Significant Accident

On Sun, 19 Sep 2004 20:23:43 GMT,
wrote:

Terry Morse writes:

I've posted eight photos of the crank on a private page of the
club's web site:

http://www.ocrebels.com/CrankFailure.htm

Hmm. Note what looks like a flaw in the metal at the 9 o'clock
position, surrounded by corrosion:

http://www.ocrebels.com/images/failure/pic7.htm

It looks to my untrained eye that this could have been the original
source of the crack. I'm also curious about the shiny section, from
about the 1 o'clock position to the 7 o'clock position. The
polished surface suggests that the fracture was there quite a while,
and the relative movement of the pieces did the polishing.

The smooth region you describe is the area that failed all at once.
The upper part of that picture, the granular part, failed gradually in
fatigue. Motion in such a fracture remains the classically termed
"hairline crack" until rupture. In that crack motion is at best
fretting that may advance oxidation. Hairline crack means only that
the part had not yet separated and is meaningless for that reason.
Such cracks are at times more visible when wet because they emit a
fine slurry of oxide, in aluminum as well as steel.

If there are any materials people here, could they comment?

That's the comment.

Jobst Brandt


Dear Jobst,

"That's the comment."

How fortunate that there is nothing else to be said and no
possibility of error!

Carl Fogel

I'm working on getting photos of the spindle, but it may be infeasible as the
bike is going back to the owner soon. It is known to be at least ten years
old, gone through many, many centuries and doubles, and I'm told (by the
temporary holder) that it might date from the days when Merlin made its own
bottom bracket. The guy is also apparently a Big (300#?) individual.

Jobst or anyone: is there anything the normal rider, not privy to an Xray
crystallography setup, can do to check for incipient failures of this genre?

Art
Temporary usercode - to be deleted when spam starts. Use MyBrainHurts at this ISP to reach me

Arthur Shapiro writes:

I'm working on getting photos of the spindle, but it may be
infeasible as the bike is going back to the owner soon. It is known
to be at least ten years old, gone through many, many centuries and
doubles, and I'm told (by the temporary holder) that it might date
from the days when Merlin made its own bottom bracket. The guy is
also apparently a Big (300#?) individual.

That's too bad. I think a good picture of such a failure helps riders
realize that these things break and that there is effectively no
warning.

Jobst or anyone: is there anything the normal rider, not privy to an
Xray crystallography setup, can do to check for incipient failures
of this genre?

Especially with a heavy rider, replace the spindle every 20,000 miles
whether it needs it or not. I say that because most of my Campagnolo
spindles failed in the bearing races and didn't last long enough for
this type of failure. When one finally outlasted bearing failure mode
it broke in the taper after about 30,000 miles but that's only a guess
because that spindle survived several right and left failed cranks.

Jobst Brandt

On Sun, 19 Sep 2004 11:45:54 -0400, Sheldon Brown
wrote:

An anonymous poster quoted somebody without attribution:
I was under the impression that Shimano and Campagnolo tapers were
significantly different and not compatable.

You're not alone, many people are under this impression.

In my experience, however, it is almost never a problem in practice, if
you make allowance for the desired chainline.

Is there a rule of thumb you use? Something like, "When using a Campa
crank with a Shimano BB subtract X mm from the spec spindle length?"

Pat

Email address works as is.

Dye penetrant testing will make the crack more easily seen. A good auto
machine shop/supplier should have inexpensive kits.

Magna-flux, a bit more expensive, will reveal some flaws below the surface
of the metal.

Neither of these non-destructive tests are effective without some
understanding that may be beyond some here.

A proper visual inspection with understanding, good light, a 5X or 10X glass
and a clean field is a good beginning.

Brandt is usually correct. Note when his critics address side issues,
a.k.a. quibbles.


"Arthur Shapiro" wrote in message
news:tvm3d.329400$Oi.272261@fed1read04...
I'm working on getting photos of the spindle, but it may be infeasible as
the
bike is going back to the owner soon. It is known to be at least ten
years
old, gone through many, many centuries and doubles, and I'm told (by the
temporary holder) that it might date from the days when Merlin made its
own
bottom bracket. The guy is also apparently a Big (300#?) individual.

Jobst or anyone: is there anything the normal rider, not privy to an Xray
crystallography setup, can do to check for incipient failures of this
genre?

Art
Temporary usercode - to be deleted when spam starts. Use MyBrainHurts at
this ISP to reach me

John Forrest Tomlinson wrote:

On Sun, 19 Sep 2004 11:45:54 -0400, Sheldon Brown
wrote:

An anonymous poster quoted somebody without attribution:

Since then I have used Shimano BB's and spindles, even with my
Campagnolo cranks. That doesn't mean it can't occur again but it
hasn't for a long time.

and responded:

I was under the impression that Shimano and Campagnolo tapers were
significantly different and not compatable.

You're not alone, many people are under this impression.

In my experience, however, it is almost never a problem in practice, if
you make allowance for the desired chainline.

Is it possible to make chainline predictions just from the length of
the spindle?

Probably, if the spindles are precision, which Campy and Shimano are (at least
in the higher end stuff). Specialized also used to make very nice precision
spindles. (Some "for" Campy and some "for" Shimano.)

Or are their differences in the diameter of the tapered
section of the axle between the two brands. That is, will a
Campagonolo and Shimano spindle of, say 107mm both result in the same
chainline?

No. But I don't know the specific details -- I've typically done trial and
error rather than predictions.

jobst.brandt wrote: (clip) The smooth region you describe is the area that
failed all at once. The upper part of that picture, the granular part,
failed gradually in fatigue. (clip)
^^^^^^^^^^^^^^^^
Jobst, I am surprised that a person of your expertese would make this
layman's error. You have it backwards. A typical fatigue failure starts
with a tiny crack, which acts as a stress raiser, causing the crack to
propagate across the metal. During the load cycling, as the crack grows,
there is movement within the crack, which polishes the surfaces on both
sides. When the undamaged metal has dimished in cross section to the point
where it can no longer sustain the load, there is a sudden fracture. The
part that breaks suddenly is not worn--the crystaline surface is the exposed
metal structure. The crystaline structure was there all along.

The idea that vibration, or cyclical stress causes the metal to
"crystalize," making it weaker, is not correct. I am not a metallurgist,
but I did graduate with highest honors from the University of California in
mechical engineering. I mention this only to lend credibility to what I am
saying above.

'Layman's error' or mis-speaking? The body of Jobst's work, compared to
your claimed summa cum laude, earns him the benefit of the doubt.


"Leo Lichtman" wrote in message
...

jobst.brandt wrote: (clip) The smooth region you describe is the area
that failed all at once. The upper part of that picture, the granular
part, failed gradually in fatigue. (clip)
^^^^^^^^^^^^^^^^
Jobst, I am surprised that a person of your expertese would make this
layman's error. You have it backwards. A typical fatigue failure starts
with a tiny crack, which acts as a stress raiser, causing the crack to
propagate across the metal. During the load cycling, as the crack grows,
there is movement within the crack, which polishes the surfaces on both
sides. When the undamaged metal has dimished in cross section to the
point where it can no longer sustain the load, there is a sudden fracture.
The part that breaks suddenly is not worn--the crystaline surface is the
exposed metal structure. The crystaline structure was there all along.

The idea that vibration, or cyclical stress causes the metal to
"crystalize," making it weaker, is not correct. I am not a metallurgist,
but I did graduate with highest honors from the University of California
in mechical engineering. I mention this only to lend credibility to what
I am saying above.

On Mon, 20 Sep 2004 13:57:53 -0400, "Doug Huffman"
wrote:

'Layman's error' or mis-speaking? The body of Jobst's work, compared to
your claimed summa cum laude, earns him the benefit of the doubt.


"Leo Lichtman" wrote in message
...

jobst.brandt wrote: (clip) The smooth region you describe is the area
that failed all at once. The upper part of that picture, the granular
part, failed gradually in fatigue. (clip)
^^^^^^^^^^^^^^^^
Jobst, I am surprised that a person of your expertese would make this
layman's error. You have it backwards. A typical fatigue failure starts
with a tiny crack, which acts as a stress raiser, causing the crack to
propagate across the metal. During the load cycling, as the crack grows,
there is movement within the crack, which polishes the surfaces on both
sides. When the undamaged metal has dimished in cross section to the
point where it can no longer sustain the load, there is a sudden fracture.
The part that breaks suddenly is not worn--the crystaline surface is the
exposed metal structure. The crystaline structure was there all along.

The idea that vibration, or cyclical stress causes the metal to
"crystalize," making it weaker, is not correct. I am not a metallurgist,
but I did graduate with highest honors from the University of California
in mechical engineering. I mention this only to lend credibility to what
I am saying above.


Dear Doug,

I imagine that Jobst receives the same benefit of the doubt
and general good will that he exercises--or better.

Carl Fogel

"Leo Lichtman" writes:

The idea that vibration, or cyclical stress causes the metal to
"crystalize," making it weaker, is not correct. I am not a metallurgist,
but I did graduate with highest honors from the University of California in
mechical engineering. I mention this only to lend credibility to what I am
saying above.

When I see "University of California" all by itself like that I think
Berkeley, and ISTR that the college of engineering at Berkeley does
not award honors. Maybe I am remembering incorrectly, or maybe you
graduated from a different campus. I don't mean to question your
statement, it just looked funny when I read it.