On Assessment Of Low Cycle Fatigue Damage Of Large Steam Turbine Rotor

Large steam turbine unit is an important engine equipment related to nationaleconomic construction，accurately assessing the damage of steam turbine rotor hasimportant theoretical significance and engineering value. Initialization and develop-ment of turbine rotor damage depends not only on the quality of process equipment,material properties, operating conditions, it is also affected by the control performanceof practical controller. This paper is devoted to research about nonlinear cumulativemethods of low cycle fatigue damage of turbine rotor, online assessment for low cy-cle fatigue damage of turbine rotor, model establishment for life extending control ofturbine rotor and its performance assessment, the main innovations are as follows:(1) Aiming at the problem that linear cumulation method of low cycle fatigue dam-age is not accurate enough, this paper proposes firstly a new nonlinear cumula-tive method based on thermodynamic entropy. Using the Manson-Coffin equa-tion and cyclic stress strain relationship, it eliminates the unmeasured parame-ters of original model. Compared to the current linear cumulative damage modelused widely, this thermodynamic entropy model considers the damage accumu-lation under cyclic stress, and it solves the inaccurate problem of Miner linearcumulative method. This paper provides a new actually available damage as-sessment method in engineering.(2) This paper firstly proposes the online damage assessment method based on fi-nite element analysis technology, it determines the method and steps of estab-lishing online model of low cycle fatigue damage. Combined the advantages ofnumerical calculation method with that of analytical method, this paper studiesthe sensitivity of thermal stress transfer function. It quantitatively analyzes the impact of rotor model geometry in finite element software on the results. Thiswork derives the low cycle fatigue damage of domestic300MW steam turbineunder cold start, comparison with the result of finite element analysis validatesthe accuracy of present model.(3) Aiming at the need to estimate white noise in performance assessment and theneed to verify the online damage assessment model based on finite element anal-ysis technology, this paper establishes a predictive model of low cycle fatiguedamage of steam turbine rotor using dynamic partial least square. The correla-tions between different variables and the correlation at different time are elimi-nated. The most influential variables are determined. Data from rolling processof a domestic300MW turbine unit are used for the dynamic PLS model, a sim-ulation example validates the on-line damage assessment model based on finiteelement technology. At the same time, it supplies a new modeling approach tobe used in the damage assessment of turbine rotor.(4) In the operating process of steam turbine, aiming at the questions of using thedamage as target in performance assessment, this paper decomposes the non-linear model of low cycle fatigue into a dynamic linear part and a static non-linear part. It effectively reduces the complexity of the original model. Poly-nomial model and assessment benchmark are established based on the domestic300WM steam turbine unit rolling process data. Simulation result shows thatperformance is nonlinear with the change of proportional factor. This work sup-plies a new method of performance assessment for damage control of steamturbine unit.