Coming from a gov't economist, nice use of "ceteris paribis!" A better translation would be, "all else being equal" but you got the idea of it.

---

---

Portside Athletics Blog

Hmm...does that mean we'll be seeing a different frame underneath Jens next season?

---

http://bikeblather.blogspot.com/

I think you have hit it right on the head.

Back in March, Trek got Pezcyclingnews to write an article on the February tunnel testing with Basso on a Equinox TTX and a P3C and the difference was "within the margin of error". According to several people who were at the tunnel, it took 2 days of changing Basso's position before they could get his drag as low as on the P3C in his original position, so the TTX was only as fast as the P3C because they improved his position, not because the bikes were equally fast. According to my sources, there was a significant difference in favor of the P3C when the rider positions and the parts used were the same.

Now they show tunnel tests with the same February date, but all of a sudden the rider is Levi. Yet the CdA of Levi is supposedly the same as the CdA of Basso, and way above where Levi tested previously. Could it be that they would rather not use Basso's name anymore, and Levi was never tested on a P3C? Or do they really buy a P3C for every Disco rider to test him in the tunnel on our bikes? Or did they squeeze Levi on Basso's P3C, which would obviously be way too big for him? I can't make sense out of it.

And to top it all off, the already unsustainable "the bikes are the same" from the pez article has turned into a significant gap in favor of the TTX on bikeradar, even though they are supposedly comparing the same frames? The SSL name of the bikeradar makes no difference, that just changes the weight, the shape is the same (especially since this test is from February when there wasn't even an SSL to begin with). In the most benevolent case, it shows that tests with a rider on the bike are completely unreliable and unrepeatable (we have done tons of tests with riders on the bike, and you just cannot get riders to keep the same position when switching bikes - which is why we use a test dummy, see our latest eNews (https://customercare.cervelo.com/esupport/enews.aspx). In a less benevolent way, maybe the inaccuracy of the test was used to skew the results in favor of the one paying for the test.

---

Gerard Vroomen
3T.bike
OPEN cycle

The outline on the screen method is next to useless. We did that in the past, but the line is so thick relative to the rider size that you can easily have a 1-2cm change in rider position without noticing, and that can make quite a big difference in drag. The one thing you can see with the outline is that a rider never keeps his position during a test, and therefore that it is impossible to make the sort of equipment comparisons that Trek is trying to make. It's not that the movement of the rider is a problem per se for the aerodynamics, it is that you cannot guarantee that the rider will make the same movements from one test to the next, and so that comparison is invalid. That is again why we use the test dummy, so that we can make the comparisons between equipment without having to wonder if we are really measuring equipment changes or rider changes.

And it is not even that the overall equipment changes are small, the thing is that we want to isolate changes. So we will do test where the entire bike and rider are the same, the only difference is the shape of the seattube between the front derailleur and the seatstays. We want to analyze only that part, and once we have found the best shape there, move on to the next part of the bike. You cannot do that with a rider on the bike due to accuracy, but you also cannot do that without a rider due to relevance. So the solution is to have a test dummy. We just spent another three full days at the tunnel doing just that, analyzing the bike detail by detail, one yaw angle at a time. Four hours on 10cm of tube, maybe a lot of work, but at least something that will eventually prove that we are not, as some people assume, getting to a point where all bikes are converging and the differences are getting smaller. In fact, if anything the differences seem to be getting bigger, as some bikes get dumber and dumber shapes.

---

Gerard Vroomen
3T.bike
OPEN cycle

Boy, that wind tunnel seems about as leaky as a certain drug testing lab in France...

---

----------------------------------
"Go yell at an M&M"

THAT's pretty funny.

---

Gerard Vroomen
3T.bike
OPEN cycle

the plot thickens. Or the soup smells fishy ... something like that ...

In grad school, I had the good fortune of getting the noted economist Jagdish Bhagwati as the professor for one of my econ courses. With his Indian accent, it sounded like he was saying "get her asparagus." It took us days to figure out what he was talking about. Great course, though, and this is coming from someone who is abysmal in economics.

It's no big deal, this sort of stuff happens all the time.

---

Gerard Vroomen
3T.bike
OPEN cycle

Now, with all of the different brand and fit options available it is largely coming down to warranty and service.

This is the un-sexy part of success in this business or any business for that matter. People always love to drone on and on about product, and innovation and the science of it all ( but given what Gerard said, even that can be grossly flawed). And in the post Interbike time-frame these discusssions always seem to reach a peak. This is not to say that innovation is NOT important - it is. However, assuming that the product is close, what REALLY drives a business forward are those less glamourous and un-sexy things that you do behind the scenes with your customers to service and support them and build solid working relationships with them to the absolute best of your ability. That's hard to measure in a wind tunnel :)

---

---

Steve Fleck @stevefleck | Blog

True enough. Elite is a good example of this. As far as appearances go, their bikes have nothing to distinguish them from the competition (IMO) and I they appear to be MOP as far as aero is concerned. Yet, they continue to build a very solid growing business, due in large part for their reputation for excellent customer service. Having an excellent bike is now table stakes for most of the market. As someone else stated earlier, it is indeed a great time to be a buyer.

Well, I would think that service always counts, whether product is similar or not. But the notion that somehow product quality and performance is converging is not yet the case in the bike industry, quite the opposite. The ultitude of possibilities has increased so much, and the degree to which new design tools are used as well, that the differences are increasing. 20 years ago, everybody bought a tubeset from one of three tube makers. The resulting frames really didn't differ that much in performance. Now, with the endless options in carbon material, lay-up, shapes combined with the intense lack of know-how in most of the industry means that the variation is much bigger in the overall spectrum. That said, it is true that within that spectrum there is a big cloud of frames that are all very similar, simply because they are made by the same manufacturer using the same materials., lay-ups and similar shapes.

We see something similar in aerodynamics. There are quite a few frames out there that are worse than an old style steel frame, thanks to bigger frontal area and stupid shapes. But there are now also (though very, very few :-) frames with less drag than the Lotus. So the size of the spectrum has actually increased.

---

Gerard Vroomen
3T.bike
OPEN cycle

Of course the options for meaningless differentiation are greater than the options for significant gains. That's why you see diamond-shaped "aero" tubes and nonsense like that. But the options for significant gains now are still larger than they were 20 years ago. I agree that "significant" is still relative, the diamond-shaped frame still gets you to the finish, and if you are strong enough it may even get you there first. But any bike with two wheels will do that, and it doesn't change the fact that different equipment would get you there faster, be it 3, 5 or 10 minutes. Is that a significant gain? Depends on the person.

So basically, the spectrum in options at any point in time is greater than the spectrum in significant gains. But that spectrum of significant gains is bigger now than in the past. It may be small now (or big, depending on your own concept of what is significant) but it would have been even smaller in the past, not bigger.

---

Gerard Vroomen
3T.bike
OPEN cycle

[reply]
We see something similar in aerodynamics. There are quite a few frames out there that are worse than an old style steel frame, thanks to bigger frontal area and stupid shapes. But there are now also (though very, very few :-) frames with less drag than the Lotus. So the size of the spectrum has actually increased.[/reply]

This is funny yet true. Even without optimizing tube shapes, a lot of companies could learn from the leaders in the industry...

Headtubes where the widest point is the bearing cups
3:1 aspect ratios
Bladed seatstays

It seems like everyone is always testing against the P3C.
Why is that?:)

Can Gerard tell us which are the very few frames faster than the Lotus???Thanks

Interesting. This implies that wind-tunnel testing of individuals is not that worthwhile.

We're creatures of habit: maybe a static pose will come out different. But I suspect when you get on a bike and pedal for five minutes, on average, your movements are going to be very close to identical. In my own field testing experience, I get very consistent numbers in the same position. Also, I can click my camera timer, then run and hop on the bike and get identical photos nearly every time -- and that only gives me about 15seconds. In the tunnel, the adjustments needed to match the outline were quite minimal.

BTW, does your dummy move its legs? If not, are you sure there are no interactions between moving legs and the frame? John Cobb has certainly suggested there are -- at least when it comes to water bottles.


- jens

Aren't you really comparing apples and oranges here, though? Gerard's point is that it is difficult to use the video images to be sure that the rider is in the exact same position on different bikes, fitted with different equipment. What you're talking about, though, is degree to which the rider assumes the same position each time when riding the same bike. I agree with you re. the latter assertion, at least for an experienced rider with good "body awareness", but I also know from experience just how difficult it is to set up two bikes exactly the same way, regardless of how many measurements you make, templates you devise, etc. That's partially why the P2T vs. Javelin comparison that I did worked out so nicely: the bikes essentially had the same head tube lengths and top tube lengths, such that I didn't need to change stem height or length to replicate my position, all I had to do was transfer the bars and stem together.

BTW, just for clarity, it certainly isn't the tunnel staff that leaks. In fact, they have always been super-professional, we never learn anything from them about other companies they test for.

---

Gerard Vroomen
3T.bike
OPEN cycle

Thanks for all the replies to this thread.

Mostly on topic....It seems that no testing is done at 15 or 20 mph, always at 30mph. You could model the results and calcualte for slower speeds, but how do you know your perfect 30mph shape doesn't stall out at 18 or 20mph and is actually worse for your average Joe? Is it a limitation of the wind tunnel?

Testing at higher speeds increases the signal-to-noise ratio, i.e., the magnitude of the difference you're seeking to identify is larger relative to random errors in the drag measurements. There is no need to test at lower speeds, because drag force is highly linearly related to wind velocity^2 over the range of velocities a cyclist is likely to travel (and then some).

At what point is a PXC going to be faster than my old Softride RocketTT? How about a FasTT? (and yes I'm talking about a pedaling rider on board, not a test dummy, though I do believe a pedaling test dummy could be constructed using a motorized bottom bracket to turn the cranks). I love the work that you are doing and believe that Cervelo's expertise could improve upon the beam-style technology greatly, but trying to make a dual-use TT/Tri legal bike seems overly constrained due to your investment in CSC. I suspect this has played a significant role in the loss of Bjorn as a Cervelo athlete.

I wasn't a Softride fanboy until I rode one and spent some time using the adjustment mechanisms that they provide, but now I am a believer. Anyone doing a race like Hawaii on a double diamond frame is just wasting their watts. The Zipp 2001/3001, Lotus, Cat Cheetah, Corima Fox, Softride, and even the Kestrel KM40 (not the Airfoil, due to tube shapes) all bring something special to the table that just can't be replicated under UCI rules. I am not saying that every one of these bikes is faster than a P3C, but I know some of them are, and I know why. I suspect that Gerard and company know why as well, but I don't understand why they are not leading the way in pioneering a new generation of out-of-the-box triathlon bike design.

Chris