Go High and Get Hot

A quick physics lesson. Feel free to skip this one but with the news today that Fabian Cancellara could attack the Hour Record in August at the Aguascalientes Bicentennial Velodrome in Mexico, here’s a quick look at the effect of altitude and temperature on performance.

Within hours of the story appearing in La Gazzetta Dello Sport Cancellara’s Trek team denied the plans saying the focus is on the classics. But if the record attempt is still uncertain the physics is fixed: riding at altitude in Mexico could be just what is needed to break the hour record. It’s not just for elite records either, air temperature can have a big effect on daily riding and Strava conquests alike.

First the basics. As you’ll know aerodynamics are crucial to cycling performance. You can improve your position on the bike, you can improve your clothing and you can improve the bike itself. All three are interrelated. For example change your position on the bike and the airflow is altered and consequently you might want to change, say, your helmet or at least its angle on your head. But once you’ve nailed the position and made your bike as slippery as possible as well as working on other variables like tire choice and even the bearings then it’s time to think about travel and temperature.

Why? Because if you’ve done all you can to improve your fitness and power and you’ve got the best position possible then there are still parts of the equation left to play with, notably air density.

Air density decreases with altitude but so do other factors which affect density like humidity and temperature. But for the sake of argument, at sea level the standard air density is 1.225kg/m3. Go to 1,870 metres above sea level, the height of the Aguascalientes velodrome and everything else being equal this number drops to 0.975kg/m3. This helps bring about significant reduction in the air resistance.

Air density reduces with temperature but normally the higher you go, the colder it gets. But Aguascalientes enjoys a tropical latitude and better still the Bicentennial Velodrome is a modern covered track. This means the temperature can be adjusted: turn the heating up. The standard air density is based on a set temperature but turn up the heating to a summery 28°C (82°F) and at 1,870m the air density drops further to 0.932kg/m3.

The relative humidity is a another factory. If you’ve not thought of it before you might imagine that humid air is thicker, it can feel uncomfortable. In English they say it’s sticky, the French say “il fait lourd“, it’s heavy weather. But actually humid air is lighter because water as a gas is two hydrogen atoms and one oxygen compared to the heavier combo of paired nitrogen and oxygen atoms so the more H2O per m3 the less other heavier molecules.

Go to Aguascalientes Velodrome and install a humidifier to crank the air up to a steamy 90% relative humidity and the air density drops to 0.917kg/m3.

As you can see altitude is a prime determinant, then temperature followed by humidity. If you want more on the maths, see here.

Aero gains vs Aerobic performance losses
So far so good, just to to altitude, find a covered velodrome and turn up the heating, right? Only altitude brings other other problems. The very gains in aerodynamics from the reduced air density mean a reduction in the partial pressure of oxygen and consequently your VO2 Max or aerobic performance declines. There’s a clear trade off here. Scanning the web studies suggesting a performance drop-off at 1,870m of between 5-10%. So for all the gains in aerodynamics there are reductions in physiological performance.

Normal level
All these calculations don’t just apply to elite athletes and their world record attempts. Note back at sea level the difference in air density between a summer’s day and a winter’s day is about 10%. This alone can account for that slow feeling in winter, especially when dressed in bulky gear. In competition it’s hard to get a benefit given others share the same conditions but worth noting if you like to compare times or speed on a training loop and you need to remember the temperature and weather when comparing times. You can use the science for Strava high scores, just venture out on warm humid day for a natural advantage.

Wind adjustment
A quick mention of the wind. Last week’s piece on the aerodynamics of skiing brought an email from a reader asking whether time trial times should be adjusted for varying wind conditions. Watching the Winter Olympics you might see the ski jump and how wind conditions are included in the calculation of the total points. If you saw the recent Tour of Dubai you might remember Tony Martin’s frustration at being the last ride to start and reports say the wind had got up, slowing his ride. By contrast BMC Racing and Taylor Phinney opted for an early start, a factor behind his win.

For me no adjustment should be made. For starters the calculation would be a theoretical nightmare and a practical impossibility to measure windspeed around the course. In the ski jump alone they take five measurements, a 10km TT course might need five or maybe 50 measurements. But it’s a philosophical difference, cycling embraces the natural environment and the wind is to be exploited. Tony Martin might resent being seeded first in a time trial but it’s the burden of being a world champion.

Note the weather all year long will affect performance. Wet roads in a time trial can be slippery but the increased humidity could make for a faster ride… so long as the downpour doesn’t bring down the air temperature too much. Amateur meteorology can help picking winners for a time trial.

Will he, won’t he? It makes sense for Cancellara to keep the plans confidential for now because even injury this year could mean he “loses” in his bid for the record but there are many variables to address, notably the UCI’s review of the rules that could let him use a modern track bike, the kind allowed in a pursuit race.

If you did want to go for the record then the Aguascalientes velodrome, built in 2009, is a good pick offering reduced air density and the controlled conditions of an indoor arena but not being so high to diminish aerobic performance too much. Just ask François Pervis who smashed the kilometre record by two seconds there in December.

I didn’t wan to turn this into an algebra attack, Google air density and you’ll plenty from the raw physics to aviation data and even some cycling-specific pages; plenty for a good school project. As much as the romantics might like to escape it, cycling is underpinned by physics and air density is a big factor so bear this in mind whether it’s the hour record or that feeling of speed on a warmer spring day.

42 thoughts on “Go High and Get Hot”

  1. I have my doubts about aerodynamics testing in a wind tunnel, due to it
    being a controlled indoor environment, whereas real-world conditions are
    not controlled but constantly variable, just as you mention above.

    I don’t know how much of those variables are replicated in the wind tunnel, so cannot
    be sure whether to trust the results!

    • Don’t say that outloud around anyone with $2000+ aero wheels or an aero road helmet. They’ll shout all kinds of “scientific facts” backing up why their expensive “wind tunnel-proven” setups are far superior. Trust me, I’ve done it.

      But I’m with you on this one, and personally the Martin/Phinney example backs us up. If you test in a controlled environment to validate your claims, what happens when you can’t control the environment?

      In a semi-related note, I saw a recent article (*credit escapes me) that mentioned the last guy to break the record, Sosenka, did it on a 333 track, not a 250. Overlooking his epo history, a good point was made that the longer track means fewer turns. So for all the things that can be controlled, i wonder how important that factor would be as well? Would it be worth 10 meters? 100? Enough to set a new record?

      • I don’t know the calculations or estimations but the shorter the track, the harder the effort. Increased “g-force” on every bend means a strain on the arms and back but crucially the blood flow too, it’s a mild centrifuge where your blood is being forced out into the legs and the return to the heart is that bit harder.

        Sosenka also used heavy wheels so the longer track helped him.

      • Darren and Stephen, you may have your doubts about tunnels testing, and that’s fair enough, however there has been enough peer reviewed published science to demonstrate that wind tunnel derived aerodynamics results do translate very well to the data measured in “the real world”. e.g.:

        Of course some tunnels are better than others, and tunnel design and test process matters, but for a well set up tunnel, the results most definitely reflect real world physics, which should not surprise since the physics is the same.

        An issue with tunnel testing is whether a position set up in a tunnel is sustainable when riding for longer durations and under the sort of effort level when racing. That’s where field testing of aerodynamics has some specific advantages.

        Each method has great advantages, and is now sufficiently sophisticated to tease out quite small variations and enable performance optimisation.

        As for outdoor variability, well fortunately the technology now exists to measure with high frequency and with precision the speed, power and wind velocity (speed and direction) while on a bike, and to closely examine the actual impact to performance under variable conditions.

        Guess what we find? Yep, more aero in a tunnel (and more aero in field testing) = more aero in real world variable conditions but also that aerodynamics is impacted by the effective wind “yaw” angle, hence why some wheels for instance are better in cross winds than others which might be superior in low wind conditions.

        As to the impact of track distance, there are swings and roundabouts to consider.

        The G-forces are there but are not horrible at hour record speeds, but that little press 300-400 times per hour has some impact. So that’s some swings.

        An example of roundabouts: for instance when using a longer track, the centre of mass of the rider has to travel further on longer tracks (the lean angle in the turns is less) than on a shorter track (greater lean angle).

        It may not seem a lot, but it does add up. Keep in mind that on most UCI accredited tracks, the turns represent ~2/3rds of the total lap distance.

        In track events, it’s the wheels that must traverse the 250m/333.33m line, but the COM travels inside that line for much of the journey and covers less distance than the wheels do. So if the lean angle is reduced (i.e. as on a longer track) then the COM has to travel further (and hence faster) for the same lap distance and time travelled by the wheels. Aero drag is more a function of the rider’s COM speed than it is of the wheel speed, so increasing the difference between the two (i.e. with a shorter track) has an advantage.

        It’s partly why faster track riders, who need to lean slightly more due to their higher speed, get an additional bonus.

        There are other factors such as the greater natural variability in wheel speed and cadence between straights and turns on shorter track versus longer ones, so a rider needs to train to manage those transitions.

        And riding good lines matters. e.g. if you ride just 10cm above the black line, in an hour record you’ll dud yourself ~120m compared with the official distance (which is simply calculated based on number of laps x lap distance, plus a pro rata calculation for the final incomplete lap).

        The subtleties are endless.

    • The entire Formula 1 grid disagrees with you… obviously there are differences from wind tunnel to real world, but the technology is certainly developed enough to correctly calculate and adjust statistically for these differences.

      • F1 is a very different beast – velocity alone makes the aerodynamics so much more important. But to your point, F1 tests the whole system. The car is very easy to define and is the sum total of the elements involved. Testing two different diffusers (on the car) gives clear differences and allows off-car correlation of other diffuser results to occur (because the rest of the car is aerodynamically constant, unlike a cyclist).
        Testing an aero helmet and saying “it’s 6% faster” is nonsense by itself: you need to also explain that, of total drag, that is 6% of about 2%. Same for wheels: a 5% slipperier wheel is only 5% of 4% of the total drag (so, 0.2%, or not really very much).
        The single biggest factor is the person on the bike – wheels, frame, and helmet (in fact, all non-person aspects) make up maybe 8% (note: this is a wild-ar$e guess because I don’t have time to look it up right now), so the benefits are virtually all in the realm of ‘psychological’…

        • While I agree that the purported benefits are not as great as they are laid out (aero benefits are not additive!), they are a long way from psychological. It is a matter of degrees though. If I ride at 20kph, they are there but not significant. If I am riding at 50kph they are there and significant. For the average Joe, a gain of 1 second in a 40km TT is not a big deal … for Tony Martin it is the difference between rainbow stripes and silver … no matter how you slice that, it is significant.

          Alec makes a great point about sustainability. I have seen hundreds of riders in long tail aero helmets with their heads turtled and the tail sticking up in the air. Thus, they are taking a piece of equipment that could save them a dozen seconds or more in a TT and converting it to one that will cost them the same additional time. It is only aero if you can maintain the position. Otherwise, it is bling …

  2. The title of this piece – ‘hot and high’ – echoes the exact words of caution taught to all student pilots. The take-off run from an airfield high in the mountains in a sweltering humid summer afternoon is very much longer than what’s needed at sea level and 15ºC (standard conditions). The lower air density degrades both lift and engine power, unlike in cycling, where it works mostly in favour of the record breaker (at least for altitudes below the onset of hypoxia at ca. 5000 feet).

  3. He should have to do it on an outdoor track like Merckx, Anquetil and co. The UCI stripped back all the crazy positions and bike tech. Why not go all the way.

    I realise outdoor tracks are diminishing due to indoor tracks popping up everywhere.

  4. You didn’t complete the picture to see why a 10% difference would matter so much. Drag is proportional to the square of the speed. The greatest source of the stuff slowing a competitive cyclist down in a velodrome is aerodynamic resistance. (no turtle racing)

    The faster you go, getting the next whole Km/h is many times more difficult. For example, 25Kmh to 26Kmh requires much more power than 10Kmh to 11Kmh.

    Taking 10% of the drag away as well as heating the air definitely increases the distance traveled over an hour.

    • To clarify:

      Drag force is proportional to the square of velocity but is linearly proportional to the coefficient of drag. Aero equipment and positioning reduce the coefficient of drag, so 10% better aerodynamics result in 10% less drag force independent of speed.

      What is true is that as your speed increases, the drag force increases as a percentage of the total force resisting your movement. Total force includes things like mechanical resistance, rolling resistance, etc. in addition to aerodynamic resistance

      So, at lower speed a 10% reduction in coefficient of drag has a lower effect on total resisting forces than at higher speed but still results in only a 10% decrease in drag force.

    • Aero Force = 1/2 * density * CdA * velocity^2
      Aero Power = 1/2 * density * CdA * velocity^3

      At a given speed, multiplying density by 0.9 (a 10% reduction) multiplies power by 0.9 (a 10% reduction).

      At a given power, multiplying density by 0.9 (a 10% reduction) multiplies velocity by 0.9^1/3 (a 3.5% reduction). 3.5% of 50km is an additional 1.7km which is still very significant.

  5. Air density is primarily affected by 4 things:
    – temperature
    – altitude
    – humidity
    – barometric pressure

    The latter has a far more significant impact on air density than does humidity.

    e.g. air density drops only 0.9% with an increase in relative humidity from 0% to 100%.

    but change barometric pressure from say 1020HpA (pretty typical high pressure day) to 980HpA (a storm cell) and air density drops by over 4%.

    for comparison (ceteris paribus):
    an increase in temperature from 5C to 25C reduces air density by ~7%.
    an increase in altitude from 50m to 1000m reduces air density by ~11%.

    Just to give people an idea of impact to speed for same power (say 420W) *at same altitude* (50m) and same aerodynamics (CdA~ 0.28m^2) and rolling resistance (0.0025 – good track):

    Scenario 1:
    Temp 10C
    Air press: 1020HpA
    Rel Humidity: 20%

    Scenario 2:
    Temp 28C
    Air press: 990HpA
    Rel Humidity: 80%

    Sceanrio 2 would see a speed increase of ~3.4% over scenario 1.

    Add 1000 metres in altitude to scenario 2 and for same power you’ll get and increase in speed over scenario 1 of ~7.3%

    To attain the same speed as scenario 1 would only require 82% of the power.

    There is most definitely a trade off between aerodynamics gains and physiological losses due to altitude. Such that there is likely to be a sweet spot of altitudes for events of different durations, especially if the track has some temperature controls.

    I note that for Mexico world cup recently, most of the recent new records set were sprint events, not endurance events, with the exception of the new distance pursuit events for women.

    • Good point as the calcs were implying a static air pressure that only chaned with altitude rather than weather. With this the rider could pick the day or even the hour if a front was moving in or out. It all helps.

      • Well I’d need to go back and check the rules, but when my clients decided to attempt UCI age group hours records (I’ve coached 3 of the current age category hour records), they had to nominate the day and time of the attempt to the UCI at least 2 months in advance. In fact one had to give 3 months notice. But then others have managed attempts with less notice so it’s probably one of those not well documented or followed processes/rules.

        Obree famously made his successful attempt the day after his failed effort, so not sure if the rules regarding logistics on this have changed since, or indeed if this was just bureaucracy requirement rather than actual rules.

        Nevertheless, the relevant doping agency needs to be informed in advance as well so that they can, if they desire, perform OOC testing (for a pro this is already business as usual, so more applies to amateurs outside of whereabouts processes). There is mandatory testing on the day as well, which the athlete must pay for, as well as pay for the UCI officials, timing systems, track hire etc. Presumably a velodrome with Cancellara riding would be able to make some money though through ticket sales.

        With my clients, we looked at balancing track access with a time of year that was more likely to provide favourable environmental conditions, but you cannot guarantee a great day. Rainy seasons can also have high pressure sunny days too.

        In the case of my clients, one was in a temperature controlled track which reduces such concerns somewhat, the other two were not, and we had to balance choosing a time of year that was warm enough to be more likely to provide lower air density (and a summer storm cell is perfect), versus it being so hot that the heat became detrimental to performance. This is a track which can get well above 35C inside on bad days (I’ve raced it myself at 39C inside). That sort of heat is OK for pursuits (just ask Jack Bobridge) but not so great for long efforts over an hour.

        So I work on the basis of preparing the athlete to set a record on a not so good day, and if conditions are cooperative, well we gleefully accept the extra metres.

        For an hour record attempt, if a temp increase of 5C doesn’t affect a rider’s power, that equates to an extra ~290 metres.

        Boardman beat Merck’s record by ~ 10 metres IIRC. Equated to air density change alone (again assumption of ceteris paribus), if it had been just ~0.2C cooler, he may not have broken it.

    • From memory the day Bobridge broke the world IP at Dunc Gray was a perfect storm- stinking hot, very humid and I believe there a low pressure system moving over Sydney as well. Goes part way to showing how the aerodynamics of the rider themselves – Bobridge’s position vs the superman position- isn’t the only factor. The cocktail of conditions you’ve (and INRNG has) described has a large bearing on speed.

  6. Not sure it’s been mentioned, but it needs to pointed out that one needs to take into consideration that riding “outdoors” in Mexico City has smog/air quality working against one. As compared too an indoor attempt were presumably air quality is better!

    • Mexico City itself is very bad, but Aguacalientes lies about 500km to the northwest and the local government claims it has good air quality. It is plausible given the size of the town, its surroundings and the predominant wind directions.

  7. Given what you (valuably) assert, and reading the above comments, plus tallowing tfor he UCIs current restrictions on permissible equipment; shouldn’t there be some very restrictive standards set, e.g., venue, height above sea level, time-frame/season, in order to make any future hour attepts competive and/or meaningful by allowing for all the climactic conditions referred to?

    Such a regime might reinforce and add lasting value and legitimacy to the concept of going for the 1-hour World Record?

  8. RE: the wind adjustment. For an outdoor time trial race indeed it is not very practical. But for an hour record in a velodrome it should be pretty easy to correct the time for air density, no? I don’t think that’s a bad idea.

    • If you adjust time, then it’s no longer an hour record 😉

      Fact is, environmental conditions impact records in many sports. The intrigue of the unknown randomness is part of the fun. Some sports do set some limits on conditions for a record to be considered valid (e.g. running sprint records, long jump etc have max tailwind limits).

      There is wind in an indoor velodrome too, the rider creates a gradual swirl, like a toothpick in a bathtub. The smaller the interior of the track, the more swirl they can generate and benefit from. I have measured this effect at one indoor track and it’s worth an equivalent at hour record velocities of a little over 5W benefit, or adds approximately another 250 metres over the hour at same power once the swirl is in full effect (it takes about 10-minutes or so). Some tracks will have indoor environments that help the swirl, others will hinder it.

      Then you have tracks where doors, dock entries and windows open and close creating some issues, both with air movement and with localised air density changes.

  9. ‘A quick physics lesson. Feel free to skip this one ……’
    I nearly did skip this, but this is one fascinating thread… never considered the toothpick in a bath-tub effect! could this be the beginning of a whole ‘wind doping’ scandal with a few strategically placed electric fans…

    • I’d never thought of that either. Perhaps the UCI should mandate that the record must be set at sea level, at 20 deg C on an indoor figure-8 track…

  10. I set my 10mile TT PB on a wet windy day – but I reckon the fact that it was p*ssing down and that I wanted to finish and get back to the village hall as fast as possible was as much responsible for my speed as the reduced density of the humid air…

    • I’m with you Andy. I ride noticeably better on wet and/or rainy days. By noticeably, I mean by a considerable margin – in the area of five to ten seconds per kilometre with lesser levels of fatigue.
      Interesting then that I should read it’s more favourable on hot days – I guess perception and personal preference counts for more than science in my respect!

Comments are closed.