A note on efktive stresses 159excavation as different discontinuities are traversed. We would predict that
for almost all rock tunnels there will be lengths where there will be little
inflow, and lengths where there could be high inflow. In general, we will
not be able to predict the specific local water inflow: consideration of the
diagrams in Fig. 9.8 makes this clear. Additionally, it will generally be
unknown how such a network is connected to the regional hydrogeological
regime. Thus, engineers know that during tunnel construction they should
have defensive strategies against major local high water inflows, but that
the precise location of these flows cannot be predicted.
Given the discussion in this chapter, would the reader include
permeability measurements in a site investigation for a particular project?
If so, should the necessary tests be conducted using boreholes? If so, how
are the results to be interpreted? Questions of scale effect and permeability
anisotropy will not be answered using a borehole strategy and, as a
consequence, many engineers have used full-scale prototype excavations
to determine permeability, e.g. the water inflow into a section of tunnel.
Much, of course, depends on the engineering objective.
Of all of the subjects presented in this book, perhaps permeability and
its corollaries are the prime examples of the fact that rock engineering is
an art. We think we understand the scientific principles, we understand
the difficulties of dealing with a natural rock mass, we may have large
resources, but there is no simple procedure for establishing ’the’ perme-
ability of a rock mass.
9.6 A note on effective stresses
In soil mechanics, wide use is made of the concept of effective stress, as
developed by Terzaghi (1963). We recall the explanation in Chapter 3 that
stress is a tensor, comprised of three normal and three shear components.
If fluid is present in the material matrix, the pressure, u, exerted by the fluid
will effectively reduce the normal components of stress in the stress tensor,
because the fluid has a hydrostatic pressure acting in all directions. This
hydrostatic pressure has no effect on the shear components of the stress
tensor. Thus, when the fluid is present, we can mod@ the stress tensor to
an effective stress tensor as follows:
OAX r,, r,, 2.4 0 0 Ox, - zf r*y
rzx 7, ‘5Tz 0 0[% ‘5, ryz]-[U ]=‘ ryx rzx ‘5,;u oz :: -u 1.
Here we have considered the simple case where the full hydrostatic
pressure has been subtracted; the reader should note that many
suggestions have been made for modifying the full value by coefficients to
account for the material microstructure and degree of saturation.
In Fig. 9.10(a), the water pressure is acting within the material micro-
structure, i.e. in the context of primary permeability, resulting in the
effective stress tensor given above. In Fig. 9.10@) we show, via the Mohr
circle diagram, the effect on strength of introducing water. Before water is
introduced, the stress condition is as in Case 1 in the diagram. When a