Environmental Engineering Reference
In-Depth Information
There are many natural slope conditions where the pore-
water pressures in the unsaturated soil portion of the slope
are strongly influenced by the infiltration of precipitation.
Modeling the unsaturated soil portion becomes important
in understanding changes in matric suction in a slope and
consequential changes in the computed factor of safety.
negative pore-water pressures are eliminated in a relatively
short time following a rainfall. Sweeney (1982) presented in
situ suction measurements from a slope instrumented with
tensiometers in Hong Kong. The slope was inclined at 60
with an average height of 30 m. From the slope surface to a
depth of 5-10 m the soil consisted of a completely weath-
ered rhyolite. The saturated coefficient of permeability for
in situ completely decomposed rhyolite was in the range
of 10 5
8.3.13 Ground Surface Moisture Flux Conditions
The solution for nonisothermal flows can be obtained by
solving the partial differential equations for the air phase,
the liquid water and water vapor phases, and the heat flow
equation simultaneously. The pore-air pressure changes can
sometimes be assumed to be negligible during a transient
process (Rahardjo, 1990; Wilson, 1990). As a result, only
water and heat flow differential equations need to be solved
simultaneously.
Solving the partial differential equations for nonisothermal
moisture flow at the ground surface is closely related to mod-
eling evaporation from the ground surface. Further informa-
tion on coupled heat and moisture flow was presented in
Chapter 6 where ground surface boundary conditions where
addressed.
10 7 m/s (GEO, 1993). The stratum immediately
below the completely decomposed rhyolite was a completely
to highly weathered rhyolite layer that was 5-10 m in thick-
ness. The underlying soil was a slightly weathered rhyolite
followed by bedrock at 20-30 m.
Suction measurements were made throughout the year
of 1980 and the following observations were made from
the results. First, matric suction showed a gradual reduc-
tion near ground surface during the rainy season. However,
the pore-water pressures remained negative even during the
long rainy season. Second, matric suctions in the 5-17 m
depth remained constant throughout the year. The ground-
water table at a depth of about 30 m rose and fell by 9 m
throughout the year. The results illustrate the significant
water storage capacity of the soil and its role in maintain-
ing matric suctions throughout most of the soil profile. The
matric suctions decreased at shallow depths into the slope
but remained almost constant throughout the intermediate
depths.
A second example where in situ matric suction measure-
ments were made involved a slope consisting of colluvium
material in Hong Kong (Anderson, 1983). The saturated
coefficient of permeability of the colluvium material ranged
from 10 4 to 10 7 m/s (GEO, 1993). The measured matric
suctions in the colluviums were relatively low prior to a rain-
storm and decreased to small values after heavy rainstorms.
This example showed that it is also possible for the matric
suctions to essentially disappear under certain conditions.
The above two examples show that there are certain condi-
tions under which matric suctions in the soil might disappear
during prolonged rainfall while there are other conditions
under which matric suctions are maintained throughout the
soil profile. There are rainfall and soil conditions under
which soil suction can be maintained in a soil profile. Situ-
ations where soil suction may be lost or maintained during
rainfall can be studied using saturated-unsaturated soil infil-
tration modeling. The physical processes associated with
infiltration have been the focus of numerous studies (Zhang
et al., 2004).
The wetting-front or wetting-band concept was introduced
by Lumb (1962) in relation to the investigation of slope
failures in Hong Kong. Figure 8.70 shows the variation of
degree of saturation with depth during rainfall. The soil is
assumed to become saturated near ground surface during
heavy rainfall and essentially saturated to a depth h under
ongoing infiltration. The wetting front was assumed to be
defined by a sharp separation between the initial conditions
8.4 CONDITIONS UNDER WHICH MATRIC
SUCTION CAN BE MAINTAINED
The effect of negative pore-water pressure on the stability
of slopes is sometimes ignored in geotechnical engineer-
ing practice. While this might be a reasonable assumption
in some situations, it does not provide the geotechnical
engineer with an in-depth understanding of potential fail-
ure mechanisms. This is particularly true when the water
table is deep and a reduction in matric suction near ground
surface becomes the trigger mechanism for slope instability.
It is often rationalized that matric suction can be ignored
in slope stability studies because there is a perception that
the infiltration of rainfall will produce a wetting front that
gradually moves downward and causes the disappearance of
matric suction throughout the soil profile. Consequently, it is
prematurely concluded that matric suction cannot be relied
upon to improve the stability of a slope in the long term.
The infiltration of water into a soil slope depends not only
on the rainfall intensity and duration but also on the saturated
coefficient of permeability of the near-ground-surface soil,
the permeability function, and the water storage capacity
of the soil. An unsaturated soil can have a greatly reduced
coefficient of permeability from that of a saturated soil and
also possesses an increased capacity for water storage.
Rainfall must be sustained over a considerable time period
and the rainfall intensity must approach the saturated coef-
ficient of permeability of the soil at the ground surface in
order to eliminate matric suction from the soil profile. In
situ suction measurements of negative pore-water pressures
have shown that in some cases matric suctions do not dis-
appear following intense rainstorms while in other cases the
 
Search WWH ::




Custom Search