Fracture, Fatigue, Failure and Damage Evolution , Volume 3

62 D. Celli et al. Fig. 9.1 Academic finger disk 9.2 Experimental Procedure First, a reference data set and spectral analysis was collected via SLV or the current measurement approach used within TWE for bladed disks and blisks. The component analyzed in this study was a 16-bladed academic finger disk, shown in Fig. 9.1, and has an outer diameter of 30.48 cm, and each blade has a width and thickness of 1.91 and 0.64 cm, respectively. The academic finger disk was excited via a JODON piezoelectric shaker, and velocity data was taken with a PSV-500 scanning head laser. A sine wave chirp, sweeping through frequencies of 0.5 to 1.5 kHz was generated via Polytec Scanning Vibrometry (PSV) system and hardware. Note, within the PSV software, displacement data was calculated and then exported for direct comparison to DIC data. The frequency resolution for the SLV was 312 mHz and the sample time for each sweep was 3.2 s. There were 177 points of interest defined along the academic finger disk. However, PSV software samples points of interest multiple times, and therefore the experimental testing time was approximately 1 h and 43 min. For the DIC experimental setup, Photron SA-Z cameras were utilized in stereo to capture time domain displacement data. Note that the Correlated Solutions software was used both for image acquisition and post-processing of images. The working distance of the cameras was 81.28 cm with a stereo angle set to 25 deg. Nikon 50 mm lenses were used to obtain a field of view of 33.02 cm. This yielded a resolution of approximately 0.329 mm/pixels, and for this reason a spray paint speckle pattern was applied with an approximate facet diameter size of 1 mm. Af# of 8 was found to yield a sufficient depth of field for displacements as well as maintaining focus of entire component. Note, the minimum frame rate required for this test was guided by Nyquist criterion or (fmax/2) or 3000 fps in this case. However, a conservative frame rate of 5000 fps was selected or 1.4 fNyq with an exposure time of 1/33,333 s. Calibration was performed via a 20 mm Correlated Solutions calibration plate, and a 0.026 calibration score was obtained. Finally, image acquisition was taken over the same period as the sine wave chirp, 3.2 s. Note that for only one scan of the component utilizing DIC requires 3.2 s, whereas the same scan to collect all 177 points of interest using SLV requires 177∗3.2 or 566.4 s. 9.3 Experimental Results In order to obtain mode shape data via DIC, correlation of DIC was first performed in Correlated Solutions VIC-3D with a subset size of 19 and a step size of 3. Eight points spanning the length from the hub to the tip of the top most blade, shown in Fig. 9.1, and out-of-plane displacement data W, along with in-plane spatial coordinates, were sampled for each image. Next an FFT analysis was performed on the displacement data collected via DIC to gather magnitude (|W|) and phase (φ) data

RkJQdWJsaXNoZXIy MTMzNzEzMQ==