A5034893508- Magnetic properties of thin films
- Machine Learning in Materials Science
- Theoretical and Computational Physics
- Magnetic Properties and Applications
- Nuclear Materials and Properties
- Fusion materials and technologies
- High-pressure geophysics and materials
- X-ray Diffraction in Crystallography
- Physics of Superconductivity and Magnetism
- Advanced materials and composites
- Electron and X-Ray Spectroscopy Techniques
- Advanced Thermodynamics and Statistical Mechanics
- Thermal properties of materials
- Geomagnetism and Paleomagnetism Studies
- Nuclear reactor physics and engineering
- Force Microscopy Techniques and Applications
- Quantum chaos and dynamical systems
- Multiferroics and related materials
- Chaos control and synchronization
- Energetic Materials and Combustion
- Characterization and Applications of Magnetic Nanoparticles
- Material Dynamics and Properties
- Quantum and electron transport phenomena
- Molten salt chemistry and electrochemical processes
- Geophysical and Geoelectrical Methods
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
2015-2025
CEA Cadarache
2022-2025
Sandia National Laboratories California
2018-2021
Sandia National Laboratories
2020-2021
CEA Le Ripault
2014-2019
Laboratoire de Mathématiques et Physique Théorique
2014-2018
Centre National de la Recherche Scientifique
2015-2017
Université de Tours
2015-2017
Since the classical molecular dynamics simulator LAMMPS was released as an open source code in 2004, it has become a widely-used tool for particle-based modeling of materials at length scales ranging from atomic to mesoscale continuum. Reasons its popularity are that provides wide variety particle interaction models different materials, runs on any platform single CPU core largest supercomputers with accelerators, and gives users control over simulation details, either via input script or by...
We present a scale-bridging approach based on multi-fidelity (MF) machine-learning (ML) framework leveraging Gaussian processes (GP) to fuse atomistic computational model predictions across multiple levels of fidelity. Through the posterior variance MFGP, our naturally enables uncertainty quantification, providing estimates confidence in predictions. used density functional theory as high-fidelity prediction, while ML interatomic potential is low-fidelity prediction. Practical materials’...
We developed a systematic polarizable force field for molten trivalent rare-earth chlorides, from lanthanum to europium, based on first-principle calculations. The proposed model was employed investigate the local structure and physicochemical properties of pure salts their mixtures with sodium chloride. computed densities, heat capacities, surface tensions, viscosities, diffusion coefficients disclosed evolution along lanthanide series, filling gaps poorly studied elements, such as...
Abstract A data-driven framework is presented for building magneto-elastic machine-learning interatomic potentials (ML-IAPs) large-scale spin-lattice dynamics simulations. The ML-IAPs are constructed by coupling a collective atomic spin model with an ML-IAP. Together they represent potential energy surface from which the mechanical forces on atoms and precession of spins computed. Both ML-IAP parametrized data first-principles calculations. We demonstrate efficacy our across...
Tungsten (W) is a material of choice for the divertor due to its high melting temperature, thermal conductivity, and sputtering threshold. However, W has very brittle-to-ductile transition at fusion reactor temperatures (≥1000 K), it may undergo recrystallization grain growth. Dispersion-strengthening with zirconium carbide (ZrC) can improve ductility limit growth, but much effects dispersoids on microstructural evolution thermomechanical properties are still unknown. We present machine...
We present the development of machine-learning interatomic potentials for uranium dioxide ${\mathrm{UO}}_{2}$. Density functional theory calculations with a Hubbard $U$ correction were leveraged to construct training set atomic configurations. This was designed capture elastic and plastic deformations, as well point extended defects, it enriched through an active learning procedure. New configurations added database using multiobjective criterion based on predicted uncertainties energy...
We present a methodology based on the N\'eel model to build classical spin-lattice Hamiltonian for cubic crystals capable of describing magnetic properties induced by spin-orbit coupling like magnetocrystalline anisotropy and anisotropic magnetostriction, as well exchange magnetostriction. Taking advantage analytical solutions model, we derive theoretical expressions parametrization integrals dipole quadrupole terms that link them material. This approach allows us accurate models with...
We present the work-biased path-sampling scheme to calculate chemical potentials in atomic scale simulations. This is based on a series of chained insertion and deletion paths from N + 1 atom systems, sampling being performed themselves rather than final configurations. Equations for parallel path generations as well geometrically biased insertions or deletions are presented. then two applications our approach uranium dioxide crystal. The first test case validation Xe UO2. second explores...
In this work, we leverage atomistic spin-lattice simulations to examine how magnetic interactions impact the propagation of sound waves through a ferromagnetic material. To achieve this, characterize wave velocity in BCC iron, prototypical material, using three different approaches that are based on oscillations kinetic energy, finite-displacement derived forces, and corrections elastic constants, respectively. Successfully applying these methods within framework, find good agreement with...
We present a study on the transport and material properties of aluminum spanning from ambient to warm dense matter conditions using machine-learned interatomic potential (ML-IAP). Prior research has utilized ML-IAPs simulate phenomena in matter, but these potentials have often been calibrated for narrow range temperatures pressures. In contrast, we train single ML-IAP over wide temperatures, density functional theory molecular dynamics (DFT-MD) data. Our approach overcomes computational...
The magnetic behavior of bcc iron nanoclusters, with diameters between 2 and 8 nm, is investigated by means spin dynamics simulations coupled to molecular dynamics, using a distance-dependent exchange interaction. Finite-size effects in the total magnetization as well influence free surface surface/core proportion nanoclusters are analyzed detail for wide temperature range, going beyond cluster bulk Curie temperatures. Comparison made experimental data theoretical models based on mean-field...
A temperature-dependent approach involving Green-Kubo equilibrium atomic and spin dynamics (GKEASD) is reported to assess phonon magnon thermal transport processes accounting for phonon-magnon interactions. Using body-centered cubic (BCC) iron as a case study, GKEASD successfully reproduces its characteristic lattice conductivities. The nonelectronic conductivity, i.e., the sum of conductivities, calculated using BCC Fe, agrees well with experimental measurements. Spectral energy analysis...
Abstract Computational tools to study thermodynamic properties of magnetic materials have, until recently, been limited phenomenological modeling or small domain sizes limiting our mechanistic understanding thermal transport in ferromagnets. Herein, we the interplay phonon and spin contributions conductivity $$\alpha$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>α</mml:mi> </mml:math> -iron utilizing non-equilibrium molecular dynamics simulations. It was observed that...
We present a classical molecular-spin dynamics (MSD) methodology that enables accurate computations of the temperature dependence magnetocrystalline anisotropy as well magnetoelastic properties magnetic materials. The nonmagnetic interactions are accounted for by spectral neighbor analysis potential (SNAP) machine-learned interatomic potential, whereas contributions using combination an extended Heisenberg Hamiltonian and N\'eel pair interaction model, representing both exchange...
We show here using atomistic simulations that ${\mathrm{BiFeO}}_{3}$ films can be driven through a topological transition when increasing uniaxial anisotropy, which possibly achieved by strain. Tuning the anisotropy close to transition, we individual antiferromagnetic skyrmions reach very large protection. These entities then excited electric fields and spin torque controllably at speeds exceeding 10 km/s.
Efficient algorithms for the calculation of minimum energy paths magnetic transitions are implemented within geodesic nudged elastic band (GNEB) approach. While an objective function is not available GNEB and a traditional line search can, therefore, be performed, use limited memory Broyden-Fletcher-Goldfarb-Shanno (LBFGS) conjugate gradient in conjunction with orthogonal spin optimization (OSO) approach shown to greatly outperform previously used velocity projection dissipative...
In this study, we performed a numerical investigation of the thermophysical properties liquid (U, Zr) mixtures, which are particularly relevant in context hypothetical nuclear accidents and formation in-vessel coriums. To do so, atomistic simulations leveraging classical molecular dynamics an interatomic potential developed for solid structures performed. Our methodology is first validated by comparing predictions our model melting temperature structure factors to experimental, phase...
Frequency-dependent permeability tensor for unsaturated polycrystalline ferrites is derived through an effective medium approximation that combines both domain-wall motion and rotation of domains in a single consistent scattering framework. Thus averaged on distribution function the free energy encodes paramagnetic states anhysteretic loops. The initial computed, frequency spectra are given by varying macroscopic remanent field.
Magnetization of clusters is often simulated using atomistic spin dynamics for a fixed lattice. Coupled spin-lattice simulations the magnetization nanoparticles have, to date, neglected change in size atomic magnetic moments near surfaces. We show that introduction variable leads better description experimental data small Fe nanoparticles. To this end, we divide atoms into surface-near shell and core with bulk properties. It demonstrated both magnitude moment exchange interactions need be...
Statistical averaging theorems allow us to derive a set of equations for the averaged magnetization dynamics in presence colored (non-Markovian) noise. The non-Markovian character noise is described by finite auto-correlation time, tau, that can be identified with response time thermal bath system interest. Hitherto, this model was only tested case weakly correlated (when tau equivalent or smaller than integration timestep). In order probe its validity broader range times, model, based on...