wherefandare, respectively, the unit shear resistance which can be
mobilized and unit shear stress generated on the failure surface. The
analysis is applied to all conceivable failure surfaces, and the supposed
minimum factor of safety Fminis sought.
Stability is very sensitive to uw, which must be estimated from a
flownet or predicted on the basis of the pore pressure coefficients (Section
2.3.1) in the absence of field data. It is therefore sometimes more con-
venient in analysis to consider porewater pressures in terms of the dimen-
sionless pore pressure ratio, ru:
ruuw/ z (2.23)wherezis the depth below ground surface and zthe local vertical
geostatic stress.
Parameterrumay effectively equate to B ̄(equation (2.5)) in the case of
a saturated fill. The value of rucan often be taken as sensibly uniform within
a cohesive downstream shoulder, and equilibrium reservoir full values will
typically lie in the range 0.10–0.30. The initial porewater pressures gener-
ated in a cohesive fill develop as a result of the construction process itself,
i.e. overburden load and plant loads. Dissipation rate is a time-dependent
function of permeability and drainage path length (i.e. slope geometry).
Construction porewater pressures are partially dissipated prior to impound-
ing, after which they progressively stabilize to correspond to the advancing
seepage front and, ultimately, the reservoir full steady-state seepage or
other, varying, operational conditions (illustrated in Fig. 2.14). Abdul
Hussain, Hari Prasad and Kashyap (2004) presents the results of a study on
modelling pore water pressures and their influence on stability.
The form of the critical failure surface for Fminis controlled by
many factors, including soil type and the presence of discontinuities or
interfaces, e.g. between soft soil and rock. A number of failure surfaces
representative of different embankment and/or foundation situations
are illustrated schematically in Fig. 2.13. For most initial analyses
involving relatively homogeneous and uniform cohesive soils, two-
dimensional circular arc failure surfaces are assumed. The probable
locus of the centre of the critical circle in such cases, with ru0.3, can
be approximated by
zcHcot(0.62tan) (2.24a)and
ycHcot(0.6 tan) (2.24b)wherezcandycare coordinates with respect to the toe, measured positive
upwards and into the slope respectively, His the height and is the slope
angle.