Statistical characterization of nanostructured materials from severe plastic deformation in machining

Endowing conventional microcrystalline materials with nanometer-scale grains at the surfaces can offer enhanced mechanical properties, including improved wear, fatigue, and friction properties, while simultaneously enabling useful functionalizations with regard to biocompatibility, osseointegration, electrochemical performance, etc. To inherit such multifunctional properties from the surface nanograined state, existing approaches often use coatings that are created through an array of secondary processing techniques (e.g., physical or chemical vapor deposition, surface mechanical attrition treatment, etc.). Obviating the need for such surface processing, recent empirical evidence has demonstrated the introduction of integral surface nanograin structures on bulk materials as a result of severe plastic deformation during machining-based processes. Building on these observations, if empirically driven, process–structure mappings can be developed, it may be possible to engineer enhanced nanoscale surface microstructures directly using machining processes while simultaneously incorporating them within existing computer-numeric-controlled manufacturing systems. Toward this end, this article provides a statistical characterization of nanograined metals created by severe plastic deformation in machining-based processes that maps machining conditions to the resulting microstructure, namely, the mean grain size. A specialized designed experiments approach is used to hypothesize and test a linear mixed-effects model of two important machining parameters. Unlike standard analysis approaches, the statistical dependence between subsets of experimental grain size observations is accounted for and it is shown that ignoring this inherent dependence can yield misleading results for the mean response function. The statistical model is applied to pure copper specimens to identify the factors that most significantly contribute to variability in the mean grain size and is shown to accurately predict the mean grain size under a few scenarios.