Minimize Flux Distance X-Axis w/ Magnet Shield

[GPT]

Is it possible to...
Minimize the distance the flux travels in the x-axis from the centre of a magnet. (in 2d representation)

Ive attached an cross-section image detailing what I'd like to happen. A "shield" is wrapped around the magnet, leaving only the N and S poles exposed. The flux is attracted to the shield, and passes through it. Relatively little flux bypasses the shield.

Step 2: Increase the distance the flux travels along the y-axis. Would do this by increasing current (I'll actually be using a solenoid). Its my understanding that depending on the saturation of the shield, its thickness would need to be adjusted relative to the height of the flux.

So what do you all think. Is it do-able? Any suggestions?

EDIT: Would the use of a horseshoe magnet increase the flux distance along the y-axis?

 

[vk6kro]

If you used soft iron as shown in your diagram, I think you would make the field stronger horizontally. The soft iron becomes an induced magnet and makes the actual magnet appear bigger than it is.

You can increase the magnetic field by putting soft iron in the path of the magnetic lines of force and leaving only a small air gap. In this air gap, the field will get concentrated and may be a lot stronger than with the magnet alone.

This is pretty much what a horseshoe magnet does, but you can extend it to have a smaller gap at the end and then get a strong field in that gap.

 

[GPT]

Hm, well...
Problem is I cannot really put anything more than a few inches above the magnet. And the flux lines need to be quite strong even at about a foot from the top of the magnet.

 

[Bob S]

I have seen cylindrical permanent magnets with a soft iron shield like you describe, but they had a cylindrical layer of aluminum between the permanent magnet and the soft iron to prevent the shield from "shorting out" the permenent magnet. The one I saw also had a soft iron cap on one end so all of the magnetic flux was at the open end. The cross sectional area of the soft iron shield should be roughly the same as the permanent magnet.

 

[vk6kro]

You could consider putting rare Earth magnets above your magnet.

If you get about 10 like this:
http://www.dealextreme.com/search.dx/search.magnets#1105
the field would be stronger than it is now.

However, air is not a good conductor of magnetic fields and a foot is a long way in magnetic terms.

Larger Neodymium magnets are available at specialist stores and may be even stronger and less fragile than the ones above. They have an amazing ability to find each other and to find Iron objects and to move with great speed.

 

[GPT]

Anyone know of particular arrangements of magnets that would do what I am looking for? Some simulation software might help. Any recommendations?

 

[vk6kro]

Probably not.

You haven't said what you are trying to do, but using a magnet a foot away is not being realistic. Even strong Neodymium magnets can only turn a compass needle at that distance.

 

[Relay]

A few thoughts. A saturated shield is no longer a shield. The shield in your drawing will short out the magnetic flux and reduce it's height above the magnet. I don't know of any cases where the magnetic lines of force are thrown out away from the magnet as your drawing suggests. If you do find a solution please share it with the rest of us.

 

[Bob S]

For magnetic shields around photomultipliers (to keep magnetic field out) the shields are several coaxial layers of soft iron separated by radial gaps to keep inner layers from saturating. In your case, I would recommend several coaxial layers of soft iron, about 1.5 to 2 rimes as long as your magnet. The radial thickness of the inner layer should be thick enough to carry all the return flux without saturating. This would mean that the cross section of the innermost layer (thickness times circumference) is roughly the same as your magnet. The innermost layer should have an ID perhaps 1.5 to 2 times the diameter of your magnet. On the outer surface of the iron layers, remember that H parallel is c[STRIKE]onstant[/STRIKE] continuous across the iron-air boundary, so B outside (in air) is ~1000 times less than the B in the iron.

Bob S

 

[GPT]

Thanks Bob S.
Ill try to see if I can get that model going in a simulation.

Anyone have experience installing Ansoft's Maxwell SV (student version) on Linux? Specifically Ubuntu?