How to engineer the ultimate downhill mountain bikeComments Off on Fearless design
When Drew Pautler leads his team of professional downhill mountain bike racers this summer at the Canada-hosted International World Cup, he’ll be pushing his abilities and his equipment to their limits. Hurtling down the slopes of Mont-Sainte-Anne at speeds up to 80 kph requires quick reflexes to pick a line through the winding, densely forested track and a uniquely designed bike engineered to withstand the constant impacts of the rocky terrain.
“The closest sports I can compare it to would probably be downhill skiing or Formula One racing in the sense that it has the most variables,” says Pautler, co-manager and member of professional downhill mountain bike racing team Primary-Devinci. “You are dealing with different terrain conditions, weather conditions and grueling courses. It’s a scenario where you have to have the most dedicated equipment. That includes suspension, tires, brakes and of course frame design.”
It’s that last component, the frame, that can literally make or break a racer, he says. Not only does it have to be as light as possible, but since downhillers place their safety in the care of bike manufacturers, it also has to be exceptionally strong.
That’s why, in addition to their skill, the team will also be packing a secret weapon called the Wilson 4, a professional downhill mountain bike designed and manufactured by Cycles Devinci. The company manufactures more than 80 award-winning mountain, road and hybrid models in its Chicoutimi, Que.-based manufacturing facility, but the Wilson line, which made its debut at the Montreal bike show last year, is the latest to garner industry acclaim.
At only 8.8 pounds, including shocks, it is the lightest frame in its class but, says Bruno Gauthier, director of research and development for Cycles Devinci, it doesn’t sacrifice any of the strength or rigidity downhill racers count on to get them down the course in first place and in one piece.
“Our customers want to be able to do crazy things,” he says. “So when we do our tests, we have to put our standards even higher than what our customers are going to do with the bikes to make sure that these crazy guys won’t break them or get hurt.”
Twist of Fate
While Devinci’s 2008 Wilson model took two years to perfect, the final design is actually the culmination of nearly a decade of industry-leading product design. In 1999, the then head of Devinci’s R&D department, Érick Auger, realized that he needed to adopt a finite element analysis (FEA) approach to the company’s product development if it wanted to keep pace with its much larger competitors. Fabricating and testing multiple physical prototypes to hit upon an improved design by trial and error would simply be too expensive and time consuming.
Instead, he turned to Bruno Gauthier who, at the time, was working for Fab Concept, a Chicoutimi-based engineering consulting firm that specialized in dynamic and FEA analysis. The problem was, they no hard numbers to go on. No published data on the structural loads mountain bikes sustain was available and, if the numbers did exist, they were a tightly held industry secret that Devinci’s competitors (e.g. Giant, Certified, Cannondale) weren’t about to share with the relatively small Canadian manufacturer.
“So we had to make an educated guess,” Gauthier says. “But the results we got back were crap. We could use it to show where the weak points were, but it wouldn’t allow us to optimize the bike’s design.”
Given the difficulties of their initial attempt, the pair decided they would need to measure the forces a mountain bike encounters as it was being ridden. For help, Auger approached the Aluminium Research and Development Centre of Quebec (CQRDA), an organization that fosters partnerships between universities and SMEs in the aluminium production and processing industries.
Surprisingly, CQRDA had a possible solution for the pair almost immediately. Dr. Yvan Champoux, a professor of mechanical engineering at Sherbrooke University, had been playing with the same idea himself. A specialist in the development of force transducers, he leads Groupe d’Acoustique et Vibrations de l’Université de Sherbrooke (GAUS), the largest vibra-accoustics research group in Canada. Inspired by a lecture on the vibra-accoustics of golf clubs, the professor envisioned applying his specialty to his own passion, cycling.
“I started to do some activities in that field when I was contacted by Devinci,” he recalls. “I didn’t know them at that time, but we clicked together and started to work on the first instrumented mountain bike.”
The Instrumented Bike
The “instrumented bike,” as the company refers to it, was an industry first, and is equipped with 36 sensors (a combination of force transducers and strain gauges) positioned at key locations on the bike’s frame. Data cables from the various sensors feed into a backpack, which holds the power supply and an industrial data logger. Sampling at 1,000 times per second, the data logger records every jolt, vibration and peak load a downhiller would typically encounter during a race.
“Initially, the idea was to put only strain gauges on the frame and then use the results in the FEA model,” Champoux explains. “We realized, however, that this is an inverse problem approach where you are using the output to eventually calculate the input or loads. That’s great if you have a very simple system and there is only one force and one acceleration. But as soon as you have more than one or two inputs, you have to inverse a matrix and then, well, it gets complicated.”
Champoux says an added complication was that all the loads have to be measured in both the horizontal and vertical since the forces vary in magnitude and direction from moment to moment. Working with the bike designers at Devinci, his research group solved these problems by using a combination of commercial and custom-built force transducers.
“There is nothing on the market to measure forces on the seat stem of a bike, for instance, so we had to build them from scratch,” he says. “And we didn’t have a clue as what kind of forces we would obtain, especially because they are dynamic forces. Shocks can be very surprising. You can have very high values, but they only last a few milliseconds.”
According to Gauthier, the initial data recovered from the instrumented bike came as a shock; the reality was much worse than their initial educated guess had tried to approximate.
“You could lift a small car with the forces that we found at the rear axle of a mountain bike frame,” says Gauthier, who refuses to reveal the actual numbers. “Landing flat on a hard surface like concrete from five or more feet is the worst case scenario for us. A dual suspension system, with the front and rear shocks, will absorb most of that impact, but you are also going to force the shocks to the end of their travel. Whatever force is left over will be absorbed by the rider and the frame.”
Armed with their experimental data, the designers began working out how they would turn a mass of raw sensor data into something useful. Gauthier, who by 2001 was a full-time Devinci employee, says the R&D team used a statistical analysis algorithm to get an “average” of all the loads over time.
The processed data was then used to build virtual and physical test benches to replicate the six most critical load case scenarios a mountain bike encounters, including front and rear breaking, the maximum force at the front and rear hubs, the force at the saddle and the torsional forces imparted while pedaling. To model and analyze these load cases, the company bought seats of CosmosMotion and CosmosWorks from Montreal-based VAR SolidXperts.
Gauthier says that although CosmosWorks isn’t the most complex FEA software on the market, the seamless integration between the three SolidWorks applications makes them best solution for Devinci.
“A dual suspension bike is a complicated structure with a lot of moving parts,” he explains. “To analyze each of these parts, you must input the force data collected from the hub of the instrumented bike into the virtual bicycle modeled in SolidWorks and imported into CosmosMotion.
“After that, the software will give you the internal forces that go into the frame,” he adds. “The software then imports that data directly into CosmosWorks where you can analyze the frame overall or its parts in isolation, such as the linkage isolated from the front triangle and the front triangle from the fork. It’s all automatic.”
The new process, or “Devinci Intelligence” as the company markets it, changed their design approach in fundamental ways, Gauthier says. For instance, frame weight could be reduced by pinpointing locations at which to vary the wall thickness of the 60/61 T6 aluminium piping.
“Our designs also began to be more organic—more flowing with fewer hard edges and more rounded shapes,” he says. “You can have edge in your design, but it needs to be in places where it won’t negatively effect the structural integrity of the frame.”
“At the same time, a perfectly round tube isn’t always the best solution either. It all depends on where you put the tube on the bike. Sometimes it needs to be round, sometimes oval and sometimes square with rounded edges. The goal of our FEA analysis is to find the correct tube shape for each section of the frame.”
The changes made to the Wilson’s 2008 version are too numerous to list, but one striking example is the bike’s down pipe—the hypotenuse leg of the frame’s front triangle. Instead of a perfectly round tube found on many bikes, the Wilson’s down pipe is flared through a process called hydroforming, in which water pressurizes the inside of the aluminium tube until it deforms to the shape of an exterior mould.
“If we didn’t use hydroforming, we wouldn’t be able to place the material exactly where we want to,” Gauthier explains. “By using hydroforming, we were able to eliminate a reinforcing gusset present on the older version of the Wilson bike, thereby saving the weight of the material while maintaining the same frame strength.”
In addition, the team redesigned the suspension system to increase travel to nine inches, added a bridge to the frame’s rear triangle to increase lateral stiffness and bent the seat and top pipes inward to provide greater stand-over height and lower the bike’s centre of gravity.
Gauthier says the team ran through approximately 40 iterations for each major component to arrive at five final design candidates. In the end, however, they fabricated only one physical Wilson prototype, which was then subjected to a battery of six physical test benches—machines that validate the FEA analysis and manufacturing process by inflicting four years worth of wear in only 24 hours.
Based on the success with its mountain bikes, the company is in the process of applying the same design approach to its other products lines. In further collaboration with Sherbrooke and VélUS, a group dedicated to bike research established by professor Champoux, the team has created an instrumented road bike.
Yet for all Devinci’s technological sophistication, Gauthier says the key ingredient of the company’s design recipe isn’t the instrumented bike, the software or the extensive testing process. “In the beginning, it is the always the instinct of the designer,” he says. “After that, our Devinci Intelligence process confirms if his instinct was correct or not, but the most important engineering tool we use is ‘whatever makes the most sense’—the sum total of our accumulated engineering knowledge and experience.”