Originally Posted by
Brian Ratliff
Okay.
If that's the case, then there is no way I can see that the frame can return the energy back to the crank.
And if you were a mechanical engineer of genius, one of the worlds greatest bike designers, or an idiot savant who could run a Finite Element Analysis in your head, this would might something.
But you're not, so it doesn't.
Finite Element Analysis says that you're wrong -
http://www.bikethink.com/Frameflex.htm
Keith Bontrager's famous rant on frame flex says that you're wrong (look it up on the Internet Wayback machine - Trek had it pulled when they bought Bontrager) and so does David Kirk, who while not of Bontrager's legendary ability still seems to be pretty competent:
http://www.kirkframeworks.com/Flex.htm
It will deflect at the start of the power stroke until the force reached the peak, then it would return in proportion to the magnitude of the decreasing pedal force, but not give much of its energy back to the drivetrain because the spring force is returned during the part of the pedal stroke where it cannot very well help rotate the crank.
Well, yes: as previously stated, reality is not limited by what you think. Mr FEA says that your model is too simplistic.