13th International Conference on Fracture June 16–21, 2013, Beijing, China -7- be due to the sensor position and analysis method. The sensor was placed only based on the nominal position of the precrack tip on the specimen surface. The real, measured initial crack length across the specimen thickness is not taken into consideration. Therefore, sensor position (2) on the crack was investigated. With (2) failure of the n-th strand positioned at the post-test measured mean initial crack length a0 indicates crack initiation. Scatter of the crack initiation forces Fini can still be observed with this procedure, 0.86Fmax ≤ Fini ≤ 1.0Fmax. But it is clearly lower than with method (1), although the tests were performed at -40 °C this time. Nevertheless, the data base for method (2) is still too small to finally rate the appropriateness of the technique. Figure 8. Principle of detection of stable crack initiation by failure of crack sensors, left: (1) crack sensor position at crack tip and failure of the first strand, right: (2) crack sensor position on the crack and failure of the n-th strand positioned at the measured initial crack tip. 2.2. Large scale full blow tests A large wall thickness is typical for many applications of DCI such as casings or transport and storage cask for radioactive materials. In order to investigate how the fracture mechanics characteristics of small specimens, which could even be determined within quality assurance procedures, correspond to the results of large specimens with component-like thicknesses, series of large scale full blow tests were performed. Since large specimens do not show R-curve behavior under dynamic loading at -40 °C, a test method was developed for determination of dynamic fracture toughness values with SE(B)140 specimens (length 1350 mm, width 280 mm, thickness 140 mm, a0/W = 0.5) at -40 °C by use of a servo-hydraulic impulse loading test system (max. 1 MN and 8 ms-1). As with small scale testing, different strain gage instrumentations (Fig. 9) including as per BS and ASTM were compared with respect to their force measurement capability with SE(B)140 specimens at impact conditions and a stress intensity rate of 5·104 MPa√ms-1. All strain gages were statically calibrated before the tests. Further details are reported for instance in [9]. Fig. 10 displays an example of force–time records and crack sensor signals. The test can roughly be assorted into 3 phases. During phase I, the rubber mat between striker and specimen is compressed and finally cut. After that, in phase II, the actual loading of the specimen takes place at a significantly higher but nearly constant loading rate compared to phase I. The stress intensity rate which is characteristic for the test is calculated as differential quotient in phase II. Phase III is characterized by unstable cleavage crack growth until final fracture of the specimen. In phase III, the force signals F1–F4 and F2–F3 are not considered for further analysis with respect to the underlying test goal to determine dynamic fracture toughness at initiation of unstable cleavage fracture. The good agreement of the signals F1–F4 and F2–F3 illustrates the high symmetry of
RkJQdWJsaXNoZXIy MjM0NDE=