Engineering Rock Mechanics

202 Rock mass classification

potential failure mechanism is not identified within the classification
system. Stress is not included in the RMR system; the intact strength of rock
is not included in the Q-system. Either of these parameters could be a
fundamental cause of failure in certain circumstances. Even more
severely, a shear or fault zone in the rock could exist which dominates the
potential failure mechanism of, say, a cavern or slope.
Because the perceived main governing parameters for rock engineering
have been included in the RMR and Q-systems, their use must provide
some overall guidance. However, the use of these systems as the sole
design tool cannot be supported on scientific grounds. For example, the fact
that the measured values of discontinuity frequency and RQD depend on
the direction of measurement has been clearly explained in Chapter 7, yet
this is not accounted for in either of the systems described. Similarly,
because the rock mass modulus depends on the discontinuity stiffnesses
to a large extent, the modulus is also anisotropic, yet the predictions of E
only provide a single (i.e. isotropic) value.
We feel, therefore, that despite its past contributions, the rock mass
classification approach will be supplemented by other methods in due
course, as the correct mechanisms are identified and modelled directly.
Moreover, it is an unnecessary restriction to use the same classification
parameters without reference to either the project or the site. For exam-
ple, in a hydroelectric scheme pressure tunnel, the in situ stress and proximity
of the tunnel to the ground surface are two of the most important
parameters. The RMR system cannot help under these circumstances. The
Q-system cannot be used for predicting E below a dam if the stratified nature
of the rock mass means that there is sigruficant anisotropy of stiffness.

12.6 Extensions to rock mass classification

techniques

Given our comments in the previous section, we believe that there are two

main ways in which the rock mass classification approach can be improved.

The first is a straightforward extension of the current systems, but

incorporating fuzzy mathematics to account for variations in the individual

component parameters. The second is to choose those parameters that are

most relevant to a particular engineering objective and hence the classifi-

cation systems for different projects would involve different constituent

parameters-using the RES (rock engineering systems) approach briefly

described in Chapter 14 (Hudson, 1992).

72.6. I Use of fuzzy muthemutics

Engineers may encounter problems in using the current rock mass clas-

sification systems because the inherent variability of rock masses is difficult

to take into account-for example, if mean discontinuity spacing vanes from

0.3 to 2.0 m, what value should be used in the system? By assigning a fuzzy

number to such parameters, and then using the techniques of fuzzy arith-

metic to combine the numbers, it is possible to generate a fuzzy number

representing the classification value. Such a number then embodies the

'most certain' classification value, together with information regarding its