Date of Award

12-2017

Degree Type

Thesis

Degree Name

Master of Science in Mechanical Engineering (MSME)

Department

Mechanical Engineering

Committee Chair

Yung C. Shin

Committee Member 1

Marcial Gonzalez

Committee Member 2

David R. Johnson

Abstract

This study focuses on devising a unified multi-scale numerical framework to predict the grain size evolution by dynamic recrystallization in metals and alloys for an array of severe plastic thermo-mechanical deformation conditions. The model is developed to predict the temporal and spatial grain size evolution of the material subjected to high strain rate and temperature dependent deformation.

Dynamic recrystallization evolves by either a continuous grain refinement mechanism around room temperatures or by a discontinuous grain nucleation and growth mechanism at elevated temperatures. The multi-scale model bridges a dislocation density-based constitutive framework with microscale physics-based recrystallization laws to predict both the types of recrystallization phenomena simultaneously. The simulations are conducted within an integrated probabilistic cellular automata-finite element framework to capture the physics of the recrystallization mechanisms.

High strain rate loading experiments in conjunction with microstructural characterization tests are conducted for pure copper to characterize the dynamic grain size evolution in the material and evaluated against the model predictions. Synchrotron X-rays are integrated with a modified Kolsky tension bar to conduct in situ temporal characterization of the grain refinement mechanism operating during the dynamic deformation of copper and evaluated against the developed model kinetics.

Finally, the model is implemented to predict the grain size evolution developed during the friction stir spot welding of Al 6061-T6 for varying tool rotational speeds. The experiments show that the original microstructure is completely replaced by a recrystallized fine-grained microstructure with the final average grain size and morphology dependent on the process parameters. The model accurately predicts the process temperature rise with increasing tool rotational speeds, which results in a higher rate of grain coarsening during the dynamic recrystallization phenomenon.

Share

COinS