Geology Reference
In-Depth Information
contraction during undrained shearing and, as such, cannot model, for
example, soft clays (or dense sands) under undrained conditions. If you
use a Mohr-Coulomb model for undrained conditions, then you simply
use the undrained strength to control failure conditions and not the
effective strength parameters. The Nicoll Highway Committee of
Inquiry Report on the collapse includes a well-written section on the
problem with the Mohr-Coulomb model (Magnus
et al
., 2005).
The
finite difference program FLAC is probably the second most
generally used software for geotechnical design.
Until recently, the programwas quite daunting, requiring individual
commands to be typed in, but recent versions have a graphic interface,
which makes things easier. As for PLAXIS and other sophisticated
programs, a great deal of knowledge and understanding is needed if
reasonable results are to be achieved. For example, the model must
rst
be set up with proper boundary conditions and brought to equilibrium
as natural ground before any engineering works such as excavation are
simulated. FLAC progressively calculates and checks solutions.
Intermediate stages can be calculated, saved and expressed graphically
as a movie which can illustrate how strains are developing with time.
FLAC, like its sister programUDEC, can cope with large displacements,
more so than typical FE analyses. FLAC is used mainly for soil and rock
that can be characterised as continua. UDEC is used for fractured rock
and each fracture or set of fractures can be speci
ed individually in
terms of geometry and engineering parameters. Both UDEC and FLAC
can be used for foundation design, tunnels and slopes.
Other commonly used software include the suite produced by
Rocscience, such as Phase
2
, and their use is discussed in detail in
Hoek
et al
.
(1995) and at
http://www.rocscience.com/education/
Many authors, experienced in the development and use of soft-
ware, have recommended that sophisticated software should be
used in an investigatory way, using many simple models to
check sensitivity to assumptions rather than trying to prepare a
single complex model in an attempt to simulate all aspects of a
situation at the same time (Star
eld & Cundall, 1988). Swannell
& Hencher (1999) discuss the use of software speci
cally for
cavern design.
Many suites of software have been developed for particular purposes.
SLOPE/W and SLIDE, for example, are commonly used for routine
design of slopes. The software calculates stability employing the
method of slices, as discussed in section 6.6.2, and gives instant solu-
tions for FoS for a wide range of potential slip surfaces, the broad
geometries of which are speci
ed by the operator. Controlling factors
