Civil Engineering Reference
In-Depth Information
(b) Falling head permeability tests
In the falling head test illustrated in Fig. 7.2(b) water flows through the sample as the
level of water in the standpipe drops. Over a time interval
δ
t
the rate of flow is
a
δ
P
δ
Ak
P
L
=−
=
q
(7.7)
t
and hence, in the limit,
d
P
P
Ak
aL
d
t
−
=
(7.8)
Integrating with the limits
P
=
P
0
at
t
=
0 we have
ln
P
0
P
Ak
aL
t
=
(7.9)
and you can determine a value for
k
by plotting ln(
P
0
/
P
) against
t
and finding the
gradient. Notice that in a falling head test the effective stresses change because the
pore pressures change as the level of water in the standpipe falls. Any volume changes
that occur as a result of these changes of effective stress have to be neglected.
Values of the coefficient of permeability measured in laboratory permeameter tests
often do not represent the permeability in the ground, for a variety of reasons such as
anisotropy (i.e. values of
k
different for horizontal and vertical flow) and small samples
being unrepresentative of large volumes of soil in the ground, and in practice values
of
k
measured from
in situ
tests are much better.
7.5 Principal features of soil loading tests
Soil strength and stiffness are investigated and measured in tests in which soil sam-
ples are loaded and unloaded and the resulting stresses and strains are measured. The
requirements for testing soils are rather like those for testing metals, concrete and plas-
tics, but the special features of soil strength and stiffness impose special requirements.
The most important of these are:
1. Total stresses and pore pressures must be controlled and measured separately so
effective stresses, which govern soil behaviour, can be determined.
2. Drainage of water into, or out of, the sample must be controlled so that tests may
be either drained (i.e. constant pore pressure) or undrained (i.e. constant volume).
3. To investigate soil stiffness, measurements must be made of small strains (see
Chapter 13), but to investigate soil strength it is necessary to apply large strains,
sometimes greater than 20%.
4. Because soils are essentially frictional it is necessary to apply both normal and shear
stresses. This can be done either by applying confining pressures to cylindrical or
cubic samples or by applying normal stresses in direct shear tests (see Fig. 3.3); the
relationships between the principal stresses on cylindrical samples and the normal
and shear stresses on shear samples were discussed in Sec. 3.2.
