We get a lot of questions about how we measure the aerodynamic effectiveness of a bike and what the importance of actual wind tunnels is in an age when software can mimic some conditions we see in the wind tunnel. It's definitely interesting and we've spent the past 15 or so years developing the protocol we use in the wind tunnel and investigating CFD and other ways to dream, develop and test bicycle aerodynamics.
What is aerodynamics?
Aerodynamics is the study of forces and the resulting motion of objects through the air. In our case, the object refers to the system comprised by the bicycle and accessories plus the rider. For a detailed discussion, take a look at our Thinking and Process page on Aerodynamics.
An everyday example of the forces involved is the classic scenario when as kids, we put our hand out the car window. We could feel the force trying to drag our arm backwards and the variation in its magnitude as we changed the orientation and shape of our palm. If you do this at 50 km/hr and imagine your whole body immersed in the flow, you would feel what an elite rider experiences in a time trial stage. It is clear that the least amount of resistance there is to the motion, the faster the rider will be. Jim Martin wrote the defining paper on the topic and we've referred to it often over the past few years.
What are the tools of the trade for Cervélo?
Here at Cervélo we discovered very early on that an aerodynamically efficient bike yields good results (viz. winning.) For this reason, we have invested heavily to create a very strong in-house aero department comprised of a full-time aerodynamicist and a consultant with many years of experience in the field.
In addition to having the right personnel, part of the investment has been getting the very best tools in the market for aerodynamic research and development. The three basic tools are actual road testing, computational fluid dynamics (CFD,) and wind tunnel testing. Each method has its own advantages and disadvantages.
Common sense would lead us to believe that road testing would yield the most realistic results, since it is tested with real products, real riders, and under real conditions. However, measuring the drag on the road is very difficult given all the random variations such as rider position, system vibration, varying ambient conditions, varying road conditions, etc. In addition, during the developmental stages there is no actual bike to test - which makes road testing rather difficult when we're designing a new bike.
Consequently, road testing is limited to improving the individual rider performance once the bike has been designed. We are fortunate to have direct access to the professional athletes from the Garmin-Cervélo team and they provide constant feedback to VWD (Vroomen-White Design). That helps us speed up the development time.
The second tool used by the aerodynamics department at VWD is a relatively recent phenomenon introduced to the design process of aerodynamic bicycles - and we at Cervélo are pioneers in using this technique. It is a branch of science known as computational fluid dynamics (CFD) which uses computer power to provide simulations of fluid flow.
The primary benefit of CFD is the ability to improve and modify a bicycle's shape before it is built with the parallel use of CAD (computer-aided design) and CFD. It is very easy to test several ideas, conservative or innovative, in a virtual environment to determine which one of them justifies rapid prototyping for further investigation in the wind tunnel. The main disadvantage of CFD is that a full model simulation with a complete bike plus the rider is computationally expensive. For this reason, we have devised a methodology to utilize these two tools to their full potential, yielding better bikes in a reduced turnaround time.
The Wind Tunnel
The third tool used at Cervélo is the venerable wind tunnel. Wind tunnels allow test conditions to be highly controlled and be independent of the external atmospheric conditions since the test model stays stationary while the air moves relative to it.
What wind tunnels are used by Cervélo?
The main tunnels used by Cervélo are the San Diego Air & Space Technology Center Low Speed Wind Tunnel (LWST) and Kirsten Wind Tunnel at the University of Washington Aeronautical Laboratory (UWAL) which are closed circuit and permit full scale testing. These are low speed tunnels perfect for our application. (Low speed wind tunnels are used for operations at very low mach number, with speeds in the test section up to 400 km/h, which is about 250 mph, or mach 0.3.) In fact, these two tunnels are regularly used by other high end sports applications such as skiing, bobsledding or golfing as well as for ball designs used for soccer or volleyball.
In general, air is blown by one or two fans and the highest speed is reached at the smallest cross section, commonly called the test section. Ahead of the test section there is an inlet contraction that directs the flow smoothly into the test section, with the objective of obtaining a uniform velocity and flow characteristics in the test section. The ratio between the inlet area and the test section is called the inlet contraction ratio. Larger contractions ratios usually result in more uniform free stream conditions in the test section. The diverging section behind the test section (diffuser) reduces the speed ahead of the fan.
Similarly to car aerodynamics, there is a need to simulate the rotation of the wheels in order to mimic true flow field conditions. For this reason both LSWT and UWAL provide means to rotate both wheels. Another feature of both tunnels is a raised test plate or fixture, which is utilized to minimize the effect of the ground boundary layer, making the flow more realistic.
In addition to testing the bike on its own, we have a second way of testing the aerodynamic performance of a bike - the bike/rider system. The problem with involving real riders is that it is very hard to maintain the same position for the duration of a test. In fact, we can argue that it is practically impossible since for us a normal wind tunnel trip lasts a full two to four days. We have circumvented this problem by scanning Dave Zabriskie and manufacturing a full scale copy of him in high density rigid foam, which you see above--we call it simply DZ. With DZ we have the ability to test bike/rider models from day to day with a high level of accuracy and repeatability.
To conclude, we can say that the wind tunnel is one of the best tools out there to accurately simulate the correct flow conditions on a bike/rider system such as separation points, flow patterns, vortices and ultimately drag.
Why is the wind tunnel important?
At Cervélo we use both CFD and full-scale wind tunnel testing in a complementary fashion to develop Cervélo bicycles. Initially, we use CFD to explore designs before physical prototypes exist, or aspects that cannot be easily investigated in the wind tunnel, or to understand flow characteristics that are hard to visualize in the tunnel. Secondly, we use the wind tunnel to fine-tune the best ideas.
Once we make it to the tunnel, the key design parameters have been well studied and many possible configurations perfectly defined for each. To be able to compare the performance of the different potential configurations for each of these parameter, we use a special bike we made called a mule (a steel skeleton frame) where we attach rapid prototype plastic skins like a three-dimensional puzzle that forms the bike frame's complete surface. We then run about 25 cases per day at about 12 yaw angles each. This constitutes a tremendous amount of information that is later analyzed to determine the best possible configuration for each of the design parameters. Such test would be extremely expensive and time consuming in CFD.
In the future, we will further optimise all three primary tools into one integrated process to obtain an ever better bike in an ever shorter amount of time. Computing power is always increasing rapidly, and it is not difficult to imagine a day when the circuit between CFD and wind tunnel testing is virtually simultaneous. However, the final step will always be crucial - getting the world's best riders on the bikes to see how they ride on the road.