Chapter 3 Shear Strength of Soils ‘The property that enables a material to remain in equilibrium when its surface is not level is known as its shear strength. Soils in liquid form have virtually no shear strength and even when solid have shear strengths of relatively stnall magnitudes compared with those exhibited by steel or concrete, To appreciate this section some knowledge of the relevant strength of rmateriais is useful, A brief summary of this subject is set out below. 3.1. Friction Consider a block of weight W resting on a horizontal plane (Fig, 3.18). The vertical reaction, R, equals W, and there is consequently no tendency for the block to move, Ifa small horizontal force, H, is now applied to the block and the magnitude of H is such that the block still does not move, then the reac tion R will no Jonger act vertically but becomes inclined at some angle, @, to the vertical By considering the equilibrium of forces, first im the horizontal direction and then in the vertical direction, itis seen that: Horizontal component of R Rsina ‘Vertical component of R= W=Reosa (Fig. 3.1b) w wo oNew}\a 4 Hon a R cx{o9 atecingy (0) No nodzon (©) Rosen oe spo ‘Yorce applied ‘ce apoiod Fig. Friction, 82Shear Strength of Soils 83 ‘The angle a is called the angle of obliquity and is the angle that the reaction con the plane of sliding makes with the normal to that plane. If H is slowly increased in magnitude a stage will be reached at which siding is imminent; as His increased the value of e will also inerease until, when sliding is imminent, a has reached a limiting value, ¢. If H is now increased still further the angle of obliquity, 4, will not become greater and the block, having achieved its maximum resistance to horizontal movement, will move (é is known as the angle of friction). The frictional resistance to sliding is the horizontal com- ponent of R and, as can be seen from the triangle of forces in Fig. 3.1, equals Nian ¢ where N equals the normal force on the surface of sliding (in this ease N=W). As a only achieves the value ¢ when sliding occurs, itis seen that the frictional resistance is not constant and varies with the applied load until ‘movement occurs. The term tan is known as the coellicient of frition, 3.2 Complex stress ‘When a body is acted upon by external forces then any plane within the body will be subjected to a stress that is generally inclined to the normal to the plane. Such a stress has both a normal and a tangential component and is, known as a compound, or complex, stress (Fig, 3.2) Principal plane ‘A plane that is acted upon by a normal stress only is known as a principal plane, there is no tangential, or shear, stress present. AAs is seen in the next section dealing with principal stress, only three principal planes can exist in a stressed mass. Principal stress ‘The normal stress acting on a principal plane is referred to as a principal stress. At every point in a soil mass, the applied stress system that exists can be resolved into three principal stresses that are mutually orthogonal. The principal planes corresponding to these principal stresses are called the major, intermediate and minor principal planes and are so named from a consid- eration of the principal stresses that act upon them. The largest principal stress, 1, is known as the major principal stress and acts on the major principal plane. Similarly the intermediate principal stress, o2, acts on the intermediate principal plane whilst the smallest principal stress, o3, called the minor principal stress, acts on the minor principal plane. Critical stress values and obliquities generally occur on the two planes normal to the inter- ‘mediate plane so that the effects of 2 can be ignored and a two-dimensional solution is possible. SOSSASAN = deb bbdd + —, Fig. 32 Complex stres.84 Elements of Soil Mechanies 3.3 The Mohr circle diagram Figure 3.3a shows a major principal plane, acted upon by @ major prin- cipal stress, 0}, and @ minor principal plane, acted upon by a minor principal stress, 03. By considering the equilibrium of an clement within the stressed mass (Fig. 3.3b) it can be shown that on any plane, inclined at angle # to the direction of the major principal plane, there is a shear stress, +, and @ normat StFeS8, du, The magnitudes of these stresses are: © sin 20 (01 — acoso ‘These formutae lend themselves to graphical representation, and it can be shown that the locus of stress conditions for all planes through 2 point is @ circle (generally called a Mohr circle). In order to draw a Mohr circle diagram, © specific convention must be followed, all normal stresses Gnchuding principal stresses) being plotied along the axis OX while shear stresses are plotted along the axis OY, For most cases the axis OX is horizontal and OY is, vertical, but the diagram is sometimes rotated to give correct orientation. The convention also assumes that the direction of the major principal stress parallel to axis OY, ic. the direction of the major principal plane is parallel axis OX. ‘To draw the diagram, first lay down the axes OX and OY, then set off OA and OB along the OX axis to represent the magnitudes of the minor and major principal stresses respectively, and finally construct the circle with diameter AB. This circe is the locus of stress conditions for all planes passing through the point A, ic. 2 plane passing through A and inclined to the major principal plane at angle # cuts the circle at D, The co-ordinates of the point D are the normal and shear stresses on the planc (Fig. 3.4). Major prindipal plane Minor psinapal plane 6) Fig. 3.3. Stress induced by two principal sureses, oy and os, ox a plane inclined at P10 2).Shear Sirengih of Soils 8S be fra) Poe Fig. 34 Mohr circle diagram. Proof Normal stress = 0 = OF A + AE = 03 + AD os + ABeos?# 3+ (e1~ as)eot # DE = DCsin 80" ~ 29) = DCsin 29 a8 52 sin 2e In Fig, 3.4, OF and DE represent the normat and shear stress components ‘of the complex stress acting on plane AD. From the triangle of forces ODE it can be seen that this complex stress is represented in the diagram by the Tine OD, whilst the angle DOB represents the angle of obliquity, a, of the resultant stress on plane AD, Shear stress = Limit conditions Ithas been stated that the maximum shearing resisiance is developed when the angie of obliquity equals its limiting value, &. For this condition the line OD. ‘becomes a tangent to the stress circle, inclined at angle 6 to axis OX (Fig. 3.5). ‘An interesting point that arises from Fig. 3.5 is that the failure plane is not the plane subjected to the maximum value of shear stress. The criterion of Daaxirsir she sess ° a 5 Fig. AS Mohr circle disgram for lit shoar tesintance86 Elements of Soil Mechanics failure is maximum obliquity, not maximum shear stress. Hence, although, the plane AF in Fig. 3.5 is subjected to a greater shear stress than the plane AD, it is also subjected to a larger normal stress and therefore the angle of obliguity is tess than on AD which is the plane of failure Strength envelopes 1761s assumed constant for a certain material, then the shear strength of the ‘material can be represented by a pair of lines passing through the origin, O, at angles +4 and ~9 to the axis OX (Fig. 3.6). These lines comprise the Mohr sifength envelope for the material In Fig. 3.6 a state of stress represented by circle A is quite stable as the circle lies completely within the strength envelope. Circle B is tangential to the strength envelope and represents the condition of incipient faiure, since 2 slight increase in stress values will push the circle over the strength envelope and faiture will occur. Circle C cannot exist as it is beyond the strength envelope. Relationship between @ and 8 In Fig, 3.7, /DCO= 180" 28, Fig 6. Mobe strength envelope Fig. 3.7. Relationship between g and @Shear Strength of Soils 87 In Uiangle ODC: {DOC = 6, ODE = 90°, OCD ~ 180? — 24. These angles summate to 180°, ic. +90" + 180" — 29 = 180° hhenee So ase Seas EXAMPLE 3.4. On a faiture plane in a purely frictional mass of dry sand the total stresses at failure were: shear = 3.5kN/m?, norma! = 10.0kNim? Determine (a) by caiculstion and (b) graphically the resultant stress on the plane of failure, the angle of shearing resistance of the soil, and the angle of Inclination of the failure plane to the major principal plane. Solution (@) By calculation ‘The soil is frictional, therefore the strength envelope must go through the origin, The failure point is represented by point D in Fig. 38a with coor. dinates (10, 3.5) Resultant stress = OD = V3.5" + 10? = 10.6kN/m? 33a tang = p= 035 = 1917) 450 sarap Sas 8 (&) Graphically ‘The procedure (Fig. 3.86) is first to draw the axes OK and OY and then, toe suitable seale, set off point D with co-ordinates (10, 3.5); join OD (this is the strength envelope). The stress circie is tangential to OD at the point D; draw line DC perpendicular to OD to cut OX in C, C being the centre of the circle Fig 38° Sxample 3.488 Elements of Soil Mechanics With centre Cand radius CD draw the circle establishing the points A and Bon the x-axis, By scaling, OD = resultant stress = 55" (0.6kNim?. With protractor, ¢ 19" @ Note From the diagram we see that OA = 03 = 7.6KN/m? OB = a = 1SkN/m? 3.4. Cohesion Itis possible to make a vertical cut in silts and clays and for this cut to remain standing, unsupported, for some time. This cannot be done with a dry sand which, on removal of the cutting implement, will slump until its slope is equal to an angle known as the angle of repose. In silts and clays, therefore, some other factor must contribute to shear strength. This factor is called cohesion and results from the mutual attraction existing between fine particles that tends to hold them together in a solid mass without the application of external forces. In terms of the Mohr diagram this means that the strength envelope for the soil, for undrained conditions, no longer goes through the origin but intercepts the shear stress axis (see Fig. 3.9). The value of the intercept, to the same scale as op, gives a measure of the unit cohesion available and is given the symbols ¢ or cy. 3.5 Coulomb’s law of soil shear strength It can be seen that the shear resistance offered by a particular soil is made of the two components of friction and cohesion, Frictional resistance does not hhave a constant value but varies with the value of normal stress acting on the shear plane whereas cohesive resistance has a constant value which is inde- pendent of the value of 7», In 1776 Coulomb suggested that the equation of the strength envelope of a soil could be expressed by straight line equation: noe+otang Fig. 39 A cohesive coi, subjected to undrained conditions and zero total normal stress will sll exhibit a shear stress, eyShear Strength of Soils 89 where 77 = shear stress at failure, ie. the shear strength = unit cohesion = total normal stress on failure plane @ = angle of shearing resistance ‘The equation gave satisfactory predictions for sands and gravels, for which it was originally intended, but it was not so successful when applied to silts and clays. The reasons for this are now well known and are that the drainage conditions under which the soil is operating together with the rate of the applied loading have a considerable effect on the amount of shearing resis- tance the soil will exhibit. None of this was appreciated in the 18th century and this lack of understanding continued more or less until 1925 when ‘Terzaghi published his theory of effective stress. Note It should be noted that there are other factors that affect the value of the angle of shearing resistance of a particular soil. They include the effects, of such items as the amount of friction between the soil particles, the shape of the particles and the degree of interlock between them, the density of the soil, its previous stress history, ete. Effective stress, 0” ‘The principle of effective stress was introduced in Chapter 2, Terzaghi first presented the concept of effective stress in 1925 and, again in 1936, at the First, International Conference of Soil Mechanics and Foundation Engineering, at Harvard University, He showed, from the results of many soil tests, that when an undrained saturated soil is subjected to an increase in applied normal stress, Ac, the pore water pressure within the soil increases by Au, and the value of Au is equal to the value of Ao. This increase in u caused no ‘measurable changes in either the volumes or the strengths of the soils tested and Terzaghi therefore used the term neutral stress to describe u, instead of the now more popular term pore water pressure. Terzaghi concluded that only part of an applied stress system controls measurable changes in soil behaviour and this is the balance between the applied stresses and the neutral stress. He called these balancing stresses, the effective stresses. If soil mass is subjected to the action of compressive forces applied at its boundaries then the stresses induced within the soil at any point can be estimated by the theory of elasticity, described in Chapter 4. For most soil problems, estimations of the values of the principal stresses, a}, 0) und a3 acting at a particular point are required. Once these values have been obtained, the values of the normal and shear stresses acting on any plane through the point can be computed At any point in a saturated soil each of the three principal stresses consists of two parts: (1) _u, the neutral pressure acting in both the water and in the solid skeleton in every direction with equal intensity; (2) the balancing pressures (o1 — u), (#2 ~ u) and (23 ~ u).0 Elements of Soil Mechanics As explained above, Terzaghi’s theory is that only the balancing pressures, ic. the effective principal stresses, influence volume and strength changes in saturated soils Principal effective stress = Principal normal stress ~ Pore water pressure a} =o ~u ete where the prime represents ‘effective stress: Terzaghi explained that if a saturated sol fils by shear, the normal stress, on the plane of failure, g, also consists ofthe neutral stress, u, and an effective stress which led to the equation known to all soils engineer This equation has stood the test of time and is accepted as applicable to all saturated soils. The problem of an effective stress equation for unsaturated soils is discussed in Chapter 12. 3.6 Modified Coulomb's law Shear strength depends upon effective stress and not total stress. Coulomb's, equation must therefore be modified in terms of effective stress and becomes: mae +o'tand’ where ©’ = unit cohesion, with respect to effective stresses a = effective normal stress acting on failure plane @ = angle of shearing resistance, with respect to effective stresses. It is seen that, dependent upon the loading and drainage conditions, it is possible for a clay soil to exhibit purely frictional shear strength (ic. to act as a‘c’ =0' or “9” soil, when itis loaded under drained conditions or to exhibit only cohesive strength (ic. to act as a °} = 0" or ‘c," soil) when it is loaded under undrained conditions. (See Example 3.8, Fig. 3.26.) Obviously, at an interim stage the clay can exhibit both cohesion and frictional resistance (ie. to act asa‘e’ ~ $” soil). The same situation also applies to granular soils 3.7 The Mohr—Coulomb yield theory Over the years various yield theories have been proposed for soils. The best known ones are: the Tresca theory, the von Mises theory, the Mohr-Coulomb theory and the critical state theory. The first three theories have been described by Bishop (1966) and the critical state theory by Schofield and Wroth (1968). Only the Mohr-Coulomb theory is discussed in this chapter. The theory does not consider the effect of strains or volume changes that a soil experiences onits way to failure nor does it consider the effect of the intermediate principalShear Strength of Soils 91 stress, a2. Nevertheless satisfactory predictions of soil strength are obtained and, as itis simple to apply, the Mohr-Coulomb theory is widely used in the analysis of most practical problems which involve soil strength. The Mohr strength theory is really an extension of the Tresca theory which, in turn, was probably based on Coulomb's work, hence the title. The theory assumes that the difference between the major and minor principal stresses is a function of theit sum, ie. (oy ~ 03) = f(e1 +03). Any effect due to 072 is ignored. The Mohr circle has been discussed earlier in this chapter and a typical example of a Mohr circle diagram is shown in Fig. 3.10. The intercept on the shear stress axis of the strength envelope is the intrinsic pressure, ie. the strength of the material when under zero normal stress. As we know, this intercept is called cohesion in soil mechanics and given the symbol Fig. 3.10, In Fig. 3.10: ee Deen Glee) a =03 OC kyoto) tkrata Hence 0, — 03 = dksing + (0 + 03)sing Now ccote (o1 = 05) = 2eeosd + (01 + 95)sing which is the general form of the Mobr-Coulomb theory. The equation can be expressed in terms of either total stress (as shown) or effective stress: (0 ~ 04) = 2c cose! + (04 + 04) sing! 3.8 Determination of the shear strength parameters The shear strength of a soil is controlled by the effective stress that acts upon it and it is therefore obvious that a geotechnical analysis involving the operative strength of a soil should be carried out in terms of the effective2 3.8.1 Elements of Soil Mechanics stress parameters and c. This isthe general rule and, as you would expect, there is at least one exception. The case of a fully saturated clay subjected to undrained loading is much more simple to analyse using total stress values and g, and c than with an effective stress approach. As willbe illustrated in later chapters, such a situation can arise in both slope stability and bearing capacity problems. [tis seen therefore that both the values ofthe undrained parameters, 6y and x (or d and c), and of the drained parameters, ¢! and ¢ (or gy and ¢4) are generally required, They are obtained from the results of laboratory tests carried out on representative samples of the soil with loading and drainage conditions approximating to those in the field where possible. The tests in general use are the direct shear box test, the triaxial test and the unconfined compression test, an adaption of the triaxial test. The direct shear box test The apparatus consists ofa brass box, split horizontally at the centre of the soil specimen. The soil is gripped by perforated metal grilles, behind which porous dises can be placed if required to allow the sample to drain (see Fig. 3.11) ‘The usual plan size of the sample is 60 x 60mm?, but for testing granular ‘materials such as gravel or stony clay itis necessary to use a larger box, gener- ally 300 x 300mm? although even greater dimensions are sometimes used. A vertical load is applied to the top of the sample by means of weights. ‘As the shear plane is predetermined in the horizontal direction the vertical load is also the normal load on the plane of failure. Having applied the required vertical load a shearing force is gradually exerted on the box from an electrically driven screwjack. The shear force is measured by means of a load transducer connected to a computer. By means of another transducer (fixed to the shear box) it is possible to determine the strain of the test sample at any point during shear: Strain = Movement of box Length of sample The loud reading is taken at fixed displacements, and failure of the sol specimen is indicated by a sudden drop in the magnitude of the reading or a levelling off in successive readings. In most cases the computer plots graph rome Upper hat of Box _ Porous ise <> sheartorce of. f= ee | sheartorce Porous die Lower nat of box Fig. 341 Diagrammatic sketch ofthe shear box apparatus.Shear Strength of Soils 93 of the shearing foroe against strain as the test continues, Failure of the soil is, visually apparent from a turing point in the graph, ‘The apparatus can be used for both drained and undrained tests, although undrained tests on silts and sands are not possible. EXAMPLE 3.2 ‘Undyained shear box tests were carried out on a seties of soil samples with the following results: Test ne, ‘Total normal ‘Total shear stress ‘sess ‘a failuce (eNfa) GNI) 1 100 98 2 200 9 3 300 340. 4 400 me Determine the cohesion and the angle of friction of the soil, with respect to. total stress. Solution Jn this case both the normal and the shear stresses at feilure are known, so there is no need to draw stress circles and the four failure points may simply be plotied. These poinis must lie on the strength envelope and the best straight fine through the points will establish it (Fig. 3.12). From the plot, cy = 55kNim% , = 23", 3 te Shear strass Gin?) Fig. M2 Example 32, 100 20 “300 a0 Normal sts (kN?) EXAMPLE 33 The following results were obtained from an undrained shear box test carried out on a set of undisturbed soil samples:4 Elements of Soil Mechanics Normal load (kN) 02 oa os. Strain (%) Shearing force (N) 0 0 ° ° 1 2 3 45 2 46 n 101 3 0 0 138 4 9 139 203 5 107 164 2as 6 121 180 26 7 BI 192 308 8 16 201 330 8 138 210 351 0 138 217 370 " 1B? 28 391 2 136 20 402 a 24 410 a 231 aig 1s 26 416 6 47 1 417 8 ais ‘The cross-sectional area of the box was 3600 mm? and the test was carried out in a fully instrumented shear box apparatus. Determine the strength parameters of the soi in terms of total stress Solution ‘The plot of load transducer readings against strain is shown in Fig. 3.13a, From this plot the maximum readings for normal loads of 0.2, 0.4 and 0.8 kN were 138, 237 and 417kN. For this particular case the maximum readings could obviously have been obtained directly from the tabulated results, but viewing the plots is some- times useful to demonstrate whether one of the sets of readings differs from the other two. ‘The shear stress at each maximum load reading is calculated, Normal load Normal stress Shear force «Nn (Nim) Oo) 02% 108 0.138 x 108 a 600 na 3 04 im a0 6 08, m ar U6 The plot of shear stress to normal stress is given in Fig 3.13b.Shear Strength of Soils 95 450 400] 350] 300 280 200] 150] 100] : piu 02 4 6 8 1 12 14 16 18 20 ) Fig. 313 Example 3.3 The total stress envelope is obtained by drawing a straight line through the three points. The strength parameters are: dy = 25°; cy = 13kN/m?, 3.8.2 The triaxial test As its name implies this test (Fig, 3.14) subjects the soil specimen to three compressive stresses at right angles to each other, one of the three stresses being increased until the sample fails in shear. Its great advantage is that the plane of shear failure is not predetermined as in the shear box tes ‘The soil sample tested is cylindrical with a height equal to twice its diameter. In the UK the usual sizes are 76mm high by 38mm diameter and 200 mm high by 100 mm diameter. ‘The test sample is first placed on the pedestal of the base of the triaxial cell and a loading cap is placed on its top. A thin rubber membrane is then placed over the sample, including the pedestal and the loading cap, and made water- tight by the application of tight rubber ring seals, known as "rings, around the pedestal and the loading cap. ‘The upper part of the cell, which is cylindrical and generally made of perspex is next fixed to the base and the assembled cel is filled with water. ‘The water is then subjected to a predetermined value of pressure, known as96 Elements of Soil Mechanics Rent raneducer Cell pressure e 98 me —Sampie a HA) To pore pressure 2 =] measunng apperans Weare, ifrogutes Fig. $14 The trawl apparatus, the eall pressure, which is kept constant throughout the length of the test, tis this water pressure that subjects the sample to an all-round pressure. ‘The additional axial stress is created by an axial load applied through a toad transducer, in a similar way (0 that in whieh the horizontal sheat force is applied in the shear box apparatus. By the action of an electric motor the axial load is gradually inereased it a constant rate of strain and as the axial load is applied the sample suffers continuous compressive deformation. The amount of this vertical deformation is obiained from a deformation. trans- ducer. Throughout the test, until the sample fails, readings of the deformation transducer and corresponding readings of axial toad are taken, With this data the computer plots the variation of the axial load on the sample against its vertieal strain, Determination of the additional axial stress From the load transducer it is possible at any time during the test to determine the additional axial load that is being applied 0 the sample, During the application of this load the sample experiences shortening in the vertical direction with 3 corresponding expansion in the horizontal direction ‘This means that the cross-sectional area of the sample varies, and it has been found that very hitle error is introduced if the cross-sectional urea is evaluated fon the assumption that the volume ofthe sample remains unchanged during the test. In other words the cross-sectional area is found from: Volume of sample Griginai tength — Vertical deformation ‘Cross-sectional area = Principal stresses The intermediaie prineipal stress, oz, and the minor prineipal stress, 03, ate equal and are the radial stresses caused by the cell pressure, p,. The major principal stress, 01, consists of two parts: the cell water pressuse acting on the ends of the sample and the additional axial stress from the load trans- ducer, q. To ensure that the cell pressure acts over the whole area of the end ‘cap, the bottom of the plunger is dried so that the pressure can act on the ball seating.Shear Strength of Solis 97 . > ‘ ! i wn ig, MIS Stresses in the triaxial test. From this we sce that the triaxial test can be considered as happening in two stages (Fig. 3.15), the first being the application of the cell water pressure (pe, Le. 43), while the second is the application of a deviator siress (@. ie. for ~ ob. ‘A set of at least three sumples is tested. The deviator stress is plotted tagainst vertical strain and the point of failure of euch sample is obtained, The Mohr circies for each simpie are then drawn and the best common tangent to the circles is taken as the strength envelope (Fig. 3.16). A small curvature ‘occurs in the strength envelope of most sols, but this effect is slight and for all practical work the envelope can be taken as a straight line, Types of failure Not all soil samples will fail in pure shear: there are generally some barrelling effects as well. In a sample that fails completely by burrelling there is no oy-600 "E 3 oy-400 Fi oye 200 ‘Stain Fig. 346 Typical tcixial tes results “I OO GI Shear Barreting Barreting ana staat Fig. 317 Types of fafure im the eriauia testements of Soil Mechanics definite failure point, the deviator stress simply increasing slightly with strain. In this case an arbitrary value of the failure stress is taken as the stross value at 20 per cent strain (see Fig. 3.17). Note In the past, soil laboratories made use of dial gauges to measure dis- placement, and proving rings to measure applied leads. Some laboratories still use such equipment and any reader interested in an explanation and examples of their use is guided to earlier editions of this book 38.3. The unconfined compression test This is @ special case of the triaxial test in which the all-round pressure on the sample is zero and no rubber membrane is necessary to encase the specimen, (Fig. 3.18). 11 is often used as a simple field test, but can only be used for cohesive soils. The test specimen is loaded through a calibrated spring by a simple imanually-operated screwjack at the top of the machine. In order to test soils, of varying strengths a range of springs is supplied, generally with stfinesses in the order of 2, 4, 8 and 16Némm extension, By means of an autographic revording arm the graph of load against deformation is drawn directly on to @ sheet of paper. Knowing the vertical deformation, the area of the sample at failure ean be obtained and hence the stress; in common practice the cohesion ‘ofthe soil is taken to be one half of the stress at failure (i.e, dy is assumed = 0), Rotading handle for applying compression load reas Autographic ‘compression recording arm ‘oad Fig. 3418 The unconfined compression apparats.