Abstract

This paper presents image velocimetry measurements on turbulent flows adjacent to a permeable bed made of randomly packed glass particles. For measuring flow velocities inside the bed, the refractive index of the glass particles was matched with that of the fluid. By continuously scanning in the transverse direction, we measured the streamwise and vertical velocity components within a three-dimensional domain (3D2C-PIV), including first- and second-order turbulent statistics. We established how the scanning travel speed is associated with the laser sheet thickness and the space-time velocity fluctuations for collecting reliable measurements. The methodology was applied to free-surface flows over a sloping bed under low relative submergence and supercritical conditions. Space- and time-averaged profiles were obtained in a representative elementary volume as defined by the double-averaging procedure (Nikora et al. in J Hydraulic Eng.127(2):123–133, 2001). A turbulent boundary layer over the rough bed was observed when experiments were run at intermediate Reynolds numbers Re = O(1000). Apart from measuring subsurface velocities, this method shed light on the part played by the rough bed in the overall flow dynamics: the roughness layer was a buffer region within which porosity varied sharply and turbulent stress was rapidly dampened.Graphic abstract

Highlights

  • Knowledge about how boundary-layer flows are affected by porous boundaries is central to many fluid applications, ranging from shear flows of air over and through forest canopies or foams, to heat transfer problems (Ghisalberti and Nepf 2006; Mahjoob and Vafai 2008; Suga et al 2010)

  • In order to capture flows adjacent to permeable beds exhibiting similarities with real-world scenarios, this paper presents an experimental procedure for investigating open channel flows over and through a randomly packed bed of spherical glass beads under low relative submergence conditions, i.e. with a bed roughness that is comparable in size to flow depth

  • Scanning performance was evaluated by comparing the mean velocities and turbulent stresses obtained using the Refractive Index-Matched Scanning (RIMS) methodology and those obtained by the fixed laser sheet

Read more

Summary

Introduction

Knowledge about how boundary-layer flows are affected by porous boundaries is central to many fluid applications, ranging from shear flows of air over and through forest canopies or foams, to heat transfer problems (Ghisalberti and Nepf 2006; Mahjoob and Vafai 2008; Suga et al 2010). Mountain rivers exhibit two distinctive features not shared by lowland rivers: (i) their flow-depth and roughness scales are similar, which makes their turbulent structures far more complex, and (ii) their bed permeability is much higher. Both features help to explain why classic boundary layer assumptions exhibit strong limitations for predicting river processes such as flow resistance, hyporheic exchanges or the incipient motion of sediment (Recking 2009; Rickenmann and Recking 2011; Boano et al 2014; Prancevic and Lamb, 2015; Lamb et al 2017; Rousseau 2019). Flow paths—whether in the roughness or subsurface layer—are tortuous. (These paths are sketched as dashed arrows in Fig. 1.) All these elements make it very difficult to use classic eddy-resolving modelling methods at the river scale essentially owing to their high computational cost (Keylock 2015)

Objectives
Methods
Results
Conclusion
Full Text
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