Prognosis of COVID-19 severity using DERGA, a novel machine learning algorithm.

It is important to determine the risk for admission to the intensive care unit (ICU) in patients with COVID-19 presenting at the emergency department. Using artificial neural networks, we propose a new Data Ensemble Refinement Greedy Algorithm (DERGA) based on 15 easily accessible hematological indices. A database of 1596 patients with COVID-19 was used; it was divided into 1257 training datasets (80 % of the database) for training the algorithms and 339 testing datasets (20 % of the database) to check the reliability of the algorithms. The optimal combination of hematological indicators that gives the best prediction consists of only four hematological indicators as follows: neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase, ferritin, and albumin. The best prediction corresponds to a particularly high accuracy of 97.12 %. In conclusion, our novel approach provides a robust model based only on basic hematological parameters for predicting the risk for ICU admission and optimize COVID-19 patient management in the clinical practice.

Colloidal Shear-Thickening Fluids Using Variable Functional Star-Shaped Particles: A Molecular Dynamics Study.

Complex colloidal fluids, depending on constituent shapes and packing fractions, may have a wide range of shear-thinning and/or shear-thickening behaviors. An interesting way to transition between different types of such behavior is by infusing complex functional particles that can be manufactured using modern techniques such as 3D printing. In this paper, we perform 2D molecular dynamics simulations of such fluids with infused star-shaped functional particles, with a variable leg length and number of legs, as they are infused in a non-interacting fluid. We vary the packing fraction (ϕ) of the system, and for each different system, we apply shear at various strain rates, turning the fluid into a shear-thickened fluid and then, in jammed state, rising the apparent viscosity of the fluid and incipient stresses. We demonstrate the dependence of viscosity on the functional particles' packing fraction and we show the role of shape and design dependence of the functional particles towards the transition to a shear-thickening fluid.

Materials Informatics for Mechanical Deformation: A Review of Applications and Challenges.

In the design and development of novel materials that have excellent mechanical properties, classification and regression methods have been diversely used across mechanical deformation simulations or experiments. The use of materials informatics methods on large data that originate in experiments or/and multiscale modeling simulations may accelerate materials' discovery or develop new understanding of materials' behavior. In this fast-growing field, we focus on reviewing advances at the intersection of data science with mechanical deformation simulations and experiments, with a particular focus on studies of metals and alloys. We discuss examples of applications, as well as identify challenges and prospects.

A Molecular Dynamics Simulations Study of the Influence of Prestrain on the Pop-In Behavior and Indentation Size Effect in Cu Single Crystals.

The pop-in effect in nanoindentation of metals represents a major collective dislocation phenomenon that displays sensitivity in the local surface microstructure and residual stresses. To understand the deformation mechanisms behind pop-ins in metals, large scale molecular dynamics simulations are performed to investigate the pop-in behavior and indentation size effect in undeformed and deformed Cu single crystals. Tensile loading, unloading, and reloading simulations are performed to create a series of samples subjected to a broad range of tensile strains with/without pre-existing dislocations. The subsequent nanoindentation simulations are conducted to investigate the coupled effects of prestrain and the presence of resulting dislocations and surface morphology, as well as indenter size effects on the mechanical response in indentation processes. Our work provides detailed insights into the deformation mechanisms and microstructure-property relationships of nanoindentation in the presence of residual stresses and strains.

Λ-Invariant and Topological Pathways to Influence the Strength of Submicron Crystals.

In small volumes, sample dimensions are known to strongly influence mechanical behavior: especially strength and crystal plasticity. This correlation fades away at the so-called "mesoscale," loosely defined at several micrometers in both experiments and simulations. However, this picture depends on the "entanglement" of the initial defect configuration. In this Letter, we study the effect of dislocation topology through the use of a novel observable for dislocation ensembles (the Λ invariant) that depends only on mutual dislocation linking: It is built on the natural vortex character of dislocations, and it has a continuum-discrete correspondence that may assist multiscale modeling descriptions. We investigate arbitrarily complex initial dislocation microstructures in sub-micron-sized pillars using three-dimensional discrete dislocation dynamics simulations for finite volumes. We demonstrate how to engineer nanoscale dislocation ensembles that are independent from sample dimensions, either by biased-random dislocation loop deposition or by sequential mechanical loads of compression and torsion.

Learning to Predict Crystal Plasticity at the Nanoscale: Deep Residual Networks and Size Effects in Uniaxial Compression Discrete Dislocation Simulations.

The density and configurational changes of crystal dislocations during plastic deformation influence the mechanical properties of materials. These influences have become clearest in nanoscale experiments, in terms of strength, hardness and work hardening size effects in small volumes. The mechanical characterization of a model crystal may be cast as an inverse problem of deducing the defect population characteristics (density, correlations) in small volumes from the mechanical behavior. In this work, we demonstrate how a deep residual network can be used to deduce the dislocation characteristics of a sample of interest using only its surface strain profiles at small deformations, and then statistically predict the mechanical response of size-affected samples at larger deformations. As a testbed of our approach, we utilize high-throughput discrete dislocation simulations for systems of widths that range from nano- to micro- meters. We show that the proposed deep learning model significantly outperforms a traditional machine learning model, as well as accurately produces statistical predictions of the size effects in samples of various widths. By visualizing the filters in convolutional layers and saliency maps, we find that the proposed model is able to learn the significant features of sample strain profiles.

On the shear dilation of polycrystalline lubricant films in boundary lubricated contacts.

Shearing of a solidified polycrystalline lubricant film confined between two solid surfaces has been studied by molecular dynamics simulations. In the case of a perfect commensurate contact, we observe interlayer slips within the film and shear-induced order-to-disorder transition of lubricant molecules around grain boundaries. This process is accompanied by the nucleation, propagation, and annihilation of dislocations in the solidified film, resulting in repeated dilation and collapse of the lubricant film during the stick-slip motion. In the case of an incommensurate contact, only slips at the lubricant-solid interface happen and no dilation of the lubricant film is observed during the stick-slip friction. These observations are consistent with recent surface force balance experimental measurements. In combination with our recent work [R. G. Xu and Y. S. Leng, Proc. Natl. Acad. Sci. U. S. A. 115, 6560 (2018)], this study provides a renewed picture on the physical property of nanoconfined lubricant films in boundary lubrication.

Spatial strain correlations, machine learning, and deformation history in crystal plasticity.

Systems far from equilibrium respond to probes in a history-dependent manner. The prediction of the system response depends on either knowing the details of that history or being able to characterize all the current system properties. In crystal plasticity, various processing routes contribute to a history dependence that may manifest itself through complex microstructural deformation features with large strain gradients. However, the complete spatial strain correlations may provide further predictive information. In this paper, we demonstrate an explicit example where spatial strain correlations can be used in a statistical manner to infer and classify prior deformation history at various strain levels. The statistical inference is provided by machine-learning techniques. As source data, we consider uniaxially compressed crystalline thin films generated by two dimensional discrete dislocation plasticity simulations, after prior compression at various levels. Crystalline thin films at the nanoscale demonstrate yield-strength size effects with very noisy mechanical responses that produce a serious challenge to learning techniques. We discuss the influence of size effects and structural uncertainty to the ability of our approach to distinguish different plasticity regimes.

Universality Class of Nanocrystal Plasticity: Localization and Self-Organization in Discrete Dislocation Dynamics.

The universality class of the avalanche behavior in plastically deforming crystalline and amorphous systems has been commonly discussed, despite the fact that the microscopic defect character in each of these systems is different. In contrast to amorphous systems, crystalline flow stress increases dramatically at high strains and/or loading rates. We perform simulations of a two-dimensional discrete dislocation dynamics model that minimally captures the phenomenology of nanocrystalline deformation. In the context of this model, we demonstrate that a classic rate dependence of dislocation plasticity at large rates (>10^{3}/s) fundamentally controls the system's statistical character as it competes with dislocation nucleation: At large rates, the behavior is statistically dominated by long-range correlations of "dragged" mobile dislocations. At small rates, plasticity localization dominates in small volumes and a spatial integration of avalanche behavior takes place.

Probing Microplasticity in Small-Scale FCC Crystals via Dynamic Mechanical Analysis.

In small-scale metallic systems, collective dislocation activity has been correlated with size effects in strength and with a steplike plastic response under uniaxial compression and tension. Yielding and plastic flow in these samples is often accompanied by the emergence of multiple dislocation avalanches. Dislocations might be active preyield, but their activity typically cannot be discerned because of the inherent instrumental noise in detecting equipment. We apply alternate current load perturbations via dynamic mechanical analysis during quasistatic uniaxial compression experiments on single crystalline Cu nanopillars with diameters of 500 nm and compute dynamic moduli at frequencies 0.1, 0.3, 1, and 10 Hz under progressively higher static loads until yielding. By tracking the collective aspects of the oscillatory stress-strain-time series in multiple samples, we observe an evolving dissipative component of the dislocation network response that signifies the transition from elastic behavior to dislocation avalanches in the globally preyield regime. We postulate that microplasticity, which is associated with the combination of dislocation avalanches and slow viscoplastic relaxations, is the cause of the dependency of dynamic modulus on the driving rate and the quasistatic stress. We construct a continuum mesoscopic dislocation dynamics model to compute the frequency response of stress over strain and obtain a consistent agreement with experimental observations. The results of our experiments and simulations present a pathway to discern and quantify correlated dislocation activity in the preyield regime of deforming crystals.

Shearing a glass and the role of pinning delay in models of interface depinning.

When a disordered solid is sheared, yielding is followed by the onset of intermittent response that is characterized by slip in local regions usually labeled shear-transformation zones. Such intermittent response resembles the behavior of earthquakes or contact depinning, where a well-defined landscape of pinning disorder prohibits the deformation of an elastic medium. Nevertheless, a disordered solid is evidently different in that pinning barriers of particles are due to neighbors that are also subject to motion. Microscopic yielding leads to destruction of the local microstructure and local heating. It is natural to assume that locally a liquid emerges for a finite timescale before cooling down to a transformed configuration. For including this characteristic transient in glass depinning models, we propose a general mechanism that involves a "pinning delay" time T(pd), during which each region that slipped evolves as a fluid. The new timescale can be as small as a single avalanche time step. This is a local, effective, and dynamical in nature mechanism that may be thought as dynamical softening. We demonstrate that the inclusion of this mechanism causes a drift of the critical exponents toward higher values for the slip sizes τ, until a transition to permanent shear-banding behavior happens causing almost oscillatory, stick-slip response. Moreover, it leads to a proliferation of large events that are highly inhomogeneous and resemble sharp slip band formation.

Statistics of frictional families.

We develop a theoretical description for mechanically stable frictional packings in terms of the difference between the total number of contacts required for isostatic packings of frictionless disks and the number of contacts in frictional packings, m=Nc0 - Nc. The saddle order m represents the number of unconstrained degrees of freedom that a static packing would possess if friction were removed. Using a novel numerical method that allows us to enumerate disk packings for each m, we show that the probability to obtain a packing with saddle order m at a given static friction coefficient µ, Pm(µ), can be expressed as a power series in µ. Using this form for Pm(µ), we quantitatively describe the dependence of the average contact number on the friction coefficient for static disk packings obtained from direct simulations of the Cundall-Strack model for all µ and N.

Universal properties of magnetization dynamics in polycrystalline ferromagnetic films.

We investigate the scaling behavior in the statistical properties of Barkhausen noise in ferromagnetic films. We apply the statistical treatment usually employed for bulk materials in experimental Barkhausen noise time series measured with the traditional inductive technique in polycrystalline ferromagnetic films having different thickness from 100 to 1000 nm and determine the scaling exponents. Based on this procedure, we group the samples in a single universality class, since the scaling behavior of Barkhausen avalanches is characterized by exponents τ∼1.5, α∼2.0, and 1/σνz∼ϑ∼2.0 for all the films. We interpret these results in terms of theoretical models and provide experimental evidence that a well-known mean-field model for the dynamics of a ferromagnetic domain wall in three-dimensional ferromagnets can be extended for films. We identify that the films present an universal three-dimensional magnetization dynamics, governed by long-range dipolar interactions, even at the smallest thicknesses, indicating that the two-dimensional magnetic behavior commonly verified for films cannot be generalized for all thickness ranges.

Computational studies of the glass-forming ability of model bulk metallic glasses.

Bulk metallic glasses (BMGs) are produced by rapidly thermally quenching supercooled liquid metal alloys below the glass transition temperature at rates much faster than the critical cooling rate R(c) below which crystallization occurs. The glass-forming ability of BMGs increases with decreasing R(c), and thus good glass-formers possess small values of R(c). We perform molecular dynamics simulations of binary Lennard-Jones (LJ) mixtures to quantify how key parameters, such as the stoichiometry, particle size difference, attraction strength, and heat of mixing, influence the glass-formability of model BMGs. For binary LJ mixtures, we find that the best glass-forming mixtures possess atomic size ratios (small to large) less than 0.92 and stoichiometries near 50:50 by number. In addition, weaker attractive interactions between the smaller atoms facilitate glass formation, whereas negative heats of mixing (in the experimentally relevant regime) do not change R(c) significantly. These results are tempered by the fact that the slowest cooling rates achieved in our simulations correspond to ~10(11) K/s, which is several orders of magnitude higher than R(c) for typical BMGs. Despite this, our studies represent a first step in the development of computational methods for quantitatively predicting glass-formability.

Isostaticity at frictional jamming.

Amorphous packings of frictionless, spherical particles are isostatic at jamming onset, with the number of constraints (contacts) equal to the number of degrees of freedom. Their structural and mechanical properties are controlled by the interparticle contact network. In contrast, amorphous packings of frictional particles are typically hyperstatic at jamming onset. We perform extensive numerical simulations in two dimensions of the geometrical asperity (GA) model for static friction to further investigate the role of isostaticity. In the GA model, interparticle forces are obtained by summing up purely repulsive central forces between periodically spaced circular asperities on contacting grains. We compare the packing fraction, contact number, mobilization distribution, and vibrational density of states (in the harmonic approximation) using the GA model to those generated using the Cundall-Strack approach. We find that static packings of frictional disks obtained from the GA model are mechanically stable and isostatic when we consider interactions between asperities on contacting particles. The crossover in the structural and mechanical properties of static packings from frictionless to frictional behavior as a function of the static friction coefficient coincides with a change in the type of interparticle contacts and the disappearance of a peak in the density of vibrational modes for the GA model. These results emphasize that mesoscale features of the model for static friction play an important role in determining the properties of granular packings.

Classical topological order in Abelian and non-Abelian generalized height models.

We present Monte Carlo simulations on a new class of lattice models in which the degrees of freedom are elements of an Abelian or non-Abelian finite symmetry group G, placed on directed edges of a two-dimensional lattice. The plaquette group product is constrained to be the group identity. In contrast to discrete gauge models (but similar to past work on height models), only elements of symmetry-related subsets S∈G are allowed on edges. These models have topological sectors labeled by group products along topologically nontrivial loops. Measurement of relative sector probabilities and the distribution of distance between defect pairs are done to characterize the types of order (topological or quasi-long-range order) exhibited by these models. We present particular models in which fully local non-Abelian constraints lead to global topological liquid properties.

Quasi-periodic events in crystal plasticity and the self-organized avalanche oscillator.

When external stresses in a system--physical, social or virtual--are relieved through impulsive events, it is natural to focus on the attributes of these avalanches. However, during the quiescent periods between them, stresses may be relieved through competing processes, such as slowly flowing water between earthquakes or thermally activated dislocation flow between plastic bursts in crystals. Such smooth responses can in turn have marked effects on the avalanche properties. Here we report an experimental investigation of slowly compressed nickel microcrystals, covering three orders of magnitude in nominal strain rate, in which we observe unconventional quasi-periodic avalanche bursts and higher critical exponents as the strain rate is decreased. Our experiments are faithfully reproduced by analytic and computational dislocation avalanche modelling that we have extended to incorporate dislocation relaxation, revealing the emergence of the self-organized avalanche oscillator: a novel critical state exhibiting oscillatory approaches towards a depinning critical point. This theory suggests that whenever avalanches compete with slow relaxation--in settings ranging from crystal microplasticity to earthquakes--dynamical quasi-periodic scale invariance ought to emerge.

Minimal model of plasma membrane heterogeneity requires coupling cortical actin to criticality.

We present a minimal model of plasma membrane heterogeneity that combines criticality with connectivity to cortical cytoskeleton. The development of this model was motivated by recent observations of micron-sized critical fluctuations in plasma membrane vesicles that are detached from their cortical cytoskeleton. We incorporate criticality using a conserved order parameter Ising model coupled to a simple actin cytoskeleton interacting through point-like pinning sites. Using this minimal model, we recapitulate several experimental observations of plasma membrane raft heterogeneity. Small (r ∼ 20 nm) and dynamic fluctuations at physiological temperatures arise from criticality. Including connectivity to the cortical cytoskeleton disrupts large fluctuations, prevents macroscopic phase separation at low temperatures (T ≤ 22°C), and provides a template for long-lived fluctuations at physiological temperature (T = 37°C). Cytoskeleton-stabilized fluctuations produce significant barriers to the diffusion of some membrane components in a manner that is weakly dependent on the number of pinning sites and strongly dependent on criticality. More generally, we demonstrate that critical fluctuations provide a physical mechanism for organizing and spatially segregating membrane components by providing channels for interaction over large distances.

Avalanche spatial structure and multivariable scaling functions: sizes, heights, widths, and views through windows.

We introduce a systematic method for extracting multivariable universal scaling functions and critical exponents from data. We exemplify our insights by analyzing simulations of avalanches in an interface using simulations from a driven quenched Kardar-Parisi-Zhang (qKPZ) equation. We fully characterize the spatial structure of these avalanches--we report universal scaling functions for size, height, and width distributions, and also local front heights. Furthermore, we resolve a problem that arises in many imaging experiments of crackling noise and avalanche dynamics, where the observed distributions are strongly distorted by a limited field of view. Through artificially windowed data, we show these distributions and their multivariable scaling functions may be written in terms of two control parameters: the window size and the characteristic length scale of the dynamics. For the entire system and the windowed distributions we develop accurate parametrizations for the universal scaling functions, including corrections to scaling and systematic error bars, facilitated by a novel software environment SloppyScaling.

Dielectric breakdown and avalanches at nonequilibrium metal-insulator transitions.

Motivated by recent experiments on the finite temperature Mott transition in VO(2) films, we propose a classical coarse-grained dielectric breakdown model where each degree of freedom represents a nanograin which transitions from insulator to metal with increasing temperature and voltage at random thresholds due to quenched disorder. We describe the properties of the resulting nonequilibrium metal-insulator transition and explain the universal characteristics of the resistance jump distribution. We predict that by tuning voltage, another critical point is approached, which separates a phase of boltlike avalanches from percolationlike ones.