Advanced experiments and simulations to model coherent structures in shallow smooth and rough open channels

点击翻译按钮获取中文摘要

Laboratory scale smooth open channel flow can be considered to be the simplest flow field in hydraulic engineering. However, this flow exhibits many of the phenomena that occur in complex turbulent flows. To understand complex flow fields, researchers have studied the response of the standard smooth bed flow to changing boundary conditions which are imposed in many different forms (such as change in bed roughness). It is also noteworthy that empirical relations established from very simple flows at laboratory scales are often employed far beyond their established range of validity in the more complex flow fields. This is particularly disconcerting as there has been a proliferation in the use of commercial computational codes which use the empirical relations. For e.g., there is considerable work which attempts to relate river bed form roughness to sand grain roughness employed in classical experiments. Most practical open channel flows occur on rough (mobile/immobile) beds with or without seepage (suction/injection). While the distribution of the mean velocities and shear stress is well documented, the role of turbulent structures is less known. Several critical questions arise while modeling practical flows. These include: How does roughness geometry (height, spacing, density etc.) determine the size and frequency of the larger turbulent eddies responsible for bulk of the transport? How are the conventional laws and role of coherent structures modified in the presence of seepage and bed mobility? The present proposal will systematically address these and other important questions by conducting numerical simulation and laboratory experiments on selected roughness geometry with well defined flow conditions including seepage effects. The goal of this proposal is to study the instantaneous, time-evolving coherent structures and how they are affected by the proximity of the bed/free surface. The velocity fields will be examined using proper-orthogonal decomposition, momentum analysis, and conditional quadrant analysis. The study is made possible by the availability of state-of-the-art experimental facilities and numerical codes at the University of Windsor.