Dealing with imperfection: quantifying potential length scale artefacts from nominally spherical indenter probes

Instrumented nanoindenters are commonly employed to extract elastic, plastic or time-dependent mechanical properties of the indented material surface. In several important cases, accurate determination of the indenter probe radii is essential for the proper analytical interpretation of the experimental response, and it cannot be circumvented by an experimentally determined expression for the contact area as a function of depth. Current approaches quantify the indenter probe radii via inference from a series of indents on a material with known elastic modulus (e. g., fused quartz) or through the fitting of two-dimensional projected images acquired via atomic force microscopy (AFM) or scanning electron microscopy (SEM) images. Here, we propose a more robust methodology, based on concepts of differential geometry, for the accurate determination of three-dimensional indenter probe geometry. The methodology is presented and demonstrated for four conospherical indenters with probe radii of the order of 1-10 mu m. The deviation of extracted radii with manufacturer specifications is emphasized and the limits of spherical approximations are presented. All four probes deviate from the assumed spherical geometry, such that the effective radii are not independent of distance from the probe apex. Significant errors in interpretation of material behaviour will result if this deviation is unaccounted for during the analysis of indentation load-depth responses obtained from material surfaces of interest, including observation of an artificial length scale that could be misinterpreted as an effect attributable to material length scales less than tens of nanometres in size or extent.