This paper introduces a geostatistical approach for characterising the morphological and spatial variability of rock inclusions in bimrocks and for attempting their direct mechanical characterization through an indirect description deduced by image analysis of outcrop images. The study was carried out on the Shale–Limestone Chaotic Complex (SLCC) bimrock, which is affected by several slope stability problems in the disused Santa Barbara open-pit mine (Tuscany, Italy). Due to its complex heterogeneity, the mechanical characterization of SLCC was based on a limited number of nonconventional shear tests (BimTests). Due to the relevant influence of rock inclusions on bimrock strength, the study focused on identifying the possible image parameters characterising content and variability of rock inclusions in each sampled picture. A two-step analysis was carried out for: a) identification of an image parameter (centimetric scale) that is able to grasp some property of rock fragment shape and size spatial distribution; b) verification of the correlation at the outcrop scale (multimetric scale) between the bimrock strength parameters and the new regionalised variables (ReV) describing image properties related to rock fragment distribution. A geostatistical approach was adopted for both steps, but applied to two different problems, at two different scales, with different variables. An extensive sampling of outcrop pictures was carried out and a variogram analysis was performed on the “rock fragment indicator variable”, IB (x, y), defined at the image scale, and obtained through the digital segmentation of the collected pictures. A very good link was observed between the geostatistical model parameters and the content, size and spatial variability of rock fragments shown in every picture. The values of the total sill of the block indicator variograms were then assumed as a new ReV, Ctot (x, y), defined all over the SLLC outcrop. A cross-covariance was computed between the variogram total sill and the friction angle, φ (x, y) and cohesion c (x, y), at the scale of the whole outcrop area and a preliminary analysis showed a likely correlation.
The contribution of geostatistics to the characterisation of some bimrock properties
COLI, NICCOLO';BERRY, PAOLO;BOLDINI, DANIELA;BRUNO, ROBERTO
2012
Abstract
This paper introduces a geostatistical approach for characterising the morphological and spatial variability of rock inclusions in bimrocks and for attempting their direct mechanical characterization through an indirect description deduced by image analysis of outcrop images. The study was carried out on the Shale–Limestone Chaotic Complex (SLCC) bimrock, which is affected by several slope stability problems in the disused Santa Barbara open-pit mine (Tuscany, Italy). Due to its complex heterogeneity, the mechanical characterization of SLCC was based on a limited number of nonconventional shear tests (BimTests). Due to the relevant influence of rock inclusions on bimrock strength, the study focused on identifying the possible image parameters characterising content and variability of rock inclusions in each sampled picture. A two-step analysis was carried out for: a) identification of an image parameter (centimetric scale) that is able to grasp some property of rock fragment shape and size spatial distribution; b) verification of the correlation at the outcrop scale (multimetric scale) between the bimrock strength parameters and the new regionalised variables (ReV) describing image properties related to rock fragment distribution. A geostatistical approach was adopted for both steps, but applied to two different problems, at two different scales, with different variables. An extensive sampling of outcrop pictures was carried out and a variogram analysis was performed on the “rock fragment indicator variable”, IB (x, y), defined at the image scale, and obtained through the digital segmentation of the collected pictures. A very good link was observed between the geostatistical model parameters and the content, size and spatial variability of rock fragments shown in every picture. The values of the total sill of the block indicator variograms were then assumed as a new ReV, Ctot (x, y), defined all over the SLLC outcrop. A cross-covariance was computed between the variogram total sill and the friction angle, φ (x, y) and cohesion c (x, y), at the scale of the whole outcrop area and a preliminary analysis showed a likely correlation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
