jbriggs444 said:
Where is that free body diagram? I did not ask for the work of the treadmill on the body. I asked for the work of the body on the treadmill.
Edit: I am interested in the real work done at the foot-belt interface rather than "center-of-mass" work done on the treadmill as a whole. Neither the runner nor the treadmill are rigid bodies. Attempting to apply conservation-of-energy arguments based on center-of-mass work will lead to erroneous conclusions -- this thread being a case in point.
We have:
a) a theoretical argument that shows that the work done on a treadmill is the same as on a real hill.
b) the empirical evidence that this is not the case.
Edit: I'll leave this post in, but I don't think we need any such complicated explanation. See my later posts about gravity providing the force in the case of a treadmill, but not in the case of moving uphill.
I think I have the explanation. It's certainly more complicated than it first appeared. When you walk (on the flat or uphill) you tend to bounce up. On a inclined treadmill, it's clear that the force is not constant. You give a very definite push with each foot for a short time and then ease off. This raises your centre of gravity for a short time. The push is vertical, rather than tangential to the treadmill, but it's probably a bit of both. The greater the vertical force, the longer the weight is off your feet and the longer you do not need to apply a force against the treadmill.
By this mechanism, you avoid having to apply the tangential force against the treadmill for all the distance.
On the flat or at slow speeds, this bouncing would be inefficient. But, as the treadmill speed increases, it becomes more economical to adopt this bouncing gait. Certainly, this is what happens with me as the treadmill speeds up. It's a very definite sequence of short sharp pushes with each foot.
It's too late today to look at this, but I'll try to do some basic calculations tomorrow. It's not clear to me whether you do save energy this way, but the evidence of walking uphill suggests it must be so.