I was not involved in this testing, but I can assure you that under mounting was done to hit the fit coordinates of Kiley's chosen dimensions.

We know those were modified slightly from his actual fit, specifically so that all of the bike could hit them.

Iirc, all bikes had the same extensions except PT, as that can only use proprietary extensions.

In order to hit his fit numbers, the Felt went with an under-mount.

---

----
@adamwfurlong

I think the reason you are getting so much crap is because in the past you have dished it out quite generously. I have worked with people that behaved in that manner in the past...they would deride others for their ideas (justly or not), and people would not cross/ go along with their ideas because it feels crappy to have someone say that you suck at what you do (even if you are right). The problem is that these people turn out to not be right, and their behavior goes by the old adage of 'the best defense is a good offense'. This may work for football, but it does not work for science. Those people that I referred to were eventually fired, and nearly sunk the company (being louder does not make you more right) They were unfortunately too arrogant and insecure to admit their mistake (and research/discovery is all about making mistakes...hopefully with a couple successes in the mix).
In your absence (the thread that explained the ST forum sanctions) I referred to you as a bully based on my online interactions with you (up to that point, most in that thread had praised your contributions). I reserve judgement on whether my previous impression is still accurate (not that this should matter); but based on your results presented thus far, and your replies to questions (some of which I will admit, are somewhat combative) I am not impressed with your scientific acumen. Im not saying that is a problem; but just call a spade a spade, and don't try to pass yourself off as an expert in this arena. I would hazard a guess that is the reason that some people here are asking for the raw data...so that they can judge the meaning of the data for themselves. Good luck this coming season, and I hope that you learned a bunch from your tunnel visit and had fun. When it stops being fun is when it is time to find something else.

Stephen J

---

I believe my local reality has been violated.
____________________________________________
Happiness = Results / (Expectations)^2

I figured as much. To accomplish and apples-to-apples comparison, is it reasonable to say the Felt was too big, then?

---

Trent Nix
Owned and operated Tri Shop
F.I.S.T. Advanced Certified Fitter | Retul Master Certified Fitter (back when those were things)

I thought SuperDave always said to get the biggest IA that would fit as it would be more aero.

I think an underneath mount, even all other things being equal, is faster because it reduces the surface area of the bike presented to the wind.

To your point, the question becomes is the goal to maximize a setup or to make accurate comparisons. It's possible that one of the other bikes might exhibit a similar quality, and a larger size with an underneath mount might have been used (but maybe not).

All of that said, I don't think they had an IA to test. I think they had a B-series bike.

---

Trent Nix
Owned and operated Tri Shop
F.I.S.T. Advanced Certified Fitter | Retul Master Certified Fitter (back when those were things)

I think since the Felt was the "control" bike, it doesn't much matter.

This is a good lesson for everyone. These guys have worked very hard to come up with as neutral a bike vs. bike test as they could, but it's simply not an easy task. Each, I'm sure, has a "day job" that takes priority over sifting through data and coming up with a palatable report, so I'm sure they're giving it all the time they can, literally, afford. I would encourage patience from everyone, though I don't disagree that, perhaps, there are more highly qualified people on this forum that could help crowdsource the data review and reporting.

Still, I say Bravo for the efforts!

---

Jim Manton / ERO Sports

A general reply

When this was initially proposed - I was all in right from the start. For PremierBike Tactical to be included with well established engineering companies that are the pinnacle of our industry - I was honored.

I did not know who exactly was involved but decided that I could manage any issues that might come up.

When Kiely said it would be at A2 (never been there and don't know them personally) I was happy. They have a very good reputation in and out of the bike industry. I have been to enough wind tunnels to know that these companies will not play games and they didn't. When Kiely said he had brought in professionals to write the protocol and monitor the test - I began to feel really good about what he might be able to pull off.

These tests were done with 7 people (with vested interests) watching every aspect of the tests from start to finish. Everyone watched everything and everyone verified all the measurements and set-ups. All the deviations were noted and I believe will be disclosed in the write-up.

Kiely did an outstanding job on each bike. His position was projected, tested and monitored during each run. We all discussed the effort level required to maintain a constant for the day and Kiely was able to do it and did it extremely well.

The guys at A2 did a stellar job and will provide 100% of the data that they provide to all client customers - the normal data output. We were all watching it in real time and made notes on outcomes. My notes match the outcomes I have seen thus far.

I don't know if the final paper is written yet or not. But I do know that this group did a great job. We all tried hard to be straight-up, to note all differences, discuss issues openly amongst each other and are committed to providing the results in the same open manner.

This was a first attempt for this kind of "community test" and I hope there are more tests like this in the future. There will be additions, changes and tweaks to make future tests better – everyone will acknowledge that.

I hope that those that find error or room for improvement share those thoughts and are a part of future efforts. Maybe you will consider sharing the burden and organize a test.

This can be a great way to take ownership of general testing by those that are most affected and care the most - this community of athletes.

I walk away a supporter of the effort.

---

Dan Kennison

facebook: @triPremierBike
http://www.PremierBike.com
http://www.PositionOneSports.com

Am I interpreting this data correctly that a setup for one bike starts at 0 yaw, then goes positive, then back to zero, then negative, then back to zero to result in 3 separate 'runs' at zero yaw for each bike (and double that for the control)? And that, within those 3 runs, the variation in CDA runs from 0.222 to 0.227?

Secondly - this is probably posted somewhere else already but I missed it or forgot, but what wheels were used in all the bikes? Did you simply take off the rotors when testing rim brake bikes?

A lot of these questions have been answered here.

ETA: That post was made before the testing, and some things did change between that post and test day. These will be fleshed out in the report.

RChung has the data. Cyclenuntz / David Bowden has the data. Another individual with analytical chops and experience in this field, who asked to remain nameless, has the data.

Please tell me where I did this. If that was the perception, I'd like to address it and clarify.

And some companies do just that. Diamondback, for instance -- they aren't testing the Andean Frame with a Corsair front end and bulky Di2 brifters against the P5 with Aduro and Magura brake levers. They remove that as a variable and test with no rider. Is that better? Maybe, maybe not.

Like you basically said, when you do something like this, you pick a protocol that's good and good enough -- it will never be perfect, because no protocol is perfect -- and explain it and release the data. That's what we are doing here.

I think we executed a good test, and I'm comfortable with the decisions we made and there's not much I would do differently if we were to do it over again.

I very highly doubt it. I would not run an undermount setup if I didn't have to. It adds so much frontal area. Which looks faster to you?





Notice that the undermount setup does not "reduce the surface area of the bike presented to the wind" at all...it simply presents more surface area to the wind.

You know what's coming next from some one right? "You have too many eyeballs on the numbers!" Lol.

Keep chuggin'. No way you're going to make everyone happy, but I really appreciate what you guys have done.

---

Jim Manton / ERO Sports

nevermind

Enjoy yourselves.

---

Trent Nix
Owned and operated Tri Shop
F.I.S.T. Advanced Certified Fitter | Retul Master Certified Fitter (back when those were things)

I happen to be more of the opinion that bikes should be tested close to how they are sold, after all, bike companies say "our bike is fastest", not "our bike is fastest after you swap out the crappy bars, brakes and test without the cables which are impossible to route cleanly anyhow".

Other people who plan to heavily customize their bike likely would think differently. As long as the data is available I am happy - thats why I chipped in a bit for this test.

Don't bow out now, I feel like you're just working up your head of steam.

You say you hope the test "succeeds" (whatever that means) but then do nothing but produce criticism. Put yourself in Kiley's shoes. What would you have changed? I'd like to hear you flesh out your protocol.

I saw your post before you edited it, btw. Not super impressed.

Thanks Jim!

Not to mention some of us have moved since getting back from the tunnel. Now if I could find my cycling kit......

Heath and I are forever grateful for all the knowledge you've shared with us over the years. Without you, Andy and a few others we'd not have been able to even fathom doing something like this. I also want to say thanks publicly to everyone who helped us with this pre-test.

There sure are a lot of arm chair QB's on this thread. I get it. I was one of those in the past. Oh why didn't you just do this or just do that. Too bad it doesn't work that way. When you do this it changes that and that and that and now you've got to change this and this and this to get everything lined up just how it was. The thing most don't realize is it's probably not better, probably not worse, just different.

From day 1 we knew we'd get criticism because it's impossible to set up 5 bikes exactly the same unless they are 5 exact same bikes. Even then you're going to be off a mm here or there.

The data sheet I saw posted isn't raw data. There are many calculations going on the get you those numbers. 99% of you guys who want to see it wouldn't understand it.

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

The one that pops to the top of my list is when you requested of this newsgroup to do an independent test of superbikes. I believe that this implies that you have the expertise to carry this out; otherwise, you would have raised the money and provided it to someone else who was an expert in the field to organize the testing...For example, if you used clinical diagnostics and claimed that the manufacturers/designers were just out to make profits, and that they did not work as advertised, would YOU be the one that tested the diagnostic devices and designed the experimentation if you were not an expert?
I will grant you that the outcome of a clinical diagnostic may be a little more serious than the outcome of the aeroness of a bicycle.

Stephen J

---

I believe my local reality has been violated.
____________________________________________
Happiness = Results / (Expectations)^2

But in reality all he is the expert of in this situation is keeping his position the same during all the runs and getting the companies on board and finding the prices. He then got the experts to write the protocol and run the tests. I don't think he ever claimed he was a data expert and many times said even asked for expert opinions of anyone wanted the data.

I think you, and many others, are giving him a hard time because he was dickish about things. But he also admits it and owns it. I'm very interested in the tests regardless because I think we all know that NO data from a test is going to be perfect across the board but this test should be the best to show the differences between a $15k bike and a $3k bike on a normal person. Then people can have a better idea of watts per dollar to spend which, to me, is always the biggest factor in buying something new.

---

Twitter - Instagram

This.

I finally had some time to look into this.
First, the data seems to be corrected for beta-square. I stand corrected.

The reported data is packaged weirdly. I assume the measured data sits in the 'RAW' and 'Q' columns.
Other observations: RAW grams is scaled with dynamic pressure 'Q' (pounds/square-feet) and translated to square-feet-CxA in wind coordinates. The CxA values are then transformed to bike/body coordinates using the YAW angle. Finally the CdA in m^2 and drag&watts at 24mph is calculated.

Rant: full grown engineers chose grams, feet, inches, miles and water head as "units". FFS.

I assume the YAW data is entered, not measured. Would be nice to know more about how the YAW of the wind vs the body is dealt with. As I see it, there are eight (8!) sources of errors here: wind flow vs tunnel geometry vs load cell vs platform rotation vs rotation measurement vs wheel struts vs wheel hubs vs fork/frame vs rider position. Yikes!
Five of those eight are 'static' and can (should!) be under control by the tunnel operator. But the wheel hubs, frame and rider position are all individual to the test specimen and affected by the test procedure.
Basically, I'm positive that the wheels, fork/frame and rider are NOT perfectly aligned with the wind flow at the indicated YAW=0deg. Question is how to 'tare' the YAW after every change. We're taking millimeters here that generate significant 'noise' in YAW.

My suggestion to deal with the varying-between-runs CdA-vs-YAW phenomenon is to make use of the measured YAW "torque" and adjust the YAW 'tare' to get a neutral left-right curve. Clear, huh?
There is a Raw Side F and a RAW Side R and my guess is that the measured total side force and YAW torque is dissolved in a front-wheel and a rear-wheel side force. If you (from the yaw sweep) find the YAW offset that produces left-right-symmetric 'torque-per-YAW', chances are that the between-runs CdA graphs will line up a lot better.

What say RChung et al?

Tagging @RChung... is this how this forum works?

On a more serious note, mad respect to you and your mates for doing this. People will pick it apart regardless, although I think pre-test it was made clear what the objectives and expectations of this were. Not sure there's much of a bone to pick with it all.