Numerical modeling of deforming bubble transport related to cavitating hydraulic turbines

Abstract: Cavitation is a problem in many hydroelectric power plants since it can cause adverse effects on performance and damage to nearby solid surfaces. The concerns of this thesis are the numerical aspects of flow simulations in cavitating hydraulic turbines which contain several difficulties: turbulent and complex flows, steady and moving parts of the geometry, bubble transport and cavitation development. The focus is on the accuracy and reliability of several different aspects of these difficulties, namely the study of bubble transport without phase change. Two main strategies are chosen. Firstly the investigation of the turbulent bubble-flow interaction in a turbine geometry and secondly the investigation of the bubble deformation and the bubble-flow interaction. Consequently, different methods in order to handling these types of three dimensional multi-phase flows are presented. Volume of Fluid (VOF) is used for immersed fluid-fluid flows and improved methods are presented and evaluated for the phase transport and the interface treatment. This includes the Direction Averaged Normal model (DAN) and the Direction Averaged Curvature model (DAC). The Volume of Solid (VOS) method is also presented and evaluated. VOS is built on VOF and is a second order accurate boundary treatment method in Cartesian grids for both stationary and moving geometries of complex shape. All the methods are tested using different three dimensional cases which leads to a confirmation of the high accuracy. The high accuracy of the VOF model is verified by comparing it with the experimental data for both rising wobbling bubbles and the bubble formation for air injection in the bottom of a water channel. The real advantage of the VOS method is demonstrated for a turbulent flow past a rotating propeller placed in a square channel, where the turbulence and the bubble transport are simulated using Large Eddy Simulation (LES) and Lagrangian Particle Tracking (LPT) respectively.