Environmental Engineering Reference
In-Depth Information
and a is the Henkel pore pressure parameter. The equivalent Skempton A from Henkel's a
parameter for triaxial compression is:
1
3
2
3
A
a
(11.13)
For triaxial extension conditions:
2
3
a
2
A
3
(11.14)
The pore pressure parameters A and B can be estimated by laboratory tests on the soil
to be used in the embankment. “B” is determined by placing soil compacted to the
required water content and density ratio in a triaxial cell and observing the change in pore
pressure
u for changes in cell pressure under undrained conditions. Under these condi-
tions
1
3
and the pore pressure equation becomes:
u [
( )]
B
(11.15)
3
3
This relationship will not be linear and must be determined over a range of cell pres-
sures. B will be larger for higher cell pressures than for low cell pressures.
To determine A, the soil is placed in the triaxial cell under undrained conditions and
sheared. Knowing B from the earlier testing and observing
3
, A can be
determined from the pore pressure equation. As explained above, in most cases this will
be done under K
o
conditions. Head (1985) shows a suitable test set up and lateral strain
indicators needed to control the test using conventional triaxial equipment.
More accurate estimation of the pore pressure parameters can be achieved by more closely
following the stress history of the soil in the embankment using a Bishop-Wesley triaxial cell.
To apply this method one must determine:
u,
1
and
- What the stresses in the embankment will be;
-The pore pressure parameters A and B relevant to the water content and density ratio
at which the soil is placed and the stress conditions in the embankment.
To be conservative in the estimation of pore pressures, the laboratory tests to estimate A and
B can be carried out at the upper limit of specified water content (e.g. optimum
2%) and/or
degree of compaction (density ratio 98% to 100%, standard compaction). However it must
be remembered that at optimum
2% water content the maximum achievable density ratio
may be 97% to 99%, while at optimum water content a density ratio of 100% to 101% may
be achievable and the laboratory testing should account for this (see
Figure 11.17
)
.
The stresses in the embankment due to construction of the dam can be estimated by
finite element techniques. ICOLD (1986a) and Duncan (1992, 1996a) give descriptions of
the methods and their limitations. It is concluded that:
-A two dimensional model can be used to estimate vertical stresses in a homogeneous
dam on a rigid (rock) foundation;
-A three dimensional model is necessary to model cross valley stresses in an embank-
ment on a compressible foundation;
-
The initial stresses locked into the fill during compaction should be incorporated into
the model;
-
The models must be “built” in layers, being loaded progressively.
