Application Notes
Published: 01 Jul 2021 · Last updated: 01 Jul 2021
Tags: EBSD
The physical properties of most crystalline materials are strongly dependent on their microstructural characteristics, including grain size, texture and boundary populations. Researchers have conventionally utilised a variety of techniques to measure the full range of microstructural parameters, including optical microscopy for grain size, X-ray diffraction for texture and electron backscattered diffraction (EBSD) for boundaries. Recent advances in the speed of EBSD data collection, driven by the latest generation of fast CMOS-based EBSD detectors, now enable the characterisation of all relevant parameters in just seconds, eliminating the need for cross-platform analytical approaches.
In this brief application note, we show how rolled and deformed Ni sheets can be effectively characterised in just a few minutes using the latest high-speed Symmetry S2 EBSD detector, analysing at >4500 indexed patterns per second (pps). Direct export into AZtecCrystal provides instantaneous grain size data; however, further examination of the results provides all additional key microstructural information with sub-µm resolution, demonstrating the power of the EBSD technique for routine materials characterisation.
Two samples were taken from a commercially available rolled sheet of Nickel alloy 200, approximately 2 mm in thickness. The sheet was folded perpendicular to the rolling direction as shown in figure 1. Two specimens were cut from the folded sheet, and one was selected for subsequent heat treatment. The heat treatment involved heating at 10 °C/min up to 600 °C, holding for 10 minutes and then cooling in air [1].
Fig 1. Schematic illustration of the folded Ni sheets, showing the EBSD analysis locations.
The samples were mounted in conductive bakelite so that the sectioned surfaces (in the ND-RD plane) could be mechanically polished for EBSD analysis. EBSD characterisation took place in a field emission gun (FEG) SEM, using the Symmetry S2 fibre-optically coupled EBSD detector and the AZtecHKL acquisition software. The electron beam was set to an energy of 20 kV and a current of 16 nA, and the S2 detector was operated in Speed 3 mode (using a pattern resolution of 158 x 88 pixels). Orientation map data were acquired automatically from 5 locations across the thickness of each sample, as illustrated in figure 1. Each map covered a 0.44 x 0.33 mm field of view, with a measurement step size of 0.5 µm. The average time to collect and index the 577,943 diffraction patterns from each area was 2:05 minutes, with an average acquisition speed of 4,600 pps and indexing success rate of 95%. The EBSD datasets were directly exported into AZtecCrystal and were processed using a batch processing template to clean the data and to calculate the grain size, texture and boundary characteristics.
When EBSD datasets are directly exported into AZtecCrystal, the software immediately plots an orientation and boundary map and calculates the grain size, enabling a rapid assessment of the grain characteristics in each area. Two example orientation maps, taken from the upper edge of the polished section (i.e. the compressive regime) before and after heat treatment are displayed in figure 2. Immediately apparent is the increased grain size in the heat-treated sample, and this is confirmed by the grain size measurements in figure 3. At the edges of the sample, where the effect of the heat treatment is most pronounced, there has been a 400% increase in the mean grain size, whereas in the middle of the sample there has been no statistical change in grain size.
(a)
(b)
Fig 2. Example orientation maps collected in just 2 minutes from the uppermost (compressive) area in both samples (a) As-folded sample (b) Heat-treated sample. In both maps coincident site lattice boundaries are marked by colour lines, and the scale bar represents 100 µm.
Fig 3. Graph plotting the variation in grain size (equivalent circle diameter) across each sample, highlighting the significant grain growth caused by the 10 minute heat treatment.
Representing and interpreting textures can be challenging for non-experts: AZtecCrystal provides all the standard tools (such as pole figures and orientation distribution functions – ODFs) for a full and detailed analysis of textures, but it also enables a quick estimate of the absolute strength of texture in a dataset. This can be represented by the J-index, derived from the ODF calculations, or the M-index, measured from the boundary disorientation distributions [2]: both enable a simple assessment of whether the texture is strengthening or weakening with heat treatment.
In general, the results indicate that any change in texture following heat treatment is more subtle than the corresponding changes to grain size. The original FCC rolling texture has been modified, to some extent, by the folding of the sheet, yet the major texture components are still clearly visible in the orientation distribution function (ODF), as exemplified by the lowermost area (tensile regime) of the as-folded sample, shown in figure 4 (left). These include a strong cube and copper texture component, in addition to well-developed α and β fibre components. Following heat treatment (figure 4 - right), the cube texture component has strengthened but the other texture components have noticeably weakened.
Fig 4. Orientation Distribution Function (ODF) serial sections (Phi 2 constant) showing the textures in the lowermost (tensile) regions in both samples. (Left) As-folded sample. (Right) Heat-treated sample.
The change in the overall texture strength is shown in figure 5, plotting the variation in M-index for both samples. The M-index can vary from 0 (random texture) to 1 (single crystal), and here the results indicate that at the margins of the sample there has been either no change or, in the case of the tensile region, a noticeable weakening of the overall texture. However, away from the sample edges the texture has significantly strengthened whilst in the centre, where the heat treatment has had minimal effect, the texture strength has remained unchanged.
Fig 5. Variations in texture strength across both samples, as determined using the M-index method.
The increase in grain size and modification to the texture is also accompanied by additional microstructural developments, such as significant changes in boundary populations. Although the orientation maps (figure 2) provide a quick summary of the boundary populations, a more rigorous statistical analysis of true boundary lengths is rapidly provided by AZtecCrystal and these results are summarised in figure 6.
Towards the edge of the sample, where the heat treatment has caused complete recrystallisation, there is a dramatic decrease in the total length of low angle boundaries (with 2–10° disorientation angles), a moderate decrease in high angle boundaries (>10°) and a slight increase in the abundance of Sigma 3 coincident site lattice (CSL) boundaries. In the centre of the sample, the effect of the heat treatment on the boundary populations is minimal.
Fig 6. Chart showing the variations in lengths of specific boundary types for both the as-folded (blue) and heat-treated (red) samples, from the top of the cross section (compressive regime) to the bottom (tensile regime).
The grain, texture and boundary characteristics can also be used to classify the microstructures into areas that have been recrystallised and areas that retain the initial, deformed structures. AZtecCrystal utilises a machine learning approach to classification that enables the system to be trained on a specific sample type. In figure 7 the microstructures for the heat-treated sample have been classified using a combination of local deformation and grain information, highlighting the recrystallisation at the sample edges.
Fig 7. Maps showing the deformed (red) and recrystallised (blue) regions in the heat-treated samples, automatically classified using AZtecCrystal. Values show the recrystallised fraction.
Kernel Average Misorientation (KAM) maps showing the distribution of strain across both samples, with the as-folded sample on the left and the heat-treated sample on the right. Brighter colours indicate higher strain, grain boundaries are marked with black lines and coincident site lattice (CSL) boundaries with colour lines.
Two rolled Ni sheet samples have been folded, with one subsequently being heat treated to modify the microstructure. Each sample has been effectively characterised by mapping 5 areas across the sheet thickness using high-speed EBSD. Each map has been collected in just over 2 minutes, amounting to approximately 10 minutes total analysis time on each sample, with instantaneous information provided about the grain size (to international standards). Additional offline processing rapidly provides further detailed information about the boundary populations and the texture characteristics, not to mention additional information on the nature and extent of deformation and recrystallisation.
The results highlight the increase in grain size and decrease in low angle boundaries associated with the recrystallisation towards the edges of the heat-treated sample, as well as a more complex modification of the strength and nature of the sheet's texture. All of these changes will have profound effects on the sheet's physical properties.
The ability to collect such a breadth of key microstructural data in only a few minutes per sample, and to enable detailed characterisation of the effect of the heat-treatment process, is testament to the speed and exceptional sensitivity of the Symmetry S2 EBSD detector coupled with the power and intelligence of the AZtecCrystal data processing software. The results demonstrate that effective and comprehensive EBSD characterisation of a wide range of microstructures is now possible within just a few minutes, opening up applications of EBSD as a routine inspection tool.
© Oxford Instruments 2026