Investigation Of Fatigue Properties And Fatigue Design Diagram Of Titanium Alloy Ti-6Al-4V

Development of long last-stage blades is the one of the core technologies for modern steam turbines with large capacity and high efficiency. However, fatigue life and reliability are the critical problems for long blades design due to complicated stress state, wet steam and aggressive environment. The scope of this dissertation focuses on fatigue design issues of Ti-6A1-4V alloy blades, which has been used for long last stage in an ultra super critical steam turbine. The effects of stress ratio, steam and sodium-chloride (NaCl) aqueous environments, material microstructure and surface properties on the fatigue strength and fracture mechanisms of Ti-6A1-4V alloy have been studied. A Goodman diagram for fatigue design of the blade is presented. The main contents and conclusions are as follows:(1) Axial force-controlled fatigue tests were conducted at stress ratios of-1,-0.6,-0.3,0.1, and0.6on Ti-6A1-4V specimens in air at room temperature. The staircase method was employed to obtain accurate values of fatigue limit at each stress ratios and the Goodman diagrams corresponding to101cycles. The effect of mechanical processing method on Goodman diagrams was investigated, with particular emphasis on the hard turning plus polishing with and without vacuum stress relieve anneal carried out after polishing. Results indicate that beneficial effect of residual compressive stress decreases and vanishes finally with increasing of stress ratio. Compared to fatigue crack originating from surface for annealed specimens, the fatigue crack initiation sites are located in the interior of the specimen due to the effect of residual stress when low stress ratios are present. The mechanism of fatigue crack initiation of Ti-6A1-4V owes to the competition of slipping at free specimen surface and cyclic plasticity damage accumulation inside the specimen.(2) An environment assisting fatigue testing system was developed for investigating the fatigue properties in saturated steam with low oxygen concentration and sodium-chloride (NaCl) aqueous. In order to investigate the influence of simulated environments on Goodman diagrams, fatigue experiments were performed in saturated steam with low oxygen concentration at100℃and NaCl aqueous at80℃. For0.1stress ratio loading conditions, steam environment demonstrated the most serious effect on the endurance limit, and the reduction of fatigue strength was12.3%. The environment of NaCl aqueous also produced a9.6%drop in fatigue strength. For all corrosion environments, cracks originated from surface and no corrosion pits were observed.(3) By combining crystal finite element method with Chaboche model which is simultaneously contained kinematic hardening and isotropic hardening rules, the mechanical properties of α/β grains undergoing cyclic loadings were analysed. The simulation results indicate that the plastic strain accumulated more in a grains than that in β grains. The incompatible deformation induces stress concentration at α-β interfaces, thereby promoting fatigue crack initiation at these sites. As the volume fraction of a grains decreases, the area containing incompatible deformation shifts down, so as to increase cyclic deformation resistance of Ti-6A1-4V. The probability of fatigue crack initiation falls off rapidly asβ grain size decreases.(4) Compared to annealed specimens, the surface compressive residual stresses induced by the turning could improve the fatigue performance defined by Goodman diagram of Ti-6A1-4V. The improvement in fatigue strengths seems to decrease with stress ratio increasing. Based on X-ray measurement and finite element simulation, it is manifested that the relaxation of residual stresses at surface and in depth is associated with number of cycles, stress amplitude and stress ratio investigated. At R=-1, the residual stresses were relaxed after several cycles and then maintained stable. The improvement on fatigue strength at this stress ratio is attributed to surface crack closure induced by compressive residual stress. For R=-0.6-0.1, however, the relaxation of residual stress is very limited. The fatigue failure thus occurs as a result of internal defect and local stress concentration due to the decrease of the mean stresses by the larger compressive stress.(5) Considering the effects of environments (in air at room temperature, saturated steam low oxygen concentration at100℃and NaCl aqueous at80℃) and machining residual stresses, Goodman diagram for Ti-6A1-4V is developed for fatigue design of turbine blades by using statistics method. Haigh parameter was proposed to modify fatigue curves for titanium blades design. A comparison of different methods shows that the proposed isochronous curve approach agrees better with experimental results than Goodman and Gerber models.