I received the following post from John Buckley in an exchange we had over paralympians, leg springs, and push-off contribution at top speed:

"As to your observations regarding able-bodied sprinting – I’m not convinced! While you give a convincing example of why and how you think there is no aggressive push-off from the ankle, from a biomechanical point of view the ankle joint must do substantial work prior to push-off. But I am willing to concede that some of this might be geared towards pushing the athlete upwards rather than forwards (which would agree with your observation).

I did some work looking at joint power output in sprinting some 6 years ago. I found that there was a proximal to distal timing in the generation of peak positive power at the hip, the knee and then the ankle. That is the ankle contributed significantly just prior to push-off. I guess what it comes down to is how much of this ‘push-off’ goes to thrusting the athlete forwards and how much goes to thrusting him/her upwards? To maintain top speed there has to be some forward thrust (otherwise there would be deceleration), I guess your ice experiment suggests that the ratio of forwards to upwards thrust changes as speed increases (but I can’t see how it could ever go to zero forward thrust!)."

This debate/discussion is lively and entertaining, and will probably never be resolved to everyone's satisfaction.  What we are faced with is a 'conceptual paradigm" that most human and applied biomechanists bring to these issues.  That paradigm focuses on the assumption--perhaps a mistaken one based upon a body of contemporary research--that fast running requires forward propulsion.  One a runner is up to speed, this is simply not true.  The net requirement for forward propulsion is zero if a runner's speed does not change.   And any change in speed is fully accounted for by the interaction between the foot and the ground.

The long-held super ball analogy is that, like a runner who is up to speed, each conserves its forward momentum by bouncing.  No net input of propulsive force is required.  A close look at the Cheetah blades of the paralypians makes this process very clear.  They are not--actually cannot--be 'tuned' for anything other than 'bounce mechanics.' 

As John noted in one of his exchanges, "All a (passive) prosthesis can do is return (in an elastic manner) the energy it is able to store following foot contact.  But of course, for an amputee to benefit from this energy return, the deformation and recoil of the prosthesis needs to occur at an optimal frequency (related to contact time).  It is difficult to imagine that this could occur for all phases of the 100m sprint, i.e. both the acceleration and maximum speed phases. "

Although this is sometimes difficult to grasp, coaches should consider what would happen if we could somehow take an athlete at his or her top speed and prevent either a left or right foot from leaving the ground.  The athlete will still rotate forward over the base of support at his or her current meters-per-second.

My contention is that, at such speeds, the movement over the base of support would simply dwarf any volitional attempt to 'push forward' against the ground.  Forward speed and reduced contact time makes applying an appreciable backward force very difficult. Considering that the athlete does not have to deal with a headwind, the posture necessary for adopting the limb orientation to push backward forcefully just doesn't seem possible.  Because of balance issues at those high end speeds, the athlete would be hard pressed not to fall.

Ken Jakalski