Naa) 12 Shear Strength of Soil Introduction ee Ez . shear stent paaneter + fet of emlding on shear strength of cohesive sis + Eifect of variation of shear stensth depending on the direction of load appicaion + Us cE to obtain shear rength of saturated cohesive soils Mohr (1900) presented a theory for ruptere in materials that contended that a ma- terial fils hocause ofa critical combinatien of normal stress and shearing stress and not from either maximum normal or shear stres alone, Thus the functional elation. ship between normal stress and shear stress on failure plane ean be expressed in the following form: =f) (21) 469 in liquid form have virtually no shear strength and even when silid have shear strengths of relatively small magnitudes compared with those exhitited by steel or concrete.ws hs = HED R a (=9 atsiiding) (a) No horizontal (b) Horizontal force applied force applied Fig. 3.1. Friction,470 Chapter 12 | Shear Strength of Soil The failure envelope defined by Eq. (12.1) is a curved line. For most soil me~ chanics problems, itis sufficient to approximate the shear siress on the failure plane as a linear function of the normal stress (Coulomb, 1776). This linear function can be written as where c= cohesion 4 = angle of internal friction = normal stress on the failure plane (Ga skearsteney ‘The preceding equation is called the Mohr-Coulomb failure criterion. the total normal stress at a point is the sum of the effective stress (o") and pore water pressure (1), oF The effective stress a’ is carried by the soil solids. The Mohr—Coulomb failure crite- expressed in terms of effective stress, will be of the form. ‘Thus, re tease : 2.3) are expressions a strength based on total stress anc effective stress respectively. The value of c’ for sand and inorganic silt is 0. For normally consolidated clays, c” can be approximated at 0. Overconsolidated clays have values of c’ that are greater than 0. The angle of friction, ¢, is sometimes referred to as the drained angle of friction. Typical values of 6" for some granular soils are given in Table 12.1. rion, Table 12.1 Typical Values of Drained Angle of Friction for Sands and Sills, type ides) CBaRownted grains Loose 20) Medium 3-35 Dense 35.38) Sand: Angular grains Loose Medium Dense Gravel with some sand SitesDevelopment of a mechanical model for soil 5.3 Simple models of soil plasticity Soil as a frictional material The use of elastic models alone (as desesibed previously) implies that the soil is infinitely strong. If at a point on any plane within a soil mass the shear stress becomes equal to the shear strength of the soil, then failure will oceur at that point Coulomb originally proposed that the limiting strength of soils was frictional, imagining that if slip (plastic failure) occurred along any plane within an element of closely packed particles (soil), then the slip plane would be rough due to all ofthe individual particle-to-particle contacts, Friction is commonly described by: T=uN where Tis the limiting frictional force, Nis the normal force acting perpendicular to the slip plane and is the coefficient of friction. This is shown in Figure 5.5(a). In an element of soil, itis mote useful to use shear stress and normal stress insted of Tand N tang) where tan is equivalent to the coefficient of friction, which is an intrinsic material property related to the roughness of the shear plane (te. the shape, size and angularity of the soil particles). The frictional relationship in terms of stresses is town in Figure 5.5(b) ‘While Coulomb's frictional model represented loosely packed particle arrangements well, if the parti- cles are arranged in a dense packing then additional intial interlocking between the particles can cause the frictional resistance r, to be higher than that predicted considering frietion alone. Ifthe normal stress is increased, it can become high enough that the contact forces between the individual particles cause particle breakage, which reduces the degree of interlocking and makes slip easier. At high normal stresses, therefore, the interlocking effect disappears and the material behaviour becomes purely fiic- tional again. This is also shown in Figure 5.50). In accordance with the principle that shear stress in a soil can be resisted only by the skeleton of solid particles and not the pore water, the shear strength (,) of a soil at a point on a particular plane is nor- mally expressed as a function of effective nonnal stress (o) rather than total stress. Friction + interlocking @ o nal strength along a plane of Particles along a plane of slip. (b) strength of an assembly of 152— WDA = Luda y = = Fig. 3.2 Complex stress, Major principal plane Minor principal plane ) Fig.3.3 Stress induced by two principal stresses, a, andl a, on a plane inclined at @to a,Smt’ Elements of Sil Mechanics hear = 3.5 KPa; normal = 10.0 kPa, Determine () by calculation and (b) graphically the resultant stress on the Plate oF failure, the angle of shearing resistance of the soil, and tho angle of Inclination ofthe failure plane to the major principal plane Solution (a) By calculation Tie sol is fictional, therefore the sirength envelope must go through the origin. The failure point is represented by point D in Fig. 3.8a with coordinates (10,3:5), Resultant stress = OD = V3.5? + 10? = 106 kPa 35 ng= 25-03 tn g= 77 035 395193 0-459 546° (©) Graphically ‘The procedure (Fig. 3.8b) is first to draw the axes OX and OY and thea, to @ suitable scale, set off point D with coordinates (10, 3.5): join OD (thi i the ‘Brzneth envelope) The stress circle i tangential io OD atthe point D: drav line DC perpendicular to OD to cut OX in C, C being the centre of the circ70 5 q % Fig-38 Bample3.1 With centre C and radius CD draw the circle establishing the points A and B ‘onthe x-axis. By scaling, OD = resultant stress = 10.6 kPa, With protractor, 8= 19%; 5° ‘Note From the diagran we see that OA=6,=7.6kPa 15kPa 3.4 Cohesion ' | fe, some oe arr mast | COO 1 shear strength jad ess from the ‘mutalaractonexitingbetween fn particles that ends o hol the together in 8 4 solid mass without te application of extemal forces. In terms ofthe Mohr diagram { thismeans hat the stenginenvelope forthe sol or unained conditions, no longer | goes though he origin bc inercepsthe shear stress axis (see Fig 3.9) The aloof | — ¢ Fig.3.9 A cohesive col subjected to undtaned conditions and zero total normal ses wil il shit shearstess,Soil behaviour in shear The Mohr—Coulomb model As described in Section 5. the stae of stress in an element of soil is defined in terms of the normal and shear stresses applied to the boundaries of the soil element. States of stress in two dimensions can be represented on a plot of shear stress (r) against effective normal stress (7. The stress state for a 2-D clement of soil can be represented ether by a pair of points with coordinates (! t) and (02, ty). ot by a Sage tein’ by ys shown in Figure 5.6. The stress oi end of a diameter tlxough a Mobr cisele at an angle of 2010 the horizontal represent the stress conditions on a plane at an angle of @to the minor principal stress. The circle therefore represents the stress states on all possible planes within the soil element. The principal stress components alone are enough to fully describe the position and size of the Mohr circle and so are often used t0 describe the stress state, as it reduces the number of stress variables from three (a, , f,) to two (gj, 03). When the Figure 5.6 Mohr-Coulomb failure criterion, 153element of soil reaches failure, the circle will just touch the failure envelope, at a single point. The failure envelope is defined by the fietional model described above: however, it can be difficult to deal ‘with the non-linear past of the envelope associated with interlocking, so that it is eommoa practice to approximate the failure envelope by a straight line described by: P e ‘where c' and ¢” are shear strength parameters referred to as the cohesion intereept and the angle of shearing resistance, respectively, Failure will thus occur at any point in the soil where a critical combi- nation of shear stress and effective normal stress develops. It should be appreciated that c’ and ¢ are simply mathematical constants defining a linear relationship between shear strength and effective normal stress. Shearing resistance is developed mechanically due to inter-particle contact forces and friction, as described above: therefore, if effective normal stress is zero then shearing resistance must also be zer0 ‘(ualess thete is cementation or some other bonding between the particles) and the value of c” would be eco. This point is crucial to the interpretation of shear strength parameters (described in Section 5.4), ‘A siate of stress represented ty a stress point that plots above the failure envelope, or by @ Mohr circle, part of which lies above the envelope, is impossible With reference to the general case with c’>0 shown in Figure 5.6, the relationship between the shear strength parameters and the effective principal stresses at failure at a particular point can be deduced, ‘compressive stress being taken as postive. The coordinates of the tangent point are reand of where For a given state of stress 115 apparent that, because o{=9;—u and o{=9,-u, the Mohr exrles tor total ‘and effective stresses have the same diameter bat their centres are separated by the corresponding pore water pressure 1, as shown in Figure 5.7, Similarly, total and effective stress points are separated by the value of 154Soil behaviour in shear Mobt-Coulomb Moh circle of| Mohr circle of total effective stresses stresses Effect of drainage conditions on shear strength The shear strength of a soil under undrained conditions is different from that under drained conditions The failure envelope is defined in terms of effective stresses by 9 and c’, and so isthe same irrespective of whether the soil is under drained or undrained conditions; the difference is that under a given set of applied total stresses, in undrained loading excess pore pressures are generated which change the effec- tive stresses within the soil (ander drained conditions excess pore pressures are zero as consolidation is complete). Therefore, two identical samples of soil which are subjected to the same changes in total stress but under different drainage conditions will have different intemal effective stresses and therefore different strengths according to the Mohr-Coulomb criterion. Rather than have to determine the pore pressures and effective stresses under undrained conditions, the umdrained strength can be expressed in terms of total stress. The failure envelope will sil be linear, but will have a different gradient and inter- cept: 8 Mohr-Coulomb model can therefore still be used, but the shear strength parameters are different and denoted by ¢, and ¢, (=0, see Section 5.6), with the subseripts denoting undrained behaviour. The drained strength is expressed directly in terms of the effective stress parameters c’ and g described pre- viously. ‘When using these strength parameters to subsequently analyse geotechnical constructions in practice, the principal consideration isthe rate at which the changes in foal stress (due to construction operations) are applied in relation to the rate of dissipation of excess pore water pressure (consolidation), which in tum is related to the permeability of the soil as described in Chapter 4. loading may be fast enough to ined response in coarse-grained material is taken to be synonymous with whe Tonge’ sl eles othe ds i of he Gaston sly 155 generate an un12.8 Inclination of the Plane of Failure Caused by Shear 4: «| y 7 oe 6 2 4 Moht-Coutom faite criteria ’ te ¢ * Tetive noma eso © Figure 12.1 Mohr-Coulomb failure eriterion The significance of Eq, (12.3) can be explained by referring to Figure 12.1, which, shows an elemental soil mass. Let the effective normal stress and the shear stress on the plane ab be o” and 7, respectively. Figure 12.1b shows the plot of the failure envelope defined by Eq, (12.3). If the magnitudes of o and 7 on plane ab are such that they plot as point A in Figure 12.1b, shear failure will not occur along the plane. If the effective normal stress and the shear stress on plane ab plot as point B (which falls on the failure envelope), shear failure will occur along that plane. A state of stress on a plane represented by point C cannot exist, because it plots above the failure envelope, and shear failure in a soil would have occurred already. Inclination of the Plane of Failure Caused by Shear Asstated by the Mohr—Coulomb failure criterion, failure from shear will occur when the shear stress on a plane reaches a value given by Eq, (12.3). To determine the incli- nation of the failure plane with the major principal plane, refer to Figure 12.2, where 7 and o} are, respectively, the major and minor effective principal stresses. The fail- ure plane EF makes an angle 9 with the major principal plane. To determine the angle @ and the relationship between oj and , refer to Figure 12.3, which is a plot of the Mohr's circle for the state of stress shown in Figure 12.2 (see Chapter 10). In Figure 12.3, gi is the failure envelope defined by the relationship 7, = c’ + 0” tan. The radial line ab defines the major principal plane (CD in Figure 12.2), and the radial line ad defines the failuce plane (EF in Figure 12.2). It can be shown that Lbad = 2 = 90 + $',or (2 anFig.3.10 Mohieircle diagram 472 Chapter 12 | Shear Strength of Soll 4 # F vy © o> este normal se Figure 122 Inclination of failure plare in soil with major prneipal plane Figure 12.3 Mohe’scirele and failure envelopeThere are several laboratory methods availabe to determine the shear strength pa rameters (ie..¢, 6. €..6") of Various soil specimens in the laboratory. They are as follows rain triaxial tes Torsional ring she The direct shear test and the triaxial test are the two commonly used techniques for determining the shear strength parameters These two tests will be described in detail in the sections that follow (EEE) Direct Shear Test The direct shear test is the oldest and simplest form of shear test arrany A diagram of the direct shear test apparatus is shown in Figure 12.4. The test equipment consists of a metal shear box in which the soil specimen is place meat. The soil specimens may be square or circular in plan. The size of the specimens generally used is about SL mm X SL mm or 102 mm % 102 mm (in. X 2 i. or 4 in. X 4 im.) across and about 25 mm (1 in.) high. The box is split horizon: ly into halves. Normal force on the specimen is applied from the top of the shear box. The normal stress on the specimens can be as great as 1050 kNim (150 Thin.) Shear force is applied by moving one-balf of the box rel other to eause failure in the soil specimen, Depending on the equipment, the shear test can be either stress controlled or strain controlled, In stess-controlled tests, the shear force is applied in equal incre ‘ments until the specimen fails The failure oecirs along the plane of split of the shear box. After the application of each incremental load, the shear displacement of the top half of the box is measured by a horizontal dial gauge. The change in the height474 Chapter 12 | Shaar Strength of Soil Loading plate Bi] Porous stone owe 2 aiiidiiiieameD of the specimen (and thus the volume change of the specimen) during the test can be obtained from the readings of a dial gauge that measures the vertical movement of the upper loading plate. In strain-controlled tests, a constant rate of shear displacement is applied to ‘one-half of the box by a motor that acts through gears. The constant rate of shear displacement is measured by a horizontal dial gauge. The resisting shear force of the soil corresponding to any shear displacement can be measured by a horizontal proving ring or load cell. The volume change of the specimen during the test is ob- tained in a manner similar to that in the stress-controlled tests. Figure 12.5 shows a photograph of strain-controlled direct shear test equipment. ‘The advantage of the strain-controlled tests is that in the case of dense sand, peak shear resistance (that is, at failure) as well as lesser shear resistance (that is, at a point after failure called ultimate strength) can be observed and plotted. In stress-controlled tests, only the peak shear resistance can be observed and plotted. Note that the peak shear resistance in stress-controlled tests can be only approx- imated because failure occurs at a stress level somewhere between the prefai ure load increment and the failure load increment. Nevertheless, compared with strain-controlled tests, stress-controlled tests probably model real field situations better. Normal force (CSNOTMALSHEES = Sopatarce often “The fisting) SHGF SEES for any shear displacement can be calculated as = snes - Resisting shear force Cos ‘ectional area of the sp.125 Direct Shear Test 475 Figure 12.5 Strain-controlled direct shear equipment (Courtesy of Braja M. Das, Henderson, Nevada) Figure 12.6 shows a typical plot of shear stress and change in the height of the specimen against shear displacement for dry loose and dense sands. These observa~ tions were obtained from a strain-controlled test. The following generalizations can be developed from Figure 12.6 regarding the variation of resisting shear stress with shear displacement: 1, In loose sand, the resisting shear stress increases with shear displacement until 4 failure shear stress of 7, is reached. After that, the shear resistance remains approximately constant for any further increase in the shear displacement. 2. Indense sand, the resisting shear stress increases with shear displacement until it reaches a failure stress of 1» This 7; is called the peak shear strength. After failure stress is attained, the resisting Shear stress gradually decreases as shear displacement increases until it finally reaches a constant value called the ulti- ‘mate shear strength Since the height of the specimen changes during the application of the shear force (as shown in Figure 12.6), it is obvious that the void ratio of the sand changes (at least in the vicinity of the split of the shear box). Figure 12.7 shows the nature of Variation of the void ratio for loose and dense sands with shear displacement. At476 Chapter 12 | Shear Strength of Sol Peak shea suength, Shear stess ‘Shear displacement Es eg B Figure 126 Plot of shear 3 sfress and change in height fg ‘of specimen against shear 2 g| displacement for loose and ae dense dry sand (direct shear test) large shear displacement, the void ratios of loose and dense sands become practi- cally the same, and this is termed the critical void ratio. It is important to note that, in dry sand, ono’ Loose sand ; Ultimate z strength 3 2 Figure 12.7 Nature of variation of void Po ratio with shearing displacement Sheceing displacementSoil behaviour in shear mate as ZN Porous (or sali) pater specims oy @ Figure 5.8 Direct shear apparatus: (a) schematic, (b) standard direct shear apparatus (image courtesy of Impact Test Equipment Ltd). The triaxial apparatus is the most widely used Inboratory device for measuring soil behaviour in shear, and is suitable for all cylindrical specimen, generally this sits within a chamber of pressurised water. The sample Jena is stead silly by long tan aly by th RS SS EE ESROHERA) GRD te aoe soo in Figs 59, The moet common tes, ‘compression, involves plying shear to the soil by holding the confining pressure constant and applying compressive axial Joad through the loading ram, The main features of the apparatus are also shown in Figure 5.9. The circular base has a central ped- estal on which the specimen is placed, there being access through the pedestal for drainage and for the measurement of pore water pressure, A Perspex cylinder, sealed between a ring and the circular cell top, forms the body of the cell. The cell top has a central bush through which the loading sam passes. The ler and cell top clamp onto the base, a seal being made by means of an O-ring, ey 157| tng | Air release valve lo |___— Perspex cylinder Top drainage i ate Loading cap sees Porous disc — fer pressure Set Deemer 3 J specimen ) Figure 5.9 The triaxial apparatus: (a) schematic, (b) a standard triaxial cell (image courtesy of Impact Test Equipment Ltd,). 158Figure 12.13 Fount seit Bl Si . The shear strength along the surface of con! ‘and the foundation can be given as y 6 +o! tans (12.19) MSGR an ton hts he sila te founaton ss Note that the preceding equation is similar in form to Eq. (12.3). ie foundation material can be placed in the bottom part of the direct shear test box ‘and then the soil can be placed above it (that isin the top part of the box),as shown, in Figure 12.14, and the test can be conducted in the usual manner. Figure 12.15 shows the results of direct shear tests conducted in this manner with @ quartz sand and concrete, wood, and steel as foundation materials, with o” = 100 kN/mv (145 Ibvin2)..Di soit Bi Foundation materiat Flow 1214 Dt sear fst deternine intra rision angleSmith's Elements of Soil MechanicsNona loadikN) Moria ress (Pa) Shearfrce(N)——_ Shear sess k 02 02 x10 56 138 pee 300 3600 os m 66 The plot of shear stresétonormal stress is given in Fi. 3.130, lope is obtained by drawing a straight line through the three points. The strength parameters are: 9, = 25°; c, = 13 kPa The tol stress env om} sal Ait sei 02 4 6 B12 14 16 18 20 @ ‘Strain (6) ‘ EUR gia ‘050100 150-200 50 o Normal sess (kPa) 5.8.2 Tatiana ttDirect shear tests were performed on a dry, sandy soil The sizeof the specimen was 2 in, x 2 in, x 0.75 in, Test results are as follows! Shear sre a alures, ane ‘Normal 520 Find the shear stross parameters, stresses, obtained from the tests are plotted against the normal stresses in Figure 12.18, from which ¢” = 0, 6” = 32° I Figure 12.1812.7 General Comments on Direct Shear Test 485 I Example 12.2 l Following are the results of four drained direct shear tests on an overconsoli- dated clay: Diameter of specimen = $0 mm Height of specimen = 25 mm ™) 1515 1999 2516 303.4 1485 Determine the relationships for peak shear strength (x)) and residual shear strength (1,). Solution Area of the specimen (A) = (7/4) ( TOO a ) 0.0019634 m?, Now the following table can be prepared. Normal No Peak shear y= Sos . Test force, stress.o” force,Sy.. 7A g we (N) kN iw) UN (Nine) 1 150 764 1575 802 “42 2S 2 290 1273 1999) 1018, 566 (288 3 390 1783 257.6 1312 129 S24 4580 280.1 3634 185.1 M45 736 The variations of and z, with o” are plotted in Figure 12.19, From the plots, we find that Peak strength: — 7{kN/m? Residual strength: +,(kNim?) 40+ 0° tan 27 o' tan 14.6, (Note: For all overconsolidated clays, the residual shear strength can be expressed as o' tan $y where 6; = effective residual friction angle.)486 Chapter 12 | Shear Strength of Sail 300 8 8 Shear stress, + (kN?) z 0 0 100150200280 30380 [Effective normal stress, a (kN?) A diagra a ixfal test [ayout is shown in triaxial test in progress in the laboratory. In this test, a soil specimen ebout 36 mm (1.4 in.) in diameter and 76 mm (3 in.) Jong generally is used. The specimen is encased by a thin rubber membrane and placed inside a plastic cylindrical chamter that usually is fled with water or glycerine. The specimen is subjected to a confinirg pressure by compression of the fluid in the chamber. \Note: Air is sometimes used as a compression medium.) 1. Application of dead weights or hydraulic pressure in equal increments until the specimen fails. (Axial deformation of the specimen resulting from the load ap- plied through the ram is measured by a dial gauge.) 2. Application of axial deformation at a constant rate by means of a geared or hydraulic loading press. Thisis a strain-controlled test. The axial load applied by the loading ram corresponding to a given axial deforma- tion is measured by a proving ring or load cell attached to the ram.81-05 5 9% t+ : +t} : ! 3 9% 1-05 0y-83 (KPa) ‘Stain Fig 916 Typical ao esos. Types of failure Notall sol samples wil fil in pure shear; there are generally some barrelling effects 4s well. In asample that fails completely by barrelling there is no definite failure point, the deviator stress simply increasing slightly with srain, In this case anarbitary value ‘of the failure stress is taken a the stress value at 20 per cent strain see Fig. 3.17). ‘Note In the past, soil laboratories made use of dil gauges to measure displacement, and proving rings to measure applied loads. Some laboratories still use such equip. iment, and any reader interested in an explanation and examples oftheir uses guided tothe 6th, orearlier, editions of this book. “AW OG Shear Barring Barrting ote12.9 Consolidated-Drained Tiaxia Test Axia load ber rng Tene Flexible tube Gi water Bl Porous ise Specimen eavtose ina uber membrane Figure 12.20 Diagram of triaxial test equipment (After Bishop and Bjerrum, 1960, with ‘permission from: ASCE) 1. Consolidated-drained test or drained test (CDtest) 2. Consolidated-undrained test (CU test) 3. Unconsolidated-undrained test or undrained test (WUhtest) ‘The general procedures and implications for each of the tests in saturated soils are described in the following sections. (25) consolidated-Drained Triaxial Test In the CD test, the saturated specimen first is subjected to an all around confining pressure, ¢,. by compression of the chamber fluid (Figure 12.224). As confining pres- sure is applied, the pore water pressure of the specimen increases by 1 (if drainage is prevented). This increase in the pore water pressure can be expressed as a nondi- mensional parameter in the form where B = Skempton’s pore pressure parameter (Skempton, 1954). 487488 Chapter 12 | Shear Strength of Soil Figure 12.24 A triaxial test in progress in the laboratory (Courtesy ofS Vanapatt University of Otawa, Canad) { { ey A0 ayo t t Figuro 12.22 Censolidated-drained “ ¢ triaxial test (a) specimen under A chamber-conflnirg pressure: (b) deviator stress application @ ®[EXD] consolidated-Undrained Triaxial Test ‘The consolidated-undrained testis the most common type of triaxial test. In this test the saturated soil specimen is frst consolidated by an all-around chamber fluid pres- sure, o;, that results in drainage (Figures 12.284 and 12.28). After the pore water pressure generated by the application of confining pressure is dissipated, the devi- ator stress, Aa, on the specimer is increased to cause shear failure (Figure 12.28e). During this phase of the test, the drainage line from the specimen is kept closed. Because drainage is not permitted, the pore water pressure, Au,, will increase. During, Figure 12.29 Triaxial soil specimen at failure during a consolidated-undrained test (Courtesy ofS Varupai, University of Otawa, Canada)Efotve sts flue envelope 5 = 0" tn 6” Total stress failure envelope 5 Shear stress aon aio Nonna wes ot Kot “au uy Figure 12.90 Total and effective stres failure envelopes for consolidated undrained triaxial tests effective stress Mohr’s circles at failure obtained from consolidated-undrained triax- ial tests in sand and normally consolideted clay, Note that A and B are two total stress Mohr’s circles obtained from two tests, Cand D are the effective stress Mohr's circles corresponding to total stress circles A and B, respectively. The diameters of circles A and Care the same; similarly, the diameters of circles Band D are the same. In Figure 12.30, the total stress failure envelope can be obiained by drawing a line that touches all the total stress Mohr’s circles. For sand and normally consoli- dated clays, this will be approximately a straight line passing through the origin and may be expressed by the equation a0 tad (12.35) ihe angle that the total stress failure envelope makes with the normal stress axis.also known as the consolidated-undrained angle of shearing resistance Equation (12.35) is seldom used for practical considerations. Similar to Eq. (12.21). for sand and normally consolidated clay, we can write (2 = <2) (12.36) a, + ;,412.41 Unconsolidated-Undrained Iiaxial Test Ez Unconsolidated-Undrained Triaxial Test In unconsolidated-undrained tests, drainage from the soil specimen is not permit- ted during the application of chamber pressure o;, The test specimen is sheared to failure by the application of deviator stress, A, and drainage is prevented. Because drainage is not allowed at any siage, the test can be performed quickly. Because of the application of chamber confining pressure o,, the pore Water pressure in the soil specimen will increase by «..A further increase in the pore water pressure (Au,) will occur because of the deviator stress application. Hence, the total pore water pressure win the specimen at any stage 0° deviator stress application can be given as uu, + Au, (12.40) From Eqs. (12.20) and (12.34),u, = Bos and Au, Akoyso. ‘This test usually is conducted on clay specimens and depends on a very impor- tant strength concept for cohesive soils if the soil is fully saturated. The added axial stress at failure (Ac,), is practically the same regardless of the chamber confining pressure. This property is showr: in Figure 12.34. The failure envelope for the total stress Mohr’s circles becomes a horizontal line and hence is called a 6 = O condition. From Eq. (12.9) with @ = 0, we get (12.42) where ¢, is the undrained shear strength and is equal to the radius of the Mohr’s, circles. Note that the = 0 concept is applicable to only saturated clays and silts. ‘Total stress Mobr's circles a faire Shear stress Normal sess Figure 42.94 Total stress Mohr’s circles and failure envelope (§ = 0) obtained from tinconsoldated-undrained triaxial tests on fully saturated cohesive soil3.8.3. The unconfined compression test in this test (Fig. 3.18) no all-roand pressure is applied to the soil specimen and the results obtained give a measure ofthe unconfised compressive strength of the soil. The testis only applicable o cohesive sols ant, although not as popular asthe triaxial test, itis used where a rapid result is required. An electric motor within Fig.3.18 The unconfined compressiontest. i supporting the specimen upwards and the load carried Pe sal secret Uy set reac eyed san fesse a placetned transducer and the lond~displacement curve is plotted on a Eom nected to the system. The loud and strain readings at failure are used to give a direc measure of the unconfined compressive strength of the soil.12.12 Unconfined Compression Test on Saturated Clay 509 Ea Unconfined Compression Test on Saturated Clay _ ie confining pressure o, is 0. An axial load is rapidly applied to the specimen to cause failure. At failure, the total minor principal stress is zero and the total major principal stress is o,, (Figure 12.36). Because the undrained shear strength is independent of the confining pressure as long as the soil is fully saturated and fully undrained, we have (1245) wheres is th Table 12.6 gives the approxi- mate consistencies of clays on the basis of their unconfined compression strength. A photograph of unconfined compression test equipment is shown in Figure 12.37 Figures 12.38 and 12.39 show the failure in two specimens—one by shear and one by bulging—at the end of unconfined compression tests. “Total stress Moles rok at failure le shear stress Figure 12.36 Unconfined compression test a Normal stress ind Unconfined Table 12.6 General Relationship of Cons a Comsistene Nin toni? we SS~S™SC as soft 25-50 02505 Meaium 50-100 sti 100-200 Very si 200-400 ‘Hard a0 >A510 Chapter 12 | Shear Strength of Sol Figure 12.38 Failure by shear of an unconfined compression test specimen (Courtesy of Baja M. Das, Henderson, Nevada) Figure 12.37 Unconfined compression test equipment (Courtesy of ELE Invemationaly Figure 12.39 Failure by bul compression test specimen (Courtesy of Braja M. Das, Henderson, g of an unconfined vevada)Soil behaviour in shear The specimen is placed on either a porous or a solid disc on the pedestal of the apparatus. Typical specimen diameters (in the UK) are 38 and 100mm, A loading cap is placed on top of the specimen, and the specimen is then sealed in a rubber membrane, O-rings under tension being used to seal the mem brane to the pedestal and the loading cap to make these connections watertight. In the case of sands, the specimen must be prepared in a rubber memibrane inside a rigid former which fits around the pedestal. A small negative pressure is applied to the pore water to maintain the stability of the specimen while the former is removed prior to the apphcation of the all-round confining pressure, A connection may also be made through the loading cap to the top of the specimen, a flexible plastic tube leading from the loading cap to the base of the cell; this connection is normally used for the application. of back pressure (as described later in this section). Both the top of the loading cap and the lower end of the loading ram have coned seatings, the load being transmitted through a steel ball. The specimen is subjected to an all-round fluid pressure in the cell, consolidation is allowed to take place. if appropriate, and then the axial stress is araclually increased by the application of compressive load through the ram until failure of the specimen takes place, usually on a diagonal plane through the sample (see Figure 5.6). The load is measured by ‘means of a load ring or by a load transducer fitted to the loading ram either inside or outside the cell. The system for applying the all-round pressure must be capable of compensating for pressure changes due to cell leakage or specimen volume change. Prior to triaxial compression, sample consolidation may be permitted under equal increments of total stress normal to the end and circumferential surfaces of the specimen, ie. by increasing the confining fluid pressure within the triaxial cel. Lateral strain in the specimen is not equal to zero during consolida- tion under these conditions (unlike in the oedometer test, as described in Section 4.2). This is known as isotropic consolidation. Dissipation of excess pore water pressure takes place due to drainage through the porous dise at the bottom (or ‘op. or both) of the specimen. The drainage connection leads to an external volume gauge, enabling the volume of water expelled from the specimen to be measured. Filter paper drains, in contact with the end porous dise, are sometimes placed around the circumference of the specimen; both vertical and radial drainage then take place and the rate of dissipation of excess pore ‘water pressure is increased to reduce test time for this stage. ‘The pore water pressure within a triaxial specimen can usually be measured, enabling the results to be expressed in terms of effective stesses within the sample, rather than just the known applied total stresses: conditions of no flow either ont of or into the specimen must be maintained, otherwise the comect pressure will be modified, Pore water pressure is normally measured by means of an electronic pressure transducer. If the specimen is partially saturated, a fine porous ceramic dise must be sealed into the pedestal of the cell ifthe correct pore water pressure is to be measured. Depending on the pore size of the ceramic, only pore water can flow through the dise, provided the difference between the pore air and pore water pressures is below a certain value, known as the air entry value of the dise. Under undrained conditions the ceramic dise will remain fully soturated with water, provided the air entry value is high enough, enabling the correct pore water pressure to be measured. The use of a coarse porous disc, as normally used for a fully saturated soil, would result in the measurement of the pore air pressure in a partially saturated soil The It the original cross-sectional area ofthe specimen is 4p, the original length is /, and the original volume is V,, then, if the volume of the specimen decreases during the test, les, @o 159 4‘where 6, is the volumettie strain (AV/Y,) and 6, is the axial strain (AJ). If the volume of the specimen increases during a drained test, the sign of AV will change and the numerator in Equation 5.16 becomes, (146). Ifrequired, the radial strain (¢,) could be obtained fiom the equation +26, GID In addition, the strain conditions in the specimen are not uniform due (0 fiietional restraint produced by the loading cap and pedestal disc; this results in dead zones at each end of the specimen, which becomes, barel-shaped as the test proceeds. Non-uniform deformation of the specimen can be largely eliminated by lubrication of the end surfaces. Ithas been shown, however, that non-uniform deformation has no sig- nificant effect on the measured strength of the soil, provided the lensth/diameter ratio of the specimen is, not less than 2, The compliance of the rubber membrane must also be accounted for. Figure 5.10 Mohr circles for triaxial stress conditions.‘Many variations of test procedure are possible with the triaxial apparatus, but the three principal types of test areas follows: (CPerconsotidetet Cdrained (UU). The specimen is subjected to a specified confining presse and ‘then the principal stress difference is applied immediately, with no drainage/consolidation being permitted at any stage of the test. The test procedure is standardised in BS1377, Part 7 (UK), CEN ISO'TS 17892-8 (Eusope) and ASTM D2850 (US), (Miconsolidared-Undrained (CU). Drainage of the specimen is permitted under a specified confining pressure until consolidation is complete; the principal stress difference is then applied with no further dhainage being permitted, Pore water pressure measurements may be made during the undraied past ofthe test to determine strength parameters in terms of effective suesses. The consoldation phase is isotopic in most standard testing, denoted by CIU, Medera computer-contolled triaxial unahines (also known as stress path cells) use hydraulic pressure control units to control the cell (confining) pressure, back pressure and ram load (axial stess) independently (Figure 5.14). Sueh an apparatus can therefore apply’ a ‘no-ateral sain’ condition where suesses ate anssotpie. mimicking the one- dlmeasional compression that oceurs in an oedometer test, These tess axe often denoted by CAU (the “A” standing for anisotropic). The test procedure is standardised in BS1377, Pat 8 (UK), CEN ISO! ‘TS 17892-9 (Europe) and ASTM D4767 (US). © Consolidated-Drained (CD). Drainage of the specimen is permitted under a specified confining pressuze until consolidation is complete: with drainage sill being permitted, the prieipal stress difference is then applied ata ate slow enough to ensure thatthe exes pore water pressure is main- tained at zero. The test procedure is standardised in BS1377, Part 8 (UK), CEN ISO/TS 17892-9 (Exwope) and ASTM D7I81 (US),3.9 Determination of the total stress parameters @, and c, The undrained shear test 5 ‘The simplest method to determine values for the total shear strength parameters of a soil is to subject suitable samples ofthe soil to this test, In the test the soil sample is prevented from draining during shear and is therefore sheared immediately after the application of the normal load (in the shear box) or immediately after the application of the cell pressure (in the tiaxial apparatus). A sample can be tested in 15 minutes ‘or less, $0 that there is no tme for any pore pressures developed to dissipate or to distribute themselves evenly throughout the sample. Measurements of pore water pressure are therefore not possible and the results ofthe test can only be expressed interms of toal stress. “The unconiined compresson apparatus is only capable of carrying out an undriined test on a clay sample with noradial pressure applied. The test akes about a minute. Undrained tests on silts and sands are not possible in the shear box. ‘The following results were obtained from a series of undrained triaxial tests carried out on undisturbet samples of a compacted soil: a fe 388 465 a8 Each sample, originally 76 mm long and 38 mm in diameter, experienced a ‘vertical deformation of 5.1 mm. ‘Draw the strength envelope and determine the Coulomb equation for the shear strength ofthe sol in terns of total stresses. Solution ‘Volume of sample = ~ x 38? x 76=86 193 mm?Sits Element of Soi Mechanics & Therefore cross-sectional are at Failure = 1931916 mp2 ie 5.1 hPa) rn SRE a oa ug 1216 400 oe n9 600, SAGE AIO" S502 982 The Mohr circles for total stréss and the strength envelope are shown in cS 400 300 200 1000 KPa) Fig 3.19 bxample 3.4 Example 3.5 ‘A sample of clay was subjected to an undrained triaxial test witha cell pressure of 100 kPa and the addition axial stress necessry to cause failure was found to be 188 kPa. Assuming that 6, = 0°, determine te value of additional axial sexe that would be required to cause failure of a furher sample of the soi if it was tested undrained with a cel pressure of 200 kPa, Solution ‘The fist step is to draw the stress circle that represents the conditions for the first test i.e. = 100 KPa and o; = 188 + 100 =388 kPa, The circle is shown taShear Strength of Soils 107 1 DOS 00 on hPa) ig. 320 Baample 35. Fig. 3.20 and the strength envelope representing the condition that fy =O" is ‘now drawn asa herizc nal line tangential to the stress circle, The next step is 10 draw the stress cise with 6, = 200 kPa and tangential tothe strength envelope, ‘Where this circle ents the normal stress axis it gives the value of o, whi seen tobe 388 kPa. ress required for failure = (0, ~ 03) = (388-200) “Itcan be seen fromthe figure that ¢, = 94 kPa, This value can be obtnined ‘numerically, from theresult of either test, if itis remembered that: se BE on 4-0 js 3.10 Determination af the effective stress parameters @’ and c” ‘There are two relevant rinxial tests. 3.10.1 The drained test ‘A porous disc is placed on the pedestal before the test sample is placed in position, so that water can driineout from the soil. The triaxial cell is then assembled, filled ‘with water and pressutiae. The cell pressure creates a pore water pressure within the soil sample and the apparatus is left until the sample has consolidated, i. until the ‘pore water pressure hasfieen dissipated by water seeping out through the porous disc, into the burette (see Fig. 3.21). This process usually takes about a day but is quicker if a porous disc is insta ed beneath the loading cap and joined to the pedestal dise by connecting strips 0 vertical filter paper placed on the outside of the sample but within the rubber m=mbrane. During this consolidation stage the water level in a ‘burett half full of wate: and connected tothe base of the sample is monitored. When the water level stopsris ng then the point of full consolidation has been reached.3.10.2 The consolidated undrained test This is the most common form of triaxial test usedin soils laboratories to determine ©’ and ¢ Ithas the advantage thatthe shear part ofthe test can be carried out in only ‘wo tothree hours, ‘The sample is consolidated exactly as for the drained test, but at this stage the 0. It could also be that an initially saturated specimen has partially dried prior to testing, or has ‘been repaired. Another reason could be the entrapment of air between the specimen and the membrane. In the case of fissured clay’ the failure envelope at low values of confining pressure is curved, as shown in Figure 5.26. This is because the fissures open to some extent on sampling, resulting in a lower strength, and only when the confining pressure becomes high enough to close the fissures again does the strength become constant. Therefore. the unconfined compression test is not appropriate in the case of fissured clays. The size of a fissured clay specimen must also be large enough to represent the mass Envelope for fissured clay Effective stress cltcle Failure envelope Figure 5.26 Unconsolidated-undrained triaxial test results for saturated clay. 179Soll behaviour in shear Figure 5.27 Damage observed following the Rissa quick clay flow-slide (photo: Norwegian Geotechnical Institute - NGI). The results shown in Figure 5.28 were obtained at failure in a series of UU triaxial tests on speci- ‘mens taken from the same approximate depth within a layer of soft saturated clay. The raw data from the tests and the use of a spreadsheet to interpret the test data may be found on the ‘Companion Website. Determine the undraineé shear strength at this depth within the soil 40. 400kPa 200kPa I pressure TOOKPS 10% 20% Figure 5.28 Example 5.3. Solution ‘The maximum value of g (~q, is read from Figure 5.28 for each test, and Equation 5.37 used to determine ¢, for each sample. As the samples are all from the same depth but tested at different confining pressures, c, should theoretically be the same for all samples, so an average is taken as 16.3KPa, 181Development of a mechanical model for soil aa The results shown in Table 5.4 were obtained at failure in a series of triaxial tests on specimens of a saturated clay initially 38mm in diameter by 76mm long. Determine the values of the shear strength parameters with respect to (a) total stress and (b) effective stress. Table 5.4 Example 5.4 Typeoftet Config Al oad Axial deformation Volume change ‘pressure kPa) (N) (mm) (mi) (a) Undrained (UU) 200 222 9.83 = 400 215 10.06 - 600 226 10.28 = {b) Drained (0) 200 403 10.81 66 400 843, 12.26 82 600 1265 14.17 9s Solution The principal stress difference at failure in each test is obtained by dividing the axial load by the cross-sectional area of the specimen at failure (Table 5.5). The corrected cross-sectional area is caleulated from Equation 5.16, There is, of couse, no volume change during an undrained test on a saturated clay. The initial values of length, area and volume for each specimen are: J,= 76mm, 4, =1135mm?, V, = 86x10? mm? ‘The Mohr circles at failure and the corresponding failure envelopes for both series of tests are shown in Figure 5.29. In both cases the failure envelope is the line neatest to a common tangent to the Mohr circles. The total stress parameters, representing the undrained strength of the clay, are 6, =85KPa, gy Table 5.5 Example 5.4 (contd.) TR Oy (a) 200 0.129 - 1304 170 370 400 0.132 - 1309 164 564 600 0.135 = 1312 W ™m (b) 200 0.142 0.07 1222 330 530 400 0.16 0.095 1225 691 1091 600 0.186 0.110 1240 1020 1620 182Soil behaviour in shear a re ERs i= FAP SN 11 | Ci Go (kPa) Figure 5.29 Example 5. The effective stress parameters, representing the drained strength of the clay, are c=0, #27 5.7 The critical state framework "he SL EBEEROIE visi rset by Rosse oo (1988, cess an esation of te observed pattems of behaviour of saturated clays under applied principal stresses. However, the critical state concept applies to all soils, both coarse- (as described in Section 5.5) and fine-grained. The concept Sn (volumetric change). It was demonstrated that a characteristic surface exists which limits all pos sible states of the clay, and that all effective stress paths reach or approach a line on that surface when yielding oceurs at constant volume under constant effective stress, This line represents the eritieal states of the soil, and links p’.q and v at these states. The model was originally derived based on observations of behaviour from triaxial tests, and in this section the critical state concept will be ‘rediscovered’ in a similar way by considering the triaxial behaviour that has previously been discussed in this chapter. Volumetric behaviour during undrained shear The behaviour of a saturated fine-grained soil during undrained shearing in the triaxial cell is first consid- ered, as presented in Figure 5.23, Asthe pore pressure within the sample change during undrained shearing, the ESP camot be determined only from the TSP as for the drained test (Figure 5.19(c)). However, during ‘undrained shearing itis known that ‘here must be no change in volume (Av=0), so use can be made of the volumetric behaviour shown in Figure 5.22(b). As q is independent of u, any excess pore water pressure generated during undrained shearing must result fiom a change in mean sttess (Ap), i€. m=Ap". The nor- mally consolidated and overconsolidated clays from Figure 5.23 are plotted in Figure 5.30(a), assuming that the samples are all consolidated under the same cell pressure prior to undrained shearing. The nomnally consolidated (NC) clay has an initial state on the ICL and shows a large positive ultimate value of wat citi- cal state (ultimate strength), corresponding to a reduction in p’. The lightly overconsolidated (LOC) clay has 183Development of a mechanical model for soil a higher initial volume, having swelled a ite, with its initial state lying on the unload-reload line. At criti cal state, the excess pore water pressure is lower than the NC clay, so the change in pis less. The heavily overconsolidated (HOC) clay starts ata higher inital volume, having swelled more than the LOC sample. The sample exhibits negative excess pore water pressure at ertical state, ie. an inerease in p' due to shear ing. It will be seen from Figure 5.30) thatthe points representing the critical states of the three samples all lie on a line parallel to the ICL and slightly below it. This is the ertical state line (CSL) that has been described earlier, but represented it. v-p" space, and represents the specific volume of the soil at critical state, ie. when q is at its ultimate value. This line isa projection of the set of critical states ofa soil in the +p’ plane, just as the critical state line on the q-p" plane (g= Mp’, Figure 5.11(c)) isa projection of the same set of critical states in terms of the stess parameters q and p’ Considering the behaviour of the coarse-grained soil under triaxial compression described in Figure 5.19, the volumetric behaviour is plotted in Figure 5.30(b). Samples A and B both exhibit dilation (inesease in ¥) due to shearing; as the confining stress increases, the amount of dilation reduces. As the confining stress is increased still farther (sample C), compression occurs in place of dilation. Again, the final points at critical state lie on a straight lite (the CSL) parallel to the ICL; however, the CSL for this eoarse- grained soil will have different numerical values of the intercept and gradient compared to the fine- ‘rained soil considered in Figure 5.30(a), the CSL being an intrinsic soil property at at Nest 2 (og scale) Figure 5.30 Volumetric behaviour of soils du 184 P {logscale) \g (a) undrained tests, (b) drained tests.5.8 Residual strength Inthe drained triaxial test, is a limit to the stain ting the shear strength of clays at large strains is by means of the ring shear apparatus (Bishop er al, 1971; Brombead, 1979), an ‘aunular direct shear apparanus. The annular speciuen (Figure 5.34(a)) is sheared, under a given normal stress, on a horizontal plane by the rotation of one half of the apparatus relative to the other; there is no sestriction tothe magnitude of shear displacement berween the vo halves of the specimen. The rate of rola tion must be slow enough to ensure thatthe specimen remains ina drained conslition, Shear stress, which is calculated from the applied torque, is plotted against shear displacement as shown in Figure $34(b), 188 hc be applSoil behaviour in shear ‘Shear displacement ® © Figure 5.34 (a) Ring shear test, and (b) residual strength. ‘The shear strength falls below the peak value, and the clay in a narrow zone adjacent to the failure plane will soften and reach the critical state. However, because of non-uniform strain in the specimen, the exact point on the curve corresponding to the critical state is uncertain. With continuing shear dis- placement the shear strength contines to decrease, below the critical state value, and eventually reaches a residual value at a relatively large displacement. If the soil contains a relatively high proportion of plate-like particles, a reorientation of these particles parallel to the failure plane will oceur (in the narrow zone adjacent to the failure plane) as the strength decreases towards the residual value. However. reori- entation may not occur ifthe plate-like particles exhibit high interparticle fiiction. In this ease, and in the case of soils containing a relatively high proportion of bulky particles, rolling and translation of particles takes place as the residual strength is approached, It should be appreciated that the critical state concept envisages continuous deformation of the specimen as a whole, whereas in the residual condition there is preferred orientation or translation of particles in a narrow shear zone. The original soil structure in this narrow shear zone is completely destroyed as a result of particle reorientation, remoulded specimen can therefore be used in the ring shear apparatus if only the residual strength (and not the peak strength) is required. "The esl rom series of tet, under a range of values of uonmal sues, euable the fie exvelope for both peak and residual strength to be obtained, the residual strength parameters in terms of effective stress being denoted c! and ¢{. Residual strength data from ring shear testing for a large range of soils have been published (€ g. Lupine al, 1981; Mesri and Cepeda-Diaz, 1986; Tiwari and Marui, 2005), ‘which indicate that the value of cf can be taken to be zero. Thus, the residual strength can be expressed as tang, (542) Typical values of ¢ are given in Section 5.9 (Figure 5.39) 189Development of a mechanical model for soil properties described in Chapter 1 (Jp and J for fine-grained soils and ema. and Cn for coarse-grained soils). These simple tests can be undertaken using disturbed samples. The spoil generated from drilling a borehole provides essentially a continuous disturbed sample for testing so a large amount of information deseribing the strength of the gro.nd can be gathered without performing detailed laboratory or field (Sp >upp ene ropesties. f should be noted that the use of any of the approximate correlations presented here should be ing the amount of data from which the ation par Defore the results of laboratory tests are known (e.g, for Critical state angle of shearing resistance (6%,) Figure 5.35(a) shows data of 9%, for 65 coarse-grained soils determined fiom shearbox and triaxial testing collected by Bolton (1986), Miura eta. (1998) and Hanna (2001). The data are correlated against Con—€nan This correlation is appropriate as both ejay Ad ein are independent of the soil state/density (defined by void ratio c) just as ¢%, is an intrinsic property and independent of e. A comelation line, rep- resenting a best fit to the data is also shown in Figure 5.35(a), which suggests that of, increases as the potential for volumetric change (¢,.,~€,.) increases. The large scatter in the data may be attributed to the influence of particle shape (angularity), which is only pastally captured by eu ~Cuue and grain roughness (which isa fimetion of tke parent materials) from which the soil was produced) Figure 5.35(b) shows data of 6, for 32 undisturbed and 32 remoulded fine-grained soils from triaxial testing collected by Kenney (1959), Parry (1960) and Zhu and Yin (2000). The data are correlated against plasticity index (J,) which is also independent of the current soil state (defined by the current ‘water content 1). A correlation line, representing a best fit to the undisturbed data is also shown in Figure 5.35(b), which suggests that 6, reduces as the plasticity index increases. This may be attributed to the increase in clay fiaction (i.e fine platy particles) as Jp increases, with these particles tending to : = | ae I. } ed ® nan “nin © 1408) Figure 5.35 Correlation of ¢"y with index properties for (a) coarse-grained, and (b) fine-grained soils. 1903.13 Behaviour of soils under shear ‘Before discussing this important subject the follewing definitions must be established. nt up to the w: . Clay which, atno time in its history, has been sub- Jected to pressures greater than its existing overburden pressure. . F Clay which, during its history, has been subjected to pressures greater than its existing overburden pressure. One cause of overcon. Solidation i the erosion of material that once existed above the clay layer. BoulUee clays ar overconsolidated, asthe many tons of pressure exerted by the mass of fee above them has been removed. . The maximum value of pressure exerted on an overconsolidate fore the pressure was relieved . ‘The ratio of the value of the effective preconsolidation Pressure to the value of the presently existing effective overburden pressure. A normally consolidated clay has an OCR = 1.) whilst an overconsolidated clay has an OCR > 1.0.In this chapter, the shear strengths of ined. wnular and cohesive soils were exam= Laboratory procedures for determining the shear strength parameters were described. ‘A summary of subjects covered in this chapter is as follows: According te the Mohr-Coulomb failure criteria, the shear strength of soil can be expressed as (Diet sescanding! arto commonly id bron est thos determine the shear strensth parameters of soil Shear strength of soil is dependent on the drainage conditions, Triaxial tests can be conducted under three different drainage conditions: (Section 12.9) (Section 12.10) (Seetion 12.11) ‘The unconfined compression testis a special type of unconsolidated-undrained test (Section 12.12). Sensitivity is a loss of strength of cohesive soils due to remolding (Section 12.14). Due to the nature of disposition of clay soils the shear strength may vary de- pending on tie direction of load application (Section 12.15). This is referred to as strength anisotropy of clay.. 7 ae, is another method to determine the undrained shear siren ine laboratory anlilSection 12.10) In textbooks, determination of the shear strength parameters of cohesive soils appears tobe fairly simple, However in practic. the proper choice of these Parameters for design and stbiily cheeks of various earth earth-etaning. and farth-supporied structures is very ifleut end requires experince and an appro- pera aasirin gine oece pene er eee eenatee propria tee bjpes of stength parameters (consolidated. drained, consolideled-undrained, and imeonsoldaed-untraned) were tatoduced. Theit use. depends on. drainage down, The, f clays can be used to evaluate th fe soils With the assumption that the load caused by construction hay been applied rapidly and there has been!Soll behaviour in shear IA. 1 Knowledge of the strength and stiffness of soil (its constitutive behaviour) is fundamental for assessing the stability and performance of geotechnical constructions. It relates ground stresses (which are in equilibrium with the applied loads) to ground strains (giving compatible deformations). The constitutive behaviour of soil is highly non-linear and dependent on the level of confining stress, but for most practical problems, it can be modelled/idealised using isotropic linear elasticity coupled with Mohr-Coulomb (stress dependent) plasticity. 2 The strength and stiffness of soil may be directly measured in the laboratory by direct shear tests, triaxial tests or residual (ring) shear tests (amongst others). The method of operation, set-up and strengths/weaknesses of each test have been described. These tests may be used to derive strength properties for the Mohr-Coulomb model (c and ¢) and the shear stiffness of the soil (G). Digital ‘examplars provided on the Companion Website have demonstrated how digital data from modern computerised test apparatus may be efficiently processed. 3 For most common rates of loading or geotechnical processes, coarse-grained soils will behave in a drained way. The peak shear strength of such soils is governed by dilatancy (volumetric change) which is density (state) dependent. This behaviour may be modeled by secant peak friction angles or using a linearised Mohr-Coulomb model (7=<’t+’ tang’). If such soils are sheared to large enough strains, volume change will cease and the soil will reach a critical state (ultimate strength). Soil properties may typically be obtained from drained direct shear or drained triaxial tests. Fine-grained soils will behave in a similar way if they are allowed to drain (ie. for slow processes or long-term conditions). If loaded rapidly, they will respond in an undrained way, and the strength is defined in terms of total, rather than effective stresses using the Mohr-Coulomb ‘model (r=¢,). Drained and undrained strength properties and the shear modulus G (which is independent of drainage conditions) are typically quantified using triaxial tests (CD, CU, UU) in such soils, At very large strains, the strength of fine- grained soils may reduce below the critical state value to a residual value defined by a friction angle ¢, which may be measured using a ring shear apparatus. 4. With increasing strain with any drainage conditions, all saturated soils will move towards a critical state where they zchieve their ultimate shear strength. The Critical State Line defines the critical states for any initial state of the soil, and is therefore an intrinsic soil property. By linking volume change to shear strength, the critical state concept shows that drained and undrained strength both represent the effective stress path arriving at the critical state line (for the two extreme amounts of drainage). 5 Simple index tests (described in Chapter 1) may be used with empirical correlations to estimate values of a range of strength properties (4, ¢, and ¢,). Such data can be useful when high quality laboratory test data are unavailable and for providing additional data to support the results of such tests. While useful, empirical correlations should never be used to replace a comprehensive programme of laboratory testing. 195