Do Aerodynamics Matter?

Okay so the title is a bit sensationalist but I wanted it to balance my last post which was titled "Does Weight Matter?"  I don't think there are many people left who would argue that, as far as aerodynamics go, the more the better.  It is, however, still subject to the "all else being equal" deal.  Just because your bike may be set up in an aerodynaically sound way doesn't mean that you are aero on it - think "slam that stem" with a guy barely able to put his fingertips on his bar tops.  

In any case, the TTs at the Tour always provide a good chance to think and talk about aerodynamics and aerodynamic equipment.  Wheels are the obvious choice to focus on since they are visible and they are generally the only components that is commonly measured by themselves.  Front wheels are more constructive to look at since rear wheels are so often rebranded in ways that cloud their provenenance, and the aerodynamics of the rear wheel are accepted as playing a very second fiddle to those of the front.

The following chart (hopefully it follows - I'm having trouble formatting it on this page) shows the drag values of a bunch of wheels across various angles of incidence. Please note that this chart is very imperfect, since we're comparing data produced by different entitities with different tires in different tunnels. Nonetheless, it's the best we can do for this purpose. The wheel is identified, and the source of the data is in parentheses - so there are two series for 404s, one produced by Zipp and one that uses Bontrager's data.  The data sets all appear on the source websites.  10 grams of drag is closely equivalent to 1 watt, so if the difference between 2 wheels at a given angle is 40 grams, then the one wheel will take approximately 4 watts more to go the same speed as the other. 

Click to make huge.

As we now know, Wiggins and Froome handed out pretty substantive ass whoopings.  They both did so using the old standby HED H3 wheel, of which there are two models - the H3 and the newer H3D.  I don't know which they used, but it's somewhat immaterial.  As you can see in the chart above, the H3 models do well at low wind angles - going straight into the wind, they are great.  As the wind angle widens, they do less well.  Notably, they are crazy consistent across wind angle bands - no matter where the wind's coming from, they are what they are.  But at a lot of wind angles, the H3s apparently leave a lot of grams of drag on the table.  So why do the Sky guys use them, and how are they able to use them so well?

For one thing, Wiggins and Froome HAUL ASS.  When you go 30mph on a nearly windless day, your apparent wind angle is constantly "darn near zero."  So they were probably confident that on that day, they were using the wheels at their best. 

For another, very deep front wheels have interesting handling characteristics.  I haven't ridden a ton of deep fronts, but I've ridden enough to know that they generally have a gyroscopic effect that makes turning a bit of an adventure.  The TT course had a lot of twists and turns, and TTs are far from purely watt and aero contests.  Good TT riders display technique in corners and over rises and dips that help them conserve energy and maintain high speed. 

Yet another factor is how the wheels work with the bike.  It's all well and good to evaluate wheels in isolation but there's plenty of research now (ENVE has been a leading proponent of this) that says that the pairing of bike and wheel can be a more significant factor than the wheel itself.  Given Sky's tendency to sweat the small stuff, I'm guessing that they know that the HED H3 works just fine thanks with their Pinarello TT bikes. Anecdotally, Wiggins always seems to use this wheel, so there may be a factor of preference at play as well.

Cancellara was third in the TT, using a Bontrager Aeolus 7 front (he may have been using an Aeolus 9 - it's a bit hard to tell).  In any case, whether 9 or 7, Bontrager's data shows that the wheel Cancellara used is really good at the narrow angles that you often see when you're going fast.  Anecdotally, Cancellara seems to always use this kind of wheel (he used 808 fronts an awful lot when he was on Zipp sponsored teams), so again we see that there might be some personal preference. 

Tejay Van Garderen was fourth on what appeared to be an Easton EC90TT, which is a 56mm front wheel.  There's not a lot of aero data out there on the Easton but it has a decided "v" shape - more so than any other rim I know of, and it's narrow.  This would imply that, like the H3, it is relatively stronger at low angles of incidence and relatively less good at wider wind angles. 

Fifth was Sylvain Chavanel, on an 808 front, and between the Bontrager and the Zipp I would say you are probably splitting hairs.  The shape of their curves is quite similar, but the differences between the Zipp-produced data and the HED-produced data are interesting.  I'd love to tackle this topic now but since I am well over the word count which our data shows produces an alarming trail off in reader interest, I will instead tease for a next installment - "tune in next time for another thrilling episode, entitled what the hell do America's Cup sail trimmers have to do with how data is produced in a wind tunnel?"

Back to blog


I think I'd prefer consistency across all yaw angles rather than just being awesome in a small range…plus lets face it, how aero my wheels are is not what is keeping me from the podium :)


Wheel stabilty in crosswinds needs to be addressed as well. Two years ago I was riding on a HED H3 wheel at our local 20K TT. The early season winds were so bad that I couldn't control or feel safe riding in the aerobars and I'm a fairly large rider, 185 pounds. Needless to say the TT results were poor due to not riding much on the aerobars. Last year I sold the H3 and bought the Bontrager D3 5 and rode in early season crosswinds of 25 MPH and I felt completly in control and never left the aerobars except to brake. The same could be said with other blunt nose rims as well, Zipp Firecrest, ENVE SES, HED Stinger, etc. This year I started riding on the Bontrager D3 9. My TT times dropped considerably and I feel confident in all but the strongest crosswinds. The HED H3 has a very pointy V style edge which is not good with crosswinds. The newer blunt nose rims have much better stablity in high crosswinds.


Michael – Excellent questions. Yes, the claim is that some wheels have negative drag at some wind angles. The most popular protocol for testing is with wind speed set to 30mph. The most popular reasoning for this is that this is the slowest windspeed that gives useful data. THIS MATTERS A WHOLE LOT. In Monday's TT, over 41k, three riders (I'm pretty sure) averaged 30mph or better. So 30mph is a speed that people can go, but it's not one that very many people can go for very long. On your next ride, see if you can hold 30mph for one minute (granted it would be easier if you had a crazy aero setup). WHY DOES WIND SPEED MATTER SO MUCH? Wind FORCE increases with the SQUARE of the increase in wind SPEED. A 16 mph wind is four times as strong as an 8 mile an hour wind. A 32 mph wind is 16 times as strong as that 8 mph wind. There is an erroneous thought out there that aero gear is more important for slower riders since they spend more time on course, battling wind. The truth is that a person who goes half the speed of another will get 1/4 the aero benefits that someone going twice as fast will get, minute to minute – if you half the wind speed, you quarter the wind strength. Since the slower person going half the speed will be on course twice as long, you double those quartered benefits since you reap them for twice as long. Net result? The rider going half as fast for twice the time gets half of the benefit that the faster rider gets. Aerodynamic considerations are 2.25 as important at 30mph as they are at 20mph. Also, Michael gets +5 for using "angle of incidence" while Mike, well, Mike I'm afraid you get -5 for using "yaw."


I'm not great at reading graphs, but am I to conclude that some wheels get negative drag at certain angles of incidence? Also, at what speeds are these drags calculated? Does it even matter?


Leave a comment

Please note, comments need to be approved before they are published.