Air Entrainment and Surface Fluctuations in a Turbulent Ship Hull Boundary Layer

Abstract

Air entrainment due to turbulence in a free-surface boundary layer shear flow created by a horizontally moving vertical surface-piercing wall is studied through experiments and direct numerical simulations (DNS). In the experiments, the moving wall is created by a laboratory-scale device composed of a surface-piercing stainless steel belt that travels in a loop around two vertical rollers; one length of the belt between the rollers simulates the moving wall. The belt accelerates suddenly from rest until reaching constant speed and creates a temporally evolving boundary layer analogous to the spatially evolving boundary layer that would exist along a surface-piercing towed flat plate. We report cinematic laser-induced fluorescence measurements of water surface profile histories, cinematic observations and measurements of air entrainment events, and air bubble size distributions and motions. To complement the experiments, DNS of the temporally evolving turbulent boundary layer were conducted, considering both the air and water phases. Because of cost considerations, only a portion of the belt was simulated at a lower Reynolds number, keeping the Froude number, however, at the same levels as in the experiments. The results of the experiments and DNS are found to be in qualitative agreement and are used synergistically to explore the physics of the air entrainment process; quantitative agreement is not to be expected given the differences in setup and Reynolds numbers. In the experiments and DNS, the free-surface motion is found to consist of a region near the belt with fast-moving uncorrelated large-amplitude ripples and an outer region of small-amplitude propagating waves. Entrainment events similar to plunging breaking waves are found in the experiments, and these and other entrainment mechanisms are examined in detail in the DNS. The spatial distributions of bubble numbers and velocities are reported along with their diameter distributions.