Deformation behavior of Mg-8.5wt.%Al alloy under reverse loading investigated by in-situ neutron diffraction and elastic viscoplastic self-consistent modeling. (1st April 2016)

• Record Type: Journal Article
• Authors: Wang, H.; Lee, S.Y.; Gharghouri, M.A.; Wu, P.D.; Yoon, S.G.
• Is Part Of: Acta materialia. Volume 107(2016)
• Page Start: 404
• Page End: 414
• Publication Date: 2016-04-01
• Subjects: Magnesium alloy -- Twinning -- De-twinning -- Internal elastic strain -- Polycrystal plasticity model; Materials -- Periodicals; Materials science -- Periodicals; Materials -- Mechanical properties -- Periodicals; Metallurgy -- Periodicals; Chemistry, Inorganic -- Periodicals; 620.112
• DOI: 10.1016/j.actamat.2016.01.066
• Languages: English
• ISSNs: 1359-6454

Abstract

Abstract: The EVPSC-TDT model for polycrystal plasticity and in-situ neutron diffraction have been used to investigate the behavior of a Mg-8.5wt.%Al alloy with two starting textures: 1) a typical extrusion texture in which a majority of the grains are oriented favorably for extension twinning via compression perpendicular to the basal pole, and 2) a modified texture in which extension twinning can be activated via tension parallel to the basal pole in a majority of the grains. Using a small number of adjustable parameters, and only two macroscopic tensile stress–strain curves for calibration, the model is able to capture, quantitatively, the trends in multiple data sets, including grain-level elastic lattice strains, and diffraction peak intensity changes due to lattice re-orientation associated with twinning. For twinning, the model assumes a polar critical resolved shear stress activation criterion and assigns the stress and hardening of the parent crystal to a newly formed twin. The model allows twinning to be driven either by the stress in the parent crystal (matrix reduction), in which case all of the twin transformation strain is assigned to the matrix, or by the stress in the twin (twin propagation), in which case all of the twin transformation strain is assigned to the twin. A detailed comparison between the model predictions and the neutron diffraction data reveals that assigning all of the twin transformation strain either to the matrix or to the twin is too one-sided, leading to excessive relaxation and hardening effects. A more equitable partitioning of the twin transformation strain is necessary. It is suggested that the stress and hardening assigned to a newly formed twin is of less importance to the performance of the model than the partitioning of the twin transformation strain. Graphical abstract: