Abstract
Conventional models of power law creep assume that deformation is accommodated by the activation of dislocation glide and climb, where climb operates as a recovery mechanism that facilitates the continued glide of dislocations past obstacles. The current work explores the utility of a model in which climb can accommodate significant strain independently of glide and may explain changes in the constitutive behavior, such as texture evolution, strain anisotropy, activation energy, and strain rate sensitivity. Specifically, strain accommodation via the climb of <a> dislocations within hexagonal close-packed (HCP) alloys of Mg was modeled using a climb incorporating viscoplastic self-consistent model (VPSC-CLIMB). The parameters governing the simulation results of this model were optimized to experimental texture and strain anisotropy measurements of as-rolled and annealed Mg alloy AZ31B deformed in tension along the sheet rolling direction (RD) and the in-plane transverse direction (TD) (at temperatures ranging from 20 to 350C and strain rate ranging from 10-5 to 10-1 1/s) by genetic algorithm (GA). Broadening the experimental and computational (VPSC-CLIMB) study to another alloy revealed that climb was also an important strain accommodation mechanism in the dilute Mg alloy ZK10; however, the climb of both <a> and <c> dislocations was necessary to match the experimental strain anisotropy. X-ray Line Profile Analysis (XLPA) was performed to estimate the mobile dislocation densities for <a> climb in AZ31B and <a> and <c> climb in ZK10, and it is suggested that both alloy chemistry and texture may play a role in determining which climb modes are active.
Simulations performed in intermediate temperatures also suggest that strain accommodation via climb may explain the evolution in texture evolution and strain anisotropy observed experimentally, though the role of <c+a> slip in these temperatures is currently unknown as the loading conditions selected could not excite its activation. The assumption of independent climb and glide was investigated using predictions of the rate sensitivity and activation energy could be compared with experimentally obtained values. The agreement was acceptable over a wide range of conditions and was especially good in the power law regime. Obviously, as more climb occurs, the simulations become more rate sensitive than those that only include glide. This generally enables the model to describe the transition from thermally activated plasticity at low temperatures and higher strain rates to power law climb and glide at higher temperatures and lower strain rates. The practical utility of the model was explored in the context of forming limit diagram (FLD) prediction, which is a rather rigorous test given the fact that model parameterization was performed solely on the basis of uniaxial in-plane tensile testing, whereas FLDs explore different straining conditions (e.g., plane strain, equibiaxial stretching). The fundamental hypothesis being explored was whether or not the transition from thermally activated plasticity to climb accommodated flow might explain the transition in formability exhibited by Mg alloys, from very limited cold formability to highly formable at moderately elevated temperatures (200 - 250C) by predicting forming limit diagrams using the aforementioned VPSC-CLIMB model within the context of the Marciniak and Kuczyński (M-K) forming limit prediction. First, it had to be admitted that the initial experimental data set was inadequate to constrain the parameter (especially that governing the slip of <c+a> slip) selection over the entire range of straining paths required for forming limit prediction. Experimental forming data was used to further refine the model parameters. Second, in the conditions explored (200 to 250°C and strain rates of ~10-3 - 10-2 1/s) climb was found to be active only at low rates (~10-3 1/s) or higher temperatures (250°C). At lower temperatures (150 °C), <c+a> slip accommodated significant amounts of strain, suggesting that a transition between a regime where <c+a> slip accommodates significant strain at lower temperatures and a regime where <a> climb begins to accommodate strain at higher temperatures exists. When climb was active, particularly during uniaxial straining, there was a concomitant increase in formability. Third, analysis of forming limit simulations provides a new, though retrospectively obvious, insight regarding the onset of plastic instability within materials undergoing flow by multiple deformation mechanisms. The onset of plastic instability was found to occur when either 1) texture evolution deactivates a major slip mode due to Schmid effects or 2) increases in the defect strain rate causes the deactivation of more rate sensitive mechanisms (i.e., climb).
