SHEAR CHARACTERISTICS OF GRAVEL-SAND MIXTURES UNDER HIGH STRESSES

Abstract

The deposits adjoining the natural streams are usually formed by rivers and as such contain gravel, cobbles, boulders and :sand as main constituents with fine matrix. For designing structures on such deposits, an understanding of behaviour of such formation, particularly the shear characteristics is essential. Further, construction of high earth and rockfill dams causes high stresses on such deposits. Design of such structures necessitates an understanding of the shear charac teristics of large sized particle mixtures under high stresses. The conventional type of shear machines can not accommodate big sized particles. Thus it is essential to develop large sized shear testing machines to test the soils under high stresses, A triaxial test is considered to be better as compared to a direct shear test. However, the simplicity of test has made a direct shear testing device more popular. A large size direct shear box has been designed and fabricated to accommodate samples (square in plan), 1000 sq.cm. in area and 25.0 cm deep. The machine has a capacity of 20 2 kg/cm normal stress.The loading unit consists of a sturdy self compensating reaction frame for applying normal load. The shear force is applied by taking reaction against a concrete--U--type IV frame designed as abeam-cantilever member. It is designed to give ahorizontal reaction of 30 t. Arrangements for reversal of loading in horizontal direction have also been made. Two materials, namely gravel comprising of three .particle size ranges i.e. (i) 50 mm to 31.5 mm, (ii) 31.5 mm to 10 ma and (iii) 10 mm to 6.3 mm derived from the bed of Ganges near Hardwar and coarse sand comprising of river sand having particles ranging from 3.35 mm to 1mm were chosen for study. The finer material, namely sand was mixed with coarser material namely gravel in different percentages ranging from 0 to 100 by volume in dry condition to give samples for different series of tests. In all 16? tests were performed on various mixtures and materials, using ;six different values of normal stresses ranging from 5.097 to 20.097 kg/cm2. The variables considered for investigation are - (i) applied normal stress, (ii) percentage of finer material in a specimen, (iii) maximum particle size in specimen, (iv) hori zontal displacement of the specimen, (v) vertical displacement, (vi) spacing between two halves of the shear box, (vii) grading, (viii) void ratio, (ix) modulus of deformation and (x) particle breakage. Based on the analysis of test data, it has been observed that - (i) for gravel-sand mixtures under high applied normal stress, Mohr's envelope is curved and follows the law 7 = acrb where a and be are constants. For the materials tested relationship has been proposed between these two constants discussing in detail the various factors which influence the values of constants, (ii) shear strength of gravel-sand mixtures is governed by the finer or coarse material present in larger percentage , (iii) for poorly graded soil, the maximum parti cle size is noted to have practically no influence on shear strength. For gravel-sand mixtures, shear strength decreases with increase in maximum particle size in the specimen, (iv) percentage strain at failure remains practically constant, (v) vertical compression of the sample depends on the stress level, (vi) shear strength decreases with increase in spacing between two halves of the box. Spacing between two halves of the box equal to maximum particle size in specimen should be adopted for tests, (vii) well graded samples have higher shear strength than poorly graded ones, (viii) shear strength decreases with increase in void ratio, (ix) modulus of deformation, increases with increase in applied normal stress and (x) the particle breakage for soils with rounded particles is practically negli gible. Coarser soils show higher particle breakage than fine grained soils.