JohnDenker
commented
9 years ago Tangential remark, possibly amusing: In my search for real-world examples, I considered the black-start scenario at an electric-company power-plant. However, as it turns out, they go to a lot of trouble to avoid a rotational collision between the turbine and the load. They have a fluid-dynamic torque converter to get the load spinning, and then they use a Synchronous Self-Shifting clutch to engage the main drive shaft at the moment when the turbine and the load are spinning at the same speed.
An SSS clutch is a nifty piece of technology.
This is relevant to the book only insofar as it reinforces my suspicion that finding real-world examples of a rotational collision is not easy.
See also next comment in this thread.
In chapter 12 on page 296, I'm pretty sure example 12.6 is bogus. I've never seen a disk drive that worked this way. I'd be astonished to see one, since it is a terrible design, likely to scuff the disk. Every disk drive has a start/stop motor; why not engage the spindle before starting the motor? I realize this is a standard "textbook" example, passed down from generation to generation ... but that doesn't make it right.
A better suggestion appears below, but let me start with a not-entirely-constructive suggestion: Here's a real-world example of a rotational collision: Consider a small helicopter such as an R22. When starting the engine, or when idling, you want the engine to be disconnected from the rotor. When you want to go flying, you engage a clutch so that the engine and the rotor rotate together. Understanding how this works is interesting, but complicated. It's not immediately obvious how to turn it into a simple exercise. One wrinkle is that the inertia of the rotor is so enormous that if you just popped clutch the engine would stall. In reality the clutch is engaged gradually over a period of 20 seconds or so, during which time engine power makes a zeroth-order contribution to the energy budget. (There are additional wrinkles that make things even more complicated.)
More generally: It seems likely that in the vast majority of real-world devices, the clutch is engaged gradually. Otherwise you're going to break things, or at least put grossly unnecessary stress on things. So the search for simple yet practical rotational collisions may be difficult.
Possibly constructive suggestion: Set up a lightweight basket, suspended from the ceiling by a long thin fiber. The basket is initially non-rotating. Drop some sort of spinning object into the basket. A toy top with handle and bearings is particularly convenient. There will be a timescale long enough to observe the rotation of the combined system ... yet short enough compared to the torsional oscillation period so that torque can be neglected. You could tout this as the rotational analog of a ballistic pendulum.
Things get messy if you toss something into the basket from the side, because then there is orbital angular momentum in addition to spin angular momentum.