Yet another, free, all magnetic idea.

[gravityblock]

For those who may be interested, the ring magnets I posted earlier came from Wal-mart, in the automotive section.  It is packaged as a 6" magnetic parts tray with a rubber case to protect the magnet and to prevent marring of surfaces, for less than $6.00 each.  I'll post pictures of what the packaging looks like.

Broli,  I am going to try the idea you suggested about the roof flashing and the hole saw.  I found a set of hole saws I had previously bought.  I need to obtain the roof flashing or other similar material before I can proceed with this experiment.

I will not make the deadline of the 8th for testing.  I made the deadline for myself, so I would not procrastinate.

[Low-Q]

Oke I managed to perfect the circular idea I started at that other thread.

The most important part of this whole idea is the tube shielding on the inside of the ring magnets. I also rendered two versions for the moving magnets, one shielded and the other unshielded. I was not sure which one works better so I added both. But I highly speculate the unshielded one would work better.

Then the last rendition shows the ring magnets replaced by conductors. This is merely to prove the concept. Because it's hell of a lot simpler to just wind a bunch of wire than go buy some big magnets. Also I'm not very sure if the results would be better if the wire windings were further to the OD (outer diameter) of the shield or if it was at the center between the OD and ID. Again I speculate that when wound nearer to the OD it might give more asymmetry and thus better results. But just to be on the safe side I positioned it in the middle.

You are right about the forces in these drawings when the magnets are in the positions shown. In that very moment, there is a force in one direction.

If you push the magnets further in that direction, the rail magnet that is closing the ringmagnet, will feel a stronger and stronger attraction. My intuition says that when that closing magnet has became almost right in the middle of the ringmagnet, it will start to feel a counterforce that will prevent it to continue. It will find its equilibrium somewhere insde the ringmagnet

Regarding the shielding:
These will reduce the counterforce, but the shield does not distinguish between force and counterforce, so the force that is attracting and repelling the magnets on each side is also weaker accordingly.

I can make a simple demonstration in FEMM if you like. And we can look at the forces at any given time and distance, add them up, and see if there is a force greater than the counterforce.

EDIT: I read your later post, so I must test with unshielded moving magnets as well.

Vidar

[gravityblock]

You are right about the forces in these drawings when the magnets are in the positions shown. In that very moment, there is a force in one direction.

This is my finding on the forces.  Both sides of the OD has the same polarity and both sides of the ID has the opposite polarity of the OD.

If you push the magnets further in that direction, the rail magnet that is closing the ringmagnet, will feel a stronger and stronger attraction. My intuition says that when that closing magnet has became almost right in the middle of the ringmagnet, it will start to feel a counterforce that will prevent it to continue. It will find its equilibrium somewhere insde the ringmagnet

I agree with you Vidar.  In the middle of the ring magnet, there is a counterforce that prevents it to continue, since the force on both sides of the ID of the ring magnet are pointing towards the middle of the ring magnet.

Regarding the shielding:
These will reduce the counterforce, but the shield does not distinguish between force and counterforce, so the force that is attracting and repelling the magnets on each side is also weaker accordingly.

Eliminating the counterforce with the shielding will also eliminate the force of the magnets.  Can't have one without the other. 

The only way I think this can be done, is to use a Halbach Array ring magnet.  This type of ring magnet will have only one force coming from the same side of the ID and pointing in the same direction across the entire inside of the ring magnet with no opposition force from the other side.  With this type of setup, the front of the rail magnet will be attracted through the ring magnet and the back of the rail magnet will be pushed through the ring magnet with no counterforce.

[Low-Q]

The problem with the Halbach array is that the magnetic field is focused in the arrays length, not width. So a ringmagnet with this array would probably not "see" the railmagnet as long that magnet isn't inside the ring in the magnetic field. The good thing however with Halbach array ringmagnet is that the magnetic flux is focused where it suppose to be when we use that array in a DC-motor. Most possible flux is present where it is desired to have it, and not spread out in the surroundings.

The other thing with the Halbach array is that the poles are shifting along the array - shouldnt be a problem if the railmagnet was configured likewise. However, the railmagnet must at some time and place rotate a bit in order to shift from attracting to repelling mode. This action must be done right in the middle of the highest flux density to achieve maximum attraction and repell moments. This operation requires great force. In fact so much force it will equalize all attractive and repelling forces in front an behind the Halbach array ring magnet. So the system will halt.

what is possible to test however, is to make a perfect counterforce for the rotation inside the Halbach array ring. This system can stay stationary and perfectly balanced, so each time the railmagnet is going in the "hole" to rotate enough to change polarity, this stationary counterforce would therfor allow the railmagnet to shift polarity without cost.

To be honest, I wonder how that works in practice. I have the idea in my head, so I might make a drawing of it on the paper, to easier see how it work - if it work.

Well, allways sceptic, but still open minded

Vidar

[gravityblock]

The problem with the Halbach array is that the magnetic field is focused in the arrays length, not width. So a ringmagnet with this array would probably not "see" the railmagnet as long that magnet isn't inside the ring in the magnetic field. The good thing however with Halbach array ringmagnet is that the magnetic flux is focused where it suppose to be when we use that array in a DC-motor. Most possible flux is present where it is desired to have it, and not spread out in the surroundings.

The other thing with the Halbach array is that the poles are shifting along the array - shouldnt be a problem if the railmagnet was configured likewise. However, the railmagnet must at some time and place rotate a bit in order to shift from attracting to repelling mode. This action must be done right in the middle of the highest flux density to achieve maximum attraction and repell moments. This operation requires great force. In fact so much force it will equalize all attractive and repelling forces in front an behind the Halbach array ring magnet. So the system will halt.

what is possible to test however, is to make a perfect counterforce for the rotation inside the Halbach array ring. This system can stay stationary and perfectly balanced, so each time the railmagnet is going in the "hole" to rotate enough to change polarity, this stationary counterforce would therfor allow the railmagnet to shift polarity without cost.

To be honest, I wonder how that works in practice. I have the idea in my head, so I might make a drawing of it on the paper, to easier see how it work - if it work.

Well, allways sceptic, but still open minded

Vidar

Vidar, it all depends on how you build your Halbach cylinder array.  Take a look at the Halbach cylinders on wikipedia and you will see how you can have a field on the ID of the ringmagnet,  http://en.wikipedia.org/wiki/Halbach_cylinder#Halbach_cylinder

The poles don't shift along the array.  What you see as the poles shifting, is causing the field to be canceled on one side and strengthened on the other side. I don't think the rail magnet needs to be a Halbach cylinder, since there would only be one pole pointing in the same direction across the entire inside of the ring magnet.

Try to think of it like this.  On the front side of the ring magnet, you can have a very strong north pole on both the OD and ID.  On the back side of the ring magnet, you can have a very weak to no south pole on both the OD and ID.  I may be wrong about this, but I don't see why the array can't be built for the desired results we would like.

Here's a youtube video on one of the configurations you could have with a Halbach array cylinder,  http://www.youtube.com/watch?v=qv-9IAj_YnI&feature=related