Frame Flex on GCN

[Mark McM]

Originally Posted by cachagua

I guess so, thanks. What I hoped to throw some light on is that there is a component of frame flex that puts the rear axle closer to the BB (since there had been some divergence of opinion on that).

Yes, there is a component of frame flex that puts the rear axle closer to the BB. But of all the flex components, this is one of the smallest. Since strain energy is inversely proportional to stiffness, this flex component will be a very small part of the total strain energy in the frame. While the GCN video may not be an entirely realistic simulation of pedalling, it clearly shows that the majority of the strain energy in the frame is from the horizontal twisting of the BB, not from the (linear) compression of the rear triangle. Anyone who wants to claim otherwise, will need to come up with direct evidence to support their argument.

[Mark McM]

Originally Posted by cachagua

Can you elaborate? Which mechanism returns strain energy in a useful way? I'm still waiting for an explanation of how a lesser force can overcome a greater one. Or if you're thinking about the release of BB sway "lifting" the pedal on its upward stroke -- that gets the pedal farther off the ground but I don't think it rotates the cranks forward. It tips the cranks back towards "vertical" (with respect to the bike) but it doesn't turn them.

There doesn't need to be lesser force overcoming a greater one for strain energy to be released to the drivetrain. Even a drive force less than the drag on the bicycle acts to propel the bike. Here's why: Pedaling drive force is periodic - it oscillates larger and smaller during the crank revolution. But as long as the average drive force equals the drag force, the bike maintains a constant average speed. During the portion of the crank revolution where the drive force exceeds the drag force, the bike accelerates. During the portions of the crank revolution where the drive force is less than the drag force, the bike decelerates. But even when the drive force is less than the drag force, it acts to propel the bike, because it causes the bike to decelerate more slowly than if there was no drive force. So anytime there is a drive force (regardless of its magnitude), energy is transferred into the drivetrain.

As you know, as the rider presses down on the pedal during the down stroke, the frame flexes (by twisting at the BB), and both the pedal and the BB drop in response. The peak downward force in fact exceeds the bicycle drag force during this phase. Towards the end of the down stroke, the rider may ease off on the down force. This will allow the frame to unflex, because the flex force then exceeds to the pedal force. The pedal's downward speed may also decelerate, as the BB starts to rise - but as long as enough force is applied to the pedal so the pedal doesn't rise as fast as the BB, the the upward motion of the BB will cause the crank to rotate forward. And the energy that caused this extra crank rotation (and thus extra energy applied to the drivetrain) came from the strain energy stored in the frame. The drivetrain force due to the release of the strain force may not exceed the drag force, but it is results in the strain energy enterring the drivetrain.

Originally Posted by cachagua

Not sure what you mean here. The tests we've been looking at, and rider performance (if we believe the riders) couldn't be more contradictory. The video and the FEA both claim you get all your pedaling energy back out of the frame (note however that they're both based on the same incorrect assumption) but riders generally agree that a stiffer frame wastes less energy, i.e. that frame flex is lost energy.

Laboratory tests (like the Damon Rinard test) have yet to find energy lost due to frame flex. And rider performance isn't measured by how the rider feels (or believes) - it is measured with speedometers and stop watches. While riders may feel that a flex bike robs them of power, it largely has not shown up on the stop watch. Plenty of races have been won on flexy frames, and lost on stiff frames.

Originally Posted by cachagua

I guess you mean, in response to the different stiffnesses of different frames? Interesting... might that mean that stiffer frames of recent times aren't really faster than old ones, but today's riders think they are because they don't know how to ride the old ones? That's actually kind of tasty... but again, we'd better make sure not to say it to Strava-guy.

Lots of things have changed over the past few decades, not just frame stiffness. The biggest difference is probably the effectiveness (and prevalence) of PEDs, as demonstrated in how race speeds dropped dramatically after the Reasoned Decision, practically to pre-EPO era speeds(despite the use of "stiffer" frames). Other factors include: Better training methods, better road surfaces, more aerodynamic bikes and clothing, lower rolling resistance tires, etc. There are too many confounding factors to attribute changes in race speeds to frame stiffness.

[benb]

Originally Posted by Mark McM

Laboratory tests (like the Damon Rinard test) have yet to find energy lost due to frame flex. And rider performance isn't measured by how the rider feels (or believes) - it is measured with speedometers and stop watches. While riders may feel that a flex bike robs them of power, it largely has not shown up on the stop watch. Plenty of races have been won on flexy frames, and lost on stiff frames.

For sure. I'd imagine there is never going to be any test that proves the flex returns energy either.

I think the thing is rider placebo is a valuable effect. If a rider feels more efficient on a stiff frame, that is good for them, if they feel better on a flexier frame, that's good for them.

No different than superstitions in baseball.

[Black Dog]

Originally Posted by Mark McM

There doesn't need to be lesser force overcoming a greater one for strain energy to be released to the drivetrain. Even a drive force less than the drag on the bicycle acts to propel the bike. Here's why: Pedaling drive force is periodic - it oscillates larger and smaller during the crank revolution. But as long as the average drive force equals the drag force, the bike maintains a constant average speed. During the portion of the crank revolution where the drive force exceeds the drag force, the bike accelerates. During the portions of the crank revolution where the drive force is less than the drag force, the bike decelerates. But even when the drive force is less than the drag force, it acts to propel the bike, because it causes the bike to decelerate more slowly than if there was no drive force. So anytime there is a drive force (regardless of its magnitude), energy is transferred into the drivetrain.

As you know, as the rider presses down on the pedal during the down stroke, the frame flexes (by twisting at the BB), and both the pedal and the BB drop in response. The peak downward force in fact exceeds the bicycle drag force during this phase. Towards the end of the down stroke, the rider may ease off on the down force. This will allow the frame to unflex, because the flex force then exceeds to the pedal force. The pedal's downward speed may also decelerate, as the BB starts to rise - but as long as enough force is applied to the pedal so the pedal doesn't rise as fast as the BB, the the upward motion of the BB will cause the crank to rotate forward. And the energy that caused this extra crank rotation (and thus extra energy applied to the drivetrain) came from the strain energy stored in the frame. The drivetrain force due to the release of the strain force may not exceed the drag force, but it is results in the strain energy enterring the drivetrain.





Laboratory tests (like the Damon Rinard test) have yet to find energy lost due to frame flex. And rider performance isn't measured by how the rider feels (or believes) - it is measured with speedometers and stop watches. While riders may feel that a flex bike robs them of power, it largely has not shown up on the stop watch. Plenty of races have been won on flexy frames, and lost on stiff frames.





Lots of things have changed over the past few decades, not just frame stiffness. The biggest difference is probably the effectiveness (and prevalence) of PEDs, as demonstrated in how race speeds dropped dramatically after the Reasoned Decision, practically to pre-EPO era speeds(despite the use of "stiffer" frames). Other factors include: Better training methods, better road surfaces, more aerodynamic bikes and clothing, lower rolling resistance tires, etc. There are too many confounding factors to attribute changes in race speeds to frame stiffness.

This sums up so much. Our little sport is rife with "feelings", "beliefs", and assumptions that are untested or even more so are completely contrary to actual data. Human perception is simply not a reliable way to asses so many things. This has been shown time and time again, yet, without any evidence, we seem so sure that we can tell exactly what is going on in complex and dynamic systems based on feel and intuition.

[Kontact]

Originally Posted by benb

For sure. I'd imagine there is never going to be any test that proves the flex returns energy either.

How is a test that shows no loss due to flex not exactly the same as proving that flex energy is returned?

It takes energy to flex a frame. And it isn't hard to measure how much energy. If that isn't a measurable loss and the energy at the wheel is the same, doesn't that prove that the flex returned its energy? What else could have made up for the loss?

[kramnnim]

Originally Posted by Mark McM

Towards the end of the down stroke, the rider may ease off on the down force. This will allow the frame to unflex, because the flex force then exceeds to the pedal force. The pedal's downward speed may also decelerate, as the BB starts to rise - but as long as enough force is applied to the pedal so the pedal doesn't rise as fast as the BB, the the upward motion of the BB will cause the crank to rotate forward.

This all makes sense to me...except I think the BB may be rising when the crank is at 6 oclock, and the rise would not help with crank rotation.

[Kontact]

Originally Posted by kramnnim

This all makes sense to me...except I think the BB may be rising when the crank is at 6 oclock, and the rise would not help with crank rotation.

You are referring to one location on a structure that is bending and twisting, and then referring to the movement of that location in a single plane. But the outside of the BB is rising, twisting, swinging out and sliding forward. Meanwhile the chainstay attached to it is untwisting, expanding, curving, rising and swinging out.

I realize I'm repeating myself, but why do these analyses alway attempt to reduce a complex three dimensional movement to just what is observed perpendicular to the plane of the chainline? Is that even vaguely realistic when much of what is happening to the bike is like the bending of a bow from the middle of downtube all the way to the rear hub?

[benb]

Originally Posted by Kontact

How is a test that shows no loss due to flex not exactly the same as proving that flex energy is returned?

It takes energy to flex a frame. And it isn't hard to measure how much energy. If that isn't a measurable loss and the energy at the wheel is the same, doesn't that prove that the flex returned its energy? What else could have made up for the loss?

I know you've been arguing about this for pages and pages. What I meant no one is going to prove that the flex doesn't make you faster, or slower.

But if it (a certain degree of flex or lack of flex) makes you feel better, ride it.

[Mark McM]

Originally Posted by kramnnim

This all makes sense to me...except I think the BB may be rising when the crank is at 6 oclock, and the rise would not help with crank rotation.

As I mentioned, when the energy is returned depends on pedaling style. In an earlier post, I included 2 diagrams of 2 different pedaling styles. The first diagram showed an example of standing while pedaling, in which the maximum R/L pedal force differential occurred at 6 o'clock (i.e., the rider is standing with all their weight on the bottom pedal at the bottom of the pedal stroke). In this case, the BB doesn't rises while the pedal is already moving upward, so it doesn't directly help to rotate the crank. Instead, it helps raise the rider's foot/leg. (If you've ever stood and sprinted on a very flexy bike, you may liken it to jumping up and down on trampoline, with bike springing back up on the back of the pedal stroke.)

In the second diagram, the maximum R/L pedal force differential occurs at about 4 o'clock, so the BB begins its rise after 4 o'clock - when the rising BB does help with crank rotation.

[Mark McM]

Originally Posted by Kontact

You are referring to one location on a structure that is bending and twisting, and then referring to the movement of that location in a single plane. But the outside of the BB is rising, twisting, swinging out and sliding forward. Meanwhile the chainstay attached to it is untwisting, expanding, curving, rising and swinging out.

I realize I'm repeating myself, but why do these analyses alway attempt to reduce a complex three dimensional movement to just what is observed perpendicular to the plane of the chainline? Is that even vaguely realistic when much of what is happening to the bike is like the bending of a bow from the middle of downtube all the way to the rear hub?

The Bike Think web page referenced earlier did analyze the pedal stroke (and the flex in the frame) in 3 dimensions. This analysis showed that by far, the largest flex (and strain energy storage) is from the twisting of the frame/BB around the longitudinal axis. The other modes of flex contributed very little to the stored strain energy - and in particular, the linear deflection between axle and BB accounted for no more than a few percent of the total strain energy.

The conclusions of the analysis can easily be confirmed by eye - the twisting of the frame/BB around the longitudinal axis is large and pronounced, and can be easily seen. At the same time, any shortening of the distance between the axle and BB can not be seen by the naked eye. Why do you keep concentrating on something which is so obviously inconsequential?

[Kontact]

Originally Posted by benb

I know you've been arguing about this for pages and pages. What I meant no one is going to prove that the flex doesn't make you faster, or slower.

But if it (a certain degree of flex or lack of flex) makes you feel better, ride it.

I think you could demonstrate the a certain amount of flex is ideal, just as science has changed how we used to think rolling resistance works. It just requires isolating the variables caused by people to be accounted for.


And I'm not really advocating for flex, just somewhat against the marketing of drive train stiffness - especially when it leads us to sacrifice ergonomics like Q or ankle clearance to provide a quality that we don't benefit from. Dealing with problematically weirdo stuff like BBRight for several years will make you want to talk about the Emperor's Clothes.

My personal experience has been that stiff bikes are fun to ride, and don't necessarily have to be unpleasant. But I did still see with those bikes that they were on the edge of ideal tire traction at times and could "waste" energy doing what they were supposed to be ideal for - sprinting up hills.

I have also had the experience of riding a flexible aluminum Vitus that just seemed slow - as if the relatively poor spring qualities of narrow aluminum tubing was maybe turning energy into the work hardening that more naturally springy materials didn't suffer from. So it is not hard to see why people love stiffness if they have had similar experiences.



So my interest isn't obsessive advocacy for one design philosophy or another, but just getting cyclists to see that the problems of frame design are actually subtle and even counter-intuitive because pedaling is not nearly as simple as a 2D textbook illustration would suggest.

It sounds like enough tests have been performed to suggest that relatively average BB flex is not less efficient, and the reasonable discussion at this point is really just about the mechanism of efficiency, rather then whether it is true or not.

It is a fun and interesting conversation.

[Kontact]

Originally Posted by Mark McM

The Bike Think web page referenced earlier did analyze the pedal stroke (and the flex in the frame) in 3 dimensions. This analysis showed that by far, the largest flex (and strain energy storage) is from the twisting of the frame/BB around the longitudinal axis. The other modes of flex contributed very little to the stored strain energy - and in particular, the linear deflection between axle and BB accounted for no more than a few percent of the total strain energy.

The conclusions of the analysis can easily be confirmed by eye - the twisting of the frame/BB around the longitudinal axis is large and pronounced, and can be easily seen. At the same time, any shortening of the distance between the axle and BB can not be seen by the naked eye. Why do you keep concentrating on something which is so obviously inconsequential?

That's all good. I actually wasn't trying to say "the energy is stored here", but just that the flex is not to be seen and measured in one plane from one point in that plane.

Like I said earlier, there are also two separate flex issues - where the strain energy is stored, and how its release actually feeds back into the drivetrain.

[kramnnim]

Originally Posted by Mark McM

In this case, the BB doesn't rises while the pedal is already moving upward, so it doesn't directly help to rotate the crank. Instead, it helps raise the rider's foot/leg. (If you've ever stood and sprinted on a very flexy bike, you may liken it to jumping up and down on trampoline, with bike springing back up on the back of the pedal stroke.)

I get what you're saying, but the rising BB is only giving back what it gave away...and it may be giving it back when you don't need it?

[Kontact]

Originally Posted by kramnnim

I get what you're saying, but the rising BB is only giving back what it gave away...and it may be giving it back when you don't need it?

If there is constant, positive force going through the drivetrain at all times that you are pedaling, when would you not benefit from it?

[kramnnim]

Originally Posted by Kontact

If there is constant, positive force going through the drivetrain at all times that you are pedaling, when would you not benefit from it?

Sigh. Take your bike outside and climb out of the saddle at 60rpm, rocking the bike with your arms. Take note of how much rotational force is going to the chain at 6 oclock.