PSI - Issue 60

S.S. Satpute et al. / Procedia Structural Integrity 60 (2024) 525–534

529

S. S. Satpute et. al./ Structural Integrity Procedia 00 (2023) 000–000 5 A typical fracture test set up with an arrangement for image capturing for crack growth prediction, is shown in Fig. 3 (a) . The test component is a through-wall cracked pipe under four-point bend loading. The pipe has been tested under a high temperature of 300 deg. C, heated by an electrical thermal pad from outside as well as inside considering that higher thickness requires heating from both sides. The pipe block near crack tip has been wrapped by heat insulation material. The insulation material is not applied on the grid area and the crack tip, to ensure the visibility of these regions during the test for capturing the crack tip images. The imaging system captures images from all the channels simultaneously at predefined captures rates. The preconfigured captured rates are 1 set/min,2 sets/min, 5 sets/min, 12 sets/min and 1 set/sec. During the initial stage of the experiment, the capture rate is configured lowest. As the experiment progresses , the captured rate is increased gradually to record the crack propaging at an increasing speed. Typically a single cyclic fracture experiment records from 500 to 1000 image sets based on the applied load cycles numbers. High temperature resistant paint was used which withstood 300 deg C and the obtained images are of acceptable quality for crack growth evaluation. Fig 4 is a typical image of the GUI of the software – which displays the Load and LLD, CMOD, and crack tip of the grid areas A and B. In this experimental setup, the spatial resolution achieved was approximately 40 µm / pixel.

Fig. 4. GUI of online imaging system for capturing the images of crack tips and instantaneous test data

3.2. Crack Tip Location among Grid Points

Both crack tip images have been analysed manually for the crack growth on both crack tips. The pixel coordinates of both the crack tips are marked manually. For computing calibration factor, the neighbouring grid points just above the crack tip are considered as shown in the Fig.5 (a) and Fig. 5(b). The pixel coordinates and the 2D world coordinates(like 21, 22, etc.) of the grid points are known from the image, using which the calibration factor is computed for each image individually. The nearest 2D world coordinates are 10mm and the pixel-difference is known using the image software. The caliberation factor are in the range of 40µm/pixel and 50µm/pixel for 8-inch and 12-inch pipes respectively for present experimental set up. Computing the calibration factor for each image individually using the neighboring grid points ensures that the measurements are independent of wide crack opening

, pipe deformation and 3D to 2D projection of a curved surface. This methodology is explained in detail in the next section.