What Is Hysteresis?

Hysteresis is the property of a physical system in which its current state depends on the history of past inputs, not solely on the present input. When an external driving quantity, such as a magnetic field, mechanical stress, or electric field, is applied and then removed, the system does not return along the same response path it followed during loading. The area enclosed by the loading and unloading curves on a plot of output versus input, called the hysteresis loop, represents energy dissipated or stored within the system per cycle. The term derives from the Greek hysteron, meaning "coming after," and was applied to magnetic materials by physicist James Alfred Ewing in the 1880s.

Hysteresis appears across a broad range of physical and engineering domains, including ferromagnetism, ferroelectricity, structural mechanics, control systems, and biology. In each case the underlying mechanism differs, but the mathematical form of the input-output relationship, exhibiting path dependence and memory of past states, is analogous. Damping in mechanical systems and energy loss in transformer cores are both manifestations of hysteretic behavior, though the responsible physical mechanisms are distinct.

Magnetic Hysteresis

Magnetic hysteresis describes the lag between the magnetization of a ferromagnetic material and the applied magnetic field. As an external field increases, magnetic domains within the material align progressively until the material reaches saturation. When the applied field is then reduced to zero, a residual magnetization called remanence remains; the field that must be applied in the opposite direction to reduce magnetization to zero is called the coercive field. The area of the B-H loop quantifies the energy dissipated as heat per unit volume per cycle, which is a key design parameter for transformer and inductor cores: materials with a narrow loop are preferred for power applications to minimize losses, while materials with a wide, square loop are used in magnetic memory to maintain stable remanent states. An arXiv preprint on modeling of hysteresis loops in ferroelectric and magnetic materials surveys the mathematical frameworks used to characterize B-H curves, while IEEE Standard 180 defines reference points, including coercivity, remanence, and saturation, that allow numeric comparison of magnetic materials across ferromagnetic alloys and oxides.

Spin Valves and Magnetoresistive Behavior

Spin valve structures exploit magnetic hysteresis at the nanoscale to produce a large change in electrical resistance in response to modest applied magnetic fields. A typical spin valve consists of two ferromagnetic thin-film layers separated by a nonmagnetic metallic spacer. One layer is pinned to a fixed magnetization direction by exchange coupling with an antiferromagnetic underlayer; the other is a free layer whose magnetization switches in response to an external field. The resistance is low when both layers are magnetized parallel and high when they are antiparallel, owing to the giant magnetoresistance (GMR) effect documented in Physical Review Letters and related publications. The sharp switching transition of the free layer defines a hysteresis window whose width, set by the free layer coercivity, determines the field range over which the device transitions between its two resistance states.

Ferroelectric Hysteresis

Ferroelectric materials exhibit a polarization-electric field (P-E) hysteresis loop that parallels the B-H loop of ferromagnets. Applying an electric field aligns electric dipoles within polarization domains; after field removal, a remanent polarization persists and a coercive field is needed to reverse it. As characterized in ferroelectric hysteresis measurement frameworks from NIST-referenced standards, the loop shape provides direct evidence that a material is genuinely ferroelectric rather than merely high-k dielectric, and quantifies the polarization switching energy that governs power consumption in ferroelectric memory cells.

Applications

Hysteresis has applications in a wide range of disciplines, including:

• Magnetic data storage in hard disk drives and magnetic random-access memory
• Ferroelectric RAM and capacitive memory devices
• Transformer and inductor core design for power electronics
• Spin-valve magnetic field sensors in read heads and biosensors
• Control system deadband design for noise-immune switching
• Structural fatigue analysis in mechanical engineering