everything seemed normal, until I got to this chart. It implies that higher pressure is always better. That defies the ST wisdom on the subject, which claims that too high of a pressure isn't better than a more 'optimal' pressure. I'm sure we'd all give up some comfort for speed come race day. Should we?

>that is correct, on a smooth surface, but roads are not smooth surfaces.

Is there any evidence to suggest it's not correct on roads? Everything I've seen shows pretty good correlation between smooth and bumpy.

I think the reason to not go super high on roads is just comfort.

OK- slowtwitch just published an article on the new Conti's
http://www.slowtwitch.com/..._4000_S_II_3946.html

everything seemed normal, until I got to this chart. It implies that higher pressure is always better. That defies the ST wisdom on the subject, which claims that too high of a pressure isn't better than a more 'optimal' pressure. I'm sure we'd all give up some comfort for speed come race day. Should we?

assuming that chart is taken from roller testing, it would be valid only on perfectly smooth surfaces.

Tom A's field test here (which you can find at aeroweenie.com along with every other useful bit of data in the world) shows how ON road CRR can start getting worse at a certain point:


That single field test certainly isn't the final word on the matter, Kraig Willet has tested this too and I don't think he came to the same conclusion.

further research would be great =)

Crr per se always decreases when you increase tire pressure because the tire deforms less. What can increase due to the pressure increase is dissipation due to slip force but you need a very good testing protocol (or analyze this just using analytical models) to capture this. Roller testing wouldn't work due to the double contact and the higher contact loads. A testing bench with a belt and a wheel (that's the setup that Continental uses) driven at the same speed wouldn't work either because you need to allow slip in the contact.

The main issue about what you describe is what I like to call "energy path". The bouncing of the rider is induced by the pedalling loads but it's a movement in the vertical direction. Changing the stiffness of the tire you can modify how severe is that bouncing but I don't see an "energy path" that allows to recover some of that energy for propulsion purposes because the horizontal and vertical movement of the bike are pretty much completely decoupled.


The role of both rolling resistance moment and slip force on what it has been called Crr may not seem obvious because slip is not very common in typical road cycling situation. If you ride your CX bike on ice (something similar to increasing tire pressure a lot) it's more obvious, you can identify the two mechanisms that dissipates power: 1) Rolling resistance moment and 2) the product of slip force and wheel radius, a moment that also tries to decelerate the wheel. In that case, a traditional definition of Crr wouldn't be sufficient to capture total dissipation

Abstract

A linear model for evaluating the influence of road roughness on total vehicular rolling resistance is presented. This model considers rolling resistance to be the result of three energy dissipation mechanisms:

(a) smooth-surface rolling losses,

(b) energy dissipation in the tire due to road roughness, and

(c) losses in the suspension system due to relative motion between sprung and unsprung mass.

The first of these mechanisms is completely dependent on the properties of the tire, specifically, hysteretic losses accrued from tire deformation during rolling. The other mechanisms are dependent on road profile, vehicle velocity, and vehicle parameters, as well as tire properties.

Calculation of losses due to rolling resistance are made by means of a quarter-car model to evaluate the relative magnitudes of the various mechanisms. The results obtained indicate that for rough road surfaces, the losses due to the roughness-related mechanisms are comparable to those arising from smooth-surface hysteretic losses.

It would be interesting to know what's their definition of "losses" but my level of geekiness isn't high enough to pay for papers ;). Just reading the abstract, I can see that they aren't differencing mechanisms that tend to decelerate the wheel (the first one) thus having an effect on horizontal motion to those that doesn't (the other two) so applicability here may be limited.

I will write a post about this someday

If energy is transmitted through the tire and dissipated elsewhere, then it can't be "returned"

IIRC, I don't think Kraig's speed range was as large as my halfpipe testing, which could be a confounding factor...

plus, he doesn't seem to be able to get the same sensitivity from his field testing...

I've probably done more than 300 laps on multiple venues/days with varying speeds, all up mass, tire construction, pressures, etc, and haven't come anywhere near the ballpark of reproducing these results: http://www.aeroweenie.com/...ta/tire-pressure.jpg


My methodology produces results that are consistent with my own smooth/bumpy roller testing and wind tunnel results.

What were your findings for on-road crr? Higher pressure is faster indefinitely?