wheels vs watts

poynedexter

me and a mate did some wheels testing this morn. the idea was to see which set was faster (if at all).

i've three sets:

vision trimax 30, conti ultrasport 25mm tyres 100 psi (came in bike)
full carbon clinchers, novatec hubs, 23mm vittoria corsa 110 psi (the lightest of the 3, and my race wheels)
pacenti rims, hope hubs, sapin spokes, schwalbe one 25mm 100 psi (handbuilt and my everyday wheels)


the test was, we would ride on a mostly flat, with 2 small rises course with each set up. my mate would target 300 watts and i would ride beside him.his set up remained unchanged (the control). we would then record the difference in normalised power over the 5'30" effort.

the findings were

test 1: time 5'30" av sp 22.4mph NP 258 watts
control: NP 297 watts (diff 39 watts)

test 2: time 5'16" av sp 23.7mph NP 262 watts
control: NP 307 watts (diff 45 watts)

test 3: time 5'31 av sp 22.6 mph NP 263 watts
control: NP 312 watts (diff 49watts)


we tested at this wattage as it represented a good effort level for us and if one wheelset was significantly better it would surely show up. it tells me 3 things really. my £400 everyday handbuilts are my fastest wheels in this test. my carbon clinchers arent quite as racy as i'd thought. also the standard wheels are still good. 260 watts hurts regardless of which wheels i use. also the wind must have picked up for run 2 cause we went much faster.

it kinda pisses me off when i hear companies claim a 10 watt saving at 25mph about their products, because on a road bike (not TT) when i'm putting in 25mph efforts, i'm the bottleneck, not the wheels.


the carbon clinchers are soon to be for sale. so how much would a set of wheels cost where i'll see another +10 watts difference? £1500?

poynedexter

oh and the difference between us in this test? we were both on our good race bikes, both about the same height, and aero position etc.

about body weight 26lbs (not all fat) costs about 49 watts on this course. thats the game changer, not wheels folks.

sungod

different tyres, different rolling resistance
the affect of positional variations on drag
the 'control' here is demonstrably uncontrolled - it did three different speeds at different powers
the wind changed, perhaps the yaw angle did too, do you know what the yaw was? drag differential at 0 yaw can be very different to that at 15 degrees

it appears that you only rode once per wheel, in that case you can have no idea what really caused the variation

you rode different stuff, different ways, under different conditions and the results were different

poynedexter

no i disagree completely. it was a live no BS test.

it was a real world test on the same road, both riding on the hoods etc. yes the wheels were different for each test, that was the point. the only change was wheels and wind. but in the REAL world, we put in an honest effort, and its not so easy to maintain exactly the same power each time. you can yaw angle all you want, but it demonstrated how little difference there really is when wheels are switched: 10 watts range in NP range. and since i don't typically ride about at 260 watts, then tyres wheels etc do not play such a difference. and all my wheels are in good condition. if we had time i'd like to have tried a cheaper rs11 set up and repeated the test

i was expecting more difference tbh.

amaferanga

It demonstrates that you're most definitely not a scientist.....

poynedexter

amaferanga wrote:

It demonstrates that you're most definitely not a scientist.....

never claimed to be. but it does tell me how little difference changing from one wheelset (inc tyres tubes and pressure) to another makes, and that my so called race wheels arent quite so great.

Alex_Simmons/RST

Riding next to someone has an impact on power demand similar to the differences quoted here (the aerodynamics impact is not insignificant), let alone all of the unknown variables in play. Also NP is not a metric suitable for such analysis.

If you have the data files for both riders there might be something to salvage using better analysis methodology but it certainly helps to have a good test protocol to begin with.

Consider it a learning opportunity.

But I would be very careful drawing conclusions that you have.

Websta24

I think before jumping on the OP you should try and understand that he is not claiming to be a scientist.

Its a real world test, not an experiment in a wind tunnel. The variables are exactly that, variable!

sungod

i'm not jumping on the op, simply pointing out that this is not a sound basis to draw the conclusion the op claims

the op followed up with the statement "you can yaw angle all you want, but it demonstrated how little difference there really is when wheels are switched", this demonstrates denial of physical reality

Grill

Websta24 wrote:

I think before jumping on the OP you should try and understand that he is not claiming to be a scientist.

Its a real world test, not an experiment in a wind tunnel. The variables are exactly that, variable!

Which begs the question, what's the point?

buckmulligan

poynedexter wrote:

amaferanga wrote:

It demonstrates that you're most definitely not a scientist.....

never claimed to be. but it does tell me how little difference changing from one wheelset (inc tyres tubes and pressure) to another makes, and that my so called race wheels arent quite so great.

To summarise what the other guys are saying, the "noise" in the data caused by your less-than-rigorous testing methodology far outweighs the differences that you're trying to pick up in the different wheelsets; if you repeated this test you would not see the same results and would likely not even see the same relationships (i.e. ranking of wheels relative to each other). You need to be A LOT more controlled in your testing if you want to draw the kind of conclusions you're stating.

If you're actually interested in investigating this further have a read of the thread below. It should give a good indication of how systematic you have to be to get reliable data:

http://forum.slowtwitch.com/forum/Slowt ... _P3536905/

And as for the assertion that you're looking for 'real world effects', well all outdoors cycling is done in the real world, so these effects are always at play, whether you can pick up on them or not. What 5W or 10W translates to in terms of speed or relative exertion is a separate matter.

ajmitchell

if I understand you correctly you tried to see what difference in power is gained (or lost) by riding a course in the same time and same conditions but varying between 3 wheelsets. First congrats for trying, there is certainly no harm in that. Some people will never try and discover anything for themselves but just criticise others. Clearly if others can do it better, they should and post the results or offer constructive comments. I tried a similar protocol myself and don't claim it is perfect.I think the tricking thing is hold everything constant but the one thing (the wheels in this case) that you want to change. For example temp/air pressure should be the same and unless under study the tyres too (and of course rider, wind etc). In your test the control is a bit of a red herring as the control is not under test and obviously they shouldn't be in your draft or visa versa. Riding side--by-side is apparently a 5% aero gain but if its the same on each run that not a major issue. The only criticism I have is that the speed was different esp in run 2 (but you explain wind was diff which unfortunately pretty much invalidates that run). If you held power constant then speed gain could be linked with the wheel change. If you held speed constant then power gain could be linked with the wheel change (see https://www.youtube.com/watch?v=P63k0_A2ABk). If you vary speed *and* power its tricky to say much unless speed improves and power is also lower (double gain). Anyway good first attempt. I would suggest for future attempts vary one small thing and record temp and wind conditions each time.

Grill

ajmitchell wrote:

Clearly if others can do it better, they should and post the results or offer constructive comments.

Many of us spend a lot of time and money testing and not dispensing the results of said testing freely is a choice. Plus as there is a massive different in wheel performance with both frame/fork interaction as well as tyre/tub choice, it's unlikely it would be relevant for more than a handful of others.

Alex_Simmons/RST

I test cyclists' aerodynamics for living. I have a fair idea of what one can and cannot discern from field data.

As I said, if the data is available then it may be possible to do an assessment using more robust analysis methodology and be a little more definitive. But it's also quite possible that the testing protocol was inadequate to reliably answer the question posed. That will be evident from a better assessment of the data.

poynedexter

well, this didnt start out as a pure scientific study, more of a trial to see what might show up.

so to follow on, i fitted my "fast" wheels and lined up in the small scratch group in a local race last night. the reality for me is that it wouldnt matter which wheels i'd have used. riding through, putting in digs and generally riding well into vo2 max for long periods, i would have needed wheels to give me 30 watts advantage on av power. thats not going to happen with wheels i could afford (or any wheels).

i'll fit my carbon clinchers for my next event on sunday.