Date of Award
Senior Honors Thesis
College or School
The purpose of this project is to reduce a large statistical distribution of metal microstructure orientations to a manageable distribution to be used in metal forming simulations. Microstructure sensitive simulations at the macro-scale are impractical, because with so many state variables associated with material microstructure data, these simulations are extremely computationally expensive.
The goal was to develop a framework to accurately model plastic material response while repre- senting the material microstructure in a more compact form, reducing 106 or more microstructure orientations to a significantly smaller statistical distribution of representative orientations. This will significantly increase the computational efficiency and make the design process known as mi- crostructure sensitive design (MSD) feasible for industry applications. This framework is applied to metals with both cubic and hexagonal structure to validate this approach for slip and twinning deformation mechanisms.
Performing microstructure sensitive metal-forming simulations is widely recognized as a computational challenge because of the need to store large sets of state variables related to microstructure data. This makes the investigation of the accuracy of smaller, representative data sets in these simulations profitable.
The project accomplished two main goals; the development of an effective fitting algorithm to generate compacted data sets and validation of the framework for data compaction on metals with cubic structure, and hexagonal symmetry, with and without twinning. The research was applied to oxygen-free high-conductivity copper (OFHC Cu) and 6016 aluminum (Al-6016) for application of the framework to cubic metals. An anisotropic (clock-rolled) zirconium (Zr) texture was used to develop the framework for hexagonal metals. The minimum accurate data set for cubic was determined to be 825 orientations and for hexagonal metals, considering twinning and absence of twinning, the minimum number was 1600 orientations. This compaction method will increase the computational speed of microstructure sensitive forming simulations by several orders of magnitude, contributing to the computational feasibility of microstructure informed design.
Landry, Nicholas, "Computationally efficient representation of statistically described material microstructure for tractable forming simulations" (2014). Honors Theses and Capstones. 370.