U.S. patent application number 11/242969 was filed with the patent office on 2007-10-11 for laser-ultrasonic detection of subsurface defects in processed metals.
This patent application is currently assigned to Lasson Technologies, Inc.. Invention is credited to Jerome Eichenbergeer, Marvin Klein, Todd Sienicki.
Application Number | 20070234809 11/242969 |
Document ID | / |
Family ID | 38562021 |
Filed Date | 2007-10-11 |
United States Patent
Application |
20070234809 |
Kind Code |
A1 |
Klein; Marvin ; et
al. |
October 11, 2007 |
LASER-ULTRASONIC DETECTION OF SUBSURFACE DEFECTS IN PROCESSED
METALS
Abstract
Subsurface defects in a processed metal are detected by a
laser-ultrasonic method involving generation of a surface acoustic
wave at one location on the processed metal surface, and detection
of a scattered acoustic wave at another location on the processed
metal surface. The method can be used in-line to provide real time
monitoring of laser cladding and other metal processing
operations.
Inventors: |
Klein; Marvin; (Pacific
Palisades, CA) ; Sienicki; Todd; (Los Angeles,
CA) ; Eichenbergeer; Jerome; (Los Angeles,
CA) |
Correspondence
Address: |
D. Morgan Tench
1180 Corte Riviera
Camarillo
CA
93010
US
|
Assignee: |
Lasson Technologies, Inc.
|
Family ID: |
38562021 |
Appl. No.: |
11/242969 |
Filed: |
October 4, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60615925 |
Oct 5, 2004 |
|
|
|
Current U.S.
Class: |
73/602 ;
356/237.1; 356/502; 73/627; 73/643 |
Current CPC
Class: |
G01N 2291/0422 20130101;
G01N 29/46 20130101; G01N 2291/0426 20130101; G01N 29/2418
20130101; G01N 21/1702 20130101; G01N 2291/0423 20130101; G01N
2291/044 20130101; G01N 29/221 20130101 |
Class at
Publication: |
073/602 ;
073/627; 073/643; 356/237.1; 356/502 |
International
Class: |
G01N 29/04 20060101
G01N029/04; G01N 21/88 20060101 G01N021/88 |
Goverment Interests
U.S. GOVERNMENT RIGHTS
[0001] This invention was made with Government support under a
contract awarded by the United States Army. The Government has
certain rights in this invention.
Claims
1. A laser-ultrasonic method for detecting a subsurface defect
within a processed region of a metallic workpiece, comprising the
steps of: generating a probe acoustic Rayleigh wave by directing a
generation laser beam to a predetermined generation area on a
surface of the metallic workpiece within the processed region;
detecting a direct-arriving and a scattered acoustic Rayleigh wave
via an interferometer and a detection laser beam that impinges the
surface of the metallic workpiece at a detection spot which is
within the processed region and has a predetermined spatial
relationship to the generation area; repeating said step of
generating and said step of detecting for a plurality of
predetermined generation areas and detection spots on the surface
of the metallic workpiece within the processed region; and
comparing the scattered acoustic Rayleigh wave detected for at
least two predetermined detection spots to detect the subsurface
defect within the processed region of the metallic workpiece.
2. The method of claim 1, wherein the subsurface defect is a defect
selected from the group consisting of void, pore, bondline, disbond
and crack.
3. The method of claim 1, wherein the processed region is produced
by laser cladding.
4. The method of claim 3, wherein the processed region is produced
by friction stir processing.
5. The method of claim 1, wherein the processed region has the
shape of a straight or curved line of processed metal.
6. The method of claim 5, wherein the generation area has the shape
of a rectangle with the long sides of the rectangle substantially
perpendicular to the line of processed metal.
7. The method of claim 6, wherein the generation area has a length
that is substantially the same as the width of the line of
processed metal.
8. The method of claim 1, wherein the detection spot overlaps at
least a portion of the cross-sectional area of the defect when
viewed along a line perpendicular to the surface of the metallic
workpiece within the processed region.
9. The method of claim 1, wherein the generation laser beam has a
wavelength of 1064 nm.
10. The method of claim 1, wherein the detection laser beam has a
wavelength of 532 nm.
11. The method of claim 1, wherein the distance between the center
of the generation area and the center of the detection spot is less
than 4 mm.
12. The method of claim 1, wherein said step of comparing includes
performing a wavelet analysis.
13. The method of claim 1, wherein the subsurface defect is
detected within a processed region while another area of the
metallic workpiece is being processed.
14. A laser-ultrasonic device for detecting a subsurface defect
within a processed region of a metallic workpiece, comprising: a
generation laser providing a generation laser beam that impinges a
predetermined generation area on a surface of the metallic
workpiece within the processed region; a detection laser providing
a detection laser beam that impinges the surface of the metallic
workpiece at a detection spot which is within the processed region
and has a predetermined spatial relationship to the generation
area; an interferometer that detects direct-arriving and scattered
acoustic Rayleigh waves via the temporal displacement of the
surface of the metallic workpiece at the detection spot based on a
phase shift of a portion of the detection laser beam reflected from
the surface of the metallic workpiece; and an analyzer that
compares the direct-arriving and scattered acoustic Rayleigh waves
detected by the interferometer for at least two detection spots to
detect the subsurface defect within the processed region of the
metallic workpiece.
15. The device of claim 14, further comprising a translation stage
for scanning the metallic workpiece surface relative to the
detection spot.
Description
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The present invention is generally related to processing of
metals and alloys, and is more specifically concerned with
detection of defects generated near the surface of a workpiece
during processing.
[0004] 2. Description of the Related Art
[0005] Advanced metal processing methods are continuously being
developed to enable economical manufacture and repair of parts with
improved physical properties and often complicated shapes. For
example, laser cladding (also called laser powder deposition) is
being developed for build-up of stainless steel, titanium and other
metals (from metallic powders) to enable near net shape
manufacturing and repair of critical parts. Advanced joining
methods include laser welding and friction stir welding. The
friction stir approach, which involves passing a rotating tool
through a solid metallic material, may also be used for friction
stir processing (FSP) to locally create a fine-grain microstructure
providing improved mechanical properties [F. D. Nicholas, Advanced
Materials Processes 6/99, 69 (1999)].
[0006] Advanced metal processing typically occurs at high speeds
and often involves expensive workpieces and materials so that rapid
feedback on the quality of the processed region is critical to
controlling scrap rates and costs. Defects that may occur within
processed regions of metals include voids, pores, bondlines
(incompletely formed bonds), disbonds and cracks. Ideally, such
defects would be detected in-line during metal processing so as to
minimize scrap and improve product quality via timely corrective
action, which could include adjusting processing parameters and/or
interrupting the process. Metal processing defects often occur
below the surface of the processed region where they cannot be
detected by optical, spectroscopic or laser profilometer
techniques. Conventional ultrasonic detection methods are sensitive
to such subsurface defects but require that the inspected workpiece
be in contact with a fluid, which is not practical for in-line use.
Inspection methods requiring physical contact between the workpiece
and a probe are generally impractical for in-line defect detection.
In addition, surface irregularity and roughness typical of
processed metal surfaces tend to produce noise signals that
interfere with ultrasonic detection based on piezoelectric or EMAT
transducers, as well as other conventional methods.
[0007] Subsurface defect detection is also a critical requirement
for the inspection of cast and forged metals, including ingots and
railway rails, for example. Typical defects in ingots and castings
include pores and inclusions. Typical defects in railway rails
include cracks, which need to be detected in-service.
[0008] Laser ultrasonic methods have been developed for non-contact
detection of defects in solid parts. Since the "bottom" surface of
a part is often inaccessible during machining or processing
operations, the most useful laser ultrasonic methods involve both
generation and detection on the "top" surface of the part. In this
case, a pulsed generation laser beam incident on the part surface
at a predetermined generation spot generates ultrasonic waves that
propagate within and along the surface of the part. The propagated
ultrasonic waves, including those reflected from defects and the
bottom surface of the part, are detected via a detection laser beam
incident on the part surface at a predetermined detection spot. The
propagated ultrasonic waves produce a temporal displacement of the
part surface at the detection spot, which is measured via an
interferometer that analyzes a portion of the detection laser beam
reflected from the part surface. Ultrasonic waves reflected from
defects may be distinguished from other reflected ultrasonic waves
from the difference in time of arrival of the waves at the
detection spot.
[0009] Laser ultrasonic methods involving generation and detection
on the same surface have been applied to detection of various
defects, including voids and cracks, in parts of varied shapes and
comprising various materials. These methods have typically involved
generation and detection of bulk ultrasonic waves, namely
compressional waves, which tend to travel along the surface normal,
and shear waves, which tend to travel at angles to the surface
normal. Laser ablation produces strong compressional and shear
waves, whereas compressional waves produced thermoelastically are
relatively weak. Bulk ultrasonic waves are well-suited for
detecting defects that are relatively distant from the
generation-detection surface. However, bulk ultrasonic waves are
not well-suited for detecting near-surface (i.e., subsurface)
defects for which the delay time for waves reflected from defects
is very short, making ultrasonic measurements difficult.
[0010] In addition, application of prior art laser ultrasonic
methods has generally been limited to smooth and relatively even
surfaces to avoid speckle noise associated with surface roughness
and unevenness. In contrast, metallic surfaces processed by laser
cladding, friction stirring or other methods tend to be uneven and
relatively rough. Consequently, prior art laser ultrasonic
inspection methods cannot be directly applied to detection of
defects in processed metallic workpieces.
[0011] The limitations of prior art laser ultrasonic methods are
particularly acute for defect detection during laser cladding. The
cladding is typically applied in thin layers and each new layer
needs to be inspected for defects before it is buried under
subsequently applied layers. This requires detection of subsurface
defects that are very near the top surface, which cannot be
accomplished using the bulk ultrasonic waves generally employed in
the prior art.
[0012] The present invention utilizes Rayleigh waves (surface
acoustic waves) to detect subsurface defects in processed metallic
surfaces. Rayleigh waves have been used in the prior art for
characterization of near-surface material properties and for
detection of surface-breaking cracks. U.S. Pat. No. 5,894,092 to
Lindgren et al. describes use of transducers to generate and detect
Rayleigh waves in order to determine near-surface material
properties by measuring the Rayleigh wave velocity as a function of
frequency. U.S. Pat. No. 4,274,288 to Tittmann et al. describes use
of transducers to generate and detect Rayleigh waves in order to
determine the depth of a surface-breaking crack through analysis of
the ultrasonic frequencies contained in the detected ultrasonic
wave. The transducer-based approach described in both of these
prior art patents is unsuitable for use on processed metal
surfaces, which tend to be relatively rough and uneven, and cannot
be used for in-line monitoring during metal processing. In
addition, the Rayleigh wave velocity measurements used by Lindgren
are relatively insensitive to metal defects, and do not provide the
directional information needed for detection of localized defects.
Likewise, the ultrasonic frequency analysis used by Tittmann does
not provide the directional information needed to locate subsurface
defects.
[0013] In contrast, the present invention is based on detection and
analysis of scattered Rayleigh waves to detect subsurface defects.
The prior art provides no suggestion that scattered Rayleigh waves
might be useful for defect detection. Another important aspect of
the present invention is the use of laser generation and detection
of Rayleigh waves so that the invention can be applied to
relatively rough and uneven processed metal surfaces, and may be
used for in-line monitoring during metal processing.
SUMMARY OF THE INVENTION
[0014] The present invention provides a laser-ultrasonic method and
device that are useful for detection of defects within a processed
region of a metallic workpiece. The method involves a pitch-catch
approach whereby a surface acoustic wave (Rayleigh wave) is
laser-generated at a first location on the workpiece surface, and a
scattered portion of the generated Rayleigh wave is detected at a
second location on the workpiece surface, along with the
direct-arriving (unscattered) Rayleigh wave. The invention is
particularly useful for detecting voids in laser cladded metallic
layers and friction stir processed layers. The method of the
invention may be used for in-line monitoring of laser cladding and
friction stir processes, which generally result in a line or bead
of processed metal having an appreciable width. The line of
processed metal may be curved or straight.
[0015] In the method of the invention, a probe Rayleigh wave is
generated in the workpiece by directing a generation laser beam of
small dimensions to a predetermined generation area within the
processed region of the workpiece. When the probe Rayleigh wave is
scattered by a subsurface defect, the scattered Rayleigh wave is
detected via the temporal displacement of the workpiece surface
produced by the scattered Rayleigh wave. This surface displacement
is measured using an interferometer and a detection laser beam of
small diameter that impinges (interrogates) the workpiece at a
detection spot within the processed region of the workpiece. A
predetermined spatial relationship is maintained between the
generation area and the detection spot. High sensitivity and
resolution are attained via use of very small laser beam dimensions
and a close spacing between the generation area and the detection
laser spot. Sensitivity is typically highest when the detection
laser spot overlaps at least a portion of the cross-sectional area
of the defect when viewed along a line perpendicular to the surface
of the metallic workpiece within the processed region. Further
improvement in sensitivity may be provided via wavelet analysis of
the detection signal.
[0016] In a preferred embodiment, the invention is used to detect
defects in a line or bead of processed metal. In this case, the
laser generation area preferably has the shape of a rectangle with
the long sides of the rectangle substantially perpendicular to the
line of processed metal. The laser generation area preferably spans
the width of the line of the processed metal. In this case, the
entire width of the processed metal line may be continuously
monitored for defects in real time during the metal processing
operation.
[0017] The method of the invention may also be used to provide an
image of subsurface defects within the processed region of a
metallic workpiece. In this case, measurements of acoustic
waveforms (surface displacement magnitude vs. time) are made at
regularly spaced locations along the processed workpiece surface,
while a predetermined spatial relationship is maintained between
the laser generation area and the detection laser spot. This may
involve maintaining the generation and detection laser beams at
stationary positions while the workpiece is moved so that the laser
beams track along a line or bead of processed metal. Alternatively,
the workpiece may be maintained in a stationary position while the
laser beams are scanned along the surface of the processed metal.
In either case, x-y raster scanning may also be employed.
Preferably, the relative motion between the laser beams and the
workpiece is such that both laser beams impinge the workpiece
surface along the line of motion.
[0018] In one embodiment, a single waveform corresponding to a
defect-free location is chosen as a reference, and the overall
amplitude of each waveform is normalized to the amplitude of the
reference waveform. A computer program is preferably used to
calculate the Mean Square Error (MSE) between the reference
waveform and each of the other waveforms in the raster scan. A plot
of MSE intensity versus x-y location provides an image of defects
in the processed metal.
[0019] In a preferred embodiment, the waveform acquired at each
location on the workpiece surface within the processed metal region
is analyzed using a wavelet transform. This analysis detects the
characteristic changes in the waveform that are uniquely associated
with scattering from subsurface defects.
[0020] The device of the present invention for detecting a defect
in a processed metal comprises a generation laser, a detection
laser, an interferometer and an analyzer, and may further comprise
a translation stage.
[0021] The present invention provides significant advantages
compared to prior art methods. A key advantage is that the
laser-ultrasonic method and device of the invention can be used for
in-line detection of metal processing defects, enabling 100% parts
inspection and real-time process control. The invention may be
applied to detection of defects in metals processed by a variety of
methods, including laser cladding, laser welding, friction stir
processing and friction stir welding. The invention permits each
layer of a laser cladding process to be monitored for defects.
[0022] Further features and advantages of the invention will be
apparent to those skilled in the art from the following detailed
description, taken together with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0023] FIG. 1 depicts a side view (A) and a top view (B)
schematically illustrating preferred shapes and locations for the
laser generation area and the laser detection spot for detection of
defects in a cladded bead on a workpiece.
[0024] FIG. 2 depicts a side view (A) and a top view (B)
schematically illustrating detection of a defect in a workpiece via
Rayleigh waves scattered perpendicular to the direction of
travel.
[0025] FIG. 3 depicts a side view schematically illustrating
detection of a defect in a workpiece via backscattered Rayleigh
waves.
[0026] FIG. 4 shows acoustic waveforms obtained for a stainless
steel plate without and with overlap of the detection spot with a
simulated defect (blind hole 1.5 mm diameter and 0.4 mm deep).
[0027] FIG. 5 shows a wavelet analysis plot of the relative energy
at the fifth wavelet level for acoustic waveforms measured during a
linear scan on a titanium alloy workpiece with a simulated defect
(blind hole 1.0 mm diameter and 0.4 mm deep in the back
surface).
[0028] FIG. 6 shows a wavelet analysis plot of the relative energy
at the sixth wavelet level for acoustic waveforms measured during a
linear scan on the workpiece of FIG. 5.
[0029] These figures are not drawn to scale. Some features have
been enlarged relative to other features for ease of depiction.
DETAILED DESCRIPTION OF THE INVENTION
[0030] Technical terms used in this document are generally known to
those skilled in the art. The terms "subsurface" and "near-surface"
are used interchangeably to denote defects residing near the
workpiece surface at which the generation and detection laser beams
are incident. The terms "Rayleigh wave", "surface acoustic wave"
and "surface wave" are also used interchangeably. The term
"workpiece" denotes any structure that includes processed metal,
whether the metal processing is in progress or has been completed.
The term "spot" is widely used in the art to denote a laser
impingement area, which is typically circular and of small
diameter. The terms used to denote the laser impingement areas of
the present invention are "generation area" for the generation
laser and "detection spot" for the detection laser. As used in this
document, the terms "scattering" and "scattered" denote that a
change in the direction of an ultrasonic wave propagated in a
material occurs due to either deflection or reflection by a defect
in the material.
[0031] The invention provides a laser ultrasonic method and device
for detecting a subsurface defect within a processed region of a
metallic workpiece. Subsurface defects that may be detected include
voids, pores, bondlines, disbonds and cracks. The invention may be
applied to detection of defects in metals processed by a variety of
methods, including laser cladding, laser welding, friction stir
processing, friction stir welding, and casting. It may also be
applied to in-service inspection of railroad rails, and various
other workpieces comprising processed metal. The invention is
particularly advantageous for detecting defects in straight or
curved lines of processed metallic material, but may also be
applied to processed metal regions of other geometric shapes.
[0032] The device of the invention comprises: (1) a generation
laser providing a generation laser beam that impinges a
predetermined generation area on a surface of the metallic
workpiece within the processed region; (2) a detection laser
providing a detection laser beam that impinges (interrogates) the
surface of the metallic workpiece at a detection spot which is
within the processed region and has a predetermined spatial
relationship to the generation area; (3) an interferometer that
detects direct-arriving and scattered acoustic Rayleigh waves via
the temporal displacement of the surface of the metallic workpiece
at the detection spot based on the phase shift of a portion of the
detection laser beam reflected from the surface of the metallic
workpiece; and (4) an analyzer that compares the direct-arriving
and scattered acoustic Rayleigh waves detected by the
interferometer for at least two detection spots to detect the
subsurface defect within the processed region of the metallic
workpiece. The device of the invention may further comprise a
translation stage for scanning to detect defects over a relatively
large area of the workpiece, and optionally provide an image of the
defect areas. The generation and detection lasers may be of any
type that provides a suitable wavelength and sufficient power. A
suitable generation laser is a Nd:YAG laser operating at a
wavelength of 1064 nm and a pulse width of 10 ns (10-30 mJ per
pulse). A suitable detection laser is a continuous wave
frequency-doubled Nd:YAG laser operating at a wavelength of 532 nm
(about one Watt of power).
[0033] FIG. 1 illustrates preferred shapes and locations of the
generation area and the detection spot for detecting defects in a
linear cladded bead 101 on a workpiece 102. Generation laser beam
103 impinges cladded bead 101 in generation area 104, which is
preferably rectangular in shape with the long sides of the
rectangle substantially perpendicular to the line defined by
cladded bead 101. For a curved cladded bead (not shown), the long
sides of the rectangular generation area would preferably be
substantially perpendicular to the tangent to the curved line
defined by the cladded bead. Preferably, the length of generation
area 104 is substantially the same as the width of cladded bead
101, but may be larger (as shown) or smaller. If laser generation
area 104 spans the width of cladded bead 101, the entire width of
cladded bead 101 may be continuously monitored for defects in real
time during the metal cladding operation. If laser generation area
104 is substantially longer than the width of cladded bead 101, a
significant portion of the laser energy may be wasted. Detection
laser beam 105 impinges cladded bead 101 in detection spot 106,
which is located within cladded bead 101 and at a predetermined
fixed distance from generation area 104. Typically, detection spot
106 is circular and has a diameter that is much smaller than the
width of cladded bead 101.
[0034] During measurements according to the invention, the
predetermined fixed distance between generation area 104 and
detection spot 106 is maintained constant as workpiece 102 is moved
relative to detection spot 106, or detection spot 106 is moved
relative to workpiece 102, so that generation area 104 and
detection spot 106 move along cladded bead 101. Typically, laser
beams 103 and 105 are parallel and directed perpendicular to the
surface of workpiece 102, and the relative motion is effected
perpendicular to the laser beams so that the distance between the
lasers and the surface of workpiece 102 remains constant. In one
embodiment, the laser inspection system of the invention is
stationary and the relative motion is provided by the machine used
to perform the laser cladding operation. In another embodiment, the
device of the invention further comprises a translation stage for
scanning in one or two dimensions across the metallic workpiece
surface relative to the fixed generation area and detection laser
spot.
[0035] As also depicted in FIG. 1, generation laser beam 103
preferentially generates Rayleigh waves 107 and 108 that propagate
perpendicular to the long sides of rectangular generation area 104,
and travel in both directions along cladded bead 101. Forward
Rayleigh wave 107, which travels in the same direction as
generation area 104 and detection spot 106 relative to workpiece
102, arrives at detection spot 106 and is used to detect defects in
cladded bead 101. Backward Rayleigh wave 107 does not arrive at
detection spot 106 and is not used.
[0036] The method of the invention comprises the steps of: (1)
generating a probe acoustic Rayleigh wave by directing a generation
laser beam to a predetermined generation area on a surface of the
metallic workpiece within the processed region; (2) detecting a
direct-arriving and a scattered acoustic Rayleigh wave via an
interferometer and a detection laser beam that impinges the surface
of the metallic workpiece at a detection spot which is within the
processed region and has a predetermined spatial relationship to
the generation area; (3) repeating said step of generating and said
step of detecting for a plurality of predetermined generation areas
and detection spots on the surface of the metallic workpiece within
the processed region; and (4) comparing the scattered acoustic
Rayleigh wave detected for at least two predetermined detection
spots to detect the subsurface defect within the processed region
of the metallic workpiece. The step of comparing may include
generating acoustic waveforms, directly comparing acoustic
waveforms, calculating the Mean Square Error (MSE) between acoustic
waveforms, generating B-scans, performing wavelet analysis, and
combinations thereof. The scattered acoustic Rayleigh wave may be
scattered substantially perpendicular to the direction of travel of
the Rayleigh wave, or may be scattered at another angle.
[0037] FIG. 2 illustrates a preferred embodiment by which a defect
in a processed workpiece is detected via an acoustic wave derived
from scattering of a Rayleigh wave perpendicular to its direction
of travel. Rayleigh wave 207 is generated in workpiece 201 by
generation laser beam 203 at generation area 204 and is scattered
at defect 211 to generate acoustic wave 212. Scattered acoustic
wave 212 may be detected at detection spot 206 via a reflected
portion (not shown) of detection laser beam 205 and an
interferometer (not shown). Note that Rayleigh waves (surface
acoustic waves) are not confined strictly to the surface of the
material of propagation but extend approximately one wavelength
below the surface. This corresponds, for example, to a penetration
depth in steel of about 3 mm for a Rayleigh wave of 1 MHz
frequency. Consequently, subsurface defects at appreciable
distances from the workpiece surface can be detected via scattering
of surface acoustic waves (Rayleigh waves), according to the
invention. Rayleigh wave 207 is represented by multiple (3) arrows
in FIG. 2 to indicate its appreciable penetration depth.
[0038] The method of the invention generally provides highest
sensitivity to defects when acoustic wave 212 is scattered at right
angles to Rayleigh wave 207, and detection spot 206 at least
partially overlaps defect 211 with respect to the perpendicular to
the surface of workpiece 201 (as indicated in FIG. 1). In this
case, the distance between detection spot 206 and defect 211 is
minimized so that signal intensity loss due to dispersion of
scattered acoustic wave 212 is minimized. Preferably, detection
spot 206 is sufficiently small to enable resolution of defects that
would significantly affect the properties of the processed metal of
workpiece 201 with respect to its intended use. Larger spot sizes
may be used, for example, to enable inspection of a larger
workpiece area or processed metal line in a single inspection pass,
or to minimize the number of inspection passes required.
[0039] The generation area width, the detection spot diameter, and
the separation distance between these two laser impingement areas
should be selected to provide temporal separation between the
arrival times for the direct-arriving Rayleigh wave and the
scattered acoustic wave that is sufficient for the measurement
equipment to be used. For detection of defects in steel (Rayleigh
wave velocity of approximately 3 mm/.mu.s) with a detection
bandwidth of 1-10 MHz (corresponding to a generation/detection
pulse width of 100 ns), for example, requires that the width of the
generation area and the diameter of the detection spot be 300 .mu.m
or less. Note that a generation laser pulse of 10 ns width
generates Rayleigh waves having frequencies up to 100 MHz.
Generally, the separation distance between the laser generation
area and the laser detection spot should be as small as practical,
preferably less than 4 mm center to center. For interrogation of
relatively smooth surfaces using 300-.mu.m laser beams, a
separation distance of 3 mm is typically adequate. A typical spot
size is approximately 100 .mu.m, for which the beam separation
distance can be reduced. Smaller separation distances (.about.1.5
mm) may be needed for interrogation of relatively rough surfaces,
such as laser clad beads, for which the Rayleigh wave tends to be
attenuated and distorted by scattering from random surface
features. For separation distances less than .about.1.5 mm, an
ablation plasma produced by the generation laser may interfere with
the detection process.
[0040] Although a less preferred embodiment, subsurface defects in
processed metals may also be detected according to the invention
via Rayleigh waves scattered at angles other than 90.degree.
relative to the direction of travel. FIG. 3 illustrates detection
of a subsurface defect in a workpiece via backscattered Rayleigh
waves. In this case, Rayleigh wave 307 generated in workpiece 301
by generation laser beam 303 is reflected from defect 311 as
backscattered acoustic wave 312, which is detected at the detection
spot defined by detection laser beam 305. For this embodiment,
detection sensitivity is reduced by dispersion of backreflected
acoustic wave 312 over the greater distance between defect 311 and
the detection spot (compared to that for scattering perpendicular
to the Rayleigh wave). Scattered acoustic waves that impinge the
workpiece surface at an angle greater than 90.degree. also produce
diminished surface displacement so that the measurement signal is
further reduced. Likewise, forward scattered Rayleigh waves also
provide reduced detection sensitivity.
[0041] Subsurface defects are detected according to the invention
from the longer time required for the scattered acoustic wave,
which must travel a greater distance compared to the
direct-arriving Rayleigh wave, to arrive at the detection spot.
Typically, the amplitude of the signal from a laser interferometer
is plotted as a function of time (relative to the time at the end
of the generation laser pulse) to produce an acoustic waveform. In
the absence of scattering from defects, the acoustic waveform
typically exhibits a single, relatively-sharp peak corresponding to
the arrival time of the direct-arriving Rayleigh wave. Acoustic
waves resulting from scattering at approximately 90.degree. to the
Rayleigh wave tend to broaden this peak toward longer times.
Acoustic waves resulting from scattering at angles significantly
greater than 90.degree. to the Rayleigh wave tend to produce a
shoulder or second peak of smaller amplitude at longer times
(corresponding to a longer distance from the defect to the
detection spot).
[0042] The method of the invention may be used to detect defects
over relatively large areas of a processed metallic workpiece. In
this case, acoustic waveforms are measured at regularly-spaced
locations along the processed workpiece surface, while a
predetermined spatial relationship is maintained between the laser
generation area and the detection laser spot. This may involve
maintaining the generation and detection laser beams at stationary
positions while the workpiece is moved so that the laser beams
track along a line or bead of processed metal. Alternatively, the
workpiece may be maintained in a stationary position while the
laser beams are scanned along the surface of the processed metal.
In either case, x-y raster scanning may also be employed.
Preferably, the relative motion between the laser beams and the
workpiece is such that both laser beams impinge the workpiece
surface along the line of motion.
[0043] The method of the invention may also be used to provide an
image of subsurface defects within the processed region of a
metallic workpiece. In one embodiment, a single acoustic waveform
corresponding to a defect-free location is chosen as a reference,
and the overall amplitude of each of the other acoustic waveforms
is normalized to the amplitude of the reference waveform. A
computer program is preferably used to calculate the Mean Square
Error (MSE) between the reference waveform and each of the other
waveforms in the raster scan. A plot of MSE intensity versus x-y
location provides an image of defects in the processed metal.
[0044] An alternative approach for displaying an image of
subsurface defects is the B-scan, which is false color or grey
scale representation of the signal amplitude for acoustic waveforms
measured at regular intervals in a substantially straight line
along the workpiece surface. For the B-scan image, all of the
acoustic waveforms are placed on the same time scale and the signal
amplitude representation is displayed as a function of time
(x-axis) and detection spot location (y-axis). In this case, the
direct-arriving Rayleigh wave appears as a vertical line (or series
of vertical lines), and defects are easily identified as a "bulge"
along the time axis corresponding to the late-arriving scattered
acoustic waves.
[0045] In a preferred embodiment, the waveform acquired at each
location on the workpiece surface within the processed metal region
is analyzed using a wavelet analysis. This analysis accents the
characteristic changes in the waveform that are uniquely associated
with scattering from subsurface defects so that the effects of
extraneous signals, derived from workpiece surface roughness, for
example, are minimized. Note that acoustic signals tend to be
inherently nonstationary so that they are difficult to separate in
the time domain or the Fourier frequency domain. Wavelet analysis,
which involves an oscillating burst as the wavelet basis function,
is well-suitable to laser-ultrasonic applications, for which the
signal of interest is an oscillating burst. The discrete wavelet
transform involved in wavelet analysis may be implemented via a
bank of inexpensive finite impulse response (FIR) digital
filters.
[0046] In wavelet analysis, the signal is treated as the weighted
sum of overlapping wavelet functions. For a transient signal, most
of the signal energy is assumed to be concentrated in relatively
few of the wavelet coefficients, which contain most of the signal
information. Wavelet analysis typically involves transforming the
temporal signal into wavelet space, zeroing the coefficients that
carry a large fraction of the original undistorted waveform and
those that represent noise, and then transforming back into the
time domain. This procedure effectively discards the part of the
signal that contains no useful information about the defect. For
the present invention, the transformed temporal signal resulting
from wavelet analysis is much more sensitive to small changes
resulting from defect scattering.
[0047] The invention may be used to detect subsurface defects
within a processed region of a metallic workpiece while another
area of the metallic workpiece is being processed, by laser
cladding, for example. For laser cladded material built up of
layers 1-2 mm thick, each layer may be inspected for defects before
the next layer is deposited. Surface roughness may influence the
measurement conditions and results but typically does not increase
as layers are deposited since the surface of the previous layer is
melted during deposition of the next layer.
DESCRIPTION OF A PREFERRED EMBODIMENT
[0048] The efficacy of the method of the invention was demonstrated
by detecting simulated voids (blind holes) in stainless steel and
titanium workpieces. Acoustic waveforms were generated using a
pulsed Nd:YAG generation laser (1064 nm wavelength) that provided
10-ns pulses with an energy of 10-30 mJ per pulse (sufficient to
produce mild ablation). Such pulses generate Rayleigh waves with
frequencies up to 100 MHz. The generation area was rectangular with
a width of 280 .mu.m and a length of 3 mm. Detection was provided
by a Lasson AIR-532-TWM laser ultrasonic receiver operating at a
wavelength of 532 nm. This receiver has a bandwidth of 125 MHz and
provided about one Watt of power to the detection spot, which was
about 80 .mu.m in diameter. The separation distance between the
generation area and the detection spot was 3 mm (center to
center).
EXAMPLE 1
[0049] FIG. 4 shows acoustic waveforms illustrating detection
according to the invention of a simulated defect (blind hole 1.5 mm
diameter and 0.4 mm deep in the back surface) in a machined
stainless steel plate (6.4 mm thick). When the detection spot did
not overlap the simulated defect, the acoustic waveform exhibited a
relatively narrow peak (labeled "no overlap" in FIG. 4)
corresponding to arrival of the direct-arriving Rayleigh wave,
which was generated at a time of 1 .mu.s in the plot of FIG. 4.
When the detection spot did overlap the simulated defect, this peak
(labeled "defect overlap" in FIG. 4) was broadened by the signal
corresponding to arrival of the scattered acoustic wave at a later
time.
EXAMPLE 2
[0050] A B-scan was generated for a machined stainless steel plate
(6.4 mm thick) with a simulated defect (blind hole 1.5 mm diameter
and 0.4 mm deep in the back surface). The defect was detected at
the expected location in the B-scan as a time-delayed bulge in the
vertical line corresponding to the direct-arriving Rayleigh
wave.
EXAMPLE 3
[0051] A B-scan was generated for a machined titanium alloy 4-6
plate (6.7 mm thick) with a simulated defect (blind hole 1.0 mm
diameter and 0.4 mm deep in the back surface). The defect was
detected at the expected location in the B-scan as a time-delayed
bulge in the vertical line corresponding to the direct-arriving
Rayleigh wave. There is less grain scattering in titanium than in
steel, so that more ultrasonic features were visible in the
titanium plate. In addition to the bulge associated with direct
scattering from the simulated defect, the B-scan exhibited two
prominent diagonal lines that are associated with surface waves
reflected from this defect. These diagonal lines, whose slopes
depend on the surface wave velocity, may also be useful for defect
detection and localization.
EXAMPLE 4
[0052] The B-scan data for the titanium plate with the simulated
defect (Example 3) was subjected to wavelet analysis. For each of
201 scans, the largest wavelet coefficients, which contributed 70%
of the signal energy, were presumed to be associated with the
unperturbed signal due to the direct-arriving Rayleigh wave and
were zeroed out. Likewise, the smallest coefficients, which
contributed 1% of the signal energy, were presumed to result from
noise and were also zeroed out. Wavelet-processed temporal B-scans
were constructed using the remaining (intermediate) coefficients,
which were presumed to contain the information needed to provide
improved defect contrast. FIGS. 5 and 6 show wavelet analysis plots
of the relative energy at the fifth and sixth wavelet levels,
respectively, for acoustic waveforms measured at various locations
on the titanium alloy plate (Example 3). In both cases, a sharp
peak is evident at scan 98, which is the expected location for the
simulated defect.
[0053] The preferred embodiments of this invention have been
illustrated and described above. Modifications and additional
embodiments, however, will undoubtedly be apparent to those skilled
in the art. Furthermore, equivalent elements may be substituted for
those illustrated and described herein, parts or connections might
be reversed or otherwise interchanged, and certain features of the
invention may be utilized independently of other features.
Consequently, the exemplary embodiments should be considered
illustrative, rather than inclusive, while the appended claims are
more indicative of the full scope of the invention.
* * * * *