elapid said:
Acceleration and speed are different. Acceleration is the rate at which an object changes it's speed, and is calculated according to the following equation: acceleration = metres/speed^2.
Pedal speed does not change within the cycle of a pedal stroke because pedal speed is determined by cadence and crank length (pedal speed = 2 x pie x cadence x crank length/60). As cadence is dependent on completion of a full rotation of the crank, any difference in speed of the pedal during a single rotation will not be relevant in these calculations.
Ugh, pedal speed is always changing. Fortunately, pedal speed changes so little during a normal pedal stroke that it can be considered constant for most purposes. If it were not for that change in pedal speed the computrainer spinscan would not work as it depends upon the small change in wheel speed (which is directly related) to calculate the various pedal torques as the pedal moves around the circle.
And, the relevant equation is F=ma, where F is the difference between the pedal force (trying to speed the pedals up) and the wind/rolling resistance (trying to slow the pedal down). When the sum is positive the pedal accelerates, when it is negative it decelerates. And, right in between is the average speed of both the bike and the pedal for that rotation.
Edit: The computrainer is helped in its spinscan determination by the fact that when on the trainer the moving mass is small, being only the mass of the rotating rear wheel, so there is a larger variation in speed than when on the road where we have both the rotating mass of both wheels plus the moving mass of both the bike and the rider. Speed variation is quite small when riding on a bike but it is still there.