Al–Mn (AA3xxx) alloys are widely used in a broad range of products including beverage cans, automotive heat exchangers and packaging. In these alloys, tailoring the grain size and texture and the associated changes in microchemistry (i.e., the partition of alloying elements between solid solution, constitutive particles and dispersoids) during thermomechanical processing is key to achieve desired properties (e.g. strength and formability).

Here second-phase particles are of particular importance. During annealing, after hot or cold deformation, large constituent particles (≳1 μm) may promote recrystallization by particle-stimulated nucleation (PSN), while closely spaced fine dispersoids (sub-micron) may retard or even inhibit the transformation by pinning (sub)grain boundaries via Smith–Zener pinning. Both mechanisms may have strong influence on the final microstructure and texture and thus the final material properties. 

The paper presents the application of a recently developed experimental methodology to directly correlate (sub)grains and particles during recovery and recrystallization in an Al-Mn alloy. The methodology combines subgrain data from electron backscatter diffraction (EBSD) and particle data from backscatter electron (BSE) images. Correlated analysis of secondary phase particles down to 0.03 μm in diameter and subgrains and subgrain boundaries is enabled by the higher resolution of the BSE images. The workflow is demonstrated on a cold-rolled and recovered Al-Mn alloy, where constituent particles formed during casting and dispersoids formed during subsequent heating affect recovery and recrystallization upon annealing after cold deformation.

A strong subgrain boundary–dispersoid correlation is demonstrated, as 60% of the dispersoids are found at subgrain boundaries (closer than ~0.1μm), and a modified orientation dependent Smith-Zener drag expression has been derived, accordingly. This new and quantitative methodology provides improved insight into the mechanisms behind the orientation dependent nucleation and growth behaviour during recovery and recrystallization in Al-Mn alloys and in particular the strong P texture often obtained, compared to CubeND and Cube, in conditions of strong concurrent precipitation. 

The EBSD indexing is done by dictionary imaging (DI), using the open-source software EMsoft, followed by refinement as implemented in kikuchipy v0.6, where every experimental pattern is compared to a dictionary of dynamically simulated patterns. DI has proven more robust towards pattern noise (e.g., heavily deformed microstructures at a sub-micrometer length) compared to traditional Hough indexing, and it also allow for particle identification using EBSD.
 
The work presented in this paper is part of the PhD project of Håkon W Ånes, with Prof. Knut Marthinsen as the main supervisor and Prof. Ton van Helvoort, Department of Physics as the co-supervisor. 

Håkon is a major contributor to the software kikuchipy, which is an open-source Python library for processing and analysis of EBSD patterns. The PhD project of Håkon Wiik Ånes has been funded by the Department of Materials Science and Engineering and is related to NTNU Aluminium Product Innovation Center (NAPIC).

It should be emphasized that the methodology should be applicable to other alloy systems where quantitative correlated analysis of minor secondary phases (of sizes down to and smaller than 0.1 μm) and orientations of (sub)grains is important to understand microstructure evolution. It can, together with a lot of other options in the software kikuchpy be a valuable tool for the characterization of microstructure evolution during thermomechanical processing, as basis for alloy and process development of any alloy system, highly relevant for several SFI PhysMet research activities.

Read the paper here.

References:

¹Ånes, H.W., van Helvoort, A.T.J., Marthinsen, K. (2022). Correlated subgrain and particle analysis of a recovered Al-Mn alloy by directly combining EBSD and backscatter electron imaging. Materials Characterization, Vol. 193, 112228.