Practical demo - Ferromagnetic attraction at interpole boundary

[magnetics]

TL;DR

Experiment with flat side or pointed side of a nail and its attractive forces to an alternating pole static magnet

I was experimenting with multipolar permanent magnets (concentric alternating-polarity) and a nail and noticed something that initially seems counterintuitive.

Using a handheld gauss meter, when scanning normal to the magnet face (z-axis), the measured Bz is maximal near the centre of the magnet (furthest from the alternating pole) and approaches zero at the boundaries between adjacent poles, as expected due to axial cancellation.

However, when placing a small ferromagnetic object (e.g. a steel nail) over the magnet:

• With the flat end facing the magnet, the object consistently snaps to the interpole boundaries, precisely where Bz ≈ 0
• When the nail is flipped and the sharp tip faces the magnet, it preferentially aligns over the centre, where Bz is maximal

This suggests that the attractive force is dominated by field gradients rather than local field magnitude, and that object geometry (flat or point) strongly influences which part of the field is sampled.

Am I correct in interpreting this behaviour as a consequence of ∇(B²) being maximal at the interpole boundaries despite Bz ≈ 0 there? And is the preference of the pointed tip for the centre best explained by flux concentration and field symmetry?

I’m interested in whether there’s a clean way to describe this distinction between measured field strength and force in multipolar permanent magnet geometries.
Thank you.

 

[Charles Link]

Can you draw a diagram of where the poles are on this magnet? I had trouble following your explanations.

 

[Baluncore]

• With the flat end facing the magnet, the object consistently snaps to the interpole boundaries, precisely where Bz ≈ 0
• When the nail is flipped and the sharp tip faces the magnet, it preferentially aligns over the centre, where Bz is maximal

How do you know that the nail is not magnetised in some weird way?

 

[berkeman]

This suggests that the attractive force is dominated by field gradients rather than local field magnitude

Correct. If the B field is constant in a region enclosing a ferrous object, there is no net magnetic force on the object. There is a mild gradient heading axially away from a cylindrical magnet, which will attract an object like a nail. But there is a stronger gradient side-to-side between the coaxial alternating poles for that magnet that you show, which is why the flat head is strongly attracted to that region.

BTW, how can an alternating pole coaxial magnet have 3 exposed poles? Shouldn't it have an even number of poles? How is it actually constructed?

 

[Charles Link]

Correct. If the B field is constant in a region enclosing a ferrous object, there is no net magnetic force on the object. There is a mild gradient heading axially away from a cylindrical magnet, which will attract an object like a nail. But there is a stronger gradient side-to-side between the coaxial alternating poles for that magnet that you show, which is why the flat head is strongly attracted to that region.

BTW, how can an alternating pole coaxial magnet have 3 exposed poles? Shouldn't it have an even number of poles? How is it actually constructed?

The pictures were unclear and the OP's explanations were lacking. I think that larger surface with 3 rings is the sensor, but I don't know.

 

[magnetics]

The pictures were unclear and the OP's explanations were lacking. I think that larger surface with 3 rings is the sensor, but I don't know.

It could have been explained better, I agree. The magnets in the photo are static magnets and the interpole boundary is the neutral line between two neighboring opposite-polarity rings. Just like the images below, where the interpole boundary is the neutral line between two neighboring opposite-polarity squares. You see the same effect. The point of the nail is attracted to the centre of the magnet, where Bz in max, while the flat part of the head of the nail is attracted to the point where B(dx/dy) gradient is maximum. I was expecting both ends of the nail to be attracted to same spot.

 

[Charles Link]

You've got a very odd looking nail. I thought the head of the nail was the magnet. That is the head of the nail as I know it. You called it the flat end. Your nail has spiral grooves in it. I don't know if you can expect it to behave like an ordinary nail.

 

[berkeman]

You've got a very odd looking nail. I thought the head of the nail was the magnet. That is the head of the nail as I know it. You called it the flat end. Your nail has spiral grooves in it. I don't know if you can expect it to behave like an ordinary nail.

It looks like they are called "Spiral Shank Nails", and the grooves afford superior holding power for the nails:

I don't think the spiral grooves are affecting this experiment. It's just that the B-field gradient experienced by the flat head of the nail is stronger side-to-side given the dimensions of the nail head versus the forces experienced by the nail when the pointed end is down. I don't find it all that strange overall.

 

[Charles Link]

It is starting to make a little more sense, now that I have a clearer picture of it. I think the flat head is extensive enough that it is able to get magnetized in the x-y plane with a north pole and a south pole that will line up with the south and north poles respectively of the underlying magnet.

For the case of the point, the nail can get magnetized in the z direction, and if the point is above a north pole of the underlying magnet, a south pole will appear at its tip, etc., with the north pole at its head.