 |
Amy ClarkeShape memory effect deformation structures in a uranium-niobium alloyUnderstanding the mechanical behavior and texture evolution of the uranium-niobium alloy U-14Nb (atomic percent, or at. %) is important for the development of future predictive constitutive models that describe deformation behavior of this material. In this work, detailed microstructure evolution studies are performed on deformed material, and experimental observations are compared to predictions from a single-crystal model. Uranium’s high-temperature phase, gamma (γ)-U, has a body-centered cubic (bcc) crystal structure, while the room-temperature phase, alpha (α)-U, is orthorhombic. When U-14Nb is quenched from the γ-U phase to room temperature, it undergoes a martensite transformation to a monoclinically distorted version of α-U, referred to as α″. Martensite transformations occur by means of lattice shear, and there are twelve possible orientations (or variants) of the monoclinic α″ in the cubic γ, all with transformation strains in different directions. These variants are all equally represented in the as-transformed structure, so there is no net macroscopic strain. They are separated by twin boundaries, across which the structure is mirrored.
When the martensite is strained, variants whose transformation strain can best accommodate the imposed strain will grow at the expense of others, via the motion of the twin boundaries, or nucleation of new twins with similar orientation relationships. There is only one way the α″ → γ reverse transformation can proceed for each variant upon reheating, so any strain produced in this manner is reversed. So, if a piece of U-14Nb is bent at room temperature, it will straighten itself upon reheating. This phenomenon is known as the shape memory effect (SME). A group of Los Alamos colleagues, including Robert D. Field, Rodney J. McCabe, Donald W. Brown, Robert E. Hackenberg, Dan J. Thoma, Patricia O. Dickerson, John G. Swadener, and Carl M. Cady, has deformed U-14Nb specimens in compression or tension, followed by characterization of the deformation structures. Earlier studies employed transmission electron microscopy (TEM) as the main characterization tool. More recently, orientation imaging microscopy (OIM) has been used to characterize U-14Nb microstructures after shape memory effect deformation.
Orientation imaging microscopy results for samples strained in compression and tension within the shape memory effect regime are shown in the figures above. The colors shown in the OIM maps correspond to different orientations along the stress axis. The standard α″ stereographic projection at the upper right serves as a legend for the color scheme. A large majority of the orientations are blue after compressive deformation and green after tensile deformation, corresponding to orientations located on the left (blue) and right (green) side of the legend shown at the top right of the page. A stereographic projection that shows the location of several α″ poles is shown below. Orientation relationships between the α″ martensite and the γ phase are also highlighted by the addition of a γ standard triangle for the most favored variant (MFV) of martensite, or the variant that best accommodates the imposed strain, anticipated to dominate in compression or tension.
The most favored variant for deformation in compression was determined to be #10 of the twelve possible martensite variants (referred to as Hatt variants, or HVs, after B.A. Hatt, reported in 1966 in the Journal of Nuclear Materials), while the most favored variant anticipated to dominate in tension was determined to be HV #5. The orientations shown in blue (compression example) and green (tension example) in the OIM maps and corresponding legend fall within the γ standard triangles for HV #10 and HV #5, the most favored variants for compression and tension, respectively, as predicted. Additional analyses of the orientation imaging microscopy results were performed on individual grains from both the compression and tensile specimens. From a discrete α″ inverse pole figure (not shown), similar orientations in a grain were colored the same to distinguish the different martensite variants. Orientation relationships between the α″ martensite and the γ phase were then used to assign the appropriate number to each variant. A γ orientation for each Hatt variant was plotted on a γ standard triangle (with a circle to represent the spread in the data), and a stress axis was estimated from the locus of γ orientations for each grain analyzed. Accommodation strains for each Hatt variant were calculated from the approximated γ stress axis. Example results from a grain after compressive shape memory effect deformation are provided. Negative strain values were calculated for nearly all of the observed Hatt variants present in the compression example. HV #10, colored green, had the highest calculated compressive strain, as expected for the most favored variant for compression. Three additional Hatt variants, colored white, light gray, and dark gray, also compose much of the grain, but have comparable strain levels to that of HV #10. Note that the γ orientation is close to [001], a symmetry position at which all four of these Hatt variants would have the same calculated accommodation strain. Thus, it is not surprising that four Hatt variants, including HV #10, constitute the majority of the grain. Tensile strains were also calculated for Hatt variants located near the grain boundaries, which likely resulted from intergranular tensile strains. The light blue color highlights one such Hatt variant, which represents a very minor constituent of the grain. Twin boundaries between the variants were also identified with orientation imaging microscopy, and were in good agreement with predicted shape memory effect twinning systems. To study finer details of the twin structures, a focused ion beam (FIB) was used to prepare thin foils for further examination with transmission electron microscopy from specific regions of interest. Coarse twins identified with orientation imaging microscopy were confirmed with transmission electron microscopy by using select area diffraction, while finer details of the microstructure were only observable with transmission electron microscopy. Thus, orientation imaging microscopy was first used to analyze large areas, followed by subsequent transmission electron microscopy to provide finer microstructural details. By the combination of these two complementary techniques, a predicted shape memory effect twinning system not previously observed experimentally was recently confirmed.
In this work, the single-crystal model is used to predict Hatt variants from accommodation strains and twin relationships in polycrystalline U-14Nb after shape memory effect deformation. Extension to consider post-shape memory effect (i.e., unrecoverable) deformation processes is currently under way. In addition, micro-compression samples (~15 and ~30 microns in diameter and length, respectively) from individual γ grains of a polycrystalline sample are being tested in a nano-indenter instrument with a flat punch to study the influence of γ orientation on stress-strain response and to further test the single crystal model. The stress-strain behaviors exhibited by two preliminary test samples differed, suggesting that orientation will influence the deformation response. The preliminary results are promising and support further testing with this technique.
Next: An investigation of the reactivity of a trans-bis(imido) uranium(VI) complex |
|
The
images above are of pre- and post-test preliminary microsamples.
Initial microsample heights and diameters were ~30 microns
and ~15 microns, respectively. The microsamples were fabricated
from single γ grains
of a polycrystalline sample using a focused ion beam. Compression
testing was performed with a nano-indenter using a flat punch.
The differences in stress-strain response (right) are thought
to be associated with orientation dependence. Shear instability
(the white arrow in the image immediately to the left) is
highlighted in a post-test image for one of the preliminary
samples tested, and is also reflected (the red arrow in the
data plot above) in the stress-strain response for this test.
Amy
Clarke works in the Materials Science & Technology Division’s Metallurgy
Group (MST-6). She studied ferrous physical metallurgy at the Colorado
School of Mines, where she received her doctorate in 2006. Amy began her
postdoctoral appointment in May 2006. Her mentor is Robert Field of MST-6
and the Materials Design Institute, a collaborative research and educational
program with the University of California, Davis. Clarke recently received
the Willy Korf Award for Young Excellence for her Ph.D. work. This award
is given annually to one graduate student worldwide, to recognize the
contribution of their graduate research to the steel industry.
