Thanks Beaver! Where is this graph from?
I find interesting that it totally busts the myth that testing tires on a roller in a lab is valid. It also gives us a clue why high tpi tubulars and open tubulars, which are more flexible, posiibly have a lower crr because of that. It finally confirms our empirical knowledge that a pressure around 90-100 psi is the best for most conditions.
This only holds true for tire pressure (i.e. you can't keep increasing tire pressure to lower crr), not the rolling resistance of the tires themselves. The relative ranking of tires tested in the real world would still track the results of the roller tests because the primary drivers of energy loss is still going to be the same (e.g. hysteresis). So yes, tires with more flexible casings will perform better on the road, but they also performed better on rollers anyways (the top tires are usually supple casings, thin tread, little/no puncture protection).
like wind tunnel tests, you have to understand the results of rolling resistance tests in the context that they're done under specific conditions and how they apply to the real world is something you'd have to test on your own, for your own conditions.
Can someone explain to me how significant these differences actually are? Like 0.16 vs. 0.17 at very extreme angles, so significant is this anyway?
Real world yaw angles will be 0 to 5° unless you are going very slow or are riding in the desert or on the seashore.
In track aero tests, shallow wheels perform very well, and the biggest difference is in tire width.
Yup, looking at Mavic's data, you can see that you'll find a lot more wind near the coast (Kona) than inland (Annecy). Sec. 3:http://engineerstalk.mavic.com/en/yaw-a ... an-course/
So where you ride is also going to factor in the relative performance of wheels.
And of course no aero thread is complete without the aero rule of thumb, especially when the data is presented in CdA (yay!)
0.1 lbs (50 g) of drag (@ 30 mph) = 0.5 s/km = 5W = 0.005 m^2 CdA = 0.0005 Crr
These results would be a lot more interesting if they included rotational drag, especially compared to some of their non-disc counterparts