Civil Engineering Reference
In-Depth Information
a distinction is made between loading or unloading (corresponding to increasing or
decreasing
p
) and compression or extension (corresponding to positive or negative
values of
q
). Notice that for compression the sample becomes shorter and fatter and
for extension it becomes longer and thinner; the pathOA corresponds to a conventional
triaxial test with constant cell pressure, while path OD is like squeezing a toothpaste
tube. During drained tests where the pore pressure
u
remains constant the total and
effective stress paths are parallel, but during undrained tests in which the pore pressure
generally changes, the total and effective stress paths are different.
7.10 Comments on soil testing
Although the routine soil tests described in this chapter are relatively simple there is
a lot that can, and often does, go wrong with soil tests. Probably the most signifi-
cant sources of error in measurements of soil parameters and behaviour in laboratory
tests are:
1. Malfunctions and errors in the apparatus and in the instruments.
2. Incorrect detailed procedures in performing the tests.
3. Doing the wrong test or measuring the wrong parameter for a particular application.
The last of these is simply a matter of sound understanding of the basic theories
involved, rather than blindly following a cookery book approach. The purpose of this
book is to develop this sound understanding. The first two are largely a matter of care
and attention and experience. In assessing the quality of a set of test results it is essential
to distinguish very carefully and clearly between the accuracy and the resolution of the
instruments. The resolution (or precision) of an observation is the smallest increment
that can be discerned, while the accuracy is the limit within which you can be absolutely
confident of the data. For a typical dial gauge measuring small displacements, the
resolution and accuracy are both about 0.001 mm, but the resolution and accuracy of
electronic instruments are often very different.
For a typical electronic load cell, pressure transducer or displacement transducer the
resolution is linked to the electronics which converts an analogue signal (usually a small
voltage) to a digital signal. For a 16-bit converter, using 1 bit for the sign, the resolution
is1in2
15
(
30 000) of the full-scale reading, so for a pore pressure transducer with a
range of 0 to 1000 kPa the resolution is about 0.03 kPa. The accuracy depends on the
linearity (or non-linearity) of the calibration constant between pressure and voltage
and on the stability of the electronic signals. With most instruments commonly used
in soil testing you will be doing well to achieve an accuracy better than
≈
±
1 kPa, which
is very different from the resolution.
The most difficult measurements to make are of small strains less than about 0.1% in
triaxial and shear tests. With conventional instruments for measurement of axial and
volumetric strain like those shown in Figs. 7.3 and 7.7 errors arise due to leakage and
compliance (movements) in the apparatus and often these errors are greater than the
measurements being made. Measurement of small and very small strains using local
gauges and dynamic methods are described in Chapter 13.
Another factor is in detection of malfunctions in instruments. It is usually fairly
easy to see whether a dial gauge or proving ring is not working properly, but it is
much less easy to detect malfunctions in electronic instruments provided that they
