Field Tests - Rock Mechanics Based on an Anisotropic Jointed Rock Model - page 473

With four LFJs in line, a rock surface of ca. 4.5 m can be loaded. Also LFJ tests can
be performed inside a relatively undisturbed zone in the rock mass and the applicable
pressure up to 20 MPa is relatively high. The small range of displacement measure-
ments of less than 10 mm, and diffi culties when sawing slots in cases of high in situ
stresses and unfavorable geological conditions, may be regarded as disadvantageous
(ISRM 1986).
The estimation of the in-situ stress state using fl at jacks referred to as the “compensa-
tion method” will be dealt with in Section 16.4.

15.4.2 Transversely Isotropic Rock Mass
LFJ tests in transversely isotropic rock can be evaluated using a method proposed in
Kiehl & Wittke (1983), also described in Wittke (1990), which allows the determina-
tion of the elastic constants E 1 , E 2 and G 2 . Similarly to the procedure described for
dilatometer tests, in this method the results of LFJ tests carried out with different
orientations with respect to the isotropic plane must be combined. The orientation of
the slots and LFJs, with respect to the isotropic plane, must be arranged as illustrat-
ed in Fig. 15.13 (upper). In test arrangements LFJ1 and LFJ2, the slots are ordered
parallel and perpendicular to the isotropic plane. In test arrangement LFJ3, the slots
are inclined at an angle to the isotropic plane. This method is limited to LFJ tests in

which two LFJs are arranged in line and the angle

ϕ

= 60°,
respectively. In all three LFJ arrangements the plane between both LFJs normal to
the slots represents a plane of symmetry.

ϕ amounts to ϕ = 30° and ϕ

Test arrangements LFJ1 and LFJ2 can be realized, for example, in an exploration adit
that is driven normal to the strike direction of the isotropic plane. Test arrangement
LFJ3 requires a rock wall that is oriented parallel to the strike direction of the isotropic
plane. This can be accomplished, for example, by a cross-cut of the adit mentioned
above (Kiehl & Wittke 1983).
As was found from the results of parametric studies by means of three-dimensional
fi nite element analyses in which the tests LFJ1, LFJ2 and LFJ3 were simulated and the
elastic constants were varied (Kiehl & Wittke 1983) the slot widening at each deforme-
ter for test arrangements LFJ1 to LFJ3 can be expressed as follows (Fig. 15.13, upper)

(15.18)

where K is the same quantity as in (15.17) and Ci i are functions of m = G 2 /E 2 , n = E 1 /E 2
as well as ν 1 and ν 2 . However, the Poisson's ratios ν 1 and ν 2 have only a small infl uence

on the results.