Frame Flex on GCN

[mtechnica]

Originally Posted by cachagua

Fantastic, fantastic image! That's one of the strongest analogies we've come across. It's a good way to see the distinction I was trying to make between frame flex caused by drive force, vs. flex caused by forces that don't move the bike.

The whole point of a bow is that it stores energy that is applied slowly and releases it again quickly, is that what we’re saying is happening in the frames system? If so what is the energy that’s being released acting upon?

[cachagua]

Originally Posted by kramnnim

[The linked article] basically shows what the GCN video shows...

Yes, they operate from pretty much identical models.* And there's kind of a flaw behind that model. However I think we can use it to illustrate something:

The model is that you squeeze the frame between two forces, applied at the pedal and at the rear wheel. In response the frame flexes, and when you reduce one of the forces squeezing it, the frame unflexes, and the strain energy it had stored does some work.

Where does it do some work? It does some work where the force was reduced.

And that's exactly what happens when you're actually riding, only when you're riding, it's switched around the other way. When you're riding, the resistance at the rear wheel stays the same, like the pedal pressure did in the video -- and when you're riding, the force from your pedal is what's reduced, whenever you come around to the part of your pedal stroke where your force drops below its maximum.

That's what led me to think the released strain energy must act on the pedals, not on the bike's movement.


*From the article:
"Mount your bike to a trainer and disengage the resistance roller.
1. Put the cranks in the horizontal position.
2. Place a rigid block or stool under the forward pedal so that there is a small gap under the pedal.
3. While holding the rear brake firmly, stand on the pedal so that it is pushed down to the stool.
4. Keep holding the pedal down and release the brake.
When you pushed the pedal down, the chain did not move since the brake locked the rear wheel. Since the chain did not move, no work energy was delivered through the chain. The crank moved down with the pedal as the frame was strained. When you released the brake, the frame was able to move the center of the crank back up to relieve the strain energy..."

[cachagua]

Originally Posted by mtechnica

The whole point of a bow is that it stores energy that is applied slowly and releases it again quickly...

For our purposes, how fast the bow (the frame) is flexed and un-flexed isn't significant.

What is the energy that’s being released acting upon?

What I've said all along is that it retards the cranks' rotation.

[mtechnica]

Originally Posted by cachagua

What I've said all along is that it retards the cranks' rotation.

... because while the side you just cranked on is flexing back up, you’re applying the pedaling force downward on the opposite side which is also flexing down while you’re pedaling? Reducing the reaction force against your foot...

[Kontact]

Originally Posted by mtechnica

If the chain stay is a fixed length and the bottom bracket shell moves left and right, wouldn’t that mean that the distance between it and the hub actually increases under load since it would be the hypotenuse of the triangle it would make? The distance can never be less than in the unoaded state correct? So long as the chain stay isn’t under compression.

It is already a hypotenuse because it angles from the outboard dropout to the inboard BB shell. When you push the BB over you are making that base leg longer, tugging on the hypotenuse.

Originally Posted by mtechnica

At any rate shouldn’t energy loss be accountable with a hub based power meter?

There doesn't appear to be any energy lost when it has been measured with power meters. Some cyclists disagree with that, but everyone who has tested it in a lab all agree that frame flex doesn't eat energy.

Originally Posted by mtechnica

The whole point of a bow is that it stores energy that is applied slowly and releases it again quickly, is that what we’re saying is happening in the frames system? If so what is the energy that’s being released acting upon?

Since the frame has two crankarms and the BB can flex either way in a bow-like manner, it is more complicated than a bow. But the flex is essentially forced to zero out between right and left pedal stroke.

[mtechnica]

Interesting stuff to think about!

One thing I don’t know if anyone has mentioned yet, does anyone think that a flexy frame will increase your effective saddle height when pedaling hard versus a frame that doesn’t deflect? I wonder if that could make a difference to someone.

But yeah.. if you had a pedal power meter and a hub power meter and the lab shows no difference it must be mostly a mental thing.

I think the frame is deceptively complicated though, it’s interestinf how I’ve had some frames that felt dead and flexy to climb with (sorry 853 lemond) but some frames that flex noticeably but still feel efficient to pedal (like my Calfee).

[andrewsuzuki]

Originally Posted by mtechnica

At any rate shouldn’t energy loss be accountable with a hub based power meter?

Already done by Damon Rinard / Cannondale. They found no significant difference, but...

Unfortunately, the Powertap hubs have an error margin of 1.5%. The FEA shows ~1% of rider power goes into flex on a nice steel bike. So even if all of that flex energy is somehow lost, it still can't be measured accurately with current power meters. Even the SRM Science crank has a 0.5% error margin.

[Mark McM]

Originally Posted by Kontact

Actually, it doesn't work that way. Flexible frames flex until the tension on the chain is the same as a stiff frame. That happens because the amount of force that it takes to flex the frame increases with the distance it is flexed. At full flex a flexy frame is "stiff".

Careful there - Stiffness determines the change in deflection for a change in load - frames are linearly elastic, so they retain the same stiffness at all loads/deflections.

Originally Posted by Kontact

And the chainwheel and wheel do get closer. The BB can't swing off to one side without the chainstays following it, and being a fixed length the chainstays have to pull on the dropouts to follow the BB.

I don't know why you keep bringing this up. The rear triangle/chainstays are incredibly stiff in compression, so this flex is very, very small - particularly when compared to the stiffness at the BB in torsion. Therefore, chain/rear triangle flex stores very, very little energy. From tha web page analysis frequently cited here, the energy in this mode of flex is greatly swamped by the energy stored in other flex modes. Continually bringing this up adds nothing to the conversation.

[Kontact]

Originally Posted by Mark McM

Careful there - Stiffness determines the change in deflection for a change in load - frames are linearly elastic, so they retain the same stiffness at all loads/deflections.



I don't know why you keep bringing this up. The rear triangle/chainstays are incredibly stiff in compression, so this flex is very, very small - particularly when compared to the stiffness at the BB in torsion. Therefore, chain/rear triangle flex stores very, very little energy. From tha web page analysis frequently cited here, the energy in this mode of flex is greatly swamped by the energy stored in other flex modes. Continually bringing this up adds nothing to the conversation.

I don't know why you keep thinking that my description of how the chain and chainstay's length relationship changes is a description of where energy is being stored. I am not saying that in any way, and already explained that to you.


And I wasn't saying that the frame changes physical properties. I'm saying that once you've flexed the frame as far as the load will take it, it stops at that point and becomes rigid to that loading. No more flex past that point for that load.

[kramnnim]

Originally Posted by mtechnica

... because while the side you just cranked on is flexing back up, you’re applying the pedaling force downward on the opposite side which is also flexing down while you’re pedaling? Reducing the reaction force against your foot...

Exactly.

[kramnnim]

Originally Posted by andrewsuzuki

This graph suggests the frame is at equilibrium at the 7 o'clock position (~210deg), and the majority is returned between 135deg and 180deg.

This is all using real rider power curves, "an average of 17 road riders at 350W and 90rpm"

https://web.archive.org/web/20060214...edal_loads.htm

It's buried several pages back, but I have only been focusing on frame flex when pedaling out of the saddle, at low rpm, while rocking the bike back and forth. Like Contador, only less graceful since I'm not Contador. Or when taking off from a stop sign/light. This is the only time I've noticed a sluggish feeling from my flexiest frame. (and it might be my imagination)

[cachagua]

Originally Posted by Mark McM

Chain/rear triangle flex stores very, very little energy... the energy in this mode of flex is greatly swamped by the energy stored in other flex modes. Continually bringing this up adds nothing to the conversation.

I can't speak for anyone else, but the reason I continually bring this up is that it's the only flex that matters:

Originally Posted by cachagua

Let's restrict our attention to drive torque and the frame flex it produces... the reason we leave out flex caused by non-drive-torque forces is that, even if there were no frame flex, those forces weren't going to move you forward. Hence the name.

I don't think anyone's trying to say it's a lot of energy. It is indeed very, very little... and that's why I believe that when it goes to waste we don't even notice.

[kramnnim]

Originally Posted by cachagua

I don't think anyone's trying to say it's a lot of energy. It is indeed very, very little... and that's why I believe that when it goes to waste we don't even notice.

What makes some frames feel slow? It would take an awful lot of flex to waste a significant number of watts...but most of us have ridden those frames that feel sluggish. What causes that?

I rode a Look 586 for over 20,000 miles and loved it. Bought some other frames and the 586 was relegated to the trainer. Took it outdoors a few times and it feels dead.

[dddd]

What a frame feels like and what a frame delivers may be two very different things, no?

I wouldn't discount the effect of ANY of the flex elements that respond to pedal force, since ALL contribute to some net elasticity value that is the opposite or reciprocal of what the industry's marketing people term "power transfer", which is supposedly a benefit to their oft-quoted "reactivity" (which I interpret as equivalent to their also-quoted "change in rhythm". I didn't make these terms up, they did!

So in reality, anything that shows dimensional change in response to pedaling load can be assumed to be contributing to elasticity of the drivetrain even when it appears to be orthogonal/perpendicular to the loading. It just has to be quantified relative to the right axis to show that it isn't just an off-axis motion that isn't storing energy.

There is the chain tension elasticity that is part of a load path that includes everything from the chain to the spokes to the chainstay, and which as I have mentioned is HIGHLY dependent on the chainring size selected (to the tune of over 100%, even as the % difference in chainring size is much smaller than that to the tune of a square-root function or less than half of the % difference in elasticity as felt at the pedals).

One complicating aspect of how this flex affects the rider's metabolic/biological performance (in terms of power output efficiency) is the asymmetry of this flex and how the left and right pedal respond somewhat differently to what kontact described as a "complex" loading system/equation (what I term "load path"). It is further complicated in the scenario where the rider is out of the saddle and thus supporting body weight continuously while pedaling, and of course the rider's body is in a completely different activity mode at those times, so would be expected to benefit or suffer from flex in a somewhat different manner.
Same thing with changing cadence and loading, where efficiency may benefit or suffer from different levels (and different forms) of flexibility.
Pretty much none of this has been lab-quantified in detail as far as I know, but likely the rider is able to adapt to a great degree of variation before net efficiency deficits show any big losses.
I think that solid conclusions on this subject are still many years out.
But hey, what are we at, 23 pages?

[kramnnim]

Originally Posted by dddd

What a frame feels like and what a frame delivers may be two very different things, no?

Certainly...I'm just curious what causes the dead feeling. Though...a frame that feels dead to me may feel lively to someone else...