Enhancing corrosion-resistant alloy design through natural language processing and deep learning.

We propose strategies that couple natural language processing with deep learning to enhance machine capability for corrosion-resistant alloy design. First, accuracy of machine learning models for materials datasets is often limited by their inability to incorporate textual data. Manual extraction of numerical parameters from descriptions of alloy processing or experimental methodology inevitably leads to a reduction in information density. To overcome this, we have developed a fully automated natural language processing approach to transform textual data into a form compatible for feeding into a deep neural network. This approach has resulted in a pitting potential prediction accuracy substantially beyond state of the art. Second, we have implemented a deep learning model with a transformed-input feature space, consisting of a set of elemental physical/chemical property-based numerical descriptors of alloys replacing alloy compositions. This helped identification of those descriptors that are most critical toward enhancing their pitting potential. In particular, configurational entropy, atomic packing efficiency, local electronegativity differences, and atomic radii differences proved to be the most critical.

Accelerating the design of compositionally complex materials via physics-informed artificial intelligence.

The chemical space for designing materials is practically infinite. This makes disruptive progress by traditional physics-based modeling alone challenging. Yet, training data for identifying composition-structure-property relations by artificial intelligence are sparse. We discuss opportunities to discover new chemically complex materials by hybrid methods where physics laws are combined with artificial intelligence.

Phase-Field Modeling of Chemoelastic Binodal/Spinodal Relations and Solute Segregation to Defects in Binary Alloys.

Microscopic phase-field chemomechanics (MPFCM) is employed in the current work to model solute segregation, dislocation-solute interaction, spinodal decomposition, and precipitate formation, at straight dislocations and configurations of these in a model binary solid alloy. In particular, (i) a single static edge dipole, (ii) arrays of static dipoles forming low-angle tilt (edge) and twist (screw) grain boundaries, as well as at (iii) a moving (gliding) edge dipole, are considered. In the first part of the work, MPFCM is formulated for such an alloy. Central here is the MPFCM model for the alloy free energy, which includes chemical, dislocation, and lattice (elastic), contributions. The solute concentration-dependence of the latter due to solute lattice misfit results in a strong elastic influence on the binodal (i.e., coexistence) and spinodal behavior of the alloy. In addition, MPFCM-based modeling of energy storage couples the thermodynamic forces driving (Cottrell and Suzuki) solute segregation, precipitate formation and dislocation glide. As implied by the simulation results for edge dislocation dipoles and their configurations, there is a competition between (i) Cottrell segregation to dislocations resulting in a uniform solute distribution along the line, and (ii) destabilization of this distribution due to low-dimensional spinodal decomposition when the segregated solute content at the line exceeds the spinodal value locally, i.e., at and along the dislocation line. Due to the completely different stress field of the screw dislocation configuration in the twist boundary, the segregated solute distribution is immediately unstable and decomposes into precipitates from the start.

Effect of Twin Boundary Motion and Dislocation-Twin Interaction on Mechanical Behavior in Fcc Metals.

The interplay of interface and bulk dislocation nucleation and glide in determining the motion of twin boundaries, slip-twin interaction, and the mechanical (i.e., stress-strain) behavior of fcc metals is investigated in the current work with the help of molecular dynamics simulations. To this end, simulation cells containing twin boundaries are subject to loading in different directions relative to the twin boundary orientation. In particular, shear loading of the twin boundary results in significantly different behavior than in the other loading cases, and in particular to jerky stress flow. For example, twin boundary shear loading along 〈 112 〉 results in translational normal twin boundary motion, twinning or detwinning, and net hardening. On the other hand, such loading along 〈 110 〉 results in oscillatory normal twin boundary motion and no hardening. As shown here, this difference results from the different effect each type of loading has on lattice stacking order perpendicular to the twin boundary, and so on interface partial dislocation nucleation. In both cases, however, the observed stress fluctuation and "jerky flow" is due to fast partial dislocation nucleation and glide on the twin boundary. This is supported by the determination of the velocity and energy barriers to glide for twin boundary partials. In particular, twin boundary partial edge dislocations are significantly faster than corresponding screws as well as their bulk counterparts. In the last part of the work, the effect of variable twin boundary orientation in relation to the loading direction is investigated. In particular, a change away from pure normal loading to the twin plane toward mixed shear-normal loading results in a transition of dominant deformation mechanism from bulk dislocation nucleation/slip, to twin boundary motion.

Publisher Correction: Unveiling the Re effect in Ni-based single crystal superalloys.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Nonlinear membrane model for large amplitude vibration of single layer graphene sheets.

The nonlinear vibrational properties of single layer graphene sheets (SLGSs) are investigated using a membrane model. The nonlinear equation of motion is considered for the SLGSs by including the effects of stretching due to large amplitudes. The equation of motion is numerically solved utilizing the finite difference method for SLGSs with different initial and boundary conditions, sizes and pretensions. It is concluded that the nonlinear fundamental frequency of SLGSs increases by increasing the pretension and initial velocity. In addition, it is observed that an increase in the pretension weakens the effects of the initial velocity on the fundamental frequency, such that the fundamental frequency approximately becomes independent of the initial velocity. This is an important feature of the vibrating systems consisting of SLGSs which are used in the nano-electromechanical systems (NEMS), where resonators with a specific fundamental frequency and independent of the initial velocity are of interest.