Reliability Analysis Of Aero Engine Turbine Disk Fatigue-Creep Life

Aircraft turbine disk is one of the serious concerns for an aero engine in term of performance and maintenance cost.The research on aircraft turbine disk fatigue-creep life prediction is very critical.At aero engine working time,the failure is mostly generated by the low cycle fatigue,creep,their interactions damage and have a direct influence on the life and the reliability of aircraft turbine disk.The thermal and mechanical loading increase the stress and strain which change the aircraft turbine disk work to the plastic state.Due to the development of aviation industry,many types of research are conducted to increase any details working life and reliability of aero engine including the aircraft turbine disk.In this thesis,we do our research focusing on the combined fatigue and creep effect on the life and reliability of the turbine disk.A probabilistic calculation was made for the creep parameters.The principal works of this thesis are as follows:Described the basic concepts,theories,and methods used to predict the fatigue creep life.Analyzed in order the advantages and disadvantages of each method and select the method which will be used to estimate the fatigue-creep failure.Set up the probabilistic creep parameters according to an uncertainty of the prediction using Bayesian inference framework.Damage fracture summation combining Coffin Marson theory and Larson Miller theory was used to predict the fatigue creep failure of the turbine disk.Designed based on the geometrical data the three-dimension model of the turbine disk using SolidWorks.Computed the finite element analysis fatigue-creep in Ansys workbench on Nickel Base super alloy GH4133 B turbine disk at 650°C based on the material properties,load and performance parameter of the low cycle fatigue as a random factor.The result of Ansys simulation is used with damage fraction summation to calculate the low cycle fatigue creep life and fitted by two-parameters Weibull distribution.Monte Carlo simulation is used in random sampling from two-parameters Weibull distribution at the end to estimate the reliability assessment of turbine disk.