Experimental Investigation of the Interaction Between a Stationary Rigid Sphere and a Turbulent Boundary Layer

Time-resolved tomographic particle image velocimetry (PIV) measurements (acquisition rate 250 Hz) were performed in a turbulent boundary layer on the side wall of an open channel, water flow facility (cross section 60 × 60 cm2, W × H), 3.5m downstream of the inlet at a bulk flow velocity of Ub = 0.17m/s (Reb = UbH/ν = 97, 679, δ0.99 = 45.0 mm, Reθ = 752). The measurement volume was a horizontal slab (60 × 15 × 60mm3, l × w × h) extending from the wall, 30 cm above the bottom. The setup comprised four high-speed ImagerPro HS cameras (2016×2016 pixels), a high-speed laser (Nd:YLF, Darwin Duo 80M, Quantronix), optics/prisms, and data acquisition/processing software (LaVision, DaVis 8.2). Data were acquired with and without a stationary held sphere that had a diameter, D = 6mm (D+ = 51, “+” denotes inner wall scaling), and was positioned at x3 = 5.4 and 37.6mm ( = 43 and 306) from the wall (measured from the sphere’s center). Sphere Reynolds numbers based on D and the average streamwise velocity at the sphere’s center were 692 and 959, respectively. The mean streamwise velocity profiles of the undisturbed boundary layer clearly exhibit a canonical shape. Introducing the sphere strongly affected log layer and buffer layer mean velocity and Reynolds stress profiles. Recovery to the undisturbed boundary layer characteristics is faster with the sphere positioned closest to the wall. When positioned at h+ = 306, near-wall, uplifted, coherent vortical structures extend from the wall up to the sphere’s wake with which they interact.