1 Definition and physical meaning

Logarithmic decrement is a measure of how rapidly the amplitude of oscillations decreases in a damped system. It is defined using the ratio of two successive peak amplitudes (or peaks separated by a fixed number of cycles). The result compresses the decay behavior into a single number, making it convenient for comparing damping strength between systems.

1.1 Successive peak amplitude comparison

Consider a free, underdamped oscillation where successive local maxima decrease in magnitude. Let \(x_1\) be the amplitude at one peak and \(x_2\) the amplitude at the next peak after one period. The logarithmic decrement \(\delta\) is based on the natural logarithm of the ratio: \[ \delta=\ln\!\left(\frac{x_1}{x_2}\right). \] If the oscillation decays, then \(x_1>x_2\) and \(\delta\) is positive under the usual sign convention.

1.2 Relationship to decay of energy

For many linear damping models, the oscillation envelope decays exponentially. Because the mechanical (or electrical) energy associated with oscillation is proportional to the square of the amplitude, the logarithmic decrement provides an indirect but useful link to energy dissipation. Larger \(\delta\) corresponds to faster loss of stored energy from cycle to cycle.

1.3 Units, sign conventions, and interpretation

Logarithmic decrement is dimensionless: it is a logarithm of an amplitude ratio. In typical experimental usage for decaying oscillations, \(\delta>0\). If measurements produce \(x_2>x_1\), the computed \(\delta\) becomes negative; this indicates either amplification, sign/configuration errors, or operation outside the expected free-decay regime. The magnitude of \(\delta\) is interpreted as the “per-cycle” decay strength.

2 Damped oscillatory motion background

Logarithmic decrement is most natural in the context of free vibration of linear, underdamped systems, where the motion remains oscillatory but its envelope steadily contracts.

2.1 Free vs. forced vibration context

In free vibration, the system is displaced or released and then allowed to respond without sustained external driving. The resulting motion typically follows a decaying sinusoid, allowing peak-to-peak comparison. In forced vibration, steady-state oscillations may reach constant amplitude or exhibit phase-dependent behavior; peak ratios there are often governed by the forcing and resonance rather than by damping alone.

2.2 Under-, critically, and over-damped regimes

Damping determines whether the system oscillates:

  • Under-damped: oscillations occur with a decaying envelope; logarithmic decrement is well-defined using successive peaks.
  • Critically damped: the system returns to equilibrium as fast as possible without oscillating; peak-based decrement is not applicable because distinct oscillation maxima may not exist.
  • Over-damped: the return is monotonic and oscillations are absent; logarithmic decrement in the usual sense cannot be extracted from oscillatory peaks.

2.3 Harmonic motion model (linear viscous damping)

A common reference model is a mass–spring system with linear viscous damping, producing solutions of the form \[ x(t)=A e^{-\beta t}\cos(\omega_d t+\phi), \] where \(A\) and \(\phi\) are set by initial conditions, \(\beta\) controls the exponential decay of the envelope, and \(\omega_d\) is the damped oscillation frequency. In this setting, peak amplitudes follow \(x_{\text{peak}}(t)\propto e^{-\beta t}\), which is why logarithmic decrement becomes constant across cycles.

3 Mathematical formulation

The mathematical form of logarithmic decrement follows directly from the exponential decay of the oscillation envelope in linear underdamped motion.

3.1 Standard logarithmic decrement equation

For peaks separated by \(n\) cycles, let \(x_k\) and \(x_{k+n}\) denote amplitudes at those peaks. The decrement is generalized to \[ \delta_n=\ln\!\left(\frac{x_k}{x_{k+n}}\right). \] The case \(n=1\) yields the “per-cycle” decrement. For exponentially decaying envelopes, \(\delta_n\) grows linearly with \(n\), i.e., \(\delta_n=n\,\delta_1\).

3.2 Derivation from damped sinusoid solutions

Using \(x_{\text{peak}}(t)\propto e^{-\beta t}\), the amplitude at a time corresponding to the \(k\)-th peak is proportional to \(e^{-\beta t_k}\). If successive peaks occur one damped period \(T_d\) apart, then \(t_{k+1}=t_k+T_d\). Therefore, \[ \delta_1=\ln\!\left(\frac{e^{-\beta t_k}}{e^{-\beta (t_k+T_d)}}\right) =\ln\!\left(e^{\beta T_d}\right)=\beta T_d. \] This shows that, within the linear viscous model, logarithmic decrement is determined by the decay rate \(\beta\) and the damped period \(T_d\).

3.3 Expressing decrement in terms of time and period

For peaks separated by \(nT_d\), \[ \delta_n=\beta (nT_d)=n\,\beta T_d=n\,\delta_1. \] Equivalently, if the oscillation is monitored continuously, the envelope decay factor over a time interval \(\Delta t\) can be related to logarithmic decrement via \(\delta = \beta T_d\), with \(T_d\) connecting the time scale to the oscillation cycle count.

Logarithmic decrement is frequently used as an intermediate quantity because many engineering descriptions of damping employ other parameters.

4.1 Conversion to damping ratio (ζ)

For a single-degree-of-freedom linear underdamped oscillator, the damping ratio \(\zeta\) relates to \(\delta\) through \[ \zeta=\frac{\delta}{\sqrt{4\pi^2+\delta^2}}. \] This conversion assumes the same class of linear underdamped dynamics that yields a constant decrement across cycles.

4.2 Connection with quality factor (Q)

The quality factor \(Q\) is another dimensionless characterization, commonly used in resonance and oscillatory systems. For lightly damped underdamped motion, logarithmic decrement and quality factor are connected by \[ Q=\frac{2\pi}{\delta}. \] This relation follows from the idea that \(Q\) measures the number of oscillation cycles over which energy decays substantially.

4.3 Approximations for small damping

When damping is small, \(\delta\ll 2\pi\), leading to simpler approximations: \[ \zeta \approx \frac{\delta}{2\pi}, \qquad Q \approx \frac{2\pi}{\delta}. \] These approximations are useful for quick estimates and for interpreting experimental results where damping is weak and the measured decrement is small.

5 Measurement and experimental determination

In practice, logarithmic decrement is obtained from measured peak amplitudes during free decay.

5.1 Practical method using measured peak amplitudes

A typical procedure is:

  1. Initiate free vibration (e.g., release a mass, perturb a structure, excite then disconnect).
  2. Record the response \(x(t)\) over several cycles.
  3. Identify local maxima and measure their magnitudes \(x_1, x_2, \ldots\).
  4. Compute \(\delta=\ln(x_1/x_2)\) for one-cycle spacing, or \(\delta_n=\ln(x_k/x_{k+n})\) using a larger separation to reduce sensitivity to peak-picking errors.

5.2 Choosing the index between peaks

Selecting \(n\) involves a trade-off:

  • Using \(n=1\) uses adjacent peaks, maximizing temporal resolution but making results sensitive to noise and transient measurement errors.
  • Using larger \(n\) increases the decay contrast between peaks, improving robustness if data quality is adequate over the longer span.

Because \(\delta_n=n\,\delta_1\) holds for ideal exponential decay, using multiple peak pairs can also serve as an internal consistency check.

5.3 Dealing with noise and drift in amplitude measurements

Measured amplitudes can be affected by sensor noise, baseline drift, and changes in system conditions over time. Common mitigation strategies include:

  • using signal conditioning or filtering to improve peak detection without distorting amplitude,
  • estimating the oscillation envelope via fitting rather than selecting single-point maxima,
  • discarding early peaks if the system has not entered pure free-decay behavior,
  • checking whether the computed decrement remains stable across successive peak pairs.

6 Applications across disciplines

Logarithmic decrement appears wherever decaying oscillations are analyzed, particularly for linear underdamped dynamics.

6.1 Mechanical systems and structural dynamics

In structural dynamics, logarithmic decrement helps characterize damping in beams, frames, and support systems. By analyzing the free vibration following a disturbance, engineers can estimate damping parameters used in vibration models, stability analyses, and design comparisons.

6.2 Vibrating strings, beams, and resonators

For systems approximated by single or multiple vibration modes, decrement analysis can quantify modal damping. Experiments may focus on a dominant mode by choosing an observation location and excitation method that isolate that mode’s oscillatory component.

6.3 Electrical circuits (RLC oscillations)

In RLC circuits with resistance providing damping, the current or voltage can exhibit decaying oscillations. Peak measurements in time-domain waveforms can be converted into logarithmic decrement, enabling the extraction of effective damping and resistance-related parameters.

6.4 Material damping characterization

Material damping often manifests as internal friction that produces an effective reduction in oscillation amplitude. By testing specimens in resonance or free-decay experiments and computing decrement, researchers can compare damping across materials, treatments, temperatures, or loading conditions—while recognizing that real materials may show non-ideal, amplitude- or frequency-dependent behavior.

7 Data analysis and uncertainty

Accurate decrement estimation requires attention to both measurement uncertainty and model mismatch.

7.1 Propagating measurement uncertainty

Let \(x_1\) and \(x_2\) be measured peak amplitudes with uncertainties \(u_{x_1}\) and \(u_{x_2}\). Since \(\delta=\ln(x_1/x_2)\), uncertainty propagation yields an uncertainty that increases when peak amplitudes are small or when their ratio is close to one. In practice, this motivates collecting enough cycles so the decay is detectable while avoiding regions dominated by noise.

7.2 Regression approaches using multiple peaks

Instead of relying on a single peak pair, one can fit the logarithm of the envelope: \[ \ln(x_{\text{peak}})=\ln A-\beta t. \] A linear regression of \(\ln(x_{\text{peak}})\) versus time estimates \(\beta\), after which \(\delta=\beta T_d\) is obtained. This approach typically improves statistical robustness and naturally incorporates multiple peak measurements.

7.3 Detecting non-ideal damping behavior

Departures from constant decrement indicate that the damping model may not match the data. Indicators include:

  • decrement varying with time or with the choice of peak pairs,
  • systematic curvature in \(\ln(x_{\text{peak}})\) versus time,
  • differences between decrement estimated from multiple frequency components.

Such patterns can suggest nonlinear damping, changing environmental conditions, or modal coupling.

8 Special cases and limitations

Logarithmic decrement is straightforward under ideal assumptions, but several common complications limit its direct applicability.

8.1 Nonlinear damping effects

If damping depends on amplitude or velocity in a nonlinear manner, the envelope may not follow a single exponential law. In that case, the ratio of successive peaks may change over time, producing a decrement that is not constant. One may need alternative models (e.g., amplitude-dependent decay laws) or local estimates over restricted amplitude ranges.

8.2 Time-varying damping and environment changes

External factors such as temperature drift, changing boundary conditions, or sensor gain changes can alter the effective damping during the experiment. Even with linear dynamics, such time dependence can make \(\delta\) appear to vary across the record. Monitoring experimental conditions and using segmented analyses can help identify these effects.

8.3 When logarithmic decrement is not constant

Constant decrement relies on exponential decay with a fixed decay rate. If the measured response shows different decay phases—such as an initial transient followed by a later steady regime—then a single decrement value may not represent the entire dataset. In such cases, reporting decrement per regime or using a model-based envelope fit can be more appropriate.

8.4 Sensitivity to the choice of oscillation peaks

The decrement depends on selecting maxima consistently. Peak-picking errors, missed peaks, or incorrect identification of the oscillatory component can bias results, especially when damping is weak and peaks are close in magnitude. Increasing measurement length, using robust peak detection algorithms, and comparing multiple \(n\) values can reduce this sensitivity.