Author(s): Saad Mulahasan, Thorsten Stoesser, Safaa Kh. Al-Jumaili
Linked Author(s): Saad Mulahasan
Keywords: Hydraulic jumps, open channel flows, square bars, image processing, water surface profiles
Abstract:
This paper describes a method for the estimation of the time-averaged free surface profile measured directly from particle image velocimetry (PIV) images of a laboratory open channel flow over a transverse square bars of normalized spacings λ/k = 5.2 and λ/k = 10.4 representing transitional and k-type roughness geometries, respectively. Three different relative submergence ratios of H/k = 2.4, 2.7 and 3.3 for the transitional geometry and H/k = 2.5, 3.0 and 3.5 for the k-type geometry are investigated. 1500 images of the free surface have been recorded using a Baumer TXG14F CCD camera in conjunction with a Polytec BUS-11 Wotan Flash stroboscope and a halogen lamp. Seeding particles are used in capturing the water surface profile. The large spacing is characterized with dynamic free surface profile in which the H/k = 3.5 case is characterized by an extremely dynamic free surface with significant spanwise wandering motion of the hydraulic jumps, while the small spacing showed of less free surface dynamics. The mean longitudinal free surface profile is determined using an image processing approach by the Matlab software. The results by this method is a well-defined water surface profile showing visible hydraulic jumps the bars for all the large spacing and low submergence case of the small spacing. The PIV determined free surface profiles are compared with the point gauge measurements and showed good agreement. In addition to the mean free surface profile the standard deviation of the longitudinal free surface profiles are calculated for the six relative submergences. Small freesurface fluctuations are observed in the case of the small roughness spacing. A significant increase in free surface fluctuation is observed in the small-scale roughness at low submergence, and the free-surface fluctuations reached a maximum value for the large-scale roughness spacing. (2597, 73, 336)
Year: 2017