Optical Ultrasonic Measurement
This invention relates to a method and an apparatus for detecting ultrasonic waves at a surface of an object using optical techniques.
The use of optical techniques using lasers to detect ultrasonic waves on a surface of an object is widely known. In one such technique a continuous laser beam is split into two, one beam being incident on the surface and the other beam following a reference path, and the resulting reflected beams are arranged to interfere with each other. Changes in the position of the surface cause a phase difference between the two beams, so when the beams interfere the intensity changes.. In another technique a laser beam is incident on the surface, and is then reflected to pass through a very narrow band filter (such as a Fabry-Perot etalon) ; movement of the surface changes the frequency of the reflected beam, and so changes the intensity of the light emerging from the filter. It is also known to use a pulsed laser source to generate short pulses of ultrasonic waves within a material; for example in GB 2 172 106 B a pulsed laser is used to generate ultrasonic waves, and a laser interferometer is used to detect ultrasonic waves.
In some applications the velocity of the ultrasonic waves may be related to properties of the material which are of interest, and this velocity may be determined by measuring the time the waves take to propagate a known distance. Where the waves are surface waves, or Lamb waves, an obvious approach is to take this distance as the distance between the transmitter and the receiver. However this is not always possible when using a pulsed laser source, as such sources are known to exhibit pulse- to-pulse positional variation, and this can be a
significant source of error in such measurements. This problem can be overcome by using two separate laser detectors at a fixed separation, the continuous nature of the lasers providing inherently greater directional stability, and the times of arrival at the two detectors can be compared electronically, but this involves both complex optical equipment and complex electronic circuitry.
According to the present invention there is provided an apparatus for detecting ultrasonic waves at a surface of an object at two separate locations, the apparatus comprising a single continuous single-frequency laser source, means to split the laser beam into two out-going beams, means to cause the two out-going beams to be incident on the surface at the two separate locations, and means to cause the resulting reflected beams from the two separate locations to be incident on a common detector, wherein polarisation means are provided so that the detected beams from the two separate locations are polarised in orthogonal directions, and the detector comprises an interferometric device.
The out-going beams may each be split into two, one following a reference path and the other being incident on the surface at one of the said locations, each pair of beams then being recombined so as to interfere at the interferometric device. In this case one pair of beams, i.e. one reference beam and one beam incident on the surface, must have the same polarisation, while the other pair of beams have the orthogonal polarisation. In a preferred embodiment, however, the out-going beams are not split in that manner; each is incident on the surface, and is then returned to a common detector, the detector comprising a very narrow band
'filter, preferably a Fabry-Perot etalon. In either case the intensity of
3 rt P- Ω 3 σ rt rt TJ P> H- C α rt P-" Ω tr O ιQ co Cfl 0 s; rt TJ rt P1 O 3 rt o t rt SD SD H P. O φ 0 φ ø- 0 O Φ φ 0 0 SD φ H (D P- tr tr 0 0 p- Φ tr ø Φ CO rt SD SD CO 3 rt rt Φ Ω 3 (D TJ P- ø CO T rt 3 o P Ω TJ O SD Φ
P- 3 rt Φ TJ 3 0 P- O Ω P φ CD 3 SD P H Φ SD li Φ φ rt o SD φ Si Cfl cϊ- rt SD P- H Φ ω O rt < o SD Ω rt o CO ii vQ Ω P H SD o d rt P φ d PJ
O ≤ ^ 0 P- H 3 3 P- Φ 3 CO r+ P) Φ P- ø SD Φ SD CO P3 rt co α iQ P- TJ • H P-
H o Q SD Hi p- P- o O & O O rt rt o Hi rt tr σ rt O t O Φ φ tr O o φ iQ
P- tr H O rt rt 0 P- 3 Φ 0 ii Φ Φ P- Φ Φ 0 φ TJ rt ts Cfl &. tr
0 H Φ O Ii rt rt SD rt a P- t, Ω CO Φ O SD 3 SD P- TJ SD Φ rt fl P- m rt iQ o Hi 3 Φ Φ H ø- SD Ω O P- rt rt 3 H 3 rt 3 PJ Ω P- H P Ω t rt tr €,
Ω 0 ≤ Φ ii li Φ rt Hi O σ φ O CO O O 3 0 CD rt Φ P- Φ P- Φ ø SD CO 3 t α » O <! P- s: H Φ o rt Ω Hi Ω S TJ rt o SD o rt 3 rt φ φ P- • SD d O SD rt P- rt ør CD rt rt H SD SD φ SD P- H rt tl ts O tr φ rt P- α rt Ω P- P- P 0 0 < tr tr Ω ø O φ rt O O rt <i ts ^Q O t, P- li
1-1 O (D tr ø rt H P- O Φ φ Φ O ιQ H P- O 3 P- Φ rt SD 0 s: o 0 iQ SD i
SD 0 rt H SD Φ 51 ω 3 rt O tr O Cfl P- rt P- P- Hi ts ts P- to ω o O P- SD Φ rt H P- P- SD TJ TJ £ 0 φ O CD 13 0 Φ 0 M i- SD 3
O CO 0 Ω P- rt rt 0 TJ ii O rt O CO ø Cfl SD ts Hi o rt H Φ iQ ts 3 SD O o SD ør tr < TS P- ø1 **« P- rt Ω *"« rt Hi Ω SD 0" O P1
P- tr Φ O rt O tr 3 0 P- 0 .φ ii co SD Φ CO 0 1 O SD ii rt P- SD ts Φ Φ Hi
O rt H O 3 H tr 0 SD rt 0 Φ P- H φ SD Ω ιQ 0 rt SD O tr 0 rt Hi Ω H riφ O 1 φ P- . — , rt rt Ω ø P- CL T5 0 P- O rt ø1 CO 3 Φ ex P- d CO P. rt O
Ω O Ω O 3 TJ P- T rt P- P- u2 Cfl Φ SD P- P- Φ ω O P- O d O li 3
Cu prt 3 O iD o Φ c t O &) Φ rt H Φ ts 0 0 O rt Ω ts rt H 3 O
<! en O 0 3 rt P- CO SD 0 rt CU Φ SD Ω US iQ 0 3 ii TJ ts s: SD co p. Hi 0 rt D SD (D rt rt SD P- 0 CO rt Φ 0 O rt SD rt O Φ Hi P. φ o rt CD D rt P- ø1
CO TJ 0 SD rt O rt P- Φ 3 & P- rt Φ 0 tr ø rt SD O Φ Cfl Ω tr Ω Φ
0> ■ 3 r+ Ω Φ 0 P- 3 a P- SD P- 0 Φ CO o Φ O ør Ω TJ rt o Φ φ s: H O 0 rt H tS •TJ rt * rt l=S α P1 P- 3 SD o Φ P- 0 Φ rt 1 ts Q. - tr rt rt <! P.- t, iQ o -1 t rt o O 0 3 CO α α Ω CD Φ P- 3 TJ Φ d
P- P- " p- rt 3 SD P- O SD P- tr SD Φ ii tr Ω iQ rt CO P- O Φ SD ? o Ω o ø- rt rt
Ω Ω TJ ø O SD rt Hi φ CO o Φ rt ii rt Φ SD 1 0 Ω Hi co rt CO O < o Φ Φ tr Φ iQ 0 0 ør φ Hi t SD rt rt Φ iQ 0 P- CO P- € φ rt Ω P
P" co 0 0 TJ d H TJ α rt Φ o 3 P- ø- Ω H 3 SD (D O tr d ιQ (D 3 O rt Hi
TJ SD to - Φ H SD o tr SD CO ø1 P. ιQ CO O Φ CO SD Φ TJ 0 ts Φ Ω tl Φ O Φ
0) li O T •*» H O rt ^ rt H Φ Φ o O 0 d 0 1 H 3 P- Ω (D φ ø φ P. P fl Hi φ -> T TJ C P. rt α 3 3 Hi CO P li CO Hi P- O φ rt ø Φ rt rt O co ■< Ω Hi Φ 0 SD rt SD P- ø Φ SD ii Φ Hi P- H n tr rt O iQ CO Cfl TJ CO Φ CO 3
CO P- O Hi ii H rt tr 3 Ω s; P1 rt H O rt CO SD 0 Φ P- t CO SD Ω d Φ
(H tr d 3 li O r O P- Φ CO 0 P- rt H 3 O ø Ω iQ .Q a φ O rt rt P- Hi Cfl SD rt Ω rt
Φ h φ Ii ^ φ O 3 rt li P- α H Φ d Q Ω α ør ør J li Cfl rt O ør P
O 3 Hi 3 Φ 0 P- P- £U ø- SD Ω P- rt . tr rt rt Φ rt Φ Φ φ O φ H P-
3 Φ SD • X Φ 3 CO ø rt H CO li ø1 Φ P- SD t ø co Hi 3 Cfl rt CD Ω iQ Hi Ω SD O P1 — - rt (D o . Φ φ 3 rt Ω T flJ o O rt TJ ø- H CO
P- φ 3 SD 3 SD Φ Φ ; P 0 Φ Ω P- vQ •< rt ii rt P- φ rt Cfl Φ Φ rt Ω ^ TJ 3 P- Cfl O P. ■<: P- rt rt ø rt rt P- ø- rt 3 CD tr d TJ SD φ ør P- 3 d PJ TJ rt Φ Hi Φ •^ tr Ω P- P- S Ω ii ø' s: P1 rt α Φ Φ ? φ Ω rt P <!
Φ (D O P> Φ O Ii ø Φ O Ω O O P- Φ Φ O SD rt- s: Φ li Cfl Ω € Φ TJ P-
<3 rt Φ ii CO rt- SD 3 O ≤ Φ 3 Hi O P- O r+ • Hi *< φ O CO tr Ω
CO φ ø- P- 0 O Φ d SD • o O Φ P1 rt O Φ 13 Φ o O Φ O Φ d € 3 Φ d 0 0 Ω Hi H φ ø φ s; 0 H iQ Ω li rt co t) rt
P ø- φ P> H P iQ tr P- P- Ω Φ TJ rt Ω o rt rt tr P- rt O 3
Hi Φ 0 3 P- rt Ω φ SD SD rt 1 SD P- rt φ <! SD
CU ti rt O ø rt P- US ii O Φ ts ø1 Φ (D *<
Ω H ιQ O SD O SD 0 i Φ
Φ SD Φ 1 Hi rt ><: tr rt Φ SD Φ
for example surface waves (Rayleigh waves) , or Lamb waves .
The invention will now be further and more particularly described, by way of example only, and with reference to the accompanying drawing which shows a diagrammatic plan view of an apparatus for detecting ultrasonic waves.
Referring to the drawing, an apparatus 10 is shown for measuring the difference in the arrival times of ultrasonic surface waves in a specimen 12 at two locations A and B. Ultrasonic waves are generated by a laser 14 arranged to emit a pulsed light beam 16 which is focused onto a line 18 about 1 mm wide ' (and 10 mm long) on the surface Of the specimen 12. For example the laser 14 may produce pulses of energy 80 mJ and of duration 10 ns at a pulse repetition frequency of for example 20 Hz. Each laser pulse produces a very sharp ultrasonic pulse in the specimen 12 that includes frequencies above 1 MHz.
The apparatus 10 comprises an argon ion laser 20 that incorporates an etalon within the optical cavity to ensure it operates in a single longitudinal mode, so it emits a single frequency (at a wavelength of 514 nm) .
The laser 20 produces a continuous beam 22 of light which passes through a half wavelength plate 23 and then a polarising beam splitter 24, so generating two beams marked 1 and 2. The polarising beam splitter 24 reflects light with a vertical plane of polarisation, and transmits light with a horizontal plane of polarisation, and by adjusting the angle of the half wavelength plate 23 the relative intensities of the beams 1 and 2 can be adjusted. The transmitted beam 1 (which is initially horizontally polarised) is incident on the surface of the specimen 12 at location A, via a polarising beam splitter
26, and a quarter wave plate 27. The other beam 2 (which is initially vertically polarised) is passed through a half wavelength plate 28 arranged so the beam 2 emerges with horizontal polarisation, is reflected by a prism 29, and is then incident on the surface of 'the specimen 12 at location B, via a polarising beam splitter 30 and a quarter wave plate 31. The beams 1 and 2 reflected from the surface of the specimen 12, after returning through the quarter wave plates 27 and 31 respectively, are vertically polarised, and so are reflected by the beam splitters 26 and 30 respectively. The beam 1 is passed through a half wavelength plate 32 so it is horizontally polarised, and is then reflected by a prism 33 to pass through a beam splitter 34. The other beam 2 is reflected by the beam splitter 34 , so the beams then follow the same path. The beams 1 and 2 then pass through a confocal Fabry-Perot etalon 36, to be incident on a photo diode 38. The photo diode 38 provides electrical signals to an electronic detector, which in this embodiment is a cathode ray oscilloscope 40.
It will be appreciated that because the beams 1 and 2 received by the Fabry-Perot etalon are orthogonally polarised they do not interfere with each other. The photo diode 38 consequently provides signals which represent a superposition of the surface movements at the locations A and B. Every pulse of waves generated by the laser 14 propagates across the surface of the specimen 12, passing through locations B and A in succession. Hence the oscilloscope 40 will display two pulses at a time separation equal to the time the wave takes to propagate between those locations. The propagation time between locations B and A can therefore be measured accurately, and is unaffected by any displacement of the laser beam 16.
The ultrasonic waves to be detected may be in, the frequency range 1-15 MHz. Preferably the etalon incorporates a piezoelectric tuning device to ensure that the peak intensity from the argon laser 20 (514 ran) is to one side of the transmission peak and about half way down the peak (so as to maximise the rate of change of transmission with frequency) , and a feedback circuit is preferably provided to maintain this optimum sensitivity despite any vibrations of the apparatus 10. Such vibrations would typically be no more than 1 kHz, which is well below the frequency of the ultrasonic waves. The distance between the locations A and B might be in the range 10 mm - 20 mm, and similarly the distance between the location B and the spot 18 might also be about 20 mm.
It will be appreciated that the signal from the photo diode 38 displayed on the oscilloscope 40 may incorporate noise, or other signals which are not relevant, for example due to other ultrasonic wave modes. The time interval between corresponding pulses from the two locations A and B may be determined by signal processing techniques such as auto-correlation. Such autocorrelation is rendered easier by the fact that both the signals are on a single trace, as compared to a system with two spaced-apart independent detectors.
It will be understood that the apparatus may differ from that described, for example the laser 20 may be replaced by a different type of single-frequency laser, for example a diode-pumped neodymium-yttrium aluminium- garnet laser.
Bulk waves, both shear and compression, are generated if the specimen is thick, as opposed to Lamb waves in thin sheet specimens. If, as shown, the laser 14 is incident on the same side of the sheet as the
sensing beams 1 and 2, the times of arrival at A and B of singly-reflected bulk waves, tl and t2, enable an unknown thickness of a sheet to be eliminated, so that wave velocity can be determined in terms of the beam separation (A-B) and the distance (B-16) to the laser beam 16. By observing also the arrival times of Rayleigh waves, or of triply-reflected compression waves, the unknown distance (B-16) can also be eliminated.
It will hence be appreciated that to obtain accurate values of wave velocity we need a very accurate measurement of the beam separation (A-B). This may be obtained by using a test plate, whose thickness is accurately known or may be measured for example with a micrometer. From observed arrival times tl and t2 of singly-reflected bulk waves in this test plate, the separation A-B can be determined.
Not only do such measurements enable the wave velocity of bulk waves to be determined, but also the thickness of the. sheet. Measurements of wave velocity may be related to other material properties , such as temperature, as may measurements of thickness. Additional information may also be obtained by measuring the arrival times for surface compression waves.
Some experimental measurements made with the apparatus 10 are shown in the table. As an initial measurement, the separation between the laser beams (A-B) was found to be 29.1 mm. The measurements were made on an aluminium block of uniform thickness 20.00 mm.. The block was removed and then replaced between successive measurements. The calculated parameters in each case are the distance D to the laser (i.e. distance B-16), the velocity of Rayleigh waves (Vr) , the velocity of bulk waves (Vp) , and the apparent thickness (d) .
Parameter test 1 test 2 test 3 test 4
D/mm 25.89 25.84 25.88 25.81
-1
Vr/m s 2928 2925 6312 6288
Vr/m s"1 6304 6294 6312 • 6288
d/mm 20.05 20.01 20.06 19.98
It will be appreciated that the apparatus 10 enables accurate measurements to be made of the parameters, and in particular it is evident that the thickness can be measured to an accuracy better than 0.1 mm.
If it is desired to monitor several different wave modes at each location (A and B), then the signal traces will contain several successive peaks. Depending on the separation of the locations A and B, the later peaks from the location B may overlap with the earlier peaks at the location A. In this case the beams 1 or 2 may be alternately obstructed by a shutter, so that at any one time only the signals from one location are observed. Such a shutter may operate at the same frequency as the pulses emitted by the laser 14. Alternatively, one beam 1 may be allowed to reach the detector 38 continuously, whereas the other beam 2 is alternately obstructed. The signal traces produced with several successive pulses from the laser 14 are preferably averaged, for example using a computer, before being analyzed. If one beam (e.g. beam 2) is obstructed during alternate flashes from the laser 14, then the alternate signal traces would be averaged separately, so giving an average signal trace for all occasions in which both beams 1 and 2 were received, and another average signal trace for all occasions of which only beam 1 was received.