On acquiring data for large-scale crack bridging at high strain rates

Abstract

Numerical simulations are used to design test geometries and loading histories that are suitable for probing the large-scale bridging effects of through-thickness reinforcement that is shearing at high strain rates. The bridging effects are represented by a cohesive law and tests are sought that will determine any rate dependence in its parameters. The end-notched flexure test is studied, because it allows easy application of time-dependent loading and has proven to be an information-rich test in the quasi-static case. However, dynamic conditions greatly complicate fracture behavior, with possible regimes of hammering and multiple cracking, which should be avoided when maximum information is sought. Information content is addressed by focusing on regimes within the full computed solution space where crack growth is approximately steady state and the information content of experiments can be most easily assessed. Numerical results show that the hypothetical rate dependence in the cohesive law causes strong and measurable changes in the regime of steady-state behavior, if the tests are properly selected to vary the crack sliding speed. The estimates of information content are conservative, because the information available from all possible tests of specimens designed by analysis of the steady-state regime will necessarily exceed the information deduced by analyzing the steady-state regime alone.