Magnetic Materials

01 / UNDERSTAND & APPLY

1. A material’s response is described by magnetization

Atoms and molecules can possess magnetic dipole moments associated with electronic orbital angular momentum and intrinsic spin. Electron spin is a quantum property, not literally a tiny charged ball rotating about its axis. Whether many moments cancel, partly align or order collectively determines the macroscopic response of a material.

Magnetization M is magnetic dipole moment per unit volume. Its SI units are A m⁻¹ because dipole moment has units A m². It is a vector average: a sample can contain many microscopic moments but have zero net magnetization when their contributions cancel. Zero bulk magnetization does not prove the absence of all local magnetic structure.

For a small uniformly magnetized sample of volume V, the total magnetic moment is m_total = MV. For a nonuniform sample, integrate magnetization over volume. This moment can experience torque in an external field, while a net translational force generally requires field nonuniformity. A uniform field can turn a dipole without drawing it toward one particular point.

The elementary material classes are diamagnetic, paramagnetic and ferromagnetic. They differ in the sign, size and history dependence of their response. A classification is a model of a material in stated conditions; temperature, crystal direction and phase can change behavior. Avoid treating a single constant susceptibility as a complete description of every magnet.

02 / UNDERSTAND & APPLY

2. B, H and M are related but are not the same quantity

In SI notation, magnetic flux density B, magnetizing field H and magnetization M are related by B = μ₀(H + M). B is measured in tesla; H and M in A m⁻¹. For a simple homogeneous isotropic linear material, M = χ_mH with dimensionless susceptibility χ_m, giving B = μ₀(1 + χ_m)H.

Define relative permeability μ_r = 1 + χ_m and permeability μ = μ₀μ_r in that linear regime. The algebra μ_r = 1 + χ_m is not a statement that every ferromagnet has one constant μ_r independent of field and history. A nonlinear B–H curve can have different secant and differential permeabilities, and hysteresis can make B depend on how H was reached.

For a very long uniformly filled solenoid with negligible end effects, H is approximately nI, where n is turns per length. A finite sample’s internal H can differ from the applied field because its own magnetization creates a demagnetizing field. Shape matters, especially for strongly magnetic materials. The simple long-solenoid formula deliberately suppresses that complication.

Do not add μ₀χ_mH to B twice. Start from H, find M = χ_mH, then obtain B = μ₀(H + M). If a problem supplies B directly, converting it to H requires the appropriate material law and geometry, not simply dividing by μ₀ while ignoring magnetization.

03 / UNDERSTAND & APPLY

3. Diamagnetism is an induced response opposing the applied field

A magnetic field induces a small orbital response that opposes the applied field. When this response dominates the material’s net susceptibility, the material is called diamagnetic. In the usual weak linear regime, χ_m is small and negative and μ_r is slightly below one. There is no requirement for a pre-existing permanent atomic moment to align in order for this response to occur.

Diamagnetic response is present more broadly but can be masked by stronger paramagnetic or ferromagnetic contributions. Ordinary examples include water, copper and bismuth in familiar conditions. The induced response disappears when the applied field is removed in the simple nonhysteretic model. Do not describe it as a permanent magnetization retained after switching off the field.

A weak diamagnetic specimen in a nonuniform applied field tends toward regions of smaller field strength, subject to its other forces and geometry. A perfectly uniform field does not supply that field-strength gradient. The everyday phrase “repelled by a magnet” summarizes a spatially nonuniform interaction, not a uniform-field net-force law.

Ordinary diamagnetic susceptibilities are often only weakly temperature dependent over a limited range, but precise behavior remains material dependent. Superconductors can show strong flux exclusion under appropriate conditions; that is a distinct physical state and should not be treated merely as an ordinary weak diamagnet with every detail unchanged.

04 / UNDERSTAND & APPLY

4. Paramagnetism is partial alignment opposed by thermal disorder

A paramagnetic material has microscopic permanent moments whose orientations partly align with an applied field, producing a small positive susceptibility in the ordinary weak-field regime. Thermal motion competes with alignment, so the net moment is usually much smaller than the fully aligned limit. It is not correct to assign a universal fixed fraction of moments aligned in every paramagnet.

For an ideal collection of weakly interacting localized moments in an appropriate high-temperature weak-field regime, Curie’s law gives χ_m = C_Curie/T with T in kelvin. At fixed H, magnetization then decreases as temperature rises. The formula is approximate and can fail for strong interactions, low temperatures, saturation or itinerant-electron responses.

Paramagnetic susceptibility is positive, so μ_r is slightly above one in the simple linear model. A weak specimen tends toward stronger-field regions in a nonuniform applied field. After the field is removed, ordinary thermal randomization removes the net alignment in the simple equilibrium model, unlike persistent ferromagnetic remanence.

The energy of a magnetic dipole in an imposed field is U = −m·B. Alignment lowers its energy, while anti-alignment raises it. Comparing a moment’s field-energy scale mB with k_BT explains qualitatively why higher temperature weakens alignment. This is an intuition for the competition, not a replacement for the full quantum statistical calculation.

05 / UNDERSTAND & APPLY

5. Collective ordering creates domains and strong nonlinear response

In a ferromagnet below its ordering temperature, quantum exchange interactions favor parallel alignment of neighboring moments. A domain is a region with substantial local spontaneous magnetization. An unmagnetized bulk sample can contain domains pointing in different directions whose total moments largely cancel. Its atoms are not necessarily individually disordered in the same way as a high-temperature paramagnet.

Applying a field can move domain walls so favorably oriented domains grow and can rotate moments toward the field. This creates a much stronger response than ordinary weak paramagnetism. Defects, stresses and microstructure can hinder wall motion, making the response depend on the sample’s magnetic history.

At sufficiently strong magnetizing field, the material approaches saturation of magnetization when most available moments are aligned. The B value can still increase through the μ₀H contribution even when M is nearly saturated. It is therefore more precise to say magnetization saturates than to claim that B becomes perfectly constant for all additional H.

Above the Curie temperature, the long-range ferromagnetic order is lost and the material becomes paramagnetic in the elementary picture. The exact transition temperature is material specific. Heating a permanent magnet can alter its remanence before or across that transition, but no universal numerical temperature applies to every magnet or alloy.

06 / UNDERSTAND & APPLY

6. Remanence and coercivity describe the loop’s memory

When H is cycled through positive and negative values, a ferromagnet can trace a hysteresis loop rather than one reversible curve. Starting from positive saturation and reducing H to zero can leave nonzero B, called remanent flux density B_r in a B–H plot. This is evidence of retained magnetization, not a hidden applied current still being required in the idealized sample description.

The reverse field magnitude required to reduce B to zero on that branch is the coercive field H_c for the B–H convention. Coercivity defined from an M–H curve is a related but potentially different quantity, so label the plotted vertical axis. The two axes must not be swapped without changing the meaning of the geometric intercepts.

The area enclosed by a quasistatic B–H loop has units (A m⁻¹)(T) = J m⁻³ and represents hysteresis energy dissipated per unit volume per complete magnetization cycle. Multiplying by core volume and cycles per second gives hysteresis power loss. Dynamic eddy-current and other frequency-dependent losses may add to that quasistatic contribution.

The loop depends on how far the field is cycled. A minor loop does not have the same area or intercepts as a full major loop reaching saturation. A schematic large loop should not be used to infer an exact loss for a smaller operating excursion without corresponding material data.

07 / UNDERSTAND & APPLY

7. Choose magnetic behavior for the task rather than maximizing one number

A magnetically soft material is relatively easy to magnetize and demagnetize, commonly with low coercivity and a narrow hysteresis loop. Such behavior is useful for repeatedly cycled cores where low hysteresis loss matters. The word soft refers to magnetic response, not necessarily mechanical softness.

A magnetically hard material resists demagnetization, with substantial coercivity and suitable remanence. It is useful for permanent magnets that must retain a field despite opposing influences. High permeability alone does not establish that a material is the best permanent magnet; its full demagnetization behavior matters.

In AC cores, laminating conducting material or using high-resistivity magnetic materials can reduce eddy-current losses. Laminations interrupt large current loops; they do not directly remove the material’s intrinsic hysteresis loop. Frequency, saturation limits, thermal behavior and geometry all enter a real core choice.

All magnetic field lines remain divergence-free in the usual Maxwell description. Magnetization does not create isolated magnetic monopoles by cutting a bar magnet into smaller pieces. The new pieces have their own dipolar pole patterns. Domain pictures are microscopic explanations of response, not a claim that free north and south magnetic charges can simply be separated.

08 / UNDERSTAND & APPLY

8. Original practice with signs, units and history

Use the linear relation only where it is explicitly justified. Keep B, H and M units distinct, and read hysteresis intercepts from the actual axis labels.

09 / UNDERSTAND & APPLY

9. Name the response regime before using permeability

Begin with the sign and physical origin of the response: induced opposition, weak alignment or collective ordered domains. Then decide whether a linear χ_m is appropriate or whether a history-dependent curve is required. Specify internal H versus applied field whenever sample shape can matter.

For a hysteresis question, inspect the axes and operating loop before reading remanence, coercivity or loss. For a material application, separate permeability, coercivity, saturation and electrical resistivity instead of reducing the choice to “more magnetic.” Temperature and microstructure can matter as much as the category name.

• Magnetization is moment per volume and can vanish by cancellation.
• B = μ₀(H + M) does not make B, H and M interchangeable.
• Weak diamagnets have negative χ; ordinary paramagnets have positive χ.
• Ferromagnetic domains produce strong, nonlinear and history-dependent response.
• A loop area gives energy loss per volume per cycle, with other losses treated separately.
• Magnetic softness and hardness describe reversal behavior, not mechanical texture.