Reconstruction of internal flow field in waterjet propulsion pump based on EnKF data assimilation

To accurately predict the three-dimensional flow characteristics of the flow field inside a waterjet propulsion pump, data assimilation (DA) method based on unsteady ensemble Kalman filter (EnKF) is used for the reconstruction of the flow field of a pump at different flow rates Q/Q_opt = 0.85, 1, 1.15, where Q_opt is optimal flow rate at the design point. As a compensation to the spatial limitation of planar particle image velocimetry (PIV) measurements, dynamic delayed detached-eddy simulation (DDES) results validated by the PIV data is used to provide the observational data at the optimized probe locations. In DA procedure, the shear stress transport (SST) model constants are optimized by the EnKF approach. The model constants are subsequently rescaled and fitted to form a variation with the flow rate, which is extended to the prediction of the flow field with other flow rates in the vicinity of the design condition. The results show that the SST model with recalibrated constants has improved the prediction of the internal flow field in the waterjet propulsion pump, especially the separation flow in the diffuser section. The modified model constants mainly reduce the eddy viscosity and significantly improve the fluctuation characteristics in the flow field. This study provides a reference for the fast and accurate prediction of the flow field information in the waterjet propulsion pump.