Planing

[happycampyer]

How about this experiment. Use a pedal based power meter and a hub based power meter and collect raw force data from both. In a perfectly rigid system, they should output identical strain data at any time. Even with all the imperfections of comparing 2 different data collecting mechanisms, a stiffer bike should show a stronger correlation over time than a less stiff bike. I would certainly expect to see power at the hub have a lower maximum than at the pedals. If there is some point in the pedal stroke where power higher at the hub, that would point to some mechanism within the bike absorbing and re-releasing forward impulse.

Or so I would think.

PS Don't kill this thread yet, unless I am being a complete buffoon.

This type of experiment would be a lot better than the one conducted by GCN. In the GCN test, the brake is supposed to represent the resistance of inertia, gravity, wind, tire friction, etc., but it is exaggerated since the rider’s force on the pedal doesn’t allow the bike to move at all. It makes sense that the wheel moves when the brake is released, since the return of the frame to its neutral position returns the energy stored by the pedal moving from 3:00 to the block (i.e., the chain is under tension with the rear sprocket, and since the rear wheel hasn’t been allowed to move, the frame has absorbed the energy of the pedal moving by flexing—there’s some chain stretch too, but that is minimal compared to the frame’s ability to flex). But if the frame was so stiff that the rider wasn’t able to move the pedal at all, the wheel would also have moved when the brake is released, as the pedal would now be free to move from 3:00 to the block. The GCN experiment demonstrates that at least some of that energy is not lost, but since there is no way to quantify it, and the test parameters are so exaggerated (the brake causes all of the energy to be released in one instant), it comes off as a gimmick.

Assuming one could find accurate enough power meters (which would have to have perfectly synchronized clocks, among other things), a test along the lines of JohnnyChromatic’s would at least show what’s happening throughout the pedal stroke with the energy input into the system at the crank on a stiff frame versus various flexy frames. Until then, I agree that the debate will consist more of speculation than of scientific proof.

William Threadgill

[e-RICHIE]

Until then, I agree that the debate will consist more of speculation than of scientific proof.

William Threadgill

The discussion is about opinion rather than research.
That's where the disconnect is.

[jclay]

Jan's theories are solved by group debate, apparently 'cause it sure as hell is not science.

This is more complicated, a lot more complicated and we are trying to position where frame design fits with riders abilities. Toss one human element in there and you got yourself a roaring debate.

1) Within their constraints of budget (affecting people & apparatus that can be engaged) Jan & Co have applied principles of scientific inquiry to this subject. And vastly more than anyone else involved in this thread.

2) Adding additional elements doesn't remove the strain/stress properties of a spring. It does cloud analysis which is why you try to eliminate as many other elements as possible or handle them statistically or by other methods.

But if the frame was so stiff that the rider wasn’t able to move the pedal at all, the wheel would also have moved when the brake is released, as the pedal would now be free to move from 3:00 to the block.

The GCN experiment demonstrates that at least some of that energy is not lost, but since there is no way to quantify it, and the test parameters are so exaggerated (the brake causes all of the energy to be released in one instant), it comes off as a gimmick.

Until then, I agree that the debate will consist more of speculation than of scientific proof.

William Threadgill

1) If the frame was perfectly stiff and the crank allowed to rotate after brake release then that test would be an invalid representation of the behavior at issue! I did crash analysis for a while. We used some pretty sophisticated numerical models and Job 1 was to see if the damage profiles indicated anything that would violate basic model assumptions. If anything did than you couldn't just input the data, turn the crank and say “there you go....500 mph”!

2) This isn't really splitting hairs: The strain energy isn't teally released “in an instant”. It's quick but it does take a small bit of time and that's important. While pedaling the strain energy would start to be released as pedal force diminished from maximum.

3) It is NOT scientific proof but it IS based on application of scientific inquiry and is highly suggestive. Certainly suggestive enough for reasonable support of further inquiry. Dollars to doughnuts that the big name, mass builders who support professional racing teams are looking into it.

I'll add this: no one ever said stiffer is better unless they were of the Mad Men ilk, and few if any with experience in the sport or trade ever believed them.

We've got ourselves a logical conundrum here. If there's a demographic that never believed that the stiffest frame possible (all else equal) would maximize performance of a rider (all else equal) then that demographic must have believed that some amount of flex was (and still is) beneficial to performance.

So, performance: Performance can equal a few things that quickly come to mind (all else equal!):
1) Rider covers a fixed distance in less time.
2) Rider covers more distance in a fixed time.
3) Rider doesn't fatigue as quickly and can ride longer and deliver more total energy to the drivetrain.

If the afore mentioned demographic believes that something less than the stiffest frame possible will support any of 1-3 then it seems to me that they MUST believe that some amount of frame flex is an aid to performance. I don't see a logical option to that proposition. The mechanisms may be obscure but that experienced demographic MUST believe that something beneficial to performance is occurring. They might be incorrect, as may I and Jan and everyone else in the planing camp, but they do believe that something beneficial is happening.

Ok. I've finished my breakfast break and need to finish putting my latest frame together. It's 969 tubing. I don't plane...as far as I know. Go figure.

Pics here if you care to look: Flickr

[e-RICHIE]

I'll add this: no one ever said stiffer is better unless they were of the Mad Men ilk, and few if any with experience in the sport or trade ever believed them.

We've got ourselves a logical conundrum here. If there's a demographic that never believed that the stiffest frame possible (all else equal) would maximize performance of a rider (all else equal) then that demographic must have believed that some amount of flex was (and still is) beneficial to performance.

So, performance: Performance can equal a few things that quickly come to mind (all else equal!):
1) Rider covers a fixed distance in less time.
2) Rider covers more distance in a fixed time.
3) Rider doesn't fatigue as quickly and can ride longer and deliver more total energy to the drivetrain.

If the afore mentioned demographic believes that something less than the stiffest frame possible will support any of 1-3 then it seems to me that they MUST believe that some amount of frame flex is an aid to performance. I don't see a logical option to that proposition. The mechanisms may be obscure but that experienced demographic MUST believe that something beneficial to performance is occurring. They might be incorrect, as may I and Jan and everyone else in the planing camp, but they do believe that something beneficial is happening.

Ok. I've finished my breakfast break and need to finish putting my latest frame together. It's 969 tubing. I don't plane...as far as I know. Go figure.

Regardless of how the demographic is defined, we ride bicycles not frames. That said, no one I ever followed or believed said stiffer is better. What is stiff anyway? What is less stiff? And flexible?! How can you (one) isolate the frame in question from all the parts that are hung on it? And from how you feel that day? And that forth ride up the hill from the first or the seventh? Or the tire pressure? Or spoke tension? Or? That's where I come into this.

Also, your mention of the 969 tubing. That's adding to the conundrum. I know what a 969 (.9 X .6 X .9) tube is but the masses who read these missives have fewer tools in their experiential toolboxes. The diameter, the butt lengths, the transitions, the quality of manufacture (of the tubes), the material (chromemoly, or mangmoly, or heat-treated, or seamed, or drawn...), or did the frame maker cut the correct amount from either end, or did the size and design fit he rider superbly well or simply good enough, or - I'll stop there.

There's a lot going on when you grab a bicycle and ride it. I'm good listening to other's opinions and points of view. Calling it research or science, even IF you're a researcher or a scientist, falls short. And yet on another hand, when you (one) has an agenda, or needs produce content to fill pages, that too has to be considered.

[Too Tall]

JClay, just for giggles tell me if you do or want to ride a frame tuned to your body for planing.

[Michael Gordon]

Can you make one that goes to eleven?

[e-RICHIE]

All levity aside, I'm in this for the discussion not the divisiveness. I'll grant that if we are only talking frames, said units can be measured and data recorded. That (though) doesn't take into account the wide variations in manufacture (making a frame with the same spec once a favorite is selected). If replication can be done at all, I'd suggest it's only through the methods and procedures that make the hand-made sector look quaint and Dickensian by contrast. However, I laid it (my POVs) out in the previous comments. My overarching reaction to this issue is that the myth being dispelled was only and always always a myth and never taken to heart by those for whom a bicycle is a way to earn a living. I feel similarly about the flip-side; the recent fascination with flex and related is all content and click-through driven. Experience trumps all. If you need someone to tell you what's better, or what you need, or what's more efficient (this season), stop listening (or reading) and ride your bicycle more. That is how you'll know what works for you.

[Calgary_George]

Amen.

Coming from a non-cycling background where CUMULATIVE experience matters.


George Eden

[jclay]

JClay, just for giggles tell me if you do or want to ride a frame tuned to your body for planing.

None of my frames have been built with planing as a goal. I'm not a high performance rider and I'd rather have a bit more metal to withstand the ravages of humid environments and use beyond my lifetime. That said, I plan on replacing my current fixed gear road frame with 858 TT & DT and a ST with walls thinner than 96, all in standard diameter pipes, rather than using Jan's recipe. I don't know if that will provide any snap but it will be fun to build and ride regardless.

All levity aside, I'm in this for the discussion not the divisiveness. I'll grant that if we are only talking frames, said units can be measured and data recorded. That (though) doesn't take into account the wide variations in manufacture (making a frame with the same spec once a favorite is selected). If replication can be done at all, I'd suggest it's only through the methods and procedures that make the hand-made sector look quaint and Dickensian by contrast.

I appreciate that you aren't trying to derail conversation via wisecracks and mockery.

There will always be variations in the materials used, and different sized or proportioned frames will have slightly different response characteristics. If the planing phenomenon is real then there will come a time when the mass of experimentation and experience should provide a reasonable guide as to what works and for whom (as in general size and power of a rider). For example: Assume this stuff really does aid performance. Through experimentation Jan has found a frame recipe that works better for him than other frame recipes. His mass, power and cadence during hard efforts (which seems to be the relevant portion of the flight performance envelope) could easily be determined. It seems reasonable to expect that other folks with similar mass/power/cadence values would experience similar benefits. Replicate that over many riders and after a while some general relationships and correlations will become apparent. Knowledge tends to disseminate, thankfully, so I think small builders would be able to leverage the information to the benefit of their customers. Maybe more so than the majors given the different economics involved.

Given the myriad variations (different folks, bad day, medium day, strong day, and many others) I don't tend to think this would ever become something that was tunable right down to the exact number of carbon fibers in, say, the TT....gee, I think I need one more fiber there and add three to the DT. There's too much variation in a given individual...I'd think. I tend to think it would be more like smaller racers would find that frames with these walls work best for them while heavy ones would see benefit from other wall selections. I don't think it will be like tuning a piano string against a frequency generator, or whatever piano tuners use.

My overarching reaction to this issue is that the myth being dispelled was only and always always a myth and never taken to heart by those for whom a bicycle is a way to earn a living. I feel similarly about the flip-side; the recent fascination with flex and related is all content and click-through driven. Experience trumps all. If you need someone to tell you what's better, or what you need, or what's more efficient (this season), stop listening (or reading) and ride your bicycle more. That is how you'll know what works for you.

I've never lived in the world of professional or near-pro cycling as you have so I don't know what those folks were saying. Down here at lower altitudes, but still involving some fairly serious riders, increased stiffness was generally thought to be beneficial to performance. Well, it was more than generally; it was pretty much universal and still is. And yes, some magazines apply drama to the subject in order to increase revenue but that doesn't affect whether (or not) it's real.

Experience trumps all. If you need someone to tell you what's better, or what you need, or what's more efficient (this season), stop listening (or reading) and ride your bicycle more. That is how you'll know what works for you.

There's an awful lot of truth in that but one has to leave room for innovation and learning from others; sometimes someone figures something out that others (even experts), haven't, hence coaches, exercise physiologists and such as that. Jack Northrop and his flying wings come to mind. He saw the enormous advantages decades before the industry figured it out.

[thollandpe]

How about this experiment. Use a pedal based power meter and a hub based power meter and collect raw force data from both. In a perfectly rigid system, they should output identical strain data at any time. Even with all the imperfections of comparing 2 different data collecting mechanisms, a stiffer bike should show a stronger correlation over time than a less stiff bike. I would certainly expect to see power at the hub have a lower maximum than at the pedals. If there is some point in the pedal stroke where power higher at the hub, that would point to some mechanism within the bike absorbing and re-releasing forward impulse.

I think that's a great idea, and one way to partially isolate the frame from a whole mess of other components (a.k.a. springs) that affect the bike's (a.k.a. system's) response.

That said, I'm willing to wager that none of the commercially-available power meters have anywhere near the accuracy to make this calculation.

[waterlaz]

I am very skeptical of the two power meters test. I think we can assume that the hub power meter measures what we want but I am not so sure about the pedal power meter.
I am not sure if they use the correct displacement (accelerometers?). It may not be a huge difference when you are only interested in your power output but we are talking about numbers that are like 1-2% of that.

[Too Tall]

I am very skeptical of the two power meters test. I think we can assume that the hub power meter measures what we want but I am not so sure about the pedal power meter.
I am not sure if they use the correct displacement (accelerometers?). It may not be a huge difference when you are only interested in your power output but we are talking about numbers that are like 1-2% of that.

It does not matter as long as the power meter of choice is very consistent.

[tickdoc]

What this thread needs is some charts with math and such:
url.jpg

[e-RICHIE]

Can we get a first and last name here with your comment as per the Frame Forum User Agreement.
Thank you.

What this thread needs is some charts with math and such:
url.jpg

[Mark Kelly]

I am very skeptical of the two power meters test. I think we can assume that the hub power meter measures what we want but I am not so sure about the pedal power meter.
I am not sure if they use the correct displacement (accelerometers?). It may not be a huge difference when you are only interested in your power output but we are talking about numbers that are like 1-2% of that.

Something I thought up a while ago: I already built a frame pedalling forces fatigue tester per ISO4210 part 6 which mimics pedalling and accurately* measures the force applied to the "pedals". A third gauge to the tie rod that takes the "chain" force and a method of measuring frame displacement (strain) would probably achieve the required accuracy. I was thinking of putting in the strain displacement measurement anyway so I can assess the dynamic torsional response of changes to frame tube layup.

It will have to wait a while, we start picking this week so I'm tied up for the next three months.

* The gauges I bought are instrumentation grade units, certified and calibrated to within 0.1%.

[waterlaz]

It does not matter as long as the power meter of choice is very consistent.

I believe it does. What the power meter shows is the power P + error. My concern is that error might depend on frame flex, since measuring displacement of the pedals is so tricky.

[waterlaz]

An alternative could be measuring cyclist's heart rate vs power at the rear hub. Still requires extreme care setting up the experiment though.

[Michael Gordon]

An alternative could be measuring cyclist's heart rate vs power at the rear hub. Still requires extreme care setting up the experiment though.

A good tool for this is the Ishikawa Fishbone Diagram. It is a cause and effect brainstorming tool used to identify the root causes of an issue. It can be used for trouble shooting a manufacturing defect, order processing delays, or really anything.

Ishikawa diagram - Wikipedia

While I like the idea of setting up the test using a cyclist's heartrate, I think we would find there are too many variables to have control over the experiment... Not everything that can be measured should be measured, and not everything that should be measured can be measured.

In my opinion, this is why elements like weight and stiffness (which can be quantified) outsell comfort and efficiency (which are subjective)... and are aspects that would most likely be found in slightly less stiff frames if they could be quantified.

[tbonefitz]

Hey everyone, so, since this thread lives, I asked a buddy of mine about this. He's a mechanical engineer who does material stress with F-16s and F-15's. I explained all this to him, along with the theories. He then used lots of big words and math to explain it back to me. He then almost immediately knew how to actually test these theories. He started explaining a series of experiments on a bike outfitted with strain gauges on all the tubes - multiple strain gauges per tube. He immediately picked up on e-Ritchie's assertion that the frame was merely one part of a big system, and that if we're only talking about the frame, we'd need to isolate it in a particular way. Needless to say, if anyone wants hard facts on this, the actual hard facts are within reach - for a price. He figured that his company could do it, no problem, but not for free.

He did a bunch of calculations on the back of a napkin and seemed to think through stored energy due to the flex of the frame was negligible. He did, however, completely understand the concept that a stiffer frame would feel more tiring--exactly what Jan felt. It has to do with the power smoothing of the flexible bike. Again, strain gauges, math on a napkin, and this could be verified as well.

So, who wants to pony up $10k or so to figure this out?

Keith Fitzgerald

[jclay]

Hey everyone, so, since this thread lives, I asked a buddy of mine about this. He's a mechanical engineer who does material stress with F-16s and F-15's. I explained all this to him, along with the theories. He then used lots of big words and math to explain it back to me. He then almost immediately knew how to actually test these theories. He started explaining a series of experiments on a bike outfitted with strain gauges on all the tubes - multiple strain gauges per tube. He immediately picked up on e-Ritchie'ss assertion that the frame was merely one part of a big system, and that if we're only talking about the frame, we'd need to isolate it in a particular way. Needless to say, if anyone wants hard facts on this, the actual hard facts are within reach - for a price. He figured that his company could do it, no problem, but not for free.

He did a bunch of calculations on the back of a napkin and seemed to think through stored energy due to the flex of the frame was negligible. He did, however, completely understand the concept that a stiffer frame would feel more tiring--exactly what Jan felt. It has to do with the power smoothing of the flexible bike. Again, strain gauges, math on a napkin, and this could be verified as well.

So, who wants to pony up $10k or so to figure this out?

Keith Fitzgerald

While your friend and I obviously share the same view as to the mechanism I don't think 10k would cover it. It would (I think) take multiple frames, frame strain measurement in a time domain, driven force measurement also in a time domain (all synchronized of course), lots of test runs and then data reduction. I think adding a zero would get the cost estimate closer to the mark!

Thanks for sharing your conversation. If your friend is working with airframe structures then I'd assume he's pretty conversant with aeroelasticity, or at least his work is woven in with those who are, and that sublect area is highly relevant to this question.