standardization of mannequin position when tunnel testing. Requiring mannequin when rider interaction is significant; this would apply to handlebars, frames, and helmets.

For testing bikes/equipment, ideally test same rider, same position / setup in both places, then have a look at results plus the "average"

For wheels... we want to see the fastest tire on THAT wheel AND the competition. No more wide tires on wide rims vs. wide tires on H3's.

When testing bike vs. bike (no rider or rider) let's have the test of bikes of the closest sizes, not the actual "store box" sizes. Manufacturer sizes should be gone.

---

My Blog - http://leegoocrap.blogspot.com

As a consumer, what I prefer are shoot-off style events of commercially available products, like TomA did with Specialized. I'm not sure how that jibes with what you're doing, which sounds like providing more of an R&D service to manufacturers.

And I like shoot-off style events because I have greater confidence of consistency in protocol, calibration, etc. It's nice, if say, Zipp's wheel scores a CdA of 0.00001 and is the most aero wheel ever and kicks the shit out of a Mavic Open Pro. I still like to see a CdA or grams-drag plot overlaid with several other *good* wheels tested at the same time in the same conditions by the same people with the same protocol, same calibration parameters, etc. (and same rider(s) if rider is present).

Not sure how you're going to get all those manufacturers to agree to release their data so such a plot can be released. If all the losers take their ball and go home, we may not end up with much more than we have now.

But TomA's blog is a good start:

1) CdA estimates - over yaw/beta if wind tunnel
2) Error margins reported
3) High resolution pictures of equipment *as tested* from all angles. Detailed report on how different frames were set up to be comparable, e.g. matching contact points?
4) Repeated re-testing of control throughout

Following up on my last post...this is what I'd like!

You offer a reduced-cost service to manufacturers, but have an explicit "shoot-off" surcharge as part of that service.

Whenever the shoot-off fund gets big enough, you schedule a shoot-off day. Manufacturers are ejected from the building. You bring in someone like TomA as a neutral observer, and get loaned equipment from LBSs or riders. Then shoot everything off. Manufacturers get no say in anything.

That's a huge question. I'll have to have a think about it.

Let me clarify before this gets off track. This is not testing specifically for consumers; we have another avenue for that. We're presenting an opportunity for manufacturers to have access to reduced cost aero testing for their products. The testing will be completely independent. The results will be theirs to keep and use as they like. Whether or not they ever release them to the public is up to them. Much of the testing is simply R&D, not finished products, and will never be seen by the general public.

What we're looking to do is not only provide this access, but also to standardize test protocols. Establish a set of rules everyone will be expected to live by whether they test with us or not. For the consumer, this helps you trust the data when it is released, but it isn't otherwise for you.

---

Jim Manton / ERO Sports

So you're essentially trying to standardise the criteria against which equipment and positions are judged. This has wider implications than simply having comparable data for different manufacturers or testers (should they choose to divulge it) but rather, if adopted, it would direct much of the future development of the sport.
If a standard becomes accepted then companies will start designing products to maximise performance against those specific criteria. That can be positive or negative depending on how well those criteria have been selected and defined.

If done badly this could stifle innovation or might just not be widely adopted, but if you get it right and if it becomes accepted industry standard it could act to curb pointless "gimmick" innovation and instead get R&D focused on legitimate avenues of improvement.

I think you need to take a structured approach to this. Set out exactly what you aim to achieve in terms of future use and transferability of data. For example, there are a number of parameters that may be hard to replicate form one day to the next or one test facility to another such as wind tunnel correction factors, air temperature, pressure & humidity, any use of live subjects will not be adequately repeatable for equipment tests. Live subjects should only be used when testing the rider themselves....

Yes, this is the feedback I want. We have a lot of this covered already, and certainly understand and respect the implications, but we're looking for what we've missed. Excellent...thank you.

---

Jim Manton / ERO Sports

How about a standard control bike? No matter who/what/where you are testing - you run the same bike first. That way you always have the same thing to compare too.

**Just spitballing here - no idea if this would be practical or actually help.

http://www.tririg.com/...omega_whitepaper.pdf


Check.


Check.


Oh, crap. Missed out on that one. The example pic I provided had pretty crappy resolution, and was only from one angle.


Check.

For wind tunnels - take symmetric yaw sweeps out to 15 degrees in 5 degree increments and report the data. None of this weighted avg being the only reported value goofiness. If you want to include a weighted average in the reporting data you must include the formula to calculate the value.

A universal control bike would be a god send. I.e. ERO owns a bike that everyone tests against that is lent out for free to whatever testing group wants to publicly publish the data.

I had talked to A2 wind tunnel about this, and offering a "certified stamp" or something like that. I didn't want to be the guy who thought tests could be manipulated, but the last time I was there I had asked if they ever checked over the results people posted compared to what the wind tunnel reported. It'd be pretty easy to manipulate the raw data sheet and lie about numbers. I'd like to think we are all above that, but it's a possibility it could happen. We've all seen things on this forum where business owners fudge the truth here and there.

I'd also like to see the standard for wheels be without the tare of the stands removed. Running the stands by themselves and then subtracting those numbers isn't the most accurate method for reporting data (told to me by A2 tunnel). The stands react differently to a 30mm wheel vs. a full disc. A full disc wheel at high yaw could completely shelter one of the stands. . .yet when you subtract the value of both stands you are artificially making the numbers lower.

---

__________________________________________________________________________
http://www.boydcycling.com Handcrafted Revolution

I agree some kinda of universal control group that is what everything else is tested against.

---

2024: Bevoman, Galveston, Alcatraz, Marble Falls, Santa Cruz

The "scrubbing" of numbers is quite common. It's one issue we aim to "scrub" away.

Interesting thoughts on wheels. I'll talk to the guys at the San Diego Tunnel about that.

---

Jim Manton / ERO Sports

Agree with a lot of this. The control bike could be a universally-accepted spec. (round steel tubing, a defined stem/bar, 32-spoke Open Pros with Conti 4000s II 23c tires, etc.).

For the reporting, show all the yaw data; but also show the time savings for weighted yaw at 50, 40, and 30kph (the yaw weighting will be different for each, and the %s need to be an agreed standard). No tris are ridden at 50kph, and lots of IM bike legs are done at 30kph. None of the quoting watts saved at 50kph and having people thinking that's watt saving they will actually see.

---

ECMGN Therapy Silicon Valley:
Depression, Neurocognitive problems, Dementias (Testing and Evaluation), Trauma and PTSD, Traumatic Brain Injury (TBI)

it's helpful to also make sure the balance being used to generate data is not biased.

---

=================
Kraig Willett
http://www.biketechreview.com - check out our reduced report pricing
=================

This is why time savings over 40km or CdA savings would be preferable. Reporting drag savings in units of power is annoying. Results need to be reported in as non-dimensional a way as possible.

Also, applying weighting schemes probably adds value to the results, but complicates the issue of explaining to an individual the gains he may or may not realize in comparison with the scheme. I like the weighted average delta--that's something I use in testing cars, but the individual deltas from each yaw angle shouldn't disappear altogether.

---

__________________________

I tweet!

5 degree increments are a very coarse resolution if you want to plot a really useful polar.
You'll only end up with 7 data points and would likely miss many significant interference issues or stall points. I realise more points means more tunnel time or fewer test configurations but what's a reasonable compromise? This seems on the low side to me.

I'm open to variations on this theme, but if you set the bar too high for points taken people will avoid using the standard because it is too expensive to meet the expectations.

The important thing is to get the base data in as universal a formula as possible. That way the marketing people can come up with what ever numbers they want to present and consumers can either take the data as presented to them or take the base information and calculate out whatever best metric suits them.

CdA (m^2) at different yaw angles is the most fundamental unit of drag. Everything can be accurately and repeatably calculated from this set of data.

Agreed, although I think you may be putting too much faith in the marketing people to truthfully represent the results.

Playing devil's advocate here:

Ultimately, some bikes and some wheels will be faster than others. Holding every manufacturer's hand to the same flame benefits consumers but not the manufacturers, because the smart triathlete would buy the fastest (least draggy) bike that fits and is in their price range. Wheels are an even simpler decision than the bike (aside from handling and rider speed effects on yaw).

Maybe this is the price you as the manufacturer pay for discounted tunnel time?

---

__________________________

I tweet!

You seem to be misinterpreting my request. I am putting NO faith in the marketer's message. I want 'fundamental' data (CdA yaw sweep) that I can base my own decisions on and make my own metric calculations on. The marketers can simplify the data in any way they seem fit to come up with a single number that they can show as 'faster' and I will gladly ignore it as long as I have the fundamental data to make decisions on.

Ah okay. Yeah that makes sense.

---

__________________________

I tweet!

Its an interesting idea.

I would like all the tests to include both velodrome and tunnel testing to begin to set up correlation numbers.

Having a standard Mannequin would be very helpful.

The "Standard Bike" as a reference would be great but to start I think a Cervelo P5 should be included as well simply because so many bike numbers have been compared to that bike. It would also provide a (range) reference to those that have followed the bike vs P5 comparisons and begin to set up the "Standard Bikes'" numbers.

Wheels are a bit of an issue in that forks are designed to work with specific wheel profiles. If you use a "Zipp" type standard that may hurt the manufacturer that designs around a different profile. So - I would prefer that you use the wheel that is specified on the bike "as purchased". That poses a problem for those manufacturers that sell a nice bike with "training" wheels.

If you don't use a rider or simulated rider; I would suggest all the tests with the [seats removed] as that is in-many-instances a throw away item.

I would like to see left and right side sweeps at 2.5 degree increments from 0-15 as standard.

One additional issue is what to test - "bike as offered in standard purchase components vs bike as purchased and then new (saddle, wheels, tires, hydration, computer and mount etc..).

Are we testing to determine fastest ultimate bike with rider or to compare purchase value for consumers? Maybe two different protocols.

---

Dan Kennison

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

You keep using that word. I do not think it means what you think it means.

I disagree. Just one standard.
Once you start including exceptions you open the door for confusion and manipulation.
Besides, it should be utterly unnecessary. As long as the P5 is one of the bikes tested you can compare it's data with that for other bikes to your heart's content. That's the whole point - and doing it this way means the data may actually be trustworthy.

There are a lot of concerns to address, which some have already posted (i.e. regarding weightings and reported units etc.) but if I were tasked with drafting a proper standard, I'd do my best to understand:

• How companies/people "game" current tests (see SAE bHp, EPA mileage, etc., Emissions testing *cough* VW),
• That regulations can influence development in unintended ways,
• Language must be strict, clear, and unambiguous (shall, must, will...),
• "Stakeholders" must find it attractive, or the standard will not be adopted.

There would have to be a cost for a certification of compliance, and records kept such that an interested party could challenge the results and even issue fiscal penalties along with retractions for faulty testing or falsification of results. Otherwise we're playing house here. The end users (consumers and companies alike) would have to trust that the standard is meaningful else it's like a degree from a for-profit school nobody respects.

I think my vision is a little less formal. We will have set protocols for testing and reporting, will certify our own independent results, and will expect the publication of those results to me made in a specific manner. What I believe will happen is consumers will be more trusting of such testing and demand it from the manufacturers. Should an industry-wide set of agreed upon protocols for testing and reporting come out of that, so be it. We will not stray from our goal of completely independent testing. In fact, one of our rules will be not allowing manufacturers to be present during testing. In fact, that's one of the ways they'll save the most money; no travel expenses.

---

Jim Manton / ERO Sports

Jim, your aero pledge signature makes me ask the question; what's the point of coming up with a standard if the test results can't be universally applied to athletes? A manufacturer could publish data using a protocol standard but their results mean diddly until I test it on myself. Just because a standalone component tests faster than its competitor, we don't really know how it interacts in the complete setup. Too many variables, not enough $$ and time.

One additon: test, using two different speeds. Remember how the p-09 helmet performs faster (and differently) with "plug in" at a certain speed? Because, that.

I believe Cervelo have a P2K bike (with its own special case) for testing all their bikes against.

As Cervelo pretty much started the aero bike market, how about using their P2K as the benchmark? All bikes should be tested against a standardized bike and wheel combo, and use the standardized wheel, with and without a standardized mannequin. In an ideal world, there would be static and dynamic tests (mannequin). Rolling road becomes another issue, but for standard testingThey should be tested at standard wind speed and yaw angles. Since bikes aren't completely symmetrical, (+) and (-) yaw angles should be tested. Weighted average for performance should all use the same weighting calc.

All wheels should be tested on the same identical bike, and should be tested at standard wind speed and yaw angles. Weighted average for performance should all use the same weighting calc, preferably the same as for bikes. Mannequin can be static and/or dynamic as for bikes testing

Items like helmets should be.....wait for it.....tested at standard wind speed and yaw angles. Weighted average for performance should all use the same weighting calc, preferably the same as for bikes.

Standardization is the first hurdle, deciding what the standards should be is hurdle 2 and relevance to real world is the final step. There is a sweet spot of cost/repeatability/validity etc. The current ad hoc testing isn't very useful to us consumers, but ideal testing could end up being cost prohibitive.

---

TriDork

"Happiness is a myth. All you can hope for is to get laid once in a while, drunk once in a while and to eat chocolate every day"

This is exactly what I was thinking. Something testing faster than something else and that being applied across the board for many different riders at many different speeds is a claim that is virtually impossible to validate. That makes the claim useless.

Reporting in power is annoying, but it's basically the exact same problem as "time savings over 40km", it's relative.

I would like to see testing done at the speed an athlete actually buying the thing races at.

Ehhhh, not really. Time saved over 40km applies almost universally, while drag power saved is definitely different for riders of different speeds. 5W saved at 30mph is less at 20mph, but 10sec/40km saved at 50km/hr is essentially the same as at 35 km/hr.

---

__________________________

I tweet!

Jim,

I'd really appreciate if we could back up a bit to assess the benefits of in-velodrome vs wind tunnel testing.

i.e. what does one give that the other doesn't.

If this information is already out there, I am happy for a link to teach me.

Thanks for your help. I always appreciate your posts.

Randolph

When you double the velocity, drag increases fourfold. That seems highly unlikely.

When you ride 40km at 35kph instead of 50kph, you have a longer time period in which a smaller reduction in resistance (either expressed as a percentage or as absolute required power) can and does yield a greater absolute time saving.
e.g 28.8sec off 48 minutes is 1% faster, a mere 0.8% off 1:08:34 would be a saving of 33 seconds. Cervelo website has some fully worked examples?

when you half the velocity, the time savings increase

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

I cant speak to the specific engineering requirements of aero testing for bikes.

I do know a little about setting global standards for performance testing for an entire industry. I suspect that the difference between setting an international industry standard which is a mandatory requirement for businesses if their country is a signatory to a set of international regulations, and your own proposal, is that there is no means of compelling companies to work to a standard that they know may not benefit them.

I would think that the minimum standard or proposed standards are set out as a white paper for an industry wide user group to consult upon, that the proposals allow for both self certification and for you to conduct testing on companies behalf. This is subject to audit and verification but the signatories are not compelled to use you.

I think there are two different issues at work here. the first is setting a standard that can be accepted and adopted. The second is offering a deal as an independent test house.

The former requires consultation with stake holders. The second is simply a deal that you can offer people.

One should not be a prerequisite for the other. International standards are divorced from specific providers such as independent test houses, but well respected independent testers have solid businesses because they are respected for their independence.

I would think it would benefit the industry to work towards adopting a standard, but given that they can not agree on disc brakes, bottom brackets in terms of setting a standard, and certain manufacturers are so obtuse with respect to their wind tunnel testing results I would not hold my breath waiting for a standard thats accepted to be forthcoming