Two cylinders, one gap
Fill the narrow annular gap between two concentric cylinders with fluid, and spin one or both of them. At low rotation rates the fluid simply follows along in smooth, purely azimuthal layers — an exact analytic solution called circular Couette flow, with velocity varying smoothly from the inner cylinder's speed to the outer one's. Nothing crosses the gap radially; each thin shell of fluid just slides past its neighbors.
Centrifugal instability and the Taylor number
Spin the inner cylinder fast enough, though, and this smooth layered flow stops being stable. Lord Rayleigh had already shown, for an ideal inviscid fluid, that rotating flow is unstable wherever the specific angular momentum decreases outward — a fluid parcel displaced outward then finds itself spinning faster than its new surroundings require for equilibrium, and centrifugal force flings it further out. G. I. Taylor's landmark 1923 paper added viscosity to the analysis, producing a precise criterion for exactly when the real, damped flow becomes unstable, expressed through the dimensionless Taylor number.
Ta = (Ω₁² · r₁ · d³) / ν² (narrow-gap form) Ω₁ = angular velocity of the inner cylinder r₁ = radius of the inner cylinder d = gap width (r₂ − r₁) ν = kinematic viscosity of the fluid critical value, narrow gap, outer cylinder fixed: Ta_c ≈ 1708
It is a striking coincidence — not a deep connection, just a shared narrow-gap approximation — that this critical value comes out numerically close to the Rayleigh-Bénard critical Rayleigh number of 1708; the two instabilities are driven by entirely different physics.
Taylor vortices: stacked donuts
Above the threshold, the flow reorganizes into a neat stack of counter-rotating toroidal vortices running around the gap, each one roughly as tall as the gap is wide. Fluid spirals outward at the top of one vortex and inward at the top of its neighbor, giving the classic banded pattern visible when the flow is seeded with reflective flakes. Taylor's 1923 paper is remembered as one of the first cases in fluid mechanics where a quantitative linear stability prediction was confirmed by experiment to remarkable precision — a landmark not just for this specific flow but for the whole method of hydrodynamic stability analysis.
The cascade to turbulence
Keep increasing the rotation rate and the flow passes through a well-documented sequence of increasingly complex states: steady Taylor vortices give way to wavy vortex flow, where azimuthal waves travel around the vortex pattern; then to modulated wavy vortices, with two incommensurate frequencies (quasiperiodic motion); then to weakly chaotic flow; and finally to featureless turbulence in which the vortex structure is washed out entirely. Because each of these transitions is reproducible and can be studied one at a time, Taylor-Couette flow — alongside Rayleigh-Bénard convection — became one of the two central experimental testbeds for the dynamical-systems picture of how order breaks down into turbulence (period-doubling and quasiperiodic routes to chaos).
Why it matters beyond the lab
Taylor-Couette geometry shows up constantly in engineering: it is the natural model for lubrication flow in journal bearings, and rotational viscometers use exactly this setup, inferring a fluid's viscosity from the torque needed to spin the inner cylinder at a given rate. Chemical engineers deliberately run reactors in the Taylor vortex regime because each toroidal cell mixes its contents thoroughly while exchanging relatively little fluid with its neighbors, giving excellent local mixing with a controllable, near-uniform residence time — useful for polymerization and crystallization processes where ordinary turbulent mixing is too uncontrolled.
Frequently asked questions
What is the Taylor number and what does it represent?
It is the ratio of centrifugal destabilizing forces to viscous damping in a rotating curved flow, playing a role similar to a Reynolds number. Above a critical Taylor number of roughly 1700 for a narrow gap with the inner cylinder rotating, viscosity can no longer damp out small perturbations and Taylor vortices appear.
Why do only some rotation combinations become unstable?
Rayleigh's criterion says the flow is centrifugally unstable only where the specific angular momentum decreases outward. Rotating just the inner cylinder satisfies that condition and becomes unstable above a critical Taylor number. Rotating only the outer cylinder is stable at any speed in the inviscid limit, since angular momentum then increases outward; counter-rotating cylinders give a mixed picture with an unstable inner region and a stable outer one.
How is Taylor-Couette flow different from Rayleigh-Bénard convection?
Both are pattern-forming instabilities driven by a control parameter crossing a critical threshold, and both transition to turbulence through a similar sequence of increasingly complex states. Physically, Rayleigh-Bénard convection is driven by buoyancy from a temperature gradient, while Taylor-Couette flow is driven by the centrifugal effect of curved shear — but the mathematics of their linear stability analysis and bifurcation sequences is close enough that the two are usually taught side by side.
Try it live
Everything above runs in your browser — open Taylor-Couette Flow and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.
▶ Open Taylor-Couette Flow simulation