
> I never wrote that the crank was grease free. From my original post:
> "So I made some measurements of how far a crank slips up the taper
> on a bottom bracket spindle when the tapers are greased compared
> to when the grease is wiped off."
The term "wiped off" is not specific enough to draw any conclusion.
If the rag was greasy, as it usually is when you wipe grease, the
spindle was not without grease and the bore was just as greasy as
before. It is this part of the experiment that makes it questionable.
The effective grease film in such an interface is less than the
roughness of the spindle surface. Grease above and beyond that
thickness has no effect on the press fit. Therefore, I am unclear on
what affected your results other than unrepeatability.
> I just wiped them off, that's all. I have no intention of mounting
> cranks only after "de-greasing" them with solvents; as my measurements
> show, just a thorough wipe with a rag makes a measurable difference.
I think that is where doubt arises. You don't mention any special
care used to clean the bore. How you wiped the surface is not clear.
It seems that using clean paper wipes in a single pass, several times
would constitute a credible degree of grease removal, but that doesn't
work so well in the bore.
>> screw,
> I never said the screw was grease free. In fact, I specifically said I
> greased it: "METHOD... I greased the bolt's threads and washer."
>> and spindle
> I never said the spindle was grease free, either. As I originally
> wrote, I just wiped it off.
> I did not wish to compare a greased interface to an artificial
> "surgically cleaned" one; I compared an ordinary, wiped off
> interface to a greased one, and said so in my post. I don't know of
> anyone who recommends solvents and degreasers to be used before
> mounting cranks. My measurements show that wiping off the grease is
> all it takes to make what appears to me to be a significant
> difference in how far the crank slides up the taper.
>> If not, then I don't believe they were unlubricated.
> I never said the parts were supposed to be *un*lubricated, just
> "wiped off" as is commonly done.
In light of the instructions to install the cranks on clean dry
spindles and not to use lubricant, your experiment did not test this
condition. It seemed to me, and apparently to other readers, that you
were proving or disproving the prescribed method. Apparently you
were not.
>> Wiping with a rag does not constitute removing grease.
> Wiping with a rag is all I did. I removed all the grease that wiping
> with a rag removes, and recorded the results.
As I said, it depends on the method, and from the precision with which
you describe the process, I take it the process was similarly casual
and therefore removed the bulk of excess grease but not the effective
film that is critical to a taper fit. Therefore, it is hard to
reconcile the method with your results.
> One wonders if you read my original post carefully or not.
You seem to be unaware that you have a different visualization of what
you did than those who read your sparse description. There is room
within that description for vastly different conditions.
>> 2. How did you measure the depth of engagement?
> Dial calipers.
>> 3. When you say "slipped on 1.8mm", from where and how was this
>> measured and how?
> I machined a flat reference surface on the inner face of an old crank.
> I mounted said crank in my bench vise.
> I mounted a bottom bracket spindle to the arm.
> I mounted each test crank one at a time to the other end of the spindle.
> I measured the distance from the test arm to the reference arm, inside to
> inside.
> I measured this distance when the crank was finger tight.
> I also measured this distance after tightening the bolt to 300 in*lb.
> The "slipped on" distance is the first measurement minus the second.
I was unaware that this was not on a bicycle and the spindle was in
free space to be accessed. In that event, I find it unusual that you
used the other crank as a reference, there being a flat surface on the
vise or fixture to which you could perform the measure. This is getting
curiouser and curiouser.
>> The only valid measure of position of the crank on the spindle is with
>> a depth micrometer from the outside face of the crank to the end of
>> the spindle.
> I didn't know there was only one valid way to measure such things.
> Have you measured many cranks this way? The distance I measured
> is the distance the crank slips up the taper. The distance your
> method yields is the same slip distance, minus any retraction the
> crank makes back down the tapers since you must take the bolt out.
That would be if you were doing this on your bicycle, where measuring
the gap to the bottom bracket is cumbersome and possibly imprecise.
>> After tightening, the crank bolt must be removed to make this
>> measurement.
The crank has no life of its own and will not move when the bolt is
removed. The dimension is repeatable and accurate.
> Removal of the crank bolt is not necessary using my method, and I
> did not remove the bolt for measurement. Removing the bolt is a
> bad idea if you wish to find out where the crank sits after the bolt
> is torqued. Removing the bolt calls into question whether the crank
> remains in the same place or not.
It remains in the same spot, gremlins aside.
>> The initial position of the crank before tightening is an ill
>> defined position so it has no value for this test.
> Though inherently unreliable, the finger tight specification is surprisingly
> repeatable, especially when the same person does it carefully the
> same way each time, as I did for all the cranks in my test. To find
> out how repeatable "finger tight" is, I mounted one crank ten times.
> The variation was 0.1mm, and the standard deviation was only 0.03mm.
I assume you were using a wrench to "finger" the screw. It is not
clear what you were doing when you say finger tight and what torque
this represents.
> In spite of this apparent repeatability, the inherent vagueness of
> this finger tight position is why I measured the distance from the
> inside of the test crank to the machined reference surface on the
> inside of the opposite crank. As I wrote in my original post:
> "I recorded the position of each crank after mounting it the first
> time (dry). Then I recorded the position of each crank after mounting
> it the second time (greased). On average, the final resting place
> was 1.1mm farther on the spindle when mounted the second time
> (with grease) compared to the first time (dry). After wiping off the
> grease and re-mounting the cranks a third time (dry), they still ended
> up 0.6mm farther up the taper than they did the first time, though
> they did recover somewhat from the greased position."
> This is absolute position relative to the stationary opposite crank,
> not relative position starting from finger tight.
As you see, the control of your experiment raises questions about
cause and effect and whether these are significant. a 1mm depth of
engagement change is a 0.07mm press fit change. That's a substantial
change in an aluminum span of 12mm or so.
> So greased or dry, regardless of whether the crank was at the same
> place when finger tight, it came to rest (after torquing) farther up
> the tapers when greased than not. The greased cranks sat closer
> to the opposite crank than did the cranks I mounted dry.
This is to be expected.
> There are two things I did not measure, about which I am curious.
> One is how far the crank squirms up the taper when ridden. I
> think this must depend on the crank's preload, with a tighter fit
> (the greased crank) allowing less squirm.
The squirm is the same, because it is dependent only on the modulus of
elasticity of the crank. No lift-off in the press fit should occur
with a properly installed crank. How far it walks up the spindle is
the balance between screw preload and the slope of the taper and its
press. The residual load on the screw is proportional to the
tightness of the press fit. If the press is light, from too much
installation friction, the screw can be left entirely loose after use
because the press was too light and the crank position easily relaxed
in use to a screw free position.
> The other thing I did not measure is the strain on the bolt at initial
> installation, after torque, before the crank squirms. This must
> depend on the bolt's material and design (does it have a street of
> a flare at the base of the threads?). Does bolt strain also depend
> on the unengaged length of the bolt?
Yes it does, and if there were an interest in improving this joint,
screws would be thread-free and reduced diameter for the first part
under the head. Because cranks have more serious problems and do not
come unscrewed when properly installed, this is not a current concern.
> Have you made any measurements?
No. And I haven't had such crank problems, but I analyzed why, after
use, crank bolts are no longer as tight as they were at installation,
and that there is no lubricant in the interface during use that is
apparent from the fretting erosion in the steel spindle.
What was less apparent, is that cranks that do not fail at the pedal
eye, neck, or spider attachment, will ultimately fail in the square
hole where one would initially expect failure, because the press fit
stress is high there. In the years that I rode Campagnolo cranks, the
failure rate was so high that no crank got old enough to crack at the
square in the attachment. I expect to see this as the new failure
location assuming reports we see here are accurate... and I don't mean
during installation.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: I greased my crank and now it's split!
Date: 22 Sep 1997 16:51:23 GMT
Damon Rinards writes:
> Jobst says dry cranks and greased cranks squirm the same. If this
> is so, then the greased crank starts out farther up the taper and
> always stays "ahead" of the dry one.
With the same torque, the creep after tightening should be different
because there is a closer relationship between torque and taper
tightness than with an unlubricated joint where torque is more
strongly affected by friction that inhibits engagement of the taper.
Thus the crank can squirm farther away from the bolt than it would in
a lubricated fit.
This subject came to light when I found several crank bolts along the
route of a century ride and one that came out of a friends crank on a
trip over the sierra. All had been tightened to the appropriate
torque, or at least the shop said so. Since I had never had such a
failure, I gave it some thought and came to the conclusion that dry
installation would give a false tightness indication and that such
cranks could squirm fully clear of the retaining bolt, as these did.
I am sure the crank bolts did not unscrew until they were holding no
load. That is, the screws did not unscrew to fall out until the crank
left them under no load. Of course, these people were all riding
without caps on their cranks, or the screws could not have backed out.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: greas bb spindle??
Date: 3 Mar 1998 23:47:05 GMT
John Marcos writes:
> Not to impugn Dave's recommendations (too much), but I have some
> disagreements. 1-Why on earth would Race Face want to distribute
> erroneous information? It seems to me that the people who design,
> build, test, warranty and have to listen to complaints about creaky
> cranks would be the best source of information about how to install
> cranks. Only a few of us would be interested in the reasons why, so
> they probably haven't bothered to write it all down on the off
> chance that some of us might be inclined to read it.
That sounds good but I haven't heard from any manufacturers why grease
would compromise crank durability. Originally the argument went that
the hole would ream out because there was insufficient friction to
turn the spindle. This argument faded away and was replaced with the
crank would slide too far up the taper... and then what?
> 2-Your statement that tests show that the crank does indeed slide up
> the taper farther sends a red flag up about the advisability of
> greasing the taper. The stress put on the crank arm is in part due
> to the expanding force applied by the taper as you jam the arm up
> it. Forcing the crank up the taper farther than specce'd will lead
> to more stress and a greater chance of failure.
What you overlook with this assessment is that the crank squirms on
the spindle, with or without lubrication, and effectively elbows its
way up the taper as it pushes off from the retaining bolt. This
squirming force causes failure of aluminum and titanium bolts.
> The amount of 'Jam' is specified by the torque setting on the crank
> bolt. If the torque is listed for an un-greased spindle, the torque
> setting will correctly jam the crank up the ungreased spindle the
> correct distance, resulting in a within-tolerance level of stress on
> the crank arm.
You cannot specify a torque for an ungreased spindle unless you
specify a procedure for de-greasing that spindle. A spindle casually
handled by most mechanics is in fact lubricated. A crank spindle from
most any machine shop is lubricated by adsorbed aerosol hydrocarbons.
How well it is greased is debatable, but it is not de-greased.
> Greasing the spindle in this case will allow the crank to slide up
> the spindle farther, increasing the stress in the crank arm to the
> point that it may be out of tolerance and break prematurely.
Have you ever broken a crank? Most cranks break at the pedal eye or
at the junction of the right crank with its spider. This is not
affected by how tight the crank is pressed onto the spindle. In fact
I know of few cranks that have failed at the press fit, the most
conspicuous of which was a "dry" crank that had been repeatedly
re-tightened by a rider who followed the advice given by the shop.
Lest it was missed, this procedure can split most cranks with little
effort because there is always almost free travel after a crank has
been ridden hard. Following the crank up the taper with the bolt will
force it to squirm farther up the taper to reach equilibrium.
> This is, in effect, over-torquing the crank bolt, which no one would
> recommend. The anecdotal evidence that 'Many' or 'Lots' of people
> grease and suffer no ill effects means little.
You seem to beat around the torque bush. What would be the harm in
specifying the correct torque for a crank that was oiled or greased?
I suspect that the torque ratings given are for such an assembly
anyway because hardly any crank is de-greased. I am certain that the
fear of manufacturers are based on a misunderstanding of this joint.
First because they are unaware that torque is transmitted exclusively
by the press fit, not friction. Second, because they are unaware that
the crank-spindle interface is a dynamic joint that frets continuously
when used hard (other uses being irrelevant).
These two items become evident when considering the contact pressure
if the crank were caused by torque to lift off the "pulling" side of
the square hole. This would yield the hole into a "pin cushion" shape
and ruin the crank immediately so that it could not be kept tight.
That the interface squirms and moves to an equilibrium position
leaving the crank bolt at less than its initial tension is apparent
from loosening a used crank... or for that matter tightening it. The
interface also contains rouge, showing that fretting takes place.
> Good cranks are built with some safety margin and the greasing may
> not cause enough additional stress to prematurely fail the cranks.
> Cranks that are on the weaker side of the manufacturing tolerances
> might be adversely affected by the additional stress. If the torque
> settings are for a dry spindle, don't grease it when assembling. On
> the other hand, if the torque setting is for a greased spindle,
> grease away. Race Face is obviously specifying a non-grease torque
> setting, so follow their advice, leave the grease off for their
> cranks, and ride happy.
Getting back to the mode of failure. What failures have you seen.
I have not seen these over-tightening failures, but rather pedal eye
fractures. What is Race Face's concern. I have watched Specialized
MTB cranks break in half at mid span where they are broad and flat.
> 3- Not at all to insult you, but, so what if you're a professional
> engineer? OK, it sounds insulting, but only mildly, hopefully.
> Unless you are a materials science/mechanical engineer with
> extensive experience in taper joints in general, and bicycle crank
> taper joints in particular, you are just as capable of being wrong
> about things as any of the rest of us.
Oh pshaw! It is every good mechanical engineer's business to
understand mechanical interfaces and strength of materials. You are
making an extreme proposal that one needs an advanced degree in every
field that bears on a design problem. Typically a few engineering
jobs on which I have worked involved racing car brakes, transmissions,
steering, suspension, and tires, when I worked at Porsche. We had
plenty of these problems to solve. Working at the SLAC physics labs I
designed water cooling for hundred ton magnets and microscopic
detector elements to work in high vacuum. Here at HP I have worked
along with engineers with whom I designed laser optics, microscopes,
computer disk spindles and published papers on disk tribology. I
don't have an advanced degree in any of those subjects and have met
with great success and many patents in these fields. This is typical
among the people with whom I have worked.
I think you are drawing your conclusions from some other technical
environment than where engineering challenges are met. Certainly the
bicycle industry is not as high tech as the fields in which I have
worked. You sound like the bikies who are sure I am wrong about
bicycle wheels because I haven't built as many wheels as the mechanic
who obviously knows more about it than I because he built wheels for
"the pros".
> I happen to be an engineer too (non-practicing electrical, the
> simple engineer title hides that little fact) but the experience I
> am drawing upon for my humble opinions is based on years of
> experience in working with machined parts.
I believe you drew your conclusions from the wrong sample of
engineering.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 19 May 1998 01:17:51 GMT
Tom Kunich writes:
> When someone says that they have broken several RIMS I suspect that
> there is a reason. A friend of mine isn't very big or heavy but he
> is hard on wheels because he doesn't watch the road and hits lots of
> trash and potholes. He is a big mileage rider so it isn't because
> of experience.
That's a forced failure, not a cracked or separated rim. People who
don't understand that difference don't need an explanation. When
people talk of rims failing they generally don't mean the bent their
rim.
> BTW, I discussed your penchant for breaking BB spindles, cranks and
> all that sort of stuff with an old bicycle shop mechanic/owner. He said
> that he used to see those types of failures all the time and he described
> your riding technique to me: Mashes big gears up hills was the gist of it.
Thanks for the explanation but you missed the crank failure threads of
the past. Cranks break off at the pedal eye from the bending load at
the bottom of the stroke... when stepping off onto the other pedal.
Cranks generally break in bending at the bottom of the stroke. Hence,
the gears you ride have little effect on this load. Besides this, I
have observed my cranks and those of others break, an it isn't where
you are pushing hard. Your foot just ends up in the ground at the
bottom of the stroke.
> I've talked to people that rode with you and they confirm that as your
> style. Can you find fault with that description?
Yes.
> Anyway, he said that spinners seldom had any important gear failures and
> that people like me without very pronounced leg strength don't often have
> gear failures like that.
Well if he said that he must know. Having not broken any cranks he
is probably the person to tell you how they break.
> Well, I haven't yet and none of my equipment shows the strains that
> I would relate to near overload conditions that would accelerate
> fatigue problems.
I don't know many 180lb riders who put on 10,000 miles a year, mostly
climbing hills. My heavy 230lb+ friend who rides with me has crank
spindles, pedals, seat posts, stems and rims fail. He also turns
lower gears than I do, as if that had anything to do with it.
Tom, I don't really like to respond to your claims because you stick
your chin so far out it's hard to miss. Don't do this. I got the
impression when we met and talked awhile ago, that you were capable of
more reasoned postings than what you often post here. Just slow down,
re-read and edit.
> I think that you are a real tough rider Jobst. And I think that that
> is the source of most of your failures. Keith Bontrager tested many
> cranks and spindles on his fatigue tester and the only way he could
> duplicate your results was by really cranking the pressure onto it.
> Hey, I'm happy making it up Mt. Hamilton, I don't need to set any
> speed records on the way.
I'm not sure the tester was using a valid model for its load
characteristics. Just as with computers, garbage in = garbage out. I
have seen enough testers and all the ones I saw, assumed failures
occur when the crank was horizontal, so that's where the load was
applied. This is not the common failure mode of cranks.
>> I've had it happen to a month old bike. The crank assembly sags
>> when you stand and comes up, putting the seat tube in compression,
>> when you sit. In that case you ride home sitting and realize how
>> important standing over rough pavement is.
> I'm curious, how could you of all people get a junk frame?
It was a Cinelli and had you seen how they brazed their frames, you
would understand why they could easily fail like that.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 19 May 1998 15:57:56 GMT
Tho X. Bui writes:
>>> I'm curious, how could you of all people get a junk frame?
>> It was a Cinelli and had you seen how they brazed their frames, you
>> would understand why they could easily fail like that.
> I'll bite. How _do_ they braze their frames, and why would they fail
> easily? I'd never thought that brazing, as done on bike frames, isn't
> all that complicated. How would someone mess that up?
Brazing was done in a fire brick open-top oven into which two gas
flames were directed. The brick enclosure was about 30x15x30cm with a
slot in each of the four walls so a frame tube could protrude. The
appropriate corner of the frame was laid in the cavity, hand held,
until it glowed light orange, at which time brass was introduced in
the form af an 8mm wire. The brazing was visibly excellent as the
brass flowed and wet the metal. However, if left too long in the
caldron, the steel turned to coat-hanger 101. My frame apparently got
the heat and the BB cracked through the down tube and BB shell. There
was no separation of a brazed joint. The steel failed.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 19 May 1998 17:48:00 GMT
Tom Kunich writes:
>> Thanks for the explanation but you missed the crank failure threads
>> of the past. Cranks break off at the pedal eye from the bending
>> load at the bottom of the stroke... when stepping off onto the
>> other pedal. Cranks generally break in bending at the bottom of
>> the stroke. Hence, the gears you ride have little effect on this
>> load. Besides this, I have observed my cranks and those of others
>> break, an it isn't where you are pushing hard. Your foot just ends
>> up in the ground at the bottom of the stroke.
> Excuse me, but _when_ the cranks fail catastrophically isn't
> necessarily _where_ the fatigue loads occur. I've seen your spindles
> in Wheelsmith and I've seen spindles with the start of cracks. I'm
> not convinced that they are caused by the loads at BDC even though
> that is your opinion.
However, upon inspection, the nature of the crack, its growth and
separation are evident and the failure occurs as I described even if
you didn't notice the circumstances when you broke your crank. And
you are NOT excused.
> What I see is a crack that starts at the edge of the taper and grows
> both radially as well as longitudinally. That suggests to me that
> the twisting load is causing the growth. After the crack gets to a
> certain point I'm sure that the standing shock loads will exceed the
> remaining material strength.
Are you describing a failure at the spindle knuckle? I can't make out
from your description where this crank failed. Besides, there are no
shock loads on cranks unless the crank strikes the ground.
Most crank failures occur at the pedal eye. Specialized cranks are a
special case, so to speak, because they are about 10mm thick at
midspan, the place where they break in lateral bending. The fracture
shows a progressive separation across the crank from outside to
inside.
> In fact, I do not place much load on the pedals in that position
> even though you seem to think that you do. As I reach BDC I have
> already transfered my weight to the high pedal and am pulling up on
> the low pedal if the grade is steep. Just like walking up stairs.
I don't think you have observed how you walk up stairs any more
accurately than your pedaling forces. The crank is weakest against
the bending loads at the bottom of the stroke, the place where on
often stands when cornering.
>>> I've talked to people that rode with you and they confirm that as
>>> your style. Can you find fault with that description?
>> Yes.
> Well, I've received a couple of messages from people who ride with you
> that confirmed my opinion. I suggest that climbing hills like Kings
> Canyon in the big ring qualifies you as a pedal masher. Uhh, I expect
> that is some sort of compliment since you must be a powerful rider to
> do something like that. But it is an equipment stress test.
That's interesting. I have never ridden in Kings Canyon but if you
say so. Besides, the "big ring" doesn't do much if you don't mention
what sprocket I ride on the rear. My big ring isn't big either.
>> Well if he said that he must know. Having not broken any cranks he
>> is probably the person to tell you how they break.
> Having worked on bicycles for 40 years and built some 500 frames,
> sponsored racing teams and national caliber riders ought to qualify
> him to make knowledgeable comments without criticism from someone
> that doesn't know him.
That';s about like the old saw "Who knows more about bike, Greg Lemond
or you?" This is the response I got to "the Bicycle Wheel"from people
who have built many wheels. I suppose "knowledgeable comments" is a
euphemism to avoid the hard facts. I am not convinced that a mechanic
necessarily understands the dynamics and materials he works on.
> The conversation went like this - ME: What would you say was the cause
> of someone breaking cranks and spindles. HIM: I used to see that a lot
> when the lowest gear you could get was a 48 tooth on the front and
> everyone rode a straightblock.
That pretty much sums it up. The cause of broken cranks a few years
ago was lack of gears. Give me a break! As long as I have been aware
of derailleurs, lower gears than any racer might want were available.
Regina clusters with 32t sprockets were available in the 1950's. That
was before aluminum cranks and failures of steel Magistroni cranks and
their likes were unknown.
> Since 42 or smaller chain wheels and 24 tooth rear cogs have become
> common you don't see these failures all that often anymore, though
> the guys who carry big gears up hills and the real strong sprinters
> still break the stuff like before.
Oh! Real "strong sprinters" eh? You seem to be talking about forces
ruptures, not fatigue. Sprinters don't break cranks unless they
change their style and ride road, a place where their anaerobic
strength doesn't appear. Who is this person with whom you are having
an imaginary conversation?
>> I don't know many 180lb riders who put on 10,000 miles a year, mostly
>> climbing hills. My heavy 230lb+ friend who rides with me has crank
>> spindles, pedals, seat posts, stems and rims fail. He also turns
>> lower gears than I do, as if that had anything to do with it.
> I found some guy walking along Highway 9 that was pushing his bike. I
> was on my MTB having just dropped down to the road and I tried to offer
> him help figuring that he had a flat. Nope. He'd broken his C-Record
> crank at the insignia. He said that it was the _3rd_ one in 6 months.
> Sorry, but that is abuse of the cranks in my books. I don't pretend to
> know what he's doing wrong, but when this only happens to .01% of the
> riders, it's something that he is doing wrong.
Please explain what "abuse of the cranks" is. As I said, bicycles are
made for humans, not the other way around. Besides, I can vouch for
the failings of C-Record Campagnolo cranks. My only set failed both
cranks at the pedal in less than 200 miles. That occurred before I go
around to the first inspection. I stopped that right there and
switched to Dura Ace with a modicum of success.
>> I'm not sure the tester was using a valid model for its load
>> characteristics. Just as with computers, garbage in = garbage out. I
>> have seen enough testers and all the ones I saw, assumed failures
>> occur when the crank was horizontal, so that's where the load was
>> applied. This is not the common failure mode of cranks.
> I think I agree with you that the testers don't give valid data. They
> commonly break handlebars and stems and I've very seldom seen these
> sorts of failures.
Oops! Bars and stems, but especially aluminum bars break off often
for strong riders if they are not routinely replaced. That is why I
prefer to continue using my steel bars. Stems also fail. Mine failed
at the ends of the expander slots of Cinelli stems.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 19 May 1998 23:52:51 GMT
Keith Bontrager writes:
>>> I'm not sure the tester was using a valid model for its load
>>> characteristics. Just as with computers, garbage in = garbage out. I
>>> have seen enough testers and all the ones I saw, assumed failures
>>> occur when the crank was horizontal, so that's where the load was
>>> applied. This is not the common failure mode of cranks.
>> I think I agree with you that the testers don't give valid data. They
>> commonly break handlebars and stems and I've very seldom seen these
>> sorts of failures. Stress testing machines have broken frames about
>> as easily as cranks. You just turn up the pressure on the hydraulics
>> and you can break anything. That doesn't mean that you've 'tested'
>> anything important.
> Lads, before you get carried away. Jobst, maybe you haven't seen
> enough testers. You haven't seen ours, eh? We tested both
> horizontal and vertical loading.
No one mentioned what test and how it was conducted. My assumption
that it may have been an invalid model arises from the conttrary
results I have had and continue to have. I am still breaking cranks
at the pedal eye.
> Presumption seems to be a strong trait of yours. Bad characteristic
> for someone who purports to know so much. Kind of erodes your
> credibility.
Would you please point out specifically where you see this sort of
statement.
> Garbage in = garbage out. Hard to argue with that...
> Tom, you may agree that testing gives invalid data. I agree that
> this is the case when it is done or interpreted improperly.
> It is certainly not the case that all testing is always done
> improperly, right? Are you also claiming, based on Jobst's lack of
> knowledge of our testing methods and the (unknown to him) scientific
> merits of same, and a similar lack of knowledge on your part, that
> we were screwing it up? You cited it for your argument. I don't
> get it.
I know nothing of the tests nor what test was used for Tom's deductions.
> Curiously clairvoyant on both your parts...
I see nothing. Who is clairvoiant?
> Back to the testing we did. We did find that older road crank
> designs could be broken at the pedal eye when loaded at the bottom
> of the stroke, just as Jobst claimed. This is backed up by field
> samples that failed this way.
Well my newer road cranks, Campagnolo Record and Shimano DuraAce are
still breaking at the pedal eye.
> We did not find this to be the case with newer MTB cranks of Shimano
> LX specifications and higher of that vintage 4 or 5 years ago). We
> didn't test cranks below this specification at that time.
I don't understand why there should be a difference for MTB cranks.
Their function is the same as on any bicycle.
> At larger loads the newer designs would eventually break. But this
> was not a good simulation and we would never claim that it was. It
> was an exercise, a method to examine eventual failure modes. What
> would fail first, the crank or the pedal spindle? We were already
> satisfied that the newer cranks would not break as Jobst's older
> cranks had, so it was just for drill.
What do you call newer?
> I challenged the readership to produce a crank of the more modern
> design that had failed that way at the time. No one came forward.
> Since then I have kept an eye open for one, and I am always lurking
> around shop junk piles for broken stuff. Never saw one. Never
> heard of one.
I can't participate until I know what you are looking for. I see no
solution for the problem until the pedal attachment is changed. I
have proposed a method that could be retrofitted and could replace the
ones in use.
> There are simple explanations for this. That is what we were
> looking for. I do not think it is related to Jobst's theory of
> fretting between the pedal and crank at the threaded attachment.
> Nice theory, but it was not born out by the evidence.
I think you'll find that all pedal cranks that are stressed have a
rough "counter bore" made by pedal fretting. I have not seen one that
did not and this is the essence of stress concentrations.
> The actual cause is more subtle. It's pretty simple if one looks at
> the metallurgical and section differences between the older European
> designs and the later Japanese designs. No big deal once you do the
> work.
How recent? As I said, my recent DuraAce cracked as usual.
> A simple bottom line: If you are big, don't ride older European road
> crank arms indefinitely. No news here, right?
I am not convinced of this "Older" designation. My C-Record cranks
broke in the shortest mileage I have experienced.
> There were many and varied failures when the cranks and bb spindles
> were loaded horizontally, and the modes of failure were consistent
> with what we have as samples of field failures. The spindles failed
> in the way you describe generally, though there were some variations
> that showed up due to material and heat treatment differences.
> Crank arms cracked at the web between the spider and arm on the RH
> crank, and these would slowly, but eventually lead to a fracture
> through the section. Some cranks failed at the square drive taper
> too, but these failures took much longer and may not show up in the
> field.
They do. I have had them on DuraAce.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 20 May 1998 00:45:29 GMT
Tom Kunich writes:
> I haven't seen the cracks at the pedal end of cranks that you have
> talked about on any modern crank. You inform us that you break
> DuraAce cranks often enough to make a person nervous, but I've never
> seen one in my experience.
Well that's where most cranks fail, and I don't mean start to crack but
experience separation.
>> Are you describing a failure at the spindle knuckle? I can't make out
>> from your description where this crank failed. Besides, there are no
>> shock loads on cranks unless the crank strikes the ground.
> The cracks I've seen in spindles (two or three) all were cracks
> that radiated from the junction of the square edge of the taper where
> it met the round portion of the shaft. The cracks radiated counter-
> clockwise and started down the taper. that certainly looked like it
> was caused by rotational loadings to me.
Is this cranks or spindles? I don't know how you can construct sentences
that cannot be parsed as to the subject. What is breaking?
>> That's interesting. I have never ridden in Kings Canyon but if you
>> say so. Besides, the "big ring" doesn't do much if you don't mention
>> what sprocket I ride on the rear. My big ring isn't big either.
> Of course I meant Kings Mountain Rd. I have been told that someone
> saw you climbing that in a 50-17. Several people have told me about
> hearing your wheels screeching from the ultra-slow pedal cadence that
> you have while climbing in the big ring at moderate speeds.
So? That's about the gear that was common in the Kings Mountain hill
climb when we were still young and beautiful... on a track bike. I
may not be fast anymore but my legs still work. Think about it. The
rider's weight is on the pedal when standing, regardless of what gear
he uses. The cracks occur in lateral bending mode. This big gear
phobia is something like float in pedals. It's a recent occurrence.
> You may think that that isn't exceptional, but believe me, going
> up a 7-12% grade in a 50-17 impresses the hell out of me. And you
> don't expect me to believe that your style is so smooth that you
> aren't getting impact loads during that sort of thing?
Well maybe, but Kings Mountain road is a 7-8% grade at its steepest
and averages about 6%. Impact does not exist between metals and the
human physique. I think you are confusing impact with pulse loading.
This does not constitute impact with respect to structural metals.
>> That's about like the old saw "Who knows more about bike, Greg
>> Lemond or you?"
> BS, that's like saying that someone that has been repairing the
> breakages ought to know a little bit more about what the most common
> breakages are than someone who has a superman style of riding.
This may be what is often thought but it does not follow in most
cases. If you think how long people have been building wire spoked
bicycle wheels and how little was known about the stresses and loads
in the wheel until recently, your deduction doesn't ring true.
Besides, it was these experienced bicycle mechanics that inspired me
to research and write on the subject of wheels when I heard the myth
and lore they repeated.
> Jobst, every year we see your tour postings of all the passes you
> climb during your vacation. Every year we are amazed at your
> power and stamina. But don't then pretend that you aren't an
> exceptional rider and try to convince us that you are Joe Average.
So? Do you propose that bicyclists start molding themselves to the
failings of the bicycles and components that are mis-designed? It's
the same as telling a hiker to take smaller steps to preserve his
hiking boots because the soles don't like to bend that much.
>> That pretty much sums it up. The cause of broken cranks a few years
>> ago was lack of gears. Give me a break! As long as I have been aware
>> of derailleurs, lower gears than any racer might want were available.
>> Regina clusters with 32t sprockets were available in the 1950's. That
>> was before aluminum cranks and failures of steel Magistroni cranks and
>> their likes were unknown.
> OK Jobst, what is your low gear? I've ridden with the 42/19 low gear
> crowd enough to recognize one of the leaders. And I'm sure that you
> agreed with Alexi Grewal when he said that you could climb anything
> in the USA with a 42/19.
I'm not in to boasting competitions. Bicycles should support their
riders, assuming they are not unusually heavy, without breaking
dangerously. Cranks, pedals and spindles are life endangering
failures.
>> Oh! Real "strong sprinters" eh? You seem to be talking about forces
>> ruptures, not fatigue. Sprinters don't break cranks unless they
>> change their style and ride road, a place where their anaerobic
>> strength doesn't appear. Who is this person with whom you are having
>> an imaginary conversation?
> These guys that get out in the road and sprint off of every light, stop
> sign, city limit sign and grin by their riding partners put a hell of
> a lot of wear on the parts. Are you suggesting that that isn't a
> significant fatigue loading? That guy who had broken the cranks half
> way down the arm (remember -- 3 of them in 6 months!) had quads as
> big as your waist. Yet the average or even better than average riders
> are still riding those cranks with much higher mileage on them years
> later.
That is incorrect. These are not the people who have the failures.
It is the long distance and endurance riders. That's where half the
myth comes from. I have often heard that I was doing it all wrong
because Joe Doe who is a burly sprinter, doesn't break cranks. Well
that's natural. You don't get nearly as many stress cycles in a few
sprints as you do climbing a 10% grade for 10 or 15 miles... day after
day.
> Again, is that luck? Or riding style?
How about "riding"? Leave the style out of it.
>> Please explain what "abuse of the cranks" is. As I said, bicycles
>> are made for humans, not the other way around. Besides, I can
>> vouch for the failings of C-Record Campagnolo cranks. My only set
>> failed both cranks at the pedal in less than 200 miles. That
>> occurred before I go around to the first inspection. I stopped
>> that right there and switched to Dura Ace with a modicum of
>> success.
> Everyone that has broken DuraAce cranks please raise your hands! Can
> you tell us where the dividing line is between strength, weight and
> endurance? Do you expect every component manufactured to put up
> with any load conceivable for an infinite period of time?
Conceivable is easily achieved. Cranks break and that they are at the
ragged edge of doing so for riders in the normal distribution of
weight is also known. If the margin were reasonable, then I wouldn't
be inspecting cranks once a week. Besides, how much do you think
the individual can alter the loads on a crank, assuming he doesn't
ride the lowest gear on every hill?
> At some point you have to admit that you can break anything with
> an overload. Placing the values of that overload is what engineering
> is all about. Failures of bicycle parts are very low on the average
> and reflects some pretty decent approximations in my view.
We don't build any other vehicle that way.' Unfortunately, there are
too many Tom Kunichs who will defend the status quo for the industry,
so they don't even have to raise a finger. It is this attitude that
leaves the brunt of the corrections to the courts, where mostly poor
test cases come to trial, and horrendous settlements are paid by
insurers. The manufacturer then turns his back on the problem and
sails off into the sunset.
>> Oops! Bars and stems, but especially aluminum bars break off often
>> for strong riders if they are not routinely replaced. That is why I
>> prefer to continue using my steel bars. Stems also fail. Mine failed
>> at the ends of the expander slots of Cinelli stems.
> Where do you get steel bars?
I've been riding the same ones since the 1960's when Mr. Cinelli put
them on my bike himself.
> Again, Jobst you are not an average rider. You are an exception and
> a particularly rough rider on your equipment. Pretending that bicycle
> component manufacturers should fit your style probably ain't a good
> way to proceed.
What a bunch of weenies! I think you'll never know what it was to
ride hard when you were in your teens and you could "leap tall
buildings in a single bound". All the bikies in the mid (SF)
peninsula, and then some, used to come out on Sunday rides and on
trips to the Sierra Nevada to go on rides that spawned the "Death
Rides". That's when we were strong and rode till there was to dark to
continue. I recall in 1975 as I rode with Bill Robertson, just after
winning the District Road Championship in Carson City. I said to him
after crossing many passes that day, just think about it Bill, the
only reason we're stopping in Altdorf (CH) is because it's to dark to
ride over the Susten Pass. We did that the next day.
Ritchey, Fisher, Baldwin, Ball, Vierra, Perry, Robertson, Crawford,
Hoffacker, Garner, P. Rich, and many other champions were regulars on
the rides in the Santa Cruz mountains. The wimpy stuff we get today
for our bikes is playing to the wrong audience.
Ray Hosler has a little bit of the later rides. Last Chance Road is
one of the features. We did that one a week ago with Stella Hackel.
http://ng.netgate.net/~hosler/
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 20 May 1998 15:19:19 GMT
David Rees writes:
> Jobst, why haven't you tried steel cranks such as Bullseye,
> Synchros, Sweet, etc.? Do you think that they are as, or more,
> likely to suffer fatigue failures (at trhe welds) than cold-forged
> Al cranks?
My experience with weldments is not good. Where there hasn't been an
OEM use of the product, I feel it's too risky to use unless there are
other compelling reasons to believe the product is superior. I don't
want to be the product tester.
I rode steel cranks by Magistroni long after Camagnolo aluminum cranks
were available because the early ones broke left and right at a high
rate. A few years later, when I could no longer find good quality
crank cotters, I switched to Campagnolo and in the following years was
able to avoid crashing by detecting most of the cracked cranks shortly
before seporation. I took some dives. Two dozen cranks later I
switches to C-Record for 2000 miles or so and switched to Shimano, the
failure rate is way down but still there.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 20 May 1998 20:42:00 GMT
Keith Bontrager writes:
>> I don't understand why there should be a difference for MTB cranks.
>> Their function is the same as on any bicycle.
> Possibly. There are differences in cross section, and possibly in
> materials and processes, though the latter is a stretch. I am still
> curious about the lack of this type of field failures (in my
> experience of course). I have never witnessed a Shimano or Sugino
> MTB crank fail as you describe in all my efforts to find broken
> stuff. I would be happy to admit that there could be a problem,
> but, other than your claim, I have never seen evidence of it in
> testing or in the field.
Cranks on my bicycle have been breaking at a diminished rate but
continue to do so. This could be a sign of my getting older and not
riding as much. The type of failure has not changed visibly and I
have sufficient events to believe the statistical norm.
> This doesn't mean it won't happen of course. Anybody got one?
One what. Broken cranks are available from any volume bike shop.
>>> At larger loads the newer designs would eventually break. But
>>> this was not a good simulation and we would never claim that it
>>> was. It was an exercise, a method to examine eventual failure
>>> modes. What would fail first, the crank or the pedal spindle? We
>>> were already satisfied that the newer cranks would not break as
>>> Jobst's older cranks had, so it was just for drill.
>> What do you call newer?
> [...]
> The A204 cranks exhibit a very close, uniform grain, none of the
> refinement of a forging and is a lot more forgiving in certain areas
> of fatigue strength. A204 is also pretty strong stuff in tensile as
> well. It seems to be less fracture prone than 7075 aluminum, which
> is my guess at what is used on C Record, and possibly Dura Ace. (The
> latter is a guess and depends on the year the Dura Ace cranks were
> produced - Shimano has forged 7075 and similar alloys).
I don't suppose DuraAce cranks are made of inferior material to some
of the models you mention. I suspect that they would use their best
material there if it is the top of their product line. At least the
price and use among professional racers seems to imply that.
> All of this is bad guessing, and tests are needed to confirm it.
> I apologize, in advance.
I don;t see why DuraAce was not yet tested. It seems logical to me
to try to find the strongest crank to use as a comparative norm.
>>> I challenged the readership to produce a crank of the more modern
>>> design that had failed that way at the time. No one came forward.
>>> Since then I have kept an eye open for one, and I am always
>>> lurking around shop junk piles for broken stuff. Never saw one.
>>> Never heard of one.
>> I can't participate until I know what you are looking for. I see
>> no solution for the problem until the pedal attachment is changed.
>> I have proposed a method that could be retrofitted and could
>> replace the ones in use.
> An MTB crank with a broken pedal eye. Any will do, though if it's
> an old TA, it doesn't prove what we are looking for. I am really
> interested in seeing a broken pedal eye on an MTB crank of the circa
> noted above (or newer).
What makes it an MTB crank in contrast to others? I have a four year old
DuraAce with the pedal eye cracked half way through.
>>> There are simple explanations for this. That is what we were
>>> looking for. I do not think it is related to Jobst's theory of
>>> fretting between the pedal and crank at the threaded attachment.
>>> Nice theory, but it was not born out by the evidence.
>> I think you'll find that all pedal cranks that are stressed have a
>> rough "counter bore" made by pedal fretting. I have not seen one
>> that did not and this is the essence of stress concentrations.
> I agree that fretting can occur and that this is one of the ways
> fatigue failures can be induced in a part.
> I do not agree that it is the only mechanism.
No one said that it was but if that's the basis of your contention, I
want to make it clear that I never said that. I mentioned it as a
principal mechanism. I have had cranks break at the neck before the
pedal eye. These also broke in bottom-of-the-stroke bending from
outside to inside.
> I do not think it is the cause in this case. There are also many
> stress concentrations on the threaded surfaces, in the threaded
> pedal mounting hole. The first thread peaks (closest to the pedal
> flange, where it seats on the crank arm) are located near the point
> of peak stress and are loaded in cyclic tension in the loading case
> you cite.
Well with the solution I proposed often, the first thread would not be
at the pressure face of the crank and the skewing loads would not be
shear forces on the pedal face either. It's simple to put a 90 degree
countersink in the face of the crank, into which a mating face of the
crank would screw. There are many machinery interfaces made like
this. It can also be retrofitted to existing pedals with a split
conical washer that screws onto the pedal thread and against the
shoulder.
> I am not as certain that the fretting creates residual tensile
> stresses in the aluminum surfaces in an orientation that would lead
> to premature fatigue failures.
It's not the fretting itself but the canting that causes peak loads
on sharp surfaces of the crank. The chipping and gouging that occurs
doesn't help.
>>> The actual cause is more subtle. It's pretty simple if one looks at
>>> the metallurgical and section differences between the older European
>>> designs and the later Japanese designs. No big deal once you do the
>>> work.
>> How recent? As I said, my recent DuraAce cracked as usual.
> See above. Are these broken Dura Ace cranks at Wheelsmith too? What
> year were they made (or when did you start riding with them and were
> they new)?
Most of my Campagnolo cranks were returned to Campagnolo until they
ceased accepting them for replacement. Then I dropped one off at
Wheelsmith, I don't recall when. My Shimano cranks also were returned
except for the last few that I still have lying around.
>>> A simple bottom line: If you are big, don't ride older European
>>> road crank arms indefinitely. No news here, right?
>> I am not convinced of this "Older" designation. My C-Record cranks
>> broke in the shortest mileage I have experienced.
> Yeah, you caught me on that. I have no information on road stuff
> other than 4 or 5 year old Campy stuff. older = 5 year old +.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 20 May 1998 21:07:11 GMT
Tom Kunich writes:
>> Well that's where most cranks fail, and I don't mean start to crack
>> but experience separation.
> If you say so I'll believe it. But I haven't seen it and wonder how
> often it happens.
I thought you were telling me about crank failures and now you say you
don't know where cranks fail. Let me add that the cracks in the
spider finger trailing the crank are not failures until the crack
turns outward and crosses the crank. Most of mine had these cracks
throughout their service, but most failed at the pedal eye just the
same.
>>> The cracks I've seen in spindles (two or three) all were cracks
>>> that radiated from the junction of the square edge of the taper where
>>> it met the round portion of the shaft. The cracks radiated counter-
>>> clockwise and started down the taper. that certainly looked like it
>>> was caused by rotational loadings to me.
>> Is this cranks or spindles? I don't know how you can construct
>> sentences that cannot be parsed as to the subject. What is
>> breaking?
> So, where does it mention cranks in my writing above? You cannot
> parse two sentences adequately? Are you suggesting that "taper"
> only refers to the taper inside of the cranks and not on the outside
> ends of the spindles?
My involvement in this whole discussion, aside from the title, has
been mainly responding to your claim that my cranks break. When did
spindles get introduced?
>> Well maybe, but Kings Mountain road is a 7-8% grade at its steepest
>> and averages about 6%. Impact does not exist between metals and the
>> human physique. I think you are confusing impact with pulse loading.
>> This does not constitute impact with respect to structural metals.
> It was my understanding that you were the one that measured most of
> the grades in the listing of climbs - one entry is:
> Kings Mountain Road (at Archery Range Turn) 11.5%
That's a hairpin turn that one can choose to ride that way, but most
riders choose to ride around the outside where the grade is half that.
The inside radius it about six feet. That was a road center value
measured by Ray Hosler in his book about SF Bay Area bicycle rides.
He used a 100 inch aluminum rod and a decimal inch scale that measured
down to the street when the rod was leveled. That reads directly in
%.
>> Besides, it was these experienced bicycle mechanics that inspired
>> me to research and write on the subject of wheels when I heard the
>> myth and lore they repeated.
> Yeah, I see, everyone else is stupid. Got it.
Could you reconcile that with the information contained in that book?
Many of the proven items in that book are still not understood or
believed by many of these "experienced mechanics". You may want to
seep "everyone" into that category, but I don't believe you have any
reason to do so.
>> So? Do you propose that bicyclists start molding themselves to the
>> failings of the bicycles and components that are mis-designed?
>> It's the same as telling a hiker to take smaller steps to preserve
>> his hiking boots because the soles don't like to bend that much.
> There are steel cranks on the market as well as carbon fiber cranks.
> You ought to be able to find one that fits your needs Jobst. After
> 20,000 miles my Chorus cranks are going strong. No cracks. I just
> rode to work yesterday on my '82 Colnago with it's Super Record
> components. All original.
> Why do you break components and I don't? Is it your suggestion that
> the only variable is mileage? No? Is it because you ride over much
> more difficult terrain that I? Umm, probably not, though you may
> well do it more often.
I don't believe your bicycle experiences the amount of miles and the
terrain that mine does. In fact rider I know who is near my weight
and size and who rides similar routes, is Tom Ritchey. When out in
the sticks somewhere and I see road bike tracks, I have often guessed
whose they are.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 21 May 1998 01:09:51 GMT
Tom Ruta writes:
>> I don't believe your bicycle experiences the amount of miles and
>> the terrain that mine does. In fact rider I know who is near my
>> weight and size and who rides similar routes, is Tom Ritchey. When
>> out in the sticks somewhere and I see road bike tracks, I have
>> often guessed whose they are.
> Okay, I'll bite. How much do you weigh and where do you ride?
Tom Kunich knows where and it is to him that I made this reply after
he mentioned where I probably ride. The San Francisco Peninsula has
a coast range that is called the Santa Cruz Mountains. The highest
peak is Loma Prieta at 3808ft and there are many at 3000ft. As you
can guess, rides start at sea level. There are many local, county,
state and National parks in this region and in them many old logging
roads-turned-civilized climb steeply over many ridges. Some of the
roads are unimproved and some are downright un-maintained.
On the east side of the Santa Clara valley, Mt Hamilton 4209ft rises
gradually with a 6-7% service road to the Observatory and then descends
steeply and rises again over several ridges.
All these roads have great scenery and allow 100-130mile loops, and
of course shorter ones if one cuts the course so to speak.
In the summer I ride in the Alps every year and have done so for
a long time. The reports of this are in:
http://www-math.science.unitn.it/Bike/Countries/Europe/#Jobst
Other trips in the Sierra Nevada are listed in:
http://ng.netgate.net/~hosler/
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 21 May 1998 01:34:49 GMT
Keith Bontrager writes:
>> My experience with weldments is not good. Where there hasn't been an
>> OEM use of the product, I feel it's too risky to use unless there are
>> other compelling reasons to believe the product is superior. I don't
>> want to be the product tester.
>> I rode steel cranks by Magistroni long after Campagnolo aluminum cranks
>> were available because the early ones broke left and right at a high
>> rate. A few years later, when I could no longer find good quality
>> crank cotters, I switched to Campagnolo and in the following years was
>> able to avoid crashing by detecting most of the cracked cranks shortly
>> before separation. I took some dives. Two dozen cranks later I
>> switches to C-Record for 2000 miles or so and switched to Shimano, the
>> failure rate is way down but still there.
> I concur with your assessment of weldments - it has been our experience
> in testing and field observations as well.
> It does seem that there are stems you could take to minimize your
> risk if you choose to do so. It would be interesting, and a lot
> less maintenance for you.
> 1) I don't think the modification to the pedal spindle and threaded
> hole in the crank you proposed would be particularly expensive to
> execute. We have access to a machine shop and could probably be
> conned into making the modification with your guidance.
A 45 degree taper from the pedal shaft major diameter is sufficient.
I would also make split rings for retrofit to screw onto existing
pedals. The ring would be machined and threaded before slicing through
one side with a thin saw. This only works with pedals that do not
have a large thread relief, and most don't for good reason.
> 2) The fretting could also be reduced in other ways, with washers
> separating the pedal spindle flange from the crank outer surface.
It's not the fretting as much as it is the motion that skews the
thread in the pedal eye axis. Washers have been used in the past, only
to mask the damaging action. This joint is like a loose cylindrical
fit lined with threads. Screwing the pedal tight is like tightening
up the nut on the other end of a cylindrical shaft that doesn't have a
press fit. It squirms and frets.
> 3) The threaded hole, pedal eye region, or whole crank could be shot
> peened and re-anodized. This is overkill in your argument since the
> failures occur around the pedal eye only, but it is the more likely
> production method (the method we are using and the method Shimano
> employs on their XTR cranks), and it would not affect the experiment
> adversely to do the whole arm.
Hold the phone. I never said only. I have had cranks break in the
neck, the midspan, at the root and in the square tapered hole. The
most common failure has been at the pedal eye. Besides, don't anodize
it whatever you do. The problem here is that it is a dynamic joint
that should be static, or at least more static. I find the best way
to look at these things is to see it all as various hardnesses of
rubber. Everything deforms elastically.
> I might be able to schwag a crank set to anti up for this so you can
> keep riding. What crank model and arm length? No promises, but we
> do have some access to this kind of thing and I am interested.
> The latest Dura Ave cranks are designed to fit a splined tubular
> steel bottom bracket spindle. Is this the vintage cranks you are
> using?
I'm aware of these and had great hopes for them, but after looking at
them, carefully I don't see how they can endure, there being no press
fit on the tubular part and there is backlash in the stubby spline.
There can be little interference on the cylindrical taper because the
spline must engage before solid contact is made in the taper. Thus
there is no appreciable seating. It is this light press that is to
take up the back torque of standing on the pedals when they are
horizontal. I don't see how it can work. On top of that, no bike
shop can tell me of anyone who has used them extensively. From my
experience, there must be some strong endorsements before I want to
become a tester.
> If not, and if you want to stick with the same arms, I need a model
> number off the inner surface of the arm to make sure I get the right
> ones.
I'll have a look, they are 180mm DuraAce that fit on the standard
Campagnolo Record and Shimano square taper spindle interchangeably.
Thanks for your interest.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Crankset Safety? (Tight pedal damage)
Date: 21 May 1998 01:47:33 GMT
Rob D'Entremont writes:
>> What do you mean by overtightened pedals?
> Both cranks are damaged near the pedal hole. The damage is
> circular, concentric with the hole, and it's radius matches that of
> the pedal spindle "shoulder". The damaged area is recessed below
> that of the surrounding area by a small but noticeable amount. The
> edge is uniform and almost sharp enough all around for me to scrape
> my fingernail.
> I don't know what besides over tightening could cause this. The
> threads are fine.
All cranks develop this if the rider puts moderate force on the
pedals. This is a current discussion under the heading of rim
failures in this newsgroup. THis is a poorly designed joint and got
that way because at the time it became standard most bicycles used
steel cranks on which this gave no one cause to worry. On aluminum
cranks it is damaging and is a contributing cause in crank failure.
However, you should look for cracks radiating out from the bore to
determine if there is a hazard. The recess cased by this fretting is
standard. It is not caused by over tightening but is more likely
aggravated by under tightening.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 21 May 1998 15:47:27 GMT
David Rees writes:
> When you gave Campy and Shimano your cranks for replacement, what
> was their reaction, especially after you had broken a few?
I turn them in to the bike shop, the bike shop turns them in to the
distributor, and who knows where it goes from there. In any case,
there is never any comment. I have never met an engineer from the big
component manufacturers although I'm sure they exist somewhere. They
don't seem to come to the trade shows however. I know the ones from
Avocet, Ritchey and, formerly from Specialized, but Shimano,
Campagnolo, and Mavic remain obscure.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 21 May 1998 20:42:20 GMT
David Rees writes:
> Hell, if I worked for Campy I'd want to fly over, inspect your
> stuff, ride with you, etc., to get a handle on the problem and it's
> solution. As you've consistently broken cranks that last most people
> 20 years+, you'd think they'd be fascinated by the problem, even if
> a redesign is not a viable option.
As I said, the crank failure problem is swept under the rug and riders
are convinced that their failure is an anomaly by their dealers. In
fact they are intimidated to speak of it because they are seen as
gorillas who can't ride smoothly by their detractors, people who later
break a crank themselves and become converts a little too late.
Campagnolo knows well enough about crank failures because they get
plenty of returns to gather statistics. You see they even changed the
design of the C-Record to a four finger spider to get rid of the crack
between the crank and the first trailing finger. If Tullio were still
at the helm, I believe he would immediately change the pedal
attachment. That failure point is completely unnecessary.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 21 May 1998 21:32:59 GMT
crux writes anonymously:
>> Think about it. The rider's weight is on the pedal when standing,
>> regardless of what gear he uses. The cracks occur in lateral
>> bending mode. This big gear phobia is something like float in
>> pedals. It's a recent occurrence.
> But, as you've acknowledged in the past, a rider can apply a force
> to the crank that is larger than his weight.
Not for long. Climbing is hard enough without pulling up any
significant amount.
> Anyhow, doesn't this ignore the fact that some of us climb 17%
> grades while seated?
I don't see how it does. On the other hand the winner of the Mt
Washington NH race this year rode a middling gear at a 50-70 cadence,
common for good hill climbers on steep hills, much of it seated. I
used to climb most hills seated in the days before I started getting
mail from AARP. I also broke more cranks then.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech,rec.bicycles.misc
Subject: Re: Rim failures
Date: 22 May 1998 21:14:59 GMT
Keith Bontrager writes:
> This is interesting. Given this similarity, how do Tom's cranks
> fare?
I don't know. I know only that when he designed them we talked about
the problems and his point of view was similar to mine. That was why
he used the four finger spider to avoid cracking there and he put some
beef on the heel of the left crank to prevent the knuckle from
splitting. He is always weight conscious with all his components, and
there is where I was not so enthusiastic and have not used his cranks
because they are more slender in areas where I have had failures with
Campagnolo cranks. I am not eager to try new components that don't
appear to offer an advantage without good cause.
> I know a bit about his crank arms, made by Sugino out of their
> standard materials and processes, and they feature nothing
> remarkably different at the pedal eye in design or cross section.
> Does he have routine failures there?
I have heard from riders that they had failures with Sugino road
cranks the same as with others. I haven't tried them.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures
Date: 26 May 1998 22:02:26 GMT
Keith Bontrager writes:
>>> Bicycle parts, especially those which are designed to be light
>>> using materials with known fatigue characteristics, need to be
>>> replaced periodically. Aluminum handlebars, for example, need to
>>> be replaced to prevent catastrophic fatigue failure (which, let's
>>> face it, is not going to occur at a convenient time).
>> Oh? Exactly when do you propose one change these and where does
>> the manufacturer let the user know about this? This is not
>> analogous to anything in other consumer products, especially
>> considering that bicycle manufacturers don't offer a device to
>> monitor the aging process you propose the user should rectify.
> My guess is that some manufacturers do know how long their stuff
> lasts. We do. But, making the information available to the
> general public is not a simple thing, as some are fond of pointing
> out, like you are now.
> If I think an expert or pro off road racer needs a new set of bars
> every two years, and I do think that is appropriate for a well made
> bar, and have said so here and anywhere I thought someone would
> listen, how does one go about interpreting that for everyone else?
> It will make cycling more expensive to work out safe replacement
> intervals and not be wrong. That's clear.
You say that as though two years were a known amount of miles. My
experience is that time isn't such a hot measure of use. What's worse
is that the failure occurrence is erratic and some cranks last several
times as long as others before developing a significant crack. If I
were to replace at the shortest failure interval, I would lose
interest in this pretty soon from the cranks that broke in less than
2500 miles. The two main failure points are at the pedal and at the
spindle. Fortunately these can be inspected easily but spindles
cannot.
> There is also a thing I wrote up in the FAQ about service intervals.
-------------------------------------------------------------------
From the FAQ:
Of course, if you don't race, if you have more than one bike, if
you are a smooth rider, if you like to do "skyshots" you need to
work this in to the estimate. Getting tougher, eh? Many people
could ride on the good quality bars into the next millennium without
a problem. How do you sort it out? I don't know.
BB spindle. Hard to look at, but once a year. Look near the tapers
where the crank fits on. This is the weak spot. If the crank
feels funny when you are pedaling (hard to describe the feeling)
or if it comes loose unexpectedly, look long and hard at the spindle.
Cartridge BB's that allow you to change the bearings should be
treated with some respect. You can keep fresh bearings in them
forever, guaranteeing that they'll be in service until the
spindle fails!
Cranks. Check the right hand arm all around where the arm leaves
the spider. Also check the hub where the arm attaches to the
spindle - especially if the arm is machined from bar (CNC). The
section near the pedal threads was prone to failure on older
road cranks though I have not seen this on MTB cranks (yet!).
Look all over the arms on the light aftermarket cranks. Often.
Twice.
----------------------------------------------------------------------
That may sound like a working method, but my experience is that once a
year is not good enough to catch a fatigued spindle. I repack the BB
bearing at least twice a year and that didn't catch two broken ones.
I believe these parts should be built with as much margin as steel
forks are on better bicycles. They should be beyond failure for the
majority of riders. Today, forks of high quality fail in the blades
but not in the steer tube, although old racing stories abound with
broken steerers. Blades not breaking at the same time have been a
relatively benign failure for me on three occasions. They become
noticeable because one blade hinges forward. Steertubes on the other
hand do not reveal anything until separation, similar to cranks and
crank spindles.
>>> Cranksets, BB axles, cassettes/freewheels, brake and derailleur
>>> cables... and eventually frames and wheels... all need to be
>>> replaced periodically. If you put a lot of stress on these
>>> components, such as MTB XC racing or road racing, you'll need to
>>> replace them sooner.
>> I don't believe a word of it. You must be concocting this
>> theoretically from your keyboard. If not, how about giving some
>> values by which a rider should make these replacements to avoid
>> life threatening component failure. A failed crank, spindle, seat
>> post, handlebar or stem, or steer tube are all reasonably fatal
>> events if they occur at speed or in automotive traffic.
> Jobst - how about subtracting a month or two from the shortest
> service interval (in miles) you've had on cranks for a start!
> I recall citing this list a few months ago, and citing the
> same possible outcome of a failure, but getting my head handed to
> me when I did so.
> Sorry, but I am certain you were saying that in jest.
I am serious about the above. The randomness that I have seen in
these failures suggests that the manufacture is not well controlled
and that the interval is unpredictable. My heavy riding companion
should long ago have had some failures that I expect but he breaks
spindles while I break cranks.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures
Date: 26 May 1998 23:05:55 GMT
Tim McNamara writes:
>> We are not race car drivers and the replacement cost is significant
>> with respect to the operating budget, in contrast to a racing team.
> I wasn't referring to race car teams but to regular automobiles in
> private use, as well as commercial fleet vehicles.
>> know of no parts on a car that are prophylactically replaced if not
>> worn out
> Ever check a radiator hose or a heater hose to see if it is showing
> signs of weakness? And replace it to prevent having it blow out?
> Ever have your brake pads/discs/drums checked periodically? Or do
> you just wait until a loss of control of your car tells you
> something is wrong? Do you have your mechanic (or yourself if you
> have the knowledge) check whether your steering system is showing
> signs of wear? Personally, I'd rather replace these things *before*
> the wear has reached a critical stage.
That's what we are talking about, checking parts. Your suggestion to
replace them on some schedule is not the same as replaceing worn out
or nearly worn out parts. You cannot discover a failing crank spindle
unless you take it out and clean it every few hundred miles. This is
about once a week for me. Steer tubes are difficult to find cracks
even with disassembly because the bearing race can obscure a crack and
the crack may be under paint. Steer tubes do not have a shiny finish.
>> Oh? Exactly when do you propose one change these and where does the
>> manufacturer let the user know about this? This is not analogous to
>> anything in other consumer products, especially considering that
>> bicycle manufacturers don't offer a device to monitor the aging process
>> you propose the user should rectify.
> Nope, they don't, and there's probably no practical way to do so.
> What's your point?
The point is that the replacement by interval is a theoretical suggestion
without practical merit.
>>> Cranksets, BB axles, cassettes/freewheels, brake and derailleur
>>> cables... and eventually frames and wheels... all need to be
>>> replaced periodically. If you put a lot of stress on these
>>> components, such as MTB XC racing or road racing, you'll need to
>>> replace them sooner.
>> I don't believe a word of it. You must be concocting this
>> theoretically from your keyboard. If not, how about giving some
>> values by which a rider should make these replacements to avoid
>> life threatening component failure. A failed crank, spindle, seat
>> post, handlebar or stem, or steer tube are all reasonably fatal
>> events if they occur at speed or in automotive traffic.
> Well, this is the first time I've felt the infamous lash of Mr. Brandt in
> the six years I've been involved in this newsgroup.
Maybe you are insensitive to the low-balls I am fielding here. I am
accused of riding incorrectly, mistreating equipment and being unaware
that some simple methods exist to avoid such failures. When I return
these salvos to the originator in a direct manner, it calls out the
sympathy brigade. I know, that readers feel I should live by
different rules than my detractors, but the reader does not recognize
that I don't call people names they don't invent themselves.
> Sorry that not all of my ideas live up to your expectations. Well,
> no, actually I'm not sorry.
Make up your mind how you want it. I don;t care to trace your entire
thought process here. There is a backspace key on the upper left of
your keyboard that could spare us the details of your decision process.
> I really don't care whether you agree with me. You know more than I do
> and I'm happy to learn from you. OTOH, I think you're picking nits here.
That may be for someone who has not had mechanical failures but if you
realize that a broken left pedal, crank or spindle (right hand traffic)
when standing on acceleration, will throw you into passing traffic, it
no linger seems like nits.
> As far as values regarding replacement, I can't give you any. There are
> more variables than I can manage to deal with in such things- even if I
> know that a handlebar will crack at the stem junction at 1,000,000 cycles,
> I have no way of knowing how long that takes to occur. I know I've had
> parts break and that I don't replace them with identical parts; I look for
> more durable parts.
That is a bit inconsistent with your advice of changing equipment by
some interval of operation. First you suggest it and then you cant
give one for exactly the reasons there are none.
> I'm not an engineer, a physicist or a rocket scientist. A Mensan,
> yes, I have been that and I am reasonably intelligent; my education
> lies in other areas. Want to disuss the role of hippocampal
> deterioration and the loss of presynaptic acetylcholine in memory
> impairment associated with Alzheimer's disease? That I can do with
> some precision. In this arena, I can only discuss what seems to me
> to be reasonable general principles. Obviously they didn't seem
> reasonable to you.
And what am I to do with the jargon you offer? I don't think it
improves your credibility in relegating mechanical failures to nits or
proposing ways of replacing parts. Occasionally contributors post a
pile of arcane jargon or append academic achievements to postings here
only to make the technical "advice" given more even more suspect.
> As far as "not believing a word of it," Jobst, you yourself have
> posted such statements as riding the same pair of steel handlebars
> for years because of the fatigue characteristics of steel vs.
> aluminum. You have commented that bicycle frames have a limited
> useful life of (IIRC) about 100,000 miles or so (which means my bike
> ought to last about 20 years or so; this is the good news).
I think you made that up. I have ridden frames as long as 200,000
miles but they were wrecked. I have no number on how long a frame
should last. Besides, I don't have an of-the-shelf frame anymore, now
that road frames are made so short and with close clearances that they
cannot be ridden on dirt roads.
> Seems to me that you do in fact believe it and modify your own
> practices as a result. You also obviously know that bike parts wear
> out, even though you profess not to believe a word of it. Are you
> saying that everyone should just ride until parts fail and the
> consequences be damned?
I think you are dodging. The subject you proposed was to change parts
on a schedule to avoid failures. That worn out parts or ones that no
longer function wheel are replaced is not something new. This has
been common practice since the dawn of machinery. Even old draft
animals have been put to pasture before they die.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures
Date: 27 May 1998 00:17:43 GMT
Mark Hickey writes:
>> This is common bread and butter in most industries, but I assure
>> you, few skilled engineers are working on these problems because
>> there is a vocal majority of bicyclists who demand lighter and more
>> fragile equipment. They also don't ride the stuff much in many
>> cases. That's not bad except that durable equipment is often
>> discontinued because there is such an imbalance in demand. Its not
>> so much that the good stuff is no longer bought but more that the $
>> volume of the esoteric gear overwhelms the market (place).
> Let's face it, the marketplace is a democracy, and apparently you're
> not happy about who is getting elected.
How about "market" we don't all have to invoke the mod babble of the
press. "How are you impacted by the marketplace?" Sounds more like a
guy who took a dive off his bike. How are you affected by the market?
> Personally, I stay away from "stupid light" parts, too.
> Unfortunately, not everyone has the ability to discern what they can
> and can't get away with.
Oh? And where do you find the standard formerly durable equipment? I
am at a loss to find much of it anywhere. Most bicycle shops have
found it easier to throw it in the trash than to let it displace the
new lighter more expensive fashion equipment from shelf or storage
space.
>> An example is the rave of machined rims offered these days. There is
>> NO excuse for machining rims. None! Machining just costs extra and
>> guarantees that the wall thickness is unknown and variable... and it
>> costs more. About 1/3 the life of the rim is shaved off beforehand.
> I couldn't agree more (and have said the same thing for years).
Just you watch as the non-machined rims vanish. We have had many
contributors on this newsgroup extol the virtues of machined rims, or
for that matter ones with hard anodizing or ceramic coatings.
>>>> I also don't see why there would be a huge penalty in terms of
>>>> weight or expense to design parts properly.
>>> The laws of physics are not negotiable, unfortunately.
>> Oh don't hide behind the law. This isn't rocket science, so to speak.
> I'd be happy to discuss it - not hiding at all. The part you snipped
> was my contention that after using better material and technology
> (i.e. butting) handlebars, you soon reach the point where you have to
> make it heavier to make it stronger.
Hydraulically expanded TTT bars broke regularly until the word got out
that this was dangerous. Now most bars are sleeved in the attachment
area from what I have seen.
>>> To add strength, you have to add material, or make better use of
>>> the material.
>> How about the latter?
> Agreed - but again, you reach a point of diminishing returns and start
> making parts heavier to make them stronger.
It hasn't been done in many places on the bicycle. Unfortunately the
effort is expended mainly by designers who are trying to save weight.
They redesign the item to be lighter and thereby are forces to use the
material better just to stay even with the prior designs.
> That becomes a rather selfless pursuit before long - more expensive,
> heavier parts that far exceed the requirements of the vast majority
> of the riders in the world.
Who mentioned anything about "far exceed". I know enough bikies of
average weight who have broken cranks and they don't ride as much as
I. The way you talk, I take it you are a manufacturer defending the
status quo. If not, why do you do it?
> There's a market (HP engineers, for example) but it's never going to
> be a money-maker (something that seems to be a priority with many
> manufacturers for some reason).
Oh how clever.
> OTOH, I *do* believe in over-engineering, and would wager that even
> Jobst couldn't "wear out" one of my frames (even though it would
> probably be lighter than what he rides now, and probably less
> expensive).
What do you mean by "over engineering"? Must we wait for Shimano to
solve all these problems by hit or miss marketing? Maybe a large
lawsuit will come up that puts focus on this, somewhat like the
national helmet "save a child"crusade we have seen all last week.
I think gun control has a better chance, judging from the defense of
faulty designs I have seen here on this newsgroup.
>>> To be honest,
>> In contrast to what? Have you been less than truthful in previous
>> utterances?
> Well, to be honest....
Yes, what exactly do people mean by interjecting that phrase at the
end of some presentation? I hadn't thought that veracity was an issue
up till then. Is it?
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 29 May 1998 16:42:32 GMT
Rick Denney writes:
> I've been trying, though, to think of why Jobst's experience is so
> skewed in the direction of pedal eyes, instead of the more typical
> failures. I understand his explanation of the forces involved. I can
> only think of two things that might stress that part of the crank
> abnormally. One is if Jobst's style involves substantial leaning of
> the bike from side to side while climbing. I don't think this is it,
> because as I recall he mentions that he climbs mostly in the saddle.
> But if his ankles resist the lean, it might add to the bending
> moment at the bottom of the crank.
Let's not conjecture on this. Today I stand a lot on hills but did
not do so 10 years or so ago up to which time I broke most of the
Campagnolo cranks because I rode more and possibly because I rode
Campagnolo toeclips pedals that had more overhang than appears to
lever on SPD pedals. Although the rate of failure has decreased, the
mode has not changed and I regularly inspect for cracks.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures
Date: 30 May 1998 00:18:51 GMT
After taking some measurements of the spindles of my 535 SPD pedals,
it turns out that the thread relief is late enough to accept a split
45 degree tapered ring and that countersinking the crank appropriately
will be easy to do. The trouble is that this is unique to these
pedals, the thread relief not having a diameter or length tolerance
that permits this to be a generic solution. I will try this out but
not on my trip to the Alps. The results, of course will not be known
for at least a couple of years, because that's about how long my
DuraAce cranks have lasted in contrast to the Campagnolo Record cranks
that broke in about a year.
One of the comical pieces of advice I got on this in the past, from
racers who by their ranking felt they were in the know, was that I
should not used 180mm cranks, because the extra leverage breaks them
more easily. That the length of the crank has no effect on pedal eyes
breaking out was beyond the perception of the experts.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 4 Jun 1998 02:06:34 GMT
Dave Blake writes:
>> You break cranks. I don't; never have. If my chain breaks, I'll
>> probably be able to retain control of my bike, but I can't remember
>> ever breaking a chain.
> When you break a chain, your feet are still supported. In the worst
> case they begin to spin madly, but your center of balance is still
> the same (mostly). A broken crank is a bird of a different
> feather, since the one foot is heavily weighted and free falling, and
> close to a sharp piece of aluminum.
Well I don't concur with that scenario. Regardless of whether a chain
breaks, a three speed pops out of gear or the freewheel releases, you
go over the bars if you are standing with any pedaling effort. The
point is that you are standing over the leading pedal and suddenly it
is gone. I have done it and so have others. More telling is that
people unfamiliar with fixed gears on a track bike get thrown, even
when sitting, if they forget and decide to "coast" with moderate to
high cadence... and I don't mean their feet come out. I mean they get
thrown from the bike.
When a crank breaks, and this usually happens at the bottom of the
stroke, the rider is suddenly standing on the road at a speed at which
running would be difficult and is even more difficult with one leg and
a bicycle between the legs. That is why one usually crashes when a
crank breaks. A chain failure rotates the downward (load supporting
foot) past the bottom of the stroke so that it lies behind the rider
almost instantly, and unsupported, the rider literally flies forward,
the other leg not being under the load and not in any condition to
press downward, there being no resistance from the pedal. It's like
stepping into a deep hole while sprinting.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 8 Jun 1998 16:48:58 GMT
David Bittner writes:
> Of the two dozen cranks you have broken, which arm fail most frequently?
Both, since most of the failures were at the pedal eye. One left
crank broke at midspan and one at the neck. To make up for that a
couple of right cranks developed cracks near the spider.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures
Date: 8 Jun 1998 18:22:39 GMT
Paul Grammens writes:
> When I bought my first real bike around 1970, I broke the pawls on a
> couple of freewheels. The shop replaced them, but the second time the
> mechanic suggested I was pedaling too roughly. I thought about it and
> realized I was banging the pawls against their stops when going from
> coasting to pedaling. I changed that, and haven't broken a pawl since.
Well that may be what they got you to believe. There is no way the
impact of engagement of any freewheel, including the 16 tooth
ratchets, can be damaged that way. Most better freewheels have 18 to
36 clicks per revolution today, not quite as good as the 42 clicks of
yore with Regina. Loading in that manner is insignificant.
> In that respect, I disagree with you, but you are correct that crank
> cracking and breakage is life-threatening and should be very rare.
> My Nervar crankarm broke at the ideal time: I was riding up to my
> car and had my weight on my right foot. As I came to a stop, the
> crank broke and I was standing on the ground, pedal and crank stub
> strapped to my foot!
From this you might not immediately notice that the crank broke at the
bottom of the stroke from a lateral bending load, not from hard
pedaling. Most cranks break this way and therefore, the conjecture on
why cranks break is misplaced, especially by those who have not broken
cranks and feel they are somehow doing things right.
> I thought about the times I had stood on it to muscle over a rise in the
> road, going 30 mph or so. Ulp!
That such a failure would be hazardous is obvious. My first encounter
with this was on an early Campagnolo crank that failed just after I
sat down while sprinting down a steep curvy road. I switched back to
cottered steel cranks for several years.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 9 Jun 1998 21:07:01 GMT
Jerry Gardner writes:
>>> Of the two dozen cranks you have broken, which arm fail most frequently?
>> Both, since most of the failures were at the pedal eye. One left
>> crank broke at midspan and one at the neck. To make up for that a
>> couple of right cranks developed cracks near the spider.
> How many bones have you broken as a result?
None. In fact not even a raspberry. The ones that weren't discovered
by inspection occurred while starting or riding slowly, most were
discovered on inspection. Only the first one, one of the first
Campagnolo cranks sold to the public in Italy failed at high speed
while seated riding down from the Eibsee near Garmisch. The steep
climb to the lake had apparently done it in. It lasted about 200
miles. I returned both cranks at their request.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 10 Jun 1998 00:18:41 GMT
Tom Kunich writes:
> I've not had the chance to ever look at pedal spindles that have
> failed and so I don't know it they've been overloaded or if the ones
> that fail do so due to some manufacturing flaw. But it they do fail
> from overload this is certainly a hint that one could lever a crank
> pedal eye open with enough force.
They fail in fatigue and do so in torsion, in classic failure of a
square cross section. The cracks initiate in the middle of the flat
(half way between the corners of the square), almost uniformly from
all four faces. This may be due to having been rotated to a different
orientation during maintenance. Nearly all torsional force is
transmitted by the first 2-3mm of the large end of the press fit, that
being the area that shows fretting wear.
The cracks initiate in the middle of this fretting area that is
symmetrical to the point that you cannot tell what the direction of
the applied torque is. This shows that with the press fit, torque is
transmitted by differential loading, one half the face becoming less
loaded the other becoming more loaded. The fretting action is the
same for both events. The mid zone shuffles back and forth, being
tangent to the motion and transmitting no torque and is the widest
part of the wear track. The nose of the press, the largest part of
the taper that is in contact, shows no wear at all.
The resulting separation has a four pointed helical drill like face,
with the corners of the spindle breaking last since they are least
stressed and are the ones that ultimately experience a forced rupture.
Torsion failures generally have 45 degree features in them because
that is the orientation of the principal stress in a shaft.
> This is my suspicion.
I am not speculating on this.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Rim failures and other stories
Date: 10 Jun 1998 23:43:51 GMT
Tom Kunich writes:
> After carefully reading your posting several times I can't say that
> I have seen these types of failures _precisely_ as you describe.
> The BB's I saw failed by starting a crack at the junction of two
> faces of the taper at the inside edge. There was a half circle crack
> if memory serves, in several cases and then a more or less straight
> crack radiating in a counter-forward direction outward towards the
> end of the spindle at about the 45 degree angle you note. None of
> these failed spindles had gone to total failure. I asked the
> mechanic if this were typical failures and he pulled out a couple of
> more in the same condition and they looked virtually identical. I
> really didn't pursue the subject since it appeared to be a
> completely known failure mode.
If I understand what you say, that isn't the way I described the failure.
>> They fail in fatigue and do so in torsion, in classic failure of a
>> square cross section. The cracks initiate in the middle of the
>> flat (half way between the corners of the square), almost uniformly
>> from all four faces. This may be due to having been rotated to a
>> different orientation during maintenance. Nearly all torsional
>> force is transmitted by the first 2-3mm of the large end of the
>> press fit, that being the area that shows fretting wear.
>> The cracks initiate in the middle of this fretting area that is
>> symmetrical to the point that you cannot tell what the direction of
>> the applied torque is. This shows that with the press fit, torque
>> is transmitted by differential loading, one half the face becoming
>> less loaded the other becoming more loaded. The fretting action is
>> the same for both events. The mid zone shuffles back and forth,
>> being tangent to the motion and transmitting no torque and is the
>> widest part of the wear track. The nose of the press, the largest
>> part of the taper that is in contact, shows no wear at all.
As you see, the cracks do not originate at the junction of two taper
faces, but rather in the center of the faces in the smallest inscribed
circle of the contact with the largest part of the crank bore.
> What I was getting at was that it might be a clue to what your failures
> are if pedal axles fail in straight torsion, since a pedal axle is
> stronger in cross section than the end of a crank eye. If a person
> could break a pedal spindle he could most certainly fatigue the end
> of a crank.
Pedal spindles have no torsion. Even with failed bearings, torque is
insignificant, but you can't ride that way long enough to do anything
but have the pedal fall off the end of the spindle.
> I understood you to say that you generally catch these failures before
> they are catastrophic and that you probably have examples of the failed
> crank. These might give you a clue to the failure mode since I would
> expect there to be thread damage along the latitude due to the levering
> action I was talking about. Have you seen anything like that? Or is the
> damage already beyond tracing by that point?
There is no way of inspecting for a failing crank spindle because the
failure occurs inside the press fit and constantly removing it is
asking for trouble. The cranks can be inspected and these are the
ones that I have caught before separation. The are unexciting cracks
that go straight across the crank at the diameter of the pedal eye, as
I mentioned, from bending loads of standing on the pedal at the bottom
of the stroke. There is no thread damage, even with ones that broke
off.
> Again, I'm not saying that that is your problem. I am just suspicious
> that your failures seem to be non-typical of these sorts of things
> these days. And I'm told that in the old days when everyone rode the
> type of gearing you do, that those sorts of failures were common.
Your theoretical pursuit of this topic without any statistical
evidence seems to me to be harping on a matter to which you can't
offer much. What is the goal of this discussion? "I am told" and "I
suspect" is not much good to me. I rode in "those" days and I am
riding now. I was a mechanical engineer then and am so now and I work
in that field, although not on bicycles as an occupation. Your
conjecture is not uncovering anything new and is largely off the mark,
since you have not observed these failures.
> On the other hand, those older cranks weren't as well designed as
> today's equipment and that clouds the issue somewhat.
They were no different than the ones today other than materials
available. Although crank failures were rarer with steel cranks, they
also broke. I have a cracked Magistroni crank as does my friend whose
broke at midspan from a material flaw.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Bicycling magazine tip #13: tighten crank bolts
Date: 30 Sep 1998 16:57:50 GMT
Gordon D. Renkes writes:
> When I read Jobst's dire diagnosis of this incurable situation, as
> my commute bike showed these symptoms, I uttered my favorite
> 4-letter word I learned on the farm. My right crank had worked
> loose beyond creaking to wobbling.
> Before spending money, I decided to try to remedy it this way. I
> cleaned the tapers, and the bolt threads thoroughly, with solvent.
> I scrubbed the tapers with sandpaper, to smooth any burrs. The
> tapers were greased before reinstallation. I put Locktite 242 (the
> kind which is designed to prevent loosening, but allows the
> connection to be loosened with a wrench) on the bolt threads. I
> tightened them quite tightly with a socket wrench, (but not to the
> point of sounding like Sheldon).
> That was 2 or 3 years ago. I have ridden 2-3000 miles on that bike
> since, and pulled and replaced the cranks 2 or 3 times in the
> meantime during annual maintenance. They have survived the stop and
> go of commuting quite well.
A crank ridden loose becomes distorted so that the formerly square
tapered bore takes on a pin cushion shape (like the periphery of a
square pillow) on its flats. If this hasn't progressed far, there
will still be partial flat areas in each face that are rounded by the
corners of spindle contact.
Re-tightening can bring the reduced faces into contact again but the
most important bearing area has been permanently lost. If such a
crank is ridden forcefully up a hill or in sprinting, it will rotate
slightly on the spindle and loosen the screw, because it is only
partially supported.
If cranks had a substantial margin with the original bearing area,
they would most likely not loosen even with the most cursory
tightening by a mechanic. However, there is sufficient torsional
motion between a new crank and spindle to cause fretting wear that is
visible as rouge when disassembling cranks used by bicyclists who do
more than commute on flat streets and who stand up and sprint on
occasion.
Torque is transmitted effectively only by pressure between the flanks
of the four faces and the spindle, not by friction, and it does this
by loading and unlaoding these flanks, not only by increased pressure.
Therefore, just tightening a crank whose bore has been deformed will
not keep it from rotating, only the residual flat can do that.
Because the strength of this joint has been substantially compromised
by active use in the loose condition, the statement that the crank is
ruined is accurate with the exception that if it is not heavily
loaded, one can get further use from it. Not all riders demand the
full capability of a good crank to spindle fit.
The crank cannot be reformed to fit the spindle by tightening the
retaining screw. The screw is too weak to do this and cold forming
the crank requires that the the yield stress be exceed, something that
would soon cause a cracked crank knuckle.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Bicycling magazine tip #13: tighten crank bolts
Date: 30 Sep 1998 22:44:57 GMT
Eric Salathe writes:
> While I don't dispute the above scenario, I do not agree that
> disabling damage is assured nor that a rider using a once-wobbly
> crank does not demand full capability. As I mentioned above, I have
> had a crank become perceptably loose, one-footed it home, and
> aggressively tightened it. I have since ridden the cranks well over
> 10,000 hilly miles including twice riding RAMROD (a local 154mi,
> 10,000 vertical-foot timed event) and hundreds of out-of-the saddle
> climbs up 20-25% grades around Seattle in a 40-inch gear. I have
> been as demanding of these cranks as my weight permits, and they
> give no indication of rotating on the spindle.
Well then you didn't ride it hard while it was loose. However, this
scenario is worse than I thought. If you weren't riding hard, how did
it come loose? The only thing I can visualize is that it you forgot
to tighten it and rode off only to discover that fact. If it worked
loose from an even moderate press fit, it should have sustained major
damage.
> However, I weigh only 150 lbs and, as I said, I did not use the
> crank after becoming aware it was loose. Also, these are Campy G.S.
> (unpolished N.R.) cranks, so clearly I do not exert the kind of
> force on my cranks as heavier riders. I cannot dispute that a 200-lb
> rider might cause the cranks to rotate on the spindle. Nevertheless,
> I cannot accept the categorical statement that, if they become
> loose, `your cranks are ruined.' This is not to any degree true for
> lighter riders who cease riding the crank immediately upon the first
> indication that it is loose. Furthermore, continuing to use them
> poses no danger since the failure will be merely back to the wobbly
> state and not catastrophic.
You may be right there. I know only of bike racers who failed to
tighten their cranks properly and subsequently had them come loose
under load. These cranks were all ruined. After tightening they all
became loose again after short use for the reasons I described. If
you don't pedal hard at the time the crank works loose, the damage
will not be great and if you don't weigh much, the demand on the
crank- spindle interface also will not be great.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Lifetime of cranks
Date: 21 Jan 1999 17:02:59 GMT
Sergio Servadio writes:
>> Among the dozens of Campagnolo Record cranks have broken for me,
>> most broke at the pedal thread, some at the neck at the pedal end,
>> some at the root of the spider. None of the Campagnolo cranks
>> showed cracks at the spindle but that may be only because they
>> didn't last long enough to develop such cracks. DuraAce cranks
>> have lasted many times longer but a couple of these have developed
>> cracks at the spindle bore although none have gone farther than
>> that.
> Why have my cranks lasted so long? Since 1973-74 I have almost
> exclusively used two custom built bikes both equipped with
> Campagnolo Record. Of course, the cranks have many times been taken
> off to service the bearings and put back in place by expert hands.
> Quite remarkably they have been gnawed by the shoes, to testify
> their long life. But never did they fail. To be sure I am not at
> all a strong rider by a racer's standard, but my bikes have taken me
> up the hills an 'infinite' number of times.
I weigh 185 lbs (84 kg) and ride a lot in the mountains. In fact most
of my riding is hills both here and on my vacation in the Alps. One
of my riding companions, Ray Hosler had the same question as you about
his Campagnolo Record cranks, until about a month ago when his broke
off and he took a dive. He weight 135 lbs.
> So, what's the crux of the matter?
> Is there a minimum torque below which any given crank will last a
> truly infinite number of cycles (contrary to common expert wisdom)?
> How does the cyclenumber-life depend on the peak torque? When are
> mine going to fail (if ever)? Shouldn't the designers make them a
> little more durable to all?
As I point out in the item, cranks most frequently fail in lateral
bending at the pedal eye. It is not the torque or even the twist in
the crank, but the bending moment of the outboard pedal at the bottom
of the stroke that breaks them off. Essentially it is standing on the
pedal when at the bottom, such as standing on one foot in curves or
stepping off from one foot to the other while pedaling standing.
Cursory inspection is easy. Just look for cracks around the outside
face of the pedal end of the crank in sunlight. Had my friend taken
that precaution, he would not have fallen. The cracks take a few
hundred kilometers to reach failure.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Crankarm-->bb question
Date: 27 Feb 1999 21:56:18 GMT
Jon Isaacs writes:
> In Jobst's explanation of the crank on the spindle I believe I
> understood that the crank is never actually "tight" that it always
> works a bit.
I didn't say they are not tight, only that metals are elastic and two
metals, with a 3:1 elasticity ratio, fret under stress. That is they
move just as a baseball bat moves in your grip, the wood being far
harder than the flesh of your hands.
> I am curious how some cranks can be so hard to remove if they are
> never really "tight."
I could use the same analogy. I can hang on to a baseball bat and you
would have a hard time pulling it through my fingers, but that does
not mean the bat does not creep in my hands when I hit a ball with it.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.racing
Subject: Re: BEWARE of Sampson cranks!!!
Date: 16 Apr 1999 17:17:51 GMT
Chris Bruce writes:
> About a year ago, I got a good deal on a Sampson crankset from a
> shop that was about to go out of business. Well, turns out it
> wasn't such a good deal after all. Yesterday when I was riding to
> work, the right crank sheared in half where the pedal inserts.
This is the most common failure location for cranks and it occurs more
often than you think. A few years ago, when I reported crank failures
the response was about that, which is common today for not wearing a
helmet. No one would admit to a crank failure because it was
generally attributed to bad riding technique by bicycle shops rather
than a design flaw. A thread with a flat shoulder, as a pedal has, is
a terrible joint for a rotating outboard load.
These crank failures take a while to develop and can be discovered in
advance by inspection every 500 miles or less. It only takes a glance
in direct sunlight to detect such cracks before separation because
they take about 1000 miles from inception to failure. I have detected
a few on my bicycle although in the "old days" I had more than two
dozen Campagnolo Record crank failures, many of which were detected
before they broke off.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Record crank trashed?
Date: 27 Mar 2000 19:48:27 GMT
Tim McNamara writes:
>> Where can I buy these dust caps you're talking about here that fit
>> close enough to the bolt to keep them from unscrewing and causing
>> the crank to be damaged? Seems to me even if the crank bolt
>> unscrews even 1 turn, they no longer are holding the crank arm in
>> place so these caps you use must fit REAL close to do what you are
>> saying here.
> If I understand it correctly, the crank bolts don't unscrew. The
> crank squirms up the crank spindle, away from the crank bolt- with
> the result that the bolt is now loose. The dust cap merely serves
> to keep the crank bolt from falling out.
> This crank-squirming-up-the-spindle action is a little counter
> intuitive and I, for one, don't understand how this can happen.
Imagine that the crank and spindle are made of two different grades of
stiff elastic material (they are exactly that) and how the crank
deforms under load, bulging at compressive areas and shrinking at
tensile (or unloaded) areas around the press fit. The bolt has a
preload against one face of the crank and no restraint on the other.
On installation, frictional resistance prevented the crank from
sliding onto the taper as far as it would have were the interface
frictionless. When the crank deforms under load it relaxes the
frictional shear load on the spindle by moving farther up the spindle,
until the axial load on the bolt equals the press fit times the slope
of the taper. That's a lot less than the initial tension on the bolt.
Beyond that, the bulging of the crank at the four increased pressure
points puts more push on the bolt, again the other side of the crank
having no constraint, leaving it free to move even farther up the
spindle.
If the bolt was insufficiently tight in the first place, it can become
completely free from this action and subsequently unscrew small
fractions of a turn with each rotation of the crank. Proof that bolts
fall out exists abundantly. The interesting part is visualizing the
mechanism. I think it is apparent to most observant mechanics that
crank bolts are not as tight as when installed after active use. This
alone should make the phenomenon clear.
> I sort of think it has something to do with the pedalling force
> "rocking" the crank at the bottom of the pedal stroke, because the
> pedal acts as a lever normal to the plane of the crank arm. Is this
> correct?
It does this in the horizontal position as well.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Campagnolo Record crank fatigue prevention/cure?
Date: 22 Mar 2000 17:15:34 GMT
James Thomson writes:
> I have two pairs of original Campagnolo Record cranks which I would
> like to use on two vintage bicycles I'm restoring. One pair appears
> sound, while the right hand crank of the other pair has the
> beginnings of the classic fatigue cracks between the crank and
> the spider, extending about 1mm into the crank.
I have broken more than two dozen cranks, mostly Campagnolo Record,
over the years, most of which broke in the pedal eye, some at the
neck below the eye and one at midspan and a couple at the place you
mention... BUT these did not break at the crack of which you speak.
A new crack developed a few millimeters away and propagated straight
across the crank. These did not go to failure but were obviously
beyond safe use, having separated more than 1/3 the cross section.
The web crack that your crank has would best be removed by using a
round file and polishing it with a wooden dowel with fine sand paper.
I always threatened to do that but the cranks broke before that crack
ever got to a worrisome depth. My C-Record cranks, in contrast did
not have this web crack because they were properly designed with
respect to this problem. However, their pedal eyes lasted slightly
more than 2000 miles, dumping me on the road long before I even
considered inspecting for cracks. I had just finished 2200 miles in
the alps having just purchased them a few weeks before they both broke
after the trip, one breaking off, the other having a crack all the way
through one side of the pedal eye.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Campagnolo Record crank fatigue prevention/cure?
Date: 22 Mar 2000 22:38:20 GMT
Russell Seaton writes:
>> My C-Record cranks, in contrast did not have this web crack because
>> they were properly designed with respect to this problem. However,
>> their pedal eyes lasted slightly more than 2000 miles, dumping me
>> on the road long before I even considered inspecting for cracks. I
>> had just finished 2200 miles in the alps having just purchased them
>> a few weeks before they both broke after the trip, one breaking
>> off, the other having a crack all the way through one side of the
>> pedal eye.
> You rode Campagnolo Record cranks 2200 miles in the Alps? Not
> loaded touring? Record uses a 135mm bcd so the smallest ring is a
> 39. Coupled wth Campagnolo's biggest rear cog, a 26, you have a 40
> inch low gear. I've ridden in the Alps, Dolomites, and Rockies. 40
> inches is not low enough for me.
> I'm about 190 pounds. I have ridden 1998 Chorus cranks for about
> 4-5,000 miles. No broken pedal eyes or spiders. Rode/ride the
> 1991 Shimano Deore DX cranks for several tens of thousands of miles. No
> broken crank. I had to pedal a pretty slow cadence up some of those
> mountains. Didn't have any lower gears.
What gears you use is up to you, I am more aerobically limited than
strength and use a 46-50 with 13-15-17-19-21-24. When I was "young
and beautiful" I used a 47-50 because I started on 5-speed half step
and never changed, not using the small chainwheel but rarely.
the bike and rider are at:
http://www-math.science.unitn.it/Bike/Countries/Europe/Tour_Reports/Tour_of_the_Al
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Crank Life?
Date: 14 Dec 2000 21:18:19 GMT
Shawn Amir writes:
> What is the consensus for the life span of a crank--lets say a DA?
> I would like to avoid being neutered by my top tube when I stand.
Having broken many cranks, I can safely say that they do not break
suddenly but develop cracks slowly enough that visually checking them
at 200 mile intervals is adequate to prevent crank separation on a
ride. Most of my cranks were detected that way.
The place that requires careful inspection is around the pedal eye
where a crack can be detected under bright light. Remember, all
cracks are "hair line " cracks until they separate or are ridden n the
rain so that grey juice is generated. In fact, when water gets into
such a crack it sounds like chewing on gristle as the fissure is
pressed closed.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: alloy failure
Date: 4 Jan 2001 17:57:32 GMT
Sue Rogers writes:
>> What alloys are you interested in.
> The alloy used to make Shimano 600 cranks. I had a cycling accident
> recently where my front wheel slid away at very low speed. I came
> down very heavily on mine and the bike's right hand side, thereby
> giving the end of the pedal a good whack on the tarmac cycle path.
> Having had two alloy cranks (one a Shimano, the other probably a
> Campag) break on me back in the 1980s when I did a lot of cycle
> racing (and crashing), I know how sudden the fall from a broken
> crank while pedalling is and there's no way I want to risk that
> happening.
I think your fear is unwarranted because cranks do not break suddenly
and without "warning", the warning being that a crack has developed
through more than half the cross section. I have broken more than two
dozen Campagnolo and Shimano cranks of which a couple broke suddenly.
On inspection it was apparent from the face of the crack that it had
been there for a long time judging from oxidation. Subsequently I
inspected for cracks at reasonable intervals and detected cranks that
were no longer safe.
If you cannot see a crack under bright light, and the crank is not
bent, there is no reason for replacing it. Cranks break from metal
fatigue and a crash (single load cycle) does not affect fatigue if it
does not bend the crank.
> Therefore I am replacing the cranks on this bike before using it
> again. (I no longer cycle race, just commute). As the accident
> wasn't my fault, I'm looking to claim damages off the other party
> involved and need to prove that it's not a good idea to whack the
> end of alloy cranks onto concrete with 11.5 stone of rider on top.
That may be your perception but it is not the way metallurgy works.
> Hence my request for any official literature on alloy crank failure.
> I have the 1985 Cycling Weekly article on crank failure, written by
> a metallurgy lecturer at Derby University, but if there's anything
> more up to date, I'd appreciate it.
I don't think you'll find support for your contention and for the sake
of those who pay insurance premiums, I hope you don't press this case
further.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: alloy failure
Date: 5 Jan 2001 18:06:07 GMT
Jay Beattie writes:
>> I think your fear is unwarranted because cranks do not break
>> suddenly and without "warning", the warning being that a crack has
>> developed through more than half the cross section. I have broken
>> more than two dozen Campagnolo and Shimano cranks of which a couple
>> broke suddenly. On inspection it was apparent from the face of the
>> crack that it had been there for a long time judging from
>> oxidation. Subsequently I inspected for cracks at reasonable
>> intervals and detected cranks that were no longer safe.
> To Sue and Jobst: which Shimano cranks and where did they break? I
> have never broken a Japanese crank and always thought the Shimano
> were bullet proof. I am crestfallen.
At the pedal eye and at the web between crank and spider finger as
well as at the corner of the square hole where the Dura Ace crank
presses onto the spindle. Most failures (Campagnolo & Shimano) were
at the pedal eye, something that I fixed a few years ago by counter
sinking the thread 90 deg and installing a split collet to immobilize
the pedal spindle that otherwise frets (the reason we need left hand
thread on left pedals) and initiates cracks. I don't look for cracks
once a week anymore.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: alloy failure
Date: 8 Jan 2001 18:35:59 GMT
Bill Putnam writes:
>> Most failures (Campagnolo & Shimano) were at the pedal eye,
>> something that I fixed a few years ago by counter sinking the
>> thread 90 deg and installing a split collet to immobilize the pedal
>> spindle that otherwise frets (the reason we need left hand thread
>> on left pedals) and initiates cracks. I don't look for cracks once
>> a week anymore.
> Do you have a drawing of this you could post?
The problem is that it differs for each pedal spindle and may not even
fit without modifying some spindles, depending how the thread relief
is formed. For my SH-SPD pedals, the thread relief is cylindrical and
12.45mm in diameter with an OD of 18mm. The collet ring has a half mm
cylindrical edge at 18mm diameter, is 2.3mm thick and tapers at 45
degrees. The ID has a half mm chamfer where it butts against the
spindle (for the spindle fillet). The collet is split in two using
the thinnest reasonable saw.
The crank must be carefully countersunk to 18mm diameter exactly
concentric to the thread. When the pedal is screwed in firmly, only
the half mm of cylindrical portion of the collar is visible along with
its split. Ideally all pedals would have tapered faces on the spindle
to fit into tapered crank faces and we wouldn't hear much about pedals
coming loose anymore, right or left and crank failures at this place
would most likely vanish entirely.
The bicycle industry can't be bothered. Their opinion is that there
is no problem.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: alloy failure
Date: 8 Jan 2001 23:53:14 GMT
Karl Nelson writes:
> Despite the problem of pedal differences you mention, I would like to
> also request a drawing. I think if I could see a drawing, I could see
> what modifications are needed for my pedals. It would also be helpful
> to see a picture of the spindles that it works on and and example of one
> which would have to be modified so I could verify that as well.
> Did you design the collet on a computer (and have a CAD file you could
> post) or did you work from a hand drawing? If you don't have any
> drawing you could post, even a close-up digital photograph would be
> better than nothing.
The drawing exists in ME-10 2D Mi file and in DXF (AutoCAD drawing
transfer format). If you like I could send you the file in either
format and as I said the inside diameter is specifically for my
pedals. You would need to measure yours and make sure that the thread
relief is already cylindrical or you must machine it to be so.
> Even if the industry doesn't listen, it would be a great service to
> the individuals who do.
Bicycle science isn't high tech, the most important part of it being
marketing, as we see by the colored tires and rims on the road today.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: alloy failure
Date: 9 Jan 2001 00:04:47 GMT
Bill Putnam writes:
> Thanks for the explanation. This is a straightforward modification.
> It's hard to believe that the manufacturers don't include this in
> their designs. It could even be backward compatible if the spindle were
> tapered and a collet and keeper were used in the opposite direction to
> present a flat face to the "old" style crank arm. I've attempted an
> ASCII design to verify your description-if you have Microstation
> handy I could attach a Microstation file to an e mail for a better
> drawing. The collet chamfer is omitted.
> ______________________
> | | | | |
> | | | | | Crank cross section
> | |________| | |
> |/ \| |
> ----------------------
> | 18mm |
Yes.
> __________ _
> /| |\ 1.8 mm
> ------------ - Collet
> _ 0.5 mm
> | 18mm |
> __________
> | |
> | | pedal threads
> | |
> | |
> | |
> ----------------
> || | 12.54| ||
> |---------------|
> | | pedal flats for wrench
> | |
> | |
Yes.
> Did you use a steel collet? Did you turn this out of bar stock or
> purchase it off the shelf?
I used stainless steel and had it made to my drawing. The friend who
built my bicycle also bored the cranks, something that is not just
countersinking but must be accurately concentric with the thread.
Your suggestion of backward compatibility doesn't work because there
are no extra threads to accommodate the missing length when the washer
is inverted. Besides the overhang that would produce with so little
bearing area would be an invitation to failure. My suggestion is to
offer new cranks with the countersink and the collet as well as new
pedals with a conical face to go with the cranks. Old pedals of this
brand would need the collet new ones would fit and soon there would be
no old style pedals to bother with. If Shimano can come up with their
aero pedals with a unique thread a few years ago, this one is easy.
> This modification would have prolonged the life of one of my cranks-
> a Sugino Mighty Tour that cracked at the pedal eye. The Nervar Star,
> and two Campy Nuovo Record cranks would still be where they
> are (my bin of failed bicycle and car components) due to cracks at
> the spider, however. And there's still the fretting occurring between
> the spindle and crank taper. Maybe the cotterless steel cranks are
> the most durable cranks so far (aside from potentially Ashtabula).
All the bicycle folks I talked to say there is no problem so why mess
with it.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Why are pedals screwed in backwards?
Date: 29 Aug 2000 21:34:29 GMT
Eugene Dougherty writes:
> OK you mechanical engineers, why in the name of all that's holy are
> pedals screwed in backwards? If the bearings were to lock up, the
> pedals would unscrew themselves during the normal forward pedaling
> motion. The resistance from the bearings applies a torque on the
> pedal axle in the direction of unthreading.
Even a psychologist would deduce that the reason you mention for a
pedal to unscrew must be wrong, because in more then 100 years, the
left-left, right-right arrangement has been retained... without
exception. That ought to make you look for another effect that this
threading addresses.
> I agree it makes sense to counterthread the pedals, I just think
> they did it the wrong way!!!
I wouldn't harp on it. It doesn't make sense. The reason is shaft
precession in a housing, both rotating about the shaft axis with the
shaft loaded vertically. Making an "O" with thumb and forefinger,
rolling the index finger of the other hand in the opening twists the
finger in the opposite direction of rolling. This is the mechanism.
One thing that this underscores, is that the pedal and crank interface
is a moving joint, a joint that shouldn't move.
Jobst Brandt <jbrandt@hpl.hp.com>
From: jobst.brandt@stanfordalumni.org
Subject: Re: Cracks in my cranks
Newsgroups: rec.bicycles.tech
Message-ID: <lYsL8.8723$3w2.39832@typhoon.sonic.net>
Date: Wed, 05 Jun 2002 18:42:25 GMT
Johan Bornman writes:
>> A simple modification to a tapered face on the pedal shaft similar
>> to those on automotive lug nuts gets rid of the failure and the
>> need for left hand threads. I recall cars that had left hand lug
>> nuts on the right side of the car left over like tying and
>> soldering from an era just as before tapered-face lug nuts.
> Jobst, I'm glad this topic came up again. Last time it was under
> discussion I couldn't figure out what your modification looks like.
> This time you mention lug nuts. What are these and how do you mimic
> them on the pedal/crank interface?
"Lug nut" is automotive jargon for (typically four or five) nuts that
hold car wheels in place. These were originally common flat faced
nuts that held the steel rim on the brake drum, or before that, the
rim/tire to the periphery of a wooden spoked wheel. The need for left
hand threads on the wheels on the right side became apparent. At
about the same time someone realized that fretting motions between
nut, bolt and rim, could be arrested by the conical faced nuts all
cars, trucks and buses use today. I suspect this was not the first
application of such nuts because they are used on other machinery.
The conical seat performs two functions; that of preventing relative
motion between nut and rim, and increasing friction by its 60 degree
(120 deg cone) pressure angle to prevent unscrewing.
http://www.tirerack.com/wheels/tech/nutsandbolts.htm
Both of these features are also essential to good pedal to crank
attachment, both because pedals have radial load precession that
causes rotation in the thread, and a cantilevered load.
Main Entry: cantilever
Function: noun
Etymology: perhaps from cant + -i- + lever
Date: 1667
: a projecting beam or member supported at only one end: as a : a
bracket-shaped member supporting a balcony or a cornice b : either of
the two beams or trusses that project from piers toward each other and
that when joined directly or by a suspended connecting member form a
span of a cantilever bridge -- see BRIDGE illustration
For bicycles the proper modification would be to a tapered shoulder on
the pedal spindle instead of a flat one, and of course a matching
taper in the face of the crank. I have by installing a split conical
faced washer between the pedal shoulder and the crank in such a way
that it securely clamps the spindle on the inside as it is tightened
into the taper faced crank bore. This reduces the probability of
crank failure at the pedal eye as well as getting rid of left hand
threads on left pedals.
This is not something that can be done at home. It takes a lathe and
milling machine. I chose stainless steel for the washers because it
is not a high stress application where currently aluminum is the
contact and I didn't want these to rust.
Jobst Brandt <jobst.brandt@stanfordalumni.org> Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Cracks in my cranks
Newsgroups: rec.bicycles.tech
Message-ID: <7jSL8.9349$3w2.44112@typhoon.sonic.net>
Date: Thu, 06 Jun 2002 23:33:23 GMT
Tim McNamara writes:
>> For bicycles the proper modification would be to a tapered shoulder
>> on the pedal spindle instead of a flat one, and of course a
>> matching taper in the face of the crank. I have by installing a
>> split conical faced washer between the pedal shoulder and the crank
>> in such a way that it securely clamps the spindle on the inside as
>> it is tightened into the taper faced crank bore. This reduces the
>> probability of crank failure at the pedal eye as well as getting
>> rid of left hand threads on left pedals.
>> This is not something that can be done at home. It takes a lathe
>> and milling machine. I chose stainless steel for the washers
>> because it is not a high stress application where currently
>> aluminum is the contact and I didn't want these to rust.
> Has this eliminated your crank breakage problems?
Yes. The cranks I modified were nearly 20000 miles old so I decided
to try the modification because it would be no loss if it didn't help.
I have brand new cranks on the shelf, also modified now and am sure
that the pedal eye failures are resolved.
Considering that cranks break there (in fatigue) and that at times, I
climb some steep places, that require great torque, without crank
failure, it occurs to me that this is truly a long term fatigue
problem. The cracks initiate from the eroded face of the crank caused
by pedal shaft fretting there and in the threads.
To me this is much like the spoke failure problem that went on for
ages under the belief that someone temporarily overloaded a wheel.
The difference here is that we have a much more apparent model from
the automotive wheel nut as an example how it should be done.
Don't hold your breath. "Who knows more about crank failures, this
non-bicycle mechanic or Campagnolo/Shimano?" As you see, this
assessment recurs all the time in defense of the status quo.
Jobst Brandt <jobst.brandt@stanfordalumni.org> Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Cracks in my cranks
Newsgroups: rec.bicycles.tech
Message-ID: <Nz7M8.9698$3w2.45860@typhoon.sonic.net>
Date: Fri, 07 Jun 2002 19:11:41 GMT
Benjamin Lewis writes:
>> Don't hold your breath. "Who knows more about crank failures, this
>> non-bicycle mechanic or Campagnolo/Shimano?" As you see, this
>> assessment recurs all the time in defense of the status quo.
> Have you tried suggesting this to Shimano? It seems to me that they
> could not only put a great marketing spin on it, but it would be in
> their interest to sell high-end pedals/crank combinations that are
> incompatible with anything else.
I showed a typical DA crank to to a Shimano engineer at InterBike who
knew exactly what I was talking about and understood the solution. I
had crank with an almost cracked of pedal eye, the crank that I am
riding now with its modification and a new crank and pedal that was
hand tight so that the assembly could be inspected.
It was difficult to make this approach because Shimano does not want
to talk abut designs without a disclosure agreement. To bridge that
chasm, I first showed him the broken crank and asked if he recognized
the condition. The conversation got underway so I could tell him that
I had lots of these and discovered a solution for the problem which I
showed in a neutral manner, not asking for any agreement or
disagreement. Just the facts as I have presented them here. The man
was visibly interested and we parted without any resolving comment
other than "g'day".
> There's even the minor advantage that it really *is* better.
As I said, don't hold your breath. Neither Race Face nor Truvativ,
with whom I talked, have taken the subject any farther, even though
they have such crank failures often.
Jobst Brandt <jobst.brandt@stanfordalumni.org> Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Cracks in my cranks
Newsgroups: rec.bicycles.tech
Message-ID: <Cb8M8.9726$3w2.45898@typhoon.sonic.net>
Date: Fri, 07 Jun 2002 19:54:10 GMT
Karl Nelson writes:
>> For bicycles the proper modification would be to a tapered shoulder
>> on the pedal spindle instead of a flat one, and of course a
>> matching taper in the face of the crank. I have by installing a
>> split conical faced washer between the pedal shoulder and the crank
>> in such a way that it securely clamps the spindle on the inside as
>> it is tightened into the taper faced crank bore. This reduces the
>> probability of crank failure at the pedal eye as well as getting
>> rid of left hand threads on left pedals.
>> This is not something that can be done at home. It takes a lathe
>> and milling machine. I chose stainless steel for the washers
>> because it is not a high stress application where currently
>> aluminum is the contact and I didn't want these to rust.
> I am curious why you didn't simply use the lathe to machine a taper
> into the pedal spindle itself. I have your design for the collar
> but keep putting off making it because I keep wondering why that
> wouldn't be better.
Obviously, that is the manufacturing solution but if you look at it
you'll see there isn't enough "meat" on an existing pedal to do it.
The taper would run into the wrench-flats and leave these two flanks
unsupported, besides not leaving enough space for a good wrench. It
also reduces foot-to-crank clearance, that for my size is close
anyway.
> I also wonder about letting the steel taper (whether collar or
> spindle) mash its own taper into the softer aluminum of the crank.
Why not just drive the pedal in with a hammer! Aluminum is not
modeling clay and would not deform more than a fraction of the contact
face needed to restrain the loads in question.
> What would happen then? As you have said, centering that taper in
> the hole is a problem, and it's the other thing holding me up.
Put a dummy spindle in the bore, uses a center finder on the milling
machine and you have the setup.
Jobst Brandt <jobst.brandt@stanfordalumni.org> Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Pedal wear groove on crank.
Newsgroups: rec.bicycles.tech
Message-ID: <1LCt9.43567$Ik.980188@typhoon.sonic.net>
Date: Wed, 23 Oct 2002 19:27:57 GMT
Tom Nakashima writes:
>>> Upon removing the drive side pedal, I noticed a wear ring at the
>>> pedal-crank interface. It is most pronounced on the side of the
>>> hole closest to the bottom bracket bore, being about 1/16 of an
>>> inch deep.
>>> Should I take my dremel and make the wear even on the other side
>>> of the pedal hole and then add a washer?
>> Probably not but if I understand correctly, you have a loose pedal
>> problem but more so than most. See the following:
http://groups.google.com/groups?q=pedal+eye+failure+group:rec.bicycles.tech&
hl=en&lr=&ie=UTF-8&selm=9bibil%24cuf%242%40hplms2.hpl.hp.com&rnum=10
> Very interesting write-up on the pedal/crank failure. You mentioned
> you haven't had a pedal failure since modifying the pedal eye. The
> modification you're talking about is beveling the eyelet? Was this
> done with an 82 degree countersink tool and drill press?
I used a 90 degree cutter to "countersink" the crank and essentially
made the pedal shaft face have a matching taper by making a split
collet. 82 degree would work too but 90 is common in the metric world
and is simpler to define. If the pedal shaft had a little more meat
between the wrench flat and its pressure face, the shaft itself could
have been tapered. I don't even inspect the crank anymore and it's
been at least 50000 miles.
Jobst Brandt <jobst.brandt@stanfordalumni.org> Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Crank crack
Newsgroups: rec.bicycles.tech
Message-ID: <R8iya.15471$JX2.948437@typhoon.sonic.net>
Date: Tue, 20 May 2003 04:27:29 GMT
David L. Johnson writes:
> Don't worry. Replace the crank immediately instead. Imagine when
> it develops another crack, and falls off on the downstroke while
> climbing. Hello, top tube.
The proposed scenarios in this thread are entirely imagined, just as
non-bicyclists have enormous fear of the handlebar stem and their
genitals. It doesn't happen that way. When a crank breaks, the foot
lands on the ground without incident but the rider, if traveling more
than 10mph takes a dive to that side. You don't land on the top tube!
However, that fear has given us 14inch seat posts and sloping down
tubes to give a SAFER stand-over clearance.
Where is bicycling going???
Jobst Brandt
jobst.brandt@stanfordalumni.org
Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: Crank crack
Newsgroups: rec.bicycles.tech
Message-ID: <FBCya.15689$JX2.960673@typhoon.sonic.net>
Date: Wed, 21 May 2003 03:43:33 GMT
John Henderson writes:
>> THat's not the way it happens. Let me tell you, when a crank
>> breaks, you are standing on the road with one foot and cannot run
>> because you have a bicycle between your legs and the other foot
>> most likely attached to the pedal. You WILL fall to the side on
>> which the crank failed and if its a left crank you will dive under
>> passing traffic if there is any.
> Is the falling part a function of speed? I've only broken one crank
> (right side, about 2" up from the pedal spindle), but didn't fall.
> I had just entered a bike path after making a sharp turn, and was
> travelling 10-15 mph, seated. All of a sudden my right foot was
> free (I thought I had broken the pedal), but the pedal and crank
> stub were still strapped to my foot when I looked down.
You were seated. Most cranks break off in a high force instance and
for most active riders, that's a standing sprint (away from a stop) or
up a hill. The crank gradually develops the crack but since there is
a wide span of forces applied, it is common for failure to occur at
high stress and that is most often while standing, the dangerous
contition.
Jobst Brandt
jobst.brandt@stanfordalumni.org
Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: broken crank
Newsgroups: rec.bicycles.tech
Message-ID: <xKCya.15692$JX2.960839@typhoon.sonic.net>
Date: Wed, 21 May 2003 03:53:01 GMT
Tom Paterson writes:
> http://technology.open.ac.uk/materials/mem/mem-ccf4.html
> Good one there (IMHO)! Looks like the common Campagnolo "spider
> crack" might have found full expression on this crank.
Nice picture. I've had cranks like that.
This is a common failure and the one most easily prevented but the
bicycle industry doesn't want to bother. One solution is to take the
Campagnolo C-Record approach and make the crank one of the five spider
arms. Ritchey cranks do that also. Shimano doesn't do that because
their cranks are hollow and might crack from the chainring bolt
entering the hollow crank. However, with the crank in the picture,
Campagnolo and others who use this design need only to machine away
the thin web between spider and both sides of the crank so that it no
longer looks like the skin between index and middle finger spread
apart.
Jobst Brandt
jobst.brandt@stanfordalumni.org
Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: broken crank
Newsgroups: rec.bicycles.tech
Message-ID: <8qSya.15901$JX2.971919@typhoon.sonic.net>
Date: Wed, 21 May 2003 21:43:32 GMT
anonymous snipes:
>> However, with the crank in the picture, Campagnolo and others who
>> use this design need only to machine away the thin web between
>> spider and both sides of the crank so that it no longer looks like
>> the skin between index and middle finger spread apart.
> Hmmm... just curious. Your statement seems to suggest that the web
> can contribute to a crank's failure rather than its strength. How
> is this so?
Thin webs that are at the extremes of a bending element carry many
times the stress of the main member. That is why I-beams have thick
flat "caps" connected by a thin web less than half the thickness of
the caps. Cutting the beam in half lengthwise through the web and
putting the caps back to back so that the half-webs are outermost is
what we have in the crank. Structural engineers are aware of this but
the bicycle industry is not.
In the collection of failed cranks, we also see evidence of failures
at embossed or cast-in logos. Formerly, most cranks had a lengthwise
vanity groove on the outside that reduced crank stiffness by at least
30%. There are still I-beam like cranks on the market even though
this shape has notoriously poor torsional stiffness. It may be a
fashion statement because many people recognize the I-bean shape and
think cranks are stressed mainly in bending. Failures are as much
from torsion as bending. Where do these guys go to school, if at all?
http://pardo.net/pardo/bike/pic/fail/FAIL-001.html
Jobst Brandt
jobst.brandt@stanfordalumni.org
Palo Alto CA
From: jobst.brandt@stanfordalumni.org
Subject: Re: BB and pedal thread direction - WHY?
Newsgroups: rec.bicycles.tech
Message-ID: <_Q9kc.8748$Fo4.112492@typhoon.sonic.net>
Date: Thu, 29 Apr 2004 16:13:46 GMT
George White writes:
>>> From my discussions with manufacturers I get the impression it is
>>> partly NIH (not invented here) and partly like the disc brake
>>> problem. No one wants to be the first to admit there is a
>>> problem.
>> And of course, nobody wants to stick their neck out and try to
>> foist a new standard on pedals when they have one that has existed
>> for what, maybe 100 years in some form now?
> I agree with you, I don't see any "problem" with left-handed
> threading on the left pedal. The collet/cone may allow a
> right-handed threaded left pedal to be installed, and still not
> unscrew, but so what? Just taking away a _specific_ requirement
> doesn't mean it should be replaced by some other arbitrary style.
> Colleting it and simply leaving the left-handed threads would be
> fine too. Moreover, if the whole colleted crank arm thing came to
> be, but the threading was left alone, after-market cones (split
> cones?) would allow people to keep using the existing huge stock of
> pedals.
No one can offer a collet for existing pedals because the thread
relief on those spindles are undefined. The width and diameter of
that feature (the cylindrical part between shoulder and thread) must
fit snugly into a collet. THat is why this takes a drastic departure
of a new pedal standard like that of Shimano, that I believe also was
a response to the crank failure problem but didn't address the basic
fretting problem and left hand thread.
> The collet is the right thing to do because the cranks can break in
> the current design. On the other hand, changing the threading is
> jerking around. I'm willing to pay for the collet and two
> oppositely threaded spindles.
The difference is that manufacturers would not need to have left hand
thread tooling and could make one kind of pedal spindle that would
work as well on tandems that presently use left cranks on the right
with right hand thread and the converse. Tandem folks can speak for
that problem.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Just snapped 20-yr-old crank arm...
Newsgroups: rec.bicycles.tech
Message-ID: <uu9tc.13383$Fo4.187335@typhoon.sonic.net>
Date: Wed, 26 May 2004 23:11:22 GMT
Jay Beattie writes:
>>> As I have related often, failures at the pedal eye occurred on my
>>> bicycle about every 10,000 miles on the average for more than 25
>>> years. That's a lot of cranks and they were mostly Campagnolo and
>>> later Shimano.
>> good lord, that scares the crap out of me. would you venture a
>> guess at the standard deviation of those failures? (not that I'm
>> worried with 16,000 miles on my current crank and memories of a
>> catastrophic failure behind me).
> The Campy NR cranks broke with some frequency. I broke at least
> four in the 70s and 80s (and a few other brands, too). Jobst's
> experience, IMO, represents the extreme end of the spectrum. I
> have never broken a Shimano crank, and I raced at about 200lbs
> for many years on them. I still ride an older Dura Ace crank
> (1983) with probably 100K on it.
> There is not a whole lot to fear for the average cyclist. Another
> interesting thing about breaking a crank is that it usually does not
> result in a serious crash. The OP's experience is the common one --
> your foot hits the ground and you wobble to a stop. Much like when
> you break a pedal spindle. I would be more concerned with a broken
> bar, which seems to be the popular thing to break these days (based
> on the NG posts).
I disagree. If you have broken a crank while standing, you fall to
the side of the failure, there is no recovery if you are traveling at
any reasonable speed. The effect is like trying to run with one leg
with a bicycle between your legs. If you are sprinting across the
intersection when the light turns green the readymix truck next to you
will make a pancake out of you and your bicycle.
You only imagine that a crank will break while sitting but the active
rider puts his peak force into the pedals while standing and therefore
the fracture generally occurs in that mode, the sitting mode being too
low a force to cause failure. Most of my cranks were discovered by
inspection but some were not. Some of the pictures in the Hjertberg
collection are mine as are two crank spindles. When a crank spindle
breaks there is no warning and no way to check for it because it
breaks inside the press fit of the crank.
These failures are probable fatalities for those who ride adjacent to
auto traffic... or a cliff.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Another Broken Crank
Newsgroups: rec.bicycles.tech
Message-ID: <6EwJd.3584$m31.51278@typhoon.sonic.net>
Date: Tue, 25 Jan 2005 18:57:06 GMT
Andrew Muzi writes:
> I'm sorry for any confusion but they are here:
http://www.yellowjersey.org/jayscranks.html
I find the third picture interesting because it shows that the crack
was there and possibly subjected to water before it broke the final
cross section all at once. Also, Hollowtech is a sham in this model,
as it is elsewhere, no matter whether it has a slightly larger void.
What has come to my attention over time is that these fractures appear
to result from bending loads primarily when standing with the crank
straight down. None of the failures I have seen had a torsion failure
or for that matter, a bending failure in the direction of crank
rotation. That is, none failed across the long dimension of the crank
cross section, crank bending in the rotational plane is where design
emphasis is being directed.
http://dura-ace.shimano.com/publish/content/duraace/en/home.html
This page starts with a crank showing a mesh that could be a FEA
network. Either this is pure imagination or they are not loading the
analysis with a real rider input, something that can easily be done
with a few strain gauges on a crank.
Bending stress being related to the third power of beam height, this
failure mode seems logical. What is bothersome about this is that
this problem is apparently being addressed by stylists rather than
stress analysts. This is true for failures as this one and those that
occur at the pedal eye.
For those who have not had a crank fail, this may seem remote and non
threatening, but a bit of visualization could make that different.
When a crank fails, it most often occurs while standing, and if it
occurs at more than about 10mph, the rider falls to the side of the
failure. When adjacent to motor traffic this can be fatal. Crank
failure puts the rider in the position of trying to run with one leg
wile having a bicycle between his legs. It does not work unless speed
is near walking. I have avoided crashing by one footed running on a
few incidents that occurred when starting from a stop light.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Another Broken Crank
Newsgroups: rec.bicycles.tech
Message-ID: <XoGJd.3713$m31.51931@typhoon.sonic.net>
Date: Wed, 26 Jan 2005 06:03:35 GMT
Tom Sherman writes:
>> For those who have not had a crank fail, this may seem remote and
>> non threatening, but a bit of visualization could make that
>> different. When a crank fails, it most often occurs while
>> standing, and if it occurs at more than about 10mph, the rider
>> falls to the side of the failure. When adjacent to motor traffic
>> this can be fatal. Crank failure puts the rider in the position of
>> trying to run with one leg wile having a bicycle between his legs.
>> It does not work unless speed is near walking. I have avoided
>> crashing by one footed running on a few incidents that occurred
>> when starting from a stop light.
> Do the pedal and crank "stump" generally stay attached to the
> rider's shoe during crank failure when either clips and straps or
> "clipless" retention systems are being used?
The pedal, even with clips and straps remained on the foot when it hit
the road and that is the critical step, the first one. With SPD's it
was the same and I can recall the pedal hitting the ground under my
foot.
That's all after the fact. What's interesting is when will crank
manufacturers analyze the failure mode and do something about it. As
you can see, cranks are getting weaker in the axis of interest and
stronger in the wrong direction. Is there no hope? As we go on to
making everything of carbon fiber, even before understanding why the
old stuff failed, we are entering a new failure mode that is so common
that people here are asking how often they should get a new xxx.
When you see Shimano not recognizing why their Octalink didn't work
and making the spline deeper for a second try, you've got to get a
bare spot your head from scratching, so to speak. Campagnolo going
through two or more redesigns of the Delta brake, never recognizing
its basic flaws that cannot be fixed, you have to wonder why they
don't read wreck.bike or hire people skilled in the art.
Shimano proudly replied to my mail recently that they have never
accepted design information from outside the company. They don't even
want to know well known solutions to old problems that are a mystery
to the bicycle industry but solved elsewhere. Ugh!
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Another Broken Crank
Newsgroups: rec.bicycles.tech
Message-ID: <XOgKd.4156$m31.55742@typhoon.sonic.net>
Date: Fri, 28 Jan 2005 01:45:27 GMT
Mike DeMicco writes:
>> I find the third picture interesting because it shows that the
>> crack was there and possibly subjected to water before it broke the
>> final cross section all at once. Also, Hollowtech is a sham in
>> this model, as it is elsewhere, no matter whether it has a slightly
>> larger void.
> This crank has 2 built in stress risers (the points at the edges of
> the void). Are we sure the crack didn't start there? Why did the
> break occur right at the start of the void space if, as Jim Beam
> said, the crack started from the outside from abrasion? Those
> pictures are pretty blurry...
The direction of crack propagation is from the outside of the crank,
nearest the rider's foot. Successive waves of separation are clear in
the picture. The internal void had nothing to do with initiation
because it lies in the neutral zone for this failure.
Even in the orthogonal axis, it would have no effect because it is
normal to the stress gradient.
Whether the material is as dense as it would be with forging a solid
cross section is not apparent. I have no idea how this thing is made.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Another Broken Crank
Newsgroups: rec.bicycles.tech
Message-ID: <_kmKd.4208$m31.56760@typhoon.sonic.net>
Date: Fri, 28 Jan 2005 08:03:06 GMT
Tim McNamara writes:
>> The direction of crack propagation is from the outside of the
>> crank, nearest the rider's foot. Successive waves of separation
>> are clear in the picture. The internal void had nothing to do with
>> initiation because it lies in the neutral zone for this failure.
>> Even in the orthogonal axis, it would have no effect because it is
>> normal to the stress gradient.
> Never having had a crank break, I wonder whether the "successive
> waves of separation" made noise. Did this crank creak as the crack
> propagated? (Apologies for the alliteration). Would the rider have
> had any warning something was amiss before the crank broke, or was
> this a silent failure?
I have heard it when water gets in the crack. It makes a noise
similar to tendon sounds when turning your neck or chewing celery, for
instance. It is not noticeably loud and if you aren't listening for
it probably won't catch your attention. If do hear it, you probably
already saw that there is a crack.
The important point is that cranks break almost exclusively from
lateral bending from standing force on the pedal with the crank
straight down. This is apparent when reviewing many crank fractures
in which the crank broke in two (not knuckle splits) or broke in the
pedal eye, yet no cranks have been shaped to address this failure.
Today's cranks are weaker than ever in that direction and wider in the
other, becoming broad flat failure blades.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Unusual chainrings and cranks wear
Newsgroups: rec.bicycles.tech
Message-ID: <azsPd.6523$m31.79855@typhoon.sonic.net>
Date: Sat, 12 Feb 2005 19:13:10 GMT
Rado bladteth Rzeznicki writes:
> I've recently bought a road bike from a guy who got it from a
> triathonist. The bike (Avanti Team Corsa) is 2000 year model, the
> crankset and chainrings are still original and I was told they have
> made about 20 000 km. Yet, the wear that they show is unusual to me.
> Here is a picture of them:
> http://coogeeplaza.republika.pl/cranks.jpg
> Looks like the nickel plating is flaking off exposing white, grainy
> aluminium alloy and cranks have a lot of marks on them, but the
> shine is still there. Previous owner had no idea what could have
> caused this, but suspected sweat. Is this possible? Have anybody
> seen something like this before? The crankset itself works great.
The chainwheels look like almost no wear but the corrosion on the
cranks has lifted some of the anodizing. This bicycle was exposed to
a lot of salt air at the seashore no doubt. The trouble is there is
possibly more serious corrosion at other places like in various
threads in the cranks. I would take it apart and clean it. Fine
steel wool will not break through the anodizing but it would remove
the flowers of aluminum oxide. There are also shoe wear marks on the
crank that could be polished out if done carefully. Sharp
discontinuities present crack initiation locations and are best
polished out for that reason rather than for appearance.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Unusual chainrings and cranks wear
Newsgroups: rec.bicycles.tech
Message-ID: <hIAPd.6549$m31.80698@typhoon.sonic.net>
Date: Sun, 13 Feb 2005 04:29:01 GMT
dianne who? writes:
>>> http://coogeeplaza.republika.pl/cranks.jpg
>> Sharp discontinuities present crack initiation locations and are
>> best polished out for that reason rather than for appearance.
> Here's a similar model crank that had unusual failure:
> http://pardo.net/pardo/bike/pic/fail/FAIL-010.html
> I wonder if corrosion also caused crack initiation locations on this
> one? It also has shoe rub marks...
The shoe rub marks in that picture are about as smooth as any
polishing one might consider in crack initiation and are not, in my
estimation related to the failure. I don't see any corrosion on the
crank in the area of crack initiation on this one. However, as you
see in this picture, the failure mode is that of lateral bending while
standing on the pedal, crank straight down.
This has been the failure mode for the many cranks failures I have
had, whether at the pedal eye or elsewhere. Fortunately, failures
other than at the pedal have fairly ceased as have those at the pedal
since my modification. Inspecting for cracks is tedious because they
can occur anywhere along the crank.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Unusual chainrings and cranks wear
Newsgroups: rec.bicycles.tech
Message-ID: <l%BPd.6560$m31.80788@typhoon.sonic.net>
Date: Sun, 13 Feb 2005 05:57:37 GMT
Tom Sherman writes:
>> This has been the failure mode for the many cranks failures I have
>> had, whether at the pedal eye or elsewhere. Fortunately, failures
>> other than at the pedal have fairly ceased as have those at the
>> pedal since my modification. Inspecting for cracks is tedious
>> because they can occur anywhere along the crank.
> Would the use of a dye penetrant be worthwhile, or is visual
> examination adequate?
Dye penetrant requires a clean (degreased) sample and one that does
not have scratches where dye could lodge. Typically polished
automotive connecting rods, that do not scrape on anything, are ideal
for that. You often hear "it was only a hairline crack" but then all
cracks are hairline until separation unless it is in a plastic medium.
If you have been riding in rain, black ooze will emanate from a crack
and thereafter, the line will probably be more visible. My experience
is that climbing, where there is no wind noise, a wet crank makes the
sound of chewing on gristle. That is rare in my experience.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Unusual chainrings and cranks wear
Newsgroups: rec.bicycles.tech
Message-ID: <m3BPd.6551$m31.80644@typhoon.sonic.net>
Date: Sun, 13 Feb 2005 04:53:38 GMT
Rado bladteth Rzeznicki writes:
>> The chainwheels look like almost no wear but the corrosion on the
>> cranks has lifted some of the anodizing.
> Are chainrings nickel plated just for aesthetic reason? The wear on
> the teeth is minimal indeed.
I don't believe either the cranks or chainrings in question are plated
but are instead anodized, a conversion of the aluminum itself into a
ceramic coating. Anodizing is about half as dense as aluminum, much
harder, and has micro pores that allow dye coloring. As you see it
does not protect where the coating has been mechanically penetrated or
has some other flaw allowing liquid penetration.
>> This bicycle was exposed to a lot of salt air at the seashore no
>> doubt.
> That's correct. The place is Sydney, right on the shore. Paint on
> the seat tube, just above derailleur, where bottle cage is, looks
> swollen and is flaking off in the same way.
You might also check to see if the bar stem is permanently swollen
into the steer tube, a common failure even without salt air (sweat
and rain). That is why bar stems for threadless steer tubes are a
great advance in bicycle mechanics. Not only do they not corrode
solid, they require only a hex key to secure the bars, the bars can be
removed without un-taping or removing a brake lever, the head bearing
is adjusted with the same hex key and the whole assembly is amazingly
stiffer and more robust than the old quill stem.
>> The trouble is there is possibly more serious corrosion at other
>> places like in various threads in the cranks. I would take it
>> apart and clean it.
> No problems here. Cranks easily came off the BB and loud 'ping'
> accompanied unscrewing chainring bolts, which are completely dry. I
> don't believe they forgot to grease them.
You are on the right track. A complete disassembly and cleaning is
the best way to assess if anything is in need of replacement.
>> Sharp discontinuities present crack initiation locations and are
>> best polished out for that reason rather than for appearance.
> I did that and the surface feels smooth now, but mirror finish is
> gone.
Use finer steel wool and follow up with an aluminum polish like
Simichrome:
http://www.newmantools.com/simi.htm
This polishing paste is excellent for the job. Of course after the
anodizing is gone, salt air will have an easier time getting at the
parts but then you don't need to leave the bicycle out in the fog and
surf spray.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Unusual chainrings and cranks wear
Newsgroups: rec.bicycles.tech
Message-ID: <DS8Qd.6798$m31.82991@typhoon.sonic.net>
Date: Mon, 14 Feb 2005 21:37:39 GMT
Matt O'Toole writes:
>> If you have been riding in rain, black ooze will emanate from a
>> crack and thereafter, the line will probably be more visible. My
>> experience is that climbing, where there is no wind noise, a wet
>> crank makes the sound of chewing on gristle. That is rare in my
>> experience.
> Geez, that's more experience with cracked cranks than I ever hope to
> have! How many have you broken over the years? I've broken a
> couple -- experiences I do not want to repeat.
It was more then two dozen when I stopped counting. I got to thinking
that the DuraAce cranks I was riding would soon fail like the others I
had changed out about every year or so before, so there being no loss
in machining the ones I was riding because they wouldn't last much
longer anyway, I tried my modification.
This had two effects. One was that it machined away the fretting
erosion in the face of the crank where cracks initiate, and two, it
prevented motion at that interface and therefore no fretting. Those
cranks served five more years before I discovered cracks starting
inside the corners of the square taper. The cranks I have are DuraAce
ones that I bought the 1980's and are giving me greater piece of mind
now.
The cause for these failures I suspect, is that I ride mostly in
mountains and with weak back muscles, have always tended to stand on
climbs. As I pointed out, it is not torque that cause crank failures
but rather standing loads. The failure mode for practically all
cranks I have seen is from standing one the pedal at the bottom of the
stroke. If it had much to do with pedaling torque, failures would be
spiral rather than straight across from outside across the crank.
As I mentioned the left hand thread on left cranks is not for nothing.
One might ask, why don't you tighten the pedal and be done with it.
Well we all know that tandems that do not use tandem cranks, loosen
pedals in cranks with conventional threading and they are tight. The
reason is that there is always motion in this joint and it causes
fretting and crack initiation. The auto industry faced up to this
same problem by switching to conical lug nuts. Most people have never
seen the previous type. Losing a wheel was still an anecdote often
hear when I started looking at these things. That's where I got the
idea to do that at the pedal/crank interface. The bicycle industry
doesn't want to hear of it. I suppose someone must first sue the
gabillions out of Shimano or Campagnolo to catch their attention.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Pedal Torque Spec.
Newsgroups: rec.bicycles.tech
Message-ID: <VCIVd.9165$m31.108276@typhoon.sonic.net>
Date: Thu, 03 Mar 2005 18:23:49 GMT
Bruce Graham writes:
>>> The reason Shimano hasn't done anything along these lines is that
>>> they don't know about it. They have no way of knowing that the
>>> pedal attachment is flawed since they do not discuss any technical
>>> matters with anyone who doesn't have a patent disclosure and they
>>> do not read these newsgroups. I got a basically rude letter from
>>> Shimano making this clear. The end of the letter underscored that
>>> by saying that Shimano has never used an outside idea in anything
>>> they have brought to market. In other words, don't bother us
>>> further. Although I don't believe that claim, the sentiment
>>> expressed made clear how self impressed the company is.
>> My take on their sentiment is different than yours. While it may
>> appear arrogant, it's more likely a legal stance to protect patents
>> and products. They're a changed company since essentially losing
>> the SPD system to the public domain. They don't want to pay
>> royalties (wonder why there's no threadless Shimano headset?) nor
>> do they wish to give anything away.
> Hey Jobst, HP was like that (I worked there a long time ago) and
> that attitude was legally driven too. Coming from HP Labs I thought
> that process would have driven you nuts on a daily basis and would
> have been almost inescapable - but maybe being a labs insider you
> had your ways. (I realise you can't easily comment :)
That's not been my experience. I have seen plenty of outside ideas
presented and license agreements made for products. Unfortunately
Carly sold off a bunch of these things and got rid of Agilent, the
basic strength of HP. This trend started earlier with emphasis on
consumer computing products instead of high tech instruments.
> Besides, if Shimano was actually doing something about that aspect
> of pedal design, the shut-out response you got would be especially
> necessary from their perspective.
Not at all. This concept is fully in the public domain and they could
prove that easily... if they were aware of it. I have talked to
Shimano employees who told me at the end of our conversation that they
were unable to do anything. Even mentioning the idea could get them
fired so everything we talked about and the hardware they saw was
essentially bulk erased.
Jobst Brandt
jobst.brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Octalink cranks on a fixed wheel
Newsgroups: rec.bicycles.tech
Message-ID: <ZfO0e.12217$m31.125734@typhoon.sonic.net>
Date: Fri, 25 Mar 2005 06:32:25 GMT
I think there is a resistance to the idea that threads do not center
or restrain radial motion. First, one must realize that surface
friction on the crank face is not enough to keep a pedal from moving
radially in its threaded bore. Once that is established a thread
experiment should furnish the remaining evidence.
Advance a screw about ten turns into a threaded hole on some piece of
machinery that is convenient and notice that it can be wiggled
radially without appreciably changing it's depth of engagement.
Pedals do this in cranks but only in tiny displacements with fretting
motions of a few thousandths of an inch. This is enough to erode the
face of the crank and to generate radial cracks from the edge of that
eroded counterbore.
Pictures of failed cranks show these eroded recesses that this motion
caused. Failure occurred when the initiated cracks propagated through
the crank.
Jobst.Brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Octalink cranks on a fixed wheel
Newsgroups: rec.bicycles.tech
Message-ID: <aTX0e.12246$m31.125901@typhoon.sonic.net>
Date: Fri, 25 Mar 2005 17:28:38 GMT
Carl Fogel writes:
>> I think there is a resistance to the idea that threads do not
>> center or restrain radial motion. First, one must realize that
>> surface friction on the crank face is not enough to keep a pedal
>> from moving radially in its threaded bore. Once that is
>> established a thread experiment should furnish the remaining
>> evidence.
>> Advance a screw about ten turns into a threaded hole on some piece
>> of machinery that is convenient and notice that it can be wiggled
>> radially without appreciably changing it's depth of engagement.
>> Pedals do this in cranks but only in tiny displacements with
>> fretting motions of a few thousandths of an inch. This is enough
>> to erode the face of the crank and to generate radial cracks from
>> the edge of that eroded counterbore.
>> Pictures of failed cranks show these eroded recesses that this
>> motion caused. Failure occurred when the initiated cracks
>> propagated through the crank.
> I appreciate how you and Luns are trying to explain the obvious, but
> I've missed the point so often that I want to check.
> If I'm following you, the pedal shaft bends down a tiny bit under my
> foot like a beam bending (not rotating) and its flat face frets
> against the flat face of the crank, leading to bad things.
The pedal shaft need not bend and effectively does not bend relative
to its displacement under pedaling forces, that is its flexing is not
part of the motion in the aluminum crank. By the way, this also
occurs in steel cranks. That is where I first encountered it.
Besides, the left hand thread came along long before anyone thought of
anything but steel cranks.
> The wear is between the two flat surfaces, not around the
> corner and down inside the crank where the threads engage.
That is correct. However, I am sure there is also fretting inside the
thread but nowhere near the damage that is caused at the face of the
crank. A steel crank would show this more dramatically with rouge and
a recessed worn area.
> | || |
> | || |
> ---======--- <---> two horizontal faces slide a little
> | |
> | |
>
> <-pedal shaft rocks->
> Your conical lug-nut modification changes two flat surfaces (which
> can slide a little past each other and fret) into one surface
> pressing almost directly against the other:
> | |
> | | | |
> | // \\ | now the rocking motion pushes against
> |__// \\__| the sloped face instead of sliding
> | |
> | |
> <-pedal shaft still rocks->
> Is this the idea, or am I still missing the obvious?
Good graphics! You can also visualize the pedal spindle as swiveling
about a point in the center of the thread engagement and the tapered
faces as being perpendicular to that motion instead of parallel.
Jobst.Brandt@stanfordalumni.org
From: jobst.brandt@stanfordalumni.org
Subject: Re: Octalink cranks on a fixed wheel
Newsgroups: rec.bicycles.tech
Message-ID: <3VY0e.12257$m31.125976@typhoon.sonic.net>
Date: Fri, 25 Mar 2005 18:38:55 GMT
Carl Fogel writes:
> The light may be dawning.
> So the movement is not so much a bending of the pedal shaft as it is
> a microscopic rocking or even rattling of the threaded pedal shaft
> inside the threaded pedal eye?
> That is, given really coarse threads and a pedal shaft running all
> the way through the eye, the whole pedal shaft would rock at a very
> slight angle under my foot, with the near side pushing down and the
> far end coming up, like a see-saw?
> Or maybe a loose pencil cocking in a funnel?
> The idea is that the pedal shaft stays pretty much straight instead
> of bending, but tips downward because of the minute amount of play
> in the threads.
> foot pushes down
> |
> _______ V
> this | |
> end | |::::::::::::::::
> tilts \\\\\\\\\\\\\:::pedal shaft: this end tilts down
> up | Xx:::::::::::::::
> | |
> crank Xx = sliding faces & wear
> If I'm finally understanding things, it's the faint play in the
> threads that lets the flat faces of the crank and shoulder of the
> pedal shaft rub and fret.
Got it!
Jobst.Brandt@stanfordalumni.org
Index
Home
About
Blog