Does stiffness really matter on the bike? (Page 2)

Try putting in this disclaimer in the Subject - "1TT - SHUT IT!"

My post is a bit misleading. Most loss of power is due to friction in your drivetrain, so keeping your drivetrain alligned reduces friction and therefor power loss in the drivetrain. Power loss due to flex of the rest of the bike frame and pedal to crank seems secondary.

I have no scientific papers to back this up, it's just and educated guess.

 

The chain is going to eliminate that friction effect.

This is one of those problems where it is really hard to get a grasp on the magnitude unless you've made some measurements. 

A few further comments.

"Laterally stiff but vertically compliant" is a real phenomenon.  Road vibration is absorbed mostly by the ability of the bike to 'flex' (in this case, very little, but just enough) in the vertical plane).  Lateral torsion in the bottom bracket and the cranks and so forth is where some energy can be lost.

Yes, the torsion is caused by a side to side force on the BB, and so yes, the reaction (restorative) force will be in that direction too.  But the 'levers' (cranks, pedal spindles) that transferred your (mostly*) vertical downward force into the initial horizontal force that torques the BB will also transfer the horizontal restorative force back to vertical (this time upward of course).

I happen to work about 3 feet from my trainer, and the bike (the same Orbea that Andrew mentioned before - so yes, not the stiffest bike out there!) is on the trainer. So I hopped on.  (Should I log that 3 minutes?)  It seems to me (but really this would require some careful and difficult measurement to be sure) that the return happens at about the same time that the opposite foot begins the downstroke.  If that's right, then the return is going straight into (well, not 'straight' -- see above) into the opposite leg.  How much that leg is able to store and put back into the pedals I have no idea.

--
* The force that the rider puts on the pedals is unlikely to be completely vertical; any lateral component of that force will also contribute, somewhat more directly, to torsion of the BB and whatever else.

Sounds like you should be able to measure this loss if you have a pedal based power meter and a power tap and them compare results.  A Crank based system and a power tap might also work, assuming the loss between the pedal and crank are minimal.  Although I am sure there is some discrepancy due to accuracy of the devices.  Could be an interesting experiment to try on various frames.

I would expect that any losses due to the frame would be lost within the drivetrain losses that are normally accepted to be ~5%.

Shane

IMHO- Non-issue among top quality road bikes, or even on good tri bikes there's since little full-power sprinting in typical tri course.  On most decent quality road race bikes I think what most riders perceive as frame flex (or crank/BB flex) is really due to wheels.  Or maybe bearing adjustment issues.  I currently own 3 roadies (CDale CAAD5, CDale DA Synapse with BB30, and Specialized Tarmac).  All are generally felt to have excellent pedaling stiffness/power transfer, inc. that old stiff AL CAAD5.  I'm 'only' 6ft/168#, but I cannot induce perceptible frame or crank flex in any of them.  However I do notice differences in lateral stability/handling when changing to different wheels (at same tires/pressures).  I'm not saying that ALL road bike frames are equally stiff, just that differences among top bikes are too small for most non-elite racers to really notice.  OTOH- I have ridden some lesser quality frames which do seem to flex on me when mashing out of the saddle.

Bottom line is that Mark Cavendish still kicks a$$ on a Wal-Mart special

The displacements of flexion in the frame will be undiscernable to the eye. If they were in fact visualy apparent, then the bike would soon plastically deform (like that guy said earlier about a paperclip bending too far) from the amount of cycles a bycycle frame experiences.  And if your frame is aluminum or carbon, microfractures would develop and your frame won't live long.

The amount of power that is lost due to frame and component flexion is not transferred back into power generation and is lost in the form of heat. If you can measure the exact amount each component flexes, then you would have to use a series of equations to turn that power absorbtion number into an actual temperature increase.  If you really care, these equations can be found in a heat transfer text book combined with a mechanics of solids textbook.

Breaking the process down to specifics, your foot for simplicity can act as a point load on the pedal. Depending on the Q factor of the crank, a moment (torque) will be applied to your frame acting perpendicular to crank rotation.  THis is the force absorbed by the crank.  The smaller the q factor, the smaller the moment (torque). the majority of power generated is in the top part of the stroke.  And the only power that can reach your wheels is a moment acting parallel to your crank. So when your frame is experiencing its highest moment, your frame is in its highest bending state.  And during this, the force applied by your foot does not make for a moment that is perfectly parallel to the crank. So that angle that is now created between the non vertical plane of the chainrings and the vertical plane of the moment being applied by your legs can be used in a simple equation.  Cos(x) * power at the pedals = your drivetrain force.  Multiply by coefficient of drivetrain losses and you should have the power rating a power tap would read.

also to note, the higher the q factor, the larger the angle will be as mentioned above.  Does stiffness matter??  Yeah i'm sure it does.  But i bet only a pro could notice a diff between high end frames. I'm sure an experienced cyclist can notice between entry level and high end frames. I can't speak much to personal experience as my bike is entry level, and is made of 7005 aluminum. Its also the only bike i've ever ridden.

also i wanted to mention, that sound that your chainrings make when they're not aligned properly....  thats energy lost in the form of sound waves.   Definatley not anything significant

Kinda like a string on a cello.  Energy input is lost through heat and sound. 

When climbing on my old fuji and standing on the pedals, I bounce.
When climbing on my lightspeed and standing on the pedals, I surge

I'm not going to address this entire post (I disagree with a lot of it, but this isn't a physics forum), but I will say something regarding this:



No.  The flex is very apparent visually.  You underestimate the strength of these materials.  When I put my Orbea on a trainer, I can watch the BB move back and forth as I pedal.  As long as the flexion is within certain limits (and it sounds like you have underestimated those limits), there will be no plastic deformation.  What those limits are depends, of course, on materials and design.  (In any case, carbon -- at least in the form of the tubes that are use on bikes -- won't deform plastically hardly at all.  If it flexes too far it'll likely just crack.)


and this: 



It is very easy to feel (and see!) the difference in stiffness between various frames.  You don't have to be a pro. 

Right, and just like that cello, it actually takes a 'significant' (by many riders' standards) amount of wattage to produce that sound.  Even if the sound is relatively quiet (say, 30 dB, and I doubt that a misaligned chainring produces a sound that quiet) then (if I've got the conversion right -- I didn't double check this) you are using 1W of power to produce the sound.  (There's a reason that stereo speakers are often rated in Watts!)  It goes up pretty quickly from there (as decibels are logarithmic).

Bottom line:  a lot of power is lost through the drivetrain.  A noisy drivetrain is a pretty good indication that you are losing a lot more power than necessary.

Edited by Experior 2010-10-29 7:47 AM

< -- Mechanical Engineer (did a LOT of work on material stifness and strengths)

Lots of misinformation (and some good nuggets buried) in this thread. A few random thoughts reading this..

1.  Flexion IS visible.  Put your bike on a trainer and hammer away.  Look at the rear skewer and axle and BB.  You'll see flexion.

2.  Any energy that is put into the bike to make it flex is energy that is NOT going into a drivetrain.  That energy is lost.  It is not driven back into your legs to use again.  It's gone.  (Where does it go, rather unimportant but some is translated to heat, some to sound, but a majority is returned like a spring to your body and to the ground via the tires.  But it is lost energy that is not making you any faster)

3.  There is an important factor that I did not see mentioned.  Comfort.  You have have the stiffest bike in the world but if you are uncomfortable on it you are not going to be as fast.  I test road a Tarmac and a Roubaix for my road bike and I felt a LOT more comfortable on the Robaix.  It however was not as stiff as the Tarmac.

4.  A more STIFF bike is actually going to have a HIGHER rolling resistance (all things being equal) as the bumps in the road need to be absorbed somewhere.  If the bike is too stiff to absorb those bumps then that energy slows you down.  So there is a fine balance here.

5.  There have been several studies that show that tire pressure and width and not the most important factors in the determination of the tire's rolling resistance.  A consistent contact path is the most important factor.  Also considering the surfaces you ride on vary greatly (from new asphalt to bad chip seal roads) and that vaerying the pressure by 10-20 PSI makes a minuscule change in the rolling resistance it's really best not to dwell on this too much.




Edited by TriRSquared 2010-10-29 8:07 AM

On my old carbon frame, I saw significant flex around the BB.  Even in moderate riding the flex is clearly visible, let alone when I was pushing hard. 

As far as your comment that only pro's can feel a difference, I'm not sure where you're coming from on this.  Even a complete beginner can feel the difference.  It doesn't take much.

Outstanding info.  #2 is really what I was looking for on my original question.

That being said, the only issue I take with points 3-5 is that stiffness can be the same for different material bikes (carbon, alumn, steel and Ti) but have very different dampening and comfort characteristics.  Not to mention different weights.

My Ti bike is extremely stiff, but it's both comfortable and dampens the road vibration very well.  An Alumn bike of the same stiffness would do neither of those things. 

I know you know this, but I'm just clarifying the fact that "stiffness" doesn't tell the whole picture.  I'm a physicist, not a materials engineer, but I know that strength, dampening, resistance to deformation, etc, are all equally as important when determining the comfort, rolling resistance and other factors that you mentioned. 

You are absolutely correct.  There are other factors at play.  (Doesn't invalidate points 3-5 but it needs to be considered)

Stiffness is the main factor in static loading, but riding a bike is dynamic.  There are factors such as dampening that come into play.  How much they contribute is a question for the lab or simulation but they are indeed factors.

The problem with your example is that you'll never have a Ti and AL frame with the same stiffness.  They have differently material characteristics.  (A AL and Ti frame made exactly the same will not have the same stiffness)

Dampening characteristics come from things other than materials however.  Much of it is in the design of the frame itself.

Edited by TriRSquared 2010-10-29 8:31 AM

You could make bikes of almost identical stiffness using different metals, but the amount of material would be different, impacting the bike in other ways. 

This is true.  I test composites for a living (yet another engineer here). 

Not to add too much, as TriR did a good job, but you'd be surprised at how much deflection you can put into a material in flexure and still have it be in the elastic range.  (Pure tension and compression, not so much, in this type of material).

That is the crux of the benefit of carbon fiber and composite materials.  I can design my material to have the properties in the axis I choose that best suits my needs.  Metals are have the same response in all directions.  Based on the layup, I can create a lighter frame with composite that will be stiff in the direction it needs to be and compliant/damp in another axis.  Of course, all the engineering that goes into that is very expensive.

That's what I said.  The geometry would be different. 

Stiffness is a material characteristic.

Dampening is a system characteristic.  There is not property for the "dampening" of a material.

Not to pick, but this isn't correct.  Dampening is both a material and a system characteristic.  (For that matter, so is stiffness).


Edited by sand101 2010-10-29 11:18 AM

Geometry doesn't neccessarily have to be different, but tube thickness would be.  And dampening absolutly is a characteristic of material. 

I pretty much agree with all of this, but in #2, some of the energy that is returned to the body could, potentially, be stored and returned to the pedal.  This is what happens when you run, for example -- some of the energy from the impact with the ground is stored in your Achilles tendon, which then springs back to provide some propulsion.  How much, if any, is stored in this way when you ride I have no idea, but it is at least possible.

And on #5, really you need to define 'miniscule'.  As I'm sure you know, cyclists and tire companies have done studies of the crr of tires at various pressures, on various surfaces, and I, at least, find the differences to be enough to motivate me to take it into account when considering my tire pressure.

The speed of wave propagation is a material characteristic.  I guess that could be considered dampening.  Typically the viscous dampening ratio is applied more so to liquids but I guess you could apply to it to solids as well.