It's like the classical problem in motion in a straight line: If you have two masses undergoing an elastic collision, find a relationship between the velocities. And you do that by playing around with two simultaenous equations that you get from conservation of momentum and energy.
In this case:
Initial Kinetic Energy of cube=Final kinetic energy of cube + Final Rotational energy of rod. (1)
Initial Momentum of cube=Final Momentum of Cube + Final Momentum of Rod (2)
Equation (2) is a bit funny, because how can you equate linear and angular momentum?
The easiest way is to consider the collision that happens just as the cube is at the rod. It's path is tangential to the rod and so it's kind of like it is rotating with radius d/2, so has angular momentum of
, where
. The initial angular momentum of the rod is 0. The Final angular momentum of cube is
and the final angular momentum of the cube is
.
The thing I have in bold I don't like, but Knight actually uses it if you look at example12.23 on page 374. (In fact this info helps you with question 94 a bit more, but it will help u on this one too).
If you don't like the thing in bold just like I don't myself, I can provide an alternative using Newton's THird law which I find more rigorous (but still gives the same answer but only takes up more time).
Edit: Okay so i did the question now: Once you get those equations set up you cancel out the m and d in the momentum equation. Cancel out the m in the energy equation. The rest of the algebra should be handeled well.
2nd Edit: d/2 is radius