McBTC

Good point and may be irrelevant if you consider, a bow, for example, probably doesn't return 100% of the energy put into it but it doesn't work well any other way. And, hitting a pothole provides energy to the frame too but, a little dissipation can be desirable.

Andrew R Stewart 

A few data points not yet mentioned.

Gary Klein wrapped his bike's chain stays with Boron fiber tow in the early 1980s (IIRC the years properly). Much like carbon fiber wrapped tubing.

A certain older track Kilo rider designed and rode a frame that used tensioned cables to "increase" the stiffness of his bike. Maybe the 1988 games?

We don't see either design used these days (or really for any days or with other companies right after) for a bunch of reasons. That these attempts didn't make a difference that mattered in the market place or in real competition is the biggie one. Andy

TallRider

I've said for years that if manufacturers had evidence that stiffer frames --> more power transfer, they'd present the evidence. Instead, manufacturers just relied on "stiffer must be faster" until recent years when Jan Heine's notion of "planing" started to gain traction. (Note: this argument, with an actual blind test to validate it, is that a more flexible frame can better get in sync with pedal strokes. I'm not 100% sold, but the evidence for this is stronger than evidence that stiffer --> faster.)

The incredibly easy way to test stiffer=faster is to use a pedal or crank-based power meter, along with a rear-hub-based power-meter, and see whether the measures get further apart when using a more flexible frame. The fact that nobody is reporting this tells you that the difference either doesn't exist or isn't meaningful. (Not that Jan Heine's concept of planing is a bit different: he argues that a more flexible frame actually allows the rider to put out more power.)

EXTRA: if you want to read more about planing, here are links from Jan Heine's Blog 1, Jan Heine's Blog 2 (both 2014). 2011 podcast, 2017 CyclingTips podcast. The latter podcast is a broad discussion of the value of frame stiffness vs flexibility.

wphamilton

Well some on BF might dispute the "sane" part, but I did study Physics long ago (minor concentration for my Mathematics degree) and I've considered this issue since about 8 or 10 years ago.

HTupolev

I've never seen a bicycle frame that sweats, or whose vascular muscles locally relax in response to temperature increases. Actually, I'm not sure that I've ever seen a bicycle frame with any vascular system at all!

It's clear that energy is lost in flex, but it's unclear how much. Considering that nobody has ever published numbers despite it being a hot marketing point for several decades, it seems likely that it's either so small that nobody has managed to measure it, or so small that sales teams don't want to tell anyone what the significance is. Actual attempts to distinguish it certainly aren't mentioned much by industry folks... Josh Poertner said that his testing was never able to distinguish the effects of frame flex.

As far as where the flex goes, I'd be curious to know how you ruled out forward motion. It seems like it would depend on where and how the legs were bracing the cranks while the frame unwound.

If all you've controlled for is groupset and "almost" geometry, it could be just about anything.

AnkleWork

If it's so clear, then it must be easy to show, and easy to measure. Where is the mechanism for elastic flexural energy loss in a steel or aluminum metallic frame, or a covalent carbon frame? Please advise.

HTupolev

Not necessarily. A bicycle is clearly gravitationally attracted to its rider, but if someone wasn't convinced of the existence of gravity, it would be quite difficult to use a bicycle to demonstrate this to them.

Elastic damping. When you flex these materials, some energy is lost to heat. When a real-world spring is released, it doesn't let out quite as much energy as was used to compress it.

Spoonrobot 

The available research makes it clear that there is essentially no energy lost to frame flex during a complete pedal rotation under power.

To the point of this thread; a fitting similar to a rear rack could be designed to provide increased stiffness to the rear triangle as well as the top tube and downtubes but would most likely need additional brazeons added to the frame.

AnkleWork

Gravitational interaction between a bike and rider is only "clearly" implied in Newton's equations -- very far from "clear" empirically. I guess you are saying that calculated gravitational interaction makes energy loss from elastic deformation and release "clear." OK, it's your rabbit hole.


According to the Bouschinger effect, damping occurs during plastic strain, not elastic.

unterhausen

I think you probably could stiffen up the rear triangle with the right rack, even if it had to be attached with something a little stiffer than p clamp.

But I really don't think it's worth the extra weight.

HTupolev

I'm not saying that gravity has anything directly to do with frame flex. I was just making an analogy.

I'm saying that the energy loss from elastic deformation is clear in the same way that gravitational interaction between bicycle components is clear: our knowledge of physics makes it obvious that it happens, but that doesn't mean it's easy to demonstrate or measure on the spot with a bicycle.

Could you educate me? I'm not familiar with what you're referring to. The Bauschinger Effect, to my understanding, refers to the effects of the micro-stresses that are left in a material that has been plastically deformed, i.e. how a metal bar that's been plastically stretched under tensile forces will plastically deform under less compression than would have been required before.

This may be my unfamiliarity with material science speaking (I'm EE...), but it's not clear to me what that has to do with whether or not materials experience internal frictions during elastic deformations.

wphamilton

To the extent that the body damps the reflex motion (thereby absorbing the energy of flex) the system does utilize all of these things. The metal itself conducts heat pretty well without capillary actions. But the main thing that you're overlooking, is that air flow over the surface is what cools the body, and the bike frame. Aside from evaporation of course.

The only justification for it depends on bad reasoning. People ask "where does the spring energy go?", can't think of an answer, and conclude "it must go to forward motion". Two answers of "where it goes" are mentioned even in this thread, and I can think of more in fact.

Kapusta

I am seeing an argument here that when a frame flexes, it is like a spring that stores that energy and then gives it back, and therefore you are not losing efficiency.

The flaw in that argument as I see it is that even if the frame does act as a perfect spring, and returns all the energy stored in it when flexed, it does not necessarily mean it returns the energy in the most useful way. For example, since the frame is releasing the energy at the same time as you are easing up on the force on the pedals, it may go into slowing your pedal stroke down more than making the bike go faster. In other words, the energy may get transferred from a very efficient spring (the frame) to a terribly inefficient spring (your legs and body).

On the other hand, the flex of the frame might actually let the frame store energy that might otherwise be wasted in the imperfect system that is our bodies, and if released at just the right time could lead to an overall increase in efficiency. I am thinking of something analogous to running blades.

And that is as much physics theory as I care to entertain on this.

squirtdad

OP what bike do you have (pics), how much doe you way, how hard do you ride, do you feel flex or noodly? Is this a real world issue you are having or this hypothetical? curious bike geeks want to know

Kapusta

Rear-end flex leading to poor tracking and tires rubbing chainstays under hard pedaling is an issue that often plagued lightweight full suspension mountain bikes in the past (they have gotten a lot better).

One solution I found to have some positive effect on the rear stiffness of my old MKIII was to convert the QR hub to 10mm thru-bolt and using this Hadley 10mm thru-bolt:



It did not result in any miracles, but It turned the rear end from feeling like an over-cooked, soggy noodle to an al-dante noodle.

woodcraft

I filled a flexy basketball backboard pole with concrete to stiffen it

& it worked, so you could do that...

HTupolev

The point I was making is that the human body avoids changes in temperature by dynamically adjusting its cooling systems. A human body in a 50-degree dim calm ambient environment will be at about the same temperature as a human body in an 80-degree dim calm ambient environment; but a bicycle frame in the latter case will be 30 degrees warmer than in the former. Similarly, extra power dissipation in a human body will have basically zero effect on the internal temp until it overwhelms the body's adjustment range per the conditions*, but extra power dissipation in a bicycle frame will pretty much always raise its temperature.

*In which case we can cook. Climbing in hot weather can become a serious heat sickness risk, for instance.

I'm not sure that this rebuts Heine's argument. He's pretty clear that he thinks that what he calls "planing" is dependent on the timing of the cyclist's power stroke and the characteristics of the frame.

OneIsAllYouNeed

Everything in the load path from your hands-to-feet and feet-to-ground contributes to the flex felt while riding. Specifically, the hands-to-feet stiffness is felt while sprinting or low cadence efforts. The feet-to-ground stiffness is felt during most pedaling. This load path includes shoes, pedals, cranks, BB, BB shell, chainstays, dropouts, hub, spokes, rim, tire, and sometimes the seat tube and seatstays. The flex you feel is dominated by the least stiff element in that series. You should figure out which part is the culprit before willy-nilly stiffening things. Likely candidates could be thin plate dropouts, overly crimped chainstays, cheap crankarms, loose spokes, or too low tire pressure. Which is it?
A tensioned steel cable parallel to the chainstays won't significantly increase the bending stiffness of the chainstays. Popsicle sticks (or fiberglass, carbon fiber, etc) epoxied to the sides of the chainstays will significantly increase their bending stiffness.
An extra set of stays from the seat tube to the rear dropouts adds lateral stiffness, as long as they're the same size (or larger) than the original chainstays. As long as they're kept low, there wouldn't be any impact on vertical compliance (ride quality).

wphamilton

I'm not addressing all that, except to note that the salient fact is that the heat energy is dissipated by wind blowing across the surface.

Heine doesn't make an argument. He speculates, and uses poor reasoning in his speculation.

HTupolev

There's evidence - and not just from BQ (it's been corroborated i.e. by Damon Rinard at Cannondale) - that something interesting is happening with the rider's pedaling and flex about the cranks. Heine's "planing" hypothesis didn't arise as an answer to the question of "where does frame flex go?", but of "why did reducing frame stiffness appear to increase the power output of the rider?"

I don't think that the "planing" hypothesis has been robustly demonstrated, and I have no particular amount of faith that it's correct. But I also don't see how it's "clear", as you claimed, that frame flex does not return to forward motion as the frame springs back.

I didn't miss that wind blowing across the surface of the cyclist and the frame cools them off; I thought it was too obvious to bother mentioning.

The difference - and what I pointed out - is that a human body uses this cooling mechanism very differently from a bicycle frame. The human body is constantly changing its surface properties and blood flow to alter the amount of heat that gets dissipated (in order to keep temperature constant), while the bicycle frame is basically passive.
Introduce a heat power source to a human body, and it will modify itself to dissipate heat better so that the temperature does not change; introduce a heat power source to a bicycle frame, and it'll become warmer by some amount that depends on the amount of heat and the conditions (such as ambient temperatures and the airflow over the frame).

wphamilton

The fact is that whatever method is to transfer heat inside the body - whether capillary or the more efficient conduction in metal - the actual cooling occurs at the surface, from convection of the wind. Some radiant, some evaporative, but mostly from the wind. ALL of the heat energy dissipated, no matter HOW it moves it moves around inside, occurs at the surface. We're mainly talking about convection.

Why is this the important thing to consider? It's because your body generates far more excess heat than you'd see in flexing the frame. After all, ALL of the energy originates there, and most of that is NOT utilized to flex the frame. And yet it is all drawn off by convection in the wind. Just like at the frame. Of course there is heat generated, the temperature does rise in the frame. But you should note that I said you would not "feel the frame getting hotter or otherwise note an obvious heating". The cooling will prevent that, just like it does with your body, for the same reasons (convection), because it's just not that much energy.

HTupolev

Okay, so your point was actually just that the power is too low relative to the heat dissipation ability of the frame to make a noticeable difference.

This is obviously true in the real-world case, although whether it would remain true in a hypothetical case where internal frame friction caused significant power loss depends on what we'd consider "significant." Note that one person you've been agreeing with in this thread - raria - has previously pondered whether flex about the cranks was slowing them down by three miles per hour, and has seemingly even purchased a new crankset in an effort to address the issue.

I agree that, with the frame being metal, there's probably a large range of hypothetical dissipative power loss in the frame which would be considered a meaningful performance loss to many people, while not being significant enough to cause frame heating that would be very apparent to a person.

wphamilton

Correct.

Any difference of that magnitude, if due to the frame or cranks, would necessarily be from some other effect of the flexing.

Reynolds 531 

I did this on a carbon fiber frame.

woodcraft

When you guys get this worked out,

you can look into the amount of heat generated when a hair is split.