Frame Flex on GCN

[Kontact]

Originally Posted by cachagua

Even if you could hold your leg that still, it wouldn't prove what you're trying to make it prove. The problem here is the same flaw that invalidates both the video and the FEA: they depend on the resistance at the rear wheel decreasing. That doesn't happen when you're riding.



Yes, people are suggesting that, but it's a misconception.

The flex in your frame does not ever make you go forward, period. That's because for this to happen, a lesser force would have to overcome a greater one. It is that simple.

As you're riding along, your legs vary how much force they're exerting between a maximum equal to (overcoming your inertia)+(frame flex), and some amount less than that. Now here comes the crucial point: your inertia never lessens.

The only time when the frame can un-flex is when your legs ease off, and the suggestion that when that happens, the flex energy is going to overcome your inertia -- the greater force, now that your legs are applying a lesser force -- is incorrect.

That is even vaguely true. The frame un-flexes when the load shifts from right to left. And that happens without the pedal forces going to zero.

What's getting left out of your analysis is that there are actually two ways for the frame to store some pedaling energy - by lowering the chainring (right pedal load), or raising the chainring (left pedal load). Moving between them is when the stored energy is pushed down the stays.


All the various static examples are demonstrating is that the flex is still in line with the drivetrain, and isn't just flying out into space. Then you take that understanding and apply it to the fact that the tension on the chain never goes to zero and the BB flex automatically alternates from right-to-neutral-to-left.

[johnniecakes]

But what about the crank arm and pedal spindle? Wouldn't they have also flexed or deflected and contributed in some way?

[kramnnim]

Originally Posted by Fivethumbs

But it did. And the wheel turned.

To me, it looks like the trainer is flexing more than either frame.

[cachagua]

That is even vaguely true.

I apologize; I intended to be anything but vague.

The frame un-flexes when the load shifts from right to left. And that happens without the pedal forces going to zero.

Who said anything about the pedal force going to zero? All I said was it has to drop from its maximum.

What's getting left out of your analysis is that there are actually two ways for the frame to store some pedaling energy - by lowering the chainring (right pedal load), or raising the chainring (left pedal load). Moving between them is when the stored energy is pushed down the stays.

In other words, what we see when the BB "sways" from one side to the other and back? Again: the inertia at the rear wheel is pushing harder than your pedals, because your pedaling force has decreased in relation. In response to the decreased force at the pedals, the frame discharges some of its strain energy as we see, but there ain't no way that can get out the back wheel and move you along the road. The softer push is not going to overcome the harder one. Do you dispute that?

I don't know how I can make this simpler. If you and I push our hands together and push equally hard, our hands will stay still. If you relax a little and push softer, and I keep on pushing just as hard, will our hands move toward me as a result?

All the various static examples are demonstrating is that the flex is still in line with the drivetrain, and isn't just flying out into space. Then you take that understanding and apply it to the fact that the tension on the chain never goes to zero and the BB flex automatically alternates from right-to-neutral-to-left.

Flex is in line with the drivetrain -- not sure what you mean. Does it matter?
Tension on the chain never goes to zero -- doesn't need to; not part of my account.
BB flex alternates from right to neutral to left -- okay, if it goes back to neutral in the middle then manifestly there's nothing carried over to the next pedal stroke.

[Kontact]

Originally Posted by cachagua

I don't know how I can make this simpler. If you and I push our hands together and push equally hard, our hands will stay still. If you relax a little and push softer, and I keep on pushing just as hard, will our hands move toward me as a result?

No, but that isn't what the drivetrain is doing. You are describing something that would happen in a straight line. The drivetrain is twisting, and the untwisting is what is adding to the forward motion. If downforce on the pedal was constant, that flex would never go away and the stored flex energy would never get used. But on bicycles the flex oscillates from left to right, putting force back into the drive line every time it goes through the center.

Flex is in line with the drivetrain -- not sure what you mean. Does it matter?
Tension on the chain never goes to zero -- doesn't need to; not part of my account.
BB flex alternates from right to neutral to left -- okay, if it goes back to neutral in the middle then manifestly there's nothing carried over to the next pedal stroke.

Nothing is carried to the next pedal stroke - that's the point. The right side flex is used up by the BB returning to center, and then new left side flex is produced, then it is used up as the BB returns to center for the right pedal stroke.

It isn't like a spring that is simply being allowed to uncompress. Pedaling actively untwists the drivetrain, putting the work used to flex the frame to use. And that would be true even if the frame had a pivoting BB that required no real force to "flex". If the distance from chainring to cassette varied throughout the pedal stroke because the crank pivoted laterally on a bearing, then you'd still get the effect of force going through the chain from pedaling through neutral.

[ergott]

So hypothetically the twisting of the chainstays could be isolated and provide forward motion by themselves? A jig could be made to illustrate that movement and demonstrate how it could propel a bike or at the very least rotate the rear wheel.

[Kontact]

Originally Posted by ergott

So hypothetically the twisting of the chainstays could be isolated and provide forward motion by themselves? A jig could be made to illustrate that movement and demonstrate how it could propel a bike or at the very least rotate the rear wheel.

Yes. I'd have to think about it awhile to figure out how to do it, but the twisting of the stays and BB act very much like a step cycle in miniature:

[cachagua]

The twisting of the stays and BB act very much like a step cycle in miniature...

We've covered that already:

Originally Posted by cachagua

Let's look at the bike on the trainer again. You keep the brake squeezed, but ease up on the cranks, the right pedal rises up from the block -- but at the same time, the left pedal also makes a counter-rotating motion. The release of the frame flex does not advance the other pedal but moves it backwards, and thus doesn't increase the force of the next stroke, but decreases it, too.

I think I see what you're driving at -- that when the frame releases the strain energy that one pedal stroke put into it, that could move the other pedal a little farther around, or make it easier for it to rotate. It's a beautiful idea! It would be great if it were true! But as we see, what retards one pedal retards the other equally.

[Kontact]

Originally Posted by cachagua

We've covered that already:



I think I see what you're driving at -- that when the frame releases the strain energy that one pedal stroke put into it, that could move the other pedal a little farther around, or make it easier for it to rotate. It's a beautiful idea! It would be great if it were true! But as we see, what retards one pedal retards the other equally.

It has nothing to do with the pedals or advancing them. It is that the stays get shorter from twisting when the BB sways, and when the sway goes through the center point the distance between chainring and cassette gets longer. When that happens either the crank is going to stop your legs or the chain will pull the wheel. Since we know your legs don't hiccup at the bottom of the pedal stroke, the elongation goes toward moving the bike.

Again, this isn't because chain tension has relaxed, it is because the lateral displacement of the BB has relaxed into a position that requires more chain. And even if the frame didn't want to relax to that position, the other half of your pedal stroke doesn't give it any choice.

Right pedal, neutral, left pedal, neutral.
Short stays, long stays, short stays, long stays.
Less chain, more chain, less chain, more chain.

[ergott]

That movement wouldn't be enough to advance the bike even one click of the freehub even on the finest of engagement mechanisms. The bike would just be sitting there flexing in both directions. If the bike were a fixed gear, it would still just rock back forth. It's not just energy output, it's energy over time and there's no way you could propel a bike with that tiny of a movement by itself. The movement of the BB is over the time of a riders cadence so it's too gradual.

A spring can launch something in the air, but if you slow down the release of that energy (energy over time) you could have a car spring that couldn't even launch a penny.

Seems more like trying to row a boat leaving the oars in the water the whole time.

[Black Dog]

Originally Posted by ergott

That movement wouldn't be enough to advance the bike even one click of the freehub even on the finest of engagement mechanisms. The bike would just be sitting there flexing in both directions. If the bike were a fixed gear, it would still just rock back forth. It's not just energy output, it's energy over time and there's no way you could propel a bike with that tiny of a movement by itself. The movement of the BB is over the time of a riders cadence so it's too gradual.

A spring can launch something in the air, but if you slow down the release of that energy (energy over time) you could have a car spring that couldn't even launch a penny.

Seems more like trying to row a boat leaving the oars in the water the whole time.

This is the million dollar question. Where exactly is the spring's energy going? Like your penny example, slowing down the spring will transfer the energy to the device that is slowing down the spring and not the penny. With hub crank and pedal based power meters it would not be too hard to do some empirical testing with frames of various stiffness.

[ergott]

Originally Posted by Black Dog

With hub crank and pedal based power meters it would not be too hard to do some empirical testing with frames of various stiffness.

I'm willing to bet that anything going in here is way under the accuracy tolerances of the best power meters available (±1%).

[fa63]

Originally Posted by Black Dog

This is the million dollar question. Where exactly is the spring's energy going? Like your penny example, slowing down the spring will transfer the energy to the device that is slowing down the spring and not the penny. With hub crank and pedal based power meters it would not be too hard to do some empirical testing with frames of various stiffness.

If you listen to the Cycling Tips podcast which was linked earlier in the thread, Damon Rinard (Road Engineering Manager at Cannondale) talks about how he was unable to detect any loss when looking at simultaneous power measured at both the crank and the hub.

Using a simpler approach (strain energy), he also mentions in the podcast that he calculated about 40W for power which goes into flexing a frame under an all-out sprint (I think it was for a 1500-2000W effort). This was meant as an upper-bound value, but even if it is true, it is about 2% of the power input, which is more or less the accuracy of most strain-gage based power meters on the market today.

[Kontact]

Originally Posted by ergott

That movement wouldn't be enough to advance the bike even one click of the freehub even on the finest of engagement mechanisms. The bike would just be sitting there flexing in both directions. If the bike were a fixed gear, it would still just rock back forth. It's not just energy output, it's energy over time and there's no way you could propel a bike with that tiny of a movement by itself. The movement of the BB is over the time of a riders cadence so it's too gradual.

A spring can launch something in the air, but if you slow down the release of that energy (energy over time) you could have a car spring that couldn't even launch a penny.

Seems more like trying to row a boat leaving the oars in the water the whole time.

If the movement is so little to power the bike, then it is also so little to be a loss.

That's the only real point here. Either the energy stays in the drivetrain either way, or there is no energy difference.

[johnniecakes]

Me wonders how much unrecoverable energy has been spent on this subject.