A Machine Learning Approach Towards Quantitative Structure-Property Relationships for Metallic Interfaces

Abstract

In the proposed research program, the PI will develop machine-learning tools and will construct reliable reduced-order models for GB energies and temperature-dependent mobilities as a function of the macroscopic five-parameter space. The GB atomistic structure will be quantified using two-point correlation functions and robust statistical models will be built by utilizing dimensionality reduction techniques. Experimentally obtained GB property measurements will be investigated through a regression based analysis, made possible by the development of interpolating basis functions in the five-parameter space of GBs. The basis functions will not only facilitate a comparison of the simulated data with experimental measurements, but will enable the prediction of atomistic structures as a function of GB crystallography. The machine-learning techniques developed in this research program will transform the quantification of GB crystallography-structure-property relationships as they can be easily extended to the analysis of a wide range of material systems, such as multi-component alloys and functional ceramics, and to the quantification of complex properties, such as diffusivity, conductivity, corrosion resistance, and defect-interface interactions.

Document Details

Document Type
DoD Grant Award
Publication Date
May 02, 2017
Source ID
FA95501710145

Entities

People

  • Srikanth Patala

Organizations

  • Air Force Office of Scientific Research
  • North Carolina State University
  • United States Air Force

Tags

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Materials Science and Engineering.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms
  • Space