's Forum Archives - General

Archive Home >> General(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 )

Best Idea Ever: Standardized Bike/comp testing! hello...?(16 posts)

Best Idea Ever: Standardized Bike/comp testing! hello...?curtybirdychopper
Jan 9, 2004 11:05 AM
I had this thought the other day, since so many people talk about how they think their carbon wheels are stiff, or their alum frame is harsh, their fork is dampening, etc....

why is there no standardized bike and component testing service or institution or whatever? I'm sure I'm not the first with this idea, so why has it not been done? who would not like to compare the lateral stiffness of available wheelsets on the market, or the stiffness of frames bottome braket, the stiffness of bar/stem combos, etc? It would be a consumer reports for the cycling/mtb world. I'd subscribe to this type of service.

Btw, another there a common refernce point on the saddle from which to measure seat tube angle? maybe the middle of the saddle, or 2/3 back? thanks.

So who wants to go in with me and startup the standardized bike testing service?
Jan 9, 2004 11:10 AM

Their methods often stir up some controversy though...
Jan 9, 2004 11:14 AM
The Euro mags (esp. German ones...) seem to publish data much more willingly then the US publications...

Examples from a article on MTB tires courtesy of the WW board over on

Jan 11, 2004 12:20 PM
For exaple German 2004/1 issue of "TOUR" mogazine (Germany) is quite quantitative test results for carbon framesets (I have alredy mentioned them in post concernig Orbe frames). these results include both results of instrumental measurements for several paremeters (stiffness, lateral stiffness, lateral stiffness of fork ...) and feeling of test team (three memebers, each of them has tested each bike on the same 75km route with flat and mountain parts - up to 16% climbs and descents).
yeah, the anvil bikes thing is good, more on components though.curtybirdychopper
Jan 9, 2004 11:21 AM
i'd like to see how components stack up too. any info on this?

and, uh, the german stuff was good too. i only understood "gut, sehr gut" though. but i think thats the idea.

still, everyone has heard of consumer reports, but why no bicycling reports, a complete, comprehensive, independent testing service?
You said it: "Their methods often stir up some controversy"Kerry Irons
Jan 9, 2004 7:02 PM
There are plenty of people willing to do tests and propose testing methods. The challenge is in finding tests that are 1) meaningful, 2) sensitive enough to tell the difference between choices, 3) reproducible by any competent person/group, 4) statistically valid, 5) unambiguous, 6) etc. International and national standards groups spend years and MANY hours trying to sort out these issues in virtually every technical field. In relative size, the bike industry is pretty small, so developing the consensus around testing could be expected to take a LONG time, as standards bodies are by definition "volunteer" efforts. Don't hold your breath waiting for standardized tests that meet all of the above criteria.
Well said and alsobigrider
Jan 9, 2004 7:16 PM
besides biking being a relatively small industry the only major benefit would be for the consumer to choose a model that would be slightly more comfortable or stiffer.

If someone is going to run tests lets find out the mean time until failure so I wouldn't have had to face plant because of a Cook Bros. racing crank snapping on me.

It seems the big drivers in the testing industry is a large market or safety.
It would be a disaster for most bike companiesContinental
Jan 9, 2004 11:33 AM
If someone could develop a quantitive set of measurable parameters to define bicycle performance, and if these parameters were widely accepted and used, then all of the companies selling hype would fail like an over-stressed ultra-light handlebar. A few standard designs would be proven superior for the various bike usages and rider sizes. These generic designs would become very inexpensive, and product differentiation would be reduced to artwork and finish. Brand value and profit margin would plummet.
Frame Vibration Translation TestJPRider
Jan 9, 2004 12:00 PM
I completely agree (though also I don't know if it has been done before) on this. That's why in college me and two other guys stripped 3 different frame types (aluminum, chromo, and carbon fiber), hooked up accelerometers to certain points (headtube, near the seatpost clamp) and induced an identical, repeatable vibration at the rear dropouts. We let the signal die out and recorded all the data. It was really fascinating to see how the different materials responded, but unfortunately due to budget constraints (we asked our professor to help us get {at least borrow} a brand new ti frame, a cf frame, an aluminum frame, and an 853 frame and he said, "Sure - what do you need? $100?") we had to use old frames that we owned. We all swore if we got the chance we'd do it again in full detail and publish the reports - thus all the banter about "those frames are harsh" or "the ride quality is way smoother" could be quantified and backed up with hard data.

As far as smashing wheels, snapping forks and bars, etc., the only useful thing I've come across in a while is on Reynolds' website, and even in that they don't disclose what forks they tested besides their own, just calling them "Competitor B" or whatever.
Pretty extensive testing in EU mags....philippec
Jan 9, 2004 1:09 PM
As in Germany, 2 out of the three main cycling magazines in France do pretty extensive standardised testing on frames and components. Here's a junky digital cam pic of the test results for a frame measuring deflection at the BB, HT and reat triangle. The tests go back several years with identical protocols so it is easy to do comparisons. Sorry for the picture quality, but you get the general idea. And the protocols are published so that you can do your own tests if you have the right measuring equipment.


because you can't quantify "feel"terry b
Jan 9, 2004 1:09 PM
you could do lots of work on reliability if you could get enough MFGRS to give enough samples and I'd guess they already do this work themselves, but don't share it.

once you start analyzing the things we all carry on about - harshness, flex, lateral stiffness, damping, you cross over from quantitative to qualitative. of course, your could publish reams of vibration data and tell people "this one is more harsh at that end of the vibration spectrum" but it would largely be an academic exercise - most consumers wouldn't be smart enough to understand the data, and the ones that were would still maintain that their Starship frame was just as comfortable as their EOM16.5 frame (like I do.) it's all too personal, and I think we all agree that one person's laterally stiff wheelset is going to be the next person's noodly wheelset. I think it would be impossible to generate results data that would be broadly applicable to all the uncontrolled input variables (rider weight, rider style, road conditions, altitude, phase of the moon.) Trying to do it would require an experimental design with 10,000 cases.

Now, to the Consumer Reports model - taking cars as an example - reliability data can be valid but "comfort" and "ease of use" items are only for the tester doing the tests.

Besides, RBR is already a Consumer Reports for bikes.
our test came closeJPRider
Jan 9, 2004 1:24 PM
First off, the test that we ran used time-domain analysis, nothing to do with the vibration spectra across a range of frequencies. Secondly, what we measured was the amount of energy translated through a frame and to the components (seatpost, handlebar, etc.). Measuring the amount of energy to those points could be argued as a means to determine how much of an impact a rider will feel. And given that different materials translate energy differently, it could be argued from our results that one frame will be "softer" than another.

You're definitely right in saying that it would a cumbersome task to perform tests on literally thousands of components, but we thought it would be possible to sample 5 frames of common materials (al, ti, chromo, cf), but now manufacturers have started meshing all these materials together, along with many other variables. I'm still interested in doing this study though, for my own geeky knowledge if nothing else.

My $.02.
I am not sure your test is closebigrider
Jan 9, 2004 4:29 PM
I want to start by saying what you were doing was a great idea but here are some observations:

You can feel the difference in accelerations at different frequencies. We all know the difference in feel holding a high speed grinder verses a low speed buffer.

You have to take into account the length of time of the Force acting on the bike. An example is a semi-auto shotgun. The force exerted by the recoil is same as a single shot but a gas operated shotgun lengthens the time you feel the force. Makes the experience a whole less painful. Lets switch back to a bike. If you have to move your bike up 4 inches would you rather ride over a speed bump or a curb?

I wish you could sum up a bike testing with 5 end result numbers after running tests but there are just a million variables. I think it would take a million bucks just to figure out how to test and probably another 40k to test each model which would result in a complex report that Einstein would have to figure out while he was riding that steel bike around in circles.
difference in jargonJPRider
Jan 12, 2004 8:35 AM
I'm not sure by what you mean when you say "low speed grinder" or "high speed buffer". But I'm not sure that it matters either. Our study was to determine what frame material absorbs road impact energy better than others, using hard data. So is it a fair assumption to say that if a frame were to absorb more energy, it wouldn't matter what mode of impact (grinder or buffer) was used? I think it might be.

Also, what do you mean by "moving your bike up 4 inches"? You mean hopping it?

And just for one last point of clarification, we weren't testing bikes at all, just stripped frames.
why only euro mag testing?curtybirdychopper
Jan 9, 2004 2:07 PM
also, i bet it would be easier and also more difficult than it seems to make this bike review/testing work. i'm sure a few dozen quantitative comparisons could be generated fairly easily--this would be the easy part.

but maybe more important, the hard part might be saying that stiffness differences, for example, are perceptible (sp?) and/or make a difference to the rider. components, even frames have seemed to converge to the same design probably for marketing reasons as well as functional reasons as well. if someone makes an important advance, everyone else tries to reproduce their own version.

thinking of the car example in consumer reports, this gives me another idea...maybe it would be most useful to test entire new bike setups rather than individual components.

or maybe the consumer report application to bikes cannot be made because bike parts are much more interchangeable/simple than the suspension or steering in autos....?
whole bike testing...Matno
Jan 10, 2004 7:54 AM
That's what the American magazines do. It's incredibly annoying when they evaluate a bike that is actually really nice but give it low marks because of one or two easily replaceable components. I say, if they don't like, for example, the shifters on a given bike, they should swap them for whatever they like best and THEN run the test. They could mention that in their evaluation. Instead, what they seem to do is focus their reviews on some stupid thing that didn't really work (like the tires) and say that another bike in their "comparison" was much better, when in reality, switching the tires might switch their results entirely.