Bill_K said:
Under a Lorentz transformation t = γ t0, where t is the time in an arbitrary rest frame and t0 is the proper time. So a reasonable guess to make is that E behaves the same way that time behaves, namely E = γ E0 where E0 is the energy in the rest frame.
It's better not to guess but to calculate:
The change of energy is
[itex]dE = F \cdot ds[/itex]
Newton says
[itex]F = \frac{{dp}}{{dt}}[/itex]
and
[itex]p = m \cdot v[/itex]
This leads to
[itex]dE = m \cdot v \cdot dv + v^2 \cdot dm[/itex]
Now we need an expression for the inertial mass m. If there is a dependence from velocity it should be the same for all bodies. Therefore I start with
[itex]m\left( v \right): = m_0 \cdot f\left( v \right)[/itex]
where f(v) is a function of velocity independent from the body and from the frame of reference and m0 is the mass of the body in rest. So we already know
[itex]f\left( 0 \right) = 1[/itex]
Due to isotropy f must also be symmetric
[itex]f\left( { - v} \right) = f\left( v \right)[/itex]
Now let’s assume a body A with the velocity v and a body B at rest. Both bodies should have the same rest mass m0. The product C of a perfectly inelastic collision shall have the rest mass M0 and the velocity u. The conservation of momentum leads to
[itex]p = m_0 \cdot f\left( v \right) \cdot v = M_0 \cdot f\left( u \right) \cdot u[/itex]
At this point I have to make a reasonable assumption (I will check it later): If energy is conserved the mass of C shall be the sum of the masses of A and B:
[itex]M_0 \cdot f\left( u \right) = m_0 \cdot f\left( v \right) + m_0 \cdot f\left( 0 \right)[/itex]
This results in
[itex]f\left( v \right) = \frac{u}{{v - u}}[/itex]
To get the velocity u I change to a frame of reference moving with the velocity v. Now body B moves with -v and body A is at rest. As the situation is symmetric the velocity of C is
[itex]u' = - u[/itex]
The next steps depends on the transformation. Galilei transformation
[itex]u' = - u = u - v[/itex]
leads to
[itex]f\left( v \right) = 1[/itex]
Therefore in classical mechanics inertial mass is independent from the frame of reference and the change of Energy is reduced to
[itex]dE = m \cdot v \cdot dv[/itex]
The integration leads to
[itex]E = E_0 + {\textstyle{1 \over 2}}m_0 \cdot v^2[/itex]
In SRT Galilei transformation is replaced by Lorentz transformation
[itex]u' = - u = \frac{{u - v}}{{1 - \frac{{u \cdot v}}{{c^2 }}}}[/itex]
Everything else remains unchanged. This leads to
[itex]m\left( v \right) = \frac{{m_0 }}{{\sqrt {1 - \frac{{v^2 }}{{c^2 }}} }}[/itex]
Although this is often called "relativistic mass" it is still the good old inertial mass as used in Newton’s definition of momentum. With
[itex]dm = \frac{{m_0 \cdot v \cdot dv}}{{c^2 \cdot \sqrt {1 - \frac{{v^2 }}{{c^2 }}} ^3 }} = \frac{{m \cdot v \cdot dv}}{{c^2 - v^2 }}[/itex]
The change of Energy is
[itex]dE = dm \cdot c^2[/itex]
Now Integration leads to
[itex]E = E_0 + \left( {m - m_0 } \right) \cdot c^2[/itex]
With Einstein’s equation for rest mass and rest energy we get an expression for inertial mass and total energy:
[itex]E = m \cdot c^2[/itex]
Due to the additivity of energy
[itex]E = \sum\limits_i {E_i } = \sum\limits_i {\left( {m_i \cdot c^2 } \right) = \left( {\sum\limits_i {m_i } } \right)} \cdot c^2 = m \cdot c^2[/itex]
the inertial mass is additive:
[itex]m = \sum\limits_i {m_i }[/itex]
Therefore my assumption above has been correct.