Radar‐metric analysis shows that PBE + D3‐based models provide the most balanced overall performance for density, liquid structure, and diffusion, whereas LDA‐based models give the best predictions for Ga and EGaIn.
Abstract
Low‐melting liquid metals, especially Ga–In alloys, are essential for flexible electronics, soft robotics, and adaptive thermal interfaces. Predictive atomistic modeling of their melting behavior is challenging because experiments generally provide limited microscopic insight, whereas first‐principles simulations are computationally expensive. In this work, Neuroevolution Potential (NEP) potentials for the Ga–In system are developed using density functional theory (DFT) datasets of 200–700 atom supercells generated with the local density approximation (LDA), Perdew‐Burke‐Ernzerhof (PBE), and PBE + D3 functionals. Eighteen independent NEP models with different energy/force/virial weights were trained and benchmarked against density, radial distribution function (RDF), self‐diffusion coefficient (SDC), and melting temperature (
T
m
). Radar‐metric analysis shows that PBE + D3‐based models provide the most balanced overall performance for density, liquid structure, and diffusion, whereas LDA‐based models give the best
T
m
predictions for Ga and EGaIn. We further design a local Lindemann parameter for solid–liquid identification in disordered alloys, extending the two‐phase method to compositionally complex liquid‐metal systems. Together, these results provide a benchmarked NEP model set and a scalable framework for phase‐transition and transport simulations of Ga–In liquid metals.
Molecular dynamics (MD) is essential for investigating atomic‐scale processes in materials and molecular systems, but the cost of high‐accuracy machine learning force field simulations still limits accessible system sizes and timescales. Here, we propose a practical model‐switching strategy for Deep Potential (DP)‐based MD simulations that alternates between independently trained DP models with different cutoff radii: a standard 6 Å model for higher accuracy and a reduced‐cutoff 4 Å model for faster inference. The method was implemented in LAMMPS/DeePMD and evaluated using solid‐phase anatase TiO
2
and liquid‐phase polyethylene glycol (PEG). For anatase TiO
2
, the 1:3 4–6 Å switching scheme preserved radial distribution function (RDF) correlations of 0.996 or higher relative to the 6 Å baseline while achieving a 1.24‐fold speedup. For PEG, the switching scheme maintained RDF correlations of 0.996 or higher with a 1.18‐fold speedup. Additional optimization using network‐size reduction and mixed‐precision inference achieved a 2.53‐fold speedup with RDF correlations of 0.975–0.988. Constant particle‐number, pressure, and temperature (NPT) simulations remained stable, whereas constant particle‐number, volume, and energy (NVE) simulations revealed system‐dependent energy‐drift behavior, particularly for aggressively optimized models. These results demonstrate that DP model switching provides a simple and practical route for accelerating structural MD simulations while highlighting the need for validation when strict energy conservation is required.
Ryuya Kanda, Megumu Yamazaki, Yuta Yoshimoto et al.· Advanced Intelligent Discove...· 0 citations
The prediction of stable alloys forming solid-state solutions across large portions of the composition space is a serious theoretical challenge, since one has to evaluate the Gibbs free energy, including both configurational and vibrational contributions. This requires an energy theory capable of extremely high throughput. By taking the Ni-Pd system as prototype, we construct an efficient Jacobi-Legendre machine-learning potential based on density-functional-theory data, which provides accurate energies and forces across the entire composition space. Based on a cluster expansion up to three-body terms and only 873 trainable parameters, this allows us to compute the partition function by directly integrating all accessible microstates, differing for composition, atomic configuration and thermal agitation. We confirm that Ni and Pd are fully miscible, forming an $fcc$ solid-state solution. This is only metastable at room temperature, while becomes thermodynamically stable at around 600~K, with the stability achieved first at the Pd-rich end of the composition range. Interestingly, entropy and heat capacity analysis reveal a competition between the solid-state solution and two intermetallic phases with long-period L1$_0$ structure for NiPd and NiPd$_3$. All in all, our approach offers a powerful and high-throughput workflow for the study of disordered alloys, an approach that can be extended to multi-component systems such as high-entropy alloys.
Rutchapon Hunkao, U. Patil, S. Sanvito· 0 citations
For decades, atomistic simulation of chemically complex Ni‐based superalloys has remained beyond practical reach. Here, we apply the GRACE foundational machine‐learning interatomic potential to predict chemical ordering and stacking‐fault energetics in the and phases of CMSX‐4, a commercial multicomponent Ni‐based superalloy. After benchmarking against structural and thermodynamic reference data, we use hybrid Monte‐Carlo/molecular dynamics sampling to study the impact of local chemical order on planar‐fault energies. GRACE reproduces elemental equilibrium lattice parameters within of DFT references, while underestimating melting temperatures of ordered Ni–Al phases by up to . The simulations reveal local chemical ordering in the phase and the expected sublattice occupancies in the phase. In the phase, the short‐range order raises the shear barriers by approximately while leaving the intrinsic stacking fault energy of unchanged. In the phase, alloying raises the complex and superlattice intrinsic stacking fault energies by approximately relative to stoichiometric Al. These results show that pretrained foundational potentials enable atomistic simulations of chemically complex multicomponent superalloys at scales inaccessible to direct first‐principles calculations.
Aditya Vishwakarma, Sarath Menon, Fritz Körmann et al.· Advanced Engineering Materia...· 0 citations
Accurate prediction of a material's melting temperature is critical for materials design and high-temperature applications. In this work, we investigate melting behavior across a deliberately selected set of elemental metals spanning systems where cohesive-energy trends suggest that PBE and PBEsol are expected to perform differently, as well as cases where their performance is ambiguous. Melting temperatures are computed using the two-phase coexistence (TPC) approach in conjunction with a machine-learned interatomic potential based on the moment tensor potential (MTP) framework, enabling large-scale simulations that minimize finite-size effects and ensure sufficient equilibration. The TPC-MTP results reveal a clear functional dependence in the predicted melting temperatures. PBE provides good agreement for several lighter elements, whereas PBEsol gives the best overall agreement across the full dataset. However, the element-resolved trends are not governed by cohesive energy alone, indicating that liquid-phase energetics, anharmonicity, and finite-temperature phase stability also contribute to the observed functional dependence. For intermediate and structurally complex systems, both functionals exhibit less systematic performance. Overall, this study provides a systematic assessment of functional-dependent melting temperature predictions, highlighting both the strengths and limitations of the combined TPC-MTP methodology and underscoring the need for carefully selected exchange-correlation treatments in high-accuracy melting-point simulations.
Pandu Wisesa, Christopher M. Andolina, W. Saidi· Journal of Chemical Physics· 0 citations
Directed energy deposition (DED) of low-alloy steels involves strongly coupled effects among alloy composition, solidification behavior, and post-deposition heat treatment, making mechanical property prediction difficult when target-domain data are limited. This study develops a transfer-learning and continuous optimization framework for predicting heat treatment-dependent yield strength (YS), ultimate tensile strength (UTS), hardness (HV), and as-solidified phase fractions of martensite, ferrite, and austenite in DED-processed low-alloy steels. A CALPHAD-based dataset was generated for 125 low-alloy steel compositions. A multilayer perceptron (MLP) surrogate was first trained as a baseline model, then fine-tuned through transfer learning and progressively updated as staged continuous optimization; the composition pool increased from 72 to 125 compositions using Random, Greedy, and Bayesian upper-confidence-bound acquisition strategies. The heat treatment prediction accuracy improved from an average R2 of 0.757 for the baseline model to 0.929 after transfer learning and to approximately 0.997 after continuous optimization, with a nearly 78% reduction in RMSE relative to transfer learning. For the solidification outputs, the average R2 increased from 0.770 after transfer learning to approximately 0.859 after optimization. Bayesian-UCB provided the most stable and data-efficient improvement by balancing predicted performance with model uncertainty. The optimized prediction system showed low case-study errors for both solidification and heat treatment properties, demonstrating its potential as a rapid screening tool for alloy composition and tempering-condition selection in DED low-alloy steel development.
Atiqur Rahman, Sung-Heng Wu, Ranjit Joy et al.· Metals· 0 citations
Predicting the properties of multicomponent molten salts using density functional theory (DFT) remains challenging because the spatial and temporal scales required to evaluate transport properties and phase behavior are computationally prohibitive. In this work, we develop a moment tensor potential trained using a a DFT dataset of NaCl, KCl, NaCl-KCl mixtures, and the NaK alloy, enabling large-scale molecular dynamics simulations across wide ranges of temperatures and compositions. We systematically evaluate the effect of D3 dispersion corrections and apply the resulting potential to predict liquid densities, diffusion coefficients, radial distribution functions, heat capacities, thermal conductivities, and the NaCl-KCl phase diagram. The model successfully reproduces many temperature- and composition-dependent trends. However, systematic deviations in several absolute properties persist, highlighting the importance of experimental validation and calibration. These findings support a hybrid modeling framework in which first-principles-informed machine-learning potentials provide transferable predictive capability and mechanistic insight, while experimental data incorporated during model development or subsequent engineering assessments is necessary to improve quantitative accuracy.
K. Zongo, Hao Sun, Zijian Meng et al.· 0 citations