Why does Cav’s chain keep coming off after sprints?

Mark Cavendish's stage 13 win at the Tour de France saw him make history - equalling Eddy Merckx's record of 34 victories at the French grand tour. 

Interestingly, it's not just the Manx Missile's hands that have been up in the air after each of his four stage wins. We can't help but notice that Cav's chain has a habit of ejecting itself from the chainring as he crosses the line.

Image 1 of 3

And this isn't only a phenomenon taking place in France. The same thing happened back in April, at the end of stage two in the 56th Presidential Cycling Tour of Turkey in another closely fought sprint with Alpecin-Fenix's Jasper Philipsen.

Is there an issue with Cavendish's setup, is it pure coincidence that he's hit a bump - or even the timing mat - at the very moment he stops pedalling, or is there something about the way he sits up to celebrate that throws the chain off?

We asked Dov Tate of Parcours wheels, an Oxford engineering graduate who last conducted a groundbreaking study into the way wheels behave in real-world conditions.

"Interesting question!" said Tate. "First thing to note is that Deceuninck-Quick-Step are running Roval’s Rapide CLX wheels I believe. These use a DT Swiss EXP ratchet system on the freehub, so different to the pawl-based design that we use. I’m not 100 per cent sure on exactly how that freehub would respond under the (pretty extreme) conditions in a sprint.

"My instinct is that if you watch the sprint back in slo-mo, Cav is noticeably more 'violent' in his sprint body movement, then he’s the only one who has quite such an abrupt halt in his pedalling motion (when he posts up for the win)."

Tate explains: "He locks his left knee as he sits up, which probably puts a small amount of backpedal momentum into the drivetrain, whereas everyone else just rolls through the line in the drops. So his is a more jerky motion at the point of crossing the line versus the other sprinters who are smoother on their drivetrain. Given his chain will already have been 'sloshing' all over the place during the sprint, it really wouldn’t take much at all to cause it to jump and drop."

Image 1 of 2

Add to this the motion of the freehub, which will have been travelling very fast before the sudden stop. 

Tate says: "There is probably also a small amount of rotational momentum in the freehub, so as he stops pedalling, the freehub is likely still rotating forwards, along with the wheel. Plus there’s a non-zero amount of friction between freehub and wheel which will also contribute to this forward rotational momentum.

"So altogether, there’s likely a combination of some forward momentum in the chain, combined with either a complete stop at the cranks (or even a small reverse motion), which is enough to make the chain drop, given it’s already moving around violently from the sprint."

Tate sums up: "I’d say it’s not one single cause, rather the 'perfect storm' of several micro contributors."

We've asked for a comment from Deceuninck's technical staff on the ground at the Tour and will update this story as soon as we have one.

Although in slo-mo it looks as though Cav has had a lucky escape, as long as he only sits up and celebrates when there's no more pedalling to be done it shouldn't cause him a problem.