Liquid Metal Eddy Current Generator

Why?

Just pondering, trying to understand if the magnitude of the eddy current effect is useful other than servo control. I admit fully and completely my general ignorance and lack of an intuitive understanding of electro-magnetics.

Faraday invented something similar - the homopolar generator.

https://en.wikipedia.org/wiki/Homopolar_generator

There's also a magnetohydrodynamic generator:

https://en.wikipedia.org/wiki/Magnetohydrodynamic_generator

Yup. That is sort of where I was going with this. I am wondering/pondering if circulatory motion is induced in the liquid metal (in addition to the rotational motion), will that add to, have no effect, or reduce eddy current extracted?

Well, for a generator you have to move a conductor at right angles to the magnetic field, and then the induced voltage will be at right angles to both the direction of the field and direction of motion (the right hand rule if I remember correctly).

In MHD, your conducting fluid is constrained to move in one direction. It has one degree of freedom. When you start drawing current, there will be a counter-force (back pressure) on the conductive fluid, meaning that it will take work (force x displacement) to move the fluid through the field. This corresponds to the electrical energy gotten out (volts x amps x time).

So back to your question, as a general rule, the more degrees of freedom you give the conductive fluid, the less likely it will flow in the direction to generate power. I'm thinking that when you start drawing current, the forces on the fluid (back pressure) will most likely cause it to flow in a direction to avoid generating power.

Yep, that seems about right. The back emf, will cause fluid motion contrary to the E field inducing motion, especially since liquid metal is not constrained in this case by mechanical pressure to flow in one dimension.

An interesting question.

So many factors come into play, there isn't a simple worthwhile definitive answer. A direct comparison between a solid conductive rotor and a liquid metal filled conductive rotor would be difficult as an apparatus optimized for either would be very different than one optimized for the other.

.

Consider a homopolar generator with a solid conductive rotor. A uniform magnetic field encompassing the entire rotor would work well. However with a liquid metal filled rotor, a uniform field would not be beneficial because it would not allow as easy return path for the liquid flow.

So that's why I can't get a head in this world. The real reason for this: get a suitable DC output (maybe from a homopolar generator instead) from a machine with pretty high rotation speed up to 20,000 rpm.

I was trying to make an eddy current damper for a mechanical scale I made, but I think the field area needs to be intense, close proximity to the conductor, and small motion of the conductor needs to make a large change in transected field area. I think the one I made was backwards - large field area (steel plates with N48 magnets on them), and a largish aluminum plate, although much smaller than the field plates. Oscillation (as a pendulum) did not move the conductor out of the magnetic field, which was not as proximate as it likely should have been.

Like a transformer, the more windings a generator has, the higher the voltage. A homopolar generator has one winding (or 1/2, I don't know). As such it generates current at very low voltage. This is usually not very useful as power loss in a circuit is proportional to current squared. So, unless you are making one for the fun of it, it's not terribly useful.

If you want a damper, you want some strong NIB magnets as part of a magnetic circuit. (The one shown below uses an electromagnet instead of permanent magnets.) The flux will be concentrated in the air gap where your conductive disk should swing.

Silver or copper might work better as a damping disk, the thicker the better.

https://en.wikipedia.org/wiki/Magnetic_circuit

Hey thanks for the advice Rixter! What a mensch.

If low voltage at high current is desired, such as high current electrolysis (just an example), what about having multiple disks on the armature (insulated from each other), and then put the potentials in series until sufficient polarization voltage is achieved?

I am sure there are probably far more efficient DC generators already out there, but it might still be fun to build one of these for a hoot.

There's no reason you can't wire several disks in series if they're insulated from each other. The big advantage of a homopolar generator is it has very low internal resistance, so large currents may be generated.

I think you're right, it might be fun to build one of these to play around with.

https://en.wikipedia.org/wiki/Homopolar_generator

"....There's no reason you can't wire several disks in series if they're insulated from each other...."

You will have to have slip rings or liquid metal contacts for each disc series wirings...the wiring used to connect in series can't spin with the discs.

Does not each generator of this class comprise a source, cannot sources be connected in series to make a larger emf? Yes there are slip rings already required as a part and parcel of each one.

Sure, generators can be connected in series. The description of wiring discs in series seemed to me to be missing a crutial concept....but I wasn't trying to be pedantic. There have been lots of complex attempts at creating multi-winding homopolar generators, but nothing elegant, that I know of.

I guess I need to build my first one, and spin it really fast using my Tesla machine.

I still have a few kinks for work out as far as air leaks on the power section, and I have yet to build my two-stage compressor design. Heat source will have to be limited to pretty low temperatures, probably no more than 300 C. Getting 20,000 rpm out of a machine with silicon nitride roller skate bearings is basically the target.

Now, when you think you really understand it, you find something that bends your mind...

https://en.wikipedia.org/wiki/Faraday_paradox

"Blown away"! Very elegant explanation of the paradoxical behavior Faraday noted, but also one that make very clear that no work = no emf, no current.

Work (imaginary + real components) is required to get power out and VARS. Quite a lot less with VARS, but considerable work still required. Otherwise we would be turning synchronous condensers by bicycle pedal power.

Is this due to incomplete rotation into the complex plane? Or just a matter of realizable levels of over excitation for a standard three phase generator?

I need to get busy making my very own set of homopolar generators NOW.

Voltage in a homopolar generator is generally low, but current can be very high.

Spinning a good conductor (even without contacts to remove current) in a strong non-uniform magnetic field, or even a uniform field that isn't aligned just right, can induce large currents within the conductor, i.e. introducing a magnetic field may make it much more difficult to turn as rapidly and cause the rotor to heat up quickly.

Drop a rare earth magnet down a copper tube, or watch the video on YouTube of the guys dropping the aluminum place in the MRI. Things will have to be just right to allow damping without introducing problems with high rpms.

I've used the magnet falling through a copper tube many times as a demonstration of eddy currents, but I really prefer a heavy-walled aluminum U extrusion (tilted slightly away from vertical), since you can watch the magnet as it falls. Ideally the magnet would have a very small clearance between the walls, so it couldn't tumble while it falls.

It even works reasonably well in a heavy-walled aluminum 90° angle extrusion.

Years ago, I had a Bogen turntable where the diameters of the motor shaft and turntable were made such that the turntable ran just a bit faster than standard speed. The motor had an aluminum disk attached to the shaft, so it spun with the motor. A lever had a U-shaped magnet arranged so that moving the lever could bring the magnet in to straddle the rim of the disk, or out away from the disk.

Using a stroboscopic strip on the turntable, you could then adjust the lever to bring the magnet in to the correct distance so the eddy currents provided enough resistance to slow the motor down to the correct speed for that particular record, pickup cartridge friction, etc.

I suspect that the eddy currents generated in the liquid metal would simply stop the liquid from rotating with the cylinder, unless there were specific paths constraining the liquid. I can't easily picture how it would be possible to connect the path on one side of the cylinder to the other...

I was sort of thinking a coaxial wire, and one strip of conductor in contact with liquid metal inside the rotating cylinder, then make contact by use of sliding contacts. Probably the whole idea is a non-starter.

Randomly flowing eddy currents within your liquid core rotor will cause almost ALL of the energy be lost as heat. That is why all motors and generator cores are laminated, so each piece acts as an independent magnet which magnetic fields add to each other buliding a good magnetic flux, either to cause motion or transform it into electricity. Yours would be just a good heater.

How do you know (I am not so sure about this) that the eddy currents will be predominantly random in path? In understand about this in a non-laminated transformer.

I am thinking that induced physical flow of the liquid metal may serve to produce a favorable current path, but I have not worked out the mathematics at this point.