SUBJECT: Ph.D. Proposal Presentation
BY: Pierre-Alexandr Juan
TIME: Thursday, September 25, 2014, 9:30 a.m.
PLACE: Love Building, 210
COMMITTEE: Dr. Laurent Capolungo, Chair (ME)
Dr.Surya Kalidindi (ME)
Dr. Olivier Pierron (ME)
Dr. Hamid Garmestani (MSE)
Dr. Stéphane Berbenni (CNRS, France)


The main objective of this thesis is the investigation and the quantification of the influence of parent-twin and twin-twin interactions on the mechanical response of hexagonal close-packed metals. To study parent-twin interactions, a mean-field continuum mechanics approach based on a new twinning topology in which twins are embedded in twinned grains has been developed. A first model Generalizing the Tanaka-Mori scheme to heterogeneous elastic media is applied to first and second generation twinning in magnesium. In the case of first generation twinning, the model is capable of reproducing the trends in the development of backstresses within the twin domain observed experimentally. A parametric study is performed to understand the roles of grain and twin shape and of relative volume twin volume fractions on the magnitude and directions of the backstresses. Applying the methodology to the case of second-generation twinning allows the identification, in exact agreement with experimental observations, of the most likely second-generation twin variants to grow in a primary twin domain. However, internal stress level magnitudes are excessively high because of the elastic behavior assumption. Then, the first model is extended to the case of elasto-plasticity. Using a self-consistent approximation, the model, referred to as the double inclusion elasto-plastic self-consistent (DI-EPSC) scheme, is applied to Mg alloy polycrystals. It reveals that deformation system activities and plastic strain distributions within twins drastically depend on the interaction with parent domains.

The influence of twin-twin interactions on nucleation and growth of twins is being statistically studied from zirconium and magnesium electron back-scattered diffraction scans. A new twin recognition software relying on graph theory analysis has been developed to extract all microstructural and crystallographical data. It is capable of identifying all twinning modes and all twin-twin interaction types occurring in hexagonal close-packed materials. The first results obtained from high purity Zr electron back-scattered diffraction maps reveal that twin-twin interactions hinder subsequent twin nucleation. They also show that mechanisms involved in twin growth may differ significantly for each twinning mode. A second study performed on AZ31 Mg and a new probabilistic model accounting for the statistical nature of twinning incidence will then be proposed.