Tapeworm wrote:My analysis, whilst incorrect, is not totally baseless.

Care to map out for us Frank what the actual energy cost is then to move the cranks around (talking normal cranks here) with legs of different mass? We'll assume that there are no other forces other than gravity to consider.

Look, one can say that the pronouncement that the sun circles around the earth is wrong but not totally baseless because you observe the sun circling the earth every day. Wrong is wrong, regardless of how you got there and regardless of how right your observations seem.

It is pretty easy if one knows the configuration of the bike and rider and the masses of the various elements. All one need do is do the integral around an entire circle for each component and add them up. Now the integral is not that easy to do per se because we don't know the formula for the speed curve (and there can be both a linear and rotational energy component, depending upon the frame of reference. But, we can break the circle down into pieces and estimate the amount. The more pieces the more tedious the calculations but the more accurate the result. Anyhow, one will find that the only component worth worrying about is the thigh as it has the largest mass and is the part that deviates most from the circular, constant speed, motion. Know the mass of each thigh, the crank length, and the cadence and the energy variation is known. One could make the argument that the energy put into the leg could be transferred to the bike but that argument fails when pedaling unloaded. While it is possible to recover some of the energy put into the leg by transferring it to the bike when under load, it is not possible to transfer all of it because of the force directions involved (the thigh is moving up and down when it is slowing and at this time the pedal is moving mostly forward and aft). Anyhow, what is shown, if you do the analysis is this energy variation (and hence the cost of just making the pedal go around) varies with the square of the cadence. This means the power loss varies with the cube of the cadence.