Eddy current heating for the bored pickup maker

[Mike Sulzer]

wobbling disk

I picture the setup as a rotating disk where part of the disk passes between two powerful magnets, one on each side of the disk. The magnetic field points through the disk, that is, parallel to the axis of rotation. Suppose we draw a loop on the face of the disk, not one that passes around the axis but a smaller confined to one side. As the disk rotates and the loop approaches the magnets, the amount of magnetic flux passing through the loop increases, and there is a voltage around the loop. A current results and heating occurs from losses in the conductor. As the loop rotates out of the magnetic field, the flux decreases and a current is induced in the other direction, also causing heating. We can draw lots of loops, and so the whole problem is quite complicated, but this is the basic idea.

If we alter the axis of rotation so that it is no longer parallel to the field, the amount of flux passing through the loop goes down because it intercepts fewer field lines. Thus the current and the heating decrease. A wobbling disk is on average not parallel with the field, and so it is heated less than a non-wobbling disk that has its axis perfectly parallel to the field.

There are some other complicating factors for sure; again, just the basic idea.

[David Schwab]

As the disk rotates and the loop approaches the magnets, the amount of magnetic flux passing through the loop increases, and there is a voltage around the loop. A current results and heating occurs from losses in the conductor.

That was my train of thought.

[David King]

I figured the complexity of the problem would render it un-solvable but it could be tested quite easily by spinning the disk on a small, high-speed DC motor and measuring the the current passing through the motor in both circumstances. Probably just the pitch of the motor would tell us if the wobbly disc was spinning faster or slower. Slower speed implies more drag and more heat

[David Schwab]

Also if you mount the motor far enough away from the disk you should be able to tell if the disk heats up without the heat coming from the motor.

Of make a magnetic coupling device between the motor shaft and the disk so the two don't touch.

[Steve Conner]

A wobbling disk will have more drag from air resistance than one that spins true.

Mike Sulzer has a point. If the field were perfectly uniform, so that the field lines were all normal to the surface of the disk if it were true, then the heating should be decreased by wobbling.

But in practice the field won't be uniform, so wobbling should increase the eddy current losses.

[Bruce Johnson]

David S.;

I didn't mean to imply that the thermal heat in the copper disk actually comes from the motor through the shaft.....That isn't what happens. The heat in the disk is generated by the eddy currents circling around within the disk, because the copper has some resistance.

But the energy to do that comes in as mechanical energy, the torque through the shaft into the disk. The extra energy happens because the motor is trying to stay up to speed, which causes it to draw more current through its coils. That extra current also causes the motor guts to warm up. So, that extra few micro-watt-hours of power that you just paid the electric company for were burnt up in the combination of warming up the copper disk and the motor housing.

That question about whether a wobbling disk would increase or decrease the eddy current drag could be argued either way, I guess. It's kind of a moot point though. To have any decent operating efficiency, eddy current brakes usually have very tight clearances between the magnets and the disk. If there was enough clearance to allow the disk to wobble much, it wouldn't work very well in either condition.

To the readers who've stayed with us: Yes, this detailed knowledge of physics and eddy currents gives you an enormous advantage impressing women in public places! Use it!

[Mike Sulzer]

A wobbling disk will have more drag from air resistance than one that spins true.

Mike Sulzer has a point. If the field were perfectly uniform, so that the field lines were all normal to the surface of the disk if it were true, then the heating should be decreased by wobbling.

But in practice the field won't be uniform, so wobbling should increase the eddy current losses.

Outside of the space between the magnets and near the edges of this space, the field is curved. This curvature has symmetry about the axis of rotation. If you tilt the disk, you can find a region on the disk that intercepts more flux because you are aligning the face of the disk closer to perpendicularity with the field there. Located 180 degrees around the disk is a corresponding region where the same thing happens.

But you are still losing where the field is straight, and so the net effect is what? The field is weak outside of the space between the magnets. I would guess that the net effect is still a loss.

[David Schwab]

David S.;

I didn't mean to imply that the thermal heat in the copper disk actually comes from the motor through the shaft.....That isn't what happens. The heat in the disk is generated by the eddy currents circling around within the disk, because the copper has some resistance.

Right, and that's why I said friction.

But the energy to do that comes in as mechanical energy, the torque through the shaft into the disk. The extra energy happens because the motor is trying to stay up to speed, which causes it to draw more current through its coils. That extra current also causes the motor guts to warm up. So, that extra few micro-watt-hours of power that you just paid the electric company for were burnt up in the combination of warming up the copper disk and the motor housing.

Now you are talking about the motor again (see above, you wrote "That isn't what happens"), but that's indirectly warming the disk. What if you cooled the motor? What if the motor was not an electric motor? What if you were turning the disk with a hand crank? Would the disk still get warm?

Maybe instead of friction, "electron collision" is a better answer?

[Bruce Johnson]

Yes, that's it. Eddy current drag isn't really friction, but it's the electro-magnetic equivalent of friction. But, the overall result is similar. You're applying drag, slowing down the disk, and burning off the energy as heat. It's almost the same as if you were clamping on the disk with two plain steel blocks. The differences are:
1.) With an eddy current brake, only the disk heats up, not the magnets. Yeah, the magnets may warm up a little because of their close proximity, as a side effect. But there's no direct build of heat in them. With the steel blocks, the friction will build heat in both the blocks and the disk.
2.) As I said above the drag vs speed curve is hugely non-linear with the eddy current brake, slamming on at high speeds and having no effect at zero. A friction brake is also non-linear, but not as much so, and in reverse. It's strongest at zero speed and will trail off as speed goes up. Of course, that depends on the materials and heat rejection, etc.

Cooling the motor wouldn't have any effect on the operation of the eddy brake. It would just make the motor feel better and slightly reduce the current draw. The motor would still be under load, and the copper disk would still heat up.

Cranking it by hand, the copper disk would still heat up the same way. The warming of the motor housing would be replaced by the sweat on your brow. Turn it slow, and there would be no drag. Crank faster, and you'll be sweating. Try to crank it as fast as you can, and you'll find that you'll edge up to a speed where the drag gets so high that you don't have the muscle to go any faster. And the copper disk will be smokin'.

[David Schwab]

Cooling the motor wouldn't have any effect on the operation of the eddy brake. It would just make the motor feel better and slightly reduce the current draw. The motor would still be under load, and the copper disk would still heat up.

Bruce, you are making me dizzy! You wrote:

The extra energy happens because the motor is trying to stay up to speed, which causes it to draw more current through its coils. That extra current also causes the motor guts to warm up.

That has nothing to do with the disk getting warm. The disk is getting its energy from the motor, but how much current the motor is drawing has nothing to do with the disk getting warm, just with the motor getting warm. Unless we are saying that the motor is warming the disk. The motor doesn't have to be under load, it just has to be spinning the disk. Hook the disk up to a gasoline powered motor, and the disk will still heat up.

The current flowing in the disk is causing it to get warm because of resistance, which is caused by electron collision. That's the same way an electric heater works.

They also use eddy current breaking on roller coasters.

[Mike Sulzer]

Y
2.) As I said above the drag vs speed curve is hugely non-linear with the eddy current brake, slamming on at high speeds and having no effect at zero. A friction brake is also non-linear, but not as much so, and in reverse. It's strongest at zero speed and will trail off as speed goes up. Of course, that depends on the materials and heat rejection, etc.

The voltage, and thus current, that is generated depends on time rate of change of the magnetic flux. This is linear with speed and so the power dissipated increases with the square of the speed

[StarryNight]

So are eddy currents more affected by frequency or mass?

I guess an experiment would be to string a guitar with the same gauge string but each tuned differently and measure the output of each.

[Sweetfinger]

It actually ended up quite complicated. The 240V mains is rectified and filtered to feed a half bridge of ultrafast IGBTs. The RF output from this, at about 250kHz, goes through a DC blocking capacitor into a ferrite-cored transformer with a 20 turn primary and a single turn secondary.

The secondary is inserted in series with the tank circuit, which consists of the work coil and a mica capacitor rated for 200 amps of RF ($40 on Ebay, sourcing this was the hardest part of the project!)

The capacitor, work coil and transformer secondary are all water cooled, otherwise it would melt itself before melting the workpiece.

The IGBTs are driven by a phase locked loop that tracks the resonant frequency of the work circuit. I originally made that circuit for another application.

The power output is about 1kW. I tried self-oscillating circuits with MOSFETs, but I couldn't get above a few hundred watts before they died from parasitic oscillations.

A guy I knew in college around '86 was working with experimental induction furnaces for use in ore refining. After his explanation I thought of burying a giant coil underneath my parking spot. If someone else parked there, I'd flip the giant switch and melt their car. Like many of my plans for world domination, this one never came to fruition.

[Bruce Johnson]

David;

We're saying the same thing. The heat buildup inside the eddy current brake and the heat buildup inside the motor are two separate things, which happen at the same time. The eddy current brake is applying load to the motor, trying to slow it down. While doing that, it builds heat in its disk because of the resistance to the eddy currents, etc.

Meanwhile, as the load is applied to the motor, it increases its current draw in response as it tries to maintain the speed that it was designed to run at. This is due to the increased slip angle between the rotor and stator. As the current increases through the motor, the coils will also heat up somewhat due to the resistance of the copper. This heat will transfer slowly into the motor housing.

If you look at the overall energy picture, as you apply the eddy current brake, the current draw through the motor goes up, and you are consuming energy from the wall socket at a higher rate than you were before applying the brake. That increase in energy becomes two plumes of warmth, one from the motor housing, and one from the copper disk.

Is that a clearer explanation?

[David Schwab]

Is that a clearer explanation?

Sure, but the motor drawing more current still has noting to do with the disk warming.