The apparent clustering observed in localized strain maps constructed from deformation-induced surface topographies generated in AA5754-O by three in-plane deformation modes were analyzed by two different methods. One used a conventional two-point autocorrelation function, while the other combined the eigenvalue spectra associated with each map and information theory. The results from the two-point analysis were inconclusive, implying that this technique lacks the sensitivity necessary to quantify the relationships between multi-point clustering and critical strain localization. The information theory-based approach revealed that, in each strain mode, the relative spectral entropy increased monotonically, attained a maximum, and then decreased sharply to the failure strain. This behavior was attributed to two competing processes one favoring the formation of clusters, and one that suppressed them. The crossover point is a clear signature of the onset of critical strain localization and, therefore, is a precursor to failure.
Citation: Philosophical Magazine
Pub Type: Journals
Metal forming, Autocorrelation function, Information theory, Entropy, Kullback-Leibler Distance, Eigenvalue spectra