Geology Reference
In-Depth Information
Figure 3.31
Columnar joints
through rhyolite,
High Island
Reservoir, Hong
Kong. Evidently,
the traces are not
fully developed as
persistent
mechanical
fractures, otherwise
the undercut
columns would not
remain in place
(gravity being what
it is).
Joint sets are generally differentiated for rock mechanics analysis accord-
ing to their geometries. The basics of stereographic projection represen-
tation of rock joints is dealt with in
Chapter 4
and explained in detail in
Wyllie & Mah (2004). If the poles representing discontinuities plot
closely on a stereographic projection, then they are considered to com-
prise the same set. For example, when one de
nes a set in the rock mass
modelling programme, UDEC (
Chapter 6),
this is done by inputting a
mean dip and dip direction (perhaps plus or minus 5 degrees). Programs
such as Dips from Rocscience
(Chapter 6)
can be used to identify sets
statistically, according to various methods such as the Fisher distribu-
tion. This is a useful tool but a number of things must be borne in mind:
1. The original data might be biased or partial. Some joint sets may
not be fully developed at the point of observation or might be
misrepresented in terms of population, because of the geometry
of the exposure.
2. Joints of similar geometry might include different sets geologically
(in terms of time of formation) which have characteristics that are
quite different despite their parallel orientation.
3. Important but rare geological features, such as a fault, might be
overshadowed by the rest of the data and even removed from
consideration by statistical manipulation, as discussed by
Hencher (1985).
4. A better approach is
first to try to interpret the distribution and
nature of joints in terms of a model for the geological history at
the site (Rawnsley et al., 1990; Hencher &Knipe, 2007). Data such
as surface textures and mineral coating can also be very helpful
for differentiating between joint sets, especially for high-level
interpretations such as for nuclear waste studies (e.g. Bridges, 1990).
