Well, I'm not an ME, but there are a few observations I'd like to insert into this thread.
1. I have seen this type of head tube failure, but I can only specifically recall one, as I brazed a reinforcement repair to it. It was a lowly Raleigh 3-speed, ridden daily, snow & sun, by an eccentric engineer who gave me the oil bath chainguard from it. I question the mechanism used to secure the head tube during this test. The headset and fork steerer bring a great deal to the equation. Also, the frame is not rigidly held at the head tube and rear dropouts. A realistic test would allow significant dampening at these points.
2. The site is somewhat crudely translated. I interpreted the statement about the rider tiring of the color as a euphemism for stating that the level of stress incurred during this test would only occur after many, many years of hard riding, much longer than most people who ride that hard keep their bikes.
3. Ditto the statement about "design effort." I read this as construction methods, not craftsmanship. I think they were saying that steel is not inherently bad, just that lugged brazing may not be the strongest way to join it. As we know, brazing may not be the strongest, but it may be the best method, as evidenced in over a century of real-world road testing.
I agree that these statements are not of the caliber that one would expect from an engineering analysis, but I don't really know the target audience for this piece. Cycling is a bastion of untrained engineering wanna-bees--like me, for instance.
4. Finally, 25 hours of brutish out-of-the-saddle sprinting is more than any
frame I can think of would sustain. We're not talking about hill climbing
here, but all out effort, probably only maintained for 10-20 seconds at a
time. The article began by discussing the concept of fatigue life vs.
infinite life design and mention that the latter is not a useful design goal
for a bicycle. I think this is where they made their most egregious
mistake. By not designing the test to simulate the actual conditions the
frame is likely to experience, the test becomes invalid in light of their
stated objectives. It is an interesting test, as referenced by the
discussion it prompted, but not terribly useful.
> http://www.damonrinard.com/
Steve Barner, Bolton, Vermont