The cycling industry isn’t doing enough to prevent fork steerer failures.
That’s the opinion of a man who would know: Cervélo co-founder Phil White, whose company suffered a large fork recall in 2008 following fork failures that resulted in multiple crashes and injuries.
Cervélo is not alone in that sorry episode.
The cycling industry has seen a rash of fork steerer-related problems in the last few years, from nearly every major brand sold in North America. Big brands with big test labs have experienced consumer injuries and expensive recalls that the manufacturers say resulted from a mix of engineering and manufacturing flaws. Neither aluminum nor carbon steerers have been immune, and White believes the frequency of problems points to a deficiency in testing.
Specifically, a lack of testing that would replicate more than just riding forces.
Cervélo was initially unable to replicate its own problem of steerers cracking just below the stem; traditional impact tests were not producing the same results. The company set out to find the source of the issue, and then develop a test that would prevent it in the future.
“We tried all sorts of things to mimic what was happening,” White said. “We tried different headsets, different stems, and it took us a long time, and we were seeing more and more failures.
“I remember when we figured it out. We had it in a box, a frame in a box, and we took it up on a ladder, and we held it up over our head and we dropped it right on the steerer, on the top of the box. We were able to finally get it to crack. We saw a little micro crack, and then we put it on the test machine to see how long it would last. It failed really quickly.”
Such an instance is well outside the realm of normal use, but it’s a frequent enough occurrence (almost every bike is shipped inside a cardboard box at some point in its life) that White believes it should be part of the standard test regimen at every manufacturer
“When you design a product, you don’t design for just how it’s going to be used in the regular case, you design it for foreseeable abuse. You design a car for the street, but what happens if the guy isn’t paying attention and drives off the road onto the shoulder and hits a big rock? That’s foreseeable abuse. You need to design around that,” White said. “You can’t design it to never leave the road. The concept of foreseeable abuse is a pretty common part of the design. From our standpoint, this is foreseeable abuse.”
And the sort of damage Cervelo saw is not as rare as one would imagine. “You can damage [the steerer] as easily as dropping it off the top of your car, or if you put the bike in a soft airline bag and the baggage guy drops it,” White said. “If you drop it, and it falls on the steerer, that’s enough on a lightweight fork to damage it. And you won’t see it, because nobody ever takes their fork out of their bike to inspect the steerer.”
But such a test, one that replicates an impact to the top of the steerer and then combines that with a traditional fatigue test, did not exist. According to White, it still doesn’t except at Cervélo.
That test mimics the way Cervélo’s fork failures occurred in 2008. A fork is clamped into a fixture at the headset bearings, just as it would be in a frame. With spacers and a stem installed, Cervélo impacts the back of the stem, mimicking a bad crash or, more commonplace, a shipping box or an airline bag incident. A bar is then attached to the stem and twisted (like you were climbing and yanking on the bars) in a fatigue test. If the initial impact causes a crack, the fatigue test results in failure well below an acceptable number of repetitions.
“There is no international standard that addresses steerer tube strength,” explained Sam Pickman, head of Specialized’s test lab. “This is a massive gap because there is a ton of stress on the steerer tube at the stem clamp, especially when you run a long stem.”
Specialized, like most major manufacturers, performs a version of a steerer tube strength test. It installs a fork in a frame with the correct stem and all the right headset parts and then pushes down on the stem at the handlebar clamp, “as if you were stuffing your front wheel into a pot hole,” said Pickman. The steerer must withstand a minimum moment — the load applied by the stem length — to pass. The moment figure was established through a research project that determined the maximum load a rider could apply to the hoods for a given type of riding.
Trek performs a similar test, as does Giant, Felt, and most other brands. “To my knowledge all of the big U.S. manufacturers test their steerer tubes in some kind of overload case,” Pickman said.
Doug Martin, marketing manager at Felt, corroborates that claim. “We have a battery of tests we perform on top of the already stringent EN standards — standards which are actually higher than, or in addition to, CPSC standards,” he said via e-mail.
Felt already performs three different fork tests — a three-point bending load test, a steerer strength test that involves placing two ends of a 300mm steerer on blocks and adding weight to the middle, and a version of the EN frontal impact test that punishes a fork with 20% more energy than is required — and adds the extra step of randomly pulling production units to test as well. So putting an additional test in place, to take it to the next level, is “not going to inconvenience us at all,” said Felt’s director of R&D Jeff Soucek.
Other big brands, with big test labs, would likely find that another industry-standard test would take minimal effort. Soucek says he would be for the implementation of some sort of steerer tube test standard, saying, “Putting it in place and requiring it, I can’t see how that would hurt.”
But while White sees such tests as a step in the right direction, he would like to see the rest of the industry adopt the test Cervélo developed and submitted to the ASTM (American Society for Testing and Materials) four years ago that takes into account the impacts that can come from shipping and other “unforeseeable abuse.”
“I’m not going to slag anyone, I’m not trying to target anyone, but as an industry I don’t think we’re doing a good job,” White said. At the very least, he says, there needs to be a standard: whether it is Cervélo’s test or another is less important than simply making sure one exists.