'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 )

method of evaluating performance parts?(15 posts)

method of evaluating performance parts?DougSloan
Nov 30, 2003 1:42 PM
Would it be possible, even helpful, to evaluate performance parts by expressing a value in terms of watts saved or required over a "standard" part? In other words, assume a standard rider is 150 pounds, 5'8", on a 19 pound bike with box rims and 32 spokes, etc., riding at sea level, no wind, smooth road, level grade, making 200 watts. A standard for climbing could be the same, at 10% grade.

Then, for example, if you are considering buying something like Zipp 404 tubular rims, the evaluation go something like "-15 watts level, -10 watts climbing." In other words, it's like saving that amount of power under those conditions. Something like a 95 gram stem could be "-.5 watts level, -1 watts climbing." Conversely, heavy duty tires might be "+10 watts level, +20 watts climbing." (I have no idea if these numbers are even close.)

Would this be meaningful, even assuming anyone would go to the trouble of doing the analysis?

... the largest variable is the rider...Akirasho
Nov 30, 2003 3:18 PM
... suggesting that an expression in watts would only lead to confusion and anguish...

Confusion in the respect that it'd be adding yet another set of tables to those many refer to when attempting to buy performance (I'm as guilty as any).

Anguish with respect that as a human being, there are times when it all comes together and you're as textbook as the abovementioned tables... but there are also times where, for varied reasons... it ain't happenin'... which kinda leads back to confusion... cuz then you start plying through your tables and charts, HRM downloads and watts... seeking any culprit save one... the rider.

I'm guessing that there are engineers on these boards who, might be able to assemble some relative standards for some relative performance... but how would you equate that into a total package? Also, how do you get manufacturers to comply or perform independent tests that are timely, fair and accurate?

Dunno... I've tried to "weigh" several factors when making cycling purchases... but tend towards the simpler side... and as I get older... my aesthetics get the better of me (sometimes, it's better to look fast than to be fast).

Be the bike.
re: method of evaluating performance parts?lithiapark
Nov 30, 2003 3:21 PM
Doug, I think it would be possible and a lot of fun. A bike with a Power Tap hub and a well calibrated speedometer, could be used to test differences in products of all sorts for their aerodynamic benefits. On a smooth level road with little or no wind, you could measure your watt output at a given speed, say 20 mph, with the aero vs the reference piece: wheelset, fork, shoe covers-whatever. You could do the same on a long steady climb with parts of different weights. You could go to Death Valley in February and find roads like that near sea level, another useful reference point. Of course, you would need riders of different capacity re. power and flexibility. I'd be happy to test parts for the inflexible, white-haired geezers with more money than good sense category:-)
re: method of evaluating performance parts?asgelle
Nov 30, 2003 5:42 PM
This has been discussed at length at the wattage forum at which can be read and searched without subscribing. The conclusion is that except for the most closely controlled conditions (an indoor velodrome) it is unlikely that the power differences with changes in equipment will be large enough to be observed in a few tests. Riders are just not good enough at maintaining a constant speed or power. What is likely to be more useful is to use the models at
These allow you to calculate speed at a fixed power or power at a fixed speed. You can evaluate the effect of lowering aerodynamic drag, weight, or rolling resistance among other effects to see how large a change a piece of equipment is likely to produce.
Only after you get ASTM or ASME involvedKerry Irons
Nov 30, 2003 6:06 PM
The testing/professional societies take years to develop standards because it is a huge effort to come up with a test that is reproducible, meaningful, easy to conduct, hard to jigger, etc. Many tests today do not meet those requirements, so advertisers use "business physics" to calculate their numbers. ("Business physics" like "business thermodynamics" does not follow the fundamental scientific principles typically associated with terms like physics and thermodynamics.) The wind tunnel tests showing that virtually every wheel tested is the fastest shows the dangers of trying to let the manufacturer publish performance results. It would take lots of discussion from very informed people to even try to make the first version of a comprehensive test standard.
Only after you get ASTM or ASME involvedasgelle
Nov 30, 2003 6:12 PM
But there are a few independent testers getting good wheel aero data out there. Right now, Kraig Willett, is the one I would trust most.
Very helpful and very possiblegala7516
Nov 30, 2003 6:14 PM
In the November 10, issue of Velonews, there was a very subjective and comprehensive article [Black Gold Rush: Is Carbon the way to a lighter, stiffer crankset?] evaluating performance of carbon cranks.

The purpose of the article was to simulate actual riding conditions as closely as possible. "The right crank arm [driveside] was attached to a bottom-bracket that was free to rotate inside a bottom bracket shell fixed to a test bench. The driveside crankarm was then loaded at the pedal. A fixed chain wrapped around the large chainring resisted the crank's rotation (rather than fixing one crank and loading the other, which would mimic a standing condition but not riding conditions)."

The deflection of the six cranks was measured as the a 425 pound load was applied. This information allowed conclusions to be made about the stiffness. Velonews took it further by then considering weight and cost, so the index was cranks could be compared based on stiffness/wt./$ (lbs/in./g/$).

I think that laboratory conditions and well thought out tests are the only way that components can be properly evaluated. Using actual riders would cloud judgment with prejudices.

This is the first article that I have seen Velonews do like this. I would definitely like to see more.

Good question Doug.
Very helpful and very possibleasgelle
Nov 30, 2003 7:27 PM
The Velonews study does report on stiffness of the cranks measured in a realistic way, and is a useful guide on how one may choose to invest his money on cranks. However, the article does not go the extra step to show how crank stiffness impacts overall performance (we already know the impact of weight). In other words, if I'm riding at constant power, how much faster will I go on cranks with one weight/stiffness compared to cranks with another. That would allow me to decide if I want to invest in stiffer cranks or some other piece of equipment such as a lighter saddle.
Good point, but...gala7516
Nov 30, 2003 8:37 PM
Velonews should have gone the extra step and did the calculations with all else being equal how much power is saved and how much faster one can go...for reference only.

The problem is that it is going to be impossible for readers to relate that situation to themselves. Everyone's weight, power, and bike setup are totally different. The only useful figures are the ones that neglect speed and power.

A series of similar articles would allow a consumer to decide where he/she feels the money should be spend to go fast or reduce weight.
Good point, but...asgelle
Dec 1, 2003 7:30 AM
"Everyone's weight, power, and bike setup are totally different. The only useful figures are the ones that neglect speed and power."

I think this is where we may disagree, to me the only useful figures are the ones that include speed and power. It's true that everyone's weight, power, and bike set-up will be different, but I believe these variables are largely irrelevant when evaluating the effect of equipment. If the goal is to get down the road faster, the only question is does a piece of equipment let me go faster at a given power or produce more power at a given metabolic load. Only a few components try to increase power at a given load (training with power cranks for example). Most try to increase speed at a given power (more aero wheels, lighter components). To answer this type of question, we can neglect the items you mention and only look at speed given power (or power given speed). An easy example is evaluating a lighter component for speed on a steep climb. We know that on a steep climb, reducing total bike+rider weight by a given percentage will results in an increase in speed of the same percentage regardless of the factors you mention. So if we reduce the weight of the cranks by 1 lb and the bike+rider weighs 180 lbs, the rider will climb 1/180 (0.56%) faster. What does change with absolute power is the definition of "steep." The higher the power, the faster the rider and so the steeper the grade at which speed will drop low enough that the other factors besides gravity can be neglected. has user friendly models for this which makes it easy to compare speed and power for various scenarios. They also do an excellent job of presenting how the models were derived.
Good debate...gala7516
Dec 1, 2003 8:17 AM
How can you say that everyone's weight, power, and bike set-up are largely irrelevant? We are evaluating minor differences between high-end components! There needs to be some kind of controlled situation when evaluating these components. How else will you know if the component in question is making you faster? There is no way for someone to tell how much faster a component will make you.

In the Velonews example, the authors are assuming that the reader knows that a lighter stiffer crank will yield the best results. However, I think that I understand your point that a reader may want to know where to put his money. Where can one get the best bang for the buck? For example would it be more economically both financially and mechanically to buy tires, wheels, cranks, or something else? Then the next question would be which tires, wheels, or cranks? These questions could only be properly answered in with careful tests and fair comparisons.

It is an interesting debate.
Good debate...asgelle
Dec 1, 2003 9:04 AM
"There is no way for someone to tell how much faster a component will make you."

This is exactly what does. Here's an example I used in the components discussion on aero wheels.
The question was how much of a benefit do Zipp wheels provide.

We considered 2 riders one maintaining 10 mph, the other 30 mph on level ground. For an average rider with standard wheels, this would require 26.6 W at 10 mph and 409 W at 30 mph. The actual numbers would depend on the particulars such as weight, position, tire pressure, etc., but these are good for a rough estimate.

Now lets keep the power fixed and everything the same except change over to a set of Zipp wheels. These will probably be a little lighter than standard wheels but I neglected that (the change will be small and I can add it later). Most significanly, the Zipp's will drop the aerodynamic drag by about 2% (again the exact number might be a little different but we'll get a rough idea). Then the rider producing 26.8 W will increase his speed from 10 to 10.045 mph. The other rider will go from 30 to 30.20 mph. If we look at the time to cover 10 miles, the slower rider will save 16 seconds while the fasster rider saves 8.

The same analysis could be done for say tires with lower rolling resistance. We could then compare the savings from new tires with that from new wheels and decide how to spend our money.

One thing I haven't said that I should make clear is that this works as long as the changes are large enough that the trends can be seen above the uncertainty from the assumptions made. I think the above example is such a case. The example I used earlier of the choice of a lighter saddle is one where this is probably not true. The change in performance based on the difference in saddle weights might very well be smaller than changes associated with slight changes in position or comfort of one saddle versus another. The method I propose assumes power is constant. Equipment changes that might alter the ability to produce power (position changes) would not work well in this type of analysis.
there would be no measurable differencecyclopathic
Dec 1, 2003 9:04 AM
unless there's chain rub. Flexing cranks will not result into any measurable losses, period. Total losses on drivetrain is 1-2%, and most of it goes into chain and chain contact points. Second, crank as a spring and most of energy put to flex crank is returned at the end of the stroke. Truth is you'd save much more with better chainlube.
there would be no measurable differencegala7516
Dec 1, 2003 9:47 AM
So a flexible sloppy, but springy crank won't waste energy like a stiff crank?
there would be no measurable differencecyclopathic
Dec 1, 2003 10:12 AM
it will waste some energy, but it will be below measurement error threshhold. We are talking about losses less then .1wt
Now, if you have chain rub or can flex cranks enough to cause shift that would be another story.