Introduction
Shot peening is a mechanical processing surface treatment where small particles are fired at the surface of a material with enough force to cause plastic deformation. This deformation introduces a compressive stress layer that improves the strength and fatigue performance of treated material. From an electron backscatter diffraction (EBSD) characterization perspective, the level of plastic deformation can make characterization challenging with traditional approaches, as the level of plastic deformation present significantly reduces the quality of the EBSD pattern, making band detection via the Hough transform and subsequent indexing difficult.
Spherical indexing is a new approach to indexing EBSD patterns, based on a forward modeling approach to comparing experimental EBSD patterns to simulated EBSD patterns to find the correct orientation and phase. In this application note, a shot-peened titanium alloy was characterized using both conventional Hough indexing and spherical indexing. Significant improvements were observed with spherical indexing, including the ability to characterize the heavily deformed region at the surface of the material.
Discussion
In this work, shot-peening was performed on a dual-phase titanium alloy. This sample was kindly provided by Mr. Prathompoom Newyawond at the NSTDA Characterization and Testing Service Center. A cross-section of the treated surface was then prepared for EBSD analysis. EBSD data was collected using a Clarity™ direct detector with APEX™ software. Two different EBSD scans were collected, a lower magnification scan covering a 34 x 84 µm area with a 100 nm sampling step size, and a higher magnification scan covering a 4.6 x 14 µm area with a 30 nm step size.
Figure 1. a) An image quality map of lower magnification region of interest. b) An inverse pole figure (IPF) orientation map measured with Hough indexing (HI) colored relative to the normal of the shot-peened surface. c) A phase map measured with HI, with the titanium alpha phase colored red and the titanium beta phase colored blue.
Figure 1 shows the results from the lower magnification scan. Figure 1a shows the EBSD image quality (IQ) map. The shot-peened surface is on the right-hand side of the analyzed area. The image quality metric is derived from the brightness and sharpness of the bands in the EBSD pattern, as detected by the Hough transform. Stronger and sharper patterns are shaded brighter, while weaker and blurrier patterns are shaded darker in a continuous greyscale image. At the shot-peened surface, a significant reduction in image quality is observed due to the heavy deformation introduced during processing.
Figure 1b shows the inverse pole figure (IPF) orientation map derived via Hough indexing. In this map, lower confidence points have been partitioned out of the analysis and are colored black. The orientations are shown relative to the normal of the shot-peened surface (the right-hand side in this image) and not the more typical surface normal of the sample in the scanning electron microscope (SEM). This was selected to represent the surface orientation subjected to the surface treatment. Each point is colored by orientation relative to the colored reference stereographic triangle. Figure 1c shows the phase map derived via Hough indexing. Note the lamella of beta grains between the primary alpha grains and that indexing performance degrades closer to the shot-peened surface.
Figure 2. a) An IPF orientation map measured with spherical indexing (SI) colored relative to the normal of the shot-peened surface. b) A phase map measured with SI, with the titanium alpha phase colored red and the titanium beta phase colored blue.
Figure 2a shows the IPF map of the same area after spherical indexing. For spherical indexing, the patterns are saved during the initial acquisition and then reprocessed using the OIM Matrix™ module within OIM Analysis™ 9. In this work, spherical indexing was also combined with NPAR™ to improve the EBSD pattern signal-to-noise ratio and subsequent pattern matching. Note, in comparison to the Hough indexing in Figure 1b, higher quality data is found up to the edge of the cross-sectioned sample and there is significantly finer grain size at the treated surface. Figure 2c shows the phase map after spherical indexing. The beta phase has been resolved between the alpha grains throughout the analyzed area. To differentiate the alpha and beta phases, the experimental pattern is compared against simulated master patterns from both phases, and the best fit is selected for both phase and orientation.
Figure 3. a) A PRIAS map using lower PRIAS ROI. b) A PRIAS map combined with a color IPF map using HI. c) A PRIAS map combined with a color IPF map using SI.
Figure 3a shows a PRIAS™ image from the lower magnification scan, using the bottom region of interest (ROI) PRIAS channel. This image shows the sample surface morphology and edge of the sample surface. To better show the difference in indexing performance between Hough and spherical indexing, the colored IPF maps from both approaches have been overlain on the PRIAS image in Figures 3b and 3c. In these maps, lower confidence points are now transparent to show only the PRIAS signal if indexing performance was not satisfactory. Comparing Figures 3b and 3c show the spherical indexing improvements up to the edge of the sample and within the beta-lath grains.
Figure 4. a) A KAM map measured with HI. b) A KAM map measured with SI. Improvement in orientation precision with SI improves contrast between the base metal and the deformed surface layer.
The kernel average misorientation (KAM) map is well suited for characterizing the plastic deformation that develops during shot peening. The KAM maps for both Hough and spherical indexing are shown in Figures 4a and 4b, respectively. Note the orientation precision improvement with spherical indexing reduces the noise in the KAM image away from the treated surface and provides more contrast with the plastic deformation zone at the surface. With Hough indexing, the decrease in pattern quality from the deformation reduces the prevision of band detection, which increases the KAM noise level. These results show that spherical indexing provides a much clearer image of the effects of shot peening.
Figure 5. a) An image quality map of the higher magnification region of interest. b) An IPF orientation map measured with HI. c) An IPF orientation map measured with SI. Indexing improvements observed relative to HI.
Figure 5a shows the higher magnification IQ map at the shot-peened surface. The IQ map does not show a clear structure at the surface, indicating weak patterns corresponding to higher levels of plastic deformation. Figures 5b and 5c show the IPF maps with Hough and spherical indexing, respectively, with lower confidence points now again colored black. There is a significant improvement with spherical indexing within this heavily deformed region of interest, revealing a fine-grained region that has developed during the shot peening. These results show that spherical indexing is required to characterize the microstructure that develops during this processing.
Figure 6. a) A grain map derived from SI. b) A grain size map, colored according to shading used in Figure 6c. c) The grain size distribution with color coding applied to Figure 6b.
Figure 6a shows a unique grain color map derived from the spherical indexing results. In this map, the grains are determined using a 5° tolerance angle and then randomly colored to show size and morphology. The grains become much smaller near the surface. Figure 6b shows a grain size map, with Figure 6c showing a grain size distribution with the same coloring scheme applied as Figure 6b. Grains smaller than 50 nm have been resolved in the deformed region.
Conclusion
This work shows that shot peening introduces a significantly plastically deformed region at the surface of the treated material. Traditional Hough indexing is not sensitive enough to satisfactorily analyze the heavily deformed region, but the microstructure can be analyzed using the new spherical indexing approach implemented in OIM Analysis 9. This data can then be used to understand and optimize the shot peening parameters to achieve the deformation level and depth required for a given material application.