Cycling has more dimensions than any other mainstream sport. In a few moments, the deciding factor can pivot between the most brutal of physical ability to the subtlest of tactics; from a team game to an individual battle; from a pure willingness to blot out the pain to some of the world’s most complex aerodynamics. The last one was the trip step. Aerodynamics is not what the next Bradley Wiggins is fantasising about.
But he should be, because on anything other than the steepest of climbs, the air that swirls and tumbles all around a cyclist is what bike racing is about. On the flat, by the time you’re doing 25kph, more than half of the total resistance you have to overcome is from the air. Much past 30kph, and it’s the only thing worth worrying about.
To put the numbers a different way, let’s take a look at last year’s World Time Trial Championships and see where aerodynamics gets us. For purposes of handy comparison, we’ll measure it against the favourite metric of 150 years of bike riding: weight. Now, the course for the men’s TT wasn’t mountainous but it was distinctly lumpy. A couple of ten per cent climbs and numerous little jabs and growlers. An average to hard kind of day out, in other words, and certainly not one for the pure power riders.
Every kilo that a competitor could pare off himself or his bike would gain him about four seconds over the 46km. For a pro rider, on a pro bike, a kilo is a whole world of weight. But for every five watts’ worth of drag he could save, he would gain 16 seconds. Five watts of drag is trifling. It’s less than the difference between wearing a fitted helmet visor and a pair of sunglasses. There’s not a bike rider in the world, even at pro level, who couldn’t find five watts somewhere in his TT setup if he had enough money and patience to find it.
Finding it is, of course, the problem. If you want to measure aerodynamics with any sort of delicacy, you need something pretty sophisticated. You need a wind tunnel. There aren’t a lot of tunnels configured for testing bikes and riders. Of the three in the UK, the most used is Drag2Zero at the Mercedes-AMG-Petronas Formula 1Team headquarters at Brackley in Northamptonshire. The tunnel is on a quiet corner of the site, in a large, anonymous looking grey building just beside the helipad.
Inside, there is a scrupulously tidy control room full of computer screens, with windows looking out onto a big, white painted space with large ducts at each end and a bicycle bolted to a low platform in between. The concept is self explanatory. Blow wind in one duct. Suck it out the other. Measure what happens in between.
On the day we visited, the bolted down bike belonged to Julia Shaw. Julia is a British time trial specialist; a multiple national champion and a Commonwealth Games medalist. She’s worked with Simon Smart from Drag2Zero, who runs the Brackley tunnel, for several seasons. Neither could remember off hand how many tunnel sessions they’d done together.
This one was yet another exercise in refinement – to try a few small changes in position and to test a couple of bits of equipment. They weren’t looking for huge gains. After several seasons of testing, the margins have grown small. There was a simple rhythm to the work. Julia adopted her TT position and started pedalling. After a few moments, Simon, sitting in the control room, said: “Wind on, please.”
With a couple of taps on the keyboard, his colleague Paul carried out the command. He’s quite certainly done this too many times to still notice the action’s messianic overtones. There were a few moments of silence while they looked at the flickering numbers on their screens. The figures are for the most part various statistical breakdowns of the aerodynamic drag – the weight of wind pulling the rider backwards – as measured by the platform, which essentially functions as a sideways set of scales.
“Five at 48,” said Simon, and Paul tapped on his keyboard to turn the whole platform through five degrees, to simulate the effect of a slight crosswind. Forty eight is the wind speed in kph. There were a few more silent moments, then: “Ten, please.”
Finally back to zero, straight headwind, to check that the numbers had stayed consistent with the start point; that Julia hadn’t shifted position even slightly during the run. Keeping the zeros consistent requires huge concentration on the part of the rider. Finally: “Wind off, please.” Julia kept pedalling until a final measurement was taken as a zero. The whole process took seven or eight minutes.
As soon as Julia stopped and got off the bike, Wal the mechanic took her a much needed drink – the tunnel is hot and so very dehydrating – and started to change the bars for a different set.
Meanwhile Simon came into the tunnel to explain how the numbers were looking. As the discussions continued, Wal completed the handlebar change – in this case narrower on the elbows and a little higher – and then it was back to first positions. Julia remounted and Simon returned to his screen. The process was repeated.
During this run Simon went back into the tunnel. On the edge of Julia’s field of vision, and in the calm air just outside the wind, he bent over into a riding position, rounding his shoulders into his neck, to remind her to keep her position consistent. They would have another two hours of tweaking, measuring, riding and (from Julia’s point of view anyway) desiccating to do before they’d be finished. Narrower and higher, by the way, was not an improvement. It was about five watts worse.
Aerodynamics is not just hard to quantify, it’s often hard to get any sort of a grip on at all. The idea that you could ride faster if you addressed what all that air was doing is hardly new – boneshaker riders knew perfectly well that a low position got you to the bottom of a hill a lot faster and, given the the rutted roads of the 1860s and the absence of brakes, probably a fair amount deader too. The first disc wheel was marketed in the 1890s. It was fast but it was also the weight of a piano and, over a rough road surface, approximately as comfortable.
But aerodynamics stayed just a factor, not the factor. The hour record – the simple matter of how far a man can ride a bike, unpaced, on a track in one hour – was the ultimate time trial for most of the 20th century and gives you some sort of guide to the attitude to aerodynamics. Very quickly the riders switched their position from a matronly upright to a tuck on dropped handlebars. They’d got there before World War I. But then more or less nothing changed for almost 70 years. If you wanted to get aero, you got low. If you could get your chin onto the stem, good for you. Otherwise, weight remained most riders’ obsession – and on a track, weight made almost no difference whatsoever.
There were blips of innovation. In the 1920s, Swiss rider Oscar Egg, holder of the record since 1914, devised the Vélo Fusée – the rocket bike. Rather prosaically, it was just an ordinary bike with a big cone behind the saddle to reduce the low pressure area behind the rider. In fact the theory was sound, but it didn’t work in practice.
Frenchman Maurice Richard beat Egg’s record in 1933 and then, in 1936, used a wind tunnel to prepare for his attempt to better his own mark. There are a few surviving grainy photos of him in the tunnel – he’s on a bolted down bike with a huge dial behind showing the weight of wind as it pushed against him. If he made any changes in response to what he learned, they were subtle. Photos from his subsequent hour ride look indistinguishable from ones taken at a track event three years previously.
His jersey was still flapping defiantly in the breeze.
Even by 1972, when Eddy Merckx broke the record with the benefit of the thin air of Mexico City, the sum total of aerodynamic progress amounted to not much more than silk racing jerseys. They fluttered in the air with more finesse, but no less turbulence, than Richard’s wool.
Francesco Moser in 1984 was when it all changed. Moser broke Merckx’s record dressed like an extra from Flash Gordon, riding an impossibly elegant bike with a pair of beautiful, shimmering disc wheels. Within a year or two, there were aerodynamic high jinks everywhere. Disc wheels, skinsuits, comedy hats, front wheels the size of castors, bikes made out of wire… A chaos of ideas and postulation that was in no way diminished by the fact that those people with some grasp of aerodynamics knew little about bicycles, and that those people who knew about bicycles knew almost nothing of aerodynamics. You could be confident that half of any aerodynamic setup was guess work.
Gitane reinvented the Vélo Fusée for Laurent Fignon’s aborted attempt at the world hour record – it still didn’t work. Thierry Marie tried it too, and it wasn’t third time lucky. When Greg LeMond famously debuted tribars in the Tour in 1989, the general consensus was that he was an idiot because the arm position would prevent him from breathing. He ignored them and hammered Fignon in the final TT to take the Tour. But he’d have hammered him by even more if he hadn’t been wearing an ‘aero’ helmet the size of a Zeppelin.
The next landmark was the Lotus 108 that Chris Boardman used to win the Olympic individual pursuit in 1992. It was the first setup to come from a grown up tunnel testing programme; one that was focused as much on the rider’s position as on the machine itself. Even so, it was the bike that was the showstopper. It made Moser’s look like a climbing frame and demonstrated just what was possible when you combined imagination and a scientific approach.
The 108’s successor – the road going Lotus 110 that Boardman used for his Tour prologue win in 1994 – has only been overtaken in aerodynamic terms by the most recent generation of time trial ‘superbikes’.
Back when Boardman took Olympic gold, there were wild divergences in opinion about just how much difference all that tunnel testing really made. Okay, so the Lotus was the world’s most aerodynamic bike, but how much faster did it go in practical terms? Some of those involved in the development process stopped only just short of suggesting that they were king makers – they could make anyone fast enough to be Olympic champion.
Others reckoned the whole carbon shebang did nothing more than intimidate the opposition. Boardman never expressed an opinion. As far as he was concerned he had his medal, Lotus had acres of publicity, and the only loser in the whole process had been silver medalist Jens Lehmann of Germany.
So how much difference can the testing make? Well, to get back to Julia Shaw in the tunnel… Her coach Jamie Pringle was there. His reckoning was that since her first visit to the tunnel four seasons previously – a point when she was already a national champion several times over – her total drag had reduced by somewhere between ten and 15 per cent. He was basing this not on just repeating the tunnel numbers but on what he’d seen on the road in terms of speed against power. To put it another way, it’s somewhere between one and a half and two kph. To put it a third way, two to three seconds a kilometre. Most serious riders would reckon that was an awful lot.
That includes changes to her position, a different bike, different wheels, a different helmet and a different skinsuit. Most people, including Jamie and Simon, concede that Julia has found more than average in the tunnel for a rider of her level. It could be that despite her palmarès, her initial setup wasn’t that good – after all, Julia is a woman who can produce a lot of horsepower and in cycling, as in life, you can solve a lot of problems with more horses. It could be that she has been better than most at holding the ‘ideal’ position on the road. It could be all sorts of things.
The complexity of the aerodynamics involved is quite something. It’s the reason why it’s not possible to simply work out what’s best, hand it out as a general recommendation in the pages of a book or magazine, and move on. There are a few vague general rules – keep the frontal area low being the main one – but after that it’s really hard to find principles that apply universally.
In particular, the rider is a problem. The drag from a rider is something like seven times that of the equipment. This is compounded by humans being fundamentally unaerodynamic lumps, with pretty limited ranges of adjustment, at least if they’re still going to be able to actually ride a bike.
Even the basic speed of a bike is an issue. “Bikes are in a very complicated area,” explains Simon, “because at that speed there is both laminar flow and turbulent flow. Almost any other speed range would actually make life simpler.”
Laminar flow, for those of us who have forgotten our GSCE fluid dynamics, is where the air flows smoothly in parallel layers across the surface of, in this case, a bike or rider. Turbulent flow is where the air swirls and eddies at random within the few millimetres above. Crucially, both of these are flows that are ‘attached’ to the surface. They create relatively little drag – turbulent a little more than laminar – because the air outside the immediate surface layer remains undisturbed and flows round the object.
The big issue is when the flow detaches from the surface – this is what creates the wake behind a rider, and that’s where most of the drag comes from. Turbulent flow will stay attached to the bike and rider for longer than laminar, flowing further round them, meaning that when it does detach it creates a smaller wake. A lot of the more advanced aerodynamics in cycling involves managing the transition of laminar flow into turbulent flow, and consequently where the air round the rider finally detaches. The whole issue of flow over the surfaces and the attempt to manipulate it to reduce the wake is the reason for all the different textures on current skinsuits, and even the very prominent seams that some of them have.
It doesn’t help that no two riders are the same. “Julia’s back is a great shape,” Simon says. The highest point of it is forward and it then tapers smoothly, almost like a wing profile. “So the flow detaches quite low down.” If your back happens to be even a slightly different shape, just copying her position won’t get you anywhere? “Exactly.”
Simon reckons he can get someone maybe ten per cent of the way aerodynamically just by looking and tweaking. But the details always need the tunnel – it’s impossible to eyeball them.
All of that is before you introduce yaw. Most of the time in the real world, a rider isn’t dealing with air that comes from dead ahead. As soon as there is any crosswind, the effective angle starts to swing to one side. Positions that are killer into a block headwind might be a disaster when there is a crosswind – you can get over a kilometre an hour of difference. So you need to run the tests at different angles and then decide on the best compromise, depending on the rider and the anticipated conditions they’re going to race in.
You would hope that bikes and wheels would be simpler. You’re not trying to fine tune a lump of meat in a Lycra stocking and bikes don’t complain that they’re thirsty and getting cramp.
“I thought bikes would be the easy part,” admits Simon. “But you have a lot of bluff bodies [blunt bits] and a very complex wake. The total forces involved are very small – a current top end time trial bike generates very, very little drag, much less than the rider – so you’re trying to find savings on a total that’s already incredibly low.”
It may be difficult but it’s becoming an increasingly pressing part of all aspects of riding. Until recently aerodynamics was a topic for time trial or track pursuiters only. The last few seasons have seen it move to road racing. Most of the manufacturers have an aero road race bike in the line up. Teams regularly field riders in skinsuits with an eye to a break or even a sprint, and in this year’s Tour several squads used aero road helmets for flat stages.
Each step someone has taken along the road of aero-in-a-peloton has been greeted with something only marginally more polite than contempt. The problem is that physics doesn’t give up just because you laugh at it. Even in a bunch, the wind is still the enemy. A team with a little less to push all day will have a lot more with which to pull when they need to. There’s a lot more aero on the way.
And yes, the tunnel at Brackley has seen quite a few riders testing their road setups. Keep your elbows in. Keep your head low. And make sure the flow doesn’t detach too high up your back. That’s the one you really need to watch.
Julia Shaw spent slightly more than two hours in the tunnel while we were there. There were changes to bar and hand positions and a fair amount of playing around with the saddle. A couple of helmets came and went, and a different skinsuit arrived.
Some things worked at some angles but not others; some didn’t work at all. Overall she didn’t find much this time. But she and Simon were perfectly happy with that – they’d crossed off several things they’d wanted to try. The process is such that the gains get harder to find the closer you get to perfect. In that respect it’s the same as cycling has always been.
Dowsett received his Aeroad CF SLX the day before the Tour de Suisse. One of the first of the new model to leave Koblenz, it served briefly as the Movistar rider’s training bike (“It was quite novel travelling to races with a bike,” he remembers), but gained its first outing in a competitive fixture: a minor event called the Tour de Suisse. “The first ride on this was at the Tour of Switzerland. I hit 118kph on it in one of the passes. I was very confident in it. I was at home on it straight away.”