arXiv subjects

Pui-Wai Ma

Publications and source records attributed to Pui-Wai Ma.

14 recordsLinked to original sources

arXiv2025

Energy evolution in nanocrystalline iron driven by collision cascades

Nanocrystalline materials are promising candidates for future fusion reactor applications, due to their high density of grain boundaries which may serve as sinks for irradiation induced defects. We use molecular dynamics to simulate collision cascades in nanocrystalline iron and compare these to collision cascades in initially defect free single crystals. We create nanocrystalline samples via Voronoi tessellation of initially randomly placed grain seeds and via severe plastic shearing. An irradiation induced annealing is observed whereby after ~ 2 displacements per atom (dpa), irradiation drives all simulation cells to a single crystalline state. Irradiation-induced defects that distort the lattice generate elastic strain, so we use excess potential energy as a measure of defect content. At low doses, the Voronoi samples feature a few large, low energy grains, whereas the sheared samples show many small, high energy grains due to the high defect and grain boundary content caused by severe deformation. As dose increases beyond 1 dpa however, all nanocrystalline samples converge to a similar behaviour. Excess potential energy mirrors this trend, plateauing above ~ 4 dpa. We hypothesise that the initially pristine cells will also reach a similar plateau after 5 dpa, which is seemingly confirmed by running a single instance of each cell type to 10 dpa. A model is developed to explain the energy evolution.

cond-mat.mtrl-sci↗

arXiv2025

Atomistic simulations of irradiation damage on the engineering timescale: Examining the dose rate effect in tungsten

The change in materials properties subjected to irradiation by highly energetic particles strongly depends on the irradiation dose rate. Atomistic simulations can in principle be used to predict microstructural evolution where experimental data is sparse or unavailable, however, fundamental limitations of the method make it infeasible to replicate the experimental timescale spanning from seconds to hours. Here, we present an atomistic simulation method where the motion of vacancies is accelerated, while the fast degrees of freedom are propagated with standard molecular dynamics. The resulting method is free of adjustable parameters and can predict microstructural evolution under irradiation at elevated temperatures. Simulating the microstructural evolution of tungsten under irradiation at dose rates of $10^{-5}$, $10^{-4}$, and $10^{-3}$ dpa/second, we find that increasing the temperature or reducing the dose rate primarily results in a reduction of the steady-state defect concentration, in qualitative agreement with deuterium retention and post-irradiation resistivity recovery experiments. The formation of a nanoscale void is observed if a system initially containing a large dislocation loop is irradiated. We present a minimally simple rate theory model which reproduces the time-dependent defect concentration and volume swelling behaviour obtained from the simulations.

cond-mat.mtrl-sci↗

arXiv2025

Atomistic and experimental study of microstructural evolution in nanocrystalline iron subjected to irradiation

Nanocrystalline materials have been proposed for use in future fusion reactors due to their high grain boundary density that may act as a sink for irradiation-induced defects. We use molecular dynamics to model collision cascades in nanocrystalline iron and compare the damage evolution to that observed in initially perfect, single crystalline iron. The nanocrystalline material is generated either by Voronoi tessellation or severe plastic shearing. Upon irradiation, the grains in nanocrystalline simulations coarsen, with all ultimately becoming single crystalline above 2 dpa. Above a damage dose of 1 dpa, nanocrystalline cells show a lower dislocation density and lower lattice swelling than their initially pristine counterparts. Experimental X-ray diffraction data is collected on nanocrystalline iron samples subjected to self-ion irradiation. Line profile analysis data agrees with the trends observed in the atomistic simulations, revealing the presence of an irradiation induced annealing process, with a clear reduction in micro-strain with increasing dose. We attempt to determine why some grains in our atomistic simulations grow, while others shrink, by creating a Toy Model that simulates volume exchange between grains based on different hypothesised exchange mechanisms. This suggests that irradiation-induced grain growth is consistent with random growth.

cond-mat.mtrl-sci↗

arXiv2025

Review of point defect structures in hexagonal close packed metals and across the Periodic Table

We present a comprehensive ab initio dataset of formation energies and elastic properties of intrinsic point defects across all the transition and rare earth hexagonal close packed (hcp) metals, as well as metalloid elements with hcp crystal structure. Point defect properties appear weakly correlated with the c/a ratio of the hcp lattice. Instead, it is the position of an element in the Periodic Table that primarily defines the relaxation volume tensor, elastic dipole tensor and formation energy of a point defect. This suggests that the local variations in the electronic structure and interatomic bonding at the core of a defect dominate its properties, as opposed to long-range elastic deformations. Across all the metals, we find that the relaxation volumes of vacancies and self-interstitial defects are correlated with atomic volumes, with values of -0.35 and 1.46 atomic volumes for a vacancy and a self-interstitial defect, respectively, providing a universally good approximation independent of the crystal structure. This study complements and completes the existing point defect databases spanning body-centred cubic and face-centred cubic metals.

cond-mat.mtrl-sci↗

arXiv2024

Activation Functions for "A Feedforward Unitary Equivariant Neural Network"

In our previous work [Ma and Chan (2023)], we presented a feedforward unitary equivariant neural network. We proposed three distinct activation functions tailored for this network: a softsign function with a small residue, an identity function, and a Leaky ReLU function. While these functions demonstrated the desired equivariance properties, they limited the neural network's architecture. This short paper generalises these activation functions to a single functional form. This functional form represents a broad class of functions, maintains unitary equivariance, and offers greater flexibility for the design of equivariant neural networks.

cs.LG↗

arXiv2024

Negative Temperature in Spin Dynamics Simulations

A simple and computationally efficient algorithm enables implementing negative temperature values in a spin dynamics simulation. The algorithm uses a Langevin spin dynamics thermostat with a negative damping parameter, enabling the thermalization of an arbitrary interacting spin system to the Gibbs energy distribution with a given negative temperature value. Canonical spin dynamics simulations at a negative temperature are as robust as conventional positive spin temperature simulations, providing a tool for quantitative dynamic studies of the physics of highly excited magnetic states. Two simulation case studies describing spin systems with antiferromagnetic and ferromagnetic ground states are explored. The phase transitions occurring in the negative temperature range do not necessarily exhibit similarities with their positive temperature counterparts. The transition temperatures and the character of spin alignment vary depending on the spatial range and strength of spin-spin interactions.

cond-mat.mtrl-sci↗

arXiv2023

Simulations of Nanocrystalline Iron Formation under High Shear Strain

High-shear methods have long been used in experiments to refine grain structures in metals, yet the underlying mechanisms remain elusive. We demonstrate a refinement process using molecular dynamic simulations of iron, wherein nanocrystalline structures are generated from initially perfect lattices under high-shear strain. The simulation cells undergo a highly disordered state, followed by an atomic reordering and grain coarsening, resulting in nanograins. We explore the dependence on parameters such as temperature, heat dissipation rate, shear strain rate, and carbon impurity concentration. Higher temperatures lead to the formation of larger and longer grains. The faster heat dissipation sample initially yields more small grains, but their number subsequently reduces, and is lower than the slower heat dissipation sample at approximately {\gamma} = 1.5. Slower strain rates do not promote nanograin formation. The presence of carbon impurities appears to have little effect on grain formation. This detailed analysis affords insight into the mechanisms that control the formation of nanograins under high-shear conditions.

cond-mat.mtrl-sci↗

arXiv2022

A Feedforward Unitary Equivariant Neural Network

We devise a new type of feedforward neural network. It is equivariant with respect to the unitary group $U(n)$. The input and output can be vectors in $\mathbb{C}^n$ with arbitrary dimension $n$. No convolution layer is required in our implementation. We avoid errors due to truncated higher order terms in Fourier-like transformation. The implementation of each layer can be done efficiently using simple calculations. As a proof of concept, we have given empirical results on the prediction of the dynamics of atomic motion to demonstrate the practicality of our approach.

cs.LG↗

arXiv2022

A Machine-Learned Spin-Lattice Potential for Dynamic Simulations of Defective Magnetic Iron

A machine-learned spin-lattice interatomic potential (MSLP) for magnetic iron is developed and applied to mesoscopic scale defects. It is achieved by augmenting a spin-lattice Hamiltonian with a neural network term trained to descriptors representing a mix of local atomic configuration and magnetic environments. It reproduces the cohesive energy of BCC and FCC phases with various magnetic states. It predicts the formation energy and complex magnetic structure of point defects in quantitative agreement with density functional theory (DFT) including the reversal and quenching of magnetic moments near the core of defects. The Curie temperature is calculated through spin-lattice dynamics showing good computational stability at high temperature. The potential is applied to study magnetic fluctuations near sizable dislocation loops. The MSLP transcends current treatments using DFT and molecular dynamics, and surpasses other spin-lattice potentials that only treat near-perfect crystal cases.

cond-mat.mtrl-sci↗

arXiv2020

Effect of Non-Heisenberg Magnetic Interactions on Defects in Ferromagnetic Iron

Fundamental flaws in the Heisenberg Hamiltonian are highlighted in the context of its application to BCC Fe, including the particular issues arising when modelling lattice defects. Exchange integrals are evaluated using the magnetic force theorem. The bilinear exchange coupling constants are calculated for all the interacting pairs of atomic magnetic moments in large simulation cells containing defects, enabling a direct mapping of the magnetic energy onto the Heisenberg Hamiltonian and revealing its limitations. We provide a simple procedure for extracting the Landau parameters from DFT calculations, to construct a Heisenberg-Landau Hamiltonian. We quantitatively show how the Landau terms correct the exchange-energy hypersurface, which is essential for the accurate evaluation of energies and migration barriers of defects.

cond-mat.mtrl-sci↗

arXiv2017

Dynamic Simulation of Structural Phase Transitions in Magnetic Iron

The occurrence of bcc-fcc ($α$-$γ$) and fcc-bcc ($γ$-$δ$) phase transitions in magnetic iron stems from the interplay between magnetic excitations and lattice vibrations. However, this fact has never been proven by a direct dynamic simulation, treating non-collinear magnetic fluctuations and dynamics of atoms, and their coupling at a finite temperature. Starting from a large set of data generated by ab initio simulations, we derive non-collinear magnetic many-body potentials for bcc and fcc iron describing fluctuations in the vicinity of near perfect lattice positions. We then use spin-lattice dynamics simulations to evaluate the difference between free energies of bcc and fcc phases, assessing their relative stability within a unified dynamic picture. We find two intersections between the bcc and fcc free energy curves, which correspond to $α$-$γ$ bcc-fcc and $γ$-$δ$ fcc-bcc phase transitions. The maximum fcc-bcc free energy difference over the temperature interval between the two phase transition points is 2 meV, in agreement with other experimental and theoretical estimates.

cond-mat.mtrl-sci↗

arXiv2015

Hubbard-like Hamiltonians for interacting electrons in s, p and d orbitals

Hubbard-like Hamiltonians are widely used to describe on-site Coulomb interactions in magnetic and strongly-correlated solids, but there is much confusion in the literature about the form these Hamiltonians should take for shells of p and d orbitals. This paper derives the most general s, p and d orbital Hubbard-like Hamiltonians consistent with the relevant symmetries, and presents them in ways convenient for practical calculations. We use the full configuration interaction method to study p and d orbital dimers and compare results obtained using the correct Hamiltonian and the collinear and vector Stoner Hamiltonians. The Stoner Hamiltonians can fail to describe properly the nature of the ground state, the time evolution of excited states, and the electronic heat capacity.

cond-mat.str-el↗

arXiv2013

Constrained non-collinear magnetism in disordered Fe and Fe-Cr alloys

The development of quantitative models for radiation damage effects in iron, iron alloys and steels, particularly for the high temperature properties of the alloys, requires understanding of magnetic interactions, which control the phase stability of ferritic-martensitic, ferritic, and austenitic steels. In this work, disordered magnetic configurations of pure iron and Fe-Cr alloys are investigated using Density Functional Theory (DFT) formalism, in the form of constrained non-collinear magnetic calculations, with the objective of creating a database of atomic magnetic moments and forces acting between the atoms. From a given disordered atomic configuration of either pure Fe or Fe-Cr alloy, a penalty contribution to the usual spin-polarized DFT total energy has been calculated by constraining the magnitude and direction of magnetic moments. An extensive database of non-collinear magnetic moment and force components for various atomic configurations has been generated and used for interpolating the spatially-dependent magnetic interaction parameters, for applications in large-scale spin-lattice dynamics and magnetic Monte-Carlo simulations.

cond-mat.mtrl-sci↗

arXiv2010

Interatomic exchange coupling of BCC iron

We performed first-principle calculations on the exchange interaction (EI) between atoms in BCC-Fe strained volumetrically. Our results show that the volume-dependence of the EI deviates considerably from the Bethe-Slater curve. This behavior is discussed in terms of the on-site and/or inter-site direct exchange interactions between electrons.

cond-mat.mtrl-sci↗