An experimental study that contributes to the understanding of the relationships between petrographic features and engineering properties of igneous rocks is conducted. To this end, a wide range of igneous rocks were tested for their engineering properties including abrasivity (Cerchar abrasivity index), mechanical (point load strength index Is (50)), basic physical (dry density and porosity) and dynamic (P-wave velocity) characteristics. Moreover, a semi-automatic method has been developed to analyze petrographic data that relies on digital image acquisition from representative parts of representative thin sections of samples, semi-automatic image segmentation and image analysis. The method quantifies 18 petrographic features including size descrip-tors (area, perimeter, equivalent circular diameter, minimum Feret’s diameter, maximum Feret's diameter), shape descriptors (elongation, orientation, eccentricity, compactness, rectangularity, solidity, convexity), rock fabric coefficients (index of interlocking, index of grain size homogeneity, texture coefficient) and mineralogical indices (saturation index, feldspathic index, colouration index). The Pearson's correlation coefficient and multivariate regression analysis are employed to analyze the relationships between extracted petrographic features and engineering properties. In general, fine-grained and basic igneous rocks compared to the acidic and coarse-grained ones possess higher engineering quality and lower abrasiveness potential. The results imply that mineralogical composition tends to be more important than rock fabric characteristics in determining the engineering properties of igneous rocks. Furthermore, among rock fabric characteristics, size descriptors have significant influ-ence on the engineering properties. Overall, it was found that mineralogical composition and rock fabric characteristics provide a suitable complement to reliably predict engineering properties of igneous rocks.

Keywords