Preparing Test Speciment Using Undercompaction [PDF]

Zitiervorschau

R. S. L a d d t

Preparing Test Specimens Using Undercompaction

I D Inside diameter L Cyclic strength index Au Change in pore ,water pressure Aac Change in cell pressure

REFERENCE: Ladd, R. S., "Preparing Test Specimens Using Undereompaetion," Geotechnical Testing Journal, GTJODJ, Vol. 1,

No. 1, March 1978, pp. 16-23. ABSTRACT: A specimen preparation procedure is presented that

offers an improved method of preparing reconstituted sand specimens for cyclic triaxial testing. The method leads to more consistent and repeatable test results. This procedure (1) minimizes particle segregation, (2) can be used for compacting most types of sands having a wide range in relative densities, and (3) permits determination of the optimum cyclic strength of a given sand at a given dry unit weight.

Introduction The specimen preparation procedure most commonly described in the literature on cyclic triaxial strength testing [1-3] requires the sand to be saturated, poured into a water-filled forming mold (usually attached to the bottom pedestal of a triaxial cell), and then densified to the required density by some means, usually by vibrations. This method is referred to herein as the wet-pouring (pluvial) method. Several problems are associated with this wet-pouring method. The two most significant are (1) the segregation of particles when using silty and relatively well-graded sands, and (2) the difficulty of readily preparing test specimens having a prescribed dry unit weight with uniform density. A more precise means of preparing specimens is needed so that cyclic test results will be consistent, repeatable, and less influenced by specimen preparation. Presented herein is a method of reconstituting cyclic triaxial strength test specimens that minimizes most of the problems outlined previously. In addition, the concepts presented can be applied to the preparation of reconstituted test specimens for other types of tests and materials. It should be noted that there is no inference here that this method of reconstitution results in specimens which are representative of in-situ conditions. The procedure incorporates a tamping method of compacting moist coarse-grained sand in layers. Each layer is compacted to a selected percentage of the required dry unit weight of the specimen; this procedure differs from the application of a constant compactive effort to each layer required by ASTM Tests for Moisture-Density Relations of Soils, Using S.5-1b (2.5-kg) Rammer and 12-in. (304.8-mm) Drop (D 698-70) and ASTM Tests for Moisture-Density Relations of Soils, Using 10-1b (4.5-kg) Rammer and 18-in. (457-mm) Drop (D 1557-70). This new approach was selected since it is generally recognized (especially for loose- to medium-dense sands) that when a typical sand is compacted in layers, the compaction of each succeeding layer can further densify the sand below it. The method uses this fact to achieve uniform specimens by applying the concept of undercompaction. In this case, each layer is typically compacted to a lower density than the final desired value by a predetermined amount which is defined as percent undercompaction U,. The U, value in each layer is linearly varied from the bottom to the top layer, with the bottom (first) layer having the maximum U. value. The method of variation is illustrated in Fig. 1. (See

KEY WORDS: sands, compaction, triaxial tests, specimen preparation, percent undercompaction, dynamic testing Nomenclature

~3c Effective isotropic consolidation stress +--Od Cyclic axial deviator stress D10, D30, Soil diameters of which 10, 30, 50, and 60% of D50, and D6o soil weights are finer, respectively Cc Coefficient of curvature c. Coefficient of uniformity D, Relative density + tYd/2ff3c Applied cyclic stress ratio epp Peak-to-peak axial strain N Number of loading cycles Ne Number of loading cycles to obtain a given peakto-peak axial strain Number of loading cycles to failure N / N f Normalized number of cycles W T Total wet weight of material required 3/dr Required dry unit weight of test specimen Wa Average water content (as a decimal) of prepared material Vm Final volume of compacted material WL Weight of material required for each layer h. Height of compacted material at the top of the layer being considered ht Final (total) height of the specimen nt Total number of layers n Number of the layer being considered u. Percent undercompaction for layer being considered Uni Percent undercompaction selected for first layer Unt Percent undercompaction selected for final layer ni First (initial) layer u. Average percent undercompaction for layers compacted 1Associate and laboratory director, Woodward-Clyde Consultants, Clifton, N.J. Member of ASTM. 0149-6115/7810003-0016500.40

16

© 1978 bythe American Society for Testing and Materials

Copyright by ASTM Int'l (all rights reserved); Mon May 7 02:03:55 EDT 2018 Downloaded/printed by Suranaree University of Technology (Suranaree University of Technology) pursuant to License Agreement. No further reproductions authorized.

17

LADD ON SPECIMEN PREPARATION USING UNDERCOMPACTION

in this paper. In addition, to illustrate how the cyclic behavior is affected by the Uni value selected, a series of cyclic triaxial strength tests was performed on specimens of Monterey No. 0 sand in which the Uni value was varied.

MaximumValue

nder m

8 ,,=, r~ 5¸

~

~ n t l

n in la

n

,on ~x~

Material Tested

-ilerfa~tilrc=nt ii(ir_

The particle size distribution curve and the selected index properties of the Monterey No. 0 sand, obtained by Mulilis [4], are shown in Fig. 2 and Table 1, respectively. The sand is a washed uniform medium-to-fine beach sand (SP). The maximum and minimum dry unit weight determinations were performed in general accordance with ASTM Test for Relative Density of Cohesionless Soils (D 2049-69) and Kolbuszewski's method [5], respectively. The specimens tested had initial relative densities Dr of approximately 60%.

compaction

== MinimumValue (usuallyzero) n i = 1

n

nt

LAYER NUMBER Where: A. Percentunder-compactionin layerbeingconsidered,Un Un = Uni

pUni- u.tl

]

L n--~-~_ 1 x (n- 1)

-

Specimen Preparation Procedure

B. Averagepercentunder-compactionfor layerscompacted, On _ Un Un= ~

Each test specimen, 74 mm (2.9 in.) in diameter and 152 mm (6 in.) high, had an initial molding water content of approximately 6% and was compacted in eight layers in a split compaction mold not attached to the triaxial cell ("external" split compaction mold). Further details of this method of specimen preparation are given in Appendix A. After compaction, the split mold was removed and the weight, height, and diameter of the specimen were measured. The specimen was then placed in the triaxial cell and confined with a rubber membrane. The triaxial cell was filled with deaerated water, and a cell pressure o3~ of 36 k N / m z (750 psf) was applied.

Uni = Percentunder-compactionselectedfor first layer Unt = Percentunder-compactionselectedfor final layer(usuallyzero) n

= Numberof layerbeingconsidered

ni = First (initial) layer nt = Total numberof layers(final layer)

FIG. 1--Concept of undercompaction procedure.

also Appendix A.) If this method of variation is appropriate and the proper/.1, value is selected for the first layer (U,i), the end product is a specimen having a virtually uniform unit weight throughout. The method used to arrive at this proper U,i value is presented

COBBLES

COARSE

I

FINE

DIAMETER 6" I-

I-r r~

so! 70

>.

so

gD

Z F-

O rr

4" 3"

s0 4o

20

I

MEDIUM

3/4"

3/8 "

4

10

20

l [ Ill ] [ 1 1 lI l ~____ [! [ !

I~l

L 1-~

Ig]

I.

3o]

Each specimen was saturated prior to being consolidated by flushing deaerated water through the specimen under a back pressure of between 625 and 960 kN/m 2 (13 000 and 20 000 psi).

UNIFIED SOIL CLASSIFICATION SYSTEM 40

60 I

100

200

IliIl'~ I J Ill ]l[I[ ~I]lll

II

Illl

Ill

]l',I]i [ i!;i

llJ] -

-

! II J '.....

' "

'....

+ ~

i

III[~I[ I I--

.....

IIIt:[:t.:[-

200

100

10

.

--Ill!Ill- I

1,0 0.1 GRAIN SIZE iN MILLIMETERS

i i [I

I -I-.. -I ....

I

SILT OR CLAY

FINE

U,S. STANDARD SIEVE SIZE l'h"

1]:1 Ill[ I~l Igl ]~l lffl [~[

cq,ARSE [

Test Procedure

L

I[

I I I

0.01

FIG. 2--Particle size distribution curve.

Copyright by ASTM Int'l (all rights reserved); Mon May 7 02:03:55 EDT 2018 Downloaded/printed by Suranaree University of Technology (Suranaree University of Technology) pursuant to License Agreement. No further reproductions authorized.

I

0.001

18

GEOTECHNICAL TESTING JOURNAL TABLE 1--Index data for Monterey No. 0 sand.

Unified soil classification system symbol Particle size data Ds0, mm Cc~ Cu Dry unit weight data e Maximum, lb/ft 3 Minimum, lb/ft 3

Test Results and Discussion

The results of the cyclic triaxial strength tests are summarized in Table 2. A plot of the cyclic strength index versus the percent undercompaction of the first layer of each specimen is given in Fig. 3. The cyclic strength index Ic is defined as the ratio of the number of cycles to obtain a given peak-to-peak axial strain Are to the product of relative density in percent D, and applied stress ratio _+ad/2F3~, that is, Ne/Dr( +-Od/253c); Ic was used to normalize small differences in relative density and applied stress ratio from one test to another. The data show that Ic, which is directly related to the cyclic strength, varies with U,i or the uniformity of dry unit weight within a test specimen. For the U,i values evaluated (0 to 18%), the number of cycles to obtain a peak-to-peak axial strain of 10% at an applied stress ratio of 0.26 varied between 16 and 41 (see Table 2). Furthermore, a peak Ic value (optimum cyclic strength) was obtained. The U,i value where this peak occurred is defined as the optimum percent undercompaction. Another important factor in understanding the cyclic behavior of sand is its strain development characteristics. Axial strain in compression and extension versus the logarithm of the number of loading cycles is plotted in Fig. 4. The shapes of the curves vary considerably, and it was almost impossible to determine trends visually. To determine whether there was a relationship between U,i and the strain development characteristics, as was found with cyclic strength, the cyclic data were normalized. The curves of the normalized peak-to-peak strain versus the normalized number of cycles are plotted in Fig. 5. The normalized peak-to-peak strain e,p/~pp = 10% is defined as the ratio of peak-to-peak strain at a given number of cycles N to a peak-to-peak strain of 10% (selected failure criteria), while the normalized number of cycles N / N f is defined as the ratio of the number of cycles required to obtain a given e,p to the number of cycles required to obtain an Epp of 10%. This figure shows that as U,~ becomes closer to the optimum percent undercompaction, the normalized strain development curves become more concave.

SP 0.36 0.9 1.5 105.7 89.3

aCc= (1)30)2/(960 × D10). bcu = D60/D~o. c 1 lb/ft 3 = 16 kg/m 3.

During back-pressuring, an effective confining stress of 36 k N / m 2 (750 psf) was maintained. This low confining stress minimizes unrecorded volume changes during saturation; however, if the specimen has a tendency to swell, higher values should be selected. In addition, a small axial stress, sufficient to maintain the specimen in an isotropic state of stress, was applied. Saturation was assumed when the B factor (ratio of the change in pore water pressure Au to the change in cell pressure Aa~) was equal to or greater than 95%. The specimen was then consolidated to the required effective stress F3~. Changes in volume and axial height were recorded during consolidation. The relative density of the specimen prior to cyclic loading is based on these measurements. The specimens were cyclically loaded without drainage by using an eleetrohydraulic closed-loop loading system manufactured by the MTS Systems Corp. The MTS system applied a sinusoidally varying load about an ambient load at a frequency of 1 Hz. Therefore, a cyclic sinusoidally varying axial deviator stress +_oa was applied to the specimen in which the stress varied between peak compression and peak extension values. During cyclic loading, the cell pressure was kept constant, and the changes in axial load, axial deformation, and pore water pressure were recorded.

TABLE 2--Summary of results of tests preformed on Monterey sand No. O. Water Content, Dry Unit % Weight, lb/ft 3a

Test 1 2 3 4 5 6 7 8 9 10

Number of Cycles for

Dr, %

Percent After ,After After Initial UnderConsoliConsoliConsoliLiquecompaction Initial dation Initial dation Initial dation +_Od/2iY3 c faction 0 2 4 6 8 10 12 14 16 18

6.0 8.8 6.0 5.8 5.8 6.3 5.7 6.0 6.0 6.0

24.7 24.8 24.6 25.6 23.8 24.9 24.2 24.6 25.1 25.3

98.3 98.6 98.5 98.4 98.8 98.0 98.2 98.5 98.5 98.5

99.2 99.4 99.7 99.3 99.5 98.9 99.1 99.3 99.3 99.6

59.2 60.8 60.3 59.8 61.7 57.2 58.7 60.3 60.1 60.4

64.0 65.5 67.2 65.0 66.3 62.6 63.8 65.0 65.0 66.9

0.26 0.25 0.26 0.26 0.26 0.25 0.26 0.26 0.26 0.26

24 23 33 33 20 22 19 30 18 10

Peak-to-Peak Strain, % 2.5

5

10

20

Remarks b

24 22 33 33 19 22 18 28 18 9

26 24 36 36 22 24 20 30 20 11

30 28 41 40 27 29 25 35 24 16

54 42 67 57 62 47 44 64 130 43

see Note 1 see Note 1 see Note 1 see Note 1 see Note 1 see Note 1 see Note 1

a 1 lb/ft 3 = 16 kg/m 3. b Notes: 1. A significant (> 10%) decrease in peak-to-peak axial load occurred after a peak-to-peak axial strain of 10% had occurred. 2. Test specimens were 74 mm (2.9 in.) in diameter by 152 mm (6 in.) in height and were compacted in eight layers by using the moist tamping method presented in Appendix A. 3. Consolidation pressure 03c equaled 44.6 kN/m 2 (2088 lb/fl 2).

Copyright by ASTM Int'l (all rights reserved); Mon May 7 02:03:55 EDT 2018 Downloaded/printed by Suranaree University of Technology (Suranaree University of Technology) pursuant to License Agreement. No further reproductions authorized.

19

LADD ON SPECIMEN PREPARATION USING UNDERCOMPACT~ON

Symbol

Peak to Peak Axial Strain, %

O A

5 lO

Test Conditions Relative Density, Dr (%) After (~-3c ConsoliInitial Ib/ft 2 57-62

Note:

63-67

2,088

Stress Ratio +_ l:Td/2 ~3c 0.25-0.26

1 KN/m 2 = 20.88 Ib/ft 2 Number of Cycles to Obtain a Given

X ,,' E3 z -1iL9 zuJ n~

.J ~D >cD

Cyclic Strength Index = Peak to Peak Axial Strain Relative Density (%) x Stress Ratio

Optimum Cyctic Strength Index - ~

3

/0

2

\

A --. _ A l l

A

\

O

0

E)

0

1

O ~ 1 I 0

I 2

I 4

Optimum Percent Under-Compaction 6

I 8

I 10

I 12

I 14

I 16

I 18

PERCENT UNDER-COMPACTION OF FIRST LAYER

FIG. 3--Cyclic strength index versus percent undercompaction of first layer for Monterey No. 0 sand.

Conclusions

A specimen preparation procedure is presented in Appendix A that offers an improved method of preparing reconstituted sand specimens for cyclic strength testing. The method leads to more consistent and repeatable test results and a reduction in the number of uncertainties inherent in presently used procedures. This procedure, termed the undercompaction procedure, (1) minimizes particle segregation, (2) can be used for compacting most types of sands, which have a wide range in relative densities, and (3) permits determination of the optimum cyclic strength of a given sand at a given dry unit weight.

Acknowledgment Portions of this investigation were sponsored by the Professional Development Program of Woodward-Clyde Consultants (WCC). This support is acknowledged with appreciation. Special acknowledgment is given to P. Dutko of WCC who developed the percent undercompaction equations. Members of the staff of WCC who made considerable contributions are, in particular, K. Hau, H. M. Horn, Y. Kim, and J. H. Wilson. Special thanks are also due to D. Koutsoftas of Dames and Moore, M. L. Silver of the University of Illinois at Chicago Circle, and D. J. Leery of Langan Engineering Associates for their reviews of and comments on this paper.

APPENDIX A--RECONSTITUTED SPECIMEN PREPARATION PROCEDURE FOR COARSE-GRMNED SOILS A procedure is presented below for preparing coarse-grained specimens for dynamic cyclic testing or static triaxial testing.

The procedure (1) produces specimens that have a relatively uniform stress-strain response, (2) minimizes the tendency for particle segregation, and (3) can be used to compact most types of coarse-grained soils, with a relative density ranging between very loose and very dense. Although the procedure has been developed for the preparation of cohesionless test specimens, the concepts presented can be applied to the preparation of many different material types for various types of tests. Specimens can be prepared either by attaching a split mold to the bottom pedestal of the triaxial cell ("internal" split mold), as shown in Fig. 6, or in a split mold which is separate from the triaxial cell ("external" split mold), as shown in Fig. 7. A split mold is required since it eliminates many of the problems associated with the extrusion of the compacted specimen from a nonsplit mold. Most specimens, especially those containing fines, compacted in an external split mold at relative densities above about 50%, will have sufficient strength as a result of capillary force so that they may be set up in the triaxial cell without significant change in their fabric. However, extreme care is required in transferring specimens to avoid disturbing the specimen.

1. Adjust the water content of the air-dried material so that that initial degree of saturation of the compacted material will be between 20 and 70%. Oven-drying of the material is not recommended. The lower the percentage of fines in the material, the lower the degree of saturation required. A degree of saturation greater than 70% can be used if water does not bleed from the specimen during compaction. The material should be mixed with water about 16 h before use. 2. Determine the average water content of the prepared material using a minimum of t~vo determinations. 3. Assemble and check all the necessary equipment to be used in preparing the test specimen. Determine the inside diameter and the height of the mold to within _+0.02 mm ( ± 0.001

Copyright by ASTM Int'l (all rights reserved); Mon May 7 02:03:55 EDT 2018 Downloaded/printed by Suranaree University of Technology (Suranaree University of Technology) pursuant to License Agreement. No further reproductions authorized.

20

GEOTECHNICAL TESTING JOURNAL 20

15

10 uJ

O

Percent under c o m p a c t i o n value 'for first layer

o

Z < eT-

.~_

s

E

10

8

i I

~pp = 10%

i

L

@ 20

< X