*Sigh*
Ok guys, I worked this problem out during one of my breaks today while on campus. I changed things to SI units, since Imperial units tend to give me the heebie jeebies.
Let's call the original observer A and consider the entire problem in his frame i.e. we will call him the "stationary observer". I decided that the bullet would have a mass of 1.00 gram and the gun a 30.0 cm barrel. So, if the bullet's final velocity is 500 m/s (kind of fast, but ok), then its final kinetic energy is:
K_A = \frac{1}{2}mv^2 = \frac{1}{2}(0.00100\ \ \text{kg})(500\ \ \text{m}/\text{s}^2)^2 = 125\ \ \text{N}\cdot\text{m}
By the work-energy theorem, this change in the bullet's kinetic energy from zero is equal to the work done on the bullet while in the barrel. I'm not sure whether the explosive gases exert a constant force on the bullet or not, but I can still just calculate the average force exerted on the bullet by the gun over that distance:
F_{\text{av}} = \frac{W}{d} = \frac{125\ \ \text{N}\cdot\text{m}}{0.300\ \ \text{m}} = 417\ \ \text{N}
(416.6666666666666... rounded to the correct number of sig figs.)
Because we'll need it later on, I'm going to calculate the time needed for the bullet to traverse the chamber (again this assumes more or less constant force on the bullet, and therefore constant acceleration. But that's fine as long as I make this assumption about the gun consistently throughout the problem, right?)
d = \frac{1}{2}at^2
t = \sqrt{\frac{2d}{a}} = \sqrt{\frac{2d}{F_{av}/m}} = 0.00120\ \ \text{s}
The buddy, observer B, is flying along parallel to the bullet with velocity v_B = 500\ \ \text{m}/\text{s}. He grabs the bullet, puts it in his gun, and fires it in a straight line, still along its original path. The buddy's gun is of identical design to the first guy's. The key question is, what is the work done by the second gun on the bullet? Remember that we are doing this entire problem from the point of view of observer A, so the work done will be the average force acting on the bullet times its displacement
as seen by observer A . Can I justify this assertion? Well, I can at least show you a statement that it is true, later.
Let's call this displacement d_B. Clearly, it is not merely equal to the length of the barrel, but the length of the barrel plus the distance traversed by everything in frame B (ship, buddy, gun, bullet) during the time t:
d_B = 0.300\text{m} + v_{B}t
= 0.300\text{m} + (500\ \ \text{m}/\text{s})(0.00120\ \ \text{s})
= 0.300\text{m} + 0.600\text{m} = 0.900\text{m}
So the work done on the bullet when the buddy fires it is:
W = F_{\text{av}}d_B = (417\ \ \text{N})(0.900\ \ \text{m}) = 375\ \ \text{N}\cdot\text{m}
The total work done on the bullet is:
W = 125\ \ \text{J}\ \ +\ \ 375\ \ \text{J}\ \ =\ \ 500\ \ \text{J}
This result is
completely consistent with the observation that the final kinetic energy of the object as seen by A is:
K = \frac{1}{2}mv^2 = \frac{1}{2}(0.00100\ \ \text{kg})(1000\ \ \text{m}/\text{s}^2)^2 = 500\ \ \text{N}\cdot\text{m}
There is no missing energy.
I realize I haven't done anything new here, or stated any priniciple that wasn't already pointed out in the thread. But I just wanted to show the result quantitatively, to drive the point home. I hope that for this reason, you will consider it a useful contribution to the topic. Admittedly, the idea that the work done would be different as seen from observers in two different inertial frames didn't quite sit so well at first, until I dug out my first year textbook and saw this (in bold):
From University Physics by Young and Freedman, 10th ed pg. 170:
Because we used Newton's laws in deriving the work-energy theorem, we can use it only in an inertial frame of reference. The speeds that we use to compute the kinetic energies and the distances that we use to compute work must be measured in an inertial frame. Note also that the work-energy theorem is valid in any inertial frame, but the values of Wtotal and \Delta K may differ from one inertial frame to another (because the displacement and speed of a body may be different in different frames).
krab has shown us exactly why this is so, and what he did was much more elegant than my number-crunching. So as far as Slinkie's original question goes, CASE CLOSED. Notice, Eyesaw, that krab was responding directly to Slinkie's original post (he even quoted it!), so any whining that he did not address later questions is misplaced. Slinkie was
later driven to call into question the definition of work (and therefore KE) because these results were not intuitive to him. This was, misguided. I'm not trying to be mean or condescending in saying that. I'm just suggesting that when you come across a result in Newtonian Mechanics (an area of physics that, as pointed out, has firm foundations going back 300+ yrs) that is stated clearly in any introductory physics text, don't be inclined to immediately revamp Newtonian physics just because the result doesn't agree with your intuition! Be inclined always instead to revamp your intuition.
*phew*