Stochastic modelling of cavitation erosion in Francis runner
Résumé
Reaction turbines, of which Francis turbines, constitute a large proportion of low and medium head turbines installed in hydropower plants. Managing these machines represents a real challenge in terms of efficiency, competitiveness and demands on the energy market. Turbines runner blades exhibit loss of performance from damage due to several reasons. One common source of damage is erosion due to the cavitation phenomenon. Indeed, at a given operating region, rapid changes of velocity can create bubbles in the water flow due to local low pressures. When cavitation bubbles reach pressure recovery, they collapse and may induce wear or erosion in these regions. Even if this phenomenon has been intensively studied in the past decades, cavitation erosion is not fully understood as it is driven by several parameters such as flow dynamic, turbine design, environment, or material properties. Some of these parameters can be studied in laboratory to compare materials resistance between each other. This article aims to model the cavitation by a stochastic model using erosion experimental data observed in the laboratory. The benefit of such models is to consider both the uncertainties and natural fluctuations of the phenomenon. With the proposed framework, the study will highlight the differences observed in cavitation erosion experiments of two common materials used to manufacture Francis's runners. This study is the first step in a project aiming at the prediction of turbines mass loss due to cavitation erosion on actual operating Francis turbines.