U N I V E R S I T Y
O F T E C H N O L O G Y L U L E Å P
L
2 0 0 0 : 3 3T") O P T O R AT T H F S I S
A Pulsed TV Holography System for the Study of Transients in
Experimental Mechanics
PER GREN
D e p a r t m e n t o f M e c h a n i c a l E n g i n e e r i n g
Doctoral Thesis 2000:33
A pulsed T V holography system for the study of transients in
experimental mechanics
by
Per Gren
D i v i s i o n o f Experimental Mechanics Department o f Mechanical Engineering
Luleå University o f Technology SE-971 87 Luleå
Sweden
D i s s e r t a t i o n
f o r the degree o f D o c t o r o f Philosophy (Ph.D.) i n the subject area Experimental Mechanics w h i c h w i t h due permission o f the Faculty board at the Luleå University o f Technology w i l l be defended i n public, i n r o o m E243 at the Luleå University o f Technology, o n Friday the 8
t ho f December, at 10.30 am.
Extarnal examiner: Associate Professor T o r g n y Carlsson, R o y a l Institute o f Technology, Stockholm, Sweden.
A k a d e m i s k a v h a n d l i n g
för avläggande av teknologie doktorsexamen, som med vederbörligt tillstånd av Tekniska f a k u l t e t s n ä m n d e n v i d Luleå tekniska universitet kommer att offentligen försvaras, i sal E243 v i d Luleå tekniska universitet, fredagen den 8 december, k l . 10.30.
Fakultetsopponent är D o c e n t T o r g n y Carlsson, K T H , Stockholm, Sverige.
Doctoral Thesis 2000:33
A pulsed T V holography system for the study of transients in
experimental mechanics
Per Gren
D i v i s i o n o f Experimental Mechanics Department o f Mechanical Engineering
Luleå University o f T e c h n o l o g y SE-971 87 Luleå
Sweden
Luleå 2000
K e y words: optical metrology, holographic interferometry,
pulsed T V holography, pulsed digital holography,
transient wave propagation, acoustic pressure fields
Preface
This w o r k has been carried out i n parallel w i t h the w o r k as a teacher d u r i n g the years 1990-2000 at the D i v i s i o n o f Experimental Mechanics, Department o f Mechanical Engineering at Luleå University o f Technology, Sweden.
I w o u l d like to thank m y supervisors, Professor N i l s - E r i k M o l i n and Associate Professor Anders W å h l i n w h o introduced me i n t o the field o f holographic interferometry and gave me professional support and guidance, w h i c h is highly appreciated.
Special thanks also to D r Staffan Schedin, for f r u i t f u l co-operation b o t h i n the development o f the pulsed T V - h o l o g r a p h y system and the transient acoustic fields studies (paper C - G )
M a n y thanks to the f o l l o w i n g co-authors:
D r . X i d e L i , China, co-author o f paper E and guest researcher at our division Jan-June 1997. His knowledge i n tomography is highly appreciated.
D r . Marie F i n n s t r ö m , D i v i s i o n o f Physics, L T U , colleague and co-author o f paper G. H e r knowledge i n compressible fluid dynamics is highly appreciated.
D r . R o b i n Olsson, co-author o f paper A . H e is a specialist i n composite materials.
Thanks are also to D r . O l o v M a r k l u n d , Division o f Industrial Electronics L T U , f o r help w i t h an u n w r a p p i n g algorithm.
I w o u l d also like to thank m y other colleagues at the department f o r help and discussions d u r i n g this w o r k .
T h e Swedish Research C o u n c i l f o r Engineering Sciences ( T F R ) has supported
this project and is gratefully acknowledged. T h e n e w system f o r Pulsed T V
holography i n c l u d i n g the sophisticated pulsed N d : Y A G laser and auxiliary
equipment has been financed by the K n u t and Alice Wallenberg Foundation.
Abstract
A n all-electronic system f o r pulsed holographic interferometry called pulsed T V holography is developed. This is a w h o l e - f i e l d non-contacting optical measurement method suitable f o r studies o f transient events like wave propagation i n solids and fluids. Chemical w e t processing o f holographic film and optical reconstruction o f holograms are no longer needed.
T h e technique was first developed using a double pulsed ruby laser as light source. The holograms are recorded directly o n a C C D - d e t e c t o r . Quantitative data o f changes i n optical path length, caused either by a deformation o f a solid object or a change i n refractive index i n a f l u i d , are calculated directly i n a computer. T h e system f o r pulsed T V holography has recently been further developed by the purchase o f a n e w pulsed laser ( t w i n cavity, injection seeded pulsed N d : Y A G ) and a C C D camera ( P C O Sensicam) w i t h higher spatial resolution and dynamic range. I n the survey o f this thesis the increased versatility' compared to a r u b y laser based system is discussed.
D u r i n g the development o f the pulsed T V holography system a number o f experiments i n mechanics and acoustics have been accomplished. B e n d i n g waves i n impacted plates propagating at a speed o f about 2000 m/s are easily
" f r o z e n " due to the short duration laser pulses (<30 ns). These waves act as supersonic travelling acoustic sources and generate sound waves i n the surrounding air. For the first time, transient sound fields f r o m impacted plates have been visualised and measured using pulsed holographic interferometry. I n another experiment, we have demonstrated that the pulsed T V holography system is feasible i n combination w i t h tomography. B y recording a three- dimensional acoustic pressure field f r o m a number o f v i e w i n g directions f o l l o w e d by a tomographic reconstruction, the pressure i n any p o i n t can be calculated.
Finally, a method to restore fringes lost by large bulk motions is proposed.
This technique may become very attractive i n the study o f vibrations (preferable transient) o n m o v i n g or rotating objects.
I n conclusion, pulsed T V holography is proved to be a fast and reliable
m e t h o d to quantitatively study transients i n mechanics and acoustics. The
technique has a great potential i n experimental mechanics i n the future.
Thesis
This thesis comprises a survey and the following eight papers:
A G r e n P., Olsson R . , " D e f o r m a t i o n d u r i n g impact o n an orthotropic composite plate", I n K . Stetson and R . Pryputniewics, Editors, International Conference o n H o l o g r a m Interferometry and Speckle M e t r o l o g y , 429-434, Baltimore, N o v 5-8, (1990).
B A . W å h l i n , P. Gren and N - E M o l i n , " O n structure-borne sound:
Experiments showing the initial transient acoustic wave field generated by an impacted plate," J. Acoust. Soc. A m . 96(5), 2791-2797, (1994).
C S. Schedin, P. Gren and A . W å h l i n , "Transient acoustic near field i n air generated by impacted plates," J. Acoust. Soc. A m . 99(2), 700-705, (1996) .
D S. Schedin, P. Gren and A . W å h l i n , "Shock waves i n an elliptical cavity w i t h varying height," Shock waves (1997) 7: 343-350.
E S. Schedin and P. Gren, "Phase evaluation and speckle averaging in pulsed television holography," A p p l i e d Optics 36(17), 3941-3947, (1997) .
F P. G r e n , S. Schedin and X i d e L i , "Tomographic reconstruction o f transient acoustic fields recorded by pulsed T V holography," Applied Optics 37(5), 834-840, (1998).
G Staffan Schedin, Per Gren and M a r i e F i n n s t r ö m , "Measurement o f the density field around an airgun muzzle b y pulsed T V holography," i n Laser M e t r o l o g y and Inspection, Hans. J. Tiziani, Pramod K . Rastogi, Editors, Proceedings o f SPIE V o l . 3823, 13-19, (1999).
H P. Gren, "Pulsed T V holography combined w i t h D i g i t a l Speckle
Photography restores lost interference phase," Submitted to A p p l i e d
Optics.
Contents
1 Introduction 1 2 Pulsed holographic interferometry 3
2.1 Recording on film using the double-exposure method 3
2.2Electronic recording of pulsed holograms 3 2.3 Development of a pulsed T V holography system - ruby laser.. 4
3 A new system for pulsed T V holography 7 3.1 Injection-Seeded, twin Q-switched Nd:YAG laser 7
3.2 Description of our new laser system 8
3.3 The CCD camera 10 3.4 Testing of the new pulsed Nd:YAG laser 11
3.5 High sensitivity recording of bending waves using the 355nm
output 12 4 Conclusions and future work 15
5 Summary of appended papers 17
6 References 23 Appended papers
A D e f o r m a t i o n d u r i n g impact o n an orthotropic composite plate B O n structure-borne sound: Experiments showing the initial transient
acoustic wave field generated by an impacted plate
C Transient acoustic near field i n air generated by impacted plates D Shock waves i n an elliptical cavity w i t h varying height
E Phase evaluation and speckle averaging i n pulsed television holography F T o m o g r a p h i c reconstruction o f transient acoustic fields recorded b y
pulsed T V holography
G Measurement o f the density field around an airgun muzzle by pulsed T V holography
H Pulsed T V holography combined w i t h Digital Speckle Photography
restores lost interference phase
1 Introduction
T h e goal o f this project has been to develop an all-electronic measuring system f o r pulsed holographic interferometry. Such a system is o f great interest for studies o f transient events i n acoustics, solid and f l u i d mechanics.
Transient events are c o m m o n i n engineering science. Impacts o n structures, electrical discharges, explosions are examples that may cause damage to the environment or produce noise or sound. Mechanical waves are generated both i n the structures and i n the surrounding f l u i d . T h e i r time history is dependent b o t h o f the structure itself and the m e d i u m affected. M u c h research is today conducted i n order to reduce noise and vibrations i n products and increase the understanding o f h o w sound is generated f r o m a structure, f o r example a musical instrument. I n our hearing process, transient sound is important for our ability to distinguish between sound f r o m different materials [ 1 ] . Compare the differences i n the sound perceived f r o m an impact o n a wooden, metal or plastic plate. Furthermore, new materials like composites are developed w i t h n e w mechanical properties and impact responses. Demands to measure these properties and compare w i t h theoretical predictions increases.
M a n y different measurement methods are traditionally used i n mechanics and acoustics. Point measuring devises like strain gauges, accelerometers, microphones etc. are widely used. H o w e v e r , they can only provide i n f o r m a t i o n i n one point and may i n some cases disturb the measurements, especially i n cases where the structure is weak and light. Optical methods on the contrary are non-contact methods and w i l l normally not introduce any disturbances o n the field to be measured. T h e y are also often field methods.
Techniques t o study rapid events have been developed d u r i n g the last decades.
High-speed photography to record fast events like explosions and impact dynamics has advanced the understanding o f many processes [ 2 ] . C o m b i n e d w i t h the Schlieren and Shadowgraph techniques [ 3 ] , wave propagation i n transparent m e d i u m were visualised. A f t e r the discovery o f the laser i n 1960, optical holography became possible to realise. O r i g i n a l l y the principle o f holography was developed i n an attempt to increase the resolution i n electron microscopy [ 4 ] . T h e principles o f holographic interferometry were discovered soon after L e i t h and Upatnieks introduced off-axis holography i n 1961.
Powell and Stetson (1965) among others showed m u c h o f the theory, practice and potential o f this m e t h o d [5].
Holographic interferometry is an optical field m e t h o d , where t w o or more
wave fields are compared interferometrically [6,7]. W i t h this technique it is
possible to measure optical path length differences that may be caused by a
number o f physical parameters. For example, the deformation o f a solid object or the refractive index change along the light path.
Some facts about the technique are:
- T h e measurements are contactless. O n l y expanded laser light illuminates the object.
- T w o states o f an object can be compared interferometrically time separated f r o m a f e w microseconds up to weeks (depends o n the stability o f the set- up).
- D e f o r m a t i o n measurements can be performed o n rough, diffusely reflecting surfaces.
- Phase shifts caused by changes i n refractive index give rise to an interference pattern, w h i c h can be analysed to determine physical properties like density, pressure, temperature, concentration o f species. T h e phase shifts recorded o n the hologram are a projection o f the refractive index changes along the light path.
- Transient events can be studied by using pulsed lasers.
- T h e lateral dimensions o f the object may range f r o m the microscopic scale up to several meters.
I n this thesis, pulsed holographic interferometry has been developed and used to study transient wave propagation i n solids and fluids. D u r i n g the years o f this w o r k , holographic interferometry has been developed f r o m being used mostly i n research labs to be attractive also f o r industrial applications. R a p i d development o f lasers, solid state detectors like the C C D - c h i p and computers have made this possible. T h e major part i n this w o r k is the development o f a new version o f holographic interferometry, w h i c h w e call Pulsed T V holography. T h e holographic images are recorded directly o n a C C D - d e t e c t o r and stored i n a computer f o r further processing. Quantitative phase data is obtained w i t h o u t hologram w e t processing and reconstruction.
This thesis comprises a survey and eight papers. Four o f these (paper C-F) are
also included i n the thesis "Transient acoustic fields studied by pulsed T V
holography" by Staffan Schedin [8]. I n the end o f the survey a summary o f the
appended papers and future plans are presented.
2 Pulsed holographic interferometry
2.1 Recording on film using the double-exposure method
B y the use o f Q-switched pulsed ruby lasers even fast mechanical phenomenon could be studied. The ca 25ns short duration laser pulses
"freeze" most mechanical events. I n the paper by Aprahamian et.al. 1971, the use o f holographic interferometry to study propagating bending waves i n plates was demonstrated [ 9 ] . T h e y used the double exposure m e t h o d , where the first exposure records the object i n an undisturbed state and the second exposure i n a disturbed state onto the same holographic plate. I n the reconstruction o f the double-exposed hologram, cosine-shaped fringes caused by the optical path difference between the t w o recordings are superimposed o n the reconstructed image o f the object. This technique has i n this thesis been used i n paper A and B . The fringes show contours o f constant phase difference and are related to physical quantities such as displacement, rotation, strain, bending moment, temperature, pressure and mass concentration [6].
Usually, the fringes do not give i n f o r m a t i o n whether the phase change is positive or negative. B y applying the m e t h o d o f two-reference beam holographic interferometry, this sign-ambiguity can be solved [10]. The first object state is recorded by one reference beam and the second object state by another reference beam w i t h a slightly different i l l u m i n a t i o n direction (a shift o f about 0 . 0 0 5 ° ) . I n the reconstruction o f such a hologram t w o reconstruction beams via t w o mirrors resembling a Michelson interferometer set-up are used.
O n e m i r r o r is m o u n t e d o n a piezoelectric transducer and the other can be tilted to obtain the same angle between the beams as i t was i n the recording.
T h e relative phases between the t w o beams are changed i n steps o f 71/2 and f o u r phase-stepped reconstructions are recorded by a C C D camera and stored i n a computer. A four-step algorithm is then used to numerically calculate the phase difference [11]. This technique has been used i n paper B , C and D .
2.2 Electronic recording of pulsed holograms
B y using an electronic recording m e d i u m instead o f photographic film, the
t i m e consuming hologram w e t processing and optical reconstruction
procedures are avoided. T h e first attempts to record pulsed holograms were
based o n the ESPI set-up (Electronic Speckle Pattern Interferometry), where
the object and reference beams are co-linear and the object is imaged onto a
v i d i c o n o f a video camera [12,13]. T h e interference fringes were obtained
either w i t h the addition or the subtraction technique (subtraction o f t w o
images). T h e first method is experimentally easier since double pulses f r o m a
pulsed r u b y or pulsed N d : Y A G laser are recorded o n the same frame. The
second m e t h o d gives fringes w i t h higher visibility, but the images have to be recorded o n separate consecutive T V frames f o l l o w e d by a subtraction [14].
Quantitative analysis o f the fringes f r o m t w o recorded images using the spatial-carrier phase-shift m e t h o d has been proposed by Pedrini et.al. [15].
T h e images corresponding t o the t w o pulses were recorded o n separate frames o f a C C D - c a m e r a and captured by a frame-grabber. W i t h the development o f CCD-detectors w i t h more and smaller pixels, the technique to register lensless Fresnel holograms directly o n the detector has become more and more used [16-18]. I n this so-called D i g i t a l Holography technique the hologram is digitised and stored i n a computer. T h e reconstruction is performed i n the computer and contrary to the optical reconstruction b o t h the amplitude and the phase can be calculated pointwise. B y recording and reconstructing t w o states o f an object (for example before and after loading) the interference phase m o d u l o 271 can be determined by calculating the difference between the individual phases.
A n o t h e r approach is to record image-plane holograms o n a C C D -detector.
T h e reference beam is off-axis and the interference phase difference is calculated using the Fourier transform method [19]. This technique is shortly described i n the f o l l o w i n g section and i n more detail i n paper E and F.
2.3 Development of a pulsed TV holography system - ruby laser
O u r all-electronic system f o r pulsed T V holography was first developed i n combination w i t h a holographic ruby laser (Lumonics HLS3). A sketch o f the set-up is shown i n Fig. 1.
Fig. 1. T h e pulsed T V holography set-up. R L : ruby laser, N L l : negative lens,
BS: beam splitter, OBJ: object, O : object beam, L : lens system, A : aperture,
R : reference beam, N L 2 : negative lens, C C D : CCD-camera.
A negative lens ( N L l ) expands the light f o r object i l l u m i n a t i o n . T h e object (OBJ) is imaged onto the C C D detector by a lens system via the beamsplitter (BS). T h e beamsplitter allows i l l u m i n a t i o n and observation i n the same direction. A n aperture (A) i n the imaging system is used to l i m i t the spatial frequencies to be resolved by the detector. T h e small p o r t i o n o f light reflected at the flat surface o f the negative lens ( N L l ) is used as the reference beam ( R ) . T h e reference beam passes a negative lens ( N L 2 ) to spread the light u n i f o r m l y o n the detector. Thus, seen f r o m the detector the virtual image o f the reference beam appears as a p o i n t source. I t should be located at a distance f r o m the detector so that i t appears to come f r o m the same plane as that o f the aperture. I f the aperture is a rectangular single slit, the point source should laterally be positioned at a distance equal to the slit w i d t h f r o m one edge o f the slit. This is a convenient position since the interference term is separated f r o m the self-interference term o f the slit i n the Fourier domain. The dimension o f the slit must however be limited i n such a w a y that the detector resolves the interference pattern between the object and the reference beam.
T w o images o f the object are recorded o n separate frames o f the C C D - camera. T h e C C D camera ( P U L N I X T M - 9 7 0 1 ) is based o n standard T V - technique (video signal and 30 images/s) and enables progressive scanning w h i c h means that all the pixels are exposed simultaneously. I n many o f the experiments we used the vertical syncronisation o f the video signal as a trigger f o r the event and the laser. V i a a delay unit the t w o laser pulses were recorded o n either side o f the frame transfer o f the C C D - detector. I n between the t w o pulses the transient event started (for example an electrical discharge). T h e interference phase difference between the recorded images is calculated using the Fourier transform m e t h o d (see paper E, F and ref. [19]). T h e camera has the possibility o f external resetting (asynchronous reset mode), w h i c h is necessary w h e n an event is started manually (for example by releasing an impacting pendulum). H o w e v e r this input is very sensitive to electromagnetic transients and o f t e n the camera triggered at the w r o n g t i m e instant. A camera better suited f o r this application w i l l be described i n the f o l l o w i n g chapter.
O n the w h o l e the system w o r k s satisfactorily. Phase maps o f good quality can be obtained. H o w e v e r , as expected the spatial resolution and the dynamic range are n o t as good as f o r holographic recordings o n f i l m . W i t h the C C D - detector o f the P U L N I X camera (768x484 pixels) about 50 fringes can be counted over the field o f v i e w . T o get a continuous phase map, an unwrapping procedure must be used [20]. For successful u n w r a p p i n g the phase maps should contain little noise and the fringes should not be too densely spaced (<30 fringes over the field o f v i e w ) .
I n paper F it is shown that f o r a repeatable transient refractive index
experiment, recordings o f a number o f projections can be made f o l l o w e d by a
tomographic reconstruction o f the 3 D field. T h e object is rotated i n between
the recordings. Thus, each recording w i l l have the same scale w h i c h simplifies the reconstruction. Some o f these results are also published i n a Chinese j o u r n a l [21]. I n ref. [22] a symmetric sound field generated by an impacted
plate has been studied. A n air-gun bullet impacts the plate and transient bending waves propagate radially f r o m the impact point. A phase map o f the generated symmetric sound field is recorded. Quantitative values o f the pressure distribution i n any layer outside the plate can be reconstructed. T h e system has also been used to study the initial transient behaviour o f a cymbal [23]. T h e onset o f cymbal sound is thought to be strongly influenced by wave reflections at the edge, at the central dome and by the shallow concentric grooves o n the surface. Pulsed T V holography is shown to be a convenient method to study the transient behaviour o f percussian musical instruments.
The same approach is used b y a research group at the University o f Stuttgart, Germany [24]. T h e y have applied the technique f o r 3-dimensional dynamic deformation measurement using a pulsed ruby laser and one C C D camera [25]. Furthermore, experiments o n shock loaded biological tissues and the use o f endoscopes have shown n e w possible applications f o r the technique [26].
T h e ruby lasers show a number o f advantages f o r holographic interferometry.
T h e y can be double or even multiple-pulsed, by opening the electro-optical shutter (Q-switch) several times during the flash lamp p u m p i n g period [27, 28]. T h e energy output can be up to about 10 J/pulse. O u r laser can deliver 1 J i n single pulse mode and about 0.4 J+0.4 J i n double pulse mode, each pulse w i t h a duration o f about 30ns. The wavelength is 6 9 4 n m (visible, red light) w h i c h fits w e l l to the sensitivity o f photographic emulsions used i n holography. O u r experience is that the ruby laser is robust, very little maintenance is needed. There are however some drawbacks that can be listed;
- T h e limitations i n time separation between pulses f r o m the same discharge (1-800 LIS f o r our laser).
- The pulse repetition rate is l o w . I t takes about 15 seconds f o r our laser to charge the capacitors f o r the flash lamps.
- There is a d i f f i c u l t y i n maintaining equal pulse energies f r o m pulse to pulse and between pulses o f a pulse pair i n double pulsed experiments. This is annoying, since the experiments must be repeated w h e n the energies are unacceptable different.
- Reconstruction using a continuous laser cannot be made at the same wavelength as the recording.
A laser system w i t h o u t these drawbacks is described i n the next chapter.
3 A new system for pulsed T V holography
3.1 Injection-Seeded, twin Q-suntched Nd:YAG laser
A n o t h e r possible laser for pulsed holographic interferometry is a pulsed N d : Y A G laser [29]. This k i n d o f solid state laser is the most c o m m o n one, and has shown to be suitable i n applications ranging f r o m scientific research to material processing, cutting etc. i n industries. The laser is normally flash-lamp pumped and repetitively pulsed. T h e fundamental wavelength o f the laser emission is 1064 n m , but w i t h the use o f frequency conversion i n non-linear crystals, other wavelengths like 532 n m (frequency doubling), 355 n m (frequency tripling) etc. are possible.
I f a double-pulse f r o m one single N d : Y A G laser is emitted during a flash lamp cycle, the time separation between the pulses is l i m i t e d to about 200 Lis. A n approach to avoid this limitation is to use t w o independent N d : Y A G lasers w h i c h can be fired at any time separation. This solution is quite c o m m o n i n P I V (Particle Image Velocimetry), where particles f o l l o w i n g the flow are imaged at t w o time instants. T h e displacement o f the particles is determined by a cross-correlation technique and hence the velocity by d i v i d i n g w i t h the t i m e separation between the pulses. T o apply this attractive approach in holographic interferometry necessitates that the t w o lasers are mutually coherent. This is achieved by injection seeding both laser cavities t h r o u g h the rear semi-transparent mirrors using a single longitudinal mode diode-pumped continuous N d Y A G laser. I f the injected signal gives enough power o n a cavity resonance, the corresponding single axial mode w i l l prevent development o f other axial modes f r o m spontaneous emission. Thus, for resonance, the length o f the cavities must be an integer number o f half the wavelength o f the seeding laser. This is achieved by piezoelectric control o f the rear mirrors. A properly seeded pulse has a shorter b u i l d up time f r o m the Q - s w i t c h opening compared t o a non-seeded pulse. T h e mirrors are adjusted i n such a way that the pulse b u i l d up time is minimised.
This k i n d o f t w i n cavity laser has been used i n various configurations earlier by
others. I n ref. [30] they used a t w i n system and two-reference beam
holographic interferometry t o measure transient deformations o n magnetic
discs. Research groups i n Loughborough, U K , V i g o , Spain and L e o ' n ,
M e x i c o have contributed significantly to the development o f pulsed ESPI
systems, where they have used single-pulse subtraction and double-pulse
addition techniques w i t h the laser r u n n i n g at T V - r a t e [31-34].
3.2 Description of our new laser system
I n 1997 w e received grants f r o m the K n u t and Alice Wallenberg Foundation for a n e w pulsed laser system. I n the specifications f o r the purchase we tried to specify a laser system versatile for a number o f possible future applications.
T h e drawbacks o f the ruby laser listed i n section 2.3 should be avoided, f o r example the l i m i t a t i o n i n t i m e separation between double-pulses (800 LIS). Fig.
2 shows a photo o f the laser system that is characterised by the f o l l o w i n g features:
Oscillator - a m p l i f i e r
T w i n model Spectron SL804T Q-switched N d : Y A G lasers, each laser comprising a single oscillator w i t h a single power amplifier i n series ( O l , 0 2 and A l , A 2 i n Fig. 2). Each oscillator is configured w i t h a telescopic resonator w i t h intracavity m o d e - c o n t r o l l i n g aperture. This gives rise to a true T E M
00spatial profile f o r spatial u n i f o r m i t y and coherence. B o t h oscillators can be either single pulsed or double pulsed. B o t h lasers are mounted i n the same laser head o n an invar rail chassis to ensure constant alignment between the t w o beams.
Seeding
B o t h lasers are i n j e c t i o n seeded through the semi-transparent rear mirrors ( R l and R 2 i n Fig. 2) using a l O m W C W diode pumped N d : Y A G ring laser (S).
Wavelengths, polarisation
T h e 1064nm fundamental output f r o m each laser is normally combined (in the beam c o m b i n i n g cube (BC)), then frequency doubled to 532nm (in the second harmonic generator (SHG)). Alternatively, each 1064nm beam can first be frequency doubled and then the t w o green beams are combined. This approach allows flexible access to 532nm output beams w i t h either the same or orthogonal polarisation, w h i c h i n t u r n allows us to code each beam spatially. As w e l l as a green output at 532nm, the laser can also be frequency tripled to 355nm (in the t h i r d harmonic generator ( T H G ) ) .
O u t p u t energy
M a x i m u m 200 m j / p u l s e at 10 H z for the green and about 50 m j / p u l s e f o r the
U V . T h e energy is stable f r o m pulse to pulse and can be controlled by
changing the amplifier voltage and the Q - s w i t c h sync relative to the flash lamp
sync.
Fig. 2. Photo o f the new pulsed N d : Y A G laser. S: N d : Y A G seeding laser, R l
and R 2 : rear mirrors, O l and 0 2 : oscillator 1 and 2, A l and A 2 : amplifier 1
and 2, B C : beam c o m b i n i n g cube, S H G : second harmonic generator for
frequency doubling (532 n m ) , T H G : t h i r d harmonic generator for frequency
t r i p l i n g (355 n m ) , B O A , beam output apertures.
T i m i n g s
T h e pulse duration is about 13 ns. T h e t w o lasers are optimised at 10 H z each but can be operated between 5 H z and 25 H z . T h e time separation between the t w o lasers can be set f r o m zero to any t i m e . Each laser can be double pulsed w i t h time separations ranging f r o m 40 Lis to 200 Lis. Single shot operation is obtained using beam d u m p shutters. For reliable seeding, it is necessary that the oscillators are r u n repetitively. Stable single shot operation is not possible. Instead, fast solenoid-activated beam dump shutters allow access to a single, stable, single-frequency pulse. B y connecting a divider unit t o the shutters, the repetition frequency o f the emitted pulses can be divided by 2, 4, 6 or 8. This enables recordings o f repetitive double pulses at different rates.
A l i g n m e n t
For course adjustments w i t h o u t r u n n i n g the lasers, a H e - N e laser is included and adjusted co-linear w i t h the N d : Y A G beam. For more precise adjustment the lasers can be r u n i n l o n g pulse mode (fixed Q ) w i t h o u t Q-swicching the oscillators. These pulses are non-hazardous f o r the eye and f o l l o w exactly the same path as the Q-switched pulses.
3.3 The CCD camera
T h e n e w C C D camera is a P C O Sensicam double shutter w i t h a detector
containing 1024 x 1280 pixels o f size 6.7 x 6.7 (Im. Fig. 3 shows a p h o t o o f
the camera and the optics for pulsed T V holography. T h e detector is Peltier-
cooled to - 1 2 ° C and the dynamic range is 12 bits, w h i c h enables the camera
to be used at quite l o w light levels. T h e camera is controlled f r o m a computer
via a fibre optic cable, w h i c h is insensitive to electromagnetic transients. T h e
camera is externally triggered and t w o images can be captured w i t h 200 ns
t i m e separation. T h e exposure time f o r the first image is controlled by the user
via the trigger signal duration and the second is set by the read out t i m e f o r
the first (125 ms). I f the laboratory is kept dark however, the exposure o f the
C C D detector w i l l be due to the laser pulses only. Four double images per
second can be recorded at f u l l resolution. T h e camera is syncronised t o the
beam d u m p shutters o f the laser via a Stanford delay unit. This versatile
camera has also been used i n combination w i t h the ruby laser i n paper G and
H .
Fig. 3. Photo o f the C C D - c a m e r a and the optics for pulsed T V holography.
3.4 Testing of the new pulsed Nd.YAG laser
T h e laser system was first tested at the factory i n R u g b y , U K , i n combination w i t h the pulsed T V holography equipment to f u l f i l the technical specifications before delivery. T h e most important demand for a system like this is the spatial and temporal coherence between the oscillators o f the t w o lasers. The temporal coherence is achieved by i n j e c t i o n seeding f r o m a continuous N d r Y A G laser as earlier described. D i f f e r e n t resonator configurations give different spatial coherence properties.
A n oscillator w i t h an output m i r r o r w i t h a radially varying reflective profile (Gaussian resonator) is attractive i n the sense that high energy w i t h o u t a f o l l o w i n g amplifier stage can be achieved. H o w e v e r , the beam profiles, especially i n the near field, includes unwanted structures. A t w i n cavity system w i t h Gaussian resonators was first tested out. Pulsed T V holograms were recorded f o r a stationary object. Ideally n o phase differences over the field o f v i e w should appear. Due to the different spatial structures o f the t w o beams, irregular unwanted phase differences occurred i n the phase maps. These anomalies cannot be accepted, especially not w h e n the phase difference to be measured is o f the same magnitude.
N e x t , an oscillator-amplifier design was tested. Each oscillator is configured
w i t h a telescopic resonator w i t h intracavity m o d e - c o n t r o l l i n g aperture. This
gives rise to a true T E M
0o spatial profile w i t h greatly i m p r o v e d spatial u n i f o r m i t y and coherence. Each laser comprises a single oscillator w i t h a single power amplifier i n series. B o t h oscillators are i n j e c t i o n seeded t h r o u g h the semi-transparent rear mirrors o f the oscillators. Before frequency conversion the t w o beams are combined i n a beam-combining cube. I n order to avoid unwanted straight fringes superimposed o n the fringes o f interest, the t w o phase fronts must propagate i n exact same direction. A procedure to adjust this i n real time was tested out. T h e time separation between the lasers is set t o zero and the light is expanded onto a screen by a negative lens. T h e interference between the t w o wave fronts can be observed w i t h the naked eye (may be hazardous) or preferable w i t h a C C D camera and a m o n i t o r . T h e beam-combining cube is adjusted u n t i l the fringes disappear. T h e n , the pulsed T V holography equipment can be used to calculate the phase difference between the t w o wave fronts. Phase differences less than half a fringe (about 3 radians) over the cross-section o f the beam is possible to achieve, w h i c h is acceptable i n most cases. Since this phase difference is repeatable (at least o n a short term), it can be subtracted f r o m the measurements for i m p r o v e d accuracy.
3.5 High sensitivity recording of bending waves using the 355nm output
B y c o m b i n i n g the frequency doubled (532nm) and the fundamental (1064nm)
i n a second crystal frequency tripling is possible to achieve [29]. T h e use o f
shorter wavelength is attractive since smaller changes i n mechanical quantities
can be recorded. Compared to the r u b y laser (694 nm) the tripled N d : Y A G
has almost the doubled sensitivity. One o f the major problem using
holographic interferometry i n acoustics is the relatively l o w dynamic range,
only quite h i g h sound intensity levels can be recorded, see paper C . Fast
m o v i n g bending waves w i t h amplitudes less than about half the r u b y
wavelength were d i f f i c u l t t o record earlier using traditional holographic
interferometry. T h e use o f light at 355nm w i l l significantly i m p r o v e such
measurements. A n experiment showing this possibility has been p e r f o r m e d i n
our lab and a sketch o f the experimental set-up is shown i n F i g . 4.
Fig. 4. Experimental set-up for recording o f bending waves using the U V - option. P: Plate, N L : Negative lens, O : object beam, R : reference beam, BS:
beam splitter, IP: impacting laser pulse.
Some o f the mirrors and beam splitters i n the set-up are n o w changed to optics appropriate f o r 3 5 5 n m . A f t e r the t r i p l i n g crystal the light includes all three wavelengths. T h e infrared light is dumped. T h e U V light is expanded by a negative lens ( N L ) f o r illumination o f the object, i n this case a 1.3 m m thick steel plate (P), w h i c h has been polished by a sandpaper to get a diffuse surface. A fluorescent screen makes the U V - l i g h t visible i n the adjustment procedure. T h e light is reflected back f r o m the surface (object beam (O)) and an image is f o r m e d o n the C C D detector. A m i r r o r located close to the plate reflects the reference beam ( R ) . As earlier reported, see paper D and ref. [35], a focused laser pulse can be used to " i m p a c t " a plate. I n this experiment, the residual green light (IP) is focused by a lens onto the backside o f the plate.
B o t h lasers operate at 10 H z to ensure seeding w i t h closed beam dump
shutters. T h e beam d u m p shutters allow single pulses to be emitted f r o m each
cavity. A single shot trigger pulse is sent f r o m the control panel o f the laser via
a delay u n i t (Stanford) to the C C D camera. Figure 5a shows a phase map o f
the deformation 75 jXs after impact. T h e phase map can be unwrapped and
presented as a gray-scale plot o f the deformation, see Fig. 5b. T h e fastest
m o v i n g bending waves have reflected at the edges o f the plate and have a
velocity o f about 2000m/s. A phase change o f 27t radians corresponds to an
out-of-plane deformation o f 177 n m . Phase changes o f about 1/10 o f t h a t are
possible to resolve, that is vibration amplitudes d o w n to about 20 n m . I f the
experiment is repeatable, the random speckle noise is possible to reduce by
slightly m o v i n g the plate between a number o f recordings f o l l o w e d by
unwrapping and averaging [36].
O 20 40 bO öU Tuu 120 140 (mm)
4 Conclusions and future work
Pulsed holographic interferometry is a useful m e t h o d to study and measure transient events i n mechanics, acoustics and f l u i d mechanics. The short duration laser pulses o f a pulsed r u b y or N d : Y A G laser "freezes" most mechanical events, like f o r example transient bending waves i n a plate and shock waves i n a fluid. Traditionally, the recording media is photographic f i l m that has a very h i g h resolution. B y using a two-reference beam technique, quantitative data can be obtained. This technique has successfully been applied to quantitatively measure sound pressure fields f r o m impacted plates and shock waves i n an elliptical cavity.
A n all-electronic system f o r pulsed holographic interferometry called pulsed T V holography has been developed. This system is attractive i n the sense that no chemical w e t processing o f f i l m and optical hologram reconstruction are needed. T h e interference phase difference caused by a change i n object state between t w o time instants is calculated directly i n a computer. B y recording several projections o f a repeatable transient acoustic field, the pressure i n any point can be calculated by means o f a tomographic algorithm. Distortion and scaling are exactly the same i n every p r o j e c t i o n w h i c h simplifies the reconstruction procedure.
W i t h the new pulsed laser and C C D - c a m e r a several challenging projects can be listed. T h e versatility o f the laser system is high. Coherent pulses at three different wavelengths f r o m t w o mutually coherent independent lasers, each r u n n i n g at 5-25 H z , enable us to study phenomena o f a broad range i n time scales at different sensitivities. B y double pulsing b o t h lasers, a sequence o f f o u r pulses w i t h a pulse separation i n the order o f microseconds can be obtained. B y using the o p t i o n o f orthogonal polarisation f o r the t w o lasers it should be possible to record f o u r pulses o n t w o images o f one C C D camera.
B y arranging the reference beam f r o m the t w o lasers at slightly different directions, the interference terms can be spatially separated i n the Fourier plane. W i t h this technique it is thus possible t o register a transient event in f o u r steps instead o f t w o using one single camera.
C o m b u s t i o n processes have become more and more studied w i t h the aim to
reduce fuel consumption and environmental influences. B u r n i n g velocity o f
laminar and turbulent flames and the influence o f acoustic waves generated by
the flame f r o n t are fundamental problems i n combustion science. I n a Master
thesis (ref. [37]), the method o f pulsed T V holography was applied to the
study o f flames. T h e temperature field i n a laminar flame and flame speeds o f
turbulent flames was measured as pilot experiments. Future w o r k i n this field
w i l l be carried out i n co-operation w i t h theorists i n the field o f combustion.
T h e aim is to record several projections o f e.g. a temperature field and hopefully be able to reconstruct the 3 D field by computerised tomography.
A n o t h e r way w o u l d be t o calculate projections f r o m the theoretical predictions and compare w i t h the measured projections.
O u r all-electronic technique is also attractive i n non-destructive testing o f structures. Defects like debonds and delamination i n composite materials w i l l introduce anomalies i n the transient bending wave pattern [38]. H o w e v e r , to adopt the technique i n industrial applications, reliable procedures f o r automatic computer detection o f defects must be developed. T h e n o n - restricted time separation between pulses f r o m the t w o lasers allows measurements o n objects w i t h l o w flexural wave speed, e.g. biological tissues.
T h e combined pulsed T V holography and Digital Speckle Photography technique should be further developed (paper H ) . One attractive and challenging application w o u l d be to measure paper stiffness and anisotropy o n - line i n a paper machine. Knowledge o f these parameters w o u l d i m p r o v e the possibilities to control them d u r i n g p r o d u c t i o n . A laser pulse can generate bending waves and by using a theory o f bending wave propagation i n anisotropic plates the stiffness parameters i n various directions can be determined [39].
I n conclusion, pulsed T V holography has proved to be a fast and reliable
m e t h o d to quantitatively study transients i n mechanics and acoustics. T h e
technique has a great potential i n experimental mechanics i n the future.
5 Summary of appended papers
I n table 1 below, the appended papers are listed according to their content f o l l o w e d by a short summary and conclusions o f each paper.
Table 1. Appended papers
Recording Measurement Phase Phase objects, Combined Paper techniques: of transient objects, tomographic digital speckle
film, one film, two all deformations transient reconstruction photography
reference reference electronic density and Pulsed TV
beam beams fields holography