I agree with Seve on the fork/post issue. The fork affects the flow over the entire bike. The fork and seatpost are part of the frame.

Aerobars, not so much. The extensions don't matter - arms cover them anyway, so might as well stay consistent. Basebar makes a huge difference in aerodynamics, but that's all out to the side, where it doesn't really interact with the rest of the frame.

It's quite common to buy a high end bike with low end components, expecting to change the aerobars, seat, and chainring, so there's no reason that those parts should have to match the bike.

You're crazy. Almost no one I know rides a "stock" bike. Everyone rides with race wheels that didn't come with the frame as most of the time finding $10k all in one place is tough, but a bike one year and race wheels the next is pretty common. Some people upgrade the bars and cranks etc. Anyone who cares about a bike past fit and color want to know what the frame alone does to performance.

This sounds like another defense of your bad, unexplained wind tunnel tests.

Ken

---

"the trick is to keep losing weight until your friends and family ask you if you've been sick. then you know you're within 10 pounds. if they start whispering to each other, wondering if you've got cancer or aids, you're within 5. when they actually do an intervention, you're at race weight." - Slowman

EXACTLY but don't you want to know what parts work best with YOUR bike??? Why just get a set of wheels or bars if you don't know the real affect it has on the rest of the bike????

To me, a complete bike is the way to test and a rider on the bike to back up the results is our final test.

If you are testing frames, great but personally I want to know what bars, stem, etc. affect the frame so I can build the fastest bike. A simple test we did was to create 3 different forks to see how they worked with the frame and the wheels. We used the fastest of the three to compliment the frame. To me, this is incredibly important.

I also tested multiple bars with the bike. I wanted to see how each bar worked with the frame and fork.

I want to know how spoked wheels and trispoke wheels affect the frame and how deep rims or shallow rims affect the results.

My goal is to test bikes, give details on the complete bike.

All that said, I also understand that testing a frame is somewhat important and offered up the Airfoil and the 4000 to be tested.

The most important data I have learned in the tunnel? We in the industry can do all we can to make an aero bike however the percentage of athletes poorly fit on a bike counter act all the aerodynamics from the bike.

So, fire away on the tests, I'm happy to present all of the Kestrel bikes that are requested.

The problem with the methodology is that most people who care about wind-tunnel data are NOT going to use the stock set-up put together by the manufacturer. Most people in this group have their own race wheels, saddles, tires, and aerobars. In that case, how the bike tests as built up by the manufacturer is irrelevant outside of the frame.

Again, disagree.

So you are telling me if I tell you that your wheels won't be as good as XYZ brand, you don't want this information?? Or if bar XYZ works better, you don't want to consider keepig the parts that came with the bike on there?

Trust me, I get it. I owned a store for 15 years. I understand customers want to make changes but in those 15 years of ownership, I would have loved to know what parts worked best with the frame.

My guess is that the consumer would too.

So you are telling me that Kestrel tests every possible wheel and aerobar in the wind tunnel and specs the ones that, in tandem with the particular frame, produce the lowest drag?

You spec bikes with non-race wheels all the time. Are you trying to say that your bikes will be faster with those wheels than a set of Zipps I might have in the garage? Give me a break.

All bars and wheels, no, but many of the most popular ones. That's the advantage of the A2, they have most of the product there and vendors also provide me with lots of product to test.

And no, of course a Mavic wheel isn't as fast as a Zipp. Point is that the Zipp wheels I used tested faster then a 404 or Hed Stinger on my bike.

Again, the idea really is to use what we learn in the tunnel to your advantage. Many of us in the industry spend hours in the tunnel learning and it's rare to see a consumer in there who has the ability to test like we do.

Even if you independently test frames, there is still the advantage of knowing what parts compliment that frame.

And with that, it's time to go enjoy Halloween

Please do enjoy Halloween!

Also, I understand that you spend a good deal of time in the wind tunnel and here on ST dealing with us. I just want you to know that it is appreciated. I am not trying to rag on you. Just trying to get to the truth.

I agree entirely with this comment. There are some components that few people seem to change and the forks are one of them.

I don't really agree with this comment though, but there is no correct way to test a bike and as such, testing should be representative of what people want to know. For me this comes down to the question of what most people, who want to see this data, buy; F&F (plus trimmings) or complete bike. If it's the first answer, then you need to standardise components.

I think that the idea of testing with a person is a really bad one though; it's just a source of inaccuracy.

edit - you didn't run your tests with a rider and then calculate him out did you?

Are we ever going to get clear answers to the questions repeatedly posed to you in the other thread?

Able to lend: medium/56cm Fuji D-6, 54 or 56cm Ketrel Airfoil Pro

---

wovebike.com | Wove on instagram

Steve makes a good point about forks. The front end of the bike is the most important aero-wise. Because some frames sport integrated/proprietary forks I revise my first comment and say let each frame manufacturer donate whatever fork they want tested (provided it's available to the public). It's the only way to fairly represent real world/market conditions.

Tom Piszkin
UCSD Masters Triathlon Coach

If you get one of these could you test it?

Ambitious project... good luck

the idea is very interesting/intriguing.

If you have any need for a small TTX (actually about a 54), or a round tube Merlin Ariel, or need a round tube roadie I don't mind lending them out for this test.
My one question is on wheel choice for the tests.
Why test with something like a 1080 on the front when very few people actually race with a 1080 on the front?
To me the more logical choices would be:
HED Stinger 60 front, 90 rear/Stinger disk
Zipp 808 front, 1080 rear/900 disk
and then maybe a comprable clincher version of each.

I also think that each bike should be tested with the seat post that comes with it and fork rather then trying to standardize those.

---

Mike Plumb, TriPower MultiSports
Professional Running, Cycling and Multisport Coaching, F.I.S.T. Certified
http://www.tripower.org

What else do you need to know?

This will be incredibly useful information for me the next time I send my bike to a race.

Wait a minute: I don't send my bike off to races, I take it to races. I ride my bike in races.

It may be academically interesting to know which bike frame is most aerodynamic, but without data that can in some formula correlate that data to the drag from the rider/bike combo, this information is specious at best, and may be totally misleading in context of selecting a frame for the real world.

---

--------------------------------------


http://www.whydoesgodhateamputees.com/

Sigh....

---

http://bikeblather.blogspot.com/

Tom-

Most people are like goldfish - nothing but very short term memory.

Kyle published such data >15 y ago.

What wheels, handlebars, etc., were used on each of the three bikes that were tested.

If the bikes were tested "as supplied to the consumer" as claimed in the figure legend, then how/why do you think it is fair to compare framesets fitted with different components (esp. different wheels)?

If the bikes were tested fitted with the same components, then why does the legend of the figure say "as supplied to the consumer"?

Depends per wind tunnel but the ones I've worked at, I've run at 100 Hz and test 10-30 seconds for equipment tests and 40-60 seconds for athlete tests. I've run as high as 10 kHz and as low as 10 Hz and the data has averaged out well (i.e. can capture everything needed at 10 Hz depending upon the design of balance). For solid balances like SDLSWT, A2, MIT, Texas A&M, UWAL, data collection as slow as 10 Hz is good. In CO, we had an air bearing based design and we collected data at a bit higher rate (100 Hz+).

I prefer testing 100Hz+ due to the intense vortex shedding that occurs off of the tubes at yaw above 10 degrees.

When we do equipment tests, we've studied how quickly we can test each yaw point (run) at each tunnel. At A2, 10 seconds has been plenty. No difference between 10, 20, 30, or 40.

Also, so everyone knows, the A2 tunnel doesn't do "sweeps" of yaw angles at this point. They move the balance, let the wind settle to a steady state condition, then begin gathering data. The data does not change based upon the direction you change yaw angles (i.e. negative to positive yaw or positive to negative yaw).

MC

---

--
Mark Cote
MITAerobike
Specialized Bicycle Components

Do you think/find that doing "fixed" yaw angles makes a difference due to not having flow attach or reattach as the yaw angle changes? Or does it all tend to come out in the wash?

Kevin

Mark, thanks for your expertise here. And spot on as to how A2 tests yaw. I will be holding off my testing until they update the tunnel mechanism to provide faster yaw testing, as well as possible yaw sweeps. Scheduled for sometime Q1 of 2010.

Dave

This was what I was wondering about. Your preference makes sense given the typical Strouhal number for shedding cylinders (of bicycle tube size and at bicycle speeds).


Sorry, I used the term "sweep" loosely...I actually meant what you are referring to above. From the pictures, it looks like the entire setup is on an electronic turntable. Is the typically coarse sampling resolution in yaw (usually every 5 deg) due to practical concerns with tunnel time, or some other technical reason?


This is actually pretty interesting. I wonder if this is due to either 1) nothing has fully/catastrophically separated at the max tested yaw, or 2) the yaw sampling resolution is too low (for an airfoil that stalls at eg; 20 deg AoA, I have seen hysteresis confined to above 15deg AoA on the back sweep). If neither the above is happening, then something else interesting is going on...

Thanks for the info! I'll admit this stuff isn't directly relevant to the bike shootout test, but it's satisfying my curiosities and I'm learning a lot here...

Chris

---

Chris Yu
Applied Technology
Specialized Bicycle Components
@chrisyuinc