Microstructural shape characterisation of grains, cracks and crystal phases is important for linking microstructure to properties and performance in engineering materials. However, significant spatial and angular distortions are usually present in electron backscattered diffraction (EBSD) maps, a widely-used scanning electron microscopy (SEM) technique for microstructural characterisation. Multi-phase materials such as WC-Co hardmetals have additional artefacts caused by different atomic number densities between phases, where the asymmetric spatial resolution and signal intensity across a phase boundary can bias the measured phase boundary position.

I will demonstrate how to ensure accurate EBSD grain size and shape measurements, robust shape descriptor metrics, and consistent statistical treatment of these metrics in WC-Co hardmetal composites used to make cutting tools. This is required to ensure fair comparisons of material grades for industrial quality assurance, and when using them to inform material or process models.

I will also show how EBSD spatial distortions can be corrected using ‘TrueEBSD’, a correlative imaging based software method, and combined with grain boundary analysis tools in MTEX, an open-source MATLAB toolbox for handling EBSD maps, but modified for analysis of non-crystallographic shapes such as crack paths. This is used to characterise the statistical distributions of phase and grain orientations around mm-scale fatigue crack paths in WC-Co hardmetals, and then related to the mechanical behaviour of hardmetals with different carbon contents.