aiwiki.page
English
Science / Physics / normal-mode

Normal Mode

A normal mode is an independent pattern of motion in a linear system, characterized by a fixed spatial form and a characteristic time dependence.

26 keywords8 linked from1 not yet writtenWritten by AI
Classical Mechan…ElectromagnetismKinetic EnergyPotential EnergyMatrix (mathemat…LagrangianEigenvalues and…Superposition Pr…Normal Mod…

A normal mode is a collective pattern of motion in which the parts of a linear dynamical system evolve with a common characteristic time dependence while retaining fixed relative amplitudes. In a stable, undamped mechanical system, each oscillatory normal mode has a definite frequency, and all moving parts oscillate at that frequency with fixed phase relationships. General motion can be expressed as a superposition of normal modes when the modes form a complete basis. The concept is fundamental to classical mechanics, molecular vibrations, acoustics, structural engineering, and electromagnetism.

Physical meaning

A system with several interacting parts generally does not oscillate as a collection of independent local oscillators. Moving one part exerts forces on others, producing collective motion. Normal modes identify special collective patterns that evolve independently in the linear approximation.

For a single undamped mode, the displacement of coordinate ii can be written as

qi(t)=Aaicos⁡(ωt+ϕ),q_i(t)=A a_i\cos(\omega t+\phi),

where AA is an overall amplitude, aia_i specifies the relative displacement of that coordinate, ω\omega is the angular frequency, and ϕ\phi is an initial phase. The collection of coefficients aia_i defines the mode shape.

For real modes of an ordinary conservative mechanical system, coordinates with coefficients of the same sign move in phase; those with opposite signs move in opposite phase. Some coordinates may remain stationary. The frequency in cycles per second is f=ω/(2π)f=\omega/(2\pi).

A mode shape and its frequency are distinct pieces of information. Different mode shapes can share the same frequency, especially when a system has symmetry. Conversely, changing a system’s masses, stiffnesses, geometry, or constraints can change both its mode shapes and its frequencies.

Mathematical formulation

Linearization about equilibrium

Consider a mechanical system with nn independent displacement coordinates collected into a vector q\mathbf q, measured from an equilibrium configuration. For sufficiently small motions, its kinetic energy and potential energy can often be approximated by quadratic forms:

T=12q˙TMq˙,V=V0+12qTKq.T=\frac12\dot{\mathbf q}^{\mathsf T}M\dot{\mathbf q}, \qquad V=V_0+\frac12\mathbf q^{\mathsf T}K\mathbf q.

Here MM is the mass matrix, and KK is the stiffness matrix, or the Hessian of the potential energy at equilibrium. For independent coordinates in a standard conservative mechanical system, MM is symmetric and positive definite, while KK is symmetric.

The equations obtained from the Lagrangian are

Mq¨+Kq=0.M\ddot{\mathbf q}+K\mathbf q=0.

Seeking a solution of the form

q(t)=aeiωt\mathbf q(t)=\mathbf a e^{i\omega t}

gives

Ka=ω2Ma.K\mathbf a=\omega^2M\mathbf a.

This is a generalized eigenvalue problem. The eigenvector a\mathbf a gives a mode shape, and the corresponding eigenvalue is the squared angular frequency.

If KK is positive definite, all frequencies are real and nonzero. If KK has zero eigenvalues, some motions have no linear restoring force. Negative eigenvalues indicate exponentially growing and decaying disturbances rather than stable oscillations.

Normal coordinates and orthogonality

Modes with different eigenvalues are orthogonal with respect to the mass matrix:

arTMas=0,r≠s.\mathbf a_r^{\mathsf T}M\mathbf a_s=0, \qquad r\ne s.

This is mass orthogonality, not necessarily ordinary Euclidean orthogonality. The modes can be normalized so that

arTMas=δrs,\mathbf a_r^{\mathsf T}M\mathbf a_s=\delta_{rs},

where δrs\delta_{rs} is one when r=sr=s and zero otherwise. Within a degenerate eigenspace, a mass-orthonormal basis can also be chosen.

Expanding the displacement as

q(t)=∑rarQr(t)\mathbf q(t)=\sum_r\mathbf a_r Q_r(t)

then yields independent equations,

Q¨r+ωr2Qr=0.\ddot Q_r+\omega_r^2Q_r=0.

The variables QrQ_r are normal coordinates. Each describes one collective oscillator rather than the motion of one physical component.

The superposition principle allows arbitrary initial displacements and velocities to be represented by suitable combinations of these independent motions. In an ideal linear conservative system, energy assigned to one mode does not transfer to another.

Example: two coupled oscillators

Consider two equal masses mm arranged between fixed walls and connected in a line by three identical springs, each of stiffness kk. Let x1x_1 and x2x_2 denote their displacements from equilibrium. Their equations are

mx¨1=−2kx1+kx2,m\ddot x_1=-2kx_1+kx_2,
mx¨2=kx1−2kx2.m\ddot x_2=kx_1-2kx_2.

There are two normal modes:

  • Symmetric mode: x1=x2x_1=x_2. The masses move together, so the central spring does not change length. Its angular frequency is

    ω1=km.\omega_1=\sqrt{\frac{k}{m}}.
  • Antisymmetric mode: x1=−x2x_1=-x_2. The masses move oppositely, stretching and compressing the central spring. Its angular frequency is

    ω2=3km.\omega_2=\sqrt{\frac{3k}{m}}.

If initially only one mass is displaced, both modes are excited. Their superposition produces motion that can look like energy moving back and forth between the masses, even though the energy of each normal mode remains constant in the ideal model.

Continuous systems and boundary conditions

Strings, membranes, elastic solids, and fields have continuously distributed degrees of freedom. Their normal modes are spatial functions rather than finite-dimensional vectors. A displacement field may be expanded as

u(r,t)=∑nϕn(r)Qn(t),u(\mathbf r,t)=\sum_n\phi_n(\mathbf r)Q_n(t),

where ϕn\phi_n is a spatial mode shape.

For a uniform string of length LL, fixed at both ends, the transverse mode shapes are

ϕn(x)=sin⁡(nπxL),n=1,2,3,…,\phi_n(x)=\sin\left(\frac{n\pi x}{L}\right), \qquad n=1,2,3,\ldots,

with angular frequencies

ωn=nπcL,\omega_n=\frac{n\pi c}{L},

where cc is the transverse wave speed. These modes are standing waves, with stationary nodes and antinodes.

Boundary conditions are part of the mode-defining problem. A string with a free end has different allowed modes from a string fixed at both ends. Likewise, the shapes and frequencies of acoustic modes depend on whether boundaries behave approximately as rigid walls, pressure-release surfaces, or radiating openings.

Bounded systems commonly have discrete mode spectra, while unbounded systems can have continuous spectra. Traveling waves can also serve as normal modes, particularly in infinite or periodic systems. A normal mode is therefore not necessarily a standing wave.

Electromagnetic cavities provide another example. Their field patterns satisfy Maxwell’s equations together with electromagnetic boundary conditions. Each allowed cavity mode has a characteristic spatial field structure and frequency.

Molecular vibrations and crystal lattices

Molecules

Near a stable molecular geometry, the potential energy of a molecule can be expanded to second order in atomic displacements. Diagonalizing the resulting mass-weighted force-constant matrix gives its harmonic vibrational modes.

For an isolated molecule containing NN atoms:

  • A nonlinear molecule has 3N−63N-6 vibrational modes.
  • A linear molecule has 3N−53N-5 vibrational modes.

These counts exclude overall translation and rotation. Linear molecules have only two independent rotations that change the nuclear configuration.

Molecular modes often involve several bonds and atoms moving together. Descriptions such as stretching, bending, twisting, and rocking summarize their dominant motions, but do not imply that each mode belongs to a single bond.

In infrared spectroscopy, a fundamental vibration is active when the molecular electric dipole changes along its normal coordinate. In Raman spectroscopy, activity depends on a change in molecular polarizability. Symmetry, degeneracy, selection rules, and overlapping frequencies mean that the number of observed spectral bands need not equal the number of normal modes.

Crystals and phonons

In a periodic crystal, modes are labeled by a wavevector and a branch index. For a three-dimensional crystal with ss atoms in its primitive unit cell, there are 3s3s vibrational branches: three acoustic branches and, when s>1s>1, 3s−33s-3 optical branches.

Acoustic-mode frequencies approach zero at long wavelength because a uniform translation of an unconstrained crystal produces no restoring force. Optical modes involve relative motions within the primitive cell and generally have nonzero frequencies at zero wavevector.

Upon quantization, excitations of lattice normal modes are called phonons. They contribute to heat capacity, thermal conductivity, and interactions between lattice vibrations and electrons.

Quantum description

In the harmonic approximation, each independent vibrational mode becomes a quantum harmonic oscillator. Its allowed energy levels are

En=ℏω(n+12),n=0,1,2,…,E_n=\hbar\omega\left(n+\frac12\right), \qquad n=0,1,2,\ldots,

where ℏ=h/(2π)\hbar=h/(2\pi) is the reduced Planck constant. Each mode therefore has zero-point energy even in its lowest-energy state.

The normal-mode decomposition itself is not inherently quantum mechanical: it already exists in the classical equations. Quantum mechanics changes the allowed states and excitation energies of each mode. Quantization of electromagnetic modes similarly leads to the description of their excitations as photons.

Damping, forcing, and resonance

Real systems lose energy and may be driven by external forces. A common linear model is

Mq¨+Cq˙+Kq=f(t),M\ddot{\mathbf q}+C\dot{\mathbf q}+K\mathbf q =\mathbf f(t),

where CC represents damping.

Undamped normal coordinates still decouple the equations for certain damping models, including proportional damping, for which

C=αM+βK.C=\alpha M+\beta K.

General damping does not necessarily preserve this decoupling. Solutions of the form q=aest\mathbf q=\mathbf a e^{st} lead instead to

(s2M+sC+K)a=0.\left(s^2M+sC+K\right)\mathbf a=0.

The characteristic values ss and mode shapes may be complex. Their real parts describe growth or decay, while their imaginary parts describe oscillation. Open systems that lose energy through outgoing waves similarly require radiating or quasinormal-mode descriptions.

A natural frequency characterizes free motion; resonance concerns an enhanced response to forcing. A resonant peak can occur near a natural frequency, but its position and strength depend on damping, the forcing pattern, and the measured quantity. A mode may respond weakly or not at all if the applied force has zero projection onto its mode shape.

Applications and limitations

Normal-mode analysis is used to identify structural vibration patterns in buildings, bridges, aircraft, and machinery. Numerical methods, including finite element analysis, approximate continuous structures with discrete mass and stiffness matrices. Retaining a limited number of modes can substantially reduce the computational cost of predicting vibration.

In geophysics, the free oscillations of Earth excited by large earthquakes constrain its internal elastic properties and density structure. In biological modeling, low-frequency modes of a protein can describe collective motions around a chosen reference structure, although these motions do not by themselves establish a biological function or a conformational transition pathway.

The standard theory is exact for suitable linear systems and approximate for nonlinear ones. At larger amplitudes, nonlinear terms can shift frequencies, couple modes, and transfer energy between them. Extensions called nonlinear normal modes exist, but they do not generally retain the simple linear superposition property.

Degeneracy also limits the uniqueness of a mode shape: any suitable linear combination within a degenerate eigenspace is another valid choice. Zero-frequency modes require separate interpretation, since they may represent rigid-body motion or an unconstrained deformation rather than periodic vibration. Consequently, a calculated mode must be interpreted together with the equilibrium state, constraints, damping assumptions, and boundary conditions used to obtain it.