again, step out of the science world and into the real world. do you really think a P2K is not faster than a round tube bike? therefore the corelation, and yes it is opinion.

it's always funny how the guys up front are never really obsessed with gear, while the scientists and gear heads are typically wayyyy behind........

you should be thankful it's not that easy, or we'd all be riding the same damn bike! plus i'm pretty sure Cervelo would have made this abundantly clear and public by now, since i'm pretty sure they have access to all the info as us + more......

again, good on Tom for doing this, just don't think it's the golden rule

One flaw to your logic (of many): I don't think anyone is saying that the Cervelo P3C is the only fast frame out there.

That said, there are a number of "aero" frames that are slower than steel tube frames.

Rik

Nice ad hominem - really proves your point.

I'm a scientist who's normally up front, and not obsessed with gear - one of many counterexamples that show the absurdity of your statements.

Over and Out, Eric

Huh? I expect the same rigor if one is going to "prove" something. It turns out there are some studies now with controls that show statistical significance to the results regarding PowerCranks, a lot more rigor than was involved in this so-called "study". That was the point of my joining this thread if you will go back and read my first post. The same people who are accepting this data unquestionably have been extremely critical of all of the PowerCranks studies or anecdotal reports, even to the point of calling people they don't know, liars.

Anyhow, the studies that have been done do not "prove" our claims, which are simply our expectations based upon our experience. Anyone with half a brain should know these are marketing claims. If we ever get good data that proves or, even, substantially suggests our claims should be something other than what we claim we will change the claim. Until then it stays what it is. How many customers (there are a few here) have come forward and called our claims or marketing misleading? Essentially zero.

---

--------------
Frank,
An original Ironman and the Inventor of PowerCranks

[/reply]
Nice ad hominem - really proves your point.

I'm a scientist who's normally up front, and not obsessed with gear - one of many counterexamples that show the absurdity of your statements.

Over and Out, Eric[/reply]
sorry, but i'm not really referring to amateurs here.......

and let me know when Cervelo posts that their P3C is 2min faster than the P2K, until then i guess i'll just have to assume you're smarter than they are

Look, it is entirely probable that the magnitude of the results are true. This appears to be a very valid technique for assessing CdA.

The problem is in the interpretation of the results, the attributing of all the improvement to the frame. That is where the difficulty lies.

---

--------------
Frank,
An original Ironman and the Inventor of PowerCranks

Let me ask this: suppose Tom had gone into the wind tunnel with two frames. Suppose, and this is not a far-fetched supposition, that neither Tom nor the tunnel operator had a blindfold on so they both knew which frame was which.

Now here are some of the arguments you have made:

"Because the rider wasn't blinded to the bike he was riding, there is a possibility that he would hold his head a touch lower, hunch his shoulders a little more, or otherwise make subtle improvements to position while on the "faster" frame. He may deny that he did it, or he may not even be aware that he did, but he may have done it and you have no way to control for it."

"The subtle changes wouldn't have to be that large to completely invalidate any estimate of time savings over 40k."

"While the lack of blinding may not explain all of the effect, it may explain some. If it does, then any estimate of the magnitude of difference between frames has to be taken with a pinch of salt."

and

"So how much of a pinch of salt do we need to take it with? Oh that's right, we can't accurately adjust for threats to validity - that's what makes them threats to validity. So we don't actually know with a huge degree of certainty that there was any difference between the frames."

So, given these quotes, I think you're saying that you do not believe any difference between two frames could be determined as long as a rider sat on top of them. Have I misinterpreted you?

  
Quick result update:

On a "slow" day on the 37.1K TT course (20F cooler than last year and lighter winds, which is actually a disadvantage on this course), compared to the setup I raced last year which has a CdA I measured to be .225 m^2 (Soloist with borrowed 808 front), I went 1:45 faster averaging 18 more watts (248 vs. 230)

I also improved my finish from 20th out of 38 riders in the Cat 4 division to 7th out of 34 riders. I was only 10s out of 5th (who says seconds don't count!).

I haven't had a chance to fully analyze the data, but for anyone who wishes to "play along" ;-), here's the relevant stats:

• Temp. 53F vs. 73F last year
• Air density - 1.124 kg/m3 vs. 1.08 last year
• Bar. Press. - 30.31 mb vs. 29.95 last year
• Power - 248W vs. 230W last year
• Time - 52:49 vs. 54:34 last year.
  
• Wind - similar direction (W to NW) but winds stronger on downwind legs (19K total out of 37.1K on this course) last year. This is going to be a tough one to account for...
  

So, if last year my CdA was .225 m^2...what was it this year? :-)

My "quick and dirty" estimate I made in my head (can you smell the smoke?) on the drive home (taking into account changes in Crr with the lower temps, higher air density, etc.) was that I basically was...ta da!...~2 seconds per km faster than I would have been otherwise (i.e. same frame as last year)...then again I may not have been thinking straight, so I don't quote me on that ;-)

BTW...I've got a great story about being dropped off the side of the start ramp by the so-called "holder" ~30s before my start =:-0 Ended up riding a TT with a bleeding gash on my right knee. Who would've figured I'd do a MTB race last Sunday and finish without a scratch and then I'd get injured in a TT a week later...before I even started!

I'm tired...I'm going to bed...got up at 3:50 this morning to drive to the venue...fun day.

---

http://bikeblather.blogspot.com/

Congratulations! Nice work!

Also, thanks for all your hard work collecting data and posting here with it. It would have been easy (and even potentially beneficial to you) to keep your results to yourself. I'm sure I'm not the only one who really appreciates your sharing it here. Thanks!

Very nice.

Well, for starters this old Cervelo webpage shows the P3 (aluminum) as 2:14 to 3:55 faster over 40km (depending on the body position and rider level) than a "round tube frame" in "calm conditions": http://web.archive.org/...es/cervelo-calc.html

Given this data, would you care to revisit your following earlier statement?: "Likewise, if P3C is 2 min faster than P2K, the P2K is probably a good 3 min faster than round tube. And a round tube bike with the same geometry " ain't no " 5 minutes slower than a P3C no matter how you look at it!"

Is 4 minutes (for an old P3) close enough?

The site also shows the P3 (aluminum) as anywhere from 23 to 42 seconds faster than the P2K over 40km in "calm conditions". This, of course, isn't quite the difference of around 100 seconds between frames that Tom A. found in his field testing with no wind. But it is significant. One question is: how much faster is the P3C than the old P3? Another is: are there aerodynamic differences between frames that get magnified "in the field" as compared to wind tunnel testing? Tom's and Andy's field testing show very similar results in terms of the differences between a P2/P2K frame and a P3C frame.

Rik

0.210, if:
- wind impacts the same
- rolling resistance the same
- My estimated weight of bike + rider is correct (sorry, I didn't trawl back through the thread to find it so I went for 80kg)
- and relative pacing throughout the TT was the same.

Now while I have been following the thread I actually can't recall the estimated CdA difference but on the numbers above and other assumptions noted, that's a 6.7% drop in CdA or ~ 2% faster @ 250W on a flat road windless day (air density 1.2kg/m^3) which equates to 1.8 sec/km (1.7s/km @ 300W, 1.6s/km @ 400W).

---

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio

I think I shall wait for the DEFINITIVE analysis: Frank's :-)

edit: Nice going Tom!

0.210, if:
- wind impacts the same
- rolling resistance the same
- My estimated weight of bike + rider is correct (sorry, I didn't trawl back through the thread to find it so I went for 80kg)
- and relative pacing throughout the TT was the same.

Now while I have been following the thread I actually can't recall the estimated CdA difference but on the numbers above and other assumptions noted, that's a 6.7% drop in CdA or ~ 2% faster @ 250W on a flat road windless day (air density 1.2kg/m^3) which equates to 1.8 sec/km (1.7s/km @ 300W, 1.6s/km @ 400W).[/reply] Actually I updated mass to 84.1kg having realised the numbers were in the very 1st post and easy to find.
and hence get a CdA of 0.214 or a drop of 4.9% in CdA (or CxA or whatever).

Apart from the assumptions I listed (I should have written Coefficient of rolling resistance above) - was total mass the same both times?

---

_________________________________________________________________________________
Training Plans -- Power Meter Hire -- SRM Sales Australia -- cyclecoach.com -- My Blog -- Sydney Turbo Studio

[/reply]
Well, for starters this old Cervelo webpage shows the P3 (aluminum) as 2:14 to 3:55 faster over 40km (depending on the body position and rider level) than a "round tube frame" in "calm conditions": http://web.archive.org/...es/cervelo-calc.html

Given this data, would you care to revisit your following earlier statement?: "Likewise, if P3C is 2 min faster than P2K, the P2K is probably a good 3 min faster than round tube. And a round tube bike with the same geometry " ain't no " 5 minutes slower than a P3C no matter how you look at it!"

Is 4 minutes (for an old P3) close enough?

The site also shows the P3 (aluminum) as anywhere from 23 to 42 seconds faster than the P2K over 40km in "calm conditions". This, of course, isn't quite the difference of around 100 seconds between frames that Tom A. found in his field testing with no wind. But it is significant. One question is: how much faster is the P3C than the old P3? Another is: are there aerodynamic differences between frames that get magnified "in the field" as compared to wind tunnel testing? Tom's and Andy's field testing show very similar results in terms of the differences between a P2/P2K frame and a P3C frame.

Rik[/reply]
you mean that 3 year old data? you have to wonder, gee why doesn't Cervelo make it loud and clear that their P3C is nearly 5 minutes faster than a round tube frame? if that were true, and i were the marketing guru, i think i'd be shouting that to the rafters! just like the Zipp data page comparing wheels.

yeah, so my total guess wasn't that far off time wise either! do you honestly think you'd be 5 minutes faster than a similar round tube frame of identical angles? come on

i'd be willing to bet, and again just totally guessing here, that a TT specialist could get within about 5 minutes of their total time using their standard road bike, no aerobars, no aero wheels, no aero frame, etc just riding the drops. it's beyond me that you honestly think a frame makes that much difference.....

if the frame makes that much difference, then your body position must make, what, 20-30 minutes ;-)

Definitive analysis here. Nice job.

However, changes nothing regarding the arguments as to whether this frame is 2 minutes faster, as I see them. If the frame itself is actually 2 minutes faster then everyone should see it. That data doesn't exist AFAIK.

---

--------------
Frank,
An original Ironman and the Inventor of PowerCranks

Thanks. There's a bunch of problems with the following analysis, but if you look at the top times in the Cat 3 and Cat 4 groups at this event for the last 4 years it's been held on this course, they are all about 2 minutes or more faster...granted, conditions and riders are different across the years so this sort of anecdote is problematic. In fact, the time I posted yesterday would've WON the Cat 3 group in 2005. Take all that FWIW...

I sure saw a LOT of P3Cs/P2Cs, Felt DAs, and TTXs out on the course yesterday...and one guy rockin' a Lotus and another on a Hooker Elite :-)

---

http://bikeblather.blogspot.com/

Good job Tom, that is a huge improvement

Slower day for many this year, myself included

Thanks Gary! Assuming your equipment and effort was similar to last year, any speculation about how much slower timewise the conditions were this year? ~1 minute, 1.5 minutes? Just curious...

---

http://bikeblather.blogspot.com/

how deep was the field? How does the avg. speed/time stack up per category (in relation to your own this year and last)?

So, I went to analytic cycling and plugged in the numbers you gave for both last year and this year for a flat 40k time trial. It gave predicted speeds last year and this year of 11.72 and 11.9 m/s respectively. This works out to a predicted time of:
last year 56:53
this year 56:01

So, you went 1:48 faster and analytic cycling would have predicted about 50 seconds faster with no change in CdA. So, it appears aerodynamic changes accounted for about 1 minute improvement. How much of that is due to the frame is anyone's guess. But, it certainly is less than 2 minutes (although you were on a different bike last year, not a P2K, FWIW).

And, according to them, if everything were the same but the barometric conditions, the course was about 50 seconds slower.

So, I think all we can conclude from this is about half your improvement came about because you were stronger and about half occurred because you became more aerodynamic (however that was accomplished).

Would you agree?

---

--------------
Frank,
An original Ironman and the Inventor of PowerCranks

Congrats Tom, great progress!

What I'm saying is that it would be difficult to determine a difference between frames without taking a lot of care to reduce the impact of rider and observer bias. You could exclude rider bias by doing wind tunnel testing with a blindfolded rider. Another option would be to do your field test by blindfolding the rider until they were sitting on the bike preparing to ride, and ask them not to look down at the frame while riding. This relies on the rider playing by the rules and resisting the urge to look at their bike. Excluding observer bias is more difficult because you'd have to find someone who is competent to operate and supervise a wind tunnel without knowing anything about bike equipment - that person probably doesn't exist.

Obviously these approaches would be a bit tricky to implement in the real world, but it could be done if the motivation were to have a truly valid, objective comparison between pieces of equipment. My impression of the wind tunnel test data I've seen, though, is that it is the result of testing paid for and executed by people with a clear agenda. I can't claim I've ever spent any time in a wind tunnel, so maybe I'm wrong, but aren't there usually equipment sponsor reps at the tunnel with the riders when they test gear? Here's a scenario that I imagine often happens, based on descriptions of wind tunnel testing I've read. Again, I've never been in the tunnel, so let me know if I'm way off base:

1. Team rider in the tunnel on his usual gear - drag is calculated.
2. Representative of Company X, the team's new frame sponsor, shows the rider their wonderful new frame, talks him through all of it's fantastic properties, and tells him how he'll be substantially faster on this frame. He also butters the rider up, blows smoke up his ass, etc to make the rider like him. The rider now wants to please the rep.
3. Rider gets on Company X's frame, hoping that the numbers will be better. After all, he likes the rep and doesn't want to disappoint him, he bought all the fancy tech talk about the frame, and he wants to believe that he could be faster this year and gain seconds - even minutes - on his rivals in the TT. His motivation is clearly to be faster on the new frame.
4. Perhaps without even realising it, the rider makes subtle adjustments to his position on the new frame - a little hunch here, a bit of a duck there. Lo and behold, the new frame tests out faster. Everyone's happy, handshakes all around, it's Miller time. But was the frame really faster?

The whole situation is orchestrated to produce a particular outcome and, because of human nature, it probably usually does produce that outcome. In my opinion this is a big part of the reason why a lot of the fantastic results of head-to-head wind tunnel testing don't seem to pan out in the real world.

Just because the industry standard is to ignore the potential for bias introduced by the rider and observer, it doesn't mean that it's the right thing to do.

how much bias is there though when they just put a bike up on the tare without a rider??

Tom's numbers match up fairly closely to all the numbers that I've seen for bikes without riders, so while I certainly believe there is an opportunity for bias with a rider aboard, his numbers suggest that he's holding a similar position on both bikes.

Also, as someone who's done a lot of field testing, I can tell you that particularly during high-speed runs, you end up in the position in which you normally ride. I think there's actually a lot less opportunity for the sort of bias to which you refer during a field test than in a wind tunnel, which is one of the reasons I really like field testing. My .02.