Comparison of model measured runner blade pressure fluctuations with unsteady flow analysis predictions

Abstract

An accurate prediction of pressure fluctuations in Francis turbines has become more and more important over the last years, due to the continuously increasing requirements of wide operating range capability. Depending on the machine operator, Francis turbines are operated at full load, part load, deep part load and speed-no-load. Each of these operating conditions is associated with different flow phenomena and pressure fluctuation levels. The better understanding of the pressure fluctuation phenomena and the more accurate prediction of their amplitude along the hydraulic surfaces can significantly contribute to improve the hydraulic and mechanical design of Francis turbines, their hydraulic stability and their reliability.With the objective to acquire a deeper knowledge about the pressure fluctuation characteristics in Francis turbines and to improve the accuracy of numerical simulation methods used for the prediction of the dynamic fluid flow through the turbine, pressure fluctuations were experimentally measured in a mid specific speed model machine. The turbine runner of a model machine with specific speed around nq,opt = 60 min-1, was instrumented with dynamic pressure transducers at the runner blades. The model machine shaft was equipped with a telemetry system able to transmit the measured pressure values to the data acquisition system. The transient pressure signal was measured at multiple locations on the blade and at several operating conditions. The stored time signal was also evaluated in terms of characteristic amplitude and dominating frequency.The dynamic fluid flow through the hydraulic turbine was numerically simulated with computational fluid dynamics (CFD) for selected operating points. Among others, operating points at full load, part load and deep part load were calculated. For the fluid flow numerical simulations more advanced turbulence models were used, such as the detached eddy simulation (DES) and scale adaptive simulation (SAS). At the different operating conditions, distinct flow phenomena were available for evaluating the accuracy of the simulation method, e.g. rotor- stator interaction, rotating vortex rope and runner channel vortex. The experimentally obtained pressure time signals at the runner, characteristic amplitude and frequency at several operating points offered the possibility to assess the precision in the prediction of the pressure fluctuations in the Francis turbine using the numerical simulation methods.