Wind & Altitude : Effects On Power To Weight Ratio

[Cozy Beehive]

No kidding...really?

BTW, our paper in JAB was far from the first to present a mathematical model of the power requirements of cycling...for example, I remember reading this paper back when I was still an undergraduate student:

http://www.ncbi.nlm.nih.gov/pubmed/468661

However, our paper was the first to actually validate this sort of modeling approach.

Interesting. You guys just validated the model with a power meter, but it was first explored in 1979!

 

[acoggan]

Interesting. You guys just validated the model with a power meter, but it was first explored in 1979!

Nunweiler did wind tunnel tests of cyclists way back in the 1950s, but I lost my copy of that report quite some time ago, so can't say if he referenced an even earlier modeling paper.

 

[Rip:30]

Altitude. Evolution.

Hey you both should check this out re: altitude.

http://www.sciencemag.org/cgi/content/abstract/329/5987/72

Maybe the most talented bike racer exists in Nepal or the Andes, but has never thrown a leg over a bike.

 

[Cozy Beehive]

Right - but if power supply is assumed to be constant, then as Rip:30 stated wind has absolutely no affect upon it, only upon the resultant speed.

Oh I see. I'll agree there. That is correct for cycling or any other activity at low Reynold's number. Talk to the aerospace guys and you'll get a different answer.

 

[Cozy Beehive]

You can use it estimate power.

Any calculation is fundamentally an estimation. Not sure why you highlighted my wording in red.

 

[Rip:30]

Measuring diffusion capacity would probably be more telling.

In any case, it was studies like this one:

http://www.ncbi.nlm.nih.gov/pubmed/19833809

that I had in mind when I made that comment.

(Hmm: upon rereading that abstract, I'm wondering whether swimming - which is the only sport that seems to induce significant adaptations in lung structure/function, and one that my kids are currently involved in - might have a similar effect as altitude.)

We used to do intervals (short) on the trainer holding our breath. Sort of felt like swimming to me--hated it.

 

[Rip:30]

Any calculation is fundamentally an estimation. Not sure why you highlighted my wording in red.

It's not as direct as measuring torque and angular momentum at the bb or rear hub--more error prone.

 

[Cozy Beehive]

It's not as direct as measuring torque and angular momentum at the bb or rear hub--more error prone.

And I wouldn't know that?

 

[Cozy Beehive]

It's not as direct as measuring torque and angular momentum at the bb or rear hub--more error prone.

I agree with that. No one said anything about calculation as an estimation being the same as measured value. Which is why I came here to see if there was such a validation done. None as far as I can see.

 

[Cozy Beehive]

BTW, I should point out that if your focus isn't on the effect of wind per se but on how to estimate the power required to climb, e.g., l'Alpe de Huez, you shouldn't be using the Bassett et al. equation to estimate frontal area. First, it based upon only a handful of subjects, and second, it is for riders in the aero position, not sitting up and/or standing while climbing on a road bike. Better choices would be:

http://www.ncbi.nlm.nih.gov/pubmed/19199206

or

http://www.ncbi.nlm.nih.gov/pubmed/12355191

Of course, such equations only allow you estimate A, while Cd itself can also vary significantly between riders, even if they aren't using aero bars...

I respect your opinion. Not ready to spend $$ right now to find out what empirical formulas are developed in those two papers. If you have a copy of the papers, would you summarize it for us? Interestingly, I used some good assumptions and unknowingly, ended up with a drag area of 0.32 m^2. This, compared with Simmons' assumption of 0.3m^2 is just a 6% change.

By the way, how much of a "variation" is seen in Cd between riders? For all practical purposes, I consider assuming Cd constant for low Reynold's numbers valid.

 

[Cozy Beehive]

I think Alex's calculations are absolutely trustworthy...the physics are quite well-established, and the guy knows his math. IOW, with respect to the effects of wind there's nothing here to debate (except perhaps just how strong the wind is/was at a particular point on a particular climb on a particular day). The effects of altitude on aerobic power are also quite well-known. What is also known, however, it that there is significant variation between individuals in the magnitude of any effect, even if you control for the degree of acclimatization.

Andrew,

I ran a small validation check against both Alex's and Tom Compton's "Analytic Cycling" numbers. I summed them up in my conclusions in my post. Please read and see if they make sense theoretically.

 

[Rip:30]

Andrew,

I ran a small validation check against both Alex's and Tom Compton's "Analytic Cycling" numbers. I summed them up in my conclusions in my post. Please read and see if they make sense theoretically.

Numbers make sense but by setting the power as the predicted dependent variable you ignore the biological reality of a racer's body: racers don't produce more power due to wind, they just slow down. Think about your logic for a second. If it is a calm day, why would a racer hold back extra power "in reserve". No way. He goes all out. Has to compare times with people that probably raced in similar conditions. Bike racers are pinned at max power in every race. Race power should be a constant.

 

[dbrower]

Where did you attend school?

Wind absolutely affects power because it works against forward movement, particularly if its a headwind. If power changes, so does power to weight ratio.

Not going to argue with you, however. Thanks for your time.

I think you meant "Wind absolutely affects the power required to maintain a constant forward velocity", right?


-dB

 

[dbrower]

One thing that has not been discussed in any of the varied threads on various sites is the affects of the race situation.

In particular, on a number of the races, the fastest time up has been helped by the achiever having drafted in a group, or behind a teammate for varying lengths of time. With the exception of Armstrong's TT, few if any of the other times are achieved solo, which is fundamental to the assumptions in all of the models I have seen discussed.

-dB

 

[Cozy Beehive]

One thing that has not been discussed in any of the varied threads on various sites is the affects of the race situation.

In particular, on a number of the races, the fastest time up has been helped by the achiever having drafted in a group, or behind a teammate for varying lengths of time. With the exception of Armstrong's TT, few if any of the other times are achieved solo, which is fundamental to the assumptions in all of the models I have seen discussed.

-dB

Here's what I think. A CFD analysis is one of the best methods of simulating the effect of "obstacles" on wind retarding force, such as trees, spectators, switchbacks etc. Simplistic physics models are good for teaching but does not illuminate accurately the combined effects of a large number of variables, a task computationally difficult for the human mind to even fathom but extremely easy for a powerful computer, keeping errors down.

Unfortunately, there is an imbalance of use of CFD, mostly in the time trialing discipline. Not sure why but apparently, there's more devotion of research and analysis in increasing performance in time trials than there is in climbing. Did everyone forget that a stage race is won on the mountains?

Anyway, here's an example. A few years ago, a CFD software maker ran a study on a 9 man TTT scenario. The results of the simulations were illuminating. They wrote : "..compared with the lead cyclist, the drag of the rider in second place is reduced by 21% - a significant saving. The third rider feels a further small decrease in drag over the second, but from the third rider back all other cyclists experience almost identical drag. As the riders are continually progressing towards the front of the chain, taking a short turn on the front, before freewheeling to the rear of the line, on average (assuming a constant rate of rider rotation and ignoring the effect of dropping back) the drag coefficient of a rider in the TTT is around 27% lower than experienced by an individual rider. Perhaps the most surprising conclusion from the CFD simulation is that, despite feeling the full force of the oncoming air, the lead rider experiences lower drag than if he were riding an ITT at the same speed. The drag coefficient of the leading TTT rider is 0.277, while that of an individual rider is 0.285 [drag coefficient is measure of the force each rider experiences corrected for differences in size]. This rare example of "something for nothing" occurs because the second place rider reduces the influence of the lead rider's wake, increasing his base pressure and consequently reducing the drag force that he experiences."

Such is the power of CFD. Yet, sadly, they don't use this enough in studying other scenarios found in professional cycling. Beats me.

 

[Rip:30]

I don't think this has anything to do with what you're saying. The fundamental physics model of cycling is the one I described in the previous post. You can use it to calculate both power, given constant speed, or solve for speed, given constant power. Either way, you will use that equation if you are to know how wind, grade, rolling resistance etc affects power or speed.

This. ^^

Hey Cozy, I read your work on your website and it makes total sense what you are doing. By making power the dependent variable you can compare different riders' power outputs based on the variables that you have access to during races (hopefully). (Or predict rider speed, or drag, ect.--if you know power). When I used to work with cycling power stuff it was about maximizing an individuals' power output for a given duration. Race power outputs are constrained by factors such as mitochondrial density, max hr x stroke vol, atmospheric O2 concentration, CNS stuff ect. So from a biological perspective, when you said "wind effects power:weight", I couldn't see the connection--one individual should have the same power output for an identical TT course run in the wind or the calm (psychological factors aside).

Anyway looks good, interesting stuff.

 

[Rip:30]

The confusion seems to be that you are assuming constant speed, whereas Rip:30 is assuming constant power. If speed is indeed constant (as you seem to assume), then yes, wind affects the power demand, but if the power supply is constant (as Rip:30 is assuming), then it only changes the speed.

This. ^^

Hey Cozy, I read your detailed work via your website and it makes total sense now. By making power the dependent variable you can compare different riders' power outputs based on the variables that you have access to during races (hopefully). (Or predict rider speed, or drag, ect.--if you know power). When I used to work with cycling power stuff it was about maximizing an individual's power output for a given duration. Race power outputs are constrained by factors such as mitochondrial density, max hr x stroke vol, atmospheric O2 concentration, CNS, stuff ect. So from a biological perspective, when you said "wind effects power:weight"--with no additional qualifiers, I couldn't see the connection. One individual should have the same power output for an identical TT course run in the wind or the calm (psychological factors aside).

Anyway looks good, interesting stuff.

 

[acoggan]

Anyway, here's an example. A few years ago, a CFD software maker ran a study on a 9 man TTT scenario. The results of the simulations were illuminating. They wrote : "..compared with the lead cyclist, the drag of the rider in second place is reduced by 21% - a significant saving. The third rider feels a further small decrease in drag over the second, but from the third rider back all other cyclists experience almost identical drag. As the riders are continually progressing towards the front of the chain, taking a short turn on the front, before freewheeling to the rear of the line, on average (assuming a constant rate of rider rotation and ignoring the effect of dropping back) the drag coefficient of a rider in the TTT is around 27% lower than experienced by an individual rider. Perhaps the most surprising conclusion from the CFD simulation is that, despite feeling the full force of the oncoming air, the lead rider experiences lower drag than if he were riding an ITT at the same speed. The drag coefficient of the leading TTT rider is 0.277, while that of an individual rider is 0.285 [drag coefficient is measure of the force each rider experiences corrected for differences in size]. This rare example of "something for nothing" occurs because the second place rider reduces the influence of the lead rider's wake, increasing his base pressure and consequently reducing the drag force that he experiences."

http://www.ncbi.nlm.nih.gov/pubmed/2318782

http://www.ncbi.nlm.nih.gov/pubmed/10589873

http://www.trainingandracingwithapowermeter.com/2010/04/does-drafting-benefit-leading-rider_23.html

http://www.ncbi.nlm.nih.gov/pubmed/20171640

http://www.ncbi.nlm.nih.gov/pubmed/20488446

So in return, where might I learn more about the CFD study you describe above?

(BTW, did you get those two papers I sent you?)

 

[acoggan]

One thing that has not been discussed in any of the varied threads on various sites is the affects of the race situation.

Cf. the posts on the Saxobank forum by KBteammate that I cited in the thread in the clinic.

 

[Cozy Beehive]

So in return, where might I learn more about the CFD study you describe above?

(BTW, did you get those two papers I sent you?)

Dr. Coggan, the return service was already done by a reader on your blog on July 2. See http://www.trainingandracingwithapowermeter.com/2010/04/does-drafting-benefit-leading-rider_23.html . So lets see, a) you simply skim over the titles of some posts here on the forum and then subtly insult their importance b) you don't read the comments from your own blog readers.

Btw, with regard to that blog post - I have not seen anything as high as 97% efficiency in a bicycle drivetrain so you're slightly overestimating our beautiful machine. Just saying.

Yes I got all the papers you sent me so a big thank you.

By the way, I have to show you this plot :

It shows two phenomena happening as a cyclist exercises at altitude, upto 2200 m (the peak of the Tourmalet).

1) The power relief calculated for a 65 kg rider due to air density decrease - an advantage since drag is reduced. Hence, hour records are broken at altitude. This represented with red line.
2) The degradation in performance due to VO2 max decline, estimates calculated for elite runners by two people as shown in the plot (the link to these equations were tossed around by someone in the Clinic). http://midweekclub.ca/powerFAQ.htm#Q17

Might it be fair to say that these two phenomena work against each other in some fashion and the net effect of the two is a curve in between them but hinging more towards the decrease in performance due to VO2 max decline? Or am I overstepping logic?

Also, do you have a link to a study as 2) done for cyclists instead of runners?

 

[acoggan]

Dr. Coggan, the return service was already done by a reader on your blog on July 2. See http://www.trainingandracingwithapowermeter.com/2010/04/does-drafting-benefit-leading-rider_23.html . So lets see, a) you simply skim over the titles of some posts here on the forum and then subtly insult their importance b) you don't read the comments from your own blog readers.

No, I'd simply forgotten about it.

Btw, with regard to that blog post - I have not seen anything as high as 97% efficiency in a bicycle drivetrain

Then I suggest that you do some reading, perhaps starting with our wind tunnel validation paper and continuing with studies by Kyle and Berto, etc...just saying, of course.

By the way, I have to show you this plot :

It shows two phenomena happening as a cyclist exercises at altitude, upto 2200 m (the peak of the Tourmalet).

1) The power relief calculated for a 65 kg rider due to air density decrease - an advantage since drag is reduced. Hence, hour records are broken at altitude. This represented with red line.
2) The degradation in performance due to VO2 max decline, estimates calculated for elite runners by two people as shown in the plot (the link to these equations were tossed around by someone in the Clinic). http://midweekclub.ca/powerFAQ.htm#Q17

Might it be fair to say that these two phenomena work against each other in some fashion and the net effect of the two is a curve in between them but hinging more towards the decrease in performance due to VO2 max decline? Or am I overstepping logic?

If the balance were such that the reduction in VO2max were always more important, then hour records wouldn't be set at altitude, now would they?

More directly, which effect is more important depends upon:

1) the speed of the cyclist (e.g., riding on the level vs. riding uphill); and

2) the extent to which a given individual is impacted by altitude.

Also, do you have a link to a study as 2) done for cyclists instead of runners?

Lots out there, but here's one that springs to mind that made measurements at various "altitudes":

http://www.ncbi.nlm.nih.gov/pubmed/11104308

Edit: not sure if that is the right link, so try this one as well:

http://www.ncbi.nlm.nih.gov/pubmed/12840636

Edit2: here's another good one:

http://www.ncbi.nlm.nih.gov/pubmed/17882451

 

[Notso Swift]

Cozy Beehive, I haven't been on for a few days but could you do me a favour and change the tittle

 

[Notso Swift]

Cozy Beehive, I haven't been on for a few days but could you do me a favour and change the tittle because the first thing I thought when I saw it was the same as Rip. And since you seem to be well qualified you know you are wrong, so just call it something else and I will never need to bother you again

 

[Le breton]

Cozy Beehive, I haven't been on for a few days but could you do me a favour and change the tittle

That would be a good idea!
I didn't consult that thread until I saw some names appear among contributors because it sounded so stupid!

 

[ihavenolimbs]

So in return, where might I learn more about the CFD study you [Cozy] describe above?

I remember seeing it in a trade-rag a few years ago. I think it was summarised in a STAR-CD (they make CFD software that is used in F1, aerospace, and other areas) mag that they put out. I think the simulation showed about a 2% saving for the front rider, vs. riding solo?

And sorry, I've thrown away the copy we had on the work coffee-table.