Publication Date

2015

Document Type

Dissertation

Committee Members

Peter Collins (Committee Member), Ramana Grandhi (Committee Member), Nathan Klingbeil (Advisor), Raghu Srinivasan (Committee Member), Jaimie Tiley (Committee Member), Michael Uchic (Committee Member)

Degree Name

Doctor of Philosophy (PhD)

Abstract

The sampling of three dimensional (3D) mesoscale microstructural data is typically prescribed using simple rules, likely resulting in data under- or oversampling depending on the measurement(s) of interest. The first part of this work investigates one approach for determining a minimally sufficient sampling scheme for 3D microstructural data, using computer-generated phantoms of polycrystalline grain microstructures. Sources of error that are observed experimentally are modeled using phantoms, in order to determine the effect that errors have on the microstructural statistic(s)-of-interest. Minimally-sufficient sampling schemes are then established based on a required accuracy in the microstructural statistic(s). The characterization error modeling framework is subsequently demonstrated on experimentally-derived statistics from high resolution 3D serial sectioning data, in order to inform future experiments on the same material. The second part of this work lends the aforementioned approach to the additive manufacturing (AM) of Ti-6Al-4V. Statistical analysis and virtual modeling tools developed herein are used to analyze alpha and beta phase microstructures in two thin-walled Ti-6Al-4V samples. Ultimately, this research aims to provide a virtual modeling framework for analyzing uncertainty in microstructural characterization, and to produce an offering of novel solutions for addressing current issues associated with rapid qualification methods for AM of Ti-6Al-4V components.

Page Count

209

Department or Program

Ph.D. in Engineering


Included in

Engineering Commons

Share

COinS