Sie sind auf Seite 1von 11

This paper has been modified from the original to correct the pagination.

No other material has been changed.

Geotechnical Testing Journal, Vol. 28, No. 5


Paper ID GTJ12196
Available online at: www.astm.org

E. C. Leong,1 S. H. Yeo,2 and H. Rahardjo3

Measuring Shear Wave Velocity Using


Bender Elements

ABSTRACT: The use of bender element to measure shear wave velocity is popular due to its simplicity. However, there are still some uncertainties in
the interpretation of the bender element test. In this paper, the use of bender elements in the determination of shear wave velocities of dry, unsaturated,
and saturated soil specimens is examined with respect to the characteristics of the bender elements, waveform, magnitude, and frequency of the
applied voltage to the transmitter bender element and method of travel time determination. The interpretation of the bender element test can be
improved if two performance criteria are adopted: (1) a signal-to-noise ratio of at least 4 dB for the receiver signal, and (2) a wave path length to
wavelength ratio of at least 3.33. Procedures to estimate the strain level associated with the shear wave velocity measurement using bender element
are also described.

KEYWORDS: Bender element, travel time, shear wave, velocity, signal-to-noise ratio, small strain

Introduction and differences arise with respect to the method of interpretation.


The objective of the paper is to evaluate the differences and their
The original concept of using piezoelectric elements for geotech-
implications for the bender element test.
nical applications can be traced to the work of Lawrence (1963,
1965) who used piezoelectric crystals to generate one-dimensional
compression waves through sand and glass beads. Shirley (1978) Test Principles and Sources of Errors
was the first to use piezoceramic bender element for generating
In the bender element test for measurement of shear wave veloc-
and receiving shear waves in laboratory tests. Bender element con-
ity, a pair of bender elements is used whereby one of the bender
sists of two sheets of piezoelectric ceramic material such as lead
elements acts as the shear wave transmitter and the other bender
zirconate titanate, barium titanate, or lead titanate sandwiching a
element acts as the receiver. A schematic diagram of the bender
center shim of brass, stainless steel, or other ferrous nickel alloys
element test setup is shown in Fig. 1. By measuring the travel time
to add strength to it. The bender element is an electromechanical
of the wave, the wave velocity Vs is determined as follows (Dyvik
transducer capable of converting mechanical energy into electrical
and Madshus 1985; Brignoli et al. 1996; Viggiani and Atkinson
energy and vice versa. When a driving voltage is applied to the
1995):
bender element, the polarization will cause a bending displacement
and, thus, the bender element acts as a signal generator. When the
element is forced to bend, a voltage is generated and, thus, the Ltt
Vs = (1)
bender element can act as a signal receiver. t
Over the years, bender elements have found their way into a where Ltt is the tip-to-tip distance between the transmitter and
number of geotechnical testing apparatuses. Bates (1989), Brignoli receiver bender element and t is the travel time of the wave from the
et al. (1996), and Pennington et al. (2001) measured shear wave transmitter to the receiver. From the shear wave velocity, the shear
velocity in triaxial specimens using piezoceramic bender elements. stiffness Gmax can be determined from the elastic wave propagation
Dyvik and Madshus (1985) measured small strain stiffness, Gmax , theory:
of soil specimens in resonant column, oedometer, and direct simple
shear apparatuses using bender elements. Kawaguchi et al. (2001)
measured Gmax in an oedometer using bender elements. Agarwal Gmax = ρVs2 (2)
and Ishibashi (1991) used bender elements in a triaxial cubical
where ρ is the total density of the soil specimen.
box device. Blewett et al. (2000) measured shear wave velocity of
As most laboratory tests involve soil specimens of a finite size,
saturated sand using bender elements in a large container. How-
there is reflection and refraction of the wave at the specimen’s
ever, to date the bender element test has not been standardized
boundary as it travels from the transmitter bender element to the
receiver bender element. The interference of incident and reflected
Received August 21, 2003; accepted for publication January 27, 2005; pub- waves at rigid boundaries can affect the interpretation of travel
lished September 2005. time (Arulnathan et al. 1998). Due to the distortion of the wave
1 Associate Professor, School of Civil and Environmental Engineering,
during its passage through the soil specimen, four methods have
Nanyang Technological University, Singapore. been suggested for the determination of travel time, t (Viggiani and
2 Formerly Project Officer, School of Civil and Environmental Engineering,
Nanyang Technological University, Singapore.
Atkinson 1995; Arulnathan et al. 1998). They are first arrival time,
3 Professor, School of Civil and Environmental Engineering, Nanyang Tech- travel time between the characteristic points, cross-correlation of in-
nological University, Singapore. put and output signals, and cross-power of transmitter and receiver

Copyright by ASTMbyInt'l (all rights reserved);100


Mon JanHarbor
29 00:01:01 EST
PO2018
Copyright © 2005 ASTM International, Barr Drive, Box C700, West Conshohocken, PA 19428-2959. 488
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
489 GEOTECHNICAL TESTING JOURNAL

Signal
Mancuso and Vinale (1988) showed that the first deflection of
Amplifier/Filter the receiver signal may not correspond to the arrival of the wave
but to the arrival of the near-field component. The wave gener-
ated is nonuniform within the near field and, thus, results in am-
Receiver
biguity in determining arrival time. Sanchez-Salinero et al. (1986)
bender and Brignoli et al. (1996) showed experimentally the existence
element of near-field effect masking the first arrival of the wave. Arroyo
Oscilloscope Soil
Specimen et al. (2002) found that signal distortion was not due to near-
Computer for data Transmitter field effects alone and signal distortion still occurred beyond the
processing bender Stokes’ source near field. Viggiani and Atkinson (1995) concluded
element
that (1) travel time cannot be reliably determined from the first
Function arrival time of the receiver signal due to near field effects; (2)
Generator Power Amplifier
the most accurate travel time is given by the cross-correlation or
FIG. 1—Bender element test setup. the cross-power of the transmitter and the receiver signals; and
(3) the use of characteristic points to determine travel time is a
simple alternative provided that it was proven first to be consis-
signals. First arrival time is defined as the travel time to the first
tent with that given by the more rigorous cross-correlation method.
arrival of the receiver signal (ta ). Some researchers have taken first
However, Santamarina and Fam (1997) pointed out that the de-
arrival as first deflection point of the receiver signal, while others
termination of travel time using cross-correlation is only valid
have taken first arrival as first reversal point of the receiver signal.
if both the input and output signals are of the same “nature.”
Due to the difficulty in determining first arrival time, travel time
Gajo et al. (1997) found the cross-correlation method underesti-
has been computed based on characteristic points given by peaks
mated the arrival time in their bender element tests. Arulnathan
or troughs of the transmitter and receiver signals. The travel time
et al. (1998) found that travel time based on characteristic peaks
is given by the time between corresponding peaks (tp ) or troughs
or cross-correlation between transmitter and receiver signals is in-
(tt ) of the transmitter and receiver signals. Variations in determin-
correct because of wave interference at the boundaries, phase lag
ing travel time from characteristic points were also used such as
or signal distortion, and near-field effects. Arulnathan et al. (1998)
the time from the start of the transmitter signal to the first peak
further suggested that travel time should be determined from the
of the receiver signal (Viggiani and Atkinson 1995; Kawaguchi et
second wave arrival, which is less affected by wave interference at
al. 2001). In the cross-correlation between transmitter and receiver
the boundaries or the transfer functions relating electrical signals
signals, the travel time is taken as the time shift that produces the
to physical waves. However, they admitted that the second wave
peak cross-correlation between the input and output signals (tc ). It
arrival still suffers from near-field effects. Kawaguchi et al. (2001)
was originally developed for the interpretation of cross-hole tests
concluded that the use of characteristic points to determine travel
(Mancuso et al. 1989). Details of the cross-correlation method are
time is only acceptable at high frequencies.
described in Viggiani and Atkinson (1995) and Arulnathan et al.
In bender element tests, the shear wave propagates through a
(1998).
soil specimen of finite size. The effect of specimen size on shear
According to Bodare and Massarch (1984) and Mancuso et al.
wave propagation can be examined through the limitations placed
(1989), a group travel time of the wave for a range of frequencies
on specimen dimensions specified in ASTM D 2845 (1997a), Stan-
between the transmitter and receiver elements can be determined
dard Test Method for Laboratory Determination of Pulse Velocities
by linear interpolation of the absolute cross power spectrum phase
and Ultrasonic Elastic Constants of Rock. In ASTM D 2845, wave
diagram. The phase velocity for each frequency is
travel time is based on first arrival time. ASTM D 2845 recom-
mended that the ratio of the length to diameter (L/D) of the spec-
Ltt imen should not exceed 5 to obtain reliable wave velocity. Most
Vs = λf = (3)
t geotechnical laboratory test specimens have L/D ratios ranging
where λ is wavelength, f is the output frequency, and t is travel from 0.3 to 2.5. ASTM D 2845 assumes that the specimen is an
time. The wavelength can be expressed as follows: infinite medium and to eliminate wave dispersion for compression
waves, the diameter of the specimen should be greater than five
2πLtt times the wavelength (λ). Dispersion arises when the wave inter-
λ= (4) acts with the boundary of the specimen causing continual partial
angle conversion of shear wave to compression wave and vice versa.
where angle is the phase angle of the cross-power spectrum. For Silaeve and Shamina (1958) found that the limiting D/λ ratio for
each transmitter and receiver signal that is cross-powered, the cross metal is about 2. Stephenson (1978) reported that the limiting D/λ
power spectrum, GI O (f ), can be obtained. In addition, the phase ratio for any solid is 1. However, the dispersion phenomenon is ab-
angle at which the peak magnitude occurs can be found. The tests sent for shear wave propagating in the fundamental mode (Wasley
can be repeated for a range of frequencies and the phase angle is 1973), and therefore, the restriction requiring the D/λ ratio of at
plotted against the frequencies. Substituting Eq 4 into Eq 3, the least 5 does not apply. Lutsch (1959) and Thill and Peng (1969)
travel time t is given as found that a wavelength that is approximately equal to the grain
size will result in almost complete wave attenuation. Lutsch (1959)
angle recommended a wavelength of ten times the average grain size.
t= (5) For metals, Yoseph and Arjit (1989) recommended a wavelength of
2π f
100 times the average grain size. However, ASTM D 2845 recom-
Therefore, the travel time t is the gradient of the graph of phase mends that the wavelength shall be at least three times the average
angle against frequency divided by 2π. grain size. Work performed on bender elements for measurement

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
LEONG ET AL. ON BENDER ELEMENTS 490

of shear wave velocity suggests that the wave path length to the For a y-poled cantilever bender element in parallel connection,
wavelength ratio (Ltt /λ) is important in the determination of travel  2  
times (Jovičič et al. 1996; Arulnathan et al. 1998; Pennington lb ts
et al. 2001). The Ltt /λ ratio controls the shape of the receiver xf = 3d31 1+ VK (7a)
T T
signal through the degree of attenuation as the wave travels through
the specimen (Sanchez-Salinero et al. 1986; Jovičič et al. 1996).    
3 E T ts
The above discussions highlighted the ambiguities in the interpre- Fmax = Y d31 W 1+ VK (7b)
tation of travel time. The travel time is affected by the wave source, 4 11 lb T
soil type, and dimensions of the specimen. This paper examines where d31 is the piezoelectric strain constant, Y11 E
is the Young’s
the effect of the wave source as governed by the characteristics of modulus, lb is the cantilever (protrusion) length of the bender ele-
the bender element, waveform, frequency, and magnitude of the ment, W is the width of the bender element, T is the thickness of
applied voltage to the transmitter bender element. the bender element, ts is thickness of the centre shim (T ), V is the
applied voltage, and K is an empirical weighting factor (≥1). For
Characteristics of Bender Elements the receiver bender element, the important parameter is the voltage
A literature survey on shear wave velocity measurements using generated V0 . For an x-poled bender element in series connection,
bender elements showed that the type of bender element used was   
not always mentioned. Amongst the different types of piezoelectric 3 F lb ts2
V0 = g31 1− K (8)
ceramics, lead zirconate titanate (PZT) is the most commonly used. 2 WT T2
Shirley and Hampton (1978) and Bates (1989) used PZT4 bender
For a y-poled bender element in parallel connection,
elements. Brignoli et al. (1996) used PZT5A and PZT5HN bender
elements. Argawal and Ishibashi (1991) used PZT5A bender ele-   
3 F lb ts2
ments while Pennington et al. (2001) used PZT5B. Depending on V0 = g31 1− K (9)
polarization, there are two types of bender element: x-poled and y- 4 WT T2
poled (Fig. 2). From the energy point of view, there is no difference where g31 is the piezoelectric voltage constant and F is the applied
between x-poled and y-poled bender elements (Germano 2002). force. As can be seen from Eqs 6 and 7, the y-poled bender element
Both the x-poled and y-poled bender elements act similarly when in parallel connection is more suited as a transmitter as it needs a
connected in a series connection and a parallel connection, respec- lower voltage to generate motion compared to the x-poled bender
tively. The important parameters of a transmitter bender element element in series connection. The motion sensitivity of the y-poled
are the free deflection xf and the maximum force generated Fmax . bender element in parallel connection in terms of deflection per
For an x-poled cantilever bender element in series connection, unit of applied voltage is greater by a factor of two as compared
 2   with the x-poled bender element in series connection. This is ac-
3 lb ts complished by a reduction in impedance of 4:1 (Germano 2002).
xf = d31 1+ VK (6a) The x-poled bender element in series connection is more suitable
2 T T
as a receiver as it generates a higher output voltage per unit force
    applied to the tip of the bender element (Eqs 8 and 9). However, it
3 E T ts is more difficult to make a parallel connection and in most bender
Fmax = Y11 d31 W 1+ VK (6b)
8 lb T element tests, the x-poled bender element in series connection is
E
used. Typical values of Y11 , d31 , and g31 , for PZT4, PZT5A, PZT5B,
and PZT5H piezoceramic elements are summarized in Table 1. The
responses of the bender element in accordance with Eqs 6–9 vary
+ + with the bender element type and the dimensions of the bender
Applied + Applied _ +
voltage
_ voltage + element.
Deflection + Deflection The dimensions of the bender element are important as they affect
the tip deflection of the transmitter element (Eqs 6a and 7a) and
the output voltage of the receiver element (Eqs 8 and 9). Typical
dimensions of bender elements that have been used are summarized
(a) X-poled with series connection (b) Y-poled with parallel connection
in Table 2. The length of bender element used varies from 6 to
FIG. 2—Bender element types and connections. 32 mm, and the width of the bender element used varies from 6

TABLE 1—Piezoelectric constants of commercially available bender elements.

Bender Element Type


PZT4 PZT5A PZT5B PZT5H
Piezoelectric Constant SPK∗ MEC SPK MEC PSI SPK SPK MEC PSI
E
Elastic Modulus, Y11 × 1010 N/m2 7.9 8.1 7.4 6.2 6.6 7.4 5.9 5.9 6.2
Displacement Coefficient, d31 × 10−12 m/V −125 −125 −170 −177 −190 −180 −275 −264 −320
Voltage Coefficient, g31 × 10−3 Vm/N −11.4 −10.6 −10.6 −11.1 −11.6 −10.6 −9.0 −8.9 −9.5
∗ SPK: SPK Electronics Company, Ltd., Taiwan
MEC: Morgan Electro Ceramics, UK
PSI: Piezo Systems, Inc., USA.

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
491 GEOTECHNICAL TESTING JOURNAL

TABLE 2—Typical dimensions of bender elements used.


Wire
Reference Length (mm) Width (mm) Thickness (mm)

Shirley and Hampton (1978) 25.4 6.4 0.5 Top cap


Dyvik and Madshus (1986) 12.7 10.0 0.5
Bates (1989) 15.0 10.0 0.5
Argawal and Ishibashi (1991) 6.4 6.4 0.6
Brignoli et al. (1996) 20.0 10.0 0.5
Arulnathan et al. (1998) 15.0 15.0 1.0 Steel tube
Pennington et al. (2001) 12.0 10.0 1.0
Bender elements epoxy
glued into steel tube
Silicon glue
to 15 mm. For a given bender element type and applied voltage Porous
or applied force, tip deflection and output voltage of the bender stone
element are only dependent on the protrusion length of the bender Bottom
element (lb ). The protrusion length used generally ranges from 3 pedestal O-ring
to 6 mm. The width of the bender element (W ) has a significant
effect on the maximum force (Fmax ) generated by the transmitter
bender element (Eqs 6b and 7b) and the output voltage (V0 ) of the
receiver bender element (Eqs 8 and 9). A larger W increases the Wire
maximum force and reduces the output voltage. A narrow bender
element is preferred as the receiver, as a higher output voltage is FIG. 3—Setup of bender elements in triaxial platens.
generated for a given force F . The thickness of bender element
(T ) affects tip deflection (xf ) and maximum force (Fmax ) of the
transmitter bender element and output voltage (V0 ) of the receiver are shown in Table 1. The rated voltages of the bender element are
bender element. A larger T decreases xf and increases Fmax of the ±180 V and ±90 V for series and parallel connections, respectively.
transmitter bender element and reduces V0 of the receiver bender The dimensions of the bender element are 31.8 × 3.2 × 0.51 mm.
element. However, the thickness of the bender element used varies The bender element was halved to produce two strips of dimensions
in a small range (0.5–1 mm) with most of the bender elements 15.9 × 3.2 × 0.51 mm. A series connection was employed due to
used having a thickness of 0.5 mm (Table 2). When comparing the ease of wiring. The bender element was coated with epoxy glue for
performances of the bender element tests reported in the literature, waterproofing. The bender element was then positioned in a cylin-
there is a need to be aware of the bender element’s characteristics drical sleeve with a protrusion length of 5 mm and fixed in position
in relation to Eqs 6–9. The applied voltage V in Eqs 6 and 7 can using epoxy glue. The sleeve was slotted into the platen (top cap or
partially compensate for the differences in dimensions of the bender bottom pedestal) of the triaxial cell (Fig. 3). A tight-fit was achieved
element used to produce the same tip deflection xf and maximum between the sleeve and the platen by means of O-rings. The gap
force Fmax . A larger voltage applied to the transmitter element will between the sleeve and the platen was further sealed using silicon
generate a stronger wave to excite the receiver element. The stronger rubber. The other components of the test setup consist of a func-
wave provides a larger applied force F , and therefore, produces a tion generator (Hewlett Packard Model 33120A), power amplifier
larger output voltage V0 from the receiver bender element (Eqs 8 (Piezo EPA–104), and digital oscilloscope (Hewlett Packard Model
and 9). 54610B). The excitation voltage produced by the function genera-
tor (20 Vpp ) was amplified by the power amplifier and applied to
the transmitter bender element in the bottom pedestal. The power
Test Setup
amplifier, Piezo EPA-104, is capable of amplifying the function
The bender element test setup shown in Fig. 1 is the most com- generator voltage from 20 Vpp up to 200 Vpp . The wave traveling
plete as it includes a power amplifier for the applied voltage and through the soil specimen was detected by the receiver bender ele-
a signal amplifier for the receiver signal. Most bender element test ment in the top cap. Both the applied voltage and the receiver signal
setups incorporate the signal amplifier for receiver signal but not were recorded on the digital oscilloscope. The oscilloscope has two
the power amplifier for applied voltage (e.g., Shirley and Hamp- recording channels and a maximum sampling rate of 20 MHz. A
ton 1978; Bates 1989; Pennington et al. 2001). The signal ampli- sampling rate of 1 MHz was used in this study. The data recorded by
fier for the receiver signal is not necessary if the receiver signal the digital oscilloscope can be transferred to a computer for further
strength is high enough. Alternatively, signal-processing tools such signal processing.
as stacking and filtering can be used to improve the receiver sig- It is generally assumed that the response of the transmitter bender
nal quality. The bender element test setup used in this study was element is instantaneous on application of a voltage signal. Finite
incorporated in a triaxial cell and does not include the signal am- element analysis conducted by Arulnathan et al. (1998) showed
plifier. However, all the receiver signals were tapered using Han- that there was a phase lag between the applied voltage and the
ning window to eliminate Gibbs’ phenomenon (Hamming 1977) response of the transmitter. To check the phase lag between the
and filtered through two passes of a fourth order, low-pass Butter- applied voltage and the response of the bender element, the trans-
worth filter with a cut-off frequency above the dominant frequency mitter and receiver bender elements were placed in contact and
of the receiver signal to remove the phase shift due to filtering voltage was applied to the transmitter bender element. The ap-
(Li 1997). plied voltage and the receiver signal shown in Fig. 4 revealed that
The bender elements used in this study were x-poled PZT5A there is no noticeable phase lag. There is also a possibility that the
E
(Piezo PSI5A T220-A4-103X) whose values of Y11 , d31 , and g31 wave may travel through other paths other than the soil specimen

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
LEONG ET AL. ON BENDER ELEMENTS 492

30 1.5
Transmitter signal
20 1.0
Receiver signal

Voltage (V)
10 0.5

0 0.0

-10 -0.5

-20 -1.0

-30 -1.5
-200 0 200 400 600 800 1000

Time (µs)
FIG. 4—Response of receiver bender element placed in contact with transmitter bender element.

TABLE 3—Properties of soils tested. Fine


Fines: Silt & Clay Sand
Fine Medium Coarse Gravel
Mudstone
100

Percentage Passing (%)


Properties Sand Residual Soil Kaolin
Kaolin
Specific gravity, Gs 2.63 2.67 2.65 80
Maximum density (Mg/m3 ) 2.18 — — Mudstone
Minimum density (Mg/m3 ) 1.39 — — 60 residual soil
Liquid limit, LL (%) 26 31 74
Plastic limit, PL (%) 23 19 46 40
Plasticity index, PI (%) 3 12 28
Grain size distribution (%) 20
Sand 90 85 85 Sand
Silt and clay 10 15 15
USCS SP-SM SC SM 0
0.0 0.0 0.1 1.0 10.0
Diameter (mm)
FIG. 5—Grain size distributions of soils tested.
(Brignoli et al. 1996). This was checked by testing the bender el-
ement test system without a soil specimen between the platens in
the triaxial cell filled with water. The receiver bender element did Receiver Signal
not respond when a voltage was applied to the transmitter bender
element. The receiver bender element signal is dependent on the volt-
age applied to the transmitter bender element and the soil type.
The applied voltage has three parameters: waveform, magnitude,
and frequency. Table 4 summarizes the waveform, magnitude, and
Soil Samples frequency of applied voltage and soil types tested by others.
In this study, three types of soil were used: sand (dry), mudstone
residual soil (partially saturated), and kaolin (fully saturated). The
Waveform
index properties of the soils are summarized in Table 3. The grain
size distributions of the soils are shown in Fig. 5. The sand spec- Referring to Table 4, two types of waveform are commonly used:
imen was prepared in three layers using dry tamping to achieve square and sinusoidal. In this study, both square wave and sinusoidal
a total density of about 1.97 Mg/m3 . The mudstone residual soil wave were used in the bender element tests on the soil specimens.
was compacted using standard Proctor method (ASTM D 698-91 Figure 6 shows the processed receiver signal for square wave and
1997b). The compacted mudstone residual soil sample has a total sinusoidal wave of the three soil specimens. It can be seen that
density of 2.2 Mg/m3 , water content of 11.1 %, and a degree of the square wave causes greater ambiguity in arrival time than the
saturation of 85 %. The kaolin specimen was prepared by consol- sinusoidal wave. Firstly, the receiver signals for square wave do not
idating kaolin slurry at an initial water content of 200 % under a resemble the original transmitter signal. A similar observation was
vertical load of 200 kPa in a 30-cm-diameter consolidation tank. made by Jovičič et al. (1996) and Blewett et al. (2000). The sharp
The consolidated kaolin sample has a total density of 1.64 Mg/m3 and immediate rise in the signal as well as the flat portion in the sig-
and a water content of 52.8 %. The kaolin and compacted mudstone nal was lost. Secondly, the receiver signals for the square wave have
residual soil specimens were trimmed to a diameter of 50 mm using more distortion at the beginning of the receiver signal. Amongst the
a soil lathe. The length to diameter ratio of the soil specimens was three soil specimens, kaolin has the clearest receiver signal, most
approximately 2 and a confining pressure of 50 kPa was applied to probably due to its contact condition between the receiver bender
the soil specimens during the bender element test, except where it element and the soil specimen. Good contact conditions between
was specified otherwise. the sand particles and the receiver bender element and between the

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
493 GEOTECHNICAL TESTING JOURNAL

TABLE 4—Summary of waveform, magnitude, and frequency of applied voltage and soil type used by other researchers.

Voltage Applied

Reference Waveform Magnitude (Vpp ) Frequency (Hz) Soil Type Method of Interpretation

Dyvik and Madshus (1985) Square 20 5–100 Clay First arrival (first reversal)
Bates (1989) Square 10 2 Sand First arrival (?∗ )
Argawal and Ishibashi (1991) Sine 100–600 13 560 Glass spheres First arrival (?)
First arrival (first deflection,
first reversal)
Viggiani and Atkinson (1995) Square 20 50 Clay Characteristics points
Sine 1000–10 000 Cross correlation
Cross power
Brignoli et al. (1996) Sine 20 3000–10 000 Clay First arrival (first reversal)
Sand
Gajo et al. (1997) Square 150 10 000 Sand First arrival (first deflection)
Sine
First arrival (?)
Arulnathan et al. (1998) Sine 20 900–4,500 Organic soil Characteristics points
Cross correlation
Cross power
Lohani et al. (1999) Square 20 50 Clay First arrival (first reversal)
Blewett et al. (2000) Sine — 200–10 000 Sand —
Diaz-Rodriguez et al. (2001) Square 20 7 Silty clay First peak
Kawaguchi et al. (2001) Square 20 100 Clay First arrival (average of
Sine 1,2,4,8 first rise and first peak)
Pennington et al. (2001) Sine 20 8000–25 000 Clay First arrival (first deflection)
Callisto and Rampello (2002) Square 20 50 Clay Average of first deflection
Sine 10 000 and reversal point
Characteristics peaks
∗ Not specified.

Soil Square wave Sine wave


type
Sand 40 0.02 40 0.02
t1 = 240 µs
Output Voltage (V)

Output Voltage (V)


20 0.01 20 0.01
Input Voltage (V)

Input Voltage (V)

0 0.00 0 0.00

-20 -0.01 -20 -0.01


t2 = 300 µs t1 = 245 µs

-40 -0.02 -40 -0.02


0 250 500 750 1000 0 250 500 750 1000
Time (µs) Time (µs)

Mudstone 30 0.30 30 0.30


residual t1 = 140 µs

soil
Output Voltage (V)

15 0.15 15 0.15
Output Voltage (V)

Input Voltage (V)


Input Voltage (V)

0 0.00 0 0.00

t2 = 265 µs
-15 -0.15 -15 -0.15

t1 = 245 µs

-30 -0.30 -30 -0.30


0 250 500 750 1000 0 250 500 750 1000

Time (µs) Time (µs)

Kaolin 30 0.2 30 0.2


Output Voltage (V)
Input Voltage (V)

15 0.1
15 0.1
Output Voltage (V)
Input Voltage (V)

0 0.0
0 0.0
-15 -0.1
t1 = 266 µs
-15 -0.1
-30 -0.2
0 250 500 750 1000 t1 = 266 µs
Time (µs) -30 -0.2
0 250 500 750 1000

Time (µs)

Note: t1 = first deflection, t2 = first reversal

FIG. 6—Effect of waveform on receiver signals.


Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
LEONG ET AL. ON BENDER ELEMENTS 494

mudstone residual soil specimen and the receiver bender element TABLE 5—Comparison of travel times from bender element tests and
were more difficult to achieve compared with the kaolin specimen. ultrasonic tests.
Sinusoidal wave was found to be a better choice than square wave
Travel Time (µs)
as the distortion of the receiver signal at the beginning is greatly re- Confining Pressure
duced. The subsequent tests described in this paper used sinusoidal Material (kPa) Bender Element Ultrasonic∗
wave as the input signal. Sand 800 ta = 209 ta = 203
Travel times based on first deflection (ta ) (characteristic peak tp = 339
(tp ) is the travel time based on characteristic trough (tt )) and cross- tt = 326
correlation method (tc ) were determined for the soil specimens and tc = 300
Mudstone residual 800 ta = 236 ta = 226
soil tp = 401
tt = 394
0.50 tc = 375
Kaolin 800 ta = 266 ta = 263
ta tp = 335
Output Voltage (V)

tt = 330
0.25 tc = 295
∗ t was calculated from the velocity obtained from ultrasonic test as the
a
0.00 sample length was different from that in the bender element test.

-0.25 are shown in Table 5. A high confining pressure of 800 kPa was
used to eliminate the possibility of poor contact condition between
bender elements and soil specimen. In Table 5, the compacted mud-
stone residual soil specimen has a total density of 2.06 Mg/m3 and
-0.50
a water content of 10.3 %. The travel times were compared with
0 200 400 600 800 1000 the travel times obtained using an ultrasonic test system by GCTS
Time ( µs) (2002) in Table 5. The GCTS ultrasonic test system consists of
a pair of 70-mm-diameter triaxial platens containing compression
FIG. 7—Typical ultrasonic test result for mudstone residual soil speci- and shear wave piezoceramic crystals, a pulse generator with a
men at a confining pressure of 800 kPa. power amplifier, a receiver, a high-speed data acquisition system

Sand Mudstone residual soil Kaolin


25 25
Applied Voltage = 5 Vpp Applied Voltage = 5 Vpp
20 20
Intensity
Intensity

SNR = 0 dB SNR = 0 dB
15 15
10 10
5 5
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Frequency (kHz) Frequency (kHz)

25 25 25
Applied Voltage = 10 V pp Applied Voltage = 10 V pp Applied Voltage = 10 V pp
20 20 20
Intensity

Intensity
Intensity

SNR = 0 dB SNR = 1.8 dB SNR = 3.4 dB


15 15 15
10 10 10
5 5 5
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Frequency (kHz) Frequency (kHz) Frequency (kHz)

25 25 25
Applied Voltage = 20 V pp Applied Voltage = 20 Vpp Applied Voltage = 20 Vpp
20 20 20
Intensity

Intensity

Intensity

SNR = 3.4 dB SNR = 4.0 dB SNR = 6.8 dB


15 15 15
10 10 10
5 5 5
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Frequency (kHz) Frequency (kHz) Frequency (kHz)

25 25 25
Applied Voltage = 70 V pp Applied Voltage = 40 Vpp Applied Voltage = 40 Vpp
20 20 20
Intensity

Intensity

Intensity

SNR = 7.8 dB SNR = 7.4 dB SNR = 9.1 dB


15 15 15
10 10 10
5 5 5
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Frequency (kHz) Frequency (kHz) Frequency (kHz)

FIG. 8—Effect of applied voltage on power spectrum of receiver signal.

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
495 GEOTECHNICAL TESTING JOURNAL

(maximum sampling rate of 40 MHz), and a personal computer.

Relative Amplitude (1 div = 0.01V)


The difference between the ultrasonic test platens and the bender 10 Vpp , SNR = 0.0 dB
element test platens is that the piezoceramic crystals are within the
test platens and do not come into direct contact with the soil speci-
20 Vpp , SNR = 3.4 dB
men. The ultrasonic test system is equipped with signal processing
capabilities such as windowing, filtering, and stacking. In the ul-
trasonic test, the travel time of a train of shear wave pulses with a ta 70 Vpp , SNR = 7.8 dB
central frequency of about 76 kHz through the soil specimen was
determined from the first deflection of the receiver signal. A typical
set of ultrasonic test results for the mudstone residual soil specimen
is shown in Fig. 7. Table 5 shows that the travel times for both the
bender element test and the ultrasonic test are similar only if the 0 200 400 600 800 1000
first deflection of the receiver signal was used for the bender ele- Time (µs)
ment test. Travel times based on characteristic peak, characteristic
trough, and cross-correlation method are unreliable. (a) Sand (Confining pressure = 800 kPa)

Relative Amplitude (1 div = 0.1V)


10 Vpp , SNR = 1.8 dB
Applied Voltage
The magnitude of the voltage applied to the transmitter bender 20 Vpp , SNR = 4.0 dB
element does not affect the shape or appearance of the receiver
signal, but it increases the signal to noise ratio of the receiver 40 Vpp , SNR = 7.4 dB
signal. The signal to noise ratio, SNR, in decibels is given by ta
(Carlson 1986):

Signal power
SNR = 10 log (10)
Noise power
0 200 400 600 800 1000
The signal and noise power can be obtained from the power spec-
Time (µs)
trum or frequency spectrum of the receiver signal. Piersol (1996)
recommended that stationary and transient signals with spectral (b) Mudstone residual soil (Confining pressure = 50 kPa)
values less than 6 dB above the background noise should be dis-
Relative Amplitude (1 div = 0.1V).

carded unless corrective editing can be performed. The effect of 10 V pp , SNR = 3.4 dB
magnitude of applied voltage on signal quality was examined by
applying different voltages from 5 to 70 Vpp to the soil specimens.
Figure 8 shows the power spectra of receiver signals for the differ- 20 V pp , SNR = 6.8 dB
ent applied voltages. As the applied voltage increases, the intensity
or power spectrum magnitude of the receiver signal increases. At ta 40 V pp , SNR = 9.1 dB
5 Vpp , no signal was detected by the receiver bender element for
the sand specimen. Receiver signals with SNR of 4 dB or better
showed less distortion at the beginning, and it becomes easier to pick
the first deflection of the receiver signal (Fig. 9). The higher input
voltage generated a stronger shear wave relative to near-field effects
and minimized the distortion of the receiver signal. Therefore, SNR 0 200 400 600 800 1000
should be used as a criterion when selecting the appropriate voltage Time (µs)
to be applied to the transmitter bender element in bender element
tests. However, it is important to limit the applied voltage to less
(c) Kaolin (Confining pressure = 50 kPa)
than the rated voltage of the bender element to avoid depolarization FIG. 9—Effect of signal-to-noise ratio (SNR) on receiver signals.
of the bender element and eventual failure of the bender element.
The SNR can also be improved by adopting a parallel connection
for the x-poled bender element as the deflection per unit of ap- the wavelength decreases as frequency increases (Eq 3). The use of
plied voltage is greater by a factor of two when compared with the high frequency sinusoidal pulse was recommended to reduce the
x-poled bender element in series connection. near-field effect as it ensures the separation of the near field coupled
compression and shear waves (Jovičič et al. 1996; Brignoli et al.
1996). Theoretical analyses performed by Sanchez-Salinero et al.
Frequency of Applied Voltage
(1986) showed that near-field effects are significant when Ltt /λ is
Near-field effects affect the shape of the receiver signal and cause less than unity. The shape of the receiver signal at the point of
uncertainty in determining the arrival of the shear wave in bender interest will not be masked by near field effects if the Ltt /λ ratio is
element tests (Viggiani and Atkinson 1995; Jovičič et al. 1996; greater than 2 (Sanchez-Salinero et al. 1986). Jovičič et al. (1996)
Brignoli et al. 1996). The near-field effects are quantified in terms of and Brignoli et al. (1996) found that the near-field effect decreases
the wave path length to wavelength ratio, Ltt /λ (Sanchez-Salinero as Ltt /λ increases. Gajo et al. (1997) found near-field effects to
et al. 1986). Input signal frequencies from 5 Hz to 25 kHz have been be negligible in their experiments when Ltt /λ was between 6 and
used by various researchers (Table 4). For a given wave velocity, 9. However, Arulnathan et al. (1998) found that the receiver signal

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
LEONG ET AL. ON BENDER ELEMENTS 496

Frequency (kHz) Signal


30 0.15

Output Voltage (V)


L tt /λ= 0.2

Input Voltage (V)


20 Input 0.10
10 0.05
0 0.00
Output
0.5 -10 -0.05
-20 -0.10
-30 -0.15
0 500 1000 1500 2000
Time (µs)

30 0.15

Output Voltage (V)


Input Voltage (V)
20 L tt /λ= 0.6 0.10
10 0.05
0 0.00
2 -10 -0.05
-20 -0.10
-30 -0.15
0 500 1000 1500 2000
Time (µs)

30 0.18

Ouput voltage (V)


Input Voltage (V)

20 L tt /λ= 1.04 0.12


10 0.06
0 0.00
4 -10 -0.06
-20 -0.12
-30 -0.18
0 500 1000 1500 2000
Time (µs)

30 0.15

Ouput Voltage (V)


Input Voltage (V)

20 L tt /λ= 2.1 0.10


10 0.05
0 0.00
8 -10 -0.05
-20 -0.10
-30 -0.15
0 500 1000 1500 2000
Time (µs)

30 0.15

Ouput Voltage (V)


Input Voltage (V)

20 ta L tt /λ= 4.2 0.10


10 0.05
0 0.00
16 -10 -0.05
-20 -0.10
-30 -0.15
0 500 1000 1500 2000
Time (µs)

FIG. 10—Effect of frequency of applied voltage on receive signal for mudstone residual soil specimen.

deteriorated as Ltt /λ increased and attributed the deterioration to the there is a greater variation in shear wave velocity. This observation
effect of the ratio of wavelength and cantilever length of the bender is in agreement with Sanchez-Salinero et al. (1986), Arulnathan
element (λ/lb ). Near-field effects were absent if Ltt /λ is greater et al. (1998), Pennington et al. (2001), and Arroyo et al. (2002).
than 1 (Arulnathan et al. 1998). Pennington et al. (2001) found ASTM D 2845 (1997a) suggested that for reliable measurement of
that the arrival times of receiver signals can be easily interpreted wave velocity, the specimen length should be at least 10 times the
when the Ltt /λ ratio was between 2 and 10. Arroyo et al. (2002) average grain diameter and the wavelength should be at least three
showed that the Stokes’ near-field influence on shear wave velocity times the average grain diameter. These two requirements suggest
is less than 5 % if Ltt /λ ratio is greater than 1.6. However, Arroyo an Ltt /λ ratio of 3.33. Based on the above discussions, a lower limit
et al. pointed out that the criterion of imposing Ltt /λ > 1.6 is only of 3.33 for theLtt /λ ratio is reasonable. The SNRs of the receiver
useful in determining the shear wave velocity in the frequency signals in Figs. 10 and 11 were greater than 4 dB. It is also possible
domain (cross-correlation method) but not in the time domain (first to increase the Ltt /λ ratio by increasing Ltt , but the receiver signal
arrival time method), where it is signal dependent. strength and, hence, SNR may be reduced due to the longer wave
Bender element tests were performed on the soil specimens at path length.
frequencies of 0.5, 2, 4, 8, and 16 kHz. A typical set of signals for the
mudstone residual soil specimen is shown in Fig. 10. The near-field
Estimation of Strain Level
components observed before the first prominent peak reduce as the
input frequency increases. The near-field components are absent at In bender element tests, the strain levels associated with the
an Ltt /λ ratio of 4.2. The effects of Ltt /λ on the shear wave velocity shear wave velocity cannot be measured directly as the deflection
for the three soils are shown in Fig. 11. Below an Ltt /λ ratio of 2, of the bender element depends on the coupling between the soil and

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
497 GEOTECHNICAL TESTING JOURNAL

400

300

Vs (m/s)
200
L tt /λ = 3.33
100 sand
mudstone residual soil
kaolin
0
0 1 2 3 4 5 6
L tt /λ
FIG. 11—Effect of Ltt /λ on shear wave velocities of the three soils.

transmitter bender element. However, the receiver bender element voltage is used, it should be lower than the rated voltage of the
signal in a tip-to-tip test can be used to provide an estimate of the bender element to avoid depolarization. Applying common signal
maximum strain associated with the shear wave velocity measure- processing procedures (such as filtering and stacking) to the re-
ment. For the bender elements used in this study, Eq 6a reduces to ceiver signal and/or using signal amplifier for the receiver signal
(Piezo Systems, Inc. 2003): can further improve the SNR of the receiver signal. Therefore, a
power amplifier is essential in a bender element test setup, whereas
2lb2 V0 d31
xf = (11) the signal amplifier is optional. The Ltt /λ ratio can be increased
T2 by increasing the frequency of the applied voltage to the transmit-
where lb is the protrusion length of the bender element, T is the ter bender element. It is possible that further improvement to the
thickness of bender element, V0 is the output voltage, and d31 is receiver bender element signal can be achieved using a distorted si-
the piezoelectric strain constant of bender element. By assuming nusoidal pulse (Jovičič 1997) or an amplitude modulated sine burst
that the element tip deflections of the bender element given by (Popovics et al. 1999). These possibilities were not investigated
Eq 11 are equivalent to the soil particle vibration, the particle ve- in the present paper. The strain levels in the bender element tests
. can be estimated using the tip deflection equation of the bender
locity y (t) may be estimated from the time history of xf (t), i.e.,
. element.
. . 2l 2 V0 (t)d31
y(t) = xf (t) = b 2 (12)
T
. Acknowledgement
With y(t), the shear strain γ can be estimated using (White 1965):
. The work described in this paper is part of a research project
ymax
γ= (13) funded by the Defense Science and Technology Agency, Grant No.
Vs PTRC-CSE/LEO/99.02.
.
where ymax is the maximum particle velocity from Eq 12 and Vs is
the shear wave velocity. The shear strain γ is therefore dependent
References
on Vs and f .
Using this method, the maximum shear strain levels of the bender Argawal, T. K. and Ishibashi, I., 1991, “Multi-Directional Wave Ve-
element tests conducted were found to be in the order of 10−4 %. locity by Piezoelectric Crystals,” Proceedings, Recent Advances
Dyvik and Madshus (1985) and Pennington (1999) estimated max- in Instrumentation, Data Acquisition and Testing in Soil Dynam-
imum shear strains in the order of 10−3 % and 10−4 %, respectively, ics, Orlando, FL, Bhatia and Blaney, Eds., ASCE, pp. 102–117.
for their bender element tests. Arroyo, M., Wood, D. M., and Greening, P. D., 2002, “Source Near-
Field Effects and Pulse Tests in Soil Samples,” Géotechnique,
Vol. 53, No. 3, pp. 337–345.
Conclusion
Arulnathan, R., Boulanger, R. W., and Riemer, M. F., 1998, “Anal-
The uncertainties in interpretation of the bender element tests ysis of Bender Element Tests,” Geotechnical Testing Journal,
arose mainly from the quality of the receiver signal. Current un- Vol. 21, No. 2, pp. 120–131.
derstanding indicates that the sinusoidal input voltage and travel ASTM Standard D 2845-95, 1997a: Testing Materials, Labora-
time based on first deflection are preferred. The quality of the tory Determination of Pulse Velocities and Ultrasonic Elastic
receiver signal can be improved by applying the performance cri- Constants of Rock, Annual Book of ASTM Standards, ASTM
teria suggested in this study: (1) the signal-to-noise ratio (SNR) International, West Conshohocken, PA.
of the receiver signal is at least 4 dB, and (2) the wave path ASTM Standard D 698-91, 1997b: Testing Materials, Labora-
length to wavelength ratio (Ltt /λ) is at least 3.33. The SNR can tory Compaction Characteristics of Soil Using Standard Effort,
be improved by increasing the applied voltage to the transmitter Annual Book of ASTM Standards, ASTM International, West
bender element using a power amplifier, and/or adopting a par- Conshohocken, PA.
allel connection for the transmitter bender element and a series Bates, C. R., 1989, “Dynamic Soil Property Measurements During
connection for the receiver bender element. When a high-applied Triaxial Testing,” Géotechnique, Vol. 39, No. 4, pp. 721–726.

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.
LEONG ET AL. ON BENDER ELEMENTS 498

Blewett, J., Blewett, I. J., and Woodward, P. K., 2000, “Phase Lutsch, A., 1959, “The Experimental Determination of the Extent
and Amplitude Responses Associated with the Measurements and Degree of Fracture of Rock Faces by Means of an Ultra-
of Shear Wave Velocity in Sand by Bender Elements,” Canadian sonic Pulse Reflection Method,” Journal of the South African
Geotechnical Journal, Vol. 37, pp. 1348–1357. Institute of Mining and Metallurgy, Vol. 59, No. 8, pp. 412–
Bodare, A. and Massarch, K. R., 1984, “Determination of Shear 419.
Wave Velocity by Different Cross-Hole Methods,” Proceed- Mancuso, C., Simonelli, A. L., and Vinale, F., 1989, “Numerical
ings, Eighth World Conference on Earthquake Engineering, San Analysis of In-Situ S Wave Measurements,” Proceedings, Twelfth
Francisco, California, Vol. 3, pp. 39–45. International Conference on Soil Mechanics and Foundation En-
Brignoli, E. G. M., Gotti, M., and Stokoe, K. H., II, 1996, “Mea- gineering, Rio de Janeiro, Vol. 3, pp. 277–280.
surement of Shear Waves in Laboratory Specimens by Means Mancuso, C. and Vinale, F., 1988, “Propagaziane delle Onde
of Piezoelectric Transducers,” Geotechnical Testing Journal, Simiche: Teoria e Misura in Sito,” Atti Del Convegno Del
Vol. 19, No. 4, pp. 384–397. Gruppo Nazionale Di Coordinamento Per Gli Gtudi Di Ingeg-
Callisto, L. and Rampello, S., 2002, “Shear Strength and Small- naria Geotecnica, Monselice, pp. 115–138.
Strain Stiffness of a Natural Clay under General Stress Condi- Pennington, D. S., 1999, “The Anisotropic Small Strain Stiffness
tions,” Géotechnique, Vol. 52, No. 8, pp. 547–560. of Cambridge Gault Clay,” Ph.D. Thesis, University of Bristol.
Carlson, A. B., 1986, Communication Systems—An Introduction Pennington, D. S., Nash, D. F. T., and Lings, M. L., 2001, “Hor-
to Signals and Noise in Electrical Communication, New York: izontally Mounted Bender Elements for Measuring Anistropic
McGraw-Hill. Shear Moduli in Triaxial Clay Specimens,” Geotechnical Testing
Diaz-Rodriguez, J. A., Moreno-Carrizales, P., and López-Flores, Journal, Vol. 24, No. 2, pp. 133–144.
L., 2001, “A Study of Soil Microstructure Using Bender Ele- Piersol, A. G., 1996, “Validation and Editing of Shock and Vibra-
ment Tests,” Proceedings, Fourth International Conference on tion Data,” 67th Shock and Vibration Symposium Short Course,
Recent Advances in Geotechnical Earthquake Engineering and Monterey, CA.
Soil Dynamics, San Diego, CA, Prakash, Ed., Paper No. 1.39. Piezo Systems, Inc., 2003, Application Data, http://www.piezo.
Dyvik, R. and Madshus, C., 1985, Lab Measurements of Gmax com/.
Using Bender Elements, Norwegian Geotechnical Institute Pub- Popovics, J. S., Achenbach, J. D., and Song, W.-J., 1999, “Ap-
lication, No. 161, Oslo. plication of New Ultrasound and Sound Generation Method for
Gajo, A., Fedel, A., and Mongiovi, L., 1997, “Experimental Anal- Testing Concrete Structures,” Magazine of Concrete Research,
ysis of the Effects of Fluid-Solid Coupling on the Velocity Vol. 51, No. 1, pp. 35–44.
of Elastic Waves in Saturated Porous Media,” Géotechnique, Santamarina, J. C. and Fam, M. A., 1997, “Discussion on Interpre-
Vol. 47, No. 5, pp. 993–1008. tation of Bender Element Tests,” Géotechnique, Vol. 47, No. 4,
GCTS, 2002, “Ultrasonic Velocity Test System,” Commercial Lit- pp. 873–875.
erature, Tempe, AZ. Sanchez-Salinero, I., Roesset, J. M., and Stokoe, K. H., II, 1986,
Germano, C. P., 2002, “Flexure Mode Piezoelectric Transducers,” Analytical Studies of Body Wave Propagation and Attenuation,
Technical Publication TP-218, Morgan Electro Ceramics. Report GR 86-15, Civil Engineering Department, University of
Hamming, R. W., 1977, Digital Filters, Prentice Hall, Englewood Texas of Austin.
Cliffs, NJ. Shirley, D. J., 1978, “An improved shear wave transducer,” Jour-
Jovičič, V., 1997, “The Measurement and Interpretation of Small- nal of Acoustic Society of America, Vol. 63, No. 5, pp. 1643–
Strain Stiffness of Soils,” Ph.D. Thesis, City University, London. 1645.
Jovičič, V., Coop M. R., and Simič, M., 1996, “Objective Criteria for Shirley, D. J. and Hampton, L. D., 1978, “Shear Wave Measure-
Determining Gmax from Bender Element Tests,” Géotechnique, ments in Laboratory Sediments,” Journal of Acoustic Society of
Vol. 46, No. 2, pp. 357–362. America, Vol. 63, No. 2, pp. 607–613.
Kawaguchi, T., Mitachi, T., and Shibuya, S., 2001, “Evaluation Silaeva, O. I. and Shamina, O. G., 1958, “The Distribution of Elastic
of Shear Wave Travel Time in Laboratory Bender Element Pulses in Cyclindrical Specimens,” Geophysics Series, USSR
Test,” Proceedings, Fifteenth International Conference on Soil Academy of Sciences, Vol. 1, No. 1, pp. 17–24.
Mechanics and Geotechnical Engineering, Istanbul, Balkema, Stephenson, R. W., 1978, “Ultrasonic Testing for Determining Dy-
pp. 155–158. namic Soil Moduli,” Dynamic Geotechnical Testing, ASTM STP
Lawrence, F. V., 1963, Propagation of Ultrasonic Waves Through 654, ASTM International, West Conshohocken, PA, pp. 179–195.
Sand, Research Report R63-08, Boston: Massachusetts Institute Thill, R. E. and Peng, S. S., 1969, Statistical Comparison of the
of Technology. Pulse and Resonance Methods for Determining Elastic Moduli,
Lawrence, F. V., 1965, Ultrasonic Shear Wave Velocity in Sand and RI 7831, U.S. Bureau of Mines, Washington, D.C.
Clay, Research Report R65-05, Boston: Massachusetts Institute Viggiani, G. and Atkinson, J. H., 1995, “Interpretation of Bender
of Technology. Element Tests,” Géotechnique, Vol. 45, No. 1, pp. 149–154.
Li, X. S., 1997, “Discussion on Pulse Transmission System for Wasley, R. J., 1973, Stress Wave Propagation in Solids, Marcel
Measuring Wave Propagation in Soils,” Journal of Geotechnical Dekker, Inc., New York.
Engineering, ASCE, Vol. 122, No. 8, p. 883. White, J. E., 1965, Seismic Waves. McGraw-Hill Book Company,
Lohani, T. N., Imai, G., and Shibuya, S., 1999, “Determination Inc., New York.
of Shear Wave Velocity in Bender Element Test,” Proceedings, Yoseph, B. C. and Ajit, K. M., 1989, “Ultrasonic Inspection,” Met-
Second International Conference on Earthquake Geotechnical als Handbook, prepared under the direction of ASM Handbook
Engineering, Lisbon, Portugal, Vol. 1, pp. 101–106. Committee, Metals Park, OH, America Society for Metals.

Copyright by ASTM Int'l (all rights reserved); Mon Jan 29 00:01:01 EST 2018
Downloaded/printed by
Indian Institute of Technology-Bhubneshwar (Indian Institute of Technology-Bhubneshwar) pursuant to License Agreement. No further reproductions authorized.

Das könnte Ihnen auch gefallen