Laboratory Tests - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 422

The procedure for sampling and testing large-scale specimens by means of triaxial com-
pression tests was fi rst applied to cohesive soil by Wichter & Gudehus (1976). Later this
method was applied to cylindrical rock specimens to determine the deformability and
strength of jointed rock (DGEG 1979c, Natau et al. 1983, ISRM 1989b). However, var-
ious problems, related to undisturbed sampling and specimen preparation, arise when
applying this testing method. In addition, the infl uence of discontinuities leading to
anisotropic deformability and strength cannot be adequately considered by this method
since the available measuring devices do not allow the evaluating of the test results with
respect to anisotropy.

14.4.3 Brazilian Test
The Brazilian test, also referred to as the “Brazil test” (ISRM 1978c) or the “indirect
tensile test” (DGEG 1985b, DGGT 2008), is intended to estimate the uniaxial tensile
strength of intact rock indirectly by the failure stress of a diametrically loaded cylindri-
cal or disk-shaped specimen. The specimen prepared from a drill core sample should
have a diameter of D ≥ 50 mm and a length-to-diameter ratio of 0.3 ≤ L/D ≤ 1.0 (ISRM

1978c, DGEG 1985b, DGGT 2008).
The natural water content of the drill core sample should be preserved until preparation
of the specimen (ISRM 1978c).
The preparation of specimens should be carried out by means of drilling, cutting, turn-
ing or grinding. The skin surface of the specimen should be smooth and even, with a
maximum tolerance of 0.1 mm. The end faces of the specimen should be coplanar

and should form a right angle with the specimen's axis (DGEG 1985b, DGGT 2008).
As test apparatus, a testing machine with a force measuring device according to accept-
ed national requirements should be used (DGGT 2008). The load F is transmitted to
the specimen by an upper and a lower steel jaw with curved loading surfaces, the radius
of curvature

±

of which should be 1.4 to 1.6 times the specimen's radius R (Fig. 14.11,
upper left). If the specimen has a low deformability, the load may also be applied by
opposed load distribution strips (Fig. 14.11, upper right).

ρ

The specimen is inserted into the test apparatus and the strip load F is applied to the
specimen at constant loading rate until failure takes place and the failure load is record-
ed. The loading rate should be selected in such a way that the test lasts between 2 and 5
minutes (DGGT 2008).
It is assumed that the angle 2

of the loaded surface is small, that is, F is a line load
(Fig. 14.11, lower left). If linear elastic isotropic stress-strain behavior and a plane state
of stress are assumed, Fig. 14.14 (lower right) shows the horizontal and vertical normal
stress components

α

σ y , respectively, along the vertical section through the spec-
imen's axis. Accordingly, high compressive stresses arise near the loaded surfaces and
an almost constant transverse splitting tension stress

σ x and

σ x acts over the majority of the

specimen's cross-section. The stress at failure σ xf can be set equal to the tensile strength

σ tIR of the specimen:

(14.30)