The Role of Microstructure on Deformation and Damage Mechanisms in a Ni-Based Superalloy at Elevated Temperatures
Abstract
The overall objective of this research work is the development and implementation of a mechanistic based time-dependent crack growth model which considers the role of creep, fatigue and environment interactions on both the bulk and the grain boundary phase in ME3 disk material. The model is established by considering a moving crack tip along a grain boundary path in which damage events are described in terms of the grain boundary deformation and related accommodation processes. Modeling of these events was achieved by adapting a cohesive zone approach (an interface with internal singular surfaces) in which the grain boundary dislocation network is smeared into a Newtonian fluid element. The deformation behavior of this element is controlled by the continuum in both far field (internal state variable model) and near field (crystal plasticity model) and the intrinsic grain boundary viscosity which is characterized by microstructural parameters, including grain boundary precipitates and morphology, and is able to define the mobility of the element by scaling the motion of dislocations into a mesoscopic scale. Within the cohesive zone element, the motion of gliding dislocations in the tangential direction relates to the observed grain boundary sliding displacement, the rate of which is limited by the climb of dislocations over grain boundary obstacles. Effects of microstructural variation and orientation of the surrounding continuum are embedded in the tangential stress developing in the grain boundary. The mobility of the element in the tangential direction (i.e. by grain boundary sliding) characterizes the accumulation of irreversible displacement while the vertical movement (migration), although present, is assumed to alter stress by relaxation and, thus, is not considered a contributing factor in the damage process. This process is controlled by the rate at which the time-dependent sliding reaches a critical displacement and as such, a damage criterion is introduced by considering the mobility limit in the tangential direction leading to strain incompatibility and failure. This limit is diminished by environmental effects which are introduced as a dynamic embrittlement process that hinders grain boundary mobility due to oxygen diffusion. The concepts described herein indicate that implementation of the cohesive zone model requires the knowledge of the grain boundary external and internal deformation fields. The external field is generated by developing and coupling two continuum constitutive models including (i) a microstructure-explicit coarse scale crystal plasticity model with strength provided by tertiary and secondary γ' precipitates. This scale is appropriate for the representation of the continuum region at the immediate crack tip, and (ii) a macroscopic internal state variable model for the purpose of modeling the response of the far field region located several grains away from the crack path. The hardening contributions of the γ' precipitates consider dislocation/precipitate interactions in terms of γ' particles shearing and/or Orowan by-passing mechanisms. The material parameters for these models are obtained from results of low cycle fatigue tests which were performed at three temperatures; 650, 704 and 760°C. Furthermore, a series of microstructure controlled experiments were carried out in order to develop and validate the microstructure dependency feature of the continuum constitutive models. The second requirement in the implementation of the cohesive zone model is a grain boundary deformation model which has been developed, as described above, on the basis of viscous flow rules of the boundary material. This model is supported by dwell crack growth experiments carried out at the three temperatures mentioned above, in both air and vacuum environments. Results of these tests have identified the frequency range in which the grain boundary cohesive zone model is applicable and also provided data to calculate the grain boundary activation energy as well as identifying the relative contributions of creep and environment in the critical sliding displacement leading to failure. Validation of the cohesive zone model has been carried out by comparing the simulated crack growth data with that obtained experimentally. This comparison is used to optimize the different model components and to provide a route to assess the relative significance of each of these components in relation to the intergranular damage associated with dwell fatigue crack growth in the ME3 alloy. For this purpose, a set of case studies were performed in order to illustrate the sensitivity of the cohesive zone model to variations in microstructure parameters (γ' statistics and grain boundary morphology) examined within the range of temperatures utilized in this study.