Myth busters!

[bobswire]

http://janheine.wordpress.com/2011/10/03/science-and-bicycles-frame-stiffness/

[Fixed]

i knew that but i would never say it
cheers imho

[Black Dog]

It would be interesting to to get a test rig set up and see if there is a loss of power at the rear wheel from the input at the peddles between frames of various stiffness with all else being equal. You could run this at various power inputs and mimic the unequal inputs of power that a rider applies throughout the pedal stroke. The results would be very interesting.

[rain dogs]

Someone confirm this for me... or tell me I'm dense, but I read long ago that manufacturers were well aware of the modulus of elasticity retort (in other words.... the truth) but were in fact expressing "stiffness" as something of a ratio.

So, steel, sure, flexes a relatively standard amount across all steel tubing types, but the weight of the tubing types becomes less and less through SL-EL for example.

Another way of looking at it is comparing across materials.

The article doesn't really address the fact that you may create an alu. frame with the same stiffness as 531, because the alu. has super-oversize tubing.

So, let's say the stiffness is equal (the amount the frame deflects in the jig under weight), and the weight of the alu. is 300-400g less than the SLX frame.

As a ratio.... the alu. frame is "stiffer" by weight (the amount more stiff it would be if you oversized the tubing to weigh the same as the comparable frame).... although in actuality it's the same stiffness.

Maybe Ben can respond to that article? It would be great to hear from a builder.

[Chance]

Myth buster or more of the same? Seems like it’s just one more opinion without any data at all to support it.

He claims the more flexible frame is superior because it is preferred by 2 out of 3 riders in double blind tests but there is no science or data behind the assumption that performance is improved at all in any specific category. For all he and we know it could be that more flexible frames ride smoother and are therefore preferred but that doesn’t mean they are faster. Unless SPEED RELATED performance is measured the entire article is based on an opinion and nothing more.

One thing he claims indirectly is that builders were generally clueless about engineering and bought into the light weight steel hype. Most engineers have been stating for decades that all steel tubes have similar properties that affect flexibility and that making frames lighter just makes them more flexible unless tube diameters are increased substantially. For years builders kept using lighter (and therefore more flexible) tubes and kept insisting that their “latest” properties somehow made them better. Apparently the majority of buyers disagree.

[Fivethumbs]

No mention of how preferable flexible frames are during 50mph descents, which for me is just as an important aspect of a frame, as I climb and descend on just about every ride.

[Louis]

I don't know if Jan is a rat or a cat, but I'm pretty sure that he's not a structural engineer.

[palincss]

But riding in the Cascade Mountains as he does, you can be damn sure he does a lot of high speed mountain descents. And he definitely prefers flexible frames.

[Kontact]

I tend to agree with some of the article - bikes have been and many continue to be "too stiff". The only problem with that is that no one has a clear idea what the preferred stiffness should be that can be expressed mathematically. And what is fatiguing to one human might be completely different to another. After all, some people like John Tesch.

BUT, the tradeoff between diameter and liveliness isn't linear. Many oversized steel tubesets are definitely livelier than narrow, thick walled Columbus SP.

If you had to pick, too flexible is probably better than too stiff. But there are enough bikes out there now that everyone can seek (and pretend they found) the "correct" stiffness.

Combining a pedal/crank mounted power meter with a rear wheel Powertap would allow testing of some of the efficiencies of different frames, but I'd bet the results would be consistant only at the extremes.

[rice rocket]

Myth buster or more of the same? Seems like it’s just one more opinion without any data at all to support it.

You must be unaware, this is the Jan Heine way.

[learlove]

I think the industry has just about "jumped the shark" on the whole stiffness issue.

note: industry has also jumped the shark on compliance and weight. What they need to focus on is making a production bike that is not butt ass ugly. Hint: start by leveling the tt. Racing bikes are not supposed to look like my sister's 1983 huffy 10speed.

[akelman]

After all, some people like John Tesch.

Now that's an inflammatory statement. I'd like to see some evidence, sir, or I won't believe a word of it.

[rice rocket]

What they need to focus on is making a production bike that is not butt ass ugly. Hint: start by leveling the tt. Racing bikes are not supposed to look like my sister's 1983 huffy 10speed.

This has been discussed a million times over.

Do you hate your balls? More standover clearance is always welcomed. Also, Someone did a test (David Kirk?) of a level TT versus a sloping with all other things equal, and the sloping had a lower PMI, benefitting those who climb out of the saddle.

If the only downside is "learlove thinks it's ugly", I'll take it.

[learlove]

Do you hate your balls? More standover clearance is always welcomed.

If your smacking your balls your doing it wrong.

I don't know about you but I've been riding and racing since I was 13 (I'm 37 now). I've raced road, mtb, track and cross (cross - way before it was o so trendy cool) and never once in 20 plus years did my nut sack come in contact with my tt in anyway resembling a forceful manner.


And *** is PMI? Sounds like a fact or figure someone with a $10,000 bike and $2 legs would use.

[palincss]

You must be unaware, this is the Jan Heine way.

Pardon my French, but that's just plain horse$h1t.

The blind test article involved having a frame builder build three otherwise identical frames, differing only in tubing thickness, plus a number of test runs where riders swapped bikes with their times compared. That's a far cry from spouting an opinion sans data. There's plenty enough of that right here, indeed some posters make a habitual practice of it; and I don't see any of them going to that kind of trouble to back up their opinions.

[palincss]

If your smacking your balls your doing it wrong.


http://www.youtube.com/watch?v=hzPt3Y2E_Rc

[R2D2]

Columbus SLX was primarily developed because strong sprinters were cracking SL above the BB. So the primary goal was to keep the frame from breaking. SL steerer tubes had the riffled ridges.

[BillG]

Did Grant Peterson create Jan Heine in a lab to make him (Grant) seem reasonable by comparison?

[Kontact]

Did Grant Peterson create Jan Heine in a lab to make him (Grant) seem reasonable by comparison?
If that were true, wouldn't Grant's bikes be built of something nicer riding than indestructible thick wall tubing?

[jr59]

Did Grant Peterson create Jan Heine in a lab to make him (Grant) seem reasonable by comparison?


Pretty funny stuff here! :banana: :beer:

[staggerwing]

Why the daggers? He didn't say anything particularly revolutionary, and got the engineering terms correct. Yeah, he can come across as a retro-grouch, but it seems to work for him (eg his PBP time wasn't anything to laugh at).

Just a couple of things to add. Most importantly, final stiffness of the frame is significantly more dependent on diameter of tubing selected, and construction geometry, than material. Design trumps material.

Specifically, deflection of a beam is inversely proportional to the modulus of the selected material (E), and moment of inertia of the beam (I). However, moment of inertia of a tube is 1/4*(Outer radius)^4-(Inner radius)^4). That is radius to the forth power! Relatively small changes in tube diameter and wall thickness translate into significant differences in tube stiffness.

Also, deflection of that same simply supported beam is proportional to length to the third power. So once again, relatively small changes in frame size, can make for a larger than anticipated difference in overall stiffness. The ramifications can be interesting. For example, a compact geometry can likely result in stiffer frame, for a given size, while using less material.

Finally, perhaps a bit pedantic, but compliance is just the inverse of stiffness. In general, selection of term is related to which one you would prefer more of. Is your ride shaking your fillings loose, then go for something more compliant; somehow sounds more elegant than less stiff. Yeah, there is a joke in there somewhere.

Ride what works for you. :beer:

[Chance]

Pardon my French, but that's just plain horse$h1t.

The blind test article involved having a frame builder build three otherwise identical frames, differing only in tubing thickness, plus a number of test runs where riders swapped bikes with their times compared. That's a far cry from spouting an opinion sans data. There's plenty enough of that right here, indeed some posters make a habitual practice of it; and I don't see any of them going to that kind of trouble to back up their opinions.
Perhaps there is more information somewhere but the article linked in the OP does not have any supporting data nor does is reference any and as such is opinion based and not science.

Based on common use of language it appears only 3 riders were involved in testing. It doesn’t appear that two out of three meant something like 200 out of 300 preferred less stiffness because he wouldn’t have stated the “third rider” couldn’t tell a difference. Do you think the personal preferences of just three riders is enough to settle this issue? Particularly when one was neutral at best and we know nothing about their riding skills, experience, or power?

More importantly he does not mention the relative stiffness of the three bikes tested. For all we know based on article all three bikes could have been made much stiffer than normal to the point where most riders would prefer the least stiff of the three. In essence for all we know based on article the other two bikes could have been stiffer than a Cannondale CAAD3.

He may be right but what he states is not science. It’s opinion. He states a premise and then defends its value based on the statement itself.

[Chance]

Finally, perhaps a bit pedantic, but compliance is just the inverse of stiffness. In general, selection of term is related to which one you would prefer more of. Is your ride shaking your fillings loose, then go for something more compliant; somehow sounds more elegant than less stiff. Yeah, there is a joke in there somewhere.
Ride compliance (normally meaning vertically) and stiffness (normally meaning laterally and torsionally) are not the same. Check out the test rig at the top of the article and you’ll see that the frame is being tested for lateral deflection and twisting by applying a load at the bottom bracket to simulate cranking down hard on pedals, but in this rig the seatstays are not much of a factor.

It can then be concluded that if a SoftRide or Serotta DKS were tested in this rig that “compliance” versus “stiffness” data would not be proportional to that of traditional bikes. In these cases compliance is not the inverse of stiffness as these labels are normally used by cyclists to describe bikes.

This difference between compliance and stiffness also applies when materials like carbon fiber is involved due to its properties.

In essence without taking a frame’s design into account this test rig doesn’t differentiate equally between compliance and stiffness. To assume they are directly connected doesn't seem like a good idea.

[palincss]

More importantly he does not mention the relative stiffness of the three bikes tested. For all we know based on article all three bikes could have been made much stiffer than normal to the point where most riders would prefer the least stiff of the three. In essence for all we know based on article the other two bikes could have been stiffer than a Cannondale CAAD3.

We know from the article what tubing was used in the various bikes, and anyone familiar with frames made of 7/4/7 standard diameter butted steel tubing will know instantly we're not talking about frames "stiffer than a Cannondale CAAD3".

[staggerwing]

Ride compliance (normally meaning vertically) and stiffness (normally meaning laterally and torsionally) are not the same. Check out the test rig at the top of the article and you’ll see that the frame is being tested for lateral deflection and twisting by applying a load at the bottom bracket to simulate cranking down hard on pedals, but in this rig the seatstays are not much of a factor.


You are reading way more into this than I suggested, literally and figuratively.

As an engineer, when referring to a single structure, even if complex in configuration, like a bicycle frame, I would never suggest the term "compliance" for one plane and "stiffness" for another orthogonal plane. This whole vertically stiff/laterally compliant talk is more marketing drivel than anything. While properties can be tailored by direction, they cannot be decoupled. As technical terms, I stand by my original remarks.

I'm also lost by "This difference between compliance and stiffness also applies when materials like carbon fiber is involved due to its properties."

While a very unique, and technically valuable material, there is nothing magical about carbon. Due to configuration, typical material properties are very directional. One of the more popular raw configurations, a flexible sheet of carbon fibers pre-impregnated with the proper ratio of matrix (generally elevated cure epoxy), gives the designer some extra "degrees for freedom." Overall wall shape, as well as diameter, thickness, and fiber direction can be adjusted.

Might add that composite structures, of almost any fiber and matrix imaginable, tend to be somewhat viscoelastic. The unloading path is not quite the same as the loading path, and some energy is lost in the process. The difference is related to the frequency of loading. Again, just a material property, neither good nor bad.

However, just having the ability to adjust these things is not necessarily a positive. For a simple analogy, more tone controls do not equal better sound.

[MadRocketSci]

my pontifications...

i still think stiffness is "all in your head"....as long as the frame is stiff enough to not wiggle so much as to change the dynamics of the bike, fatigue the material, etc...ie, it's "stiff enough"...

what makes you accelerate on the bike? the force of the wheel contact patch against the road. What drives that force? Torque on the bottom bracket...

Torque on the bottom bracket only comes from a force on the pedal that is perpendicular to the crank. Anything else does not contribute to torque. Those non-perpendicular forces go into the bottom bracket and may cause some deformation. When released, the deformation goes away...

those non-perpendicular forces also produce a reaction force on the rider. So if these forces point forward, the rider goes backward, making it seem as if the bike is "leaping forward"...a frame that deforms less will seem to react quicker from this due to stiff coupling, whereas a more flexible frame will take a little more time due to a more delayed reaction from a softer coupling. Since the rider and bike are a closed system, the overall motion with respect to the road is not changed. Ie, you're not going faster from this....imo.

So it all depends what you like...I try to pedal smooth, so I think I don't benefit much from a stiff frame. I'll notice it sometimes when I'm trying to jerk the pedals around, but mostly, an overly stiff bike actually feels sluggish and heavy. to me. others get excited by the reaction and may be feel motivated to go faster.

So, any measured data showing differences in loop time with different bikes is probably measuring the preference of the particular rider to a level of stiffness. imo.

on a technical note, stiffness of a cylindrical thin-walled beam is proportional to the third power of its average radius and linearly with thickness. This is an engineering approximation made to the equation given above.

[schneiderrd]

I think that the discussion of stiffness vs performance comes down to frame flex especially at the bottom bracket. Before I got my Serotta, I was riding an off the shelf Univega. When I started riding my Serotta, I noticed that there was a lot less frame flex when climbing out of the saddle. I always felt that the Serotta was a better climbing bike since more of my pedaling effort went into moving the bike forward instead of flexing the frame.

[ergott]

My $0.015.

Not everyone wants the same thing out of a frame/bike. Not everyone builds bike the same way.

Find one you like and ride it. Forget everyone else. There is no one equation that will make a great bike.
:beer:

[oldpotatoe]

Columbus SLX was primarily developed because strong sprinters were cracking SL above the BB. So the primary goal was to keep the frame from breaking. SL steerer tubes had the riffled ridges.

Or because they started brazing FD tabs there, making for cracked tubes. Rifling gave some support for the tab, cuz you are right, SL tubes were breaking there.

[Gummee]

If your smacking your balls your doing it wrong.

I don't know about you but I've been riding and racing since I was 13 (I'm 37 now). I've raced road, mtb, track and cross (cross - way before it was o so trendy cool) and never once in 20 plus years did my nut sack come in contact with my tt in anyway resembling a forceful manner
I'm with you. ...and I've generally ridden with TTs right 'up there.' Starting in 87 when I bought my Rockhopper to today, every single time I've fallen I've either been going sideways or forward OVER everything.

IME standover is completely irrelevant.

M
the short legs and long-ish torso guy

[shankldu]

Rice rocket is correct anytime you shorten a tube it becomes more stiff and puts more power to the ground , no one can dispute this. If you try to bend a long tube its posible because you have mechanical advantage ,shorten the tube and you dont have that advantage and cannot bend the same tube. I have both geas on different bike and the sloping has acceleration and climbing advantages ,but is not as smooth and you feel the road more for sure. Im referencing the seatstays here even more so than the top tube .

[gdw]

"I don't know about you but I've been riding and racing since I was 13 (I'm 37 now). I've raced road, mtb, track and cross (cross - way before it was o so trendy cool) and never once in 20 plus years did my nut sack come in contact with my tt in anyway resembling a forceful manner "

Consider yourself fortunate......

[rain dogs]

Rice rocket is correct anytime you shorten a tube it become more stiff and puts more power to the ground , no one can dispute this.

Sloping top tubes are also less aerodynamic than horizontal. (that's why TT bikes still have horizontal tt). But these are design issues.

I think the article (maybe I should read it again) was talking about an "If all design aspects are relatively equal" situation.

He's not talking about how to design the stiffest bike, but that stiffness doesn't differentiate to a noticeable amount across the same material, using similar geometry, and tubing sizes, going against that idea of "531=x% stronger, x% lighter AND x% stiffer"

perhaps, and more likely, he's thinking as a tubeset and not a built frame.

I would have liked to see more data with the article though.

[MadRocketSci]

Rice rocket is correct anytime you shorten a tube it become more stiff and puts more power to the ground , no one can dispute this.

first part, sure...second part, disputable...esp for the top tube...

probably more an issue for the chain stays...more specifically, the non-drive side chainstay, which bends (the ds one is under compression). if the nds chainstay is flexing enough to change the direction of the rear wheel (non-aligned with the front)...then that is definitely not stiff enough, and you're losing power....

[Chance]

As an engineer, when referring to a single structure, even if complex in configuration, like a bicycle frame, I would never suggest the term "compliance" for one plane and "stiffness" for another orthogonal plane. This whole vertically stiff/laterally compliant talk is more marketing drivel than anything. While properties can be tailored by direction, they cannot be decoupled. As technical terms, I stand by my original remarks.

That was exactly the point of the reply: That different designs can decouple vertically compliant from laterally and torsionally stiff. In that case we are no longer comparing a “single structure” on the same basis, right? It’s not marketing, it’s engineering. What’s the drivel part?

Examples include beam bikes (SoftRide and Titanflex), curved seat stays (DKS, Terraplane, and numerous others with less curvature), collapsing seat stays (a la Moots YBB) and other lesser known designs.

You are indeed correct that if we simply double the stiffness of all tubes in a given design we will be making the bike stiffer vertically and laterally. However, the minute the design is different then it’s comparing apples and oranges. And that's why we have different designs.

Even within a given general design type there are variations. Making the down tube stiffer doesn’t have the same consequence as making the seat stays stiffer. These two options would shift vertical versus lateral stiffness ratios. One should be good and the other not.

[rice rocket]

Sloping top tubes are also less aerodynamic than horizontal. (that's why TT bikes still have horizontal tt). But these are design issues.


Cervelo, the biggest proponent of the level top tube for maximum aero efficiency, didn't make a level top tube on their latest S5. And when asked why, they said it didn't make enough of a difference.

:eek:

[forrestw]

This blog article calling itself 'science' is quite a stretch, yes the author is well respected and all but the idea that a double blind test with only 3 subjects observing the frames in question offers effectively zero statistical validity.

Yes, (lateral) frame stiffness is too often considered a holy grail, however there's clear truth that whippy frames are going to waste some of the rider's output. Whether that's significant is another matter. I would lean to probably not, however my experience with too-soft frames was experienced on my '80s bikes where the BB shells flexed so much that I had to trim the FD frequently enough to be annoying. (I eventually figured out that the underlying cause was due to poor construction techniques, not tubing choice.)

A small part of that annoyance was due to concern that some power was being lost.

Look at the limiting case, if your seat and down tubes were made of spaghetti .. ok if they flexed inches instead of thousandths of inches .. I think it's pretty clear that you would be losing most of the power of your pedal stroke. Small deflection is basically no different, every bit of BB shell flex in response to pedaling forces represents some lost energy.

But really, the need to trim the FD to avoid rubbing is the main thing I don't like about flexible frames, riding them as fixed-gear solves that problem nicely.

[e-RICHIE]

Or because they started brazing FD tabs there, making for cracked tubes. Rifling gave some support for the tab, cuz you are right, SL tubes were breaking there.


^

this is why (SLX) atmo.

ps

arrange disorder

:o :o :o
:D :D :D
;) :p :cool:

[palincss]

This blog article calling itself 'science' is quite a stretch

Calling an article published Bicycle Quarterly a "blog article" is itself quite a stretch.

[staggerwing]

That was exactly the point of the reply: That different designs can decouple vertically compliant from laterally and torsionally stiff. In that case we are no longer comparing a “single structure” on the same basis, right? It’s not marketing, it’s engineering. What’s the drivel part?

IMHO the drivel is in suggesting that vertically compliant and laterally stiff (or is it the other way around ;) ) is some sort of holy grail.

In a solid structure, bending in orthogonal planes cannot be decoupled from torsion. Significantly decreasing stiffness in one direction is going to adversely affect torsional stiffness. Adding in a mechanism, like the suspensions found on mountain bikes, changes matters entirely.

What sort of system parameters do you want to optimize? If you simply want to go faster, then perhaps all your chips should be placed in getting your body into producing maximum wattage in the most aerodynamic position obtainable, not to mention, in producing more wattage.

The beam bikes just hang the seat off of a passive suspension component The extra compliance is only in the seat.

The DKS rear is another animal, and has two features to consider. One, it is tuneable, in that you can change the bolt on bits to change ride qualities, without buying a new rig. Nice idea. Second, those add on bits are viscoelastic in response, acting a little like a shock absorber. Everything I've ever read, implies they are great descending bikes, with rearends that stay well planted under variable road conditions. That certainly seems reasonable. Haven't seen anything suggesting they are faster or more efficient under power.

Talk to people that play with racing cars. Stiff is only advantageous to a point. Too stiff, and the car simply bounces around on the surface and is ineffective at getting the horsepower applied to the pavement.

Is there also not some irony, as cyclists, in wanting a stiffer frame, yet a more supple (ie compliant) tire?

Me? I just want more time on my bike. Two days without commuting and I'm about stir crazy.

[David Kirk]

Might it be safe to say that neither the stiffest nor the softest bikes that can be made are ideal and that there is something in the middle that will work better than the ones at either end?

Can it also be said that the 'proper' stiffness for one rider might not the the proper stiffness for another rider of the same size and weight - the differences owing to riding style, strength, skill, rider preference, surface the bike is being used on.....etc?

IMO this, like most things in life, is nuanced and stiffer isn't always better nor is softer always better. Nuanced things don't really lend themselves to rules of thumb but these false rules of thumb do make good marketing so they will persist.

Maybe Goldie Locks had it right? I think so.

dave

[Fixed]

nice post dave
cheers

[staggerwing]

Might it be safe to say that neither the stiffest nor the softest bikes that can be made are ideal and that there is something in the middle that will work better than the ones at either end?

dave

You said it in far fewer words. Hats off

One fun thing about engineering, is most problems are underdetermined, and can have any number of satisfactory solutions. There can be solutions optimized for a particular set of parameters, but my solution might not look like yours, while both are perfectly functional.

Marketing paints engineering as science and its not. Many of the equations and relationships applied are merely codified and tabulated experience. These relationships also often hold only under carefully controlled conditions. I'm OK with these limitations.

I'm also perfectly comfortable stating an experienced frame craftsman, such as yourself, has a far better chance of hitting the sweet spot than a bookish engineer.

[Fixed]

that is a secret
cheers

[Chance]

In a solid structure, bending in orthogonal planes cannot be decoupled from torsion. Significantly decreasing stiffness in one direction is going to adversely affect torsional stiffness. Adding in a mechanism, like the suspensions found on mountain bikes, changes matters entirely.
On this we can agree.

It’s not engineering but rather common sense that the leaf springs on a pickup truck are vertically compliant and laterally stiff (relatively speaking). If we modified a pickup by rotating the leaf springs 90 degrees the suspension would be vertically stiff and laterally compliant. In other words it would ride harshly and steer poorly, two characteristics no one wants. The springs' torsional stiffness would remain the same in both cases.

Likewise if a bike builder uses an oval tube he can place it so that it’s stiffer in one direction versus the other. The end product should be markedly different depending on his choice. Complicating matters further is transitioning from round to oval at designated points to provide more or less stiffness as deemed desirable.

We can also agree that mountain bike suspension changes matters entirely. However, it also seems reasonable to conclude that a little suspension may also change matters to a lesser degree. Again, based on common sense rather than engineering.

Is there also not some irony, as cyclists, in wanting a stiffer frame, yet a more supple (ie compliant) tire?

Don't see irony in this. It seems completely consistent with the desire to have vertical compliance and lateral/torsional stiffness. A compliant tire is often like a pneumatic spring.

[Bob Loblaw]

Maybe I'm slow, but the crux of the article seemed to be that yield strength, not stiffness, determined which frame "performed" better. What's the correlation between high yield strength and performance (apart from frames needing to reliably withstand the power output of a cyclist)? How did they quantify "better performance?" Time trial times? Watts at the rear wheel? And were the frames compared of equal (or at least comparable) dimension, weight, and componentry?

I'd like to see more data before I toss out the window something that makes sense to my 12th grade level understanding of physics, i.e., non-flexible things (billiard balls, hammers, ping pong balls) are better energy conduits than squishy things (handballs, cardboard, springs).

BL

[learlove]

"I don't know about you but I've been riding and racing since I was 13 (I'm 37 now). I've raced road, mtb, track and cross (cross - way before it was o so trendy cool) and never once in 20 plus years did my nut sack come in contact with my tt in anyway resembling a forceful manner "

Consider yourself fortunate......

that kids obviously doing it way wrong - how do you wack your goods on a the tt on a bike that has no tt?

[Chance]

Maybe I'm slow, but the crux of the article seemed to be that yield strength, not stiffness, determined which frame "performed" better. What's the correlation between high yield strength and performance (apart from frames needing to reliably withstand the power output of a cyclist)? How did they quantify "better performance?" Time trial times? Watts at the rear wheel? And were the frames compared of equal (or at least comparable) dimension, weight, and componentry?

I'd like to see more data before I toss out the window something that makes sense to my 12th grade level understanding of physics, i.e., non-flexible things (billiard balls, hammers, ping pong balls) are better energy conduits than squishy things (handballs, cardboard, springs).

BL
Bob, it's doubtfull that is the case because he correctly identified that when a frame is made of stronger steel (having a higher yield point) it takes more force to bend it permanently (as during alignment) but the frame’s elastic stiffness is exactly the same. It then follows that when comparing two identical frames except that one is made of stronger steel there will be no difference in ride stiffness unless the rider can actually bend the frame beyond the yield point. And obviously that doesn’t happen very often short of a crash.

Think of it this way. One frame with 50,000 PSI steel and one with 100,000 PSI steel (yield strength). If a rider only stresses the frame up to 20,000 PSI then the two frames will flex exactly the same. However, when the builder tries to align the frames he has to flex the 100,000 PSI frame twice as far before it takes a permanent set. To the untrained person (the article suggests that would be builders) it would appear the stronger frame is more flexible as in less stiff.

In essence the ability to flex is not analogous to stiffness. They are related but not interchangeable.

[Chance]

BL
Your post does bring up the question of how the test bikes were made to have different stiffness. Since it’s unlikely they used different diameter tubes because then it wouldn’t be much of a blind test it follows that they probably used different wall thicknesses to vary stiffness. And obviously that would change the frame’s weight. And perhaps two out of three people testing the bikes simply liked the less stiff bikes because they weighed less.

If the test had included lighter “and” stiffer at the same time maybe the results would be different.

[staggerwing]

Maybe I'm slow, but the crux of the article seemed to be that yield strength, not stiffness, determined which frame "performed" better. What's the correlation between high yield strength and performance (apart from frames needing to reliably withstand the power output of a cyclist)? How did they quantify "better performance?" Time trial times? Watts at the rear wheel? And were the frames compared of equal (or at least comparable) dimension, weight, and componentry?

I'd like to see more data before I toss out the window something that makes sense to my 12th grade level understanding of physics, i.e., non-flexible things (billiard balls, hammers, ping pong balls) are better energy conduits than squishy things (handballs, cardboard, springs).

BL

The terminology can be a little confusing if you haven't been around it. Let me try to simplify by an example.

Get any three steel tubes, of any alloy you might dream up, Reynolds 531, 953, Columbus SLX, whatever, as long as they are identical in dimension. Inner diameter, outer diameter, butt length, and overall length must be the same. Clamp the end of each identically in a pipe clamp such that the project horizontally: a cantilever beam in mechanics of materials speak.

Now, start hanging weights off of the free end of each. It won't make any difference how you do it, as long as each setup is identical. What you will find is that each tube will deflect the same exact amount, for the same exact weight. That is because their modulus, E (basically a spring rate for a material)), is the same. And, because the deflections are identical, the wall stresses in each tube will be identical. Nothing interesting yet.

Now, keep adding weight to the ends, a little at a time. Remove them between additions to check for permanent deformations. At some point, the tube with the lowest material strength will not return to its original, straight, undeformed state. At that point, you have exceeded the strength of the material and a "plastic" deformation has taken place. Conversely, any deformation that returns to the original shape after removal of load is an "elastic" deformation.

Keep adding weight if desired. The tube that can support the most weight, without a plastic deformation, is made of the strongest steel. However, all of the steel tubes, would experience the same exact deflection, for a given weight, as long as you stayed within the elastic range.

Lets turn the tables a bit. Lets say we wanted to make the most out of our material selection, and wanted to develop a design that takes them to 50% of their yield strength (a 2x factor of safety). Now pick two more tubes, say 531 and 953 again, just because they are known quantities. Keep the outer diameter each the same, which is generally how it is done anyway. Mount as before and start hanging weights. Since we are designing for a particular application, say a bicycle, the weights are going to remain the same. However, the steel in the 953 tube is considerable stronger than that in the 531, and to reach 50% of its yield strength is going to take a larger deflection. The only way the larger deflection is going to happen, for a given weight and tube OD, is to increase the tube ID, thereby lowering wall thickness.

In the second example, the 953 TUBE is more flexible than the 531 TUBE, although the modulus of elasticity for the MATERIAL is identical. Flexibility of a STRUCTURE, such as a tube or a bicycle, is not the same as the flexibility of the MATERIAL that made the structure.

A stronger steel allows the builder to use less of it, which makes for a lighter bicycle. If tubing outer diameters were the same, then the inner diameters would increase, wall thickness would decrease, and the FRAME built of higher strength steel would be the more flexible structure.

Just to cover my backside, to really take advantage of a higher strength steel, one would go to oversize tubing. By adjusting outer diameter, as well as wall thickness, one could build a lighter frame out of higher strength steel that is every bit as stiff as the one built of lower strength steel.

Does that help in any way?

PS A flexible structure does not have to be an energy sink (as in black hole).

[Chance]

PS A flexible structure does not have to be an energy sink (as in black hole).
It is always accepted that’s not the issue because steel would return well over 99% of energy anyway. The issue that always comes up is how does the added frame flexibility affect the rider’s ability to generate pedaling power.

Some think it helps and others don’t. Like most topics in cycling it comes down to personal opinions because there is little data to substantiate claims on either side of argument.

[zap]

A stronger steel allows the builder to use less of it, which makes for a lighter bicycle. If tubing outer diameters were the same, then the inner diameters would increase, wall thickness would decrease, and the FRAME built of higher strength steel would be the more flexible structure.



Every steel frame builder I've spoken too said the exact same thing.

[Bob Loblaw]

I do get the difference between yield strength and stiffness, but the higher yield strength would offer a trade-off in performance...thinner tubing walls--->lower weight ---->loss of frame stiffness.

Sheldon Brown is one of those who maintains that stiffness is an over-emphasized issue in steel bikes, because whatever pedal stroke energy goes to flexing the frame comes back at the end of the pedal stroke.

I agree to a point. But back to my 12th grade physics, you always lose a small amount of energy when the frame flexes and then resumes its shape. That's why springs get warm when they're repeatedly flexed. The spring absorbs kinetic energy and dissipates it in the form of heat.

The loss is small in any quality bike, but it's there. In my brain, it logically follows that the less flex in a frame, the less power loss. A frame with lower stiffness and higher yield strength may have other desirable ride characteristics that supersede the infinitesimally lower energy loss of a stiffer frame, but in terms of outright mechanical efficiency, I am still convinced a stiffer frame, all else being equal, is better for turning leg power into forward motion.

It is always accepted that’s not the issue because steel would return well over 99% of energy anyway. The issue that always comes up is how does the added frame flexibility affect the rider’s ability to generate pedaling power.

Some think it helps and others don’t. Like most topics in cycling it comes down to personal opinions because there is little data to substantiate claims on either side of argument.

[rain dogs]

Cervelo, the biggest proponent of the level top tube for maximum aero efficiency, didn't make a level top tube on their latest S5. And when asked why, they said it didn't make enough of a difference.

:eek:

That's a specious argument. The S5 is a road frame, not a TT frame. The question with the S5 isn't whether or not a level top tube would be more aero, or make enough of an aero difference, it's whether or not the level top tube makes enough of a difference to take away Cervelo's high top-tube road geometry, for maximum fit and hand-positions.

We're getting a little off topic, but if I can bring it back, we're asking questions in isolation, not in a complex system.

In isolation: a level top tube is more aero than a sloping.

In isolation: two relatively equal dimensioned steel tubes w/ different yield strength and weight, are the same stiffness or the lighter less so and not more (or so the article author claims)

For everyday riders, like myself, probably none of this makes "enough of a difference". But hey, I still use a square taper Campy BB, because if it was, at one time, stiff enough for Mario Cipollini, it'll always be more than stiff enough for me. What works more for me, is the simplicity, serviceability and reliability of the design. In other words, BB30 wouldn't make enough of a difference in stiffness (to me).... although it's certainly stiffer.

The rest is just marketing and hundredths of a second (which my analog watch doesn't display). I'm not being a retro grouch, ...just a realist with my capabilities and the significance of my "best times"

[staggerwing]

It is always accepted that’s not the issue because steel would return well over 99% of energy anyway. The issue that always comes up is how does the added frame flexibility affect the rider’s ability to generate pedaling power.

While I think we agree on the technical principles; posts here and elsewhere suggest that "always accepted" certainly doesn't apply to everyone.

I really don't have a horse in this race, as to which is better, just trying to clear up some basics. There is a lot of voodoo marketing out there, and while poking holes is fun, generating a smarter customer is better.

[Louis]

thinner tubing walls--->lower weight ---->loss of frame stiffness

That's why they go to larger diameter tubes, to restore the bending stiffness, so you don't necessarily have to loose frame stiffness if you go to thinner walls. And since you get large increases in I for relatively small increases in D you can get the stiffness you're looking for without adding lots of material. (All this ignores axial loads, which I don't think are that high.)

Eventually buckling and/or crippling become an issue, but we won't bother going there.

[Chance]

The loss is small in any quality bike, but it's there. In my brain, it logically follows that the less flex in a frame, the less power loss. A frame with lower stiffness and higher yield strength may have other desirable ride characteristics that supersede the infinitesimally lower energy loss of a stiffer frame, but in terms of outright mechanical efficiency, I am still convinced a stiffer frame, all else being equal, is better for turning leg power into forward motion.
Bob, you won’t get an argument from me on that issue. However, it has to be due to the rider’s biomechanics and not the steel frame itself absorbing energy. A steel frame when acting like a spring returns practically all the energy that is put into it, so very little (insignificant amount) is turned into heat.

This entire discussion seems highly flawed from the get go. When was the last time anyone stated that their full suspension mountain bike’s frame was too stiff? It’s a crazy notion without merit. Setting aside suspension tuning how could a mountain bike’s frame be too stiff (other than it may make it too heavy)? If the ride was great due to the suspension then how would a stiff frame make it undesirable? It’s more likely that when riders bitch about a frame being too stiff it’s because it rides awful and not because it takes more effort to pedal it. Language gets in the way of discussing this as it should be.

In any case it’s hard to get over the silliness of discussing stiffness versus flexibility without first defining what is desirable and then avoiding using the terms interchangeably.

[Chance]

Eventually buckling and/or crippling become an issue, but we won't bother going there.
Not until the next time someone drops a small item on their top tube and asks about how to get rid of the dent.

[chad_k]

Your post does bring up the question of how the test bikes were made to have different stiffness. Since it’s unlikely they used different diameter tubes because then it wouldn’t be much of a blind test it follows that they probably used different wall thicknesses to vary stiffness. And obviously that would change the frame’s weight. And perhaps two out of three people testing the bikes simply liked the less stiff bikes because they weighed less.

If the test had included lighter “and” stiffer at the same time maybe the results would be different.

Obviously, you're interested in the articles. Why not actually read them?

They did a couple different tests. One used standard diameter for all bikes (2 bikes with 7-4-7 main tubes and one with 9-6-9 main tubes). The other test added a 7-4-7 OS tubeset bike. For that test, they used foam wrapping to disguise the bikes. All bikes weighed the same. Weights were added to lighter bikes.

Maybe Jan is sharper than you think?

[MadRocketSci]

the only thing turning your rear wheel is torque about the bottom bracket...everything else is doing something else...flexing or not flexing the frame, whatever, but not turning your wheels.

i imagine that most flex comes near the bottom of the down stroke. At 6 o clock you can bounce and stomp all you want on the pedal, it's not going to make you move. Your stiff bb might not flex. Your old steel bike might flex. Did you get more acceleration/energy from your stiff bike? No. You stored some useless energy in the frame and then got it back. You can store the same amount in a stiffer frame with less deflection.

torque alone will not flex your bottom bracket. All those forces that don't generate torque, ie, those that are parallel to your crank, will flex the bb. Can riders generate pure torque? I believe that is the aim of "pedaling in a circle," but when you're jerking along in a sprint you are probably generating forces that are parallel to the cranks. Especially from 3 to 6 o' clock.

So the non-bb-torquing forces going into flexing your frame, would not have produced an acceleration. with respect to the road. it might make you accelerate with respect to your bike, but the whole package (bike + rider) is unaffected.

I am challenging the notion that frame flex, less than that which noticeably alters a rider's pedaling dynamics or the relative positions and orientations of the wheels, results in a loss of forward energy. I would be very interested in arguments that can be made otherwise.

[Chance]

Obviously, you're interested in the articles. Why not actually read them?

They did a couple different tests. One used standard diameter for all bikes (2 bikes with 7-4-7 main tubes and one with 9-6-9 main tubes). The other test added a 7-4-7 OS tubeset bike. For that test, they used foam wrapping to disguise the bikes. All bikes weighed the same. Weights were added to lighter bikes.

Maybe Jan is sharper than you think?
Didn't say he wasn't in the first place, but thanks for the heads up on the other articles.

For what it's worth, lots of sharp people are often wrong so it's hard to use that as a guide for what one should believe and what we should question. Very smart people disagree all the time when seeing the same data. Just look at our government to see how differently smart people can see the same issue.:beer:

[Chance]

I am challenging the notion that frame flex, less than that which noticeably alters a rider's pedaling dynamics or the relative positions and orientations of the wheels, results in a loss of forward energy. I would be very interested in arguments that can be made otherwise.
OK. Frame flex in itself doesn't prevent greater forward motion. Instead it prevents the rider from developing as much power.

All modern bikes used in racing today are much stiffer (laterally and torsionally) than those used 30 or 40 years ago. If stiffer wasn't an advantage why would they make them? We don't need science or engineering to know that no one is racing tours on light flexy steel frames.

[Charles M]

Mods,

Please re-title this thread "myth creators"


Thank you.

[chad_k]

Didn't say he wasn't in the first place, but thanks for the heads up on the other articles.

For what it's worth, lots of sharp people are often wrong so it's hard to use that as a guide for what one should believe and what we should question. Very smart people disagree all the time when seeing the same data. Just look at our government to see how differently smart people can see the same issue.:beer:

I was responding to your comment about the experimental design. All of issues you raised were accounted for in those experiments.

I agree with you that there is often disagreement on data interpretation. That's especially relevant for the BQ articles because of the small sample size (number of bikes (including geometry, materials, sizes) and number of riders).

[Fixed]

their was an irishman that could ride very fast ,
he won a lot
liked vitus
i think cos they were stiff ?



not
cheers

[1centaur]

the only thing turning your rear wheel is torque about the bottom bracket...everything else is doing something else...flexing or not flexing the frame, whatever, but not turning your wheels.

i imagine that most flex comes near the bottom of the down stroke. At 6 o clock you can bounce and stomp all you want on the pedal, it's not going to make you move. Your stiff bb might not flex. Your old steel bike might flex. Did you get more acceleration/energy from your stiff bike? No. You stored some useless energy in the frame and then got it back. You can store the same amount in a stiffer frame with less deflection.

torque alone will not flex your bottom bracket. All those forces that don't generate torque, ie, those that are parallel to your crank, will flex the bb. Can riders generate pure torque? I believe that is the aim of "pedaling in a circle," but when you're jerking along in a sprint you are probably generating forces that are parallel to the cranks. Especially from 3 to 6 o' clock.

So the non-bb-torquing forces going into flexing your frame, would not have produced an acceleration. with respect to the road. it might make you accelerate with respect to your bike, but the whole package (bike + rider) is unaffected.

I am challenging the notion that frame flex, less than that which noticeably alters a rider's pedaling dynamics or the relative positions and orientations of the wheels, results in a loss of forward energy. I would be very interested in arguments that can be made otherwise.

No engineer here, but this post raises two points that I think are generally ignored by the "steel flexes but you get it back" arguments.

The first one is the one I have made in the past - there is no reason to believe that a frame flexed to the side returns energy to driving the wheels forward when it unflexes (nor does the energy flow back into our legs). All statements about "getting the energy back" seem to ignore that point. As I once posted to make the point, is there any reason to think the return-flex doesn't drive the wheels backward?

The second one is more interesting, which is that the energy wasted in flex may not have been spent on driving the wheels forward in the first place, so who cares other than as it relates to biomechanics? That seems to be a measurable question via a pedaling machine where identical pedaling force can be delivered consistently by an unflexing piston across a range of varyingly flexible frames and the wattage measured at the hub, but humans are not machines. Human physiology reacts to frame resistance with bones, joints and ligaments that themselves flex; the stiffer the frame perhaps the greater the human deflection, sometimes in ways that reduce power, sometimes in ways that enhance recovery, I bet. Good luck isolating those elements. Should we be seeking an ideal ratio of frame flexibility to human physiology?

While I think much of this doesn't much matter to most of us, I'm sure we've all experienced a frame that appears to not give us back all the output we think we're putting into it, an impression which we assign to lack of stiffness. I think our psychologies would benefit from the aforementioned pedaling machine answering this age-old question in a perfect theoretical sense.

[staggerwing]

I am challenging the notion that frame flex, less than that which noticeably alters a rider's pedaling dynamics or the relative positions and orientations of the wheels, results in a loss of forward energy. I would be very interested in arguments that can be made otherwise.

This is my belief, including the qualifications.

As far as why they didn't build 'em any stiffer back in the day, could it simply be there weren't as many material options available? Everything was built of steel, and if you were one of the majors, like Reynolds, and didn't have too much trouble selling your product, why would you bother making oversized or different shapes. If you were a manufacturer, changing tube ODs would require a new lug designs. If it isn't really broken, and nobody is willing to fund the change, why go there.

Advances in manufacturing, welding, computer design, and even different materials made the field a little more interesting. It pulled back with the UCI restriction to the traditional diamond-shape frame.

[MadRocketSci]

OK. Frame flex in itself doesn't prevent greater forward motion. Instead it prevents the rider from developing as much power.

All modern bikes used in racing today are much stiffer (laterally and torsionally) than those used 30 or 40 years ago. If stiffer wasn't an advantage why would they make them? We don't need science or engineering to know that no one is racing tours on light flexy steel frames.

well i tried to qualify the statement by restricting frames that flexed so much that your pedaling dynamic was affected, which would affect the rider power generated. But, even so, such flex would affect people differently...some like to climb/sprint rocking the bike all over the place, others like to keep the bike mostly upright...

probably a zillion reasons bike frames have gotten stiffer, some with perceived technical benefit. Maybe some freaks like boonen, cav, and andrew luck can really crazy-flex a mortal frame and need something stiffer. My point is, for everyone, there is a "stiff enough," beyond which there are no benefits. But mostly, i think it's a feel that is in the rider's head (specifically, their inner ear) in perceiving frame flex, how that is interpreted as good/bad, and the notion that a stiff frame seems to "accelerate better"...something I ascribe to a relative, rather than absolute, acceleration of the bike with the rider. Feeling good on the bike is the best motivator for speed....but let's not confuse it with the actual physics...

I have a Peg 8:30am, and have test ridden a Time VXR proteam, both known for their stiffness. I do notice the "rocketship, leaping ahead" feeling that everyone talks about when cranking hard, but i believe that's more between me and the bike rather than me+bike and the road...

[MadRocketSci]

1Centaur, the link between human physiology and frame stiffness if very interesting. When playing basketball, I prefer an indoor wood court to outdoor concrete - the shock and compression on the joints are higher outside, and I think I play a bit harder and more freely indoors. Similarly, i think on a bike, I prefer the feel of a tiny bit of frame give, over the feeling of my joints compressing, which is probably why I don't like an overly stiff frame...this does seem to help explain my preference, thanks for the insight!

[Bob Loblaw]

Agreed. How the bike "feels" will affect overall speed much more than stiffness or weight (assuming a competent bike). If a rider thinks the bike is sluggish and robbing him/her of speed, it is.

BL

Bob, you won’t get an argument from me on that issue. However, it has to be due to the rider’s biomechanics and not the steel frame itself absorbing energy. A steel frame when acting like a spring returns practically all the energy that is put into it, so very little (insignificant amount) is turned into heat.

This entire discussion seems highly flawed from the get go. When was the last time anyone stated that their full suspension mountain bike’s frame was too stiff? It’s a crazy notion without merit. Setting aside suspension tuning how could a mountain bike’s frame be too stiff (other than it may make it too heavy)? If the ride was great due to the suspension then how would a stiff frame make it undesirable? It’s more likely that when riders bitch about a frame being too stiff it’s because it rides awful and not because it takes more effort to pedal it. Language gets in the way of discussing this as it should be.

In any case it’s hard to get over the silliness of discussing stiffness versus flexibility without first defining what is desirable and then avoiding using the terms interchangeably.

[Charles M]

there is no reason to believe that a frame flexed to the side returns energy to driving the wheels forward when it unflexes (nor does the energy flow back into our legs).

The second one is more interesting, which is that the energy wasted in flex may not have been spent on driving the wheels forward in the first place, .



Right...

And right...

Frame deflection happens most on the down stroke from roughly 2:30 to about 4:30 - 5 oclock.

That's where the bulk of your power is created under agressive acceleration as well.


Frame return happens almost completely between 5 and 6:30...

That's where the least of your power is created under virtually all conditions but especially under agressive acceleration.



The energy is created from your legs, not a KERS equiped motor... The return energy from flex is almost completely wasted relative to driving cranks.


And yes, som people pedal in pretty nice egg shaped ovals and in some cases fairly nice circles. AND frame flex isnt created much when that type of stroke is happeneing.

Most flex is under much more peaked shapes and flex happens in the peaks...



Stiffness matters, but the frame companies dont release the high speed film and dont let me take pics and graph em side by side with power read...

If they did, It would be pretty easy to show it here.



The energy that produces bottom end frame deflection on "flexy" bikes is generally lost energy BUT several bike companies keep pressing the marketing line and we started seeing bikes that have simply gone so stiff that they're not gaining more power efficiancey than they're losing in ride quality...

That doesnt mean stiffness doesnt matter.

It means stiffness doesnt matter as much as the worst offenders are saying...


On the other hand stiffness isnt as irrelevant as some folks would like you to believe either.

[Jan Heine]

Gary Houchin-Miller once did a finite element analysis, and found that frame flex did in fact turn the rear wheel when it sprang back. His Bikethink site doesn't seem to be up any longer, but we published his study in Bicycle Quarterly Vol. 4, No. 4 (http://www.bikequarterly.com/contents.html).

When we tested bikes for Bicycle Quarterly in the University of Washington wind tunnel (http://www.bikequarterly.com/images/BQ61cover.jpg) , my collaborators noticed how much the BB of my Alex Singer moved sideways. They clearly could see the flex from 20 feet away as I pedaled with a light load. The wind tunnel personnel replied: "You should have seen Lance Armstrong last year when he was here. His bike flexed just as much!"

Since we did our initial research with the three-frame double-blind test (http://janheine.wordpress.com/2011/02/27/a-journey-of-discovery-part-5-frame-stiffness/), we looked at measuring stiffness. It's not easy to get good measurements, but we found that overall stiffness appears to be less important than the "balance" of the frame. For example, when you do a simple sideload test like the one in the drawing in our blog (http://janheine.wordpress.com/2011/10/03/science-and-bicycles-frame-stiffness/), you find that a Trek Madone and a Surly Long-Haul Trucker have about the same stiffness. However, the Trek performs wonderfully, especially under a very strong rider, whereas the Surly isn't ideal for high power outputs. The Trek has very stiff chainstays and a relatively flexible top tube (similar to traditional racing bikes). The Surly has a very flexible rear and a super-stiff top tube.

We don't know yet what mechanisms are at play, but there is no doubt that frame stiffness is a complex, but important subject. Many modern bikes are excellent, but so are many traditional ones. And some are awful.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Kontact]

No engineer here, but this post raises two points that I think are generally ignored by the "steel flexes but you get it back" arguments.

The first one is the one I have made in the past - there is no reason to believe that a frame flexed to the side returns energy to driving the wheels forward when it unflexes (nor does the energy flow back into our legs). All statements about "getting the energy back" seem to ignore that point. As I once posted to make the point, is there any reason to think the return-flex doesn't drive the wheels backward?

The second one is more interesting, which is that the energy wasted in flex may not have been spent on driving the wheels forward in the first place, so who cares other than as it relates to biomechanics? That seems to be a measurable question via a pedaling machine where identical pedaling force can be delivered consistently by an unflexing piston across a range of varyingly flexible frames and the wattage measured at the hub, but humans are not machines. Human physiology reacts to frame resistance with bones, joints and ligaments that themselves flex; the stiffer the frame perhaps the greater the human deflection, sometimes in ways that reduce power, sometimes in ways that enhance recovery, I bet. Good luck isolating those elements. Should we be seeking an ideal ratio of frame flexibility to human physiology?

While I think much of this doesn't much matter to most of us, I'm sure we've all experienced a frame that appears to not give us back all the output we think we're putting into it, an impression which we assign to lack of stiffness. I think our psychologies would benefit from the aforementioned pedaling machine answering this age-old question in a perfect theoretical sense.
If the frame is deflecting (like a spring, which it is), and neither the rider's legs or the wheels are receiving the compressed energy, where does it go?

I think all the angles, wheels, chains and lever arms make it hard for people to see something that is really simple - your legs are pushing right on the wheels. The connection is complicated looking, but 100% mechanical. When it flexes, it flexes in between the rider and the road. When it unwinds, it is still between a firm leg and the road. And like a vaulter's pole, energy not delivered to the wheel at first has to go somewhere, and goes to the road later in the stroke.

This wouldn't be true if a bicycle was made of foam rubber, but every popular bicycle frame material makes for a decent, efficient spring. When the bike flexes under load the top section of chain gets shorter, and when it uncompresses it pulls the cassette to get the chain back to full length. That simple.

A system that managed to do something else with that energy would be a lot more mysterious than that.

[Charles M]

Nah, you're wrong.

Saying that the energy isnt lost is different than thinking it goes into energy that drives the bike forward. The energy doesnt disappear, it goes back into the frame, but in a shorter duration that has no effect on drive.




You would be right if you were talking about straight vert flex (like the bobbing of full suspended mountain bikes) which could potentially return to act on the crank and chain ring...


But lateral deflection and most of the torsional move (except for that pat that lines up with the drive train) doesnt go into drive. It's returned in short duration in a movement that doesnt aid crank rotation.

[Kontact]

You would be right if you were talking about vert flex at the BB (which could potentially return to act on the crank and chain ring)...

But lateral deflection and to some degree torsional move doesnt go into drive.
Then where does it go? Just because the initial movement is in different planes that doesn't mean they are in different systems. The entire drive train is twisting and releasing, but that doesn't make the rear wheel snake around. Energy is conserved, yet no explanation is offered for where it is going, just that it couldn't be going to the wheel.

It would be very easy for someone with access to lots of neat gadgets to simply put power meters on the crank and rear wheel of a couple bikes and look at the output delta. If you don't like the discussion, that would certainly end it.

[Charles M]

You're not listening /reading...

for the x-th time,

The energy return happens in moving the frame back to center.

If that movement happened at a different time in the pedal stroke, you might be right.

But it doesnt.


And you wouldnt put a fancy gadget on the crank. It's incidental to the arguement you're making.

The bike you might use would look something like this...

http://www.pezcyclingnews.com/photos/charles/youdontknow.jpg

And you would measure the frame flex (and if you understood it's stiffness, that would give you the energy required to create a given amount of flex) in time with power to the rear wheel.

You could add a chainring / spider strain guage. You wouldn't put anything on the crank...

[Kontact]

The energry return happens in moving the frame back to center.

If that movement happened at a different time in the pedal stroke, you would be right.

But it doesnt.
If you take a simple spring and flex then release it, it doesn't return to center. It goes past center and cycles back and forth. If you release it slowly to return it to center, the energy is going back into your hand.

Which one is happening? Is the BB suddenly releasing and cycling in the plane it was deflected in?

[Charles M]

Most normal people are done at this point...


But as we're both abnormal.

You're wrong again in assuming that all springs behave like that.

[Jack Brunk]

Gag me with a spoon.

This topic is done.

[1centaur]

As a non-engineering guess, a spring anchored in multiple directions from the BB (chain stays, seat tube, DT, crank/pedal/leg) disperses the energy of its return flex in multiple directions. It is flexed away from its anchors and pulled back by them as well as in whatever direction the material/tubing wants to return to its static form. It would not stun me if SOME of that power pulled the chain forward, but I would be more surprised if NONE of the power went sideways/opposite the vaguely sideways direction of the flex (possibly putting a slight side load on the tires in the process). I have not read the BQ piece so I do not know if ANY forward motion was viewed as success rather than a measurement of watts not transferring on the down stroke showing up 100% in the return flex (might be hard to separate those two without great expense).

I will now let engineers (and those who hang around bike engineers) take over.

[bobswire]

Most normal people are done at this point...


But as we're both abnormal.

You're wrong again in assuming that all springs behave like that.


And I thought you were Abby Normal. :crap:

http://www.youtube.com/watch?v=dQ_pKqiB5Rg

[Kontact]

Most normal people are done at this point...


But as we're both abnormal.

You're wrong again in assuming that all springs behave like that.
Then just name an example of something simple made of steel, aluminum, carbon fiber, titanium, or fiberglass that doesn't.

Springs don't damp themselves. For a bike to behave differently it would have to be made of something rubbery, or have some device that slows the spring release.

[Charles M]

So rubber at different durometers is all the same and none of it is at all springy?





I think you're more at home at the forum HERE (http://www.experienceproject.com/groups/Love-My-Slinky/54763/forum)


But at a bike forum, you may want to refer back to the last page where the test subject seems maybe just that bit closer to topic.

http://www.pezcyclingnews.com/photos/charles/youdontknow.jpg


Or are you still trying to figure out what good a multi point data aquisition frame flex rig is in my understanding frame flex?


Nah,

Never mind.

Just come back with more "the earth is round so I'm right about this other thing too" type wisdom...

[Kontact]

So rubber at different durometers is all the same and none of it is at all springy?





I think you're more at home at the forum HERE (http://www.experienceproject.com/groups/Love-My-Slinky/54763/forum)


But at a bike forum, you may want to refer back to the last page where the test subject seems maybe just that bit closer to topic.

http://www.pezcyclingnews.com/photos/charles/youdontknow.jpg


Or are you still trying to figure out what good a multi point data aquisition frame flex rig is in my understanding frame flex?


Nah,

Never mind.

Just come back with more "the earth is round so I'm right about this other thing too" type wisdom...
Why are you talking about rubber? Bicycle frames aren't made of rubber, non-Newtonian fluids or pixie dust.


Look, if you don't want to have this conversation, that's cool. But when someone resorts to insults and disrespect, the "dialogue" is over.

[Jack Brunk]

All of the Mods must be drunk or suckers for studid bike stuff. This topic may have reach the low of lows concerning topics on this forum. Congrats guys.

[Charles M]

... it would have to be made of something rubbery...

Why are you talking about rubber?

http://images.cheezburger.com/completestore/2010/5/11/129180760489860600.jpg

[bobswire]

All of the Mods must be drunk or suckers for studid bike stuff. This topic may have reach the low of lows concerning topics on this forum. Congrats guys.


Jack I'm wondering what you find so offensive in this thread? I don't see any real insults, slander or hyperbole.
In fact I've read some intelligent discourse, some of it above my schooling but never the less interesting.
Of course I'd rather be looking at or discussing one of your bikes to be more appealing but variety is the spice of life.

[Kontact]

Pez: Imply that frame materials have unusual spring damping properties.
Kontact: Point out that those properties are found in elastomer substances, like rubber. Not in bicycle frame materials.
Pez: Point out that rubber can have spring properties, too.
Kontact: Wonders why Pez wants to discuss rubber spring properties since there is no rubber in the bike frames we are discussing.
Pez: Insults, further non sequitars.

At this point, there is nothing more to discuss. Durometers of rubber are, I'm sure, fascinating, but have nothing to do with the topic. Cats and trolls don't either.

Like I said, if you don't want to have a conversation, don't. No need to insult me or anyone else's intelligence with the distractors. You made a rather remarkable statement about how frames spring back to center which I would have enjoyed hearing more about.

[Charles M]

Walk over to your bike and put it in front of you facing 90 degrees in either direction. Tip the bike 5-10 degrees away from you.

Spin the crank arm to facing down and put your foot at the bottom bracket and press.

The bike will flex.

Take the pressure off.

The bike will return.



And you still won't understand ;)

[Fixed]

old bike shop snake oil sell
cheers

[zap]

snip

Then where does it go?

Do you know how many times I've asked that question?

Anyhow, as others posted, add a human to the equation and this seemingly simple matter gets a little complicated.

I think Mr. Kirk's post explains it best.

[Chance]

If the frame is deflecting (like a spring, which it is), and neither the rider's legs or the wheels are receiving the compressed energy, where does it go?

I think all the angles, wheels, chains and lever arms make it hard for people to see something that is really simple - your legs are pushing right on the wheels. The connection is complicated looking, but 100% mechanical. When it flexes, it flexes in between the rider and the road. When it unwinds, it is still between a firm leg and the road. And like a vaulter's pole, energy not delivered to the wheel at first has to go somewhere, and goes to the road later in the stroke.

This wouldn't be true if a bicycle was made of foam rubber, but every popular bicycle frame material makes for a decent, efficient spring. When the bike flexes under load the top section of chain gets shorter, and when it uncompresses it pulls the cassette to get the chain back to full length. That simple.

A system that managed to do something else with that energy would be a lot more mysterious than that.
Nah, you're wrong.
You are both wrong.

The frame doesn't consume the energy -- it can't. So where does it go? It has to end up somewhere.

The only explanation that makes sense was proposed on this forum a while back. The frame's return energy is wasted by working "against" the rider's legs in an inefficient manner.

He stated this could happen because the human body can't recover energy once spent. Unlike a hybrid car that can recover kinetic or potential energy back for future use the human body can't recover any at all. Work done on the human body by the bicycle as it unwinds (or any other outside source) doesn't lead to energy storage (as in recharging the body with more energy) and is therefore wasted.

The example he used was simple enough. If a person goes up a set of stairs he does work. If he climbs back down in theory the stairs could be doing work back to him. If instead of a human body it was a mechanical device most of that energy could be recovered for future use. However, the human body won't recover any of it. If he goes up and down 100 or 1000 times he will get tired from all the work although the sum total is ZERO. In technical terms he's done a total of zero work yet is exhausted.

This makes perfect sense for a bike too. When the bike flexes back and delivers the previously stored energy it goes to waste just like when a person goes downstairs. It doesn't disappear. The energy is wasted just like when any of us climb down a ladder or set of stairs.

[Chance]

Gary Houchin-Miller once did a finite element analysis, and found that frame flex did in fact turn the rear wheel when it sprang back. His Bikethink site doesn't seem to be up any longer, but we published his study in Bicycle Quarterly Vol. 4, No. 4 (http://www.bikequarterly.com/contents.html).

When we tested bikes for Bicycle Quarterly in the University of Washington wind tunnel (http://www.bikequarterly.com/images/BQ61cover.jpg) , my collaborators noticed how much the BB of my Alex Singer moved sideways. They clearly could see the flex from 20 feet away as I pedaled with a light load. The wind tunnel personnel replied: "You should have seen Lance Armstrong last year when he was here. His bike flexed just as much!"

Since we did our initial research with the three-frame double-blind test (http://janheine.wordpress.com/2011/02/27/a-journey-of-discovery-part-5-frame-stiffness/), we looked at measuring stiffness. It's not easy to get good measurements, but we found that overall stiffness appears to be less important than the "balance" of the frame. For example, when you do a simple sideload test like the one in the drawing in our blog (http://janheine.wordpress.com/2011/10/03/science-and-bicycles-frame-stiffness/), you find that a Trek Madone and a Surly Long-Haul Trucker have about the same stiffness. However, the Trek performs wonderfully, especially under a very strong rider, whereas the Surly isn't ideal for high power outputs. The Trek has very stiff chainstays and a relatively flexible top tube (similar to traditional racing bikes). The Surly has a very flexible rear and a super-stiff top tube.

We don't know yet what mechanisms are at play, but there is no doubt that frame stiffness is a complex, but important subject. Many modern bikes are excellent, but so are many traditional ones. And some are awful.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/
You should reconsider whether clamping the frame and fork as in the rig shown at the top of the article is a good idea.

That doesn't represent the way a bike is loaded in real life while being ridden. And that may explain why you are seeing inconsistencies between frames that test equally. Just an uneducated thought.

[Kontact]

You are both wrong.

The frame doesn't consume the energy -- it can't. So where does it go? It has to end up somewhere.

The only explanation that makes sense was proposed on this forum a while back. The frame's return energy is wasted by working "against" the rider's legs in an inefficient manner.

He stated this could happen because the human body can't recover energy once spent. Unlike a hybrid car that can recover kinetic or potential energy back for future use the human body can't recover any at all. Work done on the human body by the bicycle as it unwinds (or any other outside source) doesn't lead to energy storage (as in recharging the body with more energy) and is therefore wasted.

The example he used was simple enough. If a person goes up a set of stairs he does work. If he climbs back down in theory the stairs could be doing work back to him. If instead of a human body it was a mechanical device most of that energy could be recovered for future use. However, the human body won't recover any of it. If he goes up and down 100 or 1000 times he will get tired from all the work although the sum total is ZERO. In technical terms he's done a total of zero work yet is exhausted.

This makes perfect sense for a bike too. When the bike flexes back and delivers the previously stored energy it goes to waste just like when a person goes downstairs. It doesn't disappear. The energy is wasted just like when any of us climb down a ladder or set of stairs.
The human body is capable of recovering energy by converting the energy into motion. That's how a pogo stick or pole vauting works, as well as the basis of a tennis racket. If the drivetrain unwinds against the rider, the rider moves.

[staggerwing]

The human body is capable of recovering energy by converting the energy into motion. That's how a pogo stick or pole vauting works, as well as the basis of a tennis racket. If the drivetrain unwinds against the rider, the rider moves.

Not quite. For example with the pole vault, the vaulter temporarily stores most of his kinetic energy (1/2mV^2, V being the velocity at the end of the runup) in the pole, reconfigures his alignment and then holds on tight while that stored energy is transferred back. Energy is just moved from one system to another. Nothing special about a human being attached to the device with the stored energy. Could even be something inanimate attached.

We can go around in circles, but in the real world, there is no 100% efficient transfer or conversion of energy from one system to another. There are always pieces that can't be captured (see second law of thermodynamics).

[Jan Heine]

You should reconsider whether clamping the frame and fork as in the rig shown at the top of the article is a good idea.

That doesn't represent the way a bike is loaded in real life while being ridden. And that may explain why you are seeing inconsistencies between frames that test equally. Just an uneducated thought.

I agree, but the problem is that clamping anywhere else doesn't work, either. We tried the "Rinard" method of clamping the BB (http://www.sheldonbrown.com/rinard/rinard_frametest.html) , but we found that we were measuring flex in the BB shell. Basically, all OS frames measured the same, no matter the tubing walls, and all standard diameter frames measured the same. (OS leaves less BB shell unsupported sticking out sideways.) So that didn't work either.

So in the end, measuring stiffness is difficult. Interestingly, the simple BB deflection test matched our ride experiences of the double-blind test, but those bikes all used the same rear triangle, just varied the main tubes between 9-6-9 standard dia., 7-4-7 standard and 7-4-7 OS.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Chance]

Not quite. For example with the pole vault, the vaulter temporarily stores most of his kinetic energy (1/2mV^2, V being the velocity at the end of the runup) in the pole, reconfigures his alignment and then holds on tight while that stored energy is transferred back. Energy is just moved from one system to another. Nothing special about a human being attached to the device with the stored energy. Could even be something inanimate attached.

We can go around in circles, but in the real world, there is no 100% efficient transfer or conversion of energy from one system to another. There are always pieces that can't be captured (see second law of thermodynamics).
Thank you. As an engineer you stated it more clearly.

[Chance]

The human body is capable of recovering energy by converting the energy into motion. That's how a pogo stick or pole vauting works, as well as the basis of a tennis racket. If the drivetrain unwinds against the rider, the rider moves.
Partly true but that's not what this is about. Not really.

While it is true that the human body can serve as a vehicle to store some kinetic energy as motion we are talking about recovering work that was done by muscles. Yeah, when we ride downhill our bodies serve to store some of the energy from the climb as kinetic energy but that's due to our bodies mass and not really part of the muscles or power generating system. A concrete block could do the same and can't generate power.

In "theory" some unwinding of a bicycle frame could help accelerate the rider's legs and therefore save some of that energy as additional kinetic energy to be returned back later to help power the bike forward, but that is completely unreasonable to expect. It's not going to happen because at the speeds a bike travels the entire system's kinetic energy (or maybe it's momentum ???) is so much higher that fluctuations in speed can't occur.

In the case of the pogo stick or pole vault the energy you refer to is stored in the pogo stick's spring or in the pole itself by flexing it like a spring. In these cases and many others in sports it is possible to accomplish additional function by the person building upon the stored energy by releasing additional energy timed just right to coincide with the release of energy from the device. A trampoline works similar to a pogo stick. A gymnastics floor also to a lesser degree.

With a pogo the person jumps up as the pogo tries to spring back. Do it enough times and you build on each subsequent cycle. The same with pole vault, trampolines and other devices. It takes technique to get the timing right but can be made to work.

Bicycle frames are a different issue entirely.

[Louis]

Jan,

As an engineer myself, I'd like to thank your for bringing some science to bear on a product that is most often discussed (ad nauseam) using only intuition and seat-of-the-shorts measurements. I occasionally disagree with your conclusions but respect your search for truth.

Louis

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Chance]

I agree, but the problem is that clamping anywhere else doesn't work, either. We tried the "Rinard" method of clamping the BB (http://www.sheldonbrown.com/rinard/rinard_frametest.html) , but we found that we were measuring flex in the BB shell. Basically, all OS frames measured the same, no matter the tubing walls, and all standard diameter frames measured the same. (OS leaves less BB shell unsupported sticking out sideways.) So that didn't work either.

So in the end, measuring stiffness is difficult. Interestingly, the simple BB deflection test matched our ride experiences of the double-blind test, but those bikes all used the same rear triangle, just varied the main tubes between 9-6-9 standard dia., 7-4-7 standard and 7-4-7 OS.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/
Thanks for reply. To an untrained eye your system looks better than Rinard’s except for the fact that you appear to clamp both ends of the frameset preventing rotation. In the real world that can't happen because tires are free to rotate with minimal resistance. Your test apparatus is therefore creating an unusual loading of the frame that can’t be representative of the way we load bikes while riding.

On the other hand it's hard to tell much about your equipment from a sketch. :beer:

[Chance]

I agree, but the problem is that clamping anywhere else doesn't work, either. We tried the "Rinard" method of clamping the BB (http://www.sheldonbrown.com/rinard/rinard_frametest.html) , but we found that we were measuring flex in the BB shell.

.....cut.....


Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/


By the way, it's interesting that Rinard in the quote below seems to disagree completely with your theory. Do you have any comments on his position on this? It's just interesting that even bike experts disagree all the time.

Too Stiff?

Is it possible for a frame to be too stiff? It is tempting to write here about ride quality and vertical compliance or resilience, but since my test was designed to evaluate lateral (and a bit of torsional) stiffness only, and not vertical stiffness, I will confine my comments to this: the only drawback I can see to a frame that is infinitely stiff laterally and torsionally is the possibility of excess weight, if stiffness is achieved by overbuilding the frame.

[Fixed]

Jan Heine has a cool site thanks for the heads up
http://www.bikequarterly.com/

[merlincustom1]

Can someone please post a pic of a hot chick here now? I need my head to explode for reasons other than metallurgy.

Thank you.

[Kontact]

Partly true but that's not what this is about. Not really.

While it is true that the human body can serve as a vehicle to store some kinetic energy as motion we are talking about recovering work that was done by muscles. Yeah, when we ride downhill our bodies serve to store some of the energy from the climb as kinetic energy but that's due to our bodies mass and not really part of the muscles or power generating system. A concrete block could do the same and can't generate power.

In "theory" some unwinding of a bicycle frame could help accelerate the rider's legs and therefore save some of that energy as additional kinetic energy to be returned back later to help power the bike forward, but that is completely unreasonable to expect. It's not going to happen because at the speeds a bike travels the entire system's kinetic energy (or maybe it's momentum ???) is so much higher that fluctuations in speed can't occur.

In the case of the pogo stick or pole vault the energy you refer to is stored in the pogo stick's spring or in the pole itself by flexing it like a spring. In these cases and many others in sports it is possible to accomplish additional function by the person building upon the stored energy by releasing additional energy timed just right to coincide with the release of energy from the device. A trampoline works similar to a pogo stick. A gymnastics floor also to a lesser degree.

With a pogo the person jumps up as the pogo tries to spring back. Do it enough times and you build on each subsequent cycle. The same with pole vault, trampolines and other devices. It takes technique to get the timing right but can be made to work.

Bicycle frames are a different issue entirely.
The only thing different is scale. The BB flexes a few degrees, a vaulter's pole bends almost in half. The bicycle+rider system as a whole has plenty of momentum that smooths out the acceler/decel inherent in a cranked system. And if you were able to monitor it perfectly, you'd find that the low mass part of the system (the back end of the bike) does accel/decel constantly as it compresses and expands. The spokes allow for the same thing.

I do agree that a frame can't perfectly store and return energy. But there's lots of things that are imperfect about pedaling that leads to inefficiencies - and some studies seem to suggest that BB deflection helps address those - so it may be energy well spent.

I'm not really for or against anything, here, except extremes of stiffness or flexibility. All I was trying to get across is that there are only two outlets for the pedaling energy - the input or the output. The bike can't store it or throw it out in an unused dimension.

[Kontact]

Walk over to your bike and put it in front of you facing 90 degrees in either direction. Tip the bike 5-10 degrees away from you.

Spin the crank arm to facing down and put your foot at the bottom bracket and press.

The bike will flex.

Take the pressure off.

The bike will return.



And you still won't understand ;)
Do the same thing, but slip your foot off the side of the pedal while it is flexed, releasing the tension all at once. Then you'll understand.

[Charles M]

?


Ok...

Where does that energy go?

The crank? nope.

The pedal? nah...

The leg that isnt touching the bike any more?

Nope...


The side flex return energy doesnt translate to anything that might drive the bike forward...

That's the point on topic, but then asking another question will be your only answer...

[markie]

Walk over to your bike and put it in front of you facing 90 degrees in either direction. Tip the bike 5-10 degrees away from you.

Spin the crank arm to facing down and put your foot at the bottom bracket and press.

The bike will flex.

Take the pressure off.

The bike will return.



And you still won't understand ;)

Does the frame really flex that much? Or do you just cause the tyres to deflect causing an illusion of frame flex? To me a lot of the "flex" comes from the tyres.

[Kontact]

?


Ok...

Where does that energy go?

The crank? nope.

The pedal? nah...

The leg that isnt touching the bike any more?

Nope...


The side flex return energy doesnt translate to anything that might drive the bike forward...

That's the point on topic, but then asking another question will be your only answer...
When your weight is removed, the frame springs back (in the way you said it didn't earlier). When your weight is on it, it pushes between your leg and the wheel.

It is important to remember that the crank isn't bending between the pedal and the spindle. The chain anchors the top of the crank and the bending moment includes it. So does the release.

[staggerwing]

How much energy to you really think is being stored and released in lateral deflection?

The bicycle is an incredibly elegant mechanism for converting human muscle force into forward velocity. I've heard mechanical efficiency numbers in the high 90's, although I haven't seen direct refs to back that up. There is the start if something in the Wiki on "bicycle performance" talking to this point. It that is the case, any possible return of energy from frame flex is going to be noise.

I would love to see some real numbers, as power applied to the cranks by real humans under actual live pedaling, not a stationary rig. To really do it right would require a 6DOF load cell somewhere in the crank setup, plus a instrumented rear wheel.

Probably been done already, although as Pez alluded to, the info may be proprietary. We would still likely argue interpretation.

[jgspin]

My friend easily outweighs me 80+ lbs. and very strong on the flats; took 2nd in a recent time trial at Fiesta Island. What impresses me is that this guy can climb strong and drops me on hills 3 miles or less. He climbs in the big ring and when he gets out of the saddle I just can't keep up. When he gets in his rhythm it looks like he is flexing his bike. Could it be possible that his bike is flexible enough for him that when he times it correctly with his out of the saddle pedaling, the bike is pushing him back up on the rebound that he recovers some of the energy he used on the down stroke? If that was the case, then flexible frames would be helpful for out of the saddle climbers, right?

[Kontact]

How much energy to you really think is being stored and released in lateral deflection?

The bicycle is an incredibly elegant mechanism for converting human muscle force into forward velocity. I've heard mechanical efficiency numbers in the high 90's, although I haven't seen direct refs to back that up. There is the start if something in the Wiki on "bicycle performance" talking to this point. It that is the case, any possible return of energy from frame flex is going to be noise.

I would love to see some real numbers, as power applied to the cranks by real humans under actual live pedaling, not a stationary rig. To really do it right would require a 6DOF load cell somewhere in the crank setup, plus a instrumented rear wheel.

Probably been done already, although as Pez alluded to, the info may be proprietary. We would still likely argue interpretation.
The gain, or loss, could both be noise indeed. That's why I suggested several pages ago that the easy way to end the argument is to measure it directly with a Powertap and Quarq, or something like it. A few riders and 5 bikes should yield all the data you'd ever need to say flex is good, bad or indifferent.

[palincss]

A few riders and 5 bikes should yield all the data you'd ever need to say flex is good, bad or indifferent.

No, all you'd get would be a bunch of forumites telling you it wasn't "science" because your sample size was too small and it was all just "your opinion."

[Louis]

No, all you'd get would be a bunch of forumites telling you it wasn't "science" because your sample size was too small and it was all just "your opinion."

Depends on what you measure and what you're trying to prove.

[Jan Heine]

Depends on what you measure and what you're trying to prove.

Absolutely right. We were trying to prove
1) that frame flex is noticeable to some riders so reliably that they can tell apart frames with (slightly) different stiffness and
2) that some riders perform better on frames with certain flex characteristics than on others.

For 1), we only had to show that one person can tell the difference between three frames, two of which are identical, the third one different. We used one 9-6-9 frame and two 7-4-7 frames, both standard diameter. (Of course, the bikes had the same geometry and the same componets.) We showed that two riders could tell the "odd one out" every time. (We ran four tests, so the likelihood that it's purely chance is 1 in 6561.)

A third person couldn't tell the (small) difference between our test frames. Clearly, some people are more sensitive than others. We never determined whether the third rider could have told the difference if the frames had been more different.

2) is trickier. The only way the riders could tell the frames apart was by their performance. The bikes all tracked straight, none shimmied, we never noticed that the BB moved sideways, we never were allowed to ping the tubes, etc.

It took only a few hard pedal strokes for the two riders who could tell the difference to determine which bike they were riding. On the "favorite" bikes, the two riders were measurably faster. Their power output was higher. Even when racing against each other (they were well-matched performance-wise), the outcome usually was that the rider on the "favorite" frame came out ahead. When both were on "favorite" frames, they were evenly matched.

Before we did the double-blind test, we had ridden a number of bikes with very flexible 7-4-7 standard-diameter steel tubing, and found them to perform exceptionally well. However, a bike test has too many variables (components, geometry, fit, color) to be sure, so we designed the double-blind experiment to test our hypothesis. And in the test, the "favorite" bikes were indeed the 7-4-7 bikes. This showed that what we had been feeling during the bike tests could be replicated in a carefully controlled setting. So it wasn't in our imagination, our daily form or whatever.

Of course, there are limitations to this study. If we wanted to show that frame stiffness matters for a certain percentage of riders, then we'd need much larger samples - probably 1000 riders or so. If we wanted to show that some riders cannot tell frame stiffness at all, we'd also need a different experiment, with bikes that were more dissimilar. (All our frames were relatively flexible.)

In any case, all the blog post (http://janheine.wordpress.com/2011/10/03/science-and-bicycles-frame-stiffness/) was stating is that in the past, many riders liked flexible frames (Reynolds 753, Tange Prestige), but thought that these frames were stiff, because they were made from high-stength tuning. This "crutch" helped them reconcile the belief that "stiffer was better" with the observation that these superlight "Supersteels" performed better.

We now know that superlight steel frames are more flexible, no matter which steel is used. The next question obviously is why so many riders preferred these flexible frames over the much stiffer "drainpipe"? (If you want a stiff frame, try a 1980s mid-range Bianchi or similar made from straight-gauge, standard-diameter tubing!)

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Charles M]

When your weight is removed, the frame springs back (in the way you said it didn't earlier). When your weight is on it, it pushes between your leg and the wheel.

It is important to remember that the crank isn't bending between the pedal and the spindle. The chain anchors the top of the crank and the bending moment includes it. So does the release.



Don't put words in my mouth:

I said...






Frame deflection happens most on the down stroke from roughly 2:30 to about 4:30 - 5 oclock.

That's where the bulk of your power is created under agressive acceleration as well.


Frame return happens almost completely between 5 and 6:30...

That's where the least of your power is created under virtually all conditions but especially under agressive acceleration.



The energy is created from your legs, not a KERS equiped motor... The return energy from flex is almost completely wasted relative to driving cranks.


And yes, som people pedal in pretty nice egg shaped ovals and in some cases fairly nice circles. AND frame flex isnt created much when that type of stroke is happeneing.

Most flex is under much more peaked shapes and flex happens in the peaks...



Stiffness matters, but the frame companies dont release the high speed film and dont let me take pics and graph em side by side with power read...

If they did, It would be pretty easy to show it here.



The energy that produces bottom end frame deflection on "flexy" bikes is generally lost energy BUT several bike companies keep pressing the marketing line and we started seeing bikes that have simply gone so stiff that they're not gaining more power efficiancey than they're losing in ride quality...

That doesnt mean stiffness doesnt matter.

It means stiffness doesnt matter as much as the worst offenders are saying...


On the other hand stiffness isnt as irrelevant as some folks would like you to believe either.

_______________________


I know this after sessions using a data test rig...



And watching the test duplicated by two of the largest fram producers in the world.








Folks here can take or leave that.


You won't be able to leave it.... ;)


All done here.


Andy wins :D

[Jan Heine]

Of course, there are limitations to this study.

One other thing we didn't show in our double-blind test (http://janheine.wordpress.com/2011/02/27/a-journey-of-discovery-part-5-frame-stiffness/) was that all riders prefer more flexible frames. Some people misunderstood our article (usually without having read it) and thought we said that more flexible was better in every case.

This is of course nonsense. For two test riders, we found that the most flexible (steel) frames available today performed best, but we also showed in a different experiment that too much flex is not good, either. Basically, a frame that is too flexible apparently reacts like a full suspension bike that bobs with your pedal stroke.

Every rider needs to determine what works best for them. All we can say is that the old mantra of "stiffer is better" definitely is not true for at least some riders.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Jan Heine]

I know this after sessions using a data test rig...

And watching the test duplicated by two of the largest fram producers in the world.

Tell us more about those experiments. What was measured, and how? Which different frames were compared? What were the results? And what is the likely explanation? If energy is lost, it must be converted to heat. In the frame? The tires? Or in the rider's legs?

And finally, has that research has been published? (If it's internal data from bike makers, I suspect they'd rather keep it to themselves.)

While I am skeptical of the big makers' "research" - some of the best-riding frames we have tested came from small makers like Crumpton and Calfee, not the big ones –?there is a lot to learn, and each piece of the puzzle is helpful.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Fivethumbs]

Was there any determination made as to why the riders performed better on the more flexible frames? Was the ride less bouncy, which better allowed the riders to keep a steady rhythm? Was it more comfortable, which made them happier? Maybe the the frame flex on the down stroke returns on the upstroke assisting in pushing the pedal and leg back up.?

[Jan Heine]

Was there any determination made as to why the riders performed better on the more flexible frames? Was the ride less bouncy, which better allowed the riders to keep a steady rhythm? Was it more comfortable, which made them happier? Maybe the the frame flex on the down stroke returns on the upstroke assisting in pushing the pedal and leg back up.?

We measured speed and power. Both riders put out more power on the "favorite" bikes, and they rode faster. So it's a biomechanical issue, not an issue of efficiency in power transmission. Both riders independently recorded their observations without speaking to each other. This is somewhat subjective, but the fact that both shared the same experience is interesting.

Both riders felt that the more flexible bikes made it easier to push harder on the pedals. Their legs didn't hurt with the build-up of lactic acid, so their cardiovascular rather than their muscular system limited performance.

It appears that when you pedal hard on a stiff frame, it is a bit like pushing against a concrete wall. The wall does not move, so no work is performed. Your heart rate goes up only slightly, but your arms will get tired nonetheless. Pushing against a box that moves is easier.

So the flexible frame may allow you to input more power into the frame. (Even though we all strive to have a "round" pedal stroke, the majority of the power input occurs during the down stroke, as was mentioned by somebody else here.)

Unfortunately, it would be very difficult to test this hypothesis, because it would require measuring lactic acid buildup (or whatever makes your legs hurt) in your muscles while riding on the road.

Note that when we put out about 900 Watts, the slightly stiffer frame performed as well for one of us. When we repeated the uphill sprints, the more flexible frame became "preferred" as our power output was reduced with fatigue.

It is important to note that I am not arguing that flexible frames work best for everybody. We tested a (very stiff) Trek Madone, and found it to work very well for up to 50 miles under hard efforts. A racer who is stronger than we are might get even more out of it. But for the average rider on a century, a more flexible frame might be better.

Also note that our test riders were about 6' tall, weighed about 150-160 lb and spun at 100-110 rpm. Different body and riding styles may also prefer different flex characteristics.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Fixed]

J.H. post more i like your style imho
cheers
http://janheine.wordpress.com/2011/10/03/science-and-bicycles-frame-stiffness/

[Germany_chris]

I like reading you here more than your place because you react vs. making statements

[forrestw]

We measured speed and power. Both riders put out more power on the "favorite" bikes, and they rode faster. So it's a biomechanical issue, not an issue of efficiency in power transmission. Both riders independently recorded their observations without speaking to each other. This is somewhat subjective, but the fact that both shared the same experience is interesting.

Both riders felt that the more flexible bikes made it easier to push harder on the pedals. Their legs didn't hurt with the build-up of lactic acid, so their cardiovascular rather than their muscular system limited performance.

It appears that when you pedal hard on a stiff frame, it is a bit like pushing against a concrete wall. The wall does not move, so no work is performed. Your heart rate goes up only slightly, but your arms will get tired nonetheless. Pushing against a box that moves is easier.

So the flexible frame may allow you to input more power into the frame. (Even though we all strive to have a "round" pedal stroke, the majority of the power input occurs during the down stroke, as was mentioned by somebody else here.)

Unfortunately, it would be very difficult to test this hypothesis, because it would require measuring lactic acid buildup (or whatever makes your legs hurt) in your muscles while riding on the road.
(snip)
Also note that our test riders were about 6' tall, weighed about 150-160 lb and spun at 100-110 rpm. Different body and riding styles may also prefer different flex characteristics.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

So I repeat what I said several pages back, N=3 riders cannot imo make for a statistically valid 'blind' test. I'm also dubious that you can effectively blind (or double-blind) a stiff vs flexible frame with an experienced rider in the saddle; I can tell the stiffness differences between frames easily enough even on a fixed gear and on a bike with an FD no rider is going to miss the frame flexibility visible between chainrings and FD.

Now, I don't have a BQ subscription so my knowledge of your fest is limited to what you put in the blog. I'm working on getting the relevant articles via interlibrary loan.

That said, if the subjects knew that they were evaluating bikes of differing frame stiffness (and afaics you were one of the riders so I assume this to be true) then this is not a blind test and particularly the riders know what differences are being tested and so they know to look for stiffness (see above).

If you were metering power, good but from what I have readin prior posts your evaluations were mostly about who caught whom on ascents.

I'm strongly biased to having test numbers and of course I'm also biased to believe what I already think is true. I had concluded what Peztech wrote before he wrote it. i.e. fully meter a frame so you can measure power in, power out and frame flex.

That said, I'll repeat and amplify my thought-experiment from pages back. If you had an otherwise normal bike with a 'spaghetti-like' flexible BB, it's not hard to see that you would lose a lot of your power input, even if your pedal stroke were perfectly circular. The BB shell is always flexing away from the applied force, you're always chasing that.

I think we can agree on that, if anyone cares to make a theoretical argument that flexible frames are inherently better that's a point I think you have to refute. I think this is also Pez's point about 6 o'clock but applied to the full 360 of revolution.

This said, when we're pedaling 'round' we're not flexing the frame much, far and away the heaviest BB deflection forces are encountered when you stand to climb or sprint, and most of those are being generated up/down, not in a nice circle and anyone who's had to adjust FD trim as the big chainring moves due to BB deflection knows this.

Now I don't think this is recoverable, each stroke you sink your weight into the pedal and at the bottom of the stroke your'e quite a bit lower on a flexing BB than on a stiff one, could easily be 1/8", not huge, but in fact 1% of your pedal circle. On my cheap '83 Peugeot it could be more. 1% loss of power is nothing to snort at imo :-). You have to lift your body back up that extra height at every stroke.

The difference between that '83 Peugeot and my Serotta Legend is quite striking. The former is vertically stiff and laterally noodley, the latter is vertically compliant and laterally so stiff I can't see the big ring move wrt the FD under hard pedaling -- the only frame I've personally ridden where that's been true.

I experience significantly more riding fatigue / pain on the Peugeot than the Serotta -- granted the latter is a custom build and fits like a glove.

My hand-built frame which has the same tubing sizes as the Peugeot is a good deal stiffer in the BB because of a better BB shell and correct mitering. It's an exact copy of the Serotta's geometry and while more comfortable to ride than the Peugeot, it's not the frame I'd choose for days of more than 80 miles.

When I get to read your full article/report I may come back to say some more :-)

[Charles M]

Tell us more about those experiments. What was measured, and how? Which different frames were compared? What were the results? And what is the likely explanation? If energy is lost, it must be converted to heat. In the frame? The tires? Or in the rider's legs?

And finally, has that research has been published? (If it's internal data from bike makers, I suspect they'd rather keep it to themselves.)

While I am skeptical of the big makers' "research" - some of the best-riding frames we have tested came from small makers like Crumpton and Calfee, not the big ones –?there is a lot to learn, and each piece of the puzzle is helpful.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/


Maybe the best way to tell the info pulled was usefull was that it didnt wind up in marketing :D

The tests started out really complicated and wound up pretty simple.

strain gauges at the spider and wheel to track power at both ends of the drive train. Deflection measured in amount, duration and timing at several points on 8-15 frames.


The general outcome is posted a few times in the thread, including the post you quoted me from...



Human energy is spent both powering the drive train and deflecting the frame. The lateral and torsional deflection return happens at a point when a only fraction of the power is being applied to the drive train versus the Big power being produced along with the deflection... It's effectively lost.

I was hoping that someone would note that the other deflection that happens is in compacting the drive side chain stay under heavy load. The return of that compression is the one place where, as long as the drive train stays loaded (and it usually does), the energy stays in propulsion system.



The first testing was independent. The follow up by companies made sense, but all of the info for both company and independent is confidential.


Over a range of frames of both similar and different materials, the deflection under similar power could vary quite a bit in both volume and duration. Meaning some frames flexed further and took longer to return to center... But the timing of the deflection return was always at the same point at the bottom of the power production curve.




The one thing that a couple of frame makers are trying to do now is play with the timing.


If they can make a frame that allows for deflection but can push that return sooner (where it goes back into the power production curve of the pedal stroke), they will have a frame that should have better overall compliance but less power lost to deflection.

The thought that the best bike must be laterally stiff and vert compliant might not be as true a measure of a great performing but still comfortable bike. As long as the flex is reasonable (and some bikes are simply too sloppy), the important part is that the energy return happens in the right timing and duration to go back into driving the bike.


The part that will irk some folks at this forum... No frame maker or materials engineer involved thinks that's possible without using material with Anisotropic character...

Read this for a bit of an explanation of Ani and bikes (http://www.compositesworld.com/articles/aluminum-frame-build-incorporates-carbon-fiber-tubes)



PM me if you want a little more Jan. Would be fun to talk... There's a hell of a facility here in AZ that's comming on line shortly.



And thanks to the person that sent me the instructions on using the forum's "Ignore" list!


;)

[Jan Heine]

So I repeat what I said several pages back, N=3 riders cannot imo make for a statistically valid 'blind' test.

It depends on what you are testing. When we started talking about this, most people told us that we couldn't tell the difference between frames of different stiffness. In that case, you need only one rider to show they can. (If you wanted to prove that the Earth is round, you need one trip around the earth, not 1000.)

I'm also dubious that you can effectively blind (or double-blind) a stiff vs flexible frame with an experienced rider in the saddle

As our blog post mentioned, generations of "experienced" riders thought that Reynolds 531 was stiffer than drainpipe, and superthin Reynolds 753 tubes were stiffer yet, even though the opposite is true. Apparently, it's not easy to experience "stiffness" in a bike frame.

I can tell the stiffness differences between frames easily enough even on a fixed gear and on a bike with an FD no rider is going to miss the frame flexibility visible between chainrings and FD.

We looked at that, and found that the opposite is the case: Stiff frames more often display front derailleur rub than flexible ones. We discussed this in a sidebar "misleading indicators of stiffness" (http://www.bikequarterly.com/contents.html) once. For example, the Surly Long-Haul Trucker we tested had front derailleur rub, and so did my Mercian, which was made from 10-7-10 Reynolds "Super Tourist" tubing. None of the double-blind test bikes displayed front derailleur rub, even though they all were made from more flexible tubing. (Also, we varied only the down and top tubes, so the seat tubes were identical anyhow.)

An engineer once pointed out to us that a typical thinwall tube has much greater variations of wall thickness. The center section wall of a 7-4-7 tube is only 57% as thick as the walls at the ends. This means that the tube will flex most in the center, not at the end where the derailleur is mounted. The tube flexes in the center, but BB and derailleur both are mounted on the stiff end, and will move in unison.

A 10-7-10 tube is more uniform: the center walls are 70% as thick as the ends. This means it will flex more uniformly along its entire length, including the end where the derailleur is mounted. So you get more flex between derailleur and BB, and the derailleur rubs on the chainring.

This explanation appears plausible to me. In any case, from my experience, it's usually the stiffer frames that display front derailleur rub.

We found that we could not tell the stiffness of the frames, except by their performance. Even the superlight frames tracked straight, descended superbly, didn't shimmy or otherwise feel different. They weren't even more comfortable!

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Jan Heine]

I don't know whether it's OK to post "commercial" messages here, but with all this interest in our tests, I just put a "Tech Article Set" in our shopping basket, especially for listmembers. You get:


Vol. 5, No. 1 with our big tire test where we examined things like width, pressure, casing materials, etc. on real roads.
Vol. 6, No. 4 with the double-blind test of frame stiffness.
Vol. 7, No. 4 with more on the double-blind test, including power measurements and an OS frame.
Vol. 8, No. 1 with the experiment that quantified suspension losses. We found that comfort and speed are directly related, because vibrations are energy lost.


To order, go here (http://www.compasscycle.com/category/Books-Magazines-USA-14) and select "Tech Article Set" as the last of the drop-down options for the "Back Issues."

For more info on these issues, go here. (http://www.bikequarterly.com/contents.html)

Again, if this is inappropriate for this list, please let me know.

Jan Heine
Editor
Bicycle Quarterly
http://www.bikequarterly.com/

[Kontact]

strain gauges at the spider and wheel to track power at both ends of the drive train. Deflection measured in amount, duration and timing at several points on 8-15 frames.


Human energy is spent both powering the drive train and deflecting the frame. The lateral and torsional deflection return happens at a point when a only fraction of the power is being applied to the drive train versus the Big power being produced along with the deflection... It's effectively lost.

I was hoping that someone would note that the other deflection that happens is in compacting the drive side chain stay under heavy load. The return of that compression is the one place where, as long as the drive train stays loaded (and it usually does), the energy stays in propulsion system.




Over a range of frames of both similar and different materials, the deflection under similar power could vary quite a bit in both volume and duration. Meaning some frames flexed further and took longer to return to center... But the timing of the deflection return was always at the same point at the bottom of the power production curve.


Read this for a bit of an explanation of Ani and bikes (http://www.compositesworld.com/articles/aluminum-frame-build-incorporates-carbon-fiber-tubes)

;)
Pez,

Was there a net average (across full crank rotation, not at specific points) decrease in power transmitted between crank and wheel on the bikes with BBs that deflected more? (I assure you I understand the rest of it, but it the only important numbers are at the road, and I'm not seeing that in your explanations.)

Is it possible to make a BB deflect laterally without the chain stays twisting and effectively compacting? I had thought I pointed this out earlier but I was not explicit enough. But if it isn't possible for the BB to flex down without pulling the dropout toward it, what would happen when that tension is taken off?

Thanks.