PSI - Issue 14

Digendranath Swain et al. / Procedia Structural Integrity 14 (2019) 207–214 Author name / Structural Integrity Procedia 00 (2018) 000–000

210

4

the DIC cameras and two AE sensors can be seen. VICSNAP was used for image grabbing and VIC3D was used for DIC displacement and strain analysis following the procedures in Swain et al. (2013). Table 1. Specimen fabrication and test machine details used during the experiments Material : 15CDV6 ESR steel SCT Specimen: 6 mm by 2mm deep surface cracks at the centre of the weld with fatigue loading (ASTM E740-03) PT Specimen: Standard dog-bone without any weld Gauge area and length (GL): SCT-50 mm (w) × 8 (t) mm and 15 mm GL , PT - 12.5 mm (w) × 7.66 mm (t) and 50 mm GL Testing machine for loading : INSTRON UTM Machine with 600KN capacity Loading rate - 1mm/minute (all the samples) with hold-time of 30 s at 100 MPa steps beyond 500 MPa Specimen gripping: SCT (hydraulic grips), PT (mechanical wedge grips)

Table 2. Details of AE and DIC instrumentation used during the specimen tests.

Details of AE Instrumentation

Details of DIC setup

Sensor - PAC Single ended R15, 150kHz resonant Pre-amplifier - PAC 1220A type (20-1200khz band width)

CCD Camera – 2 Point Grey cameras in a stereo arrangement

Lens - 35 mm (Schneider make)

DAQ - PAC Express AE system

DAQ s/w - VIC SNAP with 2.5fps and 7.5 fps

Couplant - Molykote high vacuum grease

DIC random pattern : Spray brush with black on white AOI – 40 (l) × 50 (w) mm 2 (SCT), 40 (l) × 12.5 (w) mm 2 (PT)

Sensor Placement – two nos. 50mm away from the centre

Threshold applied - 40dB

DIC Analysis s/w - VIC3D Calibration grid - 3 mm

Calibration - Hsu-Neilsen B2 pencil lead break test

3. Results and Discussion The longitudinal strain fields aligning with the tensile loading axis are shown in Figs 3a and 3b for the SCT specimens just before failure. Strains were extracted at the points inside the circles in Figs 3a and 3b for correlating with AE signals. In SCT specimens it was very difficult to locate the crack-tips (evident in Fig. 1c); hence a point closer to the crack-tip was chosen for strain field correlation. The asymmetric strain field seen was due to the eccentric position of the crack. From the strain fields it can be seen that the strains near to the crack tip were fully localized. Hence, it can be suspected that the crack may be growing in radial direction (along the thickness of the specimen). A a through crack would have created severe surface strain fields around the crack tip on the surface, which was not seen here. The longitudinal displacement fields for the SCT specimens are shown in Figs. 3c and 3d corresponding to loads in Figs. 3a and 3b, wherein two black dots are placed near the centre of the cracks with arrow lines to indicate the CCOD in these specimens. The differences of the displacements of the top point and bottom point were estimated to comprehend the crack opening and correlate with AE signals. For correlating AE parameters with DIC in SCT specimens, typical data obtained from AE and DIC are assimilated in Fig 4 for SCT specimen-1. Figure 4a shows the comparison of CCOD and the cumulative hits obtained from AE. Both the curves in Fig. 4a show non-linearity beyond certain stress. Moreover, the AE cumulative hit curve shows specific jumps at certain loads correlating with DIC data. However, the non-linearity in AE cumulative hits seems more severe and prominent than CCOD. Therefore, it can be suspected that the crack growth is more severe in the bulk as compared to the surface. Similar inferences can be made from the comparison of strains near the crack-tip and the AE signal amplitude in Fig. 4b. Again, the AE cumulative energy and cumulative duration plots shown in Figs. 4c and 4d support the above inference of the internal nature of the crack growth. To gain further confidence, the second SCT specimen was tested. In the second specimen observation similar to specimen 1 was made (see Table 3). The black dotted lines in Fig. 4 are aligned with the trend changes to understand the crack growth behaviour. The nomenclatures used in Fig. 4 represent the crack growth phenomena, where it is proposed that there was an internal yielding near the crack tip beyond 800 MPa shown as-1, followed by