Search

Combined in-situ neutron diffraction measurements during post-processing heat treatment and thermo-mechanical crystal plasticity finite element (CPFE) simulations were utilized to study the residual phase stress development in the two-phase microstructure of an additively manufactured (AM) 304L stainless steel. The steel, fabricated via the laser engineering net shaping technique, has a microstructure comprising primarily of the austenite phase, with ∼ 2.5% ferrite phase by volume fraction. The post-build material was heated to greater than 1300 K and neutron diffraction data was recorded during heating and cooling. Specifically, the evolution of lattice strains in the individual phases were measured with temperature and the corresponding coefficients of thermal expansion (CTEs) calculated. The dislocation densities, phase fractions and textures, before and after heat treatment, were also measured. CPFE simulations were performed to study the interplay of the stress-free thermal strains and the mechanical strains in inducing inter-granular residual stresses in individual phases. The simulations confirmed the presence of process induced inter-granular residual stress primarily in the ferrite phase of the as-built AM material. Comparison of the relevant simulation data with experiments indicate that model predictions of the lattice strains and CTEs in both phases, as well as the inter-granular residual phase stress and pressure in the ferrite phase are in qualitative agreement with the experimental measurements and calculations.

Three 316L stainless steel materials are studied and reported upon; wrought, as-built additively manufactured (AM), and heat-treated AM material. The AM material was produced from the laser engineered net shaping (LENS) process. Macroscopic uniaxial compression stress-strain curves were obtained for all three materials. The curves were similar for the wrought and heat-treated AM materials but the as-built AM material demonstrated approximately 1.7 times greater flow stress at any given level of strain than the other two materials. Electron-backscatter diffraction analysis of the materials also showed that the microstructures of the three materials differed; with complex grain morphology for the as-built AM material. The mean grain size of each of the three materials also differed. The initial dislocation density was also measured with neutron diffraction and line-profile analysis for both wrought and as-built AM materials with the density in the AM material approximately 2.5 times greater. A single crystal model was proposed to represent the essential features of the three FCC materials accounting for dislocation interactions and representation of grain size via a simple Hall-Petch type term. The strength of this term is evaluated through independent experimental results on traditionally manufactured materials. The model was applied to each of the three materials by simulation of the uniaxial compression experiments by direct numerical simulation of electron-backscatter diffraction images. This allowed for representation of the size of each grain in the simulations. The results suggest that the difference in initial dislocation density of the three materials is the primary factor causing the difference in stress-strain response. Although the differences in grain size contribute to a higher stress for the as-built AM material, the effect is small. Other factors such as internal stress and grain morphology could play a role in mechanical behavior difference and these two factors are also discussed.

Thermomechanical cycling of a Ni49.9Ti50.1 (at.%) shape memory alloy was investigated. Combined ex situ macroscopic experiments and in situ neutron diffraction measurements were performed to relate the macroscopic evolution in behavior (e.g., dimensional instabilities) observed during thermal cycling to the responsible microscopic mechanism(s) through texture, internal strain, peak shape, and phase evolution from the neutron data. Pre-deformation in the austenite or martensite phases affected the macroscopic cyclic behavior (e.g., actuation strain), depending on the level of pre-strain and the associated microstructural changes. However, the pre-deformation did not completely stabilize the cyclic response. Subsequent thermomechanical cycling revealed that the martensite texture changed with continued thermal cycling, while the austenite texture did not. For the conditions investigated, stagnation of the martensite texture occurred around the eighth cycle, consistent with asymptotic saturation of the macroscopic transformation strains. Moreover, diffraction spectra peak shapes (broadening) were found to vary with cycling indicative of the accumulation of lattice defects, consistent with the constant increase in residual strain.

Using in situ neutron diffraction technique we investigated the temperature dependent deformation mechanisms in ultrafine grained (UFG) austenitic Fe–14Cr–16Ni alloy prepared by equal channel angular pressing. Tensile test studies show diminished ductility when testing temperature increased from 20 to 200 °C. At 200 °C, non-linear lattice strain deviation on [2 0 0] orientation proceeded plastic yielding by a large margin, accompanied by a greater distortion of crystal structure. In addition, the capability to accumulate dislocations was substantially reduced at 200 °C as evidenced by lower dislocation density than that at 20 °C. Dynamic recovery expedited at elevated temperature because of enlarged critical separation distance for annihilation of dislocation dipoles via climb. Calculations show that both high angle grain boundaries and thermal kinetic energy assisted the reduction of vacancy formation energy.

An experimental and theoretical investigation was carried out to study the strain rate dependent plastic response of beryllium over a wide range of applied compression strain rates, 10 −4 –10 4 /s. At each rate, the evolution of flow stress and the final texture with deformation was obtained from a non-textured hot-pressed (HP) sample and a textured rolled sheet. The rolled sheet material was compressed in both the in-plane (IP) and through-thickness (TT) direction for comparison. The twin volume fraction was determined from the change in texture. The activity of twinning was strongly dependent on strain rate in the IP and HP samples. We applied a multi-scale constitutive model for hexagonal close packed polycrystals that accounts for crystallographic slip and twinning on individual systems in each crystal, as well as twin reorientation. Rate effects enter the calculations only through thermally activated dislocation glide on the active slip modes. The importance of this study is that it points to the necessity of using a crystallographic model based on microstructure evolution to understand the role played by plastic anisotropy, slip–slip competition, and slip–twin competition, in the mechanical response of HCP aggregates. The model reproduces the observed flow curves and texture evolution for all tests with a unique single crystal set of parameters.