UC Davis

This study explores critical issues in the fatigue behavior and fundamental mechanisms of cyclic deformation in alloys for structural and hydrogen-energy applications using plastic strain-controlled low cycle fatigue (LCF) and constant microstructure plastic strain rate change experiments. The materials of interest are strain-hardened Type 316L austenitic stainless steel and annealed CrMnFeCoNi high entropy alloy (HEA), both in the hydrogen-precharged (H-precharged) and non-charged conditions. The LCF tests were conducted at a constant plastic strain rate of 1x10-3 s-1, probing plastic strain amplitudes between 0.1% and 0.7% in 316L and between 0.3% and 0.8% in the CrMnFeCoNi alloy. The plastic strain rate change experiments are a unique type of experiment designed to explore the fundamental mechanisms that control dislocation glide in the context of thermal activation theory. They were conducted periodically at two different regions of the lifetime, probing operational activation area values that represent the evolving and dynamic equilibrium states of the microstructure at plastic strain amplitudes between 0.3% and 0.8%. In the strain-hardened 316L austenitic stainless steel, cyclic stress response curves show continuous softening in both H-precharged and non-charged conditions at all plastic strain amplitudes. Internal hydrogen increases the cyclic strength by enhancing the effective component of the flow stresses, especially at low plastic strain amplitude. However, the increase in effective stress is accompanied by a significant reduction in initiation and total fatigue lifetimes in the H-precharged condition. At high amplitudes, back stresses become more significant in both material conditions and, although hydrogen still leads to earlier failure, the difference in LCF lifetimes is reduced. Scanning electron microscopy (SEM) observations reveal earlier onset of multiple slip in the presence of hydrogen at low strain amplitudes, indicating that the premature failure in the H-precharged condition is likely due to microcracks initiating at intersecting planar slip bands. Calculations of operational activation area during cyclic deformation suggest that dislocation glide is controlled by solute atoms, forest dislocations and cross slip in all material conditions, except at low plastic strain amplitudes, where cross slip is not active in the H-precharged condition. In the context of thermally activated deformation, hydrogen affects the dislocation glide resistance by decreasing both the dislocation activation distance and spacing. The Haasen plot also indicates the presence athermal obstacles to dislocation glide. Based on transmission scanning electron microscopy (STEM) observations, these obstacles are likely to be dislocation cell walls in the non-charged condition at all amplitudes and in the H-precharged condition at high amplitudes, and dense dislocation tangles in the H-precharged condition at low strain amplitude. Because different dislocation arrangements are favored at different plastic strain amplitudes in the presence of hydrogen, the athermal stresses also are also amplitude dependent in the H-precharged condition. Therefore, hydrogen affects both the dislocation glide kinetics and the microstructural evolution in strain-hardened 316L under cyclic straining. In the annealed CrMnFeCoNi alloy, the hysteresis loops reveal an unusual small yield point in the first cycle (monotonic behavior) at all plastic strain amplitudes. Furthermore, the work hardening rates are very low during the initial cycles. These phenomena are likely associated with the compositional complexity of this alloy and the very low dislocation density after annealing. Calculations of operational activation area reveal that dislocation interactions with the solid-solution matrix and forest dislocations control the rate of cyclic deformation in the CrMnFeCoNi alloy. Although dislocation cell structures are reported to develop during cyclic straining, cross slip is not an operative rate-controlling mechanism. The results demonstrate that, despite the different microstructural evolution under LCF testing, the kinetics of dislocation glide in this alloy are the same under monotonic and cyclic loading conditions. Furthermore, the compositional complexity does not lead to different rate-controlling obstacles compared to those observed in conventional solution strengthened alloys under similar deformation conditions. Comparing the cyclic behavior of non-charged and H-precharged CrMnFeCoNi, the addition of hydrogen leads to serrated flow during the first cycle, which indicates a hydrogen pinning effect on mobile dislocations that is overcome by an increase in dislocation density with further straining. Cyclic hardening and softening followed by a region of approximately constant peak stresses that extend to the end of life characterize the cyclic stress response in both material conditions. Hydrogen increases the peak stresses by enhancing the effective stresses in this HEA, but it does not affect the evolution of back stresses. Furthermore, severe reduction in crack initiation and total lifetimes are observed in the H-precharged condition. The analysis of fracture surfaces reveals this reduction is due to brittle intergranular failure in H-precharged CrMnFeCoNi, consistent with a reported model of hydrogen-induced strain incompatibility across grain boundaries. This dissertation work contains the first fundamental studies of dislocation glide processes during cyclic deformation of strain-hardened Type 316L austenitic stainless steel and annealed CrMnFeCoNi alloy using a unique “mechanical microscopy” approach. It is also the first study of the influence of hydrogen on dislocation glide processes in 316L stainless steel. Lastly, it is the first study to show extremely detrimental effects of hydrogen in CrMnFeCoNi during low cycle fatigue. Together, these studies help advance our understanding of the opportunities and limitations of using these materials in fatigue-critical structural and hydrogen energy related applications.