Any ferromagnetic material, like iron, contains tiny magnetic domains. These are areas where all the atoms in the sample have undergone exchange coupling so their individual magnetic fields can act as single larger magnet. These places can be visually seen (under a microscope) by allowing a colloidal suspension of iron particles to settle on the specimen’s surface. The domain boundaries leak a little magnetic flux which attracts the particles of iron. The result reveals the pattern of the domains. Normally, each domain measures around a tenth of a millimeter across
Domains are created by the exchange forces created when the kinetic energy required to move an electron to an upper free band is less than allowing two electrons to enter into a parallel orientation. Basically, that says a few metals will arrange their inner shell electrons so not every electron’s magnetic field cancels another’s out.
Normally, electrons exist in pairs with their spins opposed, so each has its magnetic pole neutralized by its mate. Remember, an electron is a spinning electric charge, and whenever a charge moves, it creates a magnetic field. If we simplify and imagine electrons as spinning tops, it would look something like this:
The green lines represent the magnetic flux. You can see how the flux would not travel very far.
But in the ferromagnetic materials, like iron, the 3d electron lacks a mate in its shell. When that happens its field will reach out into space.
If the inter-atomic distance is correct, the atoms can arrange themselves so that exchange coupling between these unpaired electrons in neighboring atoms takes place and their fields add together.
Now if all the atoms in a sample were linked into the same magnetic orientation, they would generate a powerful external field which would extend far beyond the sample. In nature a lower energy state is preferred, so domains are formed to keep the magnetic lines of force as short as possible. (A lower energy state.)
Although each small domain is a magnet, within the matrix the domains arrange themselves so that overall, an unmagnetized ferromagnetic sample shows no external magnetic field. In pure iron, because of its crystal lattice, there are six possible orientations a domain can assume. Inside a domain all of the atoms are linked and their total magnetic field points in the same direction. But, just a short tenth of a millimeter away there are another bunch of atoms all lined up and pointing their field in a different direction. Like good neighbors they establish a boundary between them. And that boundary is called a domain wall and this is the place that attracts those particles of iron. At room temperature, the boundaries’ thickness is around 1000 atomic diameters. As a sample is heated, the boundaries’ width expands.
When a sample’s temperature reaches its Curie Point, you could say the boundaries have become so wide they start to touch each other and the domains have disappeared.
This is one reason why I believe the ferromagnetic generator should operate best a few degrees below the sample’s Curie point. It needs those boundaries and the remaining still-linked exchange forces to react to the external electronic stress necessary to change the Curie Point.