Shaving Seconds

CFD and 3D scanning give racing cycles an edge

The differences between motor racing on two wheels and four go far deeper than just arithmetic. In Formula 1, for instance, races are routinely won and lost in the pits; on motorbikes, the archaic practice of overtaking still regularly occurs. It is on the drawing board, however, that the differences between the two sports really emerge.

At any particular race, a Formula 1 car represents the current state of the art, a snapshot in ongoing research and development. Teams recognize the importance of the most advanced design, simulation and testing techniques — and spend, spend, spend accordingly. The teams racing high performance bikes are every bit as keen to succeed, but here the emphasis tends to be more on fine tuning and rider skill rather than turning out machines that evolve steadily (sometimes dramatically) throughout a season. Part of this focus rests on motorcycle racing’s rules and regulations: in some races, every entrant must ride an identical bike, so room for design innovation is distinctly limited. And money is certainly an issue — that individuals still race on a reasonably competitive basis with more highly funded rivals is a reflection of the fact that cash is spread somewhat thinner than in Formula 1 racing.

Specialists at Reynard Motorsport, which manufactured many winning race cars until they closed last year, believed that some advanced techniques they used for optimizing car performance had a significant role to play in the saddled sport, and set out to prove it to the motorcycle manufacturers.

Aerodynamicists are heavily reliant on computational fluid dynamics techniques. Reynard Motorsport established a division — Advantage CFD, now owned by British American Racing — dedicated to nothing else. “We use CFD techniques to address some of the same issues that are dealt with in traditional wind tunnel testing, and have replicated real wind tunnel situations to verify that our results correlate well,” says chief engineer Dr. Rob Lewis. “Because we make simplifications in the interests of speed we do not achieve 100% accuracy, but we can quickly demonstrate the implications of design change and thus assist in optimization. Formula 1 car manufacturers recognize that a 2 or 3% performance improvement can make all the difference in a racing situation.”

As Lewis explains, CFD gains over wind tunnel testing in more than just time saving: “We can perform simulations that would be impossible with physical testing. In a wind tunnel, for example, you are basically restricted to a single car. Using CFD we can simulate a racing situation, determining the change in downforce when one car is tailing another.”

Advantage CFD works closely with its racing car clients, optimizing everything from nose cones to fuel tanks. Here, the right internal geometry can save vital fractions of a second during pitstop refueling. Approaches to motorcycle manufacturers, however, received a mixed response. “We were surprised at the range of commitment to aerodynamic analysis among bike companies,” says Lewis. “Most recognized that it had some value, but few believed that it could win them races. We were determined to change their minds.”

As with all CAE techniques, accurate geometrical representation is a prerequisite. As Lewis explains, however, this was a potential stumbling block for the motorcycle project. “We were in a bit of a Catch-22 situation. No bike manufacturer would take us seriously unless we could demonstrate the power of CFD analysis on a real bike, but there was no chance of their giving up their precious CAD data for the purpose. To get the project off the ground, we needed to find a way of generating accurate representative motorcycle geometry as quickly as possible.”

As raw CAD geometry was unavailable, a reverse engineering route seemed to be the only option. The closest thing to a racing bike that Advantage CFD could lay its hands on — very close, as it happened — was the Yamaha R1 owned by Adrian Reynard. The boss’s bike was transported to the headquarters of 3D Scanners, Coventry, UK, for a physical-to-digital makeover using ModelMaker, the company’s non-contact laser scanner. These same scanners are used as reverse engineering and inspection tools in North America by leading companies such as Boeing, Ford, Disney, Chrysler, Harley Davidson, ARC, Penske, Richard Petty Enterprises, Evernham, Lotus Cars, Porsche, Nissan, Honda, JCI, Lear and John Deere.

Model-Maker offers significant advantages over alternative technologies when it comes to 3D data capture. Among these benefits is accuracy and flexibility — unlike some systems, no prior surface treatment is required and subjects can be scanned under normal lighting conditions. However, the overriding consideration in using ModelMaker for the bike scanning was its speed.

“Although I know from past experience how accurate ModelMaker is, the priority for this job was to get a pretty good representation as quickly as possible,” explains Lewis. “There would have been no point at all in scanning a bike on its own — the rider is very much part of the racing geometry. This meant that someone had to be sitting on the bike in a convincing racing position for the duration of the scan. Realistically, this meant a number of discrete scanning sessions to avoid the ‘rider’ getting cramped, so without the speed of ModelMaker we could have been in for a long haul.”

Scanning of bike and rider turned out to be well under a day’s work, and a workable CAD model was stitched together without trouble. This model was surface-meshed in readiness for a series of CFD analyses. The geometry was used as a basis for analysis in as close to a typical racing situation as possible. Because the bike was scanned in a static position, the extension of the front suspension was unrealistic so this was tweaked before meshing. The wheels were rendered with a typical speed of rotation, and the angle of attack was based on rounding a corner at pace. The result — including the spectacular trailing billows of dirty air — was exactly the kind of thing that Lewis needed to demonstrate the power of CFD in bike racing.

“I can see why motor cycle manufacturers might approach CFD with trepidation,” he says. “The flexibility of suspension and rider position means that geometry is changing constantly throughout a race, and a complete analysis would be impossible without CAD geometry and an in-depth understanding of rider dynamics. Even with the limited number of configurations we were able to estimate from our single scanning session, however, enough useful information could be gleaned to enable us to offer sensible advice.

“By varying the bike angle slightly and measuring the resultant forces, it is possible to get a quantifiable measure of stability,” he explains. “And in bike racing, stability — and therefore rider confidence — is probably the most important factor. A slower bike with a confident rider who is prepared to race aggressively can be a far more potent combination than a more powerful machine and a rider with the wobbles.”

Lewis is confident that in time CFD will become as important in two-wheeled motor sport as it is in four. “We have already been able to recommend design changes to bike companies, and it’s conceivable that the work we have been doing may even bring radical changes to the way that motorcycles are designed in the future,” he says. “Without the 3D Scanners, though, we would have struggled to crack the chicken-and-egg situation of having no geometric data.”

—RM