Environmental Engineering Reference
In-Depth Information
-
The water content of testing e.g. soils allowed to dry from their field water content will
be stronger than those tested at the field water content. Thin-walled tubes should be
sealed as soon as they are taken to avoid drying;
-
The stress path, in particular whether tested in compression, extension, or direct sim-
ple shear (see
Figures 6.9
and
6.10
)
;
-
The test procedure e.g. UC and UU tests do not simulate the in-situ test stresses.
Sample disturbance for soft clays, such as may be present in foundations of small dams,
puddled cores, or in saturated, poorly compacted clay fill; and the sensitivity of undrained
strength to the water content for stiff clays, make UC and UU tests totally unreliable, and
they should not be used for design purposes. In particular for soft clays they may seriously
underestimate the real strength - but you never know by how much.
Tests should therefore be done by CU methods, and they should, so far as practicable,
simulate the in-situ stress conditions, and the stress path.
For testing of soft clays, the most commonly used methods are the recompression test,
and the SHANSEP technique. These are discussed in Jamiolkowski et al. (1985),
Germaine and Ladd (1988) and Lacasse and Berre (1988).
The recompression technique was developed by the Norwegian Geotechnical Institute
(NGI), and is described in Bjerrum (1973) and Jamiolkowski et al. (1985). Samples are
consolidated to the same effective stress condition as experienced in-situ, before being
sheared undrained, either in compression or extension. That is, the samples are consoli-
dated under
vo
) is maintained during consoli-
dation, but NGI have shown that the simpler procedure of consolidating isotropically
under
vo
and
ho
. Ideally a constant Ko (
ho
/
vo
and consolidating further, gives satisfactory results.
Lacasse and Berre (1988) indicate that for soils with an OCR less than 1.5, they recom-
mend a two stage application of the cell pressure, and application of the deviatoric stress
in 4 to 6 steps. For other soils, a single step consolidation is acceptable.
The SHANSEP - “Stress History And Normalised Engineering Properties” - technique
was developed by Ladd and others. It is described in Ladd and Foot (1974), Ladd et al.
(1977) and Jamiolkowski et al. (1985).
Note that for both SHANSEP and the recompression techniques, S
u
should be deter-
mined in compression or extension, whichever is appropriate. Both techniques are
claimed to give high quality strength data, which compares well with strengths deter-
mined by back analysis of failures. The recompression technique is simpler to use, requir-
ing less laboratory testing, but can give an overestimate of strength in normally and lightly
over-consolidated clays. SHANSEP involves consolidation of samples beyond their in-situ
stresses, and may lead to destruction of soil structure and cementing, giving an underesti-
mate of strength.
In many projects the SHANSEP technique would be too expensive, particularly since a
S
u
/
ho
, then increasing
v
to
vo
vs. Over Consolidated Ratio (OCR) relation has to be developed for each soil type
on the site. It is also dependent on being able to determine the OCR of the soil accurately,
and this is often difficult because the pre-consolidation pressure of the soil as determined
in oedometer tests is affected by sample disturbance. The recompression technique on the
other hand is relatively simple and is recommended. It is emphasised that a lesser number
of samples tested by recompression or SHANSEP techniques is far preferable to a large
number of cheaper UC or UU tests.
The recompression and SHANSEP methods require a knowledge of the at rest earth
pressure coefficient, Ko, in situ. This can be determined by:
-
Self boring pressuremeter (SBPM) tests;
-
Use of published relationships between Ko and effective friction angle
,
plasticity
index I
p
, and over-consolidation ratio OCR.
