- Email: [email protected]
Pulsed lasers in electronic speckle pattern interferometry
Pulsed lasers in electronic speckle pattern interferometry T.J. COOKSON,
J.N. BUTTERS,
H.C. POLLARD
This paper describes the incorporation of a pulsed ruby laser into an electronic speckle pattern interferometer. A technique is described for observing vibrational and transient events with a double pulsed laser and some typical results are given. Results of the application of the interferometer to non-destructive testing are included.
Pulsed lasers overcome one of the major disadvantages of conventional holography using a continuous-wave laser: extraneous mechanical and airborne disturbances. A Q-switched ruby laser has a pulse width of about 20-50 ns, depending on the configuration of the oscillator section, allowing transient events and objects moving at up to 2 m s-l to be observed with the ESPI unit. The use of a ruby laser to form double-pulsed holograms of a running diesel engine (Fig. 1) illustrates the ability of such a system to function in an environment where the noise level was of the order of 120 dBA and the low frequency rocking of the engine on its rubber mountings had an amplitude of approximately 3 mm. The laser pulse separation was 10 ys. Fringes can be seen on various components indicating their movement relative to the body of the engine. Electronic speckle pattern interferometry (ESPI)’ enables interferograms to be produced almost instantaneously. By coupling the very short exposure time of a pulsed laser with television processing (ESPI) we have an instrument capable of working in an industrial environment without the need for mechanical isolation and photographic processing. The development of the pulsed laser electronic speckle pattern interferometer has depended on the performance of the pulsed lasers that have been available at Loughborough. Modifications to the lasers are described. The principle of addition of speckle patterns along with electronic filtering is outlined in the following section.
The application of double pulsed interferometry television processing
to
The observation of vibrational events using a pulsed laser is performed holographically by effectively a double exposure technique. A pulsed laser can be Q-switched twice within one flash tube cycle to produce two laser pulses separated by anything from 10 ps to 1 ms. The change in vibrational amplitude between laser pulses produces fringes when the superimposed holograms are reconstructed. With television The authors are with the Department of Mechanical Engineering, University of Technology, Loughborough, UK. Received 24 October 1977.
0030-3992/78/1003-0119 OPTICS AND LASER TECHNOLOGY.
JUNE 1978
$02.00
Fig. 1 Double pulsed hologram of a single cylinder diesel engine running at 2 400 rpm. The laser pulse separation was 15 PS
processing displacement of an object has been observed by electronic subtraction of the speckle patterns produced by the two states of the object.’ This technique can be used with the pulsed laser by recording speckle patterns of first the stationary object and then the vibrating object, subtracting the two patterns to produce amplitude fringes. High stability of the object in all but the resonant vibration mode between the two recordings is necessary because the two pulses wiII need to be at least ten seconds apart with our JK System 2000 laser to allow for charging and settling time. This delay negates the advantage of the double pulse technique which is normally insensitive to rigid body move0
1978 IPC Business Press 119
ment because of the short time interval (< exposures (see Fig. 1, for example).
1 ms) between
Opto-electronic addition of speckle patterns, followed by high pass filtering, as proposed by Pollard, overcomes the above difficulty with television processing. The principle of opto-electronic addition relies on the ability of the target of the vidicon (standard EMI Type No. 9877) in the television camera to retain the light-induced distribution of charge on its surface until erased by the scanning electron beam. With the laser set to the double pulse mode, the two speckle patterns each induce a charge distribution on the surface of the target which add together. The addition of speckle patterns produces regions of high intensity modulation, where the phase of speckles is changed by integral multiples of 27r, and regions of low intensity modulation, where the phase of speckles is changed by odd multiples of rr. High pass electronic filtering of the video signal from the camera removes the regions of low intensity modulation which then appear as ‘dark’ fringes in the resulting speckle pattern interferogram. The filtering technique only produces high contrast fringes if an unmodulated reference beam is used in conjunction with an in-line focused image arrangement. An advantage of opto-electronic addition is that an expensive video store with its associated distortions of the recorded speckle pattern is not required, the two speckle patterns being perfectly superimposed on the surface of the vidicon’s target, regardless of the geometry and stability of the camera. A relatively inexpensive video tape recorder can be used to display the resulting speckle pattern interferogram. Experiments
with a modified
LU6
laser
Initially a modified Barr and Stroud LU6 pulsed ruby laser was incorporated in a standard electronic speckle pattern interferometer (Fig. 2). An 80 mm diameter disc formed the object and it was forcibly vibrated from two to five kilohertz by a piezo-electric crystal bonded to its surface. To ease alignment of the reference beam the pinhole spatial filter was omitted. Triggering of the laser was accomplished
-l--Position us,“g
of reference J.K.Loser
beom
by selecting a single field synchronization pulse, inverting it and feeding it to the laser trigger circuit. This technique ensured that the laser pulse(s) occurred at the beginning of a television frame. The first results from this speckle system gave a ‘speckle’ image but no visible fringes (laser set to the double pulse mode). A short period of holographic work with this laser had shown it capable of producing holograms but its beam shape was poor, indicating a lack of spatial coherence due to filamentary lasing. The cure for this problem is to aperture the ruby rod, so limiting lasing to a smaller section of the rod. An aperture of 2 mm diameter was found to be the best compromise between beam shape and output energy, the latter now being at least a factor of ten below the unapertured value. After this modification to the laser, the first pulsed laser speckle pattern interferograms were produced. Figure 3 is a photograph of a typical interferogram of the vibrating disc, as displayed on a television monitor. The spatial filter had now been included in the reference beam optics. Experiments
with a IJ ruby laser
The modified LU6 laser produced an output of 5 to 10 mJ (with the apertured ruby rod), enabling an object of 80 mm by 60 mm to be observed with the camera operating at its limit of gain. To study larger areas and to evaluate the system with another laser, a 1J ruby laser (JK Lasers System 2000) was included in the electronic speckle pattern interferometer. This laser is an oscillator-amplifier combination consisting of an oscillator section producing 25 to 30 mJ and two amplifiers to raise the output to 1J. The first experiments with this laser showed that intensity variations across the beam (after spatial filtering) were too great to give the smooth reference wavefront required by the interferometer. Two observations led to this conclusion. Firstly, the video signal from the camera indicated that variations in intensity of the reference beam exceeded the dynamic range of the vidicon. This made the observation of fringes difficult when testing the system with the
when
Object Lens/aperture
Video
sagnot
,
Video tape recorder or storage tube
High filter
poss ----O
I
Fig. 2 Schematic diagram of the optical and electronic arrangements of the pulsed laser electronic speckle pattern interferometer
120
OPTICS
AND
LASER
TECHNOLOGY.
JUNE
1978
vibrating disc. Secondly, tests with burn paper indicated that the shape of the laser’s output beam was of a rectangular form.‘The outputs of the oscillator and first amplifier were found to be Gaussian in shape, suggesting that interference was occurring within the final amplifier rod to produce the uneven output. The interference effect is due to the beam, which passes through a diverging lens before the amplification stages, filling the final rod so that parts are reflected by the sides of the rod and interfere with the undeviated parts of the beam. In order to maintain the output of the laser it was decided to split off part of the oscillator output before it entered the amplifiers. This technique has been used by previous workerszY 3 to provide a smooth reference beam for holographic purposes. However, with this arrangement they observed that the peak brightness of their holograms did not occur at zero path length difference (path length difference = [length of object beam path] - [length of reference beam path] ) but at about + 1.5 m path length difference. Siebertj suegests that this uhenomena is due to a freauencv shift withii:he amplifiers: He calculates that a shift of ’ 17 MHz (allied with a frequency sweep of the oscillator output) is necessary to give the peak in hologram brightness at a path length difference of + 1.5 m for the particular ruby laser he was using. To determine this possible path length mismatch for our ruby laser a holographic arrangement using a reference beam split off the oscillator output was constructed. A series of holograms was made with path length difference ranging from 0 to + 2 m. On measuring the relative brightness of these holograms the peak was found to occur at a path length difference of zero, contrary to the results of Siebert.3 Repeated experiments gave the same result. It was therefore concluded that no net phase shift was occurring between the oscillator output and main output of ,..._ TV I,.“.,..
Fig. 3 Double pulsed speckle pattern interferogram of a circular plate resonating at 4.3 kHz. Laser pulse separation was 20 Ps
modes of vibration in the side of the steel cabinet. Photographs of the interferograms produced appear in Figs 4a, 4b and 4c. The fringes in the interferograms represent the vibrational modes of the cabinet side occurring at motor speeds of 2000,300O and 7000 rpm respectively. The laser pulse separation was 270 ~_ls.The interferogram reproduced in Fig. 4d represents the vibrational mode of the side of the cabinet when struck centrally and the laser fired a short time later (transient event). Pulse separation was 155 ps. The area of view of the cabinet side was approximately 400 mm bvd 300 mm.
The electronic speckle pattern interferometer was now reassembled with the reference beam derived from the intercavity beam splitter. A force vibrated disc 150 mm in diameter was used as a test object. No difficulties were encountered in obtaining high contrast fringes of the vibrating disc. To further confirm the results of the holographic tests described above, concerning a possible frequency shift of the object beam, a series of interferograms were made with a range of path length differences (’ 0.5 m). Again we found that maximum fringe contrast occurred at zero path length difference. Our conclusion that this laser does not exhibit a detectable frequency shift was thus again confirmed. Results of tests using a pulsed laser speckle interferometer In order to allow a realistic assessment of the performance of the interferometer the objects under investigation have not been specially made or adapted. Surface treatment usually consisted of a light coat of matt white paint to eliminate specular reflections which would otherwise flood the television camera. Vibrational
tests
The vibrational behaviour of the side of a steel cabinet was studied. The cabinet was force vibrated by an electric motor standing on top of the cabinet. The laser pulses were synchronized with a signal picked up from the shaft of the electric motor whose speed was varied to excite different
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
’
The transient behaviour of the cone of a 200 mm loudspeaker has been investigated with the interferometer. A 2.4 V pulse was applied to the loudspeaker a few microseconds after the first laser pulse had illuminated the cone. The second laser pulse was timed to arrive some tens of microseconds later. In this way the transient behaviour of the cone could be studied by observing the fringe pattern variation with time after the voltage pulse had been applied to the loudspeaker. Figure 5a is an interferogram of the loudspeaker cone where the second laser pulse occurred 75 /.LSafter the voltage pulse was applied to the loudspeaker. The outer black circle, common to the two other interferograms, is the cone suspension which had not been treated-with whitener and hence does not show up on the interferograms. The circular fringe close to the centre of the cone shows that the latter isnot moving in a piston-like manner in response to the input voltage pulse. Increasing the time delay of the second laser pulse from 75 ~_tsto 83 /JS gave rise to the interferogram shown in Fig. 5b. In this interferogram we see that most of the outer part of the cone is covered by a dark fringe, showing that this area is moving bodily. The circular (light) fringe in the central region of the cone indicates the departure of the cone movement from the expected bodily motion (piston action). The remainder of this series of interferograms suggest that not only is the cone moving bodily but that also a low amplitude transverse wave is travelling out from the centre of the cone, somewhat analogous to the behaviour of a metal plate when struck centrally.4 At a time of 140 /.LSafter applying the voltage
121
Fig. 4 Double pulsed interferograms showing the vibrational modes of the side of a steel cabinet being force vibrated by an electric motor running at: a - 2000 rpm; b - 3000 rpm; c - 7000 rpm. Laser pulse separation was 270 ps. Figure 4d shows the vibrational mode of the side of the cabinet when struck centrally. Laser pulse separation was 155 ~1s
pulse (Fig. 5e) the transverse wave has been reflected by the edge of the cone and has interfered with the outgoing disturbance. This is illustrated by the distortion of the shape of the outer fringe. The appearance of these interferograms is different from those of Fig. 4 (the cabinet side) because the early experiments utilized different forms of video frame hold. The former were recorded on an experimental video storage tube whereas the latter were recorded directly onto a video tape recorder which was played back in its still frame mode for viewing and photography. The video storage tube records one frame of the video signal and is arranged to fire the laser at the start of the recording cycle.
122
Non-destructive tests
For non-destructive tests where a static load or a thermal transient is applied to an object the short duration double pulsed technique cannot be used. The speckle patterns of the object in the reference state and the stressed state are recorded on adjacent charnels of a video disc recorder (see Fig. 2). The two speckle patterns can then be subtracted to give the interferogram. The firing of the laser is synchronized with the recording cycle at the beginning of a television frame. The distortion of a sheet steel cabinet door was investigated with the subtraction technique. A small force was applied to the top left hand corner of the door between the two firings of the laser. On subtraction of the two recorded speckle patterns the fringes depicted in Fig. 6a were obtained. These regular fringes indicate a smooth displacement of the door. A spot weld was then drilled out of a box section reinforcing panel which was welded to the rear of the door. The test was then repeated giving the result in Fig. 6b. The contours of the fringes indicate a change in behaviour of the panel under stress. The almost circular fringes in the pattern are approximately coincident with the
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
position of the missing spot weld. The area of the door observed by the speckle system was 400 mm x 300 mm.
Conclusions The experimental work carried out has identified the optical characteristics of a pulsed ruby laser necessary for its inclusion in an electronic speckle pattern interferometer. With the ruby laser set to double pulse operation the principle of speckle pattern addition has been shown to produce high contrast fringes when observing vibrational and transient events. The application of the interferometer to non-destructive testing has been demonstrated. Using both addition and subtraction modes the ESPI equipped with a pulsed ruby laser forms a powerful tool for a wide range of optical NDT. By incorporating an ESPI in a pulsed laser holocamera the range of application of the camera is considerably increased because interferograms are displayed in 40 ms with the ability to update as fast as the laser will pulse (typically one test every ten seconds).
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
Fig. 5 Double pulsed interferograms showing the displacement of the cone of a loudspeaker: a - 75 IIS; b - 83 us; c - 110 ps; d - 120 ps; e - 140 ps after the application of a 2.4 V step function to the input of the loudspeaker. The first laser pulse was timed to occur before the voltage step and the second laser pulse at the time delays stated above
123
lnterferograms of a steel door produced by the subtraction technique. a - shows fringes produced by applying a small force to Fig. 6 the top left hand corner of the door. To produce interferogram b a spot weld has been drilled out of a box reinforcing panel welded to the rear of the door and the same test performed as in photograph a
Acknowledgements We should like to thank the technical and photographic staff of the Department of Mechanical Engineering for their valuable assistance. The work has been supported in part by the National Engineering Laboratory and in part by the Science Research Council.
References 1
Butters, J.N., Leendertz, (1971) 349-354
2 3
Ansley,
D.A. Appl Opt 9 (1970)
Siebert, L.D. Appl Opt 10
(1971)
and Control 4
815-821 632-637
Aprahamian, R., Evenson, D.A., Mixon, J.S., Jacoby, ExperimentalMechanics 11 (1971) 357-362
4
To:
J.A. J Measurement
David Burt, IPC House, GU13EW. Telephone:
J.L.
IPC Science and Technology Press Ltd. 32 High Street, Guildford. Surrey, England. (0483) 71661 Telex: 859556 Scitec G.
Please send me further details on COMPUTER COMMUNICATIONS
I would like a sample copy 0 Name Organisation
124
and address
OPTICS
AND
LASER
TECHNOLOGY.
JUNE
1978
J.N. BUTTERS,
H.C. POLLARD
This paper describes the incorporation of a pulsed ruby laser into an electronic speckle pattern interferometer. A technique is described for observing vibrational and transient events with a double pulsed laser and some typical results are given. Results of the application of the interferometer to non-destructive testing are included.
Pulsed lasers overcome one of the major disadvantages of conventional holography using a continuous-wave laser: extraneous mechanical and airborne disturbances. A Q-switched ruby laser has a pulse width of about 20-50 ns, depending on the configuration of the oscillator section, allowing transient events and objects moving at up to 2 m s-l to be observed with the ESPI unit. The use of a ruby laser to form double-pulsed holograms of a running diesel engine (Fig. 1) illustrates the ability of such a system to function in an environment where the noise level was of the order of 120 dBA and the low frequency rocking of the engine on its rubber mountings had an amplitude of approximately 3 mm. The laser pulse separation was 10 ys. Fringes can be seen on various components indicating their movement relative to the body of the engine. Electronic speckle pattern interferometry (ESPI)’ enables interferograms to be produced almost instantaneously. By coupling the very short exposure time of a pulsed laser with television processing (ESPI) we have an instrument capable of working in an industrial environment without the need for mechanical isolation and photographic processing. The development of the pulsed laser electronic speckle pattern interferometer has depended on the performance of the pulsed lasers that have been available at Loughborough. Modifications to the lasers are described. The principle of addition of speckle patterns along with electronic filtering is outlined in the following section.
The application of double pulsed interferometry television processing
to
The observation of vibrational events using a pulsed laser is performed holographically by effectively a double exposure technique. A pulsed laser can be Q-switched twice within one flash tube cycle to produce two laser pulses separated by anything from 10 ps to 1 ms. The change in vibrational amplitude between laser pulses produces fringes when the superimposed holograms are reconstructed. With television The authors are with the Department of Mechanical Engineering, University of Technology, Loughborough, UK. Received 24 October 1977.
0030-3992/78/1003-0119 OPTICS AND LASER TECHNOLOGY.
JUNE 1978
$02.00
Fig. 1 Double pulsed hologram of a single cylinder diesel engine running at 2 400 rpm. The laser pulse separation was 15 PS
processing displacement of an object has been observed by electronic subtraction of the speckle patterns produced by the two states of the object.’ This technique can be used with the pulsed laser by recording speckle patterns of first the stationary object and then the vibrating object, subtracting the two patterns to produce amplitude fringes. High stability of the object in all but the resonant vibration mode between the two recordings is necessary because the two pulses wiII need to be at least ten seconds apart with our JK System 2000 laser to allow for charging and settling time. This delay negates the advantage of the double pulse technique which is normally insensitive to rigid body move0
1978 IPC Business Press 119
ment because of the short time interval (< exposures (see Fig. 1, for example).
1 ms) between
Opto-electronic addition of speckle patterns, followed by high pass filtering, as proposed by Pollard, overcomes the above difficulty with television processing. The principle of opto-electronic addition relies on the ability of the target of the vidicon (standard EMI Type No. 9877) in the television camera to retain the light-induced distribution of charge on its surface until erased by the scanning electron beam. With the laser set to the double pulse mode, the two speckle patterns each induce a charge distribution on the surface of the target which add together. The addition of speckle patterns produces regions of high intensity modulation, where the phase of speckles is changed by integral multiples of 27r, and regions of low intensity modulation, where the phase of speckles is changed by odd multiples of rr. High pass electronic filtering of the video signal from the camera removes the regions of low intensity modulation which then appear as ‘dark’ fringes in the resulting speckle pattern interferogram. The filtering technique only produces high contrast fringes if an unmodulated reference beam is used in conjunction with an in-line focused image arrangement. An advantage of opto-electronic addition is that an expensive video store with its associated distortions of the recorded speckle pattern is not required, the two speckle patterns being perfectly superimposed on the surface of the vidicon’s target, regardless of the geometry and stability of the camera. A relatively inexpensive video tape recorder can be used to display the resulting speckle pattern interferogram. Experiments
with a modified
LU6
laser
Initially a modified Barr and Stroud LU6 pulsed ruby laser was incorporated in a standard electronic speckle pattern interferometer (Fig. 2). An 80 mm diameter disc formed the object and it was forcibly vibrated from two to five kilohertz by a piezo-electric crystal bonded to its surface. To ease alignment of the reference beam the pinhole spatial filter was omitted. Triggering of the laser was accomplished
-l--Position us,“g
of reference J.K.Loser
beom
by selecting a single field synchronization pulse, inverting it and feeding it to the laser trigger circuit. This technique ensured that the laser pulse(s) occurred at the beginning of a television frame. The first results from this speckle system gave a ‘speckle’ image but no visible fringes (laser set to the double pulse mode). A short period of holographic work with this laser had shown it capable of producing holograms but its beam shape was poor, indicating a lack of spatial coherence due to filamentary lasing. The cure for this problem is to aperture the ruby rod, so limiting lasing to a smaller section of the rod. An aperture of 2 mm diameter was found to be the best compromise between beam shape and output energy, the latter now being at least a factor of ten below the unapertured value. After this modification to the laser, the first pulsed laser speckle pattern interferograms were produced. Figure 3 is a photograph of a typical interferogram of the vibrating disc, as displayed on a television monitor. The spatial filter had now been included in the reference beam optics. Experiments
with a IJ ruby laser
The modified LU6 laser produced an output of 5 to 10 mJ (with the apertured ruby rod), enabling an object of 80 mm by 60 mm to be observed with the camera operating at its limit of gain. To study larger areas and to evaluate the system with another laser, a 1J ruby laser (JK Lasers System 2000) was included in the electronic speckle pattern interferometer. This laser is an oscillator-amplifier combination consisting of an oscillator section producing 25 to 30 mJ and two amplifiers to raise the output to 1J. The first experiments with this laser showed that intensity variations across the beam (after spatial filtering) were too great to give the smooth reference wavefront required by the interferometer. Two observations led to this conclusion. Firstly, the video signal from the camera indicated that variations in intensity of the reference beam exceeded the dynamic range of the vidicon. This made the observation of fringes difficult when testing the system with the
when
Object Lens/aperture
Video
sagnot
,
Video tape recorder or storage tube
High filter
poss ----O
I
Fig. 2 Schematic diagram of the optical and electronic arrangements of the pulsed laser electronic speckle pattern interferometer
120
OPTICS
AND
LASER
TECHNOLOGY.
JUNE
1978
vibrating disc. Secondly, tests with burn paper indicated that the shape of the laser’s output beam was of a rectangular form.‘The outputs of the oscillator and first amplifier were found to be Gaussian in shape, suggesting that interference was occurring within the final amplifier rod to produce the uneven output. The interference effect is due to the beam, which passes through a diverging lens before the amplification stages, filling the final rod so that parts are reflected by the sides of the rod and interfere with the undeviated parts of the beam. In order to maintain the output of the laser it was decided to split off part of the oscillator output before it entered the amplifiers. This technique has been used by previous workerszY 3 to provide a smooth reference beam for holographic purposes. However, with this arrangement they observed that the peak brightness of their holograms did not occur at zero path length difference (path length difference = [length of object beam path] - [length of reference beam path] ) but at about + 1.5 m path length difference. Siebertj suegests that this uhenomena is due to a freauencv shift withii:he amplifiers: He calculates that a shift of ’ 17 MHz (allied with a frequency sweep of the oscillator output) is necessary to give the peak in hologram brightness at a path length difference of + 1.5 m for the particular ruby laser he was using. To determine this possible path length mismatch for our ruby laser a holographic arrangement using a reference beam split off the oscillator output was constructed. A series of holograms was made with path length difference ranging from 0 to + 2 m. On measuring the relative brightness of these holograms the peak was found to occur at a path length difference of zero, contrary to the results of Siebert.3 Repeated experiments gave the same result. It was therefore concluded that no net phase shift was occurring between the oscillator output and main output of ,..._ TV I,.“.,..
Fig. 3 Double pulsed speckle pattern interferogram of a circular plate resonating at 4.3 kHz. Laser pulse separation was 20 Ps
modes of vibration in the side of the steel cabinet. Photographs of the interferograms produced appear in Figs 4a, 4b and 4c. The fringes in the interferograms represent the vibrational modes of the cabinet side occurring at motor speeds of 2000,300O and 7000 rpm respectively. The laser pulse separation was 270 ~_ls.The interferogram reproduced in Fig. 4d represents the vibrational mode of the side of the cabinet when struck centrally and the laser fired a short time later (transient event). Pulse separation was 155 ps. The area of view of the cabinet side was approximately 400 mm bvd 300 mm.
The electronic speckle pattern interferometer was now reassembled with the reference beam derived from the intercavity beam splitter. A force vibrated disc 150 mm in diameter was used as a test object. No difficulties were encountered in obtaining high contrast fringes of the vibrating disc. To further confirm the results of the holographic tests described above, concerning a possible frequency shift of the object beam, a series of interferograms were made with a range of path length differences (’ 0.5 m). Again we found that maximum fringe contrast occurred at zero path length difference. Our conclusion that this laser does not exhibit a detectable frequency shift was thus again confirmed. Results of tests using a pulsed laser speckle interferometer In order to allow a realistic assessment of the performance of the interferometer the objects under investigation have not been specially made or adapted. Surface treatment usually consisted of a light coat of matt white paint to eliminate specular reflections which would otherwise flood the television camera. Vibrational
tests
The vibrational behaviour of the side of a steel cabinet was studied. The cabinet was force vibrated by an electric motor standing on top of the cabinet. The laser pulses were synchronized with a signal picked up from the shaft of the electric motor whose speed was varied to excite different
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
’
The transient behaviour of the cone of a 200 mm loudspeaker has been investigated with the interferometer. A 2.4 V pulse was applied to the loudspeaker a few microseconds after the first laser pulse had illuminated the cone. The second laser pulse was timed to arrive some tens of microseconds later. In this way the transient behaviour of the cone could be studied by observing the fringe pattern variation with time after the voltage pulse had been applied to the loudspeaker. Figure 5a is an interferogram of the loudspeaker cone where the second laser pulse occurred 75 /.LSafter the voltage pulse was applied to the loudspeaker. The outer black circle, common to the two other interferograms, is the cone suspension which had not been treated-with whitener and hence does not show up on the interferograms. The circular fringe close to the centre of the cone shows that the latter isnot moving in a piston-like manner in response to the input voltage pulse. Increasing the time delay of the second laser pulse from 75 ~_tsto 83 /JS gave rise to the interferogram shown in Fig. 5b. In this interferogram we see that most of the outer part of the cone is covered by a dark fringe, showing that this area is moving bodily. The circular (light) fringe in the central region of the cone indicates the departure of the cone movement from the expected bodily motion (piston action). The remainder of this series of interferograms suggest that not only is the cone moving bodily but that also a low amplitude transverse wave is travelling out from the centre of the cone, somewhat analogous to the behaviour of a metal plate when struck centrally.4 At a time of 140 /.LSafter applying the voltage
121
Fig. 4 Double pulsed interferograms showing the vibrational modes of the side of a steel cabinet being force vibrated by an electric motor running at: a - 2000 rpm; b - 3000 rpm; c - 7000 rpm. Laser pulse separation was 270 ps. Figure 4d shows the vibrational mode of the side of the cabinet when struck centrally. Laser pulse separation was 155 ~1s
pulse (Fig. 5e) the transverse wave has been reflected by the edge of the cone and has interfered with the outgoing disturbance. This is illustrated by the distortion of the shape of the outer fringe. The appearance of these interferograms is different from those of Fig. 4 (the cabinet side) because the early experiments utilized different forms of video frame hold. The former were recorded on an experimental video storage tube whereas the latter were recorded directly onto a video tape recorder which was played back in its still frame mode for viewing and photography. The video storage tube records one frame of the video signal and is arranged to fire the laser at the start of the recording cycle.
122
Non-destructive tests
For non-destructive tests where a static load or a thermal transient is applied to an object the short duration double pulsed technique cannot be used. The speckle patterns of the object in the reference state and the stressed state are recorded on adjacent charnels of a video disc recorder (see Fig. 2). The two speckle patterns can then be subtracted to give the interferogram. The firing of the laser is synchronized with the recording cycle at the beginning of a television frame. The distortion of a sheet steel cabinet door was investigated with the subtraction technique. A small force was applied to the top left hand corner of the door between the two firings of the laser. On subtraction of the two recorded speckle patterns the fringes depicted in Fig. 6a were obtained. These regular fringes indicate a smooth displacement of the door. A spot weld was then drilled out of a box section reinforcing panel which was welded to the rear of the door. The test was then repeated giving the result in Fig. 6b. The contours of the fringes indicate a change in behaviour of the panel under stress. The almost circular fringes in the pattern are approximately coincident with the
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
position of the missing spot weld. The area of the door observed by the speckle system was 400 mm x 300 mm.
Conclusions The experimental work carried out has identified the optical characteristics of a pulsed ruby laser necessary for its inclusion in an electronic speckle pattern interferometer. With the ruby laser set to double pulse operation the principle of speckle pattern addition has been shown to produce high contrast fringes when observing vibrational and transient events. The application of the interferometer to non-destructive testing has been demonstrated. Using both addition and subtraction modes the ESPI equipped with a pulsed ruby laser forms a powerful tool for a wide range of optical NDT. By incorporating an ESPI in a pulsed laser holocamera the range of application of the camera is considerably increased because interferograms are displayed in 40 ms with the ability to update as fast as the laser will pulse (typically one test every ten seconds).
OPTICS AND LASER TECHNOLOGY.
JUNE 1978
Fig. 5 Double pulsed interferograms showing the displacement of the cone of a loudspeaker: a - 75 IIS; b - 83 us; c - 110 ps; d - 120 ps; e - 140 ps after the application of a 2.4 V step function to the input of the loudspeaker. The first laser pulse was timed to occur before the voltage step and the second laser pulse at the time delays stated above
123
lnterferograms of a steel door produced by the subtraction technique. a - shows fringes produced by applying a small force to Fig. 6 the top left hand corner of the door. To produce interferogram b a spot weld has been drilled out of a box reinforcing panel welded to the rear of the door and the same test performed as in photograph a
Acknowledgements We should like to thank the technical and photographic staff of the Department of Mechanical Engineering for their valuable assistance. The work has been supported in part by the National Engineering Laboratory and in part by the Science Research Council.
References 1
Butters, J.N., Leendertz, (1971) 349-354
2 3
Ansley,
D.A. Appl Opt 9 (1970)
Siebert, L.D. Appl Opt 10
(1971)
and Control 4
815-821 632-637
Aprahamian, R., Evenson, D.A., Mixon, J.S., Jacoby, ExperimentalMechanics 11 (1971) 357-362
4
To:
J.A. J Measurement
David Burt, IPC House, GU13EW. Telephone:
J.L.
IPC Science and Technology Press Ltd. 32 High Street, Guildford. Surrey, England. (0483) 71661 Telex: 859556 Scitec G.
Please send me further details on COMPUTER COMMUNICATIONS
I would like a sample copy 0 Name Organisation
124
and address
OPTICS
AND
LASER
TECHNOLOGY.
JUNE
1978