STP 599-1976 (Soil Specimen Preparations Laboratory) [PDF]

Zitiervorschau

SOIL SPECIM EN PREPARATION FOR LABORATORY TESTING A symposium presented at the Seventy-eighth Annual Meeting AMERICAN SOCIETY FOR TESTING AND MATERIALS Montreal, Canada, 22-27 June 1975

ASTM SPECIAL TECHNICAL PUBLICATION 599 D. A. Sangrey, symposium co-chairman R. J. Mitchell, symposium co-chairman List Price $35.00 04-599000-38

,4N~L ~L~/~AMER~CAN SOCIETY FOR TESTING AND MATERIALS 1916 Race Street, Philadelphia, Pa. 19103

qi]|lY

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

(~) BY AMERICAN SOCIETY FOR TESTING AND MATERIALS 1976 Library o f Congress Catalog Card Number; 76-704

NOTE The society is not responsible, as a body, for the statements and opinions advanced in this publication.

Printed in Bahimore, Md. June 1976

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Foreword The symposium on Soil Specimen Preparation for Laboratory Testing was presented at the Seventy-eighth Annual Meeting of the American Society for Testing and Materials held in Montreal, Canada, 22-27 June 1975. Committee D-18 on Soil and Rock for Engineering Purposes sponsored the symposium. D. A. Sangrey, Cornell University, and R. J. Mitchell, Queen's University of Kingston, presided as symposium cochairmen.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Related ASTM Publications Performance Monitoring for Geotechnical Construction, STP 584 (1975), $14.00, 04-584000-38 Field Testing and Instrumentation of Rock, STP 554 (1974), $18.75, 04-554000-38 Analytical Methods Developed for Application to Lunar Sample Analysis, STP 539 (1973), $15.00, 04-539000-38

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

A Note of Appreciation to Reviewers This publication is made possible by the authors and, also, the unheralded efforts of the reviewers. This body of technical experts whose dedication, sacrifice of time and effort, and collective wisdom in reviewing the papers must be acknowledged. The quality level of ASTM publications is a direct function of their respected opinions. On behalf of ASTM we acknowledge with appreciation their contribution.

A S T M Committee on Publications

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions aut

Editorial Staff Jane B. Wheeler, Managing Editor Helen M. Hoersch, Associate Editor Charlotte E. DeFranco, Senior Assistant Editor Ellen J. McGlinchey, Assistant Editor

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Contents Introduction Effect of Water Saturation History on the Strength of Low-Porosity Rocks--G. BALLIW, B. LADANYI,AND D. E. GILL Testing Equipment Rock Types Specimen Preparation Testing Procedures Experimental Results Conclusions

4 5 7 8 11 12 19

Four Factors Influencing Observed Rock Properties-P. G. CHAMBERLAIN,E. M. VAN EECKHOUT,AND E. R. PODNIEKS Discussion of Critical Factors Summary

21 22 34

Trimming Device for Obtaining Direct Shear Specimens from Samples of Stiff Fissured Clay Shale--G. N. DURHAM Residual Shear Test Procedures Waterways Experiment Station Residual Shear Testing WES Direct Shear Trimming Device Specimen Preparation Discussion

37 38 38 39 40 42

Effects of Specimen Type on the Residual Strength of Clays and Clay Shales--F. C. TOWNSENDAND P. A. GILBERT Previous Investigations Materials and Equipment Specimen Preparation Test Results and Analyses Conclusions

43 44 45 47 49 63

Effects of Storage and Extrusion on Sample Properties-ARA ARMANAND S. L. MCMANIS Literature Survey Sampling and Field Testing Laboratory Tests and Results Selection of Representative Specimens General Conclusions

66 67 68 69 80 85

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Transportation, Preparation, and Storage of Frozen Soil Samples for Laboratory Testing--T. H. W. BAKER Factors Affecting Laboratory Tests on Frozen Soils Frozen Soil Samples Machining and Preparation of Specimens for Testing Rough Cutting Methods Finishing Methods Storage and Protection During Laboratory Testing Conclusions

88 89 89 97 98 98 104 111

Temperature-Controlled Humid Storage Room-MICItAEL BOZOZUK Design Closed Flow Conditioning System Handling and Preparation of Samples for Storage Effect of Storage Time on Test Results Summary

113 115 119 122 122 125

Effect of Storage and Reconsolidation on the Properties of Champlain Clays--P. LA ROCHELLE, J. SARRAILH,AND F. A. TAVENAS Characteristics of the Cemented Clays Water Migration Following Sampling Influence of Reconsolidation Influence of Storage Time Conclusion

126 128 130 137 140 144

Pore Water Extraction and the Effect of Sample Storage on the Pore Water Chemistry of Leda Ciay--J. K. TORRANCE Soil Material Storage Procedures Pore Water Extraction Results and Discussions Conclusions and Recommendations

147 149 149 150 151 155

Variation in Atterberg Limits of Soils Due to Hydration History and Specimen PreparationmD. A. SANGREY,D. K. NOONAN, AND G. S. WEBB Test Program Conclusions

158 160 167

Effect of Specimen Preparation Method on Grain Arrangement and Compressibility in SandmARsHUD MAHMOOD,J. K. MITCHELL, AND ULF LINDBLOM

Soil Fabric One-Dimensional Compressibility

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

169 170 171

Experimental Investigation Fabric Results Compression Test Results Conclusions

171 178 180 190

A Technique for the Preparation of Specimens of Loose Layered Silts--V. A. NACCIAND R. A. D'ANDR~A Soil Description Specimen Preparation Typical Testing Procedure and Result Conclusions

193 195 195 198 200

Shrinkage of Soil Specimens During Preparation for Porosimetry Tests--T. F. ZIMMIEAND L. J. ALMALEH Equipment Experimental Work Conclusions

202 204 211 214

Compaction and Preparation of Soil Specimens for Oedometer Testing--A. R. BOOTH Choice of Compaction Method Construction of Mold Method of Compaction Adjustment of the Degree of Saturation Comparison of Specimens Effect on Results Conclusions

216 217 218 219 221 223 224 225

Laboratory Preparation of Specimens for Simulating Field Moisture Conditions of Partially Saturated Soils--T. Y. CHH AND S. N. CHEN Review of Current Methods for Pretesting Treatment Development of Equipment and Procedures for Pretesting Treatment Test Results and Discussion General Conclusion

232 236 243

Scalping and Replacement Effects on the Compaction Characteristics of Earth-Rock MixturesmR. T. DONAGHE AND F. C. TOWNSEND Procedure Test Results and Discussion Conclusions

248 249 257 274

Study of Irregular Compaction Curves--P. Y. LEE Laboratory Investigation

278 281

229 230

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Discussion of Test Results Conclusions

Importance of Specimen Preparation in Microscopy-J. E. GILLOTT Microscopic Methods Specimen Preparation for Fabric Analysis Specimen Preparation for Analysis of Particle Size and Shape Ion Bombardment Replication, Shadowing, and Coating Discussion Conclusions

282 287

289 291 293 299 300 302 304 305

Use of Ultrasonic Energy for Disaggregation of Soil SamplesmA. I. JOHNSONAND R. P. MOSTON Ultrasonic Equipment Testing Methods Summary

308 308 311 312

Soil Drying by Microwave Oven--P. V. LADEAND H. NEJADI-BABADAI Heating with Microwaves Effects of Heating Clay Mineral Systems Preliminary Investigations Determination of Water Content Effects of Microwave Heating on Soil Characteristics Summary and Conclusions Discussion

320 321 322 323 324 330 333 335

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

S T P 5 9 9 - E B/J u n. 1976

Introduction

A laboratory test run on an inappropriate specimen is often worse than no test at all. Certainly there are tests for which the preparation does not significantly change the measured soil property; but the far more common situation is to have a real or potential variation in the measured soil property as a result of alternative specimen preparation techniques. The objective of this symposium was to collect and exchange information on this problem. Hopefully this will lead to improvements in our specimen preparation methods or at least a better understanding of the influence of our preparation methods on final test results. The entire question of adverse effects on test results through specimen preparation needs to be examined in the context of the use being made of the test results. In some cases, accepted practice or a certain design method are based on a test result involving a particular specimen preparation technique. If newer, and clearly better, specimen preparation methods are proposed for this test, there will often be reluctance on the part of users to change, simply because they are accustomed to the older methods and have a strong empirical experience to account for the poorer specimen preparation. Another common situation is that there are some test parameters which can be used in different design methods. For some of these design methods, the specimen preparation is very important, while for others it is much less important. What general principles, if any, can be applied to the preparation of soil specimens for laboratory testing? In general, laboratory soil testing should be done on specimens as nearly identical to field deposits as possible. For natural soils, this means a minimum of disturbance, contamination, and alteration. For artificially prepared, Or reconstructed, soils, the objective is to duplicate the in situ structure and state of the soil, at least in those ways that would influence test results. An overall objective should be to define methods of specimen preparation and testing which achieve the smallest variability in the end result. Some methods are inherently less variable than others, and these will produce more accurate and more predictable end results. An objective of the ASTM symposium is to provide a forum for the exchange of information on a topic of interest or concern. The morning session of this symposium was separated into two major topic areas. The first group of papers was concerned with rock as an engineering material, with stress-strain and strength behavior being the main subjects. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 9 Downloaded/printed byby ASTM International www.astm.org Copyright 1976 University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

2

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

From the interest shown for this session it is clear that geotechnical engineers are becoming increasingly sensitive to the problems of rock engineering and the role of geotechnical engineering methods in solving rock mechanics problems. Storage, extrusion, and predrying effects were covered during the second part of the morning session. The overall conclusion to be drawn from this group of papers is that there exists a great potential for change in soil and rock specimens and their measured properties if samples are handled poorly and stored for long periods of time. Results from this group of papers have a direct bearing on present ASTM standard methods and, in some cases, clearly indicate a need for reconsideration of existing specifications. Three major topic areas were included in the afternoon session. Methods for preparing reconstructed loose cohesionless soil specimens in the laboratory were discussed in the first group of papers. These specimens were intended for studies of liquefaction potential and similar large deformation response. This very current subject was of particular interest to a large part of the symposium audience. Preparation of compacted soil specimens was a second topic area dealing with reconstructing soil specimens in the laboratory. Papers in this part of the symposium were primarily concerned with the problems of preparing a laboratory specimen which represented the field situation. As in the case of the second half of the morning session, there were some direct implications for present ASTM standard methods indicated in these papers. The final session of the symposium was appropriately concerned with recent techniques applied to laboratory preparation of soil specimens. All of the papers described new equipment or new techniques for preparing and testing soil specimens. None of the methods described are presently covered by ASTM standard methods, but it is reasonable to expect a need for standards in the near future if there is more widespread use of these new techniques. A number of present ASTM standard methods were included in the studies reported in this special technical publication. In several cases, the results of these research studies indicated a need to reconsider the present specification, or at least some of its details. Whether it is appropriate to change an existing standard method or add a method is an important decision which cannot be based on a single research study; however, users of specifications should be aware of potential problems even if the specification is not changed. The listing in Table 1 summarizes the papers included in this special technical publication and the ASTM standard methods to which they apply. Only the major associations are noted and there are numerous minor specification references which have not been included. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

INTRODUCTION

3

TABLE 1--ASTM standards and relevant papers. ASTM Designation D 421 D 422 D 423 D 698 D 1140 D 1557 D 1587

D 2049 D 2216 D 2217 D D D D

2664 2936 2938 3080

RelevantPapers from This Symposium

Sangrey, Noonan, and Webb Johnson and Moston Johnson and Moston Sangrey, Noonan, and Webb Chu and Chen Donaghe and Townsend Lee Johnson and Moston Chu and Chen Donaghe and Townsend Lee Arman and McManis Bozozuk LaRochelle, Sarrailh, Roy, and Tavenas Torrance Mahmood,Mitchell, and Lindblom Lade and Nejadi-Babadai Sangrey, Noonan, and Webb Johnson and Moston Ballivy, Ladanyi, and Gill Chamberlain,Van Eeckhout, and Podnieks Durham Townsend and Gilbert

This ASTM special technical publication contains a group o f symposium papers addressing a broad range of materials and testing methods. It is clearly shown that in most cases the methods of specimen preparation have a pronounced influence on the subsequent test results. In a few cases, the opposite conclusion is drawn, for example, in the paper by Townsend and Gilbert, but it is equally important to know about minor effects as it is major ones. Several o f the papers present results and conclusions which have a direct bearing on present ASTM standard methods. Collecting such information is a major reason for having a symposium and special technical publication supported by ASTM. The responsibility for critically reviewing these research studies and, where appropriate, making modifications to existing standards rests with the A S T M committee structure.

D. A . S a n g r e y Cornell University, Ithaca, N.Y.; symposium co-chairman.

R . J. M i t c h e l l Queen's University at Kingston, Ontario, Canada; symposium co-chairman.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

G. Ballivy, ~ B. Ladanyi, 2 a n d D. E. Gill 2

Effect of Water Saturation History on the Strength of Low-Porosity Rocks

REFERENCE: Ballivy, G., Ladanyi, B., and Gill, D. E., "Effect of Water Saturation History on the Strength of Low-Porosity Rocks," Soil Specimen Preparation for Laboratory Testing, ASTM STP 599, American Society for Testing and Materials, 1976, pp. 4--20. ABSTRACT: The purpose of the tests described in this paper was to investigate how the mechanical properties of rock observed in the tests are influenced by the whole saturation history of the specimen prior to testing. Three aspects of the saturation history were studied in this paper: the effect of drying and resaturating the specimen prior to testing, the effect of resaturation method, and the effect of the chemical nature of the resaturating fluid. Three rock types were used in the tests: a gneiss, a cemented sandstone, and a fine grained limestone. All three rocks had apparent porosities below 2 percent. Results of triaxial and splitting tests are reported in the paper. One series of specimens was brought from the site in its natural saturated state and tested without drying while the others were either air or oven dried and then resaturated prior to testing. The resaturation was performed either by immersing the specimen in water under vacuum, or by injecting the saturation fluid, under pressure, through a thin channel drilled along the specimen axis. Either distilled or seawater were used as the resaturating fluid. The results show that the inclusion of a drying and wetting cycle prior to testing has a clear overconsolidation effect on the rock behavior, that is, it increases its apparent strength. On the other hand, the channel saturation technique gives a better saturation of the specimen and results in a strength decrease. Finally, the results show that the chemical composition of the saturation fluid has also a significant effect on the measured rock strength. The practical conclusion to be drawn from this study is that representative rock samples, taken in connection with a given project, should, from the moment of coring until they are tested, be held under environmental conditions that are as close as possible to those which will prevail after the completion of the project. This implies that no drying and wetting cycles should be included if they are not expected to occur in practice. If this condition cannot be met, specimens should be saturated using a natural saturation fluid and using an efficient saturation technique such as the described axial channel saturation method. KEY WORDS: soils, rock mechanics, rock sampling, splitting tests, triaxial tests, saturation methods, pore pressure ~Geotechnical engineer, Lalonde, Girouard, Letendre and Associates, Montreal, P.Q., Canada. ZProfessor and associate professor, respectively, Department of Mineral Engineering, Ecol6 Polytechnique, Montreal, P.Q. Canada. 4 Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 9 Downloaded/printed byby ASTM International www.astm.org Copyright 1976 University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

BALLIVY ET AL ON LOW-POROSITY ROCKS

5

It is well known [1-5] 3 that the strength of rock depends, in large measure, on the degree of saturation of specimens at the monent of testing. In fact, it has been stated, in ASTM Test for Triaxial Compressive Strength of Undrained Rock Core Specimens Without Pore Pressure (D 2664-67) "that the field moisture condition of the specimen should be preserved until time of test . . . or should be tailored to the problem at hand." The wetting of specimens prior to testing has a strength reduction effect which is generally very large for all kinds of rocks. This effect results essentially from a reduction of the free surface energy of the material [3]. This phenomenon is well known, but its physical and thermodynamic aspects have not been completely clarified to date [6]. The intent of the present study is to illustrate more specifically the effect of water saturation history on the strength of low apparent porosity rocks. Three types of rocks have been tested, namely, a cemented cambrian sandstone, a lithographic ordovician limestone, and a charnokite (archaean granito-gneiss). As it is not usual to preserve the field moisture conditions of the specimen until laboratory testing, this paper also examines the effects of various resaturation processes on rock strength. Study of these processes included development of new resaturation equipment, special specimen preparation, and the use of two different pore fluids.

Testing Equipment The rock cutting was performed with a circular watercooled diamond saw blade. Whenever required, the ends of the specimens were ground flat on a lathe. The specimens were weighed on electronic balances, and calipers were used to measure their final dimensions. The oven used for drying the specimens was built in such a way that the air, heated to 40.5~ as it entered the oven, was forced to circulate throughout it; the total volume of air in the oven was renewed every minute. Resaturation by fluid injection was done with the apparatus shown schematically in Fig. 1. It consists essentially of a pressure vessel (A), through the cover of which eight specially prepared specimens can be connected to eight tubes; these tubes are all connected to a second pressure vessel (B) which acts as a saturation fluid reservoir. A nitrogen gas bottle (C) pressurizes, through a regulator, the saturation fluid contained in the vessel (B); this pressurized fluid is injected, by means of the tubes, through a channel drilled along the axis of each of the specimens. A second nitrogen gas bottle (D) applies a pressure, also 3 The italic numbers in brackets refer to the list o f references appended to this paper.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

6

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

FIG. l--Apparatus for saturation of rock specimens by radial divergentflow from a central channel; (a) view of the apparatus; (b) scheme of the saturation system. through a regulator, on the saturation fluid contained in the vessel (A), thus providing a constant fluid pressure at the outside surface of the specimens. The saturation fluid pressure gradient results from the difference between the pressures produced by the nitrogen gas bottles C and D. This gradient produces radial divergent flow within the specimens. The apparatus was constructed in such a way that all steel surfaces coming into contact with the saturation fluid were lined with plastic material. For the tests results reported in this paper, the pressure in C was 300 Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

BALLIVY ET AL ON LOW-POROSITY ROCKS

7

psi (2.07 MPa), and in D, 50 psi (0.34 MPa). The photograph in Fig. 1 shows the apparatus just described. Both a standard testing machine and, more often, a programmable universal testing machine (Tinius Olsen) were used for loading the specimens to failure. Diametral splitting tests were conducted between rigid plattens. Triaxial tests were performed in a modified Hock cell, in which the confining pressure was supplied by a pump (Structural Behavior Engineering Laboratories, Model 100 LP + 100 P). To test the specimens which had been resatured by fluid injection, the plattens were modified as shown in Fig. 2b. As far as the pore pressure is concerned, a nitrogen gas bottle, combined with a pore fluid reservoir, was used whenever the back pressure to be maintained during testing was less than 500 psi (3.45 MPa); otherwise, the back pressure was provided by the pump. Figure 2a is a photograph of the Hoek cell with modified plattens. Specimen deformations were measured with electromechanical extensometers, in which sensors were linear potentiometers. The longitudinal deformation of the triaxially tested specimens was measured outside the cell; the signal output by the measuring devices was recorded against the load applied by the testing machine on a standard X - Y recorder. In the case of the splitting tests, the changes in both the vertical and the horizontal diameter were measured and recorded, as for the triaxial tests.

Rock Types The tests were performed on three types of rocks from various locations in the province of Quebec.

Cambrian Cemented Sandstone (Potsdam Group) It is predominantly a white to off-white orthoquartzite with dolomitic cement; the apparent porosity of this bed is less than 2 percent.

Ordovician Sublithographic Limestone (Trenton Group, Tdtreauville Formation) It consists of beds of dense bluish-black limestone up to 6 in. in thickness, separated by shale partings; it has a lithographic stone appearance [7]. The total porosity of this rock is low (1.2 to 1.9 percent) [8], and its apparent porosity is less than 1 percent.

Charnokite This is an Archaean granito-gneiss from the Quebec City area. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

8

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

FIG. 2--Hoek's cell with modified plattens; (a) view of the cell and the plattens; (b) modified bottom platten.

Specimen Preparation General

All the specimens were prepared from NX (diameter: 2~ in. or 5.38 cm) core samples. The present study involved two types of samples which have been cored below the water table. Saturated Samples--Samples selected were kept immersed in water at the drilling site and delivered to the laboratory, where they were submitted to various procedures, including drying and resaturation. Air-Dried Samples--In the other cases, the samples selected at the drilling site were kept under ambient conditions and delivered to the laboratory where they were submitted to various procedures. In all cases, the core specimens were cut to the desired lengths a short time after delivery, and the effect of wetting caused by the cooling water during the cutting was considered insignificant as far as mechanical properties are concerned. The control of specimen saturation was made by a periodical weighing. Usually, a specimen was considered to have reached a given saturation degree when periodical weighings showed constant weight for at least three consecutive days. Specimens f o r Tensile Splitting (diametral compression) Tests

As a rule, specimen disks submitted to splitting tests were about 90 of Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

BALLIVY ET AL ON LOW-POROSITY ROCKS

9

an in. (1.90 cm) thick. The following groups of specimens were prepared from the two previously described types of samples (see Table 1). TABLE l--Groups of specimens prepared by different methods

and number of tests in each group.

Group

Tension Splitting Tests

Triaxial Compression Tests

Air-Dried Specimens A AO AS AOS AOSC AOC

42 9 19 17

50 97 80 8

Saturated Specimens S SO SOS

57 23 42

From saturated samples (S): Group S--The specimens of this group, prepared from saturated samples, were obtained by cutting specimens of core into disks which had been temporarily removed from their water bath. The specimens were then measured, weighed, and tested. Group SO--The specimens of this group were the same as in group S, except that they were oven dried before testing. Group SOS--This group is the same as group S, except that the specimens, after being measured, were oven dried and resaturated by immersion before testing. From air-dried samples (A): Group A--The specimens of core, selected from air-dried sample lots, were cut into disks. The specimens were measured, weighed, and tested. Group AO--The specimens of this group were prepared as were those in group A, except that they were oven dried before testing. Group AS--These specimens were prepared as those in group A, except that they were resaturated by immersion prior to testing. Group AOS--The specimens in this group were prepared as were those of group A, except that they were oven dried and subsequently resaturated by immersion prior to testing.

Specimensfor Triaxial Compression Tests Specimens submitted to triaxial compression tests were about 4 88 long (10.80 cm) cylinders. They were all prepared from the samples of

C opyri ght by ASTM Int ' l ( a l l r i g h t s r e s e r v e d ) ; W e d De c 2 2 1 4 : 0 0 : 0 0 E S T 2 0 1 0 D ownloaded/printed by U ni vers ity of Britis h Co l u m b i a Li b r a r y p u r s u a n t t o Li c e n se A g r eem en t . N o f u r t h er r ep r o d u ct i o n s au

10

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

Type A, that is, the air-dried samples, and they fall into one of the following groups: Group AO--This group is the same as AO, described for specimens for tensile splitting. Group AOS--This group is the same as AOS, described for specimens for tensile splitting. Group AOSC--The specimens in this group were prepared as were those in the splitting tests, except that a 89 hole was drilled along each of their axes for about 80 percent of their length [11]. No cooling fluid was used during this operation. Figure 3 shows a specimen into which a hole has been drilled, as

FIG. 3--Specimens with central channel; (a) radially saturated sandstone specimens with cut fitting, ready for testing; (b) section of a specimen with complete fitting.

described previously. This figure shows also the brass fitting that was cemented to the collar of the channel, in order to enable the specimen to be mounted on the resaturation apparatus described previously. This fitting covers the hole wall for a length equal to about 30 percent the specimen length, leaving an unlined cylindrical channel in the central portion of the specimen; the length of this cavity is then about 60 percent of that of the rock specimen. Figure 3a is a photograph of such a specimen. Note that the threaded part of the fitting was cut away before testing. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

BALLIVY ET AL ON LOW-POROSITY ROCKS

11

Group AOC--The specimens in this group were as in group AOSC, except that they were not submitted to resaturation, although they were provided with a central channel. A review o f the groups and the tests made in each group is presented in Table 1.

Testing Procedures Splitting Tests Specimens were mounted on the testing machine in such a way that the loading could be performed along two diametrically opposite lines on the lateral surface o f the disks. The electromechanical extensometers were then mounted and set to zero. The loading proceeded at such a speed that the minor principal stress increased at a rate o f 100 psi/s. Tensile strengths were calculated from the usual formula 2P To

-

nDL

(1)

where

To = tensile strength, P = maximum load applied, D = diameter of the specimen, and L = thickness o f the specimen. Only the tests in which failure started at the center o f the cross section o f the specimen, and in which failure plane coincided with the loaded diametrical plane, were considered to be valid.

Triaxial Compression Tests Jacketed specimens were mounted in the triaxial cell with proper spherically seated plattens. The cell was placed subsequently into the testing machine, and the extensometer was installed. The loads were applied at such a speed t h a t the major principal stress within the specimens increased at a rate of 100 psi/s. Prior to the test, the confining pressure was applied to the specimen by increasing simultaneously the axial and the cell pressure. Whenever required, back pressure was raised to the desired level (200 psi = 1.38 MPa with limestone specimens, 300 psi = 2.07 M P a with charnokite specimens, and up to 5000 psi = 34.5 MPa with sandstone specimens). Pore pressure was applied from the bottom plattens, and the testing was started only when the same pore pressure could be read at the top plattens. Similar procedures have been reported in the literature already [8]. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

12

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

The extensometer was then set to zero, and the specimens were brought to failure by increasing the axial stress.

Experimental Results Check o f Pore Pressure Distribution in Triaxial Specimens Before starting the study of the effect of specimen preparation history on its strength, a check of the system of pore pressure application in the triaxial cell was made by a series of triaxiai tests on cemented Potsdam sandstone. All the specimens in these tests were of the AOSC group, that is, they had a central hole and were resaturated by immersion after being air and oven dried. The triaxial tests were conducted according to the Heck's procedure [9], but the pore pressure u was kept constant during the tests. Such a system, if working well, that is, if resulting in a uniform distribution of the applied pore pressure (back pressure) throughout the specimen at failure, should result in essentially drained test conditions. According to the concept of effective stress, failure strengths of such a series of tests, plotted against the effective confining pressure, should fall on a single failure line [10]. Figure 4 shows that, with the usual scatter of results, this assumption was found to be valid in the tests with the Potsdam sandstone. These results illustrate that the pore pressure application system used in the tests was quite effective, even for rocks of such a low porosity.

Effect o f Specimen Preparation History on the Results o f Tension Splitting Tests Figures 5 and 6 show the results of a large series of tension splitting tests carried out on specimens of Trenton limestone, prepared according to various procedures. The results show clearly that tensile strength is affected very much by the specimen preparation history. In fact, three different groups of results can be seen in Figs. 5 and 6. The lowest strengths were found for specimens saturated without over drying (groups S and AS), the highest, for those that were air and oven dried (group AO), while the strengths of those tested air or oven dry (A and SO), as well as those tested after having been saturated following air or oven drying (AOS and SOS), were located between the two extremes. These results lead to the conclusion that an oven drying, or a severe air drying, produces a clear overconsolidation effect on the strength of rock, and that the effect is irreversible, that is, it cannot be eliminated by subsequent resaturation of specimens. On the other hand, Fig. 5 shows that

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authoriz

BALLIVY ET AL ON LOW-POROSITY ROCKS

ksi

' MPo

13

/

~,o -

4O0

/

I-z w

40.

~35~( LEGEND:

t,~176 ,o] , L,o

Test with opplied e pressure u= 8 0 0 psi Specimen:Cemented sondstone~ group AOSC

20 3

MPo ksi

EFFECTIVE CONFINING PRESSURE O''5

FIG. 4--Results of triaxial compression tests with Potsdam sandstone.

the overconsolidation effect seems to affect much less the modulus of elasticity of rock, because the peak strength points, shown in Fig. 5, are distributed around a mean straight line without any systematic trend.

Effect of Specimen Preparation History on the Results of Triaxial Tests Figure 7 shows, in terms of principal stresses at failure, a summary of all triaxial test results obtained with specimens of Trenton limestone and charnokite prepared by various described methods. In addition, the median failure line for AOSC specimens of Potsdam sandstone is shown also in the figure for comparison. The tests with Trenton limestone were made at different confining pressures varying from 500 to 5000 psi, while those with charnokite were all at an effective confining pressure of 1000 psi (6.9 MPa). Some results for the latter are summarized also in Table 2, together with the corresponding water contents after immersion in the distilled and the seawater, respectively. The test results lead to a number of interesting conclusions concerning Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

14

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

psi

MPa

1500

/

I0

/

A0 bJ Ico I000,

'AOS,

-6 w

I

-4

/

500-

-2

/ 0

/

LEGEND

IM P a [ ~ e d l o n

/

vol~e of To ond F~v

0.5% 19 ~ ' ~ - s t a n d o r d

deviotion

number of tests

0'5

Ii0 VERTICAL

STRAIN

1',5 210 8v = A.~I

,

25

, II, per cent

I

5--Effect of mode o f saturation on splitting tensile strength and failure strain of Trenton limestone. For the definition o f symbols, see Table 1. FIG.

the effect of preparation history on the failure behavior of rocks under triaxial test conditions. The results show clearly that: 1. The strength of dry specimens is from 20 to 30 percent higher than that of comparable specimens when tested saturated, at any confining pressure. 2. The presence of a central hole in the specimen improves and accelerates considerably the resaturation of the specimen. This is seen clearly in the results obtained for the charnokite. This method, however, could not have been applied to the specimens of Trenton limestone, which had a tendency to fracture along bedding planes during pressure saturation. The limestone, therefore, was soaked in water, and a back pressure of 200 psi was applied through the central channel. This saturation method was clearly less effective, which explains the much smaller difference between Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

BALLIVY

Initml conditions:

ET A L O N

S

or

LOW-POROSITY

A -~...._....~

ROCKS

15

A

@l To. (29Opsi) ,,o~s, S

,

TO= 740psi

@

~e,ts

AS

/ on saturated /4-- samples ~

(160psi)

TO = 780psi

~

(260psi)

SO

AO ~

tests on oven dried samples

TO = 1150psi

(;500 psi)

ID @

TO= 1290psi (~90psi)

AOS

SOS 4---

TO= 1090psi

L

tests on resaturated samples

(280psi)

LEGEND=

5~

9

(150psi)

(160psi) =standard

number of tests

deviation

To = median tensile strength

Ips== 6.9 kPo

FIG. 6--Effect of drying and saturation history on tensile splitting strength To of Trenton limestone. For the definition of symbols, see Table 1. /,

psi

-~-

%,o / 9 AOS(sea

I-

=-

water)

/

50

-300

AOSC j POTSDAMSANDSTONE

/

MPa - 400

~z 9 A/0SC (sea water)

(~

. . Q.~

3o-L2oo/ I" ~ o _ ~ ~

~''

t,,-

"EOENO

~oo

,

J l J

/ O

-

-

~T~,~"5 =median effective stresses ~. R,H.=relotive humidity 5 ~ number of tests

ioJ

L/~30(016~1 :median water content=0.30% ~ standard devJation=O.16%

,o

~ 2

5

4

~

5

,-~-=~ ksi

EFFECTIVE CONFINING PRESSURE 0"5

FIG. 7--Effect of mode of saturation on triaxial compression strengths of Trenton limestone and charnokite. For the definition of symbols, see Table 1.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

24

13

16

AOS

AOSC

Oven-dried specimens 53 800 (12 000)

43 100 (7 600) 35 700 (8 300)

ol ult psi

s

3.17 (0.6)

2.9 (0.4) 2.7 (0.5)

070

0.22 (0.9) 1.11 (0.5)

W , 070

14

24

n

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

NOTE--n = number o f tests. (7 600 psi) = standard deviation. 1 psi = 6.9 kPa. Period of saturation: 8 to 42 days. ol ult = median ultimate axial strength. o3 = 1 000 psi for all the tests. Eult = median peak deformation. W, % = water content in percent.

n

Mode of Saturation

Distilled Water

50 500 (7 100) 43 800 (5 900)

s

~0

3.1 (0.4) 2.8 (0.3)

Seawater Ol ult, psi

Fluid of Saturation

strength of charnokite.

0.28 (0.11) 0.97 (0.37)

W , 070

TABLE 2--Influence of mode of saturation and composition of pore fluid on the triaxial compression

z

m (D

O .
500.0 240.0 640.0

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

LA ROCHELLE ET AL ON STORAGE AND RECONSOLIDATION

129

however, the difference (Pc - Po) between the preconsolidation pressure Pc and the effective overburden pressure Po, as given in Table 1, indicates that the clay has gained a pseudo-preconsolidation which may be attributed to a delayed consolidation during the aging of the deposit [10] and which is locked into the clay structure by the cementation bonds.

Saint-Louis The properties of the d a y at Saint-Louis, Yamaska, which is located 160 km southwest of Quebec City, have also been discussed in detail in previous papers [6,11]. The typical values given in Table 1 correspond to the elevation at which block samples were cut out from a trench dug in a landslide crater. The clay plasticity on that site is more uniform with depth than in Saint-Alban, and the vane strength at block level is 43 kPa (900 lb/ftz). The clay is slightly overconsolidated by previous loads, but the major part of (pc - po) is estimated to be due to delayed consolidation and bonding. In terms of undrained strength and bonding, this clay may be considered representative of the average clay encountered in the SaintLaurent lowlands.

Saint-Jean. Vianney The village of Saint-Jean-Vianney, which was located approximately 200 km north of Quebec City on the northern side of the Saguenay river valley, was the site of a disastrous landslide in 1971, which was reported in the literature [12]. In many regards, the clay on that site is similar to the Toulnustouc clay reported on by Conlon [4]. The typical properties listed in Table 1 correspond to the elevation of block samples cut out from a trench dug in the crater of the landslide. The clay is very stiff, the vane strength being 240 kPa (5000 lb/ft2); it has a low plasticity and is extremely sensitive. When compared to the average Champlain clay deposits, the clay at Saint-Jean-Vianney is highly overconsolidated; the measured preconsolidation pressure at the level of the block samples is 900 kPa (8.4 T/ft 2) and it is estimated that half the value of (Pc - Po) given in Table 1 is due to delayed consolidation and bonding. When comparing the properties measured on the three sites, it becomes obvious that the clay of Saint-Alban is the least cemented; as a matter of fact, of all the Champlain clay deposits studied, that clay probably lies close to the lower end of the scale in terms of intensity of cementation. As the cementation bonds are believed to prevent swelling following a stress release, the clay from Saint-Alban was chosen for the study since it was considered to be the most susceptible to water migration within the clay samples after sampling. On the other hand, the effect of storage time on the bonds was evaluated on block samples from the two other sites where the cementation is more important. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorize

130

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

Water Migration Following Sampling In order to find out whether any water migration takes place within the samples o f cemented clays following sampling, two types of checks were made: the first one consisted o f a direct comparison of the results of unconfined compression tests made at different times after sampling, and the second check consisted o f measuring the water contents across the clay samples a few days after sampling. The samples were taken at the site o f Saint-Alban by means o f a NGI thin-wall stationary piston sampler; the stainless steel tubes used had a diameter of 73 mm, a length of 1 m, an area ratio o f 9 percent, and an internal clearance of 0.25 percent. Tube samples were taken in adjacent boreholes over an area of 4 by 4 m at two different elevations (Fig. 1); 0

I0

Cu-kPo 20 30

o

40

50

T S T-AL BAN

I

~ 1~o 2

I tube somples (upper level)

3 4 E

6

~~c

I tube somples 1 (lower level )

7

8

o vane I o vane 2 e,--e average

9 -

,o FIG.

I

i

l--Undrained strength profile at Saint-Alban.

the useable length o f sample was on the order of 80 cm per tube. Five tubes were taken at depths between 2.1 and 3.0 m, and nine tubes were between 6.1 and 7.0 m.

Unconfined Compression Tests The undrained strength at different times after sampling was measured by unconfined compression tests made on specimens trimmed to a diCopyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

LA ROCHELLE El" AL ON STORAGE AND RECONSOLIDATION 131

ameter of 3.8 cm and a height of 7.6 cm. A press, a triaxial cell, and the necessary equipment for extruding and trimming the samples were installed in a shanty on the site, so as to make compression tests immediately after sampling. In order to study the effect of short storage time on the undrained strength, four series of tests were made on the samples taken at the two different levels:

Series A--The samples were extruded, trimmed, and tested immediately after sampling on the field. Series B--The samples were extruded, trimmed, and paraffined on the field and tested in the laboratory about one week later. Series C--The samples were extruded and paraffined on the field, and they were trimmed and tested in the laboratory about one week later. Series /)---The tube samples were transported to the laboratory and stored in the humid room; the samples were extruded, trimmed, and tested the following week. All compression tests in the field and in the laboratory were made by the same operator, using the same apparatus. It may be worthwhile to mention that the so-called paraffined samples were, in fact, wrapped in plastic films which were sandwiched between layers of a mixture of paraffin and vaseline; this technique was found to be very efficient in preventing any measurable loss or gain of moisture during storage periods of more than three years in a humid room. Comparison o f the Results--When comparing the results in such a study, problems arise due to the variation of strength of the specimens, even when tested under exactly the same conditions. In the present case, two types of variations are encountered: the first one is the natural variation of the undrained strength profile along the length of the tubes, and the second is the variability in the quality of the tube samples and also of the specimens within one tube, resulting from the disturbance due to the sampling operation. Strength Profile--From previous studies in Saint-Alban, the strength profiles are known to be fairly uniform throughout the site; nevertheless, two vane profiles were determined 2 m apart in the central part of the sampling area, and the results are given in Fig. 1. It is seen that both profiles coincide fairly well. On the same figure, the positions of the tube samples have been drawn for the two levels studied; it may be observed that the vane strength varies by an appreciable amount in the depth increase, corresponding to the length of the tubes. At the upper level, the vane strength is 10 kPa (210 lb/ft 2) at the elevation of the top of the tube and 12.5 kPa (260 lb/ft 2) at the elevation of the bottom, thus increasing by 25 percent. At the lower level, there is an increase in strength of about Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

132

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

15 percent from the top to the middle of the tube sample and then a slight decrease in the bottom half. In order to make the strength results of the compression tests comparable, they were all referred to the vane strength measured at the elevation of the specimens within the tube and were expressed by the ratio C~

c~, = c . ~ / c .

(l)

where C~r is the undrained shear strength given by the unconfined compression test at failure, and C,v is the vane strength measured at the same elevation. The vane strength was taken as a reference because it indicated, in a very consistent manner, the variation of undrained strength along the profile. Any elaborate discussion on the significance and value of the vane strength is beyond the scope of the present paper. Variability o f the Quality o f Samples--In spite of all the care taken during the sampling operations in the field, there is an appreciable variation in the quality of the samples obtained in such deposits of soft sensitive clays. Figure 2 gives the results of unconfined compression tests top 0

,

0.2

4

1.4

1.6

1,8

f ~ I

I

f i

--

strength ratio, Cfv 0.6 0.8 1.0 1.2

!

2

}-

0.4

tube

/

I '

/

[

bo~'m

FIG. 2--Variability of the quality of tube samples. made on specimens trimmed from two tubes taken in two adjacent holes at the lower depth (Fig 1). It may be seen that the results from Tube I are much higher than those from Tube H, although the field technique, the sampling apparatus, the operators, and the testing method were the same for both tubes. Moreover, the strength of the specimens within the same tube sample varies appreciably, depending on the position of the

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized

LA ROCHELLE ET AL ON STORAGE AND RECONSOLIDATION 133

specimens in the tube; this variability may be attributed to sampling disturbance [13]. Of the eight 10-cm-long specimens, which can be cut from each tube, Specimen I, located at the top, was systematically eliminated as being disturbed; compression tests on Specimen 1 would give only 20 to 30 percent of the vane strength at that elevation. Even Specimens 2 and 3 (Table 2) give results which are questionable. Hence, the variability of the quality of the tube samples and of the specimens within each tube requires that the results be analysed carefully. Analysis of the Results--The results of the comparative study of the four test series are given in Table 2. Three different approaches were used to analyse the results. In a first approach, all tests were included to compute the average CM thus, the corresponding values of N given in Table 2 represent the total number of tests made for each series in this program. As some tests were giving exceptionally low values of C~, and others, exceptionally high values, the second approach, called the statistical approach in Table 2, consisted in eliminating the results lying outside _+1 standard deviation from the average, and a new average value was recalculated with the results that were not rejected; their number and the corresponding average C~ are given in Table 2, for each case. The third approach was based on an arbitrary rule whereby only the best results were used, up to a number equal to half the total number of tests available in each case. When comparing the results of the three different approaches given in Table 2, it is seen that, as expected, the first two approaches yield average values of C~ which are nearly identical; as for the third approach, it is quite normal that the average values of C~ be appreciably higher. However, for the purpose of the present study, it is not so much the absolute magnitude of Civ which is of interest, but rather its variation from one series to the other. In this respect, it is interesting to note that the three approaches used to analyze the results give essentially the same tendency. As the number of tests at either of the sampling levels is rather small in some of the series, and as the ratio C~ takes into account the variation of strength with depth, it is believed that the averages computed for both levels combined are more representative and should be used for the comparison of the different series. Moreover, as the three approaches are equally justifiable and show essentially the same tendency, the overall averages of the values of the three different approaches may be used for the sake of simplicity in the discussion; these averages are given at the bottom of Table 2. Discussion--It is seen that the same strength values are obtained from the unconfined compression tests in both Series A and B; in both cases, the samples were trimmed on the site, immediately after sampling, thus preventing water migration from the disturbed zone to the undisturbed core of the sample. As the compression tests were made on the field for Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authori

upper lower

Statistical

both

Overall average

4 14 18

8 15 23

9 28 37

Na

0.96

1.14 1.12 1.12

0.78 0.95 0.89

0.85 0.87 0.87

C/~ avg

4 3 7

5 4 10

8 6 14

N

B

0.97

1.17 1.14 1.16

0.88 0.83 0.85

0.92 0.87 0.90

C~ avg

3 3 6

3 3 6

5 6 11

N

C

0.80

0.85 0.86 0.85

, 0.73 0.82 0.77

0.73 0.81 0.78

C/v avg

4 10 14

4 14 19

7 20 27

N

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

aN is the number of test results used to c o m p u t e the corresponding average value of Cir.

upper lower both

Best tests

both

upper lower both

Sampling Level

All tests

Choice of Tests

A

Series

T A B L E 2--Comparison of unconfined compression tests.

D

0.91

1.01 1.06 1.05

0.81 0.84 0.85

0.81 0.84 0.83

C/~ avg

-I

--I m

O -
o I0

09

0.8

0'75 6 789

I0

2

3

4 5 6 789

I00

pressure,

2

3

4

5 6 789 I000

2

3

kPo

FIG. 6----Comparison of consolidation curves from block samples before and after storage

period.

Conclusion The present paper is a report on a study o f the influence o f reconsolidation and storage time on the strength and consolidation characteristics o f sensitive cemented clays from eastern Canada. From the data presented, the following tentative conclusions are suggested. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

LA ROCHELLE ET AL ON STORAGE AND RECONSOLIDATION 145

1. In the case of tube samples of soft and weakly cemented clays, a decrease of undrained strength on the order of 15 percent takes place during the first "few days after sampling, under conditions of complete stress release. 2. The decrease of strength may possibly be attributed to water migration from the disturbed outer zone of the sample towards the central part. From the water content measurements across samples, it is reasoned that larger diameter samplers would attenuate this detrimental effect. It is also suggested, pending further studies, that, in more strongly cemented clays, the effect of water migration might not be as pronounced. 3. The reconsolidation of samples under the field stresses does restore at least part of the lost strength and rigidity. However, it is also shown that, on good quality samples, the effect of reconsolidation is negligible. 4. During long periods of storage in humid rooms, block samples of medium and strongly-cemented clays have suffered a reduction of undrained shear strength on the order of 10 to 20 percent. However, the preconsolidation pressure and the general shape of the consolidation curve have not been affected by storage. Many of the observations presented in this paper stress the detrimental effect of sampling disturbance and the importance of obtaining good quality samples for elaborate programs of testing.

Acknowledgments The field and laboratory work reported in this paper were carried out by J. Sarrailh, graduate student. The help of J. P. Dussault, J. Y. Julien, S. Par6, and M. Pouliot, technicians, is appreciated greatly. This investigation was carried out with the financial support of the Ministry of Education of the Province of Quebec and the National Research Council of Canada. References [1] Hvorslev, J., "Subsurface Exploration and Sampling of Soils for Civil Engineering Purposes," Waterways Experiment Station, Vicksburg, Miss., 1949, pp. 163-164. [2] Bjerrum, L. "Problems of Soil Mechanics and Construction on Soft Clays," Stateof-the-Art Report to Session IV, 8th International Conference on Soil Mech~ics and Foundation Engineering, Moscow, Vol. 3, 1973, pp. 1l 1-159. [3] Crawford, C. B., Geotechnique, Vol. 13, No. 2, 1963, pp. 132-146. [4] Conlon, R. J., Canadian Geotechnical Journal, Vol. 3, No. 3, 1966, pp. 113-144. [5] Mitchell, R. J., Canadian Geotechnical Journal, Vol. 7, No. 3, 1970, pp. 297-312. [6] La Rochelle, P. and Lefebvre, G. in Sampling of Soil and Rock, ASTM STP 483, American Society for Testing and Materials, 1971, pp. 143-163. [7] Sangrey, D. A., Geotechnique, Vol. 22, No. 1, pp. 139-152. [8] Tavenas, F. A., Chapeau, C., La Rochelle, P., and Roy, M., Canadian Geotechnical Journal, Vol. 11, No. 1, 1974, pp. 109-141. [9] La Rochelle, P., Trak, B., Travenas, F. A., and Roy, M., Canadian Geotechnical Journal, Vol. 11, No. 1, 1974, pp. 142-164.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

146

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

[10] Bjerrum, L., Geotechnique, Vol. 17, No. 2, 1967, pp. 83-119. [11] Lefebvre, G. and La Rochelle, P., Canadian Geotechnical Journal, Vol. 11, No. 1, 1974, pp. 89-108. [12] Tavenas, F. A., Chagnon, J. Y., and La Rochelle, P., Canadian Geotechnical Journal, Vol. 8, No. 3, 1971, pp. 463-478. [13] La Rochelle, P., discussion on the State-of-the-Art Report to Session IV, 8th International Conference on Soil Mechanics and Foundation Engineering, Vol. 4.2, Moscow, 1973, pp. 102-108. [14] Tavenas, F. A., Blanchette, G., Leroueil, S., Roy, M., and La Rochefie P., "Difficulties in the In Situ Determination of Ko in Soft Sensitive Clays," Specialty Conference on In Situ Measurement of Soil Properties, American Society of Civil Engineers, Raleigh, N.C., June 1975, Vol. 1, pp. 450--476. [15] Raymond, G. P., Townsend, D. L., and Lojkasek, M. J., Canadian Geotechnical Journal, Vol. 8, No. 4, 1971, pp. 546-557. [16] Brooker, E. W. and Ireland, H. O., Canadian Geotechnical Journal, Vol. 2, No. 1, 1965, pp. 1-15. [17] Bozozuk, M. in Sampling of Soil and Rock, A S T M STP 483, American Society for Testing and Materials, 1971, pp. 121-131. [18] Lo, K. Y. and Morin, J. P., Canadian Geotechnical Journal, Vol. 9, No. 3, 1972, pp. 261-277. [19] Lefebvre, G., "Contribution h l'~tude de la stabilit6 des pentes dans les argiles ciment6es," Ph.D. thesis, Universitd Laval, Qu6bec, Canada, 1970.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

J. K. Torrance'

Pore Water Extraction and the Effect of Sample Storage on the Pore Water Chemistry of Leda Clay

REFERENCE: Torrance, J. K., "Pore Water Extraction and the Effect of Sample Storage on the Pore Water Chemistry of Leda Clay," Soil Specimen Preparation for Laboratory Testing, ASTM STP 599, American Society for Testing and Materials, 1976, pp. 147-157. ABSTRACT: In recent years, the potential significance of chemical and mineralogical influences on the behavior of the post-glacial marine clays has been increasingly recognized by the soils engineering community. The precise relationships are not known, but it seems probable that, in some cases, small differences in chemical factors may explain the differences in behavior between otherwise similar samples. In connection with an investigation of these relationships, the experiments reported in this paper were undertaken to examine the magnitude of chemical change which may occur in low-salinity Leda clay during periods o f storage. Pore water extraction devices are described, and the effects on the pore water chemistry of three months storage, under a variety of standard and modified storage procedures, are reported. It is concluded that none of the storage procedures tested is entirely satisfactory, in that potentially significant changes in the pore water chemistry occurred. Finally, it is recommended that pore water chemistry be assessed more often than is the present practice when soils engineering tests are performed on Leda clay and that this be done as soon as possible after the sample is obtained from the field. KEY WORDS: soils, tests, clays, water chemistry, moisture content, storage procedures

In most soils engineering studies, it is considered sufficient, in addition to the standard engineering measurements, to describe the geological origin of the material, certain physical properties such as texture and density, and the dominant minerals present. Rarely is any description of the chemical state of the soil system included. This is acceptable for many soil materials, but there are instances where the lack of chemical information may lead to incomplete understanding of results or to misinterpretation. An example of the latter possibility occurs when one is working with the 'Associate professor, Department of Geography, Carleton University, Ottawa, Ont. Canada.

147 Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by ASTM International www.astm.org Copyright 9 1976by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

148

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

post-glacial marine clays. The influence of chemical factors on marine clay behavior has been examined in Norway by various investigators working at the Norwegian Geotechnical Institute [1-3]. 2 In Canada, the main work to date has been published by Penner [4], Sangrey and Paul [5], and Torrance [6]. Bjerrum [1] found that, for the Norwegian marine clays, a relationship exists between the sensitivity and the salinity; namely, the sensitivity increases as the salinity decreases. In these materials, it also has been found [2,3] that relatively small differences in the concentrations of certain ions, at low but essentially constant pore water salinity, can explain differences in behavior between otherwise similar soils. The investigation of the role of chemical factors in the Canadian Leda clay has been less extensive. Penner [4] observed a relationship between electrokinetic potential and sensitivity for the Leda clays and noted that certain chemicals added to the soil affected its behavior. Sangrey and Paul [5] investigated the in situ pore water chemistry at sites in the Ottawa area and suggested that the sodium/calcium ratio in the pore water was an index of the depositional origin of the material and its susceptibility to a certain type of landslide. Further work at the Geotechnical Science Laboratories at Carleton University in Ottawa has extended these investigations and has shown, for individual samples of Leda clay, that a salinity-sensitivity relationship exists, similar to that observed in Norway [6]. The relationship, however, may vary greatly from one sample to another. These differences are thought to be related to other differences, that is, physical, mineralogical, and chemical, between samples. It was noted that, as in Norway, the various cations have different effectiveness in altering the soil behavior. It also has been found that the soduim/calcium ratio is inadequate as an index of origin since it depends on both the depositional origin and the degree of leaching and weathering of the Leda clay [7]. The problem of determining the importance of chemical factors in Leda clay is Complicated by the extreme variability of other factors known to influence its behavior. Experiments to date indicate that the relationships to be expected, at least for certain variants of Leda clay, will follow a pattern similar to that observed in Scandinavia. Among the most interesting of the Norwegian findings is the effect of low concentrations of various ions on the soil behavior [2,3,8]. If similar effects are important in Canada, then, in view of the nonuniformity of Leda clay, care must be taken to obtain the most reliable chemical information possible. In many engineering investigations there is a period of storage, usually under controlled conditions, before the laboratory investigation is undertaken or before all tests are completed. This storage period represents a time during which chemical changes may occur within the sample. The present investigation was under2Theitalic numbers in brackets refer to the list of referencesappended to this paper. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

TORRANCE ON PORE WATER CHEMISTRY OF LEDA CLAY

149

taken to determine the possible magnitude of changes in the pore water chemistry of a low-salinity Leda clay during storage under a variety of conditions.

Soil Material The soil material used in this investigation was taken in September 1974 from the site of the landslide of May 1973, north of Chelsea, Quebec. The Shelby tube samples were taken by hand from the base of the scarp at the side of the landslide scar. All samples were obtained from 10 to 12 ft below the surface and from an area measuring approximately 1 by 3 ft. The natural water content of the soil was normally between 55 and 65 percent (although higher and lower values were observed), the liquid limit ranged from 37 to 41 percent, and the plastic limit from 21 to 25 percent. Sensitivity, determined by the fall cone method, was between 10 and 20. This low value for the sensitivity was the result of the soil material coming from a near-surface location, in which it had been mildly affected by weathering. This weathering has caused an increase in the remolded shear strength.

Storage Procedures Standard storage techniques and some modifications of standard methods were tested to assess the amount of change which occurred in the pore water chemistry. The following storage procedures were used: 1. Left in the original Shelby tube. 2. Aluminum foil wrapped and waxed. 3. Plastic film wrapped and waxed. 4. Waxed only. 5. Extruded and placed in sealed plastic container with a nitrogen atmosphere. 6. Extruded, wrapped in plastic film, and placed in a sealed plastic container with a nitrogen atmosphere. 7. Extruded, and placed in a sealed plastic container with an air atmosphere. For all treatments except Treatment 1, the soil was extruded from the sample tube Within four days of sampling. Four inch samples of soil were used in each test. Before being prepared for storage, a slice of soil was taken along the full length of each segment. The surface of this slice, which had been in contact with the tube, was removed by scraping, and the remainder was subdivided lengthwise to allow water content determination on one portion and pore water extraction from the other. This procedure allowed the initial water content and pore water chemistry of the sample to be determined for comparison with the results obtained after storage. Samples prepared in the manner described were stored under conditions Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

150

SOIL SPECIMEN PREPARATION

FOR LABORATORY TESTING

o f room temperature (20~ and under conditions of high humidity at a temperature approximating the mean annual temperature of the sampling site (7 ~ Treatment 1 was carried out only under the latter conditions. The storage period reported in this paper was three months. While longer duration storage periods are being studied, three months is considered applicable to most commercial situations and is, therefore, of practical interest. Pore Water Extraction

The quality o f a pore water sample depends on the method by which it is obtained. Pore water extraction apparatus should meet certain basic requirements as to the amount of pressure that can be applied, and, at the same time, evaporation of water from the soil during extraction and from the extract while it is being collected should be minimal. Two different devices which are used in our laboratories to extract pore water from Leda clay will be described. The choice o f device depends mainly on the remolded shear strength of the soil. When samples with a remolded shear strength below approximately 0.5 t o n / m ~ are encountered, an air-activated pore water press has proven satisfactory (Fig. 1). This device consists o f a plastic sample chamber into which a disturbed sample of the soil is placed. At one end, there is a 1 mm pressure

applied

f r o m air c y l i n d e r

tcm I

I

/ I/ -metal

top plate

-rubber

gasket

- dental

dam

- s a m p l e chamber -connecting

. . . . . . . . . . . . . . . .

bolt

-protective

filter paper

-fine-pored

filter paper

plastic

outflow tube

T to

test

tube

FIG. l--Air-activated pore water press (schematic).

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

TORRANCE ON PORE WATER CHEMISTRY OF LEDA CLAY

151

diameter opening which, during operation, is covered by a fine-pored filter paper. This, in turn, is protected physically by one or more standard filter papers. A fine plastic tube (as used with atomic absorption photometers) is connected to the opening and leads to the collecting test tube. The other end of the chamber is covered by a thin rubber membrane (dental dam) and a rigid top plate (which is fitted with connections to allow application of air pressure). A rubber gasket between the dental dam and the top plate has been found to lengthen the membrane life. The apparatus is assembled and pressure applied (slowly at first to prevent puncture of the filter paper) from a cylinder of compressed air. The thin rubber membrane applies pressure to the sample, while preventing the air from passing through the sample and evaporating an unknown quantity of pore water. The present apparatus resists air leakage for pressures up to 6 to 7 atm. If the sample chamber which measures 1 in. depp by 2 in. diameter is half full or more, 1 to 5 ml of pore water can be usually obtained in less than 1 h. This apparatus in inexpensive and fast, and it is possible to use the same air cylinder to apply pressure to a number of sample chambers simultaneously. When pore water is required from stiffer samples, a mechanical pore water press is used (Fig. 2). The design is a minor modification of a pore water press that the author first saw in operation at the Norwegian Geotechnical Institute. The soil chamber is constructed of steel, the 1-mmdiameter exit is protected by layers of filter paper, and pressure is applied to the disturbed soil sample through a steel piston. This device is slow to operate but is necessary with stiff samples and in cases when only a small amount of soil is available. The concentrations of sodium, calcium, magnesium, and potassium in the extracted pore water were determined using a Jarrell-Ash atomic absorption spectrophotometer. Results and Discussions

The concentrations of sodium, calcium, magnesium, and potassium found in the pore water and the water contents at the start of the experiments and those found after three months storage for each treatment are presented in Table 1. The results show that considerable change in the pore water chemistry can occur over this time period. The differences in water contents before and after storage create a problem in making comparisons with the aluminum foil wrapped and waxed samples at 7 ~ and with the waxed only samples at both temperatures. In these cases, the data suggest that significant water loss occurred during storage. The samples, however, showed no visual evidence of major water loss, and it seems probable that the differing water contents before and after storage represent, in large measure, natural water content variations. The substantial increase in water content for one of the plastic wrapped and waxed samples and the

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized

152

i 1 cm

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

I

~

~. . . . . . . . . . . . .

plunger

[

~'~::'J" . . . .

O-rings

.........

sample chamber

..........

protective filter paper

~----i~------fone'iP;~ ed f'lter paper

...............

FIG.

plastic

outflow tube

2--Mechanical pore water press--all materials are steel unless otherwise indicated.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

7 20

7 20

7 20

7 20

7 2O

Plastic wrapwaxed(duplicate samples)

Waxed

Container N~

Container, plastic wrap I',I2

Container, air

48.4 45.0 51.3 61.4

50.3 50.9 56.0

44.0 50.4

58.7 65.5

57.1 53.5

49.4 49.9

55.5 55.1 59.3

50.7 58.5

58.0 64.2

58.7 56.4

51.9 55.8

49.6 48.7 70.5 64.2

57.5 57.1 53.0 58.4

Initial

3 Months

34 24

22 26

18 20

28 30

24 29 34

24 16 23 16

16 to 34

Initial, ppm

32 61

20 30

27 34

39 20

45 21 29

20 49 35 56

18.8 19.5 22.5 16.5

3 Months, ppm

Na

-6 150

-9 15

50 70

40 - 33

88 - 27 - 15

-17 200 52 250

Change, ~

18 22

25 19

23 17

18 20

18 18 20

23 21 18 14

14 to 25

Initial, ppm

65 107

77 77

129 107

31 31

30 26 26

28 72 25 21

48 57 55 36

3 Months, ppm

Ca

260 390

210 300

460 530

72 55

67 44 30

22 240 39 50

Change, 07o

7.9 11.5

11 8.2

11.5 8.2

7.4 8.2

7.2 7.8 8.1

9.7 10 8.2 7.9

7 to 12

Initial, ppm

40 68

43 45

93 67

16.5 16

12 13 13

13 41 12 13

26 31 27 15

3 Months, ppm

Mg

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

7 20

7

Storage Temperature

Water Content, o7o

400 490

290 450

7(30 720

110 95

67 40 38

34 310 45 65

Change, ~

21 21

26 20

20 19

20 22

20 Ig 20

19 17 19 17

17 to 22

Initial, ppm

42 42

34 36

49 48

24 29

32 24 22

20 33 17 18

40 33 31 24

Months, ppm

3

K

100 100

30 80

145 150

20 32

60 33 lO

5 95 - lO 6

Change, ~

analysis of pore waterfor Leda clay samples from Chelsea, Quebec at beginning of experiments and after three months storage.

Aluminum foil, waxed (duplicate samples)

Shelby tube, ends waxed (4 samples from tube)

Treatment

TABLE l--Chemical

ol

..k

r-

0

"11 rm o

0

.
--~" ,,__~, s ...C;.~-

-

I Specimen No. 58-13 ---~...~~ ~'" ~=~ . . . . . . . . . . . . . . . X--"" v

~

.... .. .. . . . .

-

58-17 -

l

No.

Specimen

~

Specimen

I

!

I b

I

L

\ . ~o---'" "~ 10 { l'_~'-_L'-f_-2"~-~ -v-

=

~"

9~

30

O1

~O

09

F"

o

09

c -n > -4 m

09

.


>

-u

O z

z

m

O -r

O I C >

246

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

partially saturated soils presented in this paper indicate that the soil water suction in a test specimen is an important factor affecting the stressstrain characteristics of the soil. Consequently, one of the primary requirements in the preparation of specimens of a partially saturated soil is to simulate the suction in the soil under prevalent field conditions. To satisfy this requirement, it is necessary to include the step of conditioning or pretesting treatment during the preparation of soil specimens for laboratory testing. The method of treatment as described in this paper is believed to be suitable for the aforementioned purpose, provided that the desired suction level during treatment is relatively low. If extremely high suction levels are to be maintained during treatment, apparatus and procedures much more complicated than those reported in this paper would be required.

Acknowledgments The investigations in this study were conducted at the University of South Carolina and were mostly in connection with a subgrade moisture research project sponsored by the South Carolina State Highway Department and the Federal Highway Administration. The authors wish to express their appreciation to all who have assisted in the research project and to some of the staff members in the soils department, TippettsAbbett-McCarthy-Stratton Engineers and Architects, New York, for their assistance during the preparation of this paper. References [1] Aitchison, G. D. and Richards, B. G. in Moisture Equilibria and Moisture Changes in Soils Beneath Covered Areas, Butterworth, Australia, 1965. [2] Russam, K., "Subgrade Moisture Studies by the British Road Research Laboratory," Highway Research Record No. 301, 1970, pp. 5-17. [3] Chu, T. Y., Humphries, W. K., and Chen, S. N. in Proceedings, 3rd International Conference on the Structural Design of Asphalt Pavements, 1972, pp. 53-66. [4] Chu, T. Y. et al, "Soil Moisture as a Factor in Subgrade Evaluation," to be published in Conference Proceedings, American Society of Civil Engineers Pavement Design Specialty Conference, 1975. [5] Kansas State Highway Commission, "Design of Flexible Pavements Using the Triaxial Compression Test," Bulletin 8, Highway Research Board, 1947. [6] McDowell, C., "Road Test Findings Utilized in Analysis of Texas Triaxial Method of Pavement Design," The AASHO Road Test, Proceedings of a Conference held May 1962, St. Louis, Mo., Special Report 73, Highway Research Board, 1962. [7] Seed, H. B., Chan, C. K., and Lee, C. E. in Proceedings, 1962 International Conference on Structural Design of Asphalt Pavements, University of Michigan, Ann Arbor, Mich., 1963, pp. 611-636. [8] Monismith, C. L., Seed, H. B., Mitry, F. G., and C'han, C. K. in Proceedings, Second International Conference on the Structural Design of Asphalt Pavements, 1968, pp. 109-140. [9] Escario, V. in Proceedings, Second International Research and Engineering Conference on Expansive Clay Soils, 1969, pp. 207-217.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

CHU AND CHEN ON PARTIALLY SATURATED SOILS

247

[10| Alpah, I. in Proceedings, 4th Imernational Conference Soil Mechanics and Foundations Engineering, Vol. I, 1957, p. 3. [11] Burland, J. B., Moisture Equilibria and Moisture Changes in Soils Beneath Covered Areas, Butterworths, Australia, 1965, p. 270. [12] Aitchison, G. D. and Richards, B. D. in Proceedings, Second International Research and Engineering Conference on Expansive Clay Soils, 1969, pp. 66-84. [13] Hunt, J. E. in Special Procedures for Testing Soil and Rock for Engineering Purposes, A S T M STP 479, American Society for Testing and Materials, 1970, pp. 192-197. [14] Croney, D., Coleman, J. D., and Black, W. P. M., "Movement and Distribution of Water in Relation to Highway Design and Performance," Special Report 40, Highway Research Board, 1958, pp. 226-252. [15] Aitchison, G. D. et al in Moisture Equilibria and Moisture Changes in Soils Beneath Covered Area, Butterworths, Australia, 1965, p. 7. [16] Chu, T. Y. and Mou, C. H. in Proceedings, Third International Conference on Expansive Soils, 1973, pp. 177-185. [17] Taylor, S. A., Evans, D. D., and Kemper, W. D., "Evaluating Soil Water," Bulletin 426, Agricultural Experiment Station, Utah State University, Logan, 1961, pp. 29-33.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

R. T. Donaghe' and F. C. Townsend'

Scalping and Replacement Effects on the Compaction Characteristics of Earth-Rock Mixtures

REFERENCE: Donaghe, R. T. and Townsend, F. C., "Scalping and Replacement Effects on the Compaction Characteristics of Earth-Rock Mixtures," Soil Specimen

Preparation for Laboratory Testing, ASTM STP 599, American Society for Testing and Materials, 1976, pp. 248-277. ABSTRACT: This investigation examines the validity of a scalping and replacement procedure used by many laboratories to determine compaction characteristics of earth-rock mixtures containing oversized particles. Two commonly used methods of computing densities of full-sized specimens, based upon results from tests performed on minus No. 4 fractions of the actual total sample, were examined. Compaction tests were performed on full-scale and scalped and replaced specimens, in which both gravel and fines content were varied. The test results indicate that the scalping and replacement procedure results in significantly lower maximum dry unit weights and higher optimum water content than are obtained for full-scale specimens. The use of the theoretical methods provided better approximations of experimental results on full-scale specimens having gravel contents up to 70 percent than did the relationships developed using the scalping and replacement procedure. For gravel contents above 70 percent, better approximations were obtained using the scalping and replacement procedure. KEY WORDS: soils, compaction, density, moisture content, earth-rock mixtures L a b o r a t o r y tests t o d e t e r m i n e c o m p a c t i o n characteristics o f e a r t h - r o c k m i x t u r e s for use in field c o n t r o l h a v e b e e n subject t o q u e s t i o n for m a n y years. D u e t o l i m i t a t i o n s o f e q u i p m e n t size, l a b o r a t o r y tests a r e g e n e r a l l y p e r f o r m e d o n s m a l l s p e c i m e n s , thus p l a c i n g a limit o n t h e m a x i m u m d i a m e t e r o f t h e particles w h i c h c a n b e used in the specimens. M a n y l a b o r a t o r i e s , i n c l u d i n g t h o s e o f t h e C o r p s o f Engineers, scalp o v e r s i z e d particles o f full-scale s p e c i m e n s a n d r e p l a c e t h e particles with an equal p e r c e n t a g e , b y weight, o f smaller p a n i c l e s , a s s u m i n g t h a t results f r o m tests p e r f o r m e d in small m o l d s o n such m a t e r i a l are c o m p a r a b l e to t h o s e U.S. Army civil engineering technician and research engineer, respectively, Soils Research Facility, Waterways Experiment Station, Vicksburg, Miss. 39180. 248 Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by by ASTM International www.astm.org Copyright9 1976 University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

DONAGHE AND TOWNSEND ON SCALPING AND REPLACEMENT

249

obtained from tests performed on full-scale specimens in large molds) Other laboratories perform compaction tests on the minus No. 4 fraction of soils in small molds and then apply theoretical correction factors, based on the influence of gravel on compaction characteristics, to compute densities of the full-scale specimens containing gravel. In both cases, however, various investigators have found that additional modifications have to be made in the small-scale test results in order to obtain agreement with the results of tests on full-scale specimens. The objectives of this investigation were to determine the validity of the Corps of Engineers scalping and replacement procedure and to evaluate theoretical methods of computing dry unit weights of full-scale specimens, using results from tests performed on minus No. 4 fractions of the total specimen. These objectives were achieved by comparing compaction curves for full-scale specimens (3 in. maximum particle size) and scalped and replaced specimens ( 90 in. maximum particle size), determined by using a mechanical compactor with 18 and 6-in.-diameter molds, respectively. Additional comparisons involved compaction curves for scalped and replaced specimens ( 90 in. maximum particle size), determined by using a hand-held hammer with the 6-in.-diameter mold. Companion compaction curves on the minus No. 4 fraction of the full-scale material were determined by using a hand-held hammer and 4-in.-diameter mold. Effects investigated are given in Table 1.

Procedure

Equipment A mechanical compactor, manufactured by Howard Company and equipped with 5.5 and 24.7-1b rammers, having face diameters of 2.0 and 6.0 in., respectively, was used to perform the testing. Large-scale tests were performed using an 18-in.-diameter mold and 24.7Ab rammer, while small-scale tests were performed using a 6-in.-diameter mold and 5.5-1b rammer. A special harness for suspending the 18-in.-diameter mold and specimen from a forklift was rigged with an electronic load cell, sensitive to within 0.1 lb, to obtain the specimen-plus-mold weights. A photograph of the compactor, the 24.7-1b rammer, the 18-in.-diameter mold, and the weighing harness is shown in Fig. 1.3 All other weights ZThe Corps of Engineers scalping and replacement procedure for materials containing particles larger than 2 in. for compactingin a 12-in.-diametermoldis givenin the appendix. 3The 12-in.-diameter mold shown with the weighing harness was not used in this investigation. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

18

6

mechanical

mechanical

Gravel content (large-scale tests)

Gravel content (small-scale tests)

No. 4 sieve 3b

0 10 20 30 40 50 60 100

0 10 20 30 40 50 60 100

40

40

90

90 3b

0

Gravel Content, 070

No. 4 sieve

Max Particle Size

No. 4 sieve a/~a

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

18

mechanical

Removal and replacement o f coarse particles

6 18 6 18

mechanical

Mold Dia, in.

Equipment size

Effect Investigated

Type o f Compactor Used

T A B L E l--Summary o f compaction data.

25

25

25

25

25

Fines Content, 070

130.9 133.5 132.5 132.3 131.1 131.9 129.5 103.6

133.9 135.0 136.1 137.2 138.0 137.1 134.9 112.0

134.1 138.0

130.9 133.9 131.1 134.1

M a x Dry Unit Weight ),a, lb/ft 3

C

8.6 7.7 8.1 8.1 7.9 7.8 9.5

...C

6.9 6.7 6.1 5.7 5.9 5.8 5.2

7.3 5.9

8.6 6.9 7.9 7.3

Optimum Water w, 07o

-4 m .--t

0

0

i--

z -11 O

---i

t'"u 11

ill z

rn

I-"

co O

ol 0

6

mechanical

mechanical

Fines content (large-scale tests)

Fines content (small-scale tests)

40

40

3b

90

0 10 20 30 40 50 60 100

No. 4 sieve 3Aa

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

a Scalped and replaced material. b Full-scale material. cSingle-point test performed on dry material.

18

hand-held

Gravel content (small-scale tests)

15 25 35

15 25 35

25

134.8 131.1 126.6

141.8 138.0 133.3

132.6 132.5 131.8 132.0 132.0 129.3 128.5 101.2

7.4 7.9 9.4

4.9 5.9 7.5

8.0 7.8 7.9 7.9 7.9 8.0 9.2 ...c

...g

bO

z ..-t

m

m

m "10 r'-

z o

z

r'-

0 z

z m Z

Z o .-I 0

32 m

z

O

252

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

FIO. l--Howard mechanical compactor with 12 and 18-in.-diameter molds and load cell harness f o r weighing.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

DONAGHE AND TOWNSEND ON SCALPING AND REPLACEMENT

253

were obtained by using a scale, having a sensitivity of 0.01 lb. Calibration data for the molds were: Nominal Mold Inside Diameter, in.

Area, ft2

Height, in.

Volume, ft 3

Weight, lb

6 18

0.196 1.767

4.582 18.005

0.075 2.651

6.21 321.40

Equipment used for tests with the hand-held hammer conformed to that used for standard compaction tests performed by the Corps of Engineers [1]4 and is similar to that specified in ASTM Tests for Moisture-Density Relations of Soils, Using 5.5-1b Rammer and 12-in. Drop (D 698-70), except the rammer is equipped with a sliding weight instead of a guidesleeve.

Material The materials tested in this investigation consisted of a subrounded to subangular washed gravel, having a maximum particle size of 3 in., a subrounded to subangular concrete mortar sand, and a clay (CL), combined according to the gradations given in Figs. 2 and 3. Classification data for the clay and sand are given in Fig. 4.

Specimen Preparation Batches for specimens were prepared by thoroughly mixing a predetermined amount of sand and CL material mixture (minus No. 4 material) with a measured quantity of water, using either a commercial kitchen mixer or a pugmill (specimens tested in 4 and 6-in.-diameter molds were mixed with the kitchen mixer). The resulting minus No. 4 material was then stored in airtight containers and allowed to cure for a period of at least 16 h. The plus No. 4 material for each batch was prepared by combining the air-dry portion (by weight) of material required for each sieve and then storing the resulting material in containers filled with water. Immediately prior to compaction, the cured minus No. 4 fraction was mixed with the saturated surface-dry aggregate. Each layer was hatched separately to prevent any variations in grading between layers. 4The italic numbers in brackets refer to the list of references appended to this paper. Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

254

SOIL SPECIMEN PREPARATION FOR LABORATORY TESTING

I

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

DONAGHE AND TOWNSEND ON SCALPING AND REPLACEMENT

255

.O

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

FIG. 4 - - G r a i n size distribution c u r v e s a n d classification data, C L material a n d sand.

z

-t m

0.'n

0

r-

0

Z

-t

"in

m z "u :I1 m -o

o_

"IJ m

r-

r 0

t~

DONAGHE AND TOWNSEND ON SCALPING AND REPLACEMENT

257

Testing Procedure Table 2 lists pertinent data concerning the rammer sizes and the compaction procedure. All specimens were compacted in three layers, using compactive efforts made approximately equal to the standard effort (12 300 ft. lb/ft 3) by adjusting the number of blows. Test Results and Discussion

Results of the compaction tests are summarized in Table 1 and presented graphically in Figs. 5 through 15.

Effects of Equipment Figure 5 presents results of tests performed, using both large- and small-scale equipment, on scalped and replaced specimens and minus No. 4 sieve specimens, having gravel contents of 40 and 0 percent, with 25 percent fines. The data indicate that, for both gradations, maximum dry unit weight and optimtun water content varied with the different sizes of equipment used. Optimum water contents of the 0 and 40 percent gravel specimens were decreased by 1.7 and 0.6 percent, respectively, when the 18-in.-diameter mold and 24.7-1b rammer were used in place of the 6-in.-diameter mold and 5.5-1b rammer. The maximum dry unit weight of both materials was increased by 3.0 lb/fP for the same change in sizes of equipment. This difference in maximum dry unit weight falls within a range comparable to that determined by previous investigators [2-4], who indicated that the variation in densities determined using various mold sizes was limited to not more than 4 lb/ft 3. Since only limited testing was performed to determine the effects of variation in equipment size, no attempt was made to correct other test results for these effects. When comparisons which involve both large- and smallscale test results are made, however, it appears safe to assume that differences in densities due to varying the size of equipment used are limitied to not more than 4 lb/ft 3.

Effect of Removal and Replacement of Oversize Particles Figure 6 shows compaction curves and data for the tests performed on the scalped and replaced specimen having 40 percent gravel and 25 percent fines and on the corresponding full-scale specimen, both of which were tested using large-scale equipment. The curves indicate that, in this case, scalping and replacing oversize particles results in a lower maximum dry unit weight and a higher optimum water content. The 3.9 lb/fP decrease in maximum dry unit weight due to scalping and replacement of

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions autho

12 420

12 299

6

18

24.7

5.5

lb

Weight,

6

2

Dia of Circular Face, in.

24

12

Drop, in.

3

3

No. of Layers

220

56

Blows per Layer

7

2

24 b

24

Complete Coverage Along of CircumferMold Area enceof Mold

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

6c

4

At Center of Specimen

Blows per Coverage

aCompaction equipment used: hand and mechanical compaction: 6-in.-dia mold, mechanical compaction: 18-in.-dia mold. bPius 6 on last coverage. Cplus 4 on last coverage.

Compactive Effort, ft'lb/ft ~

Mold Dia, in.

Rammer a

TABLE 2--Compaction tests using hand-heM sliding weight rammer and Howard mechanical compactor.

Ol

O

O -< .--I m (D -4

t-t~ O :D

z "11 O :lJ

-t

"o

m

z

m

o

fD "O m

__q I'-

t,D

Oo

DONAGHE

AND

TOWNSEND

ON

SCALPING

AND

259

REPLACEMENT

T E S T GRADATIONS 4SIEVE 3

0PENINGS=I = 3/4 4

10

SIEVE NUMBERS 40

200

0

I00

i\

-I 8 O -I -I Z E

I 60

F7

i

-I -I

I I

~4o Q.

!~---o~

cr 20

"i\[ i \

I

U

I

I

I -I

' ~

'

I 1 I I

~ I I I

6o

j

2 0 rl

w

I I I I

i' GRAVEL

: ! :

SAND

BO

Q.-

I00

: ]:

FINES~

140

135

40%Gr(18"MO

V'LD~~.~

Q. ~~ I.:I: ,,:} bJ

I--

,25 I

O~Gr( ~

xll

c}

GRAVEL

1,9.0 -

I

O P T W, % M A X . "Xd, P C F 115 2

CONTENT,

% ( F i = 25%)

0

40

6" M O L D

18" M O L D

6" M O L D

18" M O L D

8.6 130.9

6.9 133.9

7.9 131.1

7.3 134.1

I

I

I

I

.I

4

6

8

10

12

WATER CONTENT p ~

FIG. 5--Compaction curves f o r tests to determine effect o f varying mold diameter.

Copyright by ASTM Int'l (all rights reserved); Wed Dec 22 14:00:00 EST 2010 Downloaded/printed by University of British Columbia Library pursuant to License Agreement. No further reproductions authorized.

260

SOIL SPECIMEN

PREPARATION

TEST

I00

80 Z k,IZ

60

.

:

I0 I

SIEVE NUMBERS 40 200

P 0

ki \ ,,..--3/4", MAXl,

,"~,..\ ,I ~. M~.~~..~;

_,

TESTING

GRADATIONS

~51EVE 0PENINGS;I 3 3/4 4 i

\~

FOR LABORATORY

20m

,l

r~.

4o