Identification of 1-D cavitation model parameters by means of computational fluid dynamics

Abstract

Hydropower is a key energy conversion technology for stabilizing electrical power. When running at full load, a 3-D cavitation vortex rope develops in Francis turbines that acts as an internal source of energy and instability. 1-D models allow this phenomenon to be predicted and the range of safe operating points to be defined. These models involve three parameters: the mass flow gain factor, the wave speed and the second viscosity that must be calibrated. For the first time, the present work aims at identifying these parameters using 3-D URANS cavitating simulations. Two cavitation test cases are considered: a 2-D Venturi and a reduced scale model of a Francis turbine at full load operating conditions. RANS simulations allow the mass flow gain factor to be determined, whereas URANS simulations coupled with an optimization process allow the determination of the wave speed and the second viscosity. The new methodology shows its ability to identify the three parameters.