-:.- Experiment No. 17

DIRECT-SHEAR TEST

References

ASTM D3080AASHTO T236ASTM (1964), Symposium on Laboratory Shear Testing of Soils, ASTM, STP no. 36l.ASTM (1952), Direct Shear Testing of Soils, ASTM, STP no. 13l.Kiekbusch, E., M. Kiekbusch, and B. Schuppener (1979), A New Direct Simple Shear

Device, Geotechnical Testing Journal, ASTM, vol. 2, no. 4, December, pp. 190-199.Lee, K. L. (1970), Comparison of Plane Strain and Triaxial Tests on Sand, J. Soil Mech.

Found. Div., ASCE, SM 3, May, pp. 901-923.Saada, A. S., and F. C. Townsend (1981), State-of-Art: Laboratory Strength Testing of

Soils, ASTM, ST no. 740, pp. 7-77.

Objective

To familarize the student with a procedure for rapidly determining the strength parameters(<I> and c) of a soil

Equipment

Direct-shear device (see Fig. 17-1)2 Dial gages (0.01 mm or 0.001 in)CalipersSmall level

General Discussion

The direct-shear test imposes on a soil the idealized conditions shown in Fig. 17-2. That is,the failure plane is forced to occur at a predetermined location. On this plane there are twostresses acting-a normal stress due to an applied vertical load P; and a shearing stress dueto the applied horizontal load Ph. These stresses are simply computed as

(17-1)

(17-2)

where A is the nominal area of the specimen (or of the shear box) and is usually not correctedfor the change in sample area caused by the lateral displacement of the sample under theshear load Ph. These stresses should satisfy Coulomb's equation of Experiment No. 15,

'T = C + (In tan <I> (15-1)

As there are two unknown quantities (c and <1» in Eq. (15-1), two values, as a minimum, ofnormal (applied) stress (In and shear (measured) stress 'T will be required to obtain a solution.

Since the shear stress 'T and normal stress (In have the same significance as when used ina Mohr's circle construction, rather than solving a series of simultaneous equations for c and

169

Figure 17-1Direct-shear equipment.

(a) One type of direct-shear machine that may be eitherhand powered or motorized (as shown here). All machinesare similar in details. Here are shown a load ring and loadring dial (could use a load cell and digital equipment). Topdial is for vertical sample movement and to monitor con-solidation; left dial measures horizontal shear displace-ment.

(b) Two shear boxes. Both round and square boxes areavailable. Shear box firmly attaches to machine and, withtop half separated, relative movement to develop shearresistance is produced when a horizontal force is appliedto the upper half of the shear box. Serrated porous stonesare usually used on the top and bottom ofthe sample. Thesample is usually on the order of 20 to 25 mm thick.Carefully inspect your system to see how much of theloading block and top half of shear box is a sample load,which must be included in the total sample vertical load.

tan <\>, one may plot on a set of coordinate axes the values of Tversus Un from several tests(generally with Ton the ordinate) and draw a line through the resulting locus of points or theaverage locus of points. The slope of the line is <\>, and the T-axis intercept is the cohesion c.This yields a graphical solution of Eq. (15-1).

For cohesionless materials, the cohesion should be zero by definition and Eq. (15-1)becomes

(17-3)

Test inaccuracies and surface-tension effects of damp cohesionless materials often give asmall value of cohesion (apparent), which should be neglected unless it is more than 10 to15 kPa. If the cohesion value is large and the soil is a cohesionless material, the reason forthe large value should be investigated.

The direct-shear test was formerly quite popular. Then, as the state of the art ad-vanced, it tended to become less popular for several reasons:

1. The area ofthe sample changes as the test progresses but may not be very significant, asmost samples "fail" at low deformations.

2. The actual failure surface is not plane, as is assumed or as was intended from the waythe shear box was constructed, nor is the shearing stress uniformly distributed over the"failure" surface, as is also assumed.

170

'--- -- ----

Experiment Seventeen

Gap should beapprox. largerthan largestgrain size

Dial gage to measure verticalmovement and to observe con-solidation for "consolidated" test

Setscrews to fix loadhead into position

Alignment pins (be sure toremove pins before applicationof Ph)

-Ph

Serrated edgesto hold sample

Setscrews to separateshear box. Back off afterclamping setscrews againstload head

Figure 17-2Line details of the direct shear test.

3. The test uses a small sample, with the result that preparation errors become relativelyimportant.

4. The size of the sample precludes much investigation into pore-water conditions duringthe test.

5. Values of modulus of elasticity and Poisson's ratio cannot be determined.6. .Triaxial test apparatus was developed.

With a further advance in the state of the art, however, the direct-shear test appears tobe regaining much of its former popularity. Some of the reasons for this are as follows:

1. The triaxial test is relatively much more difficult to perform and interpret-especially ifpore pressures are taken.

2. The size of the sample makes it not so time-consuming to perform consolidated-un-drained and consolidated-drained tests, since drainage time is relatively short, even formaterials of low coefficient of permeability, when the drainage path is quite short.

3. Square sample boxes were introduced so that the reduction in area during the test canbe easily accounted for, if desired. The use of square boxes is relatively recent, andmany older machines currently in service still use round sample boxes.

4. The direct-shear machine is much more adaptable to electronic readout equipment, sothat an operator is not required to be on continuous duty for consolidated-drained tests,which may be of several days' duration.

5. It has been found that the soil parameters <I> and c obtained by the direct-shear-testmethod are about as reliable as triaxial values! (probably this is more the result of anoperator problem than of the equipment being equal in performance). This does notmean to imply that the triaxial test is undesirable; rather, if soil parameters are all thatare desired, the direct-shear values have usually been found to be quite acceptable. And,

IOn sands above 35° the <I> values from a direct shear test may be 1 to 4° larger than in a triaxial test [seeLee(1970)]. Below 35° the <I> values are about the same.

Experiment Seventeen 171

as has been pointed out, there is some information, such as pore-water behavior duringshear, that cannot be evaluated using the direct-shear test.

Direct-shear tests may be categorized as follows:

1. Unconsolidated-undrained or U tests. Shear is begun before the sample consolidatesunder the normal load Pv. If the soil is cohesive and saturated, excess pore pressuresmay develop. This test is analogous to the unconsolidated-drained triaxial test.

2. Consolidated-undrained tests. The normal force is applied, and the vertical dial move-ment is observed until settlement stops before the shearing force is applied. This test issomething between the consolidated-drained and the consolidated-undrained triaxialtests.

3. Consolidated-drained tests. The normal force is applied, and the shear force is delayeduntil all settlement stops; the shear force is then applied so slowly that no pore pressuresdevelop in the sample. This test is analogous to the consolidated-drained triaxial test.

For cohesionless soils, all three of the above tests will give the same results, whetherthe sample is saturated or unsaturated and, of course, if the shearing rate is not extremelyrapid. For cohesive materials, the soil parameters are markedly influenced by the testmethod and the degree of saturation, and whether the soil is normal or overconsolidated.Generally, two sets of shear-strength parameters are obtained for overconsolidated soils-one set for tests using normal loads less than the preconsolidation pressure and a second setfor normal loads greater than the preconsolidation pressure. Where a preconsolidated soil issuspected, it may be necessary to perform six or more tests to ensure that the appropriateshear-strength parameters are obtained.

Direct Simple Shear Test

The test equipment and procedures described here are for a direct shear test. The directsimple shear test (DSS) details are described in Kiebusch et al. (1979) and in Saada andTownsend (1981) together with limitations. The DSS uses a closed shear box fixed at thebase with the top free to translate under a horizontal force. The shear box may

1. Use hinged sides2. Use a wire-reinforced rubber membrane for the sides

The closed shear box configuration allows user control of pore pressures and also allowsa reversing force to be applied to the top (hinged or rubber membrane sides) so that cyclicstress effects can be measured. The test has found particular application in soil liquefactionstudies. It is also used in lieu of direct shear tests for soil stability/strength studies.

Procedure This Will Be a Class Project

A. COHESIONLESS SOIL-LABORATORY WORK

1. Weigh a large dish of dry sand (or wet sand with water content accurately known) ofsufficient quantity to do at least three tests at the same density.

2. Carefully assemble the shear box (back off any sample box separation and clampingscrews) and fix into position. Obtain sufficient dimensions that the sample density can becomputed. Obtain the sample cross-sectional area A.

3. Carefully place the sand in the shear box to about 5 mm from the top and place theloading block (including porous stone) on top of the soil. Take the small level and level \the loading block.

Weigh the container of sand to determine weight of the material making up the

172 Experiment Seventeen

sample. Obtain a reference thickness of the soil sample by marking the loading block atseveral points around the perimeter with respect to the shear box.

4. Apply the desired normal load P; and attach the vertical displacement dial gage (readingto 0.01 mm/div). Remember to include the weight of the loading block and upper half ofshear box as a part of Pv.2

For consolidated tests, observe the vertical displacement dial and commence thetest only after the settlement has halted. For cohesionless materials this will be almostimmediately after application of P".

5. Separate the two parts of the shear box by advancing the spacing screws in the upperhalf of the shear box. The space should be slightly larger (by eye) than the largest soilgrains in the sample. Now set the load block by tightening the setscrews provided forthat purpose in the sides of the upper half of the shear box. Next back off the spacingscrews so they clear the bottom half of the shear box; at this time the normal load andthe load due to the top half of shear box and the load block are all carried by the soilsample.

6. Attach the dial gage (0.01 mm/div) to measure the shear displacement.7. For a saturated test, saturate the sample by filling the shear box with water, and allow a

reasonable time for saturation to take place.Be sure the porous stones in the shear box are saturated if the soil to be tested

contains any moisture.8. Start the horizontal (shear) loading and take readings of load dial, shear displacement,

and vertical (volume-change) displacements. If a strain-controlled test is performed,take these readings at horizontal displacements of 5, 10, and every 10 or 20 horizontaldial displacement units

Use a strain rate on the order of 0.5 to not more than 2 mm/min. Do not use too fasta strain rate, or the shear load may peak between readings. The strain rate should besuch that the sample "fails" in about 3 to 5 min.

9. Remove the sand from the shear box and repeat steps 1 through 8 on at least twoadditional samples and to a density within 5 g and not more than 10 g of soil quantityused in first test. Be sure that the sand goes into the same volume by using the referencemarks of step 3.

In step 4 use a different value of P; for each test (suggest doubling the exteriorweights, say, 4, 8, and 16 kg + weight of load block for three tests or 5, 10, 20 kg, etc).

B. COHESIVE SOIL-LABORATORY WORK

1. Carefully trim three or four samples of the same size (and, hopefully, of the samedensity) from the larger sample block, tube sample, or other sample source. Use thesample cutter so that the size can be accurately controlled. Any sample with a weightappreciably different from the others should be discarded and another sample trimmed.[What constitutes "appreciable" compared to the size of the sample (order of 5 cm squarex 20 to 25 mm thick) will be a matter of judgment.]

Note: You may need six samples if the soil is undisturbed and preconsolidated. Keepthe samples in a controlled humidity while trimming, preparing the shear machine, andtaking care of other test details.

2. Loosen the separation and clamping setscrews in the top half of the shear box andassemble the two parts. Be sure that the porous stones are saturated unless you aretesting a dry soil. Measure the shear-box dimensions to compute the area.

3. Carefully place the soil sample in the shear box. It should just fit into the box and fill it toabout 5 mm from the top. Place the loading block in place, add the normal load PI" andattach the vertical dial gage.

2Some shear machines allow taring this along with load hanger or yoke so the added weight is the effectivenormal load.

Experiment Seventeen 173

174

l

For a consolidated test, monitor the vertical dial gage as for a consolidation test """""(Experiment No. 13) to determine when consolidation is complete.

4. Carefully separate the shear-box halves, using a gap slightly larger than the largest soilgrains present. Clamp the loading head in place using the setscrews for that purpose andthen back off the separation screws. Be sure the normal load reflects the applied verticalload + the weight of the load block and top half of shear box.

Be very careful in separating the shear box when testing soft clay that material isnot squeezed out between the two box halves-use of small vertical loads and/orconsolidation prior to box separation may be required.

5. Attach the shear-deformation dial gage, set both vertical and horizontal dial gages tozero. Fill the shear box with water for saturated tests and wait a reasonable time forsaturation to be complete.

6. Start the horizontal (shear) loading and take readings of load dial, shear displacement,and vertical (volume-change) displacements. If a strain-controlled test is performed,take these readings at horizontal displacements of 5, 10, and every 10 or 20 horizontaldial displacement units. Use a strain rate on the order of 0.5 to not more than 2 mm/min.Do not use too fast a strain rate or the shear load may peak between readings. The strainrate should be such that the sample "fails" in 5 to 10 min unless a CD test is being run.

The strain rate for CD tests should be such that the time for failure to occur tf is

(17-4)

where t50 is the time for 50 percent consolidation to occur under the normal load Pv' If t50

is not readily obtainable, use the formula

A plot of vertical dial reading vs. log time as for a consolidation test can be made todetermine when the soil has completed consolidation. When P; is very large, it may benecessary to apply the load in increments rather than all at once for reasons outlined inExperiment No. 13.

7. Remove the soil and take a water-content sample. Repeat steps 2 through 6 for two ormore additional tests. If soil is preconsolidated and you use six tests, be sure to use arange of normal loads for three on each side of the preconsolidation pressure.

C. COMPUTATIONS

The following computations are applicable to either cohesionless or cohesive soil.

1. Compute the nominal normal stress as

where A = cross-sectional area of shear-box soil sampleP; = total normal load including load block and top half of shear box

2. Plot the horizontal displacement flh vs. horizontal shear force Ph to obtain the best valueof ultimate shear force.f and compute the shear stress as:

T = Ph(ultimate)A

3Alternatively, plot horizontal-displacement dial units vs. load dial units as in Fig. 17-4 to obtain the maximumshear force.

Experiment Seventeen

DIRECT-SHEAR TEST (Cohesive Soil, Cohesion less Soil)

Project DI r et:. I-Shear te~f [.t1./I,J da t(J.,j

Location of Project 50; I La. be raloryDescription of Soil Med Ca4.".$~ Sad'cI

Data Sheet 21

Job No. _

Boring No. _/'V

Sample No. _

/\./Depth of Sample _

Tested by _~J'-"'E"'-=B=___...,.,''___..:.1(2....=.6_=t.=___ _ / jIZ/XXDate of Testi ng _

Soil state (.wt, dry) Soil sample (disturbed, &kdlgt6i8~)

Data to Obtain Sample Density if not an Undisturbed Sample

In iti al wt. contai ner + soi I = ---'.I-=J=-'J=-=-"_.-.:?---'!I=-- _Final wt. container+ soil = ---'/c.:2=-:J=-=~=-.c.:() _

Wt. of soil used = __ /_~__0_._7 _

Water Content DataWt. wet soil + cup = _

Wt. dry soil + cup = _

Wt. of cu p = ----:::-- _Wt. of water = _

Wt. of dry soi I = _Water content, w% = _

Shear specimen dataSample Dimensions:~at St side = S. 08)( S.08 Cm

Ht.= -3.#z Gm

Area = 2 S. 81 c.,.",~Vol. = 4". Z~

Density: Ywet = ,.-...J _ Ydry =

Loading rate = ._--,0,,--,-, 5=-:0=--__ :..."":.:...:",/m in

/9.~ j.Pa.Normal stress u" = ----y--:--::=-:----....-r;r-c;--r-----

( / . .3 'l" }lld, "Ja./~o6 }t~Load ring constant = /div.

Normal load = -=..s'----_}&-"'~'____ _

Vert. Vert. Horiz. Horiz. Load Horiz.dial displace. dial displace. Corr." ring shear Shear

reading ..W, reading .lH area dial force, stress

( 't,O.OlmllJ ( ) ( 'to.O/",'!! ( );\' reading (

N) T, kPa

0 0 0 0 - 0 0 0

+0·5 /0 /.If

rZ.S 2..0 19

.,.3.0 #,0 Z/f-

-r t..() ,"0 l5

~ Z,O ?5 Z'-+ /..5' /00 Z!'r/'S /.50 30

+,.,5 17.5' 3/

~/·S 200 .JI

.,../. .5' 250 2~.$

-rl. S .3 0 o z'-Note: See Fr' .n-» r: Plot. +y= ex.pa.n5ion of 5a.mple..10

I

Note: Insert units in column headings as necessary."For square samples. may use corrected specimen area at failure as A' = A" - b IlH to compute (T" and T.

Figure 17·3Direct shear-test data for a cohesionless soil. Cohesive soil data is similar.

9

Direct shear testMed. coarse sandby: JEB 1/15/_

112 L.---- --- r-- -./

/~

f-

f- VI 64

! v...- ..•...

/l~

V31

~[--u -

! YI

130

12

11

100

8

7X"0coo-l

6

50

4

3

2

oo 2 4 6 8 100 14 200 26 300

Horiz. displ., x 10-2 mm

4

E (+)E 2~I0 0x;::.

2<I (-)4

-/I.-~ •.... ~

Figure 17-4Plot of shear and volume change vs. horizontal displacement. Note use of load (not stress) and ~V to save time,since curve shape would be identical.

Note: One may use the residual shear force (value somewhat less than the ultimateat a displacement beyond that for ultimate shear force) in this computation for obtainingthe residual-strength parameters.

3. Plot the value of shear stress T vs. un for the tests. Construct a "best fit" straight linethrough the plotted points (Fig. 17-5). Be sure to use the same scale for both the ordi-nate (T) and the abscissa. Obtain the cohesion (if any) as the intercept with the ordinateaxis, and measure the slope of the line to obtain <\>.

You can obtain the residual shear-strength parameters by plotting the residualshear stress vs. Un'

4. On the graph of Bh vs. Ph and using the same horizontal displacement scale, make a plotof vertical displacement vs. Bh (as Fig. 17-4). This plot will display volume change vs.shear displacement. Make appropriate comments in your report concerning the magni-tude and shape of the plot.

176 Experiment Seventeen

60

Med. coarse sand 1/15/

-:V •

- 31(1.379)10 -166 kPa -: 3f0-

T - 25.81 - .

= 5(98.07) = 19 0I---Un 25.81 .

~'02V-$/

/V

16671V/ 19

V

50

40

30

20

10

10 20 30 40

un. kPa

50 60 70

Figure 17-5Plot of shear stress vs. normal stress to obtain soil parameter(s). The measured slope is <\>; any T-axis intercept iscohesion c. Here c = 0 for a cohesionless soil.

5. In your report make appropriate comments on the shear-strength parameters obtained.Consider whether you should have used a corrected area in computing the shear stress(and normal stress) or whether the results are conservative or unconservative withoutthe correction for area. Comment on why it is necessary in Experiments No. 14 throughNo. 16 to plot strain vs. stress to obtain the maximum stress when the maximum shearstress can be obtained from a plot such as Fig. 17-4 in this soil test.

Experiment Seventeen 177