arXiv subjects

Petr Vlčák

Publications and source records attributed to Petr Vlčák.

7 recordsLinked to original sources

arXiv2026

Benchmarking Universal Machine-Learning Interatomic Potentials for Temperature-Dependent Elasticity of Binary and High-Entropy Refractory Carbides

Accurate prediction of temperature-dependent elasticity is important for assessing refractory carbides under high-temperature conditions, but the computational cost of ab initio molecular dynamics (AIMD) limits systematic investigations across compositions and temperatures. Universal machine-learning interatomic potentials (uMLIPs) offer an efficient alternative, yet their accuracy for this task remains insufficiently established. Here, we benchmark nine uMLIPs against consistent AIMD reference data for five binary and two high-entropy (HE) carbides between 300 and 1200 K. Elastic constants and the corresponding bulk, shear, and Young's moduli are obtained using stress-strain molecular dynamics, explicitly sampling thermal atomic motion and anharmonic effects beyond thermal expansion alone. We assess absolute elastic properties and normalized thermal softening separately. MACE-MH-1 achieves the lowest overall mean absolute percentage error (5.7%). MACE-MH-1 and DPA4-Mini reproduce thermal softening most accurately, with mean deviations of 2.5 and 2.3 percentage points, respectively. All models generally underestimate stiffness, with larger equilibrium volumes relative to the AIMD reference likely contributing to this trend for most models. C12 exhibits the largest model-dependent errors. The HE carbides are described with accuracy comparable to that of the binary carbides, indicating no apparent accuracy penalty from chemical complexity. Accuracy instead varies with transition-metal composition, with group-V carbides, particularly TaC, presenting the greatest challenge. These results identify promising pretrained models for finite-temperature elasticity in carbides and show why accurate absolute stiffness and thermal softening must be assessed independently.

cond-mat.mtrl-sci↗

arXiv2026

Rust-accelerated powder X-ray diffraction simulation for high-throughput and machine-learning-driven materials science

High-throughput powder X-ray diffraction (XRD) simulations are a key prerequisite for generating large datasets used in the development of machine-learning models for XRD-based materials analysis. However, the widely used pymatgen powder XRD calculator, implemented entirely in Python, can be computationally inefficient for large-scale workloads, limiting throughput. We present XRD-Rust, a Rust-accelerated implementation of the pymatgen powder XRD calculator that maintains compatibility with existing Python-based workflows. The method retains pymatgen for crystal structure handling while reimplementing the computationally intensive parts of the XRD calculation in Rust, with optional further acceleration via SIMD vectorization and multi-threaded execution across reflections using the Rayon library. Performance benchmarking on two large crystallographic datasets, the Materials Cloud Three-Dimensional Structure Database (MC3D, 33 142 structures) and the Crystallography Open Database (COD, 515 181 structures), demonstrates substantial speedups. For MC3D, XRD-Rust achieves a median serial SIMD speedup of 10.8x (median absolute deviation, MAD, 1.8x), increasing to 15.1x (MAD 3.8x) with 8-thread parallel execution, with a maximum runtime reduction from 40.5 s to 0.9 s. For COD, the median serial SIMD acceleration reaches 10.7x (MAD 4.2x), while 8-thread parallel execution yields 19.5x (MAD 10.0x), with a maximum reduction from 1437 min to 1 min. SIMD vectorization alone provides additional performance gains ranging from a few percent to several tens of percent, depending on the workload and execution mode. Overall, these benchmarks demonstrate that XRD-Rust significantly accelerates powder XRD simulations compared to the original pymatgen implementation, enabling efficient high-throughput dataset generation and improving performance in interactive diffraction analysis applications.

cond-mat.mtrl-sci↗

arXiv2026

Lattice Parameters and Bulk Modulus of SrTi$_{1-\mathit{x}}$Mn$_{\mathit{x}}$O$_{3}$ Perovskites: A Comparison of Exchange-Correlation Functionals with Experimental Validation

We assessed four exchange-correlation functionals (LDA CA-PZ, GGA parametrized by PBE, PBEsol, and WC) in predicting the lattice parameters of SrTi$_{1-\mathit{x}}$Mn$_{\mathit{x}}$O$_{3}$ perovskites, assuming cubic structures. Predictions were verified using X-ray diffraction (XRD) for Mn content of $\mathit{x}$ = 0.0, 0.1, 0.2, 0.3, 0.5, 1.0, confirming cubic symmetry and a linear decrease in lattice parameters with increasing Mn. PBEsol and WC demonstrated the highest precision (deviations < 0.20%). Additionally, bulk moduli were calculated using the same functionals and verified with the experimental bulk modulus of SrTiO$_{3}$ (183 $\pm$ 2 GPa, Pulse-Echo method). The predicted bulk moduli exhibited a slow, linear increase with increasing Mn. The best correspondence with the experimental bulk modulus was achieved by PBEsol and WC (deviations < 0.7%). These findings highlight the reliability of PBEsol and WC functionals for accurately modeling structural properties of SrTi$_{1-\mathit{x}}$Mn$_{\mathit{x}}$O$_{3}$ perovskites, having better precision than commonly employed LDA and PBE functionals.

cond-mat.mtrl-sci↗

arXiv2026

Interactive Analysis of Static, Dynamic, and Crystalline SDTrimSP Simulations: Application to Nitrogen Ion Implantation into Vanadium

SDTrimSP is a widely used Monte Carlo simulation code based on the Binary Collision Approximation (BCA) for modeling ion implantation and ion-solid interaction processes. While an established graphical user interface (GUI) exists for simulation setup and execution, efficient post-processing, comparison of multiple simulations, and preparation of specific input file parameters remain limited. In this work, we present a web-based interface (sdtrimsp.streamlit.app) that complements existing SDTrimSP tools by focusing on interactive visualization and analysis of depth distribution profiles. The platform enables direct upload and comparison of static and fluence-dependent dynamic profiles, supports unit conversion, and provides an integrated calculator for determining the adjustable atomic density parameter of implanted ions required in dynamic simulations. In addition, the interface offers automated conversion of standard crystallographic file formats into the SDTrimSP-specific crystal structure input format for simulations into crystalline targets. The capabilities of the interface are demonstrated for nitrogen ion implantation into vanadium, including amorphous static and dynamic simulations and static crystalline simulations for different surface orientations. The results illustrate fluence-dependent saturation effects as well as orientation-dependent ion channeling behavior. Overall, the presented web-based tool provides a convenient and flexible extension to existing SDTrimSP workflows.

cond-mat.mtrl-sci↗

arXiv2026

k-Means Clustering in Fingerprint-Based Configuration Selection for Fitting Interatomic Potentials

In this study, we present a method for selecting an arbitrary number of distinct configurations from a larger data set by applying k-means clustering to atomistic configuration fingerprints based on the CrystalNN model and radial distribution function (RDF). This approach improves the accuracy of fitting classical molecular dynamics interatomic potentials to density functional theory (DFT) data for both energies and forces while requiring fewer configurations than random selection. We demonstrate this improvement by fitting an embedded-atom method (EAM) potential for titanium, using various configurational sizes from an initial set of 1800 configurations. The k-means clustering consistently achieves better precision and lower standard deviations for a smaller number of configurations than random selection. The results also suggest that only about 30 configurations are sufficient to obtain an EAM model that describes well the full set of 1800 configurations in terms of energies and forces. Additionally, t-distributed stochastic neighbor embedding (t-SNE) method was used to reduce the configuration fingerprints into 2D space, and it revealed an overlap between two configuration subsets with and without Ti vacancy, indicating similar atomic environments. This similarity is captured by k-means clustering but not by random selection. Furthermore, when the overlapping configurations with vacancies were excluded from the k-means algorithm and used only as a test set, their energy and force predictions showed similar precision to those when they were included. This indicates that the overlapping configurations in the 2D t-SNE space indeed imply potential information redundancy among the atomistic configurations.

cond-mat.mtrl-sci↗

arXiv2026

Revealing interstitial energetics in Ti-23Nb-0.7Ta-2Zr gum metal base alloy via universal machine learning interatomic potentials

Understanding the behavior of light interstitial elements in multicomponent alloys remains challenging due to the complexity of local chemical environments and the high computational cost of first-principles calculations. Here we demonstrate that three universal machine-learning interatomic potentials (uMLIPs) - MACE-MATPES-PBE-0, Orb-v3, and SevenNet-0 can efficiently map the energetics of C, N, O, and H interstitials in a Ti-23Nb-0.7Ta-2Zr (at.%) gum metal base alloy while being several orders of magnitude faster than density functional theory (DFT). All uMLIPs predict broad energy distributions (~1-3 eV) across the four interstitial elements, reflecting their strong sensitivity to local lattice chemistry. Despite alloy disorder, MACE-MATPES-PBE-0 and Orb-v3 reproduce the expected site preferences of the bcc structure: C, N, and O relax into octahedral sites, whereas H stabilizes in tetrahedral positions. In contrast, SevenNet-0 predicts H to be most stable in octahedral coordination, indicating a limitation of this model. Correlation analysis reveals two dominant chemical trends: Ti-rich environments strongly stabilize interstitials, whereas close proximity to Nb is destabilizing. Zr and Ta show no statistically significant influence, likely due to their low concentrations. Benchmarking representative O interstitial configurations against DFT confirms that the uMLIPs reasonably reproduce the energetic ordering of chemically distinct environments. DFT validation confirms that tetrahedral configurations are energetically more favorable than octahedral sites for H interstitials, further illustrating the SevenNet-0 limitation. Overall, we demonstrated that uMLIPs enable computationally efficient, statistically broad characterization of defect energetics in Ti-23Nb-0.7Ta-2Zr gum metal base alloy and provide insight into how local chemical environments govern interstitial stability.

cond-mat.mtrl-sci↗

arXiv2026

SimplySQS: An Automated and Reproducible Workflow for Special Quasirandom Structure Generation with ATAT

The special quasirandom structure (SQS) method is widely used for modeling disordered materials under periodic boundary conditions, with the ATAT mcsqs module being one of the most established implementations. However, SQS generation with mcsqs typically relies on manual preparation of input files, ad hoc execution scripts, and post-processing steps, which introduces user-dependent errors and limits reproducibility. Here, we present SimplySQS (https://simplysqs.com), an automated and reproducible workflow for SQS generation that is delivered through an online, interactive interface. SimplySQS guides users through structure import, compositional and supercell definition, and cluster parameter selection, while automatically generating all required ATAT input files and a single all-in-one execution script that encapsulates the complete search process. By standardizing input preparation, execution, and output analysis, the framework minimizes errors associated with manual file handling and enables consistent reproducibility of SQS searches. The workflow is demonstrated on the Pb1-xSrxTiO3 (PSTO, including PbTiO3 (PTO) and SrTiO3 (STO)) perovskite system. SQSs spanning the entire concentration range were generated using a single automated bash script produced by SimplySQS, after which all resulting structures were subjected to geometry optimization using a universal machine-learning interatomic potential (MACE MATPES-r2SCAN-0). This approach reliably reproduced the experimentally observed cubic-to-tetragonal transition near x = 0.5, with lattice parameters deviating by less than 1 % in the cubic region (x > 0.5) and less than 4 % in the tetragonal region (x < 0.5). Overall, SimplySQS transforms SQS generation with ATAT into an intuitive, reproducible, and systematic framework for modeling disordered materials.

cond-mat.mtrl-sci↗