Industrial physicists and engineers are expected to be knowledgeable in a wide range of disciplines. In addition, they are expected to be able to reduce centuries of scientific work and language to jargon comprehensible by others with less interest in technical details. This can be especially difficult when the mysteries of magnetics are involved. Magnetic fields are used for energy conversion, holding, levitation, torque transmission, and particle beam control. Modern industrial processes also use magnetic fields. For example, the sputter deposition process is enhanced by a magnetic field, and the presence of a magnetic field during deposition of magnetic thin films improves the magnetic characteristics of recording heads and switches. Explaining what goes on in these processes requires some knowledge of the magnetization / demagnetization process, and domain theory is the place to start.

In 1907 French physicist Pierre Weiss postulated that ferromagnetic materials were fully magnetized at all times. He attributed this to spontaneous "molecular fields" that caused total alignment of atomic magnetic moments. However, he had to explain the absence of an apparent magnetic field when the material was not magnetized. He theorized that macroscopic regions within the material were composed of sub regions, called domains, wherein all atomic magnetic moments are parallel, and that domains arranged themselves randomly when no external (or internal) magnetomotive force was present. This is analogous to world wide power struggles, and maybe even Pierre saw it this way.

A domain is the territory of a ruler, and a ruler is more powerful when his domain is large. Initially a country is usually a collection of very small domains run by local rulers who are content to just balance the influence of their neighbors; no ideas of conquest. When a strong king comes along, many of the domains and their rulers align with himand his direction. The strong domains grow at the expense of weakerones, and the country becomes a group of large domains aligned with the king. Strong local rulers keep the domains aligned with the king even when he relaxes, or is off on another quest. It is not enough that the domains are large, and have strong rulers, they must have a high population density to supply soldiers for the king's army. Large, weak domains are difficult to maintain even if they have a large population density.

If local rulers weaken for any number of reasons (maybe they can't stand the heat when the king is gone), then the folks in the outer domains do not get strong communications about their role, or their direction, and they are susceptible to change. These weakened domains may be changed by their own internal forces, or the influence of strong opposing forces in neighboring domains. In this way a gradual erosion of domain alignment may occur, and they may eventually relax to a neutral position, or be taken over completely by another invading king. Usually the domains on the edges or ends of a territory change hands first because they do not have strong neighborly support; they just do not feel strong attachment.

Magnetic domains function in the same way. All magnetic materials initially have many small randomly aligned domains. Rare earth materials have strong local forces but, unless influenced by some outside force, they live happily with neighbors in a state of equilibrium. Externally, the material appears to have no preferred alignment and it does not influence other objects. When a strong outside force appears, in the guise of a magnetizing pulse, then the magnetic domains all line up with the force and add to it, and the strong domains grow in size. When the strong external force goes away the domains would relax and go random again unless a strong internal force (coercivity of the material) takes control while the strong external force is present. Internal mechanisms that develop coercivity are "anisotropy". The four types of anisotropy are: (a) crystal, (b) stress, (3) shape, and (d) exchange. To learn more about them please refer to a technical text.

If the material has a high flux density and a strong internal force it can influence other domains. However, domains on the periphery of the magnet volume have fewer neighboring domains for support so they can be turned around and oppose the aligned domains if encouraged to do so by an external force, or random energy additions. The magnet's "king" is a magnetizing, or demagnetizing, force; flux density is the magnet's army; and "coercivity" is the ruler that keep domains aligned when the king goes away. A magnet's useful strength depends on how many domains remain aligned after the external force goes away and they have to face opposing kings and armies on their own. A short magnetic length is like an army with a wide advancing front; it is susceptible to the opposing force everywhere, but mostly at the edges. When the edges are neutralized by opposing external forces the rest of the magnet may be made ineffective. An army column with a good length to width ratio, depending on its internal and external resources, is the most effective and the same is true with magnets. External resources for a magnet would be other magnetic circuit components. Other factors are also involved. When an external magnetizing force is applied it "moves domain walls" to make the stronger domains larger. It takes just as much energy to move these domain walls back to where they were, so a magnet that is hard to magnetize usually is hard to demagnetize. However, magnets do not like heat; thermal energy reduces pole strength (flux density) and the ability of domains to remain aligned, depending on the temperature reached and intrinsic strength. Heat first weakens the "magnetic army"; this is a flux density loss that is recovered when things cool off. Next, the outer and weaker domains reverse and these losses are recoverable only by "the king." The process slows as heat travels into the magnet and encounters stronger internal domains. If a magnet gets too hot its chemistry actually changes, and it can be permanently damaged. Heating in air initially results in oxidation; higher temperatures may result in phase changes. The temperature at which chemistry changes may begin to occur in magnet materials is identified as the maximum operating temperature; this is usually much lower than the "Curie temperature". (Curie temperature is the temperature at which an element or alloy becomes totally non magnetic.) Ceramic magnets (strontium or barium ferrite) are the exception; they are made from oxides that do not change composition even after their magnetic temperature limit is reached. Magnets can be "thermally demagnetized" at moderate temperatures (less than Curie) to allow further processing after testing but, while little or no external field is exhibited, they are far from their virgin state. So long as they have not seen a chemistry change, these magnets will recover fully when remagnetized. Heat treating in the magnet manufacturing process uses precise heat and cool rates to precipitate the desired phase, a protective gas atmosphere to avoid oxidation, and usually an orienting magnetic field. Magnet material producers use combinations of atomic elements in their magnets to make domain walls harder to move and more resistant to heat. However, these elements take up a portion of the magnet's volume so they produce less total flux (fewer soldiers). "Domain wall pinning" is another factor in making domain walls harder to move; this is like driving a pile into the ground to keep the domain wall from moving. Domain theory is useful in understanding how magnets work. It cannot account for all observed magnetic phenomena, but it helps scientists describe the internal cause for observed external field changes. It is interesting to note that it has become possible to see magnetic domains with a variety of high power microscopes. So domain theory, once just a convenient way to explain how magnets work, has a factual basis.