Radially oriented magnetic field in a sphere

[Mildman]

Hello everyone,

I don't know much about magnetism but I'm actually doing some research to realize a small experiment involving magnetic levitation / "confinement"

My project is to build a sort of cage of spherical shape consisting of numerous magnets all having the same pole oriented toward the center of the sphere, pretty much like a geodesic dome, and inside of this "cage", put a spherical magnet with the same pole as the other magnets all over its external surface. As a result, if the magnetic force is strong enough to overcome gravity, I expect to see the internal sphere floating more or less around the center of the "cage". (I would like to specify that all the magnets involved would be permanent, not electromagnets)

Now, I came here to ask some advice about the internal sphere whose magnetic field must be radially oriented in order for my design be successful. I know that it is feasible to manufacture magnetic rings that way but I have been unable to find information about spherical designs.

So to summarize, here are my two questions:

1- is a magnetic sphere with a radially oriented magnetic field physically possible?

2- If yes, how can one manufacture such a ting?

Any ideas / suggestions / comments are very welcome.
Thanks in advance.

 

[electricpete]

Good luck coming up with a center magnet that has only 1 pole! (It has never been done in this world). With realistic 2-poles in the center magnet, one of them will be attracted to the outer hollow sphere.

=====================================
(2B)+(2B)' ?

 

[electricpete]

If you like the design, you might try doing it with dc electric charge instead of magnetism. In that case you can isolate a single pole or charge. If your center element is very very lightl, I imagine you might be able to levitate it that way. But not with magnetism. See Earnshaw's theorem which says something like you will need some additional form of restraint to perform magnetic levitation...can't do it with magnetic force alone.

=====================================
(2B)+(2B)' ?

 

[IRstuff]

Is this for school?

TTFN

FAQ731-376

 

[Mildman]

Thanks for your answers!
In fact just after I posted this message I realized that the inner sphere should be as a magnetic monopole which I know is impossible.

It's a brilliant idea to use DC charges instead of magnets. I Might give it a try.

In fact, I am trying to achieve a design which allow a small object to levitate and to NOT touch the support / container in ANY cases, which include movement and shock (acceleration => force applied on the support / container)

Of course I realize that if I put a very strong magnet near the "cage", the inner sphere will be attracted / repelled by it and will eventually go away but I can live with that.


On another subject, I read and studied a little more than i knew about Earnshaw's theorem and I realized that it doesn't say anything about diamagnetic elements is it?
because if you put a small magnet between two diamagnetic plates and then you put a second, more powerful magnet on top of the assembly, the small magnet will levitate, which violate Earnshaw's theorem since all the elements involved are static. (there are hundred videos on youtube)

What if the inner sphere was made of bismuth and small enough to weigh only a few gram? then it would be repelled by all the magnets on the "cage" and pushed by the center no?

IRstuff: to answer your question, no it's not for school, it's jut for fun (and a little bit to show off!)

 

[IRstuff]

OK, just bear in mind that this site is for work-related postings, and your posting may be pulled at any time.

TTFN

FAQ731-376

 

[Mildman]

Oh ok sorry I didn't mean to disturb. I was just asking for advices since the community seems very capable.

Again I'm sorry and thanks anyway.

 

[MagnetDave]

This is a popular test question in our field. I have actually encountered a few customers who have wanted designs based on magnetic spheres.

At the end of the day, Gauss's Law prevents any object like this from working, regardless of how clever you are with diamagnets (even very weak diamagnets like bismuth or very effective ones like superconducting material).

Since you are doing this for school, I strongly, strongly recommend studying Gauss's Law in its integral form to understand what you are up against. In particular, use a spherical surface. Remember - what goes out must come in.

(To prove this point, we once even made a cube like this, since it's easier to fabricate. When the last magnet goes into place, you get a dead device. No flux anywhere, won't even hold a paper clip.)

 

[Mildman]

Thank you for your answer MagnetDave.
However I don't understand it really well... I never meant to design the "cage" of magnet as a closed structure, but more like a geodesic dome formed with pentagons and hexagons and magnets fixed on their center (with the same pole toward the center for each magnet). In that case, Gauss's law is not violated and the flux can circulate from the inner side of the cage to the outer side and at the same time, repell a piece of bismuth in the inside assuming the magnets are powerful enough.

... Or am I missing something?

 

[MagnetDave]

One of us is.

What path is flux leaving through?

 

[Compositepro]

The net field will be zero. This will aswer all your questions:

http://ocw.mit.edu/courses/physics/8-02-electricity-and-magnetism-spring-2002/video-lectures/