Stress Measurements - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 525

rock salt very long relaxation times are required, which makes the use of stiff inclu-
sion cells inefficient. For example, the in-situ stress state in rock salt could not be pre-
dicted in less than a measuring time of around 400 days (Pahl & Heusermann 1989,
Heusermann 1995).

16.4 Compensation Method

The compensation method is intended to measure the normal stress component par-
allel to and near an exposed rock surface. The procedure will be subsequently de-
scribed by means of the method developed by Rocha et al. (1966b) where small flat
jacks (SFJ) are used. Alternatively, LFJ (Rocha & Da Silva 1970) and rectangular flat
jacks cemented into drilled slots of overlapping boreholes (Section 15.4, Fig. 15.15)
may be used for compensation tests. A detailed description of this method is given in
DGEG (1984a) and ISRM (1987a).

Between two or more pairs of measuring pins, fixed in holes drilled into the exposed
rock wall, a slot is sawn by means of a diamond disk saw. The relative displacements of
the pins during and after slot cutting which result from unloading of the surrounding
rock mass normal to the slot are monitored. Then a flat jack is inserted into the slot
and pressurized until the displacements are compensated (Figs. 16.13 and 16.14). The
corresponding pressure p c is called the “cancellation pressure”.
The cancellation pressure is related to the stress component normal to the slot existing
at the rock wall prior to the slot cutting by a shape factor Kf f (Fig. 16.13, lower). Kf f
depends on the size and shape of the flat jack, as well as the locations of the measuring
pins, numerical values of which must be provided by the supplier.
An advantage of the compensation method is that the stress components in different
directions can be determined without making assumptions on the stress-strain behav-
ior of the rock mass and the corresponding parameters such as the elastic constants.
However, each measurement determines stress in one direction only and therefore a
minimum of six measurements in different directions are required to determine the
three-dimensional stress state in the vicinity of an exposed rock surface. In addition, for
an underground opening, the stress state near the excavation contour does not corre-
spond to the in-situ stress state. Therefore the evaluated stress state must be interpreted
by means of subsequent analytical or numerical analyses.
In the simple case of a bored adit with circular cross-section located in homogeneous,
elastic, isotropic rock mass, the three-dimensional in-situ stress state can be deter-
mined from the results of compensation tests by a closed-form analytic solution (Na-
tau 1972). This method has been applied in practice to a water transmission tunnel
located in the Harz mountain range (Balthasar & Wenz 1983). Furthermore, analytic
solutions were presented for adits with square and elliptical cross-sections (Pinto &
Charrua-Graca 1983).