OK, so a 180 degree phase shift would be too much in the "short" coils at the ends of the main coils I assume. I'm obviously missing how the bucking flux at the ends can increase the overall main coil flux. The bucking current (70A in my FEMM run) produces an external magnetic field that couples to the other coil system. And there it induces a voltage that aids the current flow in that main coil. The same applies to that other coil, it's external field couples back to the first coil to induce aiding voltage there. Does that make sense? I think I will attempt to simulate the device with GC models which would allow long solenoids to be made and analyzed with various core materials. The shorted end turns could then be modeled with varying currents, phases, and couplings. The problem is our lack of models for solving the magnetic domain dynamics. I think it is possible to use classical electrical domain Spice models to do magnetic domain analysis, where voltage models mmf and current models flux. For example my C cores with a shorted turn in the middle of the C could be modeled as in the attached image. Each R core is the reluctance of a small section of the core and each Rair is the reluctance of the air path joining them. R gap is of course the reluctance of the air gap. Each Ni is the ampere turns associated with each R core section, the sum of all the Ni being the input ampere turns. L m is the "magnetic inductance" (obeying mmf = -L m*dPhi/dt) of the shorted turn and has a value of 1/R short where R short is the resistance of that short. It should be possible to model both C cores and then introduce additional coupling to represent the external magnetic coupling that I allude to above. Smudge
|