<%BANNER%>

Investigation of Laser-Induced Incandescence and Soot Vaporization Using Time Resolved Light Scattering in a Propane Dif...

HIDE
 Title Page
 Dedication
 Acknowledgement
 List of Tables
 List of Figures
 Abstract
 Introduction and background
 Experimental methods
 Results
 Analysis, discussion, and...
 Appendix A: Error analysis
 Appendix B: Zeroth-order lognormal...
 References
 Biographical sketch
 

PAGE 1

INVESTIGATION OF LASER-INDUCED INCANDESCENCE AND SOOT VAPORIZATION USING TIME RESOLVED LIGHT SCATTERING IN A PROPANE DIFFUSION FLAME By GREGORY DAVID YODER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Gregory David Yoder

PAGE 3

This document is dedicated to the memory of Carrie Lynn Yoder. She most unfortunately passed away in March 2003. She held a Bachelors degree from the University of Florida, a Masters degree from the University of Central Florida, and was pursuing a Ph.D. at Louisiana State University at the time of her death. She had a love for research and academia. More importantly, she had a love for life and her family. Her achievements were always a source of pride and encouragement.

PAGE 4

ACKNOWLEDGMENTS I would like to thank Dr. David Hahn for the guidance and leadership he has provided over the last two years. He allowed me to work for him on an undergraduate summer research project, which got me started in the area of laser-based diagnostics and doing graduate type research in general. He most certainly was one of the largest reasons I chose to pursue an advanced degree in mechanical engineering and to do so at the University of Florida. He outstanding teaching abilities were a major reason for my choice to stay at UF and pursue a masters degree under him. I would also like to thank Dr. Bill Lear and Dr. Bruce Carroll for serving on my committee. Their time and expertise have neither gone unnoticed nor unappreciated. I also want to thank all of my lab mates over the past two years: Dr. Jorge Carranza, Brian Fisher, Kibum Kim, Katie Masiello, Vince Hohreiter, and Allen Ball. Sharing lab and office space was always a pleasure. Lastly, and certainly not least, I would like to thank my loving parents for supporting me and always encouraging me to strive for any goal I wanted to achieve. iv

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................xi CHAPTER 1 INTRODUCTION AND BACKGROUND.................................................................1 1.1 Introduction............................................................................................................1 1.2 Motivation..............................................................................................................3 1.3 Background for Diagnostics..................................................................................4 1.3.1 Laser-Induced Incandescence......................................................................4 1.3.2 Laser Elastic Scattering...............................................................................9 1.3.3 Transmission Electron Microscopy...........................................................13 1.4 Objectives............................................................................................................14 2 EXPERIMENTAL METHODS.................................................................................16 2.1 Burner Design......................................................................................................16 2.2 Laser System........................................................................................................21 2.3 Data Acquisition System.....................................................................................25 2.4 Experimental Methods.........................................................................................28 2.4.1 Signal Linearity.........................................................................................28 2.4.2 Light Scattering with LII...........................................................................28 2.4.3 Variable Spatial Resolution.......................................................................30 2.5 TEM Methods......................................................................................................32 3 RESULTS...................................................................................................................35 3.1 Temporal Alignment............................................................................................35 3.2 Signal Linearity...................................................................................................39 3.3 Simultaneous Light Scattering and LII Measurements.......................................40 3.4 Variable Spatial Resolution LII...........................................................................46 3.5 Transmission Electron Microscopy.....................................................................52 v

PAGE 6

4 ANALYSIS, DISCUSSION, AND CONCLUSIONS...............................................54 4.1 Light Scattering...................................................................................................54 4.2 LII Analysis.........................................................................................................61 4.3 Conclusions..........................................................................................................68 4.4 Future Work.........................................................................................................69 APPENDIX A ERROR ANALYSIS..................................................................................................70 B ZEROTH-ORDER LOGNORMAL DISTRIBUTION..............................................74 LIST OF REFERENCES...................................................................................................76 BIOGRAPHICAL SKETCH.............................................................................................79 vi

PAGE 7

LIST OF TABLES Table page 2-1. Gas flow rates and diameters for diffusion burner....................................................17 2-2. Manufacturer specifications for compressed gases used in diffusion burner.............21 2-3. Description of experimental setup components.........................................................24 2-4. Summary of laser settings for light scattering measurements...................................29 2-5. Comparison of 1064 nm laser beam diameters and fluences for LII experiments....31 3-1. Summary of peak intensities for variable spatial resolution experiments.................52 A-1. Percent error analysis for LII laser set to 0.60 J/cm 2 ................................................71 A-2. Percent error analysis for LII laser set to 0.46 J/cm 2 ................................................72 A-3. Percent error analysis for LII laser set to 0.32 J/cm 2 ................................................73 vii

PAGE 8

LIST OF FIGURES Figure page 1-1. Soot formation schematic............................................................................................2 1-2. TEM schematic..........................................................................................................13 2-1. Top view schematic of diffusion burner....................................................................17 2-2. Side view schematic of diffusion burner...................................................................18 2-3. Propane diffusion flame with flame holder...............................................................18 2-4. Schematic of experimental layout.............................................................................23 2-5. Transmission calibration plot for 532 nm bandpass filter.........................................25 2-6. Laser Q-switch and flashlamp trigger timing............................................................27 2-7. Three individual soot particles photographed by TEM.............................................33 2-8. Agglomerated chain of soot particles photographed by TEM...................................34 3-1. Simultaneous laser pulses for LII laser at 0.60 J/cm 2 ................................................36 3-2. Simultaneous laser pulses for LII laser at 0.46 J/cm 2 ................................................36 3-3. Simultaneous laser pulses for LII laser at 0.32 J/cm 2 ................................................37 3-4. Laser pulses delayed by 24 ns for LII laser at 0.60 J/cm 2 .........................................38 3-5. Laser pulses delayed by 24 ns for LII laser at 0.46 J/cm 2 .........................................38 3-6. Laser pulses delayed by 24 ns for LII laser at 0.32 J/cm 2 .........................................39 3-7. Signal linearity verification plot................................................................................40 3-8. Sample average baseline measurement from scattering experiments.......................42 3-9. Integrated light scattering measurement....................................................................43 3-10. Extended data set for light scattering measurements...............................................44 viii

PAGE 9

3-11. Light scattering measurements for LII laser set to 0.60 J/cm 2 ................................45 3-12. Light scattering measurements for LII laser set to 0.46 J/cm 2 ................................45 3-13. Light scattering measurements for LII laser set to 0.32 J/cm 2 ................................46 3-14. LII plot for 0.61 J/cm 2 fluence and 2.0 mm PMT aperture.....................................47 3-15. LII plot for 0.61 J/cm 2 fluence and 3.6 mm PMT aperture.....................................48 3-16. LII plot for 0.47 J/cm 2 fluence and 2.0 mm PMT aperture.....................................48 3-17. LII plot for 0.47 J/cm 2 fluence and 3.6 mm PMT aperture.....................................49 3-18. LII plot for 0.60 J/cm 2 fluence and 2.0 mm PMT aperture.....................................50 3-19. LII plot for 0.60 J/cm 2 fluence and 3.6 mm PMT aperture.....................................50 3-20. LII plot for 0.46 J/cm 2 fluence and 2.0 mm PMT aperture.....................................51 3-21. LII plot for 0.46 J/cm 2 fluence and 3.6 mm PMT aperture.....................................51 3-22. Zeroth-order logarithmic distribution of modal diameter........................................53 4-1. Mie theory calibration plot........................................................................................56 4-2. Modal diameter as particles decay for 0.60 J/cm 2 LII laser fluence..........................57 4-3. Modal diameter as particles decay for 0.46 J/cm 2 LII laser fluence..........................58 4-4. Modal diameter as particles decay for 0.32 J/cm 2 LII laser fluence..........................58 4-5. Average particle volume as particles decay for 0.60 J/cm 2 fluence..........................59 4-6. Average particle volume as particles decay for 0.46 J/cm 2 fluence..........................60 4-7. Average particle volume as particles decay for 0.32 J/cm 2 fluence..........................60 4-8. Sample LII response with rise and decay of particle temperature.............................63 4-9. Particle temperature rise and decay model................................................................64 4-10. Particle volume experimental data and curve fit prediction....................................66 4-11. LII signal with model prediction.............................................................................67 4-12. LII signal with 2 predictive models.........................................................................68 ix

PAGE 10

B-1. ZOLD distribution for three values of .................................................................75 x

PAGE 11

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INVESTIGATION OF LASER-INDUCED INCANDESCENCE AND SOOT VAPORIZATION USING TIME RESOLVED LIGHT SCATTERING IN A PROPANE DIFFUSION FLAME By Gregory David Yoder December 2003 Chair: David W. Hahn Major Department: Mechanical and Aerospace Engineering Particulate matter (PM) and aerosol particles have become a top concern in both the realms of human health and the sustainability of the environment. A significant source of particulate matter is due to combustion processes, specifically the burning of coal and hydrocarbon fuels. As a result of these health concerns, there is a need to monitor and control the levels of these harmful particulate emissions. This thesis project reports on the use of a potential monitoring technology, and hence is a first step in finding a solution to PM-related problems. Novel laser-based diagnostics have been developed in recent years to monitor and characterize particulate matter, including soot. These laser-based techniques were applied to a laboratory diffusion burner, with the goal to ultimately apply them to a laboratory scale gas turbine engine where metallic based fuel additives will be used to control soot formation. Laser elastic light scattering (LES) and laser-induced incandescence (LII) were the two techniques used to characterize the particulate soot. xi

PAGE 12

Laser elastic scattering is a technique that allows for the determination of particle size and number density by measuring the amount of laser light scattered by the soot particles. Laser-induced incandescence is a technique that uses a pulsed laser beam to heat soot particles up to levels far above the background, causing them to emit radiation as essentially blackbodies, which can then be related to the total soot volume using suitable calibration schemes. However, the temperature reached by the laser-heated particles may cause the particles to begin to vaporize, thereby changing the parameter of interest, namely the total soot volume. The primary goal of this thesis is to investigate the particle vaporization due to LII using time-resolved laser light scattering. Based on the experimental measurements in a sooting propane diffusion flame, significant particle vaporization was found to occur on the time scale of the LII laser pulse. A model was developed to describe the particle temperature and volume as the particles are heated up and vaporized by the LII laser and as the particles subsequently cooled. This heat transfer model uses a fundamental energy balance along with the Planck distribution to model the LII signal as a function of time, and was then used to estimate the influence of particle vaporization on the time-resolved LII signal. xii

PAGE 13

CHAPTER 1 INTRODUCTION AND BACKGROUND 1.1 Introduction Particulate matter and aerosol particles have become a top concern in both the realms of human health and the sustainability of the environment. According to a 1996 study by the United States Environmental Protection Agency (U.S. EPA), particulate matter is a direct cause of increased morbidity and mortality in humans (U.S. EPA 1996). A significant source of particulate matter is due to combustion processes, specifically the burning of coal and hydrocarbon fuels. In 1996, approximately 85% of the energy used in the United States came from combustion sources (U.S. DOE 1996). Combustion products include soot, fine metallic species, and polyaromatic hydrocarbons (PAHs) that can be inhaled and are harmful to humans. Furthermore, harmful products of combustion such as oxides of nitrogen and sulfur (NO x and SO x ) can have a negative impact on the environment. NO x and SO x are key contributors to acid rain and also form particulates as nitrates and sulfates (Turns 2000). This thesis will focus on soot and on the use of laser-based diagnostics to monitor soot. Soot in simply solid carbon formed when hydrocarbon fuels are burned. Soot is formed through a four step process: (1) formation of precursor species, (2) particle inception, (3) surface growth and particle agglomeration, and (4) particle oxidation. During the formation of precursor species, chemical kinetics plays an important role. Chemical kinetics is the study of elementary reactions and their rates. Particle inception involves the formation of small particles of a critical size (3,000 10,000 atomic mass 1

PAGE 14

2 units) from growth by both chemical means and coagulation. In this step, large molecules are transformed into particles. When the small primary soot particles continue to be exposed to species of the pyrolizing flame, they experience surface growth and agglomeration. If all of the soot particles are oxidized, the flame is termed nonsooting, while; conversely, incomplete oxidation yields a sooting flame (Turns 2000). A conceptual schematic of this soot formation process is shown in Figure 1-1 below. Figure 1-1. Soot formation schematic

PAGE 15

3 1.2 Motivation Emissions from aircraft jet engines currently contain particulate matter and harmful aerosols. The levels of these emissions must be monitored and controlled to ensure the publics and the environments safety. As an example, crew members aboard aircraft carriers are subjected to jet engine exhaust on a constant basis, which is very dangerous as the particulate matter in the exhaust can be very harmful. The United States Navy currently is funding research efforts to decrease soot and particulate matter emissions from their engines. This thesis project is the first step in one of these efforts. The overall goal of such efforts is to decrease soot emissions from jet engines, including the use of fuel additives. Fuel additives have been used as soot suppressants for over 40 years, but the mechanism by which they actually suppress the soot is still a matter of debate. There are several theories as to how fuel additives, both metallic and non-metallic based, suppress soot. One theory focuses on the beginning stages of soot particle inception, and links the fuel-additive effects with changes on nucleation mechanisms of soot formation. Another theory concentrates on the role of additives as an enhancement for soot oxidation and burnout mechanisms, as a result of increased production of hydroxyl radicals, which rapidly remove soot or gaseous soot precursors. A third theory attributes the additive effects to the acceleration of the soot oxidation rate, possibly by occlusion in the soot phase. Some examples of fuel additives are ferrocene [(C 5 H 5 ) 2 Fe] and iron pentacarbonyl [Fe(CO) 5 ] (Zhang and Megaridis 1996). Novel laser-based techniques are to be developed to monitor and characterize the soot and particulate matter. These techniques will be applied first to a laboratory diffusion flame, and finally to a laboratory scale gas turbine engine. The first phase of

PAGE 16

4 the project is to model and characterize the laser-based diagnostics to be used to quantify the soot and particulate matter. The tools to be used for characterization are laser-induced incandescence (LII), laser elastic scattering (LES), and transmission electron microscopy (TEM). 1.3 Background for Diagnostics 1.3.1 Laser-Induced Incandescence Laser-induced incandescence (LII) is a well-researched technique for analyzing and characterizing sooting flames and combustion processes. Several previous research efforts have shown that LII signals are roughly proportional to soot volume fraction, which is the ratio of the volume of soot particles to the volume of gas (Schraml et al2000, Axelsson et al 2001, Wainner and Seitzman 1999). Schraml used LII to make two-dimensionally resolved measurements of soot primary particle sizes from the ratio of LII signals from two delay times. Schramls experiments were performed in a laminar ethene diffusion flame. Axelsson also used a ratio of LII signals from two time gate positions to measure soot particle size. Axelsson used a premixed flat ethane/air flame for LII experiments. Wainner reports detection of soot at better than ~1 part per trillion by use of LII techniques. Wainner chose to use a controllable soot-generating device that simulated a hot, low-soot-concentration environment similar to that of a jet engine. LII occurs when a high intensity laser beam encounters particulate matter like soot. A soot particle can absorb energy from the beam, which causes the particles temperature to increase. Soot has a very high absorptivity for visible and infrared light. At the same time, the soot can loose energy by heat transfer mechanisms to the surroundings. With sufficiently high laser energies, numerical models of the heat transfer processes indicate that the soot particles reach temperatures of 4000-4500 K (Snelling et al 1997). If the

PAGE 17

5 energy absorption rate is sufficiently high, the temperature will rise to levels where significant incandescence (essentially blackbody radiation) and vaporization can occur. Thus, LII can be used to obtain information about some average property of all the soot particles within a measurement volume (AIAA.org). A high-powered laser, typically a pulsed Nd:YAG (Neodimum: Yttrium-Aluminum-Garnett) laser operating at its fundamental wavelength of 1064 nm, is aimed through a control volume to interact with particles of unknown volume fraction. The laser beam heats up the interacting soot particles and a photomultiplier tube (PMT) is used to collect the light emitted from the particles. However, as the particles are being heated, they begin to transfer energy via conduction and radiation to their surroundings. Once the particles reach their vaporization temperature, they begin to lose mass as well. The three main energy loss mechanisms are blackbody radiation, conduction, and vaporization. Blackbody radiation accounts for the smallest losses, approximately two orders of magnitude less than vaporization, the greatest loss mechanism. Conduction losses are approximately one order of magnitude less than vaporization (Stephens et al 2003). By using basic thermodynamic principles, equations have been developed to describe the heat transfer processes that occur during LII. A fundamental energy balance describing the heat transfer processes in the interaction of a laser beam and a soot particle is given by Equation 1.1 below. 034)4)(()1()4)(()(324422dtdTCaaTTdtdMWHGKaTTaKqaaKssosbvnoaabs Equation 1.1 The terms are, respectively, the rate of laser energy absorbed per second, the rate of heat transfer to the medium (taken to be air at temperature T o ), the energy expended in

PAGE 18

6 vaporization of the carbon, the rate of energy loss by blackbody radiation, and the rate of internal energy rise (Melton 1984). From these basic equations, relationships have been developed to determine certain desired parameters of the system. For instance, the particle temperature after the laser pulse can be found using Equation 1.2 below. )/exp()1(1)(0tTTTtTgpgp Equation 1.2 In Equation 1.2 T p is the particle temperature for any given time after the laser pulse, T g is the gas temperature, more specifically the flame temperature, T p 0 is the maximum particle temperature which is reached immediately after the laser pulse, t is time, and is a characteristic cooling time defined in Equation 1.3 below. gtgpppcTcd3 Equation 1.3 In Equation 1.3, d p is the particle diameter, c p is the particle specific heat, p is the particle density, c t is the average thermal velocity of gas molecules defined by Equation 1.4 below, and g is the density of the ambient gas in the flame. ggtmkTc8 Equation 1.4 In Equation 1.4, k is the Boltzmann constant and m g is the mass of the gas molecules. The LII signal can be related to the particle temperature and size by use of the Planck distribution and is shown in Equation 1.5 found below (Roth and Filippov 1996). )],()0())(,()([*gPlanckvolPPlanckvolLIITFtPtTFtPNS Equation 1.5

PAGE 19

7 In Equation 1.5, S LII is the LII signal, is a constant parameter which represents the effect of the collection optics and the detector efficiency, N is the particle number density, P vol is the particle volume, and F Planck is a function of the Planck distribution which describes the radiation emitted by the particles. F Planck is defined below in Equation 1.6. 1)exp(51kThcCFoPlanck Equation 1.6 In Equation 1.6, h is the Planck constant, c o is the speed of light, is the detection wavelength, and k is the Boltzmann constant, and C 1 is the first radiation constant defined in Equation 1.7 below (Kock et al 2002). These equations are discussed in detail in Chapter 4. 212ohcC Equation 1.7 LII can be detected in more than one way. Prompt detection denotes setting the detector to begin looking at the signal at the moment of laser incidence, or even shortly before the laser pulse arrives. Prompt detection is purported to reduce any effects resulting from size-dependent vaporization of the soot particles. Delayed detection requires waiting until after the laser pulse to begin detecting the LII signal, typically ~30 ns after the laser pulse. This will eliminate unwanted elastic-scattered light or laser-induced fluorescence signals from polycyclic aromatic hydrocarbons or C 2 fragments (Witze et al 2001). Another variable in LII detection is the wavelength at which it is detected. Line filters and/or bandpass filters can be added to the experimental setup to only allow the detector to see a particular spectral band or range of wavelengths. This will eliminate broadband fluorescence from unwanted spectral bands. Witze used a long

PAGE 20

8 wave pass filter with a 570 nm cutoff (2001). Wainner used glass filters to monitor the LII signal between 570 and 850 nm (Wainner and Seitzman 1999). Van der Wal used an interference bandpass filter to monitor the LII signal between 400 and 450 nm (Van der Wal and Weiland 1994). Axelsson monitored specific wavelengths of 300 nm and 600 nm (Axelsson et al 2001). There are several conditions that need to be satisfied for the detected LII signal to be proportional to soot volume fraction: (1) the probed soot consists of single or loosely aggregated primary particles that are small compared to the wavelengths of the laser excitation (such that Rayleigh-limit for light absorption and emission is valid); (2) the peak particle temperatures reached during the laser pulse are relatively insensitive to the particle diameter; (3) the soot particle mass vaporization is either negligible or largely independent of particle diameter; and (4) the detected LII signal is dominated by thermal emission occurring during laser excitation or shortly thereafter, such that size dependent conductance cooling does not influence the signal (Witze et al 2001). The criteria for particles to be within the Rayleigh limit for detection is they must have a size parameter and the quantity |m| much less than one. The size parameter is defined in Equation 1.7 below, while m is the complex refractive index of the soot particles. a2 Equation 1.7 In Equation 1.7, a is the particle radius and is the laser wavelength. One characteristic of LII that makes it particularly advantageous for soot volume fraction detection is the rapid rise in LII signal with increasing laser fluence until a nearly constant plateau signal strength is reached. This is useful in strongly absorbing environments because it eliminates the need for corrections to the measured LII signal

PAGE 21

9 that are due to laser beam absorption, as long as the initial laser fluence is suitably far into the plateau regime (Witze et al 2001). 1.3.2 Laser Elastic Scattering Laser elastic scattering is a well-researched technique that can be used to determine particle size and concentration using laser-based diagnostics. This technique has been used many times in the past for soot characterization (Hahn and Charalampopoulos 1992, Witze el al 2001, Dasch 1984). Hahn used laser light scattering in an earlier metal additive study when looking at a premixed propane/oxygen flame and the influence of iron additives (1992). Witze also used LES when performing LII experiments in flames. Witzes study was focused on time-resolved LII measurements using a rectangular profile beam in a propane/air diffusion flame (2001). Dasch used a continuous wavelength Argon ion probe beam for light scattering in an LII study of a methane/oxygen flame (1984). Other earlier work that utilized laser scattering in flames was done by Menna (Menna and DAlessio 1982), Erickson (Erickson et al 1964),and Prado (Prado et al 1981). The scattering of electromagnetic (EM) waves is due to heterogeneities. Heterogeneities are disruptions in the field of the electromagnetic wave. At some level, all matter presents heterogeneities to the EM waves. These heterogeneities or obstacles can be atoms, molecules, or particles. The EM wave will induce an oscillation within the electrons of the obstacle, called an oscillating dipole moment. The charges on the molecules will oscillate at the same forcing frequency as that at which the EM wave is oscillating. This oscillating molecule becomes a source of radiation, which occurs at the same frequency of the induced oscillation. When scattered light is emitted at the same

PAGE 22

10 frequency as the incident light, it is called elastic scattering. There are two regimes in laser elastic scattering: Rayleigh and Mie scattering. Lord Rayleigh (1871) published a solution for light scattering by small, dielectric, spherical particles. Small particles are defines by d p <<, where d p is the particle diameter and is the wavelength of the laser light. Dielectric particles are simply non-absorbing particles. The Rayleigh solution states that the dipole moment radiates in all directions. Furthermore, the electric field is proportional to the dipole moment, which is proportional to the volume of the scatterer. Lastly, Rayleigh stated that light scattering has strong, inverse wavelength dependence. Rayleigh developed several equations to describe the scattering of light. Rayleigh scattering is defined by several differential scattering cross sections (area/steradian), which represent the amount of energy scattered into a given solid angle about a given direction. The differential scattering coefficients for a spherical particle in the Rayleigh regime are given by the expressions 222622214mmCVV Equation 1.8 2cosVVHHCC Equation 1.9 The VV and HH subscripts on the above coefficients denote vertically or horizontally polarized light of the incident and scattered light, respectively. If the light is vertically polarized incident to the scattering particle, then the light will still be vertically polarized when it is scattered. The Rayleigh total scattering and absorption cross sections may be written, respectively, as

PAGE 23

11 222622132mmCscat Equation 1.10 21Im2232mmCabs Equation 1.10 The extinction cross section, C ext represents the total energy removed by the scattering particle and is comprised of the sum of the scattering cross section and the absorption cross section. Since must be much less than one to be considered in the Rayleigh regime, the scattering cross section can be neglected since it is of the sixth order of while the absorption cross section is third order in (Kerker 1969). In 1908 Mie developed a generalized, exact solution for the scattering of light by a single, homogeneous sphere of arbitrary size, while Rayleighs solution is only valid for small Kerker 1969). The differential scattering cross section according to Mie theory is given by 2122)(cos)(cos)1(124nnnnnVVbannnC Equation 1.11 where the parameters and are the angular dependent functions, n n sin)(cos)(cos)1(nnP Equation 1.12 ddPnn)(cos)(cos)1( Equation 1.13 of the associated Legendre polynomial of argument cos. is defined as the angle of observation and is measured from the forward direction to the scattering direction. The parameters a )(cos)1(nP n and b n are defined as

PAGE 24

12 )()()()()()()()(nnnnnnnnmmmmmma Equation 1.14 )()()()()()()()(nnnnnnnnmmmmmmb Equation 1.15 The functions and n n are defined as )()2()(2/12/1zJzznn Equation 1.16 )()()()2()2(2/12/1zizzHznnnn Equation 1.17 where is given by )(zn )()2()(2/12/1zYzznn Equation 1.18 Here and Y are the half integral order Bessel functions of the first and second kind, or Bessel and Neumann functions, respectively, and is the half integral order Hankel function of the second kind. )(2/1zJn )(2/1zn )()2(2/1zHn Simply stated, LES involves sending a laser beam through a control volume containing particles of unknown size and concentration. The laser light will be scattered off the particles in all directions, as prescribed by Rayleigh or Mie theory. A detector or detectors are setup to collect the scattered light, with the signal received being proportional to both the size of the particles and the particle number density. One downfall of LES is the need for very precise alignment to avoid stray light. The optical layout must be arranged very carefully to ensure a quality signal by eliminating stray light and background noise, as described in Chapter 2.

PAGE 25

13 1.3.3 Transmission Electron Microscopy Transmission electron microscopy (TEM) is a very well researched method for viewing microscopic particles. A schematic of the inner workings of a transmission electron microscope is shown in Figure 1-2. The "Virtual Source" at the top represents the electron gun, producing a stream of monochromatic electrons. This stream is focused to a small, thin, coherent beam by the use of condenser lenses 1 and 2. The first lens (usually controlled by the "spot size knob") largely determines the "spot size"; the general size range of the final spot that strikes the sample. Figure 1-2. TEM schematic The second lens (usually controlled by the "intensity or brightness knob" actually changes the size of the spot on the sample; changing it from a wide dispersed spot to a

PAGE 26

14 pinpoint beam. The beam is restricted by the condenser aperture (usually user selectable), knocking out high angle electrons (those far from the optic axis, the dotted line down the center). The beam strikes the specimen and parts of it are transmitted. This transmitted portion is focused by the objective lens into an image. Optional Objective and Selected Area metal apertures can restrict the beam; the Objective aperture enhancing contrast by blocking out high-angle diffracted electrons, the Selected Area aperture enabling the user to examine the periodic diffraction of electrons by ordered arrangements of atoms in the sample. The image is passed down the column through the intermediate and projector lenses, being enlarged all the way. The image strikes the phosphor image screen and light is generated, allowing the user to see the image. The darker areas of the image represent those areas of the sample that fewer electrons were transmitted through (they are thicker or denser). The lighter areas of the image represent those areas of the sample that more electrons were transmitted through (they are thinner or less dense). This technique has been widely used for soot analysis. Zhang used thermophoretic sampling to acquire soot samples and used TEM to analyze the soot particles (Zhang and Megaridis 1996). Fotou also used TEM for analysis of soot in a fuel additive study when using a diffusion flame (Fotou et al 1995). Also, Dobbins used thermophoretic sampling to study the morphology of flame-generated soot (Dobbins and Megaridis 1987). 1.4 Objectives The primary objective of this project is to characterize and model the LII signal from a propane/oxygen diffusion flame. This model will then be used in future studies geared toward understanding the reduction of soot emissions using metal additives. A novel technique was developed using a two-laser setup that enabled simultaneous LII

PAGE 27

15 excitation and LES measurements of the heated soot particles. This experimental design is helpful in reproduction and implementation for further work. Below is a list of detailed objectives. 1. Understand the role of laser-particle interactions with respect to quantitative LII measurements of soot volume. 2. Quantify soot emissions with light scattering for correlation with LII data to optimize the accuracy of the LII method. 3. Explore the temporal profile of soot particle vaporization with a goal of enhancing in situ Raman spectroscopy of metallic species by peeling away soot.

PAGE 28

CHAPTER 2 EXPERIMENTAL METHODS 2.1 Burner Design The burner used in this experiment was designed to provide a stable, optically accessible flame that is consistent with the combustor of a gas turbine engine. A gas turbine combustor typically employs a fuel lean diffusion burner. The laboratory diffusion burner used in this experiment was fabricated from stainless steel tubing and Swagelok fittings. A top view schematic of the burner is shown in Figure 2-1, a side view schematic of the burner is shown in Figure 2-2, and a photograph of the burner setup with the propane diffusion flame running is shown in Figure 2-3. The burner has three separate, controllable gas flows. The innermost flow tube contains propane gas, controlled by a precision-drilled ruby orifice and a pressure regulator. This Bird Precision 0.0040 orifice maintains the propane flow rate at 0.31 LPM while the line pressure is kept at 50 psi. The annular flow surrounding the propane is oxygen. The oxygen flow rate is maintained by an Alicat Scientific precision gas flow controller. The oxygen volumetric flow rate was held constant at 1.82 LPM for all experiments. The outermost gas flow is nitrogen. The nitrogen flow rate was maintained by an Alicat Scientific precision gas flow controller. The nitrogen volumetric flow rate was held constant at 1.05 LPM for all experiments. A summary of the gas flow rates is given in Table 2-1. The nitrogen shroud flow holds two purposes. First, it shields the inner gases from the ambient air to allow precise control over the fuel stoichiometry. Secondly, the shroud flow aids in flame stability. It is very important to maintain a stable flame so the 16

PAGE 29

17 lasers will always strike the flame and strike it in the same location every laser pulse. In addition to the shroud flow, a 0.1-inch stainless steel mesh flame holder was used for flame stability. Figure 2-1. Top view schematic of diffusion burner Table 2-1. Gas flow rates and diameters for diffusion burner Flow Tube Gas Diameter Flow Rate D1 Propane 1/16" 0.314 LPM D2 Oxygen 9/32" 1.815 LPM D3 Nitrogen 1" shroud 1.052 LPM

PAGE 30

18 Figure 2-2. Side view schematic of diffusion burner Figure 2-3. Propane diffusion flame with flame holder

PAGE 31

19 The flame holder was positioned 1.6 above the burner. The flame holder grabs the flame and maintains it in a steady position with respect to the horizontal and the vertical planes. There are currently no fuel additive capabilities for this burner, but additional components can be added. A plexiglass shield was placed around the entire flame setup to protect the flame from drafts. Maintaining flame stability was a priority to ensure a repeatable laser/soot interaction. The shield had two holes cut in it to allow the laser beams to enter and exit, and a third hole at 90 o to allow the scattering and LII signal to reach the PMT. The region around the shield that was at the same height as the laser beams was coated with a black fabric to eliminate reflections and to cut down on stray light. Gas turbine engines run their combustors on the fuel lean side of stoichiometric. The stoichiometric quantity of oxidizer is just that amount needed to completely burn a quantity of fuel (Turns 2000). Fuel lean refers to having excess oxidizer or a shortage of fuel. Gas turbines run fuel lean due to the large amount of available air flowing through the engine and the high cost of fuel and the cost of carrying large amounts of fuel. Also, burning slightly fuel lean decreases the flame temperature, which decreases NO x and SO x formation. For stability and to replicate the combustion in a gas turbine engine, the diffusion flame in these experiments was run on the fuel lean side of stoichiometric. The governing combustion reaction for stoichiometric propane combustion is given by Equation 2.1 below. OdHcCObOHaC22283 Equation 2.1 This basic reaction neglects secondary products of combustion, free radicals, and dissociation effects. Nitrogen is not included in the reaction because the nitrogen shroud

PAGE 32

20 is assumed to not mix significantly with the propane and oxygen flows and does not directly affect the flame stoichiometry. The coefficients a, b, c, and d in Equation 2.1 are found based on the initial reactant parameters. Two cases will be considered. The first case is the stoichiometric case. For stoichiometric combustion, one mole of fuel is reacted with b moles of oxidizer to create c and d moles of products. A simple atom balance will result in a stoichiometric reaction shown in Equation 2.2 below, which yields an air to fuel ratio of 5 on a molar basis for propane. OHCOOHC22283435 Equation 2.2 The second case is the experimental case. In the diffusion flame, the volumetric flow rates of fuel and oxidizer were 0.314 LPM and 1.815 LPM, respectively. For an ideal gas, the volumetric ratio is equal to the molar ratio. Therefore, the experimental combustion reaction can be written as shown in Equation 2.3 below. 222283245.256.1942.0815.1314.0OOHCOOHC Equation 2.3 The fuel equivalence ratio is the determining factor as to whether a flame is fuel rich or fuel lean. The definition for fuel equivalence ratio is given in Equation 2.4 below. erimentaltricstoichiomeFAFAexp// Equation 2.4 A / F experimental is defined as the actual air to fuel ratio for the reactants. The air to fuel ratio for the stoichiometric case and the experimental case can be found by using Equations 2.2 and 2.3. These ratios are calculated in Equations 2.5 and 2.6 below. 51/5/tricstoichiomeFA Equation 2.5 78.5314.0/815.1/experimentalFA Equation 2.6

PAGE 33

21 The fuel equivalence ratio can now be found using Equation 2.4. is calculated in Equation 2.7 below. 87.078.5/5 Equation 2.7 With a fuel equivalence ratio less than one, the flame is running fuel lean. This is the desired stoichiometry to match that of a gas turbine combustor. Also of significance is that soot particles undergo significant oxidation (i.e. burnout) in actual combustors, hence the current fuel lean flame better emulates this condition. Table 2-2 provides the manufacturer specifications for the compressed gases used in the experiments. Table 2-2. Manufacturer specifications for compressed gases used in diffusion burner. Gas Supplier Description Nitrogen Praxair Industrial Grade, 99.7% N 2 H 2 O < 32 ppm O2 Balance (nominal) Oxygen Praxair Medical Grade, 99.5% O 2 H2O < 20 ppm Propane Praxair 99.5 wt% (liquid phase) H 2 O < 300 ppmw Ethane < 600 ppmw Propylene < 400 ppmw n-Butane < 20 ppmw Isobutane < 3000 ppmw Sulfer < 1 ppmw 2.2 Laser System The diagnostics for these experiments were performed using a two-laser setup. A Q-switched 1064 nm Nd:YAG laser was used for LII and soot vaporization. A frequency doubled Q-switched 532 nm Nd:YAG laser was used for the laser light scattering portion of the experiments. The 1064 nm laser light was focused using a lens with a 500 mm focal length. The 532 nm laser light was focused using a lens with a 250 mm focal length. A schematic showing the layout of the laser and optics system is shown in Figure

PAGE 34

22 2-4. The two lasers are both directed through the flame such that the paths cross when both beams are focused down to their minimum diameters and are on the central axis of the burner tube. The 1064 nm beam is considerably larger in diameter when the beams cross paths, approximately four times the diameter of the 532 nm beam. The 532 nm beam is centered in the middle of the 1064 nm beam. This allows for a large volume of soot particles to be heated up, creating the incandescent signal while vaporizing the particles, and a much smaller volume of particles within the larger incandescent volume to be used for light scattering. This technique ensures the scattering signal is only from particles that have been exposed to the high-powered 1064 nm beam. The laser light is collected using a photomultiplier tube (PMT) to create a signal for analysis. The laser light that is collected is both scattered 532 nm light and the incandescent light created by the heating of particles by the 1064 nm laser. The emitted light is collected at a 90-degree angle from the 1064 nm beam. The light is directed through two very small apertures to reduce background noise and to ensure the light collected is only from the small scattering volume in the flame. After the first aperture, the light is sent through a Newport bandpass filter centered at 532 nm. This filter effectively removes all light that is not at the 532 nm wavelength. This filter has a 10 nm full width half maximum (FWHM) and transmits approximately 55% of the light within the FWHM. The transmission versus wavelength calibration plot for this filter is shown in Figure 2-5. A summary of all equipment used is shown in Table 2-3.

PAGE 35

23 Figure 2-4. Schematic of experimental layout

PAGE 36

24 Table 2-3. Description of experimental setup components Device Manufacturer Model Description Lasers and Electronics 1064 nm Nd:YAG laser Big Sky Laser Ultra Q-switched, 5 Hz, Variable Power FWHM = 13 ns, 50 mJ max 532 nm Nd:YAG laser Continuum Minilite ML-II Q-switched, 5 Hz, 2.4 mJ/pulse FWHM = 5 ns, 20 mJ max Photomultiplier Tube Hamamatsu 1P28 PMT Ocsilloscope LeCroy LT 372 500 MHz, 4 GS/s digital oscilloscope Delay Generator Stanford Research DG 535 Programmable Delay Generator Instruments Voltage Supply Stanford Research PS325 Digital High Volage Power Supply Instruments Optics 532 nm Lens Newport KBX079AR.14 BBAR coated, 430-700 nm 25.4 mm diameter 250 mm focal length 1064 nm Lens Newport KBX082AR.18 BBAR coated, 1000-1550 nm 25.4 mm diameter 500 mm focal length 1064 nm Mirror CVI Laser Y1-2037-45UNP 45 degree, 1064 nm dielectric mirror 532 Mirror Spindler & Hoyer 34-0467-000 45 degree, 532 nm dielectric mirror Bandpass Filter Newport 10LF10-532 10nm FWHM, Transmission > 50% 25.4 mm diameter

PAGE 37

25 Figure 2-5. Transmission calibration plot for 532 nm bandpass filter 2.3 Data Acquisition System The signal from the laser light was collected using a photomultiplier tube (PMT). A PMT converts photons to an electrical signal. A PMT consists of a photocathode and a series of dynodes in an evacuated enclosure. When a photon of sufficient energy strikes the photocathode, it ejects a photoelectron due to the photoelectric effect. The photocathode material is usually a mixture of alkali metals, which make the PMT sensitive to photons throughout the visible region of the electromagnetic spectrum. The photcathode is at a high negative voltage, typically to volts. The PMT was charged at volts for all experiments in the present work. The photoelectron is accelerated towards a series of additional electrodes called dynodes. These electrodes are

PAGE 38

26 each maintained at successively less negative potentials. Additional electrons are generated at each dynode. This cascading effect creates 10 5 to 10 7 electrons for each photoelectron that is ejected from the photocathode. The amplification depends on the number of dynodes and the accelerating voltage. This amplified electrical signal is collected at an anode at ground potential (Tissue 2000). A Stanford Research Systems high voltage power supply was used to charge the PMT. Precise control of the laser flashlamps and Q-switches was required to eliminate jitter (i.e. variation between the LII laser and the LES laser). Lasers have an internal timing system that signals the flashlamp to start and signals the Q-switch to open to release the laser pulse. This timing is not always precise and the laser pulse can vary by up to 10 nanoseconds from shot to shot. This variation of the laser pulse from shot to shot is called jitter. For these experiments, in order to provide the desired temporal resolution it was necessary to completely remove the jitter by externally triggering the flashlamp and the Q-switch on both lasers. This was done using 2 separate delay generators. The first delay generator was internally triggered at a repetition rate of 5 Hz. This delay generator triggered the LII lasers flashlamp and Q-switch. The first delay generator was also used to trigger a second delay generator. The second delay generator was used to trigger the 532 nm lasers flashlamp and Q-switch. The Q-switch signal of the second delay generator was also used to trigger the oscilloscope. A schematic showing the timing details of the laser system is shown in Figure 2-6. The signal out of the PMT was sent to a LeCroy digital oscilloscope. The oscilloscope was triggered by the 532 nm laser Q-switch.

PAGE 39

27 Figure 2-6. Laser Q-switch and flashlamp trigger timing

PAGE 40

28 2.4 Experimental Methods 2.4.1 Signal Linearity Light scattering experiments can have signals that vary by more than an order of magnitude in a single measurement. Signal linearity refers to a detectors ability to accurately report signals over a wide dynamic range. If a detector can accurately decipher signals at low levels, but gets saturated when the signal increases, the reported data will be invalid. In these experiments, soot is being evaporated and the signals are decaying to levels up to 15 times less than the original signal. It is very important that the detector, the PMT in this case, be able to interpret the signals accurately. Signal linearity tests were performed before every set of data was recorded. These tests were done by first taking an average scattering measurement for 500 laser shots. A 50 % transmission neutral density filter of optical density equal to 0.3 was then placed inline with the PMT signal line of sight and another scattering measurement was recorded. If the signal were truly linear, the ratio of the two measurements would be The PMT voltage was adjusted to achieve this ratio. It is noted that PMTs are generally characterized by excellent linearity, often over many decades. However, with ~10 ns laser pulses, the intense burst of photons readily saturates PMTs, hence linearity must be carefully checked. 2.4.2 Light Scattering with LII The objective of the light scattering with LII experiments was to quantify the size of the soot particles as they are heated up and vaporized. Three different 1064 nm laser powers were used. The laser powers were 37.5, 29.0, and 20.0 mJ/pulse. These powers correspond to laser fluences of 0.60, 0.46, and 0.32 J/cm 2 respectively, based on a focal spot of 5.94 mm 2 The laser fluence is simply the laser energy in each pulse divided by

PAGE 41

29 the beam area. The beam area was found by ablating ink off of a slide and measuring the removed ink area. The 1064 nm beam spot was not quite a perfect circle, but was more elliptical and had a major axis measuring 3.0 mm, and a minor axis measuring 2.67 mm. The power of the 532 nm laser was 2.4 mJ/pulse. A summary of the laser beam power properties is given in Table 2-4. Table 2-4. Summary of laser settings for light scattering measurements Laser Control Setting Power (mJ/pulse) Fluence (J/cm 2 ) 1064 10 37.5 0.60 1064 8 29.0 0.46 1064 6 20.0 0.32 532 Constant 2.4 N/A According to previous research as described in Chapter 1, LII provides information that can be translated into soot volume-fraction data. The precise time the LII signal is captured is a very important factor when considering these experimental results. As the LII laser heats up the soot particles they begin to emit radiation, which is the LII signal, but they also begin to vaporize once their vaporization temperature is reached. This vaporization may lead to skewed results if the particles being measured are being destroyed at the same time. These current experiments track the size of the soot particles as they are heated up and vaporized such that the time scale of vaporization and the LII process is better understood, enabling a more thorough analysis of the LII process and hopefully to more accurate implementation. Monitoring the soot particle size while they are being vaporized was done by using a second laser for light scattering measurements. For convienience, the 1064 nm laser

PAGE 42

30 used for the laser heating will be refered to as the LII laser, while the 532 nm laser used for light scattering will be refered to as the LES laser. The LII laser was held fixed in time by delay generator 1. The LES laser was moved in time relative to the LII laser by use of delay generator 2. The LES laser was fired 24 nanoseconds before the LII laser to begin the experiment. This gave a measure of the size of the soot particles before any vaporization occurs. The LES laser was then moved up in time in 2 nanosecond intervals until it reached 20 ns after the LII laser pulse. The LES laser was then increased further in time by larger intervals until it eventually reached 132 ns beyond the LII laser pulse. From these measurements, the size history of the soot particles was tracked as they were heated up and vaporized by the LII laser with a temporal resolution of ~2 ns. 2.4.3 Variable Spatial Resolution Once light scattering measurements were concluded, LII measurements were taken to quantify soot vaporization and the LII response for multiple scenarios. A new aperture was added to the system to allow for a decrease in the beam diameter of the 1064 nm laser. With the new aperture in place, the beam diameter at the point of focus in the flame was 1.55 mm. To make a comparison to the laser fluences from the previous experiments, the laser power was adjusted for the new beam diameter to achieve similar fluences. Two laser powers were used in the LII experiments, 11.5 and 8.8 mJ/pulse. These laser powers correspond to laser fluences of 0.61 and 0.47 J/cm 2 respectively. These two laser fluences are very similar to the two highest laser fluences from the previous experiments. Therefore, there are comparable laser fluences for both the large beam diameter and the smaller beam diameter. A comparison showing the fluences and diameters is given in Table 2-5 below.

PAGE 43

31 Table 2-5. Comparison of 1064 nm laser beam diameters and fluences for LII experiments Beam Diameter (mm) Control Setting Power (mJ/pulse) Fluence (J/cm^2) 2.75 10 37.5 0.60 2.75 8 29.0 0.46 1.55 6 11.5 0.61 1.55 4.5 8.8 0.47 LII measurements were then taken for all four laser fluences. Another variable added to the experiment was the size of the detector aperture. This aperture controlled the amount of light allowed into the PMT to create the signal. For a larger aperture, more light was allowed into the detector creating a larger signal. Also, the scattering volume is larger for a larger aperture, thus that signal accounts for more soot particles. The first aperture that decreased the beam diameter of the 1064 nm laser was placed before any optics and before the laser beam reached the flame, therefore it only controlled the laser fluence. The second aperture was set at either 2.0 mm or 3.6 mm. The second aperture must be opened from its original position of approximately 0.5 mm from the light scattering experiments to achieve a strong LII signal due to the markedly weaker LII signal as compared to the LES signal. These LII experiments were done with the PMT charged with V. The flame was kept at the same fuel-lean condition. Signal linearity measurements were taken before each set a data was recorded. For each data set, 500 laser pulses were averaged to determine the LII signal.

PAGE 44

32 2.5 TEM Methods Samples of the primary soot particles from the flame were taken using a small copper mesh grid with a plastic coating. Several samples were taken by sweeping the grid through the flame allowing soot to deposit on the grid. This is referred to as thermophoretic sampling, where the temperature gradient between the hot flame and the cold mesh drives the particles to the surface of the mesh where they are deposited. The grids were attached to a grid holder and manually moved through the flame. Each sample had a slightly different residence time in the flame, all less than ~1 second. If the samples were kept in the flame too long, the plastic coating on the grid could be melted and destroy the sample. If the samples were swept through too fast, there might not be enough time for soot particles to deposit on the grid. The samples were taken at the same height above the burner that the laser beams cross through the flame. This ensured the TEM samples were a representative cross-section of the particles analyzed by laser-based diagnostics. By use of a transmission electron microscope (TEM), digital photographs were taken of the soot samples so that measurements could be made to determine primary particle size. Each photograph had a scale indicating length, and particle diameters could be simply measured from the pictures. The major axis length, minor axis length, and an estimate of the average diameter were all recorded for 30 soot samples. From these measurements, the statistical average and standard deviation for the soot primary particle size was determined. Two representative sample TEM photos are shown in Figures 2-7 and 2.8. The soot particles in Figure 2-7 are single loose particles. The soot in Figure 2-8 is an agglomerated chain of soot particles.

PAGE 45

33 Figure 2-7. Three individual soot particles photographed by TEM

PAGE 46

34 Figure 2-8. Agglomerated chain of soot particles photographed by TEM

PAGE 47

CHAPTER 3 RESULTS 3.1 Temporal Alignment The first task in setting up the experiments was getting precise temporal control between the two lasers. In order to eliminate jitter, both lasers were controlled using separate digital delay generators. The 532 nm (light scattering) laser was kept at a constant power of 2.4 mJ/pulse, but the 1064 nm laser power (LII laser) was varied between three different powers. For each laser, the delay between the flashlamp and Q-switch was optimized and remained constant. The first delay generator was used to trigger the flashlamp and Q-switch of the LII laser. This delay generator also triggered the second delay generator, which then was used to trigger the flashlamp and Q-switch of the 532 nm laser. By varying the delay time between the triggering of the second delay generator and triggering of the flashlamp, the pulse separation between the two pulses was controlled with jitter between the two pulses of about 1 ns. Because the LII laser pulse energy was controlled by adjusting the flashlamp voltage, the temporal pulse timing was slightly altered. Therefore, for each pulse energy, the digital delay settings were determined for zero pulse separation. These settings were stored, and recalled whenever the pulse energy was changed. Note the pulse width of the 532 nm scattering laser (~5 ns) is less than the pulse width of the LII laser (~13 ns). Figures 3-1, 3-2, and 3-3 show the two laser pulses on top of each other for their respective delay times. 35

PAGE 48

36 -0.00200.0020.0040.0060.0080.010.0120.014020406080100 532 nm 1064 nmIntensity (a.u.)Time (ns) Figure 3-1. Simultaneous laser pulses for LII laser at 0.60 J/cm 2 -0.00200.0020.0040.0060.0080.010.012020406080100 532 nm 1064 nmIntensity (a.u.)Time (ns) Figure 3-2. Simultaneous laser pulses for LII laser at 0.46 J/cm 2

PAGE 49

37 -0.00200.0020.0040.0060.0080.010.012020406080100 532 nm 1064 nmIntensity (a.u.)Time (ns) Figure 3-3. Simultaneous laser pulses for LII laser at 0.32 J/cm 2 The light scattering experiments all began with the 532 nm laser fired 24 ns before the LII laser to get a baseline recording representing the soot before any vaporization or incandescence caused by the LII laser. Figures 3-4, 3-5, and 3-6 show the difference between the two laser pulses when the 532 nm laser is set 24 ns ahead of the LII laser for the three LII laser fluences of 0.60, 0.46, and 0.32 J/cm 2 respectively.

PAGE 50

38 -0.00200.0020.0040.0060.0080.010.0120.014-20020406080100Intensity (a.u.)Time (ns) Figure 3-4. Laser pulses delayed by 24 ns for LII laser at 0.60 J/cm 2 -0.00200.0020.0040.0060.0080.010.012020406080100Intensity (a.u.)Time (ns) Figure 3-5. Laser pulses delayed by 24 ns for LII laser at 0.46 J/cm 2

PAGE 51

39 -0.00200.0020.0040.0060.0080.010.012-20020406080100Intensity (a.u.)Time (ns) Figure 3-6. Laser pulses delayed by 24 ns for LII laser at 0.32 J/cm 2 3.2 Signal Linearity Signal linearity measurements were taken to ensure the PMT was recording accurate data over a wide range of signals. The light scattering signals fluctuated by more than an order of magnitude during measurements and this variation must be recorded on a linear scale by the PMT to be able to draw fair conclusions from the data. To test the signal linearity, an LII measurement was taken without the light scattering laser running, then a 50% transmission neutral density filter was added to the collection optics and another LII measurement was taken without the light scattering laser running. The ratio of the signals was calculated to determine if the detector was operating in a linear range. This linearity test was performed before every set of experiments. A

PAGE 52

40 sample plot showing the signal with and without the filter is shown in Figure 3-7 below. The sample plot is taken from an LII experiment. -0.00500.0050.010.0150.020.025020004000600080001 104 No Filter With FilterIntensity (a.u.)Time (a.u.) Figure 3-7. Signal linearity verification plot As shown in Figure 3-7, the ratio of the peaks of the signals is 0.53. This ratio confirms that the recorded data is linear over a wide range of signals. For all experiments, the average signal linearity ratio was 0.54, with a standard deviation of 0.054. 3.3 Simultaneous Light Scattering and LII Measurements Light scattering measurements were taken for various LII laser powers to monitor the size of the soot particles, enabling determination if the soot particles were vaporized by the LII laser. Data measurements were started when the 532 nm scattering laser was 24 ns ahead of the LII laser in time. This gave a signal that represented the soot before any vaporization occurred. The 532 nm laser was moved forward in time (i.e. toward the LII laser pulse) in 2 ns increments until it reached 20 ns after the LII laser pulse. The

PAGE 53

41 time increments were then increased and measurements were taken until the 532 nm laser was 132 ns after the LII laser pulse. Since a 532 nm bandpass filter was used in front of the PMT, the incandescent signal from the LII laser did not affect the much stronger scattering signal from the 532 nm laser. This was verified by the absence of any signal for the LII laser alone. Only if the apertures entering the PMT were opened and the voltage gain on the PMT was raised significantly was the LII laser pulse detectable. Hence for these measurements, the LII laser was used simply to vaporize the soot particles and contributed no spurious signal to the light scattering signal. In order to eliminate any minimal contribution from the LII laser, mainly electromagnetic interference, baseline measurements were taken before and after the scattering measurements. These two baseline measurements were averaged and subtracted from the scattering data during data analysis. A sample averaged baseline measurement is shown in Figure 3-8.

PAGE 54

42 -0.0002-0.000100.00010.00021.4 10-71.6 10-71.8 10-72 10-72.2 10-72.4 10-72.6 10-7Intensity (a.u.)Time (ns) Figure 3-8. Sample average baseline measurement from scattering experiments Once the baseline measurements were averaged, the baseline was subtracted from the scattering data to give a corrected value. The integrated peak area was then calculated corresponding to a temporal width of about 13 data points or about 3.3 ns for the current digitization rate of 4 GS/s. The integrated values were then plotted as a function of time to show the decay of the soot particles as they were vaporized by the LII laser. Figure 3-9 shows a single set of light scattering measurements.

PAGE 55

43 -0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.10-30-20-100102030Intensity (a.u.)Time (532 nm laser pulse relative to 1064 nm laser pulse) Figure 3-9. Integrated light scattering measurement Figure 3-9 is a light scattering measurement from an experiment where the scattering laser was only increased in time until it reached 20 ns after the LII pulse. This type of experiment was repeated eight times for each of the three LII powers to create a statistically accurate average and standard deviation for the scattering experiments. It was determined that more data was required to follow the decay of the soot particles long after the LII laser pulse. Therefore, another set of experiments was performed in which the initial scattering measurements were taken beginning when the scattering laser was 24 ns ahead of the LII laser, then the scattering laser was moved in 2 ns increments until it reached 20 ns after the LII laser. The scattering laser was then moved in 2 to 16 ns increments away from the LII pulse. Measurements were continued to a delay of about 132 ns, where there appeared to be no additional changes in the signal. A sample plot of the extended data set is shown in Figure 3-10 below.

PAGE 56

44 -1-0.8-0.6-0.4-0.20-50050100150Intensity (a.u.)Time (532 nm laser pulse relative to 1064 nm laser pulse) Figure 3-10. Extended data set for light scattering measurements Because the two sets of scattering data were collected several months apart, four data points were selected to overlap between the two sets of scattering experiments. These overlapping points along with the initial data point (-24 ns) were used to match the 2 data sets and merge them into one comprehensive data set. The absolute difference in the common region between the two data sets was about a factor of 2.3. This constant scale factor was then used to match the data. The comprehensive data sets for the three LII laser powers are plotted in Figures 3-11, 3-12, and 3-13. Note that the scattering data are normalized to an initial value of unity at the initial delay of ns.

PAGE 57

45 00.20.40.60.811.2-50050100150Intensity (a.u.)Time (532 nm laser pulse relative to 1064 nm laser pulse) Figure 3-11. Light scattering measurements for LII laser set to 0.60 J/cm 2 00.20.40.60.811.2-50050100150Intensity (a.u.)Time (532 nm laser pulse relative to 1064 nm laser pulse) Figure 3-12. Light scattering measurements for LII laser set to 0.46 J/cm 2

PAGE 58

46 00.20.40.60.811.2-50050100150Intensity (a.u.)Time (532 nm laser pulse relative to 1064 nm laser pulse) Figure 3-13. Light scattering measurements for LII laser set to 0.32 J/cm 2 3.4 Variable Spatial Resolution LII As discussed above, the aperture to the PMT was closed to a minimum size (<1 mm) to prevent any LII signal or stray light while the scattering was recorded. Upon completion of the scattering measurements, LII data were recorded in the absence of the 532 nm scattering laser. The 532 nm line filter was retained, hence the LII signal corresponds to a bandwidth centered at 532 nm, as shown in Figure 2-5. LII experiments were performed for various setups. Two separate apertures were controlled to either alter the LII laser beam diameter or alter the amount of light that entered the PMT to provide a signal. With the first aperture (laser aperture) closed down to shrink the LII laser beam diameter, the laser power was decreased to keep the laser fluence at a similar level to that of the nominal diameter LII laser beam. The second aperture (PMT aperture) controlled the amount of light entering the PMT, therefore controlling the optical volume imaged onto the PMT detector.

PAGE 59

47 For each LII response recorded, the offset (no laser) was subtracted from the signal to zero the baseline of the data. Each experiment was repeated 12 to 15 times until a statistically accurate average and standard deviation was calculated for the experiments. This ensemble average of many experiments is the data reported below. The first set of experiments was run with the first aperture set to shrink the LII laser beam diameter to 1.6 mm. Two laser powers were used to create two laser fluences of 0.61 J/cm 2 and 0.47 J/cm 2 These two laser fluences were run using two different PMT aperture sizes. The first aperture diameter used was 2.0 mm, and the larger aperture diameter was 3.6 mm. The LII response for these experiments is shown in Figures 3-14 through 3-17 below. -0.00500.0050.010.0150.020.025-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-14. LII plot for 0.61 J/cm 2 fluence and 2.0 mm PMT aperture

PAGE 60

48 -0.0200.020.040.060.080.1-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-15. LII plot for 0.61 J/cm 2 fluence and 3.6 mm PMT aperture -0.00200.0020.0040.0060.0080.010.0120.014-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-16. LII plot for 0.47 J/cm 2 fluence and 2.0 mm PMT aperture

PAGE 61

49 -0.0100.010.020.030.040.050.060.07-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-17. LII plot for 0.47 J/cm 2 fluence and 3.6 mm PMT aperture After this first set of LII experiments, the first aperture was opened back up to allow the LII beam to return to its original diameter of 2.75 mm. The laser was then run using two different laser powers to create laser fluences of 0.596 J/cm 2 and 0.461 J/cm 2 These two fluences are within 2% of the fluences from the first LII experiments with the reduced beam diameter. The same PMT aperture sizes were used as in the first set of LII experiments. The LII response for these experiments is shown in Figures 3-18 through 3-21 below.

PAGE 62

50 -0.0100.010.020.030.040.050.060.07-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-18. LII plot for 0.60 J/cm 2 fluence and 2.0 mm PMT aperture -0.0500.050.10.150.2-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-19. LII plot for 0.60 J/cm 2 fluence and 3.6 mm PMT aperture

PAGE 63

51 -0.0100.010.020.030.040.050.06-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-20. LII plot for 0.46 J/cm 2 fluence and 2.0 mm PMT aperture -0.0500.050.10.150.2-5000500100015002000Intensity (a.u.)Time (ns) Figure 3-21. LII plot for 0.46 J/cm 2 fluence and 3.6 mm PMT aperture A summary of the peak intensities for each of the LII plots shown in Figures 3-14 through 3-21 is shown in Table 3-1 below.

PAGE 64

52 Table 3-1. Summary of peak intensities for variable spatial resolution experiments Fluence (J/cm^2) LII Beam Diameter (mm) PMT Aperture Size (mm) Peak Intensity Figure # 0.61 1.55 2.0 0.020 3.13 0.61 1.55 3.6 0.082 3.14 0.47 1.55 2.0 0.014 3.15 0.47 1.55 3.6 0.064 3.16 0.60 2.75 2.0 0.066 3.17 0.60 2.75 3.6 0.190 3.18 0.46 2.75 2.0 0.058 3.19 0.46 2.75 3.6 0.170 3.20 3.5 Transmission Electron Microscopy The transmission electron microscopy results gave data that could be used to determine the primary particle size. 35 individual digital photographs were taken using the TEM. Each of these photographs contained an absolute scale indicating length. Using a ruler to measure the size of the particles, a list of the particle sizes was compiled. 27 separate particle sizes were used to statistically determine the average and standard deviation of the primary particle size. A zeroth-order logarithmic distribution (ZOLD) was used to analyze the TEM data. The ZOLD function can be described as 2222)/(lnexp2)2/exp()(ommooaaaap Equation 3.1 In Equation 3.1, a is the particle diameter, a m is the modal diameter, and is a dimensionless measure of width used in the ZOLD analysis (Hahn et al 1995). It was found from the TEM data that the mean particle diameter was 101 nm with a standard deviation of 48 nm. Using the ZOLD analysis, the best fit was determined for a modal

PAGE 65

53 particle diameter of 75 nm, with =0.45. A plot of the probability of finding a particle of a given size is shown in Figure 3-22 below. For the ZOLD function, the modal diameter and are related to the mean and true standard deviation through Equations 3.2 and 3.3 shown below. )5.1exp(20maa Equation 3.2 2/12020)3exp()4exp(ma Equation 3.3 Hence it is seen that the modal value of 75 nm and yield the mean of 101 nm and the standard deviation of 48 nm as given by the TEM data. 00.0020.0040.0060.0080.010.0120100200300400500600ProbabilityModal Diameter (nm) Figure 3-22. Zeroth-order logarithmic distribution of modal diameter

PAGE 66

CHAPTER 4 ANALYSIS, DISCUSSION, AND CONCLUSIONS 4.1 Light Scattering Light scattering measurements reveal a decay in the soot particle scattering signal following the firing of the LII laser pulse. The decrease in scattering signal is attributed to soot particle vaporization. This decay in scattering signal can be correlated to a decrease in particle diameter using scattering theory. To make this correlation, several steps must be taken to deduce the particle diameter from the data. First, the TEM results must be used to determine the primary particle size before any laser interaction. Then using this baseline diameter as a starting value for the scattering measurements, the signal is related to the differential scattering cross-section to correlate the light scattering signal to the particle diameter as a function of time. As given in Chapter 3, the modal diameter for the soot particles is 75 nm based on TEM analysis and the ZOLD fit. This result came from the TEM analysis and the zeroth-order logarithmic distribution fit. The differential scattering cross-section may be calculated using Mie theory, and is a function of the modal diameter, complex refractive index, scattering wavelength, and scattering collection angle. Using the TEM result for modal diameter of 75 nm, a complex refractive index of m = 1.6 0.6i (Smyth and Shaddix 1996), scattering wavelength of 532 nm, and scattering collection angle of 90 o the scattering cross-section can be calculated, which corresponds to the soot particles before the LII laser interaction. The measured scattering signal is a function of the collection optics, solid angle, the detector efficiency, differential scattering cross-section, 54

PAGE 67

55 and particle number density. The equation for the scattering signal is given by Equation 4.1 below. NCSVVVV** Equation 4.1 In equation 4.1, is the signal,is an efficiency associated with the collection optics, solid angle, and the detector efficiency, is the differential scattering cross-section (cm VVS VVC 2 sr -1 ), and N is the number density (particles/cm 3 ). The VV subscript on the signal S and the differential scattering cross-section C refer to the polarization state of the incident laser light and the scattered light, which are both vertically polarized with respect to the horizontal scattering plane. Taking the ratio of the signal for any time versus the signal for t=0 (before any laser interaction with the soot particles) will result in Equation 4.2. VV )0(**)0()(**)(tCNtStCNtSVVVVVVVV Equation 4.2 Since the collection optics do not change between experiments, may be canceled out in Equation 4.2. Also, since the diffusion time scale is long (on the order of microseconds) compared to the temporal range of scattering data (~150 ns), the particles can be considered frozen in time during the measurements, therefore the number density is constant and can be cancelled out in Equation 4.2. The result after cancellation and rearrangement is shown in Equation 4.3 below. )()0()(*)0()(tftStStCtCVVVVVVVV Equation 4.3 Now the differential scattering cross-section of the particles is simply a function a time. The differential scattering cross-section at time t=0 was found from the TEM

PAGE 68

56 analysis, and the ratio of the signal at any time versus the signal at time t=0 can be easily found from the experimental data. With the differential scattering cross-section known for all times, Mie theory can be used to correlate the differential cross-section to a modal diameter. Using Mie theory with the given scattering wavelength, scattering collection angle, and complex refractive index, a calibration plot can be formed relating modal diameter and differential scattering cross-section for a constant value of = 0.45. This calibration plot is shown below in Figure 4-1. 02040608010012001 10-122 10-123 10-124 10-125 10-126 10-127 10-12Modal Diameter (nm)Differential Scattering Cross Section (cm^2/sr) Figure 4-1. Mie theory calibration plot Using a fourth order polynomial curve fit, an equation for the modal diameter as a function of differential scattering cross-section was calculated. This curve fit is shown in Equation 4.4 below. 432mod*48083.1*365491.8*254993.2*131057.4017.34VVVVVVVValCECECECED Equation 4.4 This curve fit equation yields a correlation coefficient of R=0.99968.

PAGE 69

57 The particle modal diameter can now be found directly from the scattering signal as the particles decay due to the LII laser induced vaporization. A plot of the particle diameter as a function of time for each of the three LII laser fluences is shown in Figures 4-2 through 4-4 below. 304050607080-50050100150Modal Diameter (nm)Time (532nm laser relative to 1064 nm laser) Figure 4-2. Modal diameter as particles decay for 0.60 J/cm 2 LII laser fluence

PAGE 70

58 404550556065707580-50050100150Modal Diameter (nm)Time (532 nm laser relative to 1064 nm laser) Figure 4-3. Modal diameter as particles decay for 0.46 J/cm 2 LII laser fluence 4550556065707580-50050100150Modal Diameter (nm)Time (532 nm laser relative to 1064 nm laser) Figure 4-4. Modal diameter as particles decay for 0.32 J/cm 2 LII laser fluence The nondimensional particle volume fraction, which represents the fraction of gas occupied by solid soot particles, can be found from Equation 4.5 below.

PAGE 71

59 03)(34avdaapaNf Equation 4.5 Substituting in the zeroth-order logarithmic distribution (ZOLD) for p(a), integrating from a=0 to infinity, and dividing both sides of Equation 4.5 by N, the particle number density, the average particle volume can be found according to Equation 4.6. )3exp(3423ovvolaNfP Equation 4.6 In Equation 4.6, P vol is the average particle volume, a is the mean particle diameter, and is a dimensionless measure of width used in the ZOLD. The mean diameter can be related to the modal diameter through Equation 4.7. )5.1exp(2omaa Equation 4.7 Converting modal diameters to average particle volume yields the plots of average particle volume versus relative time as shown in Figures 4-5 through 4-7. 1 1062 1063 1064 1065 1066 1067 1068 1069 106-50050100150Particle Volume (nm^3)Time (532 nm laser relative to 1064 nm laser) Figure 4-5. Average particle volume as particles decay for 0.60 J/cm 2 fluence

PAGE 72

60 1 1062 1063 1064 1065 1066 1067 1068 1069 106-50050100150Particle Volume (nm^3)Time (532 nm laser relative to 1064 nm laser) Figure 4-6. Average particle volume as particles decay for 0.46 J/cm 2 fluence 2 1063 1064 1065 1066 1067 1068 1069 106-50050100150Particle Volume (nm^3)Time (532 nm laser relative to 1064 nm laser) Figure 4-7. Average particle volume as particles decay for 0.32 J/cm 2 fluence

PAGE 73

61 4.2 LII Analysis LII measurements were taken for various laser fluences and PMT aperture sizes. The results of these experiments were presented in Chapter 3. One goal of this project was to be able to predict the LII signal and the influence of particle vaporization. From heat transfer theory it has been shown in Chapter 1 that the LII signal can be described as a function of soot particle volume and particle temperature. The equation describing the relationship between the LII signal and its dependent parameters is given by Equation 1.5 shown below. )],()0())(,()([*gPlanckvolPPlanckvolLIITFtPtTFtPNS Equation 1.5 In Equation 1.5, S LII is the LII signal, is a constant parameter which represents the effect of the collection optics, solid angle, and the detector efficiency, N is the particle number density, P vol is the average particle volume, and F Planck is a function of the Planck distribution which describes the radiation emitted by the particles. F Planck is defined below in Equation 1.6. 1)exp(51kThcCFoPlanck Equation 1.6 In Equation 1.6, h is the Planck constant, c o is the speed of light, is the detection wavelength, k is the Boltzmann constant, T is the temperature of the emitting particle, and C 1 is the first radiation constant defined by Equation 1.7. 212ohcC Equation 1.7 P vol (t) and F Planck (,T P (t)) in Equation 1.5 are the time dependent particle volume and Planck function. The Planck function is dependent on the detection wavelength and the temperature of the soot particles, which are treated as perfectly emitting blackbodies.

PAGE 74

62 The time dependent Planck function in Equation 1.5 uses the decaying particle temperature to model the LII signal. P vol (t=0) and F Planck (T g ) are initial values and are constant in the model. P vol (t=0) is the initial particle volume found from light scattering measurements before any LII laser interaction (i.e. t = -24 ns). F Planck (T g ) is found by using the initial soot particle temperature, which is assumed to equal the flame temperature. T g is assumed to be the constant-pressure adiabatic flame temperature for stoichiometric propane combustion, which is approximately 2260 K (Turns 2000). The time resolved particle temperature T P (t) is found from Equation 1.1 shown below. )/exp()1(1)(0tTTTtTgpgp Equation 1.2 The maximum particle temperature, T, and the characteristic cooling time, were varied to find an appropriate curve fit to model the LII signal. It was determined through an iterative process that the best values to most accurately represent the LII signal were T= 4800 K, and = 900 ns. This maximum particle temperature is close to values previously reported by Snelling (Snelling et al 1997). Snelling showed through numerical models of the heat transfer process that the maximum particle temperature should be between 4000-4500 K. Equation 1.2 was used only to describe temperature decay region of the LII signal. The portion of the LII signal modeled by Equation 1.2 is labeled Decay in a sample LII response in Figure 4-8 below. 0p 0p The particle temperature rise from the gas temperature of 2260 K to the maximum particle temperature of 4800 K was modeled as a linear rise in temperature. The rise in temperature is labeled in Figure 4-8 as rise. This temperature rise occurs over a period of about 30 ns, compared to the exponential decay of temperature, which takes

PAGE 75

63 approximately 1700 ns to return to the gas temperature from the maximum particle temperature. The constant *N used in Equation 1.5 was found to be 2550. A plot of the particle temperature rise and decay for a laser fluence of 0.47 J/cm 2 and a PMT aperture opening of 2.0 mm is shown in Figure 4-9 below. -0.00200.0020.0040.0060.0080.010.0120.014-5000500100015002000Intensity (a.u.)Time (ns) Decay Rise Figure 4-8. Sample LII response with rise and decay of particle temperature

PAGE 76

64 2000250030003500400045005000-5000500100015002000Temperature (K)Time (ns) Figure 4-9. Particle temperature rise and decay model The particle volume was found from light scattering measurements at discrete points in time before, during, and after the LII laser pulse. These points were then curve fit to be able to determine the particle volume continuously for the entire duration of the LII laser pulse, which was much longer than the initial light scattering measurements. The complex nature of the particle volume decay did not allow a simple exponential or polynomial curve fit. Therefore, the particle volume plot was broken up and analyzed piecewise to find a set of equations to describe the decay of the particles. All times used to describe the decay are relative times. More specifically, the LII laser was held constant in time as the LES laser was precisely controlled relative to the LII laser. From ns to -2 ns, a third order polynomial curve fit was used to describe the particle volume. From -2 ns to 10 ns, another third order polynomial curve fit was used to model the particle volume. From 10 ns to 24 ns, a final third order polynomial curve fit was

PAGE 77

65 used to model the particle volume. From 24 ns to the end of the LII signal, approximately 1700 ns, a power law curve fit was used to model the very slight decay of the particle volume over this additional range. After 24 ns, the data from the light scattering measurements mostly flattened out due to the absence of the LII laser pulse, which was the cause of the particle decay. However, the particles are still hotter than the gas temperature and are most likely gradually decaying via vaporization, thus a perfectly constant approximation is considered insufficient to model the particle volume. A plot of the modeled particle volume along with the light scattering data of the experimental particle volume is shown in Figure 4-10 below.

PAGE 78

66 0.0010.0020.0030.0040.0050.0060.0070.0080.0090200400600800100012001400 Experimental Data Curve FitParticle Volume (um^3)Time (ns) Figure 4-10. Particle volume experimental data and curve fit prediction Using the above models and curve fits to describe the particle temperature and particle volume, they can be applied to Equation 1.5 to predict the LII signal. A plot of the LII response for a laser fluence of 0.47 J/cm 2 and a PMT aperture opening of 2 mm along with the predicted model plot is shown in Figure 4-11 below.

PAGE 79

67 -0.00200.0020.0040.0060.0080.010.0120.014-5000500100015002000 LII Signal ModelIntensity (a.u.)Time (ns) Figure 4-11. LII signal with model prediction The model used for the prediction of the LII signal in Figure 4-11 uses the particle volume data obtained from light scattering measurements with an LII laser fluence of 0.46 J/cm 2 which is identical within experimental precision to the fluence used in the LII experiment. The LII laser effectively evaporates the soot particles during the time course of the LII laser pulse. The laser heats up the particles and they lose mass as they are heated past their vaporization temperature (~ 4000 K for carbon). One interesting application of this predictive model is to see what would happen to the LII signal if the particle volume did not change due to the LII laser. In this case, the term P vol (t) in Equation 1.5 would be a constant equal to the initial particle volume as determined from light scattering measurements, instead of a time dependent quantity found from the light scattering measurements. A plot showing this case for P vol = 0.00777 m 3 is shown in Figure 4-12 below. Clearly, the effect of soot particle vaporization is significantly decreasing the LII

PAGE 80

68 signal by about a factor of 4. Because the vaporization time scale is on the order of the LII pulse, even prompt LII will not avoid vaporization effects. -0.0200.020.040.060.08-5000500100015002000 LII Signal Variable Pvol Model Constant Pvol ModelIntensity (a.u.)Time (ns) Figure 4-12. LII signal with 2 predictive models 4.3 Conclusions In this study, a propane diffusion flame burner was constructed to develop laser-based diagnostics to monitor soot emissions. Time resolved laser light scattering (LES) and laser-induced incandescence (LII) were the two laserbased techniques, along with transmission electron microscopy (TEM), used to determine primary particle size and soot volume fraction. From these emission parameters, a model was created to predict

PAGE 81

69 the LII signal as a function of particle volume and particle temperature. The following conclusions have been drawn from the results of this study: 1. LII for the range of laser fluences used in these experiments results in the vaporization of soot particles. The time scale of vaporization is on the order of the length of the laser pulse (~13 ns). Whether prompt or delayed LII detection is used, the size of the soot particles will be affected by the LII laser, thereby changing the parameter of interest, namely the soot volume fraction. 2. LII is a technique that has a strong dependence on laser fluence and particle size, thus careful calibration is required for quantitative analysis. 3. Varying the spatial resolution of the LII laser and collection optics showed a nearly linear LII response. This indicates the LII signal is a nearly linear function of solid collection angle. Consequently, LII is a robust technique that can be utilized for various experimental conditions to measure soot volume fraction, notwithstanding the above comments. 4. Modeling of the LII signal is a complicated process, but provides an accurate representation of the particle temperature and volume. From these parameters, the soot volume fraction can be deduced. 4.4 Future Work It would be useful to extend the work to other flames with different primary particle size, and different optical properties (i.e. absorption coefficients) to see the effect of these parameters on the degree of soot vaporization and the correlation with the LII signal. In addition, it would be interesting to extend the present technique to a post-flame regime where the ambient gas temperatures are significantly lower. It is expected that such lower initial temperatures may reduce the degree of vaporization.

PAGE 82

APPENDIX A ERROR ANALYSIS The modal diameters calculated in Chapter 4 were done so by equating voltage from the light scattering signal to a differential scattering cross section, which was then converted into a modal diameter via Mie theory. These scattering signals were the average of eight separate trials taken over a period of two weeks. From the collection of data, the average and standard deviation of the signal was calculated. The percent error in the modal diameter can be calculated from using the standard deviations of the scattering signal. The percent error must be calculated for one standard deviation above, and one standard deviation below the average. The percent error is given by Equation A.1. % Error = %100*aaa Equation A.1 In Equation A.1, a is the modal diameter found using the average scattering signal, and a is either the modal diameter found from using the average scattering signal plus the standard deviation or minus the standard deviation. Thus, two percent errors were calculated and averaged to find the actual percent error. Three tables displaying the percent errors for the three different LII laser fluences are shown in Tables A-1 through A-3. 70

PAGE 83

71 Table A-1. Percent error analysis for LII laser set to 0.60 J/cm 2 Relative Time (ns) Svv Standard Deviation Modal Diameter (nm) Percent Error -22 0.964121 0.028245 100.9539 0.74 -20 1.009101 0.0648052 102.0167 1.26 -18 0.998578 0.086225 101.7925 1.78 -16 1.018703 0.0977756 102.2067 1.68 -14 0.96774 0.120399 101.0488 2.87 -12 0.93843 0.107887 100.2397 2.98 -10 0.968945 0.0693905 101.0801 1.73 -8 0.901033 0.0485525 99.09403 1.56 -6 0.794376 0.1047874 95.40911 3.90 -4 0.560083 0.1466562 86.94477 6.22 -2 0.328735 0.153763 77.51989 10.83 0 0.163637 0.0565334 66.33132 7.71 2 0.093394 0.0327758 59.12107 6.49 4 0.068699 0.0183006 56.10044 4.18 6 0.057276 0.0126266 54.60639 3.09 8 0.051171 0.0118403 53.78157 3.01 10 0.049484 0.0111555 53.55033 2.86 12 0.049203 0.0080787 53.51171 2.08 14 0.048683 0.0106221 53.44009 2.74 16 0.047629 0.0118138 53.29446 3.07 18 0.046044 0.0105041 53.07439 2.76 20 0.047447 0.0102373 53.26916 2.66 24 0.048814 0.0043942 53.45813 1.13 28 0.045849 0.0013186 53.04723 0.35 32 0.044327 0.0025912 52.83455 0.69 36 0.048723 0.0026363 53.44561 0.68 44 0.046157 0.0037812 53.09003 0.99 52 0.042222 0.0024192 52.53836 0.65 60 0.047182 0.0043895 53.23249 1.14 68 0.047508 0.0026639 53.2777 0.69 84 0.043978 0.002237 52.78555 0.59 100 0.047832 0.0030861 53.32248 0.80 116 0.048285 0.0028879 53.3851 0.75 132 0.048909 0.0020071 53.47117 0.52

PAGE 84

72 Table A-2. Percent error analysis for LII laser set to 0.46 J/cm 2 Relative Time (ns) Svv Standard Deviation Modal Diameter (nm) Percent Error -22 0.989674 0.0773909 101.5905 1.71 -20 0.974015 0.0473466 101.2099 1.17 -18 0.984151 0.0583743 101.4598 1.35 -16 0.957297 0.0906289 100.7709 2.35 -14 0.947542 0.0584807 100.5007 1.62 -12 0.939889 0.0932681 100.2821 2.59 -10 0.951412 0.0925583 100.6091 2.46 -8 0.935118 0.0842756 100.143 2.40 -6 0.891693 0.1230473 98.79191 3.87 -4 0.820649 0.075761 96.35206 2.77 -2 0.640676 0.0609746 89.82995 2.44 0 0.415088 0.055366 81.46595 2.84 2 0.253487 0.056232 73.19938 5.02 4 0.158022 0.0320715 65.82219 4.48 6 0.129522 0.032047 63.06621 5.19 8 0.110267 0.0180694 61.03023 3.25 10 0.10129 0.0203646 60.0295 3.85 12 0.097852 0.0143981 59.6373 2.77 14 0.093214 0.0163419 59.10008 3.23 16 0.091265 0.0156811 58.8715 3.14 18 0.087018 0.0097238 58.36762 1.99 20 0.089179 0.0142775 58.62507 2.89 24 0.088014 0.0074189 58.48657 1.51 28 0.084588 0.0076578 58.07578 1.59 32 0.087256 0.0053711 58.39606 1.10 36 0.087551 0.0061878 58.43133 1.26 44 0.086635 0.0075917 58.32184 1.56 52 0.086225 0.0059996 58.27273 1.23 60 0.090065 0.0087306 58.7299 1.76 68 0.08585 0.0044483 58.22766 0.92 84 0.086545 0.0062248 58.31106 1.28 100 0.0861 0.0042985 58.25771 0.89 116 0.086467 0.0056508 58.30166 1.16 132 0.090958 0.0087678 58.83537 1.76

PAGE 85

73 Table A-3. Percent error analysis for LII laser set to 0.32 J/cm 2 Relative Time (ns) Svv Standard Deviation Modal Diameter (nm) Percent Error -22 0.977224 0.0875407 101.2904 2.06 -20 1.049187 0.09297765 102.7076 1.10 -18 0.967168 0.11762424 101.0339 2.82 -16 1.085333 0.19473843 103.0692 4.39 -14 1.066238 0.21627485 102.9123 4.99 -12 0.930257 0.25733729 99.99912 5.64 -10 0.922373 0.22749247 99.76153 5.58 -8 0.970817 0.20496063 101.1284 4.03 -6 0.933981 0.19739844 100.1095 4.88 -4 0.88749 0.17978068 98.65418 5.41 -2 0.771167 0.14911987 94.56743 5.57 0 0.660326 0.09893379 90.53732 3.94 2 0.53499 0.12982527 86.0416 5.61 4 0.420302 0.0984546 81.68077 5.07 6 0.329307 0.07984077 77.54906 5.34 8 0.269421 0.06982805 74.20451 5.86 10 0.241374 0.07191752 72.39756 6.81 12 0.220468 0.0645295 70.93058 6.70 14 0.204267 0.0486971 69.71562 5.43 16 0.200596 0.05489329 69.43025 6.24 18 0.193481 0.04204589 68.86605 4.93 20 0.19106 0.04886439 68.67076 5.80 24 0.188227 0.00910239 68.43994 1.09 28 0.181295 0.00725069 67.86486 0.90 32 0.184343 0.00564552 68.11951 0.69 36 0.189649 0.00965329 68.55604 1.15 44 0.17846 0.00451513 67.62534 0.57 52 0.188507 0.00611587 68.46286 0.73 60 0.183394 0.01010787 68.04056 1.24 68 0.1872 0.01217863 68.35565 1.47 84 0.19162 0.00750837 68.7161 0.88 100 0.189216 0.01098262 68.52079 1.31 116 0.182177 0.01012132 67.93879 1.25 132 0.186291 0.01487665 68.28079 1.80

PAGE 86

APPENDIX B ZEROTH-ORDER LOGNORMAL DISTRIBUTION The zeroth-order lognormal distribution (ZOLD) is a skewed probability distribution. The ZOLD is used for modeling the distribution of aerosols and soot particles due to those particles being skewed to larger sizes. A given system of particles, each of diameter a, with true mean diameter and standard deviation, a and can be represented by the ZOLD using the parameters of the modal diameter and a dimensionless measure of width, a m and respectively. The probability distribution is given by Equation B.1 below. 2222)/(lnexp2)2/exp()(ommooaaaap Equation B.1 The mean particle diameter is found from Equation B.2 and the standard deviation is found from Equations B.3 and B.4. )5.1exp()(*20omadaapaa Equation B.2 daapaa)()(202 Equation B.3 2/122)3exp()4exp(ooma Equation B.4 In the ZOLD, p(a) is normalized, shown mathematically by Equation B.5. 01)(daap Equation B.5 A plot of the ZOLD for three different values of is given in Figure B-1 below. 74

PAGE 87

75 -0.00500.0050.010.0150.020.0250.03050100150200250300350400 0.2 0.5 0.8Modal Diameter (nm)p(a) Figure B-1. ZOLD distribution for three values of

PAGE 88

LIST OF REFERENCES American Institute of Aeronautics and Astronautics Website (2003). http://www.AIAA.org Date accessed: 8/30/2003 Axelsson, B., R. Collin, and P.E. Bengtsson (2001). Laser-Induced Incandescence for Soot Particle Size and Volume Fraction Measurements Using On-line Extinction Calibration. Applied Physics B, 72, 367-372 Dasch, C.J. (1984). Continuous-Wave Probe Laser Investigation of Laser Vaporization of Small Soot Particles in a Flame. Apllied Optics, 23, 2209-2215 Dobbins, R.A., and C.M. Megaridis (1987). Morphology of Flame-Generated Soot as Determined by Thermophoretic Sampling. Langmuir, 3, 254-259 Erickson, W.D., G.C. Williams, and H.C. Hottel (1964). Light Scattering Measurements on Soot in a Benzene-Air Flame. Combustion and Flame, 8, 127-132 Fotou, G.P., S.J. Scott, and S.E. Pratsinis (1995). The Role of Ferrocene in Flame Synthesis of Silica. Combustion and Flame, 101, 529-538 Hahn, D.W., and T.T. Charalampopoulos (1992). The Role of Iron Additives in Sooting Premixed Flames. Twenty-Fourth Symposium (Int.) on Combustion, The Combustion Institute, 1007-1014 Hahn, D.W., M.N. Ediger, and G.H. Pettit (1995). Ablation Plume Particle Dynamics During Excimer Laser Ablation of Polymide. Journal of Applied Physics, 77, 2759-2766 Kerker, M. (1969) The Scattering of Light and Other Electromagnetic Radiation. New York: Academic. Kock, B.F., Th. Eckhardt, and P. Roth (2002). In-Cylinder Sizing of Diesel Particles by Time-Resolved Laser-Induced Incandescence (TR-LII). Proceedings of the Combustion Institute, 29, 2775-2781 Melton, L.A. (1984). Soot Diagnostics Based on Laser Heating. Applied Optics, 23, 2201-2207 76

PAGE 89

77 Menna, P., and A. DAlessio (1982). Light Scattering and Extinction Coefficients for Soot Forming Flames in the Wavelength Range from 200 to 600 nm. Nineteenth Symposium (International) on Combustion, The Combustion Institute, 1421-1428 Prado, G., J. Jagoda, K. Neoh, and J. Lahaye (1981). A Study of Soot Formation in Premixed Propane/Oxygen Flames by In Situ Optical Techniques and Sampling Probes. Eighteenth Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, Pennsylvania, 1127-1136 Rayleigh, Lord (1871). Nature, 3, 234-265 Roth, P., and A.V. Filippov (1996). In Situ Ultrafine Particle Sizing by a Combination of Pulsed Laser Heatup and Particle Thermal Emission. J. Aerosol Sci., 27, 95-104 Schraml, S., S. Dankers, K. Bader, S. Will, and A. Leipertz (2000). Soot Temperature Measurements and Implications for Time-Resolved Laser-Induced Incandescence (TIRE-LII). Combustion and Flame, 120, 439-450 Smyth, K.C., and C.R. Shaddix (1996). The Elusive History of m = 1.57 0.56i for the Refractive Index of Soot. Combustion and Flame, 107, 314-320 Snelling, D.R., G.J. Smallwood, I.G. Campbell, J.E. Medlock, and O.L. Gulder (1997). Development and Application of Laser-Induced Incandescence (LII) as a Diagnostic for Soot Particulate Measurements. AGARD 90 th Symposium of the Propulsion and Energetics Panel on Advanced Non-Intrusive Instrumentation for Propulsion Engines, Brussels, Belgium Stephens, M., N. Turner, and J. Sandberg (2003). Particle Identification by Laser-Induced Incandescence in a Solid-State Laser Cavity. Applied Optics, 42, 3726-3736 Tissue, B.M. (2000). http://www.chem.vt.edu/chem-ed/optics/detector/pmt.html Date accessed: 9/3/2003 Turns, S.R. (2000). An Introduction to Combustion. Boston: McGraw-Hill U.S. Department of Energy (1996). Annual Energy Review. DOE/EIA-0384(96) U.S. Environmental Protection Agency (1996). Federal Register 61, 77, 17357-17358 Vander Wal, and K.J. Weiland (1994). Laser-Induced Incandescence: Development and Characterization Towards a Measurement of Soot-Volume Fraction. Applied Physics B, 59, 445-452 Wainner, R.T., and J.M. Seitzman (1999). Soot Measurements in a Simulated EngineExhaust Using Laser-Induced Incandescence. AIAA Journal, 37, 738-742

PAGE 90

78 Witze, P.O., S. Hochgreb, D. Kayes, H.A. Michelsen, and C.R. Shaddix (2001). Time-Resolved Laser-Induced Incandescence and Laser Elastic-Scattering Measurements in a Propane Diffusion Flame. Applied Optics, 40, 2443-2451 Zhang, J., and C.M. Megaridis (1996). Soot Suppresion by Ferrocene in Laminar Ethylene/Air Non-premixed Flames. Combustion and Flame, 105, 528-540

PAGE 91

BIOGRAPHICAL SKETCH Gregory David Yoder was born in Poughkeepsie, NY, on June 16, 1979, the second of two children. In 1982, he moved to Tampa, FL, where he resided until the completion of high school. In August 1997, he entered the University of Florida and received a Bachelor of Science in mechanical engineering in December 2001 with honors. In January 2002, he began pursuit of a Master of Science in mechanical engineering, which incorporated the efforts of this masters thesis. 79


Permanent Link: http://ufdc.ufl.edu/UFE0002661/00001

Material Information

Title: Investigation of Laser-Induced Incandescence and Soot Vaporization Using Time Resolved Light Scattering in a Propane Diffusion Flame
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0002661:00001

Permanent Link: http://ufdc.ufl.edu/UFE0002661/00001

Material Information

Title: Investigation of Laser-Induced Incandescence and Soot Vaporization Using Time Resolved Light Scattering in a Propane Diffusion Flame
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0002661:00001


This item has the following downloads:


Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
        Page v
        Page vi
    List of Tables
        Page vii
    List of Figures
        Page viii
        Page ix
        Page x
    Abstract
        Page xi
        Page xii
    Introduction and background
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
    Experimental methods
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
    Results
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
    Analysis, discussion, and conclusions
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
    Appendix A: Error analysis
        Page 70
        Page 71
        Page 72
        Page 73
    Appendix B: Zeroth-order lognormal distribution
        Page 74
        Page 75
    References
        Page 76
        Page 77
        Page 78
    Biographical sketch
        Page 79
Full Text












INVESTIGATION OF LASER-INDUCED INCANDESCENCE AND SOOT
VAPORIZATION USING TIME RESOLVED LIGHT SCATTERING IN A PROPANE
DIFFUSION FLAME














By

GREGORY DAVID YODER


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Gregory David Yoder

































This document is dedicated to the memory of Carrie Lynn Yoder. She most
unfortunately passed away in March 2003. She held a Bachelor's degree from the
University of Florida, a Master's degree from the University of Central Florida, and was
pursuing a Ph.D. at Louisiana State University at the time of her death. She had a love
for research and academia. More importantly, she had a love for life and her family. Her
achievements were always a source of pride and encouragement.















ACKNOWLEDGMENTS

I would like to thank Dr. David Hahn for the guidance and leadership he has

provided over the last two years. He allowed me to work for him on an undergraduate

summer research project, which got me started in the area of laser-based diagnostics and

doing graduate type research in general. He most certainly was one of the largest reasons

I chose to pursue an advanced degree in mechanical engineering and to do so at the

University of Florida. He outstanding teaching abilities were a major reason for my

choice to stay at UF and pursue a master's degree under him.

I would also like to thank Dr. Bill Lear and Dr. Bruce Carroll for serving on my

committee. Their time and expertise have neither gone unnoticed nor unappreciated.

I also want to thank all of my lab mates over the past two years: Dr. Jorge

Carranza, Brian Fisher, Kibum Kim, Katie Masiello, Vince Hohreiter, and Allen Ball.

Sharing lab and office space was always a pleasure.

Lastly, and certainly not least, I would like to thank my loving parents for

supporting me and always encouraging me to strive for any goal I wanted to achieve.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ................................................... vii

LIST OF FIGURES ........................................... ............................ viii

ABSTRACT .................................................... ................. xi

CHAPTER

1 INTRODUCTION AND BACKGROUND .................................................

1 .1 In tro d u ctio n ............................................................................................................ 1
1.2 M otiv action ...................................................... ............................................ . 3
1.3 B background for D iagnostics ..................................... ..................... ...............4...
1.3.1 Laser-Induced Incandescence....................................................4...
1.3.2 L aser E plastic Scattering ....................... ...............................................9...
1.3.3 Transm mission Electron M icroscopy ...................................... ............... 13
1 .4 O bje ctiv e s ............................................................................................................ 14

2 EXPERIMENTAL METHODS .........................................................16

2 .1 B u rn er D esig n ...................................................................................................... 16
2 .2 L aser S y stem ........................................................................................................ 2 1
2.3 D ata A acquisition System ....................................... ...................... ................ 25
2.4 E xperim ental M ethods......................................... ........................ ................ 28
2.4.1 Signal L inearity .............. ............. .............................................. 28
2.4.2 L ight Scattering w ith L II ...................................................... ................ 28
2.4.3 V ariable Spatial R solution .................................................. ................ 30
2.5 TEM Methods ....................... ........... ...............................32

3 R E S U L T S ................................................................................................................. .. 3 5

3 .1 T em poral A lignm ent........................................... ......................... ................ 35
3.2 Signal Linearity ........................................ ........ ............ ......................... 39
3.3 Simultaneous Light Scattering and LII Measurements .................................40
3.4 V ariable Spatial R solution LII...................................................... ................ 46
3.5 Transm mission Electron M icroscopy ................................................ ................ 52



v









4 ANALYSIS, DISCUSSION, AND CONCLUSIONS ..........................................54

4.1 Light Scattering ................................. ........ ...................... 54
4 .2 L II A n aly sis ......................................................................................................... 6 1
4 .3 C o n clu sio n s.......................................................................................................... 6 8
4 .4 F u tu re W o rk ......................................................................................................... 6 9

APPENDIX

A E R R O R A N A L Y SIS ..................................................................................................70

B ZEROTH-ORDER LOGNORMAL DISTRIBUTION........................................ 74

L IST O F R E F E R E N C E S ...................................................................................................76

BIO GR A PH ICAL SK ETCH .................................................................... ................ 79















LIST OF TABLES


Table page

2-1. Gas flow rates and diameters for diffusion burner ...............................................17

2-2. Manufacturer specifications for compressed gases used in diffusion burner.............21

2-3. Description of experimental setup components.................................... ................ 24

2-4. Summary of laser settings for light scattering measurements...............................29

2-5. Comparison of 1064 nm laser beam diameters and fluences for LII experiments ....31

3-1. Summary of peak intensities for variable spatial resolution experiments.............. 52

A-1. Percent error analysis for LII laser set to 0.60 J/cm2...........................................71

A-2. Percent error analysis for LII laser set to 0.46 J/cm2...........................................72

A-3. Percent error analysis for LII laser set to 0.32 J/cm2...........................................73















LIST OF FIGURES


Figure page

1-1. Soot form ation schem atic ........................................... ......................... ...............2...

1-2. T E M scheme atic............... .. .................. .................. .......................... ............... 13

2-1. Top view schematic of diffusion burner..................................................... 17

2-2. Side view schematic of diffusion burner............................................................. 18

2-3. Propane diffusion flame with flame holder ..........................................................18

2-4. Schem atic of experim ental layout ........................................................ ................ 23

2-5. Transmission calibration plot for 532 nm bandpass filter ....................................25

2-6. Laser Q-switch and flashlamp trigger timing .......................................................27

2-7. Three individual soot particles photographed by TEM........................................33

2-8. Agglomerated chain of soot particles photographed by TEM............................... 34

3-1. Simultaneous laser pulses for LII laser at 0.60 J/cm2 ................................................36

3-2. Simultaneous laser pulses for LII laser at 0.46 J/cm2...........................................36

3-3. Simultaneous laser pulses for LII laser at 0.32 J/cm2...........................................37

3-4. Laser pulses delayed by 24 ns for LII laser at 0.60 J/cm2 .............. ..................... 38

3-5. Laser pulses delayed by 24 ns for LII laser at 0.46 J/cm2 .............. ..................... 38

3-6. Laser pulses delayed by 24 ns for LII laser at 0.32 J/cm2 .............. ..................... 39

3-7. Signal linearity verifi cation plot........................................................... ................ 40

3-8. Sample average baseline measurement from scattering experiments ....................42

3-9. Integrated light scattering measurement...............................................................43

3-10. Extended data set for light scattering measurements..........................................44









3-11. Light scattering measurements for LII laser set to 0.60 J/cm2 .............................45

3-12. Light scattering measurements for LII laser set to 0.46 J/cm2 .............................45

3-13. Light scattering measurements for LII laser set to 0.32 J/cm2 .............................46

3-14. LII plot for 0.61 J/cm2 fluence and 2.0 mm PMT aperture .................................47

3-15. LII plot for 0.61 J/cm2 fluence and 3.6 mm PMT aperture .................................48

3-16. LII plot for 0.47 J/cm2 fluence and 2.0 mm PMT aperture .................................48

3-17. LII plot for 0.47 J/cm2 fluence and 3.6 mm PMT aperture .................................49

3-18. LII plot for 0.60 J/cm2 fluence and 2.0 mm PMT aperture .................................50

3-19. LII plot for 0.60 J/cm2 fluence and 3.6 mm PMT aperture .................................50

3-20. LII plot for 0.46 J/cm2 fluence and 2.0 mm PMT aperture .................................51

3-21. LII plot for 0.46 J/cm2 fluence and 3.6 mm PMT aperture .................................51

3-22. Zeroth-order logarithmic distribution of modal diameter.....................................53

4-1. M ie theory calibration plot ........................................ ........................ ................ 56

4-2. Modal diameter as particles decay for 0.60 J/cm2 LII laser fluence.......................57

4-3. Modal diameter as particles decay for 0.46 J/cm2 LII laser fluence.....................58

4-4. Modal diameter as particles decay for 0.32 J/cm2 LII laser fluence ..........................58

4-5. Average particle volume as particles decay for 0.60 J/cm2 fluence .....................59

4-6. Average particle volume as particles decay for 0.46 J/cm2 fluence .....................60

4-7. Average particle volume as particles decay for 0.32 J/cm2 fluence .....................60

4-8. Sample LII response with rise and decay of particle temperature........................63

4-9. Particle tem perature rise and decay m odel........................................... ................ 64

4-10. Particle volume experimental data and curve fit prediction................................66

4-11. L II signal w ith m odel prediction ........................................................ ................ 67

4-12. LII signal w ith 2 predictive m odels.................................................... ................ 68









B-1. ZOLD distribution for three values of co ............................ ................ 75















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

INVESTIGATION OF LASER-INDUCED INCANDESCENCE AND SOOT
VAPORIZATION USING TIME RESOLVED LIGHT SCATTERING IN A PROPANE
DIFFUSION FLAME

By

Gregory David Yoder

December 2003

Chair: David W. Hahn
Major Department: Mechanical and Aerospace Engineering

Particulate matter (PM) and aerosol particles have become a top concern in both the

realms of human health and the sustainability of the environment. A significant source of

particulate matter is due to combustion processes, specifically the burning of coal and

hydrocarbon fuels. As a result of these health concerns, there is a need to monitor and

control the levels of these harmful particulate emissions. This thesis project reports on

the use of a potential monitoring technology, and hence is a first step in finding a solution

to PM-related problems.

Novel laser-based diagnostics have been developed in recent years to monitor and

characterize particulate matter, including soot. These laser-based techniques were

applied to a laboratory diffusion burner, with the goal to ultimately apply them to a

laboratory scale gas turbine engine where metallic based fuel additives will be used to

control soot formation. Laser elastic light scattering (LES) and laser-induced

incandescence (LII) were the two techniques used to characterize the particulate soot.









Laser elastic scattering is a technique that allows for the determination of particle size

and number density by measuring the amount of laser light scattered by the soot particles.

Laser-induced incandescence is a technique that uses a pulsed laser beam to heat soot

particles up to levels far above the background, causing them to emit radiation as

essentially blackbodies, which can then be related to the total soot volume using suitable

calibration schemes. However, the temperature reached by the laser-heated particles may

cause the particles to begin to vaporize, thereby changing the parameter of interest,

namely the total soot volume. The primary goal of this thesis is to investigate the particle

vaporization due to LII using time-resolved laser light scattering. Based on the

experimental measurements in a sooting propane diffusion flame, significant particle

vaporization was found to occur on the time scale of the LII laser pulse.

A model was developed to describe the particle temperature and volume as the

particles are heated up and vaporized by the LII laser and as the particles subsequently

cooled. This heat transfer model uses a fundamental energy balance along with the

Planck distribution to model the LII signal as a function of time, and was then used to

estimate the influence of particle vaporization on the time-resolved LII signal.














CHAPTER 1
INTRODUCTION AND BACKGROUND

1.1 Introduction

Particulate matter and aerosol particles have become a top concern in both the

realms of human health and the sustainability of the environment. According to a 1996

study by the United States Environmental Protection Agency (U.S. EPA), particulate

matter is a direct cause of increased morbidity and mortality in humans (U.S. EPA 1996).

A significant source of particulate matter is due to combustion processes, specifically the

burning of coal and hydrocarbon fuels. In 1996, approximately 85% of the energy used

in the United States came from combustion sources (U.S. DOE 1996). Combustion

products include soot, fine metallic species, and polyaromatic hydrocarbons (PAH's) that

can be inhaled and are harmful to humans. Furthermore, harmful products of combustion

such as oxides of nitrogen and sulfur (NOx and SOx) can have a negative impact on the

environment. NOx and SOx are key contributors to acid rain and also form particulates as

nitrates and sulfates (Turns 2000).

This thesis will focus on soot and on the use of laser-based diagnostics to monitor

soot. Soot in simply solid carbon formed when hydrocarbon fuels are burned. Soot is

formed through a four step process: (1) formation of precursor species, (2) particle

inception, (3) surface growth and particle agglomeration, and (4) particle oxidation.

During the formation of precursor species, chemical kinetics plays an important role.

Chemical kinetics is the study of elementary reactions and their rates. Particle inception

involves the formation of small particles of a critical size (3,000 10,000 atomic mass






units) from growth by both chemical means and coagulation. In this step, large
molecules are transformed into particles. When the small primary soot particles continue
to be exposed to species of the pyrolizing flame, they experience surface growth and
agglomeration. If all of the soot particles are oxidized, the flame is termed nonsooting,
while; conversely, incomplete oxidation yields a sooting flame (Turns 2000). A
conceptual schematic of this soot formation process is shown in Figure 1-1 below.


w




ttt

n~o t t t
H20
H20


tit
C3H8


4.


H*
0
titt


06

002-O*

CO t
CO tit


Figure 1-1. Soot formation schematic









1.2 Motivation

Emissions from aircraft jet engines currently contain particulate matter and harmful

aerosols. The levels of these emissions must be monitored and controlled to ensure the

public's and the environment's safety. As an example, crew members aboard aircraft

carriers are subjected to jet engine exhaust on a constant basis, which is very dangerous

as the particulate matter in the exhaust can be very harmful. The United States Navy

currently is funding research efforts to decrease soot and particulate matter emissions

from their engines. This thesis project is the first step in one of these efforts. The overall

goal of such efforts is to decrease soot emissions from jet engines, including the use of

fuel additives.

Fuel additives have been used as soot suppressants for over 40 years, but the

mechanism by which they actually suppress the soot is still a matter of debate. There are

several theories as to how fuel additives, both metallic and non-metallic based, suppress

soot. One theory focuses on the beginning stages of soot particle inception, and links the

fuel-additive effects with changes on nucleation mechanisms of soot formation. Another

theory concentrates on the role of additives as an enhancement for soot oxidation and

burnout mechanisms, as a result of increased production of hydroxyl radicals, which

rapidly remove soot or gaseous soot precursors. A third theory attributes the additive

effects to the acceleration of the soot oxidation rate, possibly by occlusion in the soot

phase. Some examples of fuel additives are ferrocene [(CsHs)2Fe] and iron

pentacarbonyl [Fe(CO)5] (Zhang and Megaridis 1996).

Novel laser-based techniques are to be developed to monitor and characterize the

soot and particulate matter. These techniques will be applied first to a laboratory

diffusion flame, and finally to a laboratory scale gas turbine engine. The first phase of









the project is to model and characterize the laser-based diagnostics to be used to quantify

the soot and particulate matter. The tools to be used for characterization are laser-

induced incandescence (LII), laser elastic scattering (LES), and transmission electron

microscopy (TEM).

1.3 Background for Diagnostics

1.3.1 Laser-Induced Incandescence

Laser-induced incandescence (LII) is a well-researched technique for analyzing and

characterizing sooting flames and combustion processes. Several previous research

efforts have shown that LII signals are roughly proportional to soot volume fraction,

which is the ratio of the volume of soot particles to the volume of gas (Schraml et a12000,

Axelsson et al 2001, Wainner and Seitzman 1999). Schraml used LII to make two-

dimensionally resolved measurements of soot primary particle sizes from the ratio of LII

signals from two delay times. Schraml's experiments were performed in a laminar ethene

diffusion flame. Axelsson also used a ratio of LII signals from two time gate positions to

measure soot particle size. Axelsson used a premixed flat ethane/air flame for LII

experiments. Wainner reports detection of soot at better than ~1 part per trillion by use

of LII techniques. Wainner chose to use a controllable soot-generating device that

simulated a hot, low-soot-concentration environment similar to that of a jet engine.

LII occurs when a high intensity laser beam encounters particulate matter like soot.

A soot particle can absorb energy from the beam, which causes the particle's temperature

to increase. Soot has a very high absorptivity for visible and infrared light. At the same

time, the soot can loose energy by heat transfer mechanisms to the surroundings. With

sufficiently high laser energies, numerical models of the heat transfer processes indicate

that the soot particles reach temperatures of 4000-4500 K (Snelling et al 1997). If the









energy absorption rate is sufficiently high, the temperature will rise to levels where

significant incandescence (essentially blackbody radiation) and vaporization can occur.

Thus, LII can be used to obtain information about some average property of all the soot

particles within a measurement volume (AIAA.org). A high-powered laser, typically a

pulsed Nd:YAG (Neodimum: Yttrium-Aluminum-Garnett) laser operating at its

fundamental wavelength of 1064 nm, is aimed through a control volume to interact with

particles of unknown volume fraction. The laser beam heats up the interacting soot

particles and a photomultiplier tube (PMT) is used to collect the light emitted from the

particles. However, as the particles are being heated, they begin to transfer energy via

conduction and radiation to their surroundings. Once the particles reach their

vaporization temperature, they begin to lose mass as well. The three main energy loss

mechanisms are blackbody radiation, conduction, and vaporization. Blackbody radiation

accounts for the smallest losses, approximately two orders of magnitude less than

vaporization, the greatest loss mechanism. Conduction losses are approximately one

order of magnitude less than vaporization (Stephens et al 2003).

By using basic thermodynamic principles, equations have been developed to

describe the heat transfer processes that occur during LII. A fundamental energy balance

describing the heat transfer processes in the interaction of a laser beam and a soot particle

is given by Equation 1.1 below.

Kb (a)a2q Ka (T To)(4a 2) AH, dM 4 -T2 4 )4,dT 0
a (1 + GKJ) W dt 3 dt
Equation 1.1

The terms are, respectively, the rate of laser energy absorbed per second, the rate of

heat transfer to the medium (taken to be air at temperature To), the energy expended in









vaporization of the carbon, the rate of energy loss by blackbody radiation, and the rate of

internal energy rise (Melton 1984).

From these basic equations, relationships have been developed to determine certain

desired parameters of the system. For instance, the particle temperature after the laser

pulse can be found using Equation 1.2 below.

S(- 1 = (- 1) exp(-t /'c) Equation 1.2
Rg Rg

In Equation 1.2 Tp is the particle temperature for any given time after the laser

pulse, Tg is the gas temperature, more specifically the flame temperature, Tp is the

maximum particle temperature which is reached immediately after the laser pulse, t is

time, and -c is a characteristic cooling time defined in Equation 1.3 below.

TC = pcp Equation 1.3
3CtPg

In Equation 1.3, dp is the particle diameter, Cp is the particle specific heat, pp is the

particle density, ct is the average thermal velocity of gas molecules defined by Equation

1.4 below, and pg is the density of the ambient gas in the flame.

L8kTg
c = 8g Equation 1.4
xTmg

In Equation 1.4, k is the Boltzmann constant and mg is the mass of the gas

molecules. The LII signal can be related to the particle temperature and size by use of the

Planck distribution and is shown in Equation 1.5 found below (Roth and Filippov 1996).

SLIT = n N[Po, o(t)FPlanck (, Tp (t))- P- (t = O)FPlck (, Tg)] Equation 1.5









In Equation 1.5, SLII is the LII signal, qr is a constant parameter which represents

the effect of the collection optics and the detector efficiency, N is the particle number

density, Pvol is the particle volume, and FPlanck is a function of the Planck distribution

which describes the radiation emitted by the particles. Fplanck is defined below in

Equation 1.6.

C
Fplanck -1 Equation 1.6
k' exp( h )-1


In Equation 1.6, h is the Planck constant, Co is the speed of light, X is the detection

wavelength, and k is the Boltzmann constant, and C1 is the first radiation constant defined

in Equation 1.7 below (Kock et al 2002). These equations are discussed in detail in

Chapter 4.

C1 = 27hc2 Equation 1.7

LII can be detected in more than one way. Prompt detection denotes setting the

detector to begin looking at the signal at the moment of laser incidence, or even shortly

before the laser pulse arrives. Prompt detection is purported to reduce any effects

resulting from size-dependent vaporization of the soot particles. Delayed detection

requires waiting until after the laser pulse to begin detecting the LII signal, typically -30

-40 ns after the laser pulse. This will eliminate unwanted elastic-scattered light or laser-

induced fluorescence signals from polycyclic aromatic hydrocarbons or C2 fragments

(Witze et al 2001). Another variable in LII detection is the wavelength at which it is

detected. Line filters and/or bandpass filters can be added to the experimental setup to

only allow the detector to see a particular spectral band or range of wavelengths. This

will eliminate broadband fluorescence from unwanted spectral bands. Witze used a long









wave pass filter with a 570 nm cutoff (2001). Wainner used glass filters to monitor the

LII signal between 570 and 850 nm (Wainner and Seitzman 1999). Van der Wal used an

interference bandpass filter to monitor the LII signal between 400 and 450 nm (Van der

Wal and Weiland 1994). Axelsson monitored specific wavelengths of 300 nm and 600

nm (Axelsson et al 2001).

There are several conditions that need to be satisfied for the detected LII signal to

be proportional to soot volume fraction: (1) the probed soot consists of single or loosely

aggregated primary particles that are small compared to the wavelengths of the laser

excitation (such that Rayleigh-limit for light absorption and emission is valid); (2) the

peak particle temperatures reached during the laser pulse are relatively insensitive to the

particle diameter; (3) the soot particle mass vaporization is either negligible or largely

independent of particle diameter; and (4) the detected LII signal is dominated by thermal

emission occurring during laser excitation or shortly thereafter, such that size dependent

conductance cooling does not influence the signal (Witze et al 2001). The criteria for

particles to be within the Rayleigh limit for detection is they must have a size parameter

ca and the quantity Imcu much less than one. The size parameter ca is defined in Equation

1.7 below, while m is the complex refractive index of the soot particles.

a =- Equation 1.7


In Equation 1.7, a is the particle radius and X is the laser wavelength.

One characteristic of LII that makes it particularly advantageous for soot volume

fraction detection is the rapid rise in LII signal with increasing laser fluence until a nearly

constant plateau signal strength is reached. This is useful in strongly absorbing

environments because it eliminates the need for corrections to the measured LII signal









that are due to laser beam absorption, as long as the initial laser fluence is suitably far

into the plateau regime (Witze et al 2001).

1.3.2 Laser Elastic Scattering

Laser elastic scattering is a well-researched technique that can be used to determine

particle size and concentration using laser-based diagnostics. This technique has been

used many times in the past for soot characterization (Hahn and Charalampopoulos 1992,

Witze el al 2001, Dasch 1984). Hahn used laser light scattering in an earlier metal

additive study when looking at a premixed propane/oxygen flame and the influence of

iron additives (1992). Witze also used LES when performing LII experiments in flames.

Witze's study was focused on time-resolved LII measurements using a rectangular profile

beam in a propane/air diffusion flame (2001). Dasch used a continuous wavelength

Argon ion probe beam for light scattering in an LII study of a methane/oxygen flame

(1984). Other earlier work that utilized laser scattering in flames was done by Menna

(Menna and D'Alessio 1982), Erickson (Erickson et al 1964),and Prado (Prado et al

1981).

The scattering of electromagnetic (EM) waves is due to heterogeneities.

Heterogeneities are disruptions in the field of the electromagnetic wave. At some level,

all matter presents heterogeneities to the EM waves. These heterogeneities or obstacles

can be atoms, molecules, or particles. The EM wave will induce an oscillation within the

electrons of the "obstacle," called an oscillating dipole moment. The charges on the

molecules will oscillate at the same forcing frequency as that at which the EM wave is

oscillating. This oscillating molecule becomes a source of radiation, which occurs at the

same frequency of the induced oscillation. When scattered light is emitted at the same









frequency as the incident light, it is called elastic scattering. There are two regimes in

laser elastic scattering: Rayleigh and Mie scattering.

Lord Rayleigh (1871) published a solution for light scattering by small,

dielectric, spherical particles. Small particles are defines by dp<
particle diameter and k is the wavelength of the laser light. Dielectric particles are

simply non-absorbing particles. The Rayleigh solution states that the dipole moment

radiates in all directions. Furthermore, the electric field is proportional to the dipole

moment, which is proportional to the volume of the scatterer. Lastly, Rayleigh stated that

light scattering has strong, inverse wavelength dependence. Rayleigh developed several

equations to describe the scattering of light. Rayleigh scattering is defined by several

differential scattering cross sections (area/steradian), which represent the amount of

energy scattered into a given solid angle about a given direction. The differential

scattering coefficients for a spherical particle in the Rayleigh regime are given by the

expressions

k2 --2 1
C'w=--2 6 --2 Equation 1.8
4 m i2+2


CHH =CVw COS2 Equation 1.9

The VV and HH subscripts on the above coefficients denote vertically or

horizontally polarized light of the incident and scattered light, respectively. If the light is

vertically polarized incident to the scattering particle, then the light will still be vertically

polarized when it is scattered. The Rayleigh total scattering and absorption cross sections

may be written, respectively, as











3ita (X



Cabs ( W
7r +


Equation 1.10



Equation 1.10


The extinction cross section, Cext, represents the total energy removed by the

scattering particle and is comprised of the sum of the scattering cross section and the

absorption cross section. Since ca must be much less than one to be considered in the

Rayleigh regime, the scattering cross section can be neglected since it is of the sixth order

of ca, while the absorption cross section is third order in ca (Kerker 1969).

In 1908 Mie developed a generalized, exact solution for the scattering of light by

a single, homogeneous sphere of arbitrary size, while Rayleigh's solution is only valid for

small ca (Kerker 1969). The differential scattering cross section according to Mie theory

is given by


v2n-1 {at, cs b(CosjI


Equation 1.11


where the parameters 7t n and c n are the angular dependent functions,

PO1) (cos 0)
rT n (cos 0) = csi Equation 1.12
sin 0


TC (cos 0) = (Cos0) Equation 1.13
dO


of the associated Legendre polynomial P1) (cos 0) of argument cos 0. 0 is defined as

the angle of observation and is measured from the forward direction to the scattering

direction. The parameters an and bn are defined as


2


>









a



MW (a)W,:(ma) -n (Ma)An' (a)

The functions I' and n are defined as


n (z) = (2- Jn 12(z)
2


2


Equation 1.14


Equation 1.15





Equation 1.16


Equation 1.17


where X n (z) is given by


Xn (Z) = -( 12 Y 12 (z) .
2


Equation 1.18


Here J +l/ 2(z) and Y,+/2 (z) are the half integral order Bessel functions of the first and

second kind, or Bessel and Neumann functions, respectively, and H? 2 (z) is the half

integral order Hankel function of the second kind.

Simply stated, LES involves sending a laser beam through a control volume

containing particles of unknown size and concentration. The laser light will be scattered

off the particles in all directions, as prescribed by Rayleigh or Mie theory. A detector or

detectors are setup to collect the scattered light, with the signal received being

proportional to both the size of the particles and the particle number density. One

downfall of LES is the need for very precise alignment to avoid stray light. The optical

layout must be arranged very carefully to ensure a quality signal by eliminating stray

light and background noise, as described in Chapter 2.









1.3.3 Transmission Electron Microscopy

Transmission electron microscopy (TEM) is a very well researched method for

viewing microscopic particles. A schematic of the inner workings of a transmission

electron microscope is shown in Figure 1-2. The "Virtual Source" at the top represents

the electron gun, producing a stream of monochromatic electrons. This stream is focused

to a small, thin, coherent beam by the use of condenser lenses 1 and 2. The first lens

(usually controlled by the "spot size knob") largely determines the "spot size"; the general

size range of the final spot that strikes the sample.


Figure 1-2. TEM schematic

The second lens (usually controlled by the "intensity or brightness knob" actually

changes the size of the spot on the sample; changing it from a wide dispersed spot to a


. Projector Lens

S Main Screen phosphorr)









pinpoint beam. The beam is restricted by the condenser aperture (usually user

selectable), knocking out high angle electrons (those far from the optic axis, the dotted

line down the center). The beam strikes the specimen and parts of it are transmitted.

This transmitted portion is focused by the objective lens into an image. Optional

Objective and Selected Area metal apertures can restrict the beam; the Objective aperture

enhancing contrast by blocking out high-angle diffracted electrons, the Selected Area

aperture enabling the user to examine the periodic diffraction of electrons by ordered

arrangements of atoms in the sample. The image is passed down the column through the

intermediate and projector lenses, being enlarged all the way. The image strikes the

phosphor image screen and light is generated, allowing the user to see the image. The

darker areas of the image represent those areas of the sample that fewer electrons were

transmitted through (they are thicker or denser). The lighter areas of the image represent

those areas of the sample that more electrons were transmitted through (they are thinner

or less dense). This technique has been widely used for soot analysis. Zhang used

thermophoretic sampling to acquire soot samples and used TEM to analyze the soot

particles (Zhang and Megaridis 1996). Fotou also used TEM for analysis of soot in a fuel

additive study when using a diffusion flame (Fotou et al 1995). Also, Dobbins used

thermophoretic sampling to study the morphology of flame-generated soot (Dobbins and

Megaridis 1987).

1.4 Objectives

The primary objective of this project is to characterize and model the LII signal

from a propane/oxygen diffusion flame. This model will then be used in future studies

geared toward understanding the reduction of soot emissions using metal additives. A

novel technique was developed using a two-laser setup that enabled simultaneous LII






15


excitation and LES measurements of the heated soot particles. This experimental design

is helpful in reproduction and implementation for further work. Below is a list of detailed

objectives.

1. Understand the role of laser-particle interactions with respect to quantitative LII
measurements of soot volume.

2. Quantify soot emissions with light scattering for correlation with LII data to
optimize the accuracy of the LII method.

3. Explore the temporal profile of soot particle vaporization with a goal of enhancing
in situ Raman spectroscopy of metallic species by "peeling" away soot.














CHAPTER 2
EXPERIMENTAL METHODS

2.1 Burner Design

The burner used in this experiment was designed to provide a stable, optically

accessible flame that is consistent with the combustor of a gas turbine engine. A gas

turbine combustor typically employs a fuel lean diffusion burner. The laboratory

diffusion burner used in this experiment was fabricated from stainless steel tubing and

Swagelok fittings. A top view schematic of the burner is shown in Figure 2-1, a side

view schematic of the burner is shown in Figure 2-2, and a photograph of the burner

setup with the propane diffusion flame running is shown in Figure 2-3. The burner has

three separate, controllable gas flows. The innermost flow tube contains propane gas,

controlled by a precision-drilled ruby orifice and a pressure regulator. This Bird

Precision 0.0040" orifice maintains the propane flow rate at 0.31 LPM while the line

pressure is kept at 50 psi. The annular flow surrounding the propane is oxygen. The

oxygen flow rate is maintained by an Alicat Scientific precision gas flow controller. The

oxygen volumetric flow rate was held constant at 1.82 LPM for all experiments. The

outermost gas flow is nitrogen. The nitrogen flow rate was maintained by an Alicat

Scientific precision gas flow controller. The nitrogen volumetric flow rate was held

constant at 1.05 LPM for all experiments. A summary of the gas flow rates is given in

Table 2-1. The nitrogen shroud flow holds two purposes. First, it shields the inner gases

from the ambient air to allow precise control over the fuel stoichiometry. Secondly, the

shroud flow aids in flame stability. It is very important to maintain a stable flame so the









lasers will always strike the flame and strike it in the same location every laser pulse. In

addition to the shroud flow, a 0.1-inch stainless steel mesh flame holder was used for

flame stability.


Figure 2-1. Top view schematic of diffusion burner



Table 2-1. Gas flow rates and diameters for diffusion burner


Flow Tube Gas Diameter Flow Rate


Dl Propane 1/16" 0.314 LPM

D2 Oxygen 9/32" 1.815 LPM

D3 Nitrogen 1" shroud 1.052 LPM






18

















Oxygen





Figure 2-2. Side view schematic of diffusion burner


Figure 2-3. Propane diffusion flame with flame holder


<--- Nitrogen
(Shroud)

- Propane









The flame holder was positioned 1.6" above the burner. The flame holder "grabs"

the flame and maintains it in a steady position with respect to the horizontal and the

vertical planes. There are currently no fuel additive capabilities for this burner, but

additional components can be added. A plexiglass shield was placed around the entire

flame setup to protect the flame from drafts. Maintaining flame stability was a priority to

ensure a repeatable laser/soot interaction. The shield had two holes cut in it to allow the

laser beams to enter and exit, and a third hole at 900 to allow the scattering and LII signal

to reach the PMT. The region around the shield that was at the same height as the laser

beams was coated with a black fabric to eliminate reflections and to cut down on stray

light.

Gas turbine engines run their combustors on the fuel lean side of stoichiometric.

The stoichiometric quantity of oxidizer is just that amount needed to completely burn a

quantity of fuel (Turns 2000). Fuel lean refers to having excess oxidizer or a shortage of

fuel. Gas turbines run fuel lean due to the large amount of available air flowing through

the engine and the high cost of fuel and the cost of carrying large amounts of fuel. Also,

burning slightly fuel lean decreases the flame temperature, which decreases NOx and SOx

formation. For stability and to replicate the combustion in a gas turbine engine, the

diffusion flame in these experiments was run on the fuel lean side of stoichiometric.

The governing combustion reaction for stoichiometric propane combustion is given

by Equation 2.1 below.

aC3H8 + bO2 cCO2 + dH20 Equation 2.1

This basic reaction neglects secondary products of combustion, free radicals, and

dissociation effects. Nitrogen is not included in the reaction because the nitrogen shroud









is assumed to not mix significantly with the propane and oxygen flows and does not

directly affect the flame stoichiometry. The coefficients a, b, c, and d in Equation 2.1 are

found based on the initial reactant parameters. Two cases will be considered. The first

case is the stoichiometric case. For stoichiometric combustion, one mole of fuel is

reacted with b moles of oxidizer to create c and d moles of products. A simple atom

balance will result in a stoichiometric reaction shown in Equation 2.2 below, which

yields an air to fuel ratio of 5 on a molar basis for propane.

C3H8 +502 3CO2 + 4H20 Equation 2.2

The second case is the experimental case. In the diffusion flame, the volumetric

flow rates of fuel and oxidizer were 0.314 LPM and 1.815 LPM, respectively. For an

ideal gas, the volumetric ratio is equal to the molar ratio. Therefore, the experimental

combustion reaction can be written as shown in Equation 2.3 below.

0.314C3H8 +1.81502 0.942CO2 + 1.256H20 +.24502 Equation 2.3

The fuel equivalence ratio 4 is the determining factor as to whether a flame is fuel

rich or fuel lean. The definition for fuel equivalence ratio is given in Equation 2.4 below.

A / stozchzometnc Equation 2.4
A/Fexp enmental

A / F experimental is defined as the actual air to fuel ratio for the reactants. The air to

fuel ratio for the stoichiometric case and the experimental case can be found by using

Equations 2.2 and 2.3. These ratios are calculated in Equations 2.5 and 2.6 below.

A/ Fstozch=ometnc = 5/1= 5 Equation 2.5


A/ Fexp enmet =1.815/0.314 =5.78 Equation 2.6









The fuel equivalence ratio 4) can now be found using Equation 2.4. 4) is calculated

in Equation 2.7 below.

) = 5 / 5.78 = 0.87 Equation 2.7

With a fuel equivalence ratio less than one, the flame is running fuel lean. This is

the desired stoichiometry to match that of a gas turbine combustor. Also of significance

is that soot particles undergo significant oxidation (i.e. burnout) in actual combustors,

hence the current fuel lean flame better emulates this condition. Table 2-2 provides the

manufacturer specifications for the compressed gases used in the experiments.

Table 2-2. Manufacturer specifications for compressed gases used in diffusion burner.

Gas Supplier Description
Nitrogen Praxair Industrial Grade, 99.7% N2
H20 < 32 ppm
02 Balance (nominal)
Oxygen Praxair Medical Grade, 99.5% 02
H20 < 20 ppm
Propane Praxair 99.5 wt% (liquid phase)
H20 < 300 ppmw
Ethane < 600 ppmw
Propylene < 400 ppmw
n-Butane < 20 ppmw
Isobutane < 3000 ppmw
Sulfer < 1 ppmw

2.2 Laser System

The diagnostics for these experiments were performed using a two-laser setup. A

Q-switched 1064 nm Nd:YAG laser was used for LII and soot vaporization. A frequency

doubled Q-switched 532 nm Nd:YAG laser was used for the laser light scattering portion

of the experiments. The 1064 nm laser light was focused using a lens with a 500 mm

focal length. The 532 nm laser light was focused using a lens with a 250 mm focal

length. A schematic showing the layout of the laser and optics system is shown in Figure









2-4. The two lasers are both directed through the flame such that the paths cross when

both beams are focused down to their minimum diameters and are on the central axis of

the burner tube. The 1064 nm beam is considerably larger in diameter when the beams

cross paths, approximately four times the diameter of the 532 nm beam. The 532 nm

beam is centered in the middle of the 1064 nm beam. This allows for a large volume of

soot particles to be heated up, creating the incandescent signal while vaporizing the

particles, and a much smaller volume of particles within the larger incandescent volume

to be used for light scattering. This technique ensures the scattering signal is only from

particles that have been exposed to the high-powered 1064 nm beam.

The laser light is collected using a photomultiplier tube (PMT) to create a signal for

analysis. The laser light that is collected is both scattered 532 nm light and the

incandescent light created by the heating of particles by the 1064 nm laser. The emitted

light is collected at a 90-degree angle from the 1064 nm beam. The light is directed

through two very small apertures to reduce background noise and to ensure the light

collected is only from the small scattering volume in the flame. After the first aperture,

the light is sent through a Newport bandpass filter centered at 532 nm. This filter

effectively removes all light that is not at the 532 nm wavelength. This filter has a 10 nm

full width half maximum (FWHM) and transmits approximately 55% of the light within

the FWHM. The transmission versus wavelength calibration plot for this filter is shown

in Figure 2-5. A summary of all equipment used is shown in Table 2-3.




















Aperture
Focusing Lens
Bandpass Line Filter (532 nm)
Aperture


Diffusion
Burner


Figure 2-4. Schematic of experimental layout


532 nm
Mirror










Table 2-3. Description of experimental setup components


Device Manufacturer Model Description

Lasers and Electronics

1064 nm Nd:YAG laser Big Sky Laser Ultra Q-switched, 5 Hz, Variable Power
FWHM = 13 ns, 50 mJ max

532 nm Nd:YAG laser Continuum Minilite ML-II Q-switched, 5 Hz, 2.4 mJ/pulse
FWHM = 5 ns, 20 mJ max

Photomultiplier Tube Hamamatsu 1P28 PMT

Ocsilloscope LeCroy LT 372 500 MHz, 4 GS/s digital oscilloscope

Delay Generator Stanford Research DG 535 Programmable Delay Generator
Instruments

Voltage Supply Stanford Research PS325 Digital High Volage Power Supply
Instruments

Optics

532 nm Lens Newport KBX079AR.14 BBAR coated, 430-700 nm
25.4 mm diameter
250 mm focal length

1064 nm Lens Newport KBX082AR.18 BBAR coated, 1000-1550 nm
25.4 mm diameter
500 mm focal length

1064 nm Mirror CVI Laser Y1-2037-45UNP 45 degree, 1064 nm dielectric mirror

532 Mirror Spindler & Hoyer 34-0467-000 45 degree, 532 nm dielectric mirror

Bandpass Filter Newport 10LF10-532 10nm FWHM, Transmission > 50%
25.4 mm diameter












95 0


85.0 :
Ao C.
75 C
70.0
65.0-
600-



45.0
40.0
35.0


250

150' ,
1)-4


510.0 5150 520.C 5250 530.0 535.0 540.0 545.0 550
W;,- ,r ic-.rh *nnm



Figure 2-5. Transmission calibration plot for 532 nm bandpass filter

2.3 Data Acquisition System

The signal from the laser light was collected using a photomultiplier tube (PMT).

A PMT converts photons to an electrical signal. A PMT consists of a photocathode and a

series of dynodes in an evacuated enclosure. When a photon of sufficient energy strikes

the photocathode, it ejects a photoelectron due to the photoelectric effect. The

photocathode material is usually a mixture of alkali metals, which make the PMT

sensitive to photons throughout the visible region of the electromagnetic spectrum. The

photcathode is at a high negative voltage, typically -500 to -1500 volts. The PMT was

charged at -750 volts for all experiments in the present work. The photoelectron is

accelerated towards a series of additional electrodes called dynodes. These electrodes are









each maintained at successively less negative potentials. Additional electrons are

generated at each dynode. This cascading effect creates 105 to 107 electrons for each

photoelectron that is ejected from the photocathode. The amplification depends on the

number of dynodes and the accelerating voltage. This amplified electrical signal is

collected at an anode at ground potential (Tissue 2000). A Stanford Research Systems

high voltage power supply was used to charge the PMT.

Precise control of the laser flashlamps and Q-switches was required to eliminate

jitter (i.e. variation between the LII laser and the LES laser). Lasers have an internal

timing system that signals the flashlamp to start and signals the Q-switch to open to

release the laser pulse. This timing is not always precise and the laser pulse can vary by

up to 10 nanoseconds from shot to shot. This variation of the laser pulse from shot to

shot is called jitter. For these experiments, in order to provide the desired temporal

resolution it was necessary to completely remove the jitter by externally triggering the

flashlamp and the Q-switch on both lasers. This was done using 2 separate delay

generators. The first delay generator was internally triggered at a repetition rate of 5 Hz.

This delay generator triggered the LII laser's flashlamp and Q-switch. The first delay

generator was also used to trigger a second delay generator. The second delay generator

was used to trigger the 532 nm laser's flashlamp and Q-switch. The Q-switch signal of

the second delay generator was also used to trigger the oscilloscope. A schematic

showing the timing details of the laser system is shown in Figure 2-6. The signal out of

the PMT was sent to a LeCroy digital oscilloscope. The oscilloscope was triggered by

the 532 nm laser Q-switch.












Delay Generator 1
Internal Trigger
5 Hz


to 1064


K-50 ;s-es


to 1064 nm Q-switch


Delay Generator 2
External Trigger
5 Hz


to 532 rn flashlamp


to 532 nm Q-switch

* used to control delaybetween 532 nm and 1064 nm laser pulses


Figure 2-6. Laser Q-switch and flashlamp trigger timing









2.4 Experimental Methods

2.4.1 Signal Linearity

Light scattering experiments can have signals that vary by more than an order of

magnitude in a single measurement. Signal linearity refers to a detector's ability to

accurately report signals over a wide dynamic range. If a detector can accurately

decipher signals at low levels, but gets saturated when the signal increases, the reported

data will be invalid. In these experiments, soot is being evaporated and the signals are

decaying to levels up to 15 times less than the original signal. It is very important that

the detector, the PMT in this case, be able to interpret the signals accurately. Signal

linearity tests were performed before every set of data was recorded. These tests were

done by first taking an average scattering measurement for 500 laser shots. A 50 %

transmission neutral density filter of optical density equal to 0.3 was then placed inline

with the PMT signal line of sight and another scattering measurement was recorded. If

the signal were truly linear, the ratio of the two measurements would be 12. The PMT

voltage was adjusted to achieve this ratio. It is noted that PMT's are generally

characterized by excellent linearity, often over many decades. However, with -10 ns

laser pulses, the intense burst of photons readily saturates PMT's, hence linearity must be

carefully checked.

2.4.2 Light Scattering with LII

The objective of the light scattering with LII experiments was to quantify the size

of the soot particles as they are heated up and vaporized. Three different 1064 nm laser

powers were used. The laser powers were 37.5, 29.0, and 20.0 mJ/pulse. These powers

correspond to laser fluences of 0.60, 0.46, and 0.32 J/cm2, respectively, based on a focal

spot of 5.94 mm2. The laser fluence is simply the laser energy in each pulse divided by









the beam area. The beam area was found by ablating ink off of a slide and measuring the

removed ink area. The 1064 nm beam spot was not quite a perfect circle, but was more

elliptical and had a major axis measuring 3.0 mm, and a minor axis measuring 2.67 mm.

The power of the 532 nm laser was 2.4 mJ/pulse. A summary of the laser beam power

properties is given in Table 2-4.

Table 2-4. Summary of laser settings for light scattering measurements


Fluence
Laser Control Setting Power (mJ/pulse) (J/cm2)


1064 10 37.5 0.60

1064 8 29.0 0.46

1064 6 20.0 0.32

532 Constant 2.4 N/A

According to previous research as described in Chapter 1, LII provides information

that can be translated into soot volume-fraction data. The precise time the LII signal is

captured is a very important factor when considering these experimental results. As the

LII laser heats up the soot particles they begin to emit radiation, which is the LII signal,

but they also begin to vaporize once their vaporization temperature is reached. This

vaporization may lead to skewed results if the particles being measured are being

destroyed at the same time. These current experiments track the size of the soot particles

as they are heated up and vaporized such that the time scale of vaporization and the LII

process is better understood, enabling a more thorough analysis of the LII process and

hopefully to more accurate implementation.

Monitoring the soot particle size while they are being vaporized was done by using

a second laser for light scattering measurements. For convenience, the 1064 nm laser









used for the laser heating will be referred to as the LII laser, while the 532 nm laser used

for light scattering will be referred to as the LES laser. The LII laser was held fixed in

time by delay generator 1. The LES laser was moved in time relative to the LII laser by

use of delay generator 2. The LES laser was fired 24 nanoseconds before the LII laser to

begin the experiment. This gave a measure of the size of the soot particles before any

vaporization occurs. The LES laser was then moved up in time in 2 nanosecond intervals

until it reached 20 ns after the LII laser pulse. The LES laser was then increased further

in time by larger intervals until it eventually reached 132 ns beyond the LII laser pulse.

From these measurements, the size history of the soot particles was tracked as they were

heated up and vaporized by the LII laser with a temporal resolution of 2 ns.

2.4.3 Variable Spatial Resolution

Once light scattering measurements were concluded, LII measurements were taken

to quantify soot vaporization and the LII response for multiple scenarios. A new aperture

was added to the system to allow for a decrease in the beam diameter of the 1064 nm

laser. With the new aperture in place, the beam diameter at the point of focus in the

flame was 1.55 mm. To make a comparison to the laser fluences from the previous

experiments, the laser power was adjusted for the new beam diameter to achieve similar

fluences. Two laser powers were used in the LII experiments, 11.5 and 8.8 mJ/pulse.

These laser powers correspond to laser fluences of 0.61 and 0.47 J/cm2, respectively.

These two laser fluences are very similar to the two highest laser fluences from the

previous experiments. Therefore, there are comparable laser fluences for both the large

beam diameter and the smaller beam diameter. A comparison showing the fluences and

diameters is given in Table 2-5 below.









Table 2-5. Comparison of 1064 nm laser beam diameters and fluences for LII
experiments



Beam Fluence
Diameter (mm) Control Setting Power (mJ/pulse) F(cm^2)




2.75 10 37.5 0.60
2.75 8 29.0 0.46


1.55 6 11.5 0.61
1.55 4.5 8.8 0.47

LII measurements were then taken for all four laser fluences. Another variable

added to the experiment was the size of the detector aperture. This aperture controlled

the amount of light allowed into the PMT to create the signal. For a larger aperture, more

light was allowed into the detector creating a larger signal. Also, the scattering volume is

larger for a larger aperture, thus that signal accounts for more soot particles. The first

aperture that decreased the beam diameter of the 1064 nm laser was placed before any

optics and before the laser beam reached the flame, therefore it only controlled the laser

fluence. The second aperture was set at either 2.0 mm or 3.6 mm. The second aperture

must be opened from its original position of approximately 0.5 mm from the light

scattering experiments to achieve a strong LII signal due to the markedly weaker LII

signal as compared to the LES signal.

These LII experiments were done with the PMT charged with -850 V. The flame

was kept at the same fuel-lean condition. Signal linearity measurements were taken

before each set a data was recorded. For each data set, 500 laser pulses were averaged to

determine the LII signal.









2.5 TEM Methods

Samples of the primary soot particles from the flame were taken using a small

copper mesh grid with a plastic coating. Several samples were taken by sweeping the

grid through the flame allowing soot to deposit on the grid. This is referred to as

thermophoretic sampling, where the temperature gradient between the hot flame and the

cold mesh drives the particles to the surface of the mesh where they are deposited. The

grids were attached to a grid holder and manually moved through the flame. Each sample

had a slightly different residence time in the flame, all less than ~1 second. If the

samples were kept in the flame too long, the plastic coating on the grid could be melted

and destroy the sample. If the samples were swept through too fast, there might not be

enough time for soot particles to deposit on the grid. The samples were taken at the same

height above the burner that the laser beams cross through the flame. This ensured the

TEM samples were a representative cross-section of the particles analyzed by laser-based

diagnostics.

By use of a transmission electron microscope (TEM), digital photographs were

taken of the soot samples so that measurements could be made to determine primary

particle size. Each photograph had a scale indicating length, and particle diameters could

be simply measured from the pictures. The major axis length, minor axis length, and an

estimate of the average diameter were all recorded for 30 soot samples. From these

measurements, the statistical average and standard deviation for the soot primary particle

size was determined. Two representative sample TEM photos are shown in Figures 2-7

and 2.8. The soot particles in Figure 2-7 are single loose particles. The soot in Figure 2-

8 is an agglomerated chain of soot particles.











































Figure 2-7. Three individual soot particles photographed by TEM











































Figure 2-8. Agglomerated chain of soot particles photographed by TEM














CHAPTER 3
RESULTS

3.1 Temporal Alignment

The first task in setting up the experiments was getting precise temporal control

between the two lasers. In order to eliminate jitter, both lasers were controlled using

separate digital delay generators. The 532 nm (light scattering) laser was kept at a

constant power of 2.4 mJ/pulse, but the 1064 nm laser power (LII laser) was varied

between three different powers. For each laser, the delay between the flashlamp and Q-

switch was optimized and remained constant. The first delay generator was used to

trigger the flashlamp and Q-switch of the LII laser. This delay generator also triggered

the second delay generator, which then was used to trigger the flashlamp and Q-switch of

the 532 nm laser. By varying the delay time between the triggering of the second delay

generator and triggering of the flashlamp, the pulse separation between the two pulses

was controlled with jitter between the two pulses of about 1 ns. Because the LII laser

pulse energy was controlled by adjusting the flashlamp voltage, the temporal pulse timing

was slightly altered. Therefore, for each pulse energy, the digital delay settings were

determined for zero pulse separation. These settings were stored, and recalled whenever

the pulse energy was changed. Note the pulse width of the 532 nm scattering laser (~5

ns) is less than the pulse width of the LII laser (-13 ns). Figures 3-1, 3-2, and 3-3 show

the two laser pulses on top of each other for their respective delay times.















0.014

0.014 -----
532 nm
0.01 ------- ------ --- 1064 nm


S 0.008
00 08 --- ------------ ------------ J--------- ^--- -- ---1 6 n ------------------


S0.004 -- ------ --- -------------- L ----L ------------------- -- ---_
S 0.004 ii



0.002 -

0

-0.002
0 20 40 60 80 100
Time (ns)


Figure 3-1. Simultaneous laser pulses for LII laser at 0.60 J/cm2


Figure 3-2. Simultaneous laser pulses for LII laser at 0.46 J/cm2


0.012
0.012 ------------------------------------
0.01 ---

0.008 532 nm ---- ----1--

'- 0.006

10.004
|0 .0 0 6 - - --- - - ---- - - --- - ------^- ----\- - - -

0.002 ----- ~--- --- ------------- -- ----------- :-----\- ^-- - -
0.002 -

0
-0.002 ------------ --------------------------------

-0.002
0 20 40 60 80 100
Time (ns)














0.012

0.01 ------ ---------- .---- ------ -
0. 0 10-8 -- --- - --- --- ---- --- - -- -

0.008 -.- - 532 nm
-1064 nm
O 0.006 ---- --------------------------------------------- --. ---------..

5 0.004 -

0.002 ------------ ----- ------ -- -- ---- ------



-0.002 i
0 20 40 60 80 100
Time (ns)


Figure 3-3. Simultaneous laser pulses for LII laser at 0.32 J/cm2



The light scattering experiments all began with the 532 nm laser fired 24 ns

before the LII laser to get a baseline recording representing the soot before any

vaporization or incandescence caused by the LII laser. Figures 3-4, 3-5, and 3-6 show the

difference between the two laser pulses when the 532 nm laser is set 24 ns ahead of the

LII laser for the three LII laser fluences of 0.60, 0.46, and 0.32 J/cm2, respectively.















0.014

0.012 -----

0.01 -
0.1 ----------- ---------- --^-,---------
0.008 -

0.006 --- ----------------- ---------

a 0.004 --- - -- ---- -- -------

0.002 -------------- ----- -------- --- ----

0-
-0.0 ----------------------------------------------------------
-0.002
-20 0 20 40 60 80 100
Time (ns)


Figure 3-4. Laser pulses delayed by 24 ns for LII laser at 0.60 J/cm2


Figure 3-5. Laser pulses delayed by 24 ns for LII laser at 0.46 J/cm2


0.012

0.01 --------

0.008 --------- ------------

0.006 -- --- ------- ------ -- ------


S 0.004 ----------- -------- ----------- --------



0
5 0.002 ^---- - - - --- ---- -; ----^ -- --- \-- ---- - - -

0.002 ----------- -------- -----L ----J --- -----------\ -------- _


-0.002 I
0 20 40 60 80 100
Time (ns)













0.012

0.01 ----

0.008 ------------------------ -------- -----------

S 0.006 ----
-S 0 0 6 -- --- ---- I------ ----- -- ----- ---- :- --------
= 0.004 ---

0.002 ------------ --------------

0
-0.0 ---------------------------------------------------------
-0.002
-20 0 20 40 60 80 100
Time (ns)



Figure 3-6. Laser pulses delayed by 24 ns for LII laser at 0.32 J/cm2


3.2 Signal Linearity

Signal linearity measurements were taken to ensure the PMT was recording

accurate data over a wide range of signals. The light scattering signals fluctuated by

more than an order of magnitude during measurements and this variation must be

recorded on a linear scale by the PMT to be able to draw fair conclusions from the data.

To test the signal linearity, an LII measurement was taken without the light scattering

laser running, then a 50% transmission neutral density filter was added to the collection

optics and another LII measurement was taken without the light scattering laser running.

The ratio of the signals was calculated to determine if the detector was operating in a

linear range. This linearity test was performed before every set of experiments. A









sample plot showing the signal with and without the filter is shown in Figure 3-7 below.

The sample plot is taken from an LII experiment.





0.025
0.025 -------------------------------------------------r----
0.02 ---.......... I-
',- -- No Filter
--With Filter
0.0 15 --- -----.----------------- -------------
0.01 ----- V --------- ---------------------------


a 0.005 --------



-0.005
0 2000 4000 6000 8000 1 104
Time (a.u.)



Figure 3-7. Signal linearity verification plot

As shown in Figure 3-7, the ratio of the peaks of the signals is 0.53. This ratio

confirms that the recorded data is linear over a wide range of signals. For all

experiments, the average signal linearity ratio was 0.54, with a standard deviation of

0.054.

3.3 Simultaneous Light Scattering and LII Measurements

Light scattering measurements were taken for various LII laser powers to monitor

the size of the soot particles, enabling determination if the soot particles were vaporized

by the LII laser. Data measurements were started when the 532 nm scattering laser was

24 ns ahead of the LII laser in time. This gave a signal that represented the soot before

any vaporization occurred. The 532 nm laser was moved forward in time (i.e. toward the

LII laser pulse) in 2 ns increments until it reached 20 ns after the LII laser pulse. The









time increments were then increased and measurements were taken until the 532 nm laser

was 132 ns after the LII laser pulse. Since a 532 nm bandpass filter was used in front of

the PMT, the incandescent signal from the LII laser did not affect the much stronger

scattering signal from the 532 nm laser. This was verified by the absence of any signal

for the LII laser alone. Only if the apertures entering the PMT were opened and the

voltage gain on the PMT was raised significantly was the LII laser pulse detectable.

Hence for these measurements, the LII laser was used simply to vaporize the soot

particles and contributed no spurious signal to the light scattering signal.

In order to eliminate any minimal contribution from the LII laser, mainly

electromagnetic interference, baseline measurements were taken before and after the

scattering measurements. These two baseline measurements were averaged and

subtracted from the scattering data during data analysis. A sample averaged baseline

measurement is shown in Figure 3-8.







42





0.0002


0 0 0 0 1 -. --. --.- - --.- - - - - . . .. . . . . .. . . ,I . . . .
0.0001 --------------------------------------





-0.0001 -




-0.0002 -- -------------------


1.4 10 .7 1.6 10 .7 1.8 10 .7 2 10 .7 2.2 10 .7 2.4 10 .7 2.6 10 .7
Time (ns)


Figure 3-8. Sample average baseline measurement from scattering experiments

Once the baseline measurements were averaged, the baseline was subtracted from

the scattering data to give a corrected value. The integrated peak area was then

calculated corresponding to a temporal width of about 13 data points or about 3.3 ns for

the current digitization rate of 4 GS/s. The integrated values were then plotted as a

function of time to show the decay of the soot particles as they were vaporized by the LII

laser. Figure 3-9 shows a single set of light scattering measurements.














0 I I

-0.3
-0.2 --------------
2 -- - - - - - --- - - - - '-- - - - _

-. 3 2 -------------- ----------- /------------ ,-- ------------ ------------
-0.3 ------------------------- ----- ----

S-0.54 -------------------------------- ----- -


-0.6 --



-30 -20 -10 0 10 20 30
Time (532 nm laser pulse relative to 1064 nm laser pulse)

Figure 3-9. Integrated light scattering measurement

Figure 3-9 is a light scattering measurement from an experiment where the

scattering laser was only increased in time until it reached 20 ns after the LII pulse. This

type of experiment was repeated eight times for each of the three LII powers to create a

statistically accurate average and standard deviation for the scattering experiments. It

was determined that more data was required to follow the decay of the soot particles long

after the LII laser pulse. Therefore, another set of experiments was performed in which

the initial scattering measurements were taken beginning when the scattering laser was 24

ns ahead of the LII laser, then the scattering laser was moved in 2 ns increments until it

reached 20 ns after the LII laser. The scattering laser was then moved in 2 to 16 ns

increments away from the LII pulse. Measurements were continued to a delay of about

132 ns, where there appeared to be no additional changes in the signal. A sample plot of

the extended data set is shown in Figure 3-10 below.







44








-0.2




-0. -------- -- ----------
O-.0 2 --------------------------------------------------- ------ ---- -- --- -- --_




-0.0 6 ------------- ----------- --- ----- --- ----- ----------------------------------_


-0.8 -----------------------------------------------------------------------------------------

-1
-50 0 50 100 150
Time (532 nm laser pulse relative to 1064 nm laser pulse)


Figure 3-10. Extended data set for light scattering measurements

Because the two sets of scattering data were collected several months apart, four

data points were selected to overlap between the two sets of scattering experiments.

These overlapping points along with the initial data point (-24 ns) were used to match the

2 data sets and merge them into one comprehensive data set. The absolute difference in

the common region between the two data sets was about a factor of 2.3. This constant

scale factor was then used to match the data. The comprehensive data sets for the three

LII laser powers are plotted in Figures 3-11, 3-12, and 3-13. Note that the scattering data

are normalized to an initial value of unity at the initial delay of -24 ns.














1.2
1: ---------------- -------------------- ---------

0 ----------- -- ----- -------------- -- -----------------------------------
1 -



= 0.8 ------------------ ------------------


0------------------


0.2 _----------------

0
-50 0 50 100 150
Time (532 nm laser pulse relative to 1064 nm laser pulse)


Figure 3-11. Light scattering measurements for LII laser set to 0.60 J/cm2


1.2

1- -- ----

S0.8 -

S0 .6 -- --- -- -- --- --- -- ------ ----- -- --------------- --- -- -- -
0- -

S 0.4
0.2 --- --- -- -

0. -----------------------------------------------------------------------
0 I I
-50 0 50 100 150
Time (532 nm laser pulse relative to 1064 nm laser pulse)


Figure 3-12. Light scattering measurements for LII laser set to 0.46 J/cm2







46





1.2


0 ----------- -- -------------------- -------------------------- -------------------

S 0.8

CO
0 .6 --------- ------ \ - ------- -- -------

0.2



0.
0 .2 -- -- -- -- - -- -. .. ..^ ^
0~--1 -----------------------------


-50 0 50 100 150
Time (532 nm laser pulse relative to 1064 nm laser pulse)

Figure 3-13. Light scattering measurements for LII laser set to 0.32 J/cm2

3.4 Variable Spatial Resolution LII

As discussed above, the aperture to the PMT was closed to a minimum size (<1

mm) to prevent any LII signal or stray light while the scattering was recorded. Upon

completion of the scattering measurements, LII data were recorded in the absence of the

532 nm scattering laser. The 532 nm line filter was retained, hence the LII signal

corresponds to a bandwidth centered at 532 nm, as shown in Figure 2-5. LII experiments

were performed for various setups. Two separate apertures were controlled to either alter

the LII laser beam diameter or alter the amount of light that entered the PMT to provide a

signal. With the first aperture (laser aperture) closed down to shrink the LII laser beam

diameter, the laser power was decreased to keep the laser fluence at a similar level to that

of the nominal diameter LII laser beam. The second aperture (PMT aperture) controlled

the amount of light entering the PMT, therefore controlling the optical volume imaged

onto the PMT detector.







47


For each LII response recorded, the offset (no laser) was subtracted from the signal

to zero the baseline of the data. Each experiment was repeated 12 to 15 times until a

statistically accurate average and standard deviation was calculated for the experiments.

This ensemble average of many experiments is the data reported below.

The first set of experiments was run with the first aperture set to shrink the LII laser

beam diameter to 1.6 mm. Two laser powers were used to create two laser fluences of

0.61 J/cm2 and 0.47 J/cm2. These two laser fluences were run using two different PMT

aperture sizes. The first aperture diameter used was 2.0 mm, and the larger aperture

diameter was 3.6 mm. The LII response for these experiments is shown in Figures 3-14

through 3-17 below.


0.025

0.02 -------------- -

S0.015 -------------- ------------ -------------- -- ----- --- ---



0.005 ----0.01

0 --------- ----------------- ------------

-0.005
-500 0 500 1000 1500 2000
Time (ns)

Figure 3-14. LII plot for 0.61 J/cm2 fluence and 2.0 mm PMT aperture














0.1
0.08 -------------------------------------------------------
0.08 -

0 .06 ------------ -- ---------- ---- ---- -

S 0.04 --- ------ ----------- --

0.02 -

0
-0.0 ----------------------------------------------------------
-0.02
-500 0 500 1000 1500 2000
Time (ns)


Figure 3-15. LII plot for 0.61 J/cm2 fluence and 3.6 mm PMT aperture


0.014
0.012 ------------i--------------------------------------------
0.012 -

0.01

0.008 --------------
S 0.006

I 0.004 -----------------

0.002 -

0-
------------ -------------------------------------
-0.002
-500 0 500 1000 1500 2000
Time (ns)


Figure 3-16. LII plot for 0.47 J/cm2 fluence and 2.0 mm PMT aperture






49




0.07
0.05 |---------------------------------------- ----------------I
0 .0 6 -. ..- .-- ----
0.05 -
0.04 ---------------
S0.03 ------------- ------------ -
0.03 -
0.02 --- --------- --.-----.
0.01 -
0 .0 --- .------------- ----------------- .---------------- : ------------------1----------------
---------------
-0.01 ------------------------------------- ------------------------
-0.01
-500 0 500 1000 1500 2000
Time (ns)

Figure 3-17. LII plot for 0.47 J/cm2 fluence and 3.6 mm PMT aperture

After this first set of LII experiments, the first aperture was opened back up to

allow the LII beam to return to its original diameter of 2.75 mm. The laser was then run

using two different laser powers to create laser fluences of 0.596 J/cm2 and 0.461 J/cm2.

These two fluences are within 2% of the fluences from the first LII experiments with the

reduced beam diameter. The same PMT aperture sizes were used as in the first set of LII

experiments. The LII response for these experiments is shown in Figures 3-18 through 3-

21 below.



























Figure 3-18.


LII plot for 0.60 J/cm2 fluence and 2.0 mm PMT aperture


0.2
0.15-----

o0.1 -- --- ---- ------- ---- .
H ,0 .0 5 -- ----- -- -- ----- - -- --------- ------ ----- -- .- - -
- - : - - - - - -
'w 0.05

0--------- --------
-.5------------------------------------------- -------
-0.05
-500 0 500 1000 1500 2000
Time (ns)

Figure 3-19. LII plot for 0.60 J/cm2 fluence and 3.6 mm PMT aperture


0.07 F
0.06 ------------- ----------------------- ---------------- -------------
0.05 -
0.04 ------- ---------------------
0.03 -
S0.02 ------------------- ---------------- ------------
0.01
0 :- -
-0.0 ---------------------------------------- -----------------
-0.01
-500 0 500 1000 1500 2000
Time (ns)


































Figure 3-20. LII plot for 0.46 J/cm2 fluence and 2.0 mm PMT aperture


0.2


0.15 .---------------------




w 0.05 -

---0.1 -------- ------------- ---------
--------- ; ------ ------- .--- ----------- ------ -- .----- ------ ----- _



-0.05
-500 0 500 1000 1500 2000
Time (ns)


Figure 3-21. LII plot for 0.46 J/cm2 fluence and 3.6 mm PMT aperture

A summary of the peak intensities for each of the LII plots shown in Figures 3-14

through 3-21 is shown in Table 3-1 below.


0.06
0.05 ------------------.------------.................|-
0.05 -

0.04 --- -

0.03
a 0.02 -

0.01 -
0 -.---- -- -- --------------

-0.0------------------------------------------------------------
-0.01
-500 0 500 1000 1500 2000
Time (ns)









Table 3-1. Summary of peak intensities for variable spatial resolution experiments


Fence LII Beam PMT Peak
Fluence Peak
(J/cmA2) Diameter Aperture Size Figure #
(mm) (mm)

0.61 1.55 2.0 0.020 3.13
0.61 1.55 3.6 0.082 3.14
0.47 1.55 2.0 0.014 3.15
0.47 1.55 3.6 0.064 3.16
0.60 2.75 2.0 0.066 3.17
0.60 2.75 3.6 0.190 3.18
0.46 2.75 2.0 0.058 3.19
0.46 2.75 3.6 0.170 3.20

3.5 Transmission Electron Microscopy

The transmission electron microscopy results gave data that could be used to

determine the primary particle size. 35 individual digital photographs were taken using

the TEM. Each of these photographs contained an absolute scale indicating length.

Using a ruler to measure the size of the particles, a list of the particle sizes was compiled.

27 separate particle sizes were used to statistically determine the average and standard

deviation of the primary particle size. A zeroth-order logarithmic distribution (ZOLD)

was used to analyze the TEM data. The ZOLD function can be described as

Sexp(-- 2 /2) -(Ina/a )2
p(a) = exp /22 2 Equation 3.1


In Equation 3.1, a is the particle diameter, am is the modal diameter, and Go is a

dimensionless measure of width used in the ZOLD analysis (Hahn et al 1995). It was

found from the TEM data that the mean particle diameter was 101 nm with a standard

deviation of 48 nm. Using the ZOLD analysis, the best fit was determined for a modal










particle diameter of 75 nm, with co=0.45. A plot of the probability of finding a particle

of a given size is shown in Figure 3-22 below. For the ZOLD function, the modal

diameter and oo are related to the mean and true standard deviation through Equations 3.2

and 3.3 shown below.


Ud aexp(1. 5a0


Equation 3.2


S= am [exp(4 0 ) exp(3C ) 12 Equation 3.3


Hence it is seen that the modal value of 75 nm and Co yield the mean of 101 nm

and the standard deviation of 48 nm as given by the TEM data.


0.012

0 .0 1 ---------
0.008 ---------------



0.0 --- ---------------- -- ------------



0.002
CO 0.006 --------- ------------ - --------- -- --------- ----. ----------- L----------- -






0
0 .0 0 4 -- - ------ --------- ------------- -- ---- .-- -- - -

0 .0 02 -- --------- ------------- :- ------------- -------- I -.--- -- -- :------^ ----- _


0 100 200 300 400 500 600
Modal Diameter (nm)


Figure 3-22. Zeroth-order logarithmic distribution of modal diameter














CHAPTER 4
ANALYSIS, DISCUSSION, AND CONCLUSIONS

4.1 Light Scattering

Light scattering measurements reveal a decay in the soot particle scattering signal

following the firing of the LII laser pulse. The decrease in scattering signal is attributed

to soot particle vaporization. This decay in scattering signal can be correlated to a

decrease in particle diameter using scattering theory. To make this correlation, several

steps must be taken to deduce the particle diameter from the data. First, the TEM results

must be used to determine the primary particle size before any laser interaction. Then

using this baseline diameter as a starting value for the scattering measurements, the signal

is related to the differential scattering cross-section to correlate the light scattering signal

to the particle diameter as a function of time.

As given in Chapter 3, the modal diameter for the soot particles is 75 nm based on

TEM analysis and the ZOLD fit. This result came from the TEM analysis and the zeroth-

order logarithmic distribution fit. The differential scattering cross-section may be

calculated using Mie theory, and is a function of the modal diameter, complex refractive

index, scattering wavelength, and scattering collection angle. Using the TEM result for

modal diameter of 75 nm, a complex refractive index of m = 1.6 0.6i (Smyth and

Shaddix 1996), scattering wavelength of 532 nm, and scattering collection angle of 90,

the scattering cross-section can be calculated, which corresponds to the soot particles

before the LII laser interaction. The measured scattering signal is a function of the

collection optics, solid angle, the detector efficiency, differential scattering cross-section,









and particle number density. The equation for the scattering signal is given by Equation

4.1 below.

Sr, = 3 C' N Equation 4.1

In equation 4.1, Sr, is the signal, P3 is an efficiency associated with the collection

optics, solid angle, and the detector efficiency, C',, is the differential scattering cross-

section (cm2sr-1), and N is the number density (particles/cm3). The VV subscript on the

signal S and the differential scattering cross-section C'>, refer to the polarization state of

the incident laser light and the scattered light, which are both vertically polarized with

respect to the horizontal scattering plane. Taking the ratio of the signal for any time

versus the signal for t=0 (before any laser interaction with the soot particles) will result in

Equation 4.2.

S, (tQ) = C' (t)
Sv(t= 0)=3 *N*C'(t= 0) Equation 4.2


Since the collection optics do not change between experiments, P3 may be canceled

out in Equation 4.2. Also, since the diffusion time scale is long (on the order of

microseconds) compared to the temporal range of scattering data (-150 ns), the particles

can be considered "frozen" in time during the measurements, therefore the number

density is constant and can be cancelled out in Equation 4.2. The result after cancellation

and rearrangement is shown in Equation 4.3 below.

S (t)
C', (t) = C, (t = 0) S- = f(t) Equation 4.3
SVV (t = 0)


Now the differential scattering cross-section of the particles is simply a function a

time. The differential scattering cross-section at time t=0 was found from the TEM







56


analysis, and the ratio of the signal at any time versus the signal at time t=0 can be easily

found from the experimental data. With the differential scattering cross-section known

for all times, Mie theory can be used to correlate the differential cross-section to a modal

diameter. Using Mie theory with the given scattering wavelength, scattering collection

angle, and complex refractive index, a calibration plot can be formed relating modal

diameter and differential scattering cross-section for a constant value of oo = 0.45. This

calibration plot is shown below in Figure 4-1.





120
-------- ------------- --------- -----------------------------
100 -- --
E
80 --. . .. .. .
E 60 ---
.Cu

20 -- ---------- ----- -- -
0



0 1 10-12 2 10-12 3 10-12 4 10-12 5 10-12 6 10-12 7 10-12
Differential Scattering Cross Section (cm^2/sr)

Figure 4-1. Mie theory calibration plot

Using a fourth order polynomial curve fit, an equation for the modal diameter as a

function of differential scattering cross-section was calculated. This curve fit is shown in

Equation 4.4 below.

Dmoda = 34.017 + 4.1057E13 C, 2.4993E25 *C C2 + 8.5491E36 C'3 1.083E48 C,4
Equation 4.4

This curve fit equation yields a correlation coefficient of R=0.99968.







57


The particle modal diameter can now be found directly from the scattering signal as

the particles decay due to the LII laser induced vaporization. A plot of the particle

diameter as a function of time for each of the three LII laser fluences is shown in Figures

4-2 through 4-4 below.


70 |-----------------------------------------------------------

70 -- ------~ ~--- k -------------- -- ------ ------ -

.26 0 ----- ----- ----------------- ----------------------

8O 1
0 ------------------- ---- -- ---- ----- ------
E77






60
40 ---------------------- ---------------------------------------



-50 0 50 100 150
Time (532nm laser relative to 1064 nm laser)


Figure 4-2. Modal diameter as particles decay for 0.60 J/cm2 LII laser fluence













80

75 ----------------------
6 5 ------------ --- ---- --------------------- ------------------------------------------ -
E 70 ---- -------- ------------------- -


E 60 -
-5 5 - - - - - -. . . . . . ... - - - - - - - -

4 5 _---......-----....... i^ .....-- -- -- - -- - ---- -- -- --- -- --

40
40 ----------------------
-50 0 50 100 150
Time (532 nm laser relative to 1064 nm laser)

Figure 4-3. Modal diameter as particles decay for 0.46 J/cm2 LII laser fluence


80

75 ------------ ------------------

j^70 _----- -- -- ^--- --- ----- --------------------------------------------------

65 ------ --------------
E
76 6

o 55 --- ------------- ------------- --------------------

50 ---
450 --------------------------------------------
45 I
-50 0 50 100 150
Time (532 nm laser relative to 1064 nm laser)

Figure 4-4. Modal diameter as particles decay for 0.32 J/cm2 LII laser fluence

The nondimensional particle volume fraction, which represents the fraction of gas

occupied by solid soot particles, can be found from Equation 4.5 below.










f, =N J -ta3 p(a)da Equation 4.5
aO3

Substituting in the zeroth-order logarithmic distribution (ZOLD) for p(a),

integrating from a=0 to infinity, and dividing both sides of Equation 4.5 by N, the particle

number density, the average particle volume can be found according to Equation 4.6.


P f, -47d3 exp(3C 2) Equation 4.6
N 3

In Equation 4.6, Pvoi is the average particle volume, U is the mean particle

diameter, and co is a dimensionless measure of width used in the ZOLD. The mean

diameter can be related to the modal diameter through Equation 4.7.

S= am exp(1.5y ) Equation 4.7

Converting modal diameters to average particle volume yields the plots of average

particle volume versus relative time as shown in Figures 4-5 through 4-7.





9106
8 10 ..........
810 --------^ --------------------------------------------
S 710 ------ -----------------------
E
6 106
-- 5 106 -



3 106
21 06
24 106 -------------- --- ---- ---------------------------------------------------- -

210 16 ----------------------^--^x --
1 106
-50 0 50 100 150
Time (532 nm laser relative to 1064 nm laser)

Figure 4-5. Average particle volume as particles decay for 0.60 J/cm2 fluence

































Figure 4-6.


Average particle volume as particles decay for 0.46 J/cm2 fluence


Average particle volume as particles decay for 0.32 J/cm2 fluence


9 106
8 1 6 .......... ------ ------------------ --------- --------------------
8 106
< 7 106 . -- -

6 106
E
- 5 106
S 4 106
5 10 --- -- ----------- ------------ - -- ------------------------------------ -

a 3 10 --------------- ----------------- ---- -
2O 3 o 10- - -> __- - -------------- -------------------- -

2 10 6--------------------------- ----------------------------
1 106 I I
-50 0 50 100 150
Time (532 nm laser relative to 1064 nm laser)


9106

8 10 6 i i i i- -- -

E 7 1 0 . . ..-- -- -- -- -

S 6106 -------------- --------------


ou 6
S 410 ----------- ----- ------

3 106
3 10 6 ------------- ^ .----------- --- --- ---------------------------- -
2106------------------------------------
2 106
-50 0 50 100 150
Time (532 nm laser relative to 1064 nm laser)


Figure 4-7.









4.2 LII Analysis

LII measurements were taken for various laser fluences and PMT aperture sizes.

The results of these experiments were presented in Chapter 3. One goal of this project

was to be able to predict the LII signal and the influence of particle vaporization. From

heat transfer theory it has been shown in Chapter 1 that the LII signal can be described as

a function of soot particle volume and particle temperature. The equation describing the

relationship between the LII signal and its dependent parameters is given by Equation 1.5

shown below.

SLIT = n N[Pvo1 (t)FPlanck (k,Tp (t))- P (t = O)FPlanck (, Tg)] Equation 1.5

In Equation 1.5, SLII is the LII signal, r" is a constant parameter which represents

the effect of the collection optics, solid angle, and the detector efficiency, N is the particle

number density, Pvol is the average particle volume, and FPlanck is a function of the Planck

distribution which describes the radiation emitted by the particles. FPlanck is defined

below in Equation 1.6.

C
Fplanck = -1 Equation 1.6
5 exp( hc)-1


In Equation 1.6, h is the Planck constant, Co is the speed of light, X is the detection

wavelength, k is the Boltzmann constant, T is the temperature of the emitting particle,

and Ci is the first radiation constant defined by Equation 1.7.

C1 = 27hc2 Equation 1.7

Pvol(t) and Fplanck(k,Tp(t)) in Equation 1.5 are the time dependent particle volume

and Planck function. The Planck function is dependent on the detection wavelength and

the temperature of the soot particles, which are treated as perfectly emitting blackbodies.









The time dependent Planck function in Equation 1.5 uses the decaying particle

temperature to model the LII signal. Pvol(t=0) and Fplanck(k,Tg) are initial values and are

constant in the model. Pvol(t=0) is the initial particle volume found from light scattering

measurements before any LII laser interaction (i.e. t = -24 ns). Fplanck(k,Tg) is found by

using the initial soot particle temperature, which is assumed to equal the flame

temperature. Tg is assumed to be the constant-pressure adiabatic flame temperature for

stoichiometric propane combustion, which is approximately 2260 K (Turns 2000). The

time resolved particle temperature Tp(t) is found from Equation 1.1 shown below.

-- -1 = (P 1) exp(-t/Tc ) Equation 1.2
T T
g g

The maximum particle temperature, To, and the characteristic cooling time,

-c, were varied to find an appropriate curve fit to model the LII signal. It was determined

through an iterative process that the best values to most accurately represent the LII

signal were To = 4800 K, and c = 900 ns. This maximum particle temperature is close to

values previously reported by Snelling (Snelling et al 1997). Snelling showed through

numerical models of the heat transfer process that the maximum particle temperature

should be between 4000-4500 K. Equation 1.2 was used only to describe temperature

decay region of the LII signal. The portion of the LII signal modeled by Equation 1.2 is

labeled "Decay" in a sample LII response in Figure 4-8 below.

The particle temperature rise from the gas temperature of 2260 K to the maximum

particle temperature of 4800 K was modeled as a linear rise in temperature. The rise in

temperature is labeled in Figure 4-8 as rise. This temperature rise occurs over a period of

about 30 ns, compared to the exponential decay of temperature, which takes







63


approximately 1700 ns to return to the gas temperature from the maximum particle

temperature. The constant q*N used in Equation 1.5 was found to be 2550. A plot of the

particle temperature rise and decay for a laser fluence of 0.47 J/cm2 and a PMT aperture

opening of 2.0 mm is shown in Figure 4-9 below.




Rise > De-
< Decay >
0.014

0.012 --------

0.01

S 0.008 -

0.006 -

0.004----- -------

0.002 -

0
-0.002 -----------------------------------------

-0.002
-500 0 500 1000 1500 2000
Time (ns)


Figure 4-8. Sample LII response with rise and decay of particle temperature













5000

4500 ---
24000 ---------- I------ --- ---------- ---------------- ------- --------_


C O 3 5 0 0 - - -- - - --- - - ---- I-- -. - - -


E


3500 -------------------
F- 3000


2500 -


2000
-500 0 500 1000 1500 2000
Time (ns)



Figure 4-9. Particle temperature rise and decay model

The particle volume was found from light scattering measurements at discrete

points in time before, during, and after the LII laser pulse. These points were then curve

fit to be able to determine the particle volume continuously for the entire duration of the

LII laser pulse, which was much longer than the initial light scattering measurements.

The complex nature of the particle volume decay did not allow a simple exponential or

polynomial curve fit. Therefore, the particle volume plot was broken up and analyzed

piecewise to find a set of equations to describe the decay of the particles. All times used

to describe the decay are relative times. More specifically, the LII laser was held

constant in time as the LES laser was precisely controlled relative to the LII laser. From

-10 ns to -2 ns, a third order polynomial curve fit was used to describe the particle

volume. From -2 ns to 10 ns, another third order polynomial curve fit was used to model

the particle volume. From 10 ns to 24 ns, a final third order polynomial curve fit was









used to model the particle volume. From 24 ns to the end of the LII signal,

approximately 1700 ns, a power law curve fit was used to model the very slight decay of

the particle volume over this additional range. After 24 ns, the data from the light

scattering measurements mostly flattened out due to the absence of the LII laser pulse,

which was the cause of the particle decay. However, the particles are still hotter than the

gas temperature and are most likely gradually decaying via vaporization, thus a perfectly

constant approximation is considered insufficient to model the particle volume. A plot of

the modeled particle volume along with the light scattering data of the experimental

particle volume is shown in Figure 4-10 below.













0.009


0.008 --------


0.007 --.----











0.003 .... ........
0.006 --------






0.005 -






0.001
00.003
0.001 ----
0 "


-- --- . --



-- -






- --.. .- --


-- . ..-!-- -- -

-- -- -- -



600 800
Time (ns)


Figure 4-10. Particle volume experimental data and curve fit prediction

Using the above models and curve fits to describe the particle temperature and

particle volume, they can be applied to Equation 1.5 to predict the LII signal. A plot of

the LII response for a laser fluence of 0.47 J/cm2 and a PMT aperture opening of 2 mm

along with the predicted model plot is shown in Figure 4-11 below.


Experimental Data
- Curve Fit





















1000 1200 1400












0.014
0.012 ----------------- i-----------
0.01-------------- ---Lll Signal _
S Model
0.008 ----------------- -------- ----- ----
0.006 -
0.0 --------- -------------I ----S----
0 .0 0 4 -- --------- -- --- -- -- ------ --- -- ---- -- -- -- --
0 -- (-- ------- --- ------ -.-.-.- ....--------- _
S0.002 ----------------
0.02---------------------

-0.002
-500 0 500 1000 1500 2000
Time (ns)

Figure 4-11. LII signal with model prediction

The model used for the prediction of the LII signal in Figure 4-11 uses the particle

volume data obtained from light scattering measurements with an LII laser fluence of

0.46 J/cm2, which is identical within experimental precision to the fluence used in the LII

experiment.

The LII laser effectively evaporates the soot particles during the time course of

the LII laser pulse. The laser heats up the particles and they lose mass as they are heated

past their vaporization temperature (~ 4000 K for carbon). One interesting application of

this predictive model is to see what would happen to the LII signal if the particle volume

did not change due to the LII laser. In this case, the term Pvol(t) in Equation 1.5 would be

a constant equal to the initial particle volume as determined from light scattering

measurements, instead of a time dependent quantity found from the light scattering

measurements. A plot showing this case for Pvol = 0.00777 |tm3 is shown in Figure 4-12

below. Clearly, the effect of soot particle vaporization is significantly decreasing the LII







68


signal by about a factor of 4. Because the vaporization time scale is on the order of the

LII pulse, even prompt LII will not avoid vaporization effects.


0.08


:\ LII Signal
0 .0 6 - -- -- -
0.06 Variable Pvol Model
0 4- -- Constant Pvol Model

S0 .04 --- -- --- ---- --- --- -- ---- --- -- -------- -------- ----- ---- ---- --

0.0 ---\--- ---- ----








-0.02
-500 0 500 1000 1500 2000
Time (ns)


Figure 4-12. LII signal with 2 predictive models

4.3 Conclusions

In this study, a propane diffusion flame burner was constructed to develop laser-

based diagnostics to monitor soot emissions. Time resolved laser light scattering (LES)

and laser-induced incandescence (LII) were the two laser-based techniques, along with

transmission electron microscopy (TEM), used to determine primary particle size and

soot volume fraction. From these emission parameters, a model was created to predict









the LII signal as a function of particle volume and particle temperature. The following

conclusions have been drawn from the results of this study:

1. LII for the range of laser fluences used in these experiments results in the
vaporization of soot particles. The time scale of vaporization is on the order of the
length of the laser pulse (-13 ns). Whether prompt or delayed LII detection is
used, the size of the soot particles will be affected by the LII laser, thereby
changing the parameter of interest, namely the soot volume fraction.

2. LII is a technique that has a strong dependence on laser fluence and particle size,
thus careful calibration is required for quantitative analysis.

3. Varying the spatial resolution of the LII laser and collection optics showed a nearly
linear LII response. This indicates the LII signal is a nearly linear function of solid
collection angle. Consequently, LII is a robust technique that can be utilized for
various experimental conditions to measure soot volume fraction, notwithstanding
the above comments.

4. Modeling of the LII signal is a complicated process, but provides an accurate
representation of the particle temperature and volume. From these parameters, the
soot volume fraction can be deduced.

4.4 Future Work

It would be useful to extend the work to other flames with different primary

particle size, and different optical properties (i.e. absorption coefficients) to see the effect

of these parameters on the degree of soot vaporization and the correlation with the LII

signal. In addition, it would be interesting to extend the present technique to a post-flame

regime where the ambient gas temperatures are significantly lower. It is expected that

such lower initial temperatures may reduce the degree of vaporization.














APPENDIX A
ERROR ANALYSIS

The modal diameters calculated in Chapter 4 were done so by equating voltage

from the light scattering signal to a differential scattering cross section, which was then

converted into a modal diameter via Mie theory. These scattering signals were the

average of eight separate trials taken over a period of two weeks. From the collection of

data, the average and standard deviation of the signal was calculated. The percent error

in the modal diameter can be calculated from using the standard deviations of the

scattering signal. The percent error must be calculated for one standard deviation above,

and one standard deviation below the average. The percent error is given by Equation

A.1.


% Error = --- *100% Equation A. 1


In Equation A. 1, U is the modal diameter found using the average scattering

signal, and a is either the modal diameter found from using the average scattering signal

plus the standard deviation or minus the standard deviation. Thus, two percent errors

were calculated and averaged to find the actual percent error. Three tables displaying the

percent errors for the three different LII laser fluences are shown in Tables A-i through

A-3.










Table A-1. Percent error analysis for LII laser set to 0.60 J/cm2


Relative Standard Modal Percent
Time (ns) Deviation Diameter Error
(nm)

-22 0.964121 0.028245 100.9539 0.74
-20 1.009101 0.0648052 102.0167 1.26
-18 0.998578 0.086225 101.7925 1.78
-16 1.018703 0.0977756 102.2067 1.68
-14 0.96774 0.120399 101.0488 2.87
-12 0.93843 0.107887 100.2397 2.98
-10 0.968945 0.0693905 101.0801 1.73
-8 0.901033 0.0485525 99.09403 1.56
-6 0.794376 0.1047874 95.40911 3.90
-4 0.560083 0.1466562 86.94477 6.22
-2 0.328735 0.153763 77.51989 10.83
0 0.163637 0.0565334 66.33132 7.71
2 0.093394 0.0327758 59.12107 6.49
4 0.068699 0.0183006 56.10044 4.18
6 0.057276 0.0126266 54.60639 3.09
8 0.051171 0.0118403 53.78157 3.01
10 0.049484 0.0111555 53.55033 2.86
12 0.049203 0.0080787 53.51171 2.08
14 0.048683 0.0106221 53.44009 2.74
16 0.047629 0.0118138 53.29446 3.07
18 0.046044 0.0105041 53.07439 2.76
20 0.047447 0.0102373 53.26916 2.66
24 0.048814 0.0043942 53.45813 1.13
28 0.045849 0.0013186 53.04723 0.35
32 0.044327 0.0025912 52.83455 0.69
36 0.048723 0.0026363 53.44561 0.68
44 0.046157 0.0037812 53.09003 0.99
52 0.042222 0.0024192 52.53836 0.65
60 0.047182 0.0043895 53.23249 1.14
68 0.047508 0.0026639 53.2777 0.69
84 0.043978 0.002237 52.78555 0.59
100 0.047832 0.0030861 53.32248 0.80
116 0.048285 0.0028879 53.3851 0.75
132 0.048909 0.0020071 53.47117 0.52










Table A-2. Percent error analysis for LII laser set to 0.46 J/cm2


Relative Standard Modal Percent
Time (ns) Deviation Diameter Error
(n m)

-22 0.989674 0.0773909 101.5905 1.71
-20 0.974015 0.0473466 101.2099 1.17
-18 0.984151 0.0583743 101.4598 1.35
-16 0.957297 0.0906289 100.7709 2.35
-14 0.947542 0.0584807 100.5007 1.62
-12 0.939889 0.0932681 100.2821 2.59
-10 0.951412 0.0925583 100.6091 2.46
-8 0.935118 0.0842756 100.143 2.40
-6 0.891693 0.1230473 98.79191 3.87
-4 0.820649 0.075761 96.35206 2.77
-2 0.640676 0.0609746 89.82995 2.44
0 0.415088 0.055366 81.46595 2.84
2 0.253487 0.056232 73.19938 5.02
4 0.158022 0.0320715 65.82219 4.48
6 0.129522 0.032047 63.06621 5.19
8 0.110267 0.0180694 61.03023 3.25
10 0.10129 0.0203646 60.0295 3.85
12 0.097852 0.0143981 59.6373 2.77
14 0.093214 0.0163419 59.10008 3.23
16 0.091265 0.0156811 58.8715 3.14
18 0.087018 0.0097238 58.36762 1.99
20 0.089179 0.0142775 58.62507 2.89
24 0.088014 0.0074189 58.48657 1.51
28 0.084588 0.0076578 58.07578 1.59
32 0.087256 0.0053711 58.39606 1.10
36 0.087551 0.0061878 58.43133 1.26
44 0.086635 0.0075917 58.32184 1.56
52 0.086225 0.0059996 58.27273 1.23
60 0.090065 0.0087306 58.7299 1.76
68 0.08585 0.0044483 58.22766 0.92
84 0.086545 0.0062248 58.31106 1.28
100 0.0861 0.0042985 58.25771 0.89
116 0.086467 0.0056508 58.30166 1.16
132 0.090958 0.0087678 58.83537 1.76










Table A-3. Percent error analysis for LII laser set to 0.32 J/cm2


Relative Standard Modal Percent
Svv Diameter
Time (ns) Deviation Diameter Error
(nm)

-22 0.977224 0.0875407 101.2904 2.06
-20 1.049187 0.09297765 102.7076 1.10
-18 0.967168 0.11762424 101.0339 2.82
-16 1.085333 0.19473843 103.0692 4.39
-14 1.066238 0.21627485 102.9123 4.99
-12 0.930257 0.25733729 99.99912 5.64
-10 0.922373 0.22749247 99.76153 5.58
-8 0.970817 0.20496063 101.1284 4.03
-6 0.933981 0.19739844 100.1095 4.88
-4 0.88749 0.17978068 98.65418 5.41
-2 0.771167 0.14911987 94.56743 5.57
0 0.660326 0.09893379 90.53732 3.94
2 0.53499 0.12982527 86.0416 5.61
4 0.420302 0.0984546 81.68077 5.07
6 0.329307 0.07984077 77.54906 5.34
8 0.269421 0.06982805 74.20451 5.86
10 0.241374 0.07191752 72.39756 6.81
12 0.220468 0.0645295 70.93058 6.70
14 0.204267 0.0486971 69.71562 5.43
16 0.200596 0.05489329 69.43025 6.24
18 0.193481 0.04204589 68.86605 4.93
20 0.19106 0.04886439 68.67076 5.80
24 0.188227 0.00910239 68.43994 1.09
28 0.181295 0.00725069 67.86486 0.90
32 0.184343 0.00564552 68.11951 0.69
36 0.189649 0.00965329 68.55604 1.15
44 0.17846 0.00451513 67.62534 0.57
52 0.188507 0.00611587 68.46286 0.73
60 0.183394 0.01010787 68.04056 1.24
68 0.1872 0.01217863 68.35565 1.47
84 0.19162 0.00750837 68.7161 0.88
100 0.189216 0.01098262 68.52079 1.31
116 0.182177 0.01012132 67.93879 1.25
132 0.186291 0.01487665 68.28079 1.80














APPENDIX B
ZEROTH-ORDER LOGNORMAL DISTRIBUTION

The zeroth-order lognormal distribution (ZOLD) is a skewed probability

distribution. The ZOLD is used for modeling the distribution of aerosols and soot

particles due to those particles being skewed to larger sizes. A given system of particles,

each of diameter a, with true mean diameter and standard deviation, a and CY can be

represented by the ZOLD using the parameters of the modal diameter and a

dimensionless measure of width, am and Go, respectively. The probability distribution is

given by Equation B.1 below.

exp(-C-/2) (Ina/am)2
p(a) = exp 2G Equation B. 1


The mean particle diameter is found from Equation B.2 and the standard deviation

is found from Equations B.3 and B.4.

= a p(a)da -=am exp(1.5 2) Equation B.2


2 = (a )2 p(a)da Equation B.3


C =am [exp(4 )- exp(3 2)J /2 Equation B.4

In the ZOLD, p(a) is normalized, shown mathematically by Equation B.5.

fp(a)da = 1 Equation B.5

A plot of the ZOLD for three different values of co is given in Figure B-1 below.















0.03


0.025


0.02


p(a) 0.015

0.01


0.005


0


-0.005


1- 0.2
-----0.5
-0.8


Figure B-1. ZOLD distribution for three values of o0


0 50 100 150 200 250 300 350 400

Modal Diameter (nm)















LIST OF REFERENCES


American Institute of Aeronautics and Astronautics Website (2003).
http://www.AIAA.org Date accessed: 8/30/2003

Axelsson, B., R. Collin, and P.E. Bengtsson (2001). "Laser-Induced Incandescence for
Soot Particle Size and Volume Fraction Measurements Using On-line Extinction
Calibration." Applied Physics B, 72, 367-372

Dasch, C.J. (1984). "Continuous-Wave Probe Laser Investigation of Laser Vaporization
of Small Soot Particles in a Flame." Apllied Optics, 23, 2209-2215

Dobbins, R.A., and C.M. Megaridis (1987). "Morphology of Flame-Generated Soot as
Determined by Thermophoretic Sampling." Langmuir, 3, 254-259

Erickson, W.D., G.C. Williams, and H.C. Hottel (1964). "Light Scattering
Measurements on Soot in a Benzene-Air Flame." Combustion andFlame, 8, 127-
132

Fotou, G.P., S.J. Scott, and S.E. Pratsinis (1995). "The Role of Ferrocene in Flame
Synthesis of Silica." Combustion andFlame, 101, 529-538

Hahn, D.W., and T.T. Charalampopoulos (1992). "The Role of Iron Additives in
Sooting Premixed Flames." Twenty-Fourth Symposium (Int.) on Combustion, The
Combustion Institute, 1007-1014

Hahn, D.W., M.N. Ediger, and G.H. Pettit (1995). "Ablation Plume Particle Dynamics
During Excimer Laser Ablation of Polymide." Journal of Applied Physics, 77,
2759-2766

Kerker, M. (1969) The Scattering of Light and Other Electromagnetic Radiation. New
York: Academic.

Kock, B.F., Th. Eckhardt, and P. Roth (2002). "In-Cylinder Sizing of Diesel Particles by
Time-Resolved Laser-Induced Incandescence (TR-LII)." Proceedings of the
Combustion Institute, 29, 2775-2781

Melton, L.A. (1984). "Soot Diagnostics Based on Laser Heating." Applied Optics, 23,
2201-2207









Menna, P., and A. D'Alessio (1982). "Light Scattering and Extinction Coefficients for
Soot Forming Flames in the Wavelength Range from 200 to 600 nm." Xiiniteeith
Symposium (International) on Combustion, The Combustion Institute, 1421-1428

Prado, G., J. Jagoda, K. Neoh, and J. Lahaye (1981). "A Study of Soot Formation in
Premixed Propane/Oxygen Flames by In Situ Optical Techniques and Sampling
Probes." Eighteenth Symposium (International) on Combustion, The Combustion
Institute, Pittsburgh, Pennsylvania, 1127-1136

Rayleigh, Lord (1871). Nature, 3, 234-265

Roth, P., and A.V. Filippov (1996). "In Situ Ultrafine Particle Sizing by a Combination
of Pulsed Laser Heatup and Particle Thermal Emission." J. Aerosol Sci., 27, 95-
104

Schraml, S., S. Dankers, K. Bader, S. Will, and A. Leipertz (2000). "Soot Temperature
Measurements and Implications for Time-Resolved Laser-Induced Incandescence
(TIRE-LII)." Combustion and Flame, 120, 439-450

Smyth, K.C., and C.R. Shaddix (1996). "The Elusive History of m = 1.57 0.56i for the
Refractive Index of Soot." Combustion and Flame, 107, 314-320

Selling, D.R., G.J. Smallwood, I.G. Campbell, J.E. Medlock, and O.L. Gulder (1997).
"Development and Application of Laser-Induced Incandescence (LII) as a
Diagnostic for Soot Particulate Measurements." AGARD 90th Symposium of the
Propulsion and Energetics Panel on Advanced Non-Intrusive Instrumentation for
Propulsion Engines, Brussels, Belgium

Stephens, M., N. Turner, and J. Sandberg (2003). "Particle Identification by Laser-
Induced Incandescence in a Solid-State Laser Cavity." Applied Optics, 42, 3726-
3736

Tissue, B.M. (2000). http://www.chem.vt.edu/chem-ed/optics/detector/pmt.html
Date accessed: 9/3/2003

Turns, S.R. (2000). An Introduction to Combustion. Boston: McGraw-Hill

U.S. Department of Energy (1996). "1996 Annual Energy Review." DOE/EIA-
0384(96)

U.S. Environmental Protection Agency (1996). Federal Register 61, 77, 17357-17358

Vander Wal, and K.J. Weiland (1994). "Laser-Induced Incandescence: Development
and Characterization Towards a Measurement of Soot-Volume Fraction." Applied
Physics B, 59, 445-452

Wainner, R.T., and J.M. Seitzman (1999). "Soot Measurements in a Simulated
EngineExhaust Using Laser-Induced Incandescence." AIAA Journal, 37, 738-742






78


Witze, P.O., S. Hochgreb, D. Kayes, H.A. Michelsen, and C.R. Shaddix (2001). "Time-
Resolved Laser-Induced Incandescence and Laser Elastic-Scattering Measurements
in a Propane Diffusion Flame." Applied Optics, 40, 2443-2451

Zhang, J., and C.M. Megaridis (1996). "Soot Suppresion by Ferrocene in Laminar
Ethylene/Air Non-premixed Flames." Combustion and Flame, 105, 528-540















BIOGRAPHICAL SKETCH

Gregory David Yoder was born in Poughkeepsie, NY, on June 16, 1979, the second

of two children. In 1982, he moved to Tampa, FL, where he resided until the completion

of high school. In August 1997, he entered the University of Florida and received a

Bachelor of Science in mechanical engineering in December 2001 with honors. In

January 2002, he began pursuit of a Master of Science in mechanical engineering, which

incorporated the efforts of this master's thesis.