Sign up for our newsletter


The science of cobbles

James Witts
31 Mar 2019

As the pros gear up for Paris-Roubaix, we look at how scientists are working to make riding the cobbles easier

This year’s Paris-Roubaix boasts more than 50km of cobbles, which is precisely why this classic race in northern France is described as ‘The Hell of the North’. For manufacturers and teams it’s one of the highlights of the calendar.

It’s also a platform to showcase new technologies designed to help bikes roll smoothly over the pavé.

It’s why Pinarello created a suspension system at the top of the seatstays on its K8-S and why Specialized introduced its ‘FutureShock’ suspension betwixt head tube and stem on its latest Roubaix. 

But is there any scientific evidence to suggest whether these and other similar innovations have real merit and, if so, which is more important: damping at the front of the bike or at the back?

Monitoring vibrations

‘We’ve loaded our bikes with sensors when recceing a race like Roubaix to add empirical data to the feedback we get from riders,’ says Carsten Jeppesen, head of technical operations at Team Sky.

‘It’s why many riders will choose rear suspension and also why, for instance, they wouldn’t go for front suspension, which some have tried but have found just too foreign.’

Jeppesen also notes that riders choose a longer wheelbase, which goes to explain the thinking behind the Pinarello K8-S, with it’s springy rear end and longer wheelbase than the F10.

But it’s not very scientific. We’re after independent, verifiable evidence, and rather surprisingly for a sport where data is king, only one independent study exists examining the vibrations emanating from the Roubaix cobbles.

Scientist and cyclist Sebastien Duc took part in the 2015 Paris-Roubaix Challenge, an amateur event that takes place the day before the pro race. The 1.80m, 68kg Duc aimed to measure not only the scale of vibrations created by the cobbles but also where on the bike and body they’re at their peak.

‘I loaded my Specialized Roubaix Expert with two tri-axial accelerometers – on the stem and seatpost – and set the tyre pressure at 5 bar [about 73psi],’ Duc says. ‘I then measured the RMS, VDV and vibration level…’

OK, stop right there – explanation required. In this oscillating world, RMS (Root Mean Square, measured in m/s2) is essentially the average vibrational value, in this case of a ride on the cobbles, while VDV (Vibration Dose Value, m/s1.75) represents the cumulative value. Vibration level is oscillations per second, or hertz (Hz).

Collecting all the data after the event, Duc discovered that during his 139km-long, 15-cobbled-sector ride his speed varied from 19.1-27.8kmh; his heart rate fluctuated between 122-155bpm; his cadence was between 79-87rpm; and power output varied from 167-235W. 

‘According to RMS and VDV values, vibration exposure is more intense at the hands than at the cyclist’s seat, regardless of the speed or difficulty of the cobbles,’ Duc reveals.

For reference, ASO categorises the cobbles by difficulty from two-star (relatively easy) to five-star (bone-shattering), and this year has introduced colour-coding to make it easier for TV viewers to identify them .

‘Over the four-star sections, RMS equated to 35m/s2 at the stem compared to 28m/s2 at the seatpost.’ In the workplace – say ploughing a field on a tractor – anything over 10m/s2 is deemed dangerous. 

So that’s it. Sky’s Jeppesen is wrong and manufacturers should focus their efforts at the front of the bike instead of the rear. Of course, it’s not that simple.

‘The frequency of vibration levels was actually higher at the seatpost,’ says Duc. ‘Over the three-star stretches of cobbles, 30Hz at the seatpost compared to around 20Hz at the stem.’ 

In short, the vibrations at the rear were less intense but more frequent.

Sparseness of studies

So far, so inconclusive. We needed more data, but Cyclist had to be creative. Cue research by Paul MacDermid of Massey University, New Zealand, who compared the vibration impact of road cycling to off-road cycling.

His research wasn’t on cobbles, but MacDermid says he can make a good estimate of the results based on his own data.

Take the RMS value. On MacDermid’s tests, the left-arm accelerometers measured 18m/s2 and 27m/s2 for road and off-road; at the seatpost it was 12m/s2 and 18m/s2.

MacDermid says cobbles would generate similar vibration proportions and that subjects would have to generate around 30% more power to maintain the same speed, which is significant as power is related to vibration.

‘This is based on Training Peaks data from Mat Hayman’s 2016 win, when in the Arenberg Forest his average power went up by 44% and his heart rate by 20% [compared to earlier in the race],’ says MacDermid. 

Of course, Hayman’s increase in effort can’t entirely be attributed to the extra vibrations stimulated by those cobbles – it’s safe to assume Hayman also put the hammer down in a section notorious for making or breaking the race.

But if we broadly apply that 30% increase to cobbles and the increased vibrations, we’re looking at RMS values of over 30m/s2 through the handlebars and over 20m/s2 through the saddle.

So it would seem that again, up front is the area that requires greater damping focus. ‘You could certainly argue that, because propelling the bike over the bumps will elicit more upper-body movement,’ says MacDermid.

‘That’s supported by further data we have which shows on uphill sections at speeds of around 16.5kmh, vibrations were typically greater through the bars and arms than through the seatpost.’

His figures alter, however, when the road tilts downhill: ‘We performed another study looking at the effects on the bike when heading down over 13cm steps.

‘The results showed that the seatpost and ankle took the greatest battering with handlebars a close third.’

Built for the cobbles

It’s inconclusive whether suspension would be better employed at the front of the bike or the rear, but it could be the case that neither has quite as much influence as the riders themselves.

‘A study in France showed that the heavier the rider, the lower the Vibration Dose Value,’ says MacDermid.

Essentially the 80kg-plus guys naturally dampen the vibrations more than the 60kg guys.

And another French study showed that a rider’s posture also has a significant influence on the VDV, in particular the position of the forearms and wrist angles, and that optimised bike geometry can reduce this figure by up to 50%. 

There’s also the issue of riding style, says Jeppesen: ‘As an example, if you look at Fabian Cancellara, who’s smooth on his bike, he’d probably experience fewer vibrations than someone like Ian Stannard, who’s more about brute force and stomps the pedals.’ 

So there you have it. When the pro riders hit the Arenberg at this year’s Paris-Roubaix, some of them will appreciate extra damping at the bars, while others will be happy with suspension at the rear (and all will go for wider, softer tyres than usual).

But the winner will most likely be the one who simply knows how to ride the cobbles best. After all, last year’s winner, Mat Hayman, was on a Scott Foil – an aero bike that makes few concessions to comfort.