Probing the Strength and Fracture Toughness of Hard Second Phases at the Micron Scale

It is known that the mechanical properties of composite materials and alloys are influenced by the intrinsic mechanical properties of their reinforcements or second phases. It is however challenging to determine such local properties, given their small size and irregular shape. Here we present two newly developed methods to probe the local strength and fracture toughness combining focused ion beam (FIB) milling with microtesting techniques and finite element (FE) simulation. To probe local strength Nextel™ 610 nanocrystalline alumina fibers are used as a test bench material. Deep-etching is carried out to expose the fibres over lengths of a few tens of μm from an aluminium matrix. FIB milling is then used on selected fibres to machine a wide notch oriented perpendicular to the axis of the fibre. A nanoindenter with a flat tip is then used to apply a compressive force along the notched fibre's axis, which produces a bending moment in the ligament opposite to the notch leading to failure. The measured response coupled with the FE model of each notched fibre is used to calculate the stress state in the ligament prior to failure. The strength distribution across a series of such tests is statistically analyzed and fractographic analysis with a SEM is performed. To probe local fracture toughness we produce thin chevron-notched cantilever beams by FIB micromachining. These beams are then loaded using a nanoindenter. The method is successfully applied once the procedure to generate sufficiently thin triangular ligaments with the FIB is mastered, which is crucial to ensure that instability occurs after some stable crack growth, leading to meaningful toughness measurements which are minimally influenced by FIB-induced defects. Test data are interpreted using compliance calibration curves determined by FE simulation of each beam. The method is demonstrated on small volumes of fused quartz and alumina fibres and produces reproducible results consistent with expected values.