Consolidation Behaviour of a Cement Stabilised Marine Soil

CONSOLIDATION BEHAVIOUR OF A CEMENT STABILISED MARINE SOIL

I. Bushra

Research Scholar, Department of Civil Engineering, Indian Institute of Technology, Madras–600036, India.

E-mail:

R.G. Robinson

Associate Professor, Department of Civil Engineering, Indian Institute of Technology, Madras–600036, India.

E-mail:

ABSTRACT: Deep cement mixing is currently accepted world wide as a ground improvement technology in order to improve the strength and deformation characteristics of soft cohesive soils. Though many studies are reported in the literature, the consolidation behaviour of cement treated clays is not well understood. In the present investigation the consolidation property of cement treated marine clay was studied by performing Constant Rate of Strain (CRS) tests, as the conventional consolidation frame cannot be used to apply high consolidation pressures. A marine soil excavated at a depth of 1.5 m from Ennore, near Chennai was selected for the study. The range of cement contents selected were 2.5 %, 5 %, 7.5 %, 10 %, 15 % and 20 % with curing period of 28 days. Water-cement ratio of 0.6 at an optimum total clay water content of 1.25 times the liquid limit water content was used for the test. Consolidation properties obtained for all the tests are reported in the paper for different cement contents.


431

Consolidation Behaviour of a Cement Stabilised Marine Soil

1. INTRODUCTION

1.1 Cement–Soil Stabilisation

Many geotechnical problems are encountered when construction activities are carried out in soft clay deposits due to their low shear strength and high compressibility characteristics. These soils are widespread in coastal and low land regions. Introduction of cement into soft ground, or cement-soil stabilisation, either in the form of dry cement powder or cement slurry is a popular method of ground improvement technique. The inclusion of cement into soil-water system causes physico-chemical changes at microstructural level leading to improved mechanical behaviour of the treated soil at macroscopic level. The in situ deep mixing technique is an established means of enhancing bearing capacity and reducing settlements by placing columnar inclusions in soft ground. The short-term gain in strength is the result of primary hydration reaction, which also leads to a reduction in moisture content during the chemical reaction. Subsequent long term gain in strength is a result of secondary pozzolanic reaction between the lime produced and the clay minerals. The extent of the strength improvement depends on the mineralogy, environmental conditions of the soft ground, curing period and the type and amount of cement used (Uddin et al. 1997, and Zen & Iwataki, 2006). Many researchers (Miura et al. 2001, Tan et al. 2002, and Horpibulsuk et al. 2003, 2004) carried out experimental studies to understand the strength improvement in soft ground using cement stabilisation techniques. Lorenzo et al. (2004 & 2006) have shown the existence of an optimum clay water content at which the maximum strength can be achieved for cement treated soils.

This paper mainly aims in determining the consolidation behaviour of a marine soil collected from Ennore, Chennai at an optimum clay water content of 1.25 times the liquid limit. The optimum clay water content was determined by conducting unconfined compression tests on cement treated samples for remoulding water contents ranging from 0.8 to 1.8 times the liquid limit.

2. EXPERIMENTAL INVESTIGATION

2.1 Soil Sample

Marine soil was excavated at a depth of 1.5 m from a site near Ennore, Chennai. As per the borehole charts available at the site, the soil stratification consists of 3 m thick soft clay, 3–8 m thick sand and 8–10 m stiff clay. The water table is at a depth of 0.7 m from the ground level. Sufficient quantity of the soil sample was brought from the site for carrying out the experiments. The soil was then air dried, crushed and sieved through 4.75 mm sieve to remove shell pieces and other bigger sized particles. The index and basic properties of the soil are listed in Table 1.

The unconfined compressive strength in the field was below 20 kPa at a water content of 45 % which is about 0.8 times the liquid limit. The coefficient of consolidation (Cv) obtained from consolidation test was 6.272 × 10–5 cm2/sec and permeability was 6.40562 × 10–9 cm/sec for an applied pressure of 400 kPa. X-ray diffraction (XRD) results of the base clay shows the presence of Quartz and Feldspar in the silt fraction and traces of illite and Kaolinite in the clay fraction.

Table 1: Basic Properties of Marine Soil

Properties / Values
Liquid Limit (%)
Plastic Limit (%)
Plasticity Index (%)
Specific Gravity
Free Swell Index (%)
Activity
Classification
Grain Size Distribution
Sand (%)
Silt (%)
Clay (%)
pH value
Organic matter / 56
25
31
2.59
45
0.7
CH
9
47
44
7.2
5.47

2.2 Methodology of Testing

For given cement content, the shear strength of the cement treated soil depends on the clay water content. There exists an optimum clay water content at which the shear strength is maximum (Lorenzo et al., 2006). From the UCC tests performed on samples treated with cement contents of 10%, 15% and 20% the optimum clay water content, which is the water content at which the shear strength is maximum, is taken as 1.25 times the liquid limit irrespective of the cement content. The remoulding water content (w*) is defined as the water content prior to the addition of cement slurry. The base clay was mixed with remoulding water content in the Hobart mixer for ten minutes with a planetary speed of 61 rpm for the first five minutes and a speed of 125 rpm for the last five minutes to obtain uniform mixing of samples based on trials. The prepared remoulded clay was then mixed with cement slurry at a water-cement ratio of 0.6 for another ten minutes in the Hobart mixer until a homogeneous clay-water–cement paste was attained.

The specimens for the consolidation tests were prepared by placing the clay-water–cement paste directly into the oedometer ring by thumb kneading. Care was taken to ensure that the sample prepared is free of air voids. This technique was adapted to avoid possible disturbance during the subsequent sample cutting and fitting into the oedometer ring. The oedometer rings together with the specimen were wrapped in layers of cling film and placed in a mist room for curing for 28 days. In previous research conducted by Uddin et al. (1997), the post yield compression line was almost unaffected with curing time in excess of the optimum curing period of 28 days. After curing, each specimen was removed from the plastic covers and placed into the consolidometer cell with filter paper and porous stones at both ends of the specimen.

Consolidation characteristics of soils are commonly studied by step loading tests using conventional oedometer frames. Step load tests on stabilised soil are unsuitable as the properties of the stabilised soil changes with time. When testing stabilised soils, the time-dependent effects of stabilisation can alter the compression characteristics unless tests are relatively rapid in relation to the rate of stabilisation (Kassim & Clarke 1999). The constant rate of strain consolidation test (CRS test) is a method to reduce consolidation test time. In addition, higher consolidation pressures can be easily applied to the soil samples. In this test, the soil sample is loaded continuously, rather than incrementally, at a constant rate of strain. It was Smith & Wahls (1969) who first suggested an analysis that could be used to interpret the data from the test.

The set up used in this experimental investigation is as shown in Figure 1. Conventional digital triaxial frame was used to apply constant strain rate of strain. The conventional consolidation cell was slightly modified with a provision for measurement of pore pressures at the base of the specimen. Drainage is permitted from the top end of the specimen while the pore pressure is measured from the bottom end of the specimen. This allows the excess pore pressures to be measured and controlled at the undrained face of the specimen, enabling the estimation of coefficient of consolidation. In the present study, a deformation rate of
0.05 mm/minute was adopted that was calculated as per the procedure given by Smith Wahls (1969) based on the following equation:

where, R is the strain rate and m is a proportionality constant that would normally range between 0.6 and 0.8. The values of coefficient of consolidation (Cv) and compression index (Cc) are taken from 1-D consolidation test, previously conducted. Ho and eo are initial height and initial void ratio, respectively. The values of (ub/σ1) is taken as 0.5, where ub is the pore pressure at the undrained boundary and s1 is the applied vertical pressure.

The samples taken from the mist room after 28 days of curing were directly placed in the consolidation cell of the equipment. In order to verify the validity of CRS test for stabilised soil samples, a test was conducted on a sample using the conventional consolidation frame using step loading procedure. The tests could not be performed beyond a pressure of 800 kPa. Before carrying out the CRS tests, the pore pressure lines were properly flushed to ensure proper pore pressure measurements. As the samples were cured directly in the oedometer ring made of stainless steel, there is likely chance of bonding developed between the ring and the soil, which may resist the applied load. In order to verify this, a controlled test was conducted on a typical sample. The sample after curing was carefully ejected out of the ring and placed back after cleaning the surface and application of lubricant. The results were then compared with another sample of same conditions which was not ejected out of the ring. CRS test were conducted on samples treated with cement content of 2.5%, 5%, 7.5%, 10%, 15% and 20%.

Fig. 1: Set up for CRS Test

3. RESULTS AND DISCUSSIONS

The e-log σv¢ plots obtained by the CRS tests are compared with the conventional one-dimensional consolidation tests, by step loading, in Figure 2, for a typical case of 10% cement treatment. The results compare reasonably well within the range of pressures used in the one-dimensional consolidation tests, suggesting that the CRS tests is suitable for cement treated soils. The e-log σv¢ plots for samples ejected out and replaced in the cell and those without ejecting out are shown in Figure 3. The results are remarkably the same suggesting that the bonding stress between the wall of the cell and the soil is negligible. Therefore, further tests were conducted without ejecting the sample out of the cell.

Fig. 2: Comparison of Step Loading and CRS Test

Fig. 3: e–log σv¢ Plot Showing the Effect of Ejection of Samples

The variation of yield stress with percent cement is shown in Figure 4. The value of yield stress increases with increase in percentage of cement. However, beyond a cement content of 15%, the increase in yield stress is marginal. The Cc values increased with increase of cement content at lower percentage of cement (cement content < 10%), as seen in Figure 5. Beyond a cement content of 10%, the compression index decreases. This may be due to the sudden breakage of cementation bonds at the contacts at lower percentage of cement. At higher cement contents, the bond strength is strong that sudden reduction in compressibility has not occurred leading to lesser compression index values.

Fig. 4: Yield Stress Vs Percentage of Cement

Fig. 5: Variation of Cc with Cement Content

The e-log σv¢ plots of 2.5 % ,5 %,7.5 % 10 %, 15% and 20% cement treated samples along with the plot of untreated soil sample are shown in Figure 6. As expected, the behaviour of cement treated soils behave like a over-consolidated soils. After yield stress, linear e-log σv¢ was obtained similar to normally consolidated clays. The position of e-log σv¢ plots of cement treated soils lie above the untreated soils suggesting the formation of cementation bonds. The yield stress is obtained as the point of intersection of two straight lines extended from the linear portions on either end of the compression curve plotted as log (1 + e) against log σv¢ (Sridharan et al., 1991). The compression index Cc is calculated as the slope of the e-log σv¢ plot beyond the yield stress.

Fig. 6: e-Log σv¢ Plot for Cement Treated Samples

4. CONCLUSIONS

·  Constant Rate of strain consolidation test is best suited for stabilised soils where age is a dominant factor.

·  The role of cement admixture is to increase the yield stress. However, beyond certain percentage of cement, increase in yield stress is small.

·  The values of compression index tend to increase with increase in cement content. Beyond a cement content of 10% the compression index decreases with increase in cement content.

REFERENCES

Horpibulsuk, S., Miura, N. and Nagaraj, T.S. (2003). “Assessment of Strength Development in Cemented Clays with Abrams’ Law as a Basis”, Geotechnique, Vol. 53 (4), 439–444.

Horpibulsuk, S., Miura, N. and Bergado, D.T. (2004). “Undrained Shear Behaviour of Cement Admixed Clay at High Water Content”, J. Geotech &Geoenviron Eng., ASCE, Vol. 130 (10), 1096–1105.

Kassim, K.A. and Clarke, B.G. (1999). “Constant Rate of Strain Consolidation Equipment and Procedure for Stabilised Soils”, Geotechnical Testing Journal, Vol. 22, 13–21.

Lorenzo, G.A. and Bergado, D.T. (2004). “Fundamental Parameters of Cement-Admixed Clay–New Approach”, Geotech & Geoenviron Eng., Vol. 130 (10), 1042–1050.