Doing a quick re-arranging of some equations, I come up with the rotational energy E = 1/4 * m * r2 * w2 where m is the mass of the spindle armature, r is the radius of the spindle armature and w is the angular velocity which is about the rpm * 0.104. To work out the mass of the spindle armature it's the density (say about 5,000kg/m3) * pi * r2 * length. Alternatively I suspect there ought to be an empirical way of doing it....Measure the run-up time at full power (36V), and the current it takes at start up: now I guess slowing down should be pretty well the reverse of this. So say it takes 3 sec to get to full speed at 36V and is drawing 4 amps, a resistor of 36/4 = 9 ohms ought to slow it down in roughly 3 sec. The DC resistance of the motor will influence this, so you might want to measure that as well, but if it's considerably less than the value you come up with then it obviously won't have a huge effect.