The wind tunnel seems to think there are significant differences between tire shapes. Contis kick the crap out of the rounder vittorias... and it didn't sell well but schwalbe had a tt tire that was shaped faster.
You may want to re-read my post. I never said that any differences in tire shape were aerodynamically insignificant. I said that high-end bicycle tire casings all assume a round cross-section. To wit:
All bicycle tire casings (clinchers and tubulars alike) assume a circular cross-section (with a constant radius) when inflated. High-quality bike tires are pressure vessels with extraordinarily supple (flexible) walls. They're all round; their inflation pressure forces them to be round. On a clincher, changing the rim width changes the radius of the casing arc, but it doesn't change whether it's circular.
I further specified that different tread thickness profiles allow small variations in net tire shape, even though the underlying casing has a round cross-section.
Road tire designers can alter the shape of the tread rubber a tiny bit by playing with its thickness, but the underlying casing is round.
In my original post, I was referring to the common belief that clincher rim beads somehow manage to make the casing non-round even where the two do not touch. That situation is effectively impossible given the extremely low bending stiffness of high-quality casings and the extremely high bending stresses that inflation pressure would apply to the casing anywhere it tried to assume a non-round shape.
I say "effectively" because it's not theoretically impossible, but the effect would be so tiny as to be unnoticeable to riders and pretty difficult to measure. A wider rim and larger-diameter fire do
result in significantly increased camber stiffness and probably reduced slip angle, which is why many riders prefer this setup. They're detecting a real difference in feel; they're just mistaken about the source of that difference. But "camber stiffness" and "slip angle are esoteric tire engineering concepts, and it's pretty reasonable that most riders would ascribe the difference to something that seems more intuitive to them.
The data is real even those the differences in shape might not leap out for the naked eye. check blather about bikes site, among others.
Oh, yes! Blather 'Bout Bikes. Lovely site. I read it regularly. Oddly, the author (Tom Anhalt) and I are both bike-obsessed mechanical engineers working in the medical device industry. We've never met, but Tom certainly seems competent (at the very least).
I'd bet folding money that Tom would agree with me about tire casing shapes—that their casing cross-sections are circular between the rim beads. Pressure vessels, even composite ones like bicycle tires, are pretty simple from an engineering perspective. Predicting their deformed shape under internal pressure is trivial.
I've not seen Tom speculate on why GP4000s tend to test faster than other tires, but I'd go so far as to say he and I would both ascribe the Conti wind tunnel results to tread pattern and, to a lesser extent, tread profile. If you can find a citation wherein Tom ascribes the GP4000 results to casing
shape, I'll be mightily impressed. I'll also have to reach out to him so I can ask (politely) why the hell he thinks Conti was able to produce a non-round casing cross section on a supple bias-ply bike tire with a symmetric/quasi-symmetric layup.
It's worth noting that Flo Cycling has a blog post about trying to capture the profile of an inflated GP 4000
. Their published profile shows what looks like a non-round casing, though they don't call out where the tread begins.
IMHO, their method of acquiring the profile is deeply flawed. They evidently used modeling clay to make a mold of the mounted tire and then somehow measured that. This is weird, because it's worse than measuring the tire directly (via a Faro arm, or, better still, 3D scanning.
I suspect they sawed the hardened mold in half, scanned the cross section with a conventional scanner (as you would a photo) and then measured the cross-section in Photoshop or something. This is odd; it's inherently less accurate than a 3D scan would be and distortion of the clay as it dried would be an issue. One commenter asks "why didn't you just laser-scan it?" The answer is pretty unsatisfying: "our mold was good enough." If they already knew the shape of the profile with great precision, they could say that. But since they don't have anything to compare it to, they have no idea whether it's good enough.
Flo Cycling is pretty transparent in publishing their testing methods and results, and I appreciate that Much of what they do seems methodically sound, at least to me. But I saw one exchange wherein the Flo people clearly had never heard of the statistical concept of "error bars," used for quantifying uncertainty. I found that a little shocking. In That context, the decision not to 3D scan the tire is disappointing but less than surprising.
Edit: fixed typos