Q & A: self-reversing dynamos

Motors mewchanically and electrically connected to generators and vice versa I am trying to understand what happens if we connect an elctric motor to a generator. Both are identical machines. Suppose I have a dynamo (shunt or series field) and a motor identical to it. We couple one to the other electrically and mechanically and feed electrical power to one of the field windings. Will the rotation spaontaneously REVERSE at irregular intervals (like the earth's magnetic field reverses every 10,000 years or so)
- john (age 70)
Surrey, UK

The real motor-generator combination you mention would actually run down, because the initial energy would gradually get dissipated as heat generated in the wires, via friction, etc. If we ignore that, I don't see any reason why it would or even could reverse.

The earth's magnetic field does reverse every 150,000 years or so on the average, with very irregular timing. However, it's a much more complicated system, more like the swirling patterns formed in a pot of water on the stove than like a few organized coils. Part of the reversal process involves the current patterns breaking up, changing directions gradually, etc. The motors you described don't have big streams of currents that can break up, or any continuous range of directions to drift around gradually.

Mike W.

(published on 05/16/2011)

Thank you for your reply. Yes you are right Mike! I am trying to see that a magnetic field COULD spontaneously reverse if a dynamo and motor shared the same magnetic field (with losses replenished by steady power supply, mechanical or electrical) I want to thus see it MIGHT happen for the earth's field without wrestling with the magnetohydrodynsmics of unknown liquid and plastic phases of iron! I have a waterwheel that spontaneously reverses when fed water at the right steady rate. And of course a pendulum, rotating over the top, will spomtaneously reverse if pushed in some steady repetitive way. The circulation in thermal convection cells also spontaneously reverses when heated from the bottom at the right steady rate. Most non-linear systems are like that. The most likely way to get field reversal or direction reversals is to connect two series-wound machines in series. Such a machine, as generator, is unstable on its own and as motor would be but for windage losses. But I agree it takes considerable imagination to see quite how to couple them!
- john (age 70)
Surrey, UK

John- Yes, your question is reminiscent of pendulum problems, where chaotic behavior can be found despite the simplicity of the  system. I'm a little surprised that something as simple as two lossless dynamos hooked together could show such interesting behavior unless it were tuned to some special initial amplitude. But I guess you're describing something with losses but a steady energy input, and maybe that wouldn't have to be fine-tuned. Maybe one of our local dynamical systems types, who understand such things much better, can help out on this.

Mike W.

(published on 05/18/2011)

Many thanks Mike. Do you know you are the first expert in 50 years who has even understood what I am on about in this question. I am a physicist trying to see HOW something COULD happen, before figuring out all the details. Dynamical Systems used to be called servomechanisms and there chaos was well known to practical folk LONG BEFORE anyone was allowed to mention it! It is similar to hydraulics vs hydrodynamics, where one is what does happen and the other "what ought to happen" as limited by our education. Whenever you have two coupled resonant things, energy surges back and forth between them (giving the so-called modes of vibration of linesr theory). Real cases are always nonlinear. Thank you for unserstanding and helping.
- john (age 70)
surrey UK

John- Thanks for the background. I'd heard that some work on chaotic systems (e.g. in astronomy) long pre-dated the modern computer-assisted wave of interest. I hadn't known that the servomechanism field was familiar with chaotic behavior.

Mike W.

(published on 05/18/2011)