The next generation of advanced materials is tending toward increasingly complex compositions. Synthesizing precise composition is time-consuming and becomes exponentially demanding with increasing compositional complexity. An experienced human operator does significantly better than a novice but still struggles to consistently achieve precision when synthesis parameters are coupled. The time to optimize synthesis becomes a barrier to exploring scientifically and technologically exciting compositionally complex materials. This investigation demonstrates an active learning (AL) approach for optimizing physical vapor deposition synthesis of thin-film alloys with up to five principal elements. We compared AL-based on Gaussian process (GP) and random forest (RF) models. The best performing models were able to discover synthesis parameters for a target quinary alloy in 14 iterations. We also demonstrate the capability of these models to be used in transfer learning tasks. RF and GP models trained on lower dimensional systems (i.e., ternary, quarternary) show an immediate improvement in prediction accuracy compared to models trained only on quinary samples. Furthermore, samples that only share a few elements in common with the target composition can be used for model pre-training. We believe that such AL approaches can be widely adapted to significantly accelerate the exploration of compositionally complex materials.
A solid state phase transformation of Ti-6Al-4V was studied using high speed in situ x-ray diffraction measurements made during rapid cooling of a cold metal transfer arc weld bead deposited onto a water cooled substrate. Analysis of body centered cubic (BCC) and hexagonal close packed (HCP) lattices revealed an abrupt, nonlinear shift in the lattice parameters of both phases just after the HCP phase had nucleated. Postmortem transmission electron microscopy confirmed that V diffusion was mostly suppressed during cooling. Together, these results indicate that at this cooling rate of approximately 104 K/s, which is representative of cooling rates of many additive manufacturing and welding processes, kinematic coherency of the BCC–HCP interfaces gives rise to the anomalous lattice expansion and contraction behaviors of both phases during the initial nucleation and growth stages of (mostly) martensitic transformation from BCC to HCP; the role of diffusion in such lattice anomalies is shown to be minimal.
Correlative characterization often begins with non-destructive evaluation, such as X-ray computed tomography (XCT) and X-ray microscopy (XRM). Regions or features of interest are identified in the 3D X-ray datasets and then, after identification, are characterized in other, destructive imaging modalities. In the realm of modern X-ray imaging, a critical challenge arises: the pursuit of sub-micron-level defect detection within large volumes. Countless material applications demand statistically relevant samplings of defects, encompassing secondary phases, inclusions, phase segregation, cracks, pores, delamination, and grain structures. However, there is an inherent tradeoff between the volume that can be imaged and the resolution of the images. Imaging at a large volume may yield a statistically relevant sample size but often suffers from poor resolution, rendering it inadequate for detecting these elusive defects. Conversely, high-resolution imaging captures intricate details but confines the examination to a small volume, raising questions about statistical representativeness.