In contrast with the triaxial apparatus, in which the cell pressure also acts on the top of the sample, in the cell test the pressure in the cell only acts on the cylindrical surface of
Trang 1CELL TEST
An early version of the triaxial test was developed around 1938 by Keverling Buisman, see Figure 23.1 Actually, the triaxial test can be considered to be an improved version of this Dutch cell test The apparatus consists of a container with a cylindrical glass wall (the cell), with
.
.
.
.
.
.
.
.
.
.
.
.
.
Figure 23.1: Cell test
a fixed rubber membrane, in which the sample can be installed The water pressure in the cell can be controlled In contrast with the triaxial apparatus,
in which the cell pressure also acts on the top of the sample, in the cell test the pressure in the cell only acts on the cylindrical surface of the sample, because of the way of installing the membrane, which is glued to the top and bottom plates
of the apparatus The usual testing procedure consists of applying a vertical load, by means of a gradually increasing dead weight, and to measure the vertical deformation Note that in the standard triaxial test the rate of vertical deformation is imposed, and the vertical load is measured The cell test is stress controlled, whereas the triaxial test is strain controlled In horizontal direction there is no difference between the two tests: in both the horizontal stress is controlled by the cell pressure The size of the sample in a cell test usually is 6.5 cm diameter, which is somewhat larger than the size of the samples in a triaxial test (3.8 cm) This is of minor importance, however, and there exist larger cells, especially for tests on course material, such as gravel
The original purpose of the cell test was to investigate the stresses that would occur in a loaded sample while the strains remained small, without reaching failure For this purpose the lateral deformations should be kept to a minimum, close to zero This can be accomplished, approximately, by filling the cell with water, and to prevent volume change of this water, by closing the cell The cell pressure will then increase if the vertical stress is increased In this classical form of the cell test the cell pressure and the vertical deformation are measured as a function of the vertical stress The measured states of stress can be considered as safe, and correspond to small deformations only A disadvantage of this procedure is that the cell pressure, and thus the state of stress in the sample, depends upon the stiffness of the system of cell, water and rubber membrane An air bubble in the cell, or a fold in the membrane, may lead to greater lateral deformations and lower horizontal stresses An advantage is that the sample is never brought
to failure, so that the behavior at various stress levels can be tested on a single sample
135
Trang 2In the course of time the procedures of the triaxial test and of the cell test have been modified, and the possibilities have been increased,
so that an arbitrary combination of stresses can be applied, and all strains can be measured, or that the deformations are imposed and the
. .
σxx σzz σxz σzx σ1 σ3 σ3 σ1
.
.
.
.
.
.
.
.
.
.
.
.
c φ .
. Figure 23.2: Determination of c and φ from two cell tests corresponding stresses are measured The main difference between the triaxial test and the cell test then is that in the cell test the cell stress does not act on the top of the sample, but only in horizontal direction Many testing procedures have been developed For the determination of the shear strength parameters c and φ a popular procedure is to do two triaxial tests on two different samples of the same material, at different cell pressures, at a constant vertical deformation rate, as described in Chapter 21 Another procedure is to load a sample in a cell test, keeping the vertical stress constant, and then to carefully bring the sample almost to failure, by letting a small amount of water escape from the cell, in the form of water drops The cell pressure will decrease, and water can be drained until the cell pressure remains constant The sample then is on the verge of failure, but the deformations remain small In a second stage a new test can be done on the same sample, by first increasing the vertical stress, with the drainage tap of the cell closed The cell pressure will increase, with a value that depends upon the stiffness of the system of cell, water and membrane This new cell pressure can easily be measured, of course Next the sample can again be brought to the limit of failure by draining off some water from the cell, in small drops The Mohr circles for such a testing procedure are shown in Figure 23.2 Because in this type of test actual failure of the sample is just avoided, the values of the shear strength parameters c and φ are usually somewhat smaller than the values obtained from triaxial tests . . ε1
σ1− σ3
Figure 23.3: Some test results
It may be mentioned that very often laboratory tests are being used to determine the relation between stress and strain for the entire range of strains, from the small deforma-tions in the early stages, up to the large deformadeforma-tions at failure, and perhaps beyond, see Figure 23.3 If the vertical load is applied by imposing the strain (or the strain rate) a pos-sible decrease of the stress after reaching the maximum stress can also be detected The maximum strength is called peak strength, and the final strength, at very large strains,
is called the residual strength For certain types of soils the residual strength is much lower than the peak strength, for instance the calcareous sands that occur in offshore coastal zones of Western Australia and Brazil An example of such a result is also shown
in Figure 23.3 In this type of material the peak strength is so high, with respect to the
Trang 3residual strength, because the sand particles have been cemented together The sand will become very stiff, but brittle It appears to be very strong, and it is, but as soon as the structure has been broken, the strength falls down to a much lower value In the construction of two offshore platforms near the coast of Western Australia this has caused large problems, because the shear strength of the soil was reduced very severely after the driving of the foundation piles through the soil
Very often it is not sufficient to just determine the maximum strength of the soil, at failure Then only part of the available information is being used, and there is a definite risk of overestimating the strength In general it is much better to determine the relation between stresses and strains over the entire range of possible strains This also enables to take into account the reduction of the shear strength after a possible peak, depending on the strain of the soil
Problems
pressure remains constant, indicating failure The cell pressure then is 5.2 kPa If it is assumed that the cohesion c = 0, then what is the friction angle φ
of the material?
at 10 kPa What can be the vertical load before failure of the sample occurs?