As I understood Steven's explanation, the curved crank arm could compress like a spring at TDC (12 o'clock) when standing on the pedal and then release this energy as the pedal moved ahead.
At 11:30 AM 7/1/06 -0700, Kurt Sperry wrote:
>The problem with that theory is that the pedal isn't indexed to a round
>chainring in any particular way. You can put a 90 degree (or any other, it
>makes no functional mechanical difference) bend in the crankarm, a curve or
>a double loop-de-loop and from a mechanical standpoint nothing
>really changes except you've simply moved the 'dead spot' relative to the
>chainring and weakened the arm for a given amount of material used.
>
>Kurt Sperry
>Bellingham WA