Zeit said:
Hello,
I don't understand why I should be wrong. Rest mass, as its name suggests me, cannot increase : it is an invariant mass. The "relativistic mass" is, as I understand it, a wrong concept ; we should talk of "relativistic momentum". So, may be is it better to use E² = (mc²)²+(pc)² than E = gamma*mc², where m is the invariant mass.
But, what I see is even when an object as a kinetic energy = 0, these equations tell us there's energy. So, if I'm wrong, tell me please
Here is the fine point behind what Pete was saying, in terms of invariant mass (which is both modern and has a totally unambiguous defintion).
The invariant mass of a point mass is Lorentz invariant - it does not change when you change the frame of reference.
However, if you have a system with a non-zero volume, the invariant mass of that system is only invariant under a Lorentz boost if the system is an isolated system.
For one reference, see the paper I quoted earlier.
A vector like the energy-momentum 4-vector is covariant if and only if the norm of that vector, i.e. the mass for the energy-momentum 4-vector, is invariant under a Lorentz boost.
A system with a non-zero volume has an energy-momentum 4-vector, but that vector is not covariant when the system is not isolated. So the name "invariant mass" could be slightly misleading for distributed systems with non-zero volume. The system has a mass, calculated in the usual manner, but that number is only independent of the frame of reference used when the system is isolated, or when the system has zero volume.
If you work at it, you can eventually see why this is so due to the relativity of simultaneity. A non-isolated system is constantly exchanging energy with its surroudnings. The total amount of energy "in the system" depends on the defintion of simultaneity used in the case where the system is interacting with its surroundings. For an isolated system, this isn't a problem, the energy is constant and the energy-momentum 4-vector is covariant.
What is ALWAYS covariant is the stress-energy tensor.
Relativistic mass is wildly coordinate dependent. Invariant mass is better, but it still is coordinate/frame dependent when you have a non-isolated system. The stress-energy tensor is ALWAYS covariant and thus independent of the coordinates used. This , in my opinion, is an excellent reason to stay away from the concept of mass, and (as Einstein did) to stick with the stress-energy tensor as the fundamental description of distributed systems, i.e. systems with a non-zero volume.