INVESTIGATION OF PHYSICO-MECHANICAL PROPERTIES
AND LOCAL DISTRIBUTION OF METAL MICROSTRUCTURES
USING SCANNING ACOUSTIC MICROSCOPY
Professor Roman Gr. Maev
University of Windsor
Centre for Imaging Research and Advanced Material Characterization,
Windsor, N9B 3P4, Ontario, Canada
ABSTRACT
This presentation deals with the analysis of the potential of a fundamentally new
method – Scanning Acoustic Microscopy – for studying the physical microstructure
and the dynamic processes of failure in advanced materials of the most diverse nature,
including metals and various metal alloys, matrix metals, multi-layer systems etc.
Work in the ultrasonic microvision technology has given rise to a new experimental
means for generating images of the interior of matter with a resolution comparable to
that of light micrographs, with many practical applications.
INTRODUCTION
There are at present several effective methods for the investigation of physicomechanical properties and local distribution of microstructures in various advanced
materials. Most important among these are electron and optical microscopy, a
number of spectral techniques, including traditional ultrasonic nondestructive
materials characterization and evaluation (NDE) methods. The role of NDE is
changing and will continue to change dramatically. It has become increasingly evident
that it is both practical and cost effective to expand the role of NDE and quality
control to include all aspects of production process and to introduce it as much as
earlier in the manufacturing cycle. Today, and even increasingly in the future, using
advanced optical, thermal, ultrasonic, laser-ultrasound, acoustic emission sensors,
vibrational, electro-magnetic, X-ray technique, etc., as well as modern measurement
technique, along with signal/data processing, on-line information on the processing
conditions can be continuously generated. Real-time process monitoring for more
effective and efficient real-time control of various processes and as a result improved
vehicle manufacturing quality control inspection and reliability will now become a
practical reality.
The new materials structures, joints, parts, components made from various materials
demand the innovative applications of modern NDE techniques to monitor and
control as many stages of the production process as possible. Simply put, intelligent
advance manufacturing is impossible without integrating modern nondestructive
evaluation into the production system.
Works in the imaging systems, including optical, infrared, microwave, ultrasonic, etc.
technique, has given rise to a new experimental means for generating images of the
interior of matter with many practical applications.
694
BACKGROUND
Acoustical imaging is well-known as a powerful tool for studying the microstructure
and properties of materials, and it has attracted the efforts of research groups in
different countries. Actually, high-resolution acoustic imaging, or acoustic microscopy
is a relatively new technique. The idea of a microscope using sound rather than light was
first put forward by a Russian scientist (Sokolov, 1934). However, due to the limitations
of existing technology, it was many years before high-resolution acoustic imaging was
actually realized. The first prototype acoustic microscope was built in the USA, (C.
Quate, Stanford University) in 1974. The industrial application of acoustic microscopy is
still a developing field of study.
The activity of most groups in the world is focused on the development and
application of acoustic microscopy to surface and nearest subsurface layers of
specimens. For example, the group of Prof. A. Briggs (Oxford, UK) [1] has proposed
effective methods of characterizing surface and subsurface cracks, plastic
deformations and other surface and subsurface defects in materials. Prof. B. KhuriYakub (Stanford, USA) [1,2] and his group are successfully developing quantitative
methods for characterization of elastic properties at the surfaces of solids. Prof. J.
Kushibiki (Tohoku Univ., Japan) [1] is developing the line-focus-beam method to
study surface elastic anisotropy; and Prof. G. Wade (Santa-Barbara, USA) [1,2] has
worked on signal and image processing from surfaces and subsurface regions. Centre
for Imaging Research and Advanced Materials Characterization (Windsor, Canada)
have developed expertise in the visualization of structures deep inside objects. The
results of our research on transmission acoustic microscopy is well appreciated for the
study and NDE of films and thin specimens New methods was developed for
visualizing bulk microstructures of anisotropic materials and have successfully
applied it to investigations of defect microstructure of advanced crystalline materials,
including carbon materials, high-Tc superconducting crystals, various metals,
including austenite steel, aluminum alloys and others [2].
Our group also has developed various short-pulse acoustic microscopes for bulk
microstructure characterization of materials and currently develop next generation of
3D imaging scanning acoustic microscope [3}. In the Centre have designed and built
a few different prototypes of hand-handle acoustic imaging systems with 2D multilens and 2D matrix array transducers. Based on that knowledge and research
experience, we have studied the microstructure and properties of various advanced
materials as well as have undertaken examination of various welding NDE of casting
aluminum alloys, riveting and resistance spot-weld quality control inspections for the
automotive industry, among others.
ACOUSTIC IMAGING METHODS
The scanning acoustic imaging technique uses ultrasonic pulse for near-surface and
bulk structure examination. A single narrow pulse generated by the transducer travels
towards the sample surface where it is partially reflected back while the reminder
propagates through the specimen interacting with internal structure. The reflected
signal, which is received by the transducer generally, contains a peak corresponding
to the surface reflection succeeded by echoes reflected from the internal structure
These echoes reach a transducer after a time delay that depends on the ultrasound
velocity in a particular area of the sample. The amplitudes of the reflected and
695
transmitted waves are a function of the acoustic properties of the material at each
scanning point. Echo reflection can be caused by different kinds of non-homogeneity
or discontinuity (for example: phase and grain boundaries, voids, inclusions and
cracks) [3–9].
In our research we use two kinds of desktop acoustic microscopes and a hand-held
array transducer. The first one is ELSAM (Ernst Leitz Scanning Acoustic
Microscope) manufactured by LEITZ Ltd. The second ultrasonic system is an ultrashort pulse reflection ultrasonic microscope (SPSAM) designed by the Acoustic
Microscopy Center of the Russian Academy of Sciences (Moscow, Russia).
ELSAM
LEITZ has developed the ELSAM, with a frequency range from 0.1 to 2.0 GHz, six
magnification selections with a maximum resolution of 0.3µm [4]. It can operate both
as an optical reflection microscope and as an acoustic microscope. The scanning of
the acoustic objective is accomplished with the oscillator coil drive. The microscope
produces 512  512 pixel images with a scan area starting from 0.8  0.8 mm down to
40  40 µm.
Its acoustic focusing system uses a plano-concave sapphire (Al2O3) lens with a zinc
oxide piezo-electric transducer and a wide aperture. So-called continuous mode is
used, that is the excitation electric signal is produced by a high frequency generator
and it consists of several periods [10].
The ELSAM acoustic microscope was designed mainly for a high-resolution surface
and subsurface investigation and it does not allow for deep structure imaging.
Optical Unit
Oscilloscope
Acoustical Unit
Adjustable Table
Control Unit
Fig. 1. ELSAM acoustic microscope
Fig. 2. Desktop SPSAM acoustic
microscope
SPSAM
This wide-field, short pulse SAM operates with a frequency range between 25–100
MHz and various lens apertures [3, 5]. The frequency and design of each of the lenses
was based upon the analytical and numerical calculations of the propagation of
focused ultrasonic pulses through aluminum casting samples. Short probe pulses (3–5
oscillation periods) allow for the resolving of fine microstructural details at a depth of
5 mm, with a resolution of 50–150 µm.
The short probe pulses are partially reflected back to the microscope whenever a nonhomogenous object is encountered. The received reflected signal is stored in a
personal computer where special software can recognize and process the reflections
696
from different internal defects and generates visual output. This output will then be
analyzed and interpreted by researchers.
A mechanical scanner is used to provide one and two-dimensional movements of the
acoustic sensor for B-scan and C-scan representation of the specimen's internal
structure. A scanner with a scanning area of 80  120 mm, a maximum mean
scanning velocity of 2 mm/sec and a precision of 0.05 mm was used. A computer
received the collected data and provided material interface detection and recognition
as well as A-scan, B-scan and C-scan imaging.
The pulse method features an adjustable time-delay gate that is used to collect
individual echo-signals. All of the collected signals from a scanning area of interest,
at various depths of penetration, are used to represent the C and B acoustical images
[2]. Therefore, the delay gate is useful to study reflected signals from a particular
region. The depth of penetration of acoustic waves is inversely proportional to the
density, acoustic velocity and attenuation in the material.
With respect to image processing, current methods are based on amplitude and delay
time measurements of the demodulated reflected pulses. The processing of fully
digitized signal waveforms permits one to increase measurement accuracy and to
investigate the frequency dependence of attenuation. However, special methods of
digital processing are required if the return signals are greatly overlapped or
transformed. This occurs when the thickness of an inclusion is about one wavelength
or it is too deep to be detected using a high frequency probe. During experimental
preparations, new algorithms for digital signal processing were developed based on
non-linear parameter estimation. This method permitted us to separate pulse responses
reflected from each inhomogeneity and therefore measure their time delay and
amplitude [1, 3, 5].
2D Matrix Transducer
Matrix and array transducers offer several enhanced capabilities in imaging and beam
forming compared to a single transducer scanning system. The matrix transducer
technology makes fast data acquisition and imaging possible, while using motionless
transducers. Advanced technologies such as micro-machining and IC manufacturing
can produce matrices of thousand elements of less than 0.1 mm in size, creating an
acoustic analog of CCD camera [6]. This size is essential for phased arrays, given the
element size must be around half a wavelength for efficient beam-forming [7].
However, for matrices not using phased principles to build the acoustic beam, such
small elements represent a disadvantage. Matrix transducers having such a high
density can only operate in the pick-up mode due to power dissipation restrictions.
Moreover, acoustic wavelengths at megahertz frequencies are at least 103 times larger
than the light wavelengths, thus, there seems to be no reason to go down to the micron
range resolution [8].
One of the array transducers used in our study has the diameter of 8.0 mm and
contained 52 emitting-receiving broadband elements with the central frequency
around 20 MHz; the matrix period was 1.0 mm, individual elements being of
0.740.74 mm2 square. The elements were switched one at a time using a custom
coaxial relay multiplexer interfaced with the parallel port. Experimental data have
been acquired with a USD-15 (ultrasonic pulser Krautkramer) connected to a TDS520 (Tektronix digitizing oscilloscope), digital data were transferred to the PC via a
GPIB interface card. The sampling rate (TDS-520) used in the experiments was 500
697
MSamples/s. For image smoothing and quality enhancing various 2D interpolation
algorithms were used.
The imaging capability of the system generally depends on the penetration depth of
the transducer. The penetration depth drops with the decreasing size of the element
due to the beam divergence and the increasing electrical mismatch with 50 Ohms
transmitters and receivers. In general the image quality gets worse for smaller and
deeper reflectors. Images obtained on 1.0 and 2.0 mm flat-bottomed holes drilled
inside a 319 aluminum casting sample are given in Figures 3 and 4. The holes were
grouped by four for the depth range from 1.0 to 8.5mm separated by 0.5 mm depth
step within a group. Featured C-scans were obtained to 8.5 mm deep for 2.0 mm
flat-bottom holes, however reflectors of 1.0 mm diameter were only clearly
observed to 4.5 mm deep.
Fig. 3. C-scans of 1 mm flat-bottom
reflectors located in the depth range:
a) 1.0–2.5 mm; b) 3.0–4.5 mm
(0.5 mm step).
Fig. 4. C-scans of 2 mm flat-bottom
reflectors located in the depth range:
a) 1.0–2.5 mm; b) 7.0–8.5 mm (0.5 mm
step).
In Figure 5 example of an imaging application for the matrix transducer portable
device is given. The generated acoustic images represent laser spot weld samples of
0.75 and 1.0 mm thick. Surface burn marks were respectively around 3.0 and 7.0 mm.
(a)
(b)
(c)
(d)
Fig. 5. Acoustic images of laser spot welds generated with a desktop scanning
acoustic microscope and corresponding images obtained with the 52-element array:
(a), (b) 0.75 mm thick steel plate approximately 3.0 mm. weld; (c), (d) 1.0 mm
thick steel plate around 7.0 mm weld.
698
Fig. 6. Acoustic images of copper wires inside the flexible printed circuit
board, which is incorporated into 1.5 mm thick plastic cover.
Matrix transducer is capable to visualize location and orientation of the flexible
printed circuit board (PCB) inside the plastic. In Figure 6, images of 0.2 mm thick and
2 mm wide wires are shown. The separation between the wires in the right image is
1 mm.
HIGH-RESOLUTION ULTRASONIC INSPECTION METHODS FOR JOINT
INSPECTION
The traditional methods of any joints monitoring are combinations of visual
inspection, pry testing and destructive tests, like in the resistance spot welds, with
hammer and chisel. The only effective non-destructive method has been the ultrasonic
pulsed echo technique based on an evaluation of ultrasonic echo patterns in a
specimen. The current ultrasound technique is capable of determining the
approximate parameters of the joint. Disadvantages of that technique and method are
evident: a wide range of ambiguity in treating echograms; a high sensitivity of the
result to the alignment of the transducer perpendicular to the surface; the dependence
of the results on operator experience and skill; the impossibility of building the
method into an on-line process; etc.
The hottest NDT problem today is high-resolution inspection of welding joints. The
problem of inspection for many-layered (2 or more layers) joints formed by spotwelding or diffusion welding is under investigation. Using acoustic imaging systems
with short probe pulses provides a way to visualize small-scale failures of contact and
other defects at different depths. For spot welding acoustic imaging systems makes it
possible to inspect fine details of internal areas of joints in spite of the curved outer
surfaces of welding spots.
To show the potential of high-resolution acoustic imaging technique we used a widefield short-pulse acoustic microscope (see Fig. 2). The method was applied to
evaluation of spot-welding joints of two steels sheets 1-2 mm thick of each. In spite of
the fact that the top surface of specimen was distributed by welding, high quality
acoustic images of the nugget zone were obtained [9]. The C-scan and B-scan images
contain well-shaped defects of joints and the confirm power of the method for NDE
and QC of spot-welded joints. The combination of the C-scan and the B-scan images
in one picture can give a real 3D image of the defect distribution inside the welding
zone, which can be also very useful from a technological point of view.
The current results of this study show that high-resolution acoustic imaging technique
can be successfully applied in spot welding as an effective inspection method to
detect any kind of defects, and can be used for manufacturing welding quality control.
699
Fig. 7. C-scan Images of the interface area (depth penetration 1 mm) from the spot
welding samples (Ultrasound frequency 50 MHz).
Acoustic imaging methods based on acoustic visualization and characterization are
extremely promising also for laser welds and rivet joints quality evaluation (see figure
5). This technique can provide total bulk reconstruction of the joint zone, including
the topography of the top and bottom faces, the structure of the interface between the
sheets, and any expulsions, voids, pores, cracks or other defects in the welding zone.
BULK AND SUBSURFACE STRUCTURE ANALYSIS OF ALUMINUM
CASTING USING ACOUSTIC IMAGING METHODS
The metallurgical structure of aluminum castings (grain structure, inclusions, porosity
etc.) is usually inspected with optical and scanning electron microscopy. [10, 11].
These both methods provide information only from the surface of a sample while
requiring as-polished or etched metallographic specimens [12]. In a bulk structure
study, as a rule, a series of cross sections is required [112]. Through the image
processing some important quantitative characteristics of the casting structure may be
provided [14].
The main point of this study was to apply the method of acoustic microscopy for the
subsurface and non-destructive bulk structure visualization as well as casting
quantitative characterization.
The service properties of Al-Si-Cu castings depends on their chemical composition,
liquid metal treatment and solidification rate. In general, a small, equiaxed grain
structure, fine constituents and absence of structural discontinuities produces optimal
properties. Consequently, the casting responds better to heat treatment processes.
Porosity in aluminum casting can result from excessive amounts of hydrogen (gas
porosity) or a lack of feeding (shrinkage porosity) [15, 16]. Gas porosity takes the
form of rounded cavities while shrinkage porosity appears as elongated interdendritic
cavities. Commercial castings can contain either type or both in combination. The
service performance of automotive components, especially during high cycle fatigue,
depends on the morphological characteristics of the porosity. These characteristics
can be determined using traditional destructive microscopy techniques, however these
techniques are costly and time consuming. They do not allow process engineers to
make proactive interventions. Consequently, it is essential that quick, non-destructive
techniques for the evaluation of internal casting structure be developed that can be
utilized in an industrial environment.
Some test samples were cut out from different parts of automobile engines. The
samples employed in ELSAM study were prepared using traditional optical
microscopy techniques. Others had no such ideal surface and were prepared with
700
different surface finishes ranging from rough machining to samples with a mirror-like
surface. A few samples were studied with the as-cast rough surface.
SPSAM Results
The ability to locate the exact position of an object using the acoustic microscopy method
is based on the study of the reflections of acoustic waves from the object and the
discontinuities inside it [17].
Fig. 8 demonstrates the ability of the acoustic microscopy method to reveal the
location of flaws in the casting structure. The figure depicts a collection of acoustic
images showing the volume distribution of shrinkage porosity and non-visible cracks
inside the sample. It consists of an acoustic image of the machined surface and three
other images of the subsurface structure. All of the pictures show high-contrast
images of imperfections (the dark spots) in the bulk area of the sample. In Fig. 8a, the
C-scan image shows only the machined surface information because the electronic
gate is set at the surface of the examining sample. With adjustment of the electronic
gate, the volume information can be represented as a set of images with smaller
volume information. Fig. 8b to 8d represent volume information from three different
layers: between 0.1 to 0.5 mm, 0.5 to 1.0 mm, and 1.0 to 1.5 mm respectively. The
acoustic microscope's ability to locate flaws is demonstrated by examining those
images with smaller volume information.
(a) Surface
(d) Depth range 1.0 to 1.5 mm
C-scan
1 mm
B-scan
(c) Depth range 0.5 to 1.0 mm
1 mm
(b) Depth range 0.1 to 0.5 mm
Fig. 8. Acoustic B- and C-scan images of an aluminum casting sample
Each figure, 8b to 8d, shows a couple of results: a C-scan image in which the
researcher chooses a particular layer of aluminum casting specimen and a B-scan
image taken at the position of flaw where the dark line indicates a corresponding Cscan image. Since the velocity of acoustic wave propagating in the Al-casting is
known, by examining the time delay, the location of the flaws can be calculated (in
the case of the Fig. 8b: they are about 0.3 mm below the surface). Similar procedures
for two different gate positions are depicted in Fig. 8c and 8d. Fig. 8c shows that other
flaws are located at about 1.0 mm below the surface; Fig. 8c indicates that yet another
flaw is located at about 1.4 mm beneath the surface. Therefore, taken together, the C
and B scan images provide information about the exact location of the detected flaws.
ELSAM RESULTS
The resolution of the acoustic microscope in high frequency renders comparable
results to those shown by the light microscope. Furthermore, acoustic microscopy is
701
capable of penetrating optically opaque material. The difference in quantitative results
may be attributed to the difference in depths of images. Fig. 9 shows microstructure of
a Al-casting alloy with shrinkage porosity and the Cu-rich phase. The ability to focus
on a specified depth range (10 µm in this case) allows for better contrast images of
Cu-rich phase in acoustic image.
Fig. 10 illustrates the eutectic Si structure of the Al-casting alloy. Fig. 10a is obtained
using the light microscope; Fig. 10b shows acoustical image of higher magnification,
with area corresponding to the region marked as a box on Fig. 10a. The transducer was
focused 20 µm deep under the surface that explains why surface scratches, easily
noticeable in optical image, are almost invisible in acoustical image.
100
(a)
(b)
Fig. 9. (a) Light optical micrograph and (b) acoustic micrograph (1.4 GHz lens) of
Al-casting alloy polished surface, showing fragments of the shrinkage porosity and
Cu-rich phase.
(a)
(b)
Fig. 10. Microstructure of Al-casting alloy, (a) Light optical micrograph, showing
relatively well-modified eutectic Si structure, and (b) acoustic micrograph (1.4
GHz lens), showing details of the microstructure from (a).
CONCLUSION
Main areas to which acoustic imaging may be applied are finished product inspection
and the detection of inclusions and discontinuities. Such technique has the potential to
provide reliable, rapid and cost effective methods to visualize high contrast smallscale failures and defects at different depths within inspected parts. Provided this
technology can adapt to high volume manufacture, it has considerable promise for
application to the advance manufacturing quality control inspection.
702
REFERENCES
1. A. Briggs: Acoustic microscopy. Claredon Press, Oxford, 1992.
2. R. Gr. Maev: Scanning acoustic microscopy of polymeric and biological substances.
Tutorial Archives of Acoustics, 1988, 13: 13–43.
3. R. G.r Maev: Einsatz der Akustomicroskopie in den Materialwissenschaften. Review of
the BRD-USSR bilateral seminar “Microscopy in Material Sciences”, Moscow, 1988, 3542.
4. M. Hoppe: Design and Performance of the Leitz ELSAM High Resolution Acoustic
Microscope. Proc. The First Joint Soviet–West Germany International Symposium On
Microscope Photometry and Acoustic Microscopy in Science, Moscow, (Ed. M. Hoppe,
R. Maev), USSR, September 16–21, 1985.
5. K.I. Maslov: Acoustic Scanning Microscope for Investigation of Subsurface Defects.
Acoustical Imaging, V.20, Plenum Press, NY, 274-281, 1993.
6. K. Erikson, A. Hairston, A. Nicoli, J. Stockwell, and T. White, A 128128 ultrasonic
transducer array, in: Proc. IEEE Ultrasonics Symp., Vol. 2, IEEE, NY, 1625, 1997.
7. D. H. Turnbull, Two-Dimensional Transducer Arrays for Medical Ultrasonic Imaging,
Ph.D. Thesis, University of Toronto 1992.
8. R. Gr. Maev, A. Ptchelintsev and A. Denissov, Acoustical imaging using matrix of
piezoelectric transducers for nondestructive evaluation of plastic composites, in: Proc.
ICCE, 665-667, ICCE Orlando, 1999.
9. R.Gr. Maev, D.F. Watt, R. Pan, et all, Development of high resolution ultrasonic
inspection methods for welding microdefectoscopy, Acoustical Imaging, V.22, Plenum
Press, NY, 779-784,1996.
10. Optical Microscopy, G. F. Vander Voort, eds., ASM International, V. 9 1992.
11. Scanning Electron Microscopy, H. E. Exner, eds., ASM International, V. 9 1992.
12. V. A. Phillips: Modern Metallographic Techniques and Their Appplications. Interscience,
1971.
12. R. C. Gifkins: Optical Microscopy of Metals. Elsevier Publishing, NY, 1970.
14. R. J. Lee, W. A. Spitzig, J. F. Kelly, R. M. Fisher: Quantitative Metallography by
Computer-Controlled Scanning Electron Microscopy, Pract. Metallogr., 21: 27–41 1984.
15. Aluminum and Aluminum Alloys, J. Davis, eds., ASM International, 1993.
16. G. Powell, S-H, Cheng, C. Mobley: A Fractography Atlas of Casting Alloys. Battelle Press,
1992.
17. J. H. Sokolowski, R. G. Maev, H. T. Lee, E. Y. Maeva, and K. I. Maslov: Analysis of
Casting Subsurface Structure Using Acoustic Microscopy, Proc. QNDE, Snowbird, Utah,
July 19–24, 1998.
703