Spatial variability of rock strength by irregular lump point load testing of welded Topopah Spring tuff.

摘要

The research presented in this dissertation focuses on both the experimental investigation and theoretical analysis of the irregular lump point load test. The theoretical solutions in this study mainly emphasized both size and shape effect and stress analysis of irregular lump point load testing (ILPLT). Specimens tested were collected from four lithophysal units of the welded Topopah Spring Tuff at Yucca Mountain, Nevada: upper lithophysal zone (Tptrl, Tptul), middle nonlithophysal zone (Tptpmn), lower lithophysal zone (Tptpll), and lower middle nonlithophysal zone (Tptpln). Two types of laboratory tests have been conducted on samples from eight boreholes and from the Exploratory Studies Facility, conventional point load tests to obtain the lump point load strength index, and measurement of load and displacement. The former was conducted on 716 irregular lumps to test their deformation and strength, and the latter was performed on 396 lumps to study the stiffness of the rock.; Two applied multivariate techniques, Principal Component Analysis and Canonical Correlation Analysis, are employed to determine which one is more influential among the size variables. The better regression model for size effect is selected by regression analysis, and the relationships between the WD (cross-section), W (width) and D (distance between loading points) and the strength are studied for each borehole and stratigraphic unit, based on the best fit model. Logarithmic regression analysis is introduced for obtaining size-corrected point load strength index of irregular lumps. A t-test was performed on the data correlations to evaluate the statistical significance of the correlation. Because strength is very dependent on lump size, the effect on strength of size changes is explored by sensitivity analysis. The size and shape effects are discussed, based on the study by Brook (Brook, 1980). In addition, a shape effect, namely the edge effect, observed in the tests, is analyzed. Using the 2D FEM analysis, the detailed stress distribution induced by in point loading was explored. The fundamental relationship between strain energy and strained volume in point load testing is developed preliminarily in this dissertation to study the stiffness for the load and displacement in point load testing. The models of damage behavior in point load testing are summarized from 396 tests. A new damage constitutive process is proposed to represent the entire failure behavior. A size-corrected stiffness is proposed to obtain a relative unique stiffness value because stiffness of an elastic rock depends not only on the kind of rock but also on its geometry. (Abstract shortened by UMI.)