Environmental Engineering Reference
In-Depth Information
-
The difference between multiple reversal direct shear and field residual is only small in
weak rocks such as shales and claystones.
The authors' experience in testing clays and very weak claystone is that the difference
may be significantly greater (up to 6°) and direct shear tests on the slide surface or bed-
ding surface shear should always be done if practicable.
The authors' experience in testing soil from several landslides where soils were derived
from sedimentary rocks, tuff and basalt is that the Bromhead ring shear is very suscepti-
ble to the presence of particles of sand. These “catch” on the shear surface and lead to
higher residual strengths than obtained from multiple reversal direct shear or tests on slide
planes. Comparable results were only obtained by drying and sieving the soil on a 0.150 mm
sieve to remove any particles of sand before re-wetting it to the plastic limit for testing.
It can be seen that ideally tests should be carried out on the low strength planes if
already present, or that ring shear tests should be carried out provided that the soil is free
of sand particles. Unfortunately, ring shear machines are not widely available and revers-
ing direct shear tests have to be used. In these cases, the design strength should be selected
at the lower bound of results unless exhaustive testing shows this to be conservative. This
is valid since few (if any) of the reversing tests may have achieved a proper alignment of
particles, and the low results (provided these were not testing errors) are probably the best
indicator of residual strength. The validity of the results should also be checked by com-
parison with published data, e.g. that in
Figure 6.5
. The notable exceptions to the data on
that figure are allophane and halloysite clays which do not have classical platey structure,
so give higher values (e.g.
25° to 35°).
In many cases of slope instability, where stabilization measures are being designed, the
strength can be determined by back analysis and stabilization works designed on the basis
of increasing the factor of safety by no less than 30%. This obviates some of the need for
accurate knowledge of the shear strength from laboratory tests.
A method which has been used by the authors to determine the “lower bound” residual
strength of a soil, is to separate the clay fraction from the silt (and sand) of the soil by sedi-
mentation. The clay fraction is then tested by overconsolidating it in the direct shear machine,
and then determining the residual strength by multiple reversal direct shear testing. This pro-
cedure was suggested by Pells (1982) and has given reasonable values when used. More
recently we routinely use the ring shear on the minus 0.150 mm sieve soil as described above.
In nature, it is not uncommon to observe a clay rich zone 0.2 mm to 1 mm thick at the
slide surface, and in these circumstances the procedures outlined above can be expected to
give a good representation of the field condition.
R
6.1.5
Laboratory testing for undrained strength
The undrained strength of cohesive soils can be obtained from several laboratory tests:
(a) Unconfined compression (UC);
(b) Unconsolidated undrained (UU);
(c) Consoldiated undrained (CU).
It must be recognised (Germaine and Ladd, 1988; Jamiolkowski et al., 1985) that the
undrained strength is affected by:
-
The degree of disturbance of the soil sample. Disturbance can occur because of stress
relief on sampling; the sampling technique (disturbance as the thin wall tube or other
samples is pushed into the soil); handling (e.g. drying or disturbance during transporta-
tion, extrusion of the sample from the sampling tube, or trimming in the laboratory);
