The Most Aero Cadence? Tested! 

how to be more aero? PT: ride backwards ( thumbnail 🙂 ).

Five out and five back! Finally some testing with a sample size greater than 1 or 2. I'm looking at you, GCN does science.

I wonder how significant the position of the derailleur cage and chainline is in these tests. At 55rpm. The RD is necessarily shifted into a smaller cog and as a result the cage will be retracted.

I dreamed of this question the past few weeks ! thanks for creating this video!

Lower cadences also result in a lower demand on the cardiovascular system (e.g. lower HR). But you've got to be able to produce those higher torque levels without the muscles and joints breaking down. Time to get in the gym people!

I think this has the most applicability for time trials where if you can sustain a power that may not be optimized for human biomechanics at a lower RPM, there's some significant gains there. Conversely, I don't think riders should be rushing to lower their cadence in pursuit of aero because of the biomechanical drawbacks that come with lower cadence towards heart rate, muscle fatigue, etc. All that said, data is data. This at least empowers more informed choices.

Can it be the higher leg speed against the wind at the top of the revolution, drag being exponential?

Ullrichs secret finally revealed

Amazing detail and work. Was you able to capture crank angle. I wonder at different cadences a varying proportion of the stroke is spent in flexion, extension. i.e stompy or smooth pedalling.

The results make perfect sense to me due to the aero pumping forces generated from spinning cranks, pedals and legs just like wheel hubs and brake rotors produce. Now I just need to invest in new knees!

Cool. Looks a good strategy on shorter prologue style tt's. Don't think I could maintain low cadence for greater than 10mins but anything under that there maybe some performance gains.

Brilliant content. Idea: Compare disc vs rim brakes on the same bike. Simply swap the fork as the front will be the majority of the difference. No need to actually connect the front brake for testing. A concealed rim brake fork would be ideal. The wheel could be tricky to keep constant. Build two wheels using the same rim. Of course if you use a rim brake rim people will say you handicapped the disc setup as they claim the brake track is not aero optimized as a disc brake rim. 4 wheels with two different rims would settle the question. Of course simply choosing typical wheels with the same depth and width would give a pretty good idea of the difference without going through all the aforementioned pedantry.

You're in a class of your own when it comes to scientific analysis of cycling on youtube! Congratulations! Slightly off topic, but if I understood you correctly, you were surmising that your bike yaw was reduced with the lower cadence because you were more stable on the bike at low cadence. You mentioned that you can't tell what your body was doing, because the pitot is mounted on the bike, not your body, which is why you had to guess about body yaw. In my experience, there is more body yaw, or movement relative to a center line in plane with the wheels, at lower cadence. That's due to the greater effort necessary to turn the pedals at a lower RPM at a given speed relative to a higher cadence at the same speed. So where the bike may have less yaw at low cadence, I think the body yaw would more than compensate. At higher cadence (same speed, of course), there is less effort required to turn the pedals, and consequently less body yaw is necessary to output that effort. As with everything else cycling, every rider is different, but in general, if you observe spinners vs. mashers, the spinners will produce noticeably less upper body movement than the mashers. I am surprised that bike yaw would not coincide with body yaw, however. I would have expected them to be in sync. I would guess that the majority of the yaw being at the bike or at the body has little effect on the aerodynamics, though, because lower cadence was the clear winner in spite of the yaw or where it takes place (assuming constant wind speed and direction). In summary, I posit that the reduced bike yaw you observed at lower cadence is compensated by greater yaw of the body, while the inverse is true with the greater bike yaw you observed at higher cadence being compensated by less yaw of the body. It would have been interesting to have had video of your body movements from a follow vehicle to compare your body movement at each cadence. Maybe you're an exceptionally smooth rider regardless of your cadence!

Very interesting! I naturally rude at a slow cadence ~70rpm, and have been training myself to ride at higher cadence. Now I want to do this test myself.

Good test! The total movement of the leg is more with faster cadence, it spends the same time moving forward and backward as the slower, so it has to be faster, since cda is on the square of speed, it has to be less aero. As moves faster, probably more wake behind the rider. Question is, a shorter or a longer crank would make much difference? Especially if we dont want to kill our knees. Longer cranks makes the moves faster, so probably an aero penalty there, but is it compensated with the higher torque combined with harder gear?

Maybe Bert Grabsch was onto something back in the day! Are you sure its just aero watts though? At 90rpm you may have more drivetrain friction than at 55 rpm.

I've been doing 80-85 rpm tt's for years now. Biomechanical reasons mostly but also nice to see that it's a bit more aero.

I have a question, when you pedal in strong headwind, the wind also hit your legs. Since the power is measured at the pedals/crank/wheel, does the wind provide extra resistance for your legs that does not get measured by the powermeter? sure feels like it 😅

My cadence naturally drops at higher speeds. It's got something to do with the way I climb which is usually between 70-75rpm. When it gets fast on the flat rides my body naturally assumes a "sustained climb" cadence and it drops from 100-90 to about 80-70rpm.

I would have guessed high cadence is less aero purely because the foot is coming over 12oclock faster thus even though the time is less there is more of them so the net effect is an object moving at higher speed at 12 which is not offset by 6 as drag isn’t linear vs velocity

Still waiting for your follow-up video on crank length. Having spent the past year and a half at 160mm (including multipletypes of road racing), I've identified distinct pluses and minuses for both large and small crank arms.

Given the difficulty of aero testing, it feels like there should be two categories of tests. The first being rigid protocols in near ideal conditions, the second more realistic situation. When I'm out riding, I'm not trying to optimize my position. It's about enjoying the time and being out. For those of us that don't race, the feel and actual energy saving is probably more important. One could argue, for casual riding any wheel will do and it comes down to "do you like the wheels?" Are there sensors that measure "twitchiness" of aero wheels out on the road? My Orbea Orca came with 34mm fulcrum wheels and they feel more twitchy on windy days compared to box rims on my Tarmac. There have been days when gusts of 30mph wind almost made me crash.

Hi PT I have been testing this for a long time, also with varying cadences, and got to the same result. Now in the final slide you mention that there may be a confounding factor from cadence induced vortex shedding. My question is how such a thing would be measured, since even body position cannot be optimal for more intense wind conditions, so both varying yaw and varying wind speed .. I've been knocking my head against the wall trying to figure this out just from my speed curves, since also my kph goes up and down and up and down as well, and this is only exacerbated at higher speed ..

I feel that I move up and down with a very high cadence, perhaps that creates loss of efficiency albeit very little. I find in a group chain ride i push lower cadence high gears to take the front then switch to high cadence low torque as I move to the back. This would mean that I should be maximizing my efficiency. Dirty air from cadence should be less concerning in a sheltered pelaton or bunch post lead out position.

Really good series! Have you ever done any testing like this on helmets?

Can you do error propagation for the wattage difference? Curious to see if there is a statistically significant difference between the two rpms

I need to know which cleat position you recommed?

Fantastic video

This effect is actually potentially pretty hard to measure with a power meter. Notably, the direct aerodynamic forces on your legs are upstream of any usual power measurement, so they cannot be included. The .009 m2 difference you measured is the "linear" drag but the "rotational" drag can be pretty significant too.

I read the title and just started laughing. a burning question for all of us 😂😂

I used to work in UAE 2016 to 2019 and became a little addicted to TT’s as the cycle tracks generally are smooth and pan flat and the wind is predictable (its always there just depends on time of day for direction) so along the way improving my 50k time from 1:40 to 1:14 I found the slower cadence (which I found more natural anyway) high hands helicopter tape on the seat post (it was a Spez Shiv helmets wheels and clothing could drop those times and my ftp went from 240 to 330 - long story I was faster riding slower……….😊

Hey PT, thank you for your great videos! Can you recommend >30mm aluminium rim brake wheels since carbon brake tracks suffer from dissolving under high temps or do you think carbon disc brake wheels are superior?

In addition to the difference in aero resistance to forward velocity, there must also be a significant difference in the aero resistance to the legs' rotational work (separate to forward aero resistance to forward movement that was measured in these tests) ie. spinning fast legs must incur more aero related rotational energy loss than slow spinning legs. I'd guess this spinning aero resistance is a bigger factor than any possible physiological gain from high cadence/low torque

Excellent work 👍

How does a change in cadence change the leg-seattube interaction share? Each event is shorter, that is true, but there are inversely more events per unit of time. And if the speed is kept the same, that then means the same amount of interaction time? There has to be a transient effect at play then or something else entirely.

Might be a dumb question, but the final charts on roll, why does it not centre on 0? Is this just the effect of the cross wind, and so leaning the bike into the wind?

very interesting. That would support buying a big ass front ring for flat rides. Anything more than 88-89 cadence and i feel my form goes out the window. So a large chainring, middle of the cassette at the back, aero tuck, and 42-45kmh flat segments should be pretty doable with a slow-ish cadence. I have 53 currently and i run out of gears pretty quickly when i go above zone 2.

Nobody say that (it's my personal opinion) but having oval chainrings qives noticeable aero advantage (more time spent in more aero legs position). Wold love to see the real world test data.

so might we start seeing the hour record being ridden at lower cadences....?

Would there be a bow wave with the tube that close to the bike?

Very interesting. But... And this is a big but. How do you know what power did you generate? Powermeters are not accurate at all to calculate such little differences. If I were you, I'd try use a motor to power the bicycle, and I'd just pedal freely if at all. Obviously that's not that simple to do, otherwise I'd have tried it. :) Another point is this wind measure device. I find it a bit surprising that it's records such "noisy" (fluctuating) data. I wonder if you had two of these on the same bike, one mounted like yours and the other on the head tube (so movement of the handlebar wouldn't affect its reading) would both of them read the same?

Something in these numbers seems off to me. You're applying 260w on the pedals. Let's say your drivetrain (not perfectly clean) cost you 4% of that, i.e. 10w. And the aero watts @35 km/h are calculated as 200w. This means that the total rolling resistance is about 50w, or 25w per tire. That's seems to be quite a lot for a road tire, at least if we give value to the bicyclerollingresistance data. So either BCR data are completely off (which won't surprise me as their setup have little ressemblance to the real world), or I'm missing something obvious...

Now test how crank length affects aero!

Interesting test, I think there are some additional factors you didn't eliminate or mention though. I think you said you did all high cadence tests and then low cadence tests - this means your weight will be lower in the latter tests due to sweat/perspiration and therefore be slightly faster and have a lower estimated CdA. Since you rode both ways at the same power with a different cadence, this means you used a different gear - depending on your cassette and chainring this will change your drivetrain efficiency from crosschaining, chain angles, and maybe even the amount of dirt/wear/lube on each cog. It would also be great if you could do more of these tests in the future with your TT bike where this kind of change would be more applicable.

Always wanted to see this and totally makes sense. I guess at speed of 45kmh and rpm 95 or 80 the difference would be almost none? (Yet TT guys might look into this)

Not certain, but I believe than any object moving cross-flow creates more drag than that same object when static. Hence, no or low cadence has lower cda than high cadence....in theory as well as in practice. Once saw a video of a vibrating cable in a smoke stream...the vibrations force, or ‘fool’, the air to behave as if the cable is wider than it actually is....analagous to how 20-25 film frames per second ‘fools’ your eye to see the frames as one continuous motion. Hence, the vibrating cable, and the moving leg, cause more turbulence and drag. Calling all real experts of fluid mechanics out there, am I simply wrong??

If there’s a bustle in your hedgerow, don’t be alarmed now.

Remco rides a high cadence, he rocks a lot

How do you distinguish between drive train loss and aero drag? Higher cadence also means more drive train loss

Most aero cadence is to not have the feet at max extention for long?

Summarizing - it doesn't matter:)

When is the peak tour available?

All else equal, the drivetrain should be more efficient when spinning slower, which could be a contributing factor to measured cadence efficiency differences.

Pitot tube on a bike, that's next level! Next, a flow angle vane and integrated heating so you avoid ice build-up. 😅 Hambini could borrow some from a Typhoon or an A380 to look more serious. In any case, looks like a representative test, having a controlled/consistent real-life environment.

Other than an academic exercise, slower cadence maybe more aero isn't faster of course because its less biomechanically efficient. Riders put out more power at higher RPM to overcome small delta in drag increase and why sprints are won in the RPM range of 130's where in theory aerodynamics would be most challenged aka power being the cube of velocity. Obtusely, this is analogous to hip angle on a roadbike or TT bike. Closing the hip angle which subtracts rider power is sometimes preferred due to benefit in aero drag reduction. In the case of cadence however, the opposite is true. Cadence is more noise to aero drag increase relative to major gain in power because Watts = Torque X RPM, Torque = Pedal Force X fixed crank arm length. 'For the same lbs force' pedal pressure, RPM aka cadence is a big deal to power output...proportional to power. RPM = Power for equivalent pedal force.

Waiting for Evenepoel to lower his cadence

What about with a disc wheel? 🤔

Isn't it just (lower cadence) = (lower fan loss)?

If you're spending less on performance car tires that's scary!

dont believe you reach 30kph with 130w mate... that would be quite competetive world wide timetrialing

Haven't finished the video, so maybe you address it, the legs have their own aero drag. Imagine when stationary, the feet are whirring away, they have their own aero drag to simply go in circles. With that, it stands to reason that a higher cadence would have higher aero drag.

Wow, dangling something in the air with lower speed disrupts that said air less - truly groundbreaking!

0

Fake 100% and Fake Kask, living the dream!

Zzzzzzz Whats the answer/?

Interesting test, but totally useless. I don't think any human are able to keep a cadence on 55 rpm through a bike race.