The Role of Plasticity Index in Predicting
Research Officers, River Research Institute, P.O. Mohonpur, Dist.-Nadia, India
Plasticity and compressibility are typical properties of clays. Atterberg’s limits of a clayey soil reflect the clay content and clay type of a soil. Compression index is also a clay dependent parameter. Among different correlations between the engineering and index properties of soils, which are often used to lessen the work load of a soil investigation program, Skempton’s relationship (1944) between compression index (Cc) and liquid limit (wL) given as Cc=0.007(wL-10) for the remoulded clays is well known. Its modified form for the normally consolidated clays proposed by Terzaghi and Peck (1948) is very popular in geotechnical practice. Another popular relationship between compression index and initial void ratio (e0) has been proposed by Nishida (1956). There are similar other relationship given by different researchers, but the use of plasticity index, IP in the prediction of Cc is scarce.
On the other hand there are several correlations available of other engineering characteristics in relation to IP of a soil. Skempton (1957) has shown a linear relationship between the ratio of undrained shear strength (cu) and effective vertical stress (sv/), that is, cu / sv/ and plasticity index, IP. Seed et al. (1962) have noticed that IP of a soil may be used as a single factor for predicting swelling potential. Hence the relationship between Cc and IP remains a possible research subject in the field of geotechnical engineering.
The expression for compression of a soil shows that the soil parameter indicating compressibility is Cc/ = Cc / (1 + e0), instead of Cc. Cc/ defined as compression ratio seems to be a more useful basis for comparing the compressibility of two soils at a particular stress level. For this reason Cc/ has been considered in this study of the correlation between compression characteristics and plasticity index of clayey soils.
A series of consolidation tests on artificially mixed soil samples have been carried out, starting from an initial water content approximately equal to their liquid limit. The soil samples are different types of clays mixed with riverine sand in varying proportions. Seed et al. (1964) have already reported the use of such mixed soils in order to study the fundamental aspects of Atterberg’s limits. In this paper however, emphasis is laid on developing relationship between the compression characteristics and plasticity index of different soils.
Artificially mixed soil samples were prepared using commercially available bentonite and kaolin, and natural clays, passing 0.075 mm size sieve as the source of fines and a riverine sand passing 0.425 mm size sieve having specific gravity, Gs = 2.63 and effective size, D10 = 0.16 mm as the source of non-plastic material. These were mixed in varying proportions to get fifty numbers of different plasticities and hence index properties.
The natural soils were collected from different parts of the state of West Bengal, India. Amongst them, Purulia and Bankura soils are residual and all other soils are alluvial in nature.
The properties of the clays and proportion of those clays used to prepare different mixed soil samples are given in the Table 1.
Table 1. Clays and Mixtures used in the Study
|Clay Type/Origin||Clay Fraction
|Percentage of Clay in
Different Clay-Sand Mixtures
|No. of Mixed Soil Samples|
(Bentonite & Bankura Clay
Mixed in 1:1 Ratio)
Wet mechanical analysis by pipette method was followed to determine the clay fraction. Casagrande’s apparatus was used for determination of liquid limit. Soil slurry was prepared by adding distilled water corresponding to wL and after proper maturing it was put in the consolidation ring of size 60 mm diameter and 20 mm height. Care was taken to minimize friction of the ring wall and to expel entrapped air from the slurry. Every incremental load was placed after 24 hours during the consolidation test.
RESULTS AND DISCUSSION
It has been found (details not given here) that out of fifty soil samples, forty-nine samples are above the A-line on the plasticity chart; except one, for which plastic limit could not be determined. Again fourteen samples belong to CH group, fourteen samples belong to CI group, twenty samples belong to CL group and one sample belongs to CL-ML group.
In most of the cases the e-log sv/ curve has shown linearity approximately from sv/ = 20 kPa. So, though the tests were done in the stress range 5-640 kPa; the slope of the e-log sv/ curve in the stress range of 20-640 kPa has been taken as Cc and the initial void ratio is taken as the void ratio at 20 kPa.
Regression equations are developed between Cc and wL, e0, IP separately. Similar equations are also developed between the compression ratio, Cc/ and wL, e0, IP. The results of the statistical analysis have been presented in Table 2.
Table 2. Various Correlations Obtained
|No. of Samples||Equation||Correlation Coefficient
|Standard Error of Estimate|
|50||Cc = 0.0124wL – 0.1761||0.993||0.060|
|50||Cc = 0.5269e0 – 0.2117||0.995||0.053|
|49||Cc = 0.0150IP – 0.0198||0.996||0.046|
|50||Cc/ = 0.0021wL + 0.0587||0.930||0.035|
|50||Cc/ = 0.0888e0 + 0.0525||0.934||0.034|
|49||Cc/ = 0.0025IP + 0.0866||0.942||0.031|
Note. Cc/ = Cc / (1 + e0)
Cc vs wL relationship obtained in the present investigation is different to that presented by Skempton (1944). In this regard it may be worthy to be mentioned here, that all the tests in the present research program have been conducted under a unique stress range and the samples are above the A-line of the plasticity chart. The regression coefficients obtained in all the cases are very good. Thus, Cc vs IP relationship can also be used to predict Cc. The regression analysis of Cc/ vs wL, e0 or IP also show very good correlation.
Figure 1. Variation of Liquid Limit (wL), Plasticity Index (IP)
and Compression Index (Cc) with Kaolin Clay Fraction (C)
The wL vs C, IP vs C and Cc vs C have been plotted separately for kaolin (Fig. 1), bentonite (Fig. 2) and composite soil-sand mixtures (Fig. 3) as the number of data generated for these types of soil mixtures are suitable for plotting. In the lower clay percentage range the wL vs C curve shows a tendency to make an intercept on the ordinate for all the clays, while in the higher clay percent range the curve is linear and its backward projection passes through the origin. The IP vs C curve for kaolin mixtures (Fig. 1) is linear and passes through the origin.
Figure 2. Variation of Liquid Limit (wL), Plasticity Index (IP) and
Compression Index (Cc) with Composite Clay Fraction (C)
Figure 3. Variation of Liquid Limit (wL), Plasticity Index (IP)
and Compression Index (Cc) with Bentonite Clay Fraction (C)
But the IP vs C curves for the composite clay (Fig. 2) and bentonite (Fig. 3) comprise of roughly two straight line segments. The first segments make offsets on the abscissas and are steeper and extend approximately above C = 30%. Beyond that threshold clay percent, the second line segments have flatter slopes and if extended backwards, pass through the origins. The observations are same as those by Seed et al. (1964).
But a new thing emerges out from these plots that the nature of variations of Cc vs C plots are conspicuously similar to the nature of variations of IP vs C plots. So, it is probable that IP will relate better with Cc and this may be the cause for slightly better correlation in Cc vs IP in comparison with Cc vs wL.
Plasticity index, IP may be used in predicting Cc or Cc/ in addition to wL or e0. For remoulded clays Cc = 0.015Ip – 0.0198 and Cc/ = 0.0025IP + 0.0866. If a consistent relationship between the undisturbed and remoulded samples of normally consolidated clays is developed, the relationships proposed in the paper may come to some practical utility.
|© 2004 ejge|