Turbulent Kinetic Energy: Production and Dissipation

Core Idea

In turbulent flows, mean kinetic energy is continuously converted to turbulent kinetic energy by large-scale eddies (production), which cascade to progressively smaller scales and ultimately dissipate as heat through viscous action at Kolmogorov scales. This energy cascade explains why turbulent flows are irreversible and dissipate mechanical energy as heat far more efficiently than laminar flows, making understanding turbulence essential for minimizing pumping power.

Explainer

From your study of turbulent pipe flow, you know that turbulence is characterized by chaotic, three-dimensional velocity fluctuations superimposed on the mean flow. Reynolds decomposition separates these: u = Ū + u', where Ū is the time-averaged velocity and u' is the fluctuating component. The product of these fluctuations — terms like ρ·u'v' — gives rise to the Reynolds stresses that are responsible for the dramatically higher friction factors in turbulent flow compared to laminar. But where does this turbulent agitation come from, and where does it go? The answer is the turbulent kinetic energy budget: k = ½(u'² + v'² + w'²), the kinetic energy stored in velocity fluctuations per unit mass.

Production is the source term. The mean flow gradient (dŪ/dy near a wall, for example) acts on the Reynolds stresses to continuously extract energy from the organized mean motion and inject it into turbulent fluctuations. Physically, this is the mechanism by which shear layers become unstable: the mean velocity gradient is the engine that keeps turbulence alive against dissipation. In a fully developed pipe flow, this production is highest near the wall where the velocity gradient is steepest. Without a mean velocity gradient to sustain it, turbulence would decay — this is exactly what happens in grid turbulence experiments where flow passes through a mesh and then decelerates into a uniform mean flow, causing turbulence intensity to decay downstream.

The produced turbulent energy does not dissipate immediately. Instead, it undergoes an energy cascade: large-scale eddies — whose size is set by the geometry of the flow (pipe diameter, shear layer thickness) — break up into progressively smaller eddies through nonlinear inertial interactions. The cascade is a one-way energy transfer from large to small scales; it is not a symmetric process. At each scale, eddies are unstable and break apart, feeding their energy to smaller structures. This continues until eddies reach the Kolmogorov microscale η = (ν³/ε)^1/4, where ν is kinematic viscosity and ε is the dissipation rate per unit mass. At this scale, viscous forces dominate over inertial forces — the local Reynolds number is of order unity — and the eddy's kinetic energy is irreversibly converted to heat.

The ratio of the largest turbulent scale (integral scale L, roughly the pipe radius or boundary layer thickness) to the Kolmogorov scale scales as L/η ~ Re^3/4. This means that at Re = 10⁶, Kolmogorov eddies are roughly 10^4.5 times smaller than the energy-containing eddies. Directly simulating all these scales simultaneously (Direct Numerical Simulation) requires computational grids scaling as Re^9/4 — which is why turbulence modeling (k-ε, k-ω, etc.) is necessary for engineering calculations. These models add transport equations for k and ε (or related quantities) to the mean-flow equations, replacing the unresolved small-scale physics with empirical closure relations. The fundamental structure of the energy cascade — production at large scales, dissipation at small scales, conservative transfer between — is the physical justification for why these two-equation models work at all.