Introduction
OIM Analysis™ is the leading software program for the visualization and processing of electron backscatter diffraction (EBSD) mapping data. Users can create different maps showing a sample's crystal orientation or grain morphology. They can create charts of grain boundary disorientation or local misorientation distributions and generate pole figure plots of preferred orientation distributions. This range of data analysis gives users the tools they need to understand their samples and corresponding microstructures better. Beyond this wide range of analytical possibilities, OIM Analysis also offers features that make the software easy to use and customizable for individual users. Templates can be generated with user-defined outputs and parameters that can be saved, shared, and reused on other datasets. These templates can be set up for one-button analysis or used with the batch processor to automatically and consistently analyze a series of datasets. OIM Analysis has set the bar for EBSD data processing, and now, with OIM Analysis 9, the best has gotten better.
OIM Analysis 9 has gotten faster. This is important, as the Velocity EBSD detectors provide significantly faster data acquisition rates than previous detectors. This allows users to collect more data and larger datasets. The improved algorithms open data sets and generate maps faster, and enable users to move images in the workspace without waiting to re-render the image. The fundamental improvements enhance the user experience with OIM Analysis 9.
In addition to streamlining the system to make it easier to use, significant improvements have been made to the OIM Matrix™ module within OIM Analysis. OIM Matrix is the module that generates and uses dynamical diffraction-based simulated EBSD patterns to improve pattern indexing compared to conventional Hough transform-based approaches. Within OIM Matrix, spherical indexing has been added, significantly increasing the speed of using the simulated patterns for indexing without decreasing the resulting data quality. This makes OIM Matrix easier to use in more routine and specialized applications.
Examples and Discussion
To illustrate the performance of OIM Matrix, data was collected from a sample of rolled aluminum. This sample type can be difficult due to the amount of plastic deformation within the microstructure. The resulting EBSD patterns will be more diffuse because the periodicity of the atomic structure is disrupted enough by the deformation to reduce the intensity and sharpness of the diffraction signal within the EBSD pattern. Figure 1a shows the IPF orientation map relative to the surface normal direction using traditional Hough-based indexing. A confidence index (CI) partition, including CI values greater than 0.1, has been applied to identify the correctly indexed points within the data. Points excluded from this threshold are colored black in this image. The indexing success rate was 25%. This data was collected with a Velocity Super EBSD detector over a 72 x 56 µm area with a 70 nm step size. The EBSD patterns were saved to a disk for subsequent analysis with OIM Matrix.
Figure 1. a) The IPF orientation map from a rolled aluminum sample relative to the surface normal direction, using traditional Hough-based indexing. b) The IPF orientation map from the same sample after spherical indexing analysis.
Figure 1b shows the IPF orientation map after spherical indexing analysis. The indexing success rate improved to 72%. The same confidence indexing thresholding was applied in this and all subsequent results. A key benefit of spherical indexing is the ease of use. Users select the spherical indexing option in the Reindexer, load the master pattern of the applicable phases, and set the bandwidth parameter for indexing. Figure 2 shows the OIM Matrix user interface when using spherical indexing. When loading the master pattern, OIM Matrix includes an extensive library of precalculated master patterns for many commonly analyzed materials. OIM Matrix also provides tools for calculating new master patterns when the crystallographic information of a material’s crystal symmetry, lattice parameters, and atomic positions are available.
Figure 2. The OIM Matrix user interface when using spherical indexing.
OIM Analysis also includes other tools for improving EBSD pattern indexing. One of these tools is NPAR™, an approach for enhancing the EBSD pattern signal-to-noise ratio by local kernel averaging of acquired EBSD patterns. Figure 3a shows the IPF orientation map of this data after NPAR processing, with an indexing success rate of 52% after this analysis.
Figure 3. a) The IPF orientation map for a rolled aluminum sample after NPAR processing. b) The results after combining spherical indexing and NPAR reindexing analysis on the same sample.
OIM Analysis was built to allow users to combine the advantages of OIM Matrix, spherical indexing, and NPAR to achieve the best possible results. Figure 3b shows the results after combining spherical indexing and NPAR reindexing analysis. The resulting data has an indexing success rate of 97%. It is important to note that this improved data quality is not a cleanup operation. It is based on actual measurements from acquired EBSD patterns. At each point, users can see the pattern and indexing results.
Another key synergy within the OIM Analysis workspace is the ability to apply OIM Matrix and NPAR on a selected data partition only. One example would be only to reindex points with a low initial confidence index. Another example would be only to reindex points of a particular phase. This improves the overall efficiency of reprocessing and gives users the flexibility to optimize data quality in novel ways for specific materials. OIM Matrix is also compatible with ChI-Scan™, where EDS data simultaneously collected with the EBSD data can be used to differentiate crystallographically similar samples to improve phase differentiation.
OIM Matrix and spherical indexing can also be used to improve the orientation precision of EBSD measurements. Figure 4 shows an example from data collected from a 1 mm x 1 mm region with a 30 µm step size on a silicon single crystal comparing the kernel average misorientation (KAM) distribution from the Hough indexing, spherical indexing, and spherical indexing plus an orientation refinement. The maps and distributions show a significant improvement, allowing for improved deformed microstructure characterization.
Figure 4. An example of data collected from a 1 mm x 1 mm region with a 30 µm step size on a silicon single crystal comparing the KAM distribution from the Hough indexing, spherical indexing, and spherical indexing plus an orientation refinement.
Figure 5 shows the effects of EBSD pattern pixel resolution and signal-to-noise levels of the average KAM values obtained with orientation refinement. The noise levels correspond to the initial pattern (noise 0), patterns after a moderate level of added noise (noise 1), and patterns after a significant level of added noise (noise 2). There are two key takeaways from this data. First, even with moderate noise, excellent orientation precision values can be measured with the orientation refinement functionality. Second, these results can be obtained with 120 x 120-pixel images, easily obtained by all EDAX EBSD detectors. Additional pixel resolution does not significantly improve performance.
Figure 5. The effects of EBSD pattern pixel resolution and signal-to-noise levels of the average KAM value obtained with orientation refinement.
Conclusion
The new features and functionality in OIM Analysis 9 will provide users with better data, faster analysis, and improved results.