Using quantified uncertainty to guide efficient phase diagram determination via sequential learning
Theresa Davey1, Brandon J. Bocklund2, Zi-Kui Liu2, Ying Chen1
1. School of Engineering, Tohoku University, Japan
2. Department of Materials Science and Engineering, Pennsylvania State University, USA
MRS Fall 2020, Virtual
Contributed oral presentation in Data Science and Automation to Accelerate Materials Development and Discovery
Phase diagrams are a fundamental tool in materials design, but thorough experimental determination is challenging, expensive, and time consuming. Phase diagrams calculated entirely from first-principles may reduce time and expense, providing information at the prediction stage. Our previous work demonstrated a methodology to obtain a first-principles only CALPHAD-type solid phase diagram reproducing all major features, with little or no prior knowledge of the system [1]. This can guide reduced experiments needed for database validation.
Considering the quantified uncertainty of the first-principles phase diagram [2] using ESPEI [3], a sequential learning approach is taken to systematically add thermodynamic and phase boundary data in regions of highest uncertainty. This simulates how the first-principles only phase diagram can most efficiently guide experimental investigation towards a thorough understanding of a materials system. This convergence towards a target phase diagram will be quantified and compared to the chronological introduction of experimental data.
[1] T. Davey et al., CALPHAD XLVIII, June 2019.
[2] N. Paulson et al., Acta Mater. 174 (2019) 9–15.
[3] B. Bocklund et al., MRS Commun. (2019) 1-10.