Penetrometers for

Soil Permeability
and Chemical Detection

prepared by

Susan E. Burns, PhD, P.E.

and

Paul W. Mayne, PhD, P.E.

Geosystems Engineering Group

School of Civil and Environmental Engineering
Georgia Institute of Technology
Atlanta, Georgia 30332-0355

Funding provided by
National Science Foundation
Arlington, Virginia

Army Research Office

Raleigh, North Carolina

July 1998
Penetrometers for Soil Permeability
and Chemical Detection

By Susan E. Burns and Paul W. Mayne

Funding provided by NSF and ARO

Issued by Georgia Institute of Technology
Report No. GIT-CEEGEO-98-1, July 1998

Executive Summary

Cone penetration soundings are an expedient

and efficient means of geoenvironmental site
characterization for delineating site stratigraphy.
Standard piezocone penetrometers provide continuous
measurements of point resistance (qc), sleeve friction
(fs), and shoulder porewater pressures (ub). Several
specialized penetrometers are available with additional
sensor capabilities, notably the seismic cone outfitted
with a velocity geophone to allow downhole
measurements of shear wave velocity (Vs) with depth.
The importance of soil permeability (k) is
recognized as a commonly-sought parameter in the
varied disciplines of water resources, environmental
sciences, geotechnical engineering, groundwater
systems, and waste management. Permeability is a
fundamental characteristic parameter reflecting the
ease through which fluids flow through porous media.
The permeability (or hydraulic conductivity) relates
directly to the coefficient of consolidation (ch) and
constrained modulus (D=1/mv). An analytical solution
is presented for interpreting the coefficient of
consolidation from piezocone dissipation tests in clayey
and silty soils, based on principles of cavity expansion
theory and critical-state soil mechanics. This solution
accommodates the routine monotonic decay of
porewater pressures with time associated with soft to
firm to stiff silts and clays, as well as the noted dilatory
responses (increase then decrease with time)
observed in overconsolidated fissured materials.
A new integrated optics (IO) chemical sensor is
introduced for realtime assessments of subsurface
pore fluid chemistry. This sensor is compact,
economical, reversible, and based on the principles of
wave-interference patterns to distinguish fluid types. It
is novel in that it is essentially "tuneable" to a wide
variety of different chemistries. In trial laboratory tests,
the BTEX (benzene-related) group of fuel contaminants
was assessed in small-scale chamber tests of
quartzitic sand.
 ACKNOWLEDGEMENTS

The authors extend their gratitude to Dr. Priscilla P. Nelson who served as the program director for
the Geomechanical, Geotechnical, and Geoenvironmental Engineering Division at the National
Science Foundation (NSF) that provided the primary funding support for these efforts. In particular,
the NSF program permitted the opportunities to develop field testing skills related to cone,
piezocone, and seismic cone penetration sounding, as well as the academic freedom to conduct both
analytical and numerical evaluations of the piezocone dissipation problem.

Additional funds for research were allocated by a separate source within NSF and by the US Army
Research Office (ARO). An NSF Traineeship in Sensors Development in the Chemistry Department
at Georgia Tech was issued from a program administered by Prof. Rick Browner at Georgia Tech. A
small technology transfer research (STTR) program was funded by ARO specifically for the work
on the integrated opto-electronic (IO) chemical sensor. This aspect of the research program
involved an interdisciplinary team comprised of Nile Hartman, Dan Campbell, and Jim Suggs of the
Electro0Optics, Environment, and Materials Laboratory (EOEML) within the Georgia Tech
Research Institute (GTRI), John Edwards of Photonics Sensor Systems (PSS), and Dr. Guangxuan
Zhu and Mark Prytula of the Daniel Environmental Laboratory at Georgia Tech.

During these studies, guidance and technical committees assisted the first author, Dr. Susan E. Burns
on her residency as both a graduate research assistant and graduate teaching assistant. The
committees consisted of Dr. Paul W. Mayne, Dr. Patricia M. Dove, Dr. J. David Frost, Dr. Glenn J.
Rix, and Dr. J. Carlos Santamarina. Technical assistance on certain aspects of the work were
provided by Ken Thomas and Seth Scott in the Geosystems Laboratory at Georgia Tech. Additional
help was given by several graduate students including Dr. Barry Shiyo Chen and Dr. Yasser A.
Hegazy.

Appreciation is also due to several individuals who assisted our research program over the past
several years. Dr. Jean-Lou Chameau (currently Dean of the GT College of Engineering) provided
the funds to purchase our first seismic cone penetrometer. Equipment donations were provided by
Tom Nolan and Jay Auxt of Hogentogler & Company, Columbia, Maryland. Partners from industry
also aided in securing outside funding and actual field experiences for testing and operation of the
seismic piezocone equipment, including: Dr. Stevan Vidic of AGRA Earth & Environmental,
Phoenix; Danny Brahana of Law/Gibb Group, Atlanta; Kevin McLain of the Missouri DOT; and
Mike Palmer of Law Engineering, Houston.

ii
 TABLE OF CONTENTS

Executive Summary ............................................................................................................. i

Acknowledgements............................................................................................................. ii
Table of Contents............................................................................................................... iii
List of Symbols .................................................................................................................. vi
Mathcad Symbols Used ................................................................................................... viii

CHAPTER 1 INTRODUCTION

Cone Penetration Testing ..............................................................................................1

Tip Resistance and Sleeve Friction Measurements ...................................................3
Porewater Pressure Measurements ............................................................................3
Sensor Technology for Geoenvironmental Site Investigation ......................................7
Geotechnical Sensor Technology ..............................................................................7
Inclinometer............................................................................................................8
Resistivity ...............................................................................................................8
Vibratory Module ...................................................................................................8
Accelerometer/Geophone .......................................................................................8
Neutron/Gamma Radiation.....................................................................................9
Lateral Stress ..........................................................................................................9
Acoustic Emissions ................................................................................................9
Pressuremeter Module ............................................................................................9
Time Domain Reflectometry ..................................................................................9
Video Cone ...........................................................................................................10
Environmental Sensor Technology..........................................................................10
Resistivity .............................................................................................................10
pH .........................................................................................................................10
Redox Potential ....................................................................................................10
Temperature..........................................................................................................10
Dielectric Constant ...............................................................................................10
Site Characterization and Analysis Penetrometer System (SCAPS)....................11
Gamma Radiation .................................................................................................11
Integrated Opto-Electronic Sensors......................................................................12
Sampling Technology ..............................................................................................12
BAT Enviroprobe .................................................................................................12
Hydropunch ..........................................................................................................12
Envirocone®.........................................................................................................12
Cone Sipper™ ......................................................................................................13
Summary of Cone Sensor and Sampling Technology .............................................13
Maintaining Integrity in the Sampling of Contaminated Fluids ..............................16
Chemical Compatibility........................................................................................16
Filter and Sampling Materials ..............................................................................17
Filters Used in Cone Penetration Testing.............................................................19
Summary of Chemical Compatibility Issues During Sampling...............................20
Experiments and Modeling in Geoenvironmental Cone Penetration Testing.............20

iii
CHAPTER 2 LITERATURE REVIEW

Introduction .................................................................................................................22
Evaluation of the Coefficient of Consolidation From Piezocone Dissipation Data....22
Generation of Excess Porewater Pressures..............................................................22
Pore Pressure Dissipation Behavior.........................................................................24
Radial versus Vertical Drainage...........................................................................27
Type 1 versus Type 2 Dissipation ........................................................................27
Existing Soil Models For Pore Pressure Dissipation ...............................................28
Dissipation in Normally- and Lightly-Overconsolidated Clays...........................29
Dissipation in Heavily-Overconsolidated Clays ..................................................40
Chronological Development of Pore Pressure Dissipation Modeling .....................42
Summary ..................................................................................................................42
Evaluation of Soil Permeability ..................................................................................44
Coefficient of Permeability......................................................................................44
Influence of Soil Parameters ................................................................................45
Measurement of the Permeability of Soils...............................................................45
Laboratory Measurement of Permeability............................................................45
Field Measurement of Permeability .....................................................................46
Piezocone Testing Evaluation of Permeability.....................................................47
Summary ..................................................................................................................51

CHAPTER 3 INTEGRATED OPTO-ELECTRONIC CHEMICAL SENSOR FOR BTEX

DETECTION IN CONE PENETRATION TESTING

Introduction .................................................................................................................53
Sensor Operation......................................................................................................53
Sensor Housing ........................................................................................................55
Fluid Sampling System ............................................................................................55
Geochemical Behavior of the BTEX Chemicals.........................................................57
Laboratory Testing Program .......................................................................................57
Experimental Results ..................................................................................................61
Discussion ...................................................................................................................66
Conclusions .................................................................................................................67

CHAPTER 4 PIEZOCONE MODEL FOR EVALUATING THE COEFFICIENT OF

CONSOLIDATION IN SOFT TO HARD CLAYS

Introduction .................................................................................................................68
Components of Porewater Pressure .........................................................................68
Shoulder Position Dissipation Model ......................................................................69
Normal-Induced Pore Pressures From Cavity Expansion Theory .......................69
Shear-Induced Pore Pressures From Modified Cam Clay....................................73
Derivation of the Analytical Solution to the Consolidation Equation..................74
Model Parameters .................................................................................................79
Evaluated Results From Test Sites..............................................................................82
Reference Values of the Coefficient of Consolidation ............................................82
Dissipation Sites ......................................................................................................83
iv
 Soft Clay Sites ......................................................................................................84
Stiff Clay Sites......................................................................................................88
Error Surfaces for Model Parameters..........................................................................93
Comparison With Existing Solutions..........................................................................99
Normalized Dissipation Data ......................................................................................99
Discussion .................................................................................................................104
Conclusions ...............................................................................................................104

CHAPTER 5 INTERPRETATION OF SEISMIC PIEZOCONE RESULTS FOR THE

EVALUATION OF SOIL PERMEABILITY IN CLAYS

Introduction ...............................................................................................................106
Seismic Cone Penetration Test ..............................................................................106
Vs-qT Relationships in Soils ...................................................................................110
Geotechnical Estimation of Mass Density.............................................................113
Methodologies for Evaluating Permeability from Piezocone Data...........................115
Evaluation of the Constrained Modulus from Piezocone Data..............................116
Cone Tip Resistance Correlation........................................................................116
Janbu's Method ...................................................................................................118
Shear Wave Velocity Correlation.......................................................................120
Evaluation of Permeability ....................................................................................122
Estimated Permeability Comparison with Existing Methods ................................125
Conclusions ...............................................................................................................126

CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS

Conclusions ...............................................................................................................128
Recommendations .....................................................................................................129

BIBLIOGRAPHY ..........................................................................................................131

v
 LIST OF SYMBOLS

Af = Skempton's pore pressure parameter at failure.

Bq = penetration pore pressure ratio = ∆u/(qT-σvo).
D = constrained modulus = 1/mv.
Dr = relative density.
E = Young’s modulus.
G = shear modulus.
Gmax = maximum shear modulus = small strain shear modulus.
Gs = specific gravity.
Ir = rigidity index = G/τmax.
J0 = Bessel function of the first kind of order 0.
J1 = Bessel function of the first kind of order 1.
K = hydraulic conductivity.
Kow = octanol-water partition coefficient.
LL = liquid limit.
M = slope of the critical state line = 6sinφ'/(3-sinφ') in triaxial compression.
Nf = bearing capacity factor.
Nk = cone bearing factor.
Nm = cone resistance number.
Nq = bearing capacity factor for deep failure.
Nu = bearing capacity factor for porewater pressure.
OCR = overconsolidation ratio = σp'/σvo'.
PL = plastic limit.
R = radius.
S = saturation.
T = time factor.
T* = modified time factor.
Vs = shear wave velocity.
Y0 = Bessel function of the second kind of order 0.
Y1 = Bessel function of the second kind of order 1.
a' = attraction = c' cotφ'.
an = net area ratio for unequal cone end areas.
av = coefficient of compressibility.
c' = effective cohesion intercept.
c = coefficient of consolidation.
ch = coefficient of consolidation in the horizontal direction.
cv = coefficient of consolidation in the vertical direction.
eo = void ratio.
f = modified hyperbola parameter.
fs = cone sleeve friction.
g = modified hyperbola parameter.
g = gravitational constant = 9.81 m/s2.
hw = height of water table.
k = coefficient of permeability (hydraulic conductivity) = k*ρg/µ.
k* = intrinsic permeability.
m = modulus number.

vi
mv = coefficient of volumetric compressibility = av/(1+eo).
n = porosity.
qc = measured cone tip resistance.
qT = corrected cone tip resistance = qc + (1-u2)an.
r = radius.
ro = radius of cone or probe.
rp = radius of plasticized zone.
rshear = radius of shear zone.
su = undrained shear strength.
t = time.
to = initial time.
t100 = time at which excess pore pressure = 0 (end of primary consolidation).
uo = hydrostatic pore pressure.
u1 = pore pressure measured on the face of a cone penetrometer.
u2 = pore pressure measured behind the tip of a cone penetrometer.
u3 = pore pressure measured behind the sleeve of a cone penetrometer.
um = measured pore pressure.
umax = maximum measured pore pressure.
∆u = excess pore pressure = um - uo.
∆um = pore pressure generated in excess of hydrostatic.
∆uoct = excess pore pressure due to changes in mean octahedral stress.
∆us = excess pore pressure due to changes in mean octahedral shear stress.
wn = water content.
x = distance.
α = Henkel's pore pressure parameter associated with ∆τoct.
β = Henkel's pore pressure parameter associated with ∆σoct.
γs = shear strain.
γsat = saturated unit weight.
γt = total unit weight.
γw = unit weight of water.
εa = axial strain.
φ' = effective stress friction angle.
ρ = mass density of the material = γt/g.
ρw = mass density of water = γw/g.
σa' = 100 kPa (reference stress = 1 atm).
∆σoct = change in octahedral normal stress.
σp' = preconsolidation stress.
σvo = total overburden stress.
σvo' = effective overburden stress.
∆τoct = change in octahedral shear stress.
τ = shear stress.
τmax = shear strength.
ν = Poisson’s ratio.

vii
 MATHCAD SYMBOLS

a = cone radius.
b = radius of the plasticized zone.
c = radius of shear zone.
σvo = total overburden stress.
eσvo = effective overburden stress.
uo = hydrostatic pore pressure.
u2 = pore pressure measured behind cone tip.
qT = corrected cone tip resistance.
rcone = cone or probe radius.
rplastic = radius of the plasticized zone.
φ = effective stress friction angle.
ch = coefficient of consolidation (radial).
Ir = rigidity index.
M = slope of the critical state line.
su = undrained shear strength.

viii
 CHAPTER 1

INTRODUCTION

Cone Penetration Testing

Cone and piezocone penetration tests are a widely utilized technology for the in-situ
investigation of soil deposits, and are primarily used for profiling soil stratigraphy and for
distinguishing fine-grained clayey soils from coarse-grained sandy soils. In addition to stratigraphy,
cone tests yield information about soil strength, soil stress history, and the hydrogeologic conditions
at a site, all of which provide valuable information for site characterization. More recently,
specialized sensors have been incorporated within the cone body to detect and measure
geoenvironmental properties in order to broaden the usefulness of the cone penetrometer. Cone
testing is ideal for environmental site characterization because it generates no cuttings, uses no
drilling fluids, minimizes the potential for aquifer cross-contamination through the use of grouting,
and reduces the chances of hazardous exposure for operating personnel.
A standard cone penetrometer has a diameter of 3.57 cm, a tip angle of 60 degrees, a
projected tip area of 10 cm2, and a friction sleeve with an area of 150 cm2 (Figure 1.1 and Figure
1.2). Established standards (ASTM D5778-95) allow for the design of larger penetrometers in order
to include additional sensors or to increase the robustness of the device. The geometry of the larger
penetrometer is: 4.37 cm diameter, a tip angle of 60 degrees, a projected tip area of 15 cm2, and a
friction sleeve area of 200 to 300 cm2. The penetrometers are hydraulically pushed from the soil
surface at a rate of 20 mm/s during a cone sounding using either a conventional drill rig or a cone
truck (Figure 1.3). Both the tip and the sleeve of the penetrometer are instrumented with strain
gauges to measure stresses exerted on the soil during penetration (qc = tip resistance

Figure 1.1. View of Miniature, Four Standard-Size, and Large Penetrometers

1
 Figure 1.2. Diagram of Piezocone Sensing Elements

Figure 1.3. Fugro Cone Truck During Cone Penetration Testing

= stress exerted on cone tip and fs = sleeve friction = stress exerted on cone sleeve), while pore
pressure transducers are included in the cone to measure the pore pressure generated on the cone tip
(u1 or Type 1 pore pressure), on the cone shoulder (u2 or Type 2 pore pressure), or behind the
friction sleeve (u3 or Type 3 pore pressure). Figure 1.1 shows a photograph of a variety of
piezocone penetrometers used for site investigation. The figure shows, from left to right, a
miniaturized half-scale cone penetrometer (4.2 cm2), a Type 2 electric piezocone (10 cm2), a Type 1
electric piezocone (10 cm2), a Type 2 seismic electronic piezocone (10 cm2), a dual-element (Types
1 and 2) seismic electronic piezocone (10 cm2), and a triple-element (Types 1, 2, and 3) electronic
piezocone (15 cm2).

2
Tip Resistance and Sleeve Friction Measurements
The tip and sleeve readings of a cone penetrometer are used in combination with the
magnitude of the excess pore water pressure for determining soil stratigraphy and for the evaluation
of soil properties. Many classification methods exist for the determination of soil types, and the
methods most commonly use a combination of the measured tip resistance and sleeve friction
(Robertson and Campanella, 1983; Olsen, 1994); however, methods that utilize the reading of pore
pressure are also available (Senneset et al., 1989; Robertson, 1990).
The readings from a cone are semi-continuous, performed every two centimeters of push, and
are rapid, economical, and reproducible. The readings in a cone penetration test are actually
averaged values measured over the cone's effective zone of influence, which is a function of the soil
type being penetrated. Soil parameters derived from cone penetration data include the stress history
of clays (Chen and Mayne, 1994), unit weight of clays (Larsson and Mulabdic, 1993), relative
density of sands (Jamiolkowski et al., 1985), and the effective friction angle of sands (Robertson and
Campanella, 1983) and clays (Senneset et al., 1989). A detailed listing of the parameters derived
from piezocone testing is presented in Mayne et al. (1995).

Porewater Pressure Measurements

As discussed earlier, piezocone penetrometers are instrumented with at least one pressure
transducer to measure the penetration pore water pressures (um) generated during a sounding. This
allows for the calculation of excess pore pressure, ∆u = um - uo (Janbu and Senneset, 1974; Baligh et
al., 1981), where ∆u is the difference between the generated pore pressure and the hydrostatic
pressure (uo). The inclusion of the pore pressure reading was a significant improvement in cone
technology because the generated excess pore pressure measurements provide important insight into
the type of soil being penetrated. Soundings in fine-grained soils, which have a characteristic low
hydraulic conductivity, generate large pore pressures in excess of hydrostatic values because the
displaced water cannot move freely throughout the soil column; consequently, the sounding occurs
under undrained conditions. The interpretation of the data gathered from pore pressure
measurements is also useful for the evaluation of geoenvironmental parameters including the
coefficient of consolidation and the permeability of a soil deposit. These properties give information
about the flow of pore fluid in a soil matrix and are important in applications such as consolidation
of clay landfill liners and contaminant transport.
Coarse-grained soils which have high hydraulic conductivity do not generate large excess
pore pressures because the displaced water can easily dissipate to other parts of the soil column as
shown in Figure 1.4 and Figure 1.5. This sounding was performed in Po River sand in Italy, and the
penetration pore pressure essentially matches hydrostatic pressure throughout the profile.
In contrast, Figure 1.6 shows an example piezocone record a soft insensitive lacustrine clay
where significant excess pore pressures were generated (pore pressures measured in both the u1 and
u2 positions, with u1 always greater than u2). Figure 1.7 shows similar behavior in a soft sensitive
clay from St. Alban, Quebec; note that increasing the sensitivity of the clay decreases the difference
between the pore pressures measured in the Type 1 and 2 positions (Mayne et al., 1990). Finally,
Figure 1.8 shows the pore pressure behavior in the heavily-overconsolidated Gault clay.
In lightly-overconsolidated clays, the measured penetration pore pressures are always
positive, regardless of the location of the pore pressure transducer (Figure 1.6 and Figure 1.7).
However, in heavily-overconsolidated clays, the penetration pore pressures measured in the Type 2
or Type 3 locations can either be positive or lower than hydrostatic, or even negative, while the
penetration pore pressures measured in the Type 1 position will always remain positive. shows the
results of a dual element piezocone sounding in the heavily-overconsolidated and fissured Gault clay

3
where the measured Type 1 pore pressure is always positive, while the Type 2 pore pressure goes
lower than hydrostatic and negative during the sounding.

Penetration Pressures (MPa)

0 5 10 15 20 25 30 35
0

10
Depth (m)

u2
15 u1 u1
u2 qT
20

25 Po River Sand, Italy

uo (Jamiolkowski et al., 1985)
30

Figure 1.4. Penetration Pressures in Medium to Dense Sand,

Drained Conditions

Penetration Pore Pressures (MPa)

0 0.25 0.5 0.75 1
0
Po River Sand, Italy
5 u2 u1 (Jamiolkowski et al., 1985)

10
Depth (m)

15

20

25 uo
u2
u1
30

Figure 1.5. Penetration Pore Pressures in Medium to Dense Sand,

Drained Conditions

4
 Penetration Pressures (kPa)
-500 0 500 1000 1500 2000 2500
0

10
Depth (m)

15 u2
u1
uo
20 qT
u2 u1
25

30 Port Huron, MI
(Chen and Mayne, 1994)
35

Figure 1.6. Penetration Pressures at Port Huron, MI in a Soft Insensitive Clay

Penetration Pressures (kPa)

0 100 200 300 400 500 600
0

u2
4
Depth (m)

u1

u1 qT
6
u2
uo
8
St. Alban, Quebec
(Roy et al., 1982)
10

Figure 1.7. Penetration Pressures in a Soft Sensitive Clay

5
 Penetration Pressures (MPa)
-2 0 2 4 6
0

5
Depth (m)

10 u2
u1
qT
15
u2 uo u1

20
Gault Clay, Madingley, U.K.
(Lunne et al., 1986b)
25

Figure 1.8. Penetration Pressures in Heavily-Overconsolidated Gault Clay

The pore pressure transducer in the piezocone is in contact with a cavity within the cone
body, and pore water is brought into the cavity through a filter which is most commonly made of
polypropylene, stainless steel, or a ceramic. Before a sounding is performed, the cavity is saturated
with water and the pore pressure filter is saturated with either water, a water/glycerin mixture, or in
some cases, a heavy oil. During the initial development of the piezocone, the location of the pore
pressure filter was not standardized; early versions of the piezocone had filter locations ranging from
the apex to various positions on the shaft. Most filter positions today occur either on the
penetrometer tip or midface (Wissa et al., 1975), behind the shoulder (Jamiolkowski et al., 1985), or
behind the friction sleeve (Campanella and Robertson, 1988) (Figure 1.2). The location of the filter
element is an important consideration because the magnitude of the measured pore pressure has been
shown to decrease as the filter location is moved up the cone body, away from the cone tip
(Campanella et al., 1986a; Brown, 1993); that is, u1 > u2 > u3.
The Type 1 filter location is advantageous because the measured pore pressures are always
positive at this location which gives excellent profiling detail (Mayne et al., 1990). However, the
Type 2 position is necessary for the correction of tip resistance measured with pore pressure acting
on unequal end areas of the cone tip (Jamiolkowski et al., 1985; Lunne et al., 1986a). The correction
is given as follows:

q T = q c + (1 − a n )u2 Equation 1.1

where qT = corrected tip resistance, qc = measured cone tip resistance, an = net area ratio determined
in a calibration chamber, and u2 = Type 2 pore pressure measured at the shoulder position. The net
area ratio is a function of the cone geometry and must be measured using a pressurized triaxial

6
chamber. The reported values of an for commercially built cone penetrometers range from an = 0.38
to 0.85. It is preferable to design cones with values of an as large as possible (an > 0.8) because the
cone tip resistance correction is smaller. The correction factor is especially important in soundings
in soft to firm fine-grained soils due to the large excess pore pressures which can develop, and in
deep soundings where hydrostatic readings are large.

Sensor Technology for Geoenvironmental Site Investigation

While the traditional measuring capabilities of a cone penetrometer include tip resistance,
sleeve friction, and pore pressure, the cone makes an excellent vehicle for transporting other
measuring devices into the subsurface, with less disturbance when compared to traditional boring
methods like auger drilling which create large diameter holes and soil cuttings. Consequently,
additional sensors have been incorporated into cone penetrometers in order to measure soil
properties or geoenvironmental characteristics of a soil deposit. These sensors are most commonly
included in either the portion of the cone which is directly behind the friction sleeve, or in a separate
module which trails directly behind the cone body. An overview of the sensor technology in use
today is presented in the following paragraphs.
The main goals of a geoenvironmental site characterization include the delineation of soil
stratigraphy and hydrogeologic conditions, and the detection of the extent and magnitude of
chemical contamination. Because the investigations are complicated by the presence of hazardous
chemicals which can harm operating personnel, recent emphasis has been placed on developing
investigation techniques which perform the measurements and sampling either below the soil
surface, or within a completely contained sampling module which transfers liquid and vapor to the
surface for analysis.
Site investigation practice is beginning to de-emphasize the use of traditional boring and
sampling techniques and has begun to rely on the use of either semi-invasive techniques like cone
penetration testing, or on non-invasive techniques like seismic investigations. However, quantitative
evaluation of contaminant characteristics as a function of depth requires some minimum degree of
invasion; consequently, the use of the cone penetration test as a vehicle for the transport of
specialized sensors into the subsurface has become a common approach to site characterization.
These sensors commonly take one of two forms: either a direct non-sampling sensor which performs
measurements on the probe surface, or an indirect sampling sensor which draws an amount of pore
fluid or gas into the cone body or through transmission lines to the soil surface for analysis. Sensor
technology which does not require drawing pore fluid into the cone typically cannot provide
quantitative measurements, it can only map the relative qualitative differences in soil properties, for
example changes in bulk resistivity. To date, the quantitative measurement of chemical
concentrations in-situ can be performed only through sampling of the soil, pore fluid, or vapor. The
following section provides a review of both the sensor and sampling technology available for use in
geoenvironmental site investigation. For the purpose of review, the available sensor technologies
have been divided into applications used mainly for geotechnical or for environmental purposes;
however, the technology may be applicable to both cases in some instances.

Geotechnical Sensor Technology

One group of sensors focuses on the measurement of mechanical soil properties used in
geotechnical site characterization and design, including measurements of strength, deformability, in-
situ states of stress, and moisture content. These are discussed in the following paragraphs.

7
 Inclinometer
Inclinometers have been included in cone penetrometers in order to monitor the verticality of
the cone during a sounding (Campanella and Robertson, 1988). A cone can deviate from verticality
due to misalignment, bent push rods, soil layering, or be pushed off a vertical course during
penetration if it encounters an obstacle such as a large rock. A large deviation from the initial
alignment can cause breakage of the cone rods and loss of the cone down the hole; consequently,
many penetrometer systems are equipped with an automatic shutoff if the cone exceeds an
inclination of 9o. While the sensor measures the inclination of the cone, the data are not used to
evaluated soil parameters.

Resistivity
The resistivity cone was developed to detect changes in the electrical conductance properties
of a soil pore fluid (Horsnell, 1988; Campanella and Weemees, 1989). Resistivity, the reciprocal of
conductivity, is a measure of the mobility of ions through a medium; in general, sands exhibit high
resistivity (low conductivity), while clays, which are relatively electrically active, exhibit low
resistivity (high conductivity). Resistivity cones operate under the same principle as in surface
geophysics: four electrically-isolated electrodes are built into a module which trails the cone
penetrometer and a current is applied across the two outer electrodes, while the voltage drop across
the two inner electrodes is measured. The measured property is the resistance, which is a function of
probe geometry, and the resistivity, which is a material property, is calculated from the measured
resistance. Measurements from soil resistivity probes have been used to estimate the porosity and
formation factor of sands through use of the empirical relationship known as Archie's Law (Bellotti
et al., 1994).

Vibratory Module
Cone penetrometers have been designed with a trailing vibratory module for use in the
identification of zones of loose sands which are susceptible to liquefaction during an earthquake
(Sasaki and Koga, 1982; Sasaki et al., 1985; Koga et al., 1986; Teparaksa, 1987, 1990; Mitchell,
1988; Tokimatsu, 1988; Jamiolkowski and Robertson, 1988; Du and Wu, 1992). The method of
evaluation requires one cone sounding without the use of vibration during penetration (static
sounding), and a second cone sounding with the use of vibration (dynamic sounding). The two
soundings, one prior to vibratory densification and one after vibratory densification, are then
compared, and the liquefiable sand zones are identified by a significant increase in measured cone
tip resistance in the areas which were densified by vibration.

Accelerometer/Geophone
One of the most commonly used sensors for geotechnical site investigation is the
accelerometer or geophone, which is included in the cone for measurement of the shear wave
velocity (Vs) or corresponding small strain properties of a soil deposit (Robertson et al., 1986).
These sensors can also measure vibration and damping characteristics of a soil deposit, and the
compression wave velocity (Vp). The seismic cone test is most commonly performed as a downhole
test with a mechanical source placed at the surface to generate shear waves which propagate through
the soil deposit. The travel time from source to receiver is monitored through the detection of the
arrival time at the sensor within the cone body, and the shear wave velocity can be calculated by
division of the travel distance by the travel time. These seismic measurements are also used for
environmental applications including waste detection and landfill delineation.

8
 Neutron/Gamma Radiation
Neutron and gamma radiation sensors have been included in cone penetrometers in order to
evaluate the moisture content and relative density of a soil deposit (Marton et al., 1988; Mitchell,
1988; Sully and Echezuria, 1988; Shibata et al., 1992). Neutron sensors rely on the interaction of
the neutrons with the hydrogen present in the pore fluid, while gamma-radiation sensors rely on the
interaction of gamma rays with soil particles. These sensors are most commonly used to measure
the pore fluid and density characteristics of a soil deposit; however, studies have also been
performed which relate the natural water content measured with the radiation sensors to the relative
density and liquefaction potential of a soil deposit (Mimura et al., 1995).

Lateral Stress
Lateral stress cells have been developed in order to evaluate the in-situ state of horizontal
stress (σh) of a soil deposit (Mitchell, 1988; Sully, 1991). Lateral stress cone designs typically place
the measurement module behind the cone body in order to remove it from the variable stress regime
which occurs at the cone tip during penetration. The cell is designed for freedom of radial
movement and is instrumented with strain gauges in order to monitor the lateral stress exerted on the
cone body. The lateral stress cone provides a direct measurement of the total horizontal stress
during penetration, a property which is a function of the strength and stress history of a soil deposit
and is difficult to quantify through laboratory experimental methods.

Acoustic Emissions
Monitoring of the acoustic emissions produced during a cone sounding has been performed
using microphones included within the cone body (Villet et al., 1981; Tringale and Mitchell, 1982)
and by inclusion of a piezoelectric acoustic emission transducer located directly behind the cone tip
(Massarsch, 1986). Penetration of a cone into soil produces audible emissions due to the interaction
of the soil with the probe body and push rods. Testing using the acoustic emission sensor has shown
that layering of different soil types (sand, silt, or clay), changes in relative density, and changes in
fabric can be identified (Menge and Van Impe, 1995).

Pressuremeter Module
Pressuremeter modules have been added as components of cone penetrometers in order to
facilitate full-displacement pressuremeter testing procedures. Cone pressuremeters combine the
traditionally measured cone parameters with measurements of strength, deformation, and horizontal
stress measurements from the pressuremeter module (Houlsby and Withers, 1988; Houlsby and
Hitchman, 1988; Ghionna et al., 1995). The pressuremeter module trails the cone penetrometer and
is an inflatable probe which is expanded into the soil in a radial direction. Results from the
pressuremeter test can also be used to evaluate the complete stress-strain-strength behavior curve for
soils.

Time Domain Reflectometry

Time domain reflectometry sensors have been included in cone penetrometers for the
measurement of soil/pore fluid dielectric constant (Lightner and Purdy, 1995). The sensor operates
by monitoring the response from a pulsed electromagnetic wave sent through the surrounding soil.
The dielectric constant is then correlated to moisture content in terms of volumetric water content.
Research at Sandia National Laboratory measured moisture by volume to within + 1% when
compared to laboratory measurements.

9
 Video Cone
Video Cones have been developed which record the soil images along the edge of a cone
during a sounding (Hryciw and Raschke, 1996; Raschke and Hryciw, 1997). The imaging is
performed through a sapphire window on the body of the cone, and grain size analysis can
subsequently be performed on the recorded data using digital image analysis.

Environmental Sensor Technology

This group of sensors focuses on the evaluation of soil properties or contaminant
characteristics used in environmental site investigations. These sensors most commonly infer the
location of contamination through measurement of a property of the bulk soil/pore fluid matrix, for
example bulk resistivity. No sensor technology exists today which can provide quantitative
measurements of chemical concentration without sampling of contaminated media.

Resistivity
The resistivity cone is used in environmental investigations to delineate soil contamination
through changes in the measured bulk resistivity (Horsnell, 1988; Campanella and Weemees, 1989;
Strutynsky et al., 1991; Woeller et al., 1991a). The resistivity cone has been used to detect acidic
spills, which produce highly conductive conditions, and to detect organic non-aqueous phase liquid
(NAPL) spills, which are highly resistive contaminants. Use of the resistivity cone requires a site-
specific calibration because it is used to detect relative differences in magnitude, rather than absolute
values.

pH
Cone penetrometers have been designed with pH sensors to identify areas of acidic or basic
contamination (Olie et al., 1992). The sensor has a small flow cell within the cone body in order to
bring fluid into contact with the miniaturized glass, gel-filled Ag/AgCl pH electrode. Use of the pH
sensor has enabled mapping of pH as a function of depth, and monitoring of the progress of the
remediation of acidic waste sites.

Redox Potential
Redox potential sensors have been included in cone penetrometers to measure the reduction-
oxidation potential of a site (Olie et al., 1992). The redox potential of the bulk soil/pore fluid matrix
is useful for the monitoring of hazardous waste sites under bio-remediation because it is a measure
of the state of electron donors and acceptors, a critical control on biological activity (Pluimgraaf et
al., 1995).

Temperature
Thermocouples have been included to monitor the temperature of the cone body during a
sounding (Horsnell, 1988; Mitchell, 1988). Applications of the temperature sensor include the
mapping of zones of endothermic or exothermic temperature changes, due either to chemical
reaction such as methane generation in landfills, or to biological activity. Depth of permafrost has
also been mapped using the temperature sensor (Woeller et al., 1991b).

Dielectric Constant
Robertson et al. (1996) give details of a cone penetration device developed to measure the
dielectric constant of a soil pore fluid mixture. The dielectric constant is a measure of a molecule's
interaction with an electromagnetic field, and is a function of the applied frequency. A retractable
cone has been developed by Delft Geotechnics which operates by pushing to the depth of interest,
10
opening a soil sample chamber and pushing until the chamber is filled (Stienstra and van Deen,
1994). An electromagnetic field pulse is generated at the soil surface and transmitted through a
coaxial cable to a pin in the middle of the chamber which is the antenna; the chamber rim is the
receiver. Subsequently, the dielectric constant can be measured as a function of frequency to help
identify areas contaminated with NAPLs which have a characteristic low dielectric constant (~3 to
8), when compared to water which has a dielectric constant of 80.

Site Characterization and Analysis Penetrometer System (SCAPS)

One of the most promising cone penetration tools currently in use is the Site Characterization
and Analysis Penetrometer System (SCAPS) developed by the US Department of Defense
(Lieberman et al., 1991; Apitz et al., 1992a, 1992b; Theriault et al., 1992). The SCAPS cone was
developed as a field screening tool for the detection of the constituents of fuel products and oils. It
is also known as the laser-induced fluorescence cone (LIF-CPT). The basic operation of the cone
relies on the monitoring of a fluorescence response of aromatic hydrocarbons in aqueous solutions.
Fluorescence is a property of a group of molecules which sorb a quantum of light when exposed to
an energy source. This forces the molecules into a higher, or activated energy state, as long as the
external excitation is present. When the source is removed, the molecule returns to its normal
energy level and releases the quantum of energy as light. This release can be detected, and the
phenomenon is known as fluorescence. The cone penetrometer used with the SCAPS system is built
with a sapphire window in a slight recess on the cone body; transmission and receiving fiber optic
cables are strung down the cone rods to the window and light from an N2 laser is pulsed down the
transmission cable. When the excitation is removed, the fluorescent chemicals will emit light, which
is transmitted to the surface using the receiving fiber optic cables. Because chemicals fluoresce at
different wavelengths, the frequency of the laser light is adjustable in order to optimize detection.
While the SCAPS cone is a valuable tool for the screening of fuel contamination, it has
several limitations. These include the fact that background sources, including natural organic matter
can also fluoresce, making it necessary to acquire a site specific baseline at each testing location.
The measurement is limited to qualitative data; estimations of concentration are performed only
through correlations with measured concentrations in given soil conditions. Finally, the sensor
operation is limited to the group of compounds which fluoresce, and is not capable of detecting
many chemicals of environmental concern, including the chlorinated hydrocarbons like
trichloroethylene or perchloroethylene.
Another sensor, known as the Rapid Optical Screening Tool (ROST™) system relies on the
same concept of fluorescence but uses a neodymium-doped yttrium aluminum garnet laser operating
at wavelengths of 280 to 300 nm, instead of the N2 laser used by the SCAPS cone (Lightner and
Purdy, 1995).
Additionally, a Raman spectrograph sensor has been included in a cone and also operates
under conditions similar to the SCAPS and ROST™ cones. Use of the Raman spectrograph enables
the detection of chlorinated hydrocarbons in addition to fluorescence.

Gamma Radiation
The gamma radiation sensor was developed to identify radioactive contaminants in-situ
(Brodzinski, 1995). The sensor uses a sodium iodide crystal (NaI) in combination with spectral
analysis for the detection of radioactive 238U daughter products; however, initial trials with the
sensor indicated high detection limits (Lightner and Purdy, 1995).

11
 Integrated Opto-Electronic Sensors
A new class of sensor technology has been developed in order to quantitatively measure
chemical concentrations in-situ. These sensors rely on chemical reaction of the contaminant of
interest with a polymeric coating on an optically transparent waveguide (Hartman et al., 1988;
Hartman, 1990). This interaction produces a change in the refractive index and alters the travel of a
beam of light through the waveguide; the alteration can be measured and correlated linearly with
chemical concentration. The sensor has been laboratory tested for the measurement of pH (hydrogen
ion concentration) and ammonia, and is currently under development for the detection of the BTEX
and other chemicals. A more detailed explanation of the operation of this sensor will be presented in
Chapter 3.

Sampling Technology
Current site investigation practice requires either the drilling of a monitoring well or the use
of a discrete water sampler for the quantitative evaluation of groundwater. A monitoring well is
typically five centimeters (two inches) or larger in diameter, and requires screening at the depth from
which the sample is to be taken. While monitoring wells are the most common method for analyzing
groundwater, there are several drawbacks to their use, including expense, limited sampling depths,
generation of soil cuttings which may be contaminated, and venting of groundwater to the
atmosphere, which is especially detrimental in the detection of volatile chemicals. In addition,
sampling of monitoring wells is most commonly performed using a bailer which further increases
exposure to atmospheric oxygen. Discrete water samplers were developed in order to address some
of the problems seen in monitoring wells. The following section reviews the direct push technology
(DPT) samplers which are in use today.

BAT Enviroprobe
The BAT Enviroprobe was developed by Torstensson (1984) for the sampling of
contaminated groundwater. The push-in sampler has a similar outer configuration to a cone
penetrometer; however, glass sample vials which are sterilized and under vacuum are stored inside
the probe body. When the depth of interest is reached, an outer sleeve is retracted to expose a filter.
A double-ended hypodermic needle is then pushed through two septums which connect the filter to
the sampling vial, and pore fluid is drawn into the chamber. The BAT Enviroprobe is best suited for
taking small volume fluid samples of less than 150 mL.

Hydropunch
The Hydropunch is a liquid sampler which also has the outer form of a cone penetrometer.
This sampler is pushed to the desired sampling depth, and an outer sleeve is retracted to expose a
stainless steel filter and a miniaturized bailer is lowered through the push rods and probe body in
order to take a sample (Robertson et al., 1996). A significant difficulty associated with the use of
the Hydropunch is the high sample turbidity due to incomplete screening of fines.

Envirocone®
The Envirocone® was developed by ISMES for characterization of groundwater and vapor at
contaminated sites (O'Neill et al., 1996; Piccoli and Benoit, 1995). The device is pushed to the
sampling depth of interest and pore fluid is drawn in through either stainless steel or sintered steel
filters located above the probe tip. The device uses a positive displacement pump, contained in the
probe body, to transfer the samples through Teflon sampling lines (2 mm diameter) to the soil
surface for analysis. At the surface, the sample is analyzed for pH, redox potential, dissolved
oxygen, electrical conductivity, and temperature. Vapor samples are analyzed for dissolved oxygen,
12
total volatile organics, combustible gas, H2S, and CO2. Flushing of the system in order to minimize
the possibility of sample cross-contamination is performed using a pressurized air/water flushing
system.

Cone Sipper™
The Cone Sipper™ was developed as a mechanism to rapidly transfer ground water samples
to the soil surface (Lightner and Purdy, 1995). The probe uses a gas lift pump in combination with a
series of check valves, vacuum lines, and pressure lines to pump pore fluid from the probe intake to
analytical equipment above ground. Samples up to 100 mL in volume are obtained with the system.

Summary of Cone Sensor and Sampling Technology

The review presented here has divided the cone technology available for geoenvironmental
site investigation into two general categories: sensor equipment and sampling equipment. For the
most part, sensor technology is taken as that which can perform measurements on the sensor surface,
without bringing a sample into the probe body, while a sampling technology requires a mechanism
to draw the specimen in to the cone body.
In terms of sensors available for site investigation, the resistivity cone and the SCAPS cone
have proven to be the most viable and useful of the environmental characterization tools at the
present time. Resistivity is commonly used for the investigation of inorganic, ionic contaminants,
while the SCAPS cone is used for identification of fluorescent organic contaminants. In practice,
both of the sensors yield measurements which are qualitative indicators of the presence of
contaminants, and so are most useful as screening tools for initial phase investigations. A summary
of the available sensor technology for both geotechnical and environmental applications is presented
in Table 1.1 and Table 1.2 (modified after Bowders and Daniel, 1994).
Sampling tools have been developed in order to obtain quantitative measurements of
contaminant concentrations which can be used as the basis for remedial design. The use of in-situ
samplers tends to be slow and requires chemical analysis by laboratory instrumentation such as a gas
chromatograph, either at the soil surface, or in an off-site laboratory location. The other main
concern associated with sampling technology is the preservation of sample integrity during
acquisition. Probes which use sample transfer lines, or the same filter to obtain a series of samples
must be equipped with flushing systems and constructed of materials which are resistant to cross-
contamination. Because the issue of cross-contamination is of concern for the operation of the
BTEX sensor used in this study, a review of the materials most suited to sampling of environmental
contaminants follows.

13
 Table 1.1. Review of Geotechnical Sensors Used in Piezocone Testing
Sensor Measurement Applications Reference
Inclinometer Cone verticality • Prevent damage of Campanella et al. (1986b)
cone
Resistivity Mobility of ions in • Porosity of sands Bellotti et al. (1994);
pore fluid using • Fabric of sands Campanella and Weemees
electrically • Conductivity = (1989)
isolated electrodes 1/Resistivity
Vibratory Module Vibration of cone • Assessment of soil Sasaki and Koga (1982);
during push liquefaction potential Sasaki et al. (1985);
Mitchell (1988)
Accelerometer/ Shear wave • Measurement of small Robertson et al. (1986)
Geophone velocity strain properties of a
soil
• Site specific Gmax
• Particle velocity for
damping ratio
Neutron/ Gamma Moisture content • Soil density Marton et al. (1988);
Radiation • Moisture content Mitchell (1988); Sully and
• Correlation with Echezuria (1988);
liquefaction potential Mimura et al. (1995)
Lateral Stress Lateral stress on • Evaluation of in-situ Mitchell (1988);
cone shaft states of stress Sully (1991)
Acoustic Acoustic • Soil type Villet et al. (1981); Tringale
emissions • Soil compressibility and Mitchell (1982); Menge
• Soil fabric and Van Impe (1995)
Pressuremeter Radial • Shear strength Houlsby and Withers
Module deformation • Horizontal stresses (1988); Houlsby and
(Full • Deformability Hitchman (1988); Ghionna
displacement) et al. (1995)
Time Domain Dielectric constant • Correlated with Lightner and Purdy (1995)
Reflectometry through pulsed moisture content
electromagnetic
wave
Video Video images of • Grain size Hryciw and Raschke
soil during quantification (1996); Raschke and
penetration • Soil stratigraphy Hryciw (1997)

14
 Table 1.2. Review of Geoenvironmental Sensors Used in Piezocone Testing
Sensor Measurement Applications Reference
Resistivity Mobility of ions in • Salt water intrusion Horsnell (1988);
pore fluid using • Acid spills Campanella and Weemees
electrically isolated • Detection of water table (1989); Strutynsky et al.
electrodes in mine tailings (1991); Woeller et al.
(1991a); Malone et al.
(1992)
Temperature Temperature of • Endothermic/ Horsnell (1988);
cone body exothermic activity Mitchell (1988); Woeller et
al. (1991b)
SCAPS Laser-induced • Fuel, oil, and lubricant Lieberman et al. (1991);
fluorescence of fuel contamination capable of Apitz et al. (1992a); Apitz
contaminants; N2 fluorescing et al. (1992b); Theriault et
laser at λ = 337 nm al. (1992); Lambson and
Jacobs (1995)
Redox Potential Reduction- • Monitoring of conditions Olie et al. (1992);
Oxidation Potential during bio-remediation Pluimgraaf et al. (1995)
pH Hydrogen ion • Acid spills Brylawski (1994)
concentration • Base spills
Dielectric Dielectric constant • NAPL contamination Arulmoli (1994); Stienstra
Constant of soil/pore fluid and van Deen (1994)
mixture as a
function of
frequency
Raman Raman spectrograph • NAPL contamination Carrabba (1995)
Spectroscopy to measure argon • Chlorinated hydrocarbons Bratton and Timian (1995)
ion laser induced
fluorescence
ROST™ Laser-induced • Fuel, oil, and lubricant Naval Command (1995)
fluorescence of fuel contamination capable of
contaminants; fluorescing
neodymium-doped
yttrium aluminum
garnet laser at λ =
280 - 300 nm
Gamma Radiation Detection of • Identification of Brodzinski (1995);
Probe Uranium by- radioactive contaminants Lightner and Purdy (1995)
products using a
NaI(Tl) crystal
detector
Integrated Measurement of in- • Ammonia Hartman et al. (1988);
Optoelectronics situ chemical • pH Hartman (1990)
concentration by • BTEX This study
wave interference

15
Maintaining Integrity in the Sampling of Contaminated Fluids
Sampling of contaminated fluids for chemical analysis requires careful consideration of the
process and materials used in order to maintain specimen integrity. In order to draw ambient pore
water vapors and fluids from the soil environment into the probe chamber, it is important to have a
porous filter stone that allows the transmission of gases and fluids, yet prevents the influx of soil
particles. These aspects are complex within the scope of soil and fluid mechanics because of the
vast range in particle sizes encountered in natural geomaterials. The size classification of particles
ranges several orders of magnitude from gravel to sand ( 4.75 mm to 0.075 mm), to silt (0.075 mm
to 2 µm), to clay (< 2 µm), as shown in Figure 1.9. The porous filter must also have a permeability
high enough as to not inhibit flow and volume of fluids for testing. Finally, the porous filter must be
constructed of a material that does not interact with the groundwater chemistry and taint the desired
measurements.
Sampling of contaminated groundwaters is also complicated by the complex aqueous
chemistry often found at hazardous sites. One of the main difficulties in working with contaminated
materials is the extreme variability of chemical conditions that exist in practice. Conditions range
from low to high pH, from oxidizing to reducing conditions, and from contamination by metals to
contamination by organics; consequently, the materials used in the investigation and characterization
of a site must be suitable to resist the contaminates of concern at a particular site in order to maintain
integrity in the monitoring process.
All direct push groundwater and vapor samplers require filtering mechanisms in order to
draw a fluid sample into the probe body. Because the filter is in direct contact with the soil during
push, it must resist abrasion and clogging, and must be chemically compatible with the surrounding
groundwater. Filter materials commonly used in direct push applications in non-contaminated soils
include high density polyethylene, ceramic, brass, polyvinyl chloride, and sintered stainless steel
(Marsland and Quarterman, 1982; Campanella and Robertson, 1988; Bratton and Timian, 1995;
O'Neill et al. 1996). However, of additional concern to the issue of sampling for chemical analysis
at contaminated sites is the material or materials which will come into contact with the given sample
through filter inlets, sampling lines, or storage vials. Not only must these materials be physically
durable in order to withstand the abrasion and high pressures generated during DPT penetration, but
they must also be chemically resistant and non-reactive with the target chemicals of interest.

Chemical Compatibility
The materials most commonly used in the sampling of groundwater include stainless steel,
polytetrafluoroethylene (PTFE, Teflon®), and polyvinyl chloride (PVC). Each of these materials
possesses advantages and disadvantages and none is suited to sampling in all conditions, because the
physical and chemical properties of contaminants of concern cover a range of diverse behaviors. In
general, however, metals like stainless steel are most applicable for the sampling of organics, while
organic materials like Teflon® or rigid PVC are most applicable for the sampling of metals.

16
 Figure 1.9. Range of Particle Sizes in Nature
(Mitchell, 1993)

Parker (1992) developed suggested guidelines for the use of stainless steel, PVC, and
Teflon® for the sampling of groundwater, with a focus on chemical and physical durability, and on
material effects on sample integrity. Of the three materials, stainless steel has the greatest physical
strength, with corrosion resistance which is generally quite good, except in the long-term exposure
to highly corrosive environments with low pH, high dissolved oxygen contents, hydrogen sulfide,
carbonic acid, or high chloride concentrations. In terms of sample integrity, Parker (1992) notes that
stainless steel casing can react with or sorb metals from solution or can leach metals into solution,
especially from the weld joints. However, stainless steel shows little interaction with organic
contaminants, either in terms of sorption or outgassing into solution. These properties suggest that
stainless steel provides a better filtration and sampling material for organic contaminants than for
metallic contaminants.
Teflon® is used extensively in environmental sampling, mainly due to its corrosion and
dissolution resistance. Parker (1992) reports that Teflon® shows little interaction with metals in
solution, either in terms of sorption onto the Teflon® surface, or in terms of leaching metals into
solution. While it is known to sorb/partition some organics from solution into its matrix, Teflon®
does so at much lower concentrations than seen in flexible PVC (Tygon). Leaching of organics from
Teflon® into solution is negligible. In contrast to Teflon®, Tygon will dissolve or swell when
exposed to highly concentrated solvents. It also has a tendency to sorb metals from solution and to
leach metals from its stabilizers into solution. Tygon also sorbs/partitions organics from solution
into its matrix and can release organics from its plasticizers into solution. However, Parker (1992)
notes that rigid PVC, which is often used for well casings, shows little leaching of metals or organics
into solution and little sorption/partitioning of organics into its matrix.

Filter and Sampling Materials

A review was performed in order to evaluate the porous filters and sampling materials which
are commonly used in practice today, with references from this review compiled in Table 1.3 and
Table 1.4. The material most commonly used in DPT for filtration of both liquid and vapor samples
is stainless steel. Undoubtedly, this is due to its high abrasion resistance which is required during
push penetration of the cone penetrometer or sampling probe. While stainless steel has been shown
capable of adsorbing metals from or leaching metals into solution, depending on the geochemistry of
the strata being sampled, transfer of metal ions is strongly dependent on contact time; longer contact

17
 Table 1.3. Metallic Filter Elements for Direct Push Technology
Material and Application Reference
Stainless Steel 304/316; Argonne National Laboratory performed study of Lightner and Purdy
transfer lines made of stainless steel, nickel, aluminum, and Teflon® to quantify (1995)
interaction with samples in terms of outgassing and transfer loss of VOCs.
Temperature, flow rate and latent moisture content were varied; stainless steel,
type 304 and 316, yielded the most accurate results.
Stainless Steel; Batch water sampler; uses stainless steel wire filter screen and Auxt and Wright
all stainless steel construction for the body of the sampler. (1995)
Stainless Steel; Soil sampler; uses stainless steel 316 sample tubes with caps Auxt and Wright
and all stainless steel construction for the body of the sampler. (1995)
Stainless Steel; Envirocone uses stainless steel filter with thin, open ridges Piccoli and Benoit
when taking liquid and vapor samples in sand; used in highly abrasive soils. (1995); O'Neill et
Sample is pumped to the surface for analysis (Teflon transfer line). al. (1996)
Stainless Steel; Fine mesh stainless steel screen surrounded by perforated Konrad and
stainless steel screen for filtration of fluid used in the assessment of hydraulic Frechette (1995)
conductivity.
Stainless Steel; Hydropunch discrete water sampler. Stainless steel Strutynsky and
construction. Sainey (1992)
Stainless Steel; Monitoring well filter pack of rounded silica sand surrounded Strutynsky and
by two stainless steel wire wrapped screens. Developed to limit migration of Sainey (1992)
fines into water samples.
Stainless Steel; Filter rings on Envirocone®; used to sample both liquid and O'Neill et al.
vapor. (1996)
Stainless Steel; Filter material used for discrete water samplers like the BAT Robertson et al.
Enviroprobe and the Hydropunch (1996)
Stainless Steel; Filter elements used in piezocone penetration testing Campanella and
Robertson (1988)
Sintered Steel; Filter collars on Envirocone®; used to sample both liquid and O'Neill et al.
vapor. (1996)
Sintered Steel; Envirocone uses sintered steel filter when taking liquid and Piccoli and Benoit
vapor samples in clay; used only in soils with low abrasion. Sample is pumped (1995); O'Neill et
to the surface for analysis (Teflon transfer line). al. (1996)
Sintered Bronze; Filter element used in piezocone penetration testing Campanella and
Robertson (1988)
Carborundum; Filter element used in piezocone penetration testing Campanella and
Robertson (1988)
Aerolith-10; Filter element used in piezocone penetration testing Campanella and
Robertson (1988)
Splinted Metal Filters; Filter for vapor entry to soil gas sampling tube; Bratton et al.
tendency to clog in fine-grained soils. (1995)

18
 Table 1.4. Non-Metallic Filter Elements for Direct Push Technology
Material and Application Reference
®
Teflon ; Soil gas transfer line from cone penetrometer to gas sampling manifold Lightner and
at surface for VOC analysis. Cone push demonstration, April 1993, at DOE Purdy (1995)
Hanford Site in Washington.
Teflon®; Soil gas sampling tubing to bring soil gas from cone penetrometer to Bratton et al.
analysis equipment at surface. (1995)
Teflon®; Sampling and flushing lines in Envirocone®. O'Neill et al.
(1996)
Teflon®; Filter elements used in piezocone penetration testing Campanella and
Robertson
(1988)
PVC*; Direct push monitoring well placement; PVC used for casing and well Auxt and Wright
screen. (1995); Bratton
and Timian
(1995)
PVC; Direct push discrete water sampler; PVC filter screen used. Bratton and
Timian (1995)
PVC; Direct push monitoring well placement; PVC used for screen and casing. Bratton et al.
(1995)
HDPE**; Soil gas sampling tubing to bring soil gas from cone penetrometer to Bratton et al.
analysis equipment at surface. (1995)
Polypropylene; Filter elements used in piezocone penetration testing Campanella and
Robertson
(1988)
Plastic; Filter for vapor entry to soil gas sampling tube; tendency to clog in moist Bratton et al.
clays. (1995)
Ceramic; Cemented Quartz Sand; Stone; Filter elements used in piezocone Campanella and
penetration testing (1988) Robertson
(1988)
Note: *Polyvinylchloride
**High-density polyethylene

times lead to greater risk of contamination. In situations where contact time between the sampler
filter and the fluid being sampled is short, the risk of sample contamination is minimized.
Alternately, polymeric materials were most commonly used for the transfer of liquid and vapor from
the sampler to the ground surface. The flexibility of polymeric tubing makes it ideal for sample
transfer from depth to chemical analysis equipment located at the soil surface. As was the case with
metals, sorption/partitioning of organics into polymeric materials is a function of contact time, so
contact times between the sample and the polymeric sampling lines should be minimized in order to
preserve sample integrity.

Filters Used in Cone Penetration Testing

The filter materials used in cone penetration testing have been optimized in terms of
mechanical parameters, rather than in terms of chemical behavior. The hydraulic push of a cone into
19
the soil during piezocone penetration testing requires a filter that can withstand extreme physical
abrasion from the adjacent soil, and pore fluid pressures as high as 50 atmospheres. In addition,
piezocones measure pore pressures by filtering the fluid into a small cavity in the cone body which
houses a pore pressure transducer. In this configuration, where a pressure is measured rather than a
sample being taken, there is no concern about chemical contamination or fluid transfer, as is the case
in environmental sampling. The main purpose of the filter in the measurement of pore pressures is
the separation of fines from the fluid being measured, so the filter must efficiently separate solids
from the fluid while retaining a permeability which is high enough to not interfere with fluid flow
into and out of the cavity.
The materials used commonly as filter materials for the measurement of pore fluid pressures
in cone penetration testing include stainless steel, high density polypropylene, and ceramic. In
addition, sintered brass, aerolith-10, Carborundum, cemented quartz sand, stone, and Teflon® have
also been used (Campanella and Robertson, 1988). Smits (1982) investigated the properties of both
ceramic and sintered stainless steel filters for the measurement of dynamic pore pressures. He
reported that ceramic filters suffered from abrasion in quartz sands and also experienced pore
pressure noise in dense sands due to their low rigidity. Ceramic was also more brittle than other
materials used for filter elements and could crack during penetration in dense soil deposits. In terms
of sintered stainless steel, Smits (1982) reported that the filter surface was subject to smearing when
pushed through sands and could also clog when pushed through fine-grained soils. However,
Campanella and Robertson (1988) stated that many of the problems of abrasion and smearing have
been overcome through improvements in the material properties of both stainless steel and ceramic.
High-density plastics are commonly used because they are inexpensive and disposal, allowing
change of the filter between soundings.

Summary of Chemical Compatibility Issues During Sampling

As applications for cone penetrometry and direct push technology become more wide-
ranging, it is increasingly important to evaluate the filter and sample transfer line materials chosen
for use in a specific tool. While materials like stainless steel, ceramic, Teflon®, and polypropylene
have become standards in the industry, it is clear that both the mechanical and geochemical
properties of the site being tested must be taken into consideration. For example, Teflon® would be
the material of choice in environmental sampling for metal contamination at a site, while stainless
steel is superior for the sampling of organics. In cases where filter exchange is not possible, the
sampling contact time should be reduced to a minimum in order to limit the transfer of contaminants
into or out of the sample solution. Filter materials used during DPT or CPT penetration are also
constrained in terms of abrasion resistance, pore pressure resistance, and permeability
considerations. In summary, filters used in cone penetration testing must withstand physical abuse
without sacrificing permeability into the pore pressure transducer cavity, while filters used in geo-
environmental sampling must withstand physical abuse, must be permeable, and must be chemically
compatible with the site conditions.

Experiments and Modeling in Geoenvironmental Cone Penetration Testing

The work in this study examines two major issues in cone penetration testing, particularly for
improved site characterization in geoenvironmental engineering, through the introduction of a new
chemical sensor, and a more versatile approach to evaluating porewater dissipation tests in clays.
Specifically, the work examines the development and laboratory testing of a new class of chemical
detection sensors designed to measure the aqueous concentrations of aromatic hydrocarbons in-situ.
The device which was tested is a miniaturized integrated optoelectronics sensor developed to detect
20
the BTEX chemicals (benzene, toluene, ethylbenzene, and xylene) during cone penetration testing.
In addition, the response and interpretation of geoenvironmental data gathered from the cone pore
pressure sensor are examined. The study develops a methodology to interpret the dissipation of
piezocone pore pressure transducer measurements for the evaluation of the horizontal coefficient of
consolidation (ch) and permeability (kh) of a soil deposit, both of which are geoenvironmental
parameters important to the flow of water and transport of contaminants within the subsurface.
Chapter 2 provides a review of literature related to the dissipation behavior of the penetration
pore water pressures that are generated during cone soundings in fine-grained soils. Additionally,
current methods for the evaluation of the coefficient of consolidation and the permeability from
porewater pressure decay are reviewed.
Chapter 3 examines the behavior of a miniaturized version of an integrated opto-electronic
BTEX sensor tested in a silica sand soil/water matrix using a bench-top configuration. The sensor
was initially enclosed in a prototype module which was compatible with soil testing and was
evaluated using BTEX contaminants at a variety of concentrations. Related issues in the adaptation
of the sensor to full-scale field cone penetration testing and environmental site characterization are
discussed.
Chapter 4 focuses on the interpretation of pore pressure sensor data. It outlines the
development of a dissipation model which uses a hybrid cavity expansion and critical state soil
mechanics approach, combined with an analytical solution to the partial differential equation which
governs radial consolidation around a cone penetrometer, in order to evaluate the coefficient of
consolidation of normally-consolidated or heavily-overconsolidated clay soil deposits. The method
evaluates the monotonically decreasing porewater pressure decay with time in soft clays and silts, as
well as the dilatory dissipation behavior (an increase in pore pressure followed by a subsequent
decrease as a function of time) which has been observed in heavily-overconsolidated clays.
Chapter 5 integrates shear wave velocity measurements gathered with the seismic cone
penetrometer in order to make approximate evaluations of the in-situ void ratio of a clay deposit, and
develops a methodology that combines data from the seismic piezocone penetration test with
porewater pressure dissipation measurements to obtain estimates of the in-situ permeability. Here,
the shear wave velocity is used to evaluate the constrained modulus (D = 1/mv) and relates the
coefficient of consolidation to the permeability (k = chγw/D). The estimation is verified by
comparison with data collected at well-documented test sites.
It is hoped that the introduction of a new chemical-detection module and the improved
interpretation of porewater pressure dissipation data will further promote the utilization and
application of cone penetration testing in geoenvironmental practice.

21
 CHAPTER 2

LITERATURE REVIEW

Introduction
In addition to the direct measurement of properties, the interpretation of data gathered from
piezocone testing provides important information about the geoenvironmental properties of a soil
deposit. In particular, evaluation of the properties which control the movement of pore fluid through
a soil deposit can be performed through analysis of piezocone dissipation tests. This chapter
provides a review of the methodologies available for the evaluation of the coefficient of
consolidation and soil permeability from piezocone dissipation tests.

Evaluation of the Coefficient of Consolidation From Piezocone Dissipation Data

During a piezocone penetration test in fine-grained soils, significant pore pressures are
generated in excess of the hydrostatic pore pressure. When the piezocone test is halted during
penetration in clays, these pressures dissipate, and the measured variation of pressure with time
relates to the coefficient of consolidation and the permeability of the deposit.
The coefficient of consolidation is an important parameter in both geotechnical and
geoenvironmental engineering because it is a measure of the time-rate of response and flow
characteristics of soils. In environmental applications, the coefficient of consolidation controls the
dissipation of excess pore pressures generated in clay landfill liners when they are subjected to
repetitive overburden loading through the application of waste materials, and is also directly related
to the permeability and flow of fluid through the pore space of a soil. Consequently, it influences
the advective transport of groundwater and contaminating chemical species through the subsurface.
In traditional geotechnical applications, the coefficient of consolidation is an important parameter
for estimating rates of settlement in foundation loading (Lutenegger et al. 1988), embankment
construction (Hansbo et al. 1981), and for the design of wick drains to dewater a site (Robertson et
al. 1988a). The permeability of a soil is paramount in groundwater flow applications and mass
transport studies in geoenvironmental engineering. Because of this importance, this section reviews
the relevant literature regarding existing methodologies for the evaluation of the coefficient of
consolidation from piezocone dissipation tests.

Generation of Excess Porewater Pressures

Traditional analyses of penetration problems have been divided into the two general
categories of shallow and deep penetration problems. Shallow penetration differs from deep
penetration because the displaced volume of soil has more freedom of movement and is able to move
out and up to the soil surface, similar to a footing bearing capacity failure. However, in deep
penetration problems like cone testing, constraints on the displaced volume force deformation of the
surrounding soil in mainly lateral directions (Houlsby and Teh, 1988). Researchers have noted that
surface heave occurs in the initial driving of piles, but ceases to occur at large depths of pile
penetration below the surface (Randolph and Wroth, 1979a). Consequently, at great depths, the soil
surrounding a piezocone or a driven pile during deep penetration problems is forced to deform
within a constrained volume surrounding the penetrometer. While the complete mechanism of
failure surrounding a penetrating piezocone is not fully understood, it is generally accepted that cone
soundings produce failure under the cone tip which is similar to the expansion of a spherical cavity
in an elasto-plastic medium, and failure around the body of the cone which is similar to the
expansion of a cylindrical cavity in an elasto-plastic medium (Torstensson, 1975; 1977; Battaglio et
al., 1981).
22
 Cone penetration also induces changes in the pore fluid conditions surrounding the
penetrometer, especially in clayey soils which have a characteristic low hydraulic conductivity. For
soundings which occur under undrained conditions, the change in pore pressure is a combination of
changes in the mean octahedral normal stress, ∆σoct, and octahedral shear stress, ∆τoct. The change
in pore pressure due to the change in octahedral normal stress results from the displacement of soil
and fluid by the penetrating cone, while the change in pore pressure due to the change in octahedral
shear stress is due to the shear deformation of the soil adjacent to the cone body. Initial attempts to
estimate the magnitude of pore pressure changes from the individual components were based on the
approximate expression introduced by Henkel (1959):

∆u = β∆σ oct + α∆τ oct Equation 2.1

where ∆u = change in pore pressure, ∆σoct = change in octahedral normal stress, ∆τoct = change in
octahedral shear stress, and α and β are Henkel’s empirical pore pressure parameters, with β
typically assumed equal to 1.0 in saturated clays (Roy et al., 1981). Under the cone tip, the largest
change in magnitude of pore water pressure is due to the changes in mean normal stress and the
relative changes in shear stress are small (<20%) in comparison (Baligh, 1986). However, along the
shaft of the cone body, the shear stresses become a significant portion of the excess pore pressure
change because the large octahedral stresses acting on the cone tip undergo stress relief at this
location. Evidence suggests that the stress regime directly behind the cone tip is more significantly
affected by the shear stress than the area located directly beneath the cone tip (Wroth, 1984;
Campanella et al., 1986a).
The total magnitude of pore pressure measured in the piezocone test includes not only the
excess pore pressure generated due to changes in the mean octahedral normal stress and in the
octahedral shear stress, but also the hydrostatic pore pressure in the soil profile, and can be
represented as follows:

um = uo + ∆uoct + ∆u shear Equation 2.2

where um = the magnitude of pore pressure measured by the cone penetration test, uo = hydrostatic
pore pressure, ∆uoct = excess pore pressure due to changes in the mean octahedral stress, and ∆ushear
= excess pore pressure due to changes in the octahedral shear stress (Figure 2.1). If the prevailing
groundwater conditions are known, the hydrostatic pore pressure term can be calculated as uo =
γwhw, where γw = the unit weight of water and hw = the height to the water table above the reference
depth z. This yields the following equation for excess pore pressure:

∆um = ∆uoct + ∆u shear Equation 2.3

where ∆um = um - uo = the pore pressure generated in excess of hydrostatic pressure during a cone
sounding. It is important to note that the penetration pore pressure is a total magnitude measurement
and the contributions of the individual components (∆uoct or ∆ushear) cannot be measured separately.

23
 2ro Stress
rp=ro(Ir)0.33 Path
q ∆us

= + +
Hydrostatic
uo = hwγw
Spherical Cavity Expansion P'

um uo ∆uoct ∆ushear
Figure 2.1. Components of Measured Pore Water Pressure During Cone Penetration
Note: q = (σ1-σ3) and p' = 1/3(σ1'+σ2'+σ3')

Pore Pressure Dissipation Behavior

When cone penetration is halted, the induced excess pore pressure will dissipate to the
hydrostatic value in time. A typical pore pressure dissipation record shows magnitudes of porewater
pressure monotonically decreasing with time from the initial reading, similar to the behavior seen
during one-dimensional consolidation tests. This behavior is observed in normally- to lightly-
overconsolidated clays (note: OCR = overconsolidation ratio = preconsolidation stress / effective
vertical overburden stress = σp'/σvo'), regardless of the location of the pore pressure filter element on
the penetrometer (Figure 2.2). The same behavior is observed in heavily-overconsolidated clays
when the pore pressure filter element is located on the cone tip or face position (u1). However, when
measuring pore pressure dissipation in heavily-overconsolidated clays with the filter element located
behind the cone shoulder (u2), the pore pressures often exhibit dilatory behavior, by first increasing
in magnitude, and then subsequently decreasing with time to hydrostatic values (Figure 2.3).
Dissipation curves from piezocone testing can be modeled to predict the coefficient of
consolidation of the surrounding medium. The coefficient of consolidation incorporates the material
properties which influence the rate of consolidation in a soil deposit, where the end of primary
consolidation is defined as the dissipation of excess pore pressures to zero, or the decay of measured
pore pressures to the hydrostatic value. Consequently, the coefficient of consolidation can be used
as an indicator of soil behavior.

24
 600
Bothkennar, Scotland
2
10 cm Fugro Cone
Depth = 12 m

Pore Pressure (kPa)

(Jacobs and Coutts, 1992) u2
u1
400 u1

u2

200
uo

0
1 10 100 1000 10000
Time (sec)

Figure 2.2. Dissipation in a Lightly-Overconsolidated Clay

2500

2000
Pore Pressure (kPa)

u1 u2
Baton Rouge, LA
u1
1500 15 cm2 Fugro Cone
Depth = 37.4 m
(Chen and Mayne, 1994)
u2
1000

500 uo

0
1 10 100 1000 10000
Time (sec)

Figure 2.3. Dissipation in a Heavily-Overconsolidated Clay

25
 An important difference to note in the comparison of piezocone-determined and laboratory-
measured values of the coefficient of consolidation is that the process of cone testing significantly
remolds the soil surrounding the cone. Consequently, during a piezocone dissipation test,
consolidation takes place in a partially remolded soil, which differs from the laboratory case where
consolidation takes place in a sampled, but not remolded soil specimen (Gillespie and Campanella,
1981). Additionally, the one-dimensional consolidation test, which is used to provide reference
measurements of the coefficient of consolidation, restricts the deformation of the soil skeleton and
the movement of pore fluid to the vertical direction, while experimental evidence in pile driving,
wick drains, and cone testing shows that the movement of pore fluid corresponding to drainage tends
to flow out radially (Randolph and Wroth, 1979a), rather than the vertical movement mandated in
the Terzaghi consolidation test.
If the soil fabric is arranged isotropically, the horizontal and vertical values of the coefficient
of consolidation will be the same; however, if the soil fabric is anisotropic, as is most commonly the
case for platy clay particles, the two values will differ somewhat. In natural clays, however, the
difference between horizontal permeability (kh) and vertical permeability (kv) is typically small
(Tavenas et al., 1983b) as shown by Figure 2.4. Exceptions would be highly-stratified deposits and
varved silts and clays where kh > kv. Another important difference between the methods of
evaluation is that the one-dimensional consolidation test utilizes a very small sample of soil, while
the cone test is performed in-situ in the deposit and is affected by a large volume of soil surrounding
the cone. Finally, the one-dimensional consolidation test begins with a homogeneous initial
distribution of excess pore pressure throughout the soil specimen, while the pore pressures resulting
from a piezocone test have a radial gradient.

Figure 2.4. Relationship Between Horizontal and Vertical Permeability Measured on Natural
Marine Clays in Constant Head Laboratory Tests
(Leroueil and Jamiolkowski, 1991)

26
 Radial versus Vertical Drainage
The drainage characteristics of a soil are an important component of the consolidation
behavior which occurs during a pause in cone penetration. Clearly, in an unbounded soil medium,
drainage will occur in all directions, away from the gradients induced by the piezocone or driven
pile. Consolidation results from drainage in both the horizontal and vertical directions, and the
coefficient of consolidation can vary significantly depending on the direction of flow. Previous
studies of instrumented piles driven into soft clays showed that the major pore pressure gradients
developed along the pile shaft in the radial direction rather than in the vertical direction (Bjerrum
and Johannessen, 1961; Koizumi and Ito, 1967; Lo and Stermac, 1965; Randolph and Wroth,
1979a). Additionally, the carefully-instrumented St. Alban site has shown an increase in pore
pressure at large radial distances away from a driven pile, which has been taken as further evidence
of radial consolidation (Roy et al., 1981).
Levadoux and Baligh (1986) noted that results from the strain path method show that ch
controls the drainage behavior surrounding a cone penetrometer in all but the most extreme cases.
Houlsby and Teh (1988) performed a sensitivity analysis using the strain path method in
combination with a large strain finite element model and found that for ch > cv, the consolidation
behavior is dominated by ch, or radial drainage. Gillespie and Campanella (1981) investigated
experimentally the influence of filter location on the consolidation characteristics surrounding the
cone penetrometer. Their results showed that the dissipation around a cone with the filter located in
the Type 2 position, shoulder location, was only slightly faster than dissipation when the filter was
located 12.5, 25, or 38 cm behind the tip. This indicated that the horizontal component of drainage
contributed strongly to consolidation around the penetrometer because the Type 2 position would be
more strongly influenced by spherical drainage than would the other positions. Other authors have
also noted that radial drainage is the controlling mechanism when the pore pressure filter was
located on the shoulder (Kabir and Lutenegger, 1990; Teh and Houlsby, 1989). Consequently, the
majority of solutions for consolidation around a penetrometer or a pile assume radial drainage
dominates the dissipation behavior.

Type 1 versus Type 2 Dissipation

While dissipation curves can be recorded for pore pressure transducers located at both the
Type 1 and Type 2 positions, the recorded Type 1 profiles are not always reliable due to stress relief
on the tip during a halt in the cone sounding. A dissipation test is most commonly performed when
the sounding is paused in order to add another rod for pushing. Typically, when the sounding is
stopped, the drill rods lift slightly as the pressure from the drill rig is released. In the case of the
Type 1 pore pressure transducer, the induced pressures show a discontinuous step (Figure 2.5),
leading to inaccurate values for the induced pore pressure and dissipation profile. It is possible to
procedurally lock the rods in place; however, a standard method of performing this for dissipation
testing has not been established. The Type 2 readings are not affected by the stress release (Figure
2.5). In this study, the focus is on the pore pressure profiles obtained with Type 2 pore pressure
transducers, which are insensitive to stress relief due to rod uplift, and because the shoulder pore
pressure reading is necessary for the cone tip resistance correction.

27
 Figure 2.5. Load Release Behavior
(Campanella and Robertson, 1988)

Existing Soil Models For Pore Pressure Dissipation

Excess pore pressures produced by the insertion of a cone penetrometer, or any penetrating
probe, will in time reach equilibrium conditions, corresponding to hydrostatic pore pressure (uo)
when ∆u = 0 (Figure 2.6). Many attempts have been made to model the change in pore pressure
with time in order to obtain an in-situ estimate of the coefficient of consolidation. While the models
have varied in their degree of complexity, ranging from the simple cavity expansion theory to the
relatively complex strain path method, the predictive capabilities of the models are not significantly
different. In all cases, the soil conditions which occur during cone penetration are extremely
complicated; too complex to be modeled without significant simplifying assumptions.

28
 Pore Pressure (kPa)
0 100 200 300 400 500
0 G.W.T.

Penetration
5
Pore Pressure, u2
Depth (m)

}Layer
10 t=0
Clay

t = t100
15 ∆u
uo Hutchinson Island, GA
(Burns and Mayne, 1995b)
20

Figure 2.6. Dissipation of Excess Pore Pressures

Dissipation in Normally- and Lightly-Overconsolidated Clays

The first attempt to model the dissipation of excess pore pressures in fine-grained soils was
made by Torstensson (1977) using cavity expansion theory (Hill, 1950; Vesic, 1972). (A more
detailed review of cavity expansion theory is provided in Chapter 4.) The problem was represented
as the expansion of either a cylindrical or spherical cavity in an elastic-perfectly plastic medium. By
assuming that the soil failed in undrained shear and neglecting the volume change in the plastic
zone, the radius of the plastic zone was calculated as follows:

1
E G 1
rp = r0 [ ] 3 = ro [ ] 3 Spherical cavity Equation 2.4
2(1 + ν ) su su

1
E G 21
rp = r0 [ ] = ro [ ]
2
Cylindrical cavity Equation 2.5
2(1 + ν ) su su

where rp = radius of the plastic zone, ro = radius of the pore pressure probe, su = undrained shear
strength, E = Young’s modulus, ν = Poisson’s ratio, and G = shear modulus. Assuming an isotropic
and homogeneous soil, an impermeable cone/soil interface boundary, and no excess pore pressure
change outside the plastic zone, the maximum excess pore pressure was calculated as follows:

4 E 4 G
∆u = su ln[ ] = su ln Spherical cavity Equation 2.6
3 2 su (1 + ν ) 3 su

29
 E G
∆u = su ln[ ] = su ln Cylindrical cavity Equation 2.7
2 su (1 + ν ) su

In Torstensson's representation, the magnitude of pore pressures generated by insertion of the

piezoprobe into the clay deposit was calculated as the value predicted by cavity expansion theory.
This magnitude was then used as the initial condition for the solution of the consolidation equation.
Boundary conditions included: 1) the cone is an impermeable boundary and 2) there is no increase in
pore pressure outside the plasticized zone.

∂u 2 ∂u ∂ 2u
= ci + ci 2 Spherical Cavity Equation 2.8
∂t r ∂r ∂r

∂u 1 ∂u ∂ 2u
= ch + ch 2 Cylindrical Cavity Equation 2.9
∂t r ∂r ∂r

where ci = coefficient for isotropic drainage and ch = coefficient for radial drainage. Torstensson
solved the Terzaghi consolidation equation using the method of finite differences. The solution was
a one-dimensional, linear uncoupled analysis which predicted ch by matching field dissipation
curves with predicted dissipation curves. Linear analysis assumed that the value of the coefficient of
consolidation did not change during consolidation, and the uncoupled model assumed that during
consolidation the total stresses remained constant and the strains that occurred resulted from the
change in pore water pressure only. The value of ch was determined by matching the predicted
dissipation with the measured dissipation at 50% consolidation, or t50 which is the time at which
50% of the excess pore pressures have dissipated. Torstensson stated that his method overestimated
field values by a factor of approximately 2 for the normally-consolidated clay dissipation records
which were used to verify the model; the number of sites modeled was not reported. Sample
dissipation curves presented by Torstensson are shown in Figure 2.7; note that τf = the maximum
undrained shear strength.

30
 Figure 2.7. Spherical Cavity Dissipation Prediction
(Torstensson, 1977)

Baligh and Levadoux (1980; 1986), Baligh (1985), and Levadoux and Baligh (1986) used the
strain path method to predict the initial excess pore pressure distribution generated in normally- to
lightly-overconsolidated insensitive clays by cone soundings, in combination with a two-
dimensional linear uncoupled consolidation analysis. The strain path method was developed to
model cone penetration because cavity expansion did not consider the strain paths followed by the
soil elements surrounding the cone (Baligh, 1986). In this method, the cone was taken as a static
penetrometer with soil flowing around it like a viscous fluid. Excess pore pressures generated by
cone penetration were calculated by predicting soil velocities and strain rates using potential theory
(assuming an ideal incompressible fluid flow and neglecting shearing resistance), integrating the
strain rates along the stream lines to determine the strain history of the soil elements, and computing
the deviatoric and shear-induced pore pressures using a total stress model (Levadoux and Baligh,
1986). While the strain path method was an approximate solution, it more accurately modeled the
strain history of the soil elements surrounding the penetrometer than did cavity expansion; however,
in the solution, an inequilibrium in stresses developed near the cone tip and the cone shoulder,
making it necessary to correct the results using finite element methods (Houlsby and Teh, 1988).
The two-dimensional uncoupled consolidation solution was performed using a finite element
program. Sample dissipation curves presented by Levadoux and Baligh are shown in Figure 2.8.
The model results were verified using dissipation test results in resedimented Boston Blue Clay.
Comparisons were performed for dissipation surrounding cones with 60o and 18o tips, with pore
pressure filters on the face and shoulder of the cone, and dissipation of the normalized excess pore
pressure was determined to be a function of the time factor, T ≅ ch(t/r2), where r = cone radius.
Levadoux and Baligh (1986) also performed a finite element study which examined the
effects of using a coupled solution to the consolidation equation rather than the uncoupled solution
used in the one-dimensional consolidation tests. This solution accounted for the coupling between
pore pressures and total stresses during consolidation. Results showed that the effect was significant
for measurements made on the tip of an 18o piezocone, but as the location for measuring pore
pressure was moved up the cone body away from the tip, the effect of coupling became negligible.
31
 Figure 2.8. Strain Path Method Dissipation Prediction
(Baligh and Levadoux, 1986)

While the current study is concerned with the excess pore pressure behavior surrounding
piezocones, similar behavior has been observed in clay soils surrounding driven piles. The act of
driving a pile into soil creates much the same type of failure as has been seen in the cone penetration
test. Additionally, when driving piles in clays, excess pore pressures are generated which dissipate
as the pore water migrates away from the pile. Randolph and Wroth (1979b) developed an
analytical solution to the consolidation equation in order to model the effect of pore pressure
dissipation behavior on the pile driving process. The solution was based on cavity expansion theory
(elastic-plastic soil) and assumed that consolidation was governed by radial drainage away from the
pile. The estimated decay was compared with data from two different pile sites from the literature
and gave agreement within 10% of the measured values.
Gillespie and Campanella (1981) compared existing models from the literature with field
data obtained at two different sites in British Columbia. Their comparisons showed that the
cylindrical solution by Torstensson (1977), the analytical solution by Randolph and Wroth (1979b),
and the two dimensional solution by Baligh and Levadoux (1980) provided reasonable predictions of
ch in lightly-overconsolidated clays when compared with laboratory values. However, the spherical
solution by Torstensson was found to somewhat underpredict the value for the two sites examined
which is in disagreement with the finding of Torstensson (1977).
Battaglio et al. (1981) modeled dissipation behavior around piezocone penetrometers by
combining cavity expansion theory with Skempton's and Henkel's empirical pore pressure equations.
Initial magnitudes of pore pressure were predicted using the following formulations:

∆u = su [ln( I r ) + 0.577(3 A f − 1)] Cylindrical Cavity Equation 2.10

32
 4
∆u = su [ ln( I r ) + 0.667(3 A f − 1)] Spherical Cavity Equation 2.11
3

where Af = Skempton's pore pressure parameter at failure and Ir = the rigidity index = G/su. The
linear uncoupled consolidation equation was then solved using the method of finite differences until
the predicted curve reached a minimum deviation from the measured field curve. The model was
validated with data from two soft clays (Porto Tolle and Trieste) in Italy.
Jones and Van Zyl (1981) presented an empirical method for the evaluation of the coefficient
of consolidation. The value of t50 obtained from dissipation tests was plotted versus the value of ch
measured in laboratory oedometer testing. Linear regression of the data yielded the following
empirical relationship:

50
cv = Equation 2.12
t 50

where cv is in (m2/yr), and t50 is in (min). The authors noted that more data were needed to improve
the relationship shown in Figure 2.9.
Senneset et al. (1982) assessed the coefficient of consolidation from piezocone dissipation
tests using a logarithmic distribution of the initial excess pore pressure which was dissipated using a
solution to the one-dimensional consolidation equation. The coefficient of consolidation was
defined as:

∆u
ch = λ c ro 2 Equation 2.13
∆ui

where λc = pore pressure dissipation rate factor, ro = filter radius, ∆u = du/dt = measured rate of
dissipation, and ∆ui = initial excess pore pressure minus hydrostatic excess pore pressure = ui - uo.
The solution was presented as normalized pore pressure curves which were a function of the rate
factor for soft, medium, and stiff clays (Figure 2.10). The authors stated that the values estimated
using this method gave best agreement when measured at stresses greater than the preconsolidation
stress.

33
 Figure 2.9. cv - t50 Correlation
(Jones and Van Zyl, 1981)

Figure 2.10. Pore Pressure Dissipation Rate Curves

(Senneset et al., 1982)

Tavenas et al. (1982) drew on the relationship between ch, k, and the modulus of
deformability by developing an empirical correlation in order to evaluate t50 in Champlain Sea
Clays. The proposed method measured the permeability from permeability tests and the
preconsolidation pressure (σp') from oedometer tests. Then t50 was plotted versus kσp'/γw in order to
generate the correlation shown in Figure 2.11.

34
 Figure 2.11. t50 Correlation in Champlain Sea Clays
(Tavenas et al., 1982)

Tumay et al. (1982) evaluated the coefficient of consolidation from dissipation tests using a
combination of theoretical values and experimental measurements. The authors used the strain path
method (Levadoux and Baligh, 1980) together with experimental measurements obtained by Roy et
al. (1979) to predict the initial pore pressure distribution generated due to cone penetration.
Dissipation of these pore pressures was then performed using linear uncoupled finite element
analysis. Evaluation of ch for the soil deposit was performed using values obtained for t90, rather
than the more commonly used t50. This was because the disturbance effects, anisotropy effects, and
filter element location effects were less significant in the later consolidation phases. The influence
of anisotropy was also taken into account. The method required estimation of two factors: first T90,
the point to which the dissipation curves converge, and second, T90', the point to which straight line
portions of the dissipation curves converge as shown in Figure 2.12. The authors stated that the ratio
of the logarithms of these parameters equals a constant, A:

log T90
=A Equation 2.14
log T90 '

The time for 90% consolidation was estimated by:

t 90 = 10 exp A log t 90 ' Equation 2.15

35
where t90 = time for 90% consolidation, and t90' = time for 90% dissipation of straight line portion.
Finally, the coefficient of consolidation was calculated by:

T90 2
ch = r Equation 2.16
t 90

where r = probe radius. Values evaluated by this method were within ten to twenty times the values
measured in laboratory consolidation testing.

Figure 2.12. Pore Pressure Dissipation With Anisotropy

(Tumay et al., 1982)

Tumay et al. (1985) and Acar and Tumay (1986) performed an assessment of the steady-state
penetration of a cone penetrometer. The soil was treated as a non-compressible, inviscid fluid, and
the solution analytically modeled the strain rate field, and the strain rates, surrounding the cone
during penetration in very soft fine-grained soils. When combined with the time history of strains,
the method could be used to predict the pore pressures generated by cone penetration, although that
was not done in the study presented.
Gupta (1983) and Gupta and Davidson (1986) predicted the initial pore pressure developed
around a cone penetrometer by modeling a series of spherical cavity expansions which propagated as
the cone was advanced into the soil. Excess pore pressure was calculated using Vesic's logarithmic
distribution (Vesic, 1972) which was given as follows:

36
 rplastic
(u A − uoA )[0.943α f + 4 ln( )
rB
∆u B = Equation 2.17
rplastic
[0.943α f + 4 ln( )]
rA

where ∆uB was the excess pore pressure at any point B within the spherical cavity, uA = the pore
pressure measured at point A, uoA = the hydrostatic pore pressure at point A, rplastic = the radius of the
plastic zone = (G/su)0.333, αf = Henkel's pore pressure parameter at failure (= 0.707[3Af - 1]), and Af =
Skempton's pore pressure parameter at failure. Additionally, they accounted for the fact that
dissipation began to occur before the cone reached the location where the dissipation test was
performed. Dissipation was modeled using an alternating direction implicit finite difference solution
to the linear uncoupled consolidation equation with both radial and vertical drainage components.
The initial variables for the prediction, G/su and Af, were determined from field and lab tests, and the
dissipation curves were plotted as a function of the time factor, T50 = cht50/R2, where t50 = the time to
achieve 50% consolidation and R = the expanded cavity radius. The model was calibrated using
data from three sites and their predicted results were within approximately two to five times the
laboratory results. Sample dissipation curves presented by Gupta and Davidson are shown in Figure
2.13.

Figure 2.13. Successive Spherical Cavity Formation Dissipation Prediction

(Gupta and Davidson, 1986)

Kabir and Lutenegger (1987; 1990) performed a comparison of oedometer-measured values

of the horizontal and vertical coefficient of consolidation with piezocone predicted values using
solutions proposed by Torstensson (1977), Baligh and Levadoux (1986), and Gupta and Davidson
(1986) at two medium stiff clays sites: a soft Leda clay deposit in New York, and a soft alluvial
deposit in Nebraska. Their results showed that the two former solutions agreed well with the
laboratory values at low percentages of consolidation only, the third method gave agreement in both
cases.
37
 Houlsby and Teh (1988) modeled the dissipation of excess pore pressures around a
piezocone using the strain path method combined with a large strain finite element analysis. Their
analysis modeled the clay deposit as an incompressible elastic-perfectly plastic material and used
uncoupled consolidation theory to model pore pressure dissipation around the cone. The finite
element model was performed in order to account for the lack of stress equilibrium generated in the
strain path method; the solution from the strain path method was taken as the initial state of stress in
the model, the cone was held static and forces were applied incrementally in order to balance the
forces, and then the cone was penetrated using a large strain finite element simulation.
Results from the method were combined with Henkel's estimate of pore pressure (Equation
2.1) in order to estimate the magnitudes of pore pressure during cone penetration. The linear
uncoupled consolidation equation was then solved using an alternating direction implicit finite
difference method. They found that the shear stress contribution was minimal to the overall
magnitude of pore pressure surrounding the cone due to the fact that overconsolidation was not
taken into account. Results showed that the consolidation rate predicted at the cone shoulder was
similar to the consolidation rate predicted by cylindrical cavity expansion theory. However, at the
tip, the predicted rate was faster than cylindrical theory and along the shaft, the predicted rate was
slower than cylindrical theory due to differences in prediction of the mean stress gradients along the
length of the cone (Houlsby and Teh, 1988). Because the time factor is a function of rigidity index,
the authors chose to present the dissipation of pore pressure as a function of a modified time factor,
T*:

ch t
T* = 2
Equation 2.18
r Ir

where r = the cone radius.

Teh and Houlsby (1991) performed a combined strain path finite element method analysis of
pore pressure dissipation based on dissipation data in sensitive marine clay in Quebec reported by
Roy et al. (1981). The predicted rate of dissipation agreed quite closely with the field values for
both the tip and shaft locations. The method closely predicted the magnitude of pore pressure
developed along the shaft of the penetrometer, but underpredicted the magnitude of pore pressure
acting at the cone tip by approximately 20%.
Teh and Houlsby (1991) also analyzed cone penetration using the strain path method alone.
This analysis showed that the strain path method predictions of soil straining at large radial distances
away from the cone tip were qualitatively similar to those predicted by cylindrical cavity expansion;
however, neither method could fully represent the complex conditions at the tip of the cone. Sample
dissipation curves presented by Teh and Houlsby are shown in Figure 2.14 for a rigidity index of
equal to 100. These curves are plotted as a function of the dimensionless Time Factor, T = cht/r2,
where r = cone radius. Dissipation curves specific to a Type 2 cone are shown in Figure 2.15 where
the results are plotted against the dimensionless Time Factor, T*, where T* = T(Ir)-0.5. Incorporating
the influence of the rigidity index unified the dissipation curves for values of Ir ranging from 25 to
500 into essentially a single relationship. The proposed solution was used to evaluate the coefficient
of consolidation as a function of the time factor, the filter element location, the time needed to reach
the given level of dissipation, the cone radius, and the rigidity index of the soil.
The solution proposed by Teh and Houlsby (1991) was evaluated by Robertson et al. (1992)
through a comparison of the estimated values of coefficient of consolidation with lab measured
values at many sites throughout the world. Additionally, Schnaid et al. (1997) used the method to
38
evaluate the coefficient of consolidation at two sites in South America and in calibration chamber
results obtained at the University of Oxford. Both studies found that the most reliable predictions
resulted from pore pressure data gathered at the cone shoulder.

T
Figure 2.14. Dissipation Prediction at Multiple Locations
(Teh and Houlsby, 1988)

Figure 2.15. Dissipation Prediction for Type 2 Cone

(Teh and Houlsby, 1991)

39
 Dislocation theory has also been proposed for evaluating piezocone dissipation data
(Elsworth, 1990; 1993). In this method, cone penetration was modeled as a moving volumetric
dislocation. The advancement of the cone was treated as a point dislocation of equivalent volume
which generated accompanying pore pressure increase and the soil was modeled as a linear elastic
material. In dislocation theory, the rate of pore pressure build up at the tip was a function of the
penetration rate of the cone, and the coefficient of consolidation and soil permeability were
evaluated from the predicted pore pressure buildup and subsequent dissipation. However, the theory
was not able to predict the coefficient of consolidation for pore pressures measured on the shaft
because it predicted that the dissipation along the shaft was controlled, not by the material
parameters surrounding the cone, but by the cone penetration rate and the distance the filter was
located up the shaft from the tip. The solution was verified at two normally- to lightly-
overconsolidated clay sites and gave general agreement within a factor of four for the coefficient of
consolidation. The method overestimated the in-situ permeability by approximately two orders of
magnitude.

Dissipation in Heavily-Overconsolidated Clays

If the porewater filter elements are properly saturated, dissipation curves in normally-
consolidated or lightly-overconsolidated clays always exhibit a monotonic dissipation behavior: that
is, magnitude always decreasing with time, when the pore pressure filter is located in either the Type
1 or Type 2 location. Additionally, heavily-overconsolidated clays exhibit standard monotonic
dissipation curves when the filter is in the Type 1 location. Sully and Campanella (1994) designate
these curves as Type I curves (Figure 2.16). However, in a heavily-overconsolidated clay when the
filter is in the Type 2 location, the dissipation curves can be dilatory, showing an initial increase in
magnitude followed by a subsequent decrease in magnitude with time (Davidson, 1985; Lutenegger
and Kabir, 1988; Sully and Campanella, 1994; Campanella et al., 1986a; Chen and Mayne, 1994).
Sully and Campanella (1994) designate these curves as Type II (initially above hydrostatic,
increases, then decreases to hydrostatic); Type III (initially below hydrostatic, increases above
hydrostatic, then decreases to hydrostatic); or Type IV (initially below hydrostatic, increases to
hydrostatic) as illustrated by Figure 2.17. Assuming that the pore pressure filter was properly
saturated before the test was begun, the dilatory dissipation behavior appears to be characteristic of
dissipation behavior in heavily overconsolidated and fissured clays. Previous solutions based on
cavity expansion, strain path, and finite element methods only provide solutions for the monotonic
decay, not for the observed dilatory behavior.
Davidson (1985) reported the results of nine piezocone soundings performed in a heavily
overconsolidated clay in Florida. All of the profiles showed an increase in pore pressure, followed
by a decrease with time. Davidson attributed the phenomenon to the fact that the large normal
stresses which were developed during cone penetration were relieved behind the tip of the cone
where large shear stresses also developed. These large shear stresses forced the excess pore pressure
negative during penetration; however, when penetration was stopped, the negative values were
consumed by the inflow from the higher pressure zone developed at the tip of the cone.

40
 Figure 2.16. Pore Pressure Dissipation in Normally Consolidated Clays
(Sully and Campanella, 1994)

u2

Figure 2.17. Pore Pressure Dissipation for Shoulder Piezocone Elements in Heavily
Overconsolidated Clays (Sully and Campanella, 1994)

Sully and Campanella (1994) reported the results of dissipation tests at one stiff
overconsolidated clay site where non-standard dissipation behavior was observed. The authors
stated that the dilatory behavior was the result of pore pressure redistribution resulting from the large
gradients induced in overconsolidated soils. For the dissipation curves in which pore pressures first
increased and then decreased, they suggested an empirical offset procedure by taking the maximum
pore pressure value reached during the dissipation test as the peak value, adjusting this value to time
41
zero, and adjusting all other time values on this basis. Subsequent evaluation of ch for the remaining
monotonic decay was performed using the method proposed by Teh (1987).
Similar dilatory dissipation behavior has been observed at clays sites during the driving of
piles in clays. Coop and Wroth (1989) document pore pressures which increase and then decrease
after the driving of cylindrical steel piles in the heavily overconsolidated Gault clay deposit at
Madingley, England. The site was instrumented with a series of piezometers which recorded
negative pore pressures in addition to the overconsolidated dissipation behavior. Likewise, the
dilatory behavior during pile driving was observed by Lehane and Jardine (1994) in the stiff glacial
clay deposit at Cowden, England.

Chronological Development of Pore Pressure Dissipation Modeling

A listing of the historical developments in pore pressure dissipation modeling to date is given
in Table 2.1. In summary, numerous methods exist to evaluate the coefficient of consolidation from
piezocone dissipation data. All prior methods only address dissipation tests exhibiting the
monotonic decrease of porewater pressure with time. For the most part, the solutions can be divided
into two categories: those which use cavity expansion to predict the initial pore pressure distribution
and those which use the strain path method to predict the initial pore pressure distribution. All
solutions then dissipate the excess pore pressures using consolidation theory, either in one or two
dimensions. The strain path method was developed in order to accurately model the strain paths
which the soil elements surrounding the cone penetrometer follow; however, the method incurred an
inequilibrium of stresses at the cone tip which must be corrected. Alternately, cavity expansion
models the advance of a cone penetrometer as the expansion of either a sphere or cylinder from zero
radius to the radius of the probe. Neither the strain path method nor cavity expansion theory are
capable of completely modeling the soil conditions during cone penetration, and in reality, the two
methods yield similar results for the prediction of the coefficient of consolidation. The solution
proposed by Houlsby and Teh is the most commonly used, most likely due to the ease of use of the
normalized chart solution presented for the evaluation of ch.
The method of Sully and Campanella (1994) addressed the dilatory pore pressure profiles by
suggesting an empirical offset procedure for the evaluation of ch. The proposed method is not
supported on a theoretical basis, and has not been extensively tested in order to verify its
applicability.

Summary
This section provided a review of the current methodologies available to evaluate the
coefficient of consolidation from piezocone dissipation tests. Dissipation data recorded surrounding
driven piles in clay were also included in the review because the soil behavior in pile tests is quite
similar to that seen during the cone penetration test. While the available models differ significantly
in their level of complexity, there are no models available which incorporate all aspects of soil
behavior during a cone sounding due to the complex nature of penetration.

42
 Table 2.1. Historical Development of Piezocone Dissipation Modeling*
Reference Cavity Soil Model Initial Pore Pressure, Consoli- Comments
Type ∆ui dation
Soderburg (1962) Cylindrical, Elasto-plastic ∆u/∆ui = R/r 1-D Consolidation surrounding
radius R driven piles;
Finite Difference
Torstensson (1975, Cylindrical Elasto-plastic ∆ui = 2suln(rp/r) 1-D No shear stresses;
1977) Spherical ∆ui = 4suln(rp/r) Finite Difference
Randolph and Wroth Cylindrical Elasto-plastic ∆ui = 2suln(rp/r) 1-D Consolidation surrounding
(1979b) driven piles;
Analytical
Baligh and Levadoux Piezocone Non-linear From strain path method; 2-D
(1980); Levadoux Model Total stresss soil model
and Baligh (1986)
Battaglio et al. Cylindrical Elasto-plastic ∆ui = 2suln(rp/r) 1-D Shear by empirical method;
(1981) Spherical ∆ui = 4suln(rp/r) Finite Difference
Jones and Van Zyl n/a Empirical n/a n/a Correlation between
(1981) approach measured t50 and
oedometer measured
values of ch
Kavvadas (1982) Piezocone Non-linear From strain-path 1-D
Model method; Effective stress-
strain model
Senneset et al. (1982) Cylindrical Elasto-plastic ∆ui = 2suln(rp/r) 1-D
Tumay et al. (1982) Piezocone Linear From strain path method; 1-D
Model Experimental data
Gupta and Davidson Piezocone Elasto-plastic Modified cavity 1-D Isotropic and anisotropic
(1986) Model expansion; dissipation as
cone penetrates
Soares et al. (1987) Piezocone Non-linear Corrected by visual 2-D
Model examination and
regression analysis
Whittle (1987) Piezocone Non-linear From strain-path 1-D
Model method; Effective stress-
strain model
Houlsby and Teh Piezocone Non-linear Predicted by large strain 1-D Finite Difference
(1988); Teh and Model finite element analysis
Houlsby (1991) and strain path method
Elsworth Point Elasto-plastic From point dislocation 2-D Not applicable for u2
(1990; 1993) Dislocation theory measurements
Theory
Aubeny (1992) Piezocone Non-linear From strain path method; 2-D Coupled consolidation
Model Effective stress-strain
model
Sully and Piezocone Non-linear Predicted by large strain 1-D Empirical time shift to umax
Campanella (1994) Model finite element analysis for OC dissipation
and strain path method
Burns and Mayne Spherical Elasto-plastic ∆uoct=4suln(rp/r) 1D Incorporates shear stresses;
(1995a) ∆ushear=σvo'[1- models OC dissipation;
(OCR/2)0.8] Finite Difference
Burns (1997) Spherical Elasto-plastic ∆uoct=4suln(rp/r) 1D Incorporates shear stresses;
This Study ∆ushear=σvo'[1- models OC dissipation;
(OCR/2)0.8] Analytical
*Modified after Lunne et al. (1992) and Jamiolkowski (1995)

43
 Evaluation of Soil Permeability
In addition to the coefficient of consolidation, soil permeability is another important factor in
both geotechnical and environmental engineering which can be obtained through the interpretation
of piezocone test data. In geotechnical engineering, the permeability of the soil is used in earth dam
design, dewatering considerations, and in seepage and slope stability problems. Additionally, the
permeability controls soil behavior, in terms of drained versus undrained response with respect to
rate of loading, during in-situ and laboratory testing. In environmental applications, the
permeability of a soil controls the advective transport of both groundwater and contaminating
chemicals in the subsurface environment. Consequently, it is a critical factor in many design
instances. The permeability of a soil is commonly measured both in the laboratory using small-scale
specimens, and in the field using borehole techniques, such as pumping tests.
There are several advantages to measuring the permeability of a soil in-situ, including the
fact that testing is performed on large quantities of soil, the geochemical regime is not disturbed, the
temperature is the same, and there is no disturbance due to sampling and transportation. In-situ
devices test a large portion of the subsurface soils during measurement, as opposed to laboratory
testing methods which rely on small, high-quality specimens. There are also a number of
disadvantages to in-situ testing methods, the most significant of which include uncontrolled
boundary conditions and disturbance due to probe penetration. However, the advantages to in-situ
measurement of permeability are significant enough to warrant the use of field testing methods.
This section focuses on the review of methods to evaluate the value of the coefficient of
permeability using laboratory methods, field methods, and piezocone testing.

Coefficient of Permeability
In geotechnical engineering, the term permeability (k) has the units of (L/T) and is used to
represent the ease with which water flows through a porous medium. The coefficient of
permeability assumes that the permeant fluid is water. However, other disciplines, including
hydrogeology and environmental engineering, use the term hydraulic conductivity to refer to this
property. The term intrinsic permeability, k*, is then used to refer to the properties of the soil alone
by factoring out the effect of fluid viscosity; this term has units of (L2). The permeability is defined
as:

q
k= Equation 2.19
Ai

where k = the coefficient of permeability in geotechnical engineering terminology and q = flow rate
(L3/T), A = cross-sectional area (L2), and i= hydraulic gradient. The permeability and the hydraulic
conductivity are related by the following:

k * ρg
K= =k Equation 2.20
µ

where K = the hydraulic conductivity (L/T) in hydrogeology and environmental engineering and k*
= intrinsic permeability (L2), ρ = fluid mass density (MT2/L4), µ = dynamic viscosity (MT/L2), and g
= the acceleration due to gravity (L/T2). Throughout this chapter, the term permeability (k) will be
used and is assumed equivalent to the term hydraulic conductivity.

44
 Influence of Soil Parameters
The permeability of a soil is dependent on a variety of soil characteristics including particle
size, void ratio, mineralogy (in clays), fabric, and saturation. Additionally, the diameter of soil
particles has been related to the permeability of a soil deposit. In particular, the permeability has
been linked proportionally to the square of D10, where D10 is the equivalent diameter at which 10%
of the soil particles are finer. For sands, Hazen developed the following relationship relating the
permeability to D10 (Lambe and Whitman, 1969):

k = 100 D10 2 Equation 2.21

where k is in cm/s and D10 is in cm. Soils with a wide gradation of particle sizes can pack more
efficiently, leading to smaller diameter pore sizes which reduce the flow of fluid.
The void ratio, e, is another parameter which affects the permeability of a given soil. The
void ratio is defined as the ratio of the volume of the voids to the volume of the solids; consequently,
an increase in the void ratio, or void space, will lead to an increase in the permeability of a given soil
(Lambe, 1951).
The mineralogy and fabric of a soil deposit are most important in the behavior of clay soils.
In general, clay minerals with sodium as the exchangeable cation tend to have lower permeabilities
than other clay minerals (Lambe and Whitman, 1969). Particle size is also important for clays, with
montmorillonite showing a much lower permeability than kaolinite. Additionally, the fabric affects
the behavior, with clays which are in a flocculated state having much higher permeability values
than do clays in a dispersed state.
The other major parameter which influences the measured permeability of a soil is the degree
of saturation. In geotechnical engineering, the saturation is defined as the ratio of the volume of the
water to the volume of the voids, S = Vw/Vv. Permeability is a direct function of the degree of
saturation, with an increase in saturation causing an increase in permeability (Lambe, 1951).

Measurement of the Permeability of Soils

Many different methods, both laboratory and field, exist for the determination of the
permeability of a soil. Additionally, the permeability will vary when measured using different
testing devices or different boundary conditions. A brief review of the methods currently used to
evaluate the permeability of a soil deposit follows.

Laboratory Measurement of Permeability

Laboratory permeability testing can take many forms including testing in fixed-wall or
flexible-wall chambers, constant head or variable head testing, and oedometer or triaxial chamber
testing. Daniel et al. (1985) performed an experimental comparison between fixed-wall and flexible-
wall permeameters. It was concluded that rigid-wall permeameters were inexpensive, simple,
compatible with organic liquids, and did not need high confining stresses; however, the rigid-wall
devices did not allow complete control over stresses, did not allow measurements of deformability,
were difficult to use in terms of soil placement, and had a potential for sidewall leakage. In contrast,
flexible-wall permeameters which conform to the soil specimen minimized sidewall leakage,
allowed control of applied stresses, allowed measurements of deformation, allowed sample
backpressuring, and could test a range of sample diameters. Disadvantages to the flexible wall
permeameters included a higher cost, membrane incompatibility with organic liquids, and the need
for high confining pressures at large hydraulic gradients.
45
 Another variation in laboratory testing is the choice between falling head or constant head
testing. Falling head tests are performed with a variable pressure head exerted on the soil sample,
while constant head tests maintain the pressure head at a constant value.

Field Measurement of Permeability

Field permeability tests were developed because the soil tested in laboratory permeability
devices were conducted on small specimens. Field tests fall into four general categories: borehole,
porous probe, infiltrometer, and underdrain (Daniel, 1989). Borehole methods drill a hole from the
soil surface and measure soil permeability either by the rate of flow into or out of the borehole.
Porous probe measurements are performed by penetrating a probe to the depth of interest and
measuring the rate of flow into a sampling chamber within the device. Infiltrometer tests are
performed by placing a sealed hydraulic source at the surface and measuring the flow into the
subsurface. Finally, underdrain tests are performed by placing a hydraulic source at the surface and
measuring the rate of flow through the layer being investigated into a porous underlayer. Field
methods are advantageous over laboratory methods because they utilize a large portion of the soil
being tested; however, homogeneous conditions are assumed in the test, and the boundary conditions
cannot be controlled.
Leroueil et al. (1992) performed a comparison of three field and three laboratory techniques
to measure the value of the permeability of the clay at the Bothkennar site in Scotland. In-situ
investigation techniques used push-in piezometers (constant head), a self-boring permeameter
(constant head), and the BAT system (variable head) (Torstensson, 1984). The results from the in-
situ tests showed that the push-in piezometer and BAT system evaluated consistently lower
magnitudes of the permeability than did the self-boring permeameter. Laboratory tests performed
included measurement of the vertical permeability in oedometer (variable head) and triaxial cells
(constant head), and measurement of the horizontal permeability in oedometer (variable head) and
radial flow cells (constant head). The tests performed in the triaxial cell consistently measured
higher values of permeability than the tests performed in the oedometer, which Leroueil et al. (1992)
attributed to the fact that more coarse grained particles were contained in the triaxial specimens.
The horizontal permeabilities measured in the oedometer were also consistently lower than those
measured in the radial flow cell which the authors also attributed to differences in specimen sample
size. Values of the permeability back-calculated from settlement and pore pressure dissipation data
show that the self-boring permeameter and the radial flow cell gave the magnitudes which most
closely agreed with the actual values. However, in general, the in-situ measured values were 30%
smaller than the laboratory measured values. Leroueil et al. (1992) stated that the difference was
because the in-situ sand and silt lenses were discontinuous, while in a laboratory specimen, those
lenses could extend to the imposed drainage boundaries of the test, which would give rise to a higher
measured permeability.
Because permeability is a difficult parameter to measure accurately, many different test
configurations have been developed in order quantify the value. However, the differences in the
testing methodologies, in both the laboratory and the field, yield large variations in the measured
value of soil permeability. Figure 2.18 shows a comparison of data gathered using two field
methods and one laboratory method where the difference in the measured permeability is as large as
an order of magnitude (Leroueil and Jamiolkowski, 1991). The variability is not restricted to
differences between laboratory and field methods of measurement. Figure 2.19 shows permeability
values evaluated in the laboratory using fixed-wall and flexible-wall permeameters, and the
measured difference is again as large as an order of magnitude (Daniel, 1994).

46
 Figure 2.18. Comparison of Lab and Field Measured Permeabilities
(Leroueil and Jamiolkowski, 1991)

Figure 2.19. Comparison of Permeability Measured in Fixed versus Flexible Wall

Permeameters (Daniel, 1994)

Piezocone Testing Evaluation of Permeability

In addition to laboratory and field methods for the measurement of soil permeability, it can
also be evaluated from piezocone penetration data. To date, the methods rely either on empirical
correlations, or on the estimation of the coefficient of consolidation and evaluation of permeability
using one-dimensional consolidation theory.

47
 Schmertmann (1978) proposed an empirical correlation of permeability with piezocone
dissipation data. The correlation was based on the time for 50% and 90% dissipation of excess pore
water pressure with his proposed relationship shown in Figure 2.20. Comparison with field and
laboratory measured data was not presented; therefore, the accuracy and reliability of the approach is
uncertain.

Figure 2.20. Schmertmann's Empirical Permeability Correlation

(Schmertmann, 1978)

Baligh and Levadoux (1980, 1986) proposed the following equation for evaluating the
horizontal permeability of a soil deposit from dissipation test data:

γw
k h ( piezocone ) = RR ( piezocone ) ch ( piezocone) Equation 2.22
2.3σ vo '

where RR(piezocone) = the recompression ratio = Cr/(1+eo) = strain per log cycle of effective stress
during recompression evaluated in the laboratory. Recompression was assumed because the
increase in pore pressure caused a decrease in the mean effective stress, or contributed to an artificial
overconsolidation of the initially normally-consolidated clay. Consequently, consolidation occurred
in a recompression mode as the excess pore pressures dissipated. The approach was verified using
data from Boston Blue Clay and showed that the permeability was overpredicted by approximately
one order of magnitude in the upper 20 m, and predicted within one to two times in the lower 20 m
of the deposit. Battaglio et al. (1986) applied this method to data gathered at the Porto Tolle site and
found that the permeability evaluated for normally consolidated soil was comparable to laboratory
data.
Gupta (1983) evaluated the in-situ permeability of Gainesville, Florida clays using a site
specific correlation developed through cone penetration testing and laboratory consolidation testing.
The following correlation for Lake Alice Clay in Gainesville was given:

48
 D = 4q c Equation 2.23

where D = 1/mv = constrained modulus, and the permeability was evaluated by:

chγ w
k= Equation 2.24
D

where the coefficient of consolidation was evaluated using the method of Gupta and Davidson
(1986). This was a site specific correlation for the Lake Alice clay which was based on uncorrected
tip resistance. The field predicted values were within one order of magnitude of the lab measured
values; however, because the correction (Equation 1.1) can be significant in soundings performed in
clays, correlations based on uncorrected tip readings are not reliable.
Wroth (1984) presented the following methodology for evaluating the permeability from
self-boring pressuremeter holding tests. After evaluation of the coefficient of consolidation, the
horizontal permeability was estimated by the following relationship:

ch γ w
kh = Equation 2.25
1−ν
2G[ ]
1 − 2ν

where through elastic theory, the shear modulus, G, and the constrained modulus, D, are defined as
follows:

E
G= Equation 2.26
2(1 + ν )

E (1 − ν )
D= Equation 2.27
(1 + ν )(1 − 2ν )

and can be related:

D (1 − 2ν )
G= Equation 2.28
2(1 − ν )

No comparison with field data was presented, but the author stated that the values evaluated from
the holding tests were in agreement with the values back-calculated from the settlement of a
structure in the U.K..
Robertson et al. (1992) proposed an empirical correlation between the horizontal coefficient
of permeability and t50 observed in dissipation tests in clay soils. The correlation was based on the
values of t50 obtained with a Type 2 piezocone because that location for measuring pore pressure

49
gave the most consistent results; however, the relationship showed a significant amount of scatter
(Figure 2.21).

Figure 2.21. Empirical Permeability Correlation

Robertson et al. (1992)

Manassero (1994) performed piezocone tests in a cement-bentonite slurry wall in order to

assess the construction quality and to evaluate the hydraulic conductivity of the barrier. However,
the data gathered with the piezocone were not reliable because the pore pressure transducer used was
limited to 2 MPa, and the data collected showed obvious truncation. The cone data were classified
according to the method of Robertson et al. (1986) by choosing an equivalent soil type from that
classification, for example clay to clayey silt. The coefficient of consolidation was assessed from
piezocone dissipation tests using the method of Teh and Houlsby (1991), and the permeability was
calculated using Equation 2.24 which the author stated gave good agreement with laboratory
measured values. A procedure was proposed which assigned hydraulic conductivity values on the
basis of the equivalent soil type from the piezocone classification method.
Gaberc et al. (1995) used piezocone dissipation tests to predict the permeability of organic
marshland soils. The coefficient of consolidation was predicted according to the methods of
Levadoux and Baligh (1986), Tumay et al. (1985), Houlsby and Teh (1988), and Gupta and
Davidson (1986) with the results from Gupta and Davidson (1986) giving the results in closest
agreement with the laboratory measured values. The permeability was then evaluated according to
the method proposed by Baligh and Levadoux (1986) for normally consolidated Boston Blue Clay.
The predicted results were approximately one order of magnitude higher than the laboratory
measured values.
Parez and Fauriel (1988) presented a method for evaluation of the permeability of a soil
deposit based on correlation with t50 measured from Type 2 dissipation tests [referenced in Leroueil
50
and Jamiolkowski (1991)]. The relationship was developed for soils ranging from gravel to clay
(Figure 2.22) and the mean trend is approximately given as follows:

k = (251 ⋅ t 50 ) −1.25 Equation 2.29

where k is in cm/s and t50 is in seconds.

Figure 2.22. Permeability - t50 Correlation

(Parez and Fauriel, 1988)

Summary
In summary, multiple methods exist to evaluate the permeability of a soil deposit, either in a
laboratory or field setting. Laboratory methods offer the advantages of controlled boundary
conditions, control of permeating fluid, control of applied stresses, and control of fluid gradient,
while in-situ testing offers the advantages of minimal soil disturbance, constant geochemical
regimes, field temperature, and large volumes of representative soil being tested.
Comparative studies (Leroueil et al., 1992) in a uniform clay with silt and sand lenses
showed that in-situ methods of testing of the soil permeability measured consistently lower values
than those obtained in the laboratory, most likely due to the influence of high-permeability lenses in
the lab specimens. However, many environmental applications, including the transport of
contaminants through a soil profile, are not suited to full-scale in-situ testing, and must be confined
to laboratory measurements. In all testing methodologies, the experimental setup will affect the
51
measured values; consequently, care should be taken to guarantee that the boundary and initial
conditions are representative of the actual problem, and that the proper data are being measured, for
example, measurement of horizontal permeability versus vertical permeability. Field permeability
tests are the most common method for evaluating the permeability in-situ; however, the data
gathered during piezocone dissipation testing yield information about the flow of pore fluid through
a soil deposit and can be interpreted to assess field-measured values of the soil permeability.

52
 CHAPTER 3

INTEGRATED OPTO-ELECTRONIC CHEMICAL SENSOR

FOR BTEX DETECTION IN CONE PENETRATION TESTING

Introduction
Inclusion of an integrated optic (IO) chemical sensor, capable of quantitatively measuring
chemical concentration, within a cone penetrometer offers many advantages over the more
traditional methods of groundwater sampling and analysis. Because the IO sensor operation is fast
and reversible, contaminant concentration with depth can be mapped essentially continuously,
during the cone push. In addition, there is no exposure of the groundwater to the atmosphere during
sampling, and there is no need for transportation of the sample to a laboratory for analysis. While
the integrated optic chemical sensor will not eliminate the use of monitoring wells, it can help
optimize placement of these expensive wells and also reduce the number of wells required for
adequate coverage of a site.
This chapter details the operation of the newly developed integrated opto-electronic sensor
used for the detection of benzene, toluene, ethylbenzene, and xylene (BTEX chemicals). During this
study, the sensor was miniaturized for inclusion within a trailing cone penetrometer module. While
the sensor used in this series of experiments was developed for the detection of the BTEX chemicals,
the design can be readily adapted to detect a wide variety of other chemicals as well, including
chlorinated hydrocarbons and dissolved metals.
In this study, a series of bench-top experiments were performed, under laboratory conditions,
as a proof-of-concept for the operation of the sensor in a soil environment. This chapter outlines the
operating principle of the sensor and gives the details of the initial experimental performance in its
miniaturized form.

Sensor Operation
The Hartman interferometer used in the testing program was originally developed at the
Georgia Tech Research Institute (GTRI) for the detection of gaseous ammonia, NH3 (Hartman et al.,
1988; Hartman, 1990; Ross et al., 1991; Hartman et al., 1992). More recent versions of the
interferometer are used for the measurement of pH (hydrogen ion concentration) and for the
detection of the BTEX chemicals. Sponsorship for the integration of the sensor into a cone
penetrometer was provided by the Army Research Office (ARO) in Raleigh, North Carolina. Phase
I, which was a proof-of-concept stage was completed in June, 1997; Phase II is a joint venture
between ARO and the US Army Corps of Engineers Waterways Experiment Station (WES), and
includes the complete integration of the sensor into a trailing module of a cone penetrometer. Phase
II is scheduled for completion in February, 1999.
The basic operation of the sensor relies on the use of a chemically-selective polymeric
coating placed on an optically transparent waveguide (Figure 3.1) to interact with the chemical of
interest either through adsorption, absorption, or a chemical reaction. The measuring device
includes the glass waveguide (12 mm by 25 mm by 2 mm), polymeric coating (0.14 µm thick), a
small commercial laser light source (12.7 mm diameter by 16.5 mm length), and a data processing
electronics circuit board (127 mm long by 27 mm wide). The sensor, laser, and processing
electronics are packaged within a stainless steel casing, which is 35 mm in diameter and 100 mm
long.
When the specific chemical of interest interacts with the polymer, the effective index of
refraction is altered, and the velocity of an optical beam which is passed through the waveguide
53
changes, and becomes either slower or faster. The optical beam produces electric fields which decay
into the cover film; when the speed of this beam is altered, the electric field from the optical beam is
altered, which creates a different decay into the cover film (Figure 3.2 and Figure 3.3). The changed
optical beam from the chemically-reactive channel is then combined with an unaltered reference
channel light beam which produces an interference pattern between the two beams. This pattern can
then be correlated to the concentration of the chemical being measured. The sensor can be
configured to measure both gaseous and aqueous phases; however, the sensor used in this study was
limited to aqueous measurements of the BTEX chemicals. Additionally, the chemically-reactive
coating can be changed in order to detect different, or multiple chemicals or to optimize the
detection of specific contaminants. The original IO system was a large bench-top configuration
using lasers, mirrors, and a table array for setup. During this study, Photonic Sensor Systems and
GTRI miniaturized the layout into a compact size for incorporation into a cone penetrometer
module.

Chemical of Interest
Cover Film

Waveguide
2 mm
Detector Substrate Laser

25 mm
Figure 3.1. Integrated Optic Sensor Schematic (not to scale)

Evanescent Tails of
Electromagnetic Field

ΘL
ΘH

Low Order Mode High Order Mode

Figure 3.2. Principle of Waveguide (not to scale)

54
 Detector Array
Reactive Channel

Reference Channel

Diode Laser

Figure 3.3. Sensor Configuration (not to scale)

Sensor Housing
One of the key aspects to the integration of the BTEX sensor into a cone penetrometer is the
development of an adequate casing to carry the device into the subsurface. The following section
reviews the design aspects which are involved in the Phase II inclusion of the sensor as an integral
part of the cone penetrometer. Because the integrated optic sensing chip cannot withstand the
abrasion of direct soil contact during a cone sounding, it must be housed within an internal chamber
inside the cone module. Small groundwater samples (approximately 0.5 cc) will be drawn into the
chamber and into contact with the sensor. The key elements in the design of the system are the fluid
delivery (generation of flow) and the filtration of the fluid entering the chamber to prevent the influx
of soil particles across the waveguide.

Fluid Sampling System

Fluid can be drawn into the sensing chamber either by pumping or by the gradients flow
caused by cone motion. A moving penetrometer causes significant gradients in the vicinity of the
cone tip. Robertson et al. (1986) show that the relative pore pressure gradients decrease from the
cone tip/face to the cone shaft for different soil types (Figure 3.4). All measured penetration pore
pressures shown in the figure are normalized by the hydrostatic value established by the
groundwater table. The effect is much more dramatic for stiff to hard clays than it is for loose sands.
In fact, for very hard overconsolidated and fissured clays, the gradients are extremely large (Mayne
et al., 1990). The gradients generated during cone penetration testing may be used to provide
movement of fluids into a small cavity in the penetrometer, where exposure to the integrated optic
sensor can occur. Only small volumes are required (≈ 0.5 cm3) because of the miniature sensor size.
In typical soil deposits, the void ratio (eo) is on the order of 1.0, indicating that half the volume is
occupied by solid particles and half by void space. Voids are filled with either vapor, fluids, or both.
In saturated soils all pore space is filled with fluid; consequently, a 15 cm2 cone penetrometer (44
mm diameter) with length of 60 cm will displace a volume of soil equal to 900 cm3, of which 450
cm3 are incompressible solids and 450 cm3 are fluids and vapors (assuming eo = 1). Under these
circumstances, a normal cone penetration rate should provide adequate fluid flow for chemical
testing, without the aid of artificial pumping.

55
 Figure 3.4. Distribution of Normalized Porewater Pressures
During Penetration in Different Soil Types
(Robertson et al., 1986)

In a static condition, sampling of pore fluid will require a miniaturized pump to draw the
sample into the sensor flow cell. EPA guidelines permit the use positive displacement, peristaltic,
and gas-drive pumps for sampling of inorganic materials like dissolved metals or for non-volatile
organics; however, only positive displacement bladder, helical rotor, and gear-drive pumps are
recommended for the sampling of volatile organic compounds (EPA, 1993). Two currently
available miniaturized pump designs for groundwater sampling include a piston pump which uses
hydraulic pressure to draw the sample in and to expel the sample out of the chamber and another
device known as the Cone SipperTM which uses check valves in combination with vacuum and
pressure lines in order to sample the groundwater (Lightner and Purdy, 1995).
Future work in this ongoing study will evaluate the use of a miniaturized positive
displacement bladder pump and the Cone Sipper™ operating under a variety of physical and
chemical conditions. Physical conditions will include the simulation of variable groundwater
flowrates and variable hydrostatic pressures in order to monitor pump operation. In addition, the
study will examine the effect of pumping samples into a chamber on the measured chemical
concentrations, through comparison of sensor-measured values with known reference
concentrations. The influence of pumping is expected to be most significant in terms of the
monitoring of volatile organic chemicals.

56
 Geochemical Behavior of the BTEX Chemicals
The BTEX chemicals belong to the group of aromatic hydrocarbons (Figure 3.5) and are
constituents of fuel products. Aromatic hydrocarbons are composed of a benzene ring with
additional functional groups attached in the ortho-, meta-, and para-positions. The chemicals are
classified as light non-aqueous phase liquids (LNAPLs) because they are less dense than water and
are not miscible with water; however, they are slightly soluble (Table 3.1). Because they are the
most soluble components of gasoline, the BTEX chemicals are commonly used as diagnostics of
gasoline spills, either from leaking underground storage tanks or from surficial spills. Sampling of
the BTEX chemicals is difficult due to their volatility; consequently, in-place measurements of
BTEX concentrations, which do not expose the chemicals to atmospheric oxygen are desirable.

Laboratory Testing Program

A preliminary laboratory testing program was undertaken in order to demonstrate the
viability of the BTEX sensor in a soil environment. Initial experiments were performed in aqueous
solutions and subsequent experiments were performed with the sensor in a soil/water matrix. The
soil used in all testing was Ottawa 20/30 sand which is a relatively unreactive sand soil composed of

CH3 C2H5

Benzene Toluene Ethylbenzene

C6H6 C7H 8 C8H10

CH3 CH3 CH3

CH3

CH3

CH3

o-Xylene m-Xylene p-Xylene

C8H10 C8H10 C8H10

Figure 3.5. Structure of the BTEX Chemicals

57
 Table 3.1. BTEX Chemical Characteristics
(CRC, 1993)

Chemical Molecular Solubility in Soil Water Vapor Henry's Log Boiling Melting
Weight (g) Water Partition Pressure Constant Kow Point Point
(mg/L) Coefficient (mm Hg) (atm- (oC) (oC)
m3/mol)
Benzene 78.11 1780 97 95.2 5.53 2.12 80.1 5.5
Ethylbenzene 92.10 500 242 22.0 5.13 2.73 110.6 -95.0
Toluene 106.17 150 622 7.0 6.10 3.15 136.2 -95.0
o-Xylene 106.17 170 363 10.0 7.04 3.26 144.4 -25.3

essentially pure silica (Lambe and Whitman, 1969; Mitchell, 1993). Ottawa sand was chosen as the
soil matrix because it is a well-characterized soil and provides relatively little chemical interference
during the testing.
All tests were performed in a laboratory setting using a triaxial calibration chamber (14 cm
diameter; 61 cm height) constructed out of BTEX resistant materials: Pyrex glass, stainless steel, and
Viton o-rings (Figure 3.6). The IO sensor was enclosed in a stainless steel tube which was the
approximate diameter of a 10 cm2 cone penetrometer (35.7 mm diameter) and 400 mm length.
Inflow and outflow tubes made of Teflon® were connected to the flow cell over the waveguide in
order to draw the sample into the probe and into contact with the waveguide and the tubing was
changed between each experiment. Polyethylene filters (2 µm) were attached to the inflow section
of tubing in order to limit the migration of fines into the waveguide. Additionally, a stainless steel
filter was purchased for its better compatibility with the organic chemicals and testing conditions;
however, it was not available during the initial phase of testing. Flow through the cell was initiated
by suction using a syringe. The inflow tubing was buried approximately six to seven inches below
the sand surface, and the outflow was directed to a container outside of the testing chamber.
Calibration of the sensor was performed by preparing solutions of known concentration and
measuring the sensor response. In addition, calibration of the sensor response to toluene was
performed using a Hewlett Packard 5890 Series II Gas Chromatograph (Figure 3.7); however,
calibrations were not performed for benzene or xylene due to sensor malfunctioning which is
discussed in more detail later. For the concentration ranges tested, the sensor response was
essentially linear. All tests were performed in a static system; the IO sensor was not moved once it
was placed within the calibration chamber. A schematic of the testing setup is shown in Figure 3.8.
A series of tests were performed using benzene, toluene, and xylene as the contaminating chemicals.
Data acquisition was performed using a Pentium notebook computer (Figure 3.9) which was used to
monitor 8 channels from the waveguide. Figure 3.10 shows a picture of the miniaturized sensor
supported by an external clamp on a laboratory bench-top. The processing electronics are to the left
of the clamp, and the sensor, flow cell, and inflow and outflow tubing are to the right of the clamp.
Note also that a glass waveguide is shown at the right mid-center.

58
 Figure 3.6. BTEX Testing Chamber for IO Sensor Experiments

Figure 3.7. Gas Chromatograph Used for Reference Chemical Concentrations

59
 IO Sensor
Package
Data Acquisition
Inflow

Pyrex Glass Cylinder

5.5” o.d.; 24” in height
Gas
Chromatograph Soil Matrix

Note: Inflow and outflow

Top and
tubing made of
Bottom Platen -
stainless steel and teflon;
Stainless Steel
seals are made of viton

Figure 3.8. Testing Schematic for BTEX Detection by IO Sensor

Figure 3.9. Miniaturized Optic Sensor Configuration and Data Acquisition

60
 Figure 3.10. Complete Integrated Optics Sensor Package Configuration

Experimental Results
Seven experiments were performed with the sensor in aqueous and/or saturated sand matrices
(Table 3.2). The experiments performed were initial proof-of-concept tests done to verify the proper
operation of the miniaturized sensor in a soil matrix. The initial tests were successful; however,
malfunctioning of the prototype sensor operation prevented further experimentation. Consequently,
only the results of the initial testing program are presented.

Table 3.2. Test Series of Preliminary Experimental Program With IO Sensor

Exp Chemical Concentration Matrix Waveguide
No. (ppm) Coating
1 Benzene 250 Aqueous PBIBMA*
2 Benzene 190-380 Aqueous/Sand PBIBMA*
3 Toluene 25 Aqueous Teflon-A
4 Toluene 18 Aqueous/Sand Teflon-A
5 Xylene 6 Aqueous Teflon-A
6 Xylene 9.5 Aqueous/Sand Teflon-A
7 Xylene 7 Aqueous/Sand Teflon-A
*polybutylisobutylmethacralate (PBIBMA)

61
 The results from the seven experiments are shown in Figure 3.11 through Figure 3.17 in both
aqueous and soil environments. Figure 3.11 shows the sensor response [chemically reactive coating
polybutylisobutylmethacralate (PBIBMA)] in de-ionized water with benzene added after nine
minutes of operation at a stable baseline. The addition of benzene produced an immediate response
which measured increasing concentration up to a concentration of 250 ppm. After four minutes of
operation at a constant concentration, de-ionized water was added to the test cell in order to dilute
the concentration of benzene, and the sensor measured a decrease in concentration. Figure 3.12
shows the results of a test performed with the sensor operating in a soil/aqueous environment. This
figure shows the continuation of a previous test where the concentration in the test cell was initially
190 ppm. Additional benzene was added in order to raise the concentration in the test cell to 380
ppm. One notable difference in the response of the sensor in a soil matrix as opposed to a pure
aqueous environment is the rate at which the concentration changes. The soil matrix, as would be
expected, slows the movement of the contaminant to the inflow tubing of the sensor, producing
slower changes in concentration than are seen in pure aqueous solutions. This behavior is consistent
throughout the tests performed.

25
Benzene Concentration = 250 ppm
20
Response (π)

15 De-ionized
Water
Added
10
Benzene
added
5
Exp #1
0
0 5 10 15 20
Time (min)

Figure 3.11. Sensor Response to Benzene in Aqueous Solution

62
 35
Concentration = 380 ppm
Pure water
30 Additional added/water
benzene level lowered
Response (π) 25 added

20

15 Initial
Concentration Ottawa 20/30 sand
10 190 ppm

5
Exp # 2
0
75 100 125 150 175
Time (min)

Figure 3.12. Sensor Response to Benzene in Soil Matrix

Figure 3.13 and Figure 3.14 show the response of a different chemically reactive coating,
Teflon-A, to toluene. The Teflon-A was used as the reactive coating in these experiments because
the response was much more definitive than the response seen with the PBIBMA coating. In fact,
the sensor response time was so rapid that the data acquisition electronics could not record data fast
enough to monitor the concentration change of benzene, the smallest molecule. Consequently, these
tests were performed using toluene and xylene which are larger molecules and have slower response
time due to mass transfer effects into the coating. The toluene behavior recorded again shows rapid
response and concentration change in aqueous solutions while the response in the soil/aqueous
matrix demonstrates a significantly slower response (Figure 3.13 and Figure 3.14).

63
 12
Toluene Concentration = 25 ppm

10
De-ionized
Water

Response (π)
8
Added

4 Toluene
Added
2 No Soil Matrix Exp #3

0
10:00 10:05 10:10 10:15 10:20
Time (hr:min)

Figure 3.13. Sensor Response to Toluene in Aqueous Solution

10
9 Toluene Concentration = 18 ppm
8
7
Response (π)

6 De-ionized
5 Water
Added
4
3 Additional
Toluene Exp #4
2 Added
Ottawa 20/30 Sand
1
0
13:10 13:15 13:20
Time (hr:min)

Figure 3.14. Sensor Response to Toluene in Soil Matrix

The final set of experiments were performed using the Teflon-A reactive coating in both
aqueous and soil environments. Xylene, which is a larger molecule than toluene, shows a slower
response time than the smaller molecules, even in aqueous solution. Figure 3.15 shows the sensor's
response to xylene (concentration = 6 ppm) in an aqueous solution and Figure 3.16 and Figure 3.17
show the sensor's response in a soil matrix (concentration = 9.5 ppm and 7 ppm). The same patterns
64
are evident in the sensor response to xylene in a soil matrix as were seen with benzene and toluene:
the soil matrix slows the transport and diffusion of the chemical through the soil and into contact
with the sensor.

10
De-ionized
8 Xylene Concentration = 6 ppm Water
Added
Response (π)

6
No Soil Matrix
4
Xylene
2 Addition
Exp #5

0
10:55 11:00 11:05 11:10 11:15 11:20 11:25 11:30
Time (hr:min)

Figure 3.15. Sensor Response to Xylene in Aqueous Solution

14
Xylene Concentration = 9.5 ppm
12 De-Ionized
Water
10 Added
Response (π)

6
Xylene
4 Added Ottawa 20/30 Sand

2 Exp # 6

0
16:10 16:20 16:30 16:40 16:50 17:00
Time (hr:min)

Figure 3.16. Sensor Response to Xylene in Soil Matrix (9.5 ppm)

65
 10
De-ionized
9 Xylene Concentration = 7 ppm Water
8 Added
Response (π) 7
6
5 Xylene
Added
4
3 Exp # 7
Ottawa 20/30 Sand
2
1
0
11:55 12:00 12:05 12:10
Time (hr:min)

Figure 3.17. Sensor Response to Xylene in Soil Matrix (7 ppm)

Discussion
Two serious complications with the prototype sensor hardware prevented further
experimental testing during this phase: the waveguide mount was not machined to a sufficient
tolerance and broke the glass waveguide when it was tightened into place. Fracturing of the
waveguide occurred on three separate occasions, which required approximately two months for
repair in each instance. Additionally, there was an incompatibility of communication protocols
between the waveguide hardware and the data acquisition electronics which prevented the rapid data
transfer needed for measurement of some chemicals. The sensor development was performed at
GTRI, independently of the electronics development which was performed by General Atomics in
Florida. Ongoing work in Phase II is underway to re-design the communication systems in order to
obtain sensor-electronics compatibility. Additionally, the electronics processing hardware is being
developed under the Environmental Systems Management, Analysis, and Reporting Network (E-
SMART) in order to provide a standard for connection of environmental monitoring sensors used by
the U.S. Department of Defense.
While a complete evaluation of the sensor behavior in a variety of soil environments was not
possible due to the above reasons, enough data was gathered to demonstrate that the sensor can
measure BTEX chemical concentrations in an Ottawa sand soil matrix. A range of concentrations
were measured, from a high of 380 ppm benzene to a low of 6 ppm xylene, which demonstrates the
sensor versatility in terms of detection limits. Additionally, the only limits on the measurement
response time were imposed by the data acquisition software and hardware which is currently being
upgraded in order to improve processing time. Another key factor in the sensor performance was the
question of reversibility. In repeated experiments, the sensor demonstrated that increases as well as
decreases in chemical concentration could be measured.

66
 Experimental work with the next version of the IO sensor is currently being planned in order
to verify its field applicability, including a study of sensor response to a mixture of chemicals, the
effect of soil grain size on sensor measurements, sensor response at extremes of pH, sensor response
in high ionic strength environments, the interaction of naturally occurring soil acids with the sensor's
measurement, the influence of sampling procedure including sampling and filtering materials and
pump configurations, and the influence of dynamic penetration on the sensor readings.

Conclusions
The integrated optic sensor was miniaturized during this study and has demonstrated viability
for the detection of the BTEX chemicals in soil environments under laboratory conditions. While
the robustness of the device must be improved in order to withstand the rigors of cone penetration
testing, the BTEX sensor is clearly a promising tool in terms of providing quantitative measurement
of chemical concentrations in-situ. Of particular interest to other applications is the fact that the
waveguide and electronics packaging would be the same for sister sensors because only the
polymeric coating is changed for detecting other chemistries.

67
 CHAPTER 4

PIEZOCONE MODEL FOR EVALUATING THE COEFFICIENT OF CONSOLIDATION IN

SOFT TO HARD CLAYS

Introduction
In addition to the measurement of soil properties and contaminant concentrations, data
gathered from piezocone soundings can be used for the interpretation of soil and pore fluid
characteristics. This chapter develops a model for the interpretation of pore pressure dissipation data
measured in the Type 2 location using a hybrid model based on cavity expansion theory (Vesic,
1972) and critical-state soil mechanics (Wroth, 1984). Extending beyond prior interpretative
methods, the approach can represent the dilatory dissipation behavior observed in heavily-
overconsolidated clays, as well as the common monotonically decreasing decay of pore pressures in
soft to firm clays and silts. While it is known that the cavity expansion model for cone penetration is
a simplification in that it assumes isotropic soil and one-dimensional behavior, it offers a
straightforward and reasonable approach to the analysis of dissipation behavior in clays. Similarly,
the adopted isotropic constitutive soil model (Modified Cam Clay) is also over simplistic, yet
conveys the general nature of soil behavior. The goal of this chapter is to represent the porewater
pressure dissipation behavior of normally-consolidated to heavily-overconsolidated clays using the
hybrid formulation in combination with an analytical closed-form solution to the radial consolidation
equation.

Components of Porewater Pressure

During cone penetration, excess pore pressures develop due to changes in both normal and
shear stresses. Compression-induced pore pressures from increasing normal stresses are always
positive; however, shear-induced porewater pressure may be either positive (contractive) or negative
(dilatant). Under the cone tip, the shear stress component is derived from stress path analysis and
always has positive values (Chen and Mayne, 1994) and the magnitude of the normal-induced
response is often much larger than that of the shear-induced response. However, along the shaft, the
normal- and shear-induced components can be comparable in magnitude. While the normal
component will always remain positive, the shear-induced component can be positive at low OCRs,
or negative at high OCRs.
The zone of influence of the octahedral normal stress is a function of the rigidity index of the
soil and is large (approximately 10 to 20 diameters across), in comparison to the zone of influence of
the shear stress. The rigidity index is defined as Ir = G/su, where G = the undrained shear modulus
and su = the undrained shear strength. The shear modulus is defined as G = τ/γ, the ratio of shear
stress to shear strain. It can be related to an equivalent Young's modulus by E = 2G(1+ν), where E
= Young's modulus and ν = Poisson's ratio. In undrained loading, ν = 0.5 and typically, E might be
taken as E50, the secant modulus at one-half the undrained shear strength. This yields: G = E50/3.
In contrast, the zone affected by interface shear stresses is limited to a thin annulus next to
the body of the cone penetrometer (say 1 to 10 mm). Because the volume affected by the normal
stress is much larger than that affected by the shear stress, the dissipation of the shear-induced
pressures will occur more rapidly than will the dissipation of the normal-induced pore pressures.
The measured porewater pressures (um) are the combined magnitudes of the normal- and
shear-induced components at any given time, t, plus hydrostatic pressure. Consequently, a
dissipation test with measured um will reflect the overall summation of the individual components
with time. Figure 4.1 shows the conceptual behavior of the individual components: note that the
octahedral normal-induced pore pressures are large in magnitude and decay over a long time span,
68
while the shear-induced pore pressures are lower in magnitude and decay much more rapidly. The
hydrostatic water pressure remains essentially constant in most environments; notable exceptions to
this condition are tidally influenced areas. Figure 4.2 shows the measured behavior with the
contributions summed. The total magnitude of measured pore pressure is represented as follows:

um = ∆uoct + ∆u shear + uo Equation 4.1

Shoulder Position Dissipation Model

The core of the model presented uses cavity expansion theory to represent the octahedral
normal component of excess pore pressure combined with Modified Cam Clay to quantify the shear-
induced component of excess pore pressure and the variation of the undrained shear strength as a
function of OCR. The hybrid cavity expansion/critical-state formulation has successfully
represented penetration porewater pressures in a variety of clays with consistencies ranging from
soft to stiff to hard (Mayne, 1991). The initial piezocone dissipation model based in this approach
used a finite-difference scheme to decay these components with time (Burns and Mayne, 1995a);
however, the solution required considerable computer time and was conditionally stable at early
stages. Consequently, the latest version of the model relies on an analytical solution to the
consolidation equation which is rapid and not subject to instability. The procedure to model both
lightly- and heavily-overconsolidated clay dissipation behavior with time is outlined below with a
brief review of the concepts of Cavity Expansion, Modified Cam Clay, and the derivation of the
analytical solution to the one-dimensional consolidation equation.

Normal-Induced Pore Pressures From Cavity Expansion Theory

Cavity expansion theory was originally developed for application to the expansion of either a
spherical or cylindrical cavity indentation in metals (Bishop et al., 1945). Vesic (1972) developed
solutions to the expansion of spherical and cylindrical cavities in a soil with Mohr-Coulomb
frictional parameters, and Torstensson (1977) applied the solution to cone penetration testing. The
solution was able to estimate ultimate stresses (bearing capacity) and the generated pore water
pressures due to normal stress changes in the case of undrained cavity expansion in a soil medium
and is often used as a model for cone penetration testing. A review of the solution as applicable to
the model used in this chapter is given below.
During the expansion of a cavity in a soil medium, a plastic zone develops surrounding the
cavity and the equations of equilibrium for an element within that zone are written as:

∂σ r σ r − σθ
Spherical Cavity: +2 =0 Equation 4.2
∂r r

∂σ r σ r − σ θ
Cylindrical Cavity: + =0 Equation 4.3
∂r r

69
 250
Octahedral Normal-
Induced
200
∆uoct

∆uoct (kPa)
150

100

50

0
0.1 1 10 100 1000 10000 100000
Time (sec)

100
Shear-Induced
50 Normally Consolidated Clay ∆ushear
∆ushear (kPa)

-50

-100 Overconsolidated Clay

-150

-200
0.1 1 10 100 1000 10000 100000
Time (sec)

300
uo

200
uo (kPa)

Hydrostatic Pore Pressure

100

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.1. Dissipation Behavior of Components of Measured Pore Pressure

70
 350
Normally-Consolidated
um
300
250
umeas (kPa)

200
Overconsolidated
150
100 Hydrostatic Pressure
50
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.2. Dissipation Behavior with Summation of Components

where σr = radial normal stress, σΘ = circumferential normal stress, and r = distance to the center of
the cavity. A diagram of the soil elements is shown in Figure 4.3.
The failure condition for soils in undrained shear is given by Mohr:

σ r − σ θ = 2 su Equation 4.4

where su = the undrained shear strength.

71
 Figure 4.3. Soil Elements in Cavity Expansion
(Torstensson, 1977)

The solution used by Torstensson (1977) assumed that the soil was an ideal elastic-plastic material
and that there were no volume changes within the plastic zone, which led to the following
formulation for the width of the induced plastic zone:

rp E
Spherical Cavity =3 Equation 4.5
ro 2(1 + ν )su

rp E
Cylindrical Cavity = Equation 4.6
ro 2(1 + ν ) su

where rp = radius of the plastic zone and ro = radius of the cone.

From the preceeding equations, the ultimate cavity pressure (∆pu) was derived to be:

4 E
Spherical Cavity ∆pu = su ln Equation 4.7
3 2 su (1 + ν )

E
Cylindrical Cavity ∆pu = su ln Equation 4.8
2 su (1 + ν )

Finally, for the calculation of pore pressure, it was assumed that the change in pore pressure, ∆u,
was equal to the change in the mean stress:
72
 1
∆u = ∆σ m = ( ∆σ r + ∆σ θ + ∆σ z ) Equation 4.9
3

where ∆σm = change in the mean stress. This yielded the following equations for the generated pore
pressure:

4 E
Spherical Cavity ∆u = su ln Equation 4.10
3 2 su (1 + ν )

E
Cylindrical Cavity ∆u = su ln Equation 4.11
2 su (1 + ν )

In summary, in the present formulation, cavity expansion theory is used to represent the pore
pressures generated due to changes in the octahedral normal-induced stresses during cone
penetration testing.

Shear-Induced Pore Pressures From Modified Cam Clay

Development of shear-induced pore pressures in undrained loading is evaluated using a
constant p stress path for isotropically consolidated soil in Cambridge q-p space (Wroth, 1984). The
shear-induced pore stress can then be expressed as the difference between the initial and final
effective mean stresses:

∆ushear = σ o '−σ f ' Equation 4.12

where σo' = the initial effective normal stress and σf' = mean effective stress at failure. The stress
path representation is shown in Figure 4.4.

73
 Figure 4.4. Stress Path Representation for Shear-Induced Pore Pressures
(Chen and Mayne, 1994)

According to Modified Cam Clay, the final mean effective stress can be represented by the
following:

OCR Λ
σ f ' = σ o '( ) Equation 4.13
2

where Λ = the plastic volumetric strain ratio = 1 - Cs/Cc where Cs = the swelling index and Cc = the
compression index. Data compiled by Mayne (1988) indicated that Λ = 0.80 is sufficient for many
clays. For estimation of the shear-induced pore pressures along the cone shaft at the Type 2
position, a value of Λ = 0.80 is used in this representation.

Derivation of the Analytical Solution to the Consolidation Equation

The linear partial differential equation which controls radial consolidation surrounding a
piezocone was solved in order to obtain an analytical solution for use in the model. Previous
researchers have dealt with similar problems in relation to heat flow and consolidation surrounding
driven piles (Carslaw and Jaeger, 1959; Randolph and Wroth, 1979b). The consolidation equation
was given in Equation 2.9:

∂u ∂ 2 u c h ∂u
= ch 2 + Equation 4.14
∂t ∂r r ∂r

for a < r < b where a = rcone and b = rplastic. Assume a solution of the form:
2
u = ve − cα t
Equation 4.15

74
where v = f(r) only. Then:

∂u 2
= − cα 2 ve − cα t
Equation 4.16
∂t

∂u 2 dν
= e − cα t Equation 4.17
∂r dr

∂ 2u − cα t d ν
2
2
=e Equation 4.18
∂r 2 dr 2

Substitute:

2 2 d 2 v c − cα 2 t dν
− cα 2 ve − cα t = ce − cα t
+ e Equation 4.19
dr 2 r dr

2
Divide by ce − cα t :

d 2 v 1 dv
+ + α 2v = 0 Equation 4.20
dr 2 r dr

which is Bessel's Equation of zero order. The solution is of the form (Abramowitz and Stegun,
1964):

v (r ) = AJ 0(αr ) + BY 0(αr ) Equation 4.21

where J0 and Y0 are Bessel functions of the first and second kind, respectively and A and B are
constants. Now, use the first boundary condition which states that there is no excess pore pressure
generated outside the plasticized zone (rplastic = rcone Ir0.333) to evaluate the constants A and B (at r = rplastic,
u = 0):

v (rplastic ) = AJ 0(αrplastic ) BY 0(αrplastic ) = 0 Equation 4.22

Solve for A:

− BY 0(αrplastic )
A= Equation 4.23
J 0(αrplastic )

75
Substitute:

− BY 0(αrplastic )
J 0(αr ) + BY 0(αr ) = 0 Equation 4.24
J 0(αrplastic )

Divide by B:

− Y 0(αrplastic ) J 0(αr ) + J 0(αrplastic )Y 0(αr ) = 0 Equation 4.25

Substitute into solution to yield:

v (r ) = − J 0(αrplastic )Y 0(αr ) + Y 0(αrplastic ) J 0(αr ) Equation 4.26

Use the second boundary condition which states that the cone is an impermeable boundary (at r =
∂v
rcone, = 0 ):
∂r

Note: J0' = -J1 and Y0' = -Y1 Equation 4.27

∂v
= αJ 0(αrplastic )Y1(αr ) − αY 0(αrplastic ) J 1(αrcone ) = 0 Equation 4.28
∂r

Divide by α:

J 0(αrplastic )Y1(αrcone ) − Y 0(αrplastic ) J 1(αrcone ) = 0 Equation 4.29

In order to evaluate α, the root of the previous equation must be taken; however, this equation has an
infinite number of roots. The solution will be the sum of all possible solutions; however, evaluation
of the first 50 roots yields sufficient accuracy.

∞
u = ∑ Bn [ −Y 0(α n r ) J 0(α n rplastic ) + Y 0(α n rplastic ) J 0(α n r )] Equation 4.30
n =1

Using the initial condition imposed by spherical cavity expansion to calculate the pore pressures due
to octahedral normal stress (at t=0):

rcone I r 0.333
u = 4su ln( ) Equation 4.31
r

76
Substitute:

rcone I r 0.333 ∞
4 su ln( ) = ∑ Bn [Y 0(α n rplastic ) J 0(α n r ) − J 0(α n rplastic )Y 0(α n r )] Equation 4.32
r n =1

Set αn = α1 and multiply both sides by r[Y0(α1rplastic)J0(α1r) - J0(α1rplastic)Y0(α1r)]:

0 .333
rcone I r
r 4 su ln( )[Y 0(α1rplastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1r )] =
r
∞

∑ B r[Y 0(α r
n =1
n n plastic ) J 0(α n r ) − J 0(α n rplastic )Y 0(α n r )] ⋅ Equation 4.33

[Y 0(α1rplastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1r )]

Integrate from a to b:

b
rcone I r 0.333
4su ∫ r ln( )[Y 0(α1rplastic ) J 0(α1r) − J 0(α1rplastic )Y 0(α1r )]dr =
a
r
b ∞

∫ r∑ B [Y0(α r
a n =1
n n plastic ) J 0(α n r) − J 0(α n rplastic )Y 0(α n r )] ⋅ Equation 4.34

[Y 0(α1rplastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1r )]dr

By the properties of cylinder functions, the right hand side of the equation will vanish in all cases
except where αn = 1:

b
rcone I r 0.333
4su ∫ r ln( )[Y 0(α1 rplastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1 r )]dr =
r
a
Equation 4.35
b

B1 ∫ r[Y 0(α1rplastic ) J 0(α1 r ) − J 0(α1 rplastic )Y 0(α1r )] 2 dr

a

Solve for B1 to Bn:

b
rcone I r 0.333
4su ∫ r ln( )[Y 0(α1 rplastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1r )]dr
r
B1 = a
b
Equation 4.36
∫ r[Y 0(α r
a
1 plastic ) J 0(α1r ) − J 0(α1rplastic )Y 0(α1r )] 2 dr

77
Now all unknowns are known and the solution for u = f(r and t) is:

∞
u = ∑ Bn e − cα n t [ −Y 0(α n r ) J 0(α n rplastic ) + Y 0(α n rplastic ) J 0(α n r )]
2
Equation 4.37
n =1

This equation accounts for the pore pressure increase due to spherical cavity expansion. Because the
original equation is a linear equation, the solutions for all sets of initial conditions and boundary
conditions can be summed.
Consequently, the equation is again solved using the boundary and initial conditions for the
shear-induced pore pressure. Assuming a shear zone of radius r = rshear and by the same reasoning
∂v
as before with the following boundary conditions (at r = rshear, u = 0 and at r = rcone, = 0 ):
∂r

∞
u = ∑ An [ −Y 0( βn r ) J 0( βn rshear ) + Y 0( βn rshear ) J 0( βn r )] Equation 4.38
n =1

Using the initial condition (at t=0):

σ vo ' (1 − (0.5OCR) 0.8 )r − σ vo ' (1 − (0.5OCR) 0.8 )rshear

u= Equation 4.39
(rcone − rshear )

Solve for A1 to An:

(σvo [(1 − (05

. OCR) 0.8 )r − σvo [(1 − (05
b ' '
. OCR) 0.8 )rshear
∫ r[ (rcone − rshear )
][Y0(β1rshear ) J 0(β1r) − J 0(β1rshear )Y0(β1r)]dr
A1 = a
b

∫ r[Y0(β r
a
1 shear ) J 0(β1r) − J 0(β1rshear )Y0(β1r)]2 dr

Equation 4.40

The solution for shear-induced pore pressures is then:

∞
u = ∑ An e − cβ n t [ −Y 0( βn r ) J 0( βn rshear ) + Y 0( βn rshear ) J 0( βn r )]
2
Equation 4.41
n =1

The individual solutions can then be summed and solved for a summation of n = 1 to 50 to yield the
complete solution:

78
 ∞
u = ∑ Bn e − cα n t [ −Y 0(α n r ) J 0(α n rplastic ) + Y 0(α n rplastic ) J 0(α n r )] +
2

n =1
∞
Equation 4.42
∑Ae
n =1
n
− c βn 2 t
[ −Y 0( βn r ) J 0( βn rshear ) + Y 0( βn rshear ) J 0( βn r )]

Equation 4.41 represents the analytical solution which governs the radial consolidation of
porewater surrounding an impermeable probe. This is the form of the equation which will be used in
the evaluation of the coefficient of consolidation.

Model Parameters
In its most simple form, the dissipation model requires input values for the effective vertical
stress (σvo'), the hydrostatic pore water pressure (uo), the radius of the cone used (r), the rigidity
index (Ir), the OCR, and the effective stress friction angle (φ'). The actual OCR was reported in the
published literature for all the sites modeled in this study; however, in a situation where the OCR is
not known, it could be calculated from the following approximate equation (Chen and Mayne, 1994;
1995; 1996):

(q T − u2 )
OCR ≅ 0.46 Equation 4.43
σ vo '

where qT = corrected cone tip resistance, u2 = pore pressure measured at the cone shoulder, and σvo’
= effective vertical stress. The undrained shear strength was calculated from the critical-state
Modified Cam Clay model and corresponds to triaxial compression loading in terms of effective
stresses (Wroth and Houlsby, 1985):

su M OCR Λ
= ( ) Equation 4.44
σ vo ' 2 2

where Λ = plastic volumetric strain ratio = 1-Cs/Cc and M = the slope of the critical state line in
triaxial compression:

6 sin φ '
M= Equation 4.45
3 − sin φ '

In all the experimental case studies evaluated, the actual value of φ' was known for each clay;
however, it is interesting to note that for the case when the effective stress friction angle is equal to
30o (assumed value if φ' is unavailable), the equation for undrained shear strength reduces to:

su
= 0.345OCR 0.8 Equation 4.46
σ vo '

79
which is approximately the same result as the empirical trend given by Jamiolkowski et al. (1985).
The undrained shear strength ratio is a function of overconsolidation and will increase as the
overconsolidation increases. This is consistent except in cases where the clay becomes fissured.
Fissures are defined cracks and planes of weakness in a soil, commonly the result of passive failure
due to unloading or desiccation effects, which decrease the ultimate shear strength the clay is
capable of achieving (Marsland and Quarterman, 1982). Fissures can occur at many scales within
the soil, each of which will affect the measurement of properties differently (Figure 4.5). The effects
of fissures and joints in clay reduce the undrained strength of the soil mass. Based on experimental
work by Bishop (1972) and Skempton et al. (1969), a strength reduction factor of 0.5 was chosen as
appropriate for the very heavily fissured clays. Additional work in quantifying the strength
reduction factor is warranted.

Figure 4.5. Scale Effects of Fissuring as Related to Cone Penetration

(Marsland and Quarterman, 1982)

The size of the induced plasticized zone generated by the cone sounding was calculated
according to spherical cavity expansion using Equation 4.5. Note that:

E G
Ir = = Equation 4.47
2(1 + ν ) su su

where E = equivalent elastic Young’s modulus = 2G(1+ν), G = shear modulus, ν = Poisson’s ratio
(= 0.5 for undrained conditions), and su = undrained shear strength. The rigidity index affects both
the size of the plasticized zone as well as the increase in excess pore stress due to changes in
octahedral normal stresses. However, the appropriate value of the rigidity index is ill-defined and
generally quite difficult to assess for a given soil (Mayne, 1995). Consequently, the rigidity index
was taken as an input parameter and multiple dissipation curves are presented based on chosen
values for the rigidity index. Alternately, when the initial generated pore pressure is measured, the
rigidity index can be calculated because all other parameters (umeas, σvo', φ', OCR, and uo) are known.

80
 The next step in the model was to calculate the excess pore pressure generated due to cone
penetration. The octahedral normal component of the pore pressure was calculated using spherical
cavity expansion (Equation 4.10). Pore pressure was assumed to increase only within the plastic
zone; excess pore pressure outside the plastic zone was set equal to zero.
Finally, the magnitude of shear-induced pore pressures for a Type 2 cone was calculated
using the concepts of Cam Clay (Mayne and Bachus, 1988):

OCR Λ
∆u shear = σ vo ' [1 − ( ) ] Equation 4.48
2

where ∆ushear = excess pore pressure generated due to shear stress changes.
The determination of the width of the zone which is affected by the shear stresses is not as
straight-forward as is the determination for the octahedral normal stresses. In the analysis of shear
stresses developed during the driving of piles, Randolph and Wroth (1978, 1979a) show that shear
stress decreases as a function of radial distance away from the pile. Consequently, the increase in
excess pore pressure due to changes in shear stress were assumed to decrease linearly with distance
away from the cone body. The thickness of the shear zone was set at 2 mm; however, the error
surfaces were plotted in order to asses the effect of the width of the shear zone (from 1 mm to 10
mm) on the calculated pore pressure dissipation. The error surfaces will be discussed later in the
chapter.
The final form of the equation for calculating the generated excess pore pressure is as
follows:

4 M OCR Λ OCR Λ
at t = 0 um = [σ vo '⋅ ( ) ]ln I r + σ vo '[1 − ( ) ] + uo Equation 4.49
3 2 2 2

at t = t100 u m = uo Equation 4.50

The initial distribution of the excess pore pressure within the plastic zone was used as the initial
condition to solve the one-dimensional, uncoupled partial differential consolidation equation for
radial drainage as derived previously (Equation 2.9). Boundary conditions assume that there is no
increase in excess pore pressure outside the spherical cavity plastic zone (∆u = 0) and that the cone
body was an impermeable boundary. After the excess pore pressure dissipation was calculated, the
values were added to the hydrostatic pore pressure value at that depth. This was done because the
piezocone measures total pore pressures (hydrostatic plus excess).
The final version of the model used an analytical solution to the consolidation equation to
predict pore pressure as a function of time (Carslaw and Jaeger, 1959; Randolph and Wroth, 1979b).
All calculations were performed using the software package MathCad. The analytical solution was
an improvement over the originally used finite difference scheme for many reasons. First of all, the
forward finite difference solution is severely limited by the stability criterion which requires that
ch∆t/∆r2 < 0.5. This forces small times steps which require excessive computer time. The analytical
solution is not limited by stability concerns, is rapid because pore pressure can be calculated at any
radius and any time, and requires less memory and less time. A typical solution using the finite

81
difference approximation required one hour using a pentium 75 MHz processor, while the analytical
solution required fifteen seconds on the same computer.
The model was run interatively with an input value of ch chosen for the initial run. Based on
the results of the first iteration, a new value for ch was chosen until the deviation of the predicted
dissipation curve was at a minimum from the actual measured dissipation curve as determined using
the sum of the squared errors.

Evaluated Results From Test Sites

Documented porewater pressure dissipation measurements from piezocone soundings and
instrumented driven pile test sites in both normally-consolidated to lightly-overconsolidated and
heavily-overconsolidated clays were chosen for comparison with the prediction of the model. Six
soft clay sites which exhibited normal monotonically-decreasing dissipation curves and nine stiff to
hard clays (3 crustal) which exhibited dilatory pore pressure behavior were chosen.
A listing of the fifteen sites used for evaluation is given in Table 4.1. All model predictions
of the value of the coefficient of consolidation were performed without prior knowledge of either the
laboratory measured values or of previously predicted field values. Table 4.1 shows a compilation
of the relevant soil parameters for each site evaluated by the method including the input data
[hydrostatic pressure (uo), effective overburden stress (σvo'), overconsolidation ratio (OCR), and
effective friction angle (φ')], the cone data [corrected tip resistance (qT), measured pore pressure (u2),
and depth of dissipation test or piezometer measurements (d)], and the clay characteristics [natural
water content (wn), liquid limit (LL), and plasticity index (PI)]. References for the data are also
listed.

Reference Values of the Coefficient of Consolidation

The most common method to obtain a representative value of the coefficient of consolidation
is from the one-dimensional consolidation test. The procedure is a fitting method, based on either a
logarithmic time or Taylor square root of time plot of the rate of compression data for a clay in an
oedometer test. The measured value is the vertical coefficient of consolidation because drainage in
the consolidation test is limited to the vertical direction. In reality, the value of cv obtained from
consolidation testing is somewhat arbitrary due to the empirical steps in the fitting, to the
dependence on load increment, to the small sample size, and due to the one-dimensional limitations
imposed on the test. While measurements of the coefficient of consolidation are useful for
estimating rate of settlement and for correlating with the hydraulic conductivity, the values are only
approximate.
In an extensive study on the permeability characteristics of soft clays, Tavenas et al. (1983a;
1983b) noted that the assumptions of constant permeability, compressibility, and coefficient of
consolidation in Terzaghi's one-dimensional consolidation theory are not accurate for a soil
exhibiting a changing void ratio. Numerical modeling which accounted for the changes in
permeability and compressibility during a consolidation step showed a value of cv which was 4.4
times larger at the beginning of the load step than it was at the end of the step (Tavenas et al., 1983a;
1983b). The study results call into question the reliability of the values for the coefficient of
consolidation obtained from empirical construction techniques on one-dimensional consolidation
curves. Several researchers have shown that the measured laboratory values of the coefficient of
consolidation can vary by several orders of magnitude from the field values which were determined
from back-calculation of settlement data (Rowe and Barden, 1966; Rowe, 1972; Grisak, 1975;
Grisak and Cherry, 1975).

82
 Table 4.1. Clay Sites Used For Dissipation Modeling
Test Site Depth uo σvo' OCR wn LL PI φ' su qT u2 Reference
(m) (kPa) (kPa) (kPa) (kPa) (kPa)
Bothkennar, 12.0 107.8 96.2 1.4 55 72 50 33 48 898 499 Jacobs and
U.K. Coutts 1992
Drammen, 19.5 179.0 121.0 1.1 30 20 10 34 52 890 450 Lacasse and
Norway Lunne 1982
McDonald 20.0 181.5 178.5 1.1 36 35 10 35 78 1036 600 Sully, 1991;
Farm, B.C. Robertson et al.,
1988a
Ons∅y, 18.5 159.4 114.5 1.4 60 75 36 34 49 754 450 Lacasse and
Norway Lunne 1982
Porto Tolle, 9.3 81.4 90.6 1.2 36 52 30 30 36 580 250 Battaglio et al.
Italy* 1981
St. Alban, 4.6 40.2 42.6 1.2 54 38 18 27 15 300 160 Roy et al. 1981;
Quebec Roy et al. 1982
Amherst, 3.0 16.7 38.2 7.0 40 38 11 30.5 64 1369 80 Lutenegger
MA (crust) 1995; Lally 1993
Brent Cross, 12.0 115.0 117.8 31.0 29 75 51 20 204 2200 100 Lunne et al. 1985
U.K. Lunne et al.
1986b
Canon's 5.7 46.1 67.9 14.0 30 75 50 22.5 141 2590 436 Jardine and Bond
Park, U.K.** 1989
Cowden, 17.2 95.0 283.4 3.4 15 32 16 24 140 898 575 Lunne et al.
U.K. 1985; Lehane
and Jardine 1994
Madingley, 5.8 37.2 72.8 35.0 31 78 48 26 125 2000 200 Coop and Wroth,
U.K.** 1989; Lunne et
al. 1986b
Raquette 3.0 22.7 44.0 8.0 46 68 48 25 66 1447 317 Lutenegger and
River, NY (crust) Kabir 1987
St Lawrence 6.1 53.0 63.0 3.5 60 50 30 30 59 976 382 Lutenegger and
Seaway, NY (crust) Kabir 1987
Strong Pit, 6.7 0.0 146.3 4.4 15 26 11 35 195 2174 1200 Sully 1991;
B.C. Sully and
Campanella,
1994
Taranto, 9.0 78.5 101.5 26.0 23 60 27 28 532 5451 1818 Pane et al. 1995;
Italy*** Battaglio et al.
1986
*Piezoprobe
**Pile
***Cemented

It is also of interest to note that additional studies regarding the ratio of the horizontal
permeability of clays to the vertical permeability showed that for the natural clays tested, the ratio
ranged from 0.81 to 1.16 with an average value of 1.03 (Tavenas et al., 1983a; 1983b). The authors
and others found significant anisotropy in permeability behavior restricted to a small group of
atypical clays including varved clays and highly stratified deposits (Leroueil and Jamiolkowski,
1991).

Dissipation Sites
The model results for the clay sites are discussed below; summary data for all sites modeled
(normally-consolidated and overconsolidated) are listed in Table 4.1 and Table 4.2. Table 4.2 lists
the output information generated by the ch evaluation. While the method was developed for the
estimation of the coefficient of consolidation, it is also useful to obtain an approximate value of the
rigidity index through magnitude of the initial generated pore pressures. Table 4.2 compares
laboratory data with the method-estimated data as well. In general, the piezocone-evaluated

83
numbers are similar to the laboratory reported numbers measured in the one-dimensional
consolidation test.

Table 4.2. Comparison of Model Predictions and Laboratory Values

Site Depth Ir Piezo Lab Comment Reference
(m) Element Data Measured
(This study) cv (mm2/s)
ch (mm2/s)
Bothkennar, 12.0 100 0.2 0.32 Nash et al. (1992)
U.K. 0.08 - 0.13* Jacobs and Coutts (1992)
Drammen, 19.5 100 0.2 0.53 - 1.52 Lacasse and Lunne (1982)
Norway
McDonald 20.0 200 1.9 1.8 - 5.5* Sully (1991)
Farm, B.C. Robertson et al. (1988a)
Ons∅y, 18.5 100 0.05 0.44 - 0.79** Lacasse and Lunne (1982)
Norway 0.10 - 0.22*
Porto Tolle, 9.3 50 0.2 0.29 - 0.54 15 mm Piezo Battaglio et al. (1981)
Italy 0.20 - 0.35* Probe Battaglio et al. (1986)
Saint Alban, 4.6 200 0.6 0.30 Roy et al. (1981);
Quebec 0.10 - 0.20* Roy et al. (1982)
Amherst, MA 3.0 15 0.4 0.07 - 0.10* Crust DeGroot and Lutenegger (1994)
Lally (1993)
Brent Cross, 12.0 ~10 0.0005 0.008 - 0.03* Lunne et al. (1985)
U.K. Lunne et al. (1986b)
Canon's Park, 5.7 100 0.25 0.01 - 0.03* 102 mm Pile Jardine and Bond (1989)
U.K.
Cowden, U.K. 17.2 200 0.2 0.05 - 0.19* Lunne et al. (1985)
Lehane and Jardine (1994)
Madingley, 5.8 ~10 0.05 0.03 - 0.08* 80 mm Pile Coop and Wroth (1989)
U.K. Lunne et al. (1986b)
Raquette 3.0 50 0.5 0.035 - 0.70 Crust Lutenegger and Kabir (1987)
River, NY
St Lawrence 6.1 50 0.3 0.25 - 0.80 Crust Lutenegger and Kabir (1987)
Seaway, NY 0.07 - 0.10*
Strong Pit, 6.7 200 0.2 0.06 - 0.30* Sully (1991); Sully and Campanella
B.C. (1994)
Taranto, Italy 9.0 200 0.4 0.10 - 0.25 Cemented Battaglio et al. (1986)
Clay Bruzzi and Battaglio (1987)
*Data collected by Robertson et al. (1992) in a review of methods of coefficient of consolidation prediction from dissipation testing.
**Results from horizontal oedometer test.

Soft Clay Sites

Six soft clay sites were chosen from the literature for dissipation modeling. The sites
examined included Bothkennar, Drammen, McDonald Farm, Ons∅y, Porto Tolle, and St. Alban.
Results from the evaluation are shown in Figure 4.6 through Figure 4.11. For the most part, the
estimated value of the coefficient of consolidation reasonably approximated the laboratory measured
value, and are typically within one-half to two times the laboratory measured value of cv. An
interesting feature to note in the evaluation of the dissipation curves is the good agreement in the fit
of the estimated and the measured dissipation curves.

84
 500 Measured
uo
Ir = 50
400

Pore Pressure (kPa)

Ir = 100
Ir = 200

300
Bothkennar
Depth = 12 m
200 Shear Zone = 2 mm

100
ch = 0.2 mm2/s
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.6. Model Predictions at Bothkennar, U.K.

(Data from Jacobs and Coutts, 1992)

700 Measured
uo
600 Ir = 50
Pore Pressure (kPa)

Ir = 100
500 Ir = 200

400

300

200
Drammen
100 Shear Zone = 2 mm
Depth = 19.5 m ch = 0.2 mm2/s
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.7. Model Predictions at Drammen, Norway

(Data from Lacasse and Lunne, 1982)

85
 900 Measured
uo
800 Ir = 50

Pore Pressure (kPa)

700 Ir = 100
Ir = 200
600
500
McDonald Farm
400 Depth = 20 m
300 Shear Zone = 2 mm

200
100
ch = 1.9 mm2/s
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.8. Model Predictions at McDonald Farm, B.C.

(Data from Sully, 1991)

Measured
600
uo
Ir = 50
500
Pore Pressure (kPa)

Ir = 100
Ir = 200
400

300

200

100 Onsoy
Depth = 18.5 m
Shear Zone = 2 mm ch = 0.05 mm2/s
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.9. Model Predictions at Onsoy, Norway

(Data from Lacasse and Lunne, 1982)

86
 400 Measured
uo
Ir = 50

Pore Pressure (kPa)

Ir = 100
300 Ir = 200

200 Porto Tolle

Depth = 9.3 m
Shear Zone = 2 mm
Piezo Probe 15mm dia
100
ch = 0.2 mm2/s

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.10. Model Predictions at Porto Tolle, Italy

(Piezoprobe Data from Battaglio et al., 1981)

200 Measured
uo
Ir = 50
Pore Pressure (kPa)

Ir = 100
150 Ir = 200

100
St. Alban
Depth = 4.6 m
Shear Zone = 2 mm
50
ch = 0.6 mm2/s

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.11. Model Predictions at Saint Alban

(Data from Roy et al., 1982)

87
 Stiff Clay Sites
Nine hard to stiff clay sites were chosen from the literature for dissipation modeling in cases
where the pore pressures values demonstrated the dilatory behavior as a function of time in either
piezocone dissipation records or instrumented pile pore pressure transducer records. The sites
evaluated included Brent Cross, Canon's Park, Cowden, Madingley, Taranto and the crustal layers at
Amherst, Raquette River, and St. Lawrence Seaway. For dissipation in stiff clays, the model input
parameters were the same as in the soft clay sites.
In the case of overconsolidated clays (OCR > ~3), the shear-induced pore pressures can go
negative and force a draw-down in the value of pore pressure close to the cone body; however, the
shear-induced pore pressures are limited to a thin zone adjacent to the cone body. Because this
creates a large gradient, these pore pressures can dissipate more rapidly, allowing the pore pressure
to return to the value induced by octahedral normal stresses. In the field, this phenomenon is
observed when testing with a dual-element piezocone in overconsolidated clays. The recorded Type
1 pore pressures begin positive and decay, while the Type 2 pore pressures begin negative, then
increase, and then decay monotonically along the same line as the Type 1 pore pressures (Chen and
Mayne, 1994).
Results from the evaluation at the stiff clay sites are shown in Figure 4.12 through Figure
4.20. While the estimated values of the coefficient of consolidation at the sites are reasonable with
respect to the laboratory values, there is a bias toward estimating a faster rate of dissipation, or a
higher value of cv, in the heavily-overconsolidated clays than was measured in laboratory testing.
Additionally, the shape of the dissipation curves is also an important factor, especially for the
transient-increasing behavior. Several of the sites have dissipation curves which match the predicted
shape very well; however, some of the sites have significant variations, especially in the initial
portion of the curves, which indicates that there are other additional factors which influence the
dissipation process . Nonetheless, the results are promising and are worthy of further exploration.

250
Measured
uo
200
Pore Pressure (kPa)

Ir = 15
Ir = 50
Ir = 100
150

ch = 0.4 mm2/s
100
CVVC - Amherst, MA
Depth = 3 m
50 Shear Zone = 2 mm

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.12. Model Prediction in OC Crust at Amherst, MA

(Data from Lally, 1993)
88
 Measured
1200 Hydrostatic

1000 Brent Cross Ir = 10

Depth = 12 m Ir = 15
Pore Pressure (kPa)

800 Shear Zone = 2 mm Ir = 50

600
400
200
0
-200
ch = 0.0005 mm2/s
-400
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.13. Model Prediction at Brent Cross, U.K.

(Data from Lunne et al., 1985)

89
 Measured
600 uo
ch = 0.25 mm2/s Ir = 50
500 Ir = 100

Pore Pressure (kPa)

Ir = 200

400

300
Canon's Park
Depth = 5.7 m
200 Shear Zone = 2 mm
Pile: 102 mm dia
100

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.14. Model Prediction at Canon's Park, U.K.

(Instrumented Pile Data from Bond and Jardine, 1991)

1200 Measured
uo
Ir = 50
1000 Ir = 100
Pore Pressure (kPa)

Ir = 200
800

600

Cowden
400 Depth = 17.2 m
Shear Zone = 2 mm
ch = 0.2 mm2/s
200

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.15. Model Predictions at Cowden, U.K.

(Data from Lunne et al., 1985)

90
 900 Measured
Madingley
uo Depth = 5.8 m
800 Shear Zone = 2 mm
Ir = 10
700 Pile: 80 mm dia

Pore Pressure (kPa)

Ir = 15
600 Ir = 25

500
400
300
200
100
0 ch = 0.05 mm2/s
-100
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.16. Model Predictions at Madingley, U.K.

(Instrumented Pile Data from Coop and Wroth, 1989)

500 Measured
Hydrostatic
450 Ir = 50
400
Pore Pressure (kPa)

Ir = 100
Ir =200
350
300
250 ch = 0.5 mm2/s
200
150 Raquette River Cemetery
100 Depth = 3 m
Shear Zone = 2 mm
50
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.17. Model Prediction in Crustal Clay at Raquette River Cemetery, NY

(Data from Lutenegger, 1997)

91
 500 Measured
Hydrostatic
450 Ir = 50
400

Pore Pressure (kPa)

Ir = 100
Ir = 200
350
300 ch = 0.3 mm2/s
250
200 SLS
150 Depth = 6.1 m
100 Shear Zone = 2 mm

50
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.18. Model Prediction in Crustal Clay at Saint Lawrence Seaway, NY

(Data from Lutenegger, 1997)

1400 Measured
Ir = 50
1200 Ir = 100
Pore Pressure (kPa)

Ir = 200
1000

800 Strong Pit

Depth = 6.65 m
600 Shear Zone = 2 mm

400
ch = 0.2 mm2/s
200

0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.19. Model Prediction at Strong Pit, Canada

(Data from Sully, 1991)

92
 2000 Measured
uo
1800 Ir = 50
1600

Pore Pressure (kPa)

Ir = 100
Ir = 200
1400
1200
1000
800
600 Taranto
400 Depth = 9 m
Shear Zone = 2 mm ch = 0.4 mm2/s
200
0
0.1 1 10 100 1000 10000 100000
Time (sec)

Figure 4.20. Model Prediction in Cemented Clay at Taranto, Italy

(Data from Bruzzi and Battaglio, 1987)

Error Surfaces for Model Parameters

A study was performed in order to identify the effect of the model input parameters on the
predicted results. Specifically, the error, or the residual, was calculated between each predicted and
actual measurement by the following (Santamarina and Fratta, 1997):

ei = y i ( meas ) − yi ( pred ) Equation 4.51

where ei = the error, yi(meas) = the measured value, and yi(pred) = the model predicted value. The error
norm was then calculated as the sum of the squared errors using the following equations:

1
 n n
L n =  ∑ ei  Equation 4.52
 i 

1
 2 2
L2 = ∑ ei  Equation 4.53
 i 

where L2 = error norm with n = 2. The best fit for the model parameters was then achieved by
varying one parameter until the error converged to a minimum. The L2 error norm uses a least
squares minimization, and assumes that the measured data follow a Gaussian distribution. The input
parameters which were varied included the width of the shear zone, the coefficient of consolidation,
93
the rigidity index, the effective friction angle, and the OCR. The error norms are plotted versus a
reasonable operating range in engineering practice for the parameters investigated.
Figure 4.21 and Figure 4.22 show the resulting error norms from changing the width of the
zone of soil affected by the shearing action due to the soil and penetrometer interaction in a soft clay
(Bothkennar) and in a hard clay (Taranto), respectively. Each point on the graphs represents a
different run of the model. The results show clearly that the width of the shear zone has little effect
in the soft clay where the contribution of shear stresses are less significant than in the stiff clay. In
the soft clay, the sum of the squared errors remained essentially constant. However, the width of the
shear zone in the hard clay had a more significant effect, especially as the width of the shear zone
was increased over 2 mm. The calculated error norm reached a minimum at 2 mm, and increased as
the width of the shear zone was increased.
Variation of the shear zone width has several physical implications for the dissipation of
excess pore pressure. First, the width of the shear zone will affect the magnitude and distribution of
shear-induced excess pore pressures as a function of distance from the cone. Pile foundation studies
show that the shear-induced pore pressures decrease as a function of distance from the penetrometer
(Randolph and Wroth, 1979a); therefore, a narrow shear zone will concentrate the increase in pore
pressures due to shear stresses adjacent to the cone body. Changing the width of the shear zone also
affects the dissipation behavior surrounding a penetrometer. Figure 4.23 illustrates the zones of soil
surrounding the penetrometer which are affected by the soil/probe interaction. A narrow shear zone
will have a large increase in total pore pressure adjacent to the cone body which will drop
significantly outside the zone of shear-induced pore pressures. Consequently, a narrow shear zone
will differ from a wider one in dissipation behavior due to the initial distribution of pore pressures.
In the more narrow zone, the shear-induced pore pressures are concentrated in a small band which
creates a larger gradient (same magnitude of shear-induced pore pressure over a shorter

3000
Bothkennar
2500

2000
Error

1500

1000

500

0
0 1 2 3 4 5 6 7 8 9 10
Shear Zone (mm)

Figure 4.21. Error Norm for the Width of Shear Zone in a Soft Clay

94
 3000

2500

2000
Error

1500

1000

500 Taranto

0
0 1 2 3 4 5 6 7 8 9 10
Shear Zone (mm)

Figure 4.22. Error Norm for the Width of Shear Zone in a Hard Clay

dr ~ 1 to 10 mm

Virgin Plasticized
Zone
Region Zone due to
affected by
Cavity ro
E, ν shear
Expansion

su Shoulder Position, u2
Measurement rp = roIr0.3333
of Porewater
Pressures

Figure 4.23. Zones Affected By Cone Penetration

distance) and allows for more rapid dissipation of pore pressures. After the initial rapid dissipation
of the shear-induced pore pressures, the dissipation curves all follow essentially the same pattern.

95
Based on the results of the parametric study for the sites evaluated in this study, a shear zone with a
width of 2 mm was chosen as the value which best estimated the experimental data. However, the
width of shear is probably variable and affected by test strain rate, which would produce a different
width zone for a cone penetration sounding versus a jacked pile.
The error norms which resulted from variation of the value of the coefficient of consolidation
were also plotted. Figure 4.24 shows the effect of varying ch over approximately two orders of
magnitude in the Bothkennar clay and Figure 4.25 shows the same for the Taranto clay. In both
cases, the value of the coefficient of consolidation was easily identified; however, in the Taranto
case, steep gradients of approach of the error from both the lower and higher values of the parameter
were observed. For the same level of error, the range in the value of the coefficient of consolidation
is much larger for the soft clay than it is for the hard clay, indicating that the model is more sensitive
to variation in the parameter in the case of the hard clay.
The error norms for the rigidity index in both the soft and hard clay deposits are shown in
Figure 4.26 and Figure 4.27. In both cases, variation of the value of the rigidity index significantly
affected to calculated error norm for magnitudes less than the minimum error, but had a less
significant effect at values larger than the minimum error. Again, as was the case for the coefficient
of consolidation, the hard clay site was more sensitive to variation in the parameter than was the soft
clay site.

96
 3000
Bothkennar
2500

Error 2000

1500

1000

500

0
0.01 0.1 1 10
ch (mm2/s)

Figure 4.24. Error Norm for the Coefficient of Consolidation in a Soft Clay

3000

2500

2000
Error

1500

1000

500
Taranto
0
0.01 0.1 1 10
ch (mm2/s)

Figure 4.25. Error Norm for the Coefficient of Consolidation in a Hard Clay

97
 3000

2500
Bothkennar
Error 2000

1500

1000

500

0
0 100 200 300 400 500 600
Ir

Figure 4.26. Error Norm for the Rigidity Index in a Soft Clay

3000

2500

2000
Error

1500

1000

500
Taranto
0
0 100 200 300 400 500
Ir

Figure 4.27. Error Norm for the Rigidity Index in a Hard Clay

98
 The error norms for the effective friction angle and OCR are shown in Figure 4.28 through
Figure 4.31. Again, the data from the Bothkennar and Taranto sites were used for comparison. Of
all the parameters examined for the hard clays, the changes produced by these two parameters were
the most significant. The error norms for the friction angle and OCR in the hard clay show steep
gradients moving away from the error minimum, showing that the model is sensitive to the input
value of these parameters. Similar to the trends seen previously, the variation of the effective stress
angle and OCR are less significant for the soft clay, producing a smaller level of error for the same
variation, than was seen in the hard clay. The significant effect on the prediction of the effective
stress friction angle and OCR makes the initial evaluation of these parameters of critical importance
in the estimation of the coefficient of consolidation, primarily because they represent the undrained
strength and initial penetration porewater pressure regime used in the evaluation.

Comparison With Existing Solutions

A comparison between the model estimated values and the solution proposed by Teh and
Houlsby (1988) was performed in order to evaluate the method. The soft clay sites were used as the
basis of comparison because due to the difficulty in assessing t50 for the dilatory response in hard
clays. The method of Teh and Houlsby was chosen for comparison because it appears to be the most
widely-used solution for piezocone dissipation at present (Robertson et al., 1992). The results of the
comparison are shown in Figure 4.32; it is interesting to note that the new method predicts lower
values in each instance. In general, neither method appears to have a distinct advantage in the
prediction of ch, however.

Normalized Dissipation Data

The evaluated dissipation results can also be presented as a series of normalized curves,
similar to those given by Teh and Houlsby (1991). Figure 4.33 through Figure 4.35 show the
normalized dissipation curves estimated for values of φ' equal to 20o, 30o, and 40o. As seen in the
figures, the lower value of φ' leads to more significant differences in behavior for different values of
OCR. This is because the lower value of the friction angle leads to a smaller initial magnitude of
pore pressure, and a more rapid decay of the pressures when the values are normalized to the initial
value.

99
 3000

2500 Bothkennar

Error 2000

1500

1000

500

0
10 20 30 40 50
Effective Friction Angle (degrees)

Figure 4.28. Error Norm for the Effective Stress Friction Angle in a Soft Clay

3000

2500

2000
Error

1500

1000

500
Taranto
0
10 20 30 40 50
Effective Friction Angle (degrees)

Figure 4.29. Error Norm for the Effective Stress Friction Angle in a Hard Clay

100
 3000

Bothkennar
2500

Error 2000

1500

1000

500

0
1 1.2 1.4 1.6 1.8 2
OCR

Figure 4.30. Error Norm for the OCR in a Soft Clay

3000

2500

2000
Error

1500

1000

500
Taranto
0
0 10 20 30 40 50
OCR

Figure 4.31. Error Norm for the OCR in a Hard Clay

101
 10

This Study Estimated ch (mm2/s)

1
Bothkennar
Drammen
McDonald Farm
Onsoy
0.1 Porto Tolle
St Alban
1 to 1

0.01
0.01 0.1 1 10
2
Teh and Houlsby Estimated ch (mm /s)

Figure 4.32. Model Comparison with Existing Methods

1.6
1.4
Normalized Pore Pressure

OCR=1
1.2 OCR=2
OCR = 3
1.0 OCR = 5
0.8 OCR = 10
OCR = 20
0.6 OCR = 100
o
0.4 φ' = 20

0.2
0.0
1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01
T*= [cht / (r2Ir0.5)]

Figure 4.33. Normalized Dissipation Curves for φ' =20o

102
 1.4

1.2

Normalized Pore Pressure

OCR=1
OCR=2
1.0
OCR = 3
OCR = 5
0.8
OCR = 10
0.6 OCR = 20
OCR = 100
0.4 φ' = 30 o

0.2

0.0
1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01
T*= [cht / (r2Ir0.5)]

Figure 4.34. Normalized Dissipation Curves for φ' =30o

1.4

1.2
Normalized Pore Pressure

OCR=1
OCR=2
1.0 OCR = 3
OCR = 5
0.8
OCR = 10
0.6 OCR = 20
OCR = 100
0.4 φ' = 40 o

0.2

0.0
1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01
T*= [cht / (r2Ir0.5)]

Figure 4.35. Normalized Dissipation Curves for φ' =40o

103
 Discussion
A hybrid cavity expansion model was chosen for the evaluation of the coefficient of
consolidation. Extensive experimental evidence suggests that the dissipation of pore pressures
surrounding a Type 2 piezocone is controlled by horizontal drainage; consequently, the
consolidation equation was solved assuming only radial drainage, or a cylindrical cavity expansion
model. However, the most reliable evaluation of the coefficient of consolidation was achieved by
assuming spherical cavity expansion for the initial generated magnitude of pore pressure and for the
radius of the zone of disturbance. In actuality, the Type 2 location of the pore pressure element, just
behind the cone tip, is controlled by large stress gradients which are not modeled exactly by either
spherical or cylindrical cavity expansion.
The interesting feature of this model's approach is that pore pressures can be either positive
or negative. Negative penetration pore pressures are seen in heavily-overconsolidated clays when
the pore pressure is measured with the filter in the Type 2 position. Previous models have either
failed to provide a coherent reason for why the pore pressure might dissipate in this atypical manner,
or have chosen to neglect the phenomenon completely.

Conclusions
A cavity expansion and critical-state model was introduced for evaluating the horizontal
coefficient of consolidation in clays ranging from soft to stiff to hard clays. The scheme is able to
predict standard monotonic dissipation characteristic of soft to firm normally- to lightly-
overconsolidated clays, as well as the dilatory dissipation behavior observed in heavily-
overconsolidated clays. The formulation incorporates a component of shear-induced pore pressure
which can be either positive at low OCRs or negative at higher values of OCR. A comparison
between the predicted and lab-measured values of the coefficient was shown in Table 4.2. The fit is
within range of measured values, especially considering the approximate nature of the laboratory
estimated values of coefficient of consolidation. While the prediction of the shape of the dissipation
curve is often quite good, there still remain cases where the dissipation behavior deviates from the
predicted behavior. This indicates that there still may be some physical phenomena occurring which
the model does not take into account. These effects can include smear, soil fabric, sensitivity,
fissures, and hydraulic fracturing; consequently, future development should investigate these
aspects.
Additional complicating aspects to the prediction include the fact that reference
measurements for the coefficient of consolidation are typically obtained in the one-dimensional
consolidation test which is a measure of the vertical value, rather than the horizontal value of
coefficient of consolidation. Also, saturation of the pore pressure filter element and pore pressure
transducer chamber should be performed meticulously in order to eliminate the trapping of air
bubbles. Poorly saturated filter elements can produce pore pressure dissipation curves which mimic
the dilatory behavior of pore pressure seen in heavily-overconsolidated clays, so data integrity must
be ascertained before predictions can be considered valid.
An interesting facet is that the rigidity index (Ir) of the clay is evaluated by the analytical
method. The rigidity index was taken as an input value for the model and varied in order to yield a
series of curves for each dissipation test. At low values of OCR, the value for rigidity index which
most closely models the dissipation curve is approximately 100, while at higher OCRs (7 to 26), the
most appropriate value for rigidity index is on the order of 50 or less. This is consistent with
existing laboratory triaxial data which show Ir decreasing as a function of increasing OCR and
increasing plasticity index (Keaveny and Mitchell, 1986; Kulhawy and Mayne, 1990). For
uncemented clays, a graph of rigidity index (curves from Keaveny and Mitchell, 1986) with the
model-evaluated values superimposed is shown in Figure 4.36.
104
 It is also important to note that the model prediction is sensitive to the input values of φ' and
OCR. In this formulation, the initial magnitude of pore pressure was calculated assuming the
triaxial compression mode of failure. However, failure modes surrounding a cone penetrometer are
probably a hybrid value between triaxial compression under the tip and direct simple shear along the
sides. Nonetheless, the use of triaxial compression mode of failure is a reasonable approximation for
the test conditions. Notably, the additional influences of anisotropic stress state, stress rotation
effects, and strain rate have not been addressed here, but may influence the results. These issues
should be addresses in future studies.

300

Bothkennar
PI Drammen
10 McDonald Farm
Onsoy
Porto Tolle
St Alban
200 Amherst
Rigidity Index

Brent Cross
Canon's Park
Cowden
20 Madingley
Raquette River
St. Lawrence
Strong Pit
100 30

40

50

0
1 10 100
OCR

Figure 4.36. Model Evaluation of Rigidity Index in Non-Cemented Clays

Note: Plasticity Index Curves from Keaveny and Mitchell (1986)

The final predictive form for obtaining the coefficient of consolidation from piezocone
dissipation test data is:

0.5
T * r2Ir
ch = Equation 4.54
t

where the modified time factor T* is given as a function of both φ' and OCR (Figures 4.33 through
4.35).

105
 CHAPTER 5

INTERPRETATION OF SEISMIC PIEZOCONE RESULTS FOR THE

EVALUATION OF SOIL PERMEABILITY IN CLAYS

Introduction
One of the most critical parameters for soil characterization in geoenvironmental engineering
is the permeability of a soil deposit. Permeability controls the flow of groundwater through the
subsurface, and consequently, the advective transport of contaminating chemicals. However,
quantification of the soil permeability is a most difficult task because its value can change orders of
magnitude as a function of grain size (Mitchell, 1993). This section seeks to develop a methodology
to evaluate the in-situ soil permeability based solely on seismic piezocone test data.
The addition of an accelerometer or geophone to a cone penetrometer enables the direct
measurement of the soil shear wave velocity (Vs) with depth during a seismic piezocone sounding.
Conventional applications of shear wave velocity data include the evaluation of dynamic stiffness
(low-strain shear modulus, Gmax = Go = ρVs2, where ρ = mass density) for dealing with foundation
vibration problems and the evaluation of site-specific amplification response spectra for earthquake
events. Figure 5.1 shows seismic piezocone tests being performed for the Missouri Department of
Transportation in an evaluation of bridge stability in the event of an earthquake.
In a different application of the shear wave velocity, this section seeks to use the
measurements obtained during a seismic piezocone test to develop predictive relationships between
Vs and the mass density and constrained modulus of a soil deposit. Using these correlations in
combination with the value of the coefficient of consolidation estimated in Chapter 4, a methodology
is developed to evaluate the permeability of a soil deposit using data gathered solely from the
seismic piezocone test.

Seismic Cone Penetration Test

The seismic cone penetration test is a rapid and economical method for obtaining shear wave
velocity profiles in a downhole manner at a given site. Seismic cone penetrometers are instrumented
with either one or multiple accelerometers or geophones in the cone body to measure shear wave
arrival times emanating from a distant source.
The geophysical portion of the seismic cone test is typically performed as a downhole test
with the wave source generated at the surface (Campanella, 1994). A typical test setup is shown in
Figure 5.2. Surveys via the crosshole testing method (ASTM D-4428) are also possible (Baldi et al.,
1989). The seismic source and trigger used in the downhole surveys performed by Georgia Tech
consists of an instrumented sledge hammer which is struck against the side of a plank that is firmly
coupled to the ground by the weight of a truck. Leads from the geophone and the hammer are
connected to a Hewlett Packard Model 54601-A oscilloscope.
During the pause in cone penetration while each successive rod is added, the hammer is
struck and the travel time response is monitored on the oscilloscope. The incremental distance over
incremental time is used to calculate shear wave velocity, Vs = ∆x/∆t, providing a pseudo-interval
approach (Sully and Campanella, 1995). A true-interval approach can be obtained if two
accelerometers are located a fixed distance apart within the cone penetrometer (Burghignoli et al.,
1991). Several tests, taking only about 15 seconds each for trigger and register, are made at each
depth for the pseudo-interval approach in order to verify repeatability of the measurements. A
reverse strike may also be used for comparing polarized wave time records.

106
 Figure 5.1. Seismic Piezocone Testing for Missouri DOT in
New Madrid Earthquake Region

Oscilloscope

Source

Rods and
Cable Enlarged View

Velocity
Transducer
Horizontally Vertical
Polarized Inclinometer
Shear Waves
fs Cone Diameter
Seismic Cone 35.7 mm or
Penetrometer 43.7 mm
u2
60o
2 cm/sec qt

Figure 5.2. Configuration For Seismic Cone Test

Seismic cone penetration tests were performed at a variety of locations throughout the
country (Burns and Mayne, 1994; Burns et al., 1994; Burns and Mayne, 1995b; Mayne et al., 1996).
Figure 5.3 and Figure 5.4 show seismic cone test results obtained near San Manuel, Arizona. The
sounding was performed in mine tailings built from the spoils of a copper mining operation. These
tailings consist of approximately 20-60% fines (“rock flour”) forming silty sands and sandy silts.
The tailings are deposited in large cycloned dams for dewatering and disposal purposes. As would
107
be anticipated in these heterogeneous mining materials where the depositional process was turbulent,
the shear wave velocity profiles for the dam are somewhat variable, as are the related profiles of qc,
fs, and u2.

Cone Tip Resistance, qT Sleeve Friction, f s Pore Pressure, u2 Shear Wave Velocity, V s
(MPa) (kPa) (kPa) (m /s)
0 5 10 0 50 100 150 0 500 1000 0 50 100
0 0 0 0

5 5 5 5

10 10 10 10
Depth (m)

Depth (m)

Depth (m)

Depth (m)
15 15 15 15

20 20 20 20

25 25 25 25

30 30 30 30

Figure 5.3. Seismic Cone Test in Copper Mine Tailings at San Manuel, Arizona (Burns and
Mayne, 1994)

Figure 5.4. Seismic Cone Testing at San Manuel, AZ

108
 Figure 5.5 and Figure 5.6 show the derived shear wave velocity profile obtained on
Hutchinson Island in Savannah, Georgia (Burns and Mayne, 1995b). The soil overburden consists
of Atlantic coastal plain sediments with interbedded layers of soft plastic clay and loose clayey sand.
A rather uniform profile of shear wave velocity measurements was obtained at this site which is
indicative of the slow sedimentation process that formed these marine deposits.
Measurement of shear wave velocity using the seismic piezocone also enables the calculation
of the low-amplitude shear modulus, Gmax, if the unit weight is known:

G max = ρVs 2 Equation 5.1

where ρ = mass density = γt/g, γt = total unit weight, and g = gravitational acceleration constant (=
9.8 m/sec2; = 32.2 ft/sec2). Values of unit weight for real and natural soils are quite varied and may
range anywhere from as low as 10 kN/m3 (64 pcf) for very soft peats and organic plastic clays to as
high as 21 kN/m3 (134 pcf) for dense glacial tills. With clean sands, it may be possible to infer the unit
weight through an evaluation of relative density (Dr) from the cone tip resistance (Robertson and
Campanella, 1983; Kulhawy and Mayne, 1990; Lunne et al., 1992). However, in clays and silts, the
dependence of qc on void ratio is very small (Mayne and Rix, 1993) and therefore not directly
applicable to evaluating either unit weight or mass density of these materials. Consequently, more
systematic approaches other than estimating ρ for soils are desirable and explored in this chapter.
For this purpose, a database from all types of geomaterials (clays to sands to rocks) has been created
and a global correlation between mass density and shear wave velocity was developed.

Cone Tip Resistance, qT Sleeve Friction, fs Pore Pressure, u 2 Shear Wave Velocity, Vs
(MPa) (kPa) (kPa) (m /s)
0 10 20 0 50 100 0 250 500 0 100 200
0 0 0 0

2 2 2 2

4 4 4 4

6 6 6 6
Depth (m)

Depth (m)

Depth (m)

Depth (m)

8 8 8 8

10 10 10 10

12 12 12 12

14 14 14 14

16 16 16 16

Figure 5.5. Seismic Cone Test in Atlantic Coastal Plain Sediments at Hutchinson Island,
Georgia (Burns and Mayne, 1995b)

109
 Figure 5.6. Seismic Cone Testing at Hutchinson Island, GA

Vs-qT Relationships in Soils

Because both the cone tip resistance and shear wave velocity depend on the effective
geostatic state of stress, it is possible to correlate the two parameters; however, the relationship is
soil type dependent. For instance, Baldi et al. (1989) showed the existence of a Vs-qT correlation for
clean quartz sands is given by the following:

Vs = 277( q T ) 0.13 (σ vo ' ) 0.27 Equation 5.2

qT and Vs from SCPT tests can be used to evaluate eo in intact clays:

where Vs is in m/s, qT is in MPa, and σvo' = effective overburden stress in MPa.
Mayne and Rix (1995) compiled a database of 31 different clay sites including intact and
fissured clays with varied plasticity characteristics (8 < PI < 300), sensitivities (2 < St< 200+), and
overconsolidation stress states (1 < OCR < 100+). Approximate estimates of Vs solely as a function
of qT can be made; however, the correlation is very much improved when the void ratio is also
included as a correlative parameter. Using data from intact clays only, the following correlation
between Vs, qT, and eo was developed (n = 339, r2 = 0.832):

q T 0.435
Vs = 9.44 0.532 Equation 5.3
eo

where qT is in kPa and Vs is in m/s. Figure 5.7 shows the multiple regression relationship with data
superimposed for comparison. Through inversion of this equation, the independent measurements of

110
 1000
Vs = 9.44(qT)0.435(eo)-0.532
n = 339

In-Situ Shear Wave Velocity, Vs (m/s)

r2 = 0.832

100

Fissured (0.4 < eo < 0.9)

Intact (eo < 1.0)
Intact (1 < eo < 2)
Intact (2 < eo < 4)
Intact (4 < eo < 11)
e = 0.5
e=1
e=2
Note: Vs in meter/sec e=4
qT in kPa e=8
e = 12

10
100 1000 10000
Cone Tip Resistance, qc = qT (kPa)

Figure 5.7. Shear Wave Velocity-Cone Tip Resistance Correlations in Clay

(After Mayne and Rix, 1995)

q T 0.818
eo = 68 1.88 Equation 5.4
Vs

where qT is in kPa and Vs is in m/s. This equation provides an approximate and immediate
estimation of eo in clays where values are not normally known a priori. Furthermore, in saturated
clays, γsat = [(Gs+e)/(1+e)]γw, and a systematic assessment of total unit weight is obtained for
calculating the low-strain shear modulus, Gmax = (γt/g)Vs2.
Two case studies not included in the original clay database were examined in order to verify
the applicability of the void ratio estimations. The first site was St. Alban, Quebec (Lefebvre et al.,
1994), which is underlain by sensitive marine clays that have a relatively high void ratio profile
with values varying between 1 and 3. The groundwater table is approximately 1.2 m deep. Figure
5.8 shows the estimated eo values from the equation and, although the prediction varies from the
actual values somewhat, the increase and subsequent decrease in the eo trend with depth is evident.

111
 Void Ratio, eo
0 1 2 3
0
1
2
3
Depth (m)

4
5
6
7 Estimated
Actual
8
9

Figure 5.8. Void Ratio Prediction at St. Alban Site (Lefebvre et al., 1994)

Void ratio estimations for a site tested with the Georgia Tech seismic cone system are shown
in Figure 5.9 (Burns and Mayne, 1995). The site is located on Hutchinson Island in Savannah,
Georgia and is comprised of Atlantic coastal plain sediments consisting of interbedded layers of
plastic clay and sand (shown previously in Figure 5.5). The groundwater table is approximately 2
meters deep at this site. Again, the void ratio predictions are very approximate, but the general trend
in the prediction is in agreement.

Void Ratio, eo
0 1 2 3 4 5 6
0

5
Depth (m)

10
Estimated
Actual
15

Figure 5.9. Void Ratio Predictions for Hutchinson Island, Savannah, GA

(Burns and Mayne, 1995)

112
Geotechnical Estimation of Mass Density
Often in geotechnical analyses, the magnitude of γt is merely estimated. A more rigorous
approach requires the direct measurement of γt from undisturbed thin-wall samples taken in adjacent
borings; however, this is neither feasible nor practical when using cone penetration tests for
exploration.
A more generalized approach to the ρ estimation problem was sought by compiling data from a
wide variety of geomaterials including: intact rocks, gravels, sands, silts, and clays (Burns and Mayne,
1996). Table 5.1 summarizes the list of soils and rocks where both shear wave velocity measurements
and mass densities of these materials were known. Seismic wave velocities for soils were determined
from in-situ methods (seismic cone, crosshole, spectral analysis of surface waves, or suspension
logging techniques). The unit weights and mass densities (ρ = γ/g) of these materials were determined
from thin-wall tube, piston, or core samples taken in either the same or adjacent boreholes. Data for
clays were summarized by Mayne and Rix (1995) and data for sands were compiled by Hegazy and
Mayne (1995). For the rock materials, the data were obtained generally from laboratory testing on
small specimens, usually with ultrasonic techniques applied to determine Vs.
A correlation was observed between the mass density of the geomaterial and the measured
shear wave velocity. Figure 5.10 shows the trend and resulting statistics from a regression analysis of
the data (n = 438; r2 = 0.765; S.E. = 0.126):

ρ = 0.277 + 0.648 logVs Equation 5.5

where ρ = mass density (g/cc), Vs = shear wave velocity (m/s), n = number of data sets, r2 = coefficient
of determination, and S.E. = standard error of the independent variable (same units as ρ). Since Vs is
recognized to increase with increasing effective overburden stress, a multiple regression analysis was
also investigated with two independent variables; the results are summarized in Figure 5.11. The
statistical results showed an improved trend (n = 438; r2 = 0.820; S.E. = 0.118):

Vs 0.227
ρ = 0.701 Equation 5.6
(σ vo ' ) 0.057

where σvo' = effective vertical overburden stress (kN/m2). However, an iterative approach is required
for this equation since σvo' depends upon a knowledge of the mass density profile with depth before
calculation. As a consequence, the aforementioned equation may be of more practical usage in routine
explorations involving the need to convert Vs to Gmax. The correlation will also be of immediate value
to Vs surveys conducted using the spectral analysis of surface waves (SASW) since no other
information may be available before this test is conducted.

113
 Table 5.1. Geomaterials Database - Known Mass Density and Shear Wave Velocity
Site/Location Geomaterial Type Test Reference Sources of Data
Method
Atlanta, GA residual silty sand SASW Mayne and Harris (1993)
Bäckebol, Sweden soft postglacial clay SCPT Larsson and Mulabdic (1991)
Bagdad, AZ mine tailings SCPT Burns et al. (1994)
Berea sandstone, CA sandstone Lab Hilbert et al. (1994)
Bothkennar, UK soft organic clay SCPT Hepton (1988)
Brent Cross, UK fissured London clay SASW Lunne et al. (1986b)
Chase Group C, KS siltstone Lab Yale and Jamieson (1994)
Chase Group A, KS dolostone Lab Yale and Jamieson (1994)
Chase Group B, KS limestone Lab Yale and Jamieson (1994)
DOE site, SC clayey sand SCPT Bratton et al. (1993)
Drammen, Norway sensitive clay SCPT Campanella et al. (1986b)
Fucino, Italy soft cemented clay SCPT, CHT Burghignoli et al. (1991)
Holmen, Norway clean sand SCPT Lunne et al. (1986b)
Houston, TX stiff Beaumont clay CHT Mahar and O'Neill (1983)
Langley, BC soft glacio-marine clay SCPT Grieg et al. (1988)
Lilla Mellösa, Sweden soft organic clay SCPT Larsson and Mulabdic (1991a)
Long Beach A, CA clean sand & gravel CHT McLamore et al. (1978)
Long Beach B, CA clean sand & gravel CHT McLamore et al. (1978)
Long Beach C, CA stiff lean clay CHT McLamore et al. (1978)
Long Beach C, CA silty sand CHT McLamore et al. (1978)
Long Beach D, CA stiff plastic clay CHT McLamore et al. (1978)
Long Beach D, CA clean & silty sands CHT McLamore et al. (1978)
Long Beach E, CA limestone CHT McLamore et al. (1978)
Lower 232nd St., BC NC sensitive glacial clay SCPT Campanella et al. (1988, 1989)
Madingley, UK fissured Gault clay SASW Lunne et al. (1986b)
McDonald's Farm, BC soft clayey silt SCPT Campanella et al. (1986b)
McDonald's Farm, BC clean sand SCPT Robertson et al. (1986)
Munkedal, Sweden sensitive clay SCPT Larsson and Mulabdic (1991)
Norrköping, Sweden soft varved clay SCPT Larsson and Mulabdic (1991)
Onsöy, Sweden soft marine clay SCPT Lacasse and Lunne (1983)
Po River, Italy clean sand SCPT Baldi et al. (1988)
Såro Rd. 7/600, Sweden organic clay & gyttja SCPT Larsson and Mulabdic (1991)
Såro Rd. 6/900, Sweden soft organic clay SCPT Larsson and Mulabdic (1991)
Skå-Edeby, Sweden soft plastic clay SCPT Larsson and Mulabdic (1991)
Texcoco CAO, Mexico NC very plastic clay SPSL Jaime and Romo (1988)
Texcoco SCT, Mexico LOC very plastic clay SPSL Jaime and Romo (1988)
Tuve, Sweden organic marine clay SCPT Larsson and Mulabdic (1991)
Vålen, Sweden soft organic clay SCPT Larsson and Mulabdic (1991)
Xochimilco, Mexico soft very plastic clay SPSL Jaime and Romo (1988)
Yucca Mt., NV Paintbrush tuff Lab Price et al. (1994)
Notes: NC = normally consolidated SCPT = seismic cone penetration test.
LOC = lightly overconsolidated SPSL = suspension P & S-wave logging.
CHT = crosshole test SASW = spectral analysis of surface waves.

114
 2.6
Regression:
ρ = 0.277 + 0.648 log(Vs)

Mass Density, ρ (g/cc)

2.2 n = 438
r2 = 0.765 Rocks
Clean Sand (SP)
1.8 Silty Sand (SM)
Clayey Sand (SC)
Gravels (GP)
Mine Tailings (SM/ML)
1.4 Silt (ML and MH)
Fissured Clays (CL, CH)
Scandinavian Clays (CH, OH)
Mexican Clays (CH)

1.0 Units: ρ (g/cc) Intact Lean Clays (CL)

Intact Plastic Clays (CH)
peat Vs (m/sec) Regression Line

0.6
10 100 1000 10000
Shear Wave Velocity, Vs (m/s)

Figure 5.10. Mass Density-Shear Wave Velocity Correlations in Geomaterials

28
Least Squares Regression: 1:1
Actual Unit Weight, γT (kN/m3)

26 n = 438 r2 = 0.820
24 S.E. = 1.154 kN/m3

22
20
18
16
Units: γT (kN/m3)
14
Vs (m/sec)
12 σvo' (kPa)

10
10 12 14 16 18 20 22 24 26 28
Predicted: γT = 6.87(Vs)0.227(σvo')-0.057

Figure 5.11. Multiple Regression Evaluation for Total Unit Weight

Methodologies for Evaluating Permeability from Piezocone Data

Data gathered from seismic cone penetration tests yield information about the soil mass
density and porosity and this information can be combined with the estimated values of the

115
coefficient of consolidation (Chapter 4) in order to evaluate the permeability. In this section, three
alternate methodologies are developed to evaluate the permeability from seismic piezocone
dissipation tests. The first approach relies only on standard piezocone measurements of tip
resistance and pore pressure. In two other related schemes, readings from seismic piezocone
soundings are utilized and the independent Vs measurement is employed in the estimation of D. The
methodologies will be evaluated using the database listed in Table 5.2.

Table 5.2. Input and Reference Database for Permeability Evaluation

Site Depth ch ** qT σvo σvo' Vs Lab k Reference

(m) (mm2/s) (kPa) (kPa) (kPa) (m/s) (cm/s)
Bothkennar, 12.0 0.20 898 204 96 130 6.0 x 10-8 Jacobs and
U.K. Coutts (1992)
Drammen, 19.5 0.20 1000 300 121 200 1.0 x 10-8 Lacasse and
Norway Lunne (1982)
McDonald 20.0 1.90 1036 360 179 180 4.0 x10-7 Sully (1991)
Farm, B.C.
Onsoy, 18.5 0.05 754 296 115 130 1.3 x 10-7 Lacasse and
Norway Lunne (1982)
St. Alban, 4.6 0.60 300 137 43 100 3.0 x 10-7 Roy et al.
Quebec (1981)
Amherst, 3.0 0.40 1369 55 100* 175 2.0 x10-7 DeGroot and
MA (crust) Lutenegger
(1994)
Madingley, 5.8 0.05 2000 110 100* 256 1.4 x 10-9 Coop and
U.K. Wroth (1989)
∗σa = 100 for OC clay
**from piezocone dissipation

Evaluation of the Constrained Modulus from Piezocone Data

For one-dimensional drainage in oedometer tests, the permeability of a soil deposit is defined
as:

cv γ w
k= Equation 5.7
D

where D = the constrained modulus = ∆σz/∆εz for confined compression. Using this formulation in
combination with the estimated value of ch rather than cv developed in Chapter 4, the only unknown
remaining for the determination of the permeability is the constrained modulus. Three different
methods for the estimation of the constrained modulus using piezocone data are presented and
evaluated.

Cone Tip Resistance Correlation

A common correlation found in the literature relates the constrained modulus with cone tip
resistance in the form:

116
 D = αq c Equation 5.8

where α = an empirical correlation coefficient and qc = the measured cone tip resistance. Mitchell
and Gardner (1975) compiled various relationships from the literature and found that the value of α
ranged from 0.4 to 8 for clay soils; the large scatter was in part due to the fact that the data were not
collected using standardized cones and testing procedures, and qc was not corrected for pore pressure
effects (Lunne et al., 1986a). Kulhawy and Mayne (1990) compiled a piezocone database of 12 clay
sites with corrected tip resistances and constrained modulus data and developed the following
correlation (n = 42, r2 = 0.795, S.D. = 67 pa):

D = 8.25( q T − σ vo ) Equation 5.9

or, the constrained modulus is equal to 8.25 times the net cone resistance (Figure 5.12). Note M = to
the constrained modulus in the figure.

Figure 5.12. Constrained Modulus as a Function of Net Cone Tip Resistance

(Kulhawy and Mayne, 1990)

While the empirical correlation of tip resistance with constrained modulus yields a
reasonable approximation, there are fundamental reasons which make the relationship questionable.
The constrained modulus is a parameter assessed in a drained environment but the corrected tip
resistance in clays is an undrained parameter, which means that the two parameters are evaluated

117
under completely different effective stress conditions. Additionally, tip resistance in the cone
penetration test is essentially a measurement of soil strength, while the modulus is a measurement of
soil deformability. Correlations between the two are empirical at best. Due to these limitations, two
additional methodologies for evaluating the constrained modulus are examined.

Janbu's Method
Janbu (1963, 1985) presented a universal trend relating the constrained modulus of different
soils and rocks to the porosity (n), or alternately, the in-situ void ratio (e). The in-situ void ratio at
each clay site was evaluated using the method based on seismic cone data presented earlier in this
chapter and the constrained modulus was evaluated using Janbu's method. In most cases, seismic
piezocone data were available to provide all the necessary relevant data from a single sounding.
However, in the case of St. Alban, the shear wave velocities were gathered using surface
geophysical techniques at the same location. The formulation for the estimation of in-situ void ratio
was presented in Equation 5.4:

68q T 0.818
eo = Equation 5.10
Vs 1.88

In a soil system, the in-situ void ratio can be related to the porosity by the following
formulation:

eo
n= Equation 5.11
1 + eo

where n = porosity. Janbu's method presents the evaluation of the modulus of clays as a function of
porosity and effective overburden stress through the following relationship:

normally consolidated clays D = mσ vo ' Equation 5.12

overconsolidated clays D = mσ a ' Equation 5.13

where m = [(1+eo)/Cc]ln10 = the modulus number, Cc = the compression index, and σa' = 100 kPa.
Reference values of m for all materials are presented in Figure 5.13; values of m for normally
consolidated clays are presented in Figure 5.14. For clays, the data are presented in terms of in-situ
water content which can be related to the void ratio through the following formulation:

S
wn = e Equation 5.14
Gs o

where wn = the water content, S = the saturation, and Gs = the specific gravity of the solids. In
saturated media with S = 1, the water content is a function of the void ratio and the specific gravity.
118
 Figure 5.13. Constrained Modulus for All Materials

Figure 5.14. Constrained Modulus for Normally-Consolidated Clays

119
 The Janbu method presents a range for the value of the modulus; consequently, the data
estimated by this method are presented as a range of values, corresponding to the lower and upper
bounds for the modulus. After the void ratio is evaluated using the shear wave velocity and cone tip
resistance data, it is converted to a value of porosity or void ratio. The modulus number, m, is then
estimated and the lower and upper values for the constrained modulus are evaluated from the Janbu
method.

Shear Wave Velocity Correlation

This methodology relies on a correlation between the seismic cone measured values of shear
wave velocity and laboratory measured values of constrained modulus. A database of clay soils with
known shear wave velocity and constrained modulus was compiled in order to form the correlation.
The database for fourteen different sites, with a brief site description, is shown in Table 5.3.

Table 5.3. Shear Wave Velocity - Constrained Modulus Database

Site Site Description VS D Reference
(m/s) (kPa)
Bothkennar NC organic clay/silt 89 490 Nash et al. (1992)
96 630
103 1050
105 811
111 690
117 1050
121 1135
127 1460
130 1230
137 1585
142 1460
147 1725
160 2040
Brent Cross HOC London Clay 206 8410 Powell and Uglow (1988)
231 8180
258 9150
285 10300
Drammen Sensitive aged NC 96 2490 Lacasse and Lunne (1982)
118 2575
121 2760
130 2450
130 3350
156 250
Fucino Soft cemented clay 89 1250 Brignoli et al. (1995)
155 2880
Hutchinson Soft silty clay 142 1880 Burns and Mayne (1995)
Island 146 980
135 1880
158 582
Lilla Mellosa Soft organic clay 38 200 Larsson (1986)
43 200
51 190
59 200
67 200
75 200

120
 83 200
91 200
99 260
105 260
113 320
Madingley HOC Gault Clay 231 6570 Powell and Uglow (1988)
256 7070
279 8370
304 9170
327 11450
351 12430
Montalto Stiff plastic clay 300 18000 Holtz et al. (1985)
326 21000 Jamiolkowski et al. (1995)
316 25000
378 26000
347 28000
Onsoy Soft NC aged 76 1600 Lacasse and Lunne (1982)
86 1100
92 700
94 1000
114 2150
134 1500
Pentre Stiff silty clay 462 5925 Lambson et al. (1993)
549 5925
592 9143
627 6214
779 6660
777 14545
803 17297
821 12308
844 8767
Pisa Soft LOC clay 158 780 Mitchell et al. (1977)
186 1400 Jamiolkowski et al. (1995)
161 1325
264 5375
Saint Alban Sensitive soft aged 66 472 Leroueil (1996)
Ska Edeby Soft plastic clay 49 250 Larsson (1986)
57 240
64 250
70 250
77 310
83 310
88 400
95 400
105 500
Tilbrook Stiff fissured clay 1159 62500 Lambson et al. (1993)
959 62500
Note: NC = normally consolidated, OC = overconsolidated, LOC = lightly overconsolidated, HOC = heavily overconsolidated

The following correlation was observed between the shear wave velocity and constrained
modulus (n = 82, r2 = 0.7844, S.E. = 0.162):

D = 0.265 ⋅ Vs 1.74 Equation 5.15

121
where D is in kPa and Vs is in m/s. The compiled database is shown in Figure 5.15:

Bothkennar
100000 Drammen
Constrained Modulus (kPa)
Regression: Onsoy
Brent Cross
D = 0.265 Vs1.74
Madingley
n = 82 Pentre
10000 r2 = 0.7844 Tilbrook
Montalto
Lilla Mellosa
Ska Edeby
Hutchinson
1000 Fucino
Pisa
St Alban
Units: D (kPa)
Regress Line
Vs (m/s)

100
10 100 1000 10000
Shear Wave Velocity (m/s)

Figure 5.15. Constrained Modulus - Shear Wave Velocity Correlation in Clays

While the loading mechanisms to measure constrained modulus and shear wave velocity, or shear
modulus, are very different, the same material parameters influence both moduli, making the relation
between the two reasonable. Based on elastic theory, the constrained modulus can be related to the
elastic parameters E and ν (Equation 2.27 and Equation 2.28). Additionally, both the constrained
modulus and shear wave velocity are strong functions of the void ratio and effective overburden
stress in a soil deposit. Consequently, the correlation between the two relies on the fact that the
parameters are both dependent on the same soil properties.

Evaluation of Permeability
Using the three methods for the evaluation of the constrained modulus leads to the following
three formulations for the soil permeability:

chγ w
Cone Tip Resistance Correlation k h = Equation 5.16
8.25(q T − σ vo )

122
 ch γ w
Janbu's Method (Lower)* kh = − 4 .12
(Equation 5.17)
 68 ⋅ q T 0.818 
 
 Vs 1.88 
1112
. ⋅ ⋅ σ vo '
 68 ⋅ q T 0.818 
1 + 
 Vs 1.88 

ch γ w
(Upper)* kh = − 5.66
(Equation 5.18)
 68 ⋅ q T 0.818 
 
 Vs 1.88 
2.503 ⋅ ⋅ σ vo '
 68 ⋅ q T 0.818 
1 + 
 Vs 1.88 

ch γ w
Shear Wave Velocity Correlation k h = (Equation 5.19)
0.265 ⋅ Vs 1.74
*(Note for Janbu's method: σvo' used for normally consolidated clays, σa' used for overconsolidated clays).

Input data used for the evaluation of permeability are shown in Table 5.2. The results of the three
prediction methods are shown graphically in Figure 5.16, Figure 5.17, and Figure 5.18.
Additionally, the results are shown numerically in Table 5.4.

1.0E-05
Permeability Evaluated Using:
Predicted Permeability (cm/s)

D = 8.25(qT-σvo)
1.0E-06

1 to 1
1.0E-07 Bothkennar
Drammen
McDonald Farm
1.0E-08 Onsoy
St. Alban
Amherst
1.0E-09 ch from piezocone Madingley
dissipation

1.0E-10
1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05
Measured Permeability (cm/s)

Figure 5.16. Permeability Evaluation Using Constrained Modulus

Estimated from Net Cone Resistance

123
 1.0E-05
Permeability Evaluated Using:

Predicted Permeability (cm/s)

Janbu's Method
1.0E-06
1 to 1
1.0E-07 Bothkennar
Drammen
McDonald Farm
1.0E-08 Onsoy
St Alban
Amherst
1.0E-09 ch from piezocone Madingley
dissipation

1.0E-10
1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05
Measured Permeability (cm/s)

Figure 5.17. Permeability Evaluated Using Janbu's

Method of Modulus Determination

1.0E-05
Permeability Evaluated Using:
Predicted Permeability (cm/s)

D = 0.265Vs1.74
1.0E-06
1 to 1
1.0E-07 Bothkennar
Drammen
McDonald Farm
1.0E-08 Onsoy
St. Alban
Amherst
1.0E-09 ch from piezocone Madingley
dissipation

1.0E-10
1.0E-10 1.0E-09 1.0E-08 1.0E-07 1.0E-06 1.0E-05
Measured Permeability (cm/s)

Figure 5.18. Permeability Evaluation Using Seismic Piezocone Data

124
 Table 5.4. Estimated Values of Permeability Using Seismic Cone Data
Site Predicted k Predicted k Predicted k Predicted k Lab k Reference
(cm/s) (cm/s) (cm/s) (cm/s) (cm/s)
Method 1* Method 2* Method 2* Method 3*
(low) (high)
Bothkennar, 3.4 x 10-8 7.3 x 10-8 3.2 x 10-7 1.5 x 10-7 6.0 x 10-8 Jacobs and
U.K. Coutts (1992)
Drammen, 3.4 x 10-8 9.8 x 10-9 6.9 x 10-8 7.3 x 10-8 1.0 x 10-8 Lacasse and
Norway Lunne (1982)
McDonald 3.3 x 10-7 1.2 x 10-7 7.1 x 10-7 8.3 x 10-7 4.0 x 10-7 Sully (1991)
Farm, B.C.
Onsoy, 1.3 x 10-8 2.3 x 10-8 1.1 x 10-7 3.9 x 10-8 1.3 x 10-7 Lacasse and
Norway Lunne (1982)
St. Alban, 4.4 x 10-7 1.3 x 10-7 7.4 x 10-7 7.3 x 10-7 3.0 x 10-7 Roy et al.
Quebec (1981)
Amherst, 3.6 x 10-8 1.1 x 10-8 5.5 x 10-8 1.8 x 10-7 2.0 x 10-7 DeGroot and
MA Lutenegger
(crust) (1994)
Madingley, 3.1 x 10-9 8.1 x 10-10 5.2 x 10-9 1.2 x 10-8 1.4 x 10-9 Coop and
U.K. Wroth (1989)
* Method 1 = Evaluation of constrained modulus using net cone tip resistance correlation (Figure 5.12)
Method 2 = Evaluation of constrained modulus using Janbu's method (Figure 5.14)
Method 3 = Evaluation of constrained modulus using shear wave velocity correlation (Figure 5.15)

For the most part, the methods tend to overpredict the laboratory measured values of
permeability. However, the laboratory values themselves are not without fault and suffer from
problems of sample disturbance, smearing, stress relief, and variability. Also, it is important to note
that the reference values used for comparison are average values for the sites, not values specific to
the location of the dissipation tests.

Estimated Permeability Comparison with Existing Methods

In order to further validate the approach, a comparison was made between the empirical
method proposed by Parez and Fauriel (1988) and the permeability predicted using the direct shear
wave velocity - constrained modulus correlation. The results of the comparison for the soft clays
sites is shown in Figure 5.19. Only the soft clay sites were used for comparison due to the
ambiguity in defining a value of t50 in dissipation tests which exhibit the dilatory behavior. While
the two methods are based on very different approaches to evaluating the permeability, similar
results are produced, with neither method showing a distinct advantage in the prediction.

125
 1.E-05

This Study Estimated k (cm/s)

1.E-06

1.E-07 Bothkennar
Drammen
1.E-08 McDonald Farm
St Alban
Onsoy
1.E-09 1 to 1

1.E-10
1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05
Parez and Fauriel Estimated k (cm/s)

Figure 5.19. Comparison of Vs - D Correlation with the Method of

Parez and Fauriel (1988)

Conclusions
A general methodology has been presented for the evaluation of the permeability (k) based
on seismic cone penetration data, with three alternate approaches given in estimating the constrained
modulus (D). While all three methods yielded results which were in reasonable agreement with
laboratory measured values of permeability, two of the latter methods were based on a separate and
independent measurement of soil stiffness because the evaluation of the constrained modulus using
the method presented by Janbu (1963) and by the correlation with shear wave velocity introduces
another independent measurement into the procedure.
A summary outline of the three alternate methodologies is presented in Table 5.5. In Method
1, the constrained modulus is empirically related to the net cone resistance; in Method 2, the void
ratio is related to the shear wave velocity and the constrained modulus is evaluated using Janbu's
relationship between void ratio and modulus; and in Method 3, the constrained modulus is related to
the shear wave velocity directly. Finally in all methods, the permeability is calculated according to
one-dimensional consolidation theory.

126
 Table 5.5. Flowchart For Evaluating Permeability From Seismic Piezocone Data

qT, fs and u2

Pore pressure Shear wave velocity test:

dissipation as a low-strain properties
function of time, ∆u measurement, Vs and Gmax

Evaluate ch using Evaluate eo

cavity expansion, (or n)
modified Cam Clay, using correlation:
and 1D consolidation f(qT and Vs)
f(σvo', OCR, φ',Ir)

1. Evaluate D using
correlation: 2. Evaluate D using 3. Evaluate D using Vs-D
D=1/mv=8.25(qT-σvo) Janbu's Modulus method: correlation:
f(qT-σvo) D=0.265Vs1.74
f(qT, Vs, σvo') f(Vs)

ch γ w
k=
D

127
 CHAPTER 6

CONCLUSIONS AND RECOMMENDATIONS

Conclusions
Developments related to the usefulness of piezocone penetrometers for geoenvironmental
site characterization have been addressed in this study. These include preliminary experimentation
on the detection of the BTEX chemicals using an integrated optic sensor miniaturized for inclusion
in a cone penetrometer, and improved procedures for the determination of the coefficient of
consolidation ch, the permeability k, and the constrained modulus D, especially in heavily-
overconsolidated clays. The focus of this study concentrated on the optimization of the methods of
subsurface characterization using piezocone penetrometers.
A set of initial experiments was performed to evaluate the behavior of an integrated opto-
electronic BTEX sensor in an Ottawa sand/aqueous environment. The sensor had been miniaturized
from a bench-scale apparatus in order to fit inside a trailing module of a cone penetrometer. The
tests were performed as proof-of-concept experiments in order to demonstrate the sensor's viability
in a soil environment. Seven exploratory experiments were performed in which concentrations as
low as 6 ppm (xylene) and as high as 380 ppm (benzene) were detected by the sensor while
operating in a sand/water matrix. Malfunctioning of the sensor prevented further testing of its
operating characteristics; however, the initial tests demonstrated the sensor's viability in a soil
environment and a new version of the sensor is currently under development in order to facilitate
future testing.
The radial coefficient of consolidation (ch) was evaluated using a methodology based on
critical state soil mechanics and cavity expansion theory, in combination with an analytical solution
to the one-dimensional consolidation equation. The method was able to evaluate the standard
monotonic porewater pressure dissipation behavior in lightly-overconsolidated clays, as well as the
dilatory pore pressure behavior observed in heavily-overconsolidated clays. Because the pore
pressures which are induced by the shearing action of a cone penetrometer can be negative in
heavily-overconsolidated clays and affect only a thin zone beside the cone penetrometer, the pore
pressure dissipation profile can drawdown to magnitudes lower than hydrostatic. However, because
these shear-induced pore pressures dissipate rapidly, the effect is limited to a short period of time.
In this evaluation, the dissipation of the octahedral normal stress-induced pore pressures was
handled separately from the shear-induced pore pressures and then summed in order to achieve the
complete profile. Estimation of the coefficient of consolidation was achieved through matching of
the field-measured curve with the method-evaluated curve. The values of the coefficient of
consolidation generally agreed in soft clays and in most cases for the hard clays; however, there
were several overconsolidated clay pore pressure profiles not accurately represented by the
procedure. In these cases, there appear to be additional factors influencing the consolidation
behavior surrounding the cone. The model accounted for the effects of the width of shear zone, the
rigidity index, the OCR, and the effective friction angle. Additional parameters which should be
investigated include the influence of rate of penetration, possible smearing effects, the degree of
fissuring, the effect of possible hydraulic fracturing, and the effect of drainage along the side of the
cone penetrometer.
In order to perform an in-situ evaluation of the soil permeability, a methodology was
developed using the aforementioned coefficient of consolidation in combination with three alternate
estimates of the constrained modulus based on cone data. The proposed methods for the evaluation
of permeability (k = chγw/D) tended to over-estimate the lab-measured values for the seven case
studies examined.
128
 Recommendations
Recommendations for future research in the detection of chemicals using integrated optic
sensors in cone penetrometers includes a series of calibration chamber tests, which are currently in
progress with the BTEX sensor, for the evaluation of the sensor behavior in a variety of geochemical
environments including extremes of pH and concentrated as well as dilute BTEX solutions. The
effects of natural organic matter, effects of high temperature induced by cone penetration, and
penetration rate effects from the cone sounding should also be studied. Incorporation of the sensor
behind a cone penetrometer and upgrade of the sensor's communication electronics are also in
progress. New variations on the sensor coating will also enable detection of additional chemicals
ranging from chlorinated hydrocarbons to dissolved metals. In addition to its incorporation in a cone
penetrometer, the IO sensor shows promise as a push-in place detector which could be left to
monitor long-term concentrations of chemicals at a contaminated site.
Regarding the evaluation of the coefficient of consolidation from piezocone dissipation tests,
recommendations include the assurance of proper test procedures, especially complete saturation of
the pore pressure filter element which can produce dilatory dissipation behavior similar to that seen
in heavily-overconsolidated clays if not done properly. An important parameter is the width of shear
zone induced around the cone penetrometer during a sounding. Little experimental evidence exists
which quantifies the size of this zone or the effect of penetration rate on the width of the zone.
Because the width of the shear zone is important in the evaluation of the dissipation behavior of the
shear-induced pore pressures, further study of this parameter is needed.
In the methodology to evaluate the in-situ permeability of a soil, additional well-controlled
lab tests data should be performed for calibration and verification of improved numerical schemes.
The reference values used for comparison of the estimated permeability were average values
reported at the sites. Because permeability can vary significantly in soil deposits, more site- and
depth-specific data would provide a more reliable comparison.
Figure 6.1 illustrates a conceptual environmental seismic piezocone penetrometer which
includes the traditional cone measurement capabilities of tip resistance, sleeve friction, and pore
pressure, combined with an accelerometer to measure shear wave velocity, and a chemical detection
module. While the seismic piezocone penetrometer is currently available commercially, the
incorporation of the chemical detection module is still in development.
A listing of the measured and derived parameters used throughout this thesis is also
presented in Figure 6.1. On the left side of the figure are shown the parameters which can be
measured using the geoenvironmental cone, while the right side shows the parameters which can be
derived from those measurements. The chemical module allows for the detection and measurement
of concentration as a function of depth and of time, which is useful not only for site characterization,
but also for mass transport studies. The seismic module can measure both compression and shear
wave velocities, which can be used to detect the water table, and also to measure the stiffness of a
soil deposit, which can be correlated with soil density. The traditional cone sensors include tip
resistance, sleeve friction, and pore pressure, and are used for identification of soil type. In this
study, the pore pressure sensor was used to monitor pressure as a function of time in order to
evaluate the coefficient of consolidation in a soil deposit. As shown, an in-situ estimation of the
permeability of a soil deposit can be made through the combination of readings from the pore
pressure transducer and seismic velocity transducer. The soil permeability is a fundamental element
in the advective transport of groundwater and contaminating chemicals. The conceptual cone
penetrometer can provide significantly improved site characterization in the investigation of
geoenvironmental parameters.

129
 Measured Parameters Derived Parameters

To data acquisition system at surface

Miniaturized Sampling Pump

Chemical Concentration [X] Presence, extent, and

Integrated Optic Sensors concentration of
Chemical contamination
Detection
Module

Gmax = ρVs2
Compression Wave Velocity Vp ρ - Mass Density (correlation)
Shear Wave Velocity Vs e - Void Ratio (correlation)
D - Constrained Modulus (correlation)

Seismic-Dual
Element Cone Sleeve Friction fs Soil stratigraphy and soil type
Penetrometer

Dissipation Tests
Soil Stratigraphy
Pore Water Pressure u2 ch - Coefficient of Consolidation
k - Permeability
Pore Water Pressure u1 Degree of fissuring
Dissipation Tests = f(time)
Soil Stratigraphy
Cone Tip Resistance qc D - Constrained Modulus (correlation)
Strength

Figure 6.1. Environmental Seismic Piezocone Penetrometer Schematic

130
 BIBLIOGRAPHY

Abramowitz, M. and Stegun, I. (1964). Handbook of Mathematical Functions, U.S. Government

Printing Office, Washington, D.C., 1046 pp.
Acar, Y.B. and Tumay, M.T. (1986). Strain Fields Around Cones in Steady Penetration, Journal of
Geotechnical Engineering, ASCE, Vol. 112, No. 2, pp. 207-213.
American Society for Testing and Materials. (1996). Standard Test Methods for Crosshole Seismic
Testing (D4428), Annual Book of Standards, Vol. 4.08, Section 4, West Conshohocken, PA,
pp. 636-645..
American Society for Testing and Materials. (1996). Standard Method for Performing Electronic
Friction Cone and Piezocone Penetration Testing of Soils (D5778-95), Annual Book of
Standards, Vol. 4.08, West Conshohocken, PA.
Apitz, S.E., Theriault, G.A., and Lieberman, S.H. (1992a). Optimization of the Optical
Characteristics of a Fiber-Optic Guided Laser Fluorescence Technique for the In-Situ
Evaluation of Fuels in Soils, Proceedings, Environmental Process and Treatment
Technologies, Los Angeles, The International Society for Optical Engineering, January.
Apitz, S.E., Borbridge, L.M., Theriault, G.A., and Lieberman, S.H. (1992b). Remote In-Situ
Determination of Fuel Products in Soils: Field Results and Laboratory Investigations,
Analusis, Vol. 20, Elsevier, Paris, pp. 461-474.
Arulmoli, K. (1994). Users' Perspective - In Situ Sampling Systems, Summary Report, Workshop on
Advancing Technologies for Cone Penetration Testing for Geotechnical and
Geoenvironmental Site Characterization, Engineering and Environmental Sciences Division,
US Army Research Office, Research Triangle Park, June.
Aubeny, C.P. (1992). Rational Interpretation of In Situ Tests in Cohesive Soils, Ph.D. Thesis,
Massachusetts Institute of Technology, Cambridge, MA.
Auxt, J.A. and Wright, D. (1995). Environmental Site Characterization in the United States Using
the Cone Penetrometer, Proceedings, International Symposium on Cone Penetration Testing,
Vol. 2, Linköping, Sweden, pp. 387-392.
Baldi, G., Bellotti, R., Ghionna, V., Jamiolkowski, M., and LoPresti, D.C. (1989). Modulus of sands
from CPTs and DMTs, Proceedings, 12th International Conference on Soil Mechanics and
Foundation Engineering, Vol. 1, Rio de Janeiro, pp. 165-170.
Baldi, G., Bruzzi, D., Superbo, S., Battaglio, M., and Jamiolkowski, M. (1988). Seismic Cone in Po
River Sand, Proceedings, International Symposium on Penetration Testing, J.D. Ruiter (ed),
A.A. Balkema, Rotterdam, The Netherlands, Vol. 1, pp. 643-650.
Baligh, M.M. (1985). Strain Path Method, Journal of Geotechnical Engineering, ASCE Vol. 111,
No. 9, pp. 1108-1136.
Baligh, M.M. (1986). Undrained Deep Penetration, II: Pore Pressures, Geotechnique, Vol. 36, No. 4,
pp. 487-501.
Baligh, M.M. and Levadoux, J.N. (1980). Pore Pressure Dissipation After Cone Penetration,
Research Report R80-11, Department of Civil Engineering, Massachusetts Institute of
Technology, Cambridge, MA, 367 pp.
Baligh, M.M. and Levadoux, J.-N. (1986). Consolidation After Undrained Piezocone Penetration. II:
Interpretation, Journal of Geotechnical Engineering, ASCE, Vol. 112, No. 7, pp. 727-745.
Baligh, M.M. et al. (1981). The Piezocone Penetrometer, Proceedings, Cone Penetration Testing and
Experience, G.M. Norris and R.D. Holtz, Ed., American Society of Civil Engineers, New
York, pp. 247-263.

131
Battaglio, M., Bruzzi, D., Jamiolkowski, M., and Lancellotta, R. (1986). Interpretation of CPT's and
CPTU's: Undrained Penetration of Saturated Clays, Fourth International Geotechnical
Seminar, Field Instrumentation and In-Situ Measurements, Singapore, pp. 129-156.
Battaglio, M., Jamiolkowski, M., Lancellotta, R., and Maniscalco, R. (1981). Piezometer Probe Test
in Cohesive Deposits, Proceedings, Cone Penetration Testing and Experience, G.M. Norris
and R.D. Holtz, Ed., American Society of Civil Engineers, New York, pp. 264- 302.
Bellotti, R., Benoit, J., and Morabito, P. (1994). A Self-Boring Electrical Resistivity Probe for
Sands, Proceedings, 13th International Conference on Soil Mechanics and Foundation
Engineering, New Dehli.
Bishop, A.W. (1972). Shear Strength Parameters for Undisturbed and Remolded Soil Specimens,
Proceedings, Roscoe Memorial Symposium, Cambridge University, Foulis & Co., pp. 3-139.
Bishop, R.F., Hill, R., and Mott, N.F. (1945). The Theory of Indentation and Hardness Tests,
Proceedings of the Physical Society, London, Vol. 57, No. 3, pp. 147-159.
Bjerrum, L. and Johannessen, I. (1961). Pore Pressure Resulting from Driving Piles in Soft Clay,
Pore Pressure and Suction in Soil, Butterworth, London, pp. 108-111.
Bond, A.J. and Jardine, R.J. (1991). Effects of Installing Displacement Piles in a High OCR Clay,
Geotechnique, Vol. 41, No. 3, pp. 341-363.
Bond, A.J. and Jardine, R.J. (1995). Shaft Capacity of Displacement Piles in a High OCR Clay,
Geotechnique, Vol. 45, No. 1, pp. 3-23.
Bowders, J. and Daniel, D.E. (1994). Cone Penetration Technology for Subsurface Characterization,
Geotechnical News, Vol. 12, No. 3, September.
Bratton, J.L. and Timian, D.A. (1995). Environmental Site Applications of the Environmental CPT,
Proceedings, International Symposium on Cone Penetration Testing, Vol. 2, Linköping,
Sweden, pp. 429-434.
Bratton, W.L., Bratton, J.L., and Shinn, J.D. (1995). Direct Penetration Technology for Geotechnical
and Environmental Site Characterization, Proceedings, Geoenvironment 2000 (GSP No. 46),
Vol. 1, American Society of Civil Engineers, New Orleans, pp. 105-122.
Bratton, W.L., Timian, S.M., and Fisk, B.E. (1993). Piezoseismic-Cone-Penetration Tests at the In-
Tank Precipitation Facility, Savannah River Site, Aiken, S.C., Applied Research Associates,
South Royalton, Vermont.
Brignoli, E., Burghignoli, A., and Faiella, D. (1995). Performance of a Silo Founded on a Lacustrine
Deposit: A Case History, Proceedings, Compression and Consolidation of Clay Soils,
Yoshikuni, H. and Kusakabe, O. (eds), Balkema, Rotterdam, Vol. 1, pp. 617-622.
Brodzinski, R.L. (1995). Cone Penetrometer Deployment Demonstration for Radiation Detection
Instrumentation, Draft Copy.
Brown, D.N. (1993). Determination of Stress State of Clays by Piezocone, M.S. Thesis, School of
Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA, 168 pp.
Bruzzi, D. and Battaglio, M. (1987). Pore Pressure Measurements During Cone Penetration Tests, I
quaderni dell'ISMES, No. 229, Milan, Italy, 246 pp.
Brylawski, E. (1994). AP Van den Berg Inc, Milford, PA, 18337.
Burghignoli, A., Cavalera, L., and Chieppa, V. (1991). Geotechnical Characterization of Fucino
Clay, Proceedings, 10th European Conference on Soil Mechanics and Foundation Engineering,
Firenze, Vol. 1, pp. 27-40.
Burns, S.E., Hegazy, Y., and Mayne, P.W. (1994). Seismic Piezocone Penetration Tests, Bagdad
Copper Mines - Bagdad, AZ, Geotechnical Report, Georgia Institute of Technology, Atlanta,
GA.

132
Burns, S.E. and Mayne, P.W. (1994). Seismic Piezocone Penetration Tests, San Manuel Copper
Mine Tailings - Tucson, AZ, Geotechnical Report, Georgia Institute of Technology, Atlanta,
GA.
Burns, S.E. and Mayne, P.W. (1995a). Coefficient of Consolidation (ch) from Type 2 Piezocone
Dissipation in Overconsolidated Clays, Proceedings, International Symposium on Cone
Penetration Testing (CPT '95), Linköping, Sweden, Vol. 2, pp. 137-142.
Burns, S.E. and Mayne, P.W. (1995b). Seismic Piezocone Penetration Tests, Georgia International
Maritime Facility - Hutchinson Island, Georgia, Geotechnical Report, Georgia Institute of
Technology, Atlanta.
Burns, S.E. and Mayne, P.W. (1996). Small- and High-Strain Measurements of In-Situ Soil
Properties Using the Seismic Cone Penetrometer, Transportation Research Record, National
Research Council, Washington D.C., No. 1548, pp. 81-88.
Campanella, R.G. (1994). Field Methods for Dynamic Geotechnical Testing, Dynamic Geotechnical
Testing II, Special Technical Publication 1213, American Society for Testing and Materials,
Philadelphia, pp. 3-23.
Campanella, R.G., Baziw, E.J., and Sully, J.P. (1989). Interpretation of Seismic Cone Data Using
Digital Filtering Techniques, Proceedings, 12th International Conference on Soil Mechanics
and Foundation Engineering, Vol. 1, Rio de Janeiro, Balkema, Rotterdam, pp. 195-198.
Campanella, R.G. and Robertson, P.K. (1988). Current Status of the Piezocone, Proceedings,
International Symposium on Penetration Testing, J.D. Ruiter (ed), A.A. Balkema, Rotterdam,
Vol. 1, pp. 93-116.
Campanella, R.G., Robertson, P.K., and Gillespie, D. (1986a). Factors Affecting the Pore Water
Pressure and Its Measurement Around a Penetrating Cone, Proceedings, 39th Canadian
Geotechnical Conference, Ottawa, pp. 291-299.
Campanella, R.G., Robertson, P.K., and Gillespie, D. (1986b). Seismic Cone Penetration Test, Use
of In-Situ Tests in Geotechnical Engineering, Geotechnical Special Publication No.6, ASCE,
New York, 116-130.
Campanella, R.G., Sully, J.P., and Robertson, P.K. (1988). Interpretation of Piezocone Soundings in
Clay - A Case History, Penetration Testing in the U.K., Thomas Telford, London, pp. 203-208.
Campanella, R.G. and Weemees, I.A. (1989). Development and Use of an Electrical Resistivity
Cone for Groundwater Contamination Studies, Proceedings, 42nd Annual Canadian
Geotechnical Conference, Winnipeg, pp. 1-11.
Carrabba, M.M. (1995). Phase I Topical Report: Field Raman Spectrograph for Environmental
Analysis, DOE Contract DE-AC21-92MC29108.
Carslaw, H.D. and Jaeger, J.C. (1959). Conduction of Heat in Solids, 2nd Edition, Oxford University
Press, London, 510 pp.
Chen, B.S.Y. and Mayne, P.W. (1994). Profiling the Overconsolidation Ratio of Clays by Piezocone
Tests, Georgia Institute of Technology Internal Report, No. GIT-CEEGEO-94-1, Atlanta,
279 pp.
Chen, B.S.Y. and Mayne, P.W. (1995). Type 1 and 2 Piezocone Evaluations of Overconsolidation
Ratio in Clays, Proceedings, International Symposium on Cone Penetration Testing (CPT
'95), Vol. 2, Linköping, Sweden, pp. 143-148.
Chen, B.S.Y. and Mayne, P.W. (1996). Statistical Relationships Between Piezocone Measurements
and Stress History of Clays, Canadian Geotechnical Journal, Vol. 33, No. 3, pp. 488-498.
Coop, M.R. and Wroth, C.P. (1989). Field Studies of an Instrumented Model Pile in Clay,
Geotechnique, Vol. 39, No. 4, pp. 679-696.
CRC. (1993). Handbook of Chemistry and Physics, 74th Edition, D.R. Lide, Ed., CRC Press, Boca
Rato.
133
Daniel, D.E. (1989). In-Situ Hydraulic Conductivity Tests for Compacted Clay, Journal of
Geotechnical Engineering, ASCE, Vol. 115, No. 9, pp. 1205-1226.
Daniel, D.E. (1994). State-of-the-Art: Laboratory Hydraulic Conductivity Tests for Saturated Soils,
Hydraulic Conductivity and Waste Containment Transport in Soils, ASTM STP 1142,
Daniel, D.E. and Trautwein, S.J., Eds., American Society for Testing and Materials,
Philadelphia, pp. 30-78.
Daniel, D.E., Anderson, D.C., and Boynton, S.S. (1985). Fixed-Wall Versus Flexible-Wall
Permeameters, Hydraulic Barriers in Soil and Rock, ASTM STP 874, A.I. Johnson, R.K.
Frobel, N.J. Cavalli, and C.B. Pettersson, eds., American Society for Testing and Materials,
Philadelphia, pp. 107-126.
Davidson, J.L. (1985). Pore Pressures Generated During Cone Penetration Testing in Heavily
Overconsolidated Clays, Proceedings, 11th International Conference on Soil Mechanics and
Foundation Engineering, San Francisco, Vol. 5, pp. 2699.
Degroot, D.J. and Lutenegger, A.J. (1994). A Comparison Between Field and Laboratory
Measurements of Hydraulic Conductivity in a Varved Clay, Hydraulic Conductivity and
Waste Contaminant Transport in Soil, ASTM STP 1142, D.E. Daniel and S.J. Trautwein, eds,
American Society for Testing and Materials, Philadelphia, pp. 300-317.
Du, J. and Wu, T.H. (1992). Proceedings, US-China Workshop on Cooperative Research in
Geotechnical Engineering, Tongji University, Shanghai (NSF and NNSF/PRC), 215 pp.
Elsworth, D. (1990). Theory of Partially Drained Piezometer Insertion, Journal of Geotechnical
Engineering, ASCE, Vol. 116, No. 6, pp. 899-912.
Elsworth, D. (1993). Analysis of Piezocone Dissipation Data Using Dislocation Methods, Journal of
Geotechnical Engineering, ASCE, Vol. 119, No. 10, pp. 1601-1623.
Environmental Protection Agency. (1993). Subsurface Characterization and Monitoring Techniques,
EPA/625/R-93/003a, Office of Research and Development, Washington D.C.
Gaberc, A., Ajdiç, I., and Majes, B. (1995). Vertical and Horizontal Permeability of Marshland
Subsoils, Proceedings, 11th European Conference on Soil Mechanics and Foundation
Engineering, Copenhagen, Vol. 1, pp. 127-132.
Ghionna, V. et al. (1995). Cone Pressuremeter Tests in Po River Sand, The Pressuremeter and Its
New Avenues, Balkema, Rotterdam, pp. 471-480.
Gillespie, D. and Campanella, R.G. (1981). Consolidation Characteristics from Pore Pressure
Dissipation After Piezometer Cone Penetration, Soil Mechanics Series 47, Department of
Civil Engineering, The University of British Columbia, Vancouver, 17 pp.
Gillespie, D., Lunne, T., and Eidsmoen, T. (1985). Cone Penetration Tests in Ons∅y Clay, Internal
Report - In Situ Site Investigation Techniques and Interpretation for Offshore Practice, No.
40019-7, Norwegian Geotechnical Institute, Oslo.
Greig, J.W., Campanella, R.G., and Robertson, P.K. (1988). Comparison of Field Vane Results with
Other In-Situ Test Results, Vane Shear Strength Testing in Soils: Field and Laboratory
Studies, ASTM STP 1014, pp. 247-263.
Grisak, G.E. (1975). The Fracture and Porosity of Glacial Till, Canadian Journal of Earth Science,
Vol. 12, No. 3, pp. 513-515.
Grisak, G.E. and Cherry, J.A. (1975). Hydrologic Characteristics and Response of Fractured Till and
Clay Confining a Shallow Aquifer, Canadian Geotechnical Journal, Vol. 12, No. 1, pp. 23-
43.
Gupta, R.C. (1983). Determination of the Insitu Coefficient of Consolidation and Permeability of
Submerged Soils Using Electrical Piezoprobe Soundings, Ph.D. Thesis, University of
Florida, Gainesville, 303 pp.

134
Gupta, R.C. and Davidson, J.L. (1986). Piezoprobe Determined Coefficient of Consolidation, Soils
and Foundations, Vol. 26, No. 3, pp. 12-22.
Hansbo, S., Jamiolkowski, M., and Kok, L. (1981). Consolidation by Vertical Drains, Geotechnique,
Vol. 31, No. 1, pp. 45-66.
Hartman, N.F. (1990). Optical Sensing Apparatus and Method, U.S. Patent No. 4,940,3228,
Washington, D.C., July.
Hartman, N.F., Campbell, D.L., and Gross, M. (1988). Waveguide Interferometer Configurations,
Proceedings, Leos Annual Meeting Conference, Santa Clara, November.
Hartman, N.F., Walsh, J.L., Ross, C.C., and Campbell, D.P. (1992). An Integrated Optic Gaseous
Ammonia Sensor, Proceedings, Sensors Expo, Chicago, September.
Hegazy, Y.A. and Mayne, P.W. (1995). Statistical Correlations Between Vs and Cone Penetration Data
for Different Soil Types. Proceedings, International Symposium on Cone Penetration Testing
(CPT'95), Linköping, Sweden, Vol. 2, pp. 173-178.
Henkel, D.J. (1959). The Relationships Between the Strength, Pore Water Pressure, and Volume
Change Characteristics of Saturated Clays, Geotechnique, Vol. 9, pp. 119-135.
Hepton, P. (1988). Shear Wave Velocity Measurements During Penetration Testing, Proceedings,
Penetration Testing in the U.K., Thomas Telford LTD, London, pp. 275-278.
Hilbert, L.B., Hwong, T.K., Cook, N.G.W., Nihei, K.T., Myer, L.R. (1994). Effects of Strain
Amplitude on Static and Dynamic Nonlinear Deformation of Berea Sandstone, Rock
Mechanics: Models and Measurements, (P.P. Nelson and S.E. Laubach, eds.), Balkema
Publishers, Rotterdam, Netherlands, pp. 463-471.
Hill, R. (1950). The Mathematical Theory of Plasticity, Oxford University Press, London.
Hird, C.C., Powell, J.J.M., and Yung, P.C.Y. (1991). Investigations of the Stiffness of a Glacial Clay
Till, Proceedings, 10th International Conference on Soil Mechanics and Foundation
Engineering, Florence, Vol. 1, pp. 107-110.
Holtz, R.D., Jamiolkowski, M.B., and Lancellotta, R. (1985). Lessons from Oedometer Tests on
High Quality Samples, Journal of Geotechnical Engineering, ASCE, Vol. 112, No. 8, pp.
768-776.
Horsnell, M.R. (1988). The Use of Cone Penetration Testing to Obtain Environmental Data,
Proceedings, Penetration Testing in the U.K., Thomas Telford, London, pp. 289-295.
Houlsby, G.T. and Hitchman, R. (1988). Calibration Chamber Tests of a Cone Pressuremeter in
Sands, Geotechnique, Vol. 38, No. 1, pp. 39-44.
Houlsby, G.T. and Teh, C.I. (1988). Analysis of the Piezocone in Clay, Proceedings, International
Symposium on Penetration Testing, J.D. Ruiter (ed), A.A. Balkema, Rotterdam, The
Netherlands, Vol. 1, pp. 777-783.
Houlsby, G.T. and Withers, N.J. (1988). Analysis of the Cone Pressuremeter in Clay, Geotechnique,
Vol. 38, No. 4, pp. 575-587.
Hryciw, R.D. and Raschke, S.A. (1996). Development of Computer Vision Technique for In-Situ
Soil Characterization, Transportation Research Record, National Research Council,
Washington D.C., No. 1526, pp. 86-97.
Jacobs, P.A. and Coutts, J.S. (1992). A Comparison of Electric Piezocone Tips at the Bothkennar
Test Site, Geotechnique, Vol. 42, No. 2, pp. 369-375.
Jaime, A. and Romo, M.P. (1988). The Mexico Earthquake of September 19, 1985 - Correlations
Between Dynamic and Static Properties of Mexico City Clay, Earthquake Spectra, Vol. 4,
No. 4, pp. 787-804.
Jamiolkowski, M. (1995). Opening Address, Proceedings, International Symposium on Cone
Penetration Testing (CPT '95), Linköping, Sweden, Vol. 3, pp. 7-15.

135
Jamiolkowski, M., Ladd, C.C., Germaine, J., and Lancelotta, R. (1985). New Developments in Field
and Lab Testing of Soil, Proceedings, 11th International Conference on Soil Mechanics and
Foundation Engineering, San Francisco, Vol. 1, pp. 57-154.
Jamiolkowski, M., Lancellotta, R., and Lo Presti, D.C.F. (1995). Remarks on the Stiffness at Small
Strains of Six Italian Clays, Proceedings, International Symposium on Pre-Failure
Deformation Characteristics of Geomaterials (Sappro), Balkema Press Rotterdam, Vol. 2, pp.
817-836.
Jamiolkowski, M. and Robertson, P.K. (1988). Future Trends for Penetration Testing, Proceedings,
Penetration Testing in the U.K., Thomas Telford, London, pp. 321-342.
Janbu, N. (1963). Soil Compressibility as Determined by Oedometer and Triaxial Tests,
Proceedings, 3rd European Conference on Soil Mechanics, Wiesbaden, Vol. 1, pp. 19-25.
Janbu, N. (1985). Soil Models in Offshore Engineering, Geotechnique, Vol. 35, No. 3, pp. 241-281.
Janbu, N. and Sennesset (1974). Effective Stress Interpretation of In Situ Static Cone Penetration
Tests, Proceedings, European Symposium on Penetration Testing I, Stockholm, Vol. 2, pp.
181-195.
Jardine, R.J. and Bond, A.J. (1989). Behaviour of Displacement Piles in a Heavily Overconsolidated
Clay, Proceedings, 12th International Conference on Soil Mechanics and Foundation
Engineering, Rio de Janeiro, Vol. 2, pp. 1147-1151.
Jones, G.A. and Van Zyl, D.J.A. (1981). The Piezometer Probe - A Useful Investigation Tool,
Proceedings, 10th International Conference on Soil Mechanics and Foundation Engineering,
Stockholm, Vol. 3, pp. 489-495.
Kabir, M.G. and Lutenegger, A.J. (1987). Discussion on Piezoprobe Determined Coefficient of
Consolidation, Soils and Foundations, Vol. 27, No. 2, pp. 70-72.
Kabir, M.G. and Lutenegger, A.J. (1990). In Situ Estimation of the Coefficient of Consolidation in
Clays, Canadian Geotechnical Journal, Vol. 27, No. 1 , pp. 58-67.
Kavvadas, M. (1982). Non-linear Consolidation Around Driven Piles in Clay, Ph.D. Thesis,
Massachusetts Institute of Technology, Cambridge, MA.
Keaveny, J.M. and Mitchell, J.K. (1986). Strength of Fine-Grained Soils Using the Piezocone,
Proceedings, Use of In-Situ Tests in Geotechnical Engineering (In-Situ 86), ASCE,
Geotechnical Special Publication No. 6,New York, pp. 668-685.
Koga, Y., Sasaki, Y., Itoh, Y., and Matsuo, O. (1986). Field Investigation of Seismically Damaged
Sites By Use of Vibratory Cone Penetrometer, Proceedings, 18th Joint Meeting US-Japan
Panel on Wind and Seismic Effects, Washington, D.C.
Koizumi, Y. and Ito, K. (1967). Field Tests With Regard to Pile Driving and Bearing Capacity of
Piled Foundations, Soils and Foundations, Vol. 7, No. 3.
Konrad, J.-M. and Frechette, P. (1995). The Piezocone-Permeameter Probe: A Promising Tool?,
Proceedings, Geoenvironment 2000 (GSP No. 46), American Society of Civil Engineers,
Vol. 1, New Orleans, pp. 93-104.
Kulhawy, F.H. and Mayne, P.W. (1990). Soil Properties Manual, Report EL-6800, Electric Power
Research Institute, Palo Alto, 306 p.
Lacasse, S., Ladd, C.C., and Baligh, M.M.. (1978). Evaluation of Field Vane, Dutch Cone
Penetrometer, and Piezometer Probe Testing Devices, Final Reports, prepared for U.S.
Department of Transportation, Federal Highway Administration, Washington, DC.
Lacasse, S., and Lunne, T. (1982). Penetration Tests in Two Norwegian Clays, Proceedings, 2nd
European Symposium on Penetration Testing, Vol. 2, Amsterdam, pp. 661-669.
Lacasse, S. and Lunne, T. (1983). Dilatometer Tests in Two Soft Marine Clays, Report No. 146,
Norwegian Geotechnical Institute, Oslo, Norway.

136
Lally, M.J. (1993). A Field and Laboratory Investigation of Geotechnical Properties for Design of a
Seasonal Heat Storage Facility, Master of Science Thesis, Civil Engineering Department,
University of Massachusetts at Amherst, 224 pp.
Lambe, T.W. (1951). Soil Testing for Engineers, John Wiley and Sons, New York, 165 pp.
Lambe, T.W. and Whitman, R.V. (1969). Soil Mechanics, John Wiley and Sons, New York, 553 pp.
Lambrechts, J.R., and Leonards, G.A. (1978). Effects of Stress History on Deformation of Sand,
Journal of Geotechnical Engineering, ASCE, Vol. 104, No. GT11, pp. 1371-1388.
Lambson, M.D., Clare, D.G., Senner, D.W.F., and Semple, R.M. (1993). Investigation and
Interpretation of Pentre and Tilbrook Grange Soil Conditions, Large Scale Pile Tests in Clay,
Thomas Telford, London, pp. 134-196.
Lambson, M. and Jacobs, P. (1995). The Use of Laser Induced Fluorescence Cone for
Environmental Investigations, Proceedings, International Symposium on Cone Penetration
Testing (CPT '95), Linköping, Sweden, Vol. 2, pp. 29-34.
Larsson, R. (1986). Consolidation of Soft Soils, Report 29, Swedish Geotechnical Institute,
Linköping, Sweden, 174 pp.
Larsson, R., and Mulabdic, M. (1991). Shear Moduli in Scandinavian Clays: Measurement of Initial
Shear Modulus with Seismic Cones, Report 40, Swedish Geotechnical Institute, Linköping,
Sweden, 127 pp.
Larsson, R. and Mulabdic, M. (1993). Report 42, Swedish Geotechnical Institute, Linköping,
Sweden, 240 pp.
Lefebvre, G. Leboeuf, D., Rahhal, M., Lacroix, A., and Warde, J. (1994). Laboratory and Field
Determination of Small-Strain Shear Modulus for a Structured Champlain Clay, Canadian
Geotechnical Journal, Vol. 31, No. 1, pp. 61-70.
Lehane, B.M. and Jardine, R.J. (1994). Displacement Pile Behavior in Glacial Clay, Canadian
Geotechnical Journal, Vol. 31, No. 1, pp. 79-90.
Leroueil, S. (1996). Compressibility of Clays: Fundamental and Practical Aspects, Journal of
Geotechnical Engineering, ASCE, Vol. 122, No. 7, pp. 534-543.
Leroueil, S. and Jamiolkowski, M. (1991). General Report on Exploration of Soft Soil and
Determination of Design Parameters, Proceedings, Geo-Coast '91, Yokohama, 29 pp.
Leroueil, S., Lerat, P., Hight, D.W., and Powell, J.J.M. (1992). Hydraulic Conductivity of a Recent
Estuarine Silty Clay at Bothkennar, Geotechnique, Vol. 42, No. 2, pp. 275-288.
Levadoux, J.N. (1980). Pore Pressures in Clays Due to Cone Penetration, Ph.D. Thesis, Civil
Engineering Department, Massachusetts Institute of Technology, Cambridge, MA, Vol. 1,
386 pp.
Levadoux, J.N. and Baligh, M.M. (1980). Pore Pressure During Cone Penetration, Research Report,
R80-15, Department of Civil Engineering, Massachusetts Institute of Technology,
Cambridge, MA, 310 pp.
Levadoux, J.-N. and Baligh, M.M. (1986). Consolidation After Undrained Piezocone Penetration. I:
Prediction, Journal of Geotechnical Engineering, ASCE, Vol. 112, No. 7, pp. 707-726.
Lieberman, S.H., Theriault, G.A., Cooper, S.S., Malone, P.G., Olsen, R.S., and Lurk, P.W. (1991).
Rapid, Subsurface, In-Situ Field Screening of Petroleum Hydrocarbon Contamination Using
Laser Induced Fluorescence Over Optical Fibers, Proceedings, 2nd International Symposium
for Field Screening Methods for Hazardous Wastes and Toxic Chemicals, Las Vegas, pp. 57-
63.
Lightner, E.M. and Purdy, C.B. (1995). Cone Penetrometer Development and Testing for
Environmental Applications, Proceedings, International Symposium on Cone Penetration
Testing, Vol. 2, Linköping, Sweden, pp. 41-48.

137
Lo, K.Y. and Stermac, A.G. (1965). Induced Pore Pressures During Pile Driving Operations,
Proceedings, 6th International Conference on Soil Mechanics and Foundation Engineering,
Montreal, Vol. 2.
Lunne, T., Eidsmoen, T.E., Gillespie, D., and Howland, J. (1986a). Laboratory and Field Calibration
of Cone Penetrometers, Use of In-Situ Tests in Geotechnical Engineering, Geotechnical
Special Publication No. 6, American Society of Civil Engineers, New York, pp. 714-729.
Lunne, T. Eidsmoen, T.E., Powell, J.J.M., and Quarterman, R.S.T. (1986b). Piezocone Testing in
Overconsolidated Clays, Proceedings, 39th Canadian Geotechnical Conference, Ottawa, pp.
209-218.
Lunne, T., Lacasse, S., and Rad, N. S. (1992). General Report: Recent Developments in In-Situ
Testing, Proceedings, 12th International Conference on Soil Mechanics and Foundation
Engineering, Rio de Janiero, Vol. 4, pp. 2339-2403.
Lunne, T., Powell, J., Eidsmoen, T., and Quarterman, R. (1985). Comparison of Piezocones in
Overconsolidated Clays, Contract Report No. 84223-1, Norwegian Geotechnical Institute,
Oslo, 53 pp.
Lutenegger, A.J. (1995). Geotechnical Behavior of Overconsolidated Surficial Clay Crusts,
Transportation Research Record, No. 1479, National Academy Press, Washington D.C., pp.
61-74.
Lutenegger, A.J. (1997). Personal Communication.
Lutenegger, A.J. and Kabir, M.G. (1987). Pore Pressures Generated by Two Penetrometers in Clays,
Report No. 87-2, Department of Civil and Environmental Engineering, Clarkson University,
Potsdam, 45 pp.
Lutenegger, A.J. and Kabir, M.G. (1988). Interpretation of Piezocone Results in Overconsolidated
Clays, Proceedings, Penetration Testing in the U.K., Thomas Telford, London, pp. 147-150.
Lutenegger, A.J., Kabir, M.G., and Saye, S.R. (1988).Use of Penetration Tests to Predict Wick Drain
Performance in a Soft Clay, Proceedings, International Symposium on Penetration Testing,
J.D. Ruiter (ed), A.A. Balkema, Rotterdam, The Netherlands, Vol. 2, pp. 843-848.
Lutenegger, A.J. and Timian, D.A. (1986a). Flat-Plate Penetrometer Tests in Marine Clays,
Proceedings, 39th Canadian Geotechnical Conference, Ottawa, 9 pp.
Lutenegger, A.J. and Timian, D.A. (1986b). In-situ Tests with Ko Stepped Blade, Proceedings, Use
of In-Situ Tests in Geotechnical Engineering (In-Situ 86), ASCE, Geotechnical Special
Publication No. 6,New York, pp. 730- 751.
Mahar, .J. and O'Neill, M.W. (1983). Geotechnical Characteristics of Desiccated Clay, Journal of
Geotechnical Engineering, ASCE, Vol. 109, No. 1, pp. 56-71.
Malone, P.G., Comes, G.D., Chrestman, A.M., Cooper, S.S., and Franklin, A.G. (1992). Cone
Penetrometer Surveys of Soil Contamination, Environmental Technology, Usmen and Acar,
eds, Balkema, Rotterdam, pp. 251-257.
Manassero, M. (1994). Hydraulic Conductivity Assessment of Slurry Wall Using Piezocone, Journal
of Geotechnical Engineering, ASCE, Vol. 120, No. 10, pp. 1725-1746.
Marsland, A. (1974). Comparison of the Results from Static Penetration Tests and Large In-situ
Plate Tests in London Clay, Proceedings, European Symposium on Penetration Testing,
ESOPT, Stockholm, Vol. 2:1, pp. 245-252.
Marsland, A. and Quarterman, R.S.T. (1982). Factors Affecting the Measurements and Interpretation
of Quasi Static Penetration Tests in Clay, Proceedings, Second European Symposium on
Penetration Testing, Vol. 2, Amsterdam, pp. 697-702.
Marton, R., Taylor, L., and Wilson, K. (1988). Development of an In-Situ Subsurface Radioactivity
Detection System - the Radcone, Proceedings, Waste Management '88, University of
Arizona, Tucson.
138
Massarsch, R. (1986). Acoustic Penetration Testing, Fourth International Geotechnical Seminar on
Field Instrumentation and In Situ Measurements, Singapore, pp. 71-76.
Mayne, P.W. (1988). Determining OCR of Clays from Laboratory Strengths, Journal of
Geotechnical Engineering, ASCE, Vol. 114, No. GT1.
Mayne, P.W. (1991). Determination of OCR in Clays by Piezocone Tests Using Cavity Expansion
and Critical State Concepts, Soils and Foundations, Vol. 31, No. 2, pp. 65-76.
Mayne, P.W. (1995). Undrained Plastic Modulus from Original Cam Clay, Journal of Geotechnical
Engineering, ASCE, Vol. 121, No. 5, pp. 448-451.
Mayne, P.W. and Bachus, R.C. (1988). Profiling OCR in Clays by Piezocone Soundings,
Proceedings, International Symposium on Penetration Testing, J.D. Ruiter (ed), A.A.
Balkema, Rotterdam, The Netherlands, Vol. 2, pp. 857-864.
Mayne, P.W., Burns, S.E., Hegazy, Y., and Kates, G. (1996). Report of Seismic Piezocone and Flat
Dilatometer Tests in Pemiscot and Dunklin Counties, Geotechnical Report, Georgia Institute
of Technology, Atlanta, GA.
Mayne, P.W. and Harris, D.E. (1993). Axial Load-Displacement Behavior of Drilled Shaft
Foundations in Piedmont Residuum, Technical Report to: Federal Highway Administration,
Turner-Fairbanks Highway Lab, McLean, Virginia, FHWA Reference No. 41-30-2175,
Georgia Tech.
Mayne, P.W., Kulhawy, F., and Kay, J.N. (1990). Observations on the Development of Pore-Water
Stresses During Piezocone Penetration in Clays, Canadian Geotechnical Journal, Vol. 27,
No. 4, pp. 418-428.
Mayne, P.W., Mitchell, J.K., Auxt, J.A., and Yilmaz, R. (1995). U.S. National Report on CPT,
Proceedings, International Symposium on Cone Penetration Testing (CPT '95), Linköping,
Sweden, Vol. 1, pp. 263-276.
Mayne, P.W. and Rix, G. J. (1993). Gmax-qc Relationships for Clays, ASTM Geotechnical Testing
Journal, Vol. 16, No. 1, pp. 54-60.
Mayne, P.W. and Rix, G.J. (1995). Correlations Between Shear Wave Velocity and Cone Tip
Resistance in Natural Clays, Soils and Foundations, Vol. 35, No. 2, pp. 107-110.
Mayne, P.W. and Stewart, H. E. (1988). Pore Pressure Behavior of Ko-Consolidated Clays, Journal
of Geotechnical Engineering, ASCE, Vol. 114, No. 11, pp. 1340-1346.
McLamore, V.R., Anderson, D.G., and Espana, C. (1978). Cross Hole Testing Using Explosive and
Mechanical Energy Sources, Dynamic Geotechnical Testing, STP 654, ASTM, Philadelphia,
PA, pp. 30-55.
Menge, P. and Van Impe, W. (1995). The Application of Acoustic Emission Testing With
Penetration Testing, Proceedings, International Symposium on Cone Penetration Testing
(CPT '95), Linköping, Sweden, Vol. 2, pp. 49-54.
Mimura, M., Shrivastava, A.K., Shibata, T., and Nobuyama, M. (1995). Performance of RI Cone
Penetrometers in Sand Deposits, Proceedings, International Symposium on Cone Penetration
Testing (CPT '95), Linköping, Sweden, Vol. 2, pp. 55-60.
Mitchell, J.K. (1988). New Developments in Penetration Tests and Equipment, Proceedings,
International Symposium on Penetration Testing, J.D. Ruiter (ed), A.A. Balkema, Rotterdam,
The Netherlands, Vol. 1, pp. 245-261.
Mitchell, J.K. (1993). Fundamentals of Soil Behavior, 2nd Edition, John Wiley and Sons, New York,
437 pp.
Mitchell, J.K. and Gardner, W.S. (1975). In-Situ Measurement of Volume Change Characteristics,
Proceedings, ASCE Specialty Conference on In-Situ Measurement of Soil Properties, Vol. 2,
Raleigh, pp. 279-345.

139
Mitchell, J.K., Vivatrat, V., and Lambe, T.W. (1977). Foundation Performance of Tower of Pisa,
Journal of Geotechnical Engineering, ASCE, Vol. 103, No. GT3, pp. 227-249.
Nash, D.F.T., Powell, J.J.M., and Lloyd, I.M. (1992). Initial Investigations of the Soft Clay Test Site
at Bothkennar, Geotechnique, Vol. 42, No. 2, pp. 163-181.
Naval Command. (1995). Laser Induced Fluorometry / Cone Penetrometer Technology
Demonstration Plan at the Hydrocarbon National Test Site, Report, prepared by Naval
Command, Control and Ocean Surveillance Center, San Diego.
Olie, J.J., Van Ree, C.C.D.F., and Bremmer, C. (1992). In-Situ Measurement by Chemoprobe of
Groundwater from In-Situ sanitation of Versatic Acid Spill, Geotechnique, Vol. 42, No. 1, pp.
13-21.
Olsen, R.S. (1994). Normalization and Prediction of Geotechnical Properties Using the CPT,
Technical Report GL-94-29, U.S. Army Waterways Experiment Station, Vicksburg, 323 p.
O'Neill, D.A., Baldi, G., and Della Torre, A. (1996). The Multifunctional Envirocone Test,
Proceedings, Advances in Site Investigation Practice, Thomas Telford, London, pp. 420-437.
Pane, V., Brignoli, E., Manassero, M., and Soccodato, C. (1995). Cone Penetration Testing in Italy,
Proceedings, International Symposium on Cone Penetration Testing, Vol. 1, Swedish
Geotechnical Society, Linkoping, pp. 101-114.
Parez and Fauriel (1988). Le Piézocône Améliorations Apportées à la Reconnaissance de Sols,
Revue Française de Géotech, Vol. 44, pp. 13-27.
Parker, L.V. (1992). Suggested Guidelines for the Use of PTFE, PVC, and Stainless Steel in
Samplers and Well Casings, Current Practices in Ground Water and Vadose Zone
Investigations, (ASTM STP 1118), D. M. Nielson and M. N. Sara, Eds., American Society
for Testing and Materials, pp. 217-229.
Piccoli, S. and Benoit, J. (1995). Geo-environmental Testing Using the Environmental Cone,
Proceedings, Geoenvironment 2000 (GSP 46), New Orleans, American Society of Civil
Engineers, New York, Vol. 1, pp. 93-104.
Pluimgraaf, D., Hilhorst, M, and Bratton, W.L. (1995). CPT Sensors for Bio-Characterization of
Contaminated Sites, Proceedings, International Symposium on Cone Penetration Testing,
Vol. 2, Linköping, Sweden, pp. 569-575.
Powell, J.J.M. (1997). Personal Communication.
Powell, J.J.M., Quarterman, R.S.T., and Lunne, T. (1988). Interpretation and Use of the Piezocone
Test in U.K. Clays, Proceedings, Penetration Testing in the U.K., Thomas Telford, London,
pp. 151-156.
Powell, J.J.M. and Uglow, M. (1988). The Interpretation of the Marchetti Dilatometer Test in UK
Clays, Proceedings, Penetration Testing in the U.K., Thomas Telford, London, pp. 269-273.
Price, R.H., Boyd, P.J.., Noel, J.S., and Martin, R.J. (1994). Relationship Between Static and
Dynamic Rock Properties in Welded and Non-welded Tuff, Rock Mechanics: Models and
Measurements, (P.P. Nelson and S.E. Laubach, eds.), Balkema Publishers, Rotterdam,
Netherlands, pp. 505-512.
Randolph, M.F. and Wroth, C.P. (1978). Analysis of Deformation of Vertically Loaded Piles,
Journal of the Geotechnical Engineering, ASCE Vol. 104, No. GT12, pp. 1465-1488.
Randolph, M.F. and Wroth, C.P. (1979 a). A Simple Approach to Pile Design and the Evaluation of
Pile Tests, Behavior of Deep Foundations, ASTM STP 670, Raymond Lundgren, Ed.,
American Society for Testing and Materials, Philadelphia, pp. 484-499.
Randolph M.F. and Wroth, C.P. (1979 b). An Analytical Solution for the Consolidation Around a
Driven Pile, International Journal for Numerical and Analytical Methods in Geomechanics,
Vol. 3, pp. 217-229.

140
Raschke, S.A. and Hryciw, R.D. (1997). Vision Cone Penetrometer (VisCPT) for Direct Subsurface
Soil Observation, Journal of Geotechnical and Geoenvironmental Engineering, Vol. 123,
No. 11, in press.
Robertson, P.K. (1990). Soil Classification Using the Cone Penetration Test, Canadian Geotechnical
Journal, Vol. 27, No. 1, pp. 151-158.
Robertson, P.K. and Campanella, R.G. (1983). Interpretations of Cone Penetration Tests, Canadian
Geotechnical Journal, Vol. 20, No. 4, pp. 718-745.
Robertson, P.K., Campanella, R.G., Brown, P.T., and Robinson, K.E. (1988a). Prediction of Wick
Drain Performance Using Piezometer Cone Data, Canadian Geotechnical Journal, Vol. 25, pp.
56-61.
Robertson, P.K., Campanella, R.G., Gillespie, D., and By, T. (1988b). Excess Pore Pressures and the
Flat Dilatometer Test, Proceedings, International Symposium on Penetration Testing, J.D.
Ruiter (ed), A.A. Balkema, Rotterdam, The Netherlands, Vol. 2, pp. 567-576.
Robertson, P.K., Campanella, R.G., Gillespie, D., and Grieg, J. (1986). Use of Piezometer Cone
Data, Proceedings, Use of In-Situ Tests in Geotechnical Engineering (In-Situ '86), GSP No. 6,
American Society of Civil Engineers, New York.
Robertson, P.K., Lunne, T., and Powell, J. (1996). Applications of Penetration Tests for Geo-
Environmental Purposes, Proceedings, Advances in Site Investigation Practice, Thomas
Telford, London, pp. 407-420.
Robertson, P.K., Sully, J.P., Woeller, D.J., Lunne, T., Powell, J.J.M., and Gillespie, D.G. (1992).
Estimating Coefficient of Consolidation From Piezocone Tests, Canadian Geotechnical
Journal, Vol. 29, No. 4, pp. 539-550.
Ross, C.W., Walsh, J.W., Hartman, N.F., and Boswell, F.C. (1991). Development of a Field Worthy
Sensor System to Monitor Gaseous Nitrogen Transfer from Agricultural Cropland, Phase I
Final Report, DE-FC07-89ID12905, U.S. DOE Office of Industrial Technologies,
September.
Rowe, P.W. (1972). The Relevance of Soil Fabric to Site Investigation Practice, 12th Rankine
Lecture, Geotechnique, Vol. 22, No. 2, pp. 195-300.
Rowe, P.W. and Barden, L. (1966). A New Consolidation Cell, Geotechnique, Vol. 16, No. 2, pp.
162-170.
Roy, M., Blanchet, R., Tavenas, F., and La Rochelle, P. (1981). Behavior of a Sensitive Clay During
Pile Driving, Canadian Geotechnical Journal, Vol. 18, No. 1, pp. 67-85.
Roy et al. (1979). Behavior of a Sensitive Clay During Pile Driving, Proceedings, 32nd Canadian
Geotechnical Conference, Quebec, pp. 4.28-4.49.
Roy, M., Tremblay, M., Tavenas, F., and La Rochelle, P. (1982). Development of Pore Pressures in
Quasi-static Penetration Tests in Sensitive Clay, Canadian Geotechnical Journal, Vol. 19,
No. 1, pp. 124-138.
Santamarina, J.C and Fratta, D. (1997). Discrete Signals and Inverse Problems in Civil Engineering,
American Society of Civil Engineers Press, in press.
Sasaki, Y. and Koga, Y. (1982). Vibratory Cone Penetrometer to Assess the Liquefaction Potential
of the Ground, Proceedings, 14th Panel on US-Japan Panel on Wind and Seismic Effects,
Washington, D.C.
Sasaki, Y., Koga, Y., Itoh, Y., Shimazu, T., and Kondo, M. (1985). In Situ Tests for Assessing
Liquefaction Potential Using Vibratory Cone Penetrometer, Proceedings, 17th Joint Meeting
US-Japan Panel on Wind and Seismic Effects, Tsukuba, Japan.
Schmertmann, J.H. (1978). Guidelines for Cone Penetration Tests: Performance and Design, Report
FHWA-TS-78-209, Federal Highway Administration, Washington, D.C.

141
Schnaid, F., Sills, G.C., Soares, J.M., and Nyirenda, Z. (1997). Predictions of the Coefficient of
Consolidation from Piezocone Tests, Canadian Journal of Geotechnical Engineering, Vol.
34, pp. 315-327.
Senneset, K., Janbu, N., and Svanφ, G. (1982). Strength and Deformation Parameters from Cone
Penetration Tests, Proceedings, Second European Symposium on Penetration Testing, Vol.
2, (ESOPT-2, Amsterdam), Balkema, Rotterdam, pp. 863-870.
Senneset, K., Sandven, R., and Janbu, N. (1989). Evaluation of Soil Parameters from Piezocone
Tests, Transportation Research Record 1235, Washington, D.C., pp. 24-37.
Shibata, T., Mimura, M., Shrivastava, A.K., and Nobuyama, M. (1992). Moisture Measurement by
Neutron Moisture Cone Penetrometer: Design and Application, Soils and Foundations, Vol.
32, No. 4, pp. 58-67.
Sills, G.C., Almeida, M.S.S., and Danziger, F.A.B. (1988). Coefficient of Consolidation From
Piezocone Dissipation Tests in a Very Soft Clay, Proceedings, International Symposium on
Penetration Testing, J.D. Ruiter (ed), A.A. Balkema, Rotterdam, The Netherlands, Vol. 2, pp.
967-974.
Skempton, A.W., Schuster, R.L., and Petley, D.J. (1969). Joints and Fissures in the London Clay at
Wraysbury and Edgware, Geotechnique, Vol. 19, pp. 205-217.
Smits, F.P. (1982). Penetration Pore Pressure Measured With Piezometer Cones, Proceedings,
Second European Symposium on Penetration Testing, Vol. 2, Amsterdam, pp. 871-876.
Soares, M.M., Lunne, T., and Lacasse, S. (1987). In-Situ Site Investigation Techniques and
Interpretation for Offshore Practice. Interpretation of Dissipation Tests in Onsφy Clay,
Internal Report 40019-15, Norwegian Geotechnical Institute, Oslo, November, 45 pp.
Soderburg, L.O. (1962). Consolidation Theory Applied to Foundation Pile Time Effects,
Geotechnique, Vol. 12, pp. 217-225.
Stienstra, P. and van Deen, J.K. (1994). Field Data Collection Techniques - Unconventional
Sounding and Sampling Methods, Proceedings, 20 Year Jubilee Symposium of the
Ingeokring 1994, Delft, the Netherlands, Balkema, Rotterdam, pp. 41-56.
Strutynsky, A.I. and Sainey, T.J. (1992). Use of Piezometric Cone Penetration Testing and
Penetrometer Ground Water Sampling for Volatile Organic Contaminant Plume Detection,
Current Practices in Ground Water and Vadose Zone Investigations, (ASTM STP 1118),
David M. Neilson and Martin N. Sara, Eds., American Society for Testing and Materials, pp.
199-214.
Strutynsky, A.I., Sandiford, R.E., and Cavaliere, D. (1991). Use of Piezometer Cone Penetration
Testing with Electrical Conductivity Measurements for the Detection of Hydrocarbon
Contamination in Saturated Granular Soils, Current Practices in Ground Water and Vadose
Zone Investigations, (ASTM STP 1118), David M. Neilson and Martin N. Sara, Eds.,
American Society for Testing and Materials.
Sully, J.P. (1991). Measurement of In Situ Lateral Stress During Full-Displacement Penetration
Tests, Ph.D. Thesis, Department of Civil Engineering, University of British Columbia, 470
pp.
Sully, J.P. and Campanella, R.G. (1994). Evaluation of Field CPTU Dissipation Data in
Overconsolidated Fine-Grained Soils, Proceedings, 13th International Conference on Soil
Mechanics and Foundation Engineering, Vol. 1, New Delhi, pp. 201-204.
Sully, J. P. and Campanella. R.G. (1995). Evaluation of In-situ Anisotropy from Crosshole and
Downhole Shear Wave Velocity Measurements, Geotechnique, Vol. 45, No. 2, pp. 267-282.
Sully, J.P. and Echezuria, H.J. (1988). In-Situ Density Measurement with Nuclear Cone
Penetrometer, Proceedings, International Symposium on Penetration Testing, J.D. Ruiter (ed),
A.A. Balkema, Rotterdam, The Netherlands, Vol. 2.
142
Tavenas, F., Leblond, P., Jean, P., Leroueil, S. (1983a). The Permeability of Natural Soft Clays. Part
I: Methods of Laboratory Measurement, Canadian Geotechnical Journal, Vol. 20, No. 3, pp.
629-644.
Tavenas, F., Leblond, P., Jean, P., Leroueil, S. (1983b). The Permeability of Natural Soft Clays. Part
II: Permeability Characteristics, Canadian Geotechnical Journal, Vol. 20, No. 3, pp. 645-
660.
Tavenas, F. Leroueil, S., and Roy, M. (1982). The Piezocone Test in Clays: Use and Limitations,
Proceedings, 2nd European Symposium on Penetration Testing, Amsterdam, Vol. 2, pp. 889-
894.
Teh, C.I. (1987). An Analytical Study of the Cone Penetration Test, D.Phil. Thesis, Department of
Civil Engineering, Oxford University, UK.
Teh, C.I. and Houlsby, G.T. (1989). An Analytical Study of the Cone Penetration Test in Clay, Soil
Mechanics Report No. SM099/89, University of Oxford, Oxford, U.K., 47 pp.
Teh, C.I. and Houlsby, G.T. (1991). An Analytical Study of the Cone Penetration Test in Clay,
Geotechnique, Vol. 41, No. 1, pp. 17-34.
Teparaksa, W. (1987). Use and Application of Penetration Tests to Assess Liquefaction Potential of
Soils, PhD Thesis, Department of Civil Engineering, Kyoto University, Kyoto, Japan.
Teparaksa, W. (1990). Cone Penetration Resistance of Sandy Soil and Triaxial Liquefaction
Strength, Proceedings, 22nd Annual Meeting of the Japanese Society of Soil Mechanics and
Foundation Engineering, pp. 587-590.
Theriault, G.A., Newbery, R., Andrews, J.M., Apitz, S.E., and Lieberman, S.H. (1992). Fiber Optic
Fluorometer Based on a Dual Wavelength Laser Excitation Source, Proceedings, OE/Fibers
'92, Boston, September.
Tokimatsu, K. (1988). Penetration Tests for Dynamic Problems, Proceedings, International
Symposium on Penetration Testing, J.D. Ruiter (ed), A.A. Balkema, Rotterdam, The
Netherlands, Vol. 1, pp. 117-136.
Torstensson, B.-A. (1975). Pore Pressure Sounding Instrument, Proceedings, ASCE Specialty
Conference on In Situ Measurement of Soil Properties (ISMOSP), Raleigh, N.C., Vol. 2, pp.
48-54.
Torstensson, B.-A. (1977). The Pore Pressure Probe, Nordiske Geotekniske M∅te, Oslo, Paper No.
34, pp. 34.1 to 34.15.
Torstensson, B.-A. (1984). A New System for Ground Water Monitoring, Ground Water Monitoring
Review, Vol. 4, No. 4, pp. 131-138.
Tringale, P.T. and Mitchell, J.K. (1982). An Acoustic Cone for Site Investigations, Proceedings,
Second European Symposium on Penetration Testing, Amsterdam, pp. 909-914.
Tumay, M.T. and Acar, Y.B. (1985). Piezocone Penetration Testing in Soft Cohesive Soils, Strength
and Testing of Marine Sediments: Laboratory and In Situ Measurements, American Society
for Testing and Materials, Special Technical Publication 883, Philadelphia, pp. 72-82.
Tumay, M.T., Acar, Y.B., Cekirge, M.H., and Ramesh, N. (1985). Flow Field Around Cones in
Steady Penetration, Journal of Geotechnical Engineering, ASCE, Vol. 111, No. 2, pp. 193-
204.
Tumay, M.T., Acar, Y., and Deseze, E. (1982). Soil Exploration in Soft Clays with the Quasi-Static
Electric Cone Penetrometer, Proceedings, 2nd European Symposium on Penetration Testing,
Amsterdam.
Vesic, A.S. (1972). Expansion of Cavities in Infinite Soil Mass, Journal of the Soil Mechanics and
Foundation Division, American Society of Civil Engineers, Vol. 98, No. SM3, pp. 265-290.

143
Vesic, A.S. (1975). Bearing Capacity of Shallow Foundations, Chapter 3, Foundation Engineering
Handbook, H, Winterkorn and H.Y. Fang, eds., Van Nostrand Reinhold Company, New
York, pp. 121-147.
Villet, W.C.B., Mitchell, J.K., and Tringale, P.T. (1981). Acoustic Emission Generated During the
Quasi-Static Cone Penetration of Soils, Acoustic Emission in Geotechnical Engineering
Practice, Drnevich, V.P. and Gray, R.E., eds, ASTM STP 750, pp. 174-193.
Whittle, A.J. (1987). A Constitutive Model for Overconsolidated Clays with Application to the
Cyclic Loading of Friction Piles, Sc.D. Thesis, Department of Civil Engineering,
Massachusetts Institute of Technology.
Wissa, A.E., Martin, R.E., and Garlanger, J.E. (1975). The Piezometer Probe, Proceedings, ASCE
Specialty Conference on In-Situ Measurement of Soil Properties, Vol. 1, Raleigh, NC, pp.
536-545.
Woeller, D.J., Weemees, I., Kohan, M., Jolly, G., and Robertson, P.K. (1991a). Penetration Testing
for Groundwater Contaminants, Proceedings, Geotechnical Engineering Congress, ASCE,
Boulder, CO, pp. 76-83.
Woeller, D.J., Weemees, I., Kurfurst, P.J., and Robertson, P.K. (1991b). Penetration Testing for
Arctic Soil and Permafrost Conditions, Proceedings, 44th Canadian Geotechnical
Conference, Vol. 1, Calgary, pp. 44-1 to 44-7.
Wroth, C.P. (1984). The Interpretation of In-Situ Soil Tests (24th Rankine Lecture), Geotechnique,
Vol. 34, No. 4, pp. 449-489.
Wroth, C.P. and Houlsby, G.T. (1985). Soil Mechanics - Property Characterization and Analysis
Procedures, Proceedings, 11th International Conference on Soil Mechanics and Foundation
Engineering, San Francisco, Vol. 1, pp. 1-54.
Yale, D.P. and Jamieson, W.H. (1994). Static and Dynamic Mechanical Properties of Carbonates,
Rock Mechanics: Models and Measurements, (P.P. Nelson and S.E. Laubach, eds.), Balkema
Publishers, Rotterdam, Netherlands, pp.463-471.

144