Environmental Engineering Reference
In-Depth Information
-Air intrusion may occur at the surface of the sample;
- Dissolved gas may come out of solution;
- Air may be trapped between the sample and the membrane in the triaxial test;
- Compacted soils even if compacted wet of optimum water content, as is commonly
required in dam construction, are usually only 90% to 95% saturated.
The effects of partial saturation are:
- Pore pressure changes during shearing are less than for a saturated sample. How-
ever, the trend (whether
ve changes) is not affected (Lade, 1986);
- Because pore pressure changes are less than for a saturated soil, the undrained
strength will be affected. For over-consolidated clays,
ve or
ve, and lower than for a
saturated soil, so the undrained strength is reduced; for normally consolidated clays,
u is
ve and lower than for a saturated soil, so the undrained strength is increased;
- The magnitude of actually measured pore pressures may be incorrect, usually
underestimated, because of the presence of air in the pore system.
Provided the pore pressures are measured correctly, the effective strength envelope
will not be greatly affected even though the total stress envelope is markedly affected.
To avoid problems of partial saturation, the laboratory samples should be saturated by
percolation followed by back pressure saturation. Percolation by itself will not achieve
saturation even in permeable soils (e.g. sands). Back pressure saturation is needed to
reduce the volume of the air bubbles in the pore water (Boyle's Law) and to drive air into
solution (Henry's Law). The degree of saturation should be checked by monitoring the
increase in pore pressure for an increment in cell pressure in undrained conditions:
u is
uB
3
(6.4)
where B is defined by Skempton (1954) as:
1
B
nC
C
(6.5)
v
1
sk
where n
porosity; C
v
compressibility of pore fluid; C
sk
compressibility of the
soil skeleton; For a saturated soil C
v
1.
However, for very stiff soils, and very weak rocks, C
sk
is also small and it is difficult to
achieve B
C
water
and C
v
/C
sk
is very small, hence B
1 as shown in
Figure 6.13
. In these circumstances it is sufficient to check
B for several increments of cell pressure and, provided B is constant, the correct effec-
tive strength parameters will be obtained (Lade 1986). For very stiff soils and very
weak rocks, the stiffness of the testing system becomes significant and may lead to low
measured B values.
Typically back pressure saturation will be carried out at 300 kPa to 1000 kPa, but
may be as low as 200 kPa for soft soils. It should be noted that dissolving the air in
the water takes time (often many hours). Black and Lee (1973) give a method to cal-
culate this time.
(iii)
Testing at too high a strain rate
. CD and CUDPP triaxial tests must be sheared at a
sufficiently slow rate to allow dissipation of pore pressure in the CD test and equali-
sation of pore pressure throughout the sample in the CUDPP test.
The effect of testing at too high a strain is illustrated in
Figure 6.14
for an over-
consolidated clay.
If a CUDPP test is sheared too quickly, the pore pressure changes measured at the
ends of the sample are less than the actual changes at the centre of the sample where
