Magnetic induction; one thought experiment and questions / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Magnetic induction; one thought experiment and questions

Move Thread

LAN_403

Marko

Mon Feb 04 2008, 10:41PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Something has been bugging now in a while, and that is lack of understanding what a magnetic field is in fullest sense.

After thought-experimenting for a while I was a bit shocked.
The whole thing may be much more philosophical in nature, I'd really like to know how answerable it is.

- If I have a fully closed magnetic core with very high, but non infinite permeability, wind a primary coil around it and energize with AC;

- due to it's high permeability, I can consider that all of the magnetic flux is inside the core, and leakage flux negligible;

- Now if I take a flux density meter, and sweep around the core through all possible positions, I just won't be able to measure any magnetic flux, and would actually be unable to know whether core is powered on or not.

- Now I complete the transformer by winding a secondary coil of a diameter much larger than core's circumference, I'll see EMF induced on it, and be able to draw power, even though there's no observable alternating magnetic flux anywhere in the air space between coil and the core, it's all in the core!

- the point of the problem is, just how does the coil ''know'' that there's alternating flux somewhere around it's center, as it is obvious from the U=dfi/dt?

How is this flux affecting electrons that may be very far away from the core, and actually putting the same amount of force of them regardless how far they are?

Why does magnetic field appear to 'act at distance' like this, how does just the energy get from inside of the core to the conductor that I can put meters away?

If you like practical example, take one of those big 50Hz cored induction heaters - you could measure only weak leakage flux that carries very little energy, yet a wide pool around the core can receive massive amounts of power from it.

Discuss,

Marko

Proud Mary

Tue Feb 05 2008, 12:24AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Perhaps there is no magnetic field as a thing-in-itself, Marko, but only a very particular kind of deformity in space-time around the magnet.

Electroholic

Tue Feb 05 2008, 01:33AM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

Maybe im just not getting the picture, but if the core is fully enclosed, and leakage is 0, how does the external coil couple to the one thats inside the core.

Dr. Slack

Tue Feb 05 2008, 08:16AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Because when you take a flux density meter and sweep it around the core, you are only exploring the magnetic potential in a small region.

You can only measure the voltage induced in a circuit, which means a closed loop. In more mathematical terms, you do a line integral along a closed path. If the path does not thread the core, then the integral is always zero, from the topology of the magnetic potential (not the B field, which is zero). If the path passed through the core once, then the integral is non-zero, and proportional to the amount of flux enclosed. If the path passes through the core 500 times (say it's the mains winding), then the integral is 500x that for a single threading. Hey presto - the transformer turns ratio effect on voltage is revealed.

I suppose that just shifts the problem to one of philosophy or topology, why does the wire care about the threading number in how it interacts with the magnetic potential? However, this sort of topological thing turns up all over physics, there is one effect where electrons end up with a different phase shift depending on whether they pass to the left or the right of a particle with spin.

Another way to put it, yet again this describes what happens rather than being an explanation, is that the electric potential field has a source, a divergence (that is a surface integral enclosing the source is finite and zero if not enclosing), and has zero curl (line integrals around closed loops are always zero), but magnetic fields have zero div (always zero surface integral on a closed surface) and finite curl when the integration is done round a pole.

You may take comfort from the fact that magnetism becomes symmetric with electricty when treated relatavistically, or maybe that's not such a help after all.

Marko

Tue Feb 05 2008, 09:42AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Maybe im just not getting the picture, but if the core is fully enclosed, and leakage is 0, how does the external coil couple to the one thats inside the core.

By saying core is enclosed I just exclaimed that it has no air gaps, think of it as a toroidal core if you want.

A most normal, idealized transformer except I tried to explain step by step what's going on.

I suppose that just shifts the problem to one of philosophy or topology, why does the wire care about the threading number in how it interacts with the magnetic potential?

I simply understood this that forces acting on electrons will sum up on all turns for same flux.
Less primary turns will give more current and more flux for same input voltage and thus higher secondary voltage.

The question is rather too simple, how does the energy get from the magnetic field which is *all* inside the core to conductors around it?

How is the force carried to electrons without any apparent mediator?

I don't know if this is just too fundamental to be answered, If it is, I'd be glad to add it to list of unsolved problems.

Coronafix

Tue Feb 05 2008, 11:37AM

Registered Member #160 Joined: Mon Feb 13 2006, 02:07AM
Location: Melbourne, Australia
Posts: 938

The magnetic field will also be on the outside of the primary, only the flux density will be rather
small as compared to inside the core.
Think about the magnetic field around a single wire, now by winding a solenoid it adds up on the inside of the
coil then returns on the outside. (typical bar magnet picture)
Now by turning that solenoid into a toroid, the field is internalized but there is still a field on the outside,
albeit small, and this can be measured by the flux density meter.
I believe this would be the case regardless of how high the permeability of the core used.
It doesn't make sense to have the magnetic field only on one side of the winding but not on the outside.

Dr. Slack

Wed Feb 06 2008, 08:32AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Goldsphere - no. For a completely uniformly wound toroid, or an infinitely permeable core, the external field is zero. Even for real situations, the external field is asymptotically zero, that is becomes zero in the limit. Do a thought experiement with a coil on a core of permeability 1000, then replace it with a 10000 ur core. For the same coil geometry, the feeble external field will drop another factor of 10. However, the effect of induction into a turn threading the core will be unchanged, or maybe enhanced by a factor of (1-(9/10000))^-1, that is something very close to 1. For the external field to drop by a large factor, and the induction to remain essentially the same, shows that the external B field is not the cause of the induction.

There's another thread on antennae in which SteveConner reminded us that the energy was not in the wires but in the air, I am sure that a solution to one conundrum would be a solution to the other.

The problem remains "how does a loop of wire 'know' how much flux there is contained within the loop?" Remember that the electromagnetic force is carried by photons. Quite what sort of little dance they do between elements of the problem I'm not sure, but it's not a "totally mysterious instant action at a distance", but a "still bloody hard to visualise light-speed relatavistic photon-exchange mediated" action.

<edit>

Another way to look at it is to start from the Biot-Savat law. Do another thought experiement where your toroid is air-cored, and is uniformly, but not densely, wound. Take a large one-turn search coil, and start a long way from the toroid. Compute the mutual inductance between the two windings, by integrating along each winding with respect to the field of the other. As you bring the search coil up to the toroid, the inverse square law conspires with the geometry of the toroid such that all the elemental little bits of coupling add up to zero once we have gone round an entire circuit, there's a net zero mutual inductance between the coils. Now start to push the search coil in between the turns of the toroid. For the parts of the search coil that penetrate between the windings on the toroid, the vector distance between the wires flips 180, (this is the vector that is crossed with the current element) and so their contribution to the integral changes sign, and now the coupling is non-zero. Keep pushing the search coil in, and eventually it will thread the toroid as a simple winding. Once this happens, the mutual inductance will stop changing again.

This still only shifts the strangeness, not eliminates it.

</edit>

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob