Rémy Besnard

About quantitative EBSD analysis of deformation and recovery substructures in pure Tantalum

The aim of this work is to present a quantitative analysis of features involved in recovery during annealing of deformed Tantalum. In pure metals where crystalline defects usually have high mobility, dislocation annihilation and rearrangement occur to a great extent prior to recrystallization. Therefore a complete understanding of recrystallization cannot be accomplished without an advanced knowledge of the recovery phenomenon. Depending on whether dislocations induce a measurable curvature in the crystal lattice or not, they are called Geometrically Necessary Dislocations (GNDs) or Statistically Stored Dislocations (SSDs) respectively. In the present work only GNDs are considered. For this purpose electron backscatter diffraction (EBSD) is an advantageous technique to obtain statistically representative results when compared to Transmission Electron Microscopy (TEM). However, a quantitative analysis of GNDs from EBSD data is not straightforward. Since local misorientations are induced by the curvature of the crystal lattice caused by GNDs, GNDs analysis can be done using local misorientations. However the values obtained from this analysis are step size dependent and influenced by the measurement noise. Reasoning on the basis that when the step size tends to zero, local misorientation should also tend to zero, measurement noise can be estimated [1]. The measurement noise appears to notably be very much dependent on the amplitude of local misorientations, which must be considered in the perspective of GND density calculation

Statistical analysis of dislocations and dislocation boundaries from EBSD data

Electron BackScatter Diffraction (EBSD) is often used for semi-quantitative analysis of dislocations in metals. In general, disorientation is used to assess Geometrically Necessary Dislocations (GNDs) densities. In the present paper, we demonstrate that the use of disorientation can lead to inaccurate results. For example, using the disorientation leads to different GND density in recrystallized grains which cannot be physically justified. The use of disorientation gradients allows accounting for measurement noise and leads to more accurate results. Misorientation gradient is then used to analyze dislocations boundaries following the same principle applied on TEM data before. In previous papers, dislocations boundaries were defined as Geometrically Necessary Boundaries (GNBs) and Incidental Dislocation Boundaries (IDBs). It has been demonstrated in the past, through transmission electron microscopy data, that the probability density distribution of the disorientation of IDBs and GNBs can be described with a linear combination of two Rayleigh functions. Such function can also describe the probability density of disorientation gradient obtained through EBSD data as reported in this paper. This opens the route for determining IDBs and GNBs probability density distribution functions separately from EBSD data, with an increased statistical relevance as compared to TEM data. The method is applied on deformed Tantalum where grains exhibit dislocation boundaries, as observed using electron channeling contrast imaging.