3-D flow imaging using a MHz-rate pulse burst LASER systemltces.dem.ist.utl.pt/lxlaser/lxlaser2010/upload/... · mechanism and number of amplifiers, can vary. The master oscillator

15th Int Symp on Applications of Laser Techniques to Fluid Mechanics Lisbon, Portugal, 05-08 July, 2010

- 1 -

3-D flow imaging using a MHz-rate pulse burst LASER system

Brian S. Thurow1, Kyle P. Lynch1, Steven T. Williams1, Michael B. Melnick1

1: Department of Aerospace Engineering, Auburn University, Auburn, AL, USA, [email protected]

Abstract Over the past decade significant strides have been made in the development of pulse burst laser systems capable of producing high energy laser pulses at repetition rates in excess of 1 MHz. These unique laser systems are currently being used in experiments across a variety of fields ranging from traditional fluid dynamics to combustion to plasma physics. Described here is the use of one such system for 3-D flow visualization whereby a high-repetition rate laser sheet is scanned through the flow field using a galvanometric scanning mirror and flow images are acquired using a high-speed camera. 3-D images are then reconstructed from the resulting stack of 2-D images. Using existing instrumentation, images with 220 x 220 x 68 pixel resolution can be acquired in 136 microseconds with the primary limitation being the speed and resolution of an available high-speed camera. The strength of the technique is illustrated through 3-D flow visualization experiments conducted in a turbulent jet and a turbulent boundary layer. 3-D smoke visualization images of the near-field of a turbulent jet depict the complex 3-D interaction between ring vortices and streamwise vortices that form in axisymmetric shear layer of the jet. Turbulent boundary layer images show the 3-D nature of large-scale structures present in the outer layer of the boundary layer and lay the ground work for more detailed future experiments. 1. Introduction

Since the early success of laser based planar imaging techniques, such as particle image velocimetry (PIV), planar Doppler velocimetry (PDV) and planar laser induced fluorescence (PLIF), a continuing focus of subsequent research has been on the development of techniques that enable measurements in three-dimensions (3-D) and/or at high-speeds. For example, significant strides have been made recently in the ability to acquire 3-D velocity data using techniques based on stereographic (e.g. Guezzenec et al. 1994), holographic (e.g. Trolinger et al. 1997 and Sheng et al. 2003) and tomographic (e.g. Elsinga et al. 2006) imaging concepts. Recent advances in commercial laser and camera technologies now allow for these techniques to be applied at repetition rates on the order of 1 – 10 kHz, which is fast enough to resolve the timescales across a wide variety of turbulent flows. A common limitation of many of these techniques, however, is their dependence on particle based imaging where particle seeding quality (particle density and uniformity) requirements can be significantly more demanding than in conventional 2-D imaging experiments. Furthermore, these techniques preclude the acquisition of other types of 3-D information, such as that gained through LIF and smoke flow visualization. Thus, the overall utility of these new 3-D diagnostics, although impressive and expanding, still faces some limitations.

One method of acquiring 3-D data not subject to the same limitations is through scanning methods where a laser light sheet is translated through the flow with an image captured at each successive plane; the resulting stack of 2-D images is then reconstructed into a 3-D image of the flow. A major advantage to this approach is that each image in the stack can be treated and analyzed in a manner equivalent to its stationary 2-D technique’s counterpart, thus taking advantage of the extensive knowledge available with respect to traditional planar imaging diagnostics. The scanning method has been employed extensively in low-speed flows where the fluid time scales are relatively long and physically translating the imaging system is not too difficult. More recently, the availability of kHz rate lasers, cameras and scanning mirrors has extended this technique to moderate speed flows. Still, the main limitation to scanning has been the finite amount of time required to complete the scan, over which time the flow can experience significant changes. Thus,


- 2 -

the vast majority of scanning 3-D imaging measurements have been restricted to flows with free stream velocities below 1 m/s.

Several efforts have been made over the past few decades to dramatically improve the speed of scanning techniques using specialized laser systems and ultra high-speed cameras. Long and Yip (1988) and Yip et al. (1988) used a resonant scanning mirror to deflect the output of a dye laser with a 1.4 µsec duration pulse through the flow field of a turbulent jet. During the sweep of this single pulse, 12 images, each with a resolution of 58 x 120 pixels, were acquired with an electronic framing camera operating at a framing rate of 10 million frames per second (fps) resulting in a 3-D image with 58 x 120 x 10 resolution and acquisition time of 1.2 µsec. Patrie et al. (1994) improved upon this method by using a higher energy pulsed dye laser (10 J/pulse), a 12 sided polygonal mirror rotating at 500 Hz, and a camera capable of acquiring 20 images at a resolution on the order of 120 x 90 pixels for each image. Island et al. (1996) subsequently used the technique to perform the first nearly instantaneous 3-D flow visualization of a supersonic flow. Hult et al. (2002) deviated from the single laser pulse approach and used a cluster of four double-pulse Nd:YAG lasers, a galvanometric scanning mirror and an eight CCD (i.e. 8 frames) high-speed camera to obtain a sequence of eight images (576 x 285 pixels each) in a 15 m/s reacting flow field over a span of ~88 µsec. Their technique utilized laser-induced incandescence to measure soot volume fractions in a flame. A distinct advantage of their approach was the use of pulsed (~10 nsec duration) Nd:YAG lasers whose output can be more efficiently converted to 2nd, 3rd and 4th harmonics, and is compatible with a large number of existing planar techniques. In all cases, the speed, resolution, and maximum number of frames of the high-speed camera as well as the capabilities and flexibility of the laser system were limiting factors. As such, these systems have not experienced sustained use since their initial demonstration.

The situation has recently improved, however, with the development of new high repetition rate lasers and cameras. Specifically, several research groups have built and are using a unique MHz rate pulse burst laser system capable of producing a burst of high energy (1 – 1,000 mJ/pulse), short duration (10 – 20 nsec) laser pulses at repetition rates in excess of 1 MHz. To date, these systems have been primarily used for high-speed planar imaging. This paper discusses recent efforts by the authors to utilize one of these laser systems for 3-D imaging, whereby the output of the pulse burst laser is scanned and imaged at high speeds yielding a nearly-instantaneous image of the flow. This capability is demonstrated through 3-D flow visualization experiments performed on a turbulent jet and a turbulent boundary layer. Before proceeding with a description of the 3-D flow visualization system, a brief history of the development of the pulse burst laser system is provided. 2. Description and History of MHz Rate Pulse Burst Laser Systems

The pulse burst laser concept has experienced significant growth over the last decade as the number of unique systems built over this time has grown from a single purpose built system to eight systems now being used for a variety of purposes. These eight systems are associated with Princeton University, The Ohio State University, NASA-Glenn Research Center, University of Illinois, Auburn University, University of Wisconsin, Iowa State and The Ohio State University (2nd system under construction). Due to this growth and the instrumental role this laser has in the current work, a brief history of the development of these systems is recounted here. While not essential to understanding the 3-D flow visualization aspects of this paper, it is hoped that this history will help the reader further appreciate the broad capabilities of these laser systems and possibly motivate the development of additional systems and applications.

We begin with a description of the pulse burst laser system used in this work and shown in Figure 1. The configuration of the system is classified a master oscillator power amplifier (MOPA), where a burst of low energy pulses is formed and amplified through a chain of power amplifiers. The basic architecture (i.e. pulse slicing and amplification) employed in this system is common to all the pulse burst laser systems discussed here although the details, such as the slicing


- 3 -

mechanism and number of amplifiers, can vary. The master oscillator of the system is a low power (100 mW), single longitudinal mode, continuous-wave Nd:YAG laser operating at 1064 nm, which is “sliced” into a configurable burst of pulses using an Acousto-Optic Modulator (AOM). This device is capable of creating an arbitrary number of pulses, as short as 10 nsec each, with a maximum repetition rate of over 40 MHz and with a contrast ratio of over 2,000:1. This contrast is critical to prevent gain drainage from amplification of low intensity light between pulses. The resulting short duration, low energy pulses (1 nJ/pulse) enter a series of five flashlamp-pumped Nd:YAG rod amplifiers, with rod diameters of 4, 5, 6.3, 9.5, and 12.7 mm, respectively. The first three amplifiers operate in a double-pass configuration for maximum gain while in the small to moderate signal gain regime. Amplifiers four and five are single-passed to minimize gain depletion throughout the duration of the pulse burst. Faraday isolators are placed between each amplifier stage to reduce gain depletion due to amplified spontaneous emission in the system. The overall gain of the 5-stage amplifier chain is estimated to be on the order of 108 with individual pulse energies in excess of 100 mJ possible.

After amplification the high-energy pulses are frequency-doubled to 532 nm (green) by use of a KTP nonlinear optical crystal. Presently, we achieve conversion efficiencies on the order of 30-40% although we feel higher efficiencies are possible with more aggressive optical arrangements. The output power also depends on the number of pulses and repetition rate as one typically adjusts the amplifier settings to produce a burst of pulses with uniform energy. For example, a burst of 68 pulses at 500 kHz will typically have an output of about 15 mJ/pulse at 532 nm. We note that we have also experimented with pulse trains in excess of 100 pulses; the total number of pulses generated is typically dictated by the number of frames that can be captured by available high-speed cameras. To generate light at 266 nm (ultraviolet), a similar harmonic conversion is performed with a BBO crystal. Due to the nonlinear nature of harmonic conversion, which is strongly dependent on fundamental pulse energy, generation of UV light is around 10% efficient, resulting in pulse energies at 266nm on the order of 2 mJ/pulse at 500 kHz.

Fig. 1 Diagram of Auburn University 3rd-generation Pulse Burst Laser system, configured for

output at 532 nm.


- 4 -

The system described above evolved from the system described in detail by Lempert et al. (1996) and Wu et al. (2000a) (Princeton University), who do an excellent job at describing the technical motivation for the pulse burst concept. Essentially, the pulse burst concept is necessary to keep the overall duty cycle of the amplifier system low such that the thermal load on the amplifiers is kept to a manageable level. Cooling is provided using recirculating water chillers and would be inadequate if the amplifiers were energized for sustained periods of time. In this original system, pulse slicing was achieved after a pre-amplifier stage using a pair of electro-optic Pockel’s cell, which rapidly rotate the polarization of the incident laser beam by 90º such that individual pulses can be picked out using a thin film polarizer. The resulting burst of pulses is then double passed through another amplifier and single passed through three additional amplifiers. In their original work, they were able to achieve ~0.4 mJ/pulse of energy at 532 nm over a burst of 30 pulses at 1 MHz repetition rate. Subsequent work by Wu and Miles (2000b) further optimized the system and increased the pulse energy by over an order of magnitude. With respect to fluid dynamic applications, this system was mainly used to perform high-speed flow visualization in supersonic flows, such as the unsteady shock motion at a compression corner (Wu and Miles, 2001).

The success of this first system was quickly followed by the development of a 2nd generation system at The Ohio State University by Thurow et al. (2004) with the common denominator being the contribution of Dr. Walter Lempert. The most substantial improvement to the 2nd generation system was the incorporation of a phase conjugate mirror, which acts as a non-linear mirror and effectively isolates the high-intensity, short-duration pulses from the low-intensity, long-duration background illumination. This relatively simple addition to the system increased the system efficiency by several factors yielding higher pulse energies after frequency conversion to 0.532 microns. The initial use of this system was for flow visualization studies of supersonic jets where the frame rates were sufficient to track turbulence structures in the shear layer (Thurow et al. 2002, 2003). This was followed by the development of a MHz rate planar Doppler velocimetry technique (Thurow et al. 2005) which exploited the narrow linewidth and frequency tuning capabilities of the laser system and demonstrated the ability to acquire velocity data in supersonic flows at MHz rates.

Soon after, Wernet and Opalski (2004) (NASA-Glenn Research Center) had a similar system built and used an 8-frame high resolution CCD camera array to demonstrate the ability to perform MHz rate particle image velocimetry on a turbulent jet. Kastengren et al. (2007) used another pulse burst laser system (U. Illinois) that, notably, was built custom by a commercial laser company to perform high-speed flow visualization of the supersonic shear layer in the wake behind a blunt-base cylinder.

More recently, a 3rd generation system was built at Auburn University (described above) by Thurow et al. (2009). This system is termed 3rd generation due to two major design changes. First, pulse slicing is achieved using an acousto-optical modulator instead of a pair of Pockel’s cells. This has significant economic advantages and also has proven to provide more flexibility with respect to the timing of laser pulses. The second change is the use of variable pulse duration flashlamp power supplies which increase the portion of time over which pulses can occur and provide for a more uniform gain profile over this period of time. This results in the ability to produce longer, more uniform bursts of pulses.

The use of variable pulse duration power supplies was also incorporated at a similar time in an upgrade to the Ohio State laser system, which also included the addition of a 7th amplifier. Their recent work (e.g. Jiang and Lempert, 2008, Jiang et al. 2008a, 2008b, 2009, Miller at al. 2009) has focused on the development of high repetition rate combustion diagnostics, such as that can be realized through NO PLIF. To facilitate this, significant effort has been put into the development of an optical parametric oscillator compatible with the pulse burst laser such that the output can be tuned to absorption lines of various molecules. This tuning capability is highly valued, particularly


- 5 -

in the combustion community, and has spawned further collaborations and the development of another pulse burst laser system at Iowa State (Miller et al. 2010). This new system is currently being used for multi-line OH PLIF. In a similar fashion, a 2nd laser system is under construction at The Ohio State University (Sutton et al. 2010) and will be used for Rayleigh and Raman scattering imaging in turbulent jets and flames.

The pulse burst laser concept has also found applications in the plasma physics community where Den Hartog (2008) has built a unique system to perform Thomson scattering diagnostics on the Madison Symmetric Torus at the University of Wisconsin. Their system strives to produce a burst of 200 pulses with 1 J of energy per pulse at repetition rates ranging from 5 – 250 kHz. While the front end of their system is similar to those described above, they utilize Nd:glass ampflication stages at the back end to produce the large pulse energies necessary for Thomson scattering.

It is also worth noting that several of these systems are semi-portable and can be used for experimentation at other laboratories. Our group recently transported our pulse burst laser system to Florida State University and demonstrated the ability to perform high-speed flow visualization of a Mach 1.5 impinging jet. Similarly, Jiang et al. (2010) and Danehy et al. (2010) utilized the Ohio State University system to perform MHz rate NO PLIF in a Mach 10 hypersonic wind tunnel at NASA Langley Research Center.

Overall, it is clear that the concept of the pulse burst laser system has experienced tremendous growth over the last decade and is capable of servicing the needs of a variety of communities ranging from traditional fluid dynamics to combustion to plasma physics. As more systems are built, the number of applications and techniques utilizing these unique capabilities are expected to grow. To date, the vast majority of diagnostics utilizing the pulse burst laser system have been variations on high repetition rate planar imaging. In this work, we highlight another utilization for this technology: 3-D flow visualization. 3. 3-D Flow Visualization System

A detailed description of the development of the 3-D flow visualization technique used here, including a discussion on the choice of scanner, synchronization of the laser, scanner and camera, and the various aspects of image processing and 3-D reconstruction techniques can be found in Thurow and Lynch (2009). A brief description of the essential features is provided here. As mentioned, 3-D imaging is accomplished using a scanning technique whereby the output of the pulse laser system is scanned through the flow field using a galvanometric scanning mirror. Images are acquired for each pulse (where each pulse illuminates a separate planar slice) using a high-speed camera. The resulting stack of images can then be reconstructed to form a 3-D visualization of the flow.

The main challenge to high-speed scanning is the speed at which pulses can be deflected through the flow field and imaged using a high-speed camera. In the course of developing the technique, electro-optic deflectors (EOD), acousto-optic deflectors (AOD), rotating mirrors and galvanometric scanning mirrors were considered. While all the devices demonstrated the speed necessary for high-speed scanning, a galvanometric scanning mirror (GSI Lumonics VM500, 6 mm aperture) was chosen as it was the most economical, efficient, quite robust and capable of scanning a beam with slew rates in excess of 10 MHz. Specifically, the mirror is able to generate scan speeds equivalent to over 10 million resolvable spots per second over a relatively long period of time such that several thousand resolvable spots could be achieved in a single scan if necessary. In this scenario, the pulse burst laser is fired as the mirror is moving and after it has accelerated to high angular velocities. As each laser pulse is ~20 nsec in duration, the position of the mirror is effectively frozen for each laser pulse. A drawback to using a scanning mirror, as compared to an acousto-optic deflector, is that the random access time is quite large, although this is not an issue when one only desires a single sequential scan through the flow.


- 6 -

In a typical experimental arrangement, the laser is first passed through a long focal length positive lens such that the beam waist coincides with the desired imaging location. The beam is then deflected off of the scanning mirror and onto a sheet forming cylindrical lens, which is placed close to the scanning mirror such that the entire scan angle (typically only a few degrees) is captured within the width of the lens. The motion of the mirror is controlled using an external voltage command input and highly repeatable such that its position can be synchronized with the output of the pulse burst laser without any perceptible jitter in space or time.

Images are captures using a DRS Hadland Ultra68 intensified high-speed camera. This camera is capable of capturing 68 frames with 220 x 220 pixel resolution at a maximum repetition rate of 500,000 frames per second. Thus, 3-D images typically acquired in our lab have 220 x 220 x 68 pixel resolution and can be acquired in 136 microseconds. We have also experimented with a Shimadzu HPV-2, which can acquire 100 frames with 312 x 260 pixel resolution such that 3-D images with 312 x 260 x 100 pixel resolution can be acquired in 100 microseconds. For the flows presented here, this is fast enough to effectively freeze the flow yielding a nearly instantaneous image. We note that the both the laser and scanner are capable of much higher speeds such that the speed of the technique is limited by the available camera. We are in the process of acquiring a camera with higher speeds and spatial resolution that should improve on some of these characteristics.

3-D image reconstruction is accomplished using a combination of Matlab for image processing and various 3-D rendering packages, such as those available in TecPlot or the open source program, imagevis3D. For flow visualization, where the flow is seeded with smoke, iso-surfaces that mark the boundary between seeded and unseeded fluid yield striking 3-D visualizations of the flow. Alternatively, we also present data in exploded assembly views which display more details about the internal structure of flow features observed in the images. Perhaps most revealing, but not amenable to traditional publications, is the ability to form 3-D animations where one can rotate the flow field to look at it from various angles or create a virtual slice through the flow at any desired location.

4. Examples of 3-D Flow Visualization

4.1 Turbulent Jets

Development of the 3-D flow visualization technique was conducted using the flow field of a simple axisymmetric jet as a test bed. The jet was produced using a simple conical nozzle with exit diameter of 1.2”. The nozzle was connected to a small stagnation chamber with air flow driven by a simple cooling fan. The fan’s flow rate was controlled to produce Reynolds numbers of 6,800 and 9,500. The boundary layer at the exit of the nozzle was laminar with a thickness of approximately 4 mm. The jet was seeded with smoke to visualize the boundary between jet core fluid and the unseeded ambient fluid.

Figure 2 shows two examples of images obtained in the near-field of the jet with ReD = 9,500. The presented iso-surfaces represent the boundary between the seeded jet and ambient air, the shape of which is determined by the underlying vortex dynamics of the jet. The images in Fig. 2 were acquired with a Shimadzu HPV-2 camera yielding 3-D intensity values over a volume with 312 x 260 x 100 pixel resolution. Each 3-D image is effectively instantaneous as the jet core flow moved less than one pixel over the 100 µsec acquisition time of the system. Figure 3 presents an alternative view showing a stack of cross-sectional views. This view is helpful in visualizing areas of the jet flow where mixing has blurred the boundary between seeded and unseeded fluid. This data was acquired at a Reynolds number of 6,800 with the Ultra68 high-speed camera yielding a 3-D resolution of 220 x 220 x 68. One impressive feature of these images is that it is difficult to tell in which direction the laser sheet was scanned; this is an indication of the quality of the 3-D


- 7 -

reconstruction process. These images illustrate the rich and complex nature of this relatively simple flow field; a brief

description is provided here. These images, acquired in the near field of the jet, depict a flow which is initially characterized by ring vortices formed near the nozzle exit. As expected, the formation of the ring vortices begins as a low amplitude perturbation of the shear layer that quickly grows into a large-scale vortex. The most striking feature of these images, however, is the onset of azimuthal instabilities, which appear as fingers of fluid elongated in the streamwise direction and are periodically spaced around the periphery of the jet. These structures are thought to be the result a secondary instability that occurs in the braid region between ring vortices resulting in pairs of counter-rotating streamwise vortices. Observations based on numerous 3-D visualizations show that the azimuthal frequency of these disturbances increases over the Reynolds number range considered here. In addition, the images show that these instabilities grow in the streamwise direction and have a streamwise extent that spans across several ring vortices. Further downstream, the complex 3-D interaction of the ring vortices with the streamwise vortices ultimately leads to a more chaotic structure of the flow and the eventual break down to the fully developed turbulent flow in the far field.

Fig. 2 Examples of 3-D flow visualization iso-surface images acquired in a Re = 9500 turbulent jet. Images have a resolution of 312 x 260 x 100 pixels and are instantaneous with respect to the fluid

motion.


- 8 -

Fig 3. Exploded cross-sectional view of 3-D flow visualization of a jet with Re = 6,800. The

bottom slice is at x/D = 0.15 and top slice is at x/D = 3.125. Color represents intensity of light scattered by smoke particles seeded into jet flow.


- 9 -

4.2 Turbulent Boundary Layers In addition to turbulent jets, we have also applied 3-D flow visualization to visualize the flow of

a zero pressure gradient (ZPG) turbulent boundary layer. The long-term goal of our work is to understand how the 3-D structure of the boundary layer changes with adverse pressure gradient and unsteady free stream conditions, particularly as they relate to the phenomena of boundary layer separation. The results presented here, however, are preliminary data and restricted to a steady, ZPG boundary layer. In many ways, this preliminary data is similar to that of Delo and Smits (1997); however, their work was limited Reθ ~ 700 due to the finite speed of their scanning system whereas in this work we have the ability to increase the Reynolds number by over an order of magnitude.

Figure 4 shows several views generated from a single 3-D data cube. The free stream velocity was 24 ft/s and the image volume was approximately 4” x 4” x 4”. The boundary layer was estimated to be ~1.5” thick with Reθ=2,500 and was seeded with smoke for visualization. In Fig. 4a, the structure of the boundary layer is irregular and consists of several structures inclined and elongated in the streamwise direction. Figure 4b shows a top down view of the same flow where the heads and streamwise extent of various structures can be seen. This single image contains a wealth of information about the boundary layer that we are in the process of studying. One particularly interesting structure is highlighted by a circle in Fig. 4a as it appears to be inclined in the upstream direction. A slice through this section is shown in 4c and a close-up view in 4d. Figure 4e shows an exploded assembly view of this region of the boundary layer with different 2-D slices through this structure being shown. This type of visualization reveals many of the inner details about the structure. With this additional information, the structure appears to be an amalgamation of several smaller structures, each inclined at about a 45º angle with respect to the flow direction, but offset to give the appearance of a structure inclined backwards. Analysis of the data in this fashion reminds one the work of Head and Bandyopadhyay (1981) who used an inclined light sheet to identify unique aspects of turbulent structures in a boundary layer. With 3-D flow visualization data, as demonstrated here, it is possible to perform this type of experiment virtually using a variety of visualization tools available with modern computer visualization software (typically designed for use with CFD results!).

More recently, we have performed experiments at a higher Reynolds number, which are of more practical interest. Specifically, we have increased the free stream velocity to 100 ft/s to achieve Reθ= 5900 (Rex = 3.3 x 106). Figure 5 presents two iso-surface images and an exploded view for three instances at this higher Reynolds number. In this case, the imaging volume was 2” x 2” x 2” and the boundary layer thickness was 1.2”, which gives a slightly closer look at the structures observed in the boundary layer. Fig 5a depicts a boundary layer consisting of a multitude of scales, but, in this instance, dominated by a relatively large-scale structure whose extent is greater than the boundary layer thickness. Fig 5b shows a different instance where the boundary layer height (as determined by smoke visualization) is increasing in the downstream direction. In this case, however, the thickening is occurring over the entire span of the imaging volume and cannot be attributed to single structure. As before, a large number of smaller scale features can also be observed. Fig. 5c shows a set of cross-sectional views at another instance where one can observe a large-scale structure whose features become taller and narrower in the downstream direction.

The images shown here clearly demonstrate the ability to apply 3-D flow visualization to a turbulent boundary layer and illustrate the wealth of information that can be obtained experimentally. At this point, however, the results are quite preliminary as we are in the early stages of learning how to interpret these images and how to use them appropriately to investigate this important flow field. We expect that the true benefit of 3-D flow visualization will not occur until we are able to vary more flow parameters (e.g. adverse pressure gradient) and observe the changes in the flow structure that occur.


- 10 -

a)

b)

c)

d)

e)

Fig 4. Several views of 3-D Flow Visualization applied to a turbulent boundary layer. a) Isometric view; b) Top down view; c) Slice through middle portion of BL; d) close up view of

interesting structure; e) exploded assembly view of structure.

Flow direction


- 11 -

a)

b)

c) Fig 5. 3-D flow visualization images of a turbulent boundary layer with Reθ= 5900

5. Conclusions and Future Work

The 3-D flow imaging technique presented here and demonstrated in turbulent jets and boundary layers clearly has tremendous potential to explore 3-D phenomena in a wide variety of flows and facilities. The fundamental limitation of most scan based imaging systems, the finite time required to scan and image the laser sheet, is improved by several orders of magnitude using the MHz rate pulse burst laser system and an ultra-high-speed camera such that 3-D measurements can be considered nearly-instantaneous for many flows. Considering the rapid growth in both laser and camera technology, it can be expected that this technology will become more accessible to different groups of researchers leading to the development of new techniques and applications.

At this point, we have only demonstrated the ability to perform 3-D flow visualization, which provides inherently qualitative results. To provide further insight into the 3-D flow visualization images, we are planning to apply traditional 2-D PIV in the near future along the center plane of 3-D measurement volume simultaneous with the 3-D imaging such that quantitative velocity field data can be used to complement the observations made in 3-D. Considering the role that traditional flow visualization techniques have played historically in our understanding of turbulent flows, 3-D flow visualization is expected to play a significant role going forward. Longer term, the scanning concept is amenable for the development of quantitative techniques such as scanning PIV, PLIF and PDV with the current limitation being the image quality (particularly signal to noise ratio) and cost

Flow Direction Flow Direction

Flow Direction


- 12 -

of commercially available high-speed cameras. Fortunately, camera technology continues to grow at a rapid pace such that these issues are working themselves out as illustrated by the increasing number of high-speed PLIF experiments being conducted with different pulse burst laser systems. With respect to 3-D variations, we are in the process of developing a 3-D density measurement technique based on the laser induced fluorescence of acetone vapor seeded into a high-speed flow field. We are also in the acquisition stages of a high resolution, ultra-high-speed camera that will allow us to pursue 3-D scanning PIV. As such, the work presented here lays the ground work for future development of quantitative 3-D diagnostics that will significantly improve upon the results presented here. Acknowledgments This research has been partially supported by the U.S. Army Research Office through a Young Investigator Program grant (Program Manager: Dr. Tom Doligalski). References Danehy P M, Ivey C B, Inman J A, Bathel B F, Jones S B, McCrea A C (2010) High-speed PLIF

imaging of hypersonic transition over discrete cylindrical roughness. AIAA Paper 2010-703, 48th Aerospace Sciences Meeting Jan 4-7 Orlando, FL.

Delo C J, Smits A J (1997) Volumetric visualization of coherent structure in a low Reynolds number turbulent boundary layer. Int. J. of Fluid Dyn., Vol.1, 1997.

Den Hartog D J, Jiang N, Lempert W R (2008) A pulse-burst laser system for a high-repetition-rate Thomson scattering diagnostic. Rev. Sci. Instrum. 79:10E736

Elsinga GE, Scarano F, Wieneke B, van Oudheusden BW (2006) Tomographic particle image velocimetry. Exp Fluids

41:933–947. Guezennec YG, Brodkey RS, Trigui J, Kent, JC (1994) Algorithms for fully automated three-

dimensional particle tracking velocimetry. Exp. Fluids 17: 209-219. Head M R, Bandyopadhyay (1981) New aspects of turbulent boundary layer structure. J. Fluid

Mech. 107:297-338. Hult J, Omrane A, Nygren J, Kaminski C F, Axelsson B, Collin R, Bengtsson P-E, Alden M (2002)

Quantitative three-dimensional imaging of soot volume fraction in turbulent non-premixed flames,” Exp. Fluids 33:265-69.

Island T C, Patrie B J, Mungal M G, Hanson R K (1996) Instantaneous three-dimensional flow visualization of a supersonic mixing layer,” Exp. Fluids 20:249-256.

Jiang N, Webster M, Lempert W (2009) Advances in generation of high repetition rate burst mode laser output. Appl. Opt. 48: B23-B31.

Jiang N, Lempert W, Slipchenko M, Miller J, Meyer T, Gord J (2008a) High-power UV lasers: tunable ultraviolet burst-mode laser system produces high-energy pulses”, Laser Focus World, 44:79-83.

Jiang N, Lempert W, Switzer G, Meyer T, Gord J (2008b) Narrow-linewidth megahertz-repetition-rate optical parametric oscillator for high-speed flow and combustion diagnostics. Appl. Opt. 47: 64-71.

Jiang N, Lempert W (2008) Ultra-high frame rate nitric oxide planar laser induced fluorescence imaging, Opt. Letters 33:2236-2238.

Jiang N, Webster M, Lempert W R, Miller J D, Meyer T R, Danehy P M (2010) MHz-rate NO PLIF imaging in a Mach 10 hypersonic wind tunnel. AIAA Paper 2010-1407, 48th Aerospace Sciences Meeting Jan 4-7 Orlando, FL.


- 13 -

Kastengren A, Dutton J C, Elliott G (2007) Large-scale structure visualization and convection velocity in supersonic blunt-base cylinder wakes. Phys. Fluids 19:015103.

Lempert W R, Wu P, Zhang B, Miles R B, Lowrance J L, Mastracola V, Kosonocky W F (1996) Pulse-burst laser system for high speed flow diagnostics. AIAA Paper 96-0500, 34th AIAA Aerospace Sciences Meeting, Reno, Nevada, 15-18 Jan. 1996.

Long M B, Yip B (1988) Measurement of three-dimensional concentrations in turbulent jets and flames. Proc. of 21st Symp. on Comb., pp. 701-709.

Miller J D, Slipchenko M, Meyer T R, Jiang N, Lempert W R, Gord J R (2009) Ultrahigh-frame-rate OH fluorescence imaging in turbulent flames using a burst-mode optical parametric oscillator. Opt. Lett. 34:1309-1311.

Miller J D, Slipchenko M, Meyer T R, Gord J R (2010) High-speed multi-line OH planar laser-induced fluorescence in unsteady flames. AIAA Paper 2010-1046, 48th Aerospace Sciences Meeting Jan 4-7 Orlando, FL.

Patrie B J, Seitzman J M, Hanson R K (1994) Instantaneous three-dimensional flow visualization by rapid acquisition of multiple planar flow images. Opt. Eng. 33:975-980.

Sheng J, Malkiel E, Katz J (2003) Single beam two-views holographic particle image velocimetry. App. Opt. 42:235-250.

Sutton J A, Patton R A, Gabet K N, Jiang N, Lempert W (2010) Towards high-repetition rate Rayleigh and Raman scattering imaging in turbulent jets and flames. 48th AIAA Aerospace Sciences Meeting, Orlando, FL, January 2010.

Thurow B S, Jiang N, Samimy M, Lempert W R (2004) Narrow-linewidth megahertz-rate pulse-burst laser for high-speed flow diagnostics. Appl. Opt. 43:5064-5073.

Thurow B, Lynch K (2009) Development of a high-speed three dimensional flow visualization technique. AIAA J. 47:2857-2865.

Thurow B, Jiang N, Lempert W, Samimy M (2005) MHz rate planar Doppler velocimetry in supersonic jets. AIAA J. 43: 500-511.

Thurow B, Samimy M, Lempert W (2003) Compressibility effects on turbulence structures in axisymmetric mixing layers. Phys. Fluids 15: 1755-1765.

Thurow B, Hileman J, Samimy M, Lempert W (2002) A technique for real-time visualization of flow structure in high-speed flows. Phys. Fluids. 14:3449-3452.

Thurow B S, Satija A, Lynch K (2009) Third-generation megahertz-rate pulse burst laser system. Appl. Opt. 48:2086-2093.

Trolinger JD, Rottenkolber M, Elandaloussi F (1997) Development and application of holographic particle image velocimetry techniques for microgravity applications. Meas. Sci. Tech. 8:1573-1583.

Wernet M, Opalski A (2004) Development and application of a MHz-frame rate PIV system. AIAA Paper 2004-2184.

Wu P, Lempert W R, Miles R B (2000a) MHz pulse-burst laser system and visualization of shock-wave/boundary-layer interaction in a Mach 2.5 wind tunnel. AIAA J. 38:672-679.

Wu PP, Miles R B (2000b) High-energy pulse-burst laser system for megahertz-rate flow visualization. Opt. Lett. 25:1639-1641.

Wu P, Miles R B (2001) Megahertz Visualization of Compression-Corner Shock Structures. AIAA J. 39: 1542-1546.

Yip B, Schmitt R L, Long M B (1988) Instantaneous, three-dimensional concentration measurements in turbulent jets and flames. Opt. Letters 13:96-98.

Documents

3-D flow imaging using a MHz-rate pulse burst LASER systemltces.dem.ist.utl.pt/lxlaser/lxlaser2010/upload/... · mechanism and number of amplifiers, can vary. The master oscillator