Originally Posted by
PeteHski
Rolling resistance is a fairly major component of any cycling power model. If you ignore it then your power model will be pretty inaccurate. However, you can simply ignore gyroscopic forces and MOI without significantly affecting model predictions.
Yup. That's what we found. When predicting power for speed (or speed for power) you can either include or leave the MOI out and compare the two results. With the current resolution of power and speed measurements, you'll be hard-pressed to spot a consistent difference--so we left it out in favor of parsimony.
Originally Posted by
One Wheel
Where can I get my hands on that model? I wish I had the expertise to create one myself, but I don't. Ignoring gyroscopic forces and MOI sounds like the kind of short cut I would be forced to do if I developed that model.
https://pubmed.ncbi.nlm.nih.gov/28121252/
Originally Posted by
cyclezen
significant in terms of some percent of your power lost to it. But not very variable. Most every informed study I've read/seen shows rolling resistance varies so little in the realm of cycling and velocity.
What you bring to riding is mostly constant, whether you're going 10 mph, 20 0r more... so it is a factor of the equipment you bring and the road surface, but after that it's close to constant.
'
Rolling Resistance as a factor of velocity'
there is a study read quite some years back where bicycle rolling resistance, as a variable of velocity showed an almost flat slope to RR as a factor of velocity...
it's one of those very marginal gains, if one is willing or needing to consider...
will a difference of 4-5 watts saved make much of a difference in my or anyone's ride?
To be fair, we often can get away with assuming that Crr and CdA are fixed with speed because over *most* of the speeds and surfaces that bicycles travel, that's *almost* so. The Reynolds number changes with speed which affects CdA, and at high speeds you start to run into limitations with the "spring rate" of pneumatic tires. But for the speeds that most of us go, we can usually get away with treating them as if they were constant and our predictions will still work pretty well. That's actually a clue: when the predictions start to not match with observations, then we usually sigh and add the extra damn variables back in.