Understanding the deformation mechanisms of heterogeneous materials requires information about the microstructure as well as knowledge of the mechanical properties of the different phases. While for homogeneous materials nanoindentation is nowadays routinely and successfully used to determine the mechanical properties, the characterization of heterogeneous materials is still challenging. For homogeneous samples, a mean value and a standard deviation of the modulus or the hardness can be determined from typically 10-30 indentations. When it comes to materials containing microscale heterogeneities, determining the materials properties and the properties of each material phase requires a much larger number of indents and, thus, the use of statistical analysis methods. Such methods have proven useful to characterize the elastic behavior of cement-based materials (Constantinides et al. 2006, Ulm et al. 2007, Lura et al. 2011).
In this study, we investigated the constraints and limitations of the statistical nanoindentation technique. This includes the influence of the indentation depth on the determination of the phase properties and the number of phases during deconvolution, or the effect of choosing the Cumulative Distribution Function (CDF) over the Probability Density Function (PDF). Furthermore, the applicability of the method to the different classes of materials is addressed. We apply a probabilistic approach to analyze distributions of mechanical properties determined from grid indentations to different depths. The analysis method is applied to composites consisting of Cu and Cr phases as well as to composites of unknown composition, such as industrial mortar and polymer-modified cements.