Study On Hydraulic Loss And Noise Mechanism Induced By Vortex And Multi-objective Optimization Of A Centrifugal Pump

Efficiency and noise are important objectives in the design optimization of hydraulic rotational machines,and the vortex structures in the flow field and their strong unsteady effects are commonly considered as the main causes of hydraulic losses and noise generation.However,the current analysis of vortex structures in loss and noise studies is often limited,failing to capture vortex flow characteristics fully and resulting in a lack of deep understanding of the mechanisms behind vortex-induced hydraulic losses and noise generation.This thesis focuses on the centrifugal pump and employs Improved Delayed Detached Eddy Simulation(IDDES)along with a vorticity binary decomposition method based on rigid vortex theory and Powell’s vortex sound theory to systematically investigate the effects of shear and rigid rotation in vortex flow on viscous dissipation and noise generation mechanisms.The paper establishes quantitative correlation models between vortex flow intensity features and hydraulic losses as well as noise,and develops a centrifugal pump multi-objective rapid optimization design method based on passive control of vortex flow.The main content and conclusions of this paper are as follows:Firstly,the IDDES model is used to simulate the flow field,and the reliability of the flow field simulation is verified through external characteristic experiments and particle image velocimetry experiments.Based on flow simulation,the Powell’s vortex sound equation solving program is developed using the Fluent UDS method,and the reliability of the acoustic field simulation is verified through indirect verification based on literature results and direct verification based on centrifugal pump outlet noise measurement experiment.Secondly,the rigid vortex identification method and entropy production diagnostic method are applied based on flow field simulation to analyze the influence of shear and rigid rotation strength of fluid elements on local hydraulic losses in vortex structures and establish correlation models between shear,rigid rotation,and viscous dissipation.The results indicate that shear is the main component of the vorticity and also the dominant factor causing hydraulic losses.High energy dissipation is observed only when the vortex structure region experiences strong shear.The rigid rotation effect in the vortex flow has an inhibiting effect on hydraulic losses,and the stronger the rigid rotation under comparable shear conditions,the smaller the hydraulic losses.The linear regression model based on shear and rigid vorticity exhibits good predictive performance,with a maximum error of 4% under design conditions,meeting the requirements of engineering.Thirdly,the mechanisms of shear and rigid rotation effects of vortex induced noise generation are analyzed based on Powell’s vortex sound equation and vorticity binary decomposition.An artificial neural network method is employed to establish a rapid prediction of noise based on vortex flow characteristics of RANS results.The results show that the shear effect is the main factor contributing to noise generation.When the cosine of the angle between the rigid-vorticity vector and the shear vector is negative,a mutual cancellation effect exists between the sound source terms induced by rigid rotation and shear effects.In other words,the magnitude of the total sound source term is lower than that of the source term generated by shear,indicating the rigid rotation effect would suppress noise generation.Reducing the fluctuation amplitudes of vorticity or shear can significantly reduce sound and pseudo-sound pressure levels,although pseudo-sound pressure is less sensitive.The neural network prediction model for noise,based on features of the rigid vorticity-induced acoustic source,the shear-induced acoustic source,rigid vorticity magnitude,shear magnitude,and velocity magnitude of the time-averaged flow field,can quickly calculate noise magnitude with a maximum error of 2% under design conditions.Finally,the key structural parameters are selected through range analysis,including blade inlet angle,base circle diameter,and the eighth cross-section of the volute.By minimizing hydraulic losses and noise as optimization objectives,the established prediction models(scheme I)and unsteady numerical simulation(scheme II)are respectively used to obtain objective function values.The Kriging surrogate model and NSGA II genetic algorithm are employed to carry out the efficiency and noise optimization design of the centrifugal pump.The results show the performance is almost identical under both optimization schemes,but the computational cost of the scheme I is much lower than that of scheme II.Compared to the reference pump,the optimized pump exhibits significantly reduced shear effects and enhanced rigid rotation intensity in the internal flow field,with a 3.2% increase in head and a 3.7% increase in efficiency at the design flow condition,and an average total sound pressure level reduction of 1.07%.