Stochastic and Centrifuge Modelling of Jointed Rock. Volume 2. Centrifuge Modelling of Jointed Rock

Stochastic and Centrifuge Modelling of Jointed Rock. Volume 2. Centrifuge Modelling of Jointed Rock

Author:

Publisher:

Published: 1990

Total Pages: 282

ISBN-13:

DOWNLOAD EBOOK

The use of the centrifuge for geotechnical research has gained widespread acceptance in soil engineering, but to date, relatively few tests have been performed involving rock. One reason for the scarcity of rock mechanics experiments is the question as to whether a small scale model subject to an elevated gravitational acceleration level reasonably simulates the prototype condition involving discontinuities. The objective of this research was to find out if centrifuge modelling can be used for discontinua and to get further insight into arching mechanisms. To investigate this issue of scaling, a modelling-of-models approach was followed. This was done with a trapdoor model in the centrifuge. The model package was spun up gradually to a desired gravity level, at which the trapdoor was lowered and measurements of the force on the door with the corresponding displacement were taken. Different sizes of small- scale models and of trapdoors were used with various levels of acceleration. An initial series of experiments was conducted with granular materials, namely, New Jersey coarse sand and uniform sized glass beads. These experiments not only allowed us to make comparisons with preceding experiments on the arching mechanism under normal gravity; they also served as a benchmark for the experiments with jointed rock.


Stochastic and Centrifuge Modelling of Jointed Rock. Volume 1. Fracturing of Fractured Rock

Stochastic and Centrifuge Modelling of Jointed Rock. Volume 1. Fracturing of Fractured Rock

Author:

Publisher:

Published: 1990

Total Pages: 201

ISBN-13:

DOWNLOAD EBOOK

Rock mass failure can occur through the propagation of cracks emanating from and causing coalescence of pre-existing fractures. The objective of this research was to develop a mechanical model for fracture coalescence which can be used in conjunction with numerical methods to simulate rock mass failure. The study consisted of: (1) Monitoring growth in uniaxially compressed gypsum specimens with pre-existing fractures with various arrangements. Consistent coalescence behavior as a function of preexisting fracture geometry was noted and is documented; (2) Development of an analytical coalescence model based on correlations between computed stress/strain fields and observed fracturing in the experiments. Using this analytical model, satisfactory agreement was achieved between simulated and experimental coalescence behavior as a function of pre-existing fracture geometry.


Stochastic and Centrifuge Modelling of Jointed Rock. Volume 3. Stochastic and Topological Fracture Geometry Model

Stochastic and Centrifuge Modelling of Jointed Rock. Volume 3. Stochastic and Topological Fracture Geometry Model

Author:

Publisher:

Published: 1990

Total Pages: 307

ISBN-13:

DOWNLOAD EBOOK

A hierarchical model is developed to represent the geometry of fracture sets. The most important characteristics of this model are the possibility to model sequential fracture genesis as it usually occurs in nature, and to model clustering. This model uses the inhomogeneous or the doubly stochastic (Cox) point process model to represent the midpoints of the primary set. In a second step, the fracture trace lengths of the primary set are modelled. This is done using Maximum Likelihood Estimates (MLE) of the observed data. In a third step, the secondary set is considered; in particular the independence/dependence of the two sets. The correlation of location with trace length is considered with the line-kernel function methods and the nearest neighbor fiber distance method. The hierarchical model has been applied to two cases in which detailed fracture patterns have been observed, one with a single set and one with two sets. The validity of the model was checked by visual comparison between model prediction and mapped patterns and most importantly, by statistical tests such as the second-moment analysis and the Monte Carlo test. A satisfactory fit of the predicted pattern was obtained in both cases.