13th International Conference on Fracture June 16–21, 2013, Beijing, China -6- independent and verified method. The errors were due to slight energy losses by hidden friction which may easily occur especially in larger drop towers. Therefore, the drop tower was constructively optimized and the tests were repeated, Fig. 6, right. The calculated and the measured displacement values are now nearly equal so that dynamic fracture mechanics tests could even be performed without having special equipment for displacement measurement available. But it must be considered an absolute essential prerequisite that this procedure can only be followed when the substance of Fig. 6 had been proven in advance. Further technical information on the camera and the validation of the corresponding measurement results is given at the end of chapter 2.2. Figure 6. Force vs. calculated and measured displacement, left: drop tower with friction losses, right: constructively optimized drop tower. The question of how to detect initiation of stable crack initiation in dynamic fracture mechanics tests has not yet been resolved satisfyingly. Reported techniques like magnetic emission, acoustic emission or near crack tip strain gages seem to work in special cases but cannot be seen as robust methods nor generally be applicable. Additionally, they do not seem to work with DCI reproducibly enough. Therefore, an alternative technique based on the detection of stable crack initiation by failure of strands of a crack propagation sensor attached to the ligament was investigated here, Fig. 7. Two different sensor positions - (1) first strand right in front of the initial crack tip on the specimen surface (Fig. 8, left) as well as (2) sensor at the initial crack tip on the specimen surface (Fig. 8, right) - were investigated. It was observed that in both cases the crack sensors provide discrete signal jumps in quasistatic and dynamic tests on SE(B)25 specimens at RT and -40 °C. Nevertheless, with position (1) there is a large scatter of detected initiation forces Fini of 0.5Fmax ≤ Fini ≤ 1.0Fmax. Material scatter may be a possible reason. But it seems more likely to Figure 7. Crack sensor in front of the crack tip of a SE(B)25-specimen.
RkJQdWJsaXNoZXIy MjM0NDE=