Precisely what I was getting at: why do the "82" and "108" data differ so markedly, when all other evidence points to there being a relatively small difference between the two wheels?

A couple of thoughts come to my little pea brain... First won't drag be a function of both wind and speed? If so, I think we need a three dimensional graph. iow, the best wheel at 24 mph might not be the best wheel at 26 mph. But, I'm speculating...

Second, I'm not sure that it's valid to add the rider's drag to the drag of the bike and the wheels. My guess is that the wheel tests we've seen are done in "clean" wind; not the turbulence that is churned by the rider's legs and the airflow over the back. It also seems like front and rear tests would be necessary.

Finally, it seems like all of this could vary significantly among individuals. So, we'll likely never be able to find the "right" answer. So what this is, is a way of saying, "Mine's better than yours", when the only mine's better that counts is the results!

Did I just concoct a long winded version of HTFU? ;-))

Meaning that at lower yaw angles which is more of a front on view of what would be a wider rim (Jet C2) would yield more aero drag. Yeah that would make sense. Question is though whether the lower RR would negate that deficit and as the wind tends to be more times at yaw angles not at directly face-on 0 yaw then wouldn't the overall be a faster wheel? Just curious.

Thanks
Andy

Well...that depends on the tire, I think. AFM did tests on low Crr tires (Bonty RXLPro 23 and VF Record 20) and found little, if any, appreciable difference in Crr for a given tire on varying rim widths. Based on what Al shared below, I'm not seeing where the >10% reduction in Crr is since his data implies ~2-3% at most:

---

http://bikeblather.blogspot.com/

When was the last time you rode into a wind that had zero yaw?

FWIW, last year we tested a Zipp 999 wheelset against a competitor's set of wheels on my wife's bike with her aboard, and the results were consistent with what you'd predict based on wheel-only tests.

Thanks for weighing in, Dr. Coggan. I value your input.

While this is clearly a selling point of the bicycle that you tested (the P3c in track mode), I don't know if your data is representative of the different configurations we're likely to see around here with respect to rear wheel gap and the overall design of the rear wheel cutout. As Cervelo states - no cutout is better than a bad cutout.

Chris

Hello,

I have a few points to bring up. First, to the person who said "if the rider has X drag and the wheel has Y drag, then the system has X+Y drag", this is actually completely false. Aerodynamics is very complex, and interactions are very important. We could consider the front wheel alone (in isolation from the drag of the rider/bike system) only for supersonic flow... if you have that situation, then you are probably not needing to worry about shaving a couple extra seconds off to win.

Anyways, with regards to the pictures and data that were posted, I have several concerns:

1. The coordinate system of the force measurements is not specified. I have seen work done where the drag was reported in wind tunnel axes, not the bike axes. (When I ride my bike at least, I try to keep it pointing in the direction that it is moving.)
2. There are no error bars on that data. How significant are those fluctuations in drag with yaw angle?
3. The set up is not very clear in the picture and it is not well explained. The massive metal cylinders that make up the stand that you can partially see in the picture are worrisome. We don't even know if they were "into" or "away" from the wind. In either case, they are large enough and sufficiently blunt that they might cause systematic measurement errors.
4. Grams of drag is entirely meaningless. Ignoring all of the other issues, those values would be extremely good for 100 mph wind and extremely bad for 10 mph wind. The numbers could be a) normalized using the relative headwind velocity (in the frame of the bike, ideally) to get a drag coefficient area, or b) the wind speed could be reported. The former is actually more meaningful.
5. As I previously mentioned, a wheel alone test is really only meaningful for the case where you crash within sight of the line and your wheel travels to the finish by itself. There needs to be at least a bike there to make the data meaningful.

Bottom line is that the data, as presented, is not trustworthy. There is not even sufficient information available to evaluate the merits of the test.

Regards.

That picture of the wheel is from the HED display at Eurobike, NOT from the windtunnel test...

Once upon a time I was involved with developing aircraft, and I can tell you that yes, the real world and the tunnel are very different.


Most of the things I see on ST related to aerodynamics have me scratching my head, which is why I rarely post on the subject.

---

--------------------------------------------------
Yeah, it's a great bike but the engine needs work.

 
Actually it isn't completely false, as others have pointed out, with bicycles it does happen to work out pretty close to X+Y

Anyway it was a silly simplification to rebut a silly argument someone else had made, which you either didn't notice or didn't care, because you wanted to be like that guy who raises his hand in college, not to ask a question, but to point out how much he knows.

Thanks, for letting us know that aerodynamics is complex.

Also, fyi, the wind speed and normalization used is noted earlier in the thread.

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

And they use wind tunnels to develop aircraft too, even though an aircraft would be far far more sensitive to variations between the smoothe regular airflow of a tunnel than the turbulent air of the sky, they get useful data from it.

yet we are to believe a wind tunnel is meaningless in evaluating bike wheels!

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

I'm sorry my post came across to you as it did, but I believe you have misinterpreted it. I have some experience doing wind tunnel testing on this equipment, and enough to know that the X+Y simplification does not really work out all that well.

As to the comment about the stand, that is good to know. I'm glad that stand wasn't involved. I should have looked at it more closely. At any rate, the point that the stand can really influence the data still stands.

the wheel stand is tared out of the final numbers.

---

Andy Tetmeyer (I work at HED)

Hi Andy,

Thanks for the reply. Part of my curiosity was the actual geometry of the rig. Is there an actual paper on the test by any chance?

Ok, gotta say...wow.

1. The data was measured giving drag along the translational axis, not the wind tunnel axis. To give data otherwise would be meaningless (Nimble's weird "Total Drag" numbers). No reputable company would do otherwise, and Hed is a reputable company. If you'd been paying attention to previous aero tests and discussions you'd already know this.
2. Consistency of testing is pretty impressive with wheels on multiple runs as long as they are made in the same direction (i.e. going from 0 to 30 degrees and back rather than 30 degrees to 0 and back to 30). There's some argument as to whether you should start the sweep at 30 or at 0...Zipp advocates starting at 30 iirc. Also iirc the repeatability is within a couple of grams for non-rider stuff. It's tough to get a rider to be in exactly the same position for hours on end.
3. Give me a break. That is the DISPLAY AT A TRADE SHOW. Does it even look like a wind tunnel? Why would they put two wheels next to each other like that? Yeesh.
4. Grams of drag at a given speed is how you calculate watts. It's simple physics, I'm not going to give you a physics lesson to figure out how to calculate it. It just so turns out that 30mph (or 25mph for some tests) is almost perfectly scalable down to 10mph because 30mph is pretty darn slow in terms of aerodynamics. You just aren't going to see dramatic shifts in airflow patterns at these speeds. They test at 30mph because the forces are high enough to stay well away from the tolerances of the measurement equipment.
5. This is a ridiculous statement. You are essentially saying that the wheel's aerodynamics is irrelevant unless it's not on your bike. There are thousands of hours of windtunnel tests (some from posters in this thread) that show your statement is 100% completely wrong.

And regarding X wheel and rider Y = X+Y, in the case of a front wheel this is reasonably correct. See Andy Coggan's post above. We all know that the rider's legs, bike frame, etc all have a big effect on the performance of the rear wheel, but not a lot on the front wheel. Ideally you'd take your wheel/tire/bike/self combination into a windtunnel and spend $3000 to get all of it tested and your position optimized. Since I don't have that much money I do field tests and look at data like this to make intelligent decisions about my wheels.

---

---

Where have I missed the testing data on the Jet 90 C2 and the Hed Jet Disc? I thought these were promised but have not, to my knowledge, materialized.

.

---

Bob C.

The "science" on any matter can never be settled until every possible variable is taken into account.

I must admit that I have only been to Slowtwitch a few times, and I do not pay attention to all of the threads about aerodynamics. I have seen a few, but

To some of your points:
1. Yes, bike axis is the correct way to do it. Not everyone actually does it that way. It's a good question to ask.
2. The odd hysteresis issue you bring up is worrisome.
3. I already said that I should have looked closer at the picture. As an explanation, I was extremely tired last night. Obviously, I should have looked closer.
4. No. Giving the force without saying what speed it was calculated at is meaningless. You can give an Cd or CdA, and it will scale to other (reasonable) speeds. I don't see anywhere that it says 30 mph, but then, I wasn't there, and perhaps you were.
5. No. Read the statement again. You understood it wrong.

Right, and in case you all don't realize it, I am not saying that the data is wrong. What I am saying is that there is not enough information there to make a compelling case for the accuracy of the data. I am quite curious if there is a paper which addresses these issues, because then an informed look at the data can be made. I was legitimately interested in finding out the answers to those questions, but this seems to be very closed to such inquiry.

Gosh, I wonder what makes aircraft and bicycles so different:

http://www.humankinetics.com/...U3r8V2m67Tjay3JK8nj2

(Note: reprints available upon request.)

To expand on that: the only case/place where there seems to be evidence of significant bike-rider interaction is between/aft of the rider's legs. For example, the Trimble monocoque frame (which, like the new Fuji, was quite wide at the level of the rear brake) tested quite well without a rider but poorly with a rider, and beam bikes seem to perform quite well (although not necessarily better than well-designed double-triangle frames). OTOH, when testing aerobars as part of Project 96 Jim Martin found a very high correlation (i.e., R = ~0.9) between measurements made with and without a rider, no one (to my knowledge, anyway) has been able to demonstrate that the performance of a wheel such as the HED3 is really dependent upon the spacing of the fork blades, etc. Thus, the belief that one should ignore wind tunnel data from bicycles or parts of bicycles tested without a rider is, in my opinion, rather myopic.

It is? I asked about it...but I don't see any reply anywhere.

---

http://bikeblather.blogspot.com/