# Power Data Estimates for the climbing stages

Page 13 - Get up to date with the latest news, scores & standings from the Cycling News Community.

#### Mrs John Murphy

Would it not be useful to work out the power output for the duration of the attack rather than the base of the climb?

#### Alex Simmons/RST

Le breton said:
Thanks Alex, I wrote 1 m/s but meant 1 km/h and did the calculation for 1 km/h because I read your post too fast.

(which is a bit stupid as I chose 21.6 km/h to have exactly 6 m/s)

I rechecked everything before posting but not that
OK, well that just validates my point.

See, you wouldn't even feel a 1km/h wind - that would be rated as dead calm on the wind scales and is not even close to being strong enough to make a portable anemometer move (roughly an order of magnitude less).

Le breton said:
On the CdA, 0.32 is really VERY LOW FOR A 67 kg RACER IN CLIMBING POSITION.
0.32m^2 may or may not be low in that situation, but like I said, pushing that number up simply means the range of power to account for possible wind variation gets wider, not narrower.

Le breton said:
For the Crr, I have measured it personnally on a 5km strech of road quite similar to many roads in France and obtained 0.0036. I am quite happy to use 0.004 as suggested by analyticcycling.com.
You must have exceptionally good smooth roads where you are and use exceptional tyres/tubes. Can I come train on your roads?

It is my (and others) experience that the values Analytic Cycling suggests for Crr underestimate by quite a bit. For example, the best Crr I've measured on one of the fastest wooden velodrome tracks in the world is in the 0.0023-0.0025 range* (versus Analytic Cycling's suggestion of 0.001).

Asphalt roads are typically in the 0.004-0.005 range*. There are many of us that have made such measurements.

What methodology are you using to calculate/measure Crr?

* Assuming good fast tyre/tubes as well. In case of the track that was high end Vittoria track race singles (rated at the top end of Al Morrison's tyre rolling resistance tests).

#### Le breton

Alex Simmons/RST said:
........

What methodology are you using to calculate/measure Crr?

.

I chose two points on a mountain pass about 5km apart where the map gives the altitudes with a precision of 0.1 meter.(1108.7 and 793.0).

Went to the starting point with bike (equipped with powertap) and scale on a VERY CALM DAY. Measured kilojoules with Powertap at 6, 12 and 18 km/h.

Plotted kJ/km - after subtracting gravity PART - as a function of velocity and extrapolated to 0 km/h, WHERE YOU ONLY HAVE THE EFFECT OF Crr.

Of course, the really difficult part was the excrutiating 50 min. or so it took to make the 6 km/h measurement.

I weighted myself before and a few min. after each trial.

I used Vittoria corsa evo tubulars, glued a long time before.
What method do you use?

#### Le breton

Le breton said:
...........
Plotted kJ/km - after subtracting gravity PART - as a function of velocity and extrapolated to 0 km/h, WHERE YOU ONLY HAVE THE EFFECT OF Crr.

Correction

Plotted kJ/km - after subtracting gravity PART - as a function of velocity SQUARED ( of course, I got up to correct myself) and extrapolated to 0 km/h, WHERE YOU ONLY HAVE THE EFFECT OF Crr.

#### Alex Simmons/RST

Le breton said:
I chose two points on a mountain pass about 5km apart
....
What method do you use?
Interesting. Long post to follow split to enable posting of several images.

I do field testing with a power meter also.

I don't do a lot of testing to deliberately isolate Crr. Mostly I'm interested in the CdA-Crr pair that yields lowest overall resistance. It's pretty straightforward to do that via a few means but splitting the Cda-Crr pair requires far more care, and necessitates varying key elements of the data in testing, with speed being the obvious one to vary (although one could theoretically also vary mass, or use a known additional increase in CdA, such as "Coggan's balls" although I think the balls would have to be a fair bit bigger than Andy's ).

When I have sought to split the CdA-Crr pair, I use form of regression analysis as described by Andy Coggan in the book Training & Racing with a Power Meter, which is not unlike what you describe (analytically I mean). He showed me how to do many years back now.

Data collection runs are performed on a flat (or very nearly flat constant gradient terrain) and at different speeds. You really only need about 60-90 seconds of data for each run. Speeds vary from very slow to as fast as you like but the idea is to hold speed as constant as possible and gather the average speed and power data for each run (easy to do post-hoc in software). Adjust for any small changes in kinetic energy (speed) from start to finish of each run.

You also measure factors to calculate air density, i.e. temperature, barometric pressure, altitude and humidity.

For me this has typically been performed on a velodrome. Outdoors you would do the runs in each direction on a flat/nearly flat constant gradient course so you can net out/account for the change in potential energy and wind effects (as I have been pointing out, even imperceptible changes have significant impact on the estimates).

Example data shown below:

#### Alex Simmons/RST

That data then provides us with the relationship between force (N) required to maintain a given steady state speed (m/s) - which is curvilinear of course:

A plot of the force v speed^2 for each run should be linear, and so the data points are fitted to an equation of the form y = mx + b, where the slope equates to CdA (CdA = m*2 / air density) and the vertical axis intercept equates to Crr (b/mass/gravity):

I have a spreadsheet I made available publicly many years ago via various forums which details the methodology and does all the calculations for you.

One thing to note with this method (and these regressions in general) is that the intercept (Crr) is generally pretty sensitive to errors, more so than the slope (CdA).

There are other analytical means, such as Adam Haile's short track work per lap method (hasn't worked so well for me but others have been able to get good results) and Andy Shen's method which uses Robert Chung Virtual Elevation analysis to examine the concavity of the VE plot from a data collection run at constantly increasing (or decreasing) speed, which will be either concave, flat or convex depending on whether the CdA-Crr pair is weighted too much to one or the other or the balance is just right.

I know Robert Chung has other clever means to split the pair sometimes using VE analysis.

#### Le breton

Alex Simmons/RST said:
.............That data then provides us with the relationship between force (N) required to maintain a given steady state speed (m/s) - which is curvilinear of course:

I obviously have not had the time to read in detail everything you wrote.
So, just one remark .

Your analysis and data-taking techniques are well suited for flat land biking and well suited to measure CdA.

However to determine Crr you need to have data points at very low velocities where the weight of the rolling resistance contribution is not dwarfed by the air resistance. I have not seen low velocity data in your posts. At the lowest velocity (25.3 km/h), you have an energy expenditure of ~11 kJ/km in which the rolling resistance is only about ~3kJ.

Since I live near a mountain, using its slopes is the solution of choice. Around here october-november is the best time to carry out such tests because the air is then often very calm.

Mixed in with the rolling resistance, if on a velodrome, you must have effects linked with the curves where the rolling of the tyres must be different (than straight line)

On a road there is a resistance component that I never have seen taken into account : the effect of the granularity of the road : 3 days ago "they" spread chips (gravillonnage) on a road near here on the French side of the border. It's a cheap and dangerous way of repairing roads very much in use when money is in short supply ( the Swiss hardly ever do that)

As soon as you hit such a section your velocity goes down, specially if it's fresh. The bounciness that you experience produces an energy loss that might or might not be independent of velocity.

Others : Among the tests that I have done (not as precise as the one concerning Crr) are tests on a much deteriorated forest roads where I tried to avoid as much as possible the missing asphalt sections of the road. Much to my surprise, I started to get inconsistent results if I assumed a Crr higher than 0.005.

#### Alex Simmons/RST

Le breton said:
I obviously have not had the time to read in detail everything you wrote.
So, just one remark .

Your analysis and data-taking techniques are well suited for flat land biking and well suited to measure CdA.

However to determine Crr you need to have data points at very low velocities where the weight of the rolling resistance contribution is not dwarfed by the air resistance.

Oh I agree that you need a good spread of velocities but at very low velocities you also need to be able to ride very straight.

The data I posted was just random set for illustrative purposes only of the method. Others have collected data with much more care than that and attained better results. And of course it is possible to ride very slowly on the flat.

You can also coast down to a near stop and use that data too. In that instance you have precise knowledge of power (it is zero) and speed.

Yes, velodromes and banking do bring other factors into play. But I do the testing there mainly because that's the competitive environment being used.

End of the day, I am less concerned about being able to precisely isolate Crr, than knowing simply what is faster from a given CdA-Crr pair. We already know which tyres are faster from available published testing and we have been shown it translates well to real world road results. Besides, I can't change the surface we get to race on, just the tyres/tubes.

There are some limited circumstances where knowing the split is helpful though (e.g. a hillclimb v flat TT might suggest one tyre is better than another even though two tyres might be even as far as the total energy demand from the Crr-CdA pair, but one is more aero than the other, but has higher rolling resistance).

#### Lanark

It looks like that Chris Horner will publish his SRM data during this Tour, just like he did last year. That should make it a lot easier to make an accurate estimate of the others riders' power.

He also published his Tour of California data, he did Sierra Road with 434 Watt (6,68 W/kg), in 17 minutes. I think some estimates were posted on this forum earlier that were really close to this.

http://www.srm.de/index.php/de/srm-blog/strasse/639

#### Chuffy

Just out of interest, has this kind of exercise been done on David Moncoutie? With his reputation, wouldn't he be useful in establishing a baseline?

#### Escarabajo

Chuffy said:
Just out of interest, has this kind of exercise been done on David Moncoutie? With his reputation, wouldn't he be useful in establishing a baseline?
What about if he is sick?

I think we have some ranges where we can start judging suspicious performances. Even though we have a lot of critics of those ranges, specially the young fans in this forum that believe that evolution can take place in 10-20 years.

#### Chuffy

Escarabajo said:
What about if he is sick?

I think we have some ranges where we can start judging suspicious performances. Even though we have a lot of critics of those ranges, specially the young fans in this forum that believe that evolution can take place in 10-20 years.
I know I'm probably trying to teach you to suck eggs! It was just a thought.

#### halamala

Lanark said:
It looks like that Chris Horner will publish his SRM data during this Tour, just like he did last year. That should make it a lot easier to make an accurate estimate of the others riders' power.

He also published his Tour of California data, he did Sierra Road with 434 Watt (6,68 W/kg), in 17 minutes. I think some estimates were posted on this forum earlier that were really close to this.

http://www.srm.de/index.php/de/srm-blog/strasse/639
It looks like that my calculation was pretty close.
halamala said:
Chris Horner

Tour of California 2011, Stage 4, Final climb Sierra Rd

Elevation / Höhenmeter [m] : 532 m
Distance / Streckenlänge [Km] : 5.63 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 1004 = 16 min 44 sec = 16:44
Weight rider / Gewicht Fahrer [kg] : 63.5 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 9.4 %
Average speed / mittlere Geschwindigkeit : 20.1 Km/h
Total weight / Gesamtgewicht : 71.5 kg

Power : 434.7 Watt
Power / kg : 6.8 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]

I hope he'll release the SRM file later.

#### tobsie

Could someone explain all the power stuff to me - how much wattage is an indication for PEDS? I read somewhere that when the wattage reaches 6,1 w/kg for more than xx minutes, you should start asking questions.

(Mods, delete if this isn't okay )

#### Alex Simmons/RST

tobsie said:
Could someone explain all the power stuff to me - how much wattage is an indication for PEDS? I read somewhere that when the wattage reaches 6,1 w/kg for more than xx minutes, you should start asking questions.

(Mods, delete if this isn't okay )
I think this will be OK as an answer:

Simply put, some people think that a certain power to body mass ratio for a given duration is unobtainable without "assistance". I think setting arbitrary limits is not necessarily productive.

What we know is that the highest known power to mass ratio for 1-hour by a clean rider is 6.4W/kg.

That does not mean more isn't possible without assistance.

Theoretically, if you had a rider with a combination of:
A VO2max of 90 ml/kg/min
A gross mechanical efficiency of 24%
A threshold at 90% of VO2max

which are all values that have been legitimately recorded by clean elite cyclists, then it's theoretically feasible for a clean rider to have a 1-hour power of 6.8W/kg.

Now of course there are no known reasons why some of those numbers could not be higher again, they are just reasonable upper limits that have been recorded.

Typically however, riders don't possess both a high VO2max and a high GME. But then we are not talking about typical here.

Clearly the shorter the duration, the higher that number can be.

And of course then there is the complicating factor of power output when pre-fatigued by 150+km of hard riding daily for 2+ weeks. Some riders will be better protected than others.

Whether a rider can attain that level or more is still not an indication of doping. The only real indication of doping is finding the prohibited substance in the rider in question.

I discuss ascension rates and power here:
http://alex-cycle.blogspot.com/2010/07/ascent-rates-and-power-to-body-mass.html

and in that there are also links to forum thread discussion on the issue:
http://forum.cyclingnews.com/showpost.php?p=256590&postcount=68
http://forum.cyclingnews.com/showpost.php?p=256595&postcount=69

and it has been brought up by the guys at the Science in Sport blog, who have said they think that > 6.2W/kg (I may be wrong on that number) is suspicious.

Personally I think the number is higher (and there is no real reason to set an upper limit), given that we already know a clean rider has performed at 6.4W/kg.

The point of my blog item on ascension rates and power was to ensure we factor in errors in calculations, as it's quite possible for estimates of a rider's power to span the "limit of plausibility".

All figures quoted should really have a +/- % or W noted to allow for errors in mass, wind, drafting, road surface, distance measures etc.

#### iZnoGouD

BANNED
finally mountains, can't wait to see how much power gc guys will produce

#### halamala

Frank Schleck climbed 5.8 W/kg

Preliminary calculation

Tour de France 2011, Stage 12, Luz Ardiden

Frank Schleck

Elevation / Höhenmeter [m] : 979 m
Distance / Streckenlänge [Km] : 13.3 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 2241 = 37 min 21 sec = 37:21
Weight rider / Gewicht Fahrer [kg] : 67 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 7.4 %
Average speed / mittlere Geschwindigkeit : 21.3 Km/h
Total weight / Gesamtgewicht : 75.0 kg

Power : 394.6 Watt
Power / kg : 5.8 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]

#### Escarabajo

halamala said:
Preliminary calculation

Tour de France 2011, Stage 12, Luz Ardiden

Frank Schleck

Elevation / Höhenmeter [m] : 979 m
Distance / Streckenlänge [Km] : 13.3 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 2241 = 37 min 21 sec = 37:21
Weight rider / Gewicht Fahrer [kg] : 67 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 7.4 %
Average speed / mittlere Geschwindigkeit : 21.3 Km/h
Total weight / Gesamtgewicht : 73.0 kg

Power : 394.6 Watt
Power / kg : 5.8 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]
Thanks.

Not much to see. They rode tempo almost most of the mountain. We also need to add that it was the first day in the mountains and the muscles are not well conditioned for that.

#### halamala

You're welcome.

I'm really waiting for an article from Dr. Ross Tucker (Sportscientists.com).

So that's about 1.50 slower than Armstrong in 2003. Not really sure how encouraging this is given that Armstrong had a fall while today there was a definite lull before Szmyd took over.

#### marcozero

halamala said:
Preliminary calculation
...
Weight rider / Gewicht Fahrer [kg] : 67 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Total weight / Gesamtgewicht : 73.0 kg
...

67 + 8 = 75 kg.

#### halamala

marcozero said:
67 + 8 = 75 kg.
Oh, thanks. Fixed. It doesn't effect to power & power/kg. It's still "Power : 394.6 Watt, Power / kg : 5.8 Watt / kg".

#### davestoller

halamala said:
Oh, thanks. Fixed. It doesn't effect to power & power/kg. It's still "Power : 394.6 Watt, Power / kg : 5.8 Watt / kg".

400 watts for 37 minutes for a guy his size is a lot. I mean, not believable--how does this compare to his ascent of Alpe D'Huez is the real question.

5.8 watts/kg, not impressive.

i cant tell if the bike weight is what is throwing this off.

Didnt look all that fast to me.

#### halamala

Last 10 Km: Frank Schleck climbed 5.9 W/kg

Tour de France 2011, Stage 12, Luz Ardiden [ last 10.0 Km ]

Frank Schleck

Elevation / Höhenmeter [m] : 782 m
Distance / Streckenlänge [Km] : 10.0 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 1750 = 29 min 10 sec = 29:10
Weight rider / Gewicht Fahrer [kg] : 67 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 7.8 %
Average speed / mittlere Geschwindigkeit : 20.5 Km/h
Total weight / Gesamtgewicht : 75.0 kg

Power : 395.7 Watt
Power / kg : 5.9 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]

The Record [last 10.0 Km]

Tour de France 2003, Stage 15, Luz Ardiden [ last 10.0 Km ]

Lance Armstrong

Elevation / Höhenmeter [m] : 782 m
Distance / Streckenlänge [Km] : 10.0 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 1628 = 27 min 08 sec = 27:08
Weight rider / Gewicht Fahrer [kg] : 72 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 7.8 %
Average speed / mittlere Geschwindigkeit : 22.1 Km/h
Total weight / Gesamtgewicht : 80.0 kg

Power : 458.1 Watt
Power / kg : 6.3 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]

halamala said:
Preliminary calculation

Tour de France 2011, Stage 12, Luz Ardiden

Frank Schleck

Elevation / Höhenmeter [m] : 979 m
Distance / Streckenlänge [Km] : 13.3 Km
Time in seconds / Fahrzeit in Sekunden [sec] : 2241 = 37 min 21 sec = 37:21
Weight rider / Gewicht Fahrer [kg] : 67 kg
Weight bicycle, clothes etc. / Gewicht Fahrrad [kg] : 8 kg

Grade / mittlere Seigung : 7.4 %
Average speed / mittlere Geschwindigkeit : 21.3 Km/h
Total weight / Gesamtgewicht : 75.0 kg

Power : 394.6 Watt
Power / kg : 5.8 Watt / kg

Source: [ http://www.rst.mp-all.de/bergauf.htm ]