HAL Id: jpa-00230640
https://hal.archives-ouvertes.fr/jpa-00230640
Submitted on 1 Jan 1990
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
GUIDED INTERFACE WAVES AND THEIR ROLES IN ACOUSTIC MICROSCOPY
P. Nagy, L. Adler
To cite this version:
P. Nagy, L. Adler. GUIDED INTERFACE WAVES AND THEIR ROLES IN ACOUS- TIC MICROSCOPY. Journal de Physique Colloques, 1990, 51 (C2), pp.C2-1273-C2-1276.
�10.1051/jphyscol:19902299�. �jpa-00230640�
COLLOQUE DE PHYSIQUE
Colloque C2, supplement au n02, Tome 51, Fevrier 1990 l e r Congres Prancais d ' A C 0 u S t i q ~ e 1990
GUIDED INTERFACE WAVES
ANDTEEIR ROLES IN ACOUSTIC MICROSCOPY P.B. NAGY and L. ADLER
Department o f Welding Engineering, The Ohio S t a t e U n i v e r s i t y , Columbus, Ohio 43210,
U . S . A .Rdsumd
-
L' objectif premier des ondes guiddes par interface dans la microscopie acoustique est d' amdliorer le contraste, c'est-;-dire augmenter la sensitibilitd de la reponse correspondent aux propfi6t.d~ du matdrial et d' adhdsion 2 1' dtude. Deux nouvelles techniques basdes sur les propri6t6s uniques des ondes guiddes sont propos6es pour maximizer la detection des imperfections situees aux limites pour les casd'iriterfaces ayant une orientation normale ou transverse.
Abstract
-
The primary role of guided interface waves in acoustic microscopy is to enhance the contrast, i.e. to increase the sensitivity to material and bond properties.Two novel techniques based on the unique features of guided waves are suggested to maximize the detectability of boundary imperfections at interfaces of normal and transverse orientations.
1
-
INTRODUCTIONThe most unique feature that distinguishes acoustic microscopy from all other types of micro- scopy is the origin of the contrast in the mechanical properties of the specimen. Scanning acoustic microscopes with water couplant have been used with great success in the reflection mode of operation since the late seventies. The excellent contrast of these devices is some- what surprising since the reflection coefficient of most solids in water is not a particularly sensitive measure of material properties. For instance, the reflection coefficient from such different materials a$ copper, steel, and nickel are 93.2%, 93.5%, and 94.1%, respectively, i.e. almost the same. Therefore, when the transducer is focused at the surface of the specimen and specular reflection only is generated, the acoustic contrast is very weak and even very different materials are difficult to distinguish. On the other hand, when the transducer is focused below the surface, the contrast becomes much stronger due to an interferometric effect, providing that the aperture angle is sufficiently high to generate Rayleigh-type surface waves on the specimen.
One of the most important fields of application for acoustic microscopy is the inspection and evaluation of layered structures. Figure 1 shows the geometrical configurations for normal and parallel interface inspection. In normal inspection (e.g. grain boundary studies), the acoustic contrast depends on the reflection and transmission coefficients of the interface while the surface wave velocity is not affected. In parallel inspection (e.g. coatings), the
surface wave velocity is affected by both the elastic properties and thickness of the layer and by the interface quality between the layer and the substrate. Via the above mentioned interferometric effect, changes in the surface wave velocity result in strong amplitude variations of the detected ultrasonic signal thereby producing acoustic images of the layer and interface properties. In the following, we shall introduce two special techniques taking advantage of the presence of guided surface waves to maximize the sensitivity of the acoustic microscope to interface properties.
Acoustic Lens
Tarnpie qp Lens
,,,,,,,-,-Interface ,- Sample Interface
Fig. 1
-
Two types of interface inspection with the acoustic microscope.Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19902299
C2-1274 COLLOQUE DE PHYSIQUE
2
-
NORMAL INTERFACEFigure 2 shows the origin of grain contrast in the anisotropic elastic properties of polycrys- talline materials. At a given defocusing, different grain orientations result in different shades on the acoustic micrograph. There is a second type of anisotroplc contrast, too, i.e.
the grain boundary effect which is an even more unique feature of acoustic microscopy. As a typical example of the z-dependence of both grain and grain boundary contrasts, Fig. 3 shows the acoustic micrographs of a titanium sample taken at four different defocusing levels. The whole phenomenon of changing contrast can be easily explained by Fig. 2. Let us consider two neighboring grains and the boundary between them. Since the periodicity of the corresponding V1(z) and V2(z) curves is different, their relative contrast is alternatingly positive or negative as we increase the defocusing. There are points where the grain contrast is maximum in either direction, but the grain boundaries are not really visualized except as topological bond lines separating the neighboring domains. On the other hand, there are the intersection points of the V(z) curves, where the grains themselves appear to have the same brightness and the grain boundary between them shows up as a brighter or darker region depending on whether the intersection point is below or above the specular reflection. This is because the grain boundary scatters, and therefore attenuates the surface wave component and the detected inter- ference signal becomes offset towards the specular component. The relative contrast between the boundary region and the neighboring grains is a quantitative measure of the interface scattering and can be used to evaluate the elastic properties of the actual interface.
Fig. 2
-
Different grain and grain boundary contrasts in acoustic microscopy.Fig. 3 - Acoustic microscopic grain and grain boundary contrast at different defocuses in polycrystal titanium, (a) z = -3.2 mm; (b) z = - 3 . 8 mm; (c) z = -4.2 mm; (d) z = -4.8 mm;
frequency: f = 1.6 GHz.
3
-
PARALLEL INTERFACEGenerally a layered structure can sustain an infinite number of guided.modes, so-called generalize? Lamb modes, which are usually strongly dispersive, i.e. their velocity changes with frequency. As for thin layers, when the layer thickness is small with respect to the wavelength of the interrogating ultrasonic wave, there is only one principal mode of practical
interest, the so-called dispersive Rayleigh mode. At very low frequencies the thin layer has but a negligible effect on the surface wave propagation and the principal mode behaves like
the simple Rayleigh mode on the free surface of the substrate. At very high frequencies the substrate has less and less effect on the layer and the principal mode degenerates into the simple Rayleigh mode on the free surface of the layer material. In between, the loading or stiffening effect of the layer on the substrate can be readily measured from the frequency dependent surface wave velocity [l].
In order-to account for bond imperfections at the layer-substrate interface, the so-called finite boundary stiffness model was used [2-41. These boundary conditions can be easily incorporated into existing multi-layer programs based on ideal rigid boundary conditions by introducing an additional interface layer of negligible thickness and density [ 5 1 . Figure 4 shows the calculated dispersion curves of the lowest order generalized Lamb mode (modified Rayleigh mode) of a steel layer on aluminum substrate for different boundary conditions.
These modes are slightly leaky since the structure is assumed to be immersed in water.
In order to approximate the actual boundary conditions corresponding to kissing bonds and stainless steel-aluminum solid-state bonds [6], a given SL/ST = 2.6 extensional-to-transverse boundary stiffness ratio was assumed in the calculation [7]. The actual stiffness constants were chosen to match the wide range of experimental data obtained on inertia friction welded specimens in an earlier study [ S ] . As can be expected, at very low frequencies the interface looks perfect and the dispersion curves approximate the rigid bond case. As the frequency increases, the interface looks more and more loose and the dispersion curves approximate the completely delaminated free plate case. In between, the phase velocity seems to be a sensitive quantitative measure of the effective boundary stiffness, i.e. bond quality.
In order to estimate the resulting contrast produced by such interface imperfections under acoustic microscopy inspection, V(z) curves were calculated for different bond qualities. The reflection coefficient was calculated by a recursjve multi-layered technique which would easily accomodate the additional interface layer representing the finite boundary stiffness of the stainless steel-aluminum bond [ 9 ] . Figure 5 shows the calculated V(z) curves at
fd = 1 MHz mm (frequency-by-layer thickness) for different bond qualities. The fz = 4-14 MHz mm range offers the highest inspection sensitivity. While the actual contrast changes from negative to positive values at different focusing depths, the highest obtainable contrast is dependent on the velocity change caused by a given boundary imperfection with respect to the ideal rigidly bonded layer. Figure 6 shows this acoustic contrast for medium kissing bond as a function of fd. The fairly sharp maximum around fd = 0.4 indicates that there exists an optimal inspection frequency where the interface imperfection produces the strongest defect signal on the acoustic micrograph. This result can be easily understood from the character- istic dispersive properties of the modified Rayleigh mode shown in Fig. 4. On the other hand, it shows that acoustic microscopy has a quite unique contrast mechanism since conventional bulk (either shear or longitudinal) inspection always offers higher sensitivity at higher
frequencies. 3
7
r i s i d bondI
weak k i s s i n g bond
f r e e plate
I
Fig. 4 - Calculated dispersion curves of the modified Rayleigh mode on an aluminum substrate covered by a steel layer for different interface qualities.
COLLOQUE
DE
PHYSIQUEweak k i l f i n l bond w d i u k i l l i n g bond strong klasing bond
(a) (b )
Fig. 5
-
Calculated V(z) curves at fd = 1 for different hond qualities.Fig. 6
-
Acoustic contrast for medium kissing bond as a function of fd.4
-
CONCLUSIONSGuided interface waves propagating al0fig the fluid-loaded surface of solid specimens were shown to play an important role in the generation of image contrast in acoustic microscopy.
Different techniques were shown to increase this contrast by optimizing the effect of normal and parallel material interfaces on the ~ ( z ) curves.
ACKNOWLEDGEMENT
This work was supported by the U.S. Department of Energy, Basic Energy Grant No.
DE-FG02-84ER45057.AO005.
REFERENCES
/l/ Farnell, G.W. and Adler, E.L., Physical Acoustics (Academic Press, New York, 1972), Vol. IX, pp. 35-127.
/2/ Jones, J.P. and Whittier, J.S., J. Appl. Mech.
24,
905 (1967)./3/ Tattersall, H.G., J. Phys. 6, 819 (1973).
141 Schoenberg, J. Acoust. SOC.-Am.
68,
151.6 (1980)./ 5 / Pilarski, A. and Rose, J.L., J. Appl. Phys.
63,
300 (1988)./6/ Nagy, P.B. and Adler, L., J. Nondestr. Eval.
1,
199 (1988)./7/ Haines, N.F., The theory of sound transmission and reflection at contacting surfaces, Report RD/B/N4744, Central Electricity Generating Board, Berkeley Nuclear Lahs (1980).
/8/ Adler, L., deBilly, M., Quentin, G., Talmant, M., and Nagy, P.B., in Review of Progress in Quantitative NDE (Plenum Press, New York, 1989), Vol. 8B, pp. 1973-1980.
191 Brekhovskikh, L. H., Waves in Layered Media (Academic Press, New York, 19801, pp. 23-61.
