Wind Tunnel Showdown: 11 Super Cars Face Off (Part 1)
This time we brought eleven road bikes—including models like the S-Works Tarmac SL8, the new Trek Madone, Cervélo S5, Giant Propel, Pinarello Dogma F, Van Rysel RCR Pro, Canyon Aeroad CFR, and more—into the wind tunnel to identify which one is really the fastest.
Owning a top-notch bike is most certainly one of the key pleasures of cycling for many aficionados. Many people dream about and deliberately pursue despite the huge cost tags. Perhaps one day we will be able to purchase these motorcycles even though right now we cannot. And even if we never do, the innovative technology in these bikes will progressively find its way into more reasonably priced ones, somewhat meeting our wants.
Being lightweight was road bikes' main selling feature for a long period. Manufacturers started truly investing in "aerodynamics" in their designs in the early 2010s, and shortly afterward, aero bikes emerged. Most brands at that time offered two product lines: a lightweight climbing model and an aero model for sprinting on level roads; some even included an endurance model for classics and cobblestone portions. Thanks to developments in materials and technology, today's aero bikes are smaller, and climbing bikes are likewise lighter. Certain companies have even begun combining the two looks. Though the truth is that choosing the correct road bike is still very difficult, you would think this would make buying a bike easier.
When new bikes are launched nowadays, it's almost impossible not to see claims about aerodynamic optimization, like "10W faster at 40 km/h" or "27 seconds faster over 40 km." These numbers might be somewhat helpful, but they usually don't give us a clear picture of overall performance. Brands tend to compare their latest models to their older ones and almost never compare them to their competitors, probably due to legal reasons that make it inconvenient to disclose. That's exactly why we do these tests.
We brought the 11 best road bikes available today into the wind tunnel for a back-to-back comparison to see which one is truly the fastest, which one is average, and how much faster they are compared to a benchmark model.
Model Selection
Our target is high-end road bikes, trying to strike a balance between models that consumers might want to buy and those seen in the WorldTour peloton today. We want each brand's model to be on a similar level. While the aerodynamic differences due to different drivetrain groups are generally negligible, we sometimes find that higher-end models with Dura-Ace or SRAM Red come with integrated handlebars, while the cheaper versions have separate handlebars, which introduces slight differences.
In our test, some bikes are "all-around aero" models like the Cervélo S5 and Scott Foil, while others lean more towards being "aero all-around" models, such as the Specialized Tarmac and the new Trek Madone. Cannondale has both types, and we chose the model more commonly seen in the WorldTour.
We selected the following models in alphabetical order:
Cannondale SuperSix Evo 4 Hi-Mod Team Edition
Canyon Aeroad CFR
Cervélo S5
Factor OSTRO VAM
Giant Propel Advanced SL0
Look 795 Blade RS
Pinarello Dogma F
Scott Foil RC Pro
Specialized S-Works Tarmac SL8
Trek Madone SLR 7 Gen 8
Van Rysel RCR Pro Team Replica
To compare data, we also chose a benchmark model: the 2015 Trek Emonda ALR, equipped with non-aero wheels, rim brakes, external cable routing, and round drop bars.
All the bikes are in a size of 56cm or the closest equivalent for each brand.
Our model selection was somewhat limited. For example, we really wanted to test the Colnago V4Rs to compare it with the bikes Pogacar and Vingegaard ride, as well as the Merida Scultura, BMC Teammachine R, Bianchi Oltre, Enve Melee, the new Van Rysel, and more. But either we couldn't find these bikes, or we reached out to the brands and didn't get a response.
Test
We took these bikes to the wind tunnel at the Silverstone Sports Engineering Hub to test their aerodynamic performance, quantified in square meters (㎡) and measured using drag coefficient times area (CdA). The drag coefficient essentially indicates how difficult it is for air to pass over the surface of an object. This mainly depends on the shape of the object, but the surface material also has some effect. Area is straightforward—it’s simply the frontal projected area of the object.
We conducted separate tests on just the bikes, as well as on the rider and bike system, to explore three questions:
- Which bike is the fastest?
- If the results show that the aerodynamic differences between all modern road bikes are minimal, can you ignore aerodynamics when buying your next bike and instead focus on other factors like weight, comfort, specs, and after-sales experience?
- How much faster are they compared to our benchmark model? How much advantage do you gain by upgrading to a modern racing bike?
Test Plan
The test plan was designed by myself, with guidance from the aerodynamics experts at the Silverstone lab.
We conducted tests at seven different yaw angles: -15, -10, -5, 0, +5, +10, and +15 degrees. Simply put, yaw angle describes the direction from which the wind is hitting the rider and bike system. The faster you ride, the closer the average yaw angle is to 0 degrees, but in the lab, we can measure various data, including crosswinds.
All the tests were conducted at a speed of 40 km/h, which represents a typical amateur road race, a fast group ride, or a slower professional race (for example, the 2024 Tour de France had an average speed of 41.4 km/h).
We initially planned to test at a speed of 30 km/h as well, to reflect the average speed of regular riders. However, we found that a single day was not enough to complete all the tests. If we left the equipment for the next day, the positioning would change, making the data unreliable, so we ended up dropping the 30 km/h test.
Each yaw angle was tested for 30 seconds. We could have tested for longer, but we found that 30 seconds was a good balance—long enough to get sufficient data, but also short enough to keep the overall test duration manageable, preventing the rider from getting too fatigued and allowing them to maintain a stable cadence of around 90 RPM. For static wheel tests, we only tested for 15 seconds because airflow around the wheels is very different when stationary and doesn't represent real-world conditions, so 15 seconds was sufficient to capture the necessary data.
Standardization
The wind tunnel itself was designed with temperature and air density variations in mind. Before each test, an offset calibration is performed on the wind tunnel to ensure accurate readings.
We standardized as many variables as possible, such as bike size, rider position, tires, computer mounts, bottles, and bottle cages. I also projected my position outline onto the floor, so I could make sure my riding posture stayed consistent each time. All the bikes used in the test are size 56cm or the closest size according to the brand’s geometry chart, and each bike was adjusted to fit my body as consistently as possible. Our goal was to simulate real-world riding conditions as closely as possible, but we made some compromises when necessary to ensure consistency in the tests.
I removed all the computer mounts, even though I know 95% of riders use them, because only half of the test bikes had stock computer mounts while the others didn’t, so we decided to not use them at all.
The Madone, Propel, and SuperSix are all designed with aero bottle cages, which we included in the tests. The rest of the bikes were equipped with Elite Vico Carbon bottle cages and Elite Fly bottles, which are common gear in the WorldTour. The idea was that if you buy these bikes and they come with aero bottle cages, you’re likely to use them, so they should be treated as part of the bike. Factor OSTRO VAM also has aero bottle cages available, but unfortunately, we weren’t able to get them in time for the test.
I wore my helmet all day during the tests to make sure it stayed consistently fitted. I also marked my jersey position so that after bathroom breaks, I could put it back on the same way, keeping everything consistent.
Confidence
The above data was obtained by testing the benchmark Trek Emonda ALR with the same setup both before and after the test began. There is some debate on the best way to calculate errors. Some people take the average CdA value and calculate the difference from other values, which resulted in errors of 0.24% and 0.53%, respectively. However, we chose to compare each yaw angle separately and take the maximum difference, which provides a fairer and more practical conclusion.
Additionally, the following factors could affect the results:
The handlebar widths installed on the bikes range from 38cm to 42cm, meaning the rider's hand position varies slightly. We used the lab-provided "edge" markers to fine-tune the grip position to keep the riding posture as consistent as possible.
The different handlebar widths also have an effect during static tests. This effect is small, but not negligible.
Each bike was fixed with an additional support post, but we did not account for the drag of this post since we were interested in comparing the differences between bikes rather than their absolute values.
We did not make any adjustments to consider these factors, as they were consistent for all the bikes, and we were focused on the differences rather than the absolute values. The Pinarello's rear chainstay on the drive side has no opening, so we added extra support there, which may have influenced the data, but we chose to ignore it. Canyon and Cannondale have similar designs, but the test bikes they provided had the standard openings.
Additional Details and Disclaimer
The data from our tests is not the final verdict on bike performance but rather an independent, unbiased reference. The results only represent our test on that particular day, and the data is provided for reference to give an overview of overall performance.
The chart below shows the test data. For simplicity, we reported the test results using four significant figures obtained from the wind tunnel. Based on this data, we also did some calculations to determine the speed at different power outputs. These calculations did not take into account drivetrain friction, rolling resistance, or other losses, nor did they consider the impact of yaw angle at a given speed; they are simply meant to help readers understand the potential impact of CdA on speed.
We also listed the weight of each bike, which includes the actual weight before the test, with pedals and bottle cages. Additionally, we added some other factors to consider when buying a road bike, such as ride feel, weight, comfort, components, tire clearance, aesthetics, and after-sales service. The best road bike for you will balance these features with aerodynamic gains, with each feature weighted according to your personal preferences.
Results
Rider System Test Results
The chart above shows the relationship between CdA and yaw angle for each bike. The two lines higher than all the others are the results from the two tests on the benchmark bike. All the new bikes have a lower drag coefficient compared to the aluminum bike from 10 years ago—just as expected!
Except for the two benchmark lines, the data for the other models is quite close, with the lines overlapping at various points. This suggests that some bikes are more aerodynamic at certain yaw angles, while others are slower at those angles.
When adding in the margin of error, the results overlap significantly, and we cannot determine a clear winner.
Using only the average CdA, we cannot confidently rank the top 10 bikes. Aside from knowing that the benchmark bike is clearly slower, the only conclusion we can make is that the Look 795 Blade RS is slower than the Specialized, Trek, Factor, Cervélo, and Canyon.
This chart shows the power required for each bike at 40 km/h. The Specialized Tarmac requires 280.2W (+/- an error of 3.91W).
The Look 795 Blade RS requires 286.29W, also with an error of 3.91W. Considering this margin of error, the difference between the two bikes might be negligible, or it could be as much as 13.91W.
The benchmark Trek Emonda ALR requires 304.67W. Including the margin of error, the S-Works Tarmac is at least 16.65W faster than the benchmark, and it could be up to 32.29W faster.
Static Bike-Only Test Results
Compared to the rider-bike system above, most of the bike-only data clusters together, with the two benchmark tests clearly positioned higher. This means that, purely from the bike's perspective, aerodynamic improvements are significant.
Excluding the two benchmarks, we see greater consistency among the remaining lines. This means that the bike that is faster at a given yaw angle is generally also faster at other yaw angles. In some cases, the lines cross, but the data looks clearer overall.
Considering the margin of error, although there are still some overlaps, we can confidently say that some bikes are indeed faster than others, with the Look being noticeably slower. We still can't pick an absolute winner, but we can say that the Cervelo S5 is very fast at zero yaw, though it adds more drag in crosswinds, whereas the Factor OSTRO is slower at zero yaw but performs well at other yaw angles.
We scanned the data for each bike across all yaw angles to get the average CdA, and after adding in the margin of error, we found that the Factor OSTRO VAM is the fastest, with an average CdA of 0.0882 (+/- 0.0010).
The Scott Foil and Cervélo S5 are very close (0.0897 and 0.0900, respectively, +/- 0.0010), which means they could also take the top spot given the margin of error.
The Canyon Aeroad and Cannondale SuperSix are right behind, with a difference of less than 0.0004㎡ between them, so they both could be considered second place.
The Pinarello Dogma F and S-Works Tarmac are slightly behind, but taking into account the margin of error, they could also make it to the podium.
The Van Rysel RCR Pro, Giant Propel, and Trek Madone are in the middle of the pack, while the Look Blade 795 RS is a bit further behind. All of these bikes are significantly faster than our benchmark model, which has a CdA of 0.1370.
This chart shows the power required at a wind speed of 40 km/h.
The Factor OSTRO VAM requires 72.59W (+/- 0.85W).
The slowest, the Look Blade 795 RS, requires 84.95W, also with an error of 0.85W.
The best result for the benchmark bike is 112.74W. Considering the margin of error, switching to a top-tier bike can save between 38.45W and 41.85W.
Keep in mind that this data only reflects the bike itself. The bike is just a small part of the overall CdA of the entire riding system, so the impact will be somewhat reduced.
Conclusion
Which Bike Is the Fastest?
Of course! Here is the English translation:
It’s really hard to draw any meaningful or confident conclusion about which bike is the fastest. Additionally, the rankings change significantly at different yaw angles, and although our data shows some differences, considering other factors like weight, we can't confidently say any one bike is absolutely the fastest.
From the data, the results are quite clear: aero bikes like the Factor, Scott, and Cervélo rank at the top, but the more "all-around" models still held their own, only falling behind by about 2-5W. Considering the weighted average CdA across the seven yaw angles, the Factor OSTRO is the fastest in this test. With the margin of error, it could rank third, and considering it was also the second-lightest bike on the day at only 7.23kg, it’s a very impressive result.
Looking at the data, you might think the Cervélo S5 is the fastest bike in the real world, but its performance is less effective at certain yaw angles. It performs very well at 40 km/h, but at lower or higher speeds, the impact of yaw angles will become more significant.
The Cannondale SuperSix Evo, being an "all-rounder," also performed well, coming in fifth. But with our CdA margin of error of 0.0010, it could rank as high as second.
The only bike that lagged significantly was the Look Blade 795 RS, which was about 12W behind.
Our benchmark bike, the Trek Emonda, was around 28W behind, or roughly 40W slower compared to the OSTRO VAM.
Returning to our previous conclusion, the aerodynamic data for modern bikes is so close that you can almost ignore the manufacturers' aerodynamic claims. We may not be able to confidently say which bike is the fastest, but we can certainly identify the slowest one. All the high-end bikes we tested are clearly a step above the benchmark.
That said, aerodynamics shouldn't be entirely ignored when buying a bike. The difference between the best and the worst bikes we tested could be as much as 13.90W, which could make a noticeable difference in certain situations.
How much power can you save compared to an older bike?
If you consider the 2015 Trek Emonda as your old bike, then compared to the fastest bike in our test, you could save between 16.65 and 32.29W of power. At a more achievable solo ride speed of 30 km/h, you could save between 7.03 and 13.62W. If you push the speed to 50 km/h, the power savings would increase to between 32.54 and 63.05W, but realistically, it's hard for an individual to maintain that speed for long, and if you're in a group ride, the aerodynamic demands are different.
Of course, there are also cheaper ways to save this much power, like getting an aero helmet, a good set of cycling clothes, or a new set of wheels. If you need a reason to spend money, feel free to take this one.
To be continued.
Wind Tunnel Showdown: Head-to-Head Comparison of 11 Super Bikes (Part 2)
Recommended Reading:
- Beginner's Guide: Detailed Explanation of Front Derailleur Types, Installation, and Compatibility
- Beginner's Guide: Everything You Need to Know About Cassettes
- A Comprehensive Guide to Rear Derailleurs
- Super Boost Plus 157 Rear Axle Standard: Is It Here to Stay?
- Beginner's Guide: How to Find the Optimal Tire Pressure for a Road Bike