第一原理ビッグデータに基づいたBCC鉄原子間ポテンシャルの開発

抄録

We construct the atomic artificial neural network (ANN) potential for investigation of dislocation mechanics and dynamics in body-centered cubic (BCC) iron based on Density functional theory (DFT) reference data. The bulk properties and defect formation energies predicted by the constructed ANN potential are in good agreement with reference DFT calculations. The 1/2[111] screw dislocation core structure and its energetics predicted by the ANN potential are in excellent agreement with DFT calculations. These results clearly show the excellent reproducibility and generalization ability of the constructed ANN potential.