Probabilistic Fatigue Life Prediction And Strength Analysis Of Notched Components Under Size Effect

With the functional requirements of mechanical structures getting more and more complicated,engineering components inevitably contain various types of notches and geometric discontinuities e.g.grooves,holes,notches,etc.,and often present complex stress/strain distribution state.In addition,fatigue failures normally originate from notches where local stress concentration occurs.Hence,during fatigue assessment of engineering components,it is of vital significance to investigate the influence of the notch effect on structural integrity.Moreover,limited by full-scale test cost and test conditions,small-size specimens are normally utilized for fatigue tests,and structural fatigue strength is related to structure size.In this regard,investigation of the size effects is necessary for the transition of fatigue properties from specimens to real components.In addition,uncertainty is another important factor which cannot be ignored as well.Due to the uncertainties induced by load,material etc.,fatigue life often presents unavoidable dispersion.Recently,the critical distance theories provide effective strategies for notch fatigue analysis,nevertheless,valid methods,which combine notch effect,size effects and uncertainty are still lacking in engineering practice.Focus on above requirements,based on systematically summarization of notch effect,size effects and uncertainties and related fatigue life prediction models,this thesis preforms the following work based on existing notch fatigue analysis theories:(1)On the basis of Weibull distribution and critical distance theories,a novel fatigue life model under notch effect and uncertainty is developed;then,a correction factor accounting for statistical size effect is developed in terms of highly-stressed volume approach.Using the developed models,a versatile fatigue life prediction framework which takes notch effect,statistical size effect and uncertainty into account is developed.In the end,tested fatigue lives of Al 2024-T351 are utilized for model validation and comparison.(2)The influence of size effects on critical distance values is analyzed first,then,three methods are developed for statistical and geometric size effect modelling.Specifically,the influence of statistical size effect on critical distance values is quantified by treating the notch size as correction factor;the influence of geometrical size effect on critical distance values is quantified based on mean value method and highly-stressed volume approach.Finally,two types of materials are utilized for model validation,and results indicate that the predicted life based on highly-stressed volume method is better.(3)Finite element model of a turbine rotor structure under typical working conditions is established,and the stress distribution in the dangerous area is extracted.Then,Weibull distribution is combined with highly-stressed volume modified critical distance theories for fatigue life prediction of compressor disc.The predicted results indicate that the predicted fatigue life of the new model is in good agreement with the actual fatigue life of the compressor disc.