STUDIES ON PHYSICO-CHEMICAL AND MECHANICAL PROPERTIES OF SOME INDIAN SOILS FOR DEVELOPING HEAVY CLAY PRODUCTS

Abstract

In view of the widespread use of clays in industry and taking into consideration the basic problems of highly complex nature involved therein, soil chemists have made untiring efforts to resolve them during the last four to five decades. In recent years these efforts have further augmented with increasing knowledge and better understand ing of clay mineralogy to changing physico-chemical concept of clays. This has resulted in a re-evaluation of the experimental data in various aspects of soil science, particularly those related with clay water interactions. Variations in physico-chemical characteristics of clays are essentially due to mineral make up of the parent rock, topography, weathering conditions of the parent rocks etc. Owing to the involvement of such a large number of parameters, it is natural that any single theoretical treatment is not likely to find a universal interpretation of any single happening either physical or chemical or mechanical in soil system. Keeping in view these limita tions, attempt has been made to investigate the role of certain dominating factors such as mechanical and mineralogical make up of the soil, specific surface area, exchangeable cations etc. on the plastic and rheological behaviour of a few representative natural occuring soils used in the production of heavy clay products near flowing : 215 : consistency. These investigations, which form the main theme of the present thesis are summarised below: Chapter I which is in the form of a general intro duction gives a brief review of the studies carried out by various workers on the physico-chemical and mineralogical aspects of clays and purports the aim and object of the investigations, studies under discussion. In the second chapter, the results on physicochemical, mechanical and mineralogical properties of twenty representative natural occuring Indian soils used in the production of heavy clay products from alluvial, black, red and coastal alluvium regions, covering the entire soil zone are described. On the basis of results obtained it is concluded that alluvial clays of Indogangetic plains are basic and rich in alkaline earth minerals while the Brahamputra alluvium of Assam region are acidic and low in alkaline earth constituents and contain iron in the form of hydrated oxides invariably present in all the soils of Brahamputra valley. The Indo-gangetic alluvium is often associated v/ith patches of water soluble salts particularly in Uttar Pradesh, Punjab and Haryana. The residual soils covering a large area of penin sular region have been formed over a variety of geological formations and can be classified as black, red and lateritic soils etc. The black soils of Indore, Morwi, : 216 : Gulbarga etc. are characterised by their dark reddish brown to medium and deep black in colour, heavy texture, high plasticity and shrinkage, high proportion of clay content and are generally associated with lime nodules and organic carbon. The red soils exist from Bundelkhand to extreme south, Orissa, East Hyderabad, Karnataka and a part of Madhya Pradesh, Assam and Tamil Nadu, soils are red to yellow in colour, which is basically duo to wide diffusion of iron oxide rather than higher percentage of iron oxide. The red soils of Quilon and Pondicherry are rich in organic carbon. The coastal alluvium of Kandla are dark in colour and contain fairly high proportions of marine salts, oysters, organic carbon etc. The mineralogical examination based on the results of base exchange capacity, DTA and x-ray analysis of these Indian soils are detailed earlier reveal that alluvial soils predominantly contain mica group of clay mineral while the red and blaclc soil primarily contain kandite and montmorillonite group of clay minerals respectively. It is observed that clay minerals present in these soils are quite often a mixture of two or more along with fines from accessory minerals like silica, hydrated oxides of iron and alumina, calcite etc. These constituents cause small variations in the peak temperature of the DTA curves. : 217 : In Chapter III, the rheological properties of clay pastes have been reported. Because of the technological importance of plasticity in engineering and ceramics, the subject has been critically and more comprehensively dealt with, both from the experimental and theoretical points of view. Since the Atterberg limits (i.e. liquid limit, plastic limit and plastic Index) provide the most signifi cant basis for the commercial appraisal of clays, the determination of these limits has been a' subject of much discussion. It is reported that conventional cup method for the determination of liquid limit is based on certain emperical assumptions such as the impact of 25 blows to the standard cup leading to development of critical shear ing strength (a point at which the clay water mass yields plastically). The results of liquid limit obtained are greatly influenced by rate of impact, the thickness and material of the cup, the nature and hardness of the material v/ith which the cup is knocked etc. It is experi enced that difficulties occur in cutting a groove in a sandy, silty or organic soil and even in determining the satisfactory closure of the groove. This puts severe limitations to the validity of liquid limit data on soils. Various methods are proposed for liquid limit determina tion which are based on direct or indirect shear tests but are not used because of the complex nature of the : 218 : equipment involved. From amongst these, the one based on determining the rheological behaviour at small shearing stress near liquid limit consistency appears to be the most reliable and convenient one. The Bingham yield stress obtained by extrapolation of the hysteresis curve of clay pastes when plotted against corresponding moisture content yield two linear curves of different slopes and the point of intersection can be taken up to correspond to the liquid limit of the soil. This point represents a transition point for the change in the mechanical behaviour of the clay-water mass from the liquid to the plastic state. The results of liquid limit obtained for various soils are found in close agreement v/ith the values obtained by conventional cup method and one point method. This method, based on rheological measurements for the determination of liquid limit, is independent of emperical assumption of 25 number of blows as involved in conventional method. Furthermore the element of human error involved in the determination of liquid limit by conventional cup method is also greatly reduced here. The Bingham yield stress at liquid limit consistency as obtained by this method for all soils vary in a narrov/ range of 16 to 32 g/cm irrespective of their mechanical and mineralogical make up and are quite comparable with the reported values of other workers. The plot between liquid limit and Bingham yield stress for various soils : 219 : show an ill-defined linearity with a correlation coeffi cient (r) = 0.47 for 20 d.o.f. However, Bingham yield stress did not show any significant relationship with activity coefficient of the soils. Such a behaviour is not surprising since the activity coefficient is consider ably modified by the particle size and shape, organic matter, nature and amount of non-clay minerals present in the soil. m Chapter IV, the thixotropy in clay pastes from soils has been measured by two approached as detailed by Green and Neltmann, and Goodeve and Whitfield. The plots between thixotropic coefficient (Goodeve and Whitfield)/ thixotropic break-down (Green and Weltmann) of clay pastes at varying moisture contents yield two linear curves of different slopes for each soil. The point of intersection of these linear curves again corresponds to the liquid limit of the soil and represents a transition in behaviour of clay water mass from liquid state to the plastic state. At this point the clay water mass may show an optimum degree of thixotropy and critically small shearing strength. The thixotropic coefficient and thixotropic break down at liquid limit consistency vary in a narrow range of 0.7 to 1.5 and 3.4 to 4.6 respectively. The results of liquid limit obtained by the aforesaid rheolo gical methods are again in close agreement v/ith the values obtained by conventional cup method/ one point method and : 220 : rheological method based on the measurement of Bingham yield stress at various moisture contents. Various mathematical correlations between liquid limit and thixo tropic coefficient/thixotropic break down for all the soils have been reported. Thus an overall picture with regard to the determina tion of various rheological parameters it has been conclud ed that:- 1. The basic rheological parameters as Bingham yield stress, thixotropic coefficient and thixotropic break down at liquid limit consis tency vary within a narrow range for all the soils irrespective of their chemical and mineralogical make up. 2. At the boundry of liquid and plastic state the cohesive soils exhibit a small shearing strength. This observation is in conformity with the results of various workers such as Antterberg, Casagrande, Terzaghi, Norman, Scott. 3. Slight variations observed in Bingham yield stress, thixotropic coefficient/thixotropic break down at liquid limit consistency of different soils have been attributed to nature of exchangeable cations, organic matter, nonclay mineral component present in the soil. : 221 : 4. Factors given under (3) above also cause significant variation in the cohesive proper ties of the soils. In chapter V, the role of specific surface area on various plastic and rheological properties is reported. The high activity of the clay surface and its influence on the clay water interactions greatly modify the plastic behaviour of clay mass. in view of the great impact of surface area on the mechanical properties of clay water mass, the investigations v/ere extended to find the magnitude of variation in the physico-chemical, thixotropy, shear strength properties of few Indian soils at liquid limit consistency. Various methods for specific surface area determina tion have been reviewed with particular emphasis on the gas adsorption techniques based on BET principles. The investigations reported clearly indicate the surface area is closely related to the base exchange capacity and liquid limit of the soils. It has been observed that the plot between plastic index/surface area and (a) Bingham yield stress at liquid limit; (b) thixotropic break down at liquid limit; and (c) liquid limit of various soils yield three distinct linear curves in each case due to difference in textural behaviour of the soils. Exceptional textural behaviour observed in case of a few soils have been assigned : 222 : due to the variations in cohesive properties, size and shape of the particles, and presence of non-clay minerals present in the soils. However, no significant relation ship between activity coefficient and surface area of the soil could be obtained. It is interesting to note that thixotropic coefficient/thixotropic break down at liquid limit consistency show significant correlations with surface area particularly for soils possessing high liquid limit (more than 50, CH group of soils). It is assumed that frictional resistance is likely to pre-dominate due to larger proportions of silt and sand grains present in less plastic clays, which in turn, affect the observed values of Bingham yield stress, thixotropic break-down/ thixotropic coefficient at liquid limit consistency of natural occuring soils. m Chapter VI, the role of exchangeable cations on the physico-chemical and mechanical properties of three representative soils from Roorkee (B) (illite), Madras (montmorillonite) and Morwi (B) (Kaolinite)have been studied. It has been observed that liquid limit of various cationic forms vary with the nature of exchange able cation in all the three soils. The maximum variation in liquid limit values of Madras soil and least in Morwi soil (B) have been observed, while it has been found intermediate for Roorkee soil (B). It may be noted that an appreciable increase in liquid limit values of Madras : 223 : soil (montmorillonite) saturated with monovalent cation is more pronounced as compared to Roorkee soil (illite). In the case of Morwi soil (B) (kaolinite) the exchange able monovalent cation show a small reduction in liquid limit values compared to di and tri-valent cations present in exchangeable sites. The exceptional values observed with H form for all the three soils have been assigned as a result of partial degradation of mineral lattice and dissolution of various primary and secondary minerals present in the soil by pretreatment with hydro chloric acid solution. Further, with increasing valence of exchangeable cations (i.e. from uni-valent to tri valent) in Madras soil (montmorillonite) and Roorkee soil (illite) the Bingham yield stress, thixotropic coefficient/thixotropic break-down at liquid limit con sistency also increases, while in the case of Morwi soil (kaolinite) the trend is almost reverse. However, the magnitude of variation in all these rheological parameters at liquid limit consistency is not very large for all the three soils. This behaviour indicates that thixo tropic and flow properties at this consistency are slight ly modified as a result of increasing cohesion and inter particle attraction between clay particles. This effect -f3 is more marked when tri-vAlent Al is exchanged or adsorbed in these clays.