Is the flex of the axle spindle linear? I'm not a physics major but my understanding is that flex is non-linear and will become increasingly more difficult before the point of failure is reached.

The flex is linear over a certain range of load from zero to some percentage of it's ultimate breaking strength, after that it becomes non-linear and no longer springs back to it's original position, it yields. After it yields it also becomes much less stiff.

Any mechanical component intended to last for repeated cycles of loading and unloading, especially when the consequence of failure is injury or even death, is going to be designed within the elastic range of the material with a pretty good margin of safety. Once you yield the material you are well on your way to initiating a crack and then that crack grows and then you do a face plant in the middle of a sprint.

Also realize that putting out 2400 Watts at 60 RPM is putting down 282 foot-lbs at the crank which, given a 175 crank arm is like leg pressing almost 500 lbs with one leg. I know, I know they pull up on one crank and push down on the other but you get the idea.

The point is that even the "lowly" campy square taper, cro-moly bottom bracket axle is more than good enough for 99.9% of the riders out there.

The marketing department at Shimano would like you to believe otherwise though so that you'll buy a new drivetrain.

Properly designed most bikes should last a lifetime, which isn't very good for business. So they have to come up with "NEW AND IMPROVED BOTTOM BRACKET, WITH SCRUBBING BUBBLES" so you'll buy more stuff. How much you want to bet we see 11 speed drivetrains and 135mm road bike rear spacing with ten years. "Well waddya know, all your stuff is obsolete. WE just happen to have the latest and greatest, not like brand X which uses that OLD technology." [/END RANT]

Also, generally speaking for a given bottom bracket shell diameter it's going to be a tradeoff between axle size and bearing life. The larger the axle the smaller the balls have to be and the shorter the life span. I suppose with double row bearing and ceramics that is less of an issue though.

Good info and thanks for the explanation.

It's like biopace and powercam! you get it all back at the bottom of the pedal stroke!

for a rather humorous explanation with some good analogies that make it all easy to understand (and explain to your friends) check out: http://bontrager.com/keith/rants.asp?id=32

A little more grist for the mill. If you crunch the numbers as I have above, you'll find that the ISIS design is about 3 times more torsionally stiff than the old square taper design.

Now, I've read on this board that Cipo rides an aluminum ISIS BB.

If that's the case, since the modulus of Aluminum is about three times less than that of steel, any stiffness gains by design are negated by material choice. All you've done is made a lightweight BB that's about as stiff in torsion as a traditional campy square taper cro-mo axle.

So Super mario doesn't care about the torsional stiffness of his BB axle, he's a weight weenie.