Geology Reference
In-Depth Information
6.10.6.1 Pseudo-static load analysis
One approach is simply to include a horizontal inertial load into the
analysis (some authors argue for an inclined force but it really makes
little difference considering the inexact nature of the method) and to
determine whether or not the FoS reduces below 1.0. The problem
with this approach is that if one includes the peak predicted particle
acceleration (say from equation 6.1) then very often the slope will
be shown to fail, whereas in reality the permanent displacement
would be negligible because of the extremely short time that acceleration
would be acting. As con
rmation, many vertical slopes in quarries are
acted on by accelerations approaching or exceeding 1g during produc-
tion blasting, but landslides due to blasting are very rare. Engineers
therefore often choose to use some arbitrarily reduced acceleration,
such as a nominal 0.1g, as a pseudo-static force in the stability analysis
to check that the slope (or dam) has some degree of resistance to
horizontal loading, but this is clearly rather unsatisfactory.
6.10.6.2 Displacement analysis
As discussed earlier, given a predicted acceleration against time record, it
is straightforward to calculate the likely displacement that might be
caused in a slope during an earthquake, and there are options to do so
in software such as SLOPE/W. Those displacements will always be small,
however, no matter how large the earthquake, and what matters more is
the residual state of the slope after the earthquake
is there a situation
where the ground is strain softening or is it in unstable equilibrium?
These are considerations for the engineer, who must decide whether
additional reinforcement might be necessary or other protective mea-
sures such as nets and barriers. Other software such as FLAC and UDEC
(Itasca) can be used to study the seismic susceptibility of slopes. These
being time-stepping software, the mode of failure can be identi
-
ed,
expressed visually and perhaps as a movie. It might, for example, be
possible to test the potential failure mechanism of soil nails during an
earthquake, each nail modelled speci
cally. That said, as ever, the models
can only be as good as the input data and results will only be indicative.
Blasting causes noise, ground vibrations, air overpressure and
yrock.
All of these can be controlled
generally by using less or different types
of explosive and limiting the number of charged drillholes that are
detonated at the same time. In particular, using millisecond delays
between lines of drillholes will reduce the vibration level considerably.
Details are given in Dowding (1985) and many other publications.
-
