Frame Flex on GCN

[cachagua]

A spring is only a benefit if its return to original position is propelling the rider and bike forward. If your legs simply absorb that energy it won't matter...

Correct. Strain energy in the frame cannot "spring back" as acceleration.

I'm not arguing for maximum stiffness in frames, I very much prefer springier ones. But their advantage does not come from returning energy they store as deflection. When force you exert with your legs flexes the frame, the frame is held steady at the other end by its (and your) resistance to acceleration -- and when your legs lessen their pressure, that resistance stays the same. This means the returned energy will never go in that direction.

If it goes back into your legs, can you store it there and get it back again? Mammalian connective tissue has a great deal of stretch to it and is used for energy storage in a number of settings: the strides of kangaroos and of wolves involve stretching tendons and allowing them to spring back, and tendon was used in bows and catapults since Classical times. I don't exactly see how this would work for pedaling a bike, but I'm open to anybody's theory -- I'd like to understand it better.

[93KgBike]

Originally Posted by etu

Interesting new video on GCN with a nice illustration on how frame flex might or might not make a frame better.

https://www.youtube.com/watch?v=BH_AL4rxrp8

Retrogrouches everywhere are grumbling, "No sh*t."

So will the price of steel go up or down now?

[Kontact]

Originally Posted by 93KgBike

So will the price of steel go up or down now?

Carbon springs great, as does Ti - as long as the diameter isn't gigantic.

Aluminum is the wet blanket.

[ergott]

Originally Posted by cachagua

Correct. Strain energy in the frame cannot "spring back" as acceleration.

I should know better than to type replies on my phone. My typing/writing skills fall apart!

[Kontact]

Originally Posted by cachagua

Correct. Strain energy in the frame cannot "spring back" as acceleration.

I'm not arguing for maximum stiffness in frames, I very much prefer springier ones. But their advantage does not come from returning energy they store as deflection. When force you exert with your legs flexes the frame, the frame is held steady at the other end by its (and your) resistance to acceleration -- and when your legs lessen their pressure, that resistance stays the same. This means the returned energy will never go in that direction.

If it goes back into your legs, can you store it there and get it back again? Mammalian connective tissue has a great deal of stretch to it and is used for energy storage in a number of settings: the strides of kangaroos and of wolves involve stretching tendons and allowing them to spring back, and tendon was used in bows and catapults since Classical times. I don't exactly see how this would work for pedaling a bike, but I'm open to anybody's theory -- I'd like to understand it better.

I think you're treating this like a simple pogo stick situation, where the cyclist is pushing a spring, and the spring goes slack at times. But what's really happening is that the bicycle is storing energy by twisting - the chainstays go up and down and essentially get shorter. And then, without the pedaling force ever going to zero, they go to neutral and twist the other way. That wouldn't happen if we pedaled with our legs together like a rowing machine, but it happens because our input is 180° out of phase. So the stored energy of the twisted stays simply feed into the alternate pedal stroke, not because we are removing force, but because the direction of force changes.

[fa63]

After watching the video, the impression I got was not that a "more springy" frame stores and returns energy into the rider in a meaningful way, but that frame stiffness does not matter as much as people think it does.

[andrewsuzuki]

Originally Posted by cachagua

Correct. Strain energy in the frame cannot "spring back" as acceleration.

The whole GCN video implies that it actually does (although I'm not sure their test on the trainer is completing accurate for demonstrating it). Jan/BQ's been saying that it does for years. And I remember looking over some physics professor's finite element analysis showing that all flex is transferred into forward movement but I can't find it anymore...anyone have a link?

Originally Posted by Kontact

So the stored energy of the twisted stays simply feed into the alternate pedal stroke, not because we are removing force, but because the direction of force changes.

I think it might happen even sooner. The return is so quick that the frame probably "tracks" the current downstroke-ing foot as the power (and therefore lateral displacement) tapers off in the bottom half of the stroke. So when the foot reaches the 6 o'clock position, the frame is already in equilibrium.

[andrewsuzuki]

Originally Posted by fa63

After watching the video, the impression I got was not that a "more springy" frame stores and returns energy into the rider

Not exactly -- it returns energy directly into the drivetrain which becomes lossless forward movement.

Originally Posted by fa63

but that frame stiffness does not matter as much as people think it does.

indeed. stiffness=efficiency is the intuitive conclusion which is why it's so easy to market, which in turn further inflates its importance

[fa63]

I think I was editing my comment as you were quoting me; I meant to say return energy into the rider in a meaningful way.

[Mark McM]

Let's say that maybe some type of spring compliance to store and release energy during the pedal stroke can help deliver more of the rider's power to the rear wheel. But since much of the frame flex is lateral and torsional, which is off-axis from the drive torque, maybe there is a better way to do it. For example, the Interdrive crank, with built-in spring energy storage:






I wonder why these cranks never became popular? Maybe because there was no obvious power gain?

[cachagua]

The chainstays go up and down and essentially get shorter. And then, without the pedaling force ever going to zero, they go to neutral and twist the other way...

Mmmm... so the frame springing back sums with your next pedal stroke, and the combined force moves you forward more than your pedal stroke would alone?

Maybe you could flesh that out a little bit? I'm not quite seeing it yet.

[andrewsuzuki]

Originally Posted by cachagua

Mmmm... so the frame springing back sums with your next pedal stroke

Basically yes, although as I noted in #22 it might happen a bit sooner than the next pedal stroke.

Originally Posted by cachagua

and the combined force moves you forward more than your pedal stroke would alone?

No because the energy from the spring returning was already taken from the original pedal downstroke.

Originally Posted by cachagua

Maybe you could flesh that out a little bit? I'm not quite seeing it yet.

See 4:35 to 6:00 in the GCN video

[rain dogs]

There is no free energy. It defies the laws of physics. 100W in is 100W out.

Certainly, there is a line of thinking that looks very close to: "all flex represents total loss and that no smaller amount is 'coming back' into the frame", which is incorrect, but what's 'coming back' is less than if the frame didn't flex at all in the first place (at least in a lab environment)

The stiffer frame will always lose less energy to flex. Always. But it may be a difference that is immeasureably small and has no real world influence.

The other factor is that too stiff of a frame may be possible and totally uncomfortable, but bigger tires and lower pressure takes care of that easily.

[mt2u77]

Rest assured, even the stiffest frame out there still deflects and acts like a spring. It's "spring constant" is just greater. It will still store and return just as much energy-- e.g. twice as much force to move it half the distance.

I don't see how more flex is better unless it somehow acts as a mental cue to help our pedal cycles stay on point, or indirectly as enabling some other attribute such as lighter weight, better aerodynamics, comfort, etc.

Likewise, more stiffness is not better as long as it's above some design minimum. It's not worth trading off other desirable traits to get more stiffness than is necessary. How much is necessary? Depends on the rider/use, but somewhere north of stuff rubbing together that shouldn't. I'm guessing the lower bound has a lot more to do with psychology and confidence in the bike than it has to do with performance.

[Kontact]

Originally Posted by cachagua

Mmmm... so the frame springing back sums with your next pedal stroke, and the combined force moves you forward more than your pedal stroke would alone?

Maybe you could flesh that out a little bit? I'm not quite seeing it yet.

No. I'm saying that the net force delivered should be the same, but delivered in different ways because the springy frame is storing the peak torques and delivering them later in the pedal stroke. Much like Biopace.

That said:

Originally Posted by mt2u77

Rest assured, even the stiffest frame out there still deflects and acts like a spring. It's "spring constant" is just greater. It will still store and return just as much energy-- e.g. twice as much force to move it half the distance.

I don't see how more flex is better unless it somehow acts as a mental cue to help our pedal cycles stay on point, or indirectly as enabling some other attribute such as lighter weight, better aerodynamics, comfort, etc.

There may be a point when the frame resists flex to such an extent that the asymmetric forces of pedaling overwhelm the system and go somewhere else that is energy wasting - like sidewall flex or a lateral loss of traction.


I really like climbing on my old Cannondale 3.0. The short, stiff stays delivered the peak force of my standing pedal stroke directly to forward motion. However, the bike tolerated sprinting uphill less than other bikes because the tires broke traction more easily if you weren't careful. So while it felt efficient, it limited when and how much power I could put through it without wasting it. My Merlin Extralight doesn't feel like it climbs with as much authority because it feels 'soft', but I would be shocked if I was actually expending more energy, and I know the tires don't break traction in the same circumstances as the Cannondale.

Just an illustration from my experience, not saying the above is definitive. We all like the way a stiff bike responds to input, but we don't have a somatic way to measure efficiency.

And that's ignoring the chainstays' role in distributing fatiguing bumps.