Statistical analysis of shear cracks on rock surfaces

A set of 3873 cracks on exposed granite rock surfaces are analyzed in order to investigate possible fracture mechanisms. The fracture patterns are compared with the Mohr-Coulomb and the Roscoe fracture models, which can be combined into a single fracture scheme. A third model for comparison is based on interacting `penny-shaped' micro cracks introduced by Healy et al. [Nature <Emphasis Type="Bold">439, 64 (2006)]. The former models predict a bimodal fracture angle distribution, with two narrow peaks separated by 60<Superscript>○</Superscript>-90<Superscript>○</Superscript> symmetrically on both sides of the direction of the largest principal stress, while the latter predicts a single broader peak in the same direction with standard deviation in the range 15<Superscript>○</Superscript>-20<Superscript>○</Superscript>. The crack length distributions seem consistent with numerical simulation, whereas the fracture patterns are Euclidean rather than fractal. The statistical analyses indicate that none of the models fully describe the fracture patterns. It seems that natural shear fractures easily become a complex combination of different fracture mechanisms. Copyright EDP Sciences/Società Italiana di Fisica/Springer-Verlag 2007