The parts of the soft iron covered by the magnets will have a polarity opposite that of magnets' nearest poles. The exposed parts of the soft iron will have a polarity opposite the magnets' farthest poles. The soft iron piece will not have all the same polarity, in other words. The flux lines entering the soft iron will always equal the number leaving. Always. Draw flux lines in your diagram.

It all depends on the size of your "soft material".

If you replace it with a thin sheet, your 2 magnets will act like there is almost nothing between the poles.

The thicker soft material "absorbs" the magnetic force(s).

This is my simple explanation of your situation.

As with other things in life, size is important. See #3↑.

Your soft magnetic material is just a bunch of magnetized regions (domains) arranged in random directions so that they cancel out. When you bring the strong magnet close to it, the domains that are lined up with the field grow at the expense of the others, so that the soft material becomes a magnet temporarily.

I prefer to think that the magnetism induced in #3, was changed "re-induced" when you tested with #4. That is, it was magnetised the other direction. If you had a way to lock-in the #3 induced magnetism, you would observe the expected results in #4.

Your "soft material" wouldn't be fridge magnet perchance?

Actually what's missing from the above posts is the fact that a "soft" magnetic material will "easily" realign its magnetic domains when exposed to a much stronger magnetic field, in effect erasing its previous poles.

So when you magnetize a common nail by exposing it to a magnet and then expose it to a much stronger magnet, it will "forget" that it was previously magnetized and adopt the stronger magnet's orientation. It's how data is stored on a disc drive or magnetic tape, and why a degaussing coil will wipe the tape clean (almost). The size, i.e., relative strength, of the magnetic field is what matters.

There's a *lot* missing from the above posts, and not through any fault of their own. Magnetism is such a very broad field, after all. We could mention lots of things not covered here. Things like, why do so-called 'permanent' magnets will lose their magnetisation when heated above their Curie Points? What is a 'Curie Point', anyway? Conversely, why do such materials retain a substantial residual field if a strong external field is applied as these materials are cooled below their Curie Points? And why, when you break a magnet in half, does the break nearly always appear highly crystalline? And why do rare-earth magnets just creep down a smooth, NON-magnetic aluminium slope when an unmagnetised chunk of iron of the same weight slides down unhindered? (a fun experiment to demonstrate eddy currents, btw). Lots of fun stuff there!

I have been really looking for a good explanation. The discussions seem to be going at a tangent. Let me give you some more details of my experiment.

Magnets are 10 x10 mm sq face, 6 mm thick (Along N & S side) Neodymium magnets. The soft magnetic material is ferrite. I tried with different sizes and dimensions, but result was same- irrespective of the size (I could not get as thin as a paper) both neodymium magnets stuck to the ferrite at the same time- not that I tried to magnetise soft ferrite, removed the neodymium and later tried to place the second magnet.

If the magnets repel - then they should levitate over the soft ferrite !!!!

How many magnets did you place, total? Two? Ferrite has a fairly-to-very-high initial permeability, depending on the type of ferrite. If you're using only two magnets in the configuration shown (same poles facing material), nothing is obstructing the return flux through the ferrite in the direction perpendicular to the configuation'a axis. Have you placed a 2D array of magnets out to the edges of the sheet? What were the results? I might also point out that attempting to levitate magnets in this way results in an inherently unstable field configuration in which the slightest perturbation will result in rapidly-escalating torques being applied to the levitatee, causing it to immediately flip over. Unless some means *ALSO* exists for countering this torque (the Levitron, for example, uses gyroscopic action) you cannot levitate in this way. A much better - and inherently stable - approach is to exploit the properties of materials which tend to *exclude* magnetic fields from their interiors. Such materials are called diamagnetic, and one of the best (non-superconducting) diamagnetic materials known is pyrolytic carbon (structurally similar to graphene). Better yet, you can make pyrolytic carbon yourself. First, though, I would construct an array on that ferrite sheet and note the results. You never know for sure until you do it yourself, and who knows? You may discover something new! :)

My real intention was to see if the soft magnetic ferrite will get repelled as two neodymium magnets were held tight close together by hand. I wanted to understand the concept first before embarking on a larger experiment - as you say using a large array - by simpler, cheaper technique, using magnets available in the market easily.

For levitation using linear motors for trains - they have a series of coils as tracks. Instead- if these coils can be in the train itself, the track should be soft magnetic material - existing track !!!!!

Take a small piece of ferrite and push it against the larger pole face of a single large magnet and what happens? It sticks. Same experiment, reduced to its essentials.

Magnetic fields are *vector fields*, that is, the field at every point in the space around and within a magnet or magnetic material can be represented by a magnitude and a direction - a vector. Consequently, magnetic fields can be (and are) modeled as a set of *closed, directional loops* - lines of magnetic flux. It is why there are no magnetic monopoles (except in certain cosmological theories in which they manifest as the result of a class of 'structural' defects in the fabric of spacetime. These do not apply here).

If I were to hazard a guess, I'd say offhand you are trying to create a magnetic monopole or something that behaves like one. Thing is, the field induced in your ferrite sheet will *never* present the same polarity to the magnets that the magnets present to the ferrite sheet. The polarity will *always* be the opposite for induced fields. Always. Moreover, ninety nine percent of questions of this sort would answer themselves if just this one simple fact were kept in mind: lines of magnetic flux are closed, directional loops.

"lines of magnetic flux are closed, directional loops."

GA.

Can I say it this way (in a very simple way):?

'The ferrite is not being magnetised but simply channelling the magnetic flux back more efficiently than just AIR. Therefore, each strong magnet flux competes for part of the ferrite to chaanel its flux back to close the loop.'

Thanks

I have noted your point and trying to conduct experiment by closing the magnetic loop on one side.

Magnets with strong magnetic field can attract to magnets with low magnetic filed irrespective of there Poles.

Here are results of my next experiment.

I was expecting that the intermediate soft ferrite will be able to stay afloat in between. But worse both the magnets - did get stuck hard- though without it- they were actually repelling each other !!!!

Split your drawing into two parts at the midplane of the central ferrite sheet: each half is a complete magnetic 'circuit' unto itself. I cannot stress enough that you should also draw the magnetic flux. Showing the poles is only half the story. They're nice diagrams but they're incomplete.

The following are examples of where designers have drawn the flux paths in these various components. It is this which tells them whether they're on the right track, especially in the more complicated cases.

-----

-----

-----

-----

All but the first figure were modeled using FEA (Finite Element Analysis) on computer. You don't need a rigorous model at this stage of your design. Just sketch to give a feel for what the field is doing is enough - and very instructive!

Your focus needs to be on the shape of the magnetic field that will produce the desired results, NOT on the configuration of the magnets. The shape of the field will dictate the magnet configuration needed to produce that shape - nor is the solution unique, which gives you lots of of latitude in your design.

Your experimenting is a Good Thing, btw. Nothing like 'getting your hands dirty'.

Thank you for explaining using FEA. A closed loop of magnetic path explains how two magnets or a a magnet and a soft magnetic material get stuck. Now I do understand that for explaining repulsion also magnetic paths must be studied.

Now in the experiment - if soft ferrite bar is replaced with a non-magnetised hard ferrite (I do not have and so unable to actually conduct the experiment) - will two magnets with like poles facing each other repel? It should. What is your feeling? Is there any way this can be checked with FEA analysis? Does it have feature to simulate non-magnetized - hard magnetic material?