Several significant developments in the interpretation and analysis of in situ stress measurements performed in weak rock using overcoring techniques are described. These developments include/the critical appraisal of overcoring strain data, - the determination of stress-strain transformation parameters for cross-anisotropic nonlinear rock, and the formulation of a numerical solution for cross-anisotropic rock. By suitable comparisons, it is shown that without these refinements meaningful results to the regional stress field could not have been obtained.
On decrit des progrès significatifs dans l'interpretation et l'analyse des mesures de contraintes in situ executees dans de la roche faible par les techniques de surcarottage. Ces progrès comprennent l'evaluation critique des donnees des contraintes de surcarottage, la determination des paramètres de transformation contrainte-deformation pour de la roche transversale-anisotrope non-lineaire, et la formulation d'une solution numerique pour de la roche transversale ani sotrope. En se basant sur des comparaisons convenables, on montre que sans ces perfectionnements de la technique on n'aurait pas pu obtenir des resultats valables pour le champ de contrainte regional.
Viele wichtige Entwicklungen in der Darstellung und Analyse von in-situ Spannungsmessungen in schwachem Felsen mit Überkernungstechnik werden beschrieben. Diese Technik wurde in Verbindung mit Spannungsmeßprogrammen auf drei besonderen Lagen in dem Projektbereich entwickelt. Diese Entwicklungen schließen die kritische Abschatzung von Überkernungsausdehnungsdaten, die Bestimmung von Spannungs-Ausdehnungsumwandlungskriterien fuer kreuz-anisotropische, nichtlineare Felsen, und die Formulierung einer numerischen Lösung fuer kreuz-anisotropischen Felsen ein. Durch entsprechende Vergleichungen wird bewiesen, daß bedeutungsvolle Ergebnisse fuer das regionale Spannungsfeld ohne diese Verfeinerungen nicht erlangt werden können.
Three separate programs of stress measurements were carried out at the site of the Drakensberg Pumped Storage Scheme in South Africa to determine the representative stress field around the major underground caverns and the associated waterways. The design requirements for the large span caverns in weak rock and the concrete-lined pressure tunnels required an accurate assessment of stress conditions. The scheme is located in weak, potentially erodable, horizontally bedded sandstones, siltstones, and mudstones (Figure 1). These difficult geological conditions also compounded the problem of achieving the required measurement standards. Three types of refinements in the interpretation and analysis of in situ stress measurements were made: critical appraisal of the overcoring strain data determination of the stress-strain transformation parameters realistic development of a numerical solution to the stress field for cross-anisotropic rock. The techniques used to improve and interpret the results from in situ stress measurements in weak, nonlinear cross-anisotropic rock, e.g., shales, sandstones, schists, are described in this paper.
Various methods are available for determining the absolute in situ stress field. Because of the weak, bedded nature of rock at Drakensberg, borehole overcore methods were considered the most appropriate. Limited additional testing using the hydrofracture techniques was also carried out. Although borehole strain cell devices are generally the most accurate instruments available for measuring in situ stresses, confidence intervals of the order of ±20 percent are generally the limit of accuracy obtainable, even with rock masses that can be described as linear elastic. During the early stages of the investigations, the need for innovations and refinements in the test technique to improve the quality of results for the nonlinear anisotropic rock conditions at Drakensberg was recognized. The detailed procedure for the measurement of in situ stresses using the overcoring principle is adequately described elsewhere (e.g., CSIR, 1973). An EX hole is drilled, the strain cell is inserted into this hole, bonded to the rock and then overcored. The steps involved in the evaluation of the overcoring strain data are the selection of the strain values, the determination of a modulus factor and the calculation of stresses. Corrections may be required for temperature variations, creep of the glue and rock, faulty gages, and stress redistributions caused by proximate openings. Modulus factors E and v are usually determined from simple tests on the EX core and the computation of stress is generally carried out assuming linear elastic isotropic material. Possible sources of major error in/this approach are:
nonlinear behavior of the rock (common, particularly in weak rock)
anisotropy of the rock material behavior
inhomogeneities such as bedding, inclusions and grains on the scale of the strain gauges
inhomogeneities on the scale of the overcore resulting in different sets of properties for each actual test
errors, malfunctions and idiosyncrasies of particular gages in the cell
experimental and interpretative misrepresentations of the stress and strain test data.
The basic problem reflected by these potential errors is the indirect nature of the overcoring method; i.e., stresses are not actually measured. Two categories of error sources are recognizable - those which are statistical and experimental in nature, arising from the very problem of representative measurement of the stress field and those which concern the proper representation of the constitutive behavior of the rock material.
