Skip to main content

An Experimental and Modeling Investigation of Solidification and Microstructural Evolution in Magnesium Alloys using in situ Techniques

EMSL Project ID


The goal of this proposal is to understand the microstructural evolution during non-equilibrium, rapid solidification of a molten magnesium-aluminum-zinc (AZ91) alloy, and during its subsequent heat-treatment. The microstructure of AZ91 comprises solid-solution alpha-Mg and Mg17Al12 precipitates whose volume fraction and morphology need to be controlled during solidification and subsequent heat-treatment to produce a high-strength casting. Therefore, the objectives of this work are: (1) Understand the solidification kinetics of AZ91 melt at high cooling-rates and, (2) understand the kinetics of phase evolution of Mg17Al12 and alpha-Mg during heat-treatment. These objectives will be achieved through a combination of in situ experiments and atomistic simulations with the help of EMSL’s experimental and computational resources, respectively. In parallel, solidification modeling will be performed with the help of ESI North America, an international leader and provider of commercial solidification software ProCAST. At EMSL, the project will use the dynamic transmission electron microscope (DTEM) to interrogate the solidification behavior of Mg-9wt.% Al-1wt.% Zn (AZ91) alloy samples under high cooling rates (100-1000 °C/s). The ability to characterize rapidly moving (~m/s) liquid-solid interface and the associated dendritic and precipitate nucleation and growth phenomena requires imaging capabilities with high temporal (nano-micro second) and spatial (~nm) resolutions, such as those afforded by the DTEM. Thus, the results from in situ DTEM experiments will provide unique solidification kinetic data that is otherwise not achievable by any other technique and this data will serve as an input to the task of numerical modeling of the solidification process. The numerical modeling portion of the research will be performed by project partner, ESI North America, which will optimize its commercial solidification code, ProCAST, to enable prediction of microstructures in rapidly solidified Mg alloys. Subsequently, the post-solidification heat-treatments will be simulated using the hot-stage capability of EMSL’s aberration-corrected scanning transmission electron microscope (STEM) to obtain high-resolution diffusion and phase transformation kinetics, at high spatial resolution, as a function of temperature. The STEM experiments will help understand the discontinuous and continuous precipitation kinetics of Mg17Al12 in a non-equilibrium cast microstructure. Such information is useful for the development of high-strength low-density Mg alloys since it is the continuous precipitation that is responsible for the age hardening phenomenon in Mg-Al based alloys. Finally, the experimental and numerical modeling will be combined with EMSL’s atomistic simulation capabilities using first-principles and classical potentials approach. The atomistic modeling work will incorporate the effects of non-equilibrium vacancies, expected to form during rapid solidification, to understand the thermodynamic and mechanical properties of the phases present in the cast microstructure. Thus, this research is expected to provide valuable in situ kinetic data in Mg alloys with high spatial and temporal resolution that is generally not available to date. The experimental data will be correlated with corresponding atomistic and numerical model predictions and improve our understanding of processing of Mg alloys. Development of such fundamental understanding of Mg alloys will enable their use as high-strength low-density structural components in light-weight cars and trucks to reduce fossil-fuel consumption and reduce green-house gas emissions.

Project Details

Start Date
End Date


Principal Investigator

Aashish Rohatgi
Pacific Northwest National Laboratory


Nigel Browning
University of Liverpool

Team Members

Karen Kruska
Pacific Northwest National Laboratory

James Evans
Environmental Molecular Sciences Laboratory

Libor Kovarik
Pacific Northwest National Laboratory

Niranjan Govind
Pacific Northwest National Laboratory

Related Publications

Nandipati G, N Govind, A Andersen, and A Rohatgi. 2016. "Self-Learning Kinetic Monte Carlo Simulations of Al Diffusion in Mg." Journal of Physics: Condensed Matter 28(15):Article No. 155001. doi:10. 1088/0953-8984/28/15/155001