One of my recent blog posts (Do aerodynamics matter off-road?) looked at some typical off-road scenarios - a gravel bike rider and a mountain bike rider travelling across flat ground - to see whether aerodynamic improvements are worth making in those slower off-road scenarios. The traditional thinking is that aerodynamic improvements are only worth making at higher road bike speeds, but my study debunked that.
The surprising result was that, not only do aerodynamics matter in those off-road situations, but the time savings, in minutes and seconds, over a fixed distance are actually greater when travelling off-road at those lower speeds and lower powers. That was unexpected, so I immediately wanted to investigate that further. This blog post describes a more rigorous analysis I've done to understand those speed sensitivities better, in terms of the performance benefits brought by an aerodynamic improvement. It will show how much benefit is obtained across the entire speed range.
I'll explain how I've done the analysis and the assumptions I've made. I'll also try to explain each step of the calculations so that readers can understand what the plots are showing and why they're the shapes they are.
If you want the quick answer to the question "do aerodynamics matter at low speeds?", then feel free to skip to the summary at the bottom. I will say though, just to whet your appetites, it depends on your exact question. For example:
- Do you want to know how many Watts you save at a fixed speed from an aerodynamic improvement? Then higher speeds will give you much larger Watt savings.
- Do you want to know how much time, in minutes and seconds, you save from an aerodynamic improvement over a fixed distance? Then the answer is very different, and the largest time savings occur at low to intermediate speeds instead. It also depends on your power.
This is explained fully at the end, but first the method...
Bike performance calculation method
My calculation method uses the same method I've used previously, for example in this post from 2016, whereby I calculate the four forces acting on the bike:Power * Eta = (Fa + Fg + Fr + Fi)*V
where:
Fa = Aerodynamic drag force (N)Eta = Drivetrain efficiency (e.g. 0.97 means 97% efficient)
Fg = Gravitational component force (N)
Fr = Rolling resistance force (N)
Fi = Inertia force (N)
V = Bike speed (m/s)
and:
CdA = Drag area of the bike+rider (m^2)
Crr = Coefficient of rolling resistance
rho = Air density (kg/m^3)
m = Mass of bike+rider (kg)
t = Time (s)
rho = Air density (kg/m^3)
m = Mass of bike+rider (kg)
t = Time (s)
theta = Gradient (in degrees)
g = Acceleration due to gravity (=9.81m/s^2)
g = Acceleration due to gravity (=9.81m/s^2)
This is essentially the same as the equation published by Jim Martin et al (see his 1998 paper here). I've made a few simplifications, such as assuming zero wind, so that airspeed is always equal to ground speed. Also, I'm modelling steady state conditions (no acceleration), so the inertia force term (Fi) is zero.
The equation above allows the required rider power to be calculated for a given speed and for a given set of assumptions for the other parameters, like rider weight and air density. Alternatively, instead of calculating power for a given speed, the power can be defined/fixed, then an iterative calculation can be performed to determine the speed that would be attained for that power.
Either way, I prefer to use Microsoft Excel to do these calculations. Excel's Goal Seek function works well to iteratively calculate speed for a specified power. In a similar way, the equation can be re-arranged to calculate other variables, such as gradient, CRR etc, if both speed and power are fixed.
Assumptions for the analysis
I've made a few assumptions for the majority of the calculations I've performed:
- Rider + bike weight = 85 kg
- Drivetrain efficiency = 97%
- Air density = 1.203 kg/m3 (sea level pressure and 20 degrees Celsius)
- Baseline CdA = 0.32 m2 (a typical drag area for a medium sized rider in the drops)
- Baseline CRR = 0.004 (a typical rolling resistance for a good tyre on a good road)
- No wind
- No accelerating or decelerating
- Power outputs considered = 150, 300 & 450 Watts
Clearly different riders will have values that are different from those, depending on the size of the rider and where they ride. Although I've used the values above for most of the analyses described in this blog post, I've also done some checks to see whether the conclusions are affected by a different set of assumptions. The results of those checks will be shown later.
The overall objective of this analysis is, of course, to understand the performance benefit you'd get from aerodynamic improvements.
With that in mind, I have assumed a CdA reduction of 0.012 metres squared, from the baseline CdA of 0.320 down to 0.308. This 0.012 reduction of the drag area is a 3.75% reduction. It represents the kind of aerodynamic benefit that you'd achieve by swapping from a non-aero helmet to an aerodynamic road helmet, like the Specialized Evade. In fact, I calculated this 0.012 change in CdA from this video posted by Specialized, by reverse-engineering their quoted 40 km time trial time saving of 42 seconds, when changing from their Prevail to the Evade.
Baseline Case: Speeds on the flat
Let's first consider the baseline case using the values provided above.
- For a 150 Watt output, the speed will be 30.53 kph on a flat road.
- For a 300 Watt output, the speed will be 39.52 kph on a flat road.
- For a 450 Watt output, the speed will be 45.73 kph on a flat road.
The power losses for these three cases can be seen in the spreadsheet above. The aerodynamic losses account for between 78% and 88% of the total power for these cases.
The next step is to model a speed variation. There are a few different ways to achieve a speed variation for those fixed powers, but the two methods I have focused on are using gradient and rolling resistance coefficient to achieve that wide variation in speed we'll consider. I'll show the results for gradient variation first:
Speed variation through gradient changes
Clearly, for a fixed power output, the speed will reduce if travelling on an uphill gradient, and the speed will increase if travelling downhill.
The plot to the left shows what gradient is required to achieve a wide range of speeds, between 4 kph and 80 kph for 300 Watts. The plot shows both gradient (in blue) in degrees, and also on the secondary axis the slope is shown as a %. Many cyclists will be more familiar with % slope to indicate a hill's incline/grade.
It can be seen that at 300 Watts, for a flat road (i.e. zero gradient/slope) the speed will be 39.52 kph, as noted above for the baseline case. That 39.52 kph point is indicated on the plot with a hollow diamond, where the curves pass through the x-axis (zero gradient/slope). To achieve the fastest speed of 80 kph at 300 W, a downhill slope of 10.3% is required. At the other extreme, for a 4 kph speed, a 32.6% uphill is required for 300 W. Speeds even slower than 4 kph are theoretically possible, but only at extremely high gradients where wheel spin would likely occur, therefore those unrealistically slow speeds are shown using dashed lines.
For other power outputs, 150 W and 450 W, a similar curve can be obtained, although the gradients for any given speed will be different to the 300 W curve, as shown below:
This speed variation through gradient changes, where the coefficient of rolling resistance (CRR) is fixed at a value of 0.004, represents a road cycling situation.
Speed variation through rolling resistance changes
The alternative method I used to achieve the speed variation for a fixed power output was to vary the rolling resistance coefficient (CRR) instead. The plot below shows the CRR values that are required to achieve the speed variation on flat ground. This represents the speed reductions you'd get if riding off-road, for example on gravel, or rougher mountain bike trails. The very low speeds where the CRR values are very high would represent deep mud or soft sand.
The hollow symbols in the plot above are for the baseline case where the CRR is 0.004, representing the CRR for a good road tyre on a good tarmac road. Fast speeds in the 50-80 kph range are not possible on a zero gradient surface, because negative CRR values are impossible. Even the smoothest velodrome with the fastest tyres would have small positive values for CRR.
So those two plots above show how slow speeds are attained, either through a gradient variation or via CRR variations. The following sections will show the effects of the -0.012 CdA aerodynamic improvement, starting with the method of speed variation via gradient changes.
Gradient changes: Power Savings at a fixed speed
There are two different ways that the benefits from an aerodynamic improvement can be quantified:
1) The power saving at a fixed speed.
2) The speed improvement at a fixed power.
Let's look at #1 first, the power saving at a fixed speed. The plot below shows the power saving in watts across the speed range.
Notice how, for a given speed, the saving in Watts is identical no matter what the rider power is. Remember that for any particular speed, the 450 W rider is riding up a steeper hill than the 150 W rider. However, for any given speed, the rider's power won't affect how many Watts are saved by the aero improvement. As far as aerodynamics is concerned, it's only the speed through the air that affects how many Watts of power are saved. Notice though that for the zero gradient road (the hollow symbols) the more powerful riders are travelling faster and are therefore saving more Watts of power than the 150 W rider.
The plot above shows that power savings, in Watts, rise rapidly with speed. In fact, aerodynamic power savings such as this increase with speed cubed. That means that if the speed doubles, the Watt saving is eight times larger (i.e. 2 to the power of 3). For example, the -0.012 CdA saves about 10 Watts at 40 kph. At double that speed, 80 kph, the power saving is now 80 Watts, eight times what it was at 40 kph. A similar factor of eight change occurs going from 10 to 20 kph, or 20 to 40 kph.
The watt saving at any particular speed due to the change in CdA can be calculated quite easily using the following equation on the left. The values for 80 kph (22.22 m/s) are shown as an example, which results in a power saving of 81.7 Watts for the 0.012 CdA improvement.
As a quick aside, this is why marketing departments tend to quote the number of Watts saved at high-ish speeds, typically 40, 45 or 50 kph, when trying to sell us bikes and bike equipment. That's because if they quoted the Watt saving at 30 kph, or another more reasonable speed for an amateur cyclist, the number of Watts saved would be just 30% of the Watt saving at 45 kph, and wouldn't sound nearly as impressive. Since these Watt savings are so speed-dependent, people should be very cautious when bike equipment manufacturers quote Watt savings due to aerodynamic improvements but without also saying the speed at which they are attained.
The plot above shows that the Watt savings increase significantly with speed. That might lead people to conclude, understandably so, that only people riding at fast speeds will benefit from aerodynamic improvements. However that conclusion is not correct. Read on further to understand why...
Firstly, let's take a quick look at a plot of the same power savings but as a percentage of the power before the aerodynamic improvement:
As a percentage of the original power, the 150 W rider would save a larger percentage of their power from the aerodynamic improvement at a given speed. Taking the 40 kph speed as an example, all three riders save 10.2 Watts. However, that 10.2 Watt saving is obviously a larger percentage of 150 W than it is for 300 W & 450 W, which is why the red line is above the green and blue lines in this percentage plot.
For the zero gradient road (the hollow symbols), notice how the speeds are different for the 150, 300 & 450 Watt riders, but also notice how the percentage power savings are approximately the same for all of those three hollow symbols.
Gradient changes: Kph improvements at a fixed power
Let's now look at the second option; the speed improvements at a fixed power that are achieved by an aerodynamic improvement. There are several ways to show this.
Firstly, the plot below shows the speed improvement in kph for the three different powers.
It can be seen that the speed improvements above, in kph, increase with speed in a manner that's somewhat linear. Interestingly, it's not completely linear, because speed improvements below 20 kph are fairly small, then above 20 kph the curves are not perfectly linear either. The curves are also not cubic, like the Watt saving plot in the previous section, but clearly higher speeds will result in larger speed increases, in kph, due to an aerodynamic improvement.
As a final observation, the lower rider powers result in slightly larger speed increases for a given speed. Note that the x-axis is the speed before the aerodynamic improvement, not the speed after the aero improvements.
Gradient changes: % time (or % speed) improvements at a fixed power
An alternative to the plot above is to show the speed improvement as a percentage, instead of plotting the kph improvements. Note that the speed improvement, as a percentage value, is also exactly the same as the % time improvement over a fixed distance.
This % time improvement has a different shape again to the preceding plots. The % time (and % speed) improvements yet again increase at higher speeds, but the curves level off at the higher speeds; those higher speeds that are attained only on downhill gradients. As for the kph improvement plot, the larger % time/speed improvements are achieved by the lower power (150W) rider.
Note also that the three hollow symbols above, representing the % time saving when travelling on a flat road, are almost level with each other, indicating that similar time and speed % improvements are made on a flat road, regardless of a rider's power output. This is consistent with the 2nd bar chart in my previous blog post (which is reproduced on the left, for completeness).
Gradient changes: Seconds saved at a fixed power
A final way to show the performance improvement at a fixed power is to plot the time improvement, but in seconds saved over a fixed distance, instead of as a percentage. A distance of 40 km has been chosen, but the distance is unimportant; the shape of the curves would be exactly same for any distance and only the values on the y-axis would change if a different distance was used.
Yet again, the curves in this plot have a different shape to all the other plots. We see some interesting things from this plot:
- Slower speeds result in smaller number of seconds saved, with a gradual increase in the savings up to moderate speeds of 30-45 kph.
- However, at even higher speeds than those, 40-50 kph upwards, higher speeds do not result in larger time savings over a fixed distance. Instead the largest improvements are made when riding at moderate speeds, 30-45 kph. The speed at which the largest time improvements are made depends on the power: About 30 kph for the 150 W rider, ~40 kph for the 300 W rider and ~45 kph for the 450 W rider. This is incidentally about the speed at which those riders would ride at on a flat road (i.e. the hollow symbols).
- The largest improvements in seconds are made by the weaker 150 W rider, not by the strongest rider. The best improvement made by the 450 W rider is 38 seconds, when riding at around 45 kph. The 150 W rider would save 50 seconds when riding at that same speed. At a slower speed of 30 kph, the 450 W rider would save only 31 seconds, whereas the 150 W rider saves almost double that, 55 seconds, at the same speed of 30 kph. This result is certainly counter-intuitive.
- Another way to look at the results is the speed at which the same 38 second time saving is made. The 450 W rider achieves their maximum 38 second saving at 45 kph, due to the aerodynamic improvement. The 150 W rider achieves that same 38 second saving at a significantly slower 18 kph. Who would have guessed that?
Why does that happen? The quick answer is that the 450 W rider covers a flat 40 km road distance in just 53 minutes, whereas the slower 150 W rider covers the same flat 40 km distance in 1 hour 20 minutes. A similar percentage time improvement for both, due to the aerodynamic improvement would result in a larger number of seconds saved by the slower rider, because it's the same % saving but for a longer time.
To recap, this result is very different to the very first Watt-saving plot which showed that higher speeds give dramatically higher Watt savings. Here it's moderate speeds that give the highest time savings, in seconds. Furthermore, even slow speeds of 10-20 kph, result in a significant number of seconds saved over fixed distance, even more so for weaker riders.
The takeaway from this plot, that aerodynamic improvements are important at slow and moderate speeds, especially for lower power riders, is significantly different to the Watt saving plot, which showed that the Watt savings grow substantially at higher speeds.
Speed variations achieved via rolling resistance changes
All of the previous plots showed the performance improvements as a function of speed, where those speed variations were achieved through gradient changes. What would those plots look like if the speed variation was achieved through rolling resistance changes instead?
The good news (because this blog post is already rather long) is that it doesn't change the results at all. For completeness, I have shown below the same plots, but where the speed is achieved through a rolling resistance variation instead.
You'll notice that all the curves at higher speeds on the right hand side of the plot are all shown as dashed lines, denoting that these are unrealistic cases. That's because, unlike for the gradient cases, where it's possible to go downhill to go faster, it's impossible to have negative rolling resistances. The fastest possible speeds are achieved on a smooth flat road (or a velodrome) with very low rolling resistance values.
Nevertheless, the shapes of all these curves and the conclusions are exactly the same as for the earlier plots, where the speed variation was achieved through gradient changes. I must admit I wasn't expecting this, but it's a convenient result which simplifies the conclusions.
Sensitivity of the results to the baseline assumptions
Towards the start of this blog post I explained the assumptions I'd used for the baseline case, which were:
- Rider + bike weight = 85 kg
- Drivetrain efficiency = 97%
- Air density = 1.203 kg/m3 (sea level pressure and 20 degrees Celsius)
- Baseline CdA = 0.32 (a typical drag area for a medium sized rider on the drops)
- Baseline CRR = 0.004 (a typical rolling resistance for a good tyre on a good road)
- No wind
- No accelerating or decelerating
- Power outputs considered = 150, 300 & 450 Watts
As I mentioned earlier in this post, I wanted to check how all of the results above would change if a few of those assumptions were different. I did that by adjusting a few of those assumptions and checking the effect of the changes:
A higher altitude: Air density = 0.828 kg/m3, i.e. 10,000 ft & 20 degC (-38% lower density)
A bigger or more draggy rider: Baseline CdA = 0.40 (+25% vs CdA=0.32)
In all cases, I used rolling resistance as the means to achieve different speeds for a given power.
Heavier rider:
This is an easy one, because the time savings (in seconds) are exactly the same if the rider or bike weight is heavier. There is a small change to the position of the hollow symbol, which denotes the speed achieved for the road case (CRR=0.004), which reduces by 0.5 kph. That's because a heavier rider induces higher rolling resistance losses for a given CRR value. Apart from that, the time savings due to the -0.012 CdA improvement are identical for the heavier rider.
Higher altitude:
Here I've considered a 10,000 ft altitude instead of sea level. This will indicate the sensitivity of the aerodynamic improvement if racing at a higher altitude, such as the MTB World Cup in Leogang, or the Leadville 100 MTB race. I've kept the temperature the same at 20 degrees Celsius, but nevertheless the pressure reduction reduces the air density by 38%, resulting in a lower drag force for a given speed and CdA.
As shown below, the effect of this lower density is that the time savings (orange curve) due to the aerodynamic improvements are slightly smaller, for example 31.4 seconds at 30 kph at 10,000 ft, instead of 39.5 seconds at sea level for a 300 W rider. Notice also how the speed for the flat road with a CRR of 0.004 (hollow symbols) has become faster at 10,000 ft, due to the low air density. This is why hour record attempts are often done at high altitude velodromes.
The effect of altitude and the lower air density is likely to be similar to the effect of the reduced time saving seen for higher power riders, shown earlier in this blog post, whereby for a given speed, higher power riders are achieving that speed at a higher CRR value (or less beneficial gradient). The consequence of those harder riding conditions (for a given speed) is that any aerodynamic improvement yields less % speed benefit and lower time savings over a fixed distance.
A more draggy rider:
Here I've considered a CdA of 0.40 metres squared instead of 0.32. A value of 0.40 would be indicative of a bigger road rider in an upright position, or with baggier clothing, or alternatively it's typical for a cross-country mountain biker.
Here I've considered a CdA of 0.40 metres squared instead of 0.32. A value of 0.40 would be indicative of a bigger road rider in an upright position, or with baggier clothing, or alternatively it's typical for a cross-country mountain biker.
The plot below shows two curves for this CdA=0.40 case. The orange curve shows the aerodynamic benefits for the 0.012 change in CdA, which is the benefit of the Prevail->Evade helmet change discussed earlier. Notice however, that because the starting CdA is higher, at 0.40, the aerodynamic improvement is a smaller percentage benefit, 3% instead of 3.75%. This is the reason why the time savings for the orange curve are smaller than for the baseline case (the green curve).
The purple curve shows what time saving would be achieved if the CdA reduction percentage is kept the same at 3.75%. That means the CdA reduction is larger, 0.015, instead of 0.012. For that case, for the 3.75% CdA reduction, it can be seen that the purple curve achieves slightly better time savings for the 40 km distance.
Summary of sensitivity studies:
All of these studies, which have checked the sensitivity of the results to variations in the baseline assumptions have shown that the shape of the curves stays rather similar. There are some small changes in the size of the time savings, but it indicates that the general conclusions from the earlier work can be applied universally to most riding situations.
All of these studies, which have checked the sensitivity of the results to variations in the baseline assumptions have shown that the shape of the curves stays rather similar. There are some small changes in the size of the time savings, but it indicates that the general conclusions from the earlier work can be applied universally to most riding situations.
Overall summary of the results
This has been a long blog post, partly because it's been a lot of work. If you've read it all, well done, and thank you. I'll try to summarise the results:
The overall objective of this analysis was to understand the performance benefit you'd get from an aerodynamic improvement at different riding speeds. As I hinted at during the introduction, the answer to that seemingly simple question depends on the precise question and the specific performance benefit 'metric' that somebody is interested in:
1) Watt savings at a fixed speed due to aerodynamic improvements:
For this metric, higher speeds yield higher power savings in Watts. The relationship between those Watt savings and speed is cubic, so higher speeds yield increasingly larger Watt savings. The Watt savings are also independent of rider power, and depend only on the rider's speed.
Although this metric is used a lot by bike company marketing departments, I feel that it's not particularly relevant for many situations. Normally, we don't ride at a fixed speed. If the riding becomes easier, because we've made our aerodynamics or some other aspect of our bike set up more efficient, then we would generally go faster so that our effort is about the same, instead of backing off. An exception might be if you're on a group ride and you're not drafting, so you feel the full aerodynamic losses (for example, riding side by side, chatting with a friend).
2) Power saving percentage at a fixed speed due to aerodynamic improvements:
Again, for the same reason I gave above, this fixed speed assumption is not particularly relevant for most practical situations, I think.
3) Speed improvement (in kph) at a fixed power output:
This is a more realistic scenario, where somebody rides or races at a similar effort (similar power) after improving their aerodynamics. With this metric, larger speed improvements in kph are achieved at higher speeds, again, but now the relationship with speed is more linear than for the fixed speed metrics described in (1) and (2) above.
Is this metric relevant and useful? For some people it might be; the people who care about average speeds and try to increase their average speed for certain rides. For those people, this information is useful, because if their riding speed is higher, for whatever reason, they will see a larger increase in their speed (a larger number of additional kph) for the aerodynamic improvement they have made, or are contemplating making. For a given speed, a lower powered rider will see a slightly larger speed benefit in kph, due to that aerodynamic improvement, than a higher power rider.
4) Speed or time improvement (as a %) at a fixed power output:
I feel that this 4th metric is not particularly interesting for most cyclists. Whereas I frequently hear cyclists say "I went 2 kph faster" or "I shaved a minute off my personal best", I rarely hear people talking about percentage improvements, either % time or % speed. Percentage changes just don't seem to interest most people, it seems.
Nevertheless, if % speed or % time improvement is what interests you, the plot on the left shows that higher speeds matter once again, but the benefits level off for the higher speeds, unlike for the previous metrics. As before, for a given speed, a lower power rider will see a larger % time or % speed benefit, due to an aerodynamic improvement, than a higher power rider.
5) Time saving in seconds at a fixed power output:
Most people, when talking about their future cycling ambitions, talk most often about time savings in minutes or seconds. That might be, for example, reducing their 25-mile time trial personal best by two minutes to get below one hour. Or shaving 15 minutes off their time at Leadville to get below the 9-hour 'big buckle' time. This metric matters the most to people, I feel.
For this metric, the time savings exhibit a very interesting shape versus speed. The largest benefits occur at moderate speeds. The time savings are also highly dependent on rider power. The greatest time saving benefits due to aerodynamic improvement occur for lower power riders. The greatest benefit for lower power riders also occurs at slower speeds, at about 30 kph for the 150 W rider, as shown in the plot. Finally, significant time savings occur at lower speeds. At just 15 kph, a 150 W rider would make a smaller time saving than at 30 kph, but only about half the saving. It's certainly not the speed-cubed relationship of the first metric. So aerodynamics certainly do matter at low speeds.
Some other key takeaways:
For a certain speed, a lower power rider will always get the same, or a larger benefit due to an aerodynamic improvement, compared with a stronger more powerful rider, no matter which performance metric is used. This goes against the traditional thinking that only pros, or strong riders, should worry about aerodynamics.
Therefore, the only valid reason for not choosing to make aerodynamic improvements to your bike or yourself is because you don't care about whether you go faster. The reasons often cited, that the speeds are not high enough to justify aerodynamic improvement, or that aero mods are only for Pros riding at 400+ Watts, are simply not true.
Some final thoughts...
- Let me know in the comments what you think. I'd be interested to know if this is as surprising to other people as it was to me.
- As with my previous blog post, some of the results are really surprising and counter-intuitive. I would be happy to hear if anybody tries to re-produce these results for themselves, as a check to convince yourself. I've made every effort to check my results, but such is the nature of working alone on a blog in my spare time, errors are always a possibility. If anybody wants to have a look at my spreadsheet, I put a copy on my Google Drive here. Or alternatively you could reproduce the calculation yourself if you prefer, because the maths involved isn't too difficult.
- I've done a quick search to see whether other people have found similar conclusions to me. There are a few sources of information that are aligned with what I found, which gives me confidence that I haven't mucked something up in my Excel spreadsheet:
- Specialized: The guys at Specialized produced a video (here) explaining why they used time-saved-over-40km as their metric when quoting the benefit of any aerodynamic improvements they studied and published in their "Win Tunnel" YouTube series. They explain that the time saving, in seconds, is more similar between riders of vastly different abilities, unlike the alternative metric that quotes power savings at a fixed speed.
- AeroX: This is a company that sells a device to improve a rider's aerodynamics, by measuring frontal area using a webcam. They have an explanation on their website (here) why amateurs can make bigger improvements through aerodynamics than pros, showing an example where a slower speed results in a larger time saving, in seconds. That's the same conclusion as mine.
- Alex's Cycle Blog: This blog post from 2013 describes a similar study to mine, where a fixed CdA reduction was made for a 150 W rider and 350 W rider. The blog explains that similar % time savings are achieved for a flat TT, but the time savings in absolute terms (i.e. seconds saved) is greater for the lower power rider. This conclusion is exactly the same as mine.
- Finally, I find the shape of the curves fascinating, especially the plot showing time savings in seconds. There must be a mathematical reason why the maximum benefit for an aerodynamic improvement occurs at moderate speeds and then tails off at higher speeds. The peak seems to occur where the CRR would be zero. There is probably some advanced calculus that could be done to explain why the maxima of the curves occurs at those points, but that level of maths is beyond me. If any maths experts out there who want to try to find a mathematical explanation for the shape of the curve, be my guest. I'd be very interested to hear any explanations.



























