TO NO-ONE'S great surprise, Audi dominated last weekend's 24-hour endurance race at Le Mans, in the bucolic Loire district of France. A hybrid version of its R18 sports car took the chequered flag—the first time a hybrid, from any manufacturer, won at Le Mans. Another Audi hybrid came second, and the company claimed third and fifth with a pair of turbo-diesels. Toyota returned to the race after a 13-year absence with two hybrids of its own—only to see one crash and the other retire with mechanical problems.

Though a novelty, the hybrids were not the stars of the show. A tiny triangular-shaped car known as the DeltaWing was giving the other 55 fire-breathing machines a run for their money when it was unceremoniously bumped off the track and into the crash barrier by one of the Toyotas. So ended a brave attempt to show that a car with half the weight, half the horsepower and half the aerodynamic drag could run rings round the dreadnoughts of the sport.

It was not the first time that a radical, lightweight design has challenged conventional thinking in motor racing. Something similar happened when Colin Chapman's featherweight Lotus 23, with Jim Clark at the wheel, made its debut at the Nürburgring's infamous northern loop in 1962. With its tiny 100 horsepower motor (a third that of its rivals), the Lotus 23 shot ahead of the field of ponderous Porsches, Aston Martins and Ferraris. After one lap of the rain-soaked track, Clark was 27 seconds ahead of the leading Porsche driven by the American ace, Dan Gurney. The world of motor racing had never seen anything like it before.

The following month, when two Lotus 23 cars—one with a 750cc engine and the other with a 1,000cc unit—were entered for the Le Mans endurance race, French officials promptly banned them for being too good. Chapman swore never to enter a Lotus car for the 24-hour Le Mans race ever again—and kept his promise till the day he died.

More than in any other motor sport, engineering improvements made to meet the gruelling demands of endurance racing feed directly into everyday motoring. Unlike Formula One cars or their IndyCar cousins, where teams focus on making vehicles that run furiously for a few hours, endurance events like Le Mans are more about building sporty but reliable machines that can run flat out for a full 24 hours, with only brief pit stops to take on fuel, replace worn tyres and swap drivers.

Also, much of the Le Mans circuit (officially known as Circuit de la Sarthe) is made up of public roads that are used for the rest of the year by cars and camions. Le Mans requires drivers to cope with rough, cambered surfaces as bone-jarring as anything everyday motorists face. To compete successfully, Le Mans cars have to be not only robust, but also have good fuel economy and be extremely stable, especially at high speed.

Long, fast straights dominate the Le Mans circuit. Because drivers were reaching 250mph (400km/h) before hitting the brakes for the sharp turn at the bottom of the three-mile (five-kilometre) downhill Mulsanne straight, two sets of chicanes were added in 1990 to slow the cars down. Even so, Mulsanne remains a savage test for aerodynamic stability. In 1999, a Mercedes-Benz CLR became airborne while hurtling down it, flying spectacularly over the safety fencing before landing in the woods beyond. That was not the first time. The CLRs had flipped twice during practise. Fortunately, no one was seriously hurt.

But the stability problems were enough to make Mercedes-Benz withdraw from the race and abandon its entire sports-car programme before further accidents could tarnish its public image. Last weekend, after being hit by a Ferrari, one of the Toyota hybrids also took to the skies before slamming into a tyre barrier at the bottom of the Mulsanne straight, fracturing two vertebrae in the driver's spine.

Such are the perils of relying on aerodynamic downforce for high-speed stability. Any interference, whether a nudge from another vehicle or a bump in the track, can upset the airflow over the car, especially on the wing at the rear that creates the bulk of the downforce. The car will often "porpoise", as the rear wing stalls and air pressure builds up under the nose, sending the vehicle tumbling end over end.

The little DeltaWing car at this year's Le Mans aimed to do things differently. Conceived by Ben Bowlby, a British-born racing-car designer, and built by Dan Gurney's All American Racers as a contender for the 2012 IndyCar Series, the DeltaWing carried no external wings, because of their propensity to generate drag along with downforce. Instead, the car got most of its downforce from the underside of the body.

To reduce drag still further, the DeltaWing's frontal area was made as small as possible. The front wheels were spaced just two feet apart, while the track at the back was a more conventional five-and-a-half feet. The result was a car with superior straight line and cornering speeds, and much better fuel economy than any of the IndyCars it was designed to replace.

Another feature of the delta layout is the way the wing's leading edges generate strong vortices that energise the air flowing over the rest of the wing. That keeps the airflow attached more effectively to the surface, allowing the wing to sustain much higher angles of attack before stalling. Used on a racing car, with the lifting surface beneath the vehicle and pointing down, the vortex effect can be used to keep the vehicle glued to the ground.

Though Mr Bowlby's novel design was deemed too radical for the IndyCar Series, his car was invited to participate in this year's Le Mans race as a “Project 56” entrant. Apart from the 55 cars that qualify to race each year, the organisers reserve one extra slot on the grid for a car considered so innovative that it should be allowed to compete without having to comply with the rules.

With its 1.6 litre, four-cylinder Nissan engine, the DeltaWing was far from the most powerful machine on the grid. However, like the featherweight Lotus all those years before, it tipped the scales at just half the weight of the big Audis and other “prototype” sports cars in the race. That let it produce a similar power-to-weight ratio of 600-700 horsepower per ton, but with a drag coefficient of just 0.24 compared with the 0.47 of the all-conquering R18s. What the diminutive DeltaWing lost in outright speed, it gained by not having to go into the pit so often to change tyres and refuel.

So popular was the little DeltaWing with fans that when it was forced into the crash barrier by a Toyota driver who did not see it creeping cheekily alongside, the number of people listening to the race commentary on Radio Le Mans fell by 30%. Will it race again? Probably. But most likely under different management. Mr Bowlby designed the DeltaWing as an open-source project, so other teams could build their own versions to race.