Wheel Test in this week's Comic, Accurate?

andrewgturnbull

Hi there.

A question for our power meter boys. Have a look at this thread:

http://www.timetriallingforum.co.uk/ind ... opic=15769

My question is, is the SRM (presumably the more expensive version) accurate enough to draw the comic's conclusions? Or not?

Cheers, Andy

fossyant

Probably baloney, who knows ?

I got it as well - not sure there is that much difference really ?

SteveR_100Milers

well the question is, is a variation of 11watts down to experimental error or wheel efficiency? There must be some Exp error, so lets assume its 50% of the variation, is +/-6w going to have a significant effect on a TT result? Given that you probably vary more than that from day to day (good legs vs bad legs)?

BeaconRuth

SteveR_100Milers wrote:

well the question is, is a variation of 11watts down to experimental error or wheel efficiency? There must be some Exp error, so lets assume its 50% of the variation, is +/-6w going to have a significant effect on a TT result? Given that you probably vary more than that from day to day (good legs vs bad legs)?

Yes. Wasn't Michael Hutchinson's winning margin in this year's National 10 just 1 second? That's a winning margin of 0.08%. +/-6W is a much bigger variation in power for any rider than 0.08%.

So the question is not whether +/-6W is worth going after, but whether the 'experiment' was accurate enough for the result to be meaningful. I haven't read the article so can't comment on that.

Ruth

Alex_Simmons/RST

Ruth is bang on, was the testing appropriate for the results quoted? I doubt it.

It only takes a slight rider position change to introduce variations larger than the results shown.

Different wheels will be more or less aero depending on tyre choice.

Performance of wheels in indoor veldromes can be quite different to outdoors with highly variable wind directions.

For anyone with a power meter really interested in doing their own aerodynamic field tests to determine CdA (Coefficient of drag x Effective frontal area) and CRR (Coefficient of rolling resistance), then you can use the protocol and spreadsheet I put together here:

http://freewebs.com/trainwithpower/CdA-Crr.xls

It is very helpful when assessing the impact of new equipment and/or rider position changes. Quite the free speed can be gained with good positioning.

bahzob

The consistency required in this case simply means that a given power meter gives the same reading given the same inputs time after time. Most will do this, certainly SRMs will.

(I have a power meter (a Powertap) and have found it to be remarkably consistent over the season.)

BTW I think this thread is interesting more for the views of the "power meters are a waste of time/money etc" than the actual test results. It seems to me
> Either tyres (or wheels or position etc etc) will or wont make a difference to performance.
>> If they wont then no need to test.
>> If they will then how do you test? > Answer you need a reliable, independent measure and the best one is to measure power. At least it gives a basis for comparison.

It would be interesting to see how anybody would devise an equivalent test based on speed, HR or RPE.

Kléber

The test is a joke. Maybe a poor A-level student would try it but if you are really trying to measure efficiency, using an SRM with a +/- 1.5% margin of error is bad way to start. At 300W, this translates into a range of 304.5W - 295.5W. Given the wheels tested were separated by just a few watts, this means the test, as a means to differentiate the wheels, was bad.

Also, buy a tub or tyre and identical models can weigh in at different weights on the scales. A few grams here or there make a difference.

And above all, as suggested above, the change in rider positions where the slightest change in pedalling style or position changes the data so much.

The only way to conduct a real test would be to use a more accurate strain gages in a wind tunnel, where the wheels only are testing on a special jig/bed. A few blokes going to Newport is hardly scientific, like I say it's a poor attempt worthy of a kiddie science project, no serious magazine should test like this.

andrewgturnbull

Hi there.

I'm looking forward to Cycling Weakly's next article. So much so that I've decided to write it myself. Here's the outline for the test protocol:

"Aero Wheels - Real world testing"

Following up from last week's velodrome test, and expanding on last month's roll down body positioning tests, here's a way to find out which top dollar wheel works best in the real world.

Simply assemble a collection of your favourite wheels and make you way to the top of the biggest hill you can find - preferably one which is followed by a flat section and not too much traffic on the crossroads at the bottom.

Next just roll each wheel down the hill on it's own - give each one a wee push to see it on it's way.

The wheel that rolls the furthest is obviously the fastest wheel.

Discard any that get squashed by traffic as they are just plain unlucky wheels.

Cheers, Andy

Alex_Simmons/RST

Kléber wrote:

The test is a joke. Maybe a poor A-level student would try it but if you are really trying to measure efficiency, using an SRM with a +/- 1.5% margin of error is bad way to start. At 300W, this translates into a range of 304.5W - 295.5W. Given the wheels tested were separated by just a few watts, this means the test, as a means to differentiate the wheels, was bad.

While I agree on the test and the accuracy range you quote (SRM Pros quote 2%, PTs 1.5% IIRC) but that is for power meter to power meter difference. Usually each individual power meter (properly calibrated and zeroed) will remain at, above or below real wattage by a consistent fraction and not vary +/- 2% to itself.

bahzob

Taking the caveats about testing etc. the following link gives a pretty extensive list of tyre tests. The protocol used takes out some of the variables cited as reasons to distrust results.
http://www.biketechreview.com/tires/images/AFM_tire_testing_rev6.pdf

Also I ask again of those who dont want to use power meters for tests like these how they decide what's best?

kieranb

haven't read the article but if they tested each wheel enough times then used the mean estimate for comparison variation can be controlled and a valid comparison be made. Ideally the testers should be ignorant of which wheel they are testing - maybe blindfold them and then put them on the bike?

Kléber

Since you're testing the wheels, apply Occam's razor and just test the wheels. No need for a rider on top. As I suggested above...

..."The only way to conduct a real test would be to use a more accurate strain gages in a wind tunnel, where the wheels only are testing on a special jig/bed. A few blokes going to Newport is hardly scientific, like I say it's a poor attempt worthy of a kiddie science project, no serious magazine should test like this."

fossyant

Madness isn't it - I read it and thought what a load of carp - CW never did this sort of carp years ago....

Unless you really need a £1000 pair of wheels, I'd get a rather nice set that had great bearings, were tough, let reasonably light, and aero profiled, then spend all the rest on the other parts...

I know wheel weight is important, but you can get some nice light wheel sets for a lot less than some of those tested.... but then again, if you want the ultimate.......

Ric_Stern/RST

when you see that an 18-mm 'deep' rim is the winner of the best wheel (the clincher section) you just know the testing is wrong.

ric

SteveR_100Milers

Ric_Stern/RST wrote:

when you see that an 18-mm 'deep' rim is the winner of the best wheel (the clincher section) you just know the testing is wrong.

ric

Yep that kind of surprised me too....

flattythehurdler

I would take a few cyclists ona junket to newport, apply some dodgy science in a non random, non blind, non controlled manner, and then make the winner the company that had given me the most sponsorhip/advertising/br.....
:shock: :roll:

Down the Road

Had a pair of the Shamals and was stunned to see them rated as the most Aero clincher. They seemed to think that only tubs go on deep section wheels. Leaving out the Cosmis Carbone in the clincher section makes the test rubbish.

We all have our preferences and those are ALWAYS reflected in magazine reviews and tests.

Most who read reviews have already bought the product being tested and just want verification of their decision.

Distributors probably had no idea what the comic wanted the wheels for and just sent along random samples.