Abstract:Machine learning interatomic potentials (ML-IAPs) enable quantum-accurate, classical molecular dynamics simulations of large systems, beyond reach of density functional theory (DFT). Yet, their efficiency and ability to predict systems larger than DFT supercells are not fully explored, posing a question regarding transferability to large-scale simulations with defects (e.g. dislocations, cracks). Here, we apply a three-step validation approach to body-centered-cubic iron. First, accuracy and efficiency are assessed by optimizing ML-IAPs based on four state-of-the-art ML packages. The Pareto front of computational speed versus testing root-mean-square-error (RMSE) is computed. Second, benchmark properties relevant to plasticity and fracture are evaluated. Their average relative error Q with respect to DFT is found to correlate with RMSE. Third, transferability of ML-IAPs to dislocations and cracks is investigated by using per-atom model uncertainty quantification. The core structures and Peierls barriers of screw, M111 and three edge dislocations are compared with DFT. Traction-separation curve and critical stress intensity factor (K_Ic) are also predicted. Cleavage on the pre-existing crack plane is found to be the zero-temperature atomistic fracture mechanism of pure body-centered-cubic iron under mode-I loading, independent of ML package and training database. Quantitative predictions of dislocation glide paths and KIc can be sensitive to database, ML package, cutoff radius, and are limited by DFT accuracy. Our results highlight the importance of validating ML-IAPs by using indicators beyond RMSE. Moreover, significant computational speed-ups can be achieved by using the most efficient ML-IAP package, yet the assessment of the accuracy and transferability should be performed with care.

Comparison of Interatomic Potentials on Crack Propagation Properties in Bcc Iron

Atomic-scale Simulation of Α/γ-Iron Phase Boundary Affecting Crack Propagation Using Molecular Dynamics Method

Atomistic Scale Behaviors of Crack Propagation in Nanocrystalline Bcc Iron

Molecular Dynamics Simulation of Brittle Fracture in Bcc Iron

Atomistic fracture in bcc iron revealed by active learning of Gaussian approximation potential

Machine-learning Interatomic Potential for Radiation Damage Effects in Bcc-Iron

Molecular dynamics simulation of fracture in α-iron

Development of interatomic potentials for Fe-Cr-Al alloy with the particle swarm optimization method

Molecular Dynamics Simulation on Intergranular Crack Propagation Along ∑3 Tilt Grain Boundary in Bcc Iron

The Effect of Impact Load on the Atomistic Scale Fracture Behavior of Nanocrystalline bcc Iron

Different phase transformation behaviors of B2-CuZr crystalline phase and their associated mechanical properties by molecular dynamics using different potentials

Atomistic simulation of crack cleavage and blunting in bcc-Fe

Asymmetrical propagation mechanism of the crack in bcc iron

The Effect of Interatomic Potentials on the Nature of Nanohole Propagation in Single-Crystal Nickel: A Molecular Dynamics Simulation Study

Effects of Interatomic Potential on He Bubble Creation by Cascades in Α-Iron

Atomic behaviors of crack propagation in bcc iron under dynamic loading rate with rectangular fluctuation

Efficiency, Accuracy, and Transferability of Machine Learning Potentials: Application to Dislocations and Cracks in Iron

The Effect of Point Defects Caused by Particle Bombardment on the Deformation Behaviours of Alpha-Iron: A Molecular Dynamics Simulation

Specialising neural network potentials for accurate properties and application to the mechanical response of titanium

The Effect of Stacking Faults or Twin Formation on Crack Propagation in Bcc Iron