Dynamic crack initiation in 3PB ductile steel specimens

Conclusions

The design of the experimental set-up has exposed several advantages but also one drawback: the U-shaped hammer with fixed impact heads gives uncertain boundary conditions at the impact points and thus an uncertain stress-strain state at the crack tip. The set-up will therefore be redesigned in order to avoid this uncertainty.

The advantage with the U-shaped hammer is that the crack growth area of the specimen will only get a minor translation and therefore close up photography is possible. The high speed photos have shown to be an excellent way to evaluate the experiments. The CMOD can be measured with an accuracy within about 0.1 mm. Necking (which follows after crack growth in the interior of the specimen) and visible crack growth can be fairly well determined from the photos.

The evaluation of the experiments is made from a combination of measurements and calculations. Excessive nonlinear dynamic FEM calculations is a necessary part of the investigation.

The results indicate that the J-integral for onset of crack growth is about the same for the two lower impact velocities 15 m/sec and 30 m/sec and it is decreased when the impact velocity is increased to 45 m/sec.