Wednesday 8:30 AM

March 1, 2017

Room: 11A

Location: San Diego Convention Ctr

Five decades ago, Mura has shown that the stress field in an anisotropically elastic material generated by a periodic dislocation network can be expressed analytically in reciprocal space. One can thus obtain Peach-Koehler forces on individual dislocations and determine their velocities. However, this is not possible at the mesoscale where the uncertainty in precise positions of crystal dislocations within a given mesocell gives rise to large configurational entropy. Here, we develop minimal version of a mesoscopic model, where both enthalpic and entropic contributions are taken into account. The performance of the model is demonstrated by studying the glide of 1/2<111> dislocations on {110} planes in bcc Mo. The activation enthalpy to move screw segments on different slip planes is taken from our thermodynamic model that is based on molecular statics studies of isolated screw dislocations.

Nickel-based superalloys are used extensively in applications requiring high strength and creep and fatigue resistance at elevated temperatures, such as aircraft turbine engines and power plants. Their mechanical properties are determined by the motion and cross slip of dislocations in the Ni matrix and Ni

In computational material design/discovery, it is not uncommon to have a situation in which knowledge of one or more properties of a material system is prerequisite to the optimization/prediction of another interested property, e.g. assessing plastic deformations requires information of shear modulus. The problem is that such prior knowledge is costly to achieve for all possible states of the system, e.g. different compositions, and hence is not always available. To address this problem, we propose in this work an efficient multi-step optimization which consists of two on-the-fly predictors: The first predictor predicts the prerequisite properties for all possible states of the system and the second predictor makes use of the first predictions to predict the state that may exhibit the optimal property of interest. Upon finishing one multi-step prediction, experiment/calculation will be conducted for assessing both prerequisite and interested properties at the predicted state and resulting data will be used for updating the predictors in the next multi-step prediction. To demonstrate the efficiency of the optimization scheme, an example, in which bulk modulus is maximized among the 211-MAX phases assuming that the prerequisite a and c lattice parameters of all MAX phases are not known in advance, is given.

Hyperuniform materials, which possess unique structural and physical properties, have attracted increasing attention in biology, physics and engineering. The preponderance of known hyperuniform materials consisting of discrete tailored building blocks such as colloidal particles, and bi-continuous hyperuniform structures have not been systematically investigated. In this talk, we present an integrated procedure for designing hyperuniform two-phase material microstructure via an inverse optimization reconstruction method by controlling the long range and short range order via corresponding spatial correlation functions. We numerically generate a wide spectrum of virtual materials and apply a novel lattice-particle method to study their mechanical behavior. We find that hyperuniform materials possess superior mechanical properties over their non-hyperuniform counterpart, i.e, one with same short-range order but lacking of hyperuniformity.

Thanks to the materials project, the elastic constants of thousands of compounds are now available, allowing for the screening of all manner of properties. An anomalous material of particular interest is Gum Metal, a class of Ti-Nb alloys that exhibits superelasticity, and high strength. Previous studies have tied the behavior of Gum Metal to a high elastic anisotropy related to the dislocation core radius, which is a direct result of the material’s presence near an elastic instability. By calculating the dislocation core radius of all cubic materials in the elastic constants database Gum Metal candidates are identified. It is shown by incorporating the third-order elastic constants into the elastic stability conditions, the intrinsic ductility of these alloys can be estimated from first principles, presenting an efficient approach for identifying potential high strength, ductile materials. Funding is given by the Department of Energy's Basic Energy Sciences program, grant No. EDCBEE.

Predicting the microstructure evolution that accompanies thermomechanical processing remains a considerable challenge. We simulate rolling and recrystallization using mean field approaches. For rolling to high reductions we use the viscoplastic self-consistent (VPSC) polycrystal plasticity model that has been extended to incorporate calculations of intragranular misorientations. Experimentally, intragranular misorientations are caused by the accumulation of geometrically necessary defects that are part of the driving force for recrystallization. To predict recrystallization evolution, a mean field model is used with nucleation and growth rules based on calculated misorientation and strain energy. We compare predicted and experimental deformation and recrystallization textures for copper and uranium. We find that the extended VPSC model better predicts deformation textures than conventional VPSC, and that the recrystallization model can capture well many of the recrystallization texture features.

High strength and high ductility are two major desirable goals for metal and alloy. Most traditional strengthening techniques, such as grain size refinement in nanocrystalline metal, can introduce high strength but with the cost of losing ductility. This strength-ductility trade-off can be overcome by the introduction of ultra-fine twin boundaries in polycrystalline metal. Ultrafine growth twins are common microstructure in chemical vapor deposited Nickel metal, but with low thermal stability. Due to the limitation of the experiments, computational simulation is the only possible method to study the atomistic migration of thermal detwinning. We used atomic-scale simulations to validate detwinning through the migration incoherent twin boundary. The driving force for detwinning is found to be dependent on the length and density of the incoherent twin boundary. The remaining dislocation structures after detwinning are attributed to the existence of vacancy near the incoherent twin boundary.

First principle calculation is used to develop a Al-Zn-Ni alloy with improved hardness (via precipitation hardening) and electrical conductivity for power line cables conductor. Ab initio density functional theory (DFT) simulations are used to screen combinations of Al-Zn-TM (TM = transition metals) to efficiently establish a TM alloy candidate (Al-Zn-Ni) with promising electrical conductivity. The enthalpies of formation of this alloy system was also computed using DFT simulation to identify its most stable precipitate phase as L12. This computational screening is followed by fabrication for age hardening and characterization of the Al-Zn-Ni system. The corresponding hardness and electrical conductivity measurements show a strength and electrical conductivity comparable to the Al alloys currently used for power line transmission cables.

Structural metrics to characterize grain boundary networks (GBNs) have included cluster sizes, Betti numbers, and triple junction fractions. However, all of these methods employ a binary taxonomy of GBs as either “special” or “general”. The grain boundary character distribution considers the entire spectrum of GB types, but does not encode any topological information. Microstructure design in the context of GBNs would be greatly facilitated by a structural metric capable of encoding both topological and crystallographic structure. We present a new approach for characterizing the structure of GBNs using spectral graph theory. This method is capable of handling a continuous spectrum of GB types and naturally indicates the important features of the microstructure that govern the effective response of the polycrystal. We illustrate the correlation between this spectral metric for GBN structure and the effective properties of model polycrystals and discuss the potential for designing GB networks with tailored properties.

As the demand for low wear materials grows so does the need for accurate, fast, and efficient wear predictions. Abrasive wear occurs when a harder material is rubbing against softer materials. Predictions of the removal of material from a solid surface are useful for estimating component or device service life and can be used as a component of mechanical design. A great deal of attention has been given to predicting the topographical evolution of a wearing surface. These predictions apply physics-based models that relate geometry, pressure, and material wear properties for a given configuration of materials. However, to date, these models have never been integrated into optimization design protocols that prescribe initial configurations or topologies for optimal wear performance. This presentation will highlight the development and validation of a topology optimization tool to aid in the design of the wear surfaces of bi-material composites with prescribed properties.