So speaking of questions about the results....

Back to the rolling resistance issue: I'm still wondering if there couldn't be some differences in rolling resistance between the two setups. Forget about friction for a second--what about the potential for reduced hysteresis losses resulting from vibration damping of the frame? Something else I'm not considering?? I'll say in advance that as soon as you start talking about secondary springs, my head starts hurting.

I realize this is a stretch, but I figure it might be worth considering. Thoughts?

Many thanks.

"...as soon as you start talking about secondary springs, my head starts hurting."


This is one of those concepts that's really simple when you see it in action.

Link two springs in series, one with a lower resistance to elongation than the other. The spring with the higher level of resistance will not begin to elongate until the spring with the lower level has elongated fully, or to the point that it's resistance to elongation surpasses that of the other spring.


Vibration damping in regard to hysteresis losses strikes me as a not terribly promising line of enquiry. Wouldn't you need to see some type of significant system-level resonant frequency issues in order for damping to be much of a factor?


.

---

Tech writer/support on this here site. FIST school instructor and certified bike fitter. Formerly at Diamondback Bikes, LeMond Fitness, FSA, TiCycles, etc.
Coaching and bike fit - http://source-e.net/ Cyclocross blog - https://crosssports.net/ BJJ instruction - https://ballardbjj.com/

As long as we're talking about ROTs, at the speeds we're talking about a change in CdA of .01 has roughly the same effect as a change in Crr of .001. Tom is estimating a change in CdA of a bit more than .02 so, ballpark, you'd need a change in Crr of .002. That's huge, especially since he used the same wheels and tires. Unmeasured frame losses aren't the explanation for the difference.

If the frame geometries are not identical when comparing two frames, does it matter?


P2K


P3C

Well, my question was if changes in Crr could be responsible for a portion of the change--not all of it. That said, I'll be the first to admit it's unlikely. I still think it's worth asking though since Crr was assumed.

Gary, I'm not sure if I'm following the question...isn't the point that the frames aren't the same? In other words, the rider and his position (relative to the BB) along with the wheels and bars are duplicated. So the only thing that IS different is the frame, right?

---

http://bikeblather.blogspot.com/

Well, Crr wasn't exactly pulled out of a hat. Knowing the true elevation delta over the laps puts pretty tight bounds on it.

I think he means different tubing dimensions (namely length) may give different damping characteristics. For the frame to make a difference to rolling resistance you would have to be claiming that it acted like full suspension - keeping the tyres in contact with the road more securely.

Well, I did read the protocol. Here are three excerpts I think revealing to my point.

1. the title: "Indirect estimation of CdA using a power meter" Key word, indirect.
2. from page 3. "how good of an estimate of CdA is it possible to get?
using usual approach, not very good at all
using indirect approach, not bad at all" key word, not bad. He did not seem to choose, good, very good, or excellent. So, it is a not bad estimate.
3. from page 7. "wind, if present, should be consistent in direction and speed". A condition that hardly ever occurs.

Anyhow, it looks to be a useful technique. But, there are enough "problems" that it could hardly be used as a substitute for wind tunnel testing. Hence, the difference in the drag numbers between these two conditions, which seem exceptionally large, must be somewhat suspect as representing reality.

---

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

"I get that part (kinda), but the question I have is does the second spring elongate less (I'm guessing so) or not at all??"


Not at all, at least until a threshold level is reached.

In practice, with imperfect springs, at the transition point, you can actually see a short period where there is alternation in elongation between the two springs, but this is really, really minor...


"...a long time ago I had seen a paper which discussed frame materials and efficiency. I don't remember much about it (I think it was a manufacturer sponsored, non-peer reviewed deal), and it made some suggestions that frame material affected efficiency to the ground. However, I don't remember the specific mechanisms by which they suggested those changes occurred."

I would like to see this if you ever dig it up. My understanding of these issues is informed almost entirely by my academic background in musical instrument design and construction, and when I see terms like "efficiency" and "damping" in the same sentence, I'm seeing antagonist forces...

I would like to refer this line of thought to your steel-tube bike discussion on another thread, and harken back to a conversation I once had with a frame builder of some repute. He believed that more (in his words "excessively") rigid frames were less efficient to ride, as they were less effective as a mechanism to absorb road vibrations, and were not demonstrably superior in efficiently transferring rider energy to the drive train. In his words, "any energy lost as a result of frame material or design stiffness would manifest as heat, and would be easy to measure if it occurred at a significant level."

Earlier in the thread I seem to remember that the frame that "felt" fastest was actually slower... Hmmmm...



.

---

Tech writer/support on this here site. FIST school instructor and certified bike fitter. Formerly at Diamondback Bikes, LeMond Fitness, FSA, TiCycles, etc.
Coaching and bike fit - http://source-e.net/ Cyclocross blog - https://crosssports.net/ BJJ instruction - https://ballardbjj.com/

Namely the head tube geometries, and the shape of the P3C head tube was my question. In this case it is taller than the P2K, yet although taller, possibly more aerodynamic?

"Anyhow, blahblahblah..."



...of course, since you are completely unfamiliar with the literature on the subject, you haven't read the field-test validation paper (which I think Andy actually contributed to) which demonstrated (rather conclusively) the ability to field test at a level of confidence approximating that of the wind tunnel...


...and, as has already been mentioned on this thread, some of the people actually posting on this thread are demonstrating reproducible test results with a higher level of resolution than at least one of the academic wind tunnels can, and they are using this protocol.


So, please... go away.


.

---

Tech writer/support on this here site. FIST school instructor and certified bike fitter. Formerly at Diamondback Bikes, LeMond Fitness, FSA, TiCycles, etc.
Coaching and bike fit - http://source-e.net/ Cyclocross blog - https://crosssports.net/ BJJ instruction - https://ballardbjj.com/

Hey, if you believe it to be true, so be it. I look forward to this result being independently reproduced with the same result. When that is done I would have more confidence in this representing reality. Otherwise, I still have my skepticism there is this much difference from this simple change.

I look forward to seeing how he does in his time-trial to see if he sees it in his race. If he doesn't see the improvements that he should see based upon this result I look forward to hearing the explanations as to why.

---

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

Actually, I can't. I'm just letting the tailwind push me along ;)

And by the way, that's not lard - I have my patch kit and CO2 stuffed in a ziplock down the front of my skinsuit. ;)


The CdA calculations are from several TTs on different nights with different wind conditions. I am using the Chung method as modified by Ron Ruff for "instantaneous apparent headwind". I get very similar results using the standard Chung method calculating for "virtual elevation change" The TT course is a very flat out and back, but with a changing road bearing with several turns, so it gets a variety of wind angles.

I calculate air density based on the readings from a nearby weather station. I also temperature-correct an assumed base Crr for that course and the tire/tube combination from AFM's dataset.

Any more guesses before I reveal the results?

Rik

Aah...I see. Yes, I'm sure the headtube area is one of the "difference makers", especially since it's a "leading edge". Let's not forget also that the forks are different on the 2 bikes as well. That may play a part in the differences MEASURED (that's for Frank <grin>) between these 2 setups.

In any case, faster is faster, right? :-)

---

http://bikeblather.blogspot.com/

Well, as Robert pointed out, having a known elevation helps to tightly bound the Crr. If the Crr is "off" then when choosing a CdA that "levels" the elevation plot, the elevation difference will be either too big or too small. Don't forget that the 2 main drag sources vary differently with velocity...this allows one to "pry" the two values apart by having adequate speed variations throughout the laps. The "halfpipe" configuration helps in getting that speed variation and I also make a point of varying power (and thus speed) across each lap.

In any case, even if a change in the Crr component was measured and all the difference wasn't attributable to drag reductions, it still would have been measured to be faster, right? Who's to complain about that? :-)

---

http://bikeblather.blogspot.com/

You are forgetting that this IS an independent verification of the differences Andy measured on virtually identical frames (P2T vs. P3C Track).

---

http://bikeblather.blogspot.com/

OK, that's for explaining that in a way the this music degree equipped bike racer can understand! And yeah, whether it's Crr or drag reductions, faster is faster <tm>, they'd just scale a little differently.

Either way, I guess I should probably be happy I joined the arms race already....

Unless, of course, wind, or other uncontrolled variables, make the "faster" an illusion.

I would feel much better about the reliablity and accuracy of such a result if there were say 4 different people who said, lets do this experiment. We will each do this independently using this protocol and then compare results. If they all get the same result then I think we can feel very comfortable the protocol is pretty good and reliable. If they each get substantially different results, despite having their own internal consistency, then who is correct? The fact that one person has an internal consistency is not particularly good evidence the result is accurate.

I think the protocol is very interesting and a lot of work went into coming up with it. However, I haven't seen the work that proves to me that a result such as this can be trusted to represent reality, especially in view of the magnitude of the difference.

People here agonize over the accuracy of the various power meters yet I haven't seen one person here (except perhaps me) ask what the accuracy of this technique for determining aerodynamic drag supposedly would be. Why not?

---

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

That is you Eric. Thanks for digging the frameset out of your garage and selling it to me!

The answer is 0.0228 m^2.

We only have two entries in my little contest and RChung's "Price is Right"-style guess is closest. So I guess he has "smart guy" bragging rights over Tom A for the time being! Though he gets some style-points deducted.

OK, now that I have the data in front of me:

Photo #1. I have data for three 10-mile TTs on three separate nights from July-August 2007 for which I did a Chung method CdA analysis. Equipment the same as I described earlier and as shown in the photo, except for the 3rd TT where I had a water bottle cage mounted on the downtube (no bottle). CdA calculated at 0.2716, 0.2733, and 0.2735 . Wind speeds and direction, temperatures, and air density varied greatly on the three nights (I corrected for air density as well as the temperature effect on Crr), so I'm pretty happy that the results varied as little as they did.

Photo #2. I have data so far this this year for three 10-mile TTs from March to April on the same course as 2007. Equipment the same as I described earlier and as shown in the photo, except for the 3rd one where I had to use a 23 tire because I broke my 20. Wind speeds and direction, temperatures, and air density varied greatly on the three nights. CdA calculated at 0.2497, 0.2504, and 0.2499.

Average of the CdAs in the 1st set: 0.2728 m^2
Average of the CdAs in the 2nd set: 0.2500 m^2
Difference: 0.0228 m^2

That's the difference I found between an "aero" frameset and frameset that is actually aero. And based on the tube widths and aspect ratios, the "aero" Fuji Aloha 1.0/Motobecane Nemesis is more aero than a number of other "aero" framesets

What does this mean in terms of results? You know, that actually-going-faster part? Well, there just happen to be two TTs I did with almost identical air density and power output. Makes for an easy comparison:

TT#2 from the 1st set: June 20, 2007. Calculated air density = 1.167 kg/m^3; averaged 329 watts; time: 22:33
TT#3 from the 2nd set: May 7, 2008. Calculated air density = 1.168 kg/m^3; averaged 328 watts; time: 22:02

31 seconds over 10 miles = 1.93 seconds/km = 0.42 seconds/km per 0.005 m^2 change in CdA. Hey, somebody ought to write a rule of thumb ;)

Rik

Very cool.

so what do I win for guessing .205 correctly?

I really am sorry about the collarbone or I could have posted my data...

shame having these two identically equipped tt bikes sitting here not being ridden...


-g

---

greg
www.wattagetraining.com

Frank, this is not one of those: teach physics to Frank threads so I'm not going to respond to any questions here. This method has been tried and criticized mostly elsewhere but it's a tad more complicated than explaining why a bike speeds up going round the corners in a velodrome.

It's not a training method, nor is anyone selling anything here ... so ...

Anyhow you missed Marco Pinotti's 4th place in the Giro TT. How come?