| With the improvement of industrial technology in aerospace,rail transport,naval architecture and marine,and mechanical manufacture,the modern mechanical equipments are gradually moving in a large,high-speed,and high performance direction.Most of critical components and structures undergo complex and harsh environments in service,which often causes a series of accidents.The typical failure modes for engineering structures and mechanical parts are corrosion,wear and fatigue,in which the fatigue failure is particularly prevalent.For engineering components under cyclic loadings,fatigue and fracture are the most common cause that comprises approximately 50% to 90% of mechanical failures.Owing to the concealment and sudden failure of fatigue,it will cause a serious threat to the safe operation of mechanical equipments.Once the failure occurs,it is easier to cause major accidents,human injury,and economic loss.The fatigue life and reliability of mechanical parts and structures are crucial to the development of the whole of equipments in life and system reliability.In order to guarantee the safety and reliability of equipments during service operation,it should be essential to accurately assessing and predicting the fatigue life and reliability of mechanical structures.Thus,it can contribute to making maintenance strategies and health management plans,maximizing the use of equipments,improving economic benefits,and fatigue resistant design.Due to the complexity and stochasticity of fatigue process,the traditional life prediction models and reliability analysis methods cannot meet the demand of practical engineering.To date,there still have so many critical issues that need to be resolved in fatigue theory.According to the deficiencies and shortcomings in existing models,the objective of the dissertation is to deal with the challenges in fatigue damage mechanisms,life prediction,and reliability analysis for mechanical structures,which uses the finite life design method and cumulative fatigue damage theories.In this dissertation,the experimental data from the metallic specimens of critical components of equipments and welded structures are employed to demonstrate the effectiveness of the proposed methods.The main contributions and innovative outcomes of the dissertation are summarized as follows:(1)Development of an equivalent fatigue damage rule accounting for load interaction effects for residual fatigue life prediction.Due to the complexity of loading histories under variable amplitude loading and the insufficiencies of Miner rule,the mechanisms on load sequences and load interactions are studied systematically from a point of damage accumulation.According to the double parameter fatigue criterion,a concept of fatigue damage state is introduced to characterize the damage degree of materials.In view of the limitations of traditional equivalent method,a new equivalent fatigue damage rule is developed to account for the load interaction effects between two consecutive stress amplitudes.By applying this method to a fatigue damage model based on static toughness exhaustion theory,a modified residual life prediction model is derived,which can consider both mechanisms of load sequences and load interactions on damage evolution and fatigue life.(2)Development of a nonlinear damage accumulation model based on fatigue driving energy parameter for residual fatigue life prediction.The damage variable is typically used to describe the process of fatigue failure both qualitatively and quantitatively.In order to reveal the nature of energy dissipation on fatigue that conventional damage variable often ignores,by incorporating a fatigue driving stress model and strain energy density,a fatigue driving energy parameter is proposed to present the complete process of fatigue failure.By assessing the change of fatigue driving energy parameter,a fatigue damage quantitative method and a nonlinear damage accumulation model are thus established.According to the equivalent fatigue damage rule,a residual life prediction model as well as a modified model accounting for load interactions is also derived.Moreover,several typical nonlinear characteristics of two models are extracted and analyzed.(3)Development of a modified linear damage accumulation rule based on dynamic residual S-N curve and material memory degeneration.Due to the expensive computation of nonlinear damage theories and the advantages of Miner rule in practical engineering,the dynamic residual S-N curve and material memory degeneration behaviors are studied form the pespective of residual life and S-N diagram.By introducing a material memory degeneration coefficient to quantify the slop ratio of dynamic residual S-N curve,a modified linear damage accumulation model is then ontained.This model not only maintains the simplicity of traditional Miner rule,but also can provide a quantitative analysis of fatigue damage and life assessment.In addition,three linear damage rules are used for model comparison,and some common characteristics and linear damage behaviors are discussed in detail.(4)Development of a probabilistic cumulative damage model based on double linear damage accumulation rule and a time-dependent fatigue reliability analysis method.Fatigue failure is a dynamic process of damage accumulation in nature,while the traditional static statistical model cannot consider the time-dependent characteristics due to fatigue loading.For dynamic statistical models,the Miner rule is generally used for reliability modeling,which fails to depict the two stages of fatigue process,i.e.crack initiation and crack propagation.In order to alleviate such deficiencies,by using the double linear damage rule,a probabilistic cumulative damage model is put forward under normal and lognormal distribution assumptions.Based on the stress-strength interference model,a time-dependent fatigue reliability model accounting for cumulative damage and critical damage is also developed,which can predict the reliability variation during the entire usage life. |
PDF Full Text Request