Cavitation prediction in a Kaplan turbine using standard and optimized model parameters

J. Dragica, M. Morgut, A. Škerlavaj, E. Nobile (Turboinstitut, University of Trieste)

CHALLENGE - The accurate prediction and optimization of the performance of hydro-machinery and marine propellers such as the Kaplan turbine. None of the tested turbulence models have yet predicted the efficiency level accurately hence time dependent simulations by more advanced turbulence models such as the SAS SST and zonal LES were considered.

SOLUTION - Steady state simulation predicts a significantly small efficiency level as well as a small extent of cavitation in a 6-blade Kaplan turbine due to a high pressure level in the runner because of overestimated losses in the draft tube. Cavitating flow was simulated using the homogeneous model and mass transfer rate due to cavitation was regulated by the Zwart mass transfer model. Advanced turbulence models, predicted efficiency more accurately than previously tested turbulence models within steady state simulations, with an even greater improvement reached in the case of the bulb turbine. The effect of cavitation on machine efficiency is well reproduced by steady-state and time dependent simulations, while efficiency level is well captured only by time dependent simulations which produced the same amount of cavitation on all runner blades.

BENEFITS - Time dependent simulation results in accurate efficiency level prediction. The entire calibration process of the empirical constants for three mass transfer models (the Zwart model, the FCM model and the Kunz model) was driven by the modeFRONTIER 4.2 optimization system.