Direct observation of crack arrest after bridge notch failure: A strategy to increase statistics and reduce FIB-artifacts in micro-cantilever testing

[Display omitted] •The first experimental observation of bridge notch failure is presented.•The deviation of toughness from bridge-failure and the main cantilever failure is 7%.•Multiple toughness values can be gained from a single test improving the statistics.•Potential artifacts introduced by the Ga+ FIB pre-notches are reduced. Focused ion beam (FIB) milling has been widely used to prepare micron-sized specimens for micromechanical testing, however, at the same time, unavoidable artifacts originating from the Ga+ ion beam might alter the obtained mechanical properties. Using a bridge notch geometry, which can promote the formation of a sharp natural crack after bridge-failure rather than creating a comparably blunt FIB notch was proposed as a strategy to reduce FIB-induced artifacts more than a decade ago. Even though bridge-failure is widely assumed and predicted by finite element method (FEM) simulations, it has never been observed and quantified experimentally. This study presents the first experimental observation of bridge notch failure and crack arrest before the entire through-thickness main notch (after crack arrest) propagates, which is possible by designing thin bridges and using a stiff experimental setup with superior load resolution. Consequently, we obtained up to three corresponding fracture toughness values from one bending cantilever and significantly less scattered data. Using previously reported geometry correction factors calculated by FEM, the fracture toughness estimated from the bridge-failure was compared with the one from the failure of the main through-thickness notch in CrN/AlN multi-layered and CrN hard coatings.