Frame Flex on GCN

[etu]

a bit neutral on whether flexible frames are more efficient, but I thought the idea was that more you're in sync with your bike, the more motivated you are to put out more power. also there is the thought that being limited by our cardiovascular limits means we're closer to our potential power output than being limited by burning in our legs/butts. would love to hear other's thoughts on this latter claim.

[Mark McM]

Originally Posted by Kontact

As I said, the cranks gather velocity in the power zone and that velocity is carried into the dead zone where it is used up. The "storage" is just between power and dead.

I don't know why you don't understand what I mean.

Because it is clear that you do not understand what you mean.
If you understood what was going on, then you'd realize that the reciprocating motion of the legs already incorporates some amount of energy storage (as the cranks rise) and release (as the crank drops) without needing some type eccentricity. In fact, the rotary velocity changes with Biopace cranks probably results in extra losses in leg motion compared to constant velocity rotation.

Originally Posted by Kontact

I don't have to measure it - the GCN video quite clearly shows the "expanding" stays pulling on the chain. Instead of letting the wheel go completely, they could have let the brake out slowly and demonstrated exactly how much distance the chain is pulled for the amount of crank deflection.

You must be looking at a different video, because that is not what is shown at all. The GCM video does not show "expanding" chainstays, nor do they even claim that they do. Instead, the video shows that the frame flexes torsionally; the drive force causes the drive side of the BB to drop as the frame twists. When the rear brake is released, the frame untwists, causing the drive side of the BB to rise (return to its original position). Because the pedal is fixed in position with the crank still primarily horizontal, the rising of the BB causes a rotation of the cranks, which is what cause the wheel to rotate.

Of course, this demonstration is completely rigged, and does not reflect reality. When riding, the pedal is not stopped when the crank is horizontal - instead, the full force remains on the pedal, and the frame remains flexed, until the pedal reaches the bottom of the stroke. So in a real pedaling system, the BB is not unloaded (and the "spring" energy is not released) until the cranks are vertical. When the crank is vertical, the rising of the BB does not result in crank rotation (in fact if the BB rises when the crank is past BDC, the rising BB acts to try to rotate the crank backward). Imagine if they tried the experiment, except instead stopping the pedal on the down stroke, they stopped it at the very bottom of the stroke. If they did this, then the rise in the BB when unloaded would not result in any crank rotation, and the wheel would not spin. This gets the crux of the problem I noted earlier - the frame deflection is orthogonal to the drive force, so you need a mechanism to convert/re-direct the spring energy.

[rain dogs]

Originally Posted by Kontact

1. Flexible bikes don't squeak any more than rigid bikes. If you account for the all the PF30 bikes, it would appear that not flexing is more of a problem. Having a BB shell flex with the cranking motion is going to put less side load on the bearings, not more.

2. Bigger tires are faster if we are talking 20s vs 23s or 25s. Above 25s the rolling resistance goes up when you are riding the correct pressure for your weight.

1. First, this isn't binary. There aren't "flexy" bikes and "rigid" bikes. All bikes flex. The same things you are arguing for your "flexy" bikes... if true, would occur on ALL frames, just less so on a more stiff frame. It's not like a flexy steel bike obeys certain laws of physics and a less flexy carbon or aluminum or steel bikes obeys different laws of physics.

So, if flexy bike A absorbs 20watts and returns 80% of that (16watts) you're losing 4 watts. If stiff (read: less flexy) bike absorbs 10 watts and returns 80% of that (8 watts) you're losing 2 watts. And if very stiff (read: much less flexy) bike absorbs 5 watts and returns 80% (4 watts) you're losing 1 watt.

Numbers are "imaginary" for a simple example.

2. The lowest rolling resistance has been tested in tires that are 1.5" wide (38mm) when all other things are equal (I'll have to find the link when I have more time). AND In regard to your incorrect claims about 28's:

"according to figures from another tyre brand, Continental, a 20mm tyre with 160psi, a 23mm tyre at 123psi, a 25mm tyre at 94psi and a 28mm tyre at 80psi all have the same rolling resistance."

"Similarly, the Zipp Tangente Speed clinchers were faster in the 28mm (31.3w) than the 25mm (33.2w) on the rough-surface drum." <The fastest tire in this test was 28.5mm the Schwalbe Pro One Tubeless.

And regarding pressure for weight. The wider tire the lower pressure you can ride for your weight.

[Kontact]

Originally Posted by Mark McM

Because it is clear that you do not understand what you mean.
If you understood what was going on, then you'd realize that the reciprocating motion of the legs already incorporates some amount of energy storage (as the cranks rise) and release (as the crank drops) without needing some type eccentricity. In fact, the rotary velocity changes with Biopace cranks probably results in extra losses in leg motion compared to constant velocity rotation.

Mark, you didn't understand initially when I said "going into the dead zone", which means "has additional velocity before the dead zone" and you have been confused ever since.

The comparison I made to Biopace, which your first "correction" also reflects, is that Biopace (like a frame that flexes more) does not put as much peak power to the wheel compared to a round ring, and then causes that peak energy to be carried as additional crank velocity (again, compared to round) going into the dead zone where it is then converted to power at the wheel. All of this is only in comparison the a round ring, and none of it is a net increase or decrease in total power, but a change in the amplitude and distribution of power coming down the chain throughout the cycle. Lower peaks, but higher valleys.

I'm guessing that you got lost in "crank speeds up" vs. chain speed which is why you thought we were talking about two different things, or you got confused between "going into" and "in the dead zone".

You must be looking at a different video, because that is not what is shown at all. The GCM video does not show "expanding" chainstays, nor do they even claim that they do. Instead, the video shows that the frame flexes torsionally; the drive force causes the drive side of the BB to drop as the frame twists. When the rear brake is released, the frame untwists, causing the drive side of the BB to rise (return to its original position). Because the pedal is fixed in position with the crank still primarily horizontal, the rising of the BB causes a rotation of the cranks, which is what cause the wheel to rotate.

Of course, this demonstration is completely rigged, and does not reflect reality. When riding, the pedal is not stopped when the crank is horizontal - instead, the full force remains on the pedal, and the frame remains flexed, until the pedal reaches the bottom of the stroke. So in a real pedaling system, the BB is not unloaded (and the "spring" energy is not released) until the cranks are vertical. When the crank is vertical, the rising of the BB does not result in crank rotation (in fact if the BB rises when the crank is past BDC, the rising BB acts to try to rotate the crank backward). Imagine if they tried the experiment, except instead stopping the pedal on the down stroke, they stopped it at the very bottom of the stroke. If they did this, then the rise in the BB when unloaded would not result in any crank rotation, and the wheel would not spin. This gets the crux of the problem I noted earlier - the frame deflection is orthogonal to the drive force, so you need a mechanism to convert/re-direct the spring energy.

The GCN example loads the crank where the crank actually gets loaded during cycling - the power stroke. If we attempted to ride bicycle by only pushing down at 6 o'clock the cranks wouldn't turn - power has to be applied at something other than parallel to the orientation of the crank arm. But if you apply power at 90° to the crankarm anywhere in the pedal stroke you will get some sort of frame flex and motile power. But in the case of an upright bicycle that power is largely applied 90° from 3 o'clock of each crank, and applying power at that position will cause the frame to flex in a particular way, and that way is really the only one worth discussing. If we could pull up on the back of the stroke as hard as we can push down then the BB wouldn't flex and we wouldn't be talking about it.

But the BB does flex - bringing the chainrings down during right pedal power and up on left pedal power. And it is flexing that way not because we are standing on a pedal that is at 6 o'clock, but because we are pushing down at 3 o'clock which isn't an ideal position for pushing the BB down, but happens because the crank is being restrained from flexing in other directions by the tension of the chain.

That's the crux of all of this - the flex we see is the product of the opposition of pedal force and chain resistance, not just bending that is coincidental to the force applied to the chain.


Here's another experiment GCN could have done: Hold the crank at 3 o'clock with body weight, and push the bike backwards to simulate the resistance of applying power to the road. If the right pedal is blocked, the tension in the chain will cause the right side of the BB to drop. When the force holding the bike backwards is removed, the pedal won't move, the crank won't turn but the bike will move forward as the flex comes out of the BB. And when you do it with the left pedal the right side of the BB will rise, yet all the same things will happen.

[Kontact]

Originally Posted by rain dogs

1. First, this isn't binary. There aren't "flexy" bikes and "rigid" bikes. All bikes flex. The same things you are arguing for your "flexy" bikes... if true, would occur on ALL frames, just less so on a more stiff frame. It's not like a flexy steel bike obeys certain laws of physics and a less flexy carbon or aluminum or steel bikes obeys different laws of physics.

So, if flexy bike A absorbs 20watts and returns 80% of that (16watts) you're losing 4 watts. If stiff (read: less flexy) bike absorbs 10 watts and returns 80% of that (8 watts) you're losing 2 watts. And if very stiff (read: much less flexy) bike absorbs 5 watts and returns 80% (4 watts) you're losing 1 watt.

Numbers are "imaginary" for a simple example.

2. The lowest rolling resistance has been tested in tires that are 1.5" wide (38mm) when all other things are equal (I'll have to find the link when I have more time). AND In regard to your incorrect claims about 28's:

"according to figures from another tyre brand, Continental, a 20mm tyre with 160psi, a 23mm tyre at 123psi, a 25mm tyre at 94psi and a 28mm tyre at 80psi all have the same rolling resistance."

"Similarly, the Zipp Tangente Speed clinchers were faster in the 28mm (31.3w) than the 25mm (33.2w) on the rough-surface drum." <The fastest tire in this test was 28.5mm the Schwalbe Pro One Tubeless.

And regarding pressure for weight. The wider tire the lower pressure you can ride for your weight.

I didn't say anything is binary flex/no flex, but if you are using two discreet bikes to test, that is binary. And I was speaking of the relative results of testing two obviously different bikes.



As far as rolling resistance goes, take the chart of rolling resistance of tires vs psi from the chart in this article:
https://www.bicyclerollingresistance...0s-ii-23-25-28

And plug in the recommended PSI from this chart:


So I picked 88 lbs as the input number, which gives a recommendation of 64psi for the 28c and 91psi for the 23c. When you plug those pressures into the rolling resistance chart, you get about 13.6 rolling resistance for the 23@91psi and about 13.8 for the 28@64psi.

If we take the bike weight = tire pressure thing seriously, those are the tire pressures to use to compare rolling resistance. You can certainly pump the 28 up to 70psi to lower its rolling resistance down to where the 23 is, but then you should be increasing the 23 to be fair.

[cachagua]

How do you know she's a witch?

[rain dogs]

Originally Posted by Kontact

So I picked 88 lbs as the input number, which gives a recommendation of 64psi for the 28c and 91psi for the 23c. When you plug those pressures into the rolling resistance chart, you get about 13.6 rolling resistance for the 23@91psi and about 13.8 for the 28@64psi.

I was being very conservative when I said 2-5mm wider and 5-10psi less pressure and you disagreed implying you'd need to be putting 110psi in a 28mm tire. And now you've posted a response where you wrote that 27psi less! and 5mm wider results in a difference of 13.6 and 13.8... so 0.2 Watts and nearly 30psi less!

Now, you imagine the difference of comfort of 25-27psi less. Heck even 20psi less and a 5mm wider tire. Not to mention the better traction and road feel.

So I guess your own post/data completely backs up my original point. Thanks.

So again, a less flexy bike, wider tires at the appropriate psi will give you less loss to frame flex (no matter how small), more comfort and traction (less of your claimed 'skidding under power' ), and a lower rolling resistance. In other words greater efficiency, no?

[dddd]

The pressure drop appropriate to the use of a wider tire would seem to me to perhaps be larger than what is intuitively assumed.
Wider tires put proportionally more tension stress into their casing at equivalent pressure, which hardens the casing of the tire in terms of it's ability to yield to roughness of the road surface, so the pressure drop needs to firstly overcome this increased casing tension before we even start to see any increase in the tire's ability to conform to the road surface. So the wider tire widths really should be tested at quite-lower pressures before we start looking at rolling resistance data.


As for the issue of chainstay compression, of course this does occur, but this is part of the greater load path which encompasses all of the flexible elements that contribute to drivetrain elasticity outside of lateral flex at the bottom bracket.
This load path adds up to the sum total of:
--Chain elasticity
--Rear axle and freehub body flex and including bearings
--Chainstay elastic compression
--Bottom bracket spindle flex and including bearings
--Chainring flex in torsion along axis of bottom bracket
--Chainring+cog lateral flex in response to chain tension and cross-chain angle

Note that this entire net drivetrain load path flexing/elasticity constitutes windup flex peaking at the origin of force at the crankset, expressed as degrees rotation per unit torque delivered through the selected chainring.
Further, this entire load path loading and flex is dependent, quantitatively, on the inverse of the chainring size to the second power(!). So a 10% smaller chainring imposes an increase of a larger 23.4% increase in flex felt at the pedal eye in response to pedaling force. And this using the same overall gear ratio (as when for theoretical example where the rider changes from a 50/25t to a 34/17t combination).
And with a most-common 50-34t chainset, the flex felt at the pedal eye increases by a whopping 116% after changing from the big ring to the small ring (excluding rear hub/cassette/wheel windup, which is one place where torque isn't affected by changes in chainring size assuming same overall gear ratio).

So again, it is the entire drivetrain load path which is significant and not just chainstay compression by itself.

[earlfoss]

All that math won't save you when you're facing a selection in a race, going over the crest of the hill at max effort for 15-30 sec. I can almost guarantee you the dude on a flexy frame is going to be working much harder to keep up with those on stiffer framesets. However, after he gets dropped he'll have ample time to consider the superior comfort and dampening abilities of his rig. He'll feel fresher after the race is over, too.

The ti Hampsten I raced for a season was an awesome bike. I loved it, but when the hammer was down, the little bit of bb flex was a liability. I didn't have ghost shifting or anything that extreme. Going from that to a Giant Propel, the difference was night and day. I'd say that for modern racing, to be competitive you need to have a bike that is at least as stiff as anything the average guy is riding in your event if that makes any sense.

[dddd]

Originally Posted by earlfoss

All that math won't save you when you're facing a selection in a race, going over the crest of the hill at max effort for 15-30 sec. I can almost guarantee you the dude on a flexy frame is going to be working much harder to keep up with those on stiffer framesets. However, after he gets dropped he'll have ample time to consider the superior comfort and dampening abilities of his rig. He'll feel fresher after the race is over, too.

The ti Hampsten I raced for a season was an awesome bike. I loved it, but when the hammer was down, the little bit of bb flex was a liability. I didn't have ghost shifting or anything that extreme. Going from that to a Giant Propel, the difference was night and day. I'd say that for modern racing, to be competitive you need to have a bike that is at least as stiff as anything the average guy is riding in your event if that makes any sense.

I would say that it depends on whether the "you" needing a bike "that is at least as stiff as anything the average guy is riding" is close to the weight/strength of the other riders, and whether sprinting is "your"/their thing.

I agree though that at times, say during a short/intense uphill sprint, that I (even at my 140lbs) have felt that I would be accelerating faster with a stiffer entire bike. But whether I really would be going significantly faster (especially after having had to ride and pedal the earlier part of the ride or race on that same stiffer bike) at the end of that sprint remains a point of curiosity, since I don't really know.
I do feel that for shorter efforts, when the legs are fresh, that yes, I probably could accelerate faster on a stiffer bike, but wouldn't the higher peak forces be more likely to bring on the onset of cramping during such a sprint toward the end of a ride or race? Again, I can't say for sure, though having enough solid resistance to push against would seem to improve the muscles' generating more overall power with that resistance force not being delayed and softened within the pedal stroke.
Perhaps a frame (or just drivetrain) can be developed with a more-constant deflection curve at the bottom bracket in response to pedaling effort, either through electro-mechanical gadgetry or some sort of material property akin to thixotrophy or viscoelasticity for rough analogy here(?).
Does that take us back to BioPace? Ohhhhh-Nooooo.
(Though I confess to doing almost all of my Road, XC and CX racing over the last 20+ years using BioPace II and similar BioPace SG chainrings!!!)

[andrewsuzuki]

Originally Posted by earlfoss

All that math won't save you when you're facing a selection in a race, going over the crest of the hill at max effort for 15-30 sec. I can almost guarantee you the dude on a flexy frame is going to be working much harder to keep up with those on stiffer framesets. However, after he gets dropped he'll have ample time to consider the superior comfort and dampening abilities of his rig. He'll feel fresher after the race is over, too.

The ti Hampsten I raced for a season was an awesome bike. I loved it, but when the hammer was down, the little bit of bb flex was a liability. I didn't have ghost shifting or anything that extreme. Going from that to a Giant Propel, the difference was night and day. I'd say that for modern racing, to be competitive you need to have a bike that is at least as stiff as anything the average guy is riding in your event if that makes any sense.

Substitute "flexy frame" with "25mm tires" and "stiffer frame" with "19mm tires". You can easily imagine saying something like this ten years ago, just like 19mm tires "feel" faster

[earlfoss]

Say what you want, but at least at the P12 level you're going to want a stiffer frame these days. At the end of the day, race what you brung though.

[Kontact]

Originally Posted by rain dogs

I was being very conservative when I said 2-5mm wider and 5-10psi less pressure and you disagreed implying you'd need to be putting 110psi in a 28mm tire. And now you've posted a response where you wrote that 27psi less! and 5mm wider results in a difference of 13.6 and 13.8... so 0.2 Watts and nearly 30psi less!

Now, you imagine the difference of comfort of 25-27psi less. Heck even 20psi less and a 5mm wider tire. Not to mention the better traction and road feel.

So I guess your own post/data completely backs up my original point. Thanks.

So again, a less flexy bike, wider tires at the appropriate psi will give you less loss to frame flex (no matter how small), more comfort and traction (less of your claimed 'skidding under power' ), and a lower rolling resistance. In other words greater efficiency, no?

What an odd response. Pick any point you want on the graph using any starting weight you want. The Rollingresistance guys started at 120psi and worked their way down from there. The point is that there needs to be comparable psi for both tires, and the point of comparison is load placed on them by the bicycle.

Your response isn't math. Use the math if you want to make a mathematical statement about rolling resistance.

[Ungaro]

Originally Posted by cachagua

How do you know she's a witch?

I know! She if she floats!!!!

[Mzilliox]

Originally Posted by cachagua

How do you know she's a witch?

throw her into a pond and see if she floats?