James Barshinger

Krautkramer,
Lewistown, PA 17044
Guided Wave Resonance Tuning
Joseph L. Rose for Pipe Inspection
Mem. ASME
Engineering Science and Mechanics Tremendous interest has surfaced recently on the use of guided waves in pipe inspection
Department, in the oil, chemical, and power generating industries. Relatively long lengths of piping
The Pennsylvania State University, can be inspected for corrosion and cracking from a single probe position. This saves a
University Park, PA 16802 great deal of time and money compared to using more standard point-by-point normal
e-mail:[email protected] beam inspection procedures. Pipes can be inspected without removing insulation or tar
coatings by controlling the guided wave modes and frequencies used to carry out the
Michael J. Avioli, Jr. study. This paper will review the history and state of the art of the guided wave techniques
FBS, Inc., in piping. Benefits and limitations of the various methods will be pointed out along with
State College, PA 16801 a vision of future directions in the area of pipe inspection. DOI: 10.1115/1.1491580

Introduction Although Naviers equation is intractable in this form, it can be

reduced into scalar and vector wave equations by the method of
Interest in the guided wave inspection of pipes has peaked dur-
Helmholtz, 2, where the displacement field is assumed to be a
ing the last decade because of the possibility of inspecting long
combination of the gradient of a scalar potential field, , and the
lengths of pipe from a single probe position. The potential for
curl of a vector potential field, H, with the additional constraint
inspection of pipes under insulation, soil, or tar, those embedded
that H0 9.
in concrete, and those generally hidden from sight is becoming
realized. For general reading and a review on guided waves, see uH (2)
1 6. Textbooks containing fundamental topics in guided wave
analysis can be found in 711. Early work that serves as back- Thus, the resulting wave equations are 3 and 4, where 3
ground in pipe analysis is presented in 1230. Recent published represents the longitudinal wave motion in the structure and 4
works on pipe inspection with guided waves can be found in represents the shear wave motion; c 1 and c 2 are the longitudinal
3155. The current thinking on wave resonance tuning to opti- and shear wave velocities of the medium.
mize defect detection, location, classifications, and sizing poten- 1 2
tial is introduced in 56 59. Some computational techniques to 2 , c 1 2 / (3)
c 21 t 2
assist the inspection process for defect classification and sizing
analysis can be found in 60 63. 1 2H
Material presented in this paper goes beyond earlier works and 2 H , c 2 / (4)
emphasizes the wave resonance tuning approach to defect analysis c 22 t 2
and improved penetration power. The wave tuning exercise is il- If Eqs. 3 and 4 are solved entirely, a theoretical solution will
lustrated with a defect cluster detection experiment in a pipe fol- be developed that represents all modes of propagation in the hol-
lowed by some analytical and experimental tuning of guided low cylinder, including the axially symmetric, nonaxially symmet-
waves in a coal tar coated pipe. Excellent results were obtained. ric and torsional guided wave modes 15,17. However, the axially
symmetric guided wave solution can be isolated by assuming that
only the theta component of the vector potential field is nonzero
Theory and that both the vector and scalar potential fields are only func-
In general, there are two basic methods for constructing multi- tions of coordinates r and z 9. Assuming time harmonic motion
layer waveguide solutions, the transfer matrix method 64,65, in the z coordinate direction, the solution of the scalar and vector
and the global matrix method 66. Although, the global matrix wave equations is 5 and 6.
method tends to be more computationally inefficient, it has the
benefits of simplicity, and an inherent numerical stability that can A L H10 r A L H20 r e i kz t (5)
be difficult to achieve in the transfer matrix method 4,64 67.
H A S H11 r A s H21 r e i kz t e (6)
Thus, the global matrix method is used here to describe the
multilayer, hollow cylinder waveguide. A schematic representa- The quantities involved in the solution are defined as follows:
tion of the layer system is shown in Fig. 1. A ( L ) , A ( L ) , A ( S ) , A ( S ) , are the unknown amplitude con-
The derivation begins with Naviers equation of motion 1, stants for longitudinal and shear waves propagating in the layer in
where u is the displacement field, and are the Lame constants the outward and inward directions, is the angular frequency, k is
for the material, and is the density, where the vector operations the wave number, and and are given by 7. The functions
are performed in the cylindrical coordinate system. Due to the H 1,2
n (z) are the nth order Hankel functions of the first and second
correspondence principle, it is not necessary to distinguish be- kind. Hankel functions are used instead of Bessel functions be-
tween elastic and viscoelastic layers 68. Layers are considered cause they provide better numerical stability when seeking solu-
elastic when the material constants are purely real numbers, while tions in the range of phase velocities where individual wave com-
layers are visco-elastic when the material constants are complex. ponents become nonpropagating across the layers.
2u 2 2 /c 21 k 2 , 2 2 /c 22 k 2 (7)
2 u u (1)
t2
In order to construct the boundary conditions for the problem, it
is necessary to make the layer matrix specific to a particular layer
Contributed by the Pressure Vessels and Piping Division and presented at the
Pressure Vessels and Piping Conference, Vancouver, Canada, August 4 8 2002, of
and interface. Inserting the appropriate elastic or viscoelastic ma-
THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS. Manuscript received by terial properties into the layer matrix, and inserting the radius of
the PVP Division, April 27, 2002. Associate Editor: S. Y. Zamrik. the boundary or interface that is being considered accomplishes

Journal of Pressure Vessel Technology Copyright 2002 by ASME AUGUST 2002, Vol. 124 303
 used to prescribe the boundary conditions on the free surfaces,
where the displacement components of the layer matrix must be
omitted. The specific boundary conditions for the problem are the
vanishing of normal and shear stresses and the free surfaces of the
problem, and the continuity of stress and displacement at the in-
terfaces between layers. Using the matrix formalism of 8, each
boundary condition for the problem can be prescribed. The total
set of boundary conditions can then be constructed into a global
matrix equation. For instance, the global matrix equation for a
four-layer system is 9. This matrix contains the layer matrices
for each layer, evaluated at the interfaces where a boundary con-
dition is prescribed. The subvectors A contain the unknown am-
plitude constants for the layers. The next step towards solving the
boundary value problem for the multi-layer hollow cylinder is to
generate the equations for the displacement and stress fields in the
layer. The displacement field is generated by substituting the sca-
lar and vector potential solutions into Eq. 2. The stress field is
Fig. 1 Schematic representation of a multilayer hollow derived by consecutively applying the strain-displacement and
cylinder stress-strain constitutive equations to the displacement field equa-
tions. The displacement and stresses are expressed in a matrix
format as in 8, where is called the layer matrix.

this. The notation that is used here is to use subscripts to indicate uz A L
the layer, and radius of the layer is being evaluated, such that
ur A L
12 is the layer matrix evaluated for the first layer at its outside e i kz t (8)
radius as defined by Fig. 1. An additional matrix, , is defined as rr A S
the lower 24 submatrix of the layer matrix, . This matrix is rz A S

kH10 r kH20 r H10 r H20 r
H11 r H21 r kH11 r kH21 r

k 2 2 H10 r 2
1
H r
r 1 k 2 2 H20 r 2
2
H r
r 1 2k H10 r 2
k 1
H r
r 1 2k H20 r 2
k 2
H r
r 1

2 k H11 r 2 k H21 r k
2 2
H11 r k
2 2
H21 r

11 0 0 0
12 22 0 0 A 1
A 2
0 23 33 0 e i kz t 0 (9)
A 3
0 0 34 44 A 4
0 0 0 45

To obtain a nontrivial solution for the problem, the determinant tines such as Mullers method that rely on the slope of the func-
of the global matrix must equal zero, resulting in a dispersion tion, tend to extrapolate wildly, finding some roots, but not being
equation for the layer system. Aside from the material constants able to find all of the roots for a particular frequency of interest.
for each layer, there exists a single independent variable of fre- An alternative to traditional root finding routines, is to find the
quency, and a single dependent variable of wave number. A plot of local minima of the absolute value of the dispersion equation 4.
the wave number roots versus, frequency yields the dispersion Although the method is computationally inefficient, it tends to
curve for the structure. reliably find all of the roots of the dispersion equation.
The roots of the dispersion equation can be real, imaginary, or The real and imaginary parts of the complex wave number roots
complex 69. For elastic waveguides, the roots will be either real of the dispersion equation can be converted into phase velocity
or imaginary, where the real roots describe a wave mode that is and attenuation values from Eqs. 1012. This alternative rep-
propagating with no attenuation, and the imaginary roots describe resentation is important in that the phase velocity values can be
non-propagating wave modes. Complex roots exist when material directly tied to the means of generating the guided wave, and the
attenuation is included in the model. These roots describe a propa- attenuation values are directly useful for choosing modes to pen-
gating wave mode that is attenuating with distance from the etrate a coated structure. The plots of phase velocity versus fre-
source 69. This is the type of result that is of interest for vis- quency are termed phase velocity dispersion curves, while the
coelastic or combination elastic/viscoelastic waveguides. plots of attenuation versus frequency are attenuation dispersion
curves.
Numerical Results
c ph (10)
Re k
The complex roots of the dispersion equation can be quite dif-
ficult to find, due to the fact that the absolute value of the equation Im k (11)
changes over many orders of magnitude between roots. Thus, rou-
Attenuation dB/m 20 log10 e 1000 (12)

304 Vol. 124, AUGUST 2002 Transactions of the ASME

 Table 1 Elastic and viscoelastic material properties

In order to calculate the dispersion equation for a particular

system of layers, it is necessary to obtain the material constants
for each of the materials involved. The material constants are the
longitudinal and shear wave velocities, and the material density.
For elastic layers, the wave velocities are purely real quantities, Fig. 3 Schematic of experimental arrangement showing hole
while for viscoelastic layers, the wave velocities have an imagi- and cluster locations relative to the ensemble exit point of the
nary component. The complex wave velocities for viscoelastic array
materials are seldom published so it is necessary to measure the
quantities. This can be done by measuring velocity and attenua-
tion of longitudinal and shear waves in the material 70. The
material properties for the elastic steel and viscoelastic bitumen peak amplitude response for each frequency was measured and
coating are shown in Table 1. The phase velocity and attenuation plotted as a function of frequency to identify any preferred fre-
dispersion curves for a 4-in. schedule 40 steel pipe, coated with quency ies. To perform the experiment, a Matec Explorer II1
0.006 in. of bitumen are shown in Fig. 2. tone-burst pulser/receiver was used with a 6-ft long 8-in. schedule
The phase velocity dispersion curve looks quite typical as com- 40 stainless steel pipe.
pared with dispersion curves for a single layer, elastic hollow Figure 3 shows a schematic of the testing arrangement. Figure 4
cylinder. This is likely due to the fact that the viscoelastic coating shows the Matec Explorer II instrument and a view the 15-
is thin in comparison with the elastic layer. The attenuation dis- element array. Each array element has a 500-kHz transducer
persion curve is very interesting in that it shows drastic attenua- mounted on a Plexiglas shoe that is cut to provide a phase velocity
tion changes for the different modes over the frequency range of of 0.116 in/s or 2.95 mm/s and conforms to the curvature of
interest. The two lowest-order modes at low frequency show very the pipe. Figure 5 shows the hole defect and the cluster
good attenuation performance except at the point of the cutoff defect. Figure 6 shows the pipe, seven-hole cluster, and 15-
frequency for the L(0,2) mode. As frequency increases the attenu- element array.
ation of the modes is tending to increase, although there are many Figure 7 shows the RF waveforms obtained from the hole de-
specific points along the modes that have quite low attenuation fect and from the cluster defect. The main bang portion of the
values. These modes have great potential for performing a nonde- signals as well as the pipe end echo have been partially removed
structive test in a coated pipe. from the signals. The excitation frequency was varied from 100 to
675 kHz in 25-kHz increments. The gates indicate the hole posi-
tion 22 in., and the cluster position 46 in. The bars to left and right
Experimental Results on Defect Detection in a Clean
show the peak-to-peak amplitude within the gates selected for the
Pipe hole and for the cluster, respectively. Figures 8, and 9 show the
To illustrate the benefits of frequency tuning, a sample problem bars transposed to a frequency axis slightly magnified for bet-
is given here. A hole defect and seven-hole cluster defect in a
clean pipe were insonified with varying frequency. The peak-to- 1
Explorer II is a trademark of the Matec Corporation, Northbough, MA

Fig. 2 Phase velocity and attenuation dispersion curves for a 4-in. schedule
40 steel pipe, coated with .006 in. of bitumen, showing the first six longitudinal
modes of propagation

Journal of Pressure Vessel Technology AUGUST 2002, Vol. 124 305

 total cross-sectional area of the pipe wall. A bitumen coating was
applied to one of the test samples to an approximate thickness of
0.006 in.
Three modes were chosen from the attenuation dispersion curve
to inspect the test samples: the L(0,2) mode at 190 kHz, the
L(0,3) mode at 630 kHz, and the L(0,3) mode at 740 kHz see the
circled modes on Fig. 2. From the attenuation dispersion curve,
the expected result is that the L(0,2) mode at 190 kHz, and the
L(0,3) mode at 740 kHz would be good choices for performing an
inspection, while the L(0,3) mode at 630 kHz would have unac-
ceptably large attenuation. Figures 1012 show the test results.
In all results, the echoes shown from left to right are as follows:
1 the initial excitation of the transducer; 2 the echo from the
defect; and 3 the echo from the end of the pipe BWE. The
results are arranged in two columns, one for the clean pipe and
one for the coated pipe. The upper results were taken with the
same instrument gain to show a comparison of the actual signal
strength of the defect and backwall for the two samples. For the
lower results, the instrument gain was adjusted to bring the defect
signal to 80% of full screen height. Then the defect signal strength
can be compared with the baseline noise of the system to see if
Fig. 4 Matec tone-burst pulserreceiver shown with test pipe there was enough strength to resolve the defect from the noise.
As expected, the result showing the least amount of attenuation
was the L(0,2) mode at 190 kHz. The gain and signal to noise of
ter viewing. It is clear from these charts that frequency tuning can the defect was quite similar for both the clean and coated pipes.
optimize the response from either the hole or the cluster. For The results for the L(0,3) mode at 630 kHz are generally unac-
example, and for this experiment 450 kHz produced the best am- ceptable from a nondestructive point of view, because the signal
plitude response from the cluster and 125 kHz produced the best strength is almost entirely attenuated. Increasing the frequency of
response from the hole. excitation to 740 kHz, the L(0,3) again becomes usable, showing
some attenuation, but still having excellent signal to noise on the
defect. In fact, even though the frequency of operation has almost
Experimental Results From Defect Detection in a Coal quadrupled from 190 to 740 kHz, the attenuation and signal to
Tar-Coated Pipe noise of the tests are very similar. Additionally, an interesting
In order to apply the knowledge gained from the attenuation observation is that the back-wall is reduced in amplitude from the
dispersion curves, two test samples were prepared. Both samples uncoated case. This is most likely due to mode conversion effects
consisted of a 10.5-ft length of 4-in. schedule 40 steel pipe with a and not due to the presence of the coating. This result demon-
transverse saw cut. The depth of the saw cut was made to 0.150 strates the importance of having knowledge of the attenuation
in. that produces a notch area that is equivalent to 5% of the characteristics of guided waves propagating in a coated structure.

Fig. 5 a Round bottom-hole defect the diameter is 0.375 in. and the depth is 0.25
in.; b seven-hole cluster defect

306 Vol. 124, AUGUST 2002 Transactions of the ASME

 Fig. 6 a End-on view of pipe showing location of seven-hole cluster; b view of 15-element array

Fig. 7 Gated RF waveforms from hole left and seven-hole cluster right; medium gray bars are peak-to-peak values in the
hole gate and the darker bars are peak-to-peak values in the cluster gate on the right and left sides of the RF waveforms,
respectivelya frequency range 100, 375 kHz; b frequency range 400, 675 kHz

Journal of Pressure Vessel Technology AUGUST 2002, Vol. 124 307

 Fig. 8 Amplitude versus frequency distribution of echoes from seven-hole cluster

Fig. 9 Amplitude versus frequency distribution of echoes from lone hole

Fig. 10 Test results for a clean and coated pipe with a transverse notch of 5% cross-sectional area using the L 0,2
mode at 190 kHz for inspection. The upper waveforms show the defect and back wall results for an equivalent
instrument gain to demonstrate the attenuation difference between clean and coated test samples. The lower
waveforms show the result when the instrument gain is adjusted to bring the notch to 80% of full screen height to
demonstrate the difference in signal to noise of the test for clean and coated test samples. Despite the coating,
excellent results are obtained.

308 Vol. 124, AUGUST 2002 Transactions of the ASME

 Fig. 11 Test results for a clean and coated pipe with a transverse notch of 5% cross-sectional area using the
L 0,3 mode at 630 kHz. The upper waveforms show the defect and back wall results for an equivalent instrument
gain to demonstrate attenuation differences between clean and coated test samples. The lower waveforms show
the result when the instrument gain is adjusted to bring the notch to 80% full screen height to show the difference
in signal to noise of the test for clean and coated test samples. Poor results are obtained because of the coated
pipe and improper mode and frequency selection.

Fig. 12 Test results for a clean and coated pipe with a transverse notch of 5% cross-sectional area using the L 0,3
mode at 740 kHz for inspection. The upper waveforms show the defect and back wall results for an equivalent
instrument gain to demonstrate the attenuation difference between clean and coated test samples. The lower wave-
forms show the result when the instrument gain is adjusted to bring the notch to 80% of full screen height to
demonstrate the difference in signal to noise of the test for clean and coated test samples. Despite the coating,
excellent results are obtained.

It is now quite obvious that there are many low attenuation modes to produce a good result. Quite often higher frequencies can have
to choose beyond the L(0,2) mode at low frequency. much better penetration power than lower frequencies.

Concluding Remarks
References
Great potential exists for the guided wave inspection of piping
1 Chimenti, D. E., Nayfeh, A. H., and Butler, D. L., 1982, Leaky Waves on a
over long distances from a single probe position. Phase velocity Layered Half-Space, J. Appl. Phys., 53, pp. 170176.
and frequency wave resonance tuning, however, are often essen- 2 Ditri, J., Rose, J. L., and Chen, G., 1991, Mode Selection Guidelines for
tial. Results for a set fixed phase velocity and frequency are not Defect Detection Optimization Using Lamb Waves, Proc., 18th Annual Re-
dependable. As a result of tuning, it becomes possible to improve view of Progress in Quantitative NDE Meeting, Plenum, Vol. 11, pp. 2109
2115, Brunswick, ME.
penetration power despite coatings on the pipe and also to be able 3 Fortunko, C. M., King, R. B., and Tan, M., 1982, Nondestructive Evaluation
to detect difficult defect shapes. Theoretical and experimental re- of Planar Defects in Plates Using Low-Frequency Shear Horizontal Waves, J.
sults show that many choices of mode and frequency can be made Appl. Phys., 53, pp. 34503458.

Journal of Pressure Vessel Technology AUGUST 2002, Vol. 124 309

 4 Lowe, M. J. S., 1995, Matrix Techniques for Modeling Ultrasonic Waves in Hollow Cylinders by Wedge and Comb Type Transducers, Review of
Multilayered Media, IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 42, pp. Progress in Quantitative Nondestructive Evaluation, eds., D. O. Thompson
525542. and D. Chimenti, Rev. Prog. Quant. Nondestr. Eval., 12A, pp. 211218.
5 Rokhlin, S., 1979, Interaction of Lamb Waves with Elongated Delaminations 38 Guo, D., and Kundu, T., 2000, A New Sensor for Pipe Inspection by Lamb
in Thin Sheets, Int. Adv. Nondestr. Test., 6, pp. 263285. Waves, Mater. Eval., 58, pp. 991994.
6 Rose, J. L., 2002, Standing on the Shoulders of Giants-An Example of 39 Kley, M., Valle, C., Jacobs, L. J., Qu, J., and Jarzynski, J., 1999, Develop-
Guided Wave Inspection, Mehl Honor Lecture, Mater. Eval., 60, pp. 5359. ment of Dispersion Curves for Two Layered Cylinders Using Laser Ultrason-
7 Achenbach, J. D., 1984, Wave Propagation in Elastic Solids. North-Holland ics, J. Acoust. Soc. Am., 106, pp. 582588.
Publishing Co., New York. 40 Kwun, H., Dynes, H., 1998, Long Range Guided Wave Inspection of Pipe
8 Auld, B. A., 1990, Acoustic Fields and Waves in Solids, Vol. 1 and 2, Second Using the Magnetostrictive Sensor Technology-Feasibility of Defect Charac-
Edition.: Kreiger Publishing Co., FL. terization, International Society for Optical Engineering SPIE on Nonde-
9 Graff, K. F., 1991, Wave Motion in Elastic Solids, Dover Publications Inc., structive Evaluation of Utilities and Pipelines II, 3398, pp. 28 34.
New York, NY. 41 Li, J., and Rose, J. L., 2001a, Excitation and Propagation of Non-
10 Rose, J. L., 1999, Ultrasonic Waves in Solid Media, Cambridge University Axisymmetric Guided Waves in a Hollow Cylinder, J. Acoust. Soc. Am., 109,
Press, New York, NY. pp. 457 464.
11 Viktorov, I. A., 1967, Rayleigh and Lamb WavesPhysical Theory and Appli- 42 Liu, G., and Qu, J., 1998, Transient Wave Propagation in a Circular Annulus
cations, Plenum Press, New York, NY. Subjected to Impulse Excitation on its Outer Surface, J. Acoust. Soc. Am.,
12 Armenkaas, Gazis, D. C., Herrmann, G., 1969, Free Vibrations of Circular 104, pp. 12101220.
Cylindrical Shells, Pergamon Press, New York, NY. 43 Lowe, M. J. S., Alleyne, D. N., and Cawley, P., 1998, The Mode Conversion
13 Cooper, R. M., and Naghdi, P. M., 1957, Propagation of Nonaxially Sym- of a Guided Wave by a Part-Circumferential Notch in a Pipe, ASME J. Appl.
metric Waves in Elastic Cylindrical Shells, J. Acoust. Soc. Am., 29, pp. Mech., 65, pp. 649 656.
13651373. 44 Liu, G., and Qu, J., 1998, Guided Circumferential Waves in a Circular An-
14 Fitch, A. H., 1963, Observation of Elastic Pulse Propagation in Axially Sym- nulus, ASME J. Appl. Mech., 65, pp. 424 430.
metric and Nonaxially Symmetric Longitudinal Modes of Hollow Cylinders, 45 Mudge, P. J., 2001, Field Application of the Teletest Long-Range Ultrasonic
J. Acoust. Soc. Am., 35, pp. 706 707. Testing Technique, Insight, 43, pp. 74 77.
15 Gazis, D. C., 1959, Three-Dimensional Investigation of the Propagation of 46 Nagy, P. B., Blodgett, M., and Gulis, M., 1994, Weep Hole Inspection by
Waves in Hollow Circular Cylinders, I. Analytical Foundation, J. Acoust. Circumferential Creeping Waves, NDT&E Int. 27, pp. 131142.
Soc. Am., 31, pp. 568 573. 47 Qu, J., Berthelot, Y., and Jacobs, L., 2000, Crack Detection in Thick Annular
16 Barshinger, J. N., Rose, J. L., 2001, Ultrasonic Guided Wave Propagation in Components Using Ultrasonic Guided Waves, Proc. Instrn. Mech. Engrs. Part
Pipes With Viscoelastic Coatings, QNDE. Brunswick, ME, July 29-August 3. C, 214, pp. 11631171.
17 Gazis, D. C., 1959, Three-Dimensional Investigation of the Propagation of 48 Quarry, M., and Rose, J. L., 1999, Multimode Guided Wave Inspection of
Waves in Hollow Circular Cylinders, II. Numerical Results, J. Acoust. Soc. Piping Using Comb Transducers, Mater. Eval., 57, pp. 108990.
Am., 31, pp. 573578. 49 Rose, J. L., and Zhao, X., 2001b, Anomaly Throughwall Depth Measurement
18 Cho, Y., and Rose, J. L., 1996, Guided Waves in a Water Loaded Hollow Potential With Shear Horizontal Guided Waves, Mater. Eval., 59, pp. 1234
Cylinder, J. Nondestructive Testing & Evaluation, 12, pp. 323339. 1238.
19 Ghosh, J., 19231924, Longitudinal Vibrations of a Hollow Cylinder, Bul- 50 Rose, J. L., and Zhao, X., 2001c, Flexural Mode Tuning for Pipe Elbow
Inspection, Mater. Eval., 59, pp. 621 624.
letin of the Calcutta Mathematical Society, Bull. Calcutta Math. Soc., 14, pp.
51 Rose, J. L., Jiao, D., and Spanner, J., Jr., 1996, Ultrasonic Guided Wave NDE
31 40.
for Piping, Mater. Eval., 54, pp. 13101313.
20 Greenspon, J. E., 1961, Vibrations of Thick and Thin Cylindrical Shells
52 Rose, J. L., Rajana, K., and Carr, F., 1994b, Ultrasonic Guided Wave Inspec-
Surrounded by Water, J. Acoust. Soc. Am., 33, pp. 13211328.
tion Concepts for Steam Generator Tubing, Mater. Eval., 52, pp. 307311.
21 Greenspon, J. E., 1959, Vibrations of Thick Cylindrical Shells, J. Acoust.
53 Shin, H. J., Rose, J. L., 1998, Non-Axisymmetric Ultrasonic Guided Waves
Soc. Am., 31, pp. 16821683.
in Pipes, 7th Annual Research Symposium, 1998 ASNT Spring Conference,
22 Herrmann, G., and Mirsky, I., 1956, Three-Dimensional and Shell-Theory
Anaheim, CA, March 2327.
Analysis of Axially Symmetric Motions of Cylinders, ASME J. Appl. Mech.,
54 Valle, C., Niethammer, M., Qu, J., and Jacobs, L. J., 2001, Crack Character-
78, pp. 563568.
ization Using Guided Circumferential Waves, J. Acoust. Soc. Am., 109, pp.
23 Meeker, T. R., and Meitzler, A. H., 1964, Guided Wave Propagation in Elon-
18411847.
gated Cylinders and Plates, Phys. Acoust., 1A, pp. 111167. 55 Valle, C., Qu, J., and Jacobs, L. J., 1999, Guided Circumferential Waves in
24 Mirsky, I., and Herrmann, G., 1976, Axially Symmetric Motions of Thick Layered Cylinders, Int. J. Eng. Sci., 37, pp. 13691387.
Cylindrical Shells, ASME J. Appl. Mech. 80, pp. 97102. 56 Li, J., and Rose, J. L., 2001b, Implementing Guided Wave Mode Control by
25 Mohr, W., and Holler, P., 1976, On Inspection of Thin-Walled Tubes for Use of a Phased Transducer Array, IEEE Trans. Ultrason. Ferroelectr. Freq
Transverse and Longitudinal Flaws by Guided Ultrasonic Waves, IEEE 48, pp. 761768.
Trans. Sonics Ultrason., SU-23, No. 5, pp. 369374. 57 Shin, H. J., Rose, J. L., 1998, Guided Wave Tuning Principles for Defect
26 Pao, Y. H., and Mindlin, R. D., 1960, Dispersion of Flexural Waves in an Detection in Tubing, J. Nondestruct. Eval., 17, pp. 2736.
Elastic Circular Cylinder, ASME J. Appl. Mech., 27, pp. 513520. 58 Wooh, S. C., and Shi, Y., 2001, Synthetic Phase Tuning of Guided Waves,
27 Silk, M. G., and Bainton, K. P., 1979, The propagation in Metal Tubing of IEEE Trans. Ultrason. Ferroelectr. Freq, 48, pp. 209223.
Ultrasonic Wave Modes Equivalent to Lamb Waves, Ultrasonics, 17, pp. 59 Zhu, W., and Rose, J. L., 1999, Lamb Wave Generation and Reception With
1119. Time-Delay Periodic Linear Arrays: A BEM Simulation and Experimental
28 Thompson, R. B., Alers, G. A., and Tennison, M. A., 1972, Applications of Study, IEEE Trans. Ultrason. Ferroelectr. Freq., 46, pp. 654 664.
Direct Electromagnetic Lamb Wave Generation to Gas Pipeline Inspection, 60 Alleyne, D. N., and Cawley, P., 1992, The Interaction of Lamb Waves With
IEEE Ultrasonics Symposium, pp. 9194. Defects, IEEE Trans. Ultrason. Ferroelectr. Freq., 39, pp. 381396.
29 Viktorov, I. A., 1958, Rayleigh-Type Waves on a Cylindrical Surface, Soviet 61 Chang, Z., and Mal, A., 1998, Wave Propagation in a Plate with defects,
Phys.-Acoust., 4, pp. 131136. Review of Progress in Quantitative Nondestructive Evaluation, Rev. Prog.
30 Zemanek, J., Jr., 1972, An Experimental and Theoretical Investigation of Quant. Nondestr. Eval., 17, pp. 121128.
Elastic Wave Propagation in a Cylinder, J. Acoust. Soc. Am., 51, pp. 265 62 Cho, Y., and Rose, J. L., 1996, A Boundary Element Solution for a Mode
283. Conversion Study on the Edge Reflection of Lamb Waves, J. Acoust. Soc.
31 Alleyne, D. N., Pavlakovic, B., Lowe, M. J. S., and Cawley, P., 2001, Rapid Am., 99, pp. 20972109.
Long-Range Inspection of Chemical Plant Pipework Using Guided Waves, 63 Datta, S. K., Al-Nassar, Y., and Shah, A. K., 1991, Lamb Wave Scattering by
Insight, 43, pp. 9396. a Surface Breaking Crack in a Plate, Review of Progress in Quantitative
32 Alleyne, D. N., Lowe, M. J. S., and Cawley, P., 1998, The Reflection of Nondestructive Evaluation, eds., D. O. Thompson and D. Chimenti, Rev. Prog.
Guided Waves From Circumferential Notches in Pipes, ASME J. Appl. Quant. Nondestr. Eval., 10, pp. 97104.
Mech., 65, pp. 635 641. 64 Thomson, W. T., 1950, Transmission of Elastic Waves Through Stratified
33 Alleyne, D. N., and Cawley, P., 1997, Long Range Propagation of Lamb Solid Medium, J. Appl. Phys., 21, pp. 8993.
Wave in Chemical Plant Pipework, Mater. Eval., 45, pp. 504 508. 65 Haskell, N. A., 1953, Dispersion of Surface Waves on Multilayered Media,
34 Alleyne, D. N., and Cawley, P., 1996, The Excitation of Lamb Waves in Bull. Seismol. Soc. Am., 43, pp. 1734.
Pipes Using Dry-Coupled Piezoelectric Transducers, J. Nondestruct. Eval., 66 Knopoff, L., 1964, A Matrix Method for Elastic Wave Problems, Bull. Seis-
15, pp. 1120. mol. Soc. Am., 54, pp. 431 438.
35 Ditri, J. J., 1994, Utilization of Guided Waves for the Characterization of 67 Mal, Yin, and Bar Cohen, 1991, Comps. Engra., Vol. 1, pp. 85101.
Circumferential Cracks in Hollow Cylinders, J. Acoust. Soc. Am., 96, pp. 68 Christensen, R. M., 1981, Theory of Viscoelasticity: An Introduction, Aca-
37693775. demic Press, New York, NY.
36 Ditri, J. J., and Rose, J. L., 1992, Excitation of Guided Elastic Wave Modes 69 Watson, T. H., 1972, A Real Frequency, Complex Wave-Number Analysis of
in Hollow Cylinders by Applied Surface Tractions, J. Appl. Phys., 72, pp. Leaking Modes, Bull. Seismol. Soc. Am., 62, pp. 369384.
25892597. 70 Blanc, R. H., 1993, Transient Wave Propagation Methods for Determining
37 Ditri, J., Rose, J., and Pilarski, A., 1992, Generation of Guided Waves in the Viscoelastic Properties of Solids, ASME J. Appl. Mech., 60, pp. 763768.

310 Vol. 124, AUGUST 2002 Transactions of the ASME