330 Design and analysis of surface excavationssector method could be used to develop expressions for any of the
displacement or stress components that may be required through the use
of the expressions listed in Fig. 18.12.
Within the overall context of the design and analysis of surface
excavations, discussions of the 'first pass' studies of slope design and
foundation design have followed different approaches to slopes and
foundations. With slopes, kinematic feasibility was used and it was found
that a great deal of information could be obtained from the preliminary
overview analysis. With foundations, we considered how to estimate the
stresses and displacements beneath a non-uniformly loaded area. Again,
this type of analysis would immediately highlight areas for concern
and, if discontinuities were present, indicate the regons of potential
instability.
To extend both these approaches, one would need to consider the effects
of stress in very high slopes and the effects of discontinuities on the stress
distributions beneath foundations. Moreover, there is a range of site factors
that should be taken into account in more extended analyses. These include
overall rock anisotropy and inhomogeneity, local variation in material
properties, the effects of natural factors such as groundwater, rainfall,
seismic risk and the effects of engineering factors such as blasting. We have
chosen, therefore, to present in the next section techniques for consider-
ing these together, rather than extend each of the analyses separately to
account for the wide variety of deviations from the assumed CHILE
circumstances.
18.4 Techniques for incorporating variations in
rock and site factors into the analyses
There is a range of factors that can influence the instability of surface
excavations. The techniques presented so far do not explicitly allow these
factors to be taken into account, nor indeed can they all be taken into
account quantitatively in a direct way-because there will never be
complete knowledge of the rock mass. Also, considering the total energy
balance of a rock mass system via thermodynamics, it can be shown that
any excavation must result in an alteration of the rock properties (Hudson,
1992). This means that even with complete knowledge of the rock
properties before excavation as a result of a 'perfect' site investigation, the
post-construction properties would still not be known-because the pre-
construction properties will have been altered by the excavation process.
It can be predicted with certainty that some form of analysis that deals with
uncertainty will be required.
So far, the approach has been deterministic (i.e. estimated single values
of each of the parameters involved have been used in order to produce a
single result). This is because it is essential to understand the basic
mechanics before superimposing methods which account for variability.
There have been attempts during the development of rock mechanics to
substitute probabilistic approaches for an understanding of the mecha-
nisms, i.e. to relegate the mechanics and solve the problem by treating
some of the factors as random variables.