Specialized Builds a Wind Tunnel


So there’s this news that Specialized has built its own in-house wind tunnel. My professional reaction was nothing short of “Holy cow!” It’s a colossal investment for a facility that will do nothing for a company that produces mountain bikes, city bikes, kids bikes—even an electric bike—and none of the bikes in those categories will be affected by this new facility. For a week or two Specialized’s PR team had been posting little teases on FB and Twitter, photos that were mostly just jokes, along with the phrase “aero is everything.” I was curious, but mostly only because I could tell it would lead to some announcement. But what?

More than four years ago the road product manager at one of Specialized’s competitors told me flat-out that we had essentially reached the end of the line in terms of big gains on weight and that all the real advances in technology that would aid performance would come from aerodynamics. The point being not that bikes wouldn’t continue to get lighter, but that the gains would be so incremental and at such an incredible cost in terms of durability and expense that for most bike companies the diminishing returns wouldn’t justify the investment. Instead, the gains to be made in aerodynamics were (and are) relatively low-hanging fruit.

Computational Fluid Dynamic software has speeded up development time by giving engineers virtual wind tunnels on work stations. But that software has limitations. The work stations are crazy expensive (and thats from a guy who doesn’t find Apple products to be unreasonably priced) and the license for the software costs what an engineer does. And then there’s the fact that you can only learn so much in CFD. At a certain point, you have to go to the wind tunnel. When you consider just how expensive wind tunnel time is (it can run what a good recording studio does) and how much of it you need (eight hours is barely enough to get a fair picture of how a single bike with no rider performs), then you can see how it would be possible to keep one busy for three shifts per day.


Having been on-hand for a company’s rental of time from the San Diego Low-Speed Wind Tunnel, I’m aware that you rack up a host of other costs any time you do testing. There’s the down-time for travel, and while a SoCal-based company like Felt can drive to San Diego before rush hour, Morgan Hill is a full day’s drive; any other location requires a boarding pass. Add meals and hotels to your transportation costs, and suddenly renovating a warehouse starts to sound like a pretty good idea.

I should also add that there are a great many products on the market that bear the signifier of “wind-tunnel tested.” It’s a do-nothing claim. I don’t mean to suggest that companies lie when they report that. No, the point is that plenty of companies visit the wind tunnel after the design work is complete, molds are cut, and production units are about to ship to retailers. “Blowing” a bike or helmet or wheel after the design work is complete means you know how much drag it generates. Wind-tunnel development requires multiple visits and for longer periods of time than are required just to make the claim of tested.

Significant in this that the staff and students of SBCU will gain access to the wind tunnel to aid them un better understanding the aerodynamic implications of a given fit; it’s yet another good reason for a retailer to send staff to Morgan Hill for training.


There is a darker side to this announcement, though. This ratchets up the bike development arms race. If you’d asked me which company would be first to introduce an in-house wind tunnel, I’d have said Zipp. I gotta figure the boys in Indiana have one on the drawing board. Prior to the PON acquisition I’d have guessed that Cervelo was working on one as well. Methinks that the new owners might be a bit more conservative in their spending.

While you might have winced at the industry’s first $10k bikes, news that Specialized had developed a $20k version of their Venge in conjunction with McLaren caused more than a few people to require the Heimlich Maneuver. Marketing costs generally get spread over an entire company’s bike line, which is why it’s so important to have a popular $200 mountain bike if you want to sponsor a big pro team. However, development costs are charged to the category of bike that generated them. You can expect to see all S-Works bikes tick up with this additional expense.


I expect we’re going to see a faster Venge. I expect to see that. What I wonder is if we’re going to see a more aero Roubaix. I don’t really care for how this is going to ratchet up bike development costs. But we’re going to learn a lot because Specialized is going to learn a lot. I love learning. I’m an eternal student. That part excites me. For that, I can’t freaking wait.


, ,


    1. jessyjessy

      I am exploring this subject as part of a report I need to do on possible careers I might choose. Thank you for your post it has valuable information on this topic.
      Volumetee tshirt

  1. Richard Wharton

    I think it’s awesome as well, but remember – the cost/watt-saving or cost/gram of drag reduced is exponential. It’s sort of like the EADS vs Boeing war. You end up with improvements in the single or tenths of a percent. We’ve optimized weight, we’ve optimized frame stress and strain (and we still get cracks here & there), we’re throwing money at wheelsets and helmets and tires, and of course, we’ve placed millions in to bodies and hearts and legs and lungs. I just wonder if there’d ever be an opportunity for an “IROC” style of racing, where everyone has the exact same bike, and the best rider wins.

    That, and of course, the UCI should allow the technology to develop sans 1890’s restrictions….

  2. MCH

    I’m curious about what this will mean for Specialized’s wheel business. Aero improvements in wheels still seems to yield the biggest gains in performance. Yet when I think of the best carbon/aero wheels, Specialized/Roval doesn’t come immediately to mind, and certainly isn’t in my top 5. Perhaps the ROI for this investment will come from big marketshare share gains in aero wheelsets.

    A dedicated facility should also be great fun for the designers/engineers. In addition to frames and wheels, they could begin to explore improvements in clothing, crank arms, etc.

    Should be really interesting to watch.

  3. Q

    The closest thing to IROC racing on bikes is the Little 500. Honestly, I think waching a race like that between ProTour teams, bike exchanges and all, would be a lot of fun to watch.

    I too am fascinated by what we might learn from this, even though it has little to do with the old aluminum road bike I ride. It seems to me that in another 5 years or so, they’ll have found most of the aerodynamic low-hanging fruit, much as has already been done with regards to weight. Possibly the real benefit will be for Specialized-sponsored pros. Now that we know that a rider’s optimal TT position should take power output into account, professionals with an interest in TT performance will be able to spend all the time they need to find their optimal TT position. That will continue to be important long after the bikes themselves have been optimized as far as they can.

  4. Patrick O'Brien

    That’s neat. Maybe we could get them to be a prime sponsor of the Daytona 150 Speed Weekend? Team time trials to determine starting position, then racing on the tri-oval complete with infield camping, dining and betting.

  5. Scott G.

    If aerodynamics were the point of bike racing, recumbents would have taken over
    the TdF long ago. Fully faired bents have much less drag and much more area for advertising.

    ps. Specialized will now be able to select more aerodynamic
    people to develop into bike racers.

  6. MCH

    A quick side note about recumbents – if the Lemond’s “slam the seat all the way back” method of seat positioning is correct, taken to it’s logical conclusion we should all be riding recumbents. You can’t get any further behind the pedals than that! 😉

  7. Kc

    There is “spec” racing in Japan, Keirin. Spec wheels, frames, running gear. Pretty cool, but I like the arms race in road bikes. Not really sure how they’re going to shave the Venge any more, but they are going to have to redesign the entire S Works line to adopt disc brakes, so maybe it makes sense.

  8. Gary

    Specialized has a nearly full product line unlike Zipp or Cervelo. They can spread the sunk cost across more products.

    The marketplace battle with Trek can move forward to their advantage. Theoretically, they have a big jump start ahead of the competitors. 2 big sales knobs from the technical side are number of design turns (shortened cycle between turns) and time to market.

    The advantage for their sponsored Pro Tour teams increases as well. That means more brand visibility and buyer association with the brand.

  9. Bikelink

    The equipment arms race for racing never made sense to me. It only exists to make more exotic stuff to feed $ to the industry…I wish there was a 20 lb weight limit, round tubes only, 32 spokes limit on material and depth of wheels, round spokes all required. And I’m a cat 4 masters racer who can and does buy the expensive stuff (wheels most recently) to be a little less slow.

    If it’s about enjoyment then fine road feel and all that can be pursued but other than RKP you barely hear about that, it’s just weight, stiffness, and aerodynamics.

    There is nothing more intrinsically enjoyable to riding a bike that weighs 3 pounds less or more if you don’t swap back and forth and everyone is riding the same thing. Same for riding at 16 vs. 18 miles per hour (aero) or 26 vs. 25 in a race.

    But, since the bar seems so high to do a group ride, forget about race, we’re losing tons of people who never will do those things as a teenager or an adult.

  10. Matt

    Mike Sinyard ain’t into losing money. I’m guessing, even if it was $1M or more they’ve found a way to make that pay…that and what’s $1M or $2M to a company worth more than $600M anyway?

  11. Bikelink

    Is there an issue with the independence of the analysis? Like, will anyone wonder if their #’s from their own wind tunnel with no one looking over their shoulder are real? Will their competitors wonder? Regardless it will be a huge benefit in development as you’ve described, but will they eventually want to allow for independent assessment?

  12. Gary

    There’s no independent analysis at this point either. Every manufacturer claims their product is the most aero based on wind tunnel testing, CFD etc. That doesn’t change for Specialized going forward. If you are that concerned about what really works, you buy your own windtunnel time.

    Aerodynamics isn’t as cut and dried as, say, weights. Very subtle changes can make large differences and replicating data requires great rigor in process and setup. For an individual rider, it will always be a system perspective with THEM on the bike. That’s not to say that manufacturers can’t create fundamentally better (lower drag) items. It’s just that you can’t guarantee that every bike that has that item will be faster for all users.

  13. Adam

    I really hope this leads to practical aero testing. There’s little point in having the most aero bike in the world that doesn’t consider the rider. (I’m yet to see a bike pedal itself). With access to more time, they can begin incorporating the rider into aero development – after all, we make up more than 75% of the drag of a bike set-up.

    1. Author

      Bikelink: Years ago, when I was working as a mechanic, I used to decry all the technical advances. Once I began working for a magazine and actually riding all the new stuff, putting hundreds of miles on the newest bikes, my tune changed. I received a very stark education in something I’d known intrinsically all the time: a truly better bike makes for a better experience. I’m glad I don’t have to ride bikes with only round tubes and that at least some of my wheels don’t have 32 spokes. The Felt F1 that I’ve been riding for more than a year and I’m trying to write a review of as we speak (write?) is an extraordinary machine and I do believe it has made my enjoyment on some descents greater because it handled better than some bikes and that improved handling increased my confidence, and confidence is an important ingredient in enjoyment. The point you’ve chosen—round steel tubes, 32-spokes, 20-lbs.—is just as arbitrary as requiring the Little 500 racers to all ride those clunky Roadmaster bikes. I think there’s something to suggesting that racers ought to me more or less on the same machine, but just which machine is up for much debate.

      Gary: It’s easy to be cynical about how a company with their own wind tunnel can lie about the data they get from it. And while I’m not suggesting your view is cynical, I think it might be interesting to offer a different perspective: Building your own wind tunnel is such a colossal expense that no one would go to that much trouble to just lie about data. Consider that any number of companies have used the San Diego Low-Speed Wind Tunnel and no matter what they claim the results were, no one from SDLWT ever speaks up and calls them out as liars or says they fudged the testing. As to independent verification, there doesn’t seem to be much motivation. What sells bikes are race wins. Just witness the way Trek grew for the years Lance was winning the Tour. Making a bike that is truly better is a competitive edge out on the race course, not to mention what it does for a rider’s mindset if he believes that he’s on the best bike. When that contributes to a racer’s success, the company succeeds as well.

      Adam: I’ve talked with a number of engineers about why no one tests with riders. Engineers test bikes without riders because they have the ability to change only the bike. The problem is that every cyclist sits on the bike differently. Not only do we all have different drag coefficients, the way we generate that drag varies from rider to rider. Until we can standardize how we all sit on the bike (which flies in the face of anatomically-guided fit), there’s no point in testing bikes with riders.

  14. blacksocks

    There’s a small wind tunnel in Scotts Valley, CA as well…been up and running for a while. It didn’t cost $1M and it’s smaller in scale, but incredibly useful for helmet development…

  15. Patrick O'Brien

    It would be nice if helmet makers spent as much time on improving safety as improving aerodynamics.

  16. Full Monte

    You could put me on a $20k wind-tunnel engineered Venge and I’d still have the aerodynamic equivalency of a skateboarding rhino wearing a batman cape.

    1. Author

      Blacksocks: Not publicizing having a wind tunnel does make it harder for people to be suspicious of the purpose for the wind tunnel.

      Patrick O’Brien: Fair enough, but I’m in no position to say they aren’t.

      Full Montey: You might be right, but I’d pay to see a rhino skateboarding while wearing a batman cape. Hell, I’d buy the DVD.

  17. Patrick O'Brien

    Here is one place I found testing data about the CPSC standard testing applied to inexpensive and expensive bicycle helmets.


    I also did some research on new safety innovations on helmets, such as Multiple direction Impact Protection System (MIPS) which claims to protect from rotational forces from oblique impacts. The Specialized hemet in your previous piece had MIPS. I did wonder how a helmet with an aerodynamic shape, not rounded, would perform with MIPS. My thoughts were that a helmet with multiple snag points and MIPS would be a zero sum game.

    The Snell foundation also has tested many helmets, cheap and expensive, and found no major difference in protection. I settle for a POC Trabec because of shape, extra coverage, thicker inner foam,excellent fit, and the fiber net reinforcement of the inner foam. If the MIPS technology tests well and is validated, I may upgrade to the Trabec Race in the near future. It’s your head you’re protecting folks, especially those of you who race. I can’t understand why bicycle helmet technology has been stagnant for so long in the safety area.

    If I am wrong here, I would appreciate a correction and information.

    1. Author

      Patrick: I don’t think you’re wrong, not by a long-shot. The MIPS technology sounds promising and makes a great deal of sense given what I’ve learned about the forces at work in a “typical” (if we can use that term) crash. I’ve seen a lot of crashed helmets and I’ve yet to see signs one snagged on anything—though I admit I’m no expert on these matters. The thing I come back to is that there doesn’t seem to be a helmet that offers a notably increased level of protection out there, so I’m going to stick with helmets being marketed to roadies, except when I’m on my mountain bike and then I’ll wear a helmet marketed to mountain bikers, if only because that visor really is handy on occasion.

  18. Patrick O'Brien

    Thanks Padraig. I really want to learn more about this and see if I can find some data on MIPS equipped helmet testing. As I get older, 64 this month, I am placing more importance on those brain cells that have survived the assault of that wonderful combination of barley malt, hops, water, and yeast. I am just a recreational rider and occasional overnight bike tourist, so I wear the Trabec with visor on the road and trail. Your observations on damaged helmets not showing evidence of snagging damage is very interesting. If I discover any other interesting info, I will share it and the source.

    1. Author

      We’re all just recreational riders. I refuse to allow that to devalue our cycling experience, though. We get more fragile as we go and taking care of what we have seems like a sound investment in our future riding.

  19. Gary

    Difficult to convey true intent with text on the web. I’m not cynical at all; my comment was a response to Bikelink. Apologies if it was misconstrued.

    One not-as-obvious advantage is improvements in the actual process of testing. One current limiter is the time for changeovers etc to make runs. Faster changeovers = more runs = quicker design cycles = quicker time to market.

    The market ultimately decides the value of “improvements” (aero, weight, handling, color :-)). If you are happy with your current bike, great. Us old guys need all the help we can get.

  20. Pingback: Lift : Red Kite Prayer

  21. Patrick O'Brien

    Padraig, you are right about recreational riding. I once passed a rather large guy on a hybrid bike while going up Ramsey Canyon road about 3 years ago. It is a good climb. I yelled out “on your left,” and he replied “hell, everyone is on my left!” As I rode with him for little while, I told him at least he was on his bike and not on the couch. I see him occasionally. He is much smaller and still riding.

  22. blacksocks

    @Padraig , @Patrick O-Brien – I would say that there is no such thing as a “typical” crash. When you consider the many variables involved (rider attributes, surface, speed, angle, rotation, equipment, time window, fall vs impact vs slide, and so on…the bottom line is that every crash is truly a unique event when considered through the lens of science. That said, there are millions of dollars being spent on materials research, design, testing, injury analysis and studies that will hopefully improve understanding of the brain and how it can be protected. The helmet market will evolve from all of this activity, just like any competitive market, because riders want choices and manufacturers want to make the best products possible to serve the need. What is great in this moment is that more riders are choosing to wear helmets when they ride, awareness of cycling and infrastructure for cyclists is getting better in many areas, and current helmet designs are considered to be highly effective in many instances because regardless of whether they’re “Aero”, “lightweight” or “recreational” they have to meet common standards that are relevant to cycling. ( http://www.helmets.org/stats.htm#effectiveness). As a rider, I hope we continue to improve all of these factors to make riding as fun and safe as possible for as many people as possible.

  23. Patrick O'Brien

    Blacksocks, I agree with most of your points. But, I would like to know who, which bike helmet maker specifically, is spending millions of dollars on understanding the brain and protecting it and introducing new safer helmets to achieve that goal. Why aren’t any manufacturers advertising a helmet that exceeds the CPSC standard? The CPSC standards for bike helmets have not been changed since 1999. Mandatory use of helmets in UCI sanctioned races only occurred in 2003. I have not seen and appreciable improvement in cycling helmet design as it applies to safety in the last ten years. My point of view, which is limited to a rider looking for a safer helmet, indicates to me that reducing weight, improving aerodynamics, and improving ventilation, while still meeting the CPSC standard, has been the primary focus for helmet makers until the last year or so. Safety has not improved for the majority of the models offered, and the protections they offer remains essentially the same regardless of price. My question remains: what took so long to improve safety when the makers obviously spent money on other parts of helmet design? Why did a primarily snow sports company break into bike helmets with a better safety design? My hope is that riders demand safer helmets and that the manufacturers deliver.

  24. Full Monte

    @ Patrick O’Brien

    “Hell, everyone’s on my left.”

    Allez! Allez! That made my day. Good sport, he was. Perfect bike humor. (Been that guy plenty in my life, and I’m stealing that line.)

  25. Eric W

    Hells bells –

    I’m thinking of making my own personal wind tunnel.

    Does that sound useful and not particulaly difficult to anyone else?

    Eric W.

  26. SteelClassics

    Why do they only use the tunnel for the bike. I understand that the bike has to be aerodynamic, but it is the rider that make all the difference. I read up somewhere there was about 70% more drag.
    There is no point having an aerodynamic if you stick a large person on it!

    1. Author

      SteelClassics: As I mentioned in my previous reply, Specialized can only change the bike, not the rider. Suppose they go into the wind tunnel with Tom Boonen; whatever they learn about his aerodynamics is peculiar to him. Unless your position is identical to his, anything they learn about making him more aerodynamic by changing his position isn’t applicable to you. If you want to improve your aerodynamics on the bike, then you must visit a wind tunnel.

Leave a Reply to Gary Cancel reply

Your email address will not be published. Required fields are marked *