| With the adjustment of energy structure and the further transformation of social power consumption structure,the peak-to-valley gap is increasing,and the power grid requires large-capacity thermal power plants to have faster start-up and peak load changes.The frequent start-up and shutdown and load changes during peak shaving will cause huge temperature gradients inside the steam turbine rotor,resulting in large thermal deformation and thermal stress.After a long period of action,it will cause low-cycle fatigue damage to the rotor and threaten the safety of the unit.Therefore,real-time monitoring of the thermal stress and fatigue damage of the steam turbine rotor have great significance for ensuring the safe operation of the unit,studying the start-up process of the unit,and improving the response speed of the variable load optimization.It has a positive effect on the health management and maintenance of large-capacity thermal power units.To the question of the real-time thermal stress monitoring and the damage assessment of turbine rotors of large thermal power units,the following research works have been carried out:(1)the one-dimensional recursive algorithm,one-dimensional difference algorithm,the two-dimensional difference algorithm which can be quickly calculated,and the calculation formula of the finite element method with high calculation accuracy are derived respectively.The advantages and disadvantages and stability conditions of each algorithm are analyzed.(2)Based on finite element analysis software ANSYS?,the600 MW subcritical steam turbine high and intermediate pressure rotor of a power plant was taken as the research object.The temperature distribution and thermal stress change characteristics of the steam turbine high and intermediate pressure rotor under cold and hot start conditions are analyzed.The point with high stress value is determined as the monitoring point of online calculation.And the thermal stress of the monitoring point under the same working condition is calculated by the two-dimensional difference method,and the correctness of the algorithm is verified.(3)Based on the damage theory combined with the rapid calculation features of the two-dimensional difference method,an online evaluation model for online fatigue damage evaluation is established,and the method and steps for establishing a low-cycle fatigue damage online evaluation model are given,and compared with the loss of design life given by the manufacturer to verify the accuracy of the model.The temperature and stress changes in variable load conditions are analyzed,and the optimization method of variable load is given in combination with operating regulations and fatigue damage.(4)Based on the neural network algorithm that has been widely used in the engineering field in recent years,the LSTM neural network is used to establish a steam turbine rotor thermal stress prediction method.This method can learn the depth information of long time series and dig the correlation among the data.Taking the cold start condition as an example,the thermal stress of the rotor at the dangerous monitoring point is predicted,and the feasibility of the model is verified by comparing to the original data,which provides a new research method for studying the characteristic variables that are not easy to measure in engineering practice.On the basis of the above research,the variation characteristics of temperature distribution and stress in the rotor in cold and hot start-up and variable load of 600 MW subcritical steam turbine were analyzed in detail by two-dimensional calculus of differences and finite element method,and the accuracy of the difference model and evaluation model were verified.In addition,to further shorten time of the variable load operation to meet the demand of rapid response to the peak shafting of the power grid,the optimization scheme is presented based on the analysis of thermal stress and life loss in the process of variable load.The results obtained were satisfactory. |
PDF Full Text Request