 |
Description  |
|
|
TECHNICAL FIELD
The invention is concerned with the evaluation of physical objects and,
more specifically, with nondestructive testing and evaluation.
BACKGROUND OF THE INVENTION
As reviewed, e.g., by J. F. Hinsley, Non-destructive Testing, Macdonald &
Evans, 1959, the field of nondestructive testing and evaluation includes a
variety of methods as based on physical effects such as, e.g.,
radiological, acoustic, and magnetic interactions with test objects. More
specifically, and as particularly germane to the invention, acoustic or
ultrasonic methods involve the monitoring of an elastic wave as influenced
by flaws or inhomogeneities in a test object; see, e.g., B. Banks,
Ultrasonic Flaw Detection in Metals, Prentice-Hall, 1962.
For testing purposes, an elastic wave may be generated by means of an
electro-acoustic transducer based on electromagnetic, electrostatic,
magnetostrictive, or piezo-electric effects; also, waves have been
generated optically as described, e.g., by
C. A. Calder et al., "Noncontact Material Testing Using Laser Energy
Deposition and Interferometry", Materials Evaluation, Vol. 38 (1980), No.
1, pp. 86-91 (where, also, monitoring by interferometry is described) and
by
C. P. Burger et al., "Laser Excitation Through Fiber Optics in NDE",
Journal of Nondestructive Evaluation, Vol. 7 (1987), pp. 57-64, the latter
disclosing laser energy as transmitted to a surface of interest via a
flexible optical-fiber element.
Monitoring by interferometry involves the use of two coherent beams of
light: an object beam which passes through--or is reflected by--an object
under observation, and a reference beam which is unaffected by the object.
Superposition of the two beams results in interference and, in a
wide-aperture (full-field) system, the resulting intensity distribution
yields an interferometric fringe pattern representing a contour map of
constant optical path or optical phase difference.
While, typically in a laboratory setting, interferometric monitoring can be
carried out with line-of-sight radiation, industrial settings may require
testing in confined spaces and at hidden surfaces. Accordingly, it is a
purpose of the invention to provide for a test device and method for
detecting and characterizing flaws, such device and method being
particularly suited with respect to ease of access to test objects in
commercial practice.
SUMMARY OF THE INVENTION
In the manufacture and maintenance of industrial articles, the invention
provides for the evaluation of articles and components with respect to
inhomogeneity of a surface property such as, e.g., composition or
continuity. In accordance with the invention, energy is applied to a
surface of interest (piezoelectrically or optically, for example), and a
surface wave (Rayleigh wave) is sensed by optical-fiber interferometry as
characterized by the use of at least one optical fiber to transmit a probe
signal to a surface point as well as for transmitting returning reference
and objects signals. (An optical fiber is understood as comprising a
waveguiding core-cladding structure for guiding one or several modes of
radiation having a wavelength corresponding to ultraviolet, visible, or
infrared radiation.)
Use of optical-fiber interferometry as preferred in accordance with the
invention facilitates testing in assembly-line manufacture as well as in
the preventive maintenance of critical components and assemblies such as,
e.g., jet engines and rocket motors. The method is particularly
advantageous for the inspection of otherwise hard to reach surface
locations, and such method is advantageous also in that laser radiation
used for probing remains fiber-guided over most of the optical path,
thereby enhancing industrial safety especially when fiber ends are
shielded.
BRIEF DESCRIPTION OF THE DRAWING
FIG. 1 is a schematic representation of a first device representing a first
preferred embodiment of the invention;
FIG. 2 is a schematic representation of a second device representing a
second preferred embodiment of the invention; and
FIG. 3 is a schematic representation of a third device representing a third
preferred embodiment of the invention.
DETAILED DESCRIPTION
FIG. 1 shows laser 1, lens 3, and single-mode optical fiber 4 in optical
alignment with a first beam of variable beam splitter 2. Optically aligned
with a second beam are multi-mode optical fiber 5 and photodetector 6 with
wide-band amplification electronics electrically connected to
pulser-receiver 7. The latter is connected, via stepless gate 8, to
spectrum analyzer 9 and digital oscilloscope 11 which, in turn, are
connected to respective plotters 10 and 12. Pulser-receiver 7 is further
connected to piezoelectric transducer 13 which is physically attached to
an object 14 to be tested, and the tip of fiber 4 is in proximity of
object 14 for interferometric detection of surface waves such as, e.g.,
waves influenced by an inhomogeneity 15.
Device operation involves splitting of the unspread output from the laser
so as to produce two beams having essentially equal intensity. One of the
beams is wasted, the other is focused into the single-mode optical fiber.
A small percentage of light traveling down the fiber suffers Fresnel or
internal reflection from the output end and serves as the reference wave.
The remainder of the light emerges from the fiber and is scattered by the
diffusely reflecting surface of the test object. A small, detectable
fraction of the scattered light is reflected back into the optical fiber
as return object wave to join and interfere with the internally reflected
reference wave in its propagation back through the fiber. Upon exiting
from the optical fiber, the returning light is collimated by the focusing
lens, and deflected by the beam splitter so that approximately half of the
returning light is passed into the multimode fiber. Typically, such a
fiber has a core diameter on the order of 50 micrometers as is
advantageous in the interest of low-loss coupling; with low-intensity
signals, an additional lens may be used for light insertion into the
multimode fiber.
Since the optical-fiber interferometer is excited by coherent light, there
will be destructive interference whenever the optical path length of the
externally reflected and recaptured light beam differs from the optical
path length of the internally reflected reference beam by (an odd multiple
of) a half wavelength. This difference is due to travel of externally
reflected light back and forth across the gap or Fizeau cavity between the
fiber tip and test surface. On the other hand, if this difference is (a
multiple of) a whole wavelength, interference will be constructive. Thus,
if the intensities and polarizations of the two reflections are
approximately equal, the light traveling back along the fiber will be seen
to brighten and darken as the reflecting test surface moves towards or
away from the optical fiber tip. A displacement-induced change in cavity
length of just one quarter of an optical wavelength is sufficient to shift
the output signal intensity from a maximum to a minimum or vice versa.
One important feature of an optical-fiber interferometer is intrinsic
self-alignment of reference and object beams; furthermore, throughout most
of the interferometer both beams experience the same environment of
temperature, pressure, vibration, etc. Alignment is critical, however,
with respect to coupling of light into the fiber and separating out the
returning signal. For the sake of maximizing contrast in the returned
signal, a translation stage may be used to move the fiber tip in and out
from the specimen surface. Also, for the sake of enhancing the intensity
of the externally reflected signal e.g. in the case of a low-reflecting
test surface, a compact objective or rod lens may be added to the fiber
tip to collimate the light. Acceptable results were obtained without such
additional lens on a moderately reflective machined steel surface, with
fiber tip standoff distance of up to 1 cm.
A test arrangement as schematically shown in FIG. 1 was used to examine, as
a test object, a steel bar measuring 25.4 by 25.4 by 254 mm, with an
oblique slot machined to a depth of 2 mm from the surface and having a
length of 2.8 mm. When the machined "flaw" was located between the
piezoelectric transducer and the optical fiber tip, the delay or time of
flight of the Rayleigh wave observed in the oscilloscope trace was seen to
decrease as the fiber tip was translated towards the flaw. As the fiber
tip moved past the flaw in the direction of the point of excitation (see
FIG. 1), the signal increased markedly in amplitude and, moreover, several
additional signals appeared--ostensibly as associated with components of
the wave that had been reflected from the flaw. Such in-line input and
reflected waves are very difficult, if not impossible, to detect by means
of conventional contact transducers. Indeed, none were observed in
piezoelectric pitch-catch tests as carried out with the steel bar.
The time-domain display of the ungated optical-fiber interferometer signal
transmitted to the oscilloscope through a high-pass (>1 MHz) filter showed
spikes which were readily identified as corresponding to the Rayleigh wave
generated by the piezoelectric transducer and to the partially reflected
wave from the machined surface flaw: since reflected waves are monitored
in-line, their temporal spacing or time of travel can be used to estimate
the location of a flaw. This demonstrates an advantage of optical-fiber
interferometry as compared, e.g., with spectrometry using a piezoelectric
sensor. Since the former does not involve the attachment of an
energy-absorbing mass to the surface at the point of measurement, the
acoustic wave is not altered by the sensor, so that it becomes possible to
monitor input and reflected waves together.
By processing the signals through the stepless gate to the spectrum
analyzer instead of the oscilloscope, frequency-domain information was
obtained. Strongest reflected frequencies were found to lie near 1.5 MHz,
corresponding to a wavelength of slightly less than 2 mm - approximately
the depth of the flaw. Also, significant attenuation was found in the
reflected-wave spectrum near 0.95 MHz, corresponding to a wavelength of
approximately 3 mm--close to the 2.8-mm length of the flaw.
A difficulty which may arise in the use of the arrangement of FIG. 1 lies
with the separation of the return signal from the input-end reflection,
and FIG. 2 shows a modified arrangement which readily ensures such
separation. Specifically, FIG. 2 shows laser 1, objective lens 3, single
mode optical fiber 4, photodetector 6 connected to analyzer 21 (comprising
e.g., components 7-12 of FIG. 1), ultrasonic transducer 13, test object
14, optical fiber 22, half-power or 3dB coupler 23, and optical
termination 24, e.g., in the form of a reservoir of index-matching fluid.
Also, an additional objection lens 25 is shown between the tip of fiber 4
and the test object 14.
As in any half-fringe interferometer used to achieve high resolution,
maximum sensitivity and linear range are achieved with the cavity or
stand-off distance set to yield a single mid-way between the highest and
lowest intensities. At these extremes, which lie a quarter wavelength
apart, the sensitivity is zero. Thus, disturbances such as, e.g., thermal
expansion and low-frequency mechanical vibrations, by altering the spacing
of the fiber tip from the test surface, can shift the sensitivity from its
maximum to nothing. While low-frequency (<1 MHz) vibrations having
amplitudes significantly less than one-eighth wavelength can easily be
filtered out of the signal electronically, this is not the case for
largeramplitude disturbances.
One possible way of eliminating all long-wavelength, large-amplitude
mechanical noise involves the use of both ends ends of a bidirectional
coupler as a paired interference sensor to monitor relative surface
motions at two points slightly more than one acoustic wavelength apart
(approximately 3 mm at 1 MHz in steel). This is illustrated by FIG. 3
which shows structure as described above with respect to FIG. 2, except
that fiber 22, rather than being optically terminated, serves to feed a
return signal to analyzer 21. On the basis of the known relative position
of the fiber tips, interpretation of the two return signals can be used to
eliminate the influence of mechanical noise.
While experiments as described above involved the use of piezoelectric
transducers, the use of other means for wave generation is not precluded.
In particular, as disclosed in the paper by C. P. Burger et al. cited
above, acoustic waves can be generated by means of pulses of laser light
guided through a noncontact optical fiber. Moreover, and independent of
the type of wave generation used, a scanning mode of operation can be
employed as a noncontacting optical-fiber sensor readily can be scanned
across even complicated surfaces. And finally, in the interest of enhanced
interrogation and data gathering capabilities, such sensors readily can be
combined into sensor arrays.
* * * * *
|
|
 |
|
 |
Description |
|
