Study of Al2124-SiC nanocomposites by an improved statistical nanoindentation methodology

Abstract

In this work, we first review the previous work done on statistical nanoindentation by different researchers, highlighting the main problems that have been found and possible proposed solutions. In the second part, we study and report the statistical nanoindentation of three model samples, in the form of a soft Al2124 matrix embedded with hard SiC particles. Three different variants were selected: (1) 25% of SiC particles with 3 μm diameter; (2) 25% of SiC particles with 0.7 μm diameter; and (3) 17% of SiC particles with 0.3 μm diameter. We propose a novel heuristic wavelet technique to filter the measurement noise from the raw nanoindentation data as an attempt to obtain a more robust statistical nanoindentation methodology. Our results have shown that, when the nanoindentation data are filtered, it is not necessary to select a priori the number of peaks (phases) to be analyzed and, in some cases, a wide number of bin-sizes can be used without affecting the results. Finally, a finite element modeling have been used to analyze the response of the nanoindenter regarding the position of the hard particle. Our model shows that it is impossible to get the whole hardness value of the hard SiC particle by the statistical nanoindentation methodology.