Determination of bed shear stress by digital particle image velocimetry in turbulent open channel flow

A precise knowledge of bed shear stress (τ_o) in open channels is useful in many fluid mechanics problems, such as sediment transport modeling, channel bed design, and resuspension modeling. Direct measurement of τ_o is extremely difficult. For this reason, τ_o is often computed from analytical expressions of the velocity field using time-averaged, water column velocity data obtained from hot-media anemometry. Laser Doppler Anemometry (LDA) has also been employed for the computation of τ_o, using either the law of the wall theory or the Reynolds distribution for shear stress, τ, in the water column: τ = -ρu′v′̄. The resulting vertical profile of τ is then extrapolated down to the bed/sediment interface to yield τ_o. However, LDA is limited because the volume viewed by its laser intersection is very small; thus the measuring device must be moved several times in order to complete the measurements. In this project, Digital Particle Image Velocimetry (DPIV) is used to compute the vertical profile of τ using the Reynolds shear stress distribution for a fully developed turbulent flow field. DPIV is a non-intrusive optical technique that computes velocity vectors in a two-dimensional field using cross correlation of particle locations taken nanoseconds apart with a laser and recording camera. For this experiment, the area viewed by the DPIV was a small section inside the boundary layer, within 1 mm of the gravely-sand sediment bed. Because of experimental limitations, turbulent flow was induced by a front step in the open channel. Under these conditions, a shear stress profile should consist of two portions meeting to form a semi-parabolic shape. This pattern was observed in all profiles produced in the experiment. Copyright ASCE 2004.