Physically informed artificial neural networks for atomistic modeling of materials.

Large-scale atomistic computer simulations of materials heavily rely on interatomic potentials predicting the energy and Newtonian forces on atoms. Traditional interatomic potentials are based on physical intuition but contain few adjustable parameters and are usually not accurate. The emerging machine-learning (ML) potentials achieve highly accurate interpolation within a large DFT database but, being purely mathematical constructions, suffer from poor transferability to unknown structures. We propose a new approach that can drastically improve the transferability of ML potentials by informing them of the physical nature of interatomic bonding. This is achieved by combining a rather general physics-based model (analytical bond-order potential) with a neural-network regression. This approach, called the physically informed neural network (PINN) potential, is demonstrated by developing a general-purpose PINN potential for Al. We suggest that the development of physics-based ML potentials is the most effective way forward in the field of atomistic simulations.

Molecular dynamics simulation of the martensitic phase transformation in NiAl alloys.

Using molecular dynamics simulations with an embedded-atom interatomic potential, we study the effect of chemical composition and uniaxial mechanical stresses on the martensitic phase transformation in Ni-rich NiAl alloys. The martensitic phase has a tetragonal crystal structure and can contain multiple twins arranged in domains and plates. The transformation is reversible and is characterized by a significant temperature hysteresis. The magnitude of the hysteresis depends on the chemical composition and stress. We show that applied compressive and tensile stresses reduce and can even eliminate the hysteresis. Crystalline defects such as free surfaces, dislocations and anti-phase boundaries reduce the martensitic transformation temperature and affect the microstructure of the martensite. Their effect can be explained by heterogeneous nucleation of the new phase in defected regions.