Wind Tunnel Testing: Marketing Expense or R&D?

We get a lot of questions about whether we've wind tunnel tested our wheels, and how aerodynamic they are compared to brands like Zipp and HED and Bontrager. The answer is always No we haven't. When responding to these questions I usually choose just one of the several reasons. Now is a good time to go through all of them I guess.

With no pedals, bottle cages or rider, this bike is wind tunnel tested to determine drag in simulated shop floor conditions. First of all, it would be very easy to wind tunnel test our wheels and bikes. All it takes is about $500 an hour and a photographer. Wind Tunnel Testing is really only measuring the aerodynamic drag of something in the wind tunnel, not actually using that information to improve performance (more on that later). No small number of the claims cycling brands make about wind tunnel testing is exactly that - a trip to NC or San Diego orchestrated by the marketing department simply to be able to claim that the products have been tested in the wind tunnel, which is seen by some cyclists as a testament of legitimacy or genuine engineering prowess. The Neil Pryde to the right is a good example. The picture is courtesy of the brand (well, maybe not "courtesy" as I don't think this particular usage was their intention) from their Flickr gallery on Wind Tunnel Testing here, shot in the summer of 2010. The photos are gorgeous - bike pr0n of the highest order. Only they're not about product development at all, or even about aerodynamics. If the gallery is to be believed, the bike was tested without pedals, bottle cages or a rider - simulating shop floor or Interbike conditions far more closely than anything resembing an actual ride or race. The brand launched the bike a few months later, trading heavily on the wind-tunnel claim, and you'll still see "wind-tunnel tested" in most of their ads. When the Specialized Venge came out, I read a riotously absurd remark from one of the Specialized PR flacks that they even put two Venges in a row in a wind-tunnel, to test how well the bike drafts. Not two riders on Venges, but just a couple of bikes. You know, because the rider and pedaling legs have really no bearing on a bike's aerodynamics. "Let's just do this so we can say we did this and take some pictures," is not a directive that typically springs from R&D departments. 

For a few grand, we could do the same thing, claim our bikes and wheels are wind-tunnel tested and lure in customers who are looking for that stamp of R&D authority. Hell, if we want to differentiate (and what brand doesn't?) we could test ours on its side and claim that the Wheelhouse has been wind-tunnel tested to measure aerodynamic drag when sliding out in turn 3 (perhaps the only real world scenario where testing a riderless bike is actually fruitful). But we don't do this - in part because it's an expense that does not improve the product at all (you'll note that much of this "testing" is done after the product is finished and ready for market, not while it's still on the drawing board), but mostly because it's disingenuous, which is a fancy word for bullshit. We're not in business to bullshit our customers. Conceiving a test just to make some specious marketing claim doesn't feel like a sustainable business practice to us. 

The context we more commonly hear about wind tunnel testing is whether we have data comparing our wheels to everyone else's, so a customer can calculate how much speed they would be buying, presumably to factor in price and come out with some rigid $/watt value curve. The tacit basis for this question is that many of the brands that do their own wind tunnel testing not only claim that they have in fact been tested in the wind tunnel, but that the tests reveal that their wheels save some number of watts over some ride of a given length, which is some other number more than the watts saved by other brands' wheels. If you're one of these other brands, you've got little choice but to devise your own test which generates data that proves your wheels are actually the fastest, and on it goes.

We could do this too - take our wheels to the wind tunnel and test them using any protocol we choose. Even if we don't test them against other brands, we'd at least be able to say that they save you some amount of watts over a period of time, and could compare the aero lift of each wheelset in our own line. Trouble is, we don't believe the self-reported data from other brands, and wouldn't ask you to believe ours. It's too easy to create a test or report select data that allows your product to shine, and while we do believe our customers trust us, self-generated "research" would be trading on that trust, not deepening it. Add to that the expense of the testing - which would cause us to charge you more for our wheels - and it's an easy decision for us to sit out this arms race. 

"Testing" a riderless bike in a wind tunnel is just stupid, but we do think there is validity to some wind tunnel testing of wheels - provided that it's credible, which means that it is from an independent source not commissioned by a brand. As you can imagine, tests like this are exceedingly hard to find. But a few years ago the folks over at Roues Artisinales conducted just such a test - establishing a defensible protocol for measuring aero drag of over 50 different wheels. They only tested the front wheels (as the rears are in the wake of the frame and rider, so measuring them by themselves is meaningless), put tires on the wheels, tested at wind angles of 0 to 35 degrees, weighting by incident angle so that drag at the more common 0 to 15 degrees figures in more heavily than the outlying 20 to 35 degree angles, and simulate a riding speed of 50km/h throughout. In short - they've set up the protocol to measure aero drag of the front wheel in an environment that simulates the real world as closely as possible. If you're going to place any stock in wind tunnel testing, this is the way to figure out how much.

What the RA test measured was not how much faster each wheel allows you to go, but how little wattage the drag from each wheel absorbs. That alone is a signficant phrasing. No wheel actually makes you faster since they all generate drag - they can only make you a little less slow. RA puts it accurately, while the brands more commonly tout speed gains. The results ranged from a low of about 17 watts absorbed by the Zipp 808, to a high of about 33 watts absorbed by the Mavic Ksyrium. So the delta between the most and the least aerodynamic race wheels on the planet is about 16 watts of absorbed energy. I think we can all agree that if all we want to do is go as fast as possible in a straight line, we'd all prefer the 808s to the Ksyriums.

But those two wheels are outliers in the test. Everybody's benchmark for aerodynamic performance from a mid-depth carbon wheel is the Zipp 404, which has 18 spokes and a 58mm depth. The 404 tested at 19.9 watts absorbed. (The wheel tested here was the hybrid-toroidal rim shape with dimples that preceded the current Firecrest shape.) Only 4 wheels of similar depth tested better, and they all had fewer spokes (between 3 and 16), so it's clear that Zipp can accurately claim it has a fast wheel.

Most of the wheels tested used proprietary technology, like Zipp, Mavic, Hed, Campagnolo, Lightweight, American Classic, Shimano. One exception stands out - Sonic. You may not have heard of Sonic. They don't have a blog like us, but whatever cache that causes them to give up they earn back right quick by being Belgian. Like us, they use open mold rims, OEM hubs and standard j-bend spokes, in a 20/24 build. Their 50mm tubular wheel tested here is very like our RFSW 50 or RFSC 50, which have a very similar rim shape, the same 50mm depth, the same number of j-bend spokes, and similar OEM hubs.

The Sonic 50mm wheel tested at 20.4 watts absorbed. It has two more spokes than the Zipp 404, is 8mm shallower, and only gives up half a watt of absorbed energy. At 31mph.

Does this mean our wheels are as fast as Zipps? Draw your own conclusion. Mine is that when I toe the line on my non-proprietary, open mold, wind tunnel virgin, 20/24 spoke RFSCs and RFSWs, I don't think I'm giving anything away to the guy next to me on wheels that cost 3x as much. 

Race Smart.


Back to blog


Demonstratable excellence! Both in the blog and in the November products.




Dave helped. He edited out all of my usage of the phrase "those spoke thingies".

Mike May

With all of those technical details, I had to do a double take when I saw the author, I expected to see Dave's name :)Nice read.


Another enlightening read.Thank you.


Leave a comment

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