From: jbrandt@hpl.hp.com (Jobst Brandt)
Subject: Re: Frame lifespan
Date: 29 Aug 1997
Newsgroups: rec.bicycles.tech

Dave Blake writes:

> There are some exceptions to the rating system.  For example, Tom
> Ritchey claims to build such a high performance frame that it is
> only designed to last one race season, and he will not warranty it
> against breakage beyond a year. By my rating system that would make
> him pretty lame. Of course he does not cater to people that value
> durability in their frames.

Tom built me the lightest road bike I have owned and it served close
to 100,000 miles before I ran into a ditch too deep to come out of and
wrinkled the frame.  Knowing Tom, I can't imagine under what
circumstance this claim was made, but it is not his style.  I don't
know of any such "good for one year" frame because if it makes it
through one year of the kind of riding he and I consider standard, it
would last much longer than a year.  Because frames must withstand
high single incident stresses now and then, they automatically have a
high fatigue threshold.

Therefore, your scenario of failure sounds like the mythical "One Hoss
Shay" of Oliver Wendell Holmes, that was so well designed that all its
parts failed at the stroke of noon on the hundredth year, leaving only
a heap of dust.  The concept is not new and is part of the fabric of
engineering and human experience, but it is only a myth that we like
to believe.

Jobst Brandt      <jbrandt@hpl.hp.com> 

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Frame lifespan
Date: 1 Sep 1997 22:59:39 GMT

Dave Blake writes:

> I witnessed the broken frame on race day while I witnessed the rider
> carrying the now two piece frame out of the mud at Quantico. As he
> was a fellow racer, I saw him many times later that year and asked
> about his followup. He was told by Ritchey that the frames were
> designed to last for a year, and that they would give him a
> wholesale replacement for his frame which was in its second season.
> He expected more. The frame broke clear through its down and top
> tubes causing a crash (without injury).  The rider weight was around
> 180 lbs.

I was not aware that you were referring to a dirt bike ridden in
competition.  For these bikes most warranties are highly limited
because the bikes are ridden as hard as one dares to go and that limit
rises as the rider realizes the bike can take more.  In other words
racers in the dirt run their bikes close to yield.  I don't know of a
cyclocross rider who separated his bike on a jump, unless the bike
already had a wrinkle.  That this rider had two tubes break
simultaneously is an indication that the frame had been previously
wrinkled and cracked.  You cannot break a frame with a single overload
and to make two tubes fatigue fail at the same time.  This is so
unusual that I cannot imagine how it could occur.

It seems that you did not experience this first hand but rather "a
rider" whom you observed at a race.  I don't think, from what you've
said that you know enough about the circumstances to be sure what the
history of the frame was.  Riders who have broken equipment are not
happy about it and are prone to point fingers anyway, be that their
own making or not.

> Your melodramatic tone above will not change the facts, and I am not
> making any of this up. You should not feel that just because you
> designed the badge for the head tube of Ritchey bikes that he builds
> a durable rig. He will tell you the same. He did not warranty a
> frame replacement in this case, and the rider expected it from such
> an expensive frame maker.

This is not melodrama but the way people see their machines.  It does
not happen that way.  Bikes fail at a high stress point that usually
shows a wrinkle when closely examined, as another writer pointed out.
Two tubes do not break at once unless they had cracks in them for
some time before.  I have and enough frame tubes crack to recognize
the effects.  Frames falling apart requires multiple simultaneous
failures.

> I indeed found it highly unusual.  In the early 90s frame breaks
> were quite common in MTB racing, and in every case other than this
> one that I am aware of the company replaced the frame free of
> charge. I myself had two frames replaced in 1992.  Builders have
> gotten somewhat smarter since then about the relevant design
> criteria for an off-road rig. Maybe even TR, I don't really know.

Frames have been cracking for as long as they have been made.  It
isn't something new.  Bicycle have, by popular demand, been built as
light as possible, approaching the failure point for a long time.
Those who ride harder test that limit.  The reason this can be done,
is that most failures are restricted to parts that do not fail
catastrophically.  Fork crowns, steertubes, and stems are not allowed
to break, while other parts allow the rider to stop and seek aid.

Jobst Brandt      <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Frame lifespan
Date: 3 Sep 1997 01:20:06 GMT

Bruce Frech writes:

> Half-life: I used that term in the sense of exponential rate of
> decay.  So a "full life" is not 100,000 miles, just that only 25% of
> frames will survive 100,000 miles, while half will make it to 50,000
> before suffering a fatigue failure.  Recall that this 50,000 mile
> half-life is just a guess of mine.  I was hoping someone had more
> information than me who could give a better estimate of this or some
> similar estimate.  I am not that close to the industry.  My brother
> owns a shop, there is a frame builder that works there who regularly
> repairs broken steel frames but that is still a small sample size.

I have ridden bikes plenty long, and in that time have had more than
two dozen Campagnolo Record cranks break, two crank spindles, 2 pedal
axles, many rear axles and a few Dura Ace cranks among other lesser
components.  My Italian frames had one seat tube to BB failure in less
than 1000 miles and two fork blade failures in about 40,000 miles.
My other bikes all died a wrinkled tube death or were sold so that I
could try a new geometry.  I have ridden several bikes to near 100,000
miles but none got there.  The cracked frame was obviously an overheat
problem and the fork blades were caused by the beautiful but stupid
Cinelli "sloping" internally lugged fork crown that has a large
internal discontinuity at the transition from crown to fork blade.

> I usually replace my road bike after 30-50,000 miles "just because".
> That means I figure some parts are getting old, close to fatigue limits
> and because new stuff just is nice and ....

But one failure does not a trend make.  My BB breakout was on a new
frame that was obviously ruined in manufacture.  I draw no conclusions
about other bikes from that.  The existence proof of later frames I
rode for nearly 100,000 miles show that under the roughest conditions,
frames can and should last that long.  The short life failures only
prove that frames are not built as well as they should be.  A
beautiful finish is not the whole story.

> Stresses that might exceed the level where steel has infinite
> lifespan?  I mentioned some situations where high stresses may
> occur: riding over bumps, sprinting and climbing.  Sprinting and
> climbing (out of the saddle) both induce out-of-plane forces that
> result in twisting of the frame.

That's what a bike is made for.  I ride many rough dirt roads and
trails as bicycles did 50 years ago.  Today many frame builders have
no concept of the forces frames must withstand when descending a
mountain road, braking on asphalt washboard at high speed.  When
riding with Ritchey or Johnson, two premier frame builders, in the
mountains, they cannot help but comment on the lack of riding
experience that many frame builders have.  "I wonder how many frame
builders have experienced a descent on Sonora Pass at speed and had
their life depend on it." is their comment in so many words.

> I don't know if any of these forces are above the infinite lifespan
> limit for an average frame.  Perhaps someone with a lot more info
> could answer this.  This could either come from massive actual data
> or from modeling a frame with FEA and ...  Both are out of my
> expertise.

This is not something easily captures by an analysis program, because
the inputs are unknown.  That is a major problem with people who
profess to compute theses things.  The "I wonder how many frame
builders have experienced a descent on Sonora Pass at speed." applies
to these people.  I regularly have so called experienced riders tell
me that I can't ride a road bike where I do.  That gives me an idea
what perception they have of bicycle frames and how they work.

Jobst Brandt      <jbrandt@hpl.hp.com>