Tire Pressure vs Rollability

Per Robert Chung:
One of the problems with
uncontrolled coast downs is that you can't be sure about the entrance
speed.

I never thought about that part. Instead, I picked a spot where
my entrance speed could be zero. Position the front wheel on
the spot, do a track stand, let off on the brakes, let nature
take it's course.
--
PeteCresswell

(PeteCresswell) wrote:

I picked a spot where
my entrance speed could be zero. Position the front wheel on
the spot, do a track stand, let off on the brakes, let nature
take it's course.

Yeah, I've tried that, too, with similar results to what you're seeing:
neither max speed nor rollout distance are very repeatable. This, btw, is
the protocol that Heine uses for his BQ coastdowns, which is why I think his
results haven't been very precise.

Yesterday I happened to do a ride where I climbed and then coasted down the
same hill twice. As it happens, I pedaled a bit at the top of the hill so my
"entrance" speed differed between the two runs, as did my max speed during
the run. However, once I started coasting my power was zero and speed was
determined by aero drag, rolling drag, slope, and my mass. For each of the
two descents, the slope was the same, my mass was the same, and since my
wheels and tires were the same, the Crr was the same. *If* my position were
the same and there was no wind (that wasn't the case yesterday), my CdA
would have been the same. So using these data I could calculate the CdA and
Crr for that descent (if the wind hadn't been blowing). Then, if I had
switched tires or tire pressure but held my same position, I could do
another run and figure out what the change in Crr would have been.

I've done this sort of thing quite a bit, though usually in order to
estimate CdA rather than Crr. However, Crr scales exactly like slope, so a
Crr of .005 looks like a slope of half a percent. If it's windless and I
hold my position right, I can repeatably detect slope differences at least
an order of magnitude smaller than that. That's why if you want to do this
kind of thing, a speed recorder is the way to go.

Phil W Lee wrote:

Surely if you enter at zero, then roll down to zero after the end of
the downslope, all you need to measure is the distance.
If you want to isolate rr, take the chain off first so that nothing
but aero and rr are factors, and to eliminate aero, use some method
that give an exactly reproducible riding position, like tri-bars.

In theory, yes; but in theory there is no difference between theory and
practice and in practice there is. As you could see from Pete's data, and as
has been found in other experiments by other riders on other hills, rollout
distance isn't a particularly sensitive metric. Worse, from an analytical
point of view, with a single metric like rollout distance, when it varies
from run to run there aren't any diagnostics that help you figure out why:
was it a gust of wind, a different path, a slight movement in position,
better balance, or something else?