Login to VeloClub|Not a member?  Sign up now.
July 21, 2017
July 20, 2017
July 19, 2017
July 18, 2017
July 17, 2017
  • That was a great read, Wade. Thanks!

  • Richard Riley

    Thanks for the write up Wade. Great read – looking forward to the next Test the Best that I can get to so I can test one of them out.

  • Karl

    Great write up Wade. I’m not taken by the look of the bike, but I like the innovation and drive that has gone into it.

  • Cam

    I think the marketers are still winning here. Wouldn’t a better test have been to keep all other variables the same and simply test bike versus bike, and logically current Venge versus new Venge? If I was in the market I would want to know how much faster this bike is from the current model.

    • Marcus

      Agree. Same kit, same helmet, same wheels – just switch the frames (and handlebars if they come as a package). It’s otherwise a good comparison, but it muddies it a bit by not isolating the component that is actually being sold.

      • BenW

        x3 here. May as well have stuck the rider on a freeride bike, it’s about as valid a comparison. “Oh hey, look at us changing everything to prove what we want”. The little Summary at the end was the most telling bit.

    • Chris Yu

      HI Cam, we had actually debated whether to do the comparison with just the bike changing. At the end of the day though, we wanted to illustrate the magnitude of gain that using a full aero system could provide. That was the main point of the test – though I do realize that everyone wants to know exactly what each part of that system is worth, especially the new Venge. If you hop over to our website, we provide an average breakdown. Roughly speaking, about half the gain is from the bike and wheels and the other half is from the rider accessories (skinsuit, helmet, shoes).

      • Marcus

        That’s great Chris – and there is certainly value in illustrating the gain from the full system. But given I already run aero wheels on my Tarmac, and can use all the other aero kit, all I want to know is – what’s the gain from the frame/bars itself, if I were to switch my frame and bars for the new Venge? Given I’m not coming from alu box section rims, it’s not going to be 120secs over 40k – we’re currently left to guess probably somewhere south of 60 secs? The lack of transparency in all this makes me think the saving from the frame itself probably isn’t that much?

        Also, what about a direct comparison with other leading aero frames – S5, Propel, TMR01, Aeroad, etc? Of course, using the exact same kit and components and switching the frames/bars only. If the new Venge is all that, should be a very worthwhile test for Specialized to promote, no?

        • Chris Yu

          Marcus, understand your skepticism. We’re here to be as transparent as possible. For the competitor bikes, we have historically not posted direct comparisons and let 3rd party tests do that (honest question: would you believe us if we did?). However, we have of course tested all the current class of aero bikes, including your list. Keeping wheels and components the same, except when integrated parts are required (e.g. bars on the Venge ViAS, S5, Canyon Aeroroad, etc…), we have measured roughly 50-55 sec over 40km saved to the prior fastest in class (though all the current crop of aero frames, original Venge included, test very tightly together – typically within a 10 sec/40km band).

          • slowman

            good point Chris, no i dont believe much of the “hype” pushed by marketers. i agree with Marcus too, why not tarmac 5 and venge?

          • Ragtag

            Chris, thanks for clarifying. I do think you could have just compared with the old Venge and made this more ‘believable’. Please try to empathize how the reader, reads this: “New Spec bike saves 5 mins!!!. Wait a min, I am sorry the bike only saves 120 seconds of that time. Oh wait a min, sorry, umm, uh compared to the Tarmac. Oh and that too the SL4.” Now if you have a few Specialized sceptics out there you cannot blame them for thinking that your marketing is over the top. If you had just said, New Venge saves 50 seconds over old Venge that would have been good. But I guess if you have been taking the hype approach it must have been working earlier. I so love the wind tunnel videos and your engineers really seem to be working hard looking at the various changes and pace of changes that you guys do. The look of your bikes are really good. And the bikes are ofcourse really good since they win so much. But yet I still cannot make myself buy a Spec product because the brand to me stands for all this marketing hype and shadow tactics. This is not meant to criticize the brand, company or yourself. But merely to point what some riders think and how this prevents us from buying a spec product. Though I guess we might be in the minority. All the best for future bikes.

  • Tea_Addict

    A space is required between the quantity and the unit symbol; unit symbols are not pluralised; the symbol for the hour is ‘h’. Please re-read the SI Brochure. Oh, and your statement regarding aerodynamics in motorsport is ignorant beyond measure.

    • RayG

      I noticed that too. If a car is doing 250 km/h then aerodynamics are even more important that at 40 km/h.
      And you put in twice as many watts and only saved 143 seconds? How does that prove the claims?
      Oooh, look who’s paid for a banner advertisement on CT.

      • I was paraphrasing, but it’s what the two aerodynamics PhD’s from McLaren told me.

      • jules

        I suspect what the aerodynamicists meant is that an F1 car has a surplus of power most of the time. the limiting factors are acceleration (power-to-weight) and cornering speed (downforce, which has an inverse relationship with the slipperiness of the car – i.e. you need to disturb air to generate downforce). those do not apply to cycling.

    • Jaybo

      how is it ignorant? would love for you to actually explain what you mean by that.

      F1 cars (or any powerful car really) can do 250km/h into a 40km/h headwind without too many issues, low aero drag will help this a little bit but the real tricky bit for F1 engineers is designing car that will actually give a positive downforce benefit without impacting the aerodynamic drag too much. no point having mega levels of downforce if it generates so much drag that you can’t actually push it forward.
      Wade’s phrasing isn’t exactly the easiest to understand, but if you know what you’re talking about it can make sense.

      bikes dont need to generate downforce, but they do need to be slippery as all hell, you’ve only got ~300w to push that weight around so you want as little resistance as possible, you haven’t got 500+kw!

  • touristeroutier

    These solo tests are interesting, but I often wonder how effective they are in the real world. Consider that one is drafting, or riding with a group most of the time (even if taking their turns at the front. I suppose this test shows some of what could happen in a solo breakaway, but would there be any statistically significant difference in a sprint?


  • Robert Merkel

    I’ll give Specialized credit for actually putting their bikes with journalists in the wind tunnel – “aero bikes” without actual data to back it up are a joke. That said, I’ll be interested to see whether the results can be replicated in other wind tunnels. Even without the temptations of marketing, if you optimize in one context it’s not surprising it works well when tested in the same context.

    Some some other questions/comments about the test setup:

    * be interested to see what happens to the advantages from that skinsuit with a couple of gels in the pockets and race numbers pinned on (after all, who cares how it performs without race numbers). In fact, does it actually have pockets? If not, it’s a nonstarter for anything other than crits and TTs.
    * I’m curious about their assumptions about wind strength.
    * I notice that you’ve only got one cage and bidon. Does adding a second bidon to both bikes – which most riders will -reduce the difference between the two?
    * Is the bike really going to be rideable in crosswinds with 64mm wheels – particularly for the likes of Matt Wikstrom?

    • Chris Yu

      Hey Robert, yes the GC skinsuit has pockets (as well as a full zip and opening front). When designing the suit, we actually used gels to determine the pocket location and to negate the aero impact from carrying them.

      For the on-road testing, we stuck with 1 bottle and cage to minimize the amount of swapping between runs. The comparison data holds with two bottles as well.

      What’s the question about wind strength?

      • Robert Merkel

        Thanks for the reply, and I’d like to say I really appreciate the “Win Tunnel” video series.

        My concern about wind angles is that (as I understand it, and I’m no aerodynamicist) the biggest effects of aerodynamic shaping on bikes are when the apparent wind angles are within a certain range – from memory, about 5-15 degrees off axis. The gains when the wind is head on (which is what occurs on a still day) are much smaller, and if it’s really off-axis (in a howling crosswind) you can actually get worse performance from the aero gear.

        So without getting into every single tedious detail about how you’ve modeled wind, presumably that’s an adjustable parameter in your computer model. How sensitive are your time/wattage gains to those changes?

        • Chris Yu

          Robert, that’s a good question. Typically, the aero differences will vary a bit depending on the relative crosswind angle. With most modern aero equipment, the performance gains are actually magnified with more significant crosswind angles (the physics are similar to a sailboat harnessing crosswind to sail upwind). With the test that we did with Wade and others, we saw some moderate crosswind angles so the comparison takes that into account. In general though, the quoted time savings we post are conservative in that we focus on low to no crosswind (with more moderate crosswinds, the gains will usually be larger vs. traditional equipment).

  • jakub

    well, the marketing hype remains still there and IMHO it can’t even be more obvious. stacking up aero helmet, skinsuit AND bike and claiming that it saves 5 minutes? reasonable comparison would be same clothing/helmet and different bikes. moreover, serious scientific experiments are conducted as double-blinded, to remove any potential bias. unfortunately such biases unconsciously happen in situations you would never imagine. (there is plenty of research about this, from popular stuff see for example Kahneman’s Thinking Fast/Slow, or something by Dan Ariely). honestly, on scientific grounds this “experiment” would be laughed off. I reckon that new Venge might bring some aerodynamic advantage, but the whole thing is once again exaggerated.

    • mv1in20

      Placebo effect doesn’t matter. They have power meters to standardise effort.

      • jakub

        I am not talking about the placebo effect. I am more talking about input from Specialized/McLaren people when for example fiddling with the data they put into Midas software. As a researcher I can tell you that there is something in the back of your unconscious mind that always kind of pushes you to see the results you’d like to (and as I said, this has been proved many many times in scientific literature). The whole procedure of running the data through some proprietary software (which is normally used to simulate aerodynamics of F1 cars) and then running “simulations” seems a bit bogus to me. There is so many unobserved factors that might have an effect on the results… To give you an idea about what blinding in scientific studies is – in a proper experimental setting you would have a group of people asked to ride at some wattage over the circuit, first with Venge and then on the new model (or, possibly, randomly selecting the people into control (old Venge) and test group (new Venge)) – _without_ knowing that they are going to test super-duper new aero bike that saves 5 minutes over 40km, or any similar info that might create bias. The only information the participants would have is that they should ride at let’s say ~200W average. On the other hand, researches while evaluating the results afterwards would not know which data belong to new Venge or old Venge runs.

  • some bike guy

    A fully enclosed bike would probably save 30 minutes or more. So I don’t understand the purpose of this engineering and marketing. Just bring the “Velotilt” (search Google images) to the market and there you go. In a real world situation, but not in a race. Sadly, we do not a have a discussion about what would actually be beneficial for most people but only for a couple of rich guys in a race.

  • francis

    wade, you said you went 150w on the trad setup and 300w on the aero setup.
    wont that explain the 2:23 faster time on the aero setup and not because of aerodynamics?

    • Winky

      I think the idea is that the computer simulation can adjust for power to normalise the aero results. Still keeping the power identical (or as close as possible) would seem like the best approach to get a reliable result. If nothing else, riding harder tends to put one’s upper body into a lower position which would completely invalidate the test results.

    • The simulator can normalise the data based on whichever power profile you choose. In the end, it didn’t matter which power profile we used – both ended up being the same.

      • RayG

        Sorry, Wade. Legitimate or not, this has ruined a lot of the chance of people buying this. “The simulator can normalise the data”. Yeah, sure. Go back and do it at the same power, or give us the figures from someone who did.

        • jules

          have you ridden with a PM? you can’t hold a constant power value. it’s impossible.

          • RayG

            Yes, for 5+ years now. You can still get roughly the same average, rather than one being twice as high. And where did I say he needed to hold a constant power?

            • jules

              you didn’t, but I am. wind resistance varies with the square of velocity. you can average power, but you can’t average speed and wind load, aerodynamics, etc. a normalising algorithm is unavoidable. unless you are willing to tolerate inaccuracy.

              • RayG

                Well, you could hold constant power on a velodrome, but then everyone would be complaining it wasn’t real world conditions.

                Plus there’s no way the NP for a 150W AP effort would be anywhere near the same as the NP for a 300W AP effort. Having APs roughly the same would at least give him a chance of having similar NPs.

                Finally, if you re-read the report and the comments, you’ll see he actually went 10 mins faster for the 300W effort, but they’ve done some technical jiggery pokery to say he would have gone 143 sec faster if he had done the same AP. They haven’t just used NP to come up with that.

                So I repeat my statement. Do the tests with the same effort or show us the numbers from someone who did.

                • jules

                  I’ll bet good money they’re not using NP. what they would surely be using is individual chunks of data from each ride to fit to a profile of how power, speed, rider, bike relate to each other. I’m speculating, but I can’t believe they’re just taking a single figure – AP or NP – and extrapolating a curve from that. but this is almost meaningless to discuss as I don’t know how they’ve done it. they do sound like they know what they’re talking about though.

                  • Alex Simmons

                    It’s not that difficult to run such models yourself in a spreadsheet like Excel. The maths of the physics is relatively straightforward and well understood and published. Simply micro-segment the rider’s actual ride power data and course profile with gradient values and wind vectors, input relevant CdA, mass, Crr and other key variables (e.g. air density) and check validity of the model, then re-run the model with revised mass and CdA values accordingly.

                    The main trick is dealing with the power and other data being time based, and not location based. That’s a segmentation problem.

                    I (and others) have been doing this sort of thing for many years, and about 8 years ago I layered onto it course environmental variances captured during an actual ride using a “virtual elevation” approach. I did this to account for such things when running pacing optimisation models (to compare applies with apples).

                    What I didn’t see in this report (but I may have missed it) was how well the actual ride data matched the model based on the inputs from tunnel testing.

                    IOW what validation of the model was performed?

                • Alex Simmons

                  You can’t hold power constant in a velodrome. It’s pretty much impossible. You can ride such that you produce steady laps times, but power (and wheel speed and cadence) will fluctuate quite a bit during a lap in a pseudo-sinusoidal manner.

                  Not that it matters as holding power constant is not necessary in order to assess aerodynamics.

                • velomonkey

                  I’m totally with RayG on this one – it’s really poor journalism to buy into marketing hype and have some custom 3rd party software that you know nothing of “normalize” the data and then say “um, yea, I got about the same time savings.” Come on, you did one ride at half the power!!!!!

        • Alex Simmons

          It’s entirely feasible to assess such things from rides of different levels of effort. My main criticisms of doing so though is that when riding at significantly different levels of effort, it’s possible that CdA is impacted in various ways:
          – natural positional changes when riding hard versus easy
          – a difference in bike speed results in a difference in yaw values
          Often riding faster can result in a reduction in apparent CdA.

          One can only guess at some other possible interactions when changing velocity of relative air movement over objects, including whether an assumption of constant Reynolds number applies for all shapes and sizes of components of the bike and body.

        • Do you define normalise the same as we are in this testing? i.e. adjusting values measured on different scales to a notionally common scale. We can definitely do this.

  • Winky

    One of the things that has me thinking when it comes to aero set-ups is the effect when riding in a race or group, and one’s ability to hang in. When in the group, the advantage would not seem to matter too much, but the accumulated effect of the time spent on the front (if any!), and closing gaps (perhaps due to an attack, hill, corner or simple inattentiveness) surely adds up. Aero set-ups potentially allow you to close the gaps either a bit more quickly, a bit more easily or both, reducing the effort and time spent out of the shelter of the bunch. This saving accumulates “exponentially”, as later in the race you might be a little less fatigued and can continue to close the gaps more quickly and/or easily than otherwise. We’ve all known the feeling late in a race or ride of “only just” being able to close a gap (and taking an age to do so) and the knowledge that the next time it may not be possible!

    • jules

      if you want to win, you will need to put your nose in front at some point!

      • Winky

        True enough. The benefits of aero when in front and in the wind are clear. I was thinking of the less obvious circumstances.

  • stefanrohner

    Yes! of course, and with the 2017 model it will be 10min!

    • Tim

      That’s one thing you can guarantee with Specialized, this year’s “Wonder Bike” is next year’s “Retro Benchmark”.

      The cumulative gains of 40 years of Specialized marketing must be close to infinite by now. They are that good at this sort of thing.

    • CC

      and… 2017, will see another 300+ pictures of your bike collection.. somethings in life are inevitable ! :)

      • stefanrohner

        and probably with some new ones, hope not mass production stuff. and not hidden, always with my name.

  • Freddie Merckx

    A well written article but to answer your headline, errrr NO it CAN NOT save you five minutes over 40km!

    Even so, I’m sure there is an abundance of wannabe sportivo legends only too willing to sell their backsides for a piece of this marketing b*llocks

  • Phillip Yan

    IMHO there is a flaw with the study data …What is the power output of each rider in Aero/Traditional Frame in each trial? If we make a claim in time difference ” Time Saved” between AERO/Traditional, a power output difference between AERO/Tradtional trial also need to be considered.

    To put in perspective. I cannot claim Car A is more aero than Car B in a 19.4 km course with a 10% lap time reduction, but neglecting the fact Car A has 200 horsepower more than Car B.

  • Rob Booth

    A more valid test would be the same power for the same distance. Dr Hutch did one of these on a velodrome with the result that the aero bike was quicker. The use of a wind tunnel, fancy software and journalists just smacks of marketing BS.

    • Robert Merkel

      In outdoor conditions the apparent wind often comes from directions other than straight into your face. That can considerably affect aerodynamics, and can’t be replicated in an indoor velodrome.

  • CC

    Wade & Chris – brilliant initiative and write up. It’s hard to shake the cycling community, and push the bar a little… great job. It’s probably fair to include some awareness marketing for cafes here in Melbourne… skin suits may have a solid upside, until ordering the coffees :)

  • Daniel

    Interesting article, thanks.

    I would be interested to see the advantage of this bike when looking at the entire race situation in which this bike would likely be used (flat stage ending in a bunch sprint). Some people have already raised the question of the impact that the frame would have when drafting, and I would imagine that Cavendish is unlikely to gain a notable advantage when in the bunch and possibly only slightly more when being lead out. What impact, however, does the frame have in the all out sprint? When the bikes are being thrashed side to side and elbows are going everywhere, does the frame create any real advantage? Or is the real use for this bike at the top level for the poor workhorses who have to sit at the front for hours keeping the break in check?

  • Kieran Degan

    Interesting. I think CT should do their own independent real world testing. 5 cyclists at different levels, ride on an old venge and then with the new venge package. I’ll happily volunteer :)

  • CGradeCyclist

    Firstly – kudos to Specialized for being transparent with the wind-tunnel data and how the testing was constructed.

    Honestly, I have a real problem with how this is being sold. ALl the marketing says “The Specialized Venge ViAS saves you 5 minutes”. But then you have comments form Specialized (and I quote): “we had actually debated whether to do the comparison with just the bike changing. At the end of the day though, we wanted to illustrate the magnitude of gain that using a full aero system could provide.”
    So they directly contradict their own marketing – its a ‘full aero system’ that provides 5 minutes, not the bike at all. And based on their own figures, the bike provides 1 minute max (and that’s not even against other aero bikes).
    Again, I applaud their transparency. But I don’t think the message the figures give in the end is what they think it is…

  • Paul

    Aero Frames are not really my thing – regardless of any advantage it may give, for me at least, I need to also like the look of the bike that I am riding.

    Aspects of this do appeal to me, but the shape of the handlebar, stem and head tube area do not. I do however like the idea of no exposed cables (at least in the handlebar region – much like a track bike. If there was a way to do this on a more traditional bike, I’d be all over it!

    One questIon I do have for specialized is: have you tested the placement and shape of the quick release skewer? It sounds like a silly question, but I am serious. In all the photos of the ‘experiment’ I note that the bike seems to be affixed to a jig at both the front and rear wheel, thus (presumably) hiding or at least obscuring the skewers. I’d have thought that it would be more aero to move to a thru-axle type set up whereby there is no parts ‘hanging in the wind’.

  • James Hall

    Good to see some real engineering. I recall when I got a Soloist Carbon and and getting the sail effect in a cross headwind uphill. As more a climber (i.e. no good on the flat) I wouldn’t ever go back to non-aero

  • jules

    my question to Specialised – how much of aero drag is apportioned to rider + bike? I know that with aerodynamics you can’t superimpose drag figures of two profiles – i.e. they work together in an integral fashion – but broadly speaking, if you tested rider and bike separately, what is the relative aero drag of each?

    the reason I ask is that I’ve always assumed that the major factors in aero drag are, in descending order:
    1. rider profile (the big bulky object with large frontal area)
    2. the wheels – and particularly spokes which act like an egg beater in the wind
    3. the rest of the bike

    I’ve got to say that intuitively, I have a hard time accepting that #3 (the bike) can have such a big impact on aero drag. of course, the bike still influences:
    – #2 (wheels) – as they come with the bike. although obviously you can buy wheels separately and many people (like me) have their own race wheels, so what comes with a bike from the showroom is irrelevant other than for training on Beach Rd (ok maybe that is important)
    – #1 – this is the key one. how much of the Venge’s aero advantage is down to how it positions the rider? I can see in the photos the rider position is kept relatively similar. but I know I can impact my power figures significantly, for a given speed, by adjusting my body position. the position shown in the photos here of our model Wade is very conservative (straight arms). the important thing is that it’s consistent between both bikes – granted – but what happens if he lowers his body, as you’d expect him to do in a real TT?

    • jules

      one more question – is the Q-factor the same for both bikes?

      • Mi

        And another, are they planning to do two different wheel sizes for the same bike? Maybe a 700c and an 850b? Fat road bike anyone…..? Oh hang on, that’s their MTB marketing dept. Must make for great head of dept meetings……

  • Jamie

    One real issue here is that you don’t state the actual time it took to complete the course/average speed. This is poor – you know very well that absolute time gaps will look more impressive if the ride is slower.

    Based on the information provided I think it is possible to work out that you were normalised to be travelling around 30/32kph (Tarmac/Venge), so that 5 minute saving is only if you are riding 40km solo at pretty modest speeds and taking 1hr20 to ride 40km i.e. in sportive. Time gains at higher speeds will be lower.

    The fact that you did not point this out (something I think you would have been very much aware of) smacks of collusion with the marketing department to promote their headline figure.

    When they want to promote power saving everyone is riding at 30mph, when it’s time saving its 20mph!

  • Alex Simmons

    The reduction in CdA values is really what we should look at. Removes all the misunderstanding about models.
    Those are large differences in reported CdA.
    Aero road kit: CdA (m2 , 0 deg | 10 deg) 0.290 | 0.294
    Traditional kit: CdA (m2 , 0 deg | 10 deg) 0.359 | 0.349

    A drop in CdA of ~ 0.06 m^2.

    That’s a 40-50W saving in power at same speed, around 5-6 seconds per km.

    It would have been far more helpful to have provided the CdA values by simply swapping the bike and to use same wheels and rider kit. IOW the only change being the frame.

    Sure do the comparison with the aero wheels and aero clothes as well, but making several changes at once tells us nothing about the impact of the frame. I’ve measured 10-20W differences between skinsuits, let alone regular bike clothes. Helmets too can account for sizeable wattage savings. Wheels are pretty obvious. Making all these changes at once tells us very little about the effectiveness of the frame alone.

    • jimbo

      Agreed, thats a massive drop in CdA. The body position MUST have changed in the bike swap over. I dont believe for a minute that a frame change and misc. other aero items can result in such a dramatic drop.

      • Alex Simmons

        Helmets and skinsuits can make quite a sizeable difference depending on what the starting kit was like. It’s hard to say really. That’s the problem when you don’t isolate elements when testing – it confounds the results by making multiple changes at the same time. That’s fine for understanding the total combined effect since aero gains are not linearly additive, but it’s unhelpful when the primary test is to ascertain the improvement due to changing the bike/frame alone. It’s entirely possible (but unlikely in this instance) that the frame was actually aerodynamically worse but other changes more than made up for it. I can’t recall if the tyres and tubes were the same either, they can also account for quite a handful of watts depending on what was used in each case.

  • Notso Swift

    Very good article

    But it is still ugly


Pin It on Pinterest

July 21, 2017
July 20, 2017
July 19, 2017
July 18, 2017
July 17, 2017