"Geek", not nerd. If you don't mind.

True, but I think that is important. Less than error is a valuable learning.

We are relying on the definition of modulus. Rider is exerting same force no matter what. The strain response to that stress is all that changes. For metals at least, the region of the stress-strain curve under study is very linear. Not sure about carbon fiber composite.

The stressing force applied depends on the angle of the frame at your feet, mass (or momentum) of the frame and how much if any force is applied at the handlebars and seat. (all of these may change depending on whether the frame is bendy or stiff because the rider will alter his mechanics)

Even if the stress force were the same, the elasticity may differ. Energy loss will be at least in part to imperfect elasticity.

So there is no data. Deal with it. The crowd sourced solution (every bike frame innovation not related to weight through this history of bicycles) has led to stiffer and stiffer frames. You, sire, are the one who has more need to convince with data.

Magnitude, I have no idea. With no funding and no resources, I am unlikely to be able to lead a study on this. With powermeter error being on the order of +-2%, those in the crowd wanting to hook up $7000 worth of powermeters to a bike are not likely to be impressed with the information they get. It'll be noisy and hard to interpret.

What I do know are some engineering fundamentals:

1) the shortest loadpath is going to be the most efficient. You route a load path through energy storage (frame flex acting as a spring) and there will be more losses than if you put the energy straight to the wheel. Is this trivial? Generations of bike racers liking the "feel" (and what are elite athletes anyway but the masters of their body's "feel" and its effect on winning the game) of stiffer bikes and flocking first to aluminum and then to carbon implies this effect is not trivial. And don't give me the "advertisers are god" argument. Everyone knows the bike companies make larger margins on cheaper frames. If athletes didn't have feedback to mold marketing push, everyone would be riding Denalis and paying for the privilage.

2) we are talking about energy, not force. Energy, or "work", is force over a distance, so those that say frame flex doesn't matter because the force on the BB shell are the same for a stiff or a flexy frame are mistaken. The flexy frame "pulls" more work from the load path through the deflection distance. If the frame is perfectly stiff, no work is done on it at all, regardless of the force on the support.

3) The wheel, the cyclists' legs, and the bottom bracket are what supports the drivetrain. All three of these supports are avenues for stored energy to be released. A lot of people quote the "bucket" experiment. It is a bit bunk because it is a static experiment on a dynamic system, but it does show that energy stored in bottom bracket flex can be released to the wheel. Fair enough. Now, a slightly different experiment: take you pedals, put them vertical. Step on the lower one so you bend the frame. Hold the brake. Now release the brake. Do what you want with the brake. Bounce on the pedal. Where does this stored energy go? Probably into raising your body rather than turning your wheel. Again, a bit bunk because it is a static experiment on a dynamic system, but as in the prior experiment, it does show that energy from the bottom bracket does not have to be returned through the drivetrain to the wheel. Energy will be released through the stiffest load path. Depending on the state of the system, this is not necessarily the chain. In a system dependent on linear pistons (like a bicycle) the direction this energy is released depends totally on the piston angle. At maximum chain tension, most of the energy is transferred to the wheel. At minimum chain tension, most is put back into the piston in the form of feedback. With the bicycle crank powered by an efficient pedal stroke, maximum deflection of the bottom bracket will likely be at maximum chain tension. This energy stored in deflection is released as the chain tension falls, which implies you are storing maximum energy at the precise moment when energy transfer to the wheel is most efficient, and releasing it progressively as that load path loses it's appeal.

Now, obviously this is a thesis and remains to be tested. The magnitude of the effect is unknown and likely small in the overall scheme of things (however, a small advantage is still an advantage; the only choice then being cost/benefit), and the system is complicated and hard to simplify and model. The above three points are what I expect the answer to come out to be, given my analysis of the system. That said, crowd sourcing is actually pretty accurate under the right conditions, and generations of cyclists attest to the superiority to stiffer bike frames (and other components). Discount the crowd sourced solution at your own risk. Every time a solution appears that increases the stiffness of the bike frame without affecting weight, it is immediately and nearly universally adopted by the performance cycling community. You might say it's a big marketing conspiracy hoodwinking literally millions over time, but that's at least as speculative (and very likely moreso) as the above analysis.

BTW, anyone stop to wonder why steel frames feel "lively"? It is likely because when you push the pedals hard, energy is stored in the bottom bracket "spring" and released back into your legs pushing your body upward; hence the "springy" or "lively" sensation that steel fanatics love to quote. Does this energy return back to the pedals? Depends on how well you adapt your pedaling to put your body's increased potential (read: gravitational) energy back to use. If you are "on top" of your gear, you can likely pump that energy back into the pedals with well timed pedal strokes. If you are "slogging" the gear, then you are likely catching that energy with your arms and making your upper body tired.

It's a big number subtraction problem. 2% uncertainty in the endpoint measurements might lead to a 50% error after the subtraction (say drivetrain losses are 4%; 2% is 50% of 4%).

You don't waste force, you waste energy. Energy is, in this case, defined as "work", or force over a distance. A "flexy" bottom bracket will absorb energy proportional to the amount of the force multiplied by the distance it deflects. A perfectly stiff bottom bracket that has zero flex will absorb no energy.

In the case of a perfectly stiff bottom bracket, the bottom bracket acts as a perfect support, and the force applied to it is simply a reaction force. You don't "waste" this force any more than a coffee cup "wastes" gravitational force while being supported by a table.

While this is true it still tires your muscles and uses energy to apply a force even if it is not moving anything. As to if it is more tiring to apply a force to an immovable object or one that moves a small distance, well that is the question.

The power meter seems attractive at first. Your analysis suggests otherwise. Isn't there a precise way to measure these differences. Rear wheel speed should be able to be precisely measured, no? At constant resistance, wouldn't the wheel speed be a sensitive indicator of power imparted to it.

Is your understanding of the term "on top" of a gear literal as in having your body's gravitational force assist in the pedaling. I always thought it meant not struggling to pedal or some such thing. I can't see any real way for spring that the legs feel to be sent back to the crank.

Yup. You push against a brick wall as hard as you want with no deflection or movement, but you are still using or wasting energy. Same with a perfectly stiff bottom bracket.

Do you know what hysteresis means? I think it doesn't mean what you think it means.

The only reason I say this is because your sentence made no sense.

We do need you to keep posting though if we're going to make it to page 7--maybe you can do one to defend your word choice and another to flame me for being a "definition nazi."

The thing is, if there is no data available, then we don't know if "stiffer = better power transfer." The ubiquity of the claim is not matched by the availability of the data.

Keep in mind, by the way, that this has nothing to do with ride feel, control, or confidence when sprinting, climbing or descending. Those are completely unrelated to the issue of power transfer.


Uh, no, it really hasn't. ;)

A lot of R&D has been poured into "endurance" bikes, which dampen vibrations and absorb shock; electronic shifting; disc brakes; hydraulic road brakes; aerodynamics; GPS's, power meters, saddle designs, ovalized chainrings, tubeless tires....

Plus, a preference for stiffness by *cough* some men *cough* :D does not prove that "stiffer = more power transfer." Answering that requires quantification.


Or, we can say that any difference that can't be quantified with good-quality equipment might not be worth worrying about. ;)


Velonews did a test, where they loaded 100lbs onto a pedal and measured the flex on four endurance bikes. A reasonably stiff BB deformed by about 0.7mm, and what they described as "one of the softest bike on the road" flexed... 1.7mm.

In the absence of better data, I don't think we're talking about a massive waste of energy here, or a huge difference in terms of load paths.

Besides, the typical counter-argument is that the frame acts as a spring, and returns the energy during the upstroke. The only way energy would be lost is to friction, and that's likely to be extremely small, even with a traditional steel frame.


That might be part of it, but I'd say a far more likely reason is because the frame transmits a lot of road buzz. Lots of bikes are both stiff and "lively." Similarly, the Specialized Roubaix is very stiff, but feels "dead" because the frame is soaking up so much of the road chatter.

What's so special about Page 7? Is that like Page 2? ;)

Predicted on pg 2.

A couple of things come to mind based upon the discussion thus far.

a. what you state above. At the end of the day...I have said this before having owned both flexy steel and stiff carbon bikes...the control of a stiff bike is what may contribute the most to it being faster. I had an old steel Bianchi that was a veritable wet noodle in terms of stiffness. If you hit a dip in the road just right, the whole bike would load and spring in the air. Same for the BB. I could rattle the chain out of the saddle. To me, when I pushed on the pedals, this old steelie was 'alive' all right...there felt like a lot of wasted motion.

b. There seems to be a consensus among the top bike makers and the top pro riders in the world. Stiffer = faster. Now one can say that the hundreds of engineers that work for these top companies...or the hundreds of cycling teams and top riders throughout the world are wrong. I am waiting for a top company to come out with a flexy frame that is faster than a stiff one. Haven't seen one in the last fifteen years due to the continued evolution of carbon fiber. So I side with the engineering collective being a card carrying member myself. Riding a trampoline for a bike is slower. To me, this just makes plain sense but I do not discount the control contribution which contributes to a cleaner pedal stroke where the springiness of the frame is pushing less back on the rider in varying amounts throughout the pedal stroke making the pedal stroke harder to control. I believe it is part a control thing but also lost energy stored by the frame...perhaps due to heat...or added fatigue to the rider by the frame pushing back on the rider's leg. Placing a stop watch on a flexy bike versus a stiff one at the track or up hill climbs may be more accurate than using a power meter to quantify any difference. I don't believe a flexy frame stores a lot of energy but it likely does make a difference in competition. Some may prefer a softer feel in fact and many have stated they like the lively feel of steel or Ti which can also be achieved with carbon fiber.

Oh that You don't know wnhat it means nonsense again. Yes, I do - just take it for granted, OK?

I understand the lure of the consensus argument, but if you were really siding with the engineers here, I think you would say we can't know it without data. That is true even if the data says the effect is smaller than the variability. The complete absence of data cannot be countered by feel and opinion. Once again, in my original post nobody was being labeled as wrong. I simply asked for the proof. Not the opinion, the proof. That is what scientists do.

As for waiting for a flexy frame that faster than a stiff one, how will you know when you know when it happens? Everyone will agree that is the case? The maker will advertise it as such? This is a classic case of using the hypothesis to disprove itself. Once we have (mostly) agreed that the only evidence for stiffer = faster is opinion not supported by data, how can you ever have a case of flexier = faster?

Supposing there is an ultimate sweet spot as is so often the case? Not the flexiest. Not the stiffest. Some unpredictable point in-between. How will you ever find it without controlled experimentation. Everyone should be on this bandwagon as it is in everyone's interest to know, not think, what is best.

In my opening post for this thread, I said I wasn't taking sides, and I meant it. I just want to know the truth, not as a matter of faith like religion (A told it to B who told it to C...), but as an fact established by scientific inquiry. Those who predict the locking of this thread have a valid point: we have established a couple of things that aren't going to change with more pages of posts: 1) the data either doesn't exist or is being kept from us, and 2) there is no support for the prevailing opinions except the mass of those opinions.

Stick a fork in this thread. It's done!

So much to say. :)
Just because you have no data, doesn't mean the data doesn't exist. No you don't have it. Nor do you have the dynamic CAD model of a Specialized, Trek or Giant frame as it simulates the bell curve of road loading. Nor do you have the wind tunnel data of a Foil or a Venge. Engineers are indeed data driven. Are you saying that the hundreds of engineers that are employed by the top carbon fiber racing bike manufacturers don't have data? Laughable.

You want data? Look no farther than pro cycling. This is available to the public. Name a single flexy frame that has won anything in the last 10 years. You can't. You don't think pro riders don't test different frame, wheelset and groupset combinations? Wrong again. They test 'everything'. This is what they do for a living. If a pro is faster on flexy square taper crank or noodly steel frame or flexy wheels, they would be on them.

If you want to know the truth, racing always exposes it. This is true in auto racing as well. Pro teams in any vehicle endeavor test extensively. Everybody likes to win.

Of course the data may exist. I never said it didn't So why is it not published? Giant is very proud of its weight and stiffness data. We have all seen the video. Where is the power transfer data to complete the picture. Can you honestly say that isn't peculiar?

Your argument about pros is understandable, but it is so 1984ish. Or like classic ancient religion. Just trust the pros. The priests of cycling know what is best for us. They will lead us to the true way. That is not science. Okay, so once again, why can't I see the lab data? What are they hiding?

And another thing. I am not a pro rider. Duh! Suppose I don't want the stiffest/fastest/surely most expensive. Suppose I want to find the point of diminishing returns. How do I do that? Not by watching the TdF. And do you really think the most sensitive way to determine the best possible bike design is by looking at what is being ridden in sponsored races? Do you really think pro bike riders have any clue about scientific method, statistical significance or any other aspect of structured inquiry. "Uh, I really like the way that one feels. I think I will try it tomorrow and see how I do." Oh yes, very scientific! You know as well as I that is not the basis that the manufacturer's use. You said it yourself: "laughable".

This isn't about flogging old steel or any other outdated technology. It is just about being curious regarding the true relationship between frame stiffness and power transmission. But the real question that is now coming to the surface is this: Why are you so invested in the industry's right to keep that secret?

Yes, I'm sure the bike manufacturers have no reason to exaggerate one specification for marketing purposes. :D

The problem is that without quantification and evidence, we have no idea of the actual benefit, who would benefit, how much wattage is required to produce what theoretical power loss. We also don't know the points of comparison. Are we talking about two modern frames? A traditional steel frame to a high-end CF race bike? An entry-level modern road bike to a low-end CF frame?

It's also downright weird that the manufacturers are willing to release extensive performance data for aerodynamic advantages, but none for stiffness.


Actually we can't, because none of us civilians have found the actual data. Thus, we don't really know the engineer's criteria.


We might just be talking about the difference between, say, 0.7mm and 1.7mm of deflection, at least with modern bikes. That's not a fun trampoline, is it? :D


Again, all that indicates is "pros prefer stiff frames." We can also make a pretty solid claim that a stiff frame improves control, just based on subjective evidence.

Or: There's some evidence that ovalized chainrings offer a small but measurable advantage, and some pros have used them with good success. Why hasn't the entire peloton switched? One possible reason is conservatism of the pros; another is sponsor commitments. Similarly, it has taken years for riders to start using aero road frames. For a variety of reasons, riders did not have the option to switch en masse to Cervelo frames when they first came on the scene.

I.e. "pros prefer stiff frames" does not prove the specific claim that "a stiffer frame transfers more power."

We've also seen the allegedly advantageous frames and extras not​ produce winners every time. Consider Cavendish. For several years, he was using standard road bikes like the Addict, Tarmac and Dogma, and beating the snot out of equally experienced sprinters on aero road frames. He's now using a Venge, with essentially the same results -- in fact, he's even gotten pipped at the line in this year's Tour. I.e. in an aspect of the sport where the tiniest increments mean a win, the allegedly significant (and quantified) advantages of the aerodynamic improvements don't appear to translate into victory.

Anyone try the "bb trick." It was just something I was curious about so I'd hit the bb with my palm to see how much the bb area deflects. My Trek 1220 would show the bb oscillating like crazy doing this. When I went on a full sprint the bb would displace in a similar manner. Not just 1.7 mm...

I tried the same "test" on a old Tange 2 steel frame with lugs and saw much less deflection and oscillation. Just by absolute measures, my experience on the road seemed to relate to this very simple test I did for fun.

When I sprinted at 32 mph on the Trek 1220, people behind me would comment on how much the rear was swaying about and I could physically hear all the scrub from the rear tire. I would not be surprised if there is an increase in rolling resistance from a super flexy frame.
For this frame, I even wonder if there are aerodynamic effects that need to be considered.

Not necessarily. Doing a double power meter comparison you're really concerned about precision and linearity. Since its a relativecomparison the accuracy is almost irrelevant.

Accuracy = How much power am I putting out?
Precision = How well can a measure a CHANGE in my power meter.

Justkeeppedaling: A lot of flex comes from wheels and tires, and a lot of apparent flex during a sprint is just from natural swaying of the body & bike.

The old Tarantula machine seemed like a good way to measure flex, but it did nothing to settle the discussion of if/why/how stiffness helps performance.