Stress Measurements - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 510

Local stress measurements only allow pointwise determination of the in-situ stress
state. Therefore methods have been developed to evaluate the large-scale in-situ stress
state and its spatial variation on the basis of back analyses of stress measurements and
monitoring the results by means of numerical methods. These are dealt with in Section
16.6 and illustrated by means of case histories in Section 16.7.
An early idea of the in-situ stress state helps to check the results of stress measurements
for plausibility. Therefore all available information on rock stress, including an under-
standing of the geological setting as well as reports and papers on stress measurements
previously made from the project region, should be gathered (ISRM 2003). Particularly,
information on regional geologic history and local tectonics is useful (Chapter 9). Also
consultation of the World Stress Map (WSM) may be worthwhile (Section 16.8).
In any case, the in-situ stress state evaluated by means of stress measurements carried
out ahead of construction should be verified by monitoring and back analysis during
construction.

16.2 Stress Relief

16.2.1 Triaxial Cells
Triaxial cell measurements are carried out to determine the in-situ stress tensor around
a borehole. A detailed description of various types of triaxial cells and their use can be
found in ISRM (1987a) and ISRM (2003).
The principle of these tests, which is practically the same for all types of cells, is illustrat-
ed in Fig. 16.1. From the flattened end of an exploration drilling (main borehole) with
a diameter of at least 86 mm, a pilot hole with smaller diameter is carried out coaxially,
in which the triaxial cell is inserted. Subsequently, the pilot hole is overcored with the
diameter of the main borehole forming a hollow cylindrical sample (overcore). The
strains due to stress relief are measured using strain gauges. Then the overcore includ-
ing the cell is extracted from the surrounding rock and recovered, with the core barrel,
to the head of the borehole. Some types of triaxial cells are recoverable.

The differences of strains measured after and before overcoring are related to the undis-
turbed, three-dimensional stress state existing in the rock mass before drilling of the bore-
hole. Test evaluation is normally based on elastic stress-strain behavior of the rock mass.
The elastic constants required for stress determination from measured strains are deter-
mined either by means of on-site biaxial compression tests on the overcore or by uniax-
ial compression tests on cylindrical specimens taken from the overcore (Fig. 16.2). Both
testing methods allow the elastic constants of both isotropic and transversely isotropic
rock to be determined (Wittke 1990, Nunes 2002, Section 14.4.1).
A cell where strain gauge rosettes, each comprising three gauges, are glued or cemented
onto the wall of the pilot hole was first developed at the African Council for Scientific
and Industrial Research (CSIR) in South Africa (Leeman & Hayes 1966, Leeman 1968,
Leeman 1971, ISRM 1987a). This CSRI cell does not allow continuous data logging.
An improved cell of this type is the Borre (SSPB) probe (Sjöberg & Klasson 2003,
ISRM 2003), which has a built-in data logger for continuous data logging. Modified