Power Profiling -- good stuff

It would seem a bit more useful if there was some sort of correlation between poewr output and some
other measurement such as TT speed.

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

In another thread, Andy Coggan pointed toward his article "Power Profiling" at
www.cyclingpeaks.com containing an Excel table of W/kg for different duration efforts and a range
of performance levels

going

Quoted message said:

from world champions down to untrained athletes. IMHO, this is

really

Quoted message said:

good stuff for anyone who wants to be at all analytical or objective about his/her training and
performance. That said, I have a number

of

Quoted message said:

questions and comments for Andy or any other of the power gurus and advocates. (Warning: this may
turn out to be a long post so either move on now, or go read his article if you haven't already.)

First, it seems like the way you bracketed the groups is natural,

and

Quoted message said:

in my admittedly subjective experience and observation, the overlap

is

Quoted message said:

reasonable. By that I mean, for example, the "lower half" of the cat 1's is comparable to the
"upper half" of the 2's, etc., at least to first order. I'm curious how confident you are in the
values in the table between "World Champ" and "Untrained". I ask for two reasons. For one, my own
profile doesn't fit the way I would expect. And second, the values in the table progress linearly
(as you move

through

Quoted message said:

ability levels) and the brackets cover a constant, wide range. As we all know, assume a
relationship is linear if we don't have enough

data

Quoted message said:

or insights that say it should be otherwise!

In particular, I'm wondering if you might be a little bit low with

the

Quoted message said:

"5 min" and "20 min" W/kg values, at least in the Div. III pro

through

Quoted message said:

cat 3 range, and perhaps a little high for the "5 s" and "1 min" values. For example, I am pretty
confident that the best cat 1 climbers can put out more than 5.19 W/kg over a 20-minute effort.

My own profile, at age 42, is as follows:

Duration: W/kg:
----------------------
20 minute: 4.7 5 minute: 5.1 1 minute: 8.1 5 second: 13-15

My values were about 10% higher when I last raced seriously, about

12

Quoted message said:

years ago. I did win a couple of P12 road races then, but I was certainly not at the level that
the table indicates for the

20-minute

Quoted message said:

level. OTOH, I'm pretty sure I could sprint better than the average cat 4, which is where the
table puts me for the 5-second test.

BTW, please don't think I'm nit-picking the article and the table. I think the information is
great and really useful. I just think you might like hearing about some other data points!

Another question or two...

It would be fascinating to see some real profiles for a variety of ability levels and specialties
(and all-arounders). Does anyone have some profiles they can share? My profile shows decent
lactate tolerance and VO2Max but pretty weak short-term power. I'm curious,

is

Quoted message said:

there really anyone that is well-trained and has a flat profile?

Also, W/kg is clearly a good indicator of climbing performance. But for time trialing, it isn't
that strongly related is it? In my case, my 20-minute W/kg suggest I should be in the cat 1-2
range which I think is reasonable for climbing, but, trust me, it's way off for

time

Quoted message said:

trialing. In your experience, should someone be able to work on

their

Quoted message said:

aero position, etc. so that their climbing and tt'ing are

comparable,

Quoted message said:

at least relative to other riders in the same category? In other words, is body mass a good
indicator for the aerodynamic drag one should be able to achieve? I wouldn't think so, but the

power-profile

Quoted message said:

article and table sort of suggest that one's climbing and tt'ing can be similarly categorized
based on mass alone.

Anyway, thanks again for sharing your great information and

insights.

Quoted message said:

Mark Fennell

In article <[email hidden]>, Tom Kunich

Quoted message said:

It would seem a bit more useful if there was some sort of correlation between poewr output and
some other measurement such as TT speed.

But TT speed has only two variables (other than wind, road surface, hills), power output for the
time period, and aerodynamic drag. Aero drag should be closely related to rider weight/size so if
you know weight and power somebody like Andy could reasonably calculate your TT time and shoe size.

-WG

warren said:

In article <[email hidden]>, Tom Kunich

Quoted message said:

It would seem a bit more useful if there was some sort of correlation between poewr output and
some other measurement such as TT speed.

But TT speed has only two variables (other than wind, road surface, hills), power output for
the time period, and aerodynamic drag. Aero drag should be closely related to rider weight/size
so if you know weight and power somebody like Andy could reasonably calculate your TT time and
shoe size.

Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part doesn't.
Because of the way these things scale, W/kg is a good indicator of climbing ability, but total W is
a good indicator of TT ability.

Mark Fennell said:

Quoted message said:

... IMHO, this is really good stuff for anyone who wants to be at all analytical or objective
about his/her training and performance....

Finally. A thread where Chang will have nothing to contribute.

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

It would be fascinating to see some real profiles for a variety of ability levels and specialties
(and all-arounders). Does anyone have some profiles they can share? My profile shows decent
lactate tolerance and VO2Max but pretty weak short-term power. I'm curious, is there really anyone
that is well-trained and has a flat profile?

Also, W/kg is clearly a good indicator of climbing performance. But for time trialing, it isn't
that strongly related is it? In my case, my 20-minute W/kg suggest I should be in the cat 1-2
range which I think is reasonable for climbing, but, trust me, it's way off for time trialing. In
your experience, should someone be able to work on their aero position, etc. so that their
climbing and tt'ing are comparable, at least relative to other riders in the same category?

If you work on your aero position, you should be able to get much better than riders in
your category.

The reason is that getting more aero is not easy to quantify, unless one has access to a wind
tunnel. People will spend $5 per gram shaving weight off their bike, yet not spend any time at all
improving their time trial position, where the relative benefits are much higher.

IIRC, if the drag coefficient for riding on the hoods is 100, then getting in the drops is ~80% of
that, getting a full aero setup is ~65-70%, etc. Riders like Colby Pearce or Chris Boardman could
get down to ~50% or lower.

The proportional savings are far greater than shaving a pound or two off your bike, especially when
one considers that the weight of the rider must be factored into the proportional savings of the
system as a whole.

Quoted message said:

In other words, is body mass a good indicator for the aerodynamic drag one should be able to
achieve? I wouldn't think so,

You are correct, body mass is not a good indicator.

Quoted message said:

but the power-profile article and table sort of suggest that one's climbing and tt'ing can be
similarly categorized based on mass alone.

That may be true if one generalizes, but in specific cases (like Boardman or Pearce) it is not true.
My last year racing, I went ahead and worked on the TT position after ignoring it for 10 years. I
was still pretty crappy at it, but I could tell the value it had by taking it out on group rides.
It's pretty amazing what a difference it makes. If I were to ever race again, I'd spend a lot of
time trying to get more aero with a normal road setup.

I don't know if you watched the pro1/2 Socal races after your races (I'm assuming you're master),
but notice how Paolinetti was always forearms on the handlebar in breakaways, even through turns? He
had that figured out.

In article said:

Mark Fennell said:

Quoted message said:

... IMHO, this is really good stuff for anyone who wants to be at all analytical or objective
about his/her training and performance....

Finally. A thread where Chang will have nothing to contribute.

So what else is new?

-WG

warren said:

In article said:

Mark Fennell said:

Quoted message said:

... IMHO, this is really good stuff for anyone who wants to be at all analytical or objective
about his/her training and performance....

Finally. A thread where Chang will have nothing to contribute.

So what else is new?

Nada. Stick a fork in me, I'm done.

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

In particular, I'm wondering if you might be a little bit low with the "5 min" and "20 min" W/kg
values, at least in the Div. III pro through cat 3 range, and perhaps a little high for the "5 s"
and "1 min" values. For example, I am pretty confident that the best cat 1 climbers can put out
more than 5.19 W/kg over a 20-minute effort.

Hmm - my best sprint peak power (I don't know about holding it for 5s) is about 5 "vertical squares"
below that of my 20 min value. But I always knew I couldn't sprint :-). Wish it wasn't so critical.

Jeff

Mark Fennell said:

It would be fascinating to see some real profiles for a variety of ability levels and specialties
(and all-arounders).

Check the rowing world, they've been doing it for years. They do tests on the Concept2 rowing
machine, and, ideally, the power output should be distributed thus:

Test Hwt men Lwt women 10" 123% 118% 1' 115% 112% 2 km 100% 100% 6 km 93% 95% 1 h ? 90%

(The numbers for Lwt men and hwt women are in between these. For hwt men, 2k takes about 6' and 6k
about 19'30". Lwt women 7'15" and 23'00". "Heavyweight" really means "open class"; typically those
men are 1m95 and 90 kg. Lightweight women cannot weigh more than 59 kg and are around 1m70.)

In rowing, the goal is always to deliver a maximum performance on 2 km, that's why that score is a
reference at 100%. If some of your percentages fall behind, you should spend (extra) time training
that particular zone. Ultimately, that will benefit the 2k score.

"Robert Chung" <[email hidden]> wrote in message "]news:[email hidden]...

Quoted message said:

warren said:

In article <[email hidden]>, Tom Kunich

Quoted message said:

It would seem a bit more useful if there was some sort of correlation between poewr output and
some other measurement such as TT speed.

But TT speed has only two variables (other than wind, road surface, hills), power output for
the time period, and aerodynamic drag. Aero drag should be closely related to rider weight/size
so if you know weight and power somebody like Andy could reasonably calculate your TT time and
shoe size.

Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
doesn't. Because of the way these things scale, W/kg is a good indicator of climbing ability, but
total W is a good indicator of TT ability.

Except that Heil et al. have concluded that Cd also varies inversely with weight (with an exponent
of -0.45). The overall result is then as Warren indicates, i.e., even though mass per se has little
effect on the power requirement to travel at a fixed velocity on a completely flat road, W/kg is
still likely to be a better predictor of TT performance than W alone.

Andy Coggan

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

In another thread, Andy Coggan pointed toward his article "Power Profiling" at
www.cyclingpeaks.com containing an Excel table of W/kg for different duration efforts and a range
of performance levels going from world champions down to untrained athletes. IMHO, this is really
good stuff for anyone who wants to be at all analytical or objective about his/her training and
performance. That said, I have a number of questions and comments for Andy or any other of the
power gurus and advocates. (Warning: this may turn out to be a long post so either move on now, or
go read his article if you haven't already.)

First, it seems like the way you bracketed the groups is natural, and in my admittedly subjective
experience and observation, the overlap is reasonable. By that I mean, for example, the "lower
half" of the cat 1's is comparable to the "upper half" of the 2's, etc., at least to first order.
I'm curious how confident you are in the values in the table between "World Champ" and
"Untrained". I ask for two reasons. For one, my own profile doesn't fit the way I would expect.
And second, the values in the table progress linearly (as you move through ability levels) and the
brackets cover a constant, wide range. As we all know, assume a relationship is linear if we don't
have enough data or insights that say it should be otherwise!

As the accompanying article indicates, I fixed the upper and lower ends of each range based on known
power outputs of world champions/world record holders and untrained individuals, respectively. The
values in between were spread equally (i.e., linear relationship assumed), simply because at present
there's not enough data to justify doing otherwise. Remember, the purpose of the tables is to
compare relative ability across different exercise durations reflecting different physiological
characteristics, not to attempt to assign or describe riders of different categories (strip the
category guidelines from the tables, and they would be just as useful). This is why I didn't assume
a normal distribution and spread the values that way - that might (or might not - no one at present
has the data to say for sure) better reflect reality, but has the disadvantage of squeezing
everything together toward the middle, making anyone who isn't well above or well below average
appear to be an "all rounder".

Quoted message said:

In particular, I'm wondering if you might be a little bit low with the "5 min" and "20 min" W/kg
values, at least in the Div. III pro through cat 3 range, and perhaps a little high for the "5 s"
and "1 min" values. For example, I am pretty confident that the best cat 1 climbers can put out
more than 5.19 W/kg over a 20-minute effort.

My own profile, at age 42, is as follows:

Duration: W/kg:
----------------------
20 minute: 4.7 5 minute: 5.1 1 minute: 8.1 5 second: 13-15

My values were about 10% higher when I last raced seriously, about 12 years ago. I did win a
couple of P12 road races then, but I was certainly not at the level that the table indicates for
the 20-minute level. OTOH, I'm pretty sure I could sprint better than the average cat 4, which is
where the table puts me for the 5-second test.

Better than the average cat 4 *match sprinter*, as that notion implies for somebody with that
relative ability to compared to, e.g., Sean Eadie? Again, as described in the accompanying article,
by basing things on the performance of specialists (match sprinters for 5 s), the scales tend to be
skewed from a road racers perspective. To state it another way: compared to a true sprinter, most
people racing on the road *do* have relatively low neuromuscular power. However, I don't think one
really can or should try to develop discipline-specific tables. First, too many people cross over to
different disciplines, thus making it difficult to develop valid standards, esp. since the only
point proposing discipline-specific tables would be to improve the category guidelines - which
aren't the point of the tables in the first place. Second, discipline-specific tables would deviate
from the logic that was used to develop the tables in the first place.

Quoted message said:

BTW, please don't think I'm nit-picking the article and the table. I think the information is
great and really useful. I just think you might like hearing about some other data points!

Another question or two...

It would be fascinating to see some real profiles for a variety of ability levels and specialties
(and all-arounders). Does anyone have some profiles they can share? My profile shows decent
lactate tolerance and VO2Max but pretty weak short-term power. I'm curious, is there really anyone
that is well-trained and has a flat profile?

Also, W/kg is clearly a good indicator of climbing performance. But for time trialing, it isn't
that strongly related is it? In my case, my 20-minute W/kg suggest I should be in the cat 1-2
range which I think is reasonable for climbing, but, trust me, it's way off for time trialing. In
your experience, should someone be able to work on their aero position, etc. so that their
climbing and tt'ing are comparable, at least relative to other riders in the same category? In
other words, is body mass a good indicator for the aerodynamic drag one should be able to achieve?
I wouldn't think so, but the power-profile article and table sort of suggest that one's climbing
and tt'ing can be similarly categorized based on mass alone.

This has been addressed in other posts...I'll only add the reminder (once again) that the point of
the tables is to help evaluate somebody's relative strengths and weaknesses, not predict their
performance.

Quoted message said:

Anyway, thanks again for sharing your great information and insights.

You're welcome.

BTW, there is a ~700 member mailing list at www.topica.com/lists/wattage/read devoted to discussion
of training with power meters, etc....based on the interest you've expressed here, you might want to
give it a read.

Andy Coggan

Andy Coggan said:

"Robert Chung" <[email hidden]> wrote

Quoted message said:

Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
doesn't. Because of the way these things scale, W/kg is a good indicator of climbing ability, but
total W is a good indicator of TT ability.

Except that Heil et al. have concluded that Cd also varies inversely with weight (with an exponent
of -0.45). The overall result is then as Warren indicates, i.e., even though mass per se has
little effect on the power requirement to travel at a fixed velocity on a completely flat road,
W/kg is still likely to be a better predictor of TT performance than W alone.

Cd varies inversely with mass? Hmmm, I didn't know that. Anyway, if A varies (roughly) with m^.67
and Cd varies (roughly) with m^(-.45), then CdA would vary roughly with m^(.2). A 25% difference in
mass gives a 2% difference in CdA?

I said:

Andy Coggan said:

"Robert Chung" <[email hidden]> wrote

Quoted message said:

Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
doesn't. Because of the way these things scale, W/kg is a good indicator of climbing ability,
but total W is a good indicator of TT ability.

Except that Heil et al. have concluded that Cd also varies inversely with weight (with an
exponent of -0.45). The overall result is then as Warren indicates, i.e., even though mass per se
has little effect on the power requirement to travel at a fixed velocity on a completely flat
road, W/kg is still likely to be a better predictor of TT performance than W alone.

Cd varies inversely with mass? Hmmm, I didn't know that. Anyway, if A varies (roughly) with m^.67
and Cd varies (roughly) with m^(-.45), then CdA would vary roughly with m^(.2). A 25% difference
in mass gives a 2% difference in CdA?

Goddammit, I hit the wrong key. Anyway, my calculation was off but the point should still be the
same: A 25% difference in mass gives a 5% difference in CdA?

In article said:

I said:

Andy Coggan said:

"Robert Chung" <[email hidden]> wrote
> Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
> doesn't. Because of the way these things scale, W/kg is a good indicator of climbing ability,
> but total W is a good indicator of TT ability.

Except that Heil et al. have concluded that Cd also varies inversely with weight (with an
exponent of -0.45). The overall result is then as Warren indicates, i.e., even though mass per
se has little effect on the power requirement to travel at a fixed velocity on a completely
flat road, W/kg is still likely to be a better predictor of TT performance than W alone.

I cheated by recalling an article in VeloNews from years back about two guys who claimed they could
predict TT times based on the riders' wattage output and their aero drag. Whatever happened to
those guys?

Quoted message said:

Quoted message said:

Cd varies inversely with mass? Hmmm, I didn't know that. Anyway, if A varies (roughly) with
m^.67 and Cd varies (roughly) with m^(-.45), then CdA would vary roughly with m^(.2). A 25%
difference in mass gives a 2% difference in CdA?

Goddammit, I hit the wrong key. Anyway, my calculation was off but the point should still be the
same: A 25% difference in mass gives a 5% difference in CdA?

At Analytic Cycling he uses "effective frontal area" measured in square meters for aero drag. I'm
curious if you can calculate reasonably well this based on rider height and weight. He says typical
values are .4 to .7 m2. Last night I did a little test on a perfectly flat road, 26.5 mph, 1 minute,
back and forth once each. There was a very slight wind and my efa numbers according to his
calculations were .75 and .90 (tailwind and headwind).

-WG

warren said:

I cheated by recalling an article in VeloNews from years back about two guys who claimed they
could predict TT times based on the riders' wattage output and their aero drag. Whatever happened
to those guys?

I agree, you can do a good job predicting TT times based on wattage and aero drag; but what you
wrote upthread was that if you know wattage and *weight* someone could predict TT times. Someone
could -- I just think the prediction wouldn't be as good as wattage and aero drag.

Quoted message said:

At Analytic Cycling he uses "effective frontal area" measured in square meters for aero drag. I'm
curious if you can calculate reasonably well this based on rider height and weight. He says
typical values are .4 to .7 m2. Last night I did a little test on a perfectly flat road, 26.5 mph,
1 minute, back and forth once each. There was a very slight wind and my efa numbers according to
his calculations were .75 and .90 (tailwind and headwind).

Estimating CdA from open air field tests is kinda difficult, though Allen Lim claims to be able to
do it. I'm skeptical, though of course I'm not privy to his data. I went through a little exercise
on the wattage list a few months ago with a guy who had some power data and was trying to estimate
CdA from it. Even though the data were moderately clean the coefficient of variation was kinda high.
Andy has done some stuff on trying to calculate CdA from a rider in a velodrome (I *think* it might
have been an outdoor velodrome rather than an indoor one).

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

"Andy Coggan" <[email hidden]> wrote in message

news:<[email hidden]>...

Quoted message said:

Quoted message said:

As the accompanying article indicates, I fixed the upper and lower ends

of

Quoted message said:

Quoted message said:

each range based on known power outputs of world champions/world record holders and untrained
individuals, respectively. The values in between

were

Quoted message said:

Quoted message said:

spread equally (i.e., linear relationship assumed), simply because at present there's not enough
data to justify doing otherwise. Remember,

the

Quoted message said:

Quoted message said:

purpose of the tables is to compare relative ability across different exercise durations
reflecting different physiological characteristics,

not

Quoted message said:

Quoted message said:

to attempt to assign or describe riders of different categories (strip

the

Quoted message said:

Quoted message said:

category guidelines from the tables, and they would be just as useful).

This

Quoted message said:

Quoted message said:

is why I didn't assume a normal distribution and spread the values that way - that might (or
might not - no one at present has the data to say

for

Quoted message said:

Quoted message said:

sure) better reflect reality, but has the disadvantage of squeezing everything together toward
the middle, making anyone who isn't well

above or

Quoted message said:

Quoted message said:

well below average appear to be an "all rounder".

Fair enough! But it would still be extremely useful to see some actual profiles for a variety of
types of riders. Otherwise one doesn't really know how alarmed (or satisfied) to be with a "/"
profile or a "\" or a "^". Or, how much slope matters?

I suppose we could put up some examples on the website.

Quoted message said:

Also, I assume the Untrained subjects represented characteristics of the general population and
were not in any way predisposed to the 5-second or 20-minute abilities.

No reason to think that the values are skewed in any way.

Quoted message said:

I wonder what percentage of people have the genetics (or whatever) to excel at the 5-second versus
the 20-minute performance, if trained. It seems to me, **in theory**, that would impact whether or
not the columns are linear.

It seems to me that if you assume that there's a big enough population base fighting it out to be
"top dog" in each speciality, then that assures that the scales will align properly (the most
important part). As I indicated before, logically one would expect that each scale should *not* be
linear (thus addressing your point about the extremes), but instead be normally distributed. I
could, for example, have assumed that world champion/world record performance is, say, 5 SDs above
the mean, and the lowest level of the untrained 5 SDs below, or something like that - however, that
doesn't alter the comparison across scales, and has the disadvantage of crowding together all values
in the middle.

To reiterate: the category guidelines accompanying the scales are just rough approximations, i.e.,
most riders of a certain category can generate the specified power for the specified duration -
however, that doesn't mean that all can, nor does it mean that just because you can generate a
certain amount of power that you should be a certain category (although it does indicate that you
have that potential).

One more caveat: from a theoretical perspective, the values should really be
W/kg^0.67, not W/kg...however, I think that such allometric scaling might be a little beyond the
average person's grasp, at least w/o extensive explanation.

Andy Coggan

"Mark Fennell" <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

"Kurgan Gringioni" <[email hidden]> wrote in

message news:<[email hidden]>...

Quoted message said:

Quoted message said:

If you work on your aero position, you should be able to get much better than riders in your
category.

The reason is that getting more aero is not easy to quantify, unless one

has

Quoted message said:

Quoted message said:

access to a wind tunnel. People will spend $5 per gram shaving weight

off

Quoted message said:

Quoted message said:

their bike, yet not spend any time at all improving their time trial position, where the
relative benefits are much higher.

I agree with this concept, but I'll never tt anywhere close to guys like Thurlow Rogers or Eric
Sternlict in SoCal. I'd be pleased to add a mph or two though.

All it takes, like anything else in cycling, is work. Stretching, a lot of it, so that your lower
back and hamstrings are long enough to make power in the low position. One day/week of interval
training in the position. Gradual upgrading of the tt bike on all 1st order components (ones which
break the wind). And constant tweaking of the position, timing differences on a long straight
windless descent (coasting down, poor man's wind tunnel).

Quoted message said:

Quoted message said:

I don't know if you watched the pro1/2 Socal races after your races (I'm assuming you're
master), but notice how Paolinetti was always forearms

on

Quoted message said:

Quoted message said:

the handlebar in breakaways, even through turns? He had that figured

out.

Quoted message said:

In fact I was on Jamie's team in 1997 and had the pleasure of watching (from the comfort of the
peloton) him win the Manhattan Beach GP in a break. You are right; he slices through the air
really well.

Damn, you were on that team? Email me in private if you wish, I've got an apocryphal thing to relate
from that year.

The reason I thought you were a Master is because of the Longo Visalia story you related.

"Robert Chung" <[email hidden]> wrote in message "]news:[email hidden]...

Quoted message said:

Andy Coggan said:

"Robert Chung" <[email hidden]> wrote

Quoted message said:

Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
doesn't. Because of the way these things scale, W/kg is a good indicator of climbing ability,
but total W is a good indicator of TT ability.

Except that Heil et al. have concluded that Cd also varies inversely with weight (with an
exponent of -0.45). The overall result is then as Warren indicates, i.e., even though mass per
se has little effect on the power requirement to travel at a fixed velocity on a completely flat
road, W/kg is still likely to be a better predictor of TT performance than W alone.

Cd varies inversely with mass? Hmmm, I didn't know that. Anyway, if A varies (roughly) with m^.67
and Cd varies (roughly) with m^(-.45), then CdA would vary roughly with m^(.2). A 25% difference
in mass gives a 2% difference in CdA?

You're the one that said A varies with mass^0.67, not me (remember that we're talking about frontal
area of a rider in the aero position, not simple body surface area. Be that as it may, Heil
concluded that Cd must vary with mass^-0.45 because of experimental observations that CdA varies
with mass^0.33 (or thereabouts). Depending on what other variables were entered into the multiple
regression, A in their study varied with mass^0.408 to mass^0.762....meaning that Cd had to be
inversely related to mass to get the coefficient for mass related to CdA down around one-third.

Andy Coggan

"warren" <[email hidden]> wrote in message "]news:[email hidden]...

Quoted message said:

In article said:

I said:

Andy Coggan wrote:
> "Robert Chung" <[email hidden]> wrote
>> Not quite. The "A" part of CdA varies roughly with the 2/3 power of mass, but the "Cd" part
>> doesn't. Because of the way these things scale, W/kg is a good indicator of climbing
>> ability, but total W is

a

Quoted message said:

Quoted message said:

Quoted message said:

>> good indicator of TT ability.
>
> Except that Heil et al. have concluded that Cd also varies inversely with weight (with an
> exponent of -0.45). The overall result is then

as

Quoted message said:

Quoted message said:

Quoted message said:

> Warren indicates, i.e., even though mass per se has little effect on the power requirement to
> travel at a fixed velocity on a completely flat road, W/kg is still likely to be a better
> predictor of TT performance than W alone.

I cheated by recalling an article in VeloNews from years back about two guys who claimed they
could predict TT times based on the riders' wattage output and their aero drag. Whatever happened
to those guys?

Maybe these guys? www.pkracing.com (I hope that URL is correct.)

In any case, prediction of TT times from power and aero drag can be done with tremendous
accuracy...as long as you also know the course profile and wind speed/direction accurately enough.

Quoted message said:

Quoted message said:

Quoted message said:

Cd varies inversely with mass? Hmmm, I didn't know that. Anyway, if A varies (roughly) with
m^.67 and Cd varies (roughly) with m^(-.45),

then

Quoted message said:

Quoted message said:

Quoted message said:

CdA would vary roughly with m^(.2). A 25% difference in mass gives a

2%

Quoted message said:

Quoted message said:

Quoted message said:

difference in CdA?

Goddammit, I hit the wrong key. Anyway, my calculation was off but the point should still be the
same: A 25% difference in mass gives a 5% difference in CdA?

At Analytic Cycling he uses "effective frontal area" measured in square meters for aero drag. I'm
curious if you can calculate reasonably well this based on rider height and weight. He says
typical values are .4 to .7 m2.

FWIW, that's A, not CdA...the numbers are in the right ballpark, maybe a little low. OTOH, the
default value for Cd that Tom uses (i.e., 0.5) is too low, so the product (i.e., 0.25 m^2) is in the
right ballpark.

Quoted message said:

Last night I did a little test on a perfectly flat road, 26.5 mph, 1 minute, back and forth once
each. There was a very slight wind and my efa numbers according to his calculations were .75 and
.90 (tailwind and headwind).

You *really* need to subscribe to the wattage list...

Bottom line: while *very careful* field testing can yield fairly precise numbers for CdA (i.e.,
coefficient of variation of ~2%), it still really isn't good enough to be of much use (e.g., for
choosing equipment, or even optimizing position). That's why I'm going back to the wind tunnel this
winter...anybody want to join me?

Andy Coggan