Open Collections

Umashankar, Hariharan

University of British Columbia

Atomistic events govern processes at the microstructural length scale that ultimately affect macroscopic material behavior. For instance, solute segregation to internal interfaces such as grain boundaries (GB) can significantly affect the mechanical properties. Similarly, vacancy interactions with GBs and the diffusion along the GBs are also important phenomena that control the microstructure evolution. Experimental investigations of such processes are laborious, time-consuming, and present technical challenges. This study deals with a computational approach to model solute segregation and vacancy interactions at the atomistic scale using two different simulation methods. The first method uses Density Functional Theory (DFT) calculations to quantify the solute-GB interactions in a BCC Ti- 25 at.% Mo alloy. Electrons are considered explicitly in DFT, serving as a state-of-the-art method for segregation phenomena. A major challenge in studying alloys is the multitude of atomic arrangements of the host species. Results from solutes - Y, Zr and Nb occupying all sites in the alloy yield a spectrum of solution energies. Taking the example of the strongest segregating solute, Yttrium, segregation and solution energies at GBs and in the bulk are calculated across three compositions of Mo - 25, 50 and 75 at.% allowing trends to be mapped for the full composition space. The second approach uses a surrogate model, such as the Atomic Cluster Expansion. Initially, a workflow is established using the training data from Embedded Atom Method (EAM) calculations. Here, properties of bulk, fifteen different <100> symmetric tilt GBs in pure Pt and Au, bulk mixing enthalpies of Pt-Au alloys and vacancy formation energies in bulk Pt-Au alloys are studied. The DFT data for some structures that are chosen based on the EAM surrogate modeling are generated to build a regression model with “ab-initio” accuracy. A property of interest, such as vacancy formation energy in ~ 50 at. % Au for bulk and GB sites is studied. Using a committee-of-models approach, new structures are picked iteratively to improve the model.

Text

2025-06-17

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/91364

Materials Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/