Characterization of Soft Clay

views updated

Chapter 11
Characterization of Soft Clay


The geotechnical problems usually encountered in land reclamation and soil improvement projects include settlement, consolidation, and stability. Therefore, the characterization of soft clay in terms of compressibility, consolidation and strength is essential. However, the physical characteristics of clay is also important since physical parameters can be readily correlated with compressibility, consolidation, and strength parameters.


In order to roughly assess the consistency of soil, physical tests are usually carried out on either undisturbed or disturbed samples. From physical tests, some parameters of compression and strength can be estimated. These physical parameters will be briefly discussed in the following section, and details can be found in BS 1337 (1990) and soil laboratory testing manual by Head (1986).

11.1.1 Bulk unit weight of soil (γb)

Bulk unit weight is the first and easiest parameter to measure after obtaining the sample. From the known weight and volume of soil, the bulk density is obtained. Generally known volumes are achieved when the soil weight is measured in a certain mould. This method is called linear measurement. In most cases, an oedometer ring is used. Therefore, this measurement only requires a mould and balance, as shown in Figure 11.1.

11.1.2 Water content of soil (w)

The water content of soil can be determined from the ratio of the mass of water in its natural state to the mass of soil after drying out the water.

In order to determine the content of water, the soil sample has to be dried out. The drying method is important for some soils such as organic soil. Loss of weight may occur during the drying process. Such soil may require a long drying process at low temperatures, or by air drying. Although the test is simple, it could lead to erroneous results if the test is not carried out properly. Insufficient drying time could lead to an underestimation of the water content, as shown in Figure 11.2. The sample should also be crushed into small pieces in order to dry easily. Big lumps in the sample could lead to an underestimation of the moisture content (Figure 11.2 and 11.3).

11.1.3 Specific gravity of soil (Gs)

In order to find out other remaining parameters such as dry density, the specific gravity of the soil (Gs) is a key factor.

Generally, specific gravity is measured by using the density bottle method. Distilled water is normally used as the density bottle fluid. However, other types of fluids such as kerosene, or white spirit can be used when the

soil contains soluble salts. Although this method is simple, it requires several types of apparatus. The list of apparatus required is summarized by Head (1986) as follows:

  1. Density bottles (50ml) with stoppers, numbered and calibrated (three for each sample)
  2. Constant-temperature water-bath, with a shelf for holding density bottles, maintained at 25ºC.
  3. Vacuum desiccator.
  4. Oven and moisture-content apparatus.
  5. Analytical balance reading to 0.001g.
  6. Small rifle-box.
  7. Source of vacuum, and vacuum tubing.
  8. Chattaway spatula 150x3 mm.
  9. Wash bottle.
  10. Rubber-coated tongs, or rubber gloves.

Figure 11.4 is a photograph showing the apparatus used. It can be seen that more than half of the apparatus required are the same as those used for water content measurement. After measuring the various masses with an empty density bottle, with dry soil, with liquid, and with liquid and dry soil, the specific gravity Gs can be calculated using the following equation:

where Gl= specific gravity of the liquid used, m1= mass of density bottle, m2= mass of bottle + dry soil, m3= mass of bottle + soil + liquid, m4= mass of bottle + liquid.

Details of the test can be found in BS 1377 (1990) and Manual of Soil Laboratory Testing, volume 1 (Head 1986, vol. 1). Typical results are shown in Table 11.1.

11.1.4 Natural void ratio (eo)

After the natural water content has been determined, the natural void ratio can be readily calculated using the following equation for the case of fully saturated soil:

11.1.5 Dry unit weight of soil (γd)

The dry density of soil can be determined using the following equation if the specific gravity and natural void ratio is known.

where γw is the unit weight of water.

Table 11.1 Results of specific gravity test.
Mass of Density Bottle, g, m140.61640.479
Mass of Bottle + Dry Soil, g, m260.63160.499
Mass of Bottle + Soil + Water, g, m3153.134152.791
Mass of Bottle + Full of Water, g, m4140.523140.260
Specific Gravity of Soil Particle2.7032.673
Average Specific Gravity2.688 

11.1.6 Atterberg’s limits (Wl)

There are several limits of soil which indicate the liquid state, plastic state, and solid state. The liquid limit indicates the state when soil starts to behave like liquid. The liquid limit can be measured with two types of equipment: (i) Casagrande cup (ii) cone penetrometer.— Liquid limit by the Casagrande method

The Casagrande method is the original method of measuring liquid limit using a mechanical device called the Casagrande apparatus. The soil is placed in the cup and the surface scraped level with spatulas or palette knives. A groove is made at the center using a special grooving tool. The handle is turned at two revolutions per second which lifts the cup exactly 100 mm above the base and then dropped onto the base. This causes the groove in the soil to gradually close up. The number of blows required to bring two sides of the soil at the bottom of the groove into contact over a length of 13 mm is recorded. The tests are reported for various moisture

contents—at least four moisture contents. The linear relationship of the number of blows against the moisture content will be obtained, as shown in Figure 11.5. From these, the moisture content at 25 blows is considered the liquid limit.

Details of the test procedure can be found in BS 1377 (1990) and the Manual of Soil Laboratory Testing by Head (1986). Details of the apparatus required is summarized by Head (1986) as follows, and also shown in Figure 11.6.

Apparatus required:

  1. Casagrande device
  2. Grooving tool
  3. Flat glass plate
  4. Wash bottle containing distilled water
  5. Two palette knives, with blades of 200 mm long and 30 mm wide
  6. Spatula with blade of about 150 mm long and 25 mm wide
  7. Standard moisture content apparatus.— Liquid limit obtained by a cone penetrometer

An alternative method to determine the liquid limit is by using a cone penetrometer. In this method, the soil is placed in the cup and put onto the cone penetrometer apparatus. A cone clamp is made to touch the soil surface with its tip. It is then released for a second and clamped again. The depth of penetration is measured with the help of a dial gauge. A minimum of four tests are made for different moisture contents. A linear plot is obtained when the penetration depth is plotted against the moisture content. Moisture content at 20 mm penetration is taken as the liquid limit.

Apparatus required, summarized by Head (1986), as follows:

  1. Cone penetrometer with cone
  2. Sharpness gauge for cone
  3. Flat glass plate
  4. Metal cup
  5. Wash bottle with distilled water
  6. Metal straight edge
  7. Palette knives
  8. Standard moisture content apparatus

Details of the testing procedure can be found in BS 1377 (1990) and Manual of Soil Laboratory Testing by Head (1986).

Figure 11.7 shows the equipment required for a liquid limit test with a cone penetrometer, and Table 11.2 and Figure 11.8 show the calculation and determination of liquid limit.

Table 11.2 Example of liquid limit test results.
 Liquid Limit (%)
Wet Weight + Tare (g)28.228.928.729.4
Dry Weight + Tare (g)21.421.520.921.0
Tare Weight (g)10.310.410.010.3
Water Content (%)
Penetration (mm)10.714.417.922.1

It has been reported that the cone penetrometer method is less liable to experimental and human errors than those obtained by the Casagrande method (Sherwood and Ryley 1968). Although the same or similar values can be obtained from both methods, for a liquid limit of less than 100%, the cone penetrometer method gives slightly lower values than the Casagrande method for a liquid limit of greater than 100%, as shown in Figure 11.9.

The liquid limit of some types of soil are affected by the drying method and the type of water added to the soil. Especially when the soil has an organic content or minerals with volcanic origin, the liquid limit will be changed after oven drying. Under marine conditions, the soil deposited contains water with a significant salt content. Such soils are also affected by the drying method as well as the type of water added to the soil during testing, as shown in Figure 11.10.

11.1.7 Grain size distribution of soil

Grain size distribution of soil can be carried out by two methods depending upon the types of soil. Grain size distribution of medium to coarse grain soil, such as sand and gravel, can be determined using a set of sieves, and silt and clay by the hydrometer method.— Sieve analysis

The sieve analysis method is often used for classification of granular material. Table 11.3 shows the various sizes of sieve and the different numbers of sieves used in sieve analysis. Besides the sieves, another equipment required in this process is a balance. Soil can be passed through different sizes of sieves and based on the percentage of retained weight of

soil to the total weight of soil, the grain size distribution is obtained. Generally, the dry method is applied in sieve analysis although the wet method is used for finer content in the sample.

Figure 11.11 shows sieve analysis in progress in which the sieves are shaken on a mechanical shaker. The apparatus required for sieve analysis is also summarized by Head (1986) as follows:

  1. Set of sieves
  2. Mechanical shaker
  3. Balance
  4. Rifle box
  5. Drying oven
  6. Sieve brush
  7. Metal trays
  8. Rubber pestle and mortar
  9. Scoop
Table 11.3 Various aperture sizes and number of sieves.
Aperture Size (BS)Aperture size/sieve no. (ASTM)
75 mm3.35 mm3” (76 mm)No. 60 (200 μm)
63 mm2 mm2” (51 mm)No. 140 (106 μm)
50 mm1.18 mm1 1/2” (38 mm)Mp/ 200 (75 μm)
37.5 mm600 μm1” (25 mm) 
28 mm425 μm3/4” (9.5 mm) 
20 mm300 μm3/8” (9.5 mm) 
14 mm212 μmNo. 4 (4.75 mm) 
10 mm150 μmNo. 10 (2.0 mm) 
6.3 mm63 μmNo. 20 (850 μm) 
5 mm No. 40 (425 μm) 

Details of test procedures are found in BS 1377 (1990) and Manual for Soil Laboratory Test by Head (1986).

Sometimes dispersing agents are used to ensure separation, or dispersion of discrete soil. Generally, the dispersing agent is a compound of sodium carbonate and sodium hexametaphosphate, which is reported to be the most suitable.

Table 11.4 and Figure 11.12 show the calculation of sieve analysis data and the grain size distribution curve. After obtaining the grain size distribution curve, several other parameters such as effective grain size (D10), mean grain size (D50) and uniformity coefficient (D60/D10) can be calculated from the results.

Table 11.4 Example of sieve analysis results.
Sieve Size (mm)Cumulative %
Retained (%)
3.35100 Gravel19.80
Passing 0.02Silt and Clay0.02
D10 (mm)0.25
D50 (mm)0.79
D60 (mm)0.99
Uniformity Coefficient (U)3.96— Hydrometer tests

A hydrometer test is used to determine the grain size distribution of fine grain soil of smaller than 63 mm. The method generally applied is Stoke’s law of sedimentation. In order to carry out the tests, soil is prepared in a suspension of 1000 ml. The sedimentation cylinder is usually placed in a constant-temperature bath at 25ºC. A hydrometer is lowered into the suspension, and the density determined at predetermined time steps. From the time and the effective depth of the hydrometer, the equivalent diameter of the particle size can be determined using Stoke’s law. The amount of material in the suspension smaller than the particle size can be determined from the hydrometer reading.

Generally, pretreatment of soil is required before preparation of the suspension. Pretreatment is done in order to remove the organic matter and calcareous matter. A dispersing agent such as a compound of sodium carbonate and sodium hexametaphosphate is used to ensure the separation of discrete particles.

Several adjustments are required to obtain a correct reading. The full correct equation is:

where R is the fully correct reading.

Rh is the hydrometer reading.

Cm is the meniscus correction.

Mt is the temperature correction, and

x is the dispersing agent correction.

The apparatus required is summarized from the Manual of Soil Laboratory Testing by Head (1986) as follows:

  1. Soil hydrometer
  2. Two 1000 ml glasses
  3. Constant temperature bath
  4. Stop-clock
  5. Glass rod about 12 mm φand 400 mm length

The apparatus required is shown in Figure 11.13. The details of the procedure for testing is found in BS 1377 (1990) and Manual of Soil Laboratory Testing by Head (1986). Table 11.5 and Figure 11.14 show the test result and grain size distribution curve. Figure 11.15 shows a comparison of different grain size distribution curves, with and without corrections.

Table 11.5 Hydrometer test results.
Elapsed time
Hydrometer reading
True reading
Effective depth
Hr (mm)
Fully corrected reading
Particle diameter
Percentage finer than D

11.1.8 Chemical tests

Two common chemical tests required for reclamation and soil improvement projects are (i) organic content test for soft clay and (ii) shell content test for granular fill material.

(i) Organic Content Test: The standard method used for testing organic content is chemical oxidation. The basis of this method is to oxidize the carbon content of the soil using a solution of potassium dichromate and concentrated sulphuric acid. After oxidation, the remaining reagents are titrated against a standardized ferrous sulphate solution. This allows the quantity of reagents remaining after oxidation—hence the amount of reagents used—to be determined.

The apparatus used are all chemical laboratory apparatus, as shown in Figure 11.16. The details of the testing procedure are described in BS 1377 (1990) and Manual of Soil Laboratory Testing by Head (1986).

(ii) Shell Content Test: Shell is usually made up of calcium carbonate. Therefore, the content of calcium carbonate can again be determined by the titration method, using hydrochloric acid. The apparatus used are all available in a chemical laboratory.


Consolidation tests are carried out to determine the compressibility and consolidation characteristics of soil. There are several types of consolidation tests and these will be described briefly in the following section.

11.2.1 Oedometer test

An oedometer test is the conventional consolidation test which can determine the compressibility and consolidation parameters. Traditionally, tests are carried out under several load steps starting from a smaller load step—as small as 6 kPa—and completed at a load step of as high as 1600 kPa. The duration of the loading is usually 24 hours in order to ensure completion of primary consolidation. The load increment ratio (σ21) is usually unity. The types of apparatus required are summarized as follows:

  1. Standard oedometer equipment with dial gauge and a linear displacement transducer.
  2. Various sizes of weights.
  3. Stop clock.

Figure 11.17 shows standard oedometer equipment and weight set. Table 11.6 shows the standard test results with load increment ratio unity for a duration of 24 hours. The details of the testing procedure is found in BS 1377 (1990) and Manual of Soil Laboratory Testing by Head (1986).

Table 11.6 Example of oedometer test results.
Change in
Ht. (mm)
Kv(x 10-9m/s)
1120.0161.599  0.070 

From the oedometer tests, several compression and consolidation parameters can be determined.

(i) Compression Indices (Cc & Cr): Both the compression index within a recompression range and virgin compression range can be determined as shown in Figure 11.18.

(ii) Yield Stress (σ'y): By using the Casagrande method, yield stress (σ'y) can be determined, as shown in Figure 11.19. There are several other methods, such as those proposed by Janbu (1969), Butterfield (1979) and Sridharan et al. (1991) to determine the yield stress.

However, different methods give slightly different values of yield

stress, as shown in Figure 11.20 and Table 11.7. Even the Casagrande method is affected by scale, as discussed by Mikasa (1995) (Figure 11.21). There are some non-standard tests which are carried out with different load increment ratios. It has also been reported that different load increment ratios give different values of σ'y (Figure 11.22).

Table 11.7 Yield stress values (kPa) determined from different methods of analysis on the same sample.
Sample No.DepthCasagrandeButterfieldSridharanJanbu

(iii) Coefficient of Consolidation (Cv): Coefficient of consolidation can be determined for each loading step when the settlement rate at various time steps are available. One method was proposed by Casagrade and Fadum (1944) by plotting the settlement versus log time. From the graph, t50 is determined (Figure 11.23) and coefficient of consolidation is calculated by applying the following equation:

where Tv is 0.197

d is the thickness of the layer for a single drainage condition, and half the thickness of the layer for double drainage condition.

t50 is the time required for a 50 % degree consolidation.

Taylor (1942) has proposed an alternative method to determine settlement versus square root plot (Figure 11.24). From such a plot t90 is determined, as shown in the drawing. Cv is calculated from the following equation:

where Tv is 0.848

d is the thickness of the layer for a single drainage condition, and half the thickness of a layer for double drainage condition.

t90 is the time required for a 90 % degree consolidation.

These two methods give slightly different values for Cv, as shown in Figure 11.25. Taylor’s method gives higher Cv values.

(iv) Secondary Compression Index: From the settlement versus log time plot, another parameter which can be determined is the secondary compression index, as shown in Figure 11.26.

11.2.2 End of primary consolidation test

It is well known that 24 hours duration of loading on a thin layer of about 19 mm thickness sample will result in a significant magnitude of secondary consolidation settlement. At each step of loading, additional void ratio change occurs which causes the e-log σ'v curve to move towards the left side, as shown in the Figure 11.27. This moving of the e-log σ'v curve to the left side causes the reduction of the estimated yield stress, as shown in the

figure. Therefore, to obtain the correct e-log σ'v curve at the end of primary consolidation, the consolidation test should be carried out up to the end of primary consolidation. Primary consolidation can be detected if the consolidation test is carried out in concealed conditions with a pore pressure measurement facility. Otherwise, the end of primary consolidation can be estimated from the evs√t curve using the Taylor (1942) method. Figure 11.27 shows a comparison of e and log σ'vcurves from a conventional 24-hour oedometer test and the end of a primary consolidation test. The test procedure and sample preparation are the same except for increasing the next load at the end of primary consolidation.

11.2.3 Consolidation test with the Rowe Cell

The apparatus developed by Rowe (1966) is specially useful when a consolidation test is required to be carried out with pore pressure measurement. The apparatus is concealed and the pressure is usually applied with a hydraulic pressure. Since the pore pressure can be measured, the end of primary consolidation can be easily determined. Rowe Cell also allows various combinations of drainage to be carried out, as shown in Figure 11.31. Therefore, the coefficient of consolidation due to horizontal flow can be determined from a test with radial flow.

Figures 11.28 and 11.29 show a typical design of the Rowe Cell and its photograph respectively. Figure 11.30 shows pore pressure measurement and e-log σ'v curve obtained from a Rowe Cell test with radial drainage. Tests with several combinations of drainage can be carried out, as shown in the Figure 11.31. Based on the type of drainage, the coefficient of consolidation can be calculated (Table 11.8).

11.2.4 Constant Rate of Loading (CRL) test

An alternative way of carrying out a consolidation test is the Constant Rate of Loading test. Different test methods for applying and controlling the axial load have been described by Aboshi et al. (1970), Irwin (1975), and Burghignoli (1979). The Rowe Cell, together with the geodetic hydraulic system, is a suitable equipment for constant rate of loading test. The advantage of the constant rate of loading test is the short duration required to obtain a full set of results. However, e-log σ'v curves obtained from different rates of loading are different, as shown in Figure 11.32. Therefore, a suitable constant rate of loading is necessary to determine a reasonable rate of loading.

The rate of loading can be determined with the help of pore pressure measurement. The rate of loading with a pore pressure ratio (ub/ sv) of less than 60% is reasonable, when ub is the base pore pressure and v is the applied load. An average effective stress gain can be determined using the following equation proposed by Smith and Wahls (1969):

where is the average effective stress gain and σv is the applied stress,

and α is the ratio of the average pore pressure to pore pressure at the base. It is usually taken as 0.667 or 2/3.

Compression indices can be easily determined from a constant rate of loading. Figure 11.33 shows a comparison of a constant rate of loading test carried out under various rates of loading with 24-hour conventional test results.

Yield stress can be determined in the same way as discussed in the 24-hour loading tests. It should be noted that the magnitude of yield stress increases with the rate of loading. Therefore, it is deemed necessary to select the correct rate of loading. The coefficient of consolidation can be determined by applying the following equation:

where δP'/δt is the loading rate, H is the mean height of the sample in meters, and δu is the excess pore pressure.

Details of test procedures are found in the Manual of Soil Laboratory Testing, vol. 3 by Head (1986).

11.2.5 Constant Rate of Strain (CRS) test

The constant rate of strain test on natural soil has been described by a number of researchers, such as Smith and Wahls (1969), Anwar, Wissa, Christian, Davis and Heiberg (1971), and Gorman, Hopkins, Deen and Drnevich (1978), and Sheahan and Watters (1996). Its application to very soft clayey soils, such as dredged material and slurry, was reported by Umehara and Zen (1980), Carrier III and Beckman (1984). Tests on gaseous soils were reported by Wichman (1999).

The apparatus was developed by Wissa in 1971. An alternative apparatus is the Rowe Cell. With the Rowe Cell, a constant rate of strain can be achieved by pumping water into the diaphragm cell at a specified rate.

Various strain rates based on liquid limit, coefficients of consolidation and excess pore pressure ratio have been proposed by several researchers. Wissa et al. (1971) used a strain rate which generated a base excess pore pressure (Δub) of less than 30% of applied total vertical stress (σv). The strain rate has been standardized in ASTM D4186-82 where values ranging between 0.0001 and 0.04% per minute have been recommended.

Smith and Wahls (1969) proposed the strain rate for natural soil given in the following equation based on Cv and Cc:

Sheahan and Watters (1996) proposed the following equation for computing the coefficient of consolidation and permeability from the CRS tests for natural soil in non-transient conditions where the dimensionless time factor T is greater than 5.

where H is the current specimen height, γ is the strain rate and γw is the unit weight of water. σv1 and σv2 are the total stresses at two difference times, Dt and are the average effective stress obtained from:

where ub is the pore pressure at the base.

Smith and Wahls (1969) proposed the following equation to obtain cv from the CRS test:

where av is the coefficient of compressibility and b/r is the dimensionless ratio in which b is a constant that depends on the variation in the void ratio with the depth. The practical range of b/r is between zero and two.

Anwar et al. (1971) proposed the following equation for cv based on a non-linear theory assumingfor small ub.

Figure 11.34 shows a comparison of CRS tests with various rates of strain and 24 hours consolidation test. Figure 11.35 shows a constant rate of strain testing device developed by the Massachusetts Institute of Technology (MIT) and manufactured at Wykeham Farrance. Details of the testing procedure with the Wykeham Farrance equipment can also be found in the Manual of Soil Laboratory Testing by Head (1986).

Details of various types of consolidation tests carried out on Singapore Marine Clay and ultra-soft soil can be found in Bo et al. (1986) and Bo et al. (2003b) respectively.


The shear strength of clay can be determined in the laboratory in several ways. Depending on the type of strength required, undrained or drained, and mode of failure, test methods and equipment can be selected.

11.3.1 Laboratory vane

This is a quick method and shearing is done by rotating the soil cylinder. The method of testing and mode of failure are similar to field vane testing. The apparatus is shown in Figure 11.36. Depending on the expected magnitude of shear strength, various types of spring are used. Among the four available springs, no. 1 is the stiffest and can measure shear stress up to 90 kN/m2, no. 4 is the weakest and can measure only up to 20 kN/m2. The available types of spring are shown in Table 11.9. The resulting tongue values are converted to vane strength using the following formula.

where θf is the relative angular deflection (angle) at failure and K is the torsion constant which can be obtained from the supplier of the equipment.

Details of the testing procedure can be found in the Manual of Soil Laboratory Testing by Head (1986).

Table 11.9 Types of spring for laboratory vane test.
General descriptive term for strengthSuggested spring no.Maximum shear stress (kN/m2)
Very soft420
Soft to firm260

11.3.2 Direct simple shear test

Direct simple shear equipment is developed by the Norwegian Geotechnical Institute. With this equipment both drained and undrained shear tests on clay can be carried out.

In the direct simple shear apparatus, a cylindrical sample, either 50 or 70 mm in diameter by 10 mm thick, is confined in a rubber membrane. This rubber membrane is reinforced with thin metal washers that allow vertical and horizontal displacement with little change in diameter. This equipment also allows a consolidation test to be carried out on the sample. A drainage valve is provided to allow water to flow out for consolidation. The rate and magnitude of consolidation can be measured either by vertical displacement or measuring the volume of water drained out. When a consolidation test is being carried out lateral strain is completely prevented by a consolidation clamp which is fitted to the outside of the metal washers. Before shearing is carried out, the clamp has to be removed. In the system, vertical loading is provided by an air piston and the vertical load can be measured either by a load cell or pressure gauge.

The horizontal shear force is applied by a motor drive with 25 different speeds ranging from 0.0005 mm/min to 1.2 mm/min, as shown in Table 11.10. While the load is measured on the load cell, displacement is measured either by a dial gauge or a transducer. Shearing is generally achieved by displacing the sample horizontally at a certain displacement rate. Details of the test procedure can be found in the Handbook for Direct Simple Shear by Wykeham Ferrance. The test can be carried out in quick undrained, or consolidated undrained, or drained conditions. Figure 11.37 shows the direct simple shear equipment, and Figure 11.38 shows an example of test results obtained from direct simple shear tests for various types of clay.

Table 11.10 Different gear cogs and lever positions for speed selection.
Gear Lever Position60–30 *54–36 *45–45 *36–45 *30–60 *
Speed = mm/minute * number of gear cogs

11.3.3 Triaxial test

Triaxial tests are usually carried out on undisturbed soil or rock and reconstituted sand. There are several methods of testing; the choice of testing procedure is dependent upon the loading and drainage conditions of the problem being investigated and the method of analysis to be used.

The test apparatus consists of a loading frame with a loading machine at the base. The loading machine is generally motorized, with various speeds. The base of the cell, which includes the cell pressure, pore pressure, and back pressure gauges and the drainage part (Figure 11.39), is placed on the piston of the loading machine. Both pore pressure and back pressure can be measured with a pressure transducer. The cell, which is designed for water pressure, sits on the base plate during the test. Various sizes of base plates and cells can be found in the market (Figure 11.40). The sample, normally with a length to diameter ratio of 2, is made in a cylindrical shape

and placed on the base pedestal which is connected to the piston of the loading machine, and the load is generally measured either by a load cell or a probing ring with a dial gauge. Vertical displacement is also measured by a dial gauge or a displacement transducer. Figure 11.41 shows the general arrangement of a triaxial cell and load frame. Details on setting up testing and sample preparation are found in the Manual of Soil Laboratory Testing by Head (1986).— consolidated undrained test (UU)

When quick measurements of undrained shear strength of soil are required, unconsolidated undrained tests are carried out. Normally, the tests are carried with three identical samples under various cell pressures. The selection of cell pressure depends upon the undisturbed condition of the soil. In the second and third steps of the tests, the cell pressures are increased to double that of the earlier step.

Vertical pressure is provided by a loading machine with a suitable strain rate and the vertical load is measured with a load cell or probing ring. Graphical plots of stress-strain curves can be obtained from the test (Figure 11.42). The tests are carried out usually until failure or up to 20% of strain. From the applied pressure and measured vertical pressure, Mohr circles can be produced (Figure 11.43). Usually in UU tests the Mohr circles are almost identical. Unconsolidated undrained shear strength is obtained from

maximum deviator stress, as shown in Figure 11.43. Details of testing procedures are found in BS 1377 (1990), Manual of Soil Laboratory Testing by Head (1986), and Bishop and Hankel (1962).—Co nsolidated undrained triaxial compression test with measurement of pore pressure

Consolidated undrained tests are carried out when effective stress parameters are required for long-term stability analysis of shore protection works, retaining structures, and dredging and excavation works. Consolidated undrained triaxial test with pore pressure measurement will provide both total and effective stress parameters. By consolidating soil to in-situ stress, a certain degree of sample disturbance can be eliminated. However, consolidation is normally carried out by cell pressure, called isotropic consolidation.

Usually tests are carried out on three identical samples with the first sample consolidated to in-situ effective stress, and the second and third samples consolidated to the effective stress double that of the earlier sample. Undrained tests are carried out in the same manner as UU tests and pressure and displacement are measured. Pore pressure is also measured during the tests. Details of the test procedure are found in the Manual of Soil Laboratory Testing by Head (1986) and Bishop and Hankel (1962). From the obtained displacement, stress and pore pressure measurements, the total and effective stress parameters of soil are determined, as shown in Figure 11.44.— Consolidated drained triaxial compression test

When drained parameters are required for long-term stability analysis, consolidated drained triaxial tests are carried out. The apparatus used is the same as the triaxial apparatus; however, the drainage system is open during the test. The consolidation procedure is same as the consolidated undrained test. Consolidation stresses are also selected based on the same criteria as the CIU test. However, in the shearing stage the drainage valve is open and therefore there is no pore pressure. Details of the test procedures are found in the Manual of Soil Laboratory Testing by Head (1986), Bishop and Hankel (1962).

From the data obtained from displacement and stresses, stress-strain curves and Mohr circles can be produced. In turn, drained parameters can be determined, as shown in Figure 11.45.— Triaxial extension test

Extension shear stress is required when excess pressure needs to be removed or unloaded, or the excavation process needs to be analyzed. Extension can be achieved in two ways: (i) while maintaining vertical stress, horizontal stress is increased; and (ii) while horizontal stress is being maintained, and the vertical stress is reduced. Tests can be carried out for both consolidated drained and undrained conditions. Consolidation procedures are same as that explained earlier.

Generally, in the laboratory, extension is achieved by reducing the vertical stress by slowly moving the cell downward at a controlled rate. The base of the cell is clamped to the machine platen using a G-clamp. The arrangement of an extension triaxial test is shown in Figure 11.46. Details of the test procedures are found in the Manual of Soil Laboratory Testing by Head (1986), and Bishop and Henkel (1962).

From the measurement of stresses and displacement stress-strain curves, extension shear strength parameters can be obtained, as shown in Figure 11.47.— Stress path test

Shear strength of soil is stress and strain dependent. If the path of the stress is different, the shear strength varies. Moreover, in some geotechnical construction procedures, the paths of stress vary. Therefore, depending upon the path of stress, the relevant shear stress needs to be found. Even to find the realistic in-situ shear strength of soil, consolidation of soil needs to follow the k0 path, not the isotropic path. To carry out the shear strength test with various stress paths, a computer controlled triaxial testing system

is required. The GDS triaxial system with a digital hydraulic controller developed by GDS UK is one suitable equipment for testing stress path (Figure 11.48). Most of the time, this system consists of a Bishop & Wesley-type hydraulic cell. A digital hydraulic controller can provide controlled applied pressure and can also measure pore pressure, back pressure, and volume change. With the help of the software provided, tests can be carried out for various stress paths, both at the consolidation stage and shearing stage.

Details of the Bishop & Wesley cell (Figure 11.49) can be found in the Manual of Soil Laboratory Testing by Head (1986). The design and

features of the digital hydraulic controller (Figure 11.50) can be found in the handbook of GDS instruments. Figure 11.51 shows the layout of stress path testing. Details of the testing procedure are found in the Manual of Soil Laboratory Testing by Head (1986). Figure 11.52 shows typical CK0U results, and Figure 11.53 shows several stress paths.