SR-922
REPETITIVELY PULSED RUBY LASERS AS LIGHT SOURCES
FOR HIGH-SPEED PHOTOGRAPHY
Jeffery M. Grace, Peter E. Nebolsine, Charles L. Goldey
Physical Sciences Inc.
20 New England Business Center
Andover, MA 01810
Gurdaver Chahal, James Norby, and Jean-Marc Heritier
Continuum
3150 Central Expressway
Santa Clara, CA 95051
Optical Engineering 37(8) 2205-2212 (August 1998), pp. 2205-2205
Copyright © 1998 Society of Photo-Optical Instrumentation Engineers
This paper was published in Optical Engineering and is made avaialable as an electronic reprint
with permission of SPIE. Single print or electronic copies for personal use only are allowed
Systematic or multiple reproduction, or distribution to multiple locations through an electronic
listserver or other electronic means, or duplication of any material in this paper for a fee for
commercial purposes is prohibited. By choosing to view or print this document, you agree
to all the provisions of the copyright law protecting it.
Repetitively Pulsed Ruby Lasers as Light Sources for High-Speed Photography
Jeffery M. Grace, Peter E. Nebolsine, Charles L. Goldey
Physical Sciences Inc.
20 New England Business Center
Andover, MA 01810
grace@psicorp.com
nebolsine@psicorp.com
Gurdaver Chahal, James Norby, Jean-Marc Heritier
Continuum
3150 Central Expressway
Santa Clara, CA 95051
Abstract
Two ruby laser systems employing multiple Q-switching technology have been
developed to provide ruby laser light for high-speed photography. Potential applications include
ballistics and flow visualization as well as non-destructive test evaluation using laser imaging
diagnostics such as photography, holography, and various interferometric techniques. The laser
systems produce more than 50 pulses at repetition rates up to 500 kHz with nearly constant
pulse-to-pulse energies. One system, based on a commercial laser, provides multiple pulses of
holographic quality light with individual pulse energies on the order of 10 mJ, pulse widths of 50
ns, and a pulse train length of more than 300 µs. A new ruby laser system has been developed to
provide higher pulse energies, on the order of 350 mJ per pulse, with 10 ns pulse widths and 140
µs pulse train length.
The method for multiple Q-switching by modulating the Pockels cell’s quarter wave
voltage and the formation of an individual Q-switched pulse has been investigated. The energy
within the individual pulses formed in the oscillator cavity has been successfully increased by
propagating through an amplification section, without degradation of the temporal or pulse-to1
pulse amplitude stability. Etalons for longitudinal mode selection and an iris for spatial mode
selection have been incorporated into the lower energy system and an image of a reconstructed
hologram is presented. Camera capabilities and the implications of higher pulse energies have
also been investigated.
Keywords: multiple Q-switching, Pockels cell, ruby laser, amplification, high-speed
photography, holography, high repetition rate laser
1.
INTRODUCTION
Pulsed laser light is an ideal illumination source for use in the imaging of high-speed
events such as ballistics, material fractures and fluid combustion/flows. A pulsed light source
has been developed based on repetitively Q-switched ruby laser technology and soon will be
commercially available. The newly developed source is capable of providing illumination for
large area front lit photographic and cylindrical holographic applications with < 100 ns exposure
times. For these applications it is desired to have a pulsed source which emits tens of pulses,
each having a time duration on the order of tens of ns to freeze the motion of the image subject.
In addition a high pulse repetition rate of hundreds of kHz is desired with low pulse-to-pulse
timing jitter to synchronize with the camera and low energy variation to maintain proper
exposures. The imaging of such large photographic subject requires that the energy contained in
a single pulse be 100's of mJ when using high-speed framing cameras, for example, the Cordin
330A.
The technique for obtaining multiple pulses from a single laser rod by repetitive
Q-switching the cavity while the flashlamp pumps the rod has been previously demonstrated and
2
an extensive review of this technology is given by Huntley.1 This work demonstrated the largest
amount of pulse energy (~ 20 mJ) at the desired repetition rates (~ 500 kHz). In addition, use of
ruby lasers for holographic purposes is common in the imaging community. An alternative
approach of the chopping up of a cw beam and amplifying the pulse segments was not chosen
because the high energy pulse requirement would add an excessive amount of amplification
stages and result in a very complex system.
Individual pulse energies as high as tens of mJ at repetition rates of hundreds of kHz have
been obtained by Huntley.1 This pulse energy is within a factor of ten for the desired pulse
energy but the laser systems used in Huntley’s study and in previous works were all custom built
devices of various sizes containing one-of-a-kind components. The first laser system (Section 2)
was then used to study scaling factors and operational parameters on an existing commercial
laser system that could then be applied to the higher energy version (Section 5). The formation
of an individual pulse in a pulse train has been investigated experimentally. This affects the
pulse-to-pulse amplitude and timing jitter stability. Such information is critical for high-speed
photography when the transmission function of a camera varies significantly over the exposure
time or for very short exposure times using a gated intensifier. Increasing the energy of the
individual pulses and the effect this amplification has on the system performance has been
explored. The low energy system has been operated with components for single mode selection
and the systems applicability to making shadowgraphic holograms has been demonstrated
(Section 7).
3
2.
LOW ENERGY LASER SYSTEM
The ruby laser system for lower energy applications, e.g., low f/# CCD camera
applications, is a Model 22HD manufactured by Apollo Lasers Inc. The laser is constructed in
an oscillator/amplifier configuration along a single optical rail, Figure 1. When new it was rated
for a maximum energy output of 3 J in a single Q-switched pulse. The oscillator cavity is
composed of a rear reflector mirror, Pockels cell, etalon, 1/4 waveplate, polarizing beamsplitter
cube, aperture, rod head, and resonant reflector output coupler. The system also contains a rod
head for amplification and a HeNe laser for alignment purposes. All of the components are
original to the system except for the polarizing beamsplitter cube and 1/4 waveplate. The cube
was purchased from CVI Laser, Inc. for polarization selection and replaced the original brewster
stacks that were damaged. The 1/4 waveplate was introduced into the cavity to allow the laser to
be operated in a mode where the cavity has a high loss, low Q, when the voltage was not applied
to the Pockels cell. The Pockels cell is a KD*P type and was made by INRAD, Inc. The internal
cavity etalon is of the solid type. The output coupler to the system is an etalon of the double
plate air gapped type (resonant reflector) with a reflectivity of approximately 30%. The aperture
is a variable iris.
The oscillator rod has a diameter of 3/8 in. and a length of 3 in. The amplifier rod also
has a diameter of 3/8 in. but is 6 in. in length. Both the oscillator and the amplifier flashlamps
are of the helical style. The oscillator and amplifier sections were both operated at a flashlamp
energy of 2.2 kJ during the experiments. Figure 2 is a photodetector trace of the oscillator
flashlamp light and resulting non-Q-switched, free-running, laser emission from the oscillator
section of the system. The energy output from the oscillator section of the system when freerunning with the 1/4 waveplate removed is 2.65 J. The energy output from the oscillator section
4
in a single Q-switched pulse is 1.35 J. The amplification section of the laser increases the output
energy from the oscillator section by approximately x2.5.
3.
FORMATION OF PULSES
Q-switching of a laser cavity is normally done once to obtain a single pulse. This is
usually done near the end of the flashlamp pump duration when the population inversion in the
rod is at its highest. To obtain multiple pulses the laser cavity is repetitively Q-switched over a
period of time for which the inversion, and resultant radiation, is greater than the losses
associated with the laser cavity. It has been found that the laser cavity can be repetitively
Q-switched with resulting pulses formed, for approximately the duration in which the laser will
emit when free-running as shown in Figure 2. However, for the system to yield multiple
Q-switched pulses, the laser’s Pockels cell has to be modulated at the desired pulse repetition
frequency.
The pulsing source, obtained from Medox Electro-Optics, Inc., is capable of rapidly
applying and removing 2.5 kV from the laser’s Pockels cell. The rise and fall times for the
voltage application are 10 ns or less. This voltage is the 1/4 wave voltage of the Pockels cell at
694.3 nm. The Medox system is capable of producing from 1 to 999 high voltage pulses of
between 50 and 300 ns in duration at a repetition rate up to 500 kHz. This system was integrated
into the laser system. A trigger output from the laser system was used to start the Medox unit’s
modulation of the Pockels cell. The Medox unit outputs a TTL logic voltage level signal that
rises with the leading edge of the first high voltage output pulse.
We have successfully obtained data for repetition rates from 100 to 500 kHz. Figures 3
through 5 show plots of the photodetector signal data for the laser oscillator section emission
5
only for operation at the repetition rates 500 kHz. Figures 3 and 4 were taken with a photodetector which had its response slowed with a load resistor to act as an energy integrator. The
pulse-to-pulse energy, not necessarily the pulse form, is important in photographic/holographic
applications as long as the pulses are short in duration. Figure 5 shows a single pulse acquired
without the load resistor thereby using the full bandwidth of the detector.
The pulse train amplitude shown in Figure 3 follows the flashlamp discharge envelope
shown in Figure 2. In addition the duration over which pulses are formed is nearly the same as
the 400 us duration of free-running lasing in Figure 2. The pulse-to-pulse stability in these
figures is excellent. This is more apparent in Figure 4 which shows pulses plotted on an
expanded time scale. The individual pulse width in Figure 5 is about 50 ns FWHM. The width
of the individual pulses are consistent with those measured by both Huntley and Rowlands.1,2,3 In
addition it was found that the individual pulse widths did not vary greatly during a train of pulses
or for pulse trains formed at different repetition rates. This is also consistent with previous
measurements made by Huntley.1
The noise on the baseline in Figure 3 is due to the close proximity of the photodetector to
the Pockels cell and is electrical pickup from the high voltage switching. Huntley found in his
previous investigation that in order to prevent giant pulse formation the Pockels cell modulation
should start at or before when lasing threshold has been achieved.1 The noise can be more
clearly seen at the base of the start of the pulse in Figure 4 and corresponds to the application of
the high voltage pulse to the Pockels cell. Noise pickup by the photodetector due to the high
voltage switching is also evident in Figure 5. The ripple in the signal baseline and at the peak of
the pulse are considered RF pickup and not as a result of the detector sensing photons.
6
We calculated the energy contained in a single pulse from the total energy in a pulse
train. A computer program was written which takes a data file, as displayed in Figure 3, picks off
the peak values, equates this to the total energy measured, and determines what fraction of the
total energy was contained in an individual pulse. The bar graphs in Figures 6 and 7 are the
result of this calculation when applied to the data in Figure 3. Figure 6 is a bar graph for the
entire pulse train and Figure 7 is data from the same test but plotted on an expanded time scale.
The total energy measured for this test was 0.83 J. The first 100 pulses contain a nearly constant
individual pulse energy of about 5 mJ. The energy contained in an individual pulse was seen to
increase as the pulse repetition frequency for the tests was decreased from 500 to 100 kHz.
However, the total pulse train energy was measured to be nearly the same over this same range
of pulse repetition rates. This constant pulse train energy is approximately 30% of the energy
emitted by the laser oscillator when free-running and 60% of the energy contained in a single
Q-switched pulse.
Some of the trends in the data presented in Figures 3 through 5 can be better understood
through a closer examination of the formation of the individual pulses within a pulse train. The
output signal from the Medox unit which marks the rising edge of the first output pulse was fed
to a precision digital delay generator. A delay was set on the digital delay generator to
correspond with the application of the high voltage pulse to the Pockels cell within the pulse
train. The logic level output signal from the delay generator rises in time with the application of
the high voltage and decreases when the high voltage is removed. Figure 8 is a photodetector
trace of a single pulse from the oscillator section of the laser when operated at a pulse repetition
frequency of 500 kHz. Figure 9 is the output signal from the digital delay generator which
corresponds to the application of the high voltage to the laser’s Pockels cell for this single pulse.
7
The high voltage was applied to the Pockels cell for a duration of 250 ns during these
tests. It was found that varying the time the high voltage was applied to the Pockels cell, from
200 to 300 ns, had no effect on the laser’s ability to form stable pulse trains at repetition rates
from 100 to 500 kHz. This finding agrees with observations made by Rowlands2 however
Huntley1 in previous work found a dependence of shorter high voltage pulse widths at greater
repetition rates for good pulse formation. Huntley did vary the pump rate while it was held
constant during these tests. All three studies, Rowlands, Huntley and this one, did find that good
pulses were formed with a high voltage pulse duration in the range of 200 to 300 ns. This is an
interesting result because all three systems had large variations in cavity lengths, rod sizes and
pump energies.
A critical result from Figures 8 and 9 is that the peak of the laser pulse is defined by the
falling edge of the high voltage. For this system, the falling edge of the high voltage is determined by a precision digital clock that controls the timing within the Medox unit, resulting in a
timing jitter for the laser pulses that is much less than a laser pulse width. The peak of the laser
pulse in Figure 8 coinciding with the falling edge of the trace in Figure 9 provides some
information on the pulse formation and characteristics of each pulse. When the pulse is formed
relative to the Q-switching of the system has been an issue of speculation in prior investigations.
From these figures it can be seen that the pulse builds up during the application of the high
voltage pulse, cavity at high Q, and then is cut off by the removal of the high voltage and the
resulting return of the cavity to a low Q state. This cut off of the pulse is a likely reason for the
decrease in energy seen from a pulse train as compared to a single Q-switched pulse. Every time
a pulse is formed the Pockels cell rotates the polarization of the light when the cavity is returning
to a low Q state and a portion of the pulse energy is rejected from the cavity by the polarizer.
8
A numerical model developed by Huntley (Ref. 4), based on the low energy laser system
here, i.e., cavity length, rod size, cavity losses, pump rate, has simulated this experiment and
found that the laser pulse occurs at the end of the Pockels cell pulse. Precision timing of the
laser pulse would prove to be beneficial if the laser system were to be used as a light source for a
digital camera employing a narrow imaging gate duration.
4.
AMPLIFICATION OF PULSES
It is desired to uniformly amplify the energy of the individual pulses in the train without
distorting the temporal profile of the train. Neither Huntley nor Rowlands investigated the use of
an amplifier section with their repetitively Q-switched lasers. Although amplification of a train
of pulses formed by a laser has been done in the past it is not known if it was done with a ruby
laser for similar conditions.5 One concern with implementation of amplification is that the
amplification section would preferentially amplify the larger energy pulses emitted from the
oscillator cavity giving a non-uniform amplitude pulse train. An additional concern is the
possibility that the first pulse or pulses into the amplifier section would deplete the gain in the
section to the extent that the remaining pulses would pass through unamplified.
The output emission from the oscillator section was allowed to pass through the amplifier
section (Figure 1). Photodetectors were used to monitor the amplifier input and output
emissions. The amplification of pulses successfully produced a train of pulses with temporal
characteristics of the oscillator train. The energy per pulse for this low energy system was
increased x2.5 to about 12 mJ per pulse.4 Demonstration of successful operation with
amplification is covered in more detail in the following section on the high energy laser system.
9
5.
HIGH ENERGY LASER SYSTEM
A new high energy laser system based on the technology described above has been
developed by Continuum in partnership with PSI. This system consists of an oscillator plus two
amplification stages. The oscillator section contains the same essential Q-switching elements as
described in Figure 1. The work reported here describes the illumination capability of this
advanced laser system; therefore, mode selection elements are not included in this discussion of
the oscillator cavity.
Figures 10 and 11 show the photodetector traces and energy per pulse results for pules
repetition frequencies (PRF) at 500 kHz and 250 kHz respectively. These figures exhibit nearly
ideal pulse-to-pulse stability. Figure 10a,b and Figure 11a,b were again acquired using a load
resistor to integrate the photodetector response. An active, optical element placed between the
oscillator cavity and first amplification stage artificially terminated the macro-pulse. This
element shaped the macro-pulse envelope to achieve even better stability. The macro-pulse
duration was designed to ensure the laser produces tens of pulses per shot. Figure 3 shows that
this macro-pulse time is not a fundamental limit to the laser. A modest modification has been
made to the experimental method described in Section 3 above to reduce the effects of EMI from
the Pockels cell switch and the impedance mismatch in the integration circuit.
Figures 10 and 11 show that the individual pulse energies are on the order of 350 mJ for
500 kHz and more than double that to a peak of nearly 1 J/pulse at 250 kHz. Pulse energy scales
inversely with pulse repetition frequency (PRF) for this operational envelope of the laser, as seen
in Figure 12. This figure shows the macro-pulse energy as a function of PRF for two pump
energies of amplifier stage 1 (2,900 and 3,600 J) and a fixed pump energy of the oscillator cavity
(3,600 J). This operational envelope is clearly below saturation, indicating that even higher
10
pulse energies can be obtained. The optical gain for these amplification stages is roughly x25 for
stage 1 and x2 for stage 2. Stage 1 is a double-passed amplifier and most of the gain (~x9)
occurs on the first pass while the second pass is about x2.5. Since the ruby rods in the oscillator
and the amplifier stage 1 are the same size, clearly the oscillator does not tax the ruby rod.
The pump energies given in Figure 12 correspond to a reasonably cautious operational
envelope. It was not the goal to investigate the full operational envelope of the laser. However,
some of these conditions can be defined. The minimum PRF is not know for this laser system,
though PRFs of 167 kHz have been successfully tested in the current configuration while
100 kHz exhibited oscillation greater than 1 f-stop. A successful test is one where the individual
pulse energies vary by less than a factor of 2 (1 f-stop) over the macro-pulse envelope. The
maximum PRF reported here is 500 kHz due to the Pockels cell high-voltage pulser limit
(500 kHz).
Figure 13 shows a single laser pulse trace isolated from the center region of a macropulse fired with the settings of Figure 11 (250 kHz). This method uses the full bandwidth of the
detector and oscilloscope (no load resistor). The FWHM of this laser pulse is approximately
10 ns. This fact coupled with the low jitter due to fact that the pulse forms at the close of the
Pockels cell yields a highly desirable combination for stop motion photography. In essence,
motion is frozen regardless of the open shutter time because the illumination is restricted to a
narrow time window (e.g. an object moving at 5 km/s travels 50 µm in 10 ns). This pulse shows
some interesting features that have not yet been understood. The central, large pulse contains
most of the energy.
Again, the features of the pulse are not critical for photographic applications as long as
the pulse is narrow enough.
11
6.
CAMERA INTEGRATION
Many different types of high speed framing cameras are commercially available. In all
cases there is the effective f/# of the lens system to form the image, an optical transmission of the
camera system and a fluence(energy per unit area) requirement of the image recording medium.
We will take a broad definition of transmission of the optical system to include the effects of
gain of an image intensified camera systems. Hence, transmission can be greater than unity for
such cameras. Photographic film and CCD units are the two most common capture methods
leading to storage of the final image.
The optimum choice of which high speed framing camera is best suited for a specific
application depends on many factors, including type of illumination source (pulsed or continuous), number of frames desired, f/#, field of view, image format dimensions, transmission,
and fluence requirement of recording medium. When utilizing either laser system described in
this article, the effective exposure time to “stop” the action will be the laser pulse time (except
for some very specialized cameras). Thus, in some cases, this may relax the framing rate
requirement as compared to use with a continuous illumination source. A number of mechanical
framing cameras have an exposure time that is a set fraction of the interframe time. When using
mechanical framing cameras with CW sources, some operators intentionally go to higher
framing rates than necessary to record the dynamic motion of the object. This method reduces
the exposure time to an acceptable level in order to “stop” the action. Setting a framing rate
solely by this requirement is alleviated when using short laser pulses as described above.
Sample cases have been computed for several types of framing cameras including a
rotating mirror (Cordin 330A), gated intensified CCD (Imacon 468 by Haadland Photonics) and
a fast framing CCD (Silicon Mountain Design, Inc.), see Table 1. The Cordin 330A camera has
12
been chosen for illustrating a sample calculation. This 80-frame camera system uses conventional 35 mm film. It has an effective f/# of 33 when taking into account the collection cone
angle and transmission of the optical system. Kodak High Speed Infrared 2481 film is an appropriate film to use in conjunction with ruby laser pulses. A fluence of 2.5 x 10 -5 mJ/cm2 is
sufficient to expose the film based on nominal exposure numbers. The object plane fluence
requirements are then calculated as for any camera system. Given the format size, of 18 x
25 mm, a f/33 optical system, the illumination laser spot diameter in the object plane was
calculated. The results are shown in Figure 14 using the laser performance results described in
Section 5. This figure shows that the linear object dimension scales with the square root of the
laser pulse energy.
7.
APPLICATION TO HOLOGRAPHY
One type of imaging to which the multi-pulsed laser system is to be applied to is
holography. For good holograms, both temporal and spatial coherence of the individual pulses is
needed. Previous investigators of this rapid, repetitive Q-switching of a ruby laser, did not
include mode selection in their systems. To demonstrate that the laser system pulses contained
sufficient coherence, holograms were made. The experimental data previously presented was
generated with the manufacturer’s supplied etalons installed for longitudinal mode selection and
the oscillator cavity aperture set to the full rod diameter. To obtain a single transverse mode the
cavity aperture was reduced to a diameter of 2 mm. This method employed by Apollo Lasers,
Inc. for producing coherence in a ruby laser is discussed in depth by Pitlak.6 For exposure
control purposes, only a single Q-switched pulse emitted from the oscillator section of the laser
was used. This pulse contained approximately 10 mJ of energy. This single pulse is considered
13
to contain representative coherence properties of a train of pulses if the laser was repetitively
Q-switched.
The shadowgraphic hologram shown in Figure 15 was able to be made of the object in
Figure 16. The hologram was reconstructed using a HeNe laser and a photograph of the
reconstructed image was taken with a 35 mm camera. A shadowgraphic type hologram was
chosen as a test case because it is commonly applied to ballistics testing. The experimental
apparatus consisted of a PVC cylinder 5.75 in. in diameter with the target (Figure 16) placed at
its center. Holographic film (Agfa type 8E75) lined the interior of the cylinder from 180 to
360 deg for about 6 in. along the cylinder length. The other half of the cylinder interior (0 to
180 deg) was lined with a diffuse reflector for the same length. The laser beam was split into a
reference and object beam. These two beams then entered the cylinder on axis from opposite
ends and reflected off 2 ball bearings symmetrically located about 1 in. off the cylinder axis.
These ball bearings were chrome plated, “off-the-shelf” components of good optical quality. In
this configuration, the bearings acted like convex mirrors to diverge and reflect the two beams
onto the film/reflector lining.
8.
CONCLUSIONS
A commercial holographic quality ruby laser system has been modified to provide a train
of pulses by repetitively Q-switching the oscillator cavity. The pulse trains have been formed at
repetition rates up to 500 kHz with tens of consecutive pulses having nearly constant energy.
The formation of an individual pulse within the train was investigated. It has been determined
that the pulse forms at the end of the high Q duration and is actually cut off by the transition of
the cavity to low Q by the rotation of the cavity polarization by the laser’s Pockels cell. A low
14
pulse-to-pulse timing jitter follows from this result. The energy contained within the individual
pulses of a pulse train emitted by the oscillator section have been successfully increased by an
amplification section without significant degradation of the pulse train temporal profile.
A high energy system was developed through a technology transfer based on the results
of the low energy system. This high energy laser has generated pulse trains with individual
pulse energies of nearly 1 J at 250 kHz and on the order of 350 mJ at 500 kHz for tens of pulses.
The precise timing, low jitter, short pulse duration and excellent pulse-to-pulse energy stability
meet the requirements for high-speed photographic applications.
9.
ACKNOWLEDGMENTS
This work was performed under Contract No. F08630-96-C-0031 for the Air Force
Research Laboratory, US Air Force, Eglin AFB, FL.
10.
REFERENCES
1.
Huntley, J.M., “High-Speed Laser Speckle Photography. Part 1: Repetitively Q-switched
Ruby Laser Light Source,” Optical Engineering, Vol. 33, No. 5, May 1994.
2.
Rowlands, R.E., Taylor, C.E., Daniel, I.M., “Ultrahigh-Speed Framing Photography
Employing a Multiply-Pulsed Ruby Laser and a ‘Smear-Type’ Camera: Application to
Dynamic Photoelasticity,” Proceedings of the 8th International Congress on High-Speed
Photography, June 1968.
3.
Rowlands, R.E., Wentz, J.L., “A Low-Voltage Pockels Cell having High Repetition
Rates and Short Exposure Durations,” Proceedings of the 9th International Congress on
High-Speed Photography, August 1970.
15
4.
Goldey, C.L., Grace, J.M., Nebolsine, P.E. and Huntley, J.H., “An Investigation of a
Ruby Laser Repetitively Pulsed at Rates up to 500 kHz Containing Etalons and
Amplification,” SPIE Proc. Solid State Lasers VII, January 1998.
5.
Lempert, W.R., Wu, P-F., Zhang, B., Miles, R.B., “Pulse-Burst Laser System for HighSpeed Flow Diagnostics,” presented at 34th Aerospace Sciences Meeting and Exhibit,
AIAA 96-0179, January 1996.
6.
Pitlak, R.T., Page, R., “Pulsed Lasers for Holographic Interferometry,” Optical
Engineering, Vol. 24, No. 4, July 1985.
16
Table 1. Camera Specifications
Camera
Cordin 330A
Imacon 468
SMD-64k1M
Frame Rate
(MHz)
Number of
Frames
1
100
1
80
8
16
a
Minimum Image
Fluence
(mJ/cm2)
For an OD=1 using Kodak high-speed infrared film
Assuming a 22 µm2 pixel, f/11 and 1000 cm2 object.
b
17
a
2.5 x 10-5
b
70
b
350
Resonant
Reflector
Ruby
Amplifier
Head
Aperture
1/4
Wave Solid
Plate Etalon
Mirror
Ruby
Oscillator
Head
Polarizing
Beamsplitter
Cube
Pockels
Cell
Figure 1. Ruby laser configuration.
Figure 2. Oscillator flashlamp envelope and laser emission.
18
Alignment
HeNe
Laser
Figure 3. Q-switched laser pulse train at 500 kHz.
Figure 4. Expanded time scale for Figure 3.
19
Figure 5. Temporal profile of a single pulse formed at 500 kHz.
Figure 6. Pulse energy for Q-switched laser pulses at 500 kHz (data from Figure 3).
20
Figure 7. Expanded time scale for Figure 3.
Figure 8. Single pulse within a 500 kHz pulse train.
21
Figure 9. Delay generator signal corresponding to the high voltage applied to the Pockels cell.
22
Photodetector
Voltage [V]
Osc/Amp1/Amp2 = 2.0/2.0/1.95 kV, PS = 2 µs, PW = 180 ns (sc2.083)
10
8
(a)
6
4
2
0
Photodetector
Voltage [V]
25
50
75
100
125
Time [µs]
150
175
10
8
200
(b)
6
4
2
0
90
92
94
96
98
100
Time [µs]
Pulse Energy [J]
0.5
Figure 10.
(c)
0.4
0.3
0.2
0.1
0.0
0
20
40
Pulse Number
60
80
D-8129
500 kHz pulse train, total energy = 24.87 J. (a) macro-pulse envelope; (b) expanded
time scale of (a); (c) energy per pulse in (a).
23
Photodetector
Voltage [V]
Osc/Amp1/Amp2 = 2.0/2.0/1.95 kV, PS = 4 µs, PW = 180 ns (Sc2.096) E = 26.04 J
(a)
3
2
1
0
Photodetector
Voltage [V]
25
50
75
100
125
Time [µs]
150
175
200
(b)
3
2
1
0
Pulse Energy [J]
100
Figure 11.
102
104
106
108
110
112
Time [µs]
114
116
118
1.0
0.8
0.6
0.4
0.2
0.0
120
(c)
0
10
20
Pulse No.
30
40
D-8130
250 kHz pulse train, total energy = 26.04 J. (a) macro-pulse envelope; (b) expanded
time scale of (a); (c) energy per pulse in (a).
28
26
Macro-Pulse Energy [J]
24
Amp1=1.8 kV, 500 kHz
Amp1=1.8 kV, 250 kHz
Amp1=2.0 kV, 500 kHz
Amp1=2.0 kV, 250 kHz
22
20
18
16
14
12
OSC = 2.0 kV
10
8
1.65
1.70
1.75
1.80
1.85
1.90
1.95
Amplifier Stage2 Pump Voltage [kV]
Figure 12.
2.00
2.05
D-8128
Total macro-pulse energy for 500, 250 kHz at a fixed oscillator pump energy and two
separate stage 1 amplifier pump energies.
24
4.5
Photodetector Voltage [V]
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0
50
100
Time [ns]
150
200
250
D-8260
Figure 13. Fast time response trace of a single pulse from 250 kHz data.
Illuminated Circle Diam (cm)
200
Phase II Scale
Performance
100
Phase I
Baseline Design
f/33
High Speed
Infrared 2481 Film
ER = 350 mJ 500 kHz
PRF
10
40
100
1000
PRF (kHz)
D-8133
Figure 14. Object dimension scaling with individual pulse energy.
25
Figure 15. Photograph of a reconstructed shadowgraphic hologram.
Figure 16. Holographic subject.
26