Bayesian Estimation of Grain Scale Elastic-Plastic Intrinsic Material Properties via Spherical Indentation Measurements and the Exploration of Design of Experiments Strategies

By Castillo, Andrew R.

Georgia Institute of Technology

Download (PDF)

Licensed according to this deed.

Published on

Abstract

Advisors: Surya R. Kalidindi, Joseph Roshan, David L. McDowell, Hamid Garmestani, Shreyes N. Melkote

This thesis is focused on establishing and demonstrating a statistical framework for the objective fusion of data acquired from multiscale experiments and multiscale models performed to understand and predict the intrinsic material behavior. What makes this difficult is that the experimental data often provides information only on derived quantities from the material response (only these can be measured at present) and not directly the parameters present in the physics-based multiscale materials constitutive models. Consequently, one has to use sophisticated statistical theories to estimate the values of the critically needed material parameters and quantify rigorously the implicit uncertainty in this quantification. A mathematical framework that addresses this gap and its unique capabilities are demonstrated through the extraction of single crystal elastic-plastic constants for thermodynamic phases present in the microstructure of a metallic alloy and the extraction of laminate level properties for multi-laminate composite system.

Cite this work

Researchers should cite this work as follows:

  • Castillo, Andrew R. (2021), "Bayesian Estimation of Grain Scale Elastic-Plastic Intrinsic Material Properties via Spherical Indentation Measurements and the Exploration of Design of Experiments Strategies," https://matin.gatech.edu/resources/4093.

    BibTex | EndNote

Submitter

MATIN Development Team

Tags