Additional points 1 73Chapter 5, we saw how the elastic strain compliance matrix can rep-
resent differing types of anisotropy, and especially transverse isotropy
and orthotropy. In Chapter 9, we discussed permeability and hydraulic
conductivity in terms of the same tensorial representation and in this
chapter we evaluated a directional hydraulic conductivity problem.
When a parameter does not directly incorporate a directional depend-
ency, e.g. the uniaxial compressive strength or the fracture frequency, it
is tempting to ignore any anisotropy that might be present. This tempta-
tion should be resisted: isotropy should only be assumed once there is
evidence that the degree of anisotropy is not significant.
The rock mass will always be 'Not Elastic' because there will al-
ways be some time-dependent component to the induced strains,
and loading/unloading curves will always exhibit hysteresis. So, even
though the elastic stress and strain distributions can often be helpful in
understanding the mechanics of a rock mass (e.g. in highly stressed deep
South African gold mines), it should always be remembered that the
rock mass response to engineering perturbations is Not Elastic. Estab-
lishing to what extent the inelastic rock mass response can be adequately
modelled for engineering purposes by an elastic model is part of the art
of engineering rock mechanics.