Answers are grounded in this paper and cite their sources.
We investigate small changes to a two‐dimensional turbulent boundary layer on a rough surface caused by arbitrary variations, in the streamwise direction, of the surface roughness length, z 1 ( x ). the linear changes to the flow are calculated as asymptotic sequences in the limit ϵ = u * / U 0 → 0 ( u * is the upwind friction velocity, and U 0 is a typical value of the approach wind speed). the flow is divided into two regions, each of which is subdivided into two layers; the previous analyses may then be extended by calculating the second‐order effects. Throughout the bulk of the inner region, of thickness l , the Reynolds shear‐stress gradient balances the acceleration, but very close to the surface (in the inner surface layer), up to a height O (√ lz 0 ), the shear‐stress gradient is constant and matches with its surface value. the outer region (where the leading‐order perturbations are inviscid) also divides into two layers: the middle layer extending to a height h m , where the shear in the upwind profile determines the perturbations; and the upper layer, in which the perturbations decay in potential flow, and a weak perturbation pressure develops which produces second‐order changes in the inner region. At second order the normal Reynolds stresses also accelerate the mean‐flow perturbations, but their effect is numerically small. the solutions in the inner surface layer satisfy the full non‐linear equations, and by matching prudently, the layers above also contain a non‐linear correction. We show that this means that the theory is valid when Mu * / U 0 ≪ 1 (so that it may be applied to atmospheric problems even when M = ln( z 0 / z 1 ,) is of order one). the theory agrees well with experiments and numerical simulations (which use complex models for the turbulence closure) and improves on the previous analytic theories, especially in the prediction of the surface shear stress.
No takes yet. Share an insight, caveat, or question.
Belcher et al. (1990) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: