BEHAVIOR OF AH INSTRUMENTED MODELED PIER
Jonathan Paul Friedman
B.S. Geological Engineering, Colorado School of Mines, 1983
A thesis submitted to the Faculty of the Graduate School at the University of Colorado in partial fulfillment of the requirements for the degree of Masters of Science Department of Civil Engineering 1993
This thesis for the Masters of Science degree by Jonathan Paul Friedman has been approved for the Department of Civil Engineering by
Nlen Y. Chang
Fu Hua Chen
Friedman, Jonathan Paul (M. S., Civil Engineering) Behavior of an instrumented, modeled pier in expansive clay.
Thesis directed by Professor Nien Y. Chang.
In recognition of continued residential and commercial growth in areas of swelling soils and bedrock units, alternative foundation designs have been developed over the years with varying degrees of success in preventing structural distress. One foundation type that has gained widespread popularity in providing suitable load bearing capacity and uplift resistance in swelling soils is the drilled, cast-in-place, reinforced concrete pier. This paper documents the behavior of a modeled, instrumented, reinforced concrete pier installed in swelling soil.
The model design and instrumentation was developed to measure several reactive variables which included: lateral earth pressures, strain distribution within the reinforcing bar of the pier, and the total uplifting force created by the swelling soil on the concrete pier.
The total uplifting force created by the swelling soil on the concrete pier was however, the primary objective of this experiment.
Two long term tests were generated. The initial test, which lasted 50 days, consisted of "ponding water at the model surface to simulate near surface infiltration of moisture that may result from localized landscape irrigation. The second test, which is continuing at the time this manuscript was written, consists of introducing moisture into the model via small diameter well casing to simulate a seasonal fluctuation in the groundwater level.
In comparing the results of background tests conducted on the soil and those to the pier model test, it has been shown that much greater reactions should have been registered through the data acquisition system.
This was a time dependant reactive experiment yet it was uncertain as to how much time would be required for the infiltration of water into the model without permeability rates obtained from a permeability test. However, the measured results from the second long term test will most likely provide much needed data for verification of
analysis methodologies in the design of cast-in-place reinforced concrete piers.
The form and content of this thesis are approved. I recommend its publication.
ACKNOWLEDGEMENTS ..................................... xiv
1. INTRODUCTION ...................................... 1
1.1 Engineering Problems Associated With
Expansive Soils .............................. 1
1.2 Objective............................... 3
1.3 Scope of This Study..................... 4
2. ORIGINS AND PROPERTIES OF EXPANSIVE SOILS ... 6
2.1 General................................. 6
2.2 Geologic Environments ........................ 6
2.3 Distribution of Expansive Soils .............. 7
2.3.1 National Distribution of Expansive
Soils ................................. 8
2.3.2 Local Distribution of Expansive
Soil Along The Colorado Front Range 16
2.4 Mineralogical Properties Conducive to
Swelling .................................. 23
2.4.1 Clay Mineralogy ...................... 23
2.4.2 Clay Mineral Classification .... 26
2.4.3 Clay Mineral and Water Interaction . 35
2.5 Physical Properties Affecting Swelling . 39
2.5.1 Intrinsic Properties ................. 40
22.214.171.124 Soil Composition ............ 40
126.96.36.199 Dry Density.................. 41
188.8.131.52 Soil Fabric.................. 45
184.108.40.206 Pore Water Properties ... 46
220.127.116.11 Confinement ................. 47
18.104.22.168 Time......................... 48
22.214.171.124 Permeability ................ 49
126.96.36.199 Temperature ................. 50
188.8.131.52 Structure ................... 50
184.108.40.206 Cementation ................ 51
220.127.116.11 Atterberg Limits ........... 53
18.104.22.168 Soil Suction................ 54
2.6 Environmental Factors Affecting
2.6.1 Soil Profile.......................... 55
2.6.2 Climate............................... 56
2.6.3 Vegetation............................ 58
2.6.4 Surface Drainage ..................... 58
3. STRAIGHT SHAFT SINGLE PIERS; HISTORY AND DESIGN 60
3.1 Historical Development Of Piers ............. 60
3.2 Types of Drilled Piers....................... 62
3.3 Drilled Pier Construction ................... 63
3.4 Drilled Pier Design Criteria ................ 65
3.5 Drilled Pier Field Testing................... 71
3.6 Uplift Force and Interface
4. SOIL AND CONCRETE TESTING PROGRAM................. 76
4.1 General...................................... 76
4.2 Soil Characterization........................ 77
4.2.1 Soil Sample Preparation............... 77
4.2.2 Soil Index Properties.......... 78
4.2.3 Soil Swell Potential........... 79
22.214.171.124 Swell Potential of Initial
Soil Sample................. 85
126.96.36.199 Swell Potential of Soil
Containing Bentonite ... 88
4.2.4 Soil Shear Strength............ 97
4.3 Concrete Properties ........................ 109
4.3.1 Aggregate Grain Size Distribution
and Cement Content...................110
4.3.2 Unconfined Compressive Strength . 110
4.4 Soil/Concrete Inter-face Strength .......... Ill
5. INSTRUMENTED PIER MODEL...........................116
5.2 Components and Design of the Pier Model . 116
5.3 Model Instrumentation...................... . 125
5.3.2 Data Acquisition System...............126
188.8.131.52 Instrumentation ........... 127
184.108.40.206 Data Acquisition Software
220.127.116.11 Sampling Rate...............137
5.3.3 Software Operation ................137
18.104.22.168 Test Duration ..............138
22.214.171.124 Data Management and
5.3.4 Instrument Calibration ............. 140
126.96.36.199 Tension Load Cell...........142
188.8.131.52 Total Earth Pressure
184.108.40.206 Strain Gages................143
5.4 Soil Sample Preparation......................148
5.5 Model Construction and Testing
5.6 Predicted Results .......................... 156
6. ANALYSIS AND DISCUSSION OF RESULTS................159
6.1 General Statement .......................... 159
6.2 Moisture Variation With Time.................161
6.3 Swell Pressure Measurements ................ 161
6.4 Load Cell Measurements.......................162
6.5 Stress Distribution Determined from Strain
6.6 Uplift Force and Soil/Concrete Interface
Relationship .............................. 167
7. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
FOR FUTURE RESEARCH.............................170
7.3 Recommendations for Future Research .... 172
Figure 2.1 Schematic representation of particle orientation and the 'cardhouse' structure in clays (after Rosenquist,
Figure 2.2 Distribution of potentially expansive
materials in the United States: FHWA Regions 1,
3 & 5, (After Smith, 1975)........................ 10
Figure 2.3 Distribution of potentially expansive
materials in the United States: FHWA Region 4,
(After Smith, 1975)............................... 11
Figure 2.4 Distribution of potentially expansive
materials in the United States: FHWA Region 6,
(After Smith, 1975)............................... 12
Figure 2.5 Distribution of potentially expansive
materials in the Unite States: FHWA Regions 7 &
8, (After Smith, 1975)............................ 13
Figure 2.6 Distribution of potentially expansive materials in the United States: FHWA Regions
9 & 10, (After Smith, 1975)....................... 14
Figure 2.7 Area location map showing the Colorado
Front Range Urban Corridor, (After Hart, 1974). 17
Figure 2.8 Generalized geologic map of the Colorado Front Range Urban Corridor, (after Hart,
Figure 2.9 Stratigraphic section of bedrock units that occur within the Colorado Front Range
Urban Corridor, (after Hart, 1974)................ 20
Figure 2.10 Silica Tetrahedron Sheet (after Grim,
Figure 2.11 Alumina Octahedral Sheet (after Grim,
Figure 2.12 Structure of the kaolinite clay
particle (after Grim, 1953)....................... 28
Figure 2.13 Structure of the illite clay particle
(after Grim, 1953)................................ 29
Figure 2.14 Structure of the montmorillonite clay
particle (after Grim, 1953)....................... 30
Figure 2.15 Molding water content versus dry
density and particle orientation for Boston
Blue clay, (after Pacey, 1956).................... 43
Figure 2.16 Molding water content versus dry
density and particle orientation for kaolinite
(after Seed and Chan, 1959)..................... 44
Figure 4.1 Swell potential curve for soil sample
S-l. Dry Density 95.2 pcf, Moisture Content 11.9%
(Max. Dry Density 102.6pcf., OMC 19.7).......... 86
Figure 4.2 Swell potential curve for soil sample
S-2. Dry Density 98.9 pcf, Moisture Content 14.5%
(Maximum Dry Density 102.6 pcf, OMC 19.7% ... 87
Figure 4.3 Swell potential curve for soil sample
S6B89. Dry Density 89.6 pcf, Moisture content 12.5% (Maximum Dry Density 103.0pcf, OMC
Figure 4.4 Swell potential curve for soil sample S6B108. Dry Density 108.7 pcf, Moisture Content 14.6% (Maximum Dry Density 103.0 pcf,
OMC 20.1%)......................................... 91
Figure 4.5 Swell potential curve for soil sample S6B107. Dry Density 107.2 pcf, Moisture Content 13.7% (Maximum Dry Density 1003.0 pcf,
OMC 20.1%)......................................... 92
Figure 4.6 Swell potential curve for soil sample S6B115. Dry Density 115.1 pcf, Moisture content 16.8% (Maximim Dry Density 103.0 pcf,
OMC 20.1%)......................................... 93
Figure 4.7 Swell potential curve for soil sample S6B116. Dry Density 116.0 pcf, Moisture Content 15.6% (Max. Dry Density 103.0 pcf, OMC
Figure 4.8 Swell potential curve for soil sample S6BG112. Dry Density 111.9 pcf, Moisture Content 17.2% (Maximum Dry Density 104.0 pcf,
OMC 20.5%)......................................... 95
Figure 4.9 Swell potential curve for soil sample S6BG114. Dry Density 113.9 pcf, Moisture Content 17.6% (Max. Dry Density 104.0 pcf, OMC
Figure 4.10 Swell pressure versus dry density for
samples containing 6% bentonite.................... 98
Figure 4.11 Dry density versus swell pressure for
samples containing 6% bentonite.......... 99
Figure 4.12 Volume change versus water content for
samples containing 6% bentonite..........100
Figure 4.13 Swell pressure versus relative
compaction for samples containing 6% bentonite. Figure 4.14 Dry density versus volume change for
samples containing 6% bentonite..................102
Figure 4.15 Swell pressure versus percent swell for
samples containing 6% bentonite..................103
Figure 4.16 Shear stress versus normal stress for
direct shear tests of soil samples prepared and tested at approximately 15 % moisture content. 108 Figure 4.17 Shear stress versus normal stress for direct shear tests of soil samples prepared at approximately 15 % moisture content and
concrete under dry conditions......................114
Figure 4.18 Shear stress versus normal stress for direct shear tests of soil samples prepared at approximately 15 % moisture content
and concrete under submerged conditions............115
Figure 5.1 Photograph of the threaded section of the
no. 8 reinforcing bar..............................119
Figure 5.2 Section type # 1 of model test
Figure 5.3 Section type # 2 of model test
Figure 5.4 Section type # 3 of model test
Figure 5.5 Section type # 4 of model test
Figure 5.6 Schematic diagram of data acquisition
Figure 5.7 Strain gage wiring schematic for full
Wheatstone bridge as applied to rebar..............132
Figure 5.8a Photograph of initial stage of strain gage application to rebar and wiring
Figure 5.8b Photograph of final stage of strain gage application to rebar and wiring
configuration with environmental protection. . 134
Figure 5.9 Photograph of strain gage calibration
Figure 5.10 Calibration curves of strain gage calibration Test No. 1. The average
calibration factor is 6.627xlO_2mV/lb..............146
Figure 5.11 Calibration curves of strain gage calibration Test No. 2. The average
calibration factor is 6.627xlO'2mV/lb..............147
Figure 5.12 Schematic of the model container and
Figure 5.13 Photograph of the model container,
instrumentation and data acquisition system. . 155
Figure 6.1 Graphical presentation of data recorded
by the data acquisition system.....................160
Figure 6.2 Graphical presentation of data recorded by the data acquisition system. Graph represents data from strain gage set #1 plotted at a reduced scale......................................165
Table 2.1 Typical values of free swell for common clay
minerals (after SnetHen, 1975; Grim, 1962)... 38
Table 4.1 Comparison between soil index properties of the initial soil and samples prepared by
adding 6% by dry weight............................ 80
Table 4.2 Swell potential characteristics of the
initial soil and samples prepared by adding 6%
bentonite by dry weight............................105
Table 4.3 Summary of direct shear tests performed
on soil under dry and submerged conditions. . 109
Table 4.4 Summary of direct shear tests performed on soil/concrete interface under dry and
Table 5.1 Comparison of the effect of cross
sectional area on strain for a given load using the diameters of the # 8 rebar smoothed
Table 5.2 Summary of pressure cell calibration
Table 5.3 Resulting uplift forces generated by the swelling pressures developed from various depths of wetting using a constant swelling pressure of 20,000 psf, and a 4 inch diameter
Table 6.1 Load determination as recorded by strain gage set number 1 during the Phase I
Sincere appreciation and thanks are extended to Professor Nien-Yin Chang, my advisor, for sharing his knowledge, financial assistance, his continued encouragement and optimism on this research. Professor Adjunct Fu Hua Chen and Dr. Judith Stalnaker are acknowledged for serving on my thesis review committee.
I would like to thank the members of my family, especially my parents, for their continued and unconditional encouragement during the course of the research. I can only hope that my friends and family understand the time commitment and concentrated labor that was demanded. The only means by which I can repay or perhaps justify my disposition is to offer them all a copy of this manuscript. For this is the only documentation I have that substantiates the amount of effort that has been required to complete this task.
Thanks and appreciation are extended to Mr. Rodney Rupp and Thompson Steel & Pipe Company for their contribution, fabrication and recommendations concerning the test model materials.
A special thanks is extended to Mr. Jan Chang for his most appreciated assistance in the laboratory. His expertise with soil laboratory testing equipment, and data reduction, were an immense help.
I would like to thank Mr. Apinop Suwintawong for his assistance with the data acquisition. If it were not for his computer programming skills, the automation of the data acquisition would not have been accomplished.
I would like to thank both Mr. Clyde Anderson of CTL Thompson, Inc. and CTL Thompson, Inc. for their assistance and generosity in providing power compactive tools, offering the use of a moisture density gage, and general construction suggestions. Mr. Anderson's enthusiasm was definitely encouraging.
I would like to thank Mr. Louis Lopez, Mr. Rick Flemming and Mr. Paul Lane of the Colorado State Highway Department for their general assistance in the model construction and the removal of the pier form in my absence. Their efforts are most appreciated.
The efforts of Mr. Terry MacCagnan, Mr. Jan Chang, and Professor N.Y. Chang in placing the concrete for the pier are greatly appreciated.
I would like to thank Mr. Craig Colby of GTG Fox Inc., for his efforts in arranging the delivery of the soil to the laboratory.
1.1 Engineering Problems Associated With Expansive Soils
Structural engineering problems due to swelling soils were first recognized in the United States during the 1930's. Frame type construction was, at that time, the common construction practice, and since these structures are not as rigid as concrete or brick, foundation movements caused by soils were not as obvious. With the increased use of stiffer, more rigid building materials, structures became less flexible and more brittle and the effects of differential movement became more apparent. Little was known about expansive soils at that time and many problems caused by swelling were erroneously diagnosed as being due to settlement (Fenner, Hamberg, and Nelson, 1983).
Swelling potential in bedrock units and soils derived from bedrock units containing certain clay minerals are prevalent throughout the southern and mid-western United States. Although the distribution of these soil types are well documented, it is impossible for modern civilization to deter development in these
areas. As reported by Jones and Holtz, 1973:
"Each year, shrinking soils inflict at least $2.3 billion in damages to houses, buildings, roads, and pipelines more than twice the damage from floods, hurricanes, tornadoes, and earthquakes1"
Even though this comment was made twenty years ago, the technical progress that has been made since that time is insufficient considering our present inability to properly engineer a structure against the potential damage that expansive soils create.
Structures founded on expansive clay soils can often induce heave caused by inhibited transpiration of moisture by vegetation and evaporation. Conversely, shrinkage can be induced if moisture is inhibited from entering the soil. For example, heat from the structures can cause foundation soils to lose moisture. The amount of heave that may occur in soils beneath structures is determined primarily by climatic conditions, water table depth, and the amount and activity of the clay present. The potential heave of highly expansive type of clay such as bentonite, can be considerable in arid areas where the water table is not near the ground surface. Climate conditions include such variables as the rate of
evapotranspiration and the amount and frequency of rainfall.
The gradual and uniform swelling of soils is generally not responsible for major structural damage, but rather it is differential heave of the foundation subsoil that can cause cracking in buildings and, often an apparent low in the corner areas. Differential heave can result from variations in lateral thickness of the clay strata, and variations in the water content beneath the structure.
The objective of this thesis is to provide a better understanding of the behavior of expansive soil material as a result of an increase in moisture content and the horizontal and vertical stress effects this has on straight shaft concrete piers. The results of the experiment will hopefully provide valuable information regarding; i) soil-pier interface reactions, ii) change in stress at depth within the pier, iii) change in stress with increased moisture content within the soil, and iv)
overall uplift forces created by expansive soil material on a concrete pier.
1.3 Scope of This Study
In this study, a modeled pier was designed, instrumented and constructed. The test container is cylindrical, measuring 10 feet in height and 2.5 feet in diameter. The concrete pier measured 10 feet in length and 4 inches in diameter. Four sets of strain gages mounted to the inner reinforcing bar of the pier, measured strain as a result of pressures created by the swelling soil. Earth pressure cells placed vertically measured the horizontal swelling forces within the soil.
A load cell placed at the base of the concrete pier measured the total uplift force.
The focus of this study was to document the reaction of an expansive clayey soil to induced moisture and to determine the distribution of the vertical stress created by the swelling soil on the reinforced concrete pier.
The scope of this study consisted of the following general steps:
1) Design a pier test apparatus that could be used inside the laboratory and be durable enough for several long term tests.
2) Obtain an adequate amount of soil for the test and determine the index properties and swell potential characteristics.
3) Determine the interface cohesion and friction between the concrete pier and soil.
4) Review available literature to determine characteristics of expansive soil and determine various methods of pier design and construction.
5) Design an instrumentation system capable of recording the swelling forces of the expansive clay and reactive mechanisms.
6) Calibrate the instruments used to monitor the reactions created by the swelling forces of the expansive clay.
6) Construct the pier test model.
7) Interpret and present the test results.
ORIGINS AND PROPERTIES OF EXPANSIVE SOILS
The origins and properties of expansive soils are well documented yet reactions in the field are only moderately understood. Expansive soils are located worldwide, yet are generally more reactive in semi-arid countries. Virtually every state in the United States has expansive soils however, the degree to which a soil will react is dependent on the interaction between the climate and soil conditions.
2.2 Geologic Environments
A common source of clay mineral deposits is the chemical weathering of such materials as feldspars, micas, and limestone. The mineral that forms depends on factors such as the parent rock, topography, climate, neighboring vegetation, and the weathering period. Bentonite, for example, is a clay composed primarily of montmorilUnite formed by the chemical weathering of volcanic ash. Illite is a weathering product of micas aided by the presence of potash.
The mineralogy of the clay particles governs the degree of expansive behavior. In addition, the geologic environment in which the clay particles are deposited
i.e., alluvial, colluvial or residually weathered bedrock, influences its behavior. Rosenquist, 1959, observed through electron microscopy, that the mineral arrangement in undisturbed clays does, in-fact, correspond to the cardhouse structure illustrated in Figure 2.1.
If the expansive material is a bedrock unit, its behavior will also be governed, to some degree, by the geologic properties of the formation i.e., inclined bedding, structural folds, and varying thicknesses of the deposit.
2.3 Distribution of Expansive Soils
Expansive soils are located worldwide, but are generally more troublesome in semi-arid regions.
Countries in which considerable research efforts and
design innovations have been developed include the United
States, Canada, Israel, South Africa, and Australia.
2.3.1 National Distribution of Expansive Soils
Soils of varying degrees of expansiveness are present throughout the entire United States. The areal distribution and degree of expansiveness of the materials within the United States are shown on Figure 2.2 through Figure 2.6, (Smith, 1975).
UNDlS l URBlD Salt Wafer Deposit
UNDISTURBED Fresh Wafer Deposit
Figure 2.1 Schematic representation of particle orientation and the 'cardhouse' structure in clays (after Rosenquist, 1959).
Ulllllj HIGH HIGHLY Â£ PAN Si WE ANO/OR HIGH ULULU FREQUENCY OF OCCURENCE
L0 GENERALLY OF LON EXPANSIVE CHARACTER ANO.'OR LOW FREQUENCY of OCCURRENCE
I ] NONEIPANjJVE: THE OCCURRENCE OF
--1 EXPANSIVE MATERIALS EXTREMELY
NOTE: FOUR CATCGORlCSOF EXPAN WEnÂ£SS ARE ViON ON THE OlSTRIBUTlON MAPS THESE CATEGOH'ES ARE GENERALIZED ANO QUALITATIVE AND ARE BASED UPON lE PRESENCE OF ONTUORlLLONl TE ANO THE RELATIVE FREOuEncy OF OCCURRENCE OF ARGILLACEOUS MATERIAL IN THE AREA. MAJOR CATEGORIZATIONS ARE BASED UPON STRATIGRAPHY AND PHYSIOGRAPHY. DESCRIPTIONS OF ThE PREDOMINANT GEOLOGIC FORMATIONS ARE &VEN In TABLE I. THE BASIS FOR THE CATEGORIZATION IS EXPLAINED IN PARAGRAPH 22 OF ThE TEAT.
Hap compiled t*f 0. H. Patiick. H. K. Hoods, and fiedtuct L. 1*itti. Enjincnrof Geo-logy and Rock Uecbaaici Division. U. I. Ainry Ingram laipwiys Â£ ipe* merit llalion. Vicks tug. Us.
Distribution of potentially States: FHWA Regions 1, 3 &
expansive materials in the United
5, (After Smith, 1975).
MEDIUM. MODERATELY EXPANSIVE AND/OR MODERATE FREQUENCY OF OCCURRENCE
LOW. GENERALLY OF LOW EXPENSIVE CHARACTER ANO/OR LOW FREOUENCY OF OCCURRENCE
NONEXPANSIVE: THE OCCURRENCE OF EXPANSIVE MATERIALS EXTREMELY LIMITED
NOTE: FOUR CATEGORIES OF EXPANSIVENESS ARE SHOWN ON THE DISTRIBUTION MAPS. THESE CATEGORIES ARE GENERALIZED AND QUALITATIVE AND ARE BASED UPON THE PRESENCE OF UONTMOHILLONITE ANO THE RELATIVE FREQUENCT OF OCCURRENCE OF ARGILLACEOUS MATERIAL IN THE AREA. MAJOR CATEGORIZATIONS ARE BASED UPON STRATlGRAPHT AND PHYSIOGRAPHY. DESCRIPTIONS OF THE PREDOMINANT GEOLOGIC FORMATIONS ARE GIVEN IN TABLE I. THE BASIS FOR THE CATEGORIZATION IS EXPLAINED IN PARAGRAPH Z2 OF THE TEXT.
Map compiled ttyD.M. Patrick, H.K. Woods, and Fiedenck L Swim, Engineering Geology and Nock Mechanics Division. U. S. Aiwy Engineer Valeiways E*peiimenl Station, Vicksbmg. Ms.
Distribution States: FHWA
of potentially expansive materials in the United
Region 4, (After Smith, 1975).
HIGH: HIGHLY EXPANSIVE AND/OR HIGH FREQUENCY OF OCCURRENCE
MEOIUM: MOOERATELY EXPANSIVE ANO/OR MODERATE FREQUENCY OF OCCURRENCE
LOW: GENERALLY OF LOW EXPANSIVE CHARACTER AND/OR LOW FREQUENCY OF OCCURRENCE
NONEXPANSlVÂ£: THE OCCURRENCE OF EXPANSIVE MATERIALS EXTREMELY LIMITED
Distribution of potentially expansive materials in the United
States: FHWA Region 6, (After Smith, 1975).
UCO'UU MOOF.RA If L V EXPANSIVE ANO/OR MODEftATF FREQuEnC* of OCCURRENCE
LOW. GENERAULf 0*' LOw E*pan$ivC CHARACTER ANtrOR LOW FRC 3UENCV OF OCCURRENCE
nONExpanSivE fur OCCURRENCE OF
Expansive ma ji-riai.s EatrcmEl^ Limited
FOUR CATEGORIES OF empanSivCnESS arc shown On The DiSTRieunON maps T^ESE categories ARE GEnERaliZEO ano Ouau'TativE anO are QASEOUPON TmF PRESENCE OF mOnTMORilLOniTE ANO THE RELATwE FREOuENCr OF OCCURRENCE OF ARGILLACEOUS MA TÂ£Rial in ThÂ£ AREA. MAJOR CATEGORIZATIONS ARE 3ASE0 UPON STRA TiGRAPHV ano PHYSiOGRAPHr. DESCRIPTIONS OF Tug PREOOm-INANT GEOLOGIC FORMA riONS ARE GIVEN IN TABLE I THE BASIS FOR T-iF CATEGORIZATION IS EXPLAINED IN PARAGRAPH 2t OF TmF TEXT.
Map compiled by 0. H. Patnck. H. K. woods, and Fiedenck L. Smiih. Engineenng Geology and Rock Mechanics Division. U. S. Aimy Engmeei Walerways Experiment Slahon. Vicksbutg. Ms.
Distribution of potentially expansive materials Shahes: FHWA Recrions 7 & 8. I After Smith. 1975).
in the United
Figure 2.6 Distribution of potentially expansive
materials in the United States: FHWA Regions 9 & 10, (After Smith, 1975) .
California and Texas collectively account for about 35 percent of the nation's expansive soil damage to buildings. It was estimated in 1978 that half of the states annually suffer building damage in excess of 20 million dollars from the hazard. Chen (personal comm., 1993) estimates that by the year 2000, building damage will approach 4.5 billion dollars.
The distribution of the expansive material shown in Figures 2.2 through 2.6 was categorized on two bases:
(1) degree of expansiveness and (2) expected frequency of occurrence of expansive materials. The information reviewed was correlated to provide four mapping categories that reflect the degree of expansiveness and expected frequency. The four categories are:
1. High highly expansive and /or high frequency of occurrence.
2. Medium moderately expansive and/or moderate frequency of occurrence.
3. Low generally of low expansive character and/or of low frequency of occurrence.
4. Non-Expansive These areas are primarily underlain by materials which, due to their physical
properties, do not exhibit expansive properties.
As can be seen from these figures, the highest degree of frequency of both, moderate and high potential swelling soils exist in areas west of the Mississippi River, in the great Plains, and the mountainous states.
2.3.2 Local Distribution of Expansive Soil Along The Colorado Front Range
Of the approximately 15 sedimentary bedrock formations in Colorado that contain swelling clay, 4 underlie most of the Colorado Front Range Urban Corridor Figures 2.7 and 2.8.
Geologic hazard maps of the Denver area generally indicate the same degree of concern for vertical beds as the flat lying claystones to the east of the Front Range, which underlie the majority of metropolitan Denver and surrounding areas. The near vertical sedimentary beds of the Pierre Shale however, present a much more difficult foundation situation than the flatter-lying, relatively undeformed beds. The potential for differential swelling is greater in the near vertical dipping strata, whereas
Figure 2.7 Area location map showing the Colorado Front Range Urban Corridor, (After Hart, 1974).
Dooeon Arlote, Dtnvtr-Arapohoe For motion,end Coil It RtCk Conglomtrolt undiffortnllaltd.
Otology modified after U.S.Ctolagicol Sorvtv, 1933,Otologic mop of Colorado. Sea lt:300,000
O 10 2 0 Mile*
0 IO 20 30 Kilomtlere
Figure 2.8 Generalized geologic map of the Colorado Front Range Urban Corridor, (after Hart, 1974) .
in the flatter lying beds to the east, the chance for more uniform wetting is greater and hence more uniform swelling occurs.
Geologic characterization and general distribution of these four formations have been described by the Colorado Geological Survey, 1974. Detailed lithologic descriptions of these and other bedrock units are shown in the stratigraphic section presented in Figure 2.9.
CWCLOaitft AXD CUfttr H^U* wf It } (ut:
ltt of tat ta urtot allcatoaa: <*nl7 (tnt tictrtiMtii Wat tettltfW near tm of Cattle Wet; (MartUf Im avail potential. COaCLCTOUTli IftlMt aod atltfDlpMc Wuldefa
UJtDtTWl tm SILTTTOKX: (H to light bfM faldapethlc. flar* to eedlvo-gteloed aaadetoaa, alUtiMt. u4 elayacoeo; eniltMtita at boat: Table touateto *Waalt" (low io upper part; vail erpoeed oa
SaUDSTOm Am CUTSTOMf i eaadetonee floe-grained, (aldapatUe to euertaoee, vKlta to light to. compact: cldpatoaaa dart gray. cerboeaceove; this Ug<lc coal bada U lovar part; aapoaad la pita vaat of School af Jttaea caapuei iov avail potao-tlal la aaodatooa; aodarata to vary Mgh avail potaatlal la alayatoaa.
UUKTOn: light tan. vary f loe-grelead. arto*U. triable; upper pare gredoe laterally lato alley ahalaa; bounderlaa gradatlaaal eed inter!legating, aquifer; lav avail potaatlal (a aaodatone! oodar-aca to high avail potaatlal la allty ahala.
SHAHi dart gray, allty. cadbooaceoue, clay atone. ahala* and alltitna; (racornt thin layara of bentonitei (IntegraLoad aaAdataaaa to alddta part; lover boundary gradational to uAContoraable; veil aapoaad aloog 1-13 from Colorado Springe to Puebla; generally aodarata to vary high avail potaatlal; lav avail potaatlal la aandataoa aad oan alltatoea.
ouu am nui.li dark gray, vaathara light too. highly ealcaraova; lover boundary gradational; vail aapoaad along Arhaaaaa Klvar bat*ero Plaraaca and Pueblo; gaoarally lev avail pot ant loll aodarata avail potaatlal la aaaa uppar analee.
lUCSrOMi light gray thla-baddad. atgllle-caauat eoaeroue l-loch layara o( caleoraoua ahala; rainly fonu a loot ridge aaat of Oahota hogback; vail aspaaed along Arhaneae llf aaor noraaca; lov avail aetcatUl*
UPMTOR: light gray, flao- to aadtingrained. allghtly ealearaawa. lov avail patrntWl. SHALXi dark gray. aanatUMa allty; oecaalaaal caleoraoua concretion# and thin layara af baatonlta; vail aapoaad In Canon City araa; genet-ally aodarata to vary high avail potential. UKISTOflC AJ*D SNALC: lloaatoaa light gray. argiUaeaoua. danaa. anal# dare gray; vail da*alepad tantan lea Wda near baaa; lover boundary gaoarally Mrpar chan upper; vail aapoaad to Canon City araa; lov awll potentUl la liMetona; aodarata to high avail potaatlal la ahala; vary high avail potaatlal la baatonlta.
SHAH: dark gray* thin-bedded, oecaeioaal cona-la-coaa coacrationa la lover poet; baotoolta bad# (up to O throughout; occasional chta eeadataoea la lover part: oil aapoaad lo Caaoe City mod Parry Park araa#; aodarata to vary high avail potaatlal*
Figure 2.9 Stratigraphic section of bedrock units that occur within the Colorado Front Range Urban Corridor, (after Hart, 1974).
Pierre Shale. The Pierre Shale contains montmorillonitic shale and numerous white or yellow "bentonite" beds that range in thickness from 1/4 inch to 6 inches. This formation underlies the area from central Colorado Springs south through eastern Pueblo and from Canon City to Florence and Wetmore. North of Denver it extends from Roxborough Park to just west of Green Mountain, from Golden to Boulder west of Colorado 93, and from Boulder northeast to Longmont, Loveland, Fort Collins, and Windsor.
Swell potential may range from low to very high, yet the swell potential within specific parts of the formation is generally predictable.
Laramie Formation. A less predictable bedrock unit that underlies a large part of the Urban Corridor is the Laramie Formation. This formation is composed of thick, white to yellowish-gray sandstone beds alternating with greenish-gray claystone beds. Some of these claystone beds are montmorillonitic, particularly in the middle 1/3 of the unit.
Dawson and Denver-Arapahoe Formations. The Dawson Arkose and Denver-Arapahoe Formation underlie most of the area from northern Colorado Springs on the south to Golden, Broomfield, and Brighton on the north. These formations consist of extremely variable beds of sandstone, conglomerate, siltstone, and claystone. In Colorado Springs the Dawson Arkose includes a potentially highly swelling zone that trends south-easterly from the Air Force Academy to Peterson Field, along the northeast side of Austin Bluffs. Other significant areas underlain by swelling clays in the Dawson include the Parker and Cherry Creek Reservoir areas. The Denver-Arapahoe Formation is the principle bedrock unit underlying metropolitan Denver. Some areas of this unit contain very high swelling clays that have caused millions of dollars in damage. No part of the Denver metropolitan area is completely free of potentially swelling soils. Other sedimentary bedrock units in the Urban Corridor, such as the Graneros, Carlile, and Smokey Hill Shales, Greenhorn Limestones, and Fox Hills Sandstone, normally
contain some swelling clay. Since these units generally underlie only small parts of the Urban Corridor, no specific areas of potential hazard have been described.
2.4 Mineraloqical Properties Conducive to Swelling
Expansive, argillaceous rocks, sediments, and soils generally owe their expansive character to their constituent clay mineral suite, past and present loading history, and to their natural and imposed physical environments. As reported by Porter and Liu, 1965, the mineralogical composition of a soil is paramount in its ability to exhibit swelling characteristics.
2.4.1 Clay Mineralogy
Clay minerals are composed of a group of hydrous alumino-silicate minerals belonging to the phyllosilicate and double chain, inosilicate groups. The minerals in these groups are characterized by small grain size (arbitrarily defined as being 2 microns or less in diameter), large surface area, and unbalanced electrical charges. Structurally, the phyllosilicates are mainly
platelike whereas the inosilicates are tubular in shape. The clay minerals that are of most concern with respect to volume change are in the phyllosilicate group. The extent to which water is imbibed is a function of the structural configuration, clay mineral size and to a lesser degree water chemistry, (Grim, 1968; Gillot, 1968; Weaver, 1973; Grimshaw, 1971; Snethen, 1975).
The molecular structure of the phyllosilicates consist of three general configurations which are distinguished by the arrangements of the aluminum octahedral and silica tetrahedral layers, Figure 2.10 and Figure 2.11. The aluminum octahedral layer (termed the gibbsite layer) consists of aluminum and or magnesium ions in sixfold coordination with hydroxyl of oxygen.
The silica tetrahedral layer has silicon ions in fourfold coordination with oxygen. These three configurations are further subdivided on the basis of ionic substitutions within both structural layers, e.g., aluminum for silicon, and iron or magnesium for aluminum. Clay minerals which have aluminum or other trivalent ions in
Figure 2.10 Silica Tetrahedron Sheet (after Grim, 1953).
Figure 2.11 Alumina Octahedral Sheet (after Grim, 1953)
the octahedral layers are termed dioctahedral, whereas those which contain magnesium or divalent ions are termed trioctadedral, (Snethen, 1975).
2.4.2 Clav Mineral Classification
Snethen, 1975 reports that the clay minerals have been classified in the following fashion:
i) "Two-layer clays consist of one silica tetrahedral layer bonded to one aluminum octahedral layer. Kaolinite is the common mineral in which the octahedral layer contains mainly aluminum; serpentine consists of a magnesium-rich octahedral layer."
ii) "Three layer clays have one octahedral layer bonded between two tetrahedral layers; examples of this type are illite, vermiculite, and montmorillonite. The term montmorilionite, as used here, indicates the dioctahedral magnesium bearing member of the smectite group. These minerals may occur as di- or trioctohedral."
iii) "Mixed-layer clays consist of
interstratifications of the two- and three-layer clay minerals previously described. The mixing may be regular ar random. Examples of regular mixing include chlorite, a three-layer plus octahedral layer repetition. Another common regular mixed-layer clay is montmorillonite-chlorite. The randomly mixed-layer clays consist of any of many possible combinations."
The molecular structural configurations of these three
classes of clay minerals are shown in Figure 2.12, Figure 2.13 and Figure 2.14.
The small grain size and unusually large surface area are due to the clay mineral's origin by weathering or diagenetic alteration from preexisting minerals. In these processes, alteration begins at very small centers or points on the mineral grain surfaces and eventually spreads throughout the grain. Due to the lack of physical continuity, the size of the clay mineral is inherited from the size of the initial weathering center.
Figure 2.12 Structure of the kaolinite clay particle
(after Grim, 1953).
Figure 2.13 Structure of the illite clay particle
(after Grim, 1953).
O Oxygens Hydroxyls Aluminums & silicons not shown
Figure 2.14 Structure of the montmorilIonite clay
particle (after Grim, 1953).
Clay minerals that exhibit high volume change do so because of electrical charge characteristics, degree of crystallinity (purity), and particle size. Clay minerals possessing internally unbalanced electrical charges due to crystal lattice substitutions maintain electrical balance by cations located on either the surface or the edges of the minerals. Since these cations are easily hydrated, water easily affects the development of the double-layer. Generally, the effects of crystallinity and particle size are such that volume change is increased when clay mineral size is decreased, (Snethen, 1975).
Montmorillonite. The clay mineral montmorillonite, although dioctahedral, usually contains some magnesium substituted for aluminum in the octahedral layer. This substitution results in a lattice charge deficiency which is neutralized by the presence of cations such as Na+, Ca++, or Mg++ on inter layer positions. If these ions are hydrated, the ionic radii increases resulting in an increase in space greater than the un-hydrated tetrahedral layers. Such a position tends to push adjacent
layers apart and permits access of more water to interlayer positions. Since the interlayer ions balance the molecular charge in the octahedral layer, the ions are weakly held and thus they can be removed by ion exchange. MontmorilIonite ordinarily exists as extreamly small particles having dimension on the order of a few tens of Angstrom units.
Vermiculite. Vermiculite clays are the product of fine-grained weathering and diagenetic alteration products that have formed from preexisting mica, illite, chlorite and vermiculite. Fine-grained chlorites and vermiculites possess properties similar to those of montmorillonite, particularly with respect to swelling and cation exchange. The vermiculite and chlorite minerals often occur as mixed-layer interstratifications with montmorillonite or other clay minerals. The clay vermiculites are a three-layer clay minerals exhibiting a wide variety of physical properties and variable chemical constituencies. Charge deficiencies or excesses may exist in both
tetrahedral and octahedral layers. The net charge is negative and usually balanced by interlayer Mg++, Na+, or K+ ions. Aluminum may substitute for silicon in some vermiculite, whereas other vermiculites contain no tetrahedral aluminum. The former varieties are more similar to the coarsegrained vermiculite, whereas the latter resemble montmorillonite. The interlayer cations are hydrated and control the extent to which the mineral expands (Snethen, 1975).
Chlorite. Fine-grained chlorite is considered a regular mixed-layer interstratification of a di- or trioctahedral three-layer clay and one octahedral-type layer containing magnesium. Apparently, the amount of swelling exhibited by this material is dependent upon the continuity of this octahedral-type layer. As with vermiculite, the swelling chlorites often occur in mixed-layer associations with other clay minerals (Snethen, 1975).
Mixed-layer types. Regular and random mixed-layer
combinations of montmorillonite, chlorite, and vermiculite with other clays may be of importance in contributing to expansiveness. Generally, the amount of expansion would be in proportion to the amount of montmorillonite or other expansive clay minerals present in the mixed layer association (Snethen, 1975).
Kaolinite. The clay mineral kaolinite exhibits very minor interlayer swelling. This is explained by the virtual absence of ionic substitution in either the tetra- or octahedral layers which results in more or less complete electrical neutrality and the absence of compensating cations. Also, the individual two-layer structures are more tightly bonded together by the opposing electrical charges on the adjacent octa- and tetrahedral layers. Therefore, the volume change exhibited by this mineral is mainly due to water sorbed on the periphery of individual grains (Snethen, 1975).
Illite. This three-layer clay mineral also exhibits
very minor interlayer swelling. This results from the presence of nonhydrated K+ ions in interlayer positions within the hexagonal openings of the tetrahedral layer. The K+ satisfies charge deficiencies residing mainly on the tetrahedral layer and is thus tightly bonded. These characteristics effectively preclude the admission of significant amounts of water between the unit layer (Snethen).
2.4.3 Clay Mineral and Water Interaction
The electrical charges exhibited by clay mineral grains are caused by the following: i) charge deficiencies due to the ionic substitution within the lattice, ii) broken bonds at grain edges, iii) imperfections within the lattice, and iv) the polar nature of ions exposed at clay surfaces, (Grim, 1968; Van Olphen, 1963). Lattice imperfections and broken bonds may produce either a positive or negative charge, whereas ionic substitution usually results in a negative charge.
The magnitude and location of these electrical charges are different for the various clay minerals and
are fundamental in explaining the ability of some minerals to imbibe significantly more water than others. Water associated with the clay minerals consists of three types (Snethen, 1975) :
a. Hvdroxvl or bound water. This water forms a part of the octahedral layer and cannot be removed by heating at temperatures below 400<>c for most clay minerals.
b. Interlayer water. This double-layer water which occurs between clay mineral surfaces in some clays. It is gradually removed by heating up to 150-200oC.
c. Pore water. This water occurs in the open spaces between grains and also constitutes the more spaces between grains and also constitutes the more tightly bound double-layer water on grain surfaces. This water is essentially removed by drying at room temperatures and completely removed by heating at approximately lOOoC.
The clay minerals which exhibit appreciable expansion or shrinkage are called expansive clay minerals and include montmorilIonite, vermiculite, chlorite, and
mixed-layer combinations of these minerals with each other or with other clay minerals. Halloysite, the tubular, hydrous member of the kaolinite group may also exhibit expansive properties. Kaolinite and illite generally do not exhibit volume change to the extent of montmorilIonite, vermiculite, or chlorite and are called nonswelling clay minerals. Table 2.1 lists some representative free swell data for the common clay minerals.
It has been observed that the double layer water adsorbed between clay layers in expandable clays and the water adsorbed on the surfaces of other clays possess properties which are somewhat different from those of the water in pore spaces. The double-layer water exhibits a certain degree of crystallinity which is not a property of the pore water. The crystallinity is greatest adjacent to the clay mineral itself and decreases outward from the mineral surface. The volume of the oriented water and whether the decrease in crystallinity is gradual or abrupt appears to be dependent upon the nature of the clay mineral and the type of cation present. MontmorilIonite exhibits greater amounts of oriented
Free Swell %
Mixed layer type
Test data based on swell in water of 10 cc of dried, crushed material passing No. 30 sieve and retained on the No. 50 sieve.
Free swell is variable and dependent on size and crystallinity.
Free swell is variable and dependent on amount of expandable clay minerals present.
Table 2.1 Typical values of free swell for common clay minerals (after Snethen, 1975; Grim, 1962).
water than the other clay minerals. Those cations which enhance the orientation are those whose hydrated or nonhydrated size can be accommodated within the water structure, for example, sodium and lithium are compatible, whereas calcium and magnesium are not, (Grim, 1968; Van Olphen, 1963; Snethen, 1975).
2.5 Physical Properties Affecting Swelling
Physical properties of expansive soils which determine the behavioral characteristics of the material have been enumerated and defined in a multitude of publications. In many cases, attempts have been made to isolate the individual properties and explain the behavior on the basis of a single property or a combination of single property contributions (Snethen, 1975). However, in both the laboratory and field situations, the actual behavior appears to be a function of combinations and interrelationships among the properties due to the unpredictability of test results. The following discussions are presented in a twofold categorization of the physical properties in order to point out and explain some of the interrelationships.
The categories are: 1) Intrinsic properties of the materials which contribute to or influence the actual volume change and 2) Environmental factors that effect the reaction of the physical properties (Snethen, 1975).
2.5.1 Intrinsic Properties
Intrinsic properties which influence the behavior of expansive materials are presented in the following paragraphs. Confinement, time, and temperature are not intrinsic properties as defined in the previous section; however, they are factors which influence the role of the intrinsic properties in determining the amount and rate of volume change in both laboratory and insitu conditions and will be discussed in this general section.
220.127.116.11 Soil Composition
This includes the type and amount of clay mineral within the soil and the size and specific surface area of the clay minerals. The type and amount of clay mineral are intrinsic factors which determine whether or not the material will expand. The potential for volume change rests on the mineralogic composition; and the remaining intrinsic factors, combined with the ambient environmental conditions. This combination determines the magnitude of volume change (Snethen, 1975).
"The size of the clay mineral particles in expansive materials affects the volume change by controlling the development of double-layer water on the particle periphery. Generally, small particle sizes result in
large effective surface areas which permit considerable thicknesses of double-layer water to surround the individual particles. This is particularly important for clay minerals which do not exhibit interlayer swelling since the expansivity of the materials is almost entirely due to sorption of peripheral water" (Snethen, 1975).
Clay mineral size is not an independent parameter, but often is a characteristic of the specific clay mineral. For example, montmorillonite can occur as extremely small particles which may be considered colloid. Particle sizes on the order of a few unit cells may be present if the particles are dispersed or in the deflocculated condition. Kaolinite on the other hand may occur as rather large particles which may be on the order of fine silt size. Chlorite, vermiculite, illite, and mixed-layer clays are generally intermediate in size between montmorillonite and kaolinite. In summary, the relation between clay mineral size and specific surface area is inversely proportional such that surface area increases with decreasing mineral particle size from kaolinite to montmorillonite, (Parcher, 1965; Ladd, 1959; Chen, 1973; Blomquist, 1961)
18.104.22.168 Dry Density
The dry density is an important factor in
determining the magnitude of volume change. The swell or swelling pressure of an expansive soil increases with increasing dry density for constant moisture content.
The reason for this is that higher densities result in closer particle spacing, therefore causing greater particle interaction. This particle interaction, or more precisely, double layer water interaction, results in higher osmotic repulsive forces and a greater volume change. This holds true for both remolded and undisturbed materials. Another important and somewhat indirect influence of dry density on volume change is its interrelationships with some of the other intrinsic factors. For example, the dry density of a material, particularity compacted soils, will influence the soil fabric (interparticle arrangement). Details of this influence have been described by (Pacey, 1956, and Seed et al, 1959) and are depicted in Figure 2.14 and Figure
.15 Molding water content versus dry density and particle orientation for Boston Blue clay, (after Pacey, 1956).
Figure 2.16 Molding water content versus dry density and particle orientation for kaolinite (after Seed and Chan, 1959).
As indicated in Figures 2.14 and Figure 2.15, for a given compaction effort and at low initial moisture contents, a less oriented fabric is obtained. As the moisture content increases, the soil fabric is more oriented, (Holtz, 1959; Russel, Worsham, and Andrews, 1946;
22.214.171.124 Soil Fabric
Soil fabric refers to the orientation or arrangement in space of the constituent particles. In the case of argillaceous sediments and rocks, the fabric consists of the arrangement of the plate-like clay minerals with each other and with the non-clay components. The type of clay mineral arrangement present will influence the amount and to some degree the direction (lateral or vertical) of volume change exhibited by an expansive material, (Van Olphen, 1963).
The fabrics exhibited by argillaceous sediments and rocks are complex, variable, difficult to observe and have not been specifically categorized in an acceptable manner. Individual clay mineral platelets generally occur in either agglomerated or non-agglomerated
arrangements. Agglomerated arrangements consist of independent groups of platelets which may be arranged in a chaotic manner. Nonagglomerated arrangements exhibit no association with discernable groups and the fabric appears to be uniform throughout (Snethen, 1975).
The individual clay mineral platelet within either of these two arrangements may exist as individual units of the smallest size (dispersed) or as small booklets of individual units with face-to face contacts of the individual units (aggregate). If the dispersed or aggregated units exist with no points of contact with other units and are surrounded by double-layer water, the association is denoted as deflocculated. A flocculated association is one in which the dispersed or aggregated units are in contact with adjacent units (Snethen, 1975).
126.96.36.199 Pore Water Properties
The phenomenon of volume change in expansive soils is the direct result of the availability and variation in the quantity of water in the soil, (Gupta, 1967; Alpan, 1970; Mitchell, 1973). Therefore, the water properties will have an influence on the expansive behavior. The
volume change of expansive soils is primarily due to the hydration of the clay minerals or, more precisely, the adsorption of water molecules to the exterior and interior surfaces of the clay mineral. This happens due to the tendency of the particle to balance the inherent charge deficiency. The degree of hydration is influenced by the amount and type of ions adsorbed on the particle and the amount and type of ions in the pore fluids. Pore fluids containing high concentrations of cations, i.e., soluble salts, tend to reduce the magnitude of volume change of an expansive soil. Conversely, pore fluids with low ionic concentrations may actually leach out the charge balancing cations and cementing agents and render the soil more susceptible to volume change as observed by Baver, and Winterkorn, 1935.
Surcharge or external load to an expansive material tends to reduce the amount of volume change that is likely to occur. In the laboratory measurement of swelling pressure, less than 1 percent deformation of the testing device may result in large errors in magnitude of
the swelling pressure. For in situ conditions, the presence of a layer of nonexpansive overburden material may eliminate the probability of damage from the underlying expansive material. It may be noted that confinement has its greatest influences on expansive soils in a stress related sense (swelling pressure). The greater the confinement, the greater the stress and the smaller the deformation, (Chen, 1988; Colby, 1989). Generally, the load applied by pavement is far less than that required to maintain minimal deformation; therefore, problems with expansive clays in highway subgrades are more related to deformation (Snethen, 1975, Donaldson, 1969) .
The effects of time on volume change is another interrelated property which has its major impact on the rate at which expansion occurs. The time to the first occurrence of volume change and the rate of expansion are usually a function of the permeability of the soil and the amount and source of water. Expansion occurs as soon as moisture is made available and continues until an
equilibrium condition is reached with regard to the source of water or the rate of hydration of the clay minerals (Snethen, 1975; Matyas and Radhakrishna, 1968).
Permeability is a function of the initial moisture content, dry density, and soil fabric. For compacted soils, the permeability is greater at the lower moisture contents and dry densities and decreases to some relatively constant value at about the optimum moisture content. Above optimum, the permeability is essentially constant. The obvious reason for this minimum permeability near the optimum moisture content and maximum dry density is that the voids available for moisture movement are at a minimum because of the close particle spacing. Above optimum moisture, the interaction of the double-layer water also minimizes the voids necessary for moisture movement. Insitu expansive soils, exhibiting structural voids such as fissures, fractures, and desiccation cracks will obviously have higher a permeability than normal (Jennings, 1969,
Snethen, 1975, reports that the influence of temperature is primarily limited to its effect on the viscosity and specific gravity of the adsorbed water. Increases in temperature tend to depress the double-layer water, while temperature decreases result in double-layer expansion. Of more importance is the influence of temperature on the movement of moisture, both vapor and liquid, as a result of thermal gradients within the soil mass.
"Water vapor at higher temperature will migrate toward a cooler area in an effort to equalize the thermal energy in the system. Liquid moisture movement by thermal gradients occurs as a thermo-osmotic film analogous to electro-osmotic flow." (Krazynski, 1973).
The structure of argillaceous sediments and rocks includes those features or discontinuities which contribute to the nonhomogeneity of the material. Of most concern with respect to volume change are fracture zones, fissures, cracks, and micro- and macrofaults. The structural discontinuities may exhibit variable orientations in space and originate as a result of stress
conditions which have developed in the natural sediments or rock mass. The conditions which contribute to fracturing and faulting include desiccation, stress release during unloading, and possibly tectonic loading. The structures, if not cemented, provide avenues for moisture to be introduced into the expansive soil. This occurrence is generally restricted to the upper layers within a few feet of the surface (Snethen, 1975).
Cementation refers to the bonding of mineral cements which coat particular constituents together in sedimentary rocks. The presence or absence of these cements may determine whether a particular material should be classified as a rock, sediment, or a soil. It seems logical that materials exhibiting a high degree of cementation would possess less expansive properties than materials lacking cements. As described by Snethen, 1975, the presence of cement produces two effects: the development of bonds between points of contact which decrease the likelihood of the displacement of adjacent particles and the coating of individual particles which
reduces the ability of the clay minerals to imbibe water.
The common cementing agent may either be crystalline (conforming to a particular structure) or amorphous (shapeless). Some cementing agents are: calcium carbonate, CaC03 (calcite), iron oxides or hydroxides, Feo or FeOH (hematite or goethite), and various forms of silica, Si02. The degree of resistance to weathering and strength decreases in the order of silica, iron, and carbonate. The carbonates, however, probably comprise the most common cement in sedimentary rocks. Siliceous cements are commonly associated with bentonite and other rocks which contain montmorillonite derived from the devitrification of volcanic ash. In these cases the devitrification of the ash produces silica in excess of that necessary to produce montmorillonite. The excess silica may be removed from the zone of alteration and redeposited elsewhere in the system by groundwater. The redeposited silica produces indurated zones in the sedimentary sequence.
Some argillaceous sedimentary rocks are indurated to such a degree that would be indicative of cementation, but do not contain appreciable mineral cements. These
have been referred to as compaction shales (as apposed to cementation shales). The induration is apparently derived from bonds which have developed at contact points between individual clay mineral particles. These bonds are probably time-related and have developed by diagenesis (diagenetic bonds). This is a characteristic of older rocks and occurs during and partly due to the recrystallization of the clay minerals, i.e., montmorilIonite altering to illite. These bonds, as well as the concomitant changes, tend to decrease the possibility of volume change in the material (Snethen, 1975; Mielenz and King, 1955; Obermeier, 1973).
188.8.131.52 Atterbera Limits
It is well documented that volume change behavior correlates reasonably well with liquid limit, plasticity index and shrinkage limit. These indices tend to indicate the amount of clay particles, which then has an effect on the amount of swell. The liquid limit and plastic index correlate reasonably well with swell potential, primarily because of the relationship between those indices and the type and amount of clay minerals
present. In general, the higher the liquid limit, plasticity index and shrinkage limits, the greater the potential volume change (Livneh, Kassif, and Wiseman, 1969; Nayak and Christensen, 1974; Ring, 1965; Chen 1988) .
184.108.40.206 Soil Suction
Soil suction can be described as a soil property which indicates the tendency of a soil to attract water. The concept of soil suction is very complex, and only a very brief discussion is presented. More complete descriptions of soil suction can be found in the references.
Moisture migration will occur in soils due to variations in soil suction. Moisture typically migrates from areas of low suction to areas of high suction. Moisture is redistributed until a new state of suction equilibrium is obtained. This rate may be very fast or very slow, depending upon the individual soil properties.
It is possible that two different soils in contact may be in suction equilibrium even though their moisture contents may differ. It is also possible for two
portions of the same soil to be in suction equilibrium but have different moisture contents. This occurs when one soil portion is becoming wetter while an adjacent portion is drying.
2.6 Environmental Factors Affecting Swelling
The environmental conditions which influence volume change in expansive type soils are presented in the following paragraphs.
2.6.1 Soil Profile
The properties of the soil profile which enhance or influence volume change include the total soil layer thickness, variations in the thickness, depth below ground surface, and the presence of lenses and layers of more permeable materials. Obviously, the thicker the layer of expansive soil, the greater the total potential volume change, providing a source of moisture is available. Variations in thickness of the layer will result in variations of the magnitudes of volume change, or more precisely, differential volume change.
Differential expansion, just like differential
settlement, is the major problem with regard to structural damage. The depth of the layer below ground surface may actually be a positive influence since the deeper the material, the greater the confinement on the expansive soil. In addition, the deeper the material, the less likely the expansive soil will be affected by seasonal moisture variations. Lenses or layers of higher permeability will provide avenues for the ingression of water. In fact, a mass of soil which requires that moisture must move from its extremities will take much longer to develop its total volume because as the moisture is introduced and expansion occurs, the avenues of moisture transfer are somewhat reduced (Snethen,
Expansion or shrinkage of a soil is directly related to an increase or decrease in moisture content respectively. Climatic conditions, primarily rainfall and evaporation, have a significant influence on the magnitude and time rate of heaving.
Expansive soils are most prevalent in semi-arid
regions where evaporation exceeds rainfall. In these climates, the soil is generally dry and increases in moisture content are accompanied by an increase in volume. The amount of moisture variation within the soil is affected by the length of the evaporation and the maximum rainfall duration. Where there is a definite wet and dry season, there is a large seasonal variation in the moisture content. In this type of climate, there are definite seasonal variations in moisture content of the soil. Seasonal moisture variations have been reported to depths of 10 to 12 feet; however, in semi-arid climatic conditions, the depth is normally between 5 to 7 feet (Snethen, 1975). Seasonal moisture variations are relatively constant for given climatic conditions; however, there is a general trend towards an accumulation of total moisture content. This is particularly the case when the majority of the rainfall occurs in the summertime when there is also a high evaporation rate.
This decreases the amount of seasonal variations, but tends to increase the expansion with each season until an equilibrium condition develops (Huzjak, 1988; Fenner, Hamberg and Nelson, 1983).
The effect of vegetation, both prior to and after construction, is a very important factor influencing moisture transfer. Vegetation such as trees, shrubs, and grasses affect the pre-construction moisture content of the soils. Tree roots typically tend to dry out the subsurface soils to depths of approximately 15 feet because of the moisture which that have pulled out of the soil. When the vegetation is removed and pavements or structures are places, the moisture that was being used by the vegetation will tend to accumulate beneath the pavement structure. This accumulation of moisture tends to enhance volume change and, in particular, differential heave due to the equalization of moisture content upon the density of the root structures in various locations, (Huzjak, 1988).
2.6.4 Surface Drainage
Poor surface drainage leads to the accumulation or ponding of water which can provide a source of moisture for expansive subgrades by infiltration. The extent of
the infiltration is a function of the transverse and longitudinal gradients of the soil.
STRAIGHT SHAFT SINGLE PIERSHISTORY AND DESIGN
3.1 Historical Development: Of Piers
One of the earliest known uses of foundation piers appears to have been by the Phoenicians who used sheet piles for dock and shore construction in connection with the progress of the sea trade. Sheet piles appear to have been derived from the development of boat builders in the planking of ships.
It was in the nineteenth century that, as in many engineering fields, large changes began to take place both in terms of materials and motive power. Steam power had first been applied to the driving of piles in 1801. Metal piles, mostly in the form of cast-iron pipes, became available in the mid-1830*s. Screw piles were first used for the foundations of the Maplin Sands Lighthouse in the Thames Estuary in 1838. In 1824, Portland cement was introduced and by the middle of the century its use had become increasingly popular. Before the end of the 19th century, reinforced concrete had made a dramatic impact on the construction industry.
a dramatic impact on the construction industry.
Bored piles or piers were developed most likely as the result of 'well foundations' which were used in many countries for the support of major structures. This type of foundation was built with stones being placed from the bottom up of each boring or excavation. The Taj Mahal, which was built in the period 1632 to 1650, made extensive use of this type of foundation.
In the early days bored piles offered certain advantages over driven types in that the machinery and the steel tubing, which were necessary to keep the boring open, could be removed after concrete placement was complete. Bored piles were probably less reliable than the methods used to install driven piles but as knowledge and experience grew, the reliability improved.
Today, the majority of bored piers are formed using rotary augering equipment. This has resulted from rapid development since about 1950. Machines are now available to bore a wide range of pile diameters from about (150mm) 6 inches to over (2m) 6 feet. Equipment has also been designed that enables concrete or grout to be injected to form a pile through the hollow stem of a continuously
flighted boring auger as the augers are withdrawn from the soil.
The design of piers and pier groups has advanced steadily in modern times with much of this work being carried out by engineers who specialize in foundation engineering. Effective stress methods are being developed for individual pile designs, while computer-based techniques for settlement determinations of large and complex pile groups are being used with increasing popularity.
3.2 Types of Drilled Piers
As earlier stated, the principle of the use of drilled piers is to provide a relatively inexpensive way of transferring the structural loads down to a stable material or stable zone where variations in the moisture are non-existent. Drilled piers can be straight-sided for the entire depth, or they may be constructed with a belled or 1 underreamed' base if placed in cohesive soil, in order to increase the end bearing pressure. Bored foundations have been recognized under a variety of names in the past, such as drilled piers, piles, or caissons,
and cast-in-place or cast-in-situ piles. When the base is enlarged or belled, the foundation may be referred to as a drilled-and-belled type. The designation currently receiving general acceptance is drilled-shaft foundation (McCarthy, 1988). The subject of this thesis deals with straight shafted single piers and the term pier will be used consistently in reference to this type of foundation.
3.3 Drilled Pier Construction
Bored or drilled piers refer to the type of deep foundations that are constructed by drilling into the earth and placing concrete in the boring, usually directly against the soil. The concrete is commonly reinforced to resist potential tensile forces. Drilled-shaft foundations are constructed by utilizing mechanical auger drill equipment to excavate the boring into the earth. The boring can be drilled dry if sufficiently high strength, cohesive soil are encountered. In this instance, the cast-in-place concrete comes in direct contact with the soil forming the walls of the boring. If cohesionless soils are
present or the water table is encountered, two methods are usually employed to properly install the pier; i) a bentonite slurry may be circulated into the hole as it is being drilled to prevent soil cave-in and to assist in flushing soil cuttings to the surface, ii) a reusable protective casing may be used in conjunction with the bentonite slurry if a particularly bad groundwater condition is encountered. Once the required depth is reached, concrete is placed at the bottom of the boring by use of a tremie and worked upward in a continuous pour to force displacement of the bentonite slurry. If casing is used, it is pulled as the concrete advances up the boring in such a manner as to prevent the boring walls from caving into the excavation and mixing with the concrete. This can be accomplished by using a high-slump concrete inside the casing and keeping the height of concrete above the bottom of the withdrawing casing at all times. The use of a bentonite slurry however, leaves a slippery film on the soil surrounding the excavation, which in effect, reduces the skin friction between the soil and hardened concrete (McCarthy, 1988; Reese, Owens, and Hoy, 1981).
3.4 Drilled Pier Design Criteria
Design criteria for drilled-shaft foundations in
clay have evolved primarily from i) instrumented studies
performed on full-scale foundations, (Fuller and Hoy,
1970; Newman, Salver, and Baker, 1981; Jubenville and
Hepworth 1981; McCarthy, 1988), and ii) empirical
considerations derived from experience, and from the
behavior of existing structures, (Woodward, Gardner, and
Greer, 1972; 0,Neil, 1987; Chen, 1988; Poulos and Davis,
1990). Woodward, Gardner and Greer, 1972, commented on
the practical design of piers:
"Many piers, particularity where rock bearing is used, have been designed using strictly empirical considerations which are derived from regional experience."
They further state that,
"Where subsurface conditions are well established and are relatively uniform, and the performance of past construction is well documented, the design by experience approach is usually found to be satisfactory."
McCarthy, 1988 has recommended for pier foundations embedded in relatively homogeneous clay, that the ultimate capacity is calculated using the resistance provided from the end bearing and skin friction. In
calculating the foundation depth that effectively provides skin friction, the lower 5-ft section are neglected because of disturbance and loss of strength caused by construction. Skin friction should be neglected in zones where a strong possibility of disturbance and loss of strength occurs.
The basic concept for the design of a single pier to resist upward swelling forces along the pier surface should consist of the following:
1. The pier should have structural strength to resist these upward forces.
2. The uplift resistance to the pier should be provided from the soil below the zone that is not subjected to soil moisture (i.e., below the active zone).
Prakash and Sharma, 1990, suggest that the magnitude of uplift forces, Q^, to be resisted by a pile can be approximated from the following equation as follows:
Qup = P 2 cuala
where p = pile diameter.
cu= the minimum undrained shear
strength of clay at pile point level (i.e., cohesion of the bearing stratum (c=cu=Su=qu/ 2)) La= length of pile in active zone.
In this equation, Qp should be resisted by the length of pile below the active zone. This would require estimation of pullout capacities of a single pile and pile groups, as the case may be.
One of the more favorable methods in analyzing uplifting forces on a concrete pier is by the application of effective stress against the drilled pier shaft during swelling. This method assumes that the coefficient of uplift between the concrete and soil is a function of residual angle of the internal friction of the soil. The swelling pressure is normally obtained in a oedometer in which zero lateral strain is permitted, making the swelling pressure (u) larger than what would occur in the field where lateral strain is not zero.
O'Neill and Poormoayed, 1980, have suggested the following formula, based on effective stress for calculating uplifting force:
U = 27rruz tan0 'ps
where U = total uplift force
r = radius of pier shaft
u = zero horizontal strain swell pressure z'= thickness of active expansive clay stratum
0'p) = effective angle of pier/soil friction (typically O.2
In this analysis, O'Neill assumed the coefficient of uplift is equal to the residual angle of internal friction. The residual angle of internal friction for stiff clay and clay shale is on the order of 5 to 10 degrees. This would yield an uplift coefficient on the order of 9% to 18% of the swelling pressure in the expansive soil.
Chen, 1988 conducted a model test in an attempt to determine the coefficient of uplift on a concrete pier and soil. Two cases were studied. The first case simulated a friction pier and the second case simulated an end bearing pier.
According to Chen, the predictive equation for total uplifting force for a friction pier is written as:
U = 27rrfu(D-d)
where U = total uplift force r = radius of the pier
d = depth of the soil zone unaffected by
D = total length of the pier u = lateral swelling pressure f = coefficient of uplift force
The predictive equation for an end bearing pier free from the surrounding soils is:
U = 7rr2u
where u = vertical swelling pressure
The model test concluded that the uplifting pressure along the surface of the concrete exerted by the expansive soil in the soil-pier system is approximately 15% of the vertical swelling pressure. Comparing this with the swelling pressure of the pile between 9% to 18% as suggested by O'Neill and Poormoayed, both relationships appear to agree.
The depth of soil that contributes to swelling and shrinking at a particular site depends on (1) the thickness of swelling and shrinking clay layer(s), (2) the depth of water table, and (3) the local environmental conditions that will influence the depth of seasonal changes. The depth of seasonal changes in soil moisture is mainly responsible for swelling and shrinking behavior of the clays. This depth, termed the 'active zone1 can also be affected by the existence of a structure. For example, the excavation of soil below a structure and/or
the heat transmitted by the structure to the underlying soil may alter the depth of the active zone. The depth of the active zone is generally evaluated and identified during the soils investigations work and based on local experience.
It is common engineering practice to utilize pile foundations in swelling and shrinking soils so that the foundations develop suitable bearing capacity in stable ground conditions below the active zone. Design considerations for situations in which the active zone varies in thickness consists of either one or a combination of the following two methods.
Preventative Methods These methods consist of eliminating uplift forces along the pile surface by isolating piles from the swelling clays in the active zone. The following methods can be used for such purposes:
1. Coating the pile surface in the active zone with bitumen.
2. Separating the pile from swelling soils in the active zone by the use of floating sleeves that move up and down with the surrounding soil.
Standard practice in the Denver, Colorado area calls for structures constructed on expansive soils to be founded on drilled piers. Minimum depths for piers are on the order of 10 to 15 feet with minimum penetrations of 5 feet into bedrock for lightly loaded residential homes. For larger buildings with greater load requirements, pier depths can be as great as 100 feet.
Two to four inch void spaces are usually provided beneath walls and grade beams. Positive gravity drainage and regional drainage patterns are usually employed and are directed away from the structures.
3.5 Drilled Pier Field Testing
The amount of resistance to penetration which develop between a concrete pile and the soil it penetrates can only be determined by loading tests. Pile load tests are normally executed in one of two methods; axial compression pile load tests or pullout pile load tests. Experimental and final design load tests are conducted on full scale piers in the field in which loads are applied to the pier head. The load is increased
until failure occurs. The ultimate failure load for a pile is defined as the load at which the pier plunges or the settlements occur rapidly under sustained load.
Other failure definitions consider arbitrary settlement limits such as the pier is considered to have failed when the pier head has moved 10 percent of the pile end diameter or the gross settlement of 1.5 in. (38 mm) and net settlement of 0.75 in. (19mm) occurs under two times the design load (Prakash and Sharma, 1991).
More recently, a new concept in drilled shaft load test systems was introduced by Dr. Jorj Osterberg, Emeritus Professor of Civil Engineering at Northwestern University. The device, which is installed at the bottom of a drilled shaft before concrete placement, consists of a bellows with top and bottom plates only slightly less than the diameter of the shaft. It is constructed so that an inner pipe attached to the bottom extends up to the surface through an outer pipe which is pressurized to apply the load. After the concrete has cured, the device is pressurized internally, creating an upward force on the shaft and an equal but opposite downward force in end bearing. As pressure is applied, the inner pipe moves
downward as the downward load applies compression in end bearing. The concrete shaft moves upward as skin friction on the shaft perimeter is mobilized. From a previous load-internal pressure calibration, the total load for a given internal pressure is known. Failure occurs in either skin friction or end bearing. The load-downward deflection curve in end bearing and the load-upward movement curve in skin friction can be plotted from the test data.
3.6 Uplift Force and Interface Reactions
The interaction of the soil and the concrete of the pier is best studied with soil failure criteria. In particular, the shear strength of the soil is important since it is expected that transfer of load from the soil to the pier continues until either the soil or pier fails, allowing for a slip surface of lower shear strength to be initiated.
Sorochan, 1980 determined from field tests conducted in the USSR, that the soil near the base of the pier may not fail because the pier will be moving with the soil.
In actuality, a failure surface may not be the best
explanation of the reactive mechanism. From strain gages mounted on a drilled pier, a stress distribution indicates that all sections of the pier are influenced by each layer of the soil. The base and very top may be slightly in compression and the remaining mid-sections are in tension. The tangential heave or uplift forces on the lateral pier surface increase with depth and generally reach their maximum at about half the pier length. Sorochan, 1980, concluded that the total pier uplift was more than the soil swelled at the pile base in the tests which definitely indicated tension existed in the pile and slip occurred within the soil.
Results of expansive soil-concrete pier interface strength using the ring shear test device conducted by Hoosack, 1984, indicate definite relationships exist between the variables that affect swelling soils in drilled pier applications. These variables include moisture content, clay content, and normal stress. In practical applications for piers in expansive soils, the normal force is the lateral swelling pressure against the pier. Hoosack concluded that friction coefficients and expected vertical swelling forces can be extrapolated
from test data to the extent that expected behavior can be qualitatively analyzed. The trends observed from Hoosack's data are:
1) As normal pressure is increased, the interface shear strength increases, but less than proportionally. This implies lower friction coefficients at higher normal stress.
2) Average friction coefficients at 1 tsf normal stress, range from 0.60 to 0.80, which is much higher than indicated by existing literature.
3) As water content increases, shear strength and friction values decrease.
SOIL AND CONCRETE TESTING PROGRAM
Different phases of the laboratory testing program were divided into separate sections of this report for organizational and presentation purposes even though the actual chronological order of testing and instrumentation did not always follow the order in which it is presented herein.
The soil and concrete used for the model test were characterized by conducting index tests to determine the soil properties and the qualities of the concrete.
The soil that was used for the modeled pier analysis was removed from a housing construction site. The construction site is located in southeastern area of Denver, Colorado. The soil was hauled to the laboratory in a 5 cubic yards (yd3) end dump truck and placed on the concrete ramp which provides access to the laboratory facility. The soil was then temporarily stored in a
steel trash bin
4.2 Soil Characterization
The soil was characterized by a series of index test to help quantify representative properties. Since swell potential and shear strength test are generally more complicated and aspects of each test vary, separate discussions and results of each test are presented in independent sections.
4.2.1 Soil Sample Preparation
The soil that was delivered to the laboratory was temporarily stored in a 3 yd3 trash bin until the necessary equipment was available for processing. A case rubber tired, 1825 uniloader was made available and was used to crush the larger clods of soil until a fairly uniform and consistent distribution of the soil particles was obtained. This was done by running the rubber tired machine, back and forth over small piles of soil that had been placed on the concrete ramp just outside of the laboratory. Once the soil reached an uniform consistency it was placed back into the trash bin temporarily.
4.2.2 Soil Index Properties
Soil index testing procedures performed on the soil followed the general guidelines of those found in the American Standards for Testing Materials (ASTM) soils testing manual.
Several index property tests were performed on the natural/initial soil before swell tests were completed. Once the swell tests of the natural soil were complete, it was realized that due to the very low swell potential, an additive would need to be mixed with the soil to enhance the swelling activity or an alternate source would have to be found. It was determined that it would be logistically less complicated to enhance the soil rather than to arrange for another load of soil to be delivered to the laboratory.
Bentonite was used as the additive due to the i) availability, ii) ease of its use, and iii) the high expansive nature. It was found that the bentonite powder was too light and would easily disperse, therefore granulated bentonite was eventually used. The granulated bentonite was added and mixed with the natural soil as needed for laboratory testing.
The index properties of the initial soil and samples containing 6% bentonite are summarized in Table 4.1.
4.2.3 Soil Swell Potential
The measurement of expansive swell pressure most commonly used is based upon measuring pressures exerted in the vertical direction while restraining the soil laterally. This testing method is most likely due to the assumption that swelling pressure is an isotropic property of the soil, in addition to the fact that most problems associated with swelling soils appear to have been attributed to vertical movements.
Research conducted by Parcher and Liu, 1965, at Oklahoma State University was some of the first research on lateral swelling using natural and remolded clay. In general, the results of the study found that under conditions of free vertical and lateral strain, horizontal swelling was greater than vertical swelling.
In their study, lateral swelling measurements were made by use of a modified triaxial cell.
At about the same time, in Haifa, Israel, (Komornik and Zeitlen, 1965) initiated a study of lateral swelling
Sample Type uses Class. % Passing # 200 Sieve Atterberg Limits ^d(max) ASTM O.M.C.
LL PL PI D698 (pcf) (%)
S-la SM 11* 34.4 16.4 17.9 102.6 19.7
S-lb 35.2 18.5 16.7
S-2a ML-CL 63.5** 33.6 16.3 17.3 102.6 19.7
S-2b 36.8 20.1 14.3
S6B-1 ML-CL 55.0 20.7 34.3 103.0 20.1
S6B-2 50.0 22.1 27.9
Before #200 wash ** After #200 wash
Table 4.1 Comparison between soil index properties of the initial soil and a sample that was prepared by adding 6% bentonite by dry weight.
pressure. They developed a modified odometer ring which could measure small lateral strains using electrical wire strain gages, while utilizing the variable surcharge-swell method for vertical strains.
Kormornik and Zeitlen, 1965 reported that for conditions of essentially zero volume change, horizontal swelling pressures were less than vertical pressures.
For free vertical swell the ratio of lateral swelling pressure to vertical surcharge was approximately 13. At zero volume change the ratio was approximately 0.72 to
0.80. Komornik and Zeitlen, 1965, concluded that lateral swell pressures can exceed applied vertical stress (i.e. overburden stress) except when vertical swell is close to zero for higher density materials. Hence, lateral swelling pressures would be the major principal stress and the ratio exceeds 4 for conditions of near free vertical swell and restrained lateral movement, (Colby, 1990) Although these factors are recognized, a conventional, professionally accepted testing procedure was used to test the swell potential.
The Standard Oedometer Cell was utilized for the swell potential testing program because of the ease in
preparation of the soil sample, the equipment used for loading and measurement, and the general local practice. The Standard Oedometer Cell apparatus consists of a steel loading frame with a mechanical lever-arm system to apply pressure to the specimen by hanging weights at one end of the lever-arm. The mechanical advantage of the particular system used was 11:1. The test specimens were remolded in a stainless steel confining ring with dimensions that measured 2.50 inches in diameter by 0.75 inches in height. The oedometer cell is a fixed ring type. Porous sintered corundum stones were placed directly on the top and bottom of the specimen without using filter paper in order to reduce the overall compressibility of the system. Each sample specimen and ring were placed in a clear lucite cell that is open on the top. The specimen and ring were clamped in place and stainless steel loading cap placed on the top porous stone.
Each specimen was remolded directly in the confining
ring and carefully trimmed flush to the top and bottom of
the ring. Trimmings from the top and bottom were placed
in metal dishes for moisture content determination. The specimen and ring were weighed and then carefully placed in the oedometer cell on a cool, oven dried, bottom porous stone. The porous stones were oven dried to avoid having the specimen adsorb moisture prematurely initiating swelling prior to complete test setup.
The specimen and ring were next locked into place with the retaining ring and clamping screws. The loading cap and attached oven dry porous stone were then placed on the top of the specimen. The assembled oedometer cell was carefully placed in the loading frame.
The lever-arm was made level and balanced as the specimen loading cap just made contact with the loading ram. A small seating load was applied to the specimen by adjusting small sliding weights on the lever arm. The dial gage (calibrated to 0.0001 inches) used to measure vertical deformation of the sample was set, locked in place and an initial reading taken.
An initial load of 0.5 tons per square foot was applied to each specimen and allowed to consolidate until no further signs of compression were registered by the dial gage. Dial readings were taken at intervals of 1
second, 1 minute, 2 minutes, 4 minutes, 8 minutes, 15 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, 8 hours,
16 hours, and 24 hours, until primary consolidation was complete. (The dial reading at the end of the consolidation period was the target reading at which to return the sample after total volume change had occurred during the absorption of water).
The oedometer cell was then filled with deionized water to inundate the specimen. The specimens were each allowed to fully swell until no further expansion could be registered by the dial gage. (It is this reading that allows for the determination of swell potential or % swell). Once the specimen had reached full swell potential, the load was increased in 0.5 ton per square foot (1000 pounds per square foot, 6.9 pounds per square inch) increments until the target dial gage reading was reached. The effect of each load increment was allowed to compress the specimen until no further consolidation could be detected by the dial gage. The total load required to return the sample to the target dial gage reading determined the swell pressure.
Once the target dial gage was reached, the water