Abstract
The equations for double-mean, form-induced and spatially averaged turbulent energy budgets are employed to analyse data from direct numerical simulations of turbulent open-channel flows over transitionally rough mobile beds with intermediate flow submergence. Two scenarios were considered related to (i) near-critical bed condition, and (ii) fully mobile bed condition. The bed was composed of a layer of mobile spherical particles moving on the top of one layer of fixed particles of the same size. Data analysis showed the leading energy exchanges between double-mean, form-induced, turbulent flow field contributions as well as particle motions. Above the fixed particles tops, the turbulent flow receives kinetic energy directly from the mean flow as well as from moving bed particles, which in turn also receive energy from the mean flow. For near-critical bed condition, particle aggregations enhanced mean-flow heterogeneity, strengthened turbulent stresses and their effects on the flow, while at increased bed-mobility, energy transport mechanisms became weaker and conversions induced by viscous stresses and pressure became stronger.
Highlights
Sediment transport processes and the associated morphological changes to the river bed occur across a wide range of temporal and spatial scales (e.g. Dey, 2014; García, 2008; Graf & Altinakar, 1998; Julien, 2018)
The equations for double-mean, form-induced and spatially averaged turbulent energy budgets are employed to analyse data from direct numerical simulations of turbulent open-channel flows over transitionally rough mobile beds with intermediate flow submergence
The objective of this study was to identify the key energy exchanges involved in mobileboundary flows by assessing the terms of the balance equations for double-mean kinetic energy (DMKE), form-induced kinetic energy (FKE) and spatially averaged turbulent kinetic energy (TKE)
Summary
Sediment transport processes and the associated morphological changes to the river bed occur across a wide range of temporal and spatial scales (e.g. Dey, 2014; García, 2008; Graf & Altinakar, 1998; Julien, 2018). Ancey & Heyman, 2014; García, 2008; Singh, Foufoula-Georgiou, PortéAngel, & Wilcock, 2012) To advance this problem, a procedure is needed for the transfer of information on fluidmechanical and sediment transport processes from small scales (starting with a sub-grain scale) to larger scales. Two options are considered in the selection of the averaging volume: the “global’ and “local” averaging domains, where spatial integration (Eq 1) is performed over the volumes V0,1 = 12H × 6H × z and V0,2 = 12H × 2D × z, respectively, where z is the bed-normal cell size of the simulation grid. Averaging over the “local” volume V0,2 smooths streamwise heterogeneity, but resolves flow heterogeneity in the spanwise and bed-normal directions at scales larger than 2D and z, respectively. The following subsections summarize the key findings of Vowinckel et al (2017a, 2017b) as the underpinning background for the current analysis
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call