Study on Data Assimilation and Multi-condition Generalization of Slope Flow Separation
The Ensemble Kalman Filter method is used to optimize the data assimilation of the shear stress transport model constants to accurately predict the characteristics of the flow field on the slope surface with different inclination angles.A set of flow field calculation and prediction scheme is obtained under multiple working conditions.To this end,particle image velocimetry is applied to measure the separation flow of slope models with high,middle and low gradients at Reynolds number Re=38 100,Re=57 200 and Re=85 800.The measured time-averaged velocity distribution at several downstream locations is used for data assimilation.The influence of observed data at different locations on flow field of a single model is analyzed.In this paper,the model constant of assimilation at medium slope and Re is extended to calculate the flow field at other slopes and Reynolds numbers.The generalization ability of the model augmented by data assimilation under multi-condition parameters is analyzed and verified.The results show that the model constant obtained by assimilation under medium gradient and Re=57 200 is most suitable for generalization compared with other gradients and Reynolds numbers.It can be considered that this set of model constants is suitable for all conditions and parameters studied in this paper.At the same time,compared with the original model,the eddy viscosity distribution of the assimilated flow field is mainly concentrated in the middle and downstream of the slope.The generalized calculation shows that the eddy viscosity distribution in the middle and downstream of the slope is lower than that of the calculation result of the default model.
Data assimilationEnsemble Kalman FilterModel constant optimizationFlow separationPIV
万方数据
维普
