Glossary term

Rayleigh Damping

Proportional damping model that represents the damping matrix as a weighted sum of the mass and stiffness matrices.

Branch: Mechanical Engineering
Glossary type: model
Content: Glossary term
Updated: Jun 24, 2026
Revision: v1.0.0 · reviewed

Definition

model

Rayleigh damping is a proportional viscous damping model in which the damping matrix is expressed as a linear combination of the mass and stiffness matrices.

Rayleigh damping approximates modal damping in a finite element or lumped-parameter dynamic model by setting C = alpha M + beta K. The mass-proportional coefficient mainly affects low-frequency modes, while the stiffness-proportional coefficient mainly affects high-frequency modes. Engineers use it when a dynamic analysis needs a practical damping matrix, but the assumed coefficients must be calibrated and checked over the frequency range that matters.

Rayleigh damping is a proportional viscous damping model used in finite element and modal dynamic analysis. It defines the damping matrix as:

C=\alpha M+\beta K

where $C$ is the damping matrix, $M$ is the mass matrix, $K$ is the stiffness matrix, $\alpha$ is the mass-proportional coefficient and $\beta$ is the stiffness-proportional coefficient.

For an undamped mode with circular natural frequency $\omega_r$ , the equivalent modal damping ratio is:

\displaystyle \zeta_r=\frac{1}{2}\left(\frac{\alpha}{\omega_r}+\beta\omega_r\right)

This equation shows why Rayleigh damping is useful but easy to misuse. The $\alpha/\omega_r$ term contributes more damping at low frequency. The $\beta\omega_r$ term contributes more damping at high frequency. Between the fitted frequencies the damping may be lower than either target value; above the fitted range it can become artificially high.

Engineering Role

Rayleigh damping gives a dynamic finite element model a damping matrix without measuring or modeling every physical dissipation mechanism. It is common in transient response, harmonic response, seismic analysis, structural qualification, support motion studies and preliminary aeroelastic or vibration screening.

The model is useful when damping data are sparse and the response band is limited. It is weak when the real damping comes from friction, joints, elastomers, material hysteresis, fluid films, nonlinear contact or active control. In those cases, a simple proportional damping matrix may reproduce one response metric while misrepresenting phase, decay rate, resonance amplitude or high-frequency content.

A defensible Rayleigh damping choice is tied to the engineering decision. The fitted frequencies should bracket the modes that drive the response, not arbitrary low modes selected because they are easy to extract. If the analysis depends on local stress, fatigue, payload equipment response or flutter margin, the damping ratio should be checked mode by mode over the relevant band.

Worked Example: Fit Two Target Damping Ratios

A structural model is used for vibration response between the first global mode and a higher equipment-support mode. The analyst wants:

\zeta_1=2.0\% \quad \text{at} \quad f_1=5\ \text{Hz}

and:

\zeta_2=2.0\% \quad \text{at} \quad f_2=30\ \text{Hz}

Convert to circular frequency:

\omega_1=2\pi(5)=31.416\ \text{rad/s}

\omega_2=2\pi(30)=188.496\ \text{rad/s}

For equal target damping $\zeta=0.02$ , the Rayleigh coefficients are:

\displaystyle \alpha=\frac{2\zeta\omega_1\omega_2}{\omega_1+\omega_2}

\displaystyle \beta=\frac{2\zeta}{\omega_1+\omega_2}

Substitute the values:

\displaystyle \alpha=\frac{2(0.02)(31.416)(188.496)}{31.416+188.496}=1.077\ \text{s}^{-1}

\displaystyle \beta=\frac{2(0.02)}{31.416+188.496}=1.82\times10^{-4}\ \text{s}

Check an intermediate mode at $10\ \text{Hz}$ :

\omega=2\pi(10)=62.832\ \text{rad/s}

\displaystyle \zeta=\frac{1}{2}\left(\frac{1.077}{62.832}+(1.82\times10^{-4})(62.832)\right)=0.0143=1.43\%

Check a high mode at $80\ \text{Hz}$ :

\omega=2\pi(80)=502.655\ \text{rad/s}

\displaystyle \zeta=\frac{1}{2}\left(\frac{1.077}{502.655}+(1.82\times10^{-4})(502.655)\right)=0.0468=4.68\%

Engineering comment: the fitted modes have 2 percent damping, but a 10 Hz mode has only about 1.43 percent and an 80 Hz mode has about 4.68 percent. That may be acceptable if the response of interest is concentrated between 5 and 30 Hz and the higher modes are not decision-critical. It is not acceptable to claim “2 percent damping everywhere” from these coefficients.

Rayleigh damping is not damping ratio. Damping ratio is the modal quantity being approximated; Rayleigh damping is one model used to generate a damping matrix and implied modal damping ratios.

Rayleigh damping is not modal analysis. Modal analysis identifies frequencies, mode shapes and damping; Rayleigh damping supplies assumed damping for a numerical dynamic model.

Rayleigh damping is not a material property. Its coefficients are model parameters that depend on the selected mass matrix, stiffness matrix, boundary conditions, frequency range and calibration targets.

Rayleigh damping is not proof of physical dissipation. A model can match a decay rate while using coefficients that have no credible connection to joints, materials, supports or measured test data.

Rayleigh damping is not always appropriate for nonlinear dynamics. When stiffness changes during contact, yielding, cracking or large deformation, the stiffness-proportional term can introduce damping that changes with the tangent or current stiffness convention used by the solver.

Validation and Common Mistakes

A defensible Rayleigh damping specification states $\alpha$ , $\beta$ , the fitted frequencies, target damping ratios, mass and stiffness matrix conventions, boundary conditions, included modes, frequency range of interest and whether the coefficients are based on test evidence or engineering assumption.

Common mistakes include:

fitting two frequencies and then reporting one constant damping ratio across all modes;
fitting modes outside the response band that drives the engineering decision;
overdamping high-frequency response with an excessive stiffness-proportional term;
applying the same coefficients after boundary conditions or stiffness modeling have changed;
using Rayleigh damping to hide poor correlation between test and finite element results;
combining physical dampers, modal damping and Rayleigh damping without checking double counting;
accepting a stable-looking transient response without checking whether the damping ratios are plausible mode by mode.

REF

Disciplines

Rayleigh Damping

Definition

Engineering Role

Worked Example: Fit Two Target Damping Ratios

Validation and Common Mistakes

See also

Engineering Role

Worked Example: Fit Two Target Damping Ratios

Distinction from Related Terms

Validation and Common Mistakes