| Under the influence of various physical factors such as flow field,structure,and electromagnetism,pumped storage units are very likely to cause vibration,deformation and damage to the component structure when they are in a severe transition process.In particular,under load rejection conditions,high head pump turbines operate at a highly distorted S curve,where the internal flow field changes drastically,resulting in various forms of vibration in different parts of the unit due to water hammer effects,rotation stall and vortex effects,causing deterioration of the unit’s operating condition.The difficulty lies in establishing numerical models that are both highly accurate and efficient,and that reflect the laws of real power station operation.To solve these problems,this study coupled multiple physical fields of pumped storage units by establishing a joint simulation physical model and applying a reduced-order model to speed up the computation.The study introduces different forms of step-down methods to construct an integrated coupled system for different physical fields,combining the strength and vibration response of the unit under the influence of multiple physical fields,and proposes a method to evaluate and optimise the operation of large fluctuating operating conditions.The study is carried out in the following three areas.(1)In view of the deterioration of performance indexes under load rejection conditions on pumped storage power stations,the mechanism of the influence of wicket gate closing laws on the key performance parameters of over-flow components was analysed.A reducedorder model(ROM)based on proper orthogonal decomposition was proposed for the fast calculation of the transient flow field of pumped hydro turbines,and the hydraulic characteristics of pumped storage power stations were solved by connecting the pressurised pipes and turbine control systems modelled by the method of characteristic in series with the ROM.A global sensitivity analysis of key points within the flow field under load shedding conditions was carried out by extending the Fourier amplitude test,combining pressure curves at measurement points,internal flow field resolution and eddy current analysis to develop new rules for screening key indicator parameters for wicket gate closing optimisation.Non-dominated Sorting Genetic Algorithms III was introduced to optimise various types of guide vane closing laws.By comparing various forms of wicket gate closing laws,a new three-stage closing law of reverse opening was proposed to maximise the operating performance of the unit under load rejection conditions.(2)To study the unbalanced magnetic pull caused by violent operating condition transitions or rotor eccentricity,Maxwell2 D transient simulation of the stator-rotor electromagnetic field of an excitation-convex three-phase synchronous generation motor was used to obtain the electromagnetic field distribution parameters,and the equivalent circuit model extraction was introduced to achieve fast calculation of the total electromagnetic field set parameters.An integrated simulation model of a power generation motor on the Twin Builder platform has been developed using the SML to achieve high fidelity and efficient simulation of motor operation under load shedding conditions and to provide an excitation source for the electromechanical coupling of the generator shaft system.A new method for solving the dynamic eccentric unbalanced magnetic pull force based on the distribution of spatial magnetic induction intensity and polynomial fitting was proposed based on the virtual work method and the electromagnetic finite element method.The accuracy of the new method for solving the electromagnetic force between the air gap at variable eccentricity was verified by comparing the finite element solution,the analytical solution and the engineering data.(3)To assess the influence of the large fluctuation transition process on the structural field of the unit shafting,a three-dimensional numerical simulation model of the dynamics of the multi-physics field unit shaft system was established,and the bearing stiffness was obtained by means of the support structure-bearing thermoset transient simulation,overcoming the limitation of using empirical formulae for estimation.A method was proposed to obtain the ROM of the shaft system using singular value decomposition and the Kriging proxy model.The sensitivity of the structural strength parameters to the flow-EMF excitation was obtained by response surface analysis,and a method was devised to quickly extract the stress time history of the shaft system.A rapid calculation system model of the shaft system was built in Twin Builder by integrating multiphysics field ROM.The transient strength response of the shaft system under various wicket gate closing laws was analysed,and the fatigue life of the shaft system was predicted by processing the stress time history through the rainflow matrix method.The study on the above provides theoretical support for the operation of large highhead pumped storage power stations during violent transitions,and solves the problem of violent operating processes and sudden changes in key parameters during load rejection conditions,enabling fast calculation of unit strength and vibration. |
