predominantly between the B4C and Si phases. Because the CTE of SiC and Si are much closer, less thermal mismatch effect was expected between these two phases. Furthermore, given that B4C dominates the thermal mismatch behavior and the low volume fraction of SiC, only the mismatch between B4C and Si was considered. 8.3 Experimental A micro-Raman spectroscope (Renishaw InVia model, Hoffman Estates, IL, USA) equipped with a 532 nm (green) laser rated at 100 mW total power was used in this investigation. Using a motorized stage, both point and grid-type scans can be made to obtain Raman spectra for various residual Si regions. This data can then be compared with the spectrum of the Si standard, whose dominant peak is well-known to be 520 cm 1 (see Fig. 8.3). All scans were acquired at 10 % laser power using a100 objective. The test parameters used were chosen to maximize signal intensity while minimizing any effects of laser heating of the sample. Prior to each set of scans, the Raman system was calibrated using the Si standard. Scan variability has been verified to be as low as 0.02 cm 1. The Raman data was evaluated to determine if it could be used effectively to identify changes in key Si peak characteristics such as peak position, peak width, and peak intensity. Peak position was used to determine the presence of peak shifts away from the stress-fee crystalline Si peak position of 520 cm 1, which is indicative of the induced stress state [5]. Peak width and peak intensity were used as metrics for the evolution of structural disorder in the Si phase due to thermal mismatch. To simplify the analysis, only small, circular regions of several microns diameter were considered. The current study did not explore the spatial evolution of residual stresses, but simply explored the ability of the technique to detect Raman peak changes. Additionally, it was determined whether this method could be used to accurately map the microstructural makeup. Phase makeup was verified by the presence of characteristic peaks in the acquired spectra. Using a signal-to-baseline analysis, the area under characteristic peaks specific to the phases present were computed and scan points were color-coded according to the associated phase. This was used to verify correspondence between the intended scan regions and the acquired data. These are useful as overlays with the stress and disorder maps to verify that accurate data interpretation is given. 8.4 Results and Discussion In Fig. 8.3, an example spectrum collected from the residual Si region is shown, which revealed that the residual Si exhibited a notable peak shift, peak width increase, and peak intensity decrease compared to the Si standard. The peak position for the residual silicon region was shifted by almost 2 cm 1, the peak width increased by 8.6 cm 1(i.e., ~200 %), and the peak intensity Fig. 8.3 (a) Experimental setup for a Raman point scan of the Si (light grey) surrounded by B4C (dark grey) and (b) an example of a typical spectrum used for analyzing the Si peak. Note the peak shift, peak width increase, and peak intensity decrease in the residual Si region compared to Si standard 8 Micro-Raman Spectroscopic Evaluation of Residual Microstresses in Reaction Bonded Boron Carbide Ceramics 41
RkJQdWJsaXNoZXIy MTMzNzEzMQ==