HomeArticlesFluid Dynamics

Rayleigh-Taylor Instability: When Heavy Sits on Light

Balance a dense fluid on top of a lighter one and the flat interface is unstable to any perturbation: small ripples grow into the mushroom-shaped plumes that drive mixing from supernovae to inertial confinement fusion.

mysimulator teamUpdated June 2026≈ 8 min read▶ Open the simulation

An interface that cannot stay flat

Stack a denser fluid of density ρ₂ directly on top of a lighter fluid of density ρ₁, with gravity pulling straight down. Intuitively this configuration wants to overturn — the flat interface between them sits at a local energy maximum, not a minimum, so any infinitesimal ripple releases potential energy and grows. The same instability appears whenever a dense fluid is accelerated toward a lighter one, even without gravity — an imploding shell in a fusion capsule, or a decelerating supernova shockwave, both count. The strength of the effect is captured by the Atwood number, A = (ρ₂ − ρ₁)/(ρ₂ + ρ₁), which ranges from 0 (matched densities, no instability at all) to 1 (dense fluid over a vacuum).

Growth rate and the dispersion relation

Linear stability analysis of an ideal, inviscid interface gives a clean result: a sinusoidal ripple of wavenumber k grows exponentially in time, with a growth rate that increases with both the Atwood number and the wavenumber.

γ(k) = √(A · g · k)      (ideal inviscid theory, surface tension neglected)

A = Atwood number = (ρ₂ − ρ₁) / (ρ₂ + ρ₁)
g = acceleration (gravity, or the deceleration of an interface)
k = perturbation wavenumber = 2π / wavelength

surface tension σ cuts off growth above a critical wavenumber:
k_c = √(g(ρ₂ − ρ₁) / σ)
live demo · a dense layer collapsing into mushroom plumes● LIVE

In the idealized inviscid theory, shorter wavelengths grow fastest without any limit, which is unphysical; real fluids are rescued by viscosity and surface tension, both of which suppress growth at the smallest scales and select a finite fastest-growing wavelength that sets the initial spacing of the plumes.

Mushroom plumes and the mixing layer

Once the perturbation grows large enough that linear theory breaks down, the interface develops its signature shapes: spikes of heavy fluid plunge downward into the light fluid, and bubbles of light fluid rise upward into the heavy fluid. Both develop the classic mushroom-cap silhouette, because the shear at the edge of a falling spike or rising bubble is itself unstable to the Kelvin-Helmholtz instability, which rolls the edges into a pair of counter-rotating vortices. Eventually, neighboring plumes interact and the whole interface breaks down into a turbulent mixing layer whose width grows self-similarly with time, roughly as h ≈ α · A · g · t², with the empirical prefactor α measured at somewhere around 0.02 to 0.07 depending on the initial conditions — a number that has been the subject of decades of careful experiments and simulations.

From supernovae to inertial confinement fusion

In a core-collapse supernova, the outgoing shockwave decelerates as it crosses the star's density gradients, and every deceleration through a density jump is an opportunity for Rayleigh-Taylor instability to mix heavy, freshly synthesized elements from deep in the star outward into the expanding envelope — observations of mixed elements in supernova remnants are a direct fingerprint of this process. The instability's most consequential modern role, though, is as the chief obstacle to inertial confinement fusion: when lasers ablate the surface of a fuel capsule to implode it, the accelerating ablation front is itself an interface between light, blown-off plasma and denser shell material, and Rayleigh-Taylor growth there disrupts the implosion's spherical symmetry and mixes cold shell material into the hot fusion fuel — the single biggest engineering challenge standing between current experiments and ignition.

A close cousin: the Richtmyer-Meshkov instability

A related instability, the Richtmyer-Meshkov instability, occurs when the interface between two fluids of different density is struck by a single passing shockwave rather than a sustained acceleration — an impulsive kick instead of a continuous push. It produces broadly similar mushroom-shaped mixing structures and is the dominant instability mechanism in the earliest, shock-crossing phase of an ICF implosion, before the sustained Rayleigh-Taylor growth of the deceleration phase takes over.

Frequently asked questions

What is the Atwood number and why does it matter?

The Atwood number A = (ρ_heavy − ρ_light) / (ρ_heavy + ρ_light) measures the density contrast, ranging from 0 (matched densities, no instability) to 1 (fluid over vacuum). It sets the linear growth rate directly, since the growth rate scales with the square root of A, so higher-A systems develop the instability faster and reach the nonlinear mushroom stage sooner.

Why do the plumes form mushroom caps instead of just spikes?

As a spike of heavy fluid falls into the light fluid, the two are sliding past each other at their shared boundary, which is itself unstable to the Kelvin-Helmholtz shear instability. That secondary instability rolls the edges of the spike into a pair of counter-rotating vortices, producing the classic mushroom cap.

Does surface tension stop Rayleigh-Taylor instability?

It can, for shallow density inversions. Surface tension stabilizes short-wavelength perturbations, so there is a critical wavelength below which ripples do not grow. For most large-scale flows — astrophysical plasmas, big containers — gravity dominates and surface tension is negligible, but in small experiments, such as a drop of dense dyed water on a light layer, the cutoff is easy to see.

Try it live

Everything above runs in your browser — open Rayleigh-Taylor Instability and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.

▶ Open Rayleigh-Taylor Instability simulation

What did you find?

Add reproduction steps (optional)