Mehran- Great question!
Actually, magnetism is ENTIRELY a relativistic effect. Some nice textbooks (like the one by Purcell) actually introduce magnetism by showing that it’s what’s needed to make the physical behavior of electricity look the same in all the different reference frames.
The force between the two electrons has both an electrical and a magnetic component in the frames which say the electrons are moving. The sum of these two forces isn’t exactly the same as the simple electrical force in the frame where they’re standing still. That’s because ALL forces change in different reference frames. The net force transforms just like any other force. In the case you specify, the total force (electrical plus magnetic), the sum is still a repulsive force, in all frames of reference.
You might find the following example amusing. Say that you have a stationary electron near a wire in which positive and negative charges with equal density are flowing opposite directions at equal speeds. Obviously, there’s no force between the electron and the wire, either electrical or magnetic. Now set the electron in motion parallel to the wire. We say there’s a force (say in the attractive direction) due to magnetism. What happens in a frame in which the electron is at rest, and cannot feel a magnetic force? The different Lorentz contractions of the positive and negative charges in the wire give them different densities in that frame. So in that frame there’s an attractive ELECTRICAL force. Whether you say the force is electrical or magnetic or some combination depends on which frame you use.
(republished on 08/02/06)
(published on 08/02/06)
That's a nice question. Remember that the stationary charge and the lab are using the same reference frame. So the stationary charge won't see a net charge on the nearby wire unless the lab also sees it. Generally, it won't.
Why not? One doesn't start with a rigid chain of electrons, then set them in motion getting a simple Lorentz-Fitzgerald contraction. The electrons are flowing about, and free to adjust their spacing. In the lab frame, easy to work with here, you can see that the electrostatic energy is minimized by having the electron charges just cancel the positive charges in the wire. So if we ignore minimizing the magnetic part of the energy, it remains neutral. What about the magnetic part of the energy? There's still the constraint that the overall circuit is neutral. So to the extent that the circuit is a symmetrical loop, the constraint of neutrality applies locally, regardless of the magnetic field energy. I guess that if there are kinks in the circuit loop where the effect of the magnetic Lorentz force in helping electrons turn the corners varies, you can get slight charges in the lab frame, still adding up to zero around the loop.
What about in the frame of the electrons in our little region of the wire, or to be more precise in the frame of a non-accelerating observer travelling along with them? This observer will see them stretched out and the positive charges contracted, compared to the lab values for the densities. So they'll see that region as positive, making an electrical force on that other charged particle. Of course, they also see that particle as being in motion, and thus experiencing a magnetic force as well. The two forces must exactly cancel.
(published on 09/28/13)
I'm not surprised that this question has shown up on different sites. Even if it turns out to have come from different people, it's a pretty natural thing for people to worry about.
It sounds as if you're attempting to use a rotating frame in which all the electrons are at rest rather than an inertial frame moving along with one little segment of the electrons. Rotating frames have centrifugal forces, coriolis forces and various general relativistic effects. I'm too lazy to work out the resolution of the problem in that sort of frame.
In the inertial frame, the explanation is simple, at least to lowest order in v/c. The electrons are stretched on one side of the loop and shrunk on the other side, compared to the positive charges. So that gives an electric field, whose effects on the test charge just cancel those of the magnetic Lorentz force. The ellipticity of the loop due to Lorentz-Fitzgerald contraction along one axis in the "moving" frame may play a role in cancelling terms of higher-order in v/c.
In the rotating frame I guess you need some net force on the test particle to cancel the pseudo-forces. For a symmetrical loop (rotation at least preserves that symmetry, unlike in our "moving" inertial frames) that force won't be electrical, so I suppose it's magnetic.
(published on 09/30/13)