Study On Fatigue Dissipated Energy And Its Application To Rapid Prediction Of Fatigue Properties

The conventional fatigue tests need lots of cost and time to obtain the accurate high-cycle fatigue parameters. Therefore, the theoretical and experimental rapid predictionmethods became the focus problem of the researchers. Energy theory is an importantmethod for rapid prediction of fatigue properties. However, there are many bottlenecks anddifficult problems in the existing methods, such as the computational accuracy of the high-cycle fatigue dissipated energy, the relationship between the macro dissipated energy andthe internal microstructure evolution and the stored energy measurements, etc. To this end,this dissertation aims to establish some valuable method to predict the fatigue properties inbasis of dissipation energy computation. First, a dissipated energy computation method wasproposed for the small thermal changes during high-cycle fatigue process. Secondly, therelationship between dissipation energy in macroscopic scale and the internal micro-structure evolution in microscopic scale was studied during fatigue process. Finally, tworapid prediction methods of high-cycle fatigue properties were established based on stable-state dissipated energy and initial transient dissipated energy. The major research works ofthis dissertation are as follows:(1) A fatigue dissipated energy computational technique for fatigue loadings based oninfrared thermal imaging technology is developed. In the thermodynamics framework, aheat conduction equation was established under fatigue loadings based on thin-planeassumption. An infrared fatigue test system was set to observe the small thermal changesusing the non-contact infrared thermal imaging technology and fatigue test machine. In thesystem, the environmental noise was reduced by setting the reference specimen and thethermal insulation equipment. The dissipation source, thermo-elastic source and thermalradiation source causing local temperature changes were isolated. A dissipated energy per-cycle computational equation then derived from the dissipation source, and the detectionthreshold was determined by dissipated energy measurements. The feasibility and accuracyof the dissipated energy computational technique was finally verified by a fatigue test.(2) Through fatigue dissipated energy measurements under different load history, themacro-dissipated energy can be used as a marker of the internal microstructure evolution.Based on energy balance equation in fatigue process, the dissipated energy, plastic strainenergy and stored energy per-cycle theoretical calculation method were developed underthe initial transient and steady state stages. The variation of dissipated energy and plastic strain energy were analyzed in elastic hysteresis domain. Through a comprehensivecomparative analysis of dissipated energy variations during tensile damage, fatigue damagethe results show that the fatigue dissipated energy is closed related with the internalmicrostructure evolution, and it could be used as a sensitive indicator to fatigue damageassessment. Fially, the constant and variable amplitude fatigue damage was in-stiumonitored by fatigue dissipated energy.(3) Based on the steady state dissipated energy, a rapid prediction method for fatigueproperties was developed. After a certain number of fatigue lifetime cycles, the materialreaches a steady state dissipation energy stage, and the internal microstructure evolutionreached a quasi-equilibrium state. The dissipated energy was remained constant duringfatigue process. One curve method and dual curve method of fatigue limit predictionwere proposed by fitting dissipated energy under different levels of fatigue loadings insteady-state stage. Meanwhile, the dissipated energy versus lifetime curves derived showsthe same rule with the stress versus lifetime curve, and the dissipated energy versus lifetimecurve was also characterized the discrete distribution of fatigue lifetime. A Miner’scumulative damage model was deduced from dissipated energy measurements. The modelwas used to predict the residual lifetime and to study the load sequence effects. The resultswere consistent with the experimental data with good accracy.(4) A high-cycle and very high-cycle fatigue lifetime prediction was studiedconsidering with initial micro-plasticity effects. The dissipated energy per-cycle wasrapidly increasing and then gradually stabilized, corresponding to the initial micro-plasticeffects at the initial transient dissipated energy stage. Variations of the initial dissipatedenergy were monitored under different levels of fatigue loadings at the initial transient andsteady state stages. A new prediction method was studied based on the initial plasticdissipated energy. The fatigue limit according to this method was close to the one predictedby the steady-state dissipated method and the experimental test. The initial plasticdissipated energy versus lifetime curve fitting was also similar to stress versus lifetime andsteady-state dissipated energy versus lifetime curve. Based on the shakedown theory ofplasticity and Dang-Van fatigue criteria, a new fatigue limit prediction method of veryhigh-cycle fatigue was studied based on the initial cumulative dissipated energy. Theexperimental results proved that the method has certain rationality.