Disco Stu

I just figure that any sprinter type who doesn't any frame flex, wouldn't want lateral flex OR vertical flex

Brian Ratliff

Nope. The crack in the paint just follows the joint line, probably because the glue holding it together is less stiff than the CF composite tubes.

Phantoj

"Vertical compliance" should be measurable. I haven't seen any bike reviews that actually measured the vertical compliance, but I think some Euro rag does.

I believe that vertical compliance of the frame is always more than an order of magnitude less than the vertical compliance of the tires and seat, and therefore has vitrually no effect on rider comfort.

Ramjm_2000

[QUOTE=AEO;6731081]paint is elastic enough that you wouldn't see cracks unless you broke the tubing.



I see it like this, since I experience this myself...

Lateral frame flex is bad when you have to control a squirrelly ride, you put more energy into co

jupiter422

There is always energy lost in displacement of a material. Bending a material takes work. Some of that work may be stored in the elasticity of the material, but some is also always lost in the form of heat.

jmntn2000

Frame flex decreases efficiency. Regardless of how much energy the frame returns you are still talking about wasted energy bouncing the bike up and down not propelling it forward.

logdrum

As I said, most bikes seem to have acceptable lateral flex and most modern ones are stiff. However just by design, it is very hard to see any vertical compliance. The Cervelo aero down tube shape. That makes it really vertically stiff -- plus it's a triangle already for God's sake.

My issue why is that that Vertical compliance is BS... Show some numbers maybe even high speed filiming of a vertically compliant a frame is.

waterrockets 

While this is true, with any frames of reasonable quality, reducing frame flex would have a small practical benefit to power transfer efficiency that it may as well be zero. I would argue that an ultra-stiff carbon fiber frame probably looses more energy to material damping than a slightly flexier steel frame. Still, neither one is going to lose enough energy to measure in a cycling application.

Regarding sprinters: I'm a big, somewhat powerful sprinter, and I enjoy the slightly springy feel of my frame. It's like dancing on a wood floor instead of dancing on a concrete slab.

Fat Boy

Then you need to ride my TT bike and my road bike back to back with the same wheels and seat. It's pretty amazing how smooth the road bike is in comparison.

Someone wrote that if it deflects it must produce heat. To this I agree, but I'm also realistic in the magnitude of the heat produced. It's VERY small. Just taking a SWAG I'd say on the order of 0.00001 of the input power. If you're putting 300 watts into the bike then maybe 0.03 watts is being given off as heat due to frame deflection. This is just a wild guess, but regardless, if you're staying below the yield strength of the material, the internal damping characteristics of any material used is miniscule in comparison to, say, tires. When you feel a smoothness or harshness on a bike, you're not feeling damping, you're feeling spring rate. As it turns out, the softer the bike (i.e. lower the natural frequency), the less damping is needed to control high frequency inputs (i.e. road buzz).

akatsuki

Frame flex is probably more like running lower pressure in tires to some extent - when you hit a bump you can either have your bike deflected up and down, wasting energy, or the frame can absorb it a bit (chatter)... A lot of people run their tires at higher pressure cause they like the firm feel and feel like they are going faster, but analysis has shown that they are actually going a little bit slower...

Brian Ratliff

No need for green paper here. Proof of decreased deficiency is that no spring is perfect; there is always a damping effect. Thus, energy lost. Energy lost is energy not transmitted to the drivetrain.

Small effect though, obviously. Probably neglegible over the long run. But, as to it's presence, a pretty obvious effect.

Now then, if you could pull out your green pad and show that there is no decrease in efficiency, then you will be famous, because you have just created a perpetual motion machine. Congrats.

Now, there might be a biomechanical decrease in efficiency. Basically, a yielding frame results in the center of rotation of the crank to be offset a bit toward the driving pedal. In other words, the crank is not rotating strictly around the axle anymore because of the deflection of the bottom bracket. This will result in basically an egg shapped chainring as the chain/chainring separation point will follow an egg shapped path rather than a circular path. This might or might not be beneficial from a biomechanical standpoint, and it might vary from person to person.

Again, though, the effect is small. Under high torque, it might be noticeable, particularly when you add the third dimension in there and get the crank swinging in the lateral direction under load. And every time you have a deflection in any direction, you definitely have losses, though the magnitude of these losses are small. All you can argue is that the magnitude is small enough to be neglegible. If your argument is that there are no losses, then congrats, you have yourself a perpetual motion machine.

Brian Ratliff

You are off by an order of magnitude or two.

Given data from this website on frame deflection, running through the numbers and assuming the spring constant is linear, the total energy put into the system is on the order of 0.54W at 90 rpm, assuming 1.6mm of lateral deflection under a cyclist standing on the pedals (lateral 23kg mass).

If half of this energy is dissipated as heat (not unreasonable as when you load the frame and then release it suddenly, the resulting oscillation damps out quite quickly), then the losses are on the order of 0.25W, or 0.08% of power at 300W input. This is for a OCLV frame, BTW.

For a "noodle" frame with a 2.2mm lateral deflection under a 23kg load, the total power into the system is on the order of .75W, and the losses are on the order of 0.12% of 300W input.

The primary assumption here is that the force is applied as a step function at 3 cycles per second (90 rpm multiplied by 2 for the two power strokes on each pedal cycle) instead of something more akin to a half sinusoidal function.

Don't believe the actual numbers, of course, but this will give you the order of magitude for the losses due to lateral BB deflection. Not a lot, but certainly not non-existent.

(methods below: )

F = k*x
E = 1/2*k*x^2
P = E*duty cycle
P_loss = 1/2*P

Assumes:
1) step function for force
2) losses 1/2 of input power
3) k is constant, i.e. spring is linear

Fat Boy

Interesting. One question, How did you determine that the loss was 1/2*P? I could just as easily say that number should be 0.05 or 0.005. I can see some pretty substantial losses on a material that is built to have internal damping, say polyurethane, but, I don't see Al, Ti, Steel, or CF having nearly this level of loss. Impact hammer testing would give us a better idea of what we're working with.

Someone mentioned a soft frame was similar to running low tire pressure. A tire's rolling resistance is exactly the sort of loss that is produced by something with a (relatively) large amount damping and even then the damping ratios of the bike as a whole are very small when compared to a full suspension bike or car. According to the tire rolling resistance test that is in a different thread (assuming the magnitude of the numbers are reasonable) the difference between 200 psi and 120 psi is 2.7 watts. I'd guess that any of us would feel a significant harshness difference between these two settings. Remember, this is comparing a components that is meant to deflect, produce heat, and provide a certain amount of damping. I just have the feeling that this sort of difference is going to be much larger than the difference in power loss due to frame deflection.

I also have the feeling that we'll never really know.

waterrockets 

The thing is that the load is not released suddenly at all. It's a smooth oscillation. I think half of the energy is actually completely unreasonable.

Look at this graph from a Computrainer SpinScan:

Brian Ratliff

@Fat Boy

It is a guess, of course. I cannot imagine that the efficiency would be much higher; otherwise, with a cyclical forcing function, the BB sway amplitude would get wider and wider with more energy input. I would guess that this is a well damped system. Remember that this is not a simple spring either - it is a frame. Energy put into the system will cause the entire frame to flex. There are dampers all over the frame to dissipate this energy; for instance, the saddle to hip connection and the hand to bar connection. Everything is connected together in a frame, so a damper anywhere in the system will absorb energy put into the system. It's not just the frame material which will dampen.

But yes, to complete the analysis, you'd have to do some sort of dynamic analysis on the frame/bike/rider system.

Homebrew01

I was thinking of riding my 1986 Cannondale one of these days. Just gotta pop a front wheel on it. Might be interesting to compare it to the CAAD 8 I'm riding now.

Brian Ratliff

@WR

Yes, the load is sinusoidal. But a damping factor for a dynamic system can be obtained by applying a step function to the system and measuring the response. The faster the amplitude of the response decays, the higher the damping factor. The damping factor is the system variable responsible for absorbing energy.

njkayaker

Obviously, not. That's your idea, not mine.

No. Frame flex doesn't go into driving power.

Think of having springs on the base of your foot. Some of the force is used to flex the spring. The rest of the force is used to turn the crank. The force of the springs relaxing doesn't go into turning the cranks either (it tends to push your feet of of the pedals).

===================================

This issue is how much or how less force is going to driving the bicycle. You really don't care where the loss is going.

Some of the loss goes to heat the frame but much of it might go to making the rider bounce up and down. For metal frame bicycles, any heat would quickly diffuse through the frame and be taken away by convection. Given that the frame flex has a low frequency, the heat could transfer out of a carbon frame fast enough to keep up with its production.

Also note it's likely that all frames don't flex very much in normal riding. It's probably more of an issue when "pounding" (eg, like in a sprint).

waterrockets 

The energy returned from the frame goes into the bottom of your powering foot, which is already pushing on the pedal, and goes straight into the drive train. It never returns more than you're delivering, so it's not going to tend to push your foot the wrong way.

Springs on the bottom of your feet wouldn't be bad.

Fat Boy

This is not correct. Draw the free body diagram. Force on one side of the spring (i.e. your foot) will be counteracted by the force on the other side (pedal). The spring deflection will give a phase shift of the force input, but it will not change the average force over a complete cycle.

Fat Boy

At least on this point, we agree. It would be an interesting thing to do if you had a 'sandbox' job at Trek, Specialized, Cannondale, etc. It wouldn't surprise me if it's already been done.

logdrum

The smooth ride may be attributed to more lateral flex on your road frame, maybe an integrated seat post on your TT and a big setback on your road bike's seat post and the rake of the fork

When I upgraded from a straight fork to a curved fork on my steelie, I felt the difference in smoothness. Then I changed to a seatpost with setback and a longer saddle that was a carbon shell -- now I have a smooth ride. My friend did that to his aluminum crit bike as well and completely a different bike which is till responsive to his pedaling input.

Unfortunately my field is semi conductor and now software, so I might not get all this static physics stuff. I just can't get myself to believe about the vertical compliance marketing. Show data and studies then maybe I'll believe it. I am equally happy with my alu bike and steel bike. If I want/need vertical compliance then I'll add suspension!

waterrockets 

I've asked Specialized and Cannondale. C'dale was at least kind enough to answer with "it's too hard to test, but it should be obvious that stiffer frames are faster." Specialized has been completely ignoring me, and they're the ones with the TV ads bragging about efficiency gains.

Fat Boy

That sounds like an answer to an engineering test where you're supposed to show your work, but don't have 1/2 an idea WTF you're doing. You just say, "It is intuitively obvious that the answer is 'B' ".

If these bike manufacturers were building cars instead, they'd know exactly what the effect is. It's probably an issue of money and the fact that 'stiffer is faster' is in everyone's head, so that's what they produce. Interestingly enough, the newer C'dale and Specialized bikes aren't terribly stiff. Trek's Madone is also quite flexible (based on reviews). If stiffer is faster, why doesn't C'dale just keep reproducing the same aluminum bridge that they build in the 80's and early 90's?

AEO

Are you talking about the Cannondale 2.8 which so stiff and harsh that it fatigued the rider way faster than other frames around at the time?