Geology Reference
In-Depth Information
5.5.2 Granular soil
The behaviour of granular soil such as silt, sand and gravel can be
examined in the laboratory but for design, geotechnical parameters are
generally determined by in situ testing, because of the dif
culties of a)
obtaining and transporting undisturbed samples and b) the problems
of scale effects in testing samples of large grain size.
The most common test for characterising silt, sand and gravel is the
SPT, as discussed in
Chapter 4.
Measured resistance needs to be
corrected for various in
uences, including overburden pressure and
the silt content of sand. Resistance may be affected by water softening
in the base of a borehole. Details are given in Clayton (1995). SPT N-
values are used to infer a range of properties, including density (unit
weight), friction angle and deformability which are then used for the
design of many types of structure, including foundations, retaining
walls and slopes. CPT tests can also be used in this way and are
particularly useful for the design of offshore structures.
Usually, properties of intact soils of sedimentary origin are assumed to
be representative of the larger soil mass layer or unit. This can be an
over-simpli
cation in that even quite recent soils can contain fractures
and systematic joints and many are layered with different layers having
different properties. In the latter case, permeability parallel to bedding
might be orders of magnitude higher than at right angles to bedding,
and there are many geotechnical situations where such a condition
would be important. McGown et al. (1980) discusses origins of frac-
tures in soil and how they might be dealt with when assessing geotech-
nical properties. London Clay, for example, contains many
ssures
that can be interpreted using structural geological techniques (Fookes
& Parrish, 1969). Chandler (2000) describes the signi
cance of bed-
ding parallel
flexural-slip surfaces extending at least 300m in London
Clay. Similar features are discussed by Hutchinson (2001).
5.6.1 Intact rock
Fresh to moderately weathered rock, by the de
nitions adopted here
(
Chapter 3
and Appendix C), cannot be broken by hand at the intact
sample scale, as in a piece of core. That being so, it has an uncon
ned
compressive strength of at least 12.5 MPa and is de
nitely rock-like in
that it could carry most structures without failing (presumed bearing
capacity of at least 1 MPa according to
Table 6.1)
and will not fail in a
