With Kona just around the corner, we thought it would be interesting to share some wind tunnel tidbits. Since Cervélo engineers have spent so much time in the wind tunnel, we’ve checked just about every combination of stuff we can think of. Of course we keep thinking of more stuff, so we keep going back to the tunnel! Since we are committed first and foremost to helping athletes go faster, here are some ideas to help you find free speed.
The above image shows some of the bikes we tested at one of our wind tunnel trips in 2007. Since then we’ve made many more trips and tested many more bikes.
Why should you listen to us? Because we’ve spent a lot of time in the wind tunnel, and what we’ve learned about aerodynamics could fill a library. One of the things we’ve learned is that you have to include the rider to truly capture the aero interactions between legs and bike, body and bottle. Over the years we’ve also learned the value of using "Foam Dave"—an exact replica of Dave Z in his aero position machined from foam. This provides more repeatable data—especially when moving between different bikes.
There are many places to carry water on a triathlon bike. We are interested in the impact each location has on performance. Cervélo has visited this line of inquiry a number of times over the years, as it pertains to bike design as much as it pertains to athlete needs. We tested the now ubiquitous bottle mounted vertically in the aero-bar, and we also looked at mounting a traditional round bottle horizontally between the arms. We looked at various behind-the-saddle mounting locations, and of course we measured drag with bottles mounted on the frame, varying location and shape. Basically: bottles in front of the rider, bottles behind the rider, and bottles on the frame.
Bottles in front of the rider
We should start with the hydration in front of the rider. Having the most aero solution for bottle storage is no good if you don’t drink. As one athlete put it, “weaving down the road with advanced dehydration is not very aero either.” The big advantage of hydration systems mounted to the aerobars is that you always see it. It’s always reminding you to drink, and you drink more because it’s easily accessible.
We compared two options for aerobar mounted drinking systems—a system that hangs down vertically in front of the head tube (see figure 2) and one that mounts a standard round bottle horizontally on the extensions between the arms (see figure 3). What the testing indicated was that the vertical bottle added some drag depending on the system and shape of the head tube—but not as much as a standard bottle mounted on the frame. The really bad part was the straw sticking up.
However, a standard round bottle mounted horizontally between the rider’s arms was a revelation! It filled in the turbulent area behind the rider’s hands and actually reduced the drag. So this set up was not only faster than the vertical bottle, but faster than no bottle at all!
Figure 2 A vertically mounted aero drink system between the aero bars increased drag, but less than adding a bottle on the frame.
Figure 3 That’s right, adding the horizontal bottle between the rider’s arms reduced aerodynamic drag by a huge 56 grams on average. That’s a power savings of about 5.6 Watts when riding at 40 km/h.
Bottles behind the rider
Behind the saddle bottle mounts have some issues—it’s not so easy to elegantly mount and dismount the bike and you have to get a good strong cage and mount to avoid the bottles being ejected over bumps like vertical missile launchers. However, the number of people using these systems and the mini industry it has spawned testifies to their popularity. The limitations mentioned above and the fact that the bottles are difficult to grab—especially from an aero position—means that these systems are not generally the first choice for hydration, and bottles are probably transferred to one of the other positions for regular use.
There are a variety of tests on behind the saddle systems that show a wide range of results. Some have shown a slight drag penalty and some a significant decrease. Rather than test every product in every condition, let’s look at some general conclusions that can be applied to most cases.
We planned a series of runs in which we varied the bottle height and distance from the rider. To speed up expensive wind tunnel testing time, we fabricated a quick-adjust bracket to mount bottles behind the seat in various heights and extensions (see figure 4). It wouldn’t survive real riding, but did the job marvelously for our testing purposes.
Figure 4 Three positions for behind the seat bottle locations. Of course the metre stick was removed before the wind was turned on.
The data for this particular design showed a significant drag reduction compared to no bottles or cages behind the seat. Of course when we measured without a rider drag went up by an astounding 350 grams. This is an obvious demonstration of why it is really important to test with a rider on the bike. It might seem less obvious to some, but it’s still true that similar aerodynamic interactions occur everywhere between body and bike, albeit often on a smaller scale. That’s why we made Foam Dave—our stable and repeatable windtunnel dummy—but he had not been conceived at that time.
You do want to keep the bottles (and whatever else you have stored with them) as narrow as possible to allow the air to come back together as fast as possible after it leaves your body. There was not a huge variation in drag depending on the position of the bottles behind the saddle, high or low, near or far. Many people use these systems for storage of tires, tubes, inflators etc. This looks to be a good location as long as they don’t stick out and grab air.
>This data was mostly collected in 2007 with a live rider at 0 degrees of yaw, but if you keep the bottles close to the rider and not too wide the results extend out to higher yaw angles. At any rate, while test results can vary somewhat, the lesson is that bottles behind your saddle do not have a large penalty and, depending on the design, may have a significant advantage.
This series of tests is illustrated by CFD simulations we’ve run subsequently, one of which is shown in Figure 5, which show that the athlete’s body has a low pressure recirculation zone behind the seat that the bottles appear to fill nicely.
Figure 5 CFD (Computational Fluid Dynamics) simulation showing recirculating air in the low pressure wake zone behind the rider’s body, a good place to carry bottles. Interestingly, the streamlines you see here represent just some of the air. Of course you’re completely surrounded by air.
Bottles on the frame
Regardless of the options mentioned above, bikes still have bosses and many athletes opt to carry hydration in the traditional manner. Let’s face it: you might want to carry a bottle on the frame.
Fast forward to January 2011 and LSWT test 1101. Now we have Foam Dave, our “stable and easy to work with” wind tunnel mannequin. As mentioned above, we test a lot. We test our own bikes and prototypes, and we also benchmark our competitors’ bikes as well. But we go farther than that: in January 2011 we tested a batch of tri bikes three ways: with no bottle, aero bottle or round bottle, once in each position, seat or down tube. Because frame shape affects drag differently depending on bottle location, one position is sometimes faster than the other. The images below show bikes set up with the bottle in the most aerodynamic configuration for each. Whether or not you ride a Cervélo, find your bike and bottle (aero or round), look at the photo and install your cage where it’s shown there: it is the fastest location for your bike.
The Cervélo P4 with its integrated P4 bottle is the only case where the water bottle does not add drag and slow you down seen below in Figure 6.
Figure 6 Fastest configuration for every bike. Note that every bike has “no bottle” as the fastest configuration, except the P4, which is fastest with the P4 bottle. If you add a bottle to the other bikes you add drag.
If you use a round bottle on these bikes it is always fastest on the down tube but still adds typically 40-50 grams of drag. (See figure 7.) Note the Giant has no bottle bosses so technically you won’t be installing one unless you’re willing to improvise. We used wind tunnel tape.
Figure 7 Fastest configuration with a round bottle. From left to right: Cervélo P4, Scott Plasma III, Trek Speed Concept, S-Works Shiv, Giant Trinity Advanced.
Next we tested an aero water bottle. We used an Arundel aero bottle for all of these tests but we would not expect the best position to change with other similarly shaped bottles. When the P4 bottle was replaced with the Arundel the drag increased by just 17 grams1, about half the penalty it added to the other bikes. Some of the lessons from the shielded water bottle on the P4 were applied to the new Cervélo S5 which is built for bottles, and it is the only other bike where the aero water bottle has zero drag and effectively “disappears”2.
The aero water bottles added some drag to all of the other bikes but it was roughly half of that of a regular bottle. The best position for the aero bottle on each frame is shown below, and you will note that the best position differs depending on the frame. As a general guideline, it is always best to eliminate as many gaps around the bottle as possible. If you custom fit the Arundel on a P4 to eliminate the gaps the added drag drops to zero3.
Figure 8 Fastest location for an aero bottle. From left to right: Cervélo P4, Scott Plasma III, Trek Speed Concept, S-Works Shiv, Giant Trinity Advanced.
So what did we learn from all of these tests? The big lesson was that a bottle mounted horizontally between your arms (but it must be between your arms, not too high or low) makes you faster- faster than no bottle at all! That should probably be your first choice for regular drinking. The aerodynamics of this position combined with the fact that it is always in front of you reminding you to drink makes it a great choice. The Cervélo P4 integrated bottle or a behind the saddle system is the next most aero choice and both of those are excellent options for storage of spares and other equipment as well. If you have a P4, please come by the Cervélo booth at Kona for free P4 bottle. Cut it open for storage, or continue to drink from it, but either way keep taking advantage of this design!
One final nugget is that we found that a Bento Box behind the stem was a solution that had no drag penalty. However, this may depend on stem & spacer details and how many gels and bars you stuff into it.
We all need hydration and we all have preferences, so we hope that regardless of which set up you favour, this article will help you maximize your triathlon performance. Good luck and Mahalo.
1 Average drag of LSWT 1101 Run 217 vs. 219
2 Average drag of UWAL 2008 Run 31 vs. 35
3 Average drag of LSWT 1011 Run 509 vs. 513 & 514