Sunday, May 18, 2008

Aero test - helmets - final

[b]Final results for helmet test 1:[/b]

OK, after a lot of analysis, playing with different stats and error methods here are the final results for Helmet Aero test 1 between the Specialised Decibel, Giro Advantage 2 and Limar Crono 05.

I've also readjusted the data to take into account that the rolling resistance should be the same for all groups of data (since it was... same wheels/tyres for every test). This was done by plotting all the helmet data collected as one series and generating a combined linear trend to get the rolling resistance offset, and using that combined offset for each dataset. (Google Wattage group was some use after all).

The graph also plots the amount of error (SE) for each set of data (plotted as +SE/-SE). (These are the little black bars extending from each data point). At slower speeds the error from each data point overlap the other trend lines (not what we want), but at higher speeds this is not the case (this is what we want).

The graph plots force against the velocity (speed) squared in order to get a linear line, no easy way to force the data through the same rolling resistance offset otherwise.
This means it's harder to read, so the results below are included in the easy to digest format that shows how much time you could expect to save, and your time on the course.

[b]Feedback[/b]

Now does anyone want to participate in Helmet test no 2, which should be able to test a much wider variety of aero helmets (providing we get loaner donations of lids for a week or so).
I'd also be interested in anyone else with a proper power meter (SRM/powertap) that would like to help run tests, can't hurt to have some different riders have a crack at the test protocol either (plus you'll get good info on your drag profile that might allow you to go faster with a few tweaks). And finally, if anyone has got any strings with indoor velodromes, now would be a good time to let us know!

[b]
Results: At 200 watts: 0 degrees yaw (ie a calm day).[/b]

[code]

KM......Decibel...Advantage2...Chrono
Av KMH.....35.5......36.1......36.9

time on course (hh:mm:ss)

KM......Decibel...Advantage2...Chrono
20......0:33:48...0:33:14...0:32:31
40......1:07:36...1:06:29...1:05:02
90......2:32:07...2:29:35...2:26:20
180.....5:04:14...4:59:10...4:52:41

time savings over using Decibel (hh:mm:ss)

KM......Decibel...Advantage2...Chrono
20......0:00:00...0:00:34...0:01:17
40......0:00:00...0:01:07...0:02:34
90......0:00:00...0:02:32...0:05:46
180.....0:00:00...0:05:03...0:11:33[/code]


[img]http://www.efirst.com.au/rob/fishboy/HelmetTest1FinalErrors.GIF[/img]

Saturday, May 17, 2008

Wattage showdown

[quote=fishboy,May 12 2008, 07:57 AM]1- Yes. Was waiting till the dust settled on Wattage for a consensus to be reached (if that is possible!). All that will happen in the end is that we should be able to say it is 95% or 99% certain that the aero differences do exist between these helmets, and the variation is not occurring by chance.
2 - No idea, but that banner seems to chase me around transitions.
[right][snapback]400604[/snapback][/right]
[/quote]
Wattage debate has degenerated into a smackdown between Robert Chung and Andy Coggan. Billed as the prize fight of the non tunnel aero testing methods, I'm not sure its going to help ...

[quote]Andy Coggan
More options May 17, 12:37 am
From: Andy Coggan
Date: Fri, 16 May 2008 07:37:06 -0700 (PDT)

On May 15, 11:57 pm, Robert Chung wrote:

> On May 15, 8:28 pm, Andy Coggan wrote:

> > So why did you essentially dispute me when I said that it wasn't?

> Cuz I can?

Well, hmmm. You can't possibly mean that you'd abuse your position of
authority to imply that you can do things that you can't actually do.
So, you must mean that you think that calculating a pseudo-elevation
profile allows you to identify sources of variability in ways that
can't be accomplished by other means. Assuming that's correct, I'm all
ears...in fact, why not start with my wife's velodrome tests, and tell
us all exactly why her power-vs-speed relationship varied slightly
across the different efforts?

Andy Coggan [/quote]

Sunday, May 11, 2008

Wattage debate (isn't there always!)

Debate currently raging over on Wattage as to how to derive some statistical measure of significance on the results. Hmmn, opened a can-o-worms...

[quote]From: Robert Chung
Date: Wed, 7 May 2008 11:09:10 -0700 (PDT)
Local: Thurs, May 8 2008 4:09 am
Subject: Re: Aero Testing Sheets - Update (Chung & Regression Methods)

> On May 7, 12:04 pm, Robert Chung wrote:

> > Andy (Coggan): with the regression method, using the proper model specification
> > always dominates.

> But (somewhat rhetorically) what is the "proper model specification"?
> The two alternatives are mathematically equivalent, the physics don't
> really help you decide, and the argument could be made that a power
> meter is just as much a force (torque) meter. To me, then, it comes
> down to how errors in the underlying measurements impact the precision
> of the CdA estimate, and that's what I can't decide (since the noise
> isn't necessarily "white", I don't think that a simple uncertainty
> analysis will suffice).

Since this is rhetorical, you almost surely already know the answer
but for those who don't, Andy's right, the physics models don't help,
and the mathematics are equivalent. What's not equivalent is the
statistical model. The underlying linear regression model is unbiased
and efficient (i.e., lowest variance) when the Gauss-Markov
assumptions are met. The "usual" regression model (where one regresses
W/v on v^2) will produce inefficient (though consistent aka
asymptotically unbiased) estimates because of heteroscedasticity.
However, over the range of v we usually see for these kinds of field
tests, I expect the efficiency loss will be relatively small. The
bottom line is that it's slightly better to use whatever the PM
reports rather than to transform the variables, though only very
slightly.

BTW, you're right that the errors aren't "white." Note, for example,
systematic holes in the SRM's speed reporting. [/quote]

Wednesday, May 7, 2008

Helmet goes to heaven

Suffered a setback with the demise of the control Specialised Decibel. It has gone to helmet heaven. Thankfully, no-one injured or otherwise damaged in the process.

Also been consorting with stats nerds to try to get a proper test established so that when we do a big run with a bunch of helmets, we can say "good stuff, there really are differences between these lids, and the variation didn't happen by chance".

Now, anyone got any strings they can pull at indoor velodromes in Melbourne?