Stokoe Et Alii 2004 PDF

Proceedings ISC-2 on Geotechnical and Geophysical Site Characterization, Viana da Fonseca & Mayne (eds.
)
© 2004 Millpress, Rotterdam, ISBN 90 5966 009 9
Some contributions of in situ geophysical measurements to solving

geotechnical engineering problems
Kenneth H. Stokoe, II
University of Texas, Austin, Texas, USA.
Sung-Ho Joh
Chung-Ang University, Seoul, South Korea
Richard D. Woods
University of Michigan, Ann Arbor, Michigan, USA
Keywords: geophysics, geotechnics, seismic testing, in situ tests, body waves, surface waves, case histories
ABSTRACT: This paper focuses on one in situ geophysical method, seismic measurements. Seismic (stress
wave) measurements have been used for more than 50 years in geotechnical engineering, primarily in the ar-
eas of soil dynamics and geotechnical earthquake engineering. In the past 30 years, their role has steadily in-
creased to the point where they also play an important part in characterizing sites, materials and processes for
non-dynamic problems. Case histories and applications are presented to highlight some examples.
methods are noninvasive which make them well

1 INTRODUCTION
suited and cost effective in profiling spatially and
The geotechnical engineer has always been faced temporally.
with the problem of characterizing near-surface ma- The geotechnical engineering profession has not
terials. The near-surface region is often within 10 to adopted other geophysical methods as rapidly as
100 m of the ground surface. Traditionally, field ex- seismic methods. One reason is that seismic methods
ploration programs have involved boring, sampling, directly measure a mechanical property, the initial
and penetration testing. In the 1960s, in situ geo- slope of the stress-strain relationship, which is used
physical measurements began to be employed in in the solution of many geotechnical engineering
geotechnical engineering. This work primarily in- problems. However, new demands in geoenviron-
volved seismic (stress wave) measurements which mental, geotechnical, and military applications are
were adapted from exploration geophysics. Seismic constantly increasing the need for improved and
measurements were used to characterize geotechni- higher-resolution site characterization methods.
cal sites (e.g. layering, top of bedrock, depth to wa- Geophysical methods have an important role to fill
ter table) and geotechnical materials (e.g. stiffnesses in these areas. Technical papers in the Proceedings
in shear and compression). The real demand for of the First International Conference on Site Charac-
seismic measurements grew out of the need to terization (Robertson and Mayne, 1998) emphasize
evaluate the dynamic properties of near-surface this point.
soils, specifically the shear-wave velocity, Vs. Vs is
a key parameter in soil dynamics and geotechnical 1.1 Organization
earthquake engineering. Today, however, in situ The information and examples presented in this pa-
seismic measurements are used in many more appli- per focus on the use of in situ seismic measure-
cations as discussed herein. ments. However, this material demonstrates the
The discipline of geophysics and in situ geo- relevance of geophysical methods in geotechnics. A
physical measurements encompass much more than brief background on stress waves is presented in
seismic methods. Other geophysical methods in- Section 2 to facilitate subsequent discussions. An
clude electrical, magnetic, electromagnetic, ground- overview is presented in Section 3 of four noninva-
penetrating radar, and gravity. All of these methods sive surface-wave methods. The reason is that the
offer the geotechnical engineer new and improved in use of this generalized method is rapidly increasing
situ techniques to characterize sites, materials and in geotechnical engineering. Case histories and ap-
processes. These opportunities arise from the strong plications are presented in Section 4 that involve
theoretical bases upon which geophysical methods geosystems loaded statically as well as dynamically.
are founded, the complementary physical principles The use of in situ Vs measurements in evaluating
that support various field tests, and the ability to per- sample disturbance and in predicting nonlinear soil
form the same basic measurement in the laboratory response is discussed in Section 5 followed by con-
and in the field. Furthermore, many geophysical clusions in Section 6.
97
2 BACKGROUND ON STRESS WAVES AND Dilatation Compression
Undisturbed Medium
TRADITIONAL SEISMIC METHODS
2.1 Types of Stress Waves • • • •
Traditionally, in situ seismic testing has been con-

ducted by initiating a mechanical disturbance at W avelength, λ P
some point in the earth and monitoring the resulting Direction of Propagation
motions (stress waves) at other points in the earth.

The modes of propagation most often used are body a. Compression (P) wave
waves, compression and shear waves, and one type
of surface wave, the Rayleigh wave. These waves, in U ndisturbed M edium
terms of their far-field particle motions, are illus-

trated in Figure 1. Compression waves (P waves)
have particle motion parallel to the direction of wave
propagation, while shear waves (S waves) have par-
ticle motion perpendicular to the direction of wave W avelength, λ S
propagation. Rayleigh waves (R waves) exist be- D irection of P ropagation
cause of the exposed ground surface. R waves have

particle motions that are a combination of vertical b. Shear (S) wave
(shear) and horizontal (compression) motions. Near
the surface of a uniform material, R waves create Undisturbed Medium
particle motion that follows a retrograde elliptical
pattern as illustrated in Figure 1c. The decay with
depth of the vertical and horizontal components of
R-wave particle motion is illustrated in Figure 2.
The depth axis is normalized by the Rayleigh wave-
Wa velength, λR
length, λR. It is interesting to see in Figure 2 that the
Direction of Propagation
horizontal component changes sign at a normalized
depth around 0.15. The meaning of this change in c. Rayleigh (R) wave
sign is that R-wave particle motion changes from a ȱ
retrograde ellipse to a prograde ellipse in a uniform Figure 1. Wave propagation modes: body waves within a uni-
half-space. form, infinite medium and Rayleigh waves along the surface
of a uniform half space (after Bolt, 1976)
2.2 Stress Wave Velocities
Amplitude at Depth z
Stress waves are non-dispersive in a uniform elas- Normalized Motion =
Amplitude at Surface
tic medium. The term non-dispersive indicates that
the propagation velocity is independent of fre-
λR
z
quency. These waves are also considered non-
dispersive in low-loss homogeneous soil and rock at
,
Wavelength
small strains and low frequencies. However, strati-

Depth
graphy and other forms of heterogeneity cause fre-

quency-dependent velocity. (This dependency is the
fundamental premise on which surface-wave testing
Normalized Depth =
is based as noted in Section 3.) The far-field veloci-

ties of stress waves depend on the stiffness and mass
density of the material as,
P-wave velocity:
4 Figure 2. Variation in normalized particle motions with nor-
B+ G
M 3 = E (1 − ν) malized depth for Rayleigh waves propagating along a uniform
VP = = (1) half space (from Richart et al., 1970)
ρ ρ ρ (1 + ν)(1 − 2ν)
where ρ is the mass density and M, B, G and E are
S-wave velocity: the constrained, bulk, shear, and Young’s moduli,
VS =
G
(2) respectively, and ν is Poisson’s ratio. For a homoge-
ρ neous, isotropic material, compression and shear
wave velocities are related through Poisson’s ratio,
ν, as,
98 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

3.0
1− ν
VP = VS (3)
0 .5 − ν
2.5 Constrained
Normalized Wave Velocity, V/VS

Compression W ave, V P
The far-field velocity of the Rayleigh wave, VR, is
related to the velocities of P and S waves as (Achen- 2.0
bach, 1975),
1.5
1 1
2 2 2 2 2 2
ª §V · º ª §V · º ª §V · º Shear W ave, V S
«2 − ¨ R ¸ » − 4 ⋅ «1 − ¨ R ¸¸ » «1 − ¨ R ¸ » =0 (4)
« ¨© VS ¸
¹ » « ¨© VP ¹ » « ¨© VS ¸
¹ » 1.0
¬ ¼ ¬ ¼ ¬ ¼
Rayleigh W ave, V R
0.5
A good approximation for VR in terms of Vs and
Poisson's ratio is (modified from Achenbach, 1975),
0.0
0.1 0.2 0.0 0.3 0.4 0.5 ȱ
0.874 + 1.117ν
VR ≅ VS (5) Figure 3. Relationship between stress wave velocities in a uni-
1 +ν form half space and Poisson’s ratio (from Richart et al., 1970)
These equations permit computing the relative 2.3 Wave Velocities and Degree of Saturation
values of VP, Vs and VR as a function of Poisson’s
ratio, as shown in Figure 3. At ν=0, VP=√2 Vs and The shear wave velocity is related to the shear stiff-
VR= 0.874 Vs. At ν= 0.5 (which theoretically repre- ness of the soil skeleton. In clean coarse sands,
sents an incompressible material; hence, an infinitely where capillary effects are negligible, the effective
stiff material), VP=∞ so that VP/Vs=∞. At ν=0.5, stress controls the shear stiffness, and the effect of
VR=0.955 Vs. The ratios of body wave velocities (VP saturation on shear wave velocity is only related to
/ Vs) typically determined with small-strain seismic changes in mass density ρ, through Vs=√(G/ρ). The
tests on unsaturated soil and rock are around ~1.5 to relevance of capillary forces at interparticle contacts
2.0, which corresponds to Poisson's ratio ~0.10 to on shear stiffness increases with fines content. And,
0.33; therefore, the small-strain Poisson's ratio is the lower the degree of saturation, the higher G and
relatively low. Vs become (Cho and Santamarina, 2001).
It is important to note that the S-wave velocity is On the other hand, P-wave velocity is controlled
the same in an infinite medium as in a rod (torsional by the constrained modulus, M=B+4G/3. Therefore,
motion). However, the longitudinal P-wave velocity the fluid and the granular skeleton contribute to VP.
is different, being VP=√(M/ρ) in an infinite medium For degrees of saturation, Sr, less than about 99 per-
and VL=√(E/ρ) in a rod. The “L” denotes a longitu- cent, P-wave velocity is controlled by the stiffness of
dinal wave. The relationship between VL and E is for the soil skeleton in constrained compression in the
tests in which wavelengths are much greater than the same fashion as shear waves; that is, the main influ-
radius of the rod. For shorter wavelengths, VL de- ence of water on VP over this range in Sr comes from
creases as frequency increases. Also, wave velocity, unsaturated conditions which impact the soil skele-
V, wavelength, λ, and frequency of excitation, f, are ton stiffness. However, if the degree of saturation
related for any type of stress wave as, equals 100 percent, the constrained modulus of this
two-phase medium is dominated by the relative in-
V=fλ (6) compressibility of the water in comparison to the
soil skeleton. The resulting value of VP varies with
It is also worth mentioning that the terms “elas- the void ratio or porosity, n, the bulk stiffness of the
tic” and “small strain” are often used to describe material that makes the grains, Bg, and the bulk
stress waves and associated propagation velocities stiffness of the fluid, Bf. The bulk stiffness of the
and moduli when dealing with in situ seismic meas- fluid phase is very sensitive to the presence of air.
urements. These terms are used because transient Therefore, when the degree of saturation Sr is about
mechanical disturbances created in situ during test- 99.5 to 100 percent, the value of VP is very sensitive
ing generate stress waves in geotechnical materials to Sr. Figure 4 shows the typical influence of degree
that have maximum strain amplitudes less than of saturation on VP over this very small change in
0.0001%. As a result, the stress waves exhibit degree of saturation (the shear wave velocity re-
propagation behavior that is independent of strain mains unaffected by such a small change in satura-
amplitude and possess only a minor amount of en- tion). For completeness, it is also noted that the im-
ergy dissipation due to material damping. pact of Sr on Vs and VP of rock is very small (only a
few percent change) for Sr going from zero to 100%.
Proceedings ISCʼ2 on Geotechnical and Geophysical Site Characterization, Viana da Fonseca & Mayne (eds.) 99
1500
Compression Wave Velocity, VP, m/s
1200
R a n g e fro m
900 V o id R a tio
C hanges a. One surface reflection arrangement: normal moveout (NMO)
600
V1
ic ic
300
V2 >V1
9 9 .4 9 9 .6 9 9 .8
ȱ
D e g re e o f S a tu ra tio n , S r , P e rc e n t b. Surface refraction testing
Figure 4. Typical variation in compression wave velocity with
degree of saturation changing from 99.4 to 100 % for sand (af-
ter Allen et al., 1980)
2.4 Traditional Field Methods

Field testing methods can be classified as active or Rayleigh Wave Propagation Particle
passive. Active-type methods are generally em- Motion
ployed in geotechnical engineering. In this case, a
wave is radiated into the medium from a source that c. Surface-wave testing
is energized as part of the test. Passive-type methods
are used less frequently. However, a passive system Figure 5. Generalized field arrangements used in noninstru-
can be selected when background noise can be used sive, active seismic methods
as the excitation source. Field testing methods can
also be classified as nonintrusive if all instrumenta- placed on the ground surface. Typically, compres-
tion is placed on the ground surface, or intrusive sion wave measurements are performed using either
when boreholes or penetrometers are used. The most mechanical sources that are vertically oriented or
common stress-wave based methods in field use to- explosive sources. Waves reflected from interfaces
day are briefly reviewed below. at depth are monitored with vertically-sensitive geo-
phones. The main purpose of testing is typically to
2.4.1 Nonintrusive, active methods identify and approximately locate key interfaces at
Nonintrusive methods have many advantages includ- depth.
ing: (1) elimination of the time and cost of drilling, Surface Refraction Method - The surface refrac-
(2) avoidance of potential environmental conse- tion method is an established geophysical method
quences of drilling, and (3) effective coverage of for nonintrusively identifying sediment stiffnesses
large areas. These methods include surface refrac- and layer interfaces at depth. The method is based
tion, surface reflection, and surface waves as illus- on the ability to detect the arrival of wave energy
trated in Figure 5. that is critically refracted from a higher velocity
Surface Reflection Method - The surface reflec- layer which underlies lower velocity sediment.
tion method is one of the oldest and most common Seismic signals are generated with an active source,
seismic methods. This method is well documented in and wave arrivals are detected on the surface with an
numerous textbooks in geophysics (e.g., Dobrin and array of receivers as shown in Figure 5b. As with the
surface reflection method, compression wave meas-
Savit, 1988, and Burger, 1992). The main principle
urements are typically performed using vertical me-
of the seismic reflection method is illustrated in Fig- chanical sources or explosives. The arrivals of re-
ure 5a which shows one arrangement of the source fracted waves on the ground surface are monitored
and receivers. Both the source and receivers are with vertically-sensitive geophones.

Surface-Wave Method – The surface-wave 1989, and Santamarina and Fratta, 1998). Third, P,
method can involve Rayleigh and Love waves, and SV and SH waves can be generated and measured.
testing has been conducted on land and offshore (SH waves are shear waves with particle motions in
(Stokoe et al., 1994, Stoll et al., 1994, Tokimatsu, the horizontal direction.) The main disadvantage in
1995, and Luke and Stokoe, 1998). Most testing in crosshole testing is the time and cost associated with
geotechnical engineering involves R waves. Several drilling boreholes; however, ongoing developments
variations of the generalized method are currently in penetrometer-deployed sources combined with ef-
being used or are under development as discussed in fectively deployed receivers promise efficient cross-
Section 3. The most common approach used on land hole implementations under the appropriate soil
is called the spectral-analysis-of-surface-waves conditions (e.g., Fernandez, 2000).
(SASW) method. This test method involves actively Downhole Method - In the downhole method, the
exciting Rayleigh wave energy at one point and times for compression and shear waves to travel be-
measuring the resulting vertical surface motions at tween a source on the surface and points within the
various distances (receiver points) away from the soil mass are measured. Wave velocities are then
source as illustrated in Figure 5c. Measurements are calculated from the corresponding travel times after
performed at multiple source-receiver spacings travel distances have been determined. Travel dis-
along a linear array. The generalized method and tances are typically based on assuming straight ray
variations under development have tremendous po- paths between the source and receivers, although the
tential in geotechnical engineering and are therefore analysis may sometimes account for refracted travel
discussed in more detail in Section 3. paths. Figure 6b shows a conventional setup which
requires the drilling of only one borehole. One of the
2.4.2 Intrusive, active methods main advantages of the downhole method in com-
Intrusive active methods have been widely used in parison to the crosshole method is the need for only
geotechnical engineering, particularly in soil dynam- one borehole, so the cost is less. However, the dis-
ics and geotechnical earthquake engineering begin- advantage is that wave energy has to travel increas-
ning in the 1960s. The crosshole and downhole ingly larger distances as the depth of testing in-
methods were initially developed/adapted for use creases. In the writers’ experience, the optimum
followed by development of the seismic cone pene- testing depths range from about 10 m to 50 m unless
trometer (SCPT) and suspension logger. Each specialized personnel are involved. This depth is, of
method is briefly discussed below. course, dependent on the energy developed by the
Crosshole Method - Shear and compression wave source (various high-energy, mechanical sources
velocities are determined from time-of-travel meas- have been constructed, e.g., Liu et al., 1988).
urements between a source and one or more receiv- Seismic Cone Penetrometer - The cone penetro-
ers in the crosshole method. Testing is generally meter (CPT) is a well established tool for character-
conducted by placing the active source and receivers izing soil properties by measuring tip and side resis-
at the same depth in adjacent boreholes, as illus- tances on a probe pushed into the soil (Lunne et al.,
trated in Figure 6a. The times of travel from the 1997). The SCPT test is a modification of the cone
source to the receivers, called direct travel times, penetrometer test that allows measurement of shear
and the times of travel between receivers, called in- wave velocities in a downhole testing arrangement
terval travel times, are measured. Vertically oriented (Campanella et al., 1986). Seismic energy is gener-
impacts with mechanical sources are usually applied ated at the surface near the insertion point of the
to the borehole wall using a wedged source. Verti- cone. Usually a horizontal impact on an embedded
cally oriented receivers are used to monitor horizon- anvil is used to generate the SH waves. Travel times
tally propagating shear waves with vertical particle of the shear wave energy, either direct or interval,
motion; hence SV waves. Radially oriented receivers are measured at one or more locations above the
are used to monitor horizontally propagating P cone tip as shown in Figure 6c. After testing at one
waves. Compression and shear wave velocities are depth, the cone is penetrated further into the soil,
determined by dividing the borehole spacings at the and the test is repeated. One of the important bene-
testing depth by the respective travel times. The test fits of this method is that the seismic data can be
is repeated at multiple depths to compile a complete combined with the cone resistance values to build a
profile of shear and compression wave velocities clearer picture of both soil type, strength, stiffness,
versus depth. and layering. This is an excellent example of using
There are several strengths associated with cross- multiple techniques to investigate sites.
hole testing. First, the source and receivers are Suspension Logger - Logging tools can be low-
placed closed to the material/target to be evaluated, ered into a borehole to determine material properties
thus enhancing resolution. Second, measurements with stress waves, electromagnetic waves, gamma
can also be gathered along multiple inclined ray radiation, and other physical principles. The main
paths which can be processed together to render a limitations in borehole logging are the effect of cas-
tomographic image of the cross section (Menke, ing and drilling fluids on the measured response and
Source
Direct P
and S
Direct P
Waves
and S
Waves
3-D
Receivers
Source 3 –D Receivers
a. Crosshole test b. Downhole test
Source
Fluid-Filled
Borehole
b. Downhole Testing Arrangement

Direct
S Wave
Various Receiver 1
Propagation
Modes (body Receiver 2
Horizontal and interface
Receiver waves)
Source
c. Seismic cone penetrometer test d. Suspension logging test
Figure 6: Field arrangements used to perform intrusive seismic tests (from Stokoe and Santamarina, 2000)
the depth scanned by the technique relative to the 2.5 Additional Information
zone affected by drilling the borehole. One of the Most of the information presented above was ex-
more recent advances in borehole shear wave meth- tracted from the article by Stokoe and Santamarina,
ods is the suspension logger (Kitsunezaki, 1980, 2000. The information is briefly presented to facili-
Toksoz and Cheng, 1991, and Nigbor and Inai, tate the following discussion. However, much more
1994). This test is performed in a single, mud-filled information is available in the article and in the lit-
borehole. The device is lowered on a wire line into erature because of the strong theoretical bases upon
the borehole, and seismic energy is generated and which seismic and other geophysical measurements
received by a receiver array in the borehole as are founded. Textbooks such as Richart et al. (1970),
shown in Figure 6d. The shear and compression Aki and Richards (1980), Ward (1990), Sharma
wave velocities of the surrounding material are de- (1997), and Santamarina et al. (2001) are excellent
termined from the arrival times of these waves fol- references. Manuals such as ASCE Press (1998) and
lowing standard travel-time procedures. One of the NRC (2000) are also good references. Many impor-
advantages of this method is that the wire-line nature tant topics such as material damping, geotechnical
of the test allows for measurements at significant spreading, near-field effects, mode conversions, ef-
depths (hundreds of meters). Two drawbacks of the fects of stress state on wave velocities, inherent and
method are that it generally can not be performed in stress-induced anisotropies could not be covered
a steel or thick plastic casing if soft soils are to be herein but can be important in properly applying
tested and it does not work well within about 7 m of these seismic methods.
the ground surface.

3 OVERVIEW OF EVOLVING SURFACE- Gucunski and Woods, 1991, Al-Hunaidi, 1994, Al-
WAVE METHODS Hunaidi and Rainer 1995, and Joh, 1996).
Recently many other methods have been devel-
There is one field seismic method that is under oped for determination of phase-velocity dispersion
active development and deployment today. That curves in the generalized surface-wave method. The
method is the surface-wave method. The great in- four most widely used methods are the spectral-
terest in this method arises, in large part, from the analysis-of-surface-waves (SASW) method, the fre-
noninstrusive nature of the method combined with quency-wave number (f-k) spectrum method, the
the capabilities of imaging softer layers beneath multi-channel analysis of surface wave (MASW)
stiffer materials and testing large areas rapidly and method, and the continuous surface wave (CSW)
cost effectively. One test configuration is described method. Currently, the SASW method is used
in Section 2.4.1 and illustrated in Figure 5c. The around the world including Asia and Europe, the
method, in terms of a generalized method, can be di- MASW method is mostly used in the America and
vided into two basic parts: (1) monitoring Rayleigh- some Asian countries, and the CSW method is ac-
wave propagation along the ground surface (“field tively used in the United Kingdom, Australia and
testing”), and (2) empirical or numerical modeling some Asian countries. In Table 1, general features of
of the field measurements to yield the subsurface Vs each method are compared. Advantages and disad-
profile. The objective of field testing is to determine vantages of each method are summarized in Table 2.
the phase-velocity dispersion curve for the test site. In the following sections, the fundamental principles
This objective is discussed in detail below as are the of these methods are described and some important
different approaches to meeting this objective. issues are discussed.
Modeling of the field measurements, either empiri-
cal or numerical, can vary significantly from one 3.1.1 SASW method
analysis procedure to another. The strengths and In the SASW method, the dispersive characteristics
limitations of the modeling procedures are discussed of Rayleigh waves propagating through a layered
in Section 3.2. However, since the forward modeling material are measured and then used to evaluate the
theory of surface-wave propagation was introduced S-wave profile of the material (Stokoe et al., 1994).
by Thomson (1950) and Haskell (1953), empirical SASW measurements involve generating waves at
procedures should no longer be used to analyze the one point on the ground surface and recording them
field measurements. as they pass by two or more locations, as illustrated
in Figure 7a. All measurement points are arranged
3.1 Surface-Wave Techniques: Determination of along a single radial path from the source. Meas-
Phase-Velocity Dispersion Curves and urements are performed with several (typically six or
Associated Modeling Approaches more) sets of source-receiver spacings. In each set,
The steady-state, Rayleigh-wave method (Richart et the distance between the source and first receiver is
al., 1970) is one of the initial surface-wave methods kept equal to the distance between receivers. The
that was used in geotechnical engineering, which di- phase shift versus frequency relationship is meas-
rectly measure the wavelengths of Rayleigh waves ured for surface waves propagating between the re-
for the determination of phase velocities (velocities ceivers for each receiver spacing. A typical phase
associated with wavelengths or frequencies). The plot is shown in Figure 7b. From each phase plot,
two-station method (Landisman et al., 1968 and the phase velocity of the surface wave is calculated
Sato, 1971) is another surface-wave method based at each frequency knowing the frequency, phase an-
on the inter-station phase difference. Sato used a gle and distance between the receivers. The result is
transfer function to determine phase velocities of a plot of phase velocity versus frequency for a given
surface waves for a range of frequencies, and Lan- receiver spacing, called an individual dispersion
disman et al. used an inter-station cross-correlogram curve (Figure 7c). This procedure is repeated for all
to eliminate the adverse effects of low-energy source-receiver spacings used at the site and typi-
noises. In the 1980s, the University of Texas at Aus- cally involves significant overlapping in the disper-
tin (Heisey, et al., 1982, Nazarian and Stokoe, 1984, sion data between adjacent receiver sets. The indi-
and Stokoe and Nazarian, 1985) established the vidual dispersion curves from all receiver spacings
SASW method to determine phase-velocities of are combined into a single composite dispersion
Rayleigh waves. The SASW method was an innova- curve called the experimental or “field” dispersion
tive method to make faster and more efficient meas- curve (Figure 8). Once the composite dispersion
urements than any previous methods. In addition, the curve is generated for the site, an iterative forward
dynamic stiffness matrix method (Kausel and modeling procedure or an inversion analysis algo-
Roësset, 1981) became the theoretical basis for rithm is used to determine a shear-wave velocity
modeling the phase-velocity dispersion curves to profile by matching the field dispersion curve with
yield the Vs profile of the site. Advances continue to the theoretically-determined dispersion curve (Fig-
occur in this aspect of the test (Roësset, et al., 1991, ure 8).
Table 1. Key features of four, widely used surface-wave methods
Surface-Wave Method Key Features
x phase velocities from phase differences
x two to four receivers typically used
x superposed-mode phase velocity (apparent phase velocity)
SASW method x global property over receiver-spread area
x shear-wave velocity profile from the apparent phase velocities
x comprehensive forward modeling or inversion analysis
x impulsive source, swept-sine source, or random vibration source
x phase velocities from frequency-wave number spectrum
x multiple receivers (e.g. 128, 256, etc. receivers)
x fundamental and higher-mode phase velocities
f-k spectrum method
x global property over receiver-spread area
x shear-wave velocity profile from fundamental and higher modes
x impulsive source
x limited number of receivers (usually 24 receivers)
x fundamental and higher-mode phase velocities
x walk-away measurement
MASW method x same measurement configuration as common-midpoint reflection survey
x shear-wave velocity profile from the fundamental mode
x impulsive source or swept-sine source
x phase velocity from the average phase-angle slope over receiver-spread area
x four to six receivers used
x superposed-mode phase velocity (apparent velocity)
CSW method
x shear-wave velocity profile from the apparent velocities
x steady-state harmonic source
Table 2. Advantages and disadvantages of four, widely used surface-wave methods

Surface-Wave
Advantages Disadvantages
Method
x good sampling of shallow material
x multiple measurements using different source-receiver
x more sensitive measurements for layer stiff-
SASW method configurations are required
ness contrast, using apparent velocity inver-
x expertise required for phase unwrapping and forward modeling
sion analysis
x aliasing problem in wave number domain
x inaccurate mode separation in case of poor resolution in f-k
xdispersion curves separated for fundamental
spectrum
and higher modes
x large number of traces required for good resolution in wave-
f-k method x body-wave effect extracted
number domain
xdispersion curve global to the receiver-spread
x limitation due to topographic constraint and instrumentation
area
capability
x long measurement time
x aliasing problem in wave-number domain
MASW method x mode separation of surface waves
x use of the fundamental mode only in inversion analysis
x the effects of local anomalies minimized with
x dedicated inversion analysis required but not used
the use of average phase-angle slope
x near-field effects included
CSW method x no expertise required to calculate phase ve-
x exploration depth limited
locity
x frequency-content of vibrator is limited
x reliable measurements with controlled source
The SASW method is a simple technique that is locities, which correspond to the superposed mode
easily implemented in terms of field testing. The re- of higher-mode surface waves and body waves. De-
quirement of several measurements using different termination of apparent phase velocities incorporates
source-receiver configurations is time and labor in- phase unwrapping. In a complicated multi-layered
tensive. However, the multiple source-receiver con- system, phase unwrapping may be non-systematic
figurations employ multiple sources which are se- and sometimes requires expertise, which is cumber-
lected appropriately for the measured wavelength some to inexperienced personnel. However, the non-
range at each source-receiver configuration, There- systematic nature of phase unwrapping can be im-
fore, time- and labor-intensive measurements pref- proved by a signal processing technique such as the
erably lead to high-quality results. impulse-response filtration technique (Joh et al.,
The SASW method measures apparent phase ve- 1997) and Gabor spectrum.

Impact,
D D Swept Sinusoidal
(a) Vibration, or
Random Noise
Receiver 2 Receiver 1
FFT FFT
Spectral Amplitude
Spectral Amplitude
3 3
Phase
Difference :
2
φ –φ2 1
2 (c)
1 1 600
Phase Vel., m/sec

Experimental
0 0 400 Dispersion Curve
0 200 400 600 800 0 200 400 600 800
Frequency, Hz Frequency, Hz
200
0
(b) 1 10 100
Wavelength, m
D=1m
150 Deleted
Unwrapped Phase, deg

Phase Angle, deg
100
50 2000
0
-50 1000
-100 Phase Unwrapping
-150 0
0 200 400 600 800
0 200 400 600 800 Frequency, Hz
Frequency, Hz
Figure 7. Spectral-analysis-of-surface-waves (SASW) method: Calculation of phase velocities

150
100 D = 1m 800
Composite
Phase Angle, deg
50
0
Experimental Dispersion Curve
Phase Velocity, m/sec
-50
600
-100
-150
0 200 400 600 800
Frequency, Hz 400
150
100 D = 2m
Phase Angle, deg
50
0
200
-50
-100
-150 0 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 7 2 3
0 100 200 300 400
1 10 100
Frequency, Hz
150
100 D = 4m Wavelength, m Inversion

Phase Angle, deg
50 Analysis
0
-50
-100 Shear Wave Velocity, m/sec

-150 0 200 400 600 800
0 50 100 150 200
0
Frequency, Hz
150
100 D = 8m
Phase Angle, deg
50
0 5
-50
-100
-150
0 20 40 60 80 100
Frequency, Hz 10
150
Depth, m
100
D = 16m
Phase Angle, deg
50
-50
15
-100
-150
0 10 20 30 40 50
Frequency, Hz
150
100 D = 32m 20
Phase Angle, deg
50
-50
-100
-150 25
0 5 10 15 20 25
Frequency, Hz
Figure 8. Spectral-analysis-of-surface-waves (SASW) method: Determination of a dispersion curve and shear-wave velocity profile
Importantly, the SASW method uses the apparent disadvantage both for repetitiveness problems and
phase velocity dispersion curve along with source for required testing time. Also, the data acquisition
and receiver locations in the forward modeling or required for a large number of traces may be expen-
inversion analysis. The dynamic stiffness matrix sive, and the topographic constraints may limit the
method (Kausel and Roësset, 1981), which is the reliability of the measurements.
forward modeling algorithm used in the matching or
inversion process, can simulate the apparent phase 3.1.3 Multi-channel analysis of surface waves
velocity specific to the source-receiver configura- (MASW) method
tion. The inversion analysis based on apparent phase In the MASW method (Park et al., 1999, and Miller
velocities and the dynamic stiffness matrix method et al., 1999), a large array of time traces is measured
are key features of the SASW method, which im- using a swept-sine vibratory source or an impulsive
proves the reliability and accuracy of the shear-wave hammer, using the walk-away method (Figure 10).
velocity profile. The basic field configuration and acquisition proce-
dure for the MASW measurements is generally the
3.1.2 Frequency-wave number (f-k) spectrum same as the one used in conventional common mid-
method point (CMP) body-wave reflection surveys. In the
The frequency-wave number (f-k) method is another MASW method, the dispersion curve can be deter-
method which has been widely used in the geo- mined in two approaches: the swept-frequency re-
physical area and recently adopted for geotechnical cord approach and the frequency-wave number spec-
engineering applications. In the f-k method, the trum approach. In the swept-frequency record
propagating surface waves are measured at a signifi- approach shown in Figure 11a, the linear slope of
cant number of locations in a line with the source. each component of a swept-frequency record is de-
The measurement of propagating surface waves at termined and used to calculate the phase velocity.
many sequential locations in a line can reveal the The frequency-wave number spectrum shown in
wavelengths of the surface waves, which are basi- Figure 11b is almost the same approach as the fre-
cally the reciprocals of the wave numbers. Along quency-wave number spectrum method described in
with the frequency information obtained from the the Section 3.2. In the frequency-wave number spec-
time-domain waveform, the wave number informa- trum method, the ridge of the frequency-wave num-
tion is used to determine phase velocities. ber contour plot is identified and used to determine
To describe the f-k spectrum method, a total of the phase velocity from the relationship among fre-
256 synthetic seismograms were generated using the quency, wave number and phase velocity. On the
dynamic stiffness matrix method. The layered sys- other hand in the MASW method, the phase veloc-
tem shown in Figure 9a was used as the model pro- ity-frequency contour plot is first determined from
file. The stacked traces in the time-space domain are the frequency-wave number contour plot and then
shown in Figure 9b. These results can be trans- the ridge of the phase velocity-frequency contour
formed to the frequency-wave number domain by plot is identified for the calculation of the phase ve-
means of a 2-D FFT or slant-stack analysis locity corresponding to a frequency.
(McMechan and Yedlin, 1981). Figure 9c displays The MASW method uses only the fundamental
the frequency-wave number (f-k) contour plot trans- mode for the inversion analysis. For the site with a
formed from the time-space domain data in Figure normally dispersive dispersion curve, in which phase
9b. The fundamental and higher modes of the sur- velocities increase with increasing wavelength, the
face-wave propagation are identified by the ridge fundamental mode alone may be enough to resolve
analysis of the frequency-wave number contour plot. the layer stiffness reliably. However, for a typical
In the frequency-wave number contour plot, the geotechnical site with a more complex stiffness pro-
modes of the surface-wave propagation refer to dif- file, where the measured dispersion curve may be
ferent wave numbers for a given frequency, and cor- inversely dispersive or heavily fluctuating with a up-
respond to the loci identified by the ridge analysis. and-down pattern, the inversion analysis using the
Figure 9d is the phase-velocity dispersion curve de- fundamental-mode only can not work well (Toki-
termined from the modes identified in Figure 9c. matsu et al., 1992). To make the MASW method a
This approach to determine phase velocities from the reliable exploration method, it is crucial to incorpo-
frequency-wave number spectrum is called the f-k rate higher modes as well as the fundamental mode
spectrum analysis (Gabriels et al., 1987). in the inversion analysis. Recently, an effort to use
The f-k spectrum method is superior to any other higher modes in the inversion analysis was made
method in characterizing the fundamental and higher (Kansas Geological Survey, 2003).
modes from the measured surface wave. However,
the required use of numerous receivers is the main

(a) Shear-Wave Vel., m/sec Mass Density, kg/m3 Poisson's Ratio
0 200 400 600 800 0 500 1000 1500 2000 2500 0.0 0.1 0.2 0.3 0.4 0.5
0 0 0
5 5 5
10 10 10
15 15 15
Depth, m
Depth, m
Depth, m
20 20 20
25 25 25
30 30 30
35 35 35
Figure 9. Numerical simulation illustrating the frequency-wave number (f-k) spectrum method: (a) layered geotechnical site,
(b) synthetic seisograms, (c) f-k contour plots, and (d) phase-velocity dispersion curve.
Figure 10. Multichannel analysis of surface waves (MASW) method: walk-away method for measuring a large array of traces
(Kansas Geological Survey, 2003)
3.1.4 Continuous surface wave (CSW) method 180 degree and 180 degree due to the nature of Fou-
The continuous surface-wave (CSW) method is a rier transformation. Usually the phase wrapping can
geophysical exploration technique to evaluate the be easily identified in the plot of the source-to-
subsurface stiffness structure using a vibrator and receiver distance versus the phase angle, which has
more than four receivers, as depicted in Figure 12. more than two parallel lines. In the case with wrap-
Since the CSW method was initiated by British re- ped phase angles, the phase unwrapping operation
searchers (Matthews et al, 1996, and Menzies and can be applied to recover the original phase angles.
Matthews, 1996), it has been used in Europe, Aus- After the phase velocities are determined for all
tralia and some Asian countries. Unlike other sur- the excitation frequencies, the shear-wave velocity
face-wave methods, the CSW testing only uses a vi- profile can be determined from an empirical rela-
brator to generate surface waves. The application of tionship or an inversion analysis like the one for the
the CSW method is limited to shallow stiffness pro- SASW method. Presently an empirical analysis is
filing like compaction-quality control, because the used. One advantage of the CSW method is to use an
vibrator source does not generate enough energy for average phase-angle slope. The average of the
sampling deep material. phase-angle slop eliminates the local anomalies
The CSW testing shown in Figure 12 uses four geo- which may mislead the evaluation of the global S-
phones to measure the particle-velocity history of wave velocity profile. The other advantage of using
the ground for sinusoidal vibration induced by the the average phase-angle slope is that expertise is not
vibrator. The geophones are placed in a linear array needed in determining the phase velocities, which
with an equal spacing. Sometimes five or six geo- enables the automation of the phase-velocity calcu-
phones are used to improve the accuracy of the lation. The controlled source like a electro-
measurement. The time history of particle velocity mechanical vibrator allows reliable measurements
that is measured at each geophone is transformed only in the frequency range compliant to the vibrator
into the frequency domain by Fourier transforma- specification, and measurements of frequencies out
tion. And the phase angle is determined for an exci- of the vibrator specification lose reliability. This in-
tation frequency at each geophone. Then, the phase dicates that very shallow and very deep materials
angles are plotted against the location of the geo- can not be sampled, which turns out to be a disad-
phone, as shown in the lower portion of Figure 12. If vantage of the CSW method. Also, the measurement
the soil is homogeneous, the phase angle should time is usually long compared with other surface-
have the tendency to linearly increase with the dis- wave methods. Finally, the CSW method needs to be
tance from the source. In some cases, the phase an- more refined in that an inversion analysis specific to
gle goes over 180 degree or below -180 degree be- this method needs to be developed for reliable use in
cause, the phase angle is wrapped to fall between - the future.

(a) Analysis procedure: Swept-frequency approach (Park, Miller and Xia, 1999)
(b) Analysis procedure: Frequency-wave number approach

Figure 11 Multichannel analysis of surface waves (MASW) method: Analysis Procedure (Kansas Geological Survey, 2003)
e
m
Ti
Harmonic-Wave
Vibration
(frequency = f)
d
d
φ
e,
d Distance
gl
from Sourc
An
d e
e
as
Ph
Best-Fit Line Phase Velocity :

∆φ
∆r ∆r
v ph = 2πf
∆φ
Distance from
Source, r
Figure 12 Continuous surface wave (CSW) method
3.2 Theoretical Aspects Associated with the ceivers and the modes are well separated. Specifi-
Surface-Wave Methods cally, the f-k spectrum and MASW methods use a
Most surface-wave methods as applied today are so- large array of receivers which helps the separation of
phisticated in measurement and analysis, and there- surface-wave modes.
fore give rise to important issues in terms of theo- Figure 13 shows the differences among the nor-
retical background aspects. In this section, two mal-mode solution, 2-D solution and 3-D solution of
major issues related to the surface-wave methods are propagating surface waves. The 2-D solution is dif-
discussed. These issues are the forward modeling ferent from the 3-D solution in that the 2-D velocity
procedure and the inversion analysis; that is, how to is a superposed-mode velocity of plane Rayleigh-
theoretically calculate phase velocities for a given wave modes without body-wave interference
layered system and how to evaluate a shear-wave (Roësset et al, 1991). The dynamic stiffness matrix
velocity profile from a measured dispersion curve. method was used to calculate theoretical phase ve-
These issues are important topics for better and more locities for the layered systems in Figure 13. Case 1
reliable profiling of subsurface stiffness. is a soil system with increasing stiffness with depth,
and Case 2 is a soil system with a soft layer trapped
3.2.1 Higher-mode velocities and apparent ve- between a harder surface layer and a half-space.
locity Phase velocities were calculated for: (1) different
Fundamental and higher modes in surface-wave modes of plane Rayleigh waves, (2) the 2-D solution
propagation are the distinctive features in a multi- of a plane Rayleigh wave, and (3) the 3-D solution,
layered system. When surface waves propagate which is an apparent dispersion curve for the cylin-
through a multi-layered system, the stiffness of each drical Rayleigh wave. The contribution of different
layer affects the propagation of the surface waves modes to the simulated dispersion curve is presented
(Gucunski and Woods, 1992, and Al-Hunaidi and in Figure 13b. In the case of the soil system with in-
Rainer, 1995). The different stiffnesses in the layers creasing stiffness with depth, the apparent dispersion
may confine the stress waves in some layers or cause curve essentially coincides with the fundamental
multiple refractions and reflections, leading to dif- mode of the Rayleigh wave over the complete fre-
ferent ray paths, which result in different propaga- quency (hence wavelength) range. However, in Case
tion velocities even for the same frequency. Both the 2, the 2-D and 3-D solutions in the higher-frequency
transfer matrix method (Thomson, 1950, and Has- region (smaller-wavelength region) are not just from
kell, 1953) and the dynamic stiffness matrix method one mode but a superposition of several modes.
(Kausel and Roësset, 1981) can determine funda- Also, it is important to realize that: (1) the 2-D solu-
mental- and higher-mode velocities. These modes tion resides between the modes of a plane Rayleigh
correspond to plane waves in 2-D space, and can be wave, and (2) the 3-D solution may become lower
calculated from the eigenvector analysis of the trans- than the fundamental mode at low frequencies. This
fer matrix or the dynamic stiffness matrix. phenomenon is probably due to the multiple reflec-
The superposed mode in surface-wave propaga- tions and refractions of body waves. The comparison
tion is also an important feature, because this mode of the normal-mode solution, 2-D solution and 3-D
is actually generated during testing. The superposed solution in Figure 13b implies that the 3-D solution
mode corresponds to the 3-D wave propagating in a may be the closest to the actual measurements con-
cylindrical pattern, not like the planar pattern of a 2- taminated with body waves and higher-mode
D wave. This propagation is often observed when Rayleigh waves.
the source is close to the receivers and the wavefront
still has a significant cylindrical pattern. The super- 3.2.2 Inversion analysis to evaluate a shear-wave
velocity profile
posed mode does not fall into specific normal
In the surface-wave methods, two different catego-
modes, but is somewhere between the normal
ries of inversion analysis are available, dependent on
modes. The superposed mode is often called an ap-
the type of experimental dispersion curve. The first
parent velocity or an effective velocity. Calculated
one is to use the normal-mode solution. The f-k
or measured apparent phase velocities are dependent
spectrum method and the MASW method belong to
on the actual locations of the source and receivers.
this category. Most of the available inversion tech-
In surface-wave measurements, there are two dif-
niques are based on this approach (Hossian and
ferent approaches in terms of using surface-wave
Drnevich, 1989, Addo and Robertson, 1992, Yuan
modes. The SASW and CSW methods use the su-
and Nazarian, 1993, and Xia, et al., 1999). In this
perposed mode, because the source is close to the
category, the most crucial step is to well separate
receivers and mode separation of the measured sur-
fundamental and higher modes. Specially in the
face waves is not practical. The f-k spectrum and
high-frequency range, the phase velocities for
MASW methods use the fundamental and higher
modes, because the source is far enough from the re-

4.5 m 4.5 m 4.5 m 4.5 m
VS = 60 m/sec 0.3 m VS = 90 m/sec 0.3 m
VS = 90 m/sec 0.3 m VS = 60 m/sec 0.3 m
VS = 120 m/sec VS = 120 m/sec
Case 1 Case 2
a. Layered systems used to model two soil profiles
140 140
120 120

100 100
80 80
60 60
40 40
Normal modes Normal modes
3-D Solution 3-D Solution
20 20
Case 1 2-D Solution 2-D Solution
Case 2
0 0
0 10 20 30 40 50 0 10 20 30 40 50
Frequency, Hz Frequency, Hz
(b) Theoretical dispersion curves for each layered system
Figure 13 Contribution of different modes of the Rayleigh wave to the 3-D solution of wave propagation
normal modes are very close to each other. There- In inverting the dispersion curves determined in
fore, if the field measurement configuration is not the SASW or CSW method, it is more beneficial to
good enough to differentiate modes, the inversion incorporate information on the source and receiver
analysis may end up with misleading results. Figure locations rather than to neglect them by assuming
9c is a good example of difficulty in resolving lower the measured phase velocities are far-field velocities.
modes in the high-frequency region. In the high- Figure 14 compares the resulting shear-wave veloci-
frequency region, the evaluated mode is not neces- ties of two approaches: (1) the global inversion
sarily the fundamental mode, but is one of the higher analysis, and (2) the array inversion analysis (Joh,
modes. In this case, it is almost impossible to iden- 1996).
tify which higher-mode the measured mode belongs In the global inversion analysis, information on
to. Therefore, the inversion analysis using the nor- the source and receiver locations is ignored, and it is
mal-mode solution needs to focus on only the low- assumed that the receivers are located in the far
frequency region to avoid the problem in miscount- field. In this inversion analysis, the theoretical phase
ing the normal-mode number. velocities are calculated for receivers deployed at
In the second category of inversion analysis, the virtual locations of 2λ and 4λ ( λ is wavelength for
apparent phase-velocity dispersion curve is (or a specific frequency) and optimized to match the
should be) used. The SASW and CSW methods be- general trend of the dispersion curve. On the other
long to this category (Gucunski and Woods, 1991, hand, the array inversion analysis uses the phase ve-
Rix and Leipski, 1991, Tokimatsu et al, 1992, Joh, locities specific to the source and receiver locations,
1996, and Ganji et al., 1998). In this case, it is very and finds the optimum shear-wave velocity profile to
important to calculate the apparent theoretical phase match all the individual experimental dispersion
velocity. The apparent theoretical phase velocity curves with theoretical dispersion curves corre-
should be calculated using the exact locations of sponding to each source-receiver configuration. As
source and receivers, which can have a significant shown in Figure 14b, the array inversion analysis
influence on the resulting phase-velocity dispersion made a fit between five experimental dispersion
curve. Several sets of the experimental dispersion curves with the corresponding theoretical dispersion
curves from different sets of receiver combinations curves, while the global inversion analysis made a fit
should be included to evaluate the layer stiffness to follow the general trend of the experimental dis-
contrast more reliably.
a. Global inversion b. Array inversion
Figure 14 Comparison of global and array inversion analyses (Joh, 1996)
persion curve. The resulting shear-wave velocities 4 CASE HISTORIES AND APPLICATIONS
also show the superiority of the array inversion
analysis. The array inversion analysis was able to The purpose of this section is to present some case
produce the shear-wave velocity profile almost the histories and applications that demonstrate the im-
same as the exact model assumed to generate the portance of in situ geophysical methods to the solu-
synthetic dispersion curves. However, in some cases, tion of geotechnical engineering problems. The ex-
environmental noise and undesirable effects due to amples focus on the use of in situ seismic
lateral geologic variability may intervene into real measurements, but demonstrate the relevance of
measurements so that this approach may not work geophysical measurements in geotechnical engineer-
perfectly and needs to be applied with care. ing. The examples include problems that involve
geosystems loaded statically as well as dynamically
and loaded in the linear (small strain) and nonlinear
ranges.

4.1 Soil Modulus for Settlement Analysis and Soil Swiger acknowledged the benefits of using shear
Structure Interaction wave velocity (Vs) over compression wave velocity
Settlement predictions/calculations based on the ul- (Vp) because Vs is unaffected by the water table and
timate strength of soils have been made for a century calculation of shear modulus from shear wave veloc-
or more. However, beginning in the late 1960s in the ity requires only an estimate or measurement of soil
construction of nuclear power plants and other very density. Shear modulus is calculated using Equation
large and heavy structures, it was evident that geo- 2. Furthermore, for isotropic materials, shear
technical engineers had to find better ways of ana- modulus can be converted to other moduli including
lyzing the deformation behavior of soils including Young’s modulus or constrained modulus using
analyses of settlement and soil structure interaction. Equation 1. The simplified equation relating shear
However, the geotechnical community has made and Young’s moduli is,
very slow progress in pursuing a rational approach
to these analyses. The following four case histories E = 2G(1+Ȟ) (7)
describe some attempts to use elastic modulus de-
rived from seismic wave velocity measurements to Swiger further pointed out that shear modulus
estimate the settlement of foundations on sands, measured at low strain can be adjusted to larger
gravels, heavily overconsolidated clays, and soft strains through relationships provided by Hardin and
rock. Drnevich (1972a and 1972b) or as shown by Seed
(1969) in Figure 15. Based on the first cycle of a
large load test at the Brookhaven National Laborato-
4.1.1 Case history 1- settlement analysis of large
and heavy structures ries and on modulus derived from crosshole shear
In an early recognition of the prevailing irrational wave velocity at the site, he found reasonable
approach to settlement prediction based on ultimate agreement in back calculated Young’s modulus and
strength, William Swiger of Stone and Webster in E determined from seismic wave velocity at a strain
1974 (Swiger, 1974) suggested an improvement to level of 4 x 10-5 as shown in Table 3 for two values
the geotechnical practice of settle-ment prediction of Ȟ. At the time, Swiger decided to use a range in Ȟ
using elastic moduli derived from seismic waves. but noted that Ȟ = 0.3 seemed more reasonable. To-
His description of the problem and potential solution day, we realize that Ȟ in the range of 0.15 to 0.35 is
is presented here as the basis for a modest advance- appropriate for the soil skeleton.
ment over the past quarter century in use of elastic Table 3 Moduli calculated from load test and crosshole shear
modulus for settlement prediction. wave velocity (from Swiger, 1974)
In a rational approach to settlement (deformation)
Poisson’s Ratio Ȟ = 0.3 Ȟ = 0.45
prediction, stress and strain should be related
Seismic Modulus (E) 3.7 x 106 psf 4.2 x 106 psf
through modulus. The complication for soils is that Load Test Modulus (E) 3.9 x 106 psf 3.4 x 106 psf
soil is a nonlinear material, starting from very low
strain levels, so application of any approach using
modulus has to recognize and accommodate the
nonlinear behavior. Swiger determined, based on
calculations for five power plant structures, that the
average strain causing settlement under large struc-
tures was on the order of 10-3 throughout a depth
about equal to the minimum dimension of the loaded
area. He then outlined an approach for determining a
soil modulus at an appropriate strain level.
Swiger pointed out correctly that methods of
modulus determination requiring sampling of soils
and testing specimens in the laboratory suffered
from a major inescapable drawback, sample distur-
bance. (The subject of sample disturbance is pre-
sented in more detail in Section 5.0). He also noted
that techniques exist by which modulus could be
measured in situ and without disturbance, namely
seismic wave velocity measurements. Crosshole and Figure 15. Modulus-strain relations used by Swiger, 1974
downhole seismic tests were well established by the (from Seed, 1969)
mid-1970s and Swiger used them to determine
small-strain (10-6) soil moduli at multiple depths in At the site of the turbine room of the Shipping-
the ground. port nuclear power station, a site underlain by about
60 ft (18m) of medium dense to dense sand and
Figure16. Modulus profiles from observed settlements and
seismic measurements (from Swiger, 1974)
Figure 17. Typical seismic survey results at power plant site
(from Konstantinidis et al., 1986)
gravel, strain-adjusted modulus determined from
seismic wave velocity was compared to modulus Moduli from all methods used except moduli from
computed from observed settlement by Swiger as seismic crosshole tests overestimated the measured
shown in Figure 16. These moduli show very good settlement by a factor of at least two. Using the
agreement. moduli from seismic crosshole tests, the estimated
Since the time of Swiger’s paper, another impor- settlements were within +/- 15 % of the measured
tant seismic method has become available permitting settlements.
modulus profiles to be determined without the bore- It is noteworthy that the laboratory tests, although
holes or any ground disturbance, namely surface- performed with special refinements designed to
wave testing as discussed in Section 3. With this eliminate sample disturbance and conducted on care-
nondestructive and nonintrusive method, elastic fully sampled specimens, consistently produced un-
moduli profiles for homogeneous and layered soil realistically low estimates of soil stiffness. It was
sites can be readily obtained for the purpose of set- postulated that the highly overconsolidated soils at
tlement analysis and soil structure interaction. Sur- this site were more susceptible to disturbance during
face-wave testing mitigates one of the high-cost sampling than “average” soils. The overconsoli-
elements of crosshole and downhole seismic testing, dated state of this site, compared to a soft soil site,
namely boreholes. may have also added to the applicability of seismi-
cally determined moduli in this case.
4.1.2 Case history 2 – settlement analysis of a
The authors conclude that field-determined
power plant
moduli produce better estimates of soil compressi-
Konstantinidis et al. (1986) reported a case study in bility than laboratory tests, and that the modified
which moduli developed for prediction of settlement version of Swiger’s suggested method based on
of a power plant were based on the approach sug- seismic wave velocity measurements produced the
gested by Swiger, 1974. The site consisted of both most comprehensive and realistic assessment of set-
sand and clay layers of about equal thicknesses to a tlement.
depth of 200 ft (61 m). Seismic wave velocities
measured at the site are presented in Figure 17. 4.1.3 Case history 3 – settlement of a water tank
Other methods of estimating moduli for settlement John Burland, in his Bjerrum Lecture, focused on
prediction considered by Konstantinidis et al. (1986) the need for small-strain soil properties for many
included the CPT, pressuremeter testing (PMT) and geotechnical problems, including soil structure in-
laboratory tests including consolidation and triaxial teraction (Burland, 1989). He emphasized the
compression. Short-term settlement measurements nonlinear behavior of soils and promoted the use of
(initial elastic settlement) showed about 1 inch (0.68 strain-appropriate strength or stiffness for analysis.
to 1.19 inch) [17 to 30 mm] of settlement for Unit 1 Since soil strains in most geotechnical problems fall
in the range of 0.1% or smaller, Burland urged that
of a two-unit plant. Unit 2 did not have the same se-
geotechnical engineers recognize the importance of
quence of settlement measurements so it could not strain-appropriate soil properties. The development
be compared. of laboratory techniques capable of precise meas-
urement of small strains convinced him that the gap
between dynamic and static measurements of soil

stiffness was being closed. Previously, dynamic University (Briaud and Gibbens, 1994). Full-scale
measurements of soil stiffness gave values so much footings of five sizes / configurations were tested to
higher than static measurements that many engineers failure with detailed measurements of load-
discounted the dynamic measurements. However, settlement. Thirty-one predictors were bold enough
Burland cited cases where accurately determined to make first class predictions of settlement based on
static small-strain values of stiffness were compati- detailed characterization of the site. Three of the 31
ble with seismically measured stiffness, giving predictors used seismic wave velocities from cross-
greater credibility to dynamically measured values. hole tests to determine modulus for settlement pre-
As an example, he presented a case where Young’s diction.
modulus deduced from a static water tank loading The predictors used 22 different settlement pre-
test on Mundford Chalk produced stiffness very diction methods based on soil properties determined
nearly the same as those determined from a seismic by five field tests and two laboratory tests. Several
refraction survey. Figure 18 shows Young’s measures of accuracy of prediction were compared
modulus versus depth determined by three methods; with predictions including settlement at several
0.86-m diameter plate loading tests at seven depths, stages of loading and the factor of safety at ultimate
finite-element back calculation, and seismic refrac- load. Table 4 is a compilation of factors of safety for
tion. the five footings computed from predictions by the
31 predictors. Those who used seismically deter-
mined moduli were numbers 22, 23 and 28. Two of
those (numbers 22 and 28) were consistently better
than the mean of all predictors. The other predictor,
number 23, was better than the mean for the three
larger footings. Although this prediction event did
not specifically showcase settlement predictions
based on seismically determined moduli, it provides
further evidence that this method of prediction has
potential for future application.
Table 4. Factors of Safety F = Qf / Qd (measured design

load*/predicted design load) [*ultimate with FS =3] (from
Briaud and Gibbens, 1994)
Figure 18. Young’s modulus determined from three methods

for settlement analysis of a water trunk on Mundford Chalk
(from Burland, 1989)
Burland concluded that these results along with

others open up the way for a whole new area of
study linking dynamic and static deformation prop-
erties of soils. He speculated that studies of this
kind would lead to wider application of geophysical
measurements for determining in situ stiffness prop-
erties of geotechnical materials. The writers agree
wholeheartedly.
4.1.4 Case history 4 – settlement analysis of

footings
In 1994, the Geotechnical Division of ASCE held a
settlement prediction symposium in conjunction
with an ASCE Specialty Conference at Texas A&M
4.1.5 Summary The plan view of the site, Figure 20, shows the
A few successful demonstrations of the use of footprint of four new foundation blocks and loca-
seismically determined soil modulus for settlement tions for compaction grouting, chemical grouting
predictions have been presented. Other case histories and boreholes for crosshole tests. Figure 21 shows
can be found in conferences dealing with the pre- the shear wave velocity versus depth profiles deter-
failure deformation characteristics of geomaterials mined from crosshole seismic tests after each of the
such as Shibuya et al. (1994), Jardine et al. (1998), two stages of grouting. Compaction grouting
Jamiolkowski et al. (2001), and Di Benedetto et al. achieved the minimum shear wave velocity at this
(2003). The potential for use of this more rational location, but chemical grouting was performed as an
approach to determining soil stiffness has been con- added factor of safety. The crosshole tests con-
firmed, but not yet widely adopted. In some cases, firmed a significant increase in shear wave velocity
engineers report that obtaining the seismic wave ve- leading to successful operation of forging machines
locities is too expensive (Konstantinidis et al, 1986), at this site. No excessive settlements were observed
but with the development of surface-wave methods, and vibration levels throughout the plant were not
the cost of boreholes has been eliminated. The writ- noticeable.
ers hope that elimination of this cost impediment
will allow broader application of seismically deter-
mined moduli for geotechnical engineering pur-
poses.
4.2 Crosshole Seismic Velocity for Grouting

Control
When soil improvement in the form of grouting is
selected in geotechnical applications, some means of
confirming the expected improvement is necessary.
Seismic wave velocity can be used to quantify soil
improvement by measuring wave velocities before
and after grouting. Following are three case histories
describing the successful use of seismic wave veloc-
ity to confirm the extent of ground improvement by
grouting.
4.2.1 Case history 1 - forge foundations

Seismic shear wave velocities determined from
crosshole tests were used to confirm the degree and
extent of soil improvement from combined compac-
tion and chemical grouting, (Woods and Partos,
1981). New forging machines were to be installed in
a forge shop located on deep beach deposits near the
Atlantic coast. Figure 19 shows blow count versus
depth for two locations at this site, B1 and B3. The
blow count ranged from 2 to about 20 in the upper 5
meters. The new forge machines were considerably Fig. 19. Composite soil profile (from Woods and Partos, 1981)
larger than the old forges and vibrations within the
plant from the new, larger installations were of con-
cern as were differential settlements of the forges.
Based on preliminary calculations, it was clear that
the loose sand deposits were susceptible to shake-
down settlement if large-amplitude vibrations oc-
curred. There was also a potential for transmission
of large-amplitude vibrations through the plant, so
soil improvement in the form of grouting was se-
lected to mitigate these concerns. Both compaction
and chemical grouting techniques were chosen to be
performed in series, with compaction grouting being
performed as a first stage and then chemical grout-
ing in a second state if sufficient stiffness had not Fig. 20. Site plan of new forges (from Woods and Partos, 1981)
been achieved in the first stage.

4.2.3 Case history 3 - old bridge support
New design loads on a railway line in Italy required
structural and ground improvements along the entire
line and particularly on two XIX century masonry
arch bridges (Volante et al., 2004). A need for care-
ful ground movement control during upgrading of
the old bridges led to design of a multistage-
multiport, low-pressure grouting technique. Careful
grout control was exercised during the injection
process, but in the long run it was important to de-
termine the strength and deformation parameters of
the newly grouted soil. Figure 22 shows P-wave and
S-wave velocities before and after grouting at one of
the bridge sites. In this case, the modulus of the
ground under the piers was increased by a factor of
about 2.5.
Wave Velocities (m/s)
0 400 800 1200 1600 2000
0
Figure 21. Shear wave velocity profiles after compaction grout- 2 V before
s
ing and then after chemical grouting (from Woods and Partos,
1981) 4 Vp after
Vs after
6
4.2.2 Case history 2 - subway construction

Depth (m)
8
One route of the subway in Pittsburgh, Pennsylvania
was constructed under a part of Sixth Avenue, a nar- 10 Vp before
row street with old, heavy masonry structures on
12
both sides (ENR, 1982). The invert of the subway
was well below the lower elevation of the spread 14
footings supporting the heavy buildings. The foun-
dation material consisted of coarse sand, gravel and 16
cobbles. Construction of the subway called for exca-
18
vation in a cut-and-cover process, but stability of the
adjacent building foundations was in question. 20
Chemical grouting was chosen to improve (stabilize)
the soil, and crosshole seismic tests were used to Figure 22. Seismic velocities before and after grouting (from
Volanta et al., 2004)
confirm achievement of sufficient improvement over
the un-grouted condition. 4.2.4 Summary
Crosshole equipment was fabricated that made The three case histories presented here clearly show
use of the grout pipes for placement of the source that seismic wave velocities can be used to advan-
and receivers. Shear wave velocities were used to tage in confirming quality and extent of grouting op-
characterize the soil before and after chemical grout- erations. While the examples cited all used the
ing. A target shear wave velocity was determined in crosshole seismic method, applications of surface
the laboratory using resonant column tests, and a test wave methods may provide economies where there
section of crosshole tests was performed at the site is sufficient lateral extent to apply them. For a
to confirm expectations from the laboratory study. broad-area dynamic compaction project, the writers
Before-grouting shear wave velocities ranged from have also successfully used the SASW method to
about 500 ft/sec to 1000 ft/sec (150 m/s to 305 m/s) confirm ground improvement by showing increased
and after-grouting velocities ranged from about 1400 shear wave velocities. Stokoe and Santamarina,
ft/sec to 3000 ft/sec (425 m/s to 915 m/s). The crite- 2000 have also shown evaluation of blast densifica-
ria for satisfactory soil improvement by grouting tion by the SASW method.
was either: (1) a doubling of the shear modulus (1.41
times increase in shear wave velocity) over the be- 4.3 Underground Cavity Detection
fore-grouting condition, or (2) a minimum of 1400
ft/sec (425 m/s). The Sixth-Avenue section of the Many engineering situations require the determina-
subway was successfully completed without distur- tion of the existence or absence of underground ob-
bance of the adjacent buildings. stacles, solid or void, as well as their locations and
depths. Probing for these obstacles with penetrome-
ters is a time consuming and expensive process.
Several currently available geophysical techniques
have been proposed and used to identify under-
ground anomalies. The following example and two
case histories describe the use of some of these tech-
niques.
4.3.1 Example 1 – use of GPR, SASW and cross-

hole testing for cavity detection
Three geophysical methods were studied for the de-
tection of buried cavities (Al-Shayea, 1994, and Al-
Shayea et al., 1994). A soil bin, 7 m in diameter and
2 m in depth, was used in this study. The soil bin is
shown in Figures 23 and 24. A three-cell cavity was
buried in the bin at a depth to center of the cavity of Figure 24. Cross section of sand bin with SASW setup for cav-
ity detection (from Al-Shayea et al, 1994)
614 mm. SASW data were collected along five lines
identified by the source and receiver symbols and
the skew lines marked a-c on Figure 23. The general
SASW test arrangement is shown in Figure 24.
Ground-penetration radar (GPR) data were also col-
lected along the grid lines identified on Figure 23.
Crosshole shear-wave tests were performed across
both long and short axes of the buried cavity and in
the free-field. GPR gave the most obvious identifica-
tion of the cavity as indicted on Figure 25. In this
figure, it is very clear where the electromagnetic
wave field produced by GPR was distorted, (Figure
25b) compared with the wave field of the free-field
(Figure 25a). Approximate depth and size of the cav-
ity were calculated from Figure 25b knowing the
frequency of the GPR source and the dielectric con-
stant of the sand.
SASW tests were performed directly over the
centerline of the cavity with various states of filled
and empty cells. (The cells were filled using the
same sand as in the bin.) Differences in cavity filling
Figure 25. GPR Scans: (a) free-field and (b) parallel to and
over the long axis of the cavity with all cells empty (from Al-
Shayea et al, 1994)
showed substantially different dispersion curves,

Figure 26. A smoothed free-field dispersion curve is
shown as a solid line while the empty and partially
empty void dispersion curves are shown with other
symbols. For this discussion, the symbols represent-
ing cavity conditions need not be identified because
the key result is that the existence of a void and the
size of the void both influenced the shape of the dis-
Figure 23. Plan view of sand bin with buried three-cell cavity persion curve.
(from Al-Shayea et al, 1994)

Figure 27. Crosshole velocity profiles along two paths, free-
Figure 26. Dispersion curves for free-field and for lines over field and along centerline of the cavity (from Al-Shayea, 1994)
the long axis of the cavity with various cells filled (from Al-
Shayea et al., 1994)
It is worth noting that some of the drawbacks of
Other SASW tests were performed directly over the SASW method that were cited above for cavity
and on lines skewed to the centerline of the cavity. detection may be mitigated by using recently devel-
These varying lines also showed substantially differ- oped wavelet theory (Shokouhi and Gucunski, 2003,
ent dispersion curves. and Gucunksi and Shokuohi, 2004). Continuous
Results of SV-wave seismic crosshole tests per- wavelet transforms (CWT) are a new class of trans-
formed on lines S-R1 (free-field) and S-R2 (over formations that can produce a time-frequency map
cavity) in Figure 23 are presented in Figure 27. Here of the ground surface from which indications of
it can be seen that the shear wave profile for the near-surface cavities can be derived. The data col-
free-field direction is quite consistent for three con- lection required for CWT can be achieved simulta-
ditions of the void, all cells empty, 1 cell filled with neously with SASW data collection.
sand, and all cells filled with sand. The cases of all
cells filled did not exactly match the free-field con- 4.3.2 Case history 1- mine collapse under a
dition because the cell filling process could not du- highway
plicate the free field density of the sand. In the case High resolution SH-wave reflection tests were pre-
of line S-R2, the average shear wave velocity in the formed along the right-of-way of Interstate Highway
depth region of the cavity was clearly reduced by the I-70 in southeastern Ohio, (Guy et al., 2003). A por-
existence of the cavity. Had the crosshole boreholes tion of the east bound lane of I-70 at this location
been closer together, shear wave velocity differences collapsed into old underground coal-mine workings.
would have been more dramatic at depths represen- A plan view of the east bound lanes of I-70, the lo-
tative of the cavity. cation of the collapsed highway, and a projection of
All three of the geophysical techniques used in the old underground mine workings are shown in
this study have characteristics that allow identifica- Figure 28. High-resolution SH-reflection surveys
tion of cavities or anomalies in the underground, but were performed along two lines straddling the east
each have their limitations. To mention just the most bound lanes of I-70, lines GUE-I70-1 and EBPas-
salient drawbacks, GPR suffers from ability to pene- sYY. The interface between soil and rock was clear
trate clay, SASW is limited by the need to run mul- for most of the lengths of both of these survey lines,
tiple lines of data, and crosshole boreholes need to but in the interval between stations 48320 and 48360
strategically located to straddle the cavity. Some of on line EBPassYY the SH-wave stacking velocity
these limitations may be minimized through future plot showed a discontinuous segment of the soil/rock
study while others are inherent to the basic princi- interface in the region indicted by the angled bars in
ples of the geophysical methods. Some of these Figure 29 and on the geologic cross section shown in
drawbacks simply point to the necessity of perform- Figure 30. The continuous soil/rock interface as in-
ing suites of complementary geophysical tests for terpreted is indicated on Figure 29 by the dash-dot
cavity detection. line across the plot. Soil borings and rock coring
Figure 28. Plan view of site showing eastbound lanes of I-70,
seismic reflection lines and projection of Murray Hill No. 2
Mine Workings (from Guy et al., 2003)
were performed at six locations selected by inter-

preting the soil/rock interface reflector in Figure 29.
These voids indicate stopping from the former coal
mine upwards, but that stopping had not progressed
to the pavement level.
This seismic method penetrated relatively deep,
about 20 meters in this case. The results of the tests
show a potential for future subsidence and sinkhole
development along this highway. Based on this kind
of information, remedial efforts could be applied to
stop progression of the stopping or stabilize the
ground above the current voids.
Because high resolution SH reflection utilizes
steady-state ground excitation, wavelengths can be Figure 30. Geologic cross section for line EBPassYY for sta-
tions 48300 thru 48360 including exploratory borings and iden-
controlled for good wavelength/cavity size ratios tification of voids (from Guy et al., 2003)
and may permit better cavity size and depth deter-
mination than other seismic wave based cavity de-
GPR and SASW testing at this site could not probe
tection methods.
deep enough to explore the soil/rock interface, GPR
4.3.3 Case history 2 – other geophysical testing at because of clay in the soil and SASW because of the
the mine collapse under highway I-70 lack of a sufficiently energetic excitation source.
All of the geophysical techniques cited in Section The crosshole test did not happen to encounter a
4.3.1 (except CWT) were applied by Hiltunen et al., void or loose soil. A major conclusion from these
(2004) at the mine-collapse site on I-70 described tests was that no single technique could unambigu-
above. The results confirmed the drawbacks cited ously detect voids or other anomalies throughout a
previously. However, the work by Hiltunen et al. wide range of depths.
confirmed that quality geophysicalmeasurements
could be made in close proximity to an active inter-
state highway with heavy truck traffic. Both
Figure 29. Interpreted stacked time record for line EBPassYY for Stations 48300 thru 48480 (from Guy et al., 2003)

4.3.4 Summary nel. SASW testing was performed with hand-held
While geophysical methods, including several seis- hammers as sources and accelerometers as receivers.
mic wave propagation methods, have considerable The accelerometers were held magnetically to metal
potential for locating and sizing buried objects, no disks attached to the liner. This general configura-
single method is appropriate for all sites. In the past, tion is shown in Figure 31b. Testing was conducted
engineers and geophysicists have often chosen one to profile along two planes into the liner-rock sys-
method in an attempt to locate cavities or buried ob- tem. One profile was along the springline, and the
jects and have been disappointed with the results. other profile was near the crown as illustrated in
Indications from the cases cited herein are that, Figure 31b.
while any one method may provide some parts of the The SASW testing program was designed to in-
identification puzzle, a suite of tests can provide vestigate the following: 1. thickness and quality of
more confidence in finding cavities or buried objects the concrete liner in the springline and crown areas,
2. thickness and quality of any grout in the area of
4.4 Tunnel Investigation the crown, 3. identification of any voids in the crown
Sometimes, Vs measurements are used to profile area, and 4. stiffness and variability of the rock be-
constructed systems and their geotechnical founda- hind the liner. (Grouting in the crown area was done
tion materials to assist in forensic studies. A forensic some time after construction of the liner.) The pro-
study of a concrete-lined tunnel in rock is described gram successfully answered these questions. Exam-
below (Stokoe and Santamarina, 2000). A general- ples showing how some of the questions were an-
ized cross section of the tunnel is shown in Figure swered follow.
31a. The tunnel is approximately 3 m in diameter, An interpreted Vs profile at one springline loca-
with a concrete liner that has a nominal thickness of tion is shown in Figure 32a. The profile shows a
30 cm. high-quality concrete liner (Vs > 2500 m/s) that is
An extensive investigation was conducted in about 35 cm thick. At this location, the liner is in di-
which SASW testing was performed at more than rect contact with the rock, and the rock is stiffer (and
100 locations along the longitudinal axis of the tun- presumably stronger) than the concrete.
Results from one crown location are shown in
Figure 32b. In this case, the liner is thicker than 40
Grout
cm, and there is grout between the liner and the
rock. Based on the Vs values, both the concrete and
grout are high quality. The concrete-grout-rock in-
tgrout terfaces have intimate contact; hence, no voids.
tconcrete Also, the rock is less stiff than the concrete at this
Concrete location.
Liner
Clearly, SASW testing was successfully and cost-
effectively applied in the tunnel investigation. The
Rock writers have had other successful projects in many
other underground applications (for instance,
Madianos et al., 1990, Olson et al., 1993, and Luke
a. Generalized tunnel cross section et al., 1998).
“Crown”
4.5 Offshore Shear-Wave Velocity Profiling Using
Investigation
Plane
Rock Seismic Interface Waves
Grout
As offshore construction moves into deeper water
Liner (depths greater than 1.6 km), traditional drill-and-
Receivers
sample geotechnical site investigations become ex-
Springline SASW
Investigation Array Axes
pensive and less reliable. The expense of drilling in
Plane deep water often dictates the extraction and testing
Hammer of only a few samples. Furthermore, the quality of
Source
these samples can be severely compromised when
extracted through great water depths. Other geotech-
nical site investigation methods, such as the seismic
cone penetrometer (e.g. Robertson et al., 1986), are
b. SASW Testing Arrangement and Planes of Investiga- effective on land and in shallow water, but become
tion (from Stokoe and Santamarina, 2000)
more difficult and costly to apply in the deep-water
Figure 31. SASW testing performed inside a concrete-lined environment. One seismic method that has potential
tunnel (from Stokoe and Santamarina, 2000) for deep-water seafloor investigation is the surface-
0 Shear Wave Velocity, m/s
0 100 200 300 400
concrete liner 0
(thickness ~ 35 cm)
2
2 VS from SASW
rock behind liner
Depth, m
stiffer than concrete

4
Depth, m
Station 1 6
(Springline)
8
0 1000 2000 3000 4000
8
Shear Wave Velocity, VS , m/sec
a. Interpreted Vs profile at a springline station
0
concrete liner
(thickness
{ 10 VS from SCPT
>40 cm) {
2 grout 12
(good stiffness
and thickness > 30 cm) Figure 33. Results from the SASW testing performed at the
Depth, m
offshore near Vancouver, B.C. (Rosenblad and Stokoe, 2001)

4 Rock behind liner
softer than concrete depths, ranging from 12.2 to 76.2 m, to demonstrate
the potential of the method. Personnel from
6 ConeTec also conducted SCPT measurements at the
Station 2
(Crown) site. The SASW results agree reasonably well with
the SCPT values at depths between 5.5 and 10 m as
8 seen in Figure 33. The differences that are observed
0 1000 2000 3000
are likely due to the localized versus global nature of
Shear Wave Velocity, m/sec the SCPT and SASW tests, respectively. Based on
b. Interpreted Vs profile at a “crown” station the writers’ experience on land, development of a
lower-frequency source and longer arrays would al-
Figure 32. Examples of Vs profiles measured inside a concrete- low SASW profiling to significantly greater depths
lined tunnel (from Stoke and Santamarina, 2000)
at this site, certainly to depths on the order of 30 to
50 m. However, the results do demonstrate the fea-
wave method for Vs measurements of the sediment.
sibility of the surface-wave method offshore. Stoll et
Shear-wave velocity is used because it is essentially
al. (1994), Luke and Stokoe (1998) Rosenblad
unaffected by the presence of water (and air if the
(2000), and Rosenblad et al. (2003) are among oth-
sediment is unsaturated) and because, in a saturated
ers who have also shown this surface-wave applica-
soil, Vs is far better correlated with shear strength
tion.
than Vp wave, which has been used to predict shear
strength in previous offshore investigations (Blake 4.6 Geotechnical Earthquake Engineering
and Gilbert, 1997).
The SASW method is being applied to the off- The importance of the shear stiffness of geotechnical
shore environment to determine shear-wave velocity materials in calculating their response during dy-
profiles of the seafloor. In this application, a soil- namic loading initially stimulated the development
water interface wave, called a Sholte wave (Wright of in situ seismic methods tailored to measure Vs.
et al., 1994), is measured. The measured Sholte- This development began in earnest in the 1960s with
wave dispersion curve is used to determine a shear- modifications/refinements to the crosshole and
wave velocity profile. Figure 33 shows the results downhole methods. It is continuing today with im-
from SASW testing performed by ConeTec, Inc. off provements to surface-wave methods. The following
the coast of Vancouver, B.C. (Rosenblad and Sto- two case histories and one application describe some
koe, 2001). The soils in this region were composed recent work in geotechnical earthquake engineering.
of loose silty-sands and sands that were unsaturated
(gaseous). Testing was conducted in shallow water

4.6.1 Case history 1 - profiling hard-to-sample al- acteristic Vs profile of the alluvium to be preserved.
luvium The four depth intervals were: 1.5 to 4.6 m (layer no.
Yucca Mountain, Nevada, was approved as the site 1), 4.6 to 9.2 m (layer no. 2), 9.2 to 18.3 m (layer no.
for development of the geologic repository for high- 3), and 18.3 to 30.5 m (layer no. 4). An example of
level radioactive waste and spent nuclear fuel in the the results from this procedure at one borehole is
United States. The U.S. Department of Energy has presented in Figure 34. Figure 34a shows the pro-
been conducting studies to characterize the site and files determined from downhole, suspension logging
assess its future performance as a geologic reposi- and SASW measurements in and near borehole RF-
tory. As part of these studies, a comprehensive pro- 19. Figure 34b shows the averaged Vs profiles.
gram of in situ seismic investigations was performed
at the proposed site of the Waste Handling Building
Shear Wave Velocity, m/s
(WHB). The purpose of these investigations was to
characterize the velocity structure of the subsurface 0 400 800 1200
for seismic design of the WHB facilities. 0
In situ seismic velocity measurements were per- SASW
formed by three different methods at this site. The
5
seismic methods include two borehole methods,
downhole and suspension logging, and one surface-
wave method, spectral-analysis-of-surface waves 10 Suspension
Depth, m
(SASW). The borehole surveys were conducted in Downhole Logging
16 cased boreholes to a maximum depth of 198 m.
SASW surveys were performed at 34 locations 15
around much of the proposed area which was about
300 m by 450 m in plan dimensions. The SASW 20
surveys were aimed at evaluating the top 50 m of the Material Profile:
site and investigating lateral variability. The SASW Alluvium (Qal)
surveys provided greater spatial coverage of the site 25 from 1.5 to 37 m
while the borehole surveys added critical deeper in-
formation. 30
Stokoe et al., 2003 presented a comparison of the
Vs profiles determined by the three seismic methods a. Vs profile measured with each seismic method
in the material where the most overlap in measure- Shear Wave Velocity, m/s
ments existed. This material is a hard-to-sample
Quaternary alluvium/colluvium (Qal) which ranges 0 400 800 1200
0
from a poorly graded gravel (GP) to a silty gravel
Downhole
(GW). The alluvium contains varying amounts of Layer #1
sand, cobbles and boulders, and it varies in thick- 5 SASW
ness, depth, and amount of cementation over the Layer #2
WHB site. The alluvium was measured in 15 of the
16 boreholes. The results form the most comprehen- 10
Suspension
Depth, m
sive set of Vs measurements, in terms of multiple Logging

seismic methods in a localized area and in one mate- Layer #3
15
rial type, that has ever been compared. In addition,
this comparison represents a “blind comparison”
overseen by URS personnel, in that each measure- 20
ment team was not aware of the other’s results until
after they were all submitted to URS. Therefore, not Layer #4
only did the seismic tests at the WHB site provide 25
the information needed for the seismic design of the
facility, but they also provided an interesting com- 30
parison of Vs profiles measured by different methods
b. Averaged Vs profile computed for each seismic method from
as described below. Figure 34a above
The comparison of the Vs profiles is presented as
an average profile for each method. The average Figure 34 Example of measured and averaged Vs profiles used
profiles were determined by first dividing the 1.5-to- in comparing the seismic methods; Measurements in or near
30-m depth range into four intervals, with the small- Borehole RF-19 (from Stokoe et al., 2003)
est near the surface where the Vs gradient was the
greatest and largest at depth where the gradient was
the smallest. This division allowed the overall char-
Shear Wave Velocity, m/s most resolution of these methods near the surface,
(2) the averaging effect of placing straight-line seg-
0 400 800 1200 ments through the measured travel times in down-
0 hole data reduction, and (3) wave refraction and lat-
eral variability in the Qal affecting each method
differently. The standard deviations determined from
Downhole the measurements and the COV values support
5
points (1) and (2) above. The COV values are about
SASW 0.21 for SASW measurements in layers no. 1 and
no. 2 and about 0.11 for the downhole measurements
10 in the same layers. It should be noted that the vari-
Depth, m
ability shown by ±σ includes both measurement un-

certainty and variability in material properties.
15 Average Suspension
Value Logging 4.6.2 Case history 2 - deep Vs profiling
Another part of the seismic investigations at the
20 Yucca Mountain site discussed above involved deep
Vs profiling along the top of Yucca Mountain (Sto-
koe et al., 2004). Deep profiling is defined as evalu-
25 ± one ating the shear-wave velocity structure to depths of
about 200 m. This work involved the SASW method
standard and required the use of a Vibroseis (see Figure 36)
deviation to generate the low-frequency (hence long wave-
30 length) waves necessary to profile to 200 m. Yucca
Figure 35. Comparison of average Vs profiles from 15 loca- Mountain consists of stacked layers of tuffs with Vs
tions in the WHB area (Stokoe et al., 2003)
generally above 900 m/s. Therefore, the lowest
Comparison of the average Vs profiles is pre- excitation frequency was in the range of 3 Hz and
sented in Figure 35. Also shown in the figure is the the farthest measurement point from the source was
variability in the Vs values measured with each around 500 m.
method, expressed by ± one standard deviation (±σ). SASW measurements were performed at 22 array
These data are presented numerically in Table 5, in- sites along the top of Yucca Mountain. These sites
cluding the coefficient of variation (COV = σ/Avg were spread over a distance of about 5 km. The
Vs). The comparison is shown in Figure 35 and SASW surveys were aimed at evaluating: (1) the top
demonstrates the strength and robustness of S-wave 150 to 200 m of the mountain, (2) an apparent Vs
velocity measurements today. First, identical trends gradient in the near surface (within about 5 to 15 m),
of increasing VS with depth were measured with all and (3) any lateral variability over the 5-km dis-
three methods. Second, differences in average Vs tance. The mean Vs profile that was determined from
values in each layer are small. The two largest dif- the 22 profiles is presented in Figure 37 along with
ferences are 16% and 12% and are found between the 16th and 84th percentile Vs values. The coefficient
the downhole and SASW measurements in layers of variation (COV) about the mean profile was cal-
no. 1 and no. 2, respectively. These layers are the culated by assuming the Vs values follow a log-
shallowest layers (less than 9.2 m deep) and are the normal distribution.
ones which should be expected to show the largest
differences due to: (1) the SASW method having the
Table 5 Numerical Analyses of Average VS Profiles (from Stokoe et al., 2003)

Depth In- Downhole Surveys SASW Surveys Suspension Logging Surveys
terval (m) St. σ St. St.
Avg. Avg. σ Avg. σ
No. of Dev., Avg. V No. of Dev., No. of Dev.,
VS V VS
Meas. σ S
Meas. S
σ Avg. VS Meas. σ Avg. VS
(m/s) (m/s) (m/s)
(m/s) (m/s) (m/s)
1.5 – 4.6 10 432 53 0.12 10 502 101 0.20 01 ⎯ ⎯ ⎯
4.6-9.2 12 526 56 0.11 12 590 126 0.21 21 549 ⎯ 2 ⎯
9.2-18.3 13 662 59 0.09 123 736 54 0.07 9 669 75 0.11
18.3-30.5 8 732 43 0.06 73 798 88 0.11 8 768 118 0.15
1 Suspension logging was not successfully performed at shallow depths due to high attenuation, backscattering and tube-wave in-
terference.
2 Insufficient data to perform meaningful calculations.
3 SASW surveys were performed at one less borehole than the downhole surveys due to surface obstructions.

Figure 36. Photograph of Vibroseis truck in operation on the top of Yucca Mountain (from Stokoe et al., 2001)
Several interesting trends are evident in Figure value of the mean Vs below the 5-m-thick, near-
37. First, it is observed that the near-surface Vs gra- surface zone. The measurements show a mean value
dient is quite abrupt. The values of Vs change from of approximately 1000 m/s in the depth range of 10
approximately 300 m/s in the top meter to over 800 to 150 m.
m/s at a depth of only 5 m. Below 5 m, the mean Vs
value increases gradually from about 900 to 1000 4.6.3 Application 1 - liquefaction resistance
m/s at a depth of 150 m. A gradient near the surface Evaluation of the liquefaction resistance of soils can
is expected due to the effects of weathering on the be a critical factor in many geotechnical engineering
near-surface rock. The abruptness of the gradient, investigations. Such an evaluation is typically per-
which has important implications in terms of the formed with field test such as the standard penetra-
ground motion hazard, can not be predicted without tion tests (SPT) or cone penetration test (CPT). The
field measurements of this kind. Another important general procedure, called the “simplified procedure,”
result shown in Figure 37 is the nearly constant
Shear-Wave Velocity, m/s COV No. of Profiles

0 500 1000 1500 2000 0.0 0.1 0.2 0.3 0.4 0 5 10 15 20 25
0
50
100
Depth, m
150
200
Mean
16th and 84th Percentile
250
Figure 37 Statistical analysis of VS profiles from SASW measurements on top of Yucca Mountain (Stokoe et al., 2004)
0.6 5 IMPACT OF DISTURBANCE FROM
Cyclic Stress or Resistance Ratio, CSR or CRR
Data Based on:

Mw = 5.9 to 8.3; adjusted by Mw = 7.5 SAMPLING ON PREDICTED NONLINEAR
dividing CSR by (M w/7.5)-2.56 >35 20 <5 Fines SOIL RESPONSE IN THE FIELD
Uncemented, Content (%)
Holocene-age soils
Average values of
VS1 and amax
One of the strengths of seismic measurements, as
0.4 well as other geophysical measurements, is that the
same basic measurement can be performed in the
field and in the laboratory. Field measurements of Vs
Liquefaction
are performed in the small-strain or elastic range as
No
Liquefaction discussed earlier. Therefore, laboratory measure-
ments of Vs also have to be evaluated in this small-
0.2 strain range if they are to be compared directly with
Fines Content
the field values. This comparison presently forms
< 5% the way sample disturbance is evaluated in geotech-
6 to 34%
> 35 % Field Performance nical earthquake engineering when dealing with
Liquefaction nonlinear deformational characteristics. Field and
No liquefaction
0.0 laboratory values of Vs at small strains are used to
0 100 200 300
adjust the nonlinear response of soil measured in the
Overburden Stress-Corrected Shear Wave laboratory to field conditions. The nonlinear re-
Velocity, VS1, m/s ȱ sponse is typically shown in terms of the nonlinear
ȱ variation in shear modulus with shearing strain (G –
Figure 38. Curves recommended by Andrus and Stokoe (2000)
for delineating liquefiable and nonliquefiable granular soils
log γ ). This comparison, the adjustment procedure,
based on field Vs measurements and the impact on the G – log γ and stress-stain (τ –
γ) curves are discussed below.
cedure,” was initiated by Seed and Idriss (1971) us- 5.1 Comparison of Small-Strain Field and
ing SPT blow counts correlated with a parameter Laboratory Values of Vs
called the cyclic stress ratio that represents the Invariably, when field and laboratory values of Vs
earthquake loading. This procedure has been up- are compared, values of Vs, lab range from slightly
dated over the years. CPT measurements have been less to considerably less than the in situ values, Vs,
added, initially by Robertson and Campanella field (Anderson and Woods, 1975, Long, 1980,
(1985). A national workshop was convened in 1996 Yasuda and Yamagushi, 1985, Yokoa and Konno,
which further updated the SPT and CPT procedures 1985, and Chiara, 2001). A just-completed project
and added an in situ geophysical method to the suite dealing with the resolution of site response issues in
of field tests (Youd et al., 2001). The geophysical the 1994 Northridge, CA earthquake, called the
method is in situ seismic measurements of Vs. Val- ROSRINE project, involved numerous field and
ues of Vs are correlated with earthquake loading in laboratory investigations. Sixty-three intact samples
the same manner as done in the SPT and CPT pro- were recovered and tested in the laboratory at the
cedures. University of Texas using combined resonant col-
The procedure involving Vs was presented by umn and torsional shear equipment (Darendeli,
Andrus and Stokoe, (2000). The procedure is based 2001, and Choi, 2003). Additionally, in situ seismic
on field performance data from 26 earthquakes and measurements were performed during the field in-
in situ Vs measurements at over 70 sites. The case vestigation phase, mainly be GeoVision Geophysical
history data from this procedure, adjusted to an Services, Corona, CA using a suspension logger.
earthquake moment magnitude (MW) of 7.5, is Therefore, the ROSRINE project afforded an excel-
shown in Figure 38. Of the 90 liquefaction case his- lent opportunity to investigate further the relation-
tories shown in the figure, only two incorrectly lie in ship between field and laboratory values of Vs.
the no-liquefaction region. These two points are, An example field Vs profile measured in this
however, very near the boundary. Clearly, the pro- study is presented in Figure 39. At this site, called
cedure based on field Vs measurements can be used La Cienega, in situ seismic tests (shallow crosshole
as a supplement or in lieu of the SPT and CPT pro- testing and deep suspension logging) were per-
cedures. The procedure is especially important for formed. A depth of nearly 300 m was logged. Intact
use with hard-to-sample soils such as soils contain- samples were recovered from depths ranging from 4
ing gravel and/or cobbles. The nonintrusive nature to about 240 m. The laboratory values of Vs, shown
of the SASW and other surface-wave methods make by the solid circular symbols, are plotted at the cor-
them especially well suited for this application in responding sample depths. There is considerable
hard-to-sample soils.

0 0
In-Situ Suspension Logger
In-Situ Shear Wave Velocity, Vs,field, m/s

In-Situ Crosshole
Laboratory Resonant Column No. of Specimens = 63
200
Trend Line
100
400
Depth, m
Shear Wave Velocity, ft/sec

400 600 800
Denotes “Average” 600
0 0 Range
Depth, m
Field Value Used in

Depth, ft
200 10
5 Comparison with
20 Laboratory Values
800
30
10
150 200
Shear Wave Velocity, m/sec
1000
300 0.0 0.2 0.4 0.6 0.8 1.0 1.2
0 200 400 600 800 1000 Shear Wave Velocity Ratio, Vs,lab / Vs, field
Shear Wave Velocity, m/sec
Figure 40. Variation in the ratio of laboratory-to-field shear

Figure 39. Example profile of small-strain field and laboratory
wave velocities (Vs, lab/ Vs, field) with the in-situ value of Vs,
shear wave velocities evaluated at a strong-motion earthquake
field determined in the ROSRINE project
site on the ROSRINE project (from Stokoe and Santamarina,
2000)
variability in the field Vs profile. The“average” field With the in situ Vs value and the laboratory G –
values associated with the laboratory values are log γ curve, the final step is to estimate the field G –
shown by the short vertical lines through the field Vs log γ curve. This step is accomplished by scaling the
profile in the vicinity of the sample depth. laboratory G – log γ curve using Gmax determined
A summary of all field-lab Vs comparisons from from the field seismic tests as,
the ROSRINE project is presented in Figure 40. A
total of 63 samples were tested in the laboratory. § G γ , lab ·
There is a clear trend in the data, with the velocity Gγ, field = ¨¨ ¸ Gmax, field
¸
(8)
ratio (Vs lab /Vs field) decreasing as the in situ value of © G max, lab ¹
Vs increases. (There was essentially no correlation where,
with sample depth.) In general terms, the velocity ra- G γ, field = in situ shear modulus at a shearing
tio is around one at Vs ≅ 160 m/s. However, at Vs ≅ strain of γ,
725 m/s, the velocity ratio is about 0.6, which means G γ, lab = shear modulus determine in the -
that the small-strain shear modulus from laboratory laboratory with an intact specimen
testing is on the order of 1/3 of the value in the field. at a shearing strain of γ,
This comparison strongly supports the need to per- Gmax, lab = small-strain shear modulus
form field seismic tests, certainly in studies dealing determined in the laboratory, and
with siting and retrofitting of important facilities. Gmax, field = in situ shear modulus measured
by seismic testing.
5.2 Estimated Field G - log γ Curves from Field It is assumed, of course, that evaluation of the G –
and Laboratory Measurements log γ curve in the laboratory was performed at a con-
finement state, excitation frequency, number of
Once the Vs profile has been determined at impor- loading cycles, drainage condition, etc. that repre-
tant or high-risk sites, the next step in the geotechni- sent the field conditions. Also, Gmax, field was calcu-
cal earthquake engineering investigation is determi- lated from Vs, field using Equation 2.
nation of the nonlinear characteristics of the soil. The estimated field G – log γ curves are shown by
This step typically involves cyclic and/or dynamic the dashed lines in Figures 41a, 41b and 41c for the
laboratory testing of intact specimens. In terms of soft, medium stiff and very stiff soil examples taken
nonlinear shear modulus, these results are presented from the ROSRINE project. Clearly, adjustment of
in the form of the variation in normalized modulus, the laboratory curve is critical to correctly predicting
G/Gmax, with shearing strain amplitude,γ, or simply the earthquake ground motions at the site.
G – log γ. Typical examples of G – log γ curves
from the ROSRINE project for soft, moderately stiff 5.3 Laboratory and Field Stress-Strain (τ – γ)
and very stiff soils are shown in Figures 41a, 41b, Curves
and 41c, respectively. The laboratory G - log γ
curves are shown by the solid lines in the figures. The laboratory shear stress-shear strain (τ – γ) curve
can be calculated from the laboratory G – log γ
curve as,
τ=G*γ (9)
100 0.08
Range in estimated Estimated field curve
field curve if no Vs, field using Vs, field
80
Estimated 0.06
Shear Stress, τ, MPa

Shear Modulus, G, MPa
field curve
60 using Vs, field
0.04
Range in estimated
40
field curve if no Vs, field
Lab Curve (Clay, CL)
σ ' = 0.61 atm 0.02 Lab Curve (Clay, CL)
20 o
σ ' = 0.61 atm
o
0 0.00
10-4 10-3 10-2 10-1 100 0.0 0.2 0.4 0.6 0.8 1.0
Shearing Strain, γ, % Shearing Strain, γ, %
a. Vs,in-situ = 180 m/s (soft soil) and Vs,lab/Vs,field = 0.92 a. Vs,in-situ = 180 m/s (soft soil) and Vs,lab/Vs,field = 0.92
500 0.6
Range in estimated
field curve if no Vs, field Range in estimated
400 0.5 field curve if no Vs, field
Estimated field
curve using Vs, field 0.4
300 Estimated field curve
0.3 using Vs, field
200
0.2
100
Lab Curve (Clay, CL) 0.1 Lab Curve (Clay, CL)
σ ' = 2.86 atm σ ' = 2.86 atm
o o
0 0.0
10-4 10-3 10-2 10-1 100 0.0 0.2 0.4 0.6 0.8 1.0
Shearing Strain, γ, % Shearing Strain, γ, %
b. Vs,in-situ = 360 m/s (moderately stiff soil) and Vs,lab/Vs,field b. Vs,in-situ = 360 m/s (moderately stiff soil) and Vs,lab/Vs,field
= 0.81 = 0.81
1200 1.5
Range in estimated Range in estimated
field curve if no Vs, field field curve if no Vs, field
1000
Estimated field Estimated field curve

800 1.0
curve using Vs, field using Vs, field
600
400 0.5
200 Lab Curve (Sand, SW)

Lab Curve (Sand, SW)
σ o' = 8.17 atm
σ´o = 8.17 atm
0 0.0
0.0 0.2 0.4 0.6 0.8 1.0
10-4 10-3 10-2 10-1 100
Shearing Strain, γ, %
Shearing Strain, γ, % c. Vs,in-situ = 540 m/s (very stiff soil) and Vs,lab/Vs,field = 0.60
c. Vs,in-situ = 540 m/s (very stiff soil) and Vs,lab/Vs,field = 0.60
Figure 42. Calculated laboratory τ – γ curves, estimated field τ
Figure 41. Measured laboratory G – log γ curves, estimated – γ curves using Vs field, and possible range in field τ – γ curves
field G – log γ curves using Vs, field, and possible range in field using range in Figure 40 but no measurement of Vs field; Ex-
G – log γ curves using range in Figure 40 but no measurement amples for soils from the ROSRINE project with a range in
of Vs field; Examples for soils from the ROSRINE project with a shear stiffness
range in shear stiffness

with companion sets of G – γ values taken from the 6 CONCLUSIONS
dynamic laboratory curve. The laboratory τ – γ
curves that were derived from the laboratory G – log Geophysical methods have an important and ever-
γ curves for the soft, medium stiff and very stiff increasing role to play in the solution of geotechni-
soils in Figures 41a, 41b and 41c are shown by the cal engineering problems. Seismic methods have
solid lines in Figures 42a, 42b and 42c, respectively. been embraced by geotechnical engineers over the
The estimated field τ – γ curves for these soils were past 50 years. They have been heavily used in the
determined following the procedure expressed by solution of soil dynamics and geotechnical engineer-
Equation 9, except that the estimated field G – log γ ing problems, especially in the evaluation of small-
curves were used. The estimated field curves are strain shear and compression stiffnesses. Today, Vs,
shown by the dashed lines in Figure 42. measurements in the field and laboratory form a
This example is presented to show the importance critical link in evaluating sample disturbance and in
of Vs, field in predicting the field τ – γ curves which predicting nonlinear G – log γ and τ – γ curves.
are used for deformational analyses like the ones The adoption of geophysical methods in the solu-
presented in Section 4.1. In such deformational tion of non-dynamic problems has occurred more
analyses, shear strains rarely exceed 1 % which is slowly in geotechnics, excluding geoenvironmental
the reason why the τ – γ curves are shown with a and military applications. Seismic testing is still the
scale of 0 to 1 %. Unfortunately, many geotechnical most widely used method, particularly for evaluating
engineers are not aware of this adjustment procedure site characteristics (layering, top of bedrock, voids,
for sample disturbance or the importance of Vs. etc.) and monitoring processes (grouting, damaged
or changed zones from construction activities, etc.).
5.4 What If No In Situ Vs Values Are Measured? The use of strain-adjusted moduli in settlement and
At times, the owner or client may elect to test only other deformational analyses offers a rational ap-
intact samples and not perform in situ Vs measure- proach to the solution of many of these problems.
ments. This decision may be based on cutting cost, This approach will continue to grow.
incomplete understanding of the importance of Vs, The seismic method that continues to evolve in
geotechnical engineering is the surface-wave
field or other reasons. In any case, the field G – log γ
and τ – γ curves can not be estimated using the ad- method. The nonintrusive nature of the method
justment procedure discussed above. Therefore, a makes its application very cost effective, and its use-
wide range in the estimated nonlinear field curve fulness will continue to increase. It would be very
will result. Figure 40 can be used in reverse to find beneficial to this method, as well as other geophysi-
the range in the expected field curves if one only had cal methods, to incorporate increased automation.
laboratory G – log γ curves. These ranges are shown Adoption by the profession would also benefit from
by the shaded zones in Figure 41 for the three differ- increased coverage of geophysical methods in the
ent soil stiffnesses. The ranges are quite large, ex- civil engineering curriculum. Finally, the engineer
ceeding factors of two and three for the moderately also needs to consider that the robustness of the so-
stiff and very stiff soils, respectively. The same rela- lution is significantly enhanced in many applications
tive comparison is shown by the shaded zones in by the use of a suite of geophysical measurements.
Figure 42 for the ranges in expected τ – γ curves.
ACKNOWLEDGMENTS
5.5 What If No Laboratory G – log γ Curves Are
Measured? The writers sincerely appreciate the opportunity
Obviously, the other situation that the geotechnical given by the organizers of this conference to present
engineer might face is having only the in situ Vs these results. The patience and understanding of
measurements. In the writers’ opinion, this situation Prof. António Viana da Fonseca is especially appre-
may not be as troublesome as the case above with no ciated.
Vs, field values. Hopefully, the engineer has a boring Support from the California Department of Trans-
log and the soil types identified. In this case, empiri- portation, the National Science Foundation, the
cal soil models can be substituted for the laboratory United States Geological Survey, the ROSRINE pro-
G – log γ curves (assuming no unusual or difficult ject, and the U.S. Department of Energy through a
soils). The empirical models should include vari- subcontract to Bechtel SAIC is gratefully acknowl-
ables such as soil type, confinement state, some edged. Interaction, encouragement and guidance
measure of uncertainty, etc. such as the model by from many colleagues is appreciated as is the work
Darendeli (2001). However, this approach would of many excellent graduate students. Several of the
only be used if no laboratory G – log γ curves were case histories involved work with geotechnical con-
measured and will result in wide ranges for the esti- sulting firms and their clients. Permission to publish
mated field G – log γ and τ – γ curves. the results is appreciated. Finally, a special thanks is
given to Ms. Alicia Zapata for assisting in prepara- Choi, W.J. 2003. Linear and nonlinear dynamic properties from
tion of this paper. combined resonant column and torsional shear tests of
ROSRINE phase-II specimens. Masters Thesis, University
of Texas.
Darendeli, M.B. 2001. Development of a new family of nor-
REFERENCES malized modulus reduction and material damping curves.
Ph.D. dissertation, University of Texas.
Achenbach, J. D. 1975. Wave Propagation in Elastic Solids. Di Benedetto, H., Doanh, T., Geoffroy, H., and Sauzeat, C.
North Holland, 425p. (Eds.) 2003. Deformation Characteristics of Geomaterials.
Addo, K.O., and Robertson, P.K. 1992. Shear-wave velocity A.A. Balkema Publishers, Tokyo, 1425 p.
measurements of soils using Rayleigh waves. Canadian Dobrin, M.B., and Savit, C.H. 1988. Introduction to Geophysi-
Geotechnical Journal, 29, pp. 558-568. cal Prospecting. Fourth Edition, McGraw-Hill Book Com-
Aki, K., and Richards, P.G. 1980. Quantitative Seismology: pany.
Theory and Methods, W.H. Freeman and Co., Vol. I. ENR. 1982. Shoehorning Pittsburgh’s subway. Cover Story,
Al-Hunaidi M.O. 1994. Analysis of dispersed multi-mode sig- Engineering News Record, McGraw-Hill, N.Y., July 22,
nals of the SASW method using the multiple fil- pp. 30-34.
ter/crosscorrelation technique. Soil Dynamics and Earth- Fernandez, A. 2000. Tomographic imaging stress fields in dis-
quake Eng., Vol. 13, Elsevier, pp. 13-24. crete media. Ph.D. Dissertation, Georgia Institute of Tech-
Al-Hunaidi M.O., and Rainer J.H. 1995. Analysis of multi- nology, Atlanta.
mode signals of the SASW method. Proc. 7th Int. Conf. Gabriels, P., Snieder, R., and Nolet, G. 1987. In situ measure-
Soil Dynamics and Earthquake Eng., pp.259-266. ments of shear-wave velocity in sediments with higher-
Allen, N. F., Richart, F.E., Jr., and Woods, R.D. 1980. Fluid mode Rayleigh waves. Geophys. Prospect., Vol. 35, pp.
wave propagation in saturated and nearly saturated sands. 187-196.
Journal of Geotechnical Engineering, ASCE, Vol. 106, No. Ganji V., Gukunski N., and Nazarian S. 1998. Automated in-
GT3, pp. 235-254. version procedure for spectral analysis of surface waves. J.
Al-Shayea, N. 1994. Detection of subsurface cavities using the Geotech. and Geoenv. Eng., Vol. 124, ASCE, pp. 757-770.
spectral-analysis-of- surface waves method. Ph.D. Disserta- Gucunski N., and Woods R.D. 1991. Inversion of Rayleigh
tion, University of Michigan, April, 269 pp. wave dispersion curve for SASW test. 5th Int. Conf. on Soil
Al-Shayea, N., Woods, R.D., and Gilmore, P. 1994. SASW and Dynamics and Earthquake Eng., Kalsruhe, pp. 127-138.
GPR to detect buried objects. Proceedings of the Sympo- Gucunski N., and Woods R.D. 1992. Numerical simulation of
sium on the Application of Geophysics to Engineering and SASW test. Soil Dynamics and Earthquake Eng., Vol. 11
Environmental Problems (SAGEEP), Vol. 1, pp. 543-560. (4), Elsevier, pp. 213-227.
Andrus, R.D. and Stokoe, K.H., II. 2000. Liquefaction resis- Gucunski, N., and Shokouhi, P. 2004. Application of Wavelets
tance of soils from shear-wave velocity. Journal of Geo- in Detection of Cavities Under Pavement by Surface
technical and Geoenvironmental Engineering, American Waves. Proceedings of the 5th International Conference on
Society of Civil Engineers, Vol. 126, No. 11, pp. 1019- Case Histories in Geotechnical Engineering, New York,
1025. April 13-17, paper 10.09.
ASCE Press. 1998. Geophysical exploration for engineering Guy, E.D., Nolen-Hoeksema, R.C., Daniels, J.J., and Lefchick,
and environmental investigations. Adapted from the US T. 2003. High-resolution SH-wave seismic reflection inves-
Army Corps of Engineers, No. 23, Technical engineering tigations near a coal mine-related roadway collapse feature.
and design guides, 204 p. Journal of Applied Geophysics, Vol. 54, pp. 51-70.
Blake, W. D., and Gilbert, R. B. 1997. Investigation of possible Hardin, B.O. and Drnevich, V.P. 1972a. Shear modulus and
relationship between undrained shear strength and shear damping in soils – I. measurement and parameter effect.
wave velocity for normally consolidated clays. Offshore Journal of Soil Mechanics and Foundation Engineering,
Technology Conf., Houston, 411-420. ASCE, Vol. 98, June, pp. 603-624.
Bolt, B.A. 1976. Nuclear Explosions and Earthquakes, W.H. Hardin, B.O. and Drnevich, V.P. 1972b. Shear modulus and
Freeman and Company. damping in soils – II. design equations and curves. Journal
Briaud, J-L., and Gibbens, R.M., Editors 1994. Predicted and of Soil Mechanics and Foundation Engineering, ASCE,
Measured Behavior of Five Spread Footings on Sand. Geo- Vol.98, June, pp. 667-692.
technical Special Publication No. 41, ASCE. Results of Haskell N.A. 1953. The dispersion of surface waves on multi-
Spread Footing Prediction Symposium, Sponsored by the layered media. Bulletin of the Seismological Society of
Federal Highway Administration, June, 255 p. America, Vol. 43 (1), pp. 17-34.
Burger, H.R. 1992. Exploration Geophysics of the shallow sub- Heisey, S., Stokoe, K.H., II and Meyer, A.H. 1982. Moduli of
surface, Prentice Hall, Englewood Cliffs, New Jersey. pavement systems from spectral analysis of surface waves.
Burland, J.P. 1989. Small is beautiful: the stiffness of soils at Transportation Research Record 852, pp. 22-31.
small strains. Canadian Geotechnical Journal, Vol. 26, No. Hiltunen, D.R., Nolen-Hoeksema, R.C., and Woods, R.D.
4, pp. 499-516. 2004. Characterization of abandoned mine sites beneath I-
Campanella, R.G., Robertson, P.K., and Gillespie, D. 1986. 70 via crosshole and SASW seismic wave methods. Pro-
Seismic cone penetration test, Use of in situ tests in geo- ceedings of 5th International Conference on Case Histories
technical engineering. Proceedings, In Situ ’86, ASCE Geo- in Geotechnical Engineering, New York, N.Y, April 13-17,
technical Specialty Publication No. 6, Samuel P. Clemence paper 7.08.
(ed.), Blacksburg, VA, June, pp. 116-130. Hossian, M.M., and Drnevich, V.P. 1989. Numerical and opti-
Chiara, N. 2001. Investigation of small-strain shear stiffness mization techniques applied to surface waves for backcal-
measured in field and laboratory geotechnical studies. M.S. culation of layer moduli. Nondestructive testing of pave-
Degree, University of Texas. ments and backcalculation of moduli. ASTM STP 1026,
Cho, G. C., and Santamarina, J. C. 2001. Unsaturated particu- A.J. Bush III, and G.Y. Baladi, eds., Am. Soc. for Testing
late materials: particle-level Studies. Journal of Geotechni- and Mat., Philadelphia, PA., pp. 649-669.
cal and Geoenvironmental Engineering, January Vol. 127, Jamiolkowski, M., Lancellotta, R., and Lo Presti, D. (Eds.).
Issue 1, pp. 84-96. 2001. Pre-Failure Deformation Characteristics of Geomate-
rials. Vols. 1 and 2, A.A. Balkema Publishers, Tokyo

Jardine, R.J., Davies, M.C.R., Hight, D.W., Smith, A.K.C., and Nigbor R.L., and Imai T. 1994. The suspension P-S velocity
Stallebrass, S.E. (Eds.). 1998. Pre-Failure Deformation Be- logging method. Geophysical Characteristics of Sites,
havior of Geomaterials. Thomas Telford Publishing, Lon- ISSMFE, Technical Committee 10 for XIII ICSMFE, Inter-
don, 417 pp. national Science Publishers, New York, pp. 57-63.
Joh, S.-H. 1996. Advances in interpretation and analysis tech- NRC. 2000. Seeing into the earth. Committee for Noninvasive
niques for spectral-analysis-of-surface-waves (SASW) Characterization of the Shallow Subsurface for Environ-
measurements. Ph.D. dissertation, University of Texas at mental and Engineering Applications, P.R. Romig, Chair,
Austin. 129 p.
Joh, S.-H., Rosenblad, B. L., and Stokoe, K. H., II. 1997. Im- Olson, L.D., Sack, D.A., and Stokoe, K.H., II. 1993. Stress-
proved data interpretation method for SASW tests at com- wave nondestructive testing of tunnels and shafts,. Trans-
plex geotechnical sites. International Society of Offshore portation Research Record, No. 1415, pp. 95-99.
and Polar Engineering. Park, C.B., Miller, R.D., and Xia, J. 1999. Multichannel analy-
Kansas Geological Survey 2003. Handout of the MASW work- sis of surface waves (MASW): Geophysics, Vol. 64, pp.
shop: Symposium on the application of geophysics to engi- 800-808.
neering and environmental problems, San Antonio, TX. Richart, F.E., Jr., Hall, J.R. and Woods, R.D. 1970. Vibrations
Kausel E., and Roësset J.M. 1981. Stiffness matrices for lay- of Soils and Foundations. Englewood Cliffs, New Jersey,
ered soils. Bulletin of the Seismological Society of Amer- Prentice Hall, 414 p.
ica, Vol. 71 (6), pp. 1743-1761 Rix G.J., and Leipski, A.E. 1991. Accuracy and resolution of
Kitsunezaki, C. 1980. A new method for shear wave logging. surface wave inversion. Recent advances in instrumenta-
Geophysics, Vol. 45, pp. 1489-1506. tion, data acquisition and testing in soil dynamics. Geo-
Konstantinidis, B., Van Riessen, G., and Schneider, J.P. 1986. technical special publication, No. 29, N.Y., ASCE 1991,
Structural settlements at a major power plant. Settlement of pp. 17-23.
Shallow Foundations on Cohesionless Soils: Design and Robertson, P.K., and Campanella, R.G. 1985. Liquefaction po-
Performance, Geotechnical Special Publication No. 5, tential of sands using the CPT. Journal of Geotechnical En-
ASCE, Seattle Washington, pp.54-73. gineering, ASCE, Vol. 111, No. 3, pp. 384-403.
Landisman, M., Dziewonski, A, Sato, Y., and Masse, R. 1968. Robertson, P.K., Campanella, R.G., Gillespie, D., and Rice, A.,
Cited in Geophys. J. Roy. Astron. Soc., Vol. 17, p.369. 1986. Seismic CPT to measure in situ shear wave velocity.
Liu, H.-P., Warrick, R.E., Westerlund, R.E., Fletcher, J.B., and Journal of Geotechnical Engineering, American Society of
Maxwell, G.L. 1988. An air-powered impulsive shear-wave Civil Engineers, Vol. 112, No. 8, pp 791-803.
source with repeatable signals. Bulletin Seismological So- Robertson, P.K., and Mayne, P.W. (Eds.). 1998. Geotechnical
ciety of America, Vol. 78, No. 1, pp 355-369. Site Characterization. Vols. 1 and 2, A.A. Balkema Pub-
Long, L.G. 1980. Comparison of field and laboratory dynamic lishers, Rotterdam.
soil properties,” M.S. Thesis, University of Texas at Austin. Roësset J.M., Chang D.W., Stokoe K.H. 1991. Comparison of
Luke, B.A., and Stokoe, K.H., II. 1998. Application of SASW 2-D and 3-D models for analysis of surface wave tests.
method underwater. Journal of Geotechnical and Geoenvi- Proc. 5th Int. Conf. on Soil Dyn. and Earthq. Eng.,
ronmental Engineering, Vol. 124, No. 6, June, pp. 523-531 Kalsruhe, Vol. 1, pp. 111-126.
Luke, B.A., Stokoe, K.H., II, Bay, J.A., and Nelson, P.P. 1999. Rosenblad, B.L. 2000. Experimental and Theoretical studies in
Seismic Measurements to investigate disturbed rock zones. support of implementing the spectral-analysis-of-surface-
3rd National Conference of Geo-Institute of ASCE, Urbana- waves (SASW). Ph.D., University of Texas.
Champaign, IL. Rosenblad, B. L., Stokoe, K.H., II, Kalinski M.E. and
Lunne, T., Robertson, P.K., and Powell, J.M. 1997. Cone Pene- Kavazanjian E., 2003 Shear wave velocity profiles of sedi-
trometer Testing in Geotechnical Practice. 1st Edition, ments determined from surface wave measurements,” 13th
London, NY, Blackie Academic & Professional Press, 312 International Offshore and Polar Engineering Conference,
p. Honolulu, Hawaii.
Madianos, M., Nelson, P.P., and Stokoe, K.H., II. 1990. Rosenblad, B.L. and Stokoe, K.H., II. 2001. Offshore shear
Evaluation of discrete discontinuities and repeated monitor- wave velocity profiling using interface waves. OTRC 2001
ing of rock mass characteristics with SASW testing,. 8th Conference honoring Prof. Wayne Dunlap, Houston, TX.
Annual Workshop, Generic Mineral Technology Center, Santamarina, J. C., and Fratta, D. 1998. Introduction to Dis-
U.S. Bureau of Mines, Reno, Nevada, pp. 15-23. crete Signals and Inverse Problems in Civil Engineering.
Matthews M.C., Hope V.S., and Clayton C.R.I. 1996. The use ASCE Press, Reston, 327p.
of surface waves in the determination of ground stiffness Santamarina, J. C., Klein, K.A., and Fam, M.A. 2001. Soils and
profiles. Geotechnical Eng., Vol. 119, Proc. Inst. Civil Eng, Waves. John Wiley & Sons, Ltd., Rexdale, Ontario, Can-
pp. 84-95 ada, 488 p.
McMechan, G.A., and Yedlin, M.J. 1981. Analysis of disper- Sato, Y. 1971. Bull. Earthq. Res. Inst. Tokyo, Vol. 33, p.33.,
sive waves by wave field transformation. Geophysics, Vol. Cited in Dziewonski A.M. and Hales, A.L., Numerical
46, pp. 869-874. Analysis of Dispersed Seismic Waves. Methods in compu-
Menke, W. 1989. Geophysical Data Analysis: Discrete Inverse tational physics, Vol. 11, pp.39-85.
Theory. Academic Press, 289 p. Seed, H.B. 1969. Influence of Local Soil Conditions on Earth-
Menzies B., and Matthews M. 1996. The continuous surface- quake Response. Proceedings of Specialty Session 2 – Soil
wave system: a modern technique for site investigation. Dynamics, Seventh International Conference on Soil Me-
Special Lecture: Indian Geot. Conf., Madras chanics and Foundation Engineering, Mexico City, Mexico.
Miller, R.D., Xia, J., Park, C.B., and Ivanov, J. 1999. Mul- Seed, H.B., and Idriss, I.M. 1971. Simplified procedure for
tichannel analysis of surface waves to map bedrock. The evaluating soil liquefaction potential. Journal of Soil Me-
Leading Edge, Vol. 18, no. 12, pp. 1392-1396. chanics and Found. Div., ASCE, Vol. 97, No. 9, pp. 1249-
Nazarian, S., and Stokoe, K.H., II. 1984. Nondestructive test- 1273.
ing of pavements using surface waves. Transportation Re- Sharma, P.V. 1997. Environmental and Engineering Geophys-
search Record 993, pp. 67-79. ics, Cambridge University Press, Cambridge, U.K., 475 pp.
Nazarian, S., Baker, M., and Crain, K. 1995. Use of seismic Shokouhi, P. and Gucunski, N. 2003. Application of Wavelet
pavement analyzer in pavement evaluation. Transportation Transform in Detection of Shallow Cavities by Surface
Research Record 1505, pp.1-8. Waves. Proceedings of the Symposium on the Application
of Geophysics to Engineering and Environmental Problems Xia, J., Miller, R.D., and Park, C.B. 1999. Estimation of near-
(SAGEEP), EEGS, San Antonio, TX. surface velocity by inversion of Rayleigh wave. Geophys-
Sibuya, S., Mitachi, T., and Miura, S. (Eds.) 1994. Symposium ics, 64, 691-700.
on Pre-Failure Deformation Characteristics of Geomateri- Yokota, K. and Konno, M. (1985). Comparison of soil con-
als, Two Volumes, Sapporo, Japan. stants obtained from laboratory tests and in situ tests. Sym-
Stokoe, K.H., II and Santamarina, J.C. 2000. Seismic-wave- posium on Evaluation of Deformation and Strength of
based testing in geotechnical engineering. Plenary Paper, Sandy Grounds. Japanese Society of Soil Mechanics and
International Conference on Geotechnical and Geological Foundation Engineering, pp. 111-114 (in Japanese).
Engineering, GeoEng 2000, Melbourne, Australia, pp. Yasuda, S. and Yamaguchi, I. 1985. Dynamic shear modulus
1490-1536. obtained in the laboratory and in situ. Symposium on
Stokoe, K.H., II, Rosenblad, B.L., Bay, J.A., Redpath, B., Evaluation of Deformation and Strength of Sandy Grounds.
Diehl, J.G., Steller, R.A., Wong, I.G., Thomas, P.A. and Japanese Society of Soil Mechanics and Foundation Engi-
Luebbers, M., 2003. Comparison of VS profiles from three neering, pp. 115-118 (in Japanese).
seismic methods at Yucca Mountain. Volume I, Soil and Youd, T.L., Idriss, I.M., Andrus, R.D., Arango, I., Castro, G.,
Rock America 2003,Cambridge, MA, pp. 299-306. Christian, J.T., Dobry, R., Finn, W.D.L., Harder, L.F., Jr.,
Stokoe, K.H.II, Rosenblad, B.L., Wong, I.G., Bay, J.A. Tho- Hynes, M.E., Ishihara, K., Koester, J.P., Liao, S.S.D., Mar-
mas, P.A., and Silva, W.J. 2004. Deep Vs profiling along cuson W.F., III, Martin, G.R., Mitchell, J.K., Moriwaki, Y.,
the top of Yucca Mountain using a vibroseis source and Power, M.S., Robertson, P.K., Seed, R.B., and Stokoe,
surface waves. 13th World Conference on Earthquake Engi- K.H., II. 2001. Liquefaction resistance of soils: summary of
neering, Vancouver, B.C., Canada (accepted for publica- report from the 1996 NCEER and 1998 NCEER/NSF
tion). workshops on evaluation of liquefaction resistance of soils.
Stokoe, K.H., II, and Nazarian, S. 1985. Use of Rayleigh waves Journal of Geotechnical and Geoenvironmental Engineer-
in liquefaction studies. Proceedings, Measurement and Use ing, Vol. 127, No. 10, pp. 817-833.
of Shear Wave Velocity for Evaluating Dynamic Soil Prop- Yuan D., and Nazarian S. 1993. Automated surface wave
erties. Geotechnical Engineering Division, ASCE, pp. 1-17. method: inversion technique. J. Geotechnical Eng., Vol.
Stokoe, K.H., II, Wright, S.G., Bay, J.A. and J.M. Roesset 119, No.7, ASCE, pp. 1112-1126
1994. Characterization of geotechnical sites by SASW
method. Geophysical Characteristics of Sites, ISSMFE,
Technical Committee 10 for XIII ICSMFE, International
Science Publishers, New York, pp. 15-25.
Stoll, R.D., Bryan, G.M., and Bautista, E.O. 1994. Measuring
lateral variability of sediment geoacoustic properties. Jour-
nal Acoustical Society of America, Vol. 96, No. 1, pp. 427-
438.
Swiger, W.F. 1974. Evaluation of Soil Moduli. Proceedings of
Conference on Analysis and Design in Geotechnical Engi-
neering, Vol. II, Austin, TX, June, pp. 79-92.
Thomson W.T. 1950. “Transmission of elastic waves through a
stratified solid medium. J. Applied Physics, Vol. 21 (1), pp.
89-93
Tokimatsu, K. 1995. Geotechnical site characterization using
surface waves. First International Conference on Earth-
quake Geotechnical Engineering, Vol. 3, Kenji Ishihara,
editor, Tokyo, pp. 1333-1368.
Tokimatsu, K., Tamura, S., and Kojima, H. 1992. Effects of
multiple modes on Rayleigh wave dispersion. J. Geotech.
Engrg., ASCE, 118(10), 1529-1543.
Toksoz, M.N., and Cheng, C.H. 1991. Wave Propagation in a
Borehole. in J.M. Hovem, M.D. Richardson, and R.D. Stoll
(eds.), Shear Waves in Marine Sediments, Kluwer Aca-
demic Publishers, Dordrecht, The Netherlands.
Volante, M., Scattolini, E., and Pizzarotti, E.M. 2004. Injection
Consolidation Under the Piers of the Railway Bridges for
the Rehabilitation of the Line Merano-Malles. Proceedings
of the 5th International Conference on Case Histories in
Geotechnical Engineering, New York, N.Y., April 13-17,
Paper No. 10.07.
Ward, S. H. 1990. Geotechnical and Environmental Geophys-
ics. Investigations in Geophysics No. 5, Society of Explora-
tion Geophysics, 3 volumes.
Woods, R. D., and Partos, A. 1981. Control of Soil Improve-
ment by Crosshole Testing. Proc. of the 10th International
Conference of the Inter. Soc. for Soil Mech. and Found.
Engrg. Stockholm, Sweden, Vol. 3, June, pp. 793-796.
Wright, S.G., Stokoe, K.H., II and Roesset, J.M. 1994. SASW
Measurements at geotechnical site overlaid by water. Dy-
namic Geotechnical Testing: Second Volume, ASTM STP
1213, R.J. Ebelhar, V.P. Drnevich and B.L. Kutter, Eds.,
American Society for Testing and Materials, Philadelphia,
pp. 39-57.

Stokoe Et Alii 2004 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stokoe Et Alii 2004 PDF

Uploaded by

Copyright:

Available Formats

Proceedings ISC-2 on Geotechnical and Geophysical Site Characterization, Viana da Fonseca & Mayne (eds.

Some contributions of in situ geophysical measurements to solving

methods are noninvasive which make them well

2.1 Types of Stress Waves • • • •

Traditionally, in situ seismic testing has been con-

motions (stress waves) at other points in the earth.

terms of their far-field particle motions, are illus-

propagation. Rayleigh waves (R waves) exist be- D irection of P ropagation

cause of the exposed ground surface. R waves have

small strains and low frequencies. However, strati-

graphy and other forms of heterogeneity cause fre-

is based as noted in Section 3.) The far-field veloci-

98 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

Normalized Wave Velocity, V/VS

2.4 Traditional Field Methods

100 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

a. Crosshole test b. Downhole test

b. Downhole Testing Arrangement

c. Seismic cone penetrometer test d. Suspension logging test

102 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

Table 2. Advantages and disadvantages of four, widely used surface-wave methods

104 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

Phase Vel., m/sec

Unwrapped Phase, deg

Figure 7. Spectral-analysis-of-surface-waves (SASW) method: Calculation of phase velocities

0 200 400 600 800

100 D = 4m Wavelength, m Inversion

-100 Shear Wave Velocity, m/sec

106 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

108 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

(b) Analysis procedure: Frequency-wave number approach

Best-Fit Line Phase Velocity :

Figure 12 Continuous surface wave (CSW) method

110 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

VS = 60 m/sec 0.3 m VS = 90 m/sec 0.3 m

VS = 90 m/sec 0.3 m VS = 60 m/sec 0.3 m

VS = 120 m/sec VS = 120 m/sec

a. Layered systems used to model two soil profiles

Phase Velocity, m/sec

(b) Theoretical dispersion curves for each layered system

Figure 14 Comparison of global and array inversion analyses (Joh, 1996)

112 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

114 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

Table 4. Factors of Safety F = Qf / Qd (measured design

Figure 18. Young’s modulus determined from three methods

Burland concluded that these results along with

4.1.4 Case history 4 – settlement analysis of

4.2 Crosshole Seismic Velocity for Grouting

4.2.1 Case history 1 - forge foundations

116 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

4.2.2 Case history 2 - subway construction

4.3.1 Example 1 – use of GPR, SASW and cross-

showed substantially different dispersion curves,

118 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

were performed at six locations selected by inter-

120 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

stiffer than concrete

offshore near Vancouver, B.C. (Rosenblad and Stokoe, 2001)

122 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9

sive set of Vs measurements, in terms of multiple Logging

ability shown by ±σ includes both measurement un-

Table 5 Numerical Analyses of Average VS Profiles (from Stokoe et al., 2003)

124 © 2004 Millpress, Rotterdam, ISBN 90 5966 009 9