Magnet system efficiency made from neodymium magnets

  • #1
Me and my friend want to make a strong as possible (made from permanent magnets) homemade metal separator for seeds. We are not good in the physics of such level, so we need consultation.
We want to try neodymium magnets for our separator. We learned that magnetic flux density could be reached 1.5T (tesla) at N52-N54 neodymium magnets. But we also learned two important things;

First that magnetic flux density of the magnets depends on the geometry (disc/block/sphere) and size of the magnet;

Second magnetic flux density could be raised if using the system of magnets with special packing of these magnets.

So I have a few questions

1) What geometry of the magnet will give the maximum magnetic flux density?

2) How can I calculate the size of magnet with highest magnetic flux density? For example efficient size of magnet blocks.

3) What packing of magnets will give the highest magnetic flux density? How can I calculate the best packing?

Thanks in advance.
Last edited by a moderator:

Answers and Replies

  • #2
Welcome to the PF. :smile:

Are you wanting to remove metal pieces from the seeds, or is there some kind of sorting that can be done on the seeds themselves with strong magnetic fields?

How big of a seed separator are you hoping to build? Do you have any links to magnetic seed separation that we can look at? Thanks.



  • seed_high_gradient_separator.jpg
    21.9 KB · Views: 456
  • #3
Greetings, Berkeman.

Thank you =)

1) We want to remove metal pieces.
2) We want to construct a separator maybe a little bigger then in the picture.

If talk about examples, we found this chinese separators very interesting

Eriez company is also interesting

Also here is the photo of magnets packing. That we are interested in.


  • upload_2018-9-20_10-59-0.png
    28.5 KB · Views: 406
  • example.jpg
    37.7 KB · Views: 460
  • #4
Be careful when working with neodymium magnets. The magnets in the photo below are 3/8" (9.5 mm) thick. Three of them were installed on a steel block side by side, all three with the north face up. They were too strong for the technician to lift and place them, so he set them one at a time on the block, and slid them into place between aluminum guides. He could get the first two magnets into position, but was not strong enough to push the third magnet into position. He had to design a jack screw mechanism to get the third magnet into position. Since all three were installed with the same pole upward, there was a strong repulsion between magnets. After the epoxy hardened, they stayed in place.


If a magnet of this size and strength gets away from you, it will shatter when it lands. It will also draw blood if your finger is in the way. A technician had one of these underneath the tool tray on his rollaway. It sucked a wrench out of the pocket of his foreman when the man walked too close.

There is another effect that I never fully wrapped my mind around. The strongest neodymium magnets are not necessarily the best. It's something to do with demagnetization under certain circumstances. I think it has something to do with open magnetic circuits, possibly during installation. Or maybe at higher temperatures.

This looks like a good place to get started designing permanent magnet circuits:


  • upload_2018-9-20_19-28-2.png
    204.1 KB · Views: 405
  • #5
Thank you for the link, jrmichler, it might help us.

Oh, we do understand how hard to work with the magnets. We have two 50x50x50 N48 magnet cubes at the moment and they was really close to stab me with the knife. Gladly, the knife was located 1 meter in front of me not behind and its landed with high speed directly on the magnet not harming me.

Anyway, our goal is to make as strong metal separation system as possible. Any kind of information about efficient geometry, size, packing will be usefull for us.

We also learned a little about ferromagnetic concentrators for magnet systems (in the patent It would be great to get information about this too.
  • #6
Is this related to your research? If so, I recommend a complete patent search. This is best done using the US Patent Office site: Start with the patent you already found, then look at all of the prior art patents plus the Referenced By patents. Then look at the prior art and Referenced By patents for each patent that is relevant to your work. Do a Google search for magnetic separators. Then search for patents assigned to those companies. And the prior art... Repeat until none of the patents are relevant.

Eriez is a well known name in magnetic separation. They have 24 patents that are worth looking at. Side note from their brochure: "The P-Rex Rare Earth model is possibly the strongest permanent magnet on the planet".

Have you looked at magnetic circuit simulation software? There are a number of packages, but you should know how to calculate a simple magnetic circuit before attempting to use magnetic simulation software.

A literature search can also be done. Since the math of magnetism was well understood a long time ago, relevant papers might well be 100 years old.

Suggested for: Magnet system efficiency made from neodymium magnets