1 Definition and basic concept

Hysteresis is a phenomenon in which the response of a system depends on its present conditions and on the path by which those conditions were reached. In practical terms, a system can show different outputs for the same input depending on whether the input is increasing, decreasing, or cycling.

The term is used broadly across science and engineering to describe lag, retention, and memory-like behavior. It is especially important when a material, device, or model does not return along the same route when an external influence is removed or reversed.

1.1 Historical background

The idea developed through nineteenth-century studies of magnetism, elasticity, and thermal behavior. Early investigators noticed that some materials did not respond instantaneously or symmetrically to changing forces. These observations led to the recognition that a system’s past state could influence its current behavior.

Over time, the concept became formalized in physics and engineering. Later work extended it to electronics, biology, economics, and other fields where path dependence can shape outcomes.

1.2 Core characteristics

Hysteresis is usually identified by three closely related features: path dependence, memory effect, and lag or irreversibility. These traits may appear together, although not every system exhibits all of them in the same way.

1.2.1 Path dependence

Path dependence means that the current output is determined not only by the present input, but also by the sequence of earlier inputs. Two systems with identical current conditions may behave differently if they followed different histories. This is a defining feature of hysteresis in many contexts.

1.2.2 Memory effect

A memory effect occurs when a system retains information about earlier states. The retained information may be physical, such as internal structural rearrangement in a material, or abstract, such as a state variable in a mathematical model. In either case, the system preserves traces of prior changes.

1.2.3 Lag and irreversibility

Lag refers to a delayed response to changing input. Irreversibility occurs when the return path differs from the forward path, so reversing the input does not immediately restore the original state. In many systems, lag and irreversibility combine to produce a closed loop rather than a single-valued curve.

1.3 Hysteresis loop

A hysteresis loop is a graph showing how output changes as input is cycled upward and downward. The loop usually indicates that the response during increasing input differs from the response during decreasing input. The enclosed area often represents energy dissipated in the cycle or another form of lost work.

Such loops are commonly used to compare materials, quantify model parameters, and visualize dynamic behavior. Their shape can reveal thresholds, saturation, residual effects, and other key features.

2 Mathematical description

Mathematical descriptions of hysteresis aim to represent systems whose outputs depend on history. Because the same input may correspond to multiple possible states, such models are typically non-unique and require internal variables or operators that encode prior evolution.

2.1 State variables and input-output relations

In many models, the current state is represented by one or more state variables that summarize the system’s internal condition. These variables evolve with the input and determine the output. The result is an input-output relation that cannot be expressed as a simple instantaneous function.

A useful description often separates external forcing from internal state. The external input may be field, stress, temperature, voltage, or another control variable, while the output may be magnetization, strain, polarization, current, or population response.

2.2 Nonlinear models

Because hysteresis is frequently nonlinear, many models use threshold behavior, switching rules, or distributed internal states. These approaches can reproduce loops, saturation, and asymmetry.

2.2.1 Preisach model

The Preisach model is a classical representation built from many simple switching elements. Each element changes state at different thresholds, and the combined effect produces a macroscopic hysteresis loop. It is widely used because it can approximate complex memory behavior with considerable flexibility.

2.2.2 Bouc-Wen model

The Bouc-Wen model is a smooth phenomenological model often applied to structural and mechanical systems. It captures gradual transition, energy dissipation, and pinched or rounded loop shapes. Its parameters can be adjusted to fit observed cyclic behavior.

2.3 Differential equations and operators

Some hysteretic systems are described by differential equations coupled with internal variables. Others use specialized hysteresis operators that map an input history to an output. These operators are designed to preserve memory, threshold effects, and the distinction between increasing and decreasing paths.

Such formulations are useful when the system’s response depends on both rate and state. They allow theoretical analysis while still accommodating experimentally observed loop behavior.

2.4 Energy loss in cyclic processes

Many hysteretic cycles dissipate energy. In a repeated cycle, part of the input energy is not recovered because it is converted into heat, structural rearrangement, or another irreversible process. The energy loss is often represented by the area inside the hysteresis loop.

This loss is important in materials, machines, and circuits because it can reduce efficiency, generate heating, and alter long-term performance. In some systems, the dissipation is small; in others, it is the central physical feature.

3 Types of hysteresis

Hysteresis appears in several forms depending on whether the response is primarily controlled by time, field, stress, temperature, or biological regulation. The different types are often related, but each has characteristic features and applications.

3.1 Rate-independent hysteresis

Rate-independent hysteresis is largely unaffected by how fast the input changes, as long as the sequence of states is similar. The loop shape is mainly determined by thresholds and internal switching rather than by speed. This type is common in idealized models and certain quasi-static experiments.

3.2 Rate-dependent hysteresis

Rate-dependent hysteresis changes when the input is applied more quickly or more slowly. Time delays, diffusion, viscosity, and relaxation processes often contribute to this behavior. The loop may widen, shrink, or shift depending on the driving rate.

3.3 Magnetic hysteresis

Magnetic hysteresis arises in materials whose magnetization depends on prior magnetic field exposure. When the external field is cycled, the magnetic response does not follow the same route in both directions. This form is one of the best-known examples of the phenomenon.

3.4 Elastic hysteresis

Elastic hysteresis appears in deformable materials that do not return all stored mechanical energy during loading and unloading. The response can involve internal friction, microstructural changes, or viscoelastic effects. It is often observed in polymers, rubbers, and biological tissues.

3.5 Thermal hysteresis

Thermal hysteresis describes situations in which a system’s state changes differently during heating and cooling. A phase may persist below or above the temperature at which it normally changes because of delayed nucleation or metastability. This behavior is common in phase-change materials and some biological systems.

3.6 Biological hysteresis

Biological hysteresis occurs when physiological responses depend on prior stimulation. Examples include signaling pathways, gene regulation, muscle response, and population dynamics. In these cases, internal feedback and biochemical switching can produce memory-like effects.

4 Physical examples

Hysteresis is observed in many physical systems, often with distinct mechanisms but similar loop behavior. The examples below are among the most studied and widely used in practice.

4.1 Ferromagnetism

Ferromagnetic materials show strong hysteresis because their internal domains can align and remain aligned after the external field changes. The observed magnetization reflects both the current field and the prior magnetic history of the sample.

4.1.1 Magnetization curves

A magnetization curve traces the relation between magnetic field and magnetization during a cycle. As the field increases and then decreases, the curve typically forms a loop rather than a single line. The shape of the loop reflects domain motion, rotation, and pinning.

4.1.2 Remanence and coercivity

Remanence is the magnetization that remains when the external field is reduced to zero. Coercivity is the field required to bring the magnetization back to zero after saturation in the opposite direction. These two quantities are central measures of magnetic hysteresis.

4.2 Mechanical materials

Mechanical hysteresis occurs when a material’s deformation depends on loading history. Under repeated stress, the loading and unloading paths may differ because of internal friction, molecular rearrangement, or structural change.

4.2.1 Stress-strain behavior

In stress-strain testing, hysteresis appears as a loop between applied stress and resulting strain. The enclosed area indicates mechanical energy lost in the cycle. The loop shape can reveal stiffness changes, plasticity, and damage accumulation.

4.2.2 Viscoelasticity

Viscoelastic materials combine elastic and time-dependent behavior. They may recover part of a deformation quickly and the rest gradually, or not fully. Hysteresis in such materials often reflects both reversible storage and delayed dissipation.

4.3 Phase transitions

Some phase transitions exhibit hysteresis because the system remains in a previous state beyond the equilibrium transition point. This is often associated with nucleation barriers and metastable states.

4.3.1 Supercooling and superheating

Supercooling occurs when a liquid remains liquid below its normal freezing point, while superheating occurs when a liquid remains liquid above its normal boiling point. In both cases, the transition is delayed because the new phase has not yet nucleated. These delays are commonly linked to hysteretic behavior.

4.3.2 Metastability

Metastability refers to a state that is stable against small disturbances but not the most stable possible state. A system can remain metastable for a time, then shift suddenly when a threshold is crossed. This persistence contributes to history-dependent transitions.

4.4 Electrical systems

Electrical hysteresis appears in materials and circuits whose voltage, current, charge, or polarization depends on past excitation. The effect is important in both functional devices and circuit analysis.

4.4.1 Ferroelectric materials

Ferroelectric materials possess a spontaneous electric polarization that can be reversed by an electric field. Their polarization-field curves often show a loop similar to magnetic hysteresis. The behavior is used in memory devices and sensing applications.

4.4.2 Circuit responses

Some circuits exhibit hysteresis through threshold devices, feedback, or component lag. A common example is a switching circuit with different turn-on and turn-off points. Such behavior can stabilize signals, prevent noise-triggered switching, or create oscillation patterns.

5 Applications

Because hysteresis captures memory and threshold behavior, it is useful in design, modeling, and interpretation across many disciplines. In applied settings, understanding hysteresis can improve accuracy, durability, and control.

5.1 Engineering design

Engineers account for hysteresis when selecting materials, designing machines, and predicting fatigue or heat loss. Recognizing loop behavior helps prevent inefficiency and reduce unexpected wear. It also supports safer design under cyclic loading.

5.2 Data storage and memory devices

Hysteresis is central to several memory technologies, especially those that store information in stable states. A material or device can encode one of two or more states and retain the chosen state until an input threshold is crossed. This makes hysteresis useful for nonvolatile memory concepts.

5.3 Sensors and actuators

In sensors and actuators, hysteresis can be both useful and problematic. Some devices exploit it to create stable switching or precise actuation, while others must minimize it to improve measurement accuracy. Careful calibration is often required.

5.4 Control theory

Control systems often use hysteresis to avoid rapid toggling and to introduce deliberate switching thresholds. However, excessive hysteresis can complicate response prediction and stability analysis. Controllers may therefore include compensation methods to manage the effect.

5.5 Geophysics and climate science

In geophysics and climate science, hysteresis helps describe systems with long-term memory and delayed transitions. Examples include ice-albedo feedback, soil moisture response, and some rock or reservoir processes. The concept is valuable when present conditions are influenced by earlier environmental states.

6 Measurement and analysis

Studying hysteresis requires observing how a system responds over a range of inputs, often through repeated cycling. The quality of the data depends on experimental design, instrumental precision, and the ability to distinguish true hysteresis from artifacts.

6.1 Experimental methods

Experimental methods usually involve slowly varying the control variable and measuring the output during both forward and reverse sweeps. The procedure may be repeated at different rates, temperatures, or amplitudes to separate persistent hysteresis from transient effects. Careful stabilization is often necessary.

6.2 Curve fitting and parameter estimation

Researchers often fit observed loops to a mathematical model in order to estimate parameters. Good fits can reveal thresholds, saturation levels, relaxation times, and dissipation properties. Because multiple models may reproduce similar curves, interpretation should be tied to physical understanding as well as numerical accuracy.

6.3 Visualization of hysteresis loops

Visualization typically uses plots of output versus input over a full cycle. Multiple cycles may be overlaid to show repeatability, drift, or training effects. Clear labeling of direction, turning points, and units is important for interpretation.

6.4 Sources of error and uncertainty

Apparent hysteresis can arise from noise, calibration drift, temperature variation, friction, or incomplete equilibration. Sampling rate and instrument lag can also distort the loop shape. Distinguishing genuine material or system hysteresis from measurement artifacts is a key part of analysis.

Several concepts are closely related to hysteresis, though they are not identical. Some describe energy loss, some describe multiple stable states, and others address the mechanisms that create history dependence.

7.1 Resilience and damping

Resilience refers to a system’s ability to recover after deformation or disturbance, while damping describes the reduction of oscillation or motion through energy dissipation. Both are often discussed alongside hysteresis because all three involve response under changing conditions. However, hysteresis specifically emphasizes path dependence.

7.2 Bistability

Bistability is the presence of two stable states under the same external conditions. A bistable system may switch between states when thresholds are crossed, which can generate hysteresis loops. The two concepts often overlap in practical models.

7.3 Feedback

Feedback occurs when a system’s output influences its future input or internal evolution. Positive and negative feedback can create thresholds, delays, and switching behavior that resemble hysteresis. In many cases, feedback is one mechanism behind the observed loop.

7.4 Nonlinearity

Nonlinearity means that outputs are not proportional to inputs. Hysteresis is commonly found in nonlinear systems because internal structure, thresholds, and saturation prevent simple proportional response. While many nonlinear systems are not hysteretic, most hysteretic systems are nonlinear.

7.5 History dependence

History dependence is the broader idea that a system’s past affects its present state. Hysteresis is a specific form of history dependence characterized by loops, delayed response, or return-path differences. The two terms are often related, but hysteresis usually implies a more structured and measurable effect.

</INTERNAL_LINK_CANDIDATES> Ferromagnetism (a form of magnetism with strong domain alignment) Viscoelasticity (combined elastic and time-dependent mechanical behavior) Metastability (a long-lived but not most stable state) Preisach model (a classical threshold-based hysteresis model) Bouc-Wen model (a phenomenological model for cyclic hysteresis) Remanence (residual magnetization after removing the field) Coercivity (field needed to reduce magnetization to zero) Magnetization curve (plot of magnetic response versus field) Stress-strain curve (plot of mechanical response under loading) Phase transition (a change between physical states) Supercooling (cooling below the normal freezing point without crystallizing) Superheating (heating above the normal boiling point without phase change) Ferroelectric material (a material with switchable electric polarization) Feedback (a process where outputs influence future behavior) Bistability (the existence of two stable states) Damping (loss of motion through dissipation) Nonlinearity (a non-proportional input-output relationship) History dependence (dependence of current state on past states) Energy dissipation (conversion of input work into lost energy) Control theory (the study of regulating system behavior)