Effects of uncertainties in positioning of PIV plane on validation of CFD results of a high-head Francis turbine model

The use of Computational Fluid Dynamics (CFD) for the design modern hydraulic turbines has increased and matured significantly in last decades. More recently, CFD is also used to understand how safely widen turbine operating ranges, avoid hazardous conditions during transient operation. accuracy such results relies on validation with experimental data which contains numerous sources uncertainties. present work focusing effects uncertainties positioning Particle Image Velocimetry (PIV) plane high-head Francis-99 model. A shutdown sequence considered, where available numerical are considered accurate according a conventional thorough procedure. part that procedure comparison spatially temporally varying velocity profiles along three lines PIV plane. this here uncertain, using translational rotational stochastic parameters uniform probability distribution functions. validated extract depends these uncertainties, while not possible data. polynomial chaos expansion method employed Uncertainty Quantification (UQ) study Sobol’ indices utilized Sensitivity Analysis (SA). UQ can be show impact extracted components field, sensitivity analysis reveals relative contribution each uncertain parameter quantity interest. For particular case it shown so-called horizontal component most sensitive plane-normal This all found literature differ from results. It function covered by standard deviation bounds, means quantified similar within range