Automated voxelization of 3D atom probe data through kernel density estimation

Srikant Srinivasan; Kaustubh Kaluskar; Santoshrupa Dumpala; Scott Broderick; Krishna Rajan

doi:10.1016/j.ultramic.2015.03.012

Automated voxelization of 3D atom probe data through kernel density estimation

Ultramicroscopy. 2015 Dec:159 Pt 2:381-6. doi: 10.1016/j.ultramic.2015.03.012. Epub 2015 Mar 18.

Authors

Srikant Srinivasan¹, Kaustubh Kaluskar¹, Santoshrupa Dumpala¹, Scott Broderick¹, Krishna Rajan²

Affiliations

¹ Institute of Combinatorial Discovery, Department of Materials Science and Engineering, Iowa State University, Ames, IA 50011-2300, USA.
² Institute of Combinatorial Discovery, Department of Materials Science and Engineering, Iowa State University, Ames, IA 50011-2300, USA. Electronic address: krajan@iastate.edu.

PMID: 25825028
DOI: 10.1016/j.ultramic.2015.03.012

Abstract

Identifying nanoscale chemical features from atom probe tomography (APT) data routinely involves adjustment of voxel size as an input parameter, through visual supervision, making the final outcome user dependent, reliant on heuristic knowledge and potentially prone to error. This work utilizes Kernel density estimators to select an optimal voxel size in an unsupervised manner to perform feature selection, in particular targeting resolution of interfacial features and chemistries. The capability of this approach is demonstrated through analysis of the γ / γ' interface in a Ni-Al-Cr superalloy.

Keywords: APT; Informatics; Interfaces; Kernel density estimation; Ni Alloy.

Publication types

Research Support, Non-U.S. Gov't
Research Support, U.S. Gov't, Non-P.H.S.