Deep machine learning potentials for multicomponent metallic melts: Development, predictability and compositional transferability

Ryltsev, R. E.; Chtchelkatchev, N. M.

doi:10.1016/j.molliq.2021.118181

Cited by 36 publications

(8 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, the maximum interface velocity obtained by Yang et al is higher than ours by a factor 1.4. An alternative test of the novel potential used in the present work and exhibited reasonable values of parameters (temperature dependent functions) can be made by extended study using a recently developed machine learning potentials as for an Al-Ni-Cu [58].…”

Section: Discussionmentioning

confidence: 99%

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*

Rozas

Ankudinov

Galenko

2022

J. Phys.: Condens. Matter

View full text Add to dashboard Cite

A revised study of the growth and melting of crystals in congruently melting Al$_{50}$Ni$_{50}$ alloy is carried out by molecular dynamics (MD) and phase field methods. An Embedded Atom Method (EAM) potential of Purja Pun and Mishin [Philosophical Magazine 89 (2009) 3245] is used to estimate the material's properties (density, enthalpy, and self-diffusion) of the B2 crystalline and liquid phases of the alloy. Using the same EAM potential, the melting temperature, density, and diffusion coefficient become well comparable with experimental data in contrast with previous works where other potentials were used. In the new revision of MD data, the kinetics of melting and solidification are quantitatively evaluated by the ``crystal-liquid interface velocity - undercooling" relationship exhibiting the well-known bell-shaped kinetic curve. The traveling wave solution of the kinetic phase field model as well as the hodograph equation of the solid-liquid interface quantitatively describe the ``velocity - undercooling" relationship obtained in the MD simulation in the whole range of investigated temperatures for melting and growth of Al$_{50}$Ni$_{50}$ crystals

show abstract

Section: Discussionmentioning

confidence: 99%

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*

Rozas

Ankudinov

Galenko

2022

J. Phys.: Condens. Matter

View full text Add to dashboard Cite

show abstract

“…One of the downsides of EAM potentials is a relatively poor reproduction of the melting point compared to the experimental value ∼3600 K (Wang et al report a melting temperature of 4150 K [33], in the current study we found a higher melting value close to 4350 K, albeit at the elevated pressure). More sophisticated forms for potentials, such as those based on machine-learning, can more accurately reproduce the properties of metallic melts [34][35][36] and liquid-solid equlibrium point [37]. On the other hand, the lattice parameters associated to the potential in [37] differ from experiments and, most importantly, the computational cost of machine-learning potentials is typically about 2-3 orders higher compared to EAM potentials, making them still prohibitive for long large-scale simulations.…”

Section: Methodsmentioning

confidence: 99%

Two-temperature molecular dynamics simulations of crystal growth in a tungsten supercooled melt

Grossi¹,

Pisarev²

2022

J. Phys.: Condens. Matter

View full text Add to dashboard Cite

In this work we use the two-temperature model coupled to molecular dynamics with sinks at the boundaries of the electronic subsystem to study crystal-growth rate in a quasi one-dimensional W crystal into a supercooled melt. The possibility of varying the extension of the electronic grid along with the sinks allows a more realistic description of the electronic thermal transport away from the system, providing a considerable heat dissipation from the crystallization front. Based on this approach, our results regarding crystal-growth rates are not affected even if the size of the system is changed. Moreover, comparisons are established with respect to molecular dynamics and standard two-temperature model simulations. For these comparisons between models, something remarkable is found, and it is that the temperature and the value of the maximum growth rate are the same. In contrast, the inclusion of sinks has a great impact with respect to the standard approaches specially reflected at low temperatures, where a frustration of the liquid-crystal interface dynamics is seen until a state of zero crystal growth is reached, which is not possible to characterize quantitatively since a kind of stochastic behavior is present.

show abstract

“…Han et al 14 DNPs replicate DFT values reliably. 15−26 For instance, DNPs have been applied to elemental metals 16−18 and binary metal alloys, e.g., Cu−Zr, 16 Au−Ag, 17 and Al−Mg, 18 as well as higher metal alloy systems such as Al−Cu−Ni 19 and even for high entropy alloys. 20 Also, DNPs have been successfully applied to several systems with covalent bonding, such as water 22−25 and Ga. 26 In contrast, fewer studies have been performed on ionic materials, such as the investigation of polymorphism in HfO 2 , 27 viscosity and electrical conductivity of MgSiO 4 , 28 metal halide perovskites, 29 and thermal conductivity of β-Ga 2 O 3 .…”

mentioning

confidence: 99%

“…Han et al used multilayered neural networks where each atom is represented by a small subnetwork, which scales with the number of atom neighbors with a prescribed cutoff radius; potentials generated with schemes involving multiple neural network layers are more commonly known as deep neural network potentials (DNPs). Previous studies have shown that DNPs replicate DFT values reliably. − For instance, DNPs have been applied to elemental metals − and binary metal alloys, e.g., Cu–Zr, Au–Ag, and Al–Mg, as well as higher metal alloy systems such as Al–Cu–Ni and even for high entropy alloys . Also, DNPs have been successfully applied to several systems with covalent bonding, such as water − and Ga …”

mentioning

confidence: 99%

Development and Validation of Versatile Deep Atomistic Potentials for Metal Oxides

Wisesa

Andolina

Saidi

2023

J. Phys. Chem. Lett.

View full text Add to dashboard Cite

Machine learning interatomic potentials powered by neural networks have been shown to readily model a gradient of compositions in metallic systems. However, their application to date on ionic systems tends to focus on specific compositions and oxidation states owing to their more heterogeneous chemical nature. Herein we show that a deep neural network potential (DNP) can model various properties of metal oxides with different oxidation states without additional charge information. We created and validated DNPs for Ag x O y , Cu x O y Mg x O y , Pt x O y , and Zn x O y , whereby each system was trained without any limitations on oxidation states. We illustrate how the database can be augmented to enhance the DNP transferability for a new polymorph, surface energies, and thermal expansion. In addition, we show that these potentials can correctly interpolate significant pressure and temperature ranges, exhibit stability over long molecular dynamics simulation time scales, and replicate nonharmonic thermal expansion, consistent with experimental results.

show abstract

Deep machine learning potentials for multicomponent metallic melts: Development, predictability and compositional transferability

Cited by 36 publications

References 59 publications

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*

Two-temperature molecular dynamics simulations of crystal growth in a tungsten supercooled melt

Development and Validation of Versatile Deep Atomistic Potentials for Metal Oxides

Contact Info

Product

Resources

About

Deep machine learning potentials for multicomponent metallic melts: Development, predictability and compositional transferability

Cited by 36 publications

References 59 publications

Kinetics of rapid growth and melting of Al50Ni50 alloying crystals: phase field theory versus atomistic simulations revisited *

Kinetics of rapid growth and melting of Al50Ni50 alloying crystals: phase field theory versus atomistic simulations revisited *

Two-temperature molecular dynamics simulations of crystal growth in a tungsten supercooled melt

Development and Validation of Versatile Deep Atomistic Potentials for Metal Oxides

Contact Info

Product

Resources

About

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*

Kinetics of rapid growth and melting of Al₅₀Ni₅₀ alloying crystals: phase field theory versus atomistic simulations revisited ^*