lunchtime pondering

[flydhest]

There is a good deal of pondering, posturing, and postulating about frames and materials and such. In describing peoples favorite rides, they say things like, "I push on the pedal and the bike just goes." you know, how "efficient" the bike is at energy transfer or how a flexy frame "robs" them of power.

Would the following be feasible? Calibrate a Powertap and an SRM power meter independently so that they are both precise and accurate(1),(2). Swap the two powermeters across bikes. Presumably, if they are both accurate and precise, then the power measured by the Powertap should be strictly less than the power measured by the SRM(3). Measuring the gap between these ought to give some idea of how efficient frames are, no? Have people done this? It would be an interesting addition to discussion of the merits of frame designs and such (including material choice) and would not suffer from someone having to describe how a ride felt. Moreover, it should be immune from tire pressure and the like. What would be interesting is if it yields different results across bikes for different cadences, for example, or "climbing styles" whatever that is.

If one can make the power meters precise but not accurate, then one could still look at the differences across bikes, but not be able to transform the numbers into some absolute loss of power. Nevertheless, to the extent that the meters were precise, but not accurate, differencing should reduce or eliminate the bias in measurement to the extent that it is linear. Thoughts?

(1) I picked Powertap and SRM as examples because one is at the crank and the other is at the hub.
(2)Precision means that the measured values tend to lie very close to each other for the same underlying value of the variable measured. Accuracy means that the measured value tends to be very close to the underlying value of the variable measured. They are not the same thing.
(3) Given that there is never a perfectly efficient machine, there must be some loss of power, however small, between the crank and the wheel, hence a strict inequality.

[BarryG]

My hunch is that those devices are incapable of the degree of accuracy that would be required to reliably & consistently quantify the differences between frames.

[flydhest]

Barry,

I imagine you have a point, but one need only extend the sample to let the random error washout in averaging, I should think. Over the course of, say 2 hours per bike, it seems like some sort of averaging should be able to take care of most of the difficulty.

(aside: I spend most of my life trying to come up with cute ways to let data tell you more that it really wants to.)

[bcm119]

It would be an interesting experiment, assuming the precision was high enough.

But, it would miss the other frame variablility, which is how wheel deflection affects forward motion. I think Mr Kirk explained his theory on this topic once, but I can't find it in the archives. Basically he explained how a wheel follows a sine curve and frame flexibility affected how efficiently the rear wheel tracked under pedaling force. It was an interesting topic...

[flydhest]

It would be an interesting experiment, assuming the precision was high enough.

But, it would miss the other frame variablility, which is how wheel deflection affects forward motion. I think Mr Kirk explained his theory on this topic once, but I can't find it in the archives. Basically he explained how a wheel follows a sine curve and frame flexibility affected how efficiently the rear wheel tracked under pedaling force. It was an interesting topic...

yeah, part of me was also thinking of a trainer with a powermeter in it to get that aspect, but the holding of the frame fixed seemed undesireable, so you'd have to get a powermeter attached to rollers. You would have to figure out how much is lost as heat to friction between the tire and the roller and if this compromises your inference about "real world" (1) riding. As I said, just some pondering, but I think it would be an interesting step to see just what sort of data one can get out and what you could make of it.

(1) "real world": a hypothetical construct that I use as a reference point but rarely visit.

[Marco]

A couple of thoughts:

I am in an industry where we often times use data to tell the story that we want it to tell; a variation on the theme of your data comment.

Your experiment sounds like decent science to me; problem is, once we have the results 70% of what is discussed on this board is no longer eligible for long winded, passionate commentary. In fact, a statistical survey would probably indicate that most of the threads that have generated the most (and most heated) responses were the ones that revolve around the types of subjects that would be scientifically addressed by the data in your study. For the love of this forum community Flydhest, don't do it !!! :)

[Climb01742]

another route would be to avoid making statements about how a frame rides. speaking personally, i've usually come to regret making such statements. such statements tend to start fruitless debates. as people ask you to back up what you've said. which, quite honestly, you can't...other than to say, well, that's what it felt like to me. more and more on this forum i'm coming to the conclusion that keeping one's opinion private makes more sense. or at least is less of a headache. so much of what we feel or experience on a bike isn't quantifiable. but that doesn't make it any less true for the rider who experienced it. i've almost never convinced a skeptic. nor has a skeptic ever shaken my belief in what i felt. which pretty defines fruitless, i think. i mean no offense by any of this. just a weary realization. :rolleyes:

[Tom]

When I started that thread on bubbles etc. I edited the post before sending to remove gradient percents, what speed I considered fast into the wind... because too much specific info lost the intent of what I was saying. Besides, on any given day who knows what some measurement to the tenth actually means?

Iris DeMent: "Let the Mystery Be."

[flydhest]

climb,

I think my post (given my . . . personality? can I call it that?) may have come across wrong. In all honesty, I put even money on such a test result being able to confirm some of what you (if I may use you as an example) talk about liking or not in a frame. I think of it sort of like how I approach everything in life. All that matters fundamentally is if I like something or not, but understading how and why things work makes it easier for me to figure out what I'm likely to like before having to drop change on it. Plus, just thinking about this kind of crap turns my crank.

Opinions are valid, I agree.

marco: pppppbbbbbbbtttttttttttttt :cool:

[Too Tall]

Seth, you can't match different powermeters without using some sort of avg.'ing and you'd lose your "fine" data. The "ergomo" BB powermeter has / had the best shot at this as it's future itteration has a lightning fast sampling rate...in theory...but fuggetaboutit the product is not catching on from what I can tell.

Maybe what you want is the SRM "online torque" analyzer.
http://www.srm.de/old/torque.html

I'd be very kieen to any assessment done blind eg. a powermeter that is taped over so the rider has no data to see and he/she rides both bikes recording his/her assessment(s). It's def. going to be a take it for what it's worth deal 'cause variables are sooooo huge...but interesting and certainly will enhance a review.

Below are some things to monge esp. Fly who has the math chops to understand.

F.W.I.I.W. I move my SRM between two bikes and I use the same "numbers" on both eg. Functional Threshold etc. and see no observable diff. between two very different bikes using nearly identical positions and weighing almost 6 pounds diff.
========================================
http://www.midweekclub.com/powerFAQ.htm

For current power, the PowerTap Standard displays only the power calculated every 1.26 seconds, and when set to record every 2.52 seconds, discards values calculated at 1.26 seconds, i.e., it records every other value without averaging. The Pro model, on the other hand, can display average over the last 1.26, 2.52, 5.04, 10.08, or 30.24 seconds for the current power value, but like the Standard, it records the instantaneous value at the selected recording interval, so for instance, when at the 10.08 second recording interval, every 8th value is stored, and the other 7 are discarded. Some have noted that displayed memory is often a couple Watts higher than what is downloaded. In fact, the “raw,” recorded data represents is the most accurate and unaltered information, coming directly from the hub. The reason the display is slightly off is that it uses lower-precision arithmetic, rounds improperly, or computes running averages using a method that is prone to accumulated errors or truncation. These corners are cut because memory and CPU computing power are at a premium.



The SRM averages torque during each pedal revolution, then multiplies the result by the average angular velocity (cadence) during the revolution, then makes calculations according to the specified interval:



0.1 second – all completed revolutions are averaged, if a revolution hasn’t been completed then the previous data is sampled again.



1 second – all completed revolutions in the previous second are averaged. One revolution will be sampled in the first sample, two revs will be used in the second sample, etc.



10 second – all completed revolutions are averaged; at 90 rpm this would mean the average of the previous 15 pedal revolutions.



Instantaneous power is estimated using the torque analysis function, which samples torque at 200 Hz, and in this way, SRM claims there are no artifacts in its power calculations, however, this is only an estimation of instantaneous power, because we don’t know instantaneous crank speed, and speed variations, though slight, do occur while pedaling. The crank torque and angular velocity that are combined to calculate power aren’t necessarily time-aligned properly, which can be an issue if cadence is changing rapidly.

[Big Dan]

Fly, even if you get numerical evidence that one material is more efficient than the other one, it comes down to feel and utility... .

Let's say ....the neck of a Strat vs an Ibanez Wizard II neck..... :cool:

The Ibanez let's you play faster, the Strat brings out the soul..... :D

[Climb01742]

fly,
the rider/bike/terrain combination is so complex, and so idiosyncratically individual, i'm not sure you could test it to understand it. even if you could, how applicable, and scaleable, could the findings be? my guess is it might be angels on the head of pin territory.

[flydhest]

TT: yeah, the fine data, I reckon, are less important to me, but it would be necessary to have a reading in two different places. Of course, for all I know, the potential power loss might be 1/10000000 of the measurement error.

Big Dan: I'm with you baby . . . did you see the article in the Times today about the patent and trademark hall of fame?

[Too Tall]

Big Dan...nice. But give props to Les Paul k'? He was inducted into the National Inventors Hall of fame today.

[Big Dan]

Fly & TT..I know, I know.....Les Paul's , Gibson's , Epi's are all great..
I need a Gibson ES175, but I'm afraid to leave it alone with my 3 Strats.... :eek:

What would Leo say???

Back to bikes. A test like that would help time trialist and racers, still there's something that is impossible to quantify... :D

[ada@prorider.or]

as i know a little about who srm work
you cannot do this
because even if you the most accuratte srm system
if you know how it works it cannot de done simply you have first to calibrate the srm on the bike
and with this you loose also the other data that you have measure't
you have to compare all exactly bikes and componenets
i have research this years ago for a world know componenent factory
cees

[zap]

Doesn't the only "loss" of power on a bicycle occur in the chain/teeth. I've heard that chain drive has 98% efficiency, but I wonder if fresh bicycle chains/cogs might be even better.

As far as measuring "snap", I think one would also need a very high sampling rate at both ends in order to measure how quickly power is transfered to the rear tire.

But I wonder if this would be necessary. It seems that stiffer frames tested by Tour are commonly referred to as "snappy" frames by a good number of riders.

Of course only one component is tested (frame) and subject to debate by some, and ride impressions by cyclists. So we are basically back to square one. But still, something to think about.

I'm sure folks like Dave K, erichie & other builders have far more insight based on what was built vs rider feedback.

[ada@prorider.or]

Doesn't the only "loss" of power on a bicycle occur in the chain/teeth. I've heard that chain drive has 98% efficiency, but I wonder if fresh bicycle chains/cogs might be even better.

As far as measuring "snap", I think one would also need a very high sampling rate at both ends in order to measure how quickly power is transfered to the rear tire.

But I wonder if this would be necessary. It seems that stiffer frames tested by Tour are commonly referred to as "snappy" frames by a good number of riders.

Of course only one component is tested (frame) and subject to debate by some, and ride impressions by cyclists. So we are basically back to square one. But still, something to think about.

I'm sure folks like Dave K, erichie & other builders have far more insight based on what was built vs rider feedback.

well there is also loss between cogs and crank and tyres
and wheel
if you compare with srm
also you have to know its nearly impossible to build 2 frame's exactly the same
also each fram has it own i guess you called eigenvalue
thats its own hz frequency
thats srm use also to calibrate

[djg]

Experimental design: (a) folks are wondering whether the objective measurement tools are up to their assigned tasks, (b) I'm not sure that there's adequate control regarding the pedaler (or nearly enough data to say "who cares"?), and (c) I'm not sure you've controlled at all for types of bikes versus tokens.

With my bike, I just push on the pedals and it goes. Except when it's on the workstand. Then, I push on the pedals and the pedals go, but the bike stays. I am confident in these generalizations and I believe that they can be applied to many other bikes.

[zap]

Cees, you are absolutly correct. Don't know what I was thinking when we discussed tire rolling resistance many times :rolleyes:

Time to get out of here and ride :banana:

[keno]

Seems to me there is an input problem in your experiment. How do you insure that the power being generated on each bike is identical at all times? Absent a Perfyrider, I'm not sure that you develop any meaningful data. BTW, you can use both the SRM and Powertap simultaneously on each of the bikes and hope that there are no BB or other problems that might result in corrupt data one bike to the other.

Inputting equal power to the cranks on each bike and measuring output via SRM and Powertap might be interesting but does not address the critical issues of how the frames respond on hills and sprints or powering up on the flat.

Just a thought, but what about a device for measuring power loss in the frame, if you know what I mean. All of those vibrations, heat and absence of light must have a cost. Knowing about what my new ride will cost there better be no power loss cost at all.

Have I missed something, the elusive obvious or something more or less so?

keno

[flydhest]

keno,

I reckon I was thinking one could easily keep track of the difference between the two measures or the ratio. That way, the fact that power input changes isn't crucial. The need for some structure, like linearity or log-linearity might have to be imposed.

The sampling rate, however, may be the biggest problem. Temporal aggregation issues are things I work with a lot, and I know they can phuck your inference.

[keno]

I think that I get your picture.

Let's assume, in order to isolate elements, that you had BB and hub power measurement devices that could either produce data in analog form or digital form measured at an extremely high sampling rate. Put feasibility aside. Here are a few questions rolling around in my gourd.

1. How do you deal with differing cadences and stroke characteristics? Two riders might produce identical wattages in very different ways, one spinning and the other mashing? Certain frames might be suitable for one type of riding styles and not the other. Or, depending upon power levels and pedaling style, frame desirability might reverse at certain points because of harmonics within the frames.

2. Is power loss in the chain/gearing system a variable in relation to the characteristics of the power put to it, not only in terms of absolute power but also in terms of dp/dt as well as cadence (as to the last, I would guess that higher cadence has some cost in terms of greater loss through the heat generated by friction)?

3. Let's say there is a theoretical "no loss" frame. Would it be one that would be rideable for any length of time?

As far as 1 and 2 are concerned, I'm thinking that there are more variables here than are immediately obvious, at least to me. I like, in theory, what you are proposing to do. I'm not there at this point assuming that you could get all of the data you like over time.

keno

[hypnospin]

once you quantify this we have to go back in time and let sean kelly know the vitus robbed him of sprint power. and let those match sprinters on the super pista steel, they should ease up so as not to flex the stays so durn much.

if you apply power in an unrefined manner the strong rider can hop both wheels up 'n down, ghost shift a cog or two 'n back again.

if there is not the wattage sufficient there and you ride a mega alu none is lost or gained.

with such a dynamic application of force as provided through the human brain (or not) via the legs, i suspect any measurement would have a long way to go from being other than an excersise in numerical extrapolation...

and shouldn't we be sneaking off to ride at lunch?





[QUOTE=flydhest]There is a good deal of pondering, posturing, and postulating about frames and materials and such. In describing peoples favorite rides, they say things like, "I push on the pedal and the bike just goes." you know, how "efficient" the bike is at energy transfer or how a flexy frame "robs" them of power.

Would the following be feasible? Calibrate a Powertap and an SRM power meter..et.al.

[Too Tall]

Fly said "phuck" oooooo I'm tellin'.

Ken's comments are excellent and Cees' you are catching up ;)

In the end we seek to quantify something that is subjective and to that purpose these watts device(s) may just do what Fly wants but only as a rough measure, gut check whatever. It won't measure the (bike) parts. It will measure the riders abilities as effected by the (bike) parts. Thoughts?

That, because ultimately the rider is affected by the sum of all parts: bike, tyres, harmonics, road conditions etc. etc. When the rider gets off the bike after say a few hrs. on two diff. bikes with powermeters and says "Sheet mahn dat's one fast freakin' bike, I had NO CHAIN" and we look at the numbers: avg. watts, HR, cadence, time spent at/above/below functional threshold...we begin to see a picture when we look at this rider on various bikes.

It is useful, it is not terribly scientific and we LIKE the notion.

Hey, here is a "send up". Suppose Cees loans me a set of his wheels and Serotta tools up a cool custom Ottrott for me to "evaluate" ok? You with me? Do you think my perceived effort on the bike will be a wee bit lower? Heck yes. Is it quantifiable? Hmmm. *Cees, PM me for my home address. **Ben, you have my home phone give me a call after lunch ;)

This is a cool and great thing reading reviews of products by folks we know. Their styles and abilities add meaning and depth to such reviews and I'd love to see some human data collected with it just to see what it means.

There are human data issues we do not ordinarily consider. For instance, if you are very light hearted, rested and happy as a clam your riding experience will reflect that. The body is an amazing thing, lower stress hormones and a mind ready to accomplish the business of having a GREAT experience net results will no doubt be more watts at lower HR and overall lower training stress on the bike. These are things that count!!!

Ooops, I am rambling. But than, I just commuted 20 miles to work and a guy named "Brent" told me "Dood, don't tell me you are commuting from too far away...cause you are riding soooo fast" ;) It's going to be a good day.

[csb]

i think at this point i shall leave the forum _ my lunch time
pondering involves x-ray glasses + thread count.

[dirtdigger88]

i think at this point i shall leave the forum _ my lunch time
pondering involves x-ray glasses + thread count.

Im sorry I didnt hear you. . .did you say something?

Jason