Answers are grounded in this paper and cite their sources.
The average properties of many turbulent flows are remarkably reproducible, yet the detailed flows are never the same. Here necessary conditions are found for the stability of a statistically steady state with ideal reproducibility. The examples explored in this paper are the symmetry breaking large scale flows observed in high Rayleigh number Boussinesq convection. A necessary condition found is that an integral function of convective and advective local averages must be a maximum, selecting among the possible solutions. Although these many solutions may not be determined in this century, an estimate from above of this maximum can be sought among the class of vector fields compatible with derived integral constraints and the boundary conditions. The condition for statistical stability found here permits a first determination of a unique relation between the large scale mean flows and the heat flux without the introduction of empirical parameters. It is found that statistical stability at large Rayleigh number is achieved by that solution with maximum momentum transport, and not by the solution with maximum heat transport. The past limited data supports this deduction that the large scale flow approaches its maximum possible value. New experiments, with a stress free boundary condition, are proposed to explore these and other unanticipated theoretical results.
No takes yet. Share an insight, caveat, or question.
Willem V. R. Malkus (1996) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: