Laboratory Tests - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 438

At least five (ISRM 1975a) or three (DGEG 1988) tests on specimens from the same
test location should be conducted with different normal stress levels. If only the residual
stress is of interest one specimen can also be consolidated and sheared off at different
normal stress levels (DGEG 1988).
Normally the so-called “constant normal load (CNL)” test mode is carried out in which
the normal stress

σ n is held constant during shearing. In this test mode F N needs to be
regulated during the test since, according to (14.48), σ n tends to increase due to the

reduction of the contact area A of the shear plane with increasing shear displacement.

In the “constant normal stiffness (CNS)” test mode the discontinuity is sheared off at
constant normal stiffness condition, that is, the applied normal force F N is increased
when dilatancy

δ n < 0 takes place:

F N = F N0 - k N ⋅ δ n , (14.50)

where F N0 is the initial normal force before shearing takes place and k N is referred to
as normal stiffness of the discontinuity. This test mode requires more refined testing
devices (Johnston et al. 1987, Brahim & Gerard 1989, Onishi & Dharmaratne 1990,
Indrahatna et al. 1999, Jiang et al. 2001, Jiang et al. 2004).
The multiple failure state test mode, described in Section 14.4.2 in connection with tri-
axial compression tests, allows the determination of both the peak and residual shear
strength in one single direct shear test provided that a sufficiently stiff test apparatus is
used (DGGT 1987). In the example illustrated in Fig. 14.20, the envelopes for peak and
residual strength can be described by linear relationships between

σ n and τ res (Wittke

1990).

After the test, the shear plane should be exposed and described. Photographs may also
be useful (ISRM 1975a).
Portable shear boxes reported by Locher (1968), Locher & Rieder (1970) and Ross-
Brown & Walton (1975) are easy and fast to handle and allow irregularly shaped spec-
imens containing a discontinuity to be investigated in a direct shear test with low cost.
However, the results of such tests cannot be considered as reliable. These testing de-
vices do not allow the normal force to be controlled with respect to constant normal
stress during shearing. In addition, it is generally difficult to mount the specimen into
the shear box and the discontinuity may easily be damaged during the preparation of
the specimen (Wittke 1990).
Conventional triaxial compression tests on intact rock specimens containing a disconti-
nuity are described in Lane & Heck (1964), Rosengren (1968) and Wolters (1970). The
shortcomings of these tests are evident. The investigated specimens normally have di-
ameters ranging from 3 to 5 cm, resulting in very small shear planes, and the boundary
conditions allow only very small shear displacements.