Research

The objective of our study is to investigate the deformation behavior of pure-Ti and the influence of interstitial oxygen and substitutional aluminum in it. The study involves computational and experimental analysis of pure-Ti, Ti-O, Ti-Al, and Ti-Al-O systems. For detailed aims and objectives, please refer to the proposal link. An overview of the project is shown graphically 

"Accelerating Generalized Stacking Fault Energy Prediction by Combining Friedel Model, Ab Initio Calculation and Machine Learning".   In this study, we have proposed a combined ab initio and ML-based model that can accelerate the computational prediction of GSFE curves for alloys by a factor of 80. The training dataset is generated using DFT calculations to find the SFE values of 106 metals and alloys using the ANNNI model. The features used for training the ML algorithms come from the physics-based Friedel model. The features are obtained from the electronic DOS, calculated using DFT. Other than accelerating the process of GSFE calculation, the present work also highlights a deep connection between the physics of d-electrons and the deformation behavior of transition metals and alloys. Our study reveals a highly non-linear dependence of shear modulus and stacking fault energies on the electronic features, which requires a combined approach involving state-of-the-art ab initio calculations and machine learning methods for complete understanding. The present model can accelerate alloy designing with targeted mechanical behavior by providing a fast method of screening materials in terms of stacking fault energies.