§§ “WWWWNW”!WWHIIHIHWIH“WI

   

THS

lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll

3 1293 01026 9748

This is to certify that the
thesis entitled

The Development of an Improved Quantitative Calibration
for an Exciplex Liquid/Vapor Visualization System
presented by

Roy Stanley Schafer

has been accepted towards fulfillment
of the requirements for

Master oi Science degree in Mechanical Engineering

Mam

Majé professor

 

Date August 11, 1993

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or bdoro data duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i-l IE
-L_]-

MSU IoAn Affirmative Action/Emu Opportunlty InstMon

 

 

 

 

 

 

 

THE DEVELOPMENT OF AN IMPROVED QUANTITATIVE CALIBRATION
FOR AN EXCIPLEX LIQUID/VAPOR FUEL VISUALIZATION SYSTEM

By

Roy Stanley Schafer

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Mechanical Engineering

1993

ABSTRACT

THE DEVELOPMENT OF AN IMPROVED QUANTITATIVE CALIBRATION
FOR AN EXCIPLEX LIQUID/VAPOR FUEL VISUALIZATION SYSTEM

By

Roy Stanley Schafer

Laser-induced fluorescence methods have recently been used for doing fuel
distribution studies in internal combustion engines. Exciplex liquid/vapor visualization
systems cause the liquid and vapor phases to fluoresce at different wavelengths when
excited with an ultraviolet laser, thereby allowing two-dimensional fluorescent images of
the two phases to be taken separately through bandpass filters.

Although the exciplex technique is not new, an improved quantitative calibration
method has been developed relating fluorescence intensity to mass concentration of liquid
and vapor fuel. An exciplex system composed of 1% N,N,N‘,N‘,-tetramethyl-p-
phenylenediamine (TMPD), 10% naphthalene, and 89% n-decane by weight was used.
The naphthalene/TMPD system is coevaporative with decane-based fuels such as diesel
fuel. Such a calibration can then be used to do quantitative liquid and vapor fuel

distribution studies.

ACKNOWLEDGMENTS

I would like to thank several people:

Dr. Harold Schock, my advisor, for his support which made it possible for me to go to
graduate school, and the opportunity to be associated with the MSU Engine Research
Laboratory.

Dr. Stanley Crouch for being on my advisory committee, for his assistance in helping me

understand the photochemistry involved in this project, and for the use of his Isolated
Droplet Generator which proved to be a reliable and stable instrument.

Dr. Craig Somerton for being on my advisory committee, for his guidance, and for the use
of his temperature controllers.

Tom Stuecken for his assistance in the design and construction of the setup, and for
teaching me how to "hog" some metal.

Dan McCarrick and Dr. Keunchul Lee for listening when I wanted to run ideas by them,
and for giving their input.

Dr. John Funkhouser for his assistance, and for use of the Chemistry Department facilities
for sample preparation.

The Glass Blowing shop in the Chemistry Department for assistance during sample
preparation.

Ron Haas for loaning me the halogen leak detector during the repair of the excimer laser.

My parents, Stanley and Rose, and family for understanding why I missed so many family
occasions.

I would also like to thank the NASA Lewis Research Center, Cleveland, Ohio and the
Environmental Protection Agency, Ann Arbor, Michigan for supporting this work.

iii

TABLE OF CONTENTS

page
LIST OF TABLES ........................................................................................... vi
LIST OF FIGURES ........................................................................................... vii
CHAPTER 1 - INTRODUCTION .................................................................... l
1.1 Introduction to Exciplex ........................................................ 1
1.2 An Example of Fluorescent Images of a Fuel Distribution ........ 3
CHAPTER 2 - THE EXCIPLEX MECHANISM ............................................ 6
2.1 System Composition ....................................................... 6
2.2 Separation of Phases ....................................................... 6
CHAPTER 3 - EXPERIMENTAL EQUIPMENT AND CALIBRATION
PROCEDURE ................................................................... 9
3.1 Major Equipment ................................................................... 9
3.1.1 UV - Visible Spectrophotometer .................... 9
3.1.2 Temperature-Controlled Sample Heating System ...... 9
3.1.3 Excimer Laser and Energy Meter .................... 11
3.1.4 Intensified CCD Detector ................................ 12
3.1.5 Isolated Droplet Generator ................................ 13
3.2 Procedure ............................................................................... 15
3.2.1 Vapor Phase ........................................................ 15

iv

3.2.1.1 Sample Preparation ................................ 16

3.2.1.2 Determination of Vapor Concentration Versus
Temperature ............................................ 16

3.2.1.3 Determination of Fluorescence Intensity Versus

Vapor Concentration ................................ 18

3.2.2 Liquid Phase ........................................................ 23

CHAPTER 4 - RESULTS AND DISCUSSION ............................................ 27
4.1 Vapor Phase ................................................................... 27

4.2 Liquid Phase ................................................................... 31

CHAPTER 5 - SUMMARY AND CONCLUSIONS ............................................ 35
CHAPTER 6 - RECOMIVIENDATIONS ........................................................ 37
LIST OF REFERENCES ............................................................................... 39
APPENDIX ...................................................................................................... 41

Al Equipment Connections for Laser Induced Fluorescence
Imaging .............................................................................. 41

A2 Operational Information for the Laser Induced Fluorescence
Imaging System ................................................................... 43

LIST OF TABLES
page
TABLE 1: Spectrophotometer settings used for making absorbance measurements. 18
TABLE 2: System components and parameter values for vapor phase calibration. 20
TABLE 3: Average intensity gain versus potentiometer setting. ............................ 43

TABLE 4: Gate pulse width versus potentiometer setting. ...................................... 44

vi

LIST OF FIGURES

FIGURE 1: Schematic of engine in which fluorescent images were taken. ...... pgge
FIGURE 2: Fluorescent images of the fuel distribution in an engine cylinder. ...... 4
FIGURE 3: The mechanism of Exciplex liquid/vapor visualization systems. ...... 7
FIGURE 4: Fluorescence emission spectra of the liquid and vapor phases. (3) ...... 8
FIGURE 5: Schematic of temperature-controlled sample heating system. ................. 10
FIGURE 6: Schematic of the droplet formation module of the isolated droplet

generator. ........................................................................................ 14
FIGURE 7: Schematic of setup for determining vapor concentration. ................. 17

FIGURE 8: Schematic of setup for determining fluorescence intensity of vapor phase. 19
FIGURE 9: Setup for determining vapor concentration. ...................................... 21
FIGURE 10: Temperature-controlled sample heating system with insulation removed. 22
FIGURE 11: Setup for determining fluorescence intensity of vapor phase. ...... 22

FIGURE 12: Schematic of setup for determining fluorescence intensity of liquid

phase. ....................................................................................... 23
FIGURE 13: Setup for determining fluorescence intensity of liquid phase. ...... 25
FIGURE 14: Droplet formation module of the isolated droplet generator. ...... 25
FIGURE 15: Fluorescent image of isolated droplets. ................................................ 26
FIGURE 16: A zoom of some individual droplets from Figure 15. ........................... 26

FIGURE 17: TMPD vapor concentration versus temperature for first step of vapor
phase calibration. .............................................................................. 28

vii

FIGURE 18: Normalized fluorescence intensity versus TMPD vapor concentration. 30
FIGURE 19: Normalized fluorescence intensity versus droplet mass. ................ 32

FIGURE 20: Normalized fluorescence intensity versus liquid mass using multiple
droplet summation. ............................................................................. 34

FIGURE 21: Connections for laser induced fluorescence equipment. ................ 41

FIGURE 22: Timing diagram for [CCD detector system configuration. ................ 42

viii

CHAPTER I

INTRODUCTION

Controlling the fuel distribution in the combustion chambers of internal combustion
engines is important for decreasing exhaust emissions and increasing efficiency. Several
methods of determining the spatial distribution of fuel during the engine cycle have been
employed in the past. High speed photography works for the liquid phase of a spray, but
is insensitive to the vapor phase. Early laser-induced fluorescence methods could not
distinguish between liquid and vapor.

A new laser-induced fluorescence method for determining fuel distribution has
been under development in the last decade. Exciplex-based liquid/vapor visualization
systems allow two-dimensional fluorescent images of the liquid and vapor phases to be
taken separately, using a diode array detector. Calibration of fluorescence intensity then
permits the determination of concentration maps of liquid and vapor finel. Developing

such a quantitative calibration was the focus of this work.

1.1 Introduction to Exciplex

Melton (1) developed the concept of spectrally separated fluorescence emission
from the liquid and vapor phases. His first application was to a spray from a hollow cone
injector (2). Bardsley et al. (3,4) were the first to use the exciplex technique in an internal
combustion engine. These earlier applications were strictly qualitative in nature.

More recent efforts have focused on developing quantitative calibrations for the
exciplex method relating fluorescence intensity to mass concentration of fuel. Such a
quantitative interpretation of the fluorescent images is possible because fluorescence
intensity is directly proportional to the mass concentration of the fluorescing compound.

Melton (5) developed calibration procedures based on the calculation of light absorption,

1

quantum yields, and other photophysical parameters. However, those procedures required
that each parameter be correctly determined, and therefore the overall accuracy depended
on the accuracy of a series of separate procedures. Rotunno et a1. (6) developed direct
calibration procedures for the liquid and vapor phases. These procedures involved direct
measurement of the fluorescence intensity of a known amount of liquid or vapor. The
direct procedures are simpler and believed to be more accurate by the developers, but they
acknowledge that the photophysics-based procedures are more useful for experimental
design of exciplexes. The calibration efforts stated above were for an exciplex system that
is coevaporative with decane-based fuels, such as diesel fiiel, whose normal boiling points
are in the 200 to 300 degree Celsius range.

Current efforts are focusing on the development of exciplex systems that are
coevaporative with automotive gasolines. Melton (7) is developing two systems that are
expected to be coevaporative with fiiels whose boiling points are in the 70 to 110 degree
Celsius range. Shimizu et al. (8) developed a system meant for gasoline also. The normal
boiling point of this system is 193 degrees Celsius, significantly higher than the boiling
point of gasoline and therefore not an accurate marker. Use of such a system will result in
underestimation of the true vapor concentration, since the file] will evaporate more quickly
than the fluorescent marker. However, Shimizu et al. took a unique approach in
developing their calibration. Instead of presenting their vapor phase calibration as a
spatial map of concentrations, they presented it in the form of a spatial map of equivalence
ratios. They did not calibrate the liquid phase, but did use its fluorescence to show
whether or not liquid fuel was present.

Murray and Melton (9) developed another use for exciplex laser-induced
fluorescence. They devised a two-dimensional method for visually determining droplet
temperatures in a fuel spray using exciplexes.

The focus of this project was to develop a quantitative calibration for an exciplex

liquid/vapor visualization system. Since this was the first work with the exciplex method

at the Michigan State University Engine Research Laboratory, the naphthalene/N,N,N',N'-
tetramethyl-p-phenylenediamine (TMPD) system was used, as this has been the most
successfiil of those under development. The direct calibration procedures of Rotunno et

al. (6) were adapted for this development work.

1.2 An Example of Fluorescent Images of a Fuel Distribution

To give an example of these fluorescent images and demonstrate how such a
calibration is used, several images were taken in a research engine fitted with a quartz
cylinder under motored conditions. The laser beam was focused into a horizontal sheet

and passed through the quartz cylinder as shown in Figure 1. The images were taken

fiber optic spark plug hole
image carrier

   

2 CM:
UV laser sheet

 

/ \— quartz cylinder
V W

FIGURE 1: Schematic of engine in which fluorescent images were taken.

 

 

perpendicular to the light sheet through a fiber optic image carrier placed in the spark plug
hole and are shown in Figure 2. These images are for an example only; they are not

exciplex. The fluorescence is due to both liquid and vapor combined.

90°
Intake

1 80°
BDC

270°
Comp.

      

..
— m at

r

nu 16% RN am 349 g

D

FIGURE 2: Fluorescent images of the fuel distribution in an engine cylinder.

The detector only sees light intensity, not wavelength. The colors were assigned
by a look-up table based on the measured intensity level. The variation in intensity across
the bore shows that the mixture is not homogeneous, even halfway through the
compression stroke. The calibration would then be used to convert the intensity levels
into mass concentrations of fuel.

The non-intrusive exciplex method could be coupled with other laser diagnostic
methods such as high speed flow visualization and laser doppler velocimetry to provide
information on fiiel motion as well as fiiel distribution. Such a combination of methods

would help explain the causes of a resultant fuel distribution pattern.

CHAPTER 2

THE EXCIPLEX MECHANISM

2.1 System Composition

In order for an exciplex system to. be an effective marker of the liquid and vapor
phases, the components must be virtually coevaporative in the boiling range of the solvent.
The naphthalene/TMPD system is coevaporative with decane based fuels, such as diesel
fiiel, whose normal boiling points are in the 200 to 300 degree Celsius range. A system
consisting of 10% naphthalene, 1% TMPD, and 89% decane by weight was used for the
development of this calibration.

Two additional exciplex systems have been developed by Melton (7) which are
expected to be coevaporative with solvents boiling in the 70 to 110 degree Celsius range.
These systems are expected to track the vaporization of automotive gasoline effectively.
The first system consists of 10% triethylamine, 0.5% fluorobenzene and 89.5% hexane,
and is expected to be coevaporative with a normal boiling point of 69 degrees. The
second system consists of 10% n-propyldiethylamine, 0.5% 4-fluorotoluene and 89.5%
isooctane, and is expected to be coevaporative with a normal boiling point of

approximately 100 degrees.

2.2 Separation of Phases

The basic chemical mechanism is the same for all exciplex-based systems, but the
specific chemical components are different. They all contain three basic components: a
monomer, a ground state reactant, and a solvent which serves as the fisel. In the system
used here, TMPD is the monomer (M), naphthalene is the ground state reactant (G), and

decane is the solvent. The exciplex is formed in the reversible equilibrium

M“ + G <=> E“ (2-1)
where M“ is the first excited singlet state of the monomer, and E“ is the exciplex, which is
short for excited state complex (6). The equilibrium of equation (2-1), which governs the
formation of the exciplex in the liquid but not the vapor, is controlled by adjusting the
concentration of the ground state reactant. The wavelength of the excitation source
determines which excited electronic state is reached during the absorption transition (10).
The monomer TMPD requires excitation at approximately 350 nm.

Figure 3 illustrates the individual steps of the exciplex mechanism. In the liquid

 

Concept
6) , @ Laser excitation of monomer in. liquid or vapor

® Exciplex (excited complex) formation, liquid phase only
@ Exciplex fluorescence occurs in green
(5) Monomer fluorescence occurs in blue

FIGURE 3: The mechanism of Exciplex liquid/vapor visualization systems.

phase, an excited TMPD molecule combines with a naphthalene molecule to form an
exciplex. The exciplex then fluoresces as the molecules return to the ground state. In the

vapor phase, the exciplex does not form. Instead, the excited TMPD molecule itself

fluoresces as it returns to the ground state. Since the formation of the exciplex results in it
having a lower energy level than the excited monomer, its emission is also at lower energy.
The energy of an emitted photon is given by

E = hc/lt (2-2)
where h is Planck's constant, c is the speed of light, and it is the wavelength. Therefore,
the exciplex emission occurs at longer wavelengths than that of the monomer. Figure 4
shows the fluorescence emission spectra of the two phases. Using appropriate bandpass

filters then allows fluorescent images of the liquid and vapor phases to be taken separately.

 

 

 

 

E GAS LIGJID
Z
LIJ
'—
Z
LU
0
2
DJ
0
(0
DJ
M
O
D
.J
U.
300 400 500 600

WAVELENGTH (nm)

FIGURE 4: Fluorescence emission spectra of the liquid and vapor phases. (3)

CHAPTER 3

EXPERIMENTAL EQUIPMENT AND CALIBRATION PROCEDURE

3.1 Major Equipment
Descriptions of the experimental equipment and their features will be given

separately to avoid interrupting the procedure with these details.

3.1.1 UV-Visible Absorption Spectrophotometer

To determine vapor concentrations, absorbance measurements were made with a
Varian model Cary 1E UV-Visible absorption spectrophotometer. A spectrophotometer,
which is generally used in analytical chemistry, can make absorbance measurements at a
specific wavelength or produce a spectrum, i.e. absorbance over a range of wavelengths.
This project only required measurements to be made at a single wavelength. The Cary IE
is a dual beam instrument. The collimated beam from the light source is chopped and
directed alternately through the sample and a reference. Absorbance is then determined by
comparison of the two beams' intensities with a photo multiplier tube. Adjustable slits in
the monochromator section control the spectral bandwidth, while gratings control the
wavelength. The spectrophotometer is controlled by a Dell 325D computer (with math
c0processor and GPIB board) through an IEEE488 interface using Varian drivers and

sofiware.

3.1.2 Temperature-Controlled Sample Heating System
A temperature-controlled sample heating system was designed and constructed to
vary the vapor concentration of a sample during the vapor phase calibration. A schematic

of the system is shown in Figure 5.

 

 

 

 

 

 

 

 

 

 

 

 

 

lightltrap
f 1 I W
F
temperature heater
controller element ‘sample
chamber
l:l°C [I
O O
l

 

 

 

 

 

 

l l
flexible ball
shaft coupler bearings

FIGURE 5: Schematic of temperature-controlled sample heating system.

Air is circulated through the sample chamber in a closed loop of 2.0 inch copper
pipe. Since the system is capable of reaching temperatures of several hundred degrees
Celsius, the fan is driven by an externally mounted motor. A 1/15 horsepower Dayton
motor and variable speed control were used. A flexible shafi coupler was used to connect
the motor to the fan blade shaft to prevent heat transfer from the fan blade shaft into the
motor. The fan blade shaft is supported by ball bearings at two points. The standard
bearing grease was removed and the bearings were repacked with Dow Corning 41
Extreme High Temperature Bearing Grease. The maximum operating temperature of the
heating system is limited to 288 degrees Celsius (550 degrees F) by this silicone lubricant.

A Master Appliance 260 degree Celsius (500 degree F) heat gun element was used
to heat the air, though higher temperature elements are available. The element was

insulated thermally and electrically from the copper pipe with mica insulation. The inside

ll

of the heating loop was painted flat black and light traps were constructed to prevent
visible radiation from the heating element from reflecting into the sample chamber and
causing error in the absorption or fluorescence intensity measurements. Air temperature is
controlled with an Omega model 4001-KC proportional controller with K-type
thermocouple input and relay output. The heating element is controlled by the relay
output via an external contactor due to its power requirement. The controller input
thermocouple is located in the air stream just above the sample holder. A second
thermocouple is placed between the sample holder block and sample cuvet to measure the
sample temperature. The mass of the holder and cuvet keeps the sample temperature
nearly constant even though the air temperature fluctuates 8 to 10 degrees per controller
cycle.

The sample chamber has optical access windows mounted on three sides for
making absorption and fluorescence measurements. The windows are T20 Suprasil clear
firsed quartz discs from Heraeus Amersil. The sample holder is mounted in the sample

chamber such that air can pass over all sides of the sample cuvet.

3.1.3 Excimer Laser and Laser Energy Meter

Ultraviolet excitation for the fluorescence was provided by a Lambda Physik
model EMG160TMSC tunable excimer laser. This is a pulsed laser which can operate at
wavelengths of 193, 248, 308, and 351 nm by changing the gas mixture and optics.
Tuning mode produces a beam with a narrower bandwidth than power mode. This model
contains two laser heads, an oscillator head and an amplifier head. In tuning mode, the
oscillator and tuning optics produce a narrow bandwidth beam which is injected into the
amplifier head for power amplification. The beam is also vertically polarized.

The laser was operated at 351 nm using a gas mixture of xenon, fluorine, and
helium. At 351 nm, it produces an 18 nanosecond FWHM pulse with a rated maximum

pulse energy of 70 m]. In actual use, the pulse energy was initially higher but then

12

dropped below this value with age and use of a gas fill. According to a Lambda Physik
service representative, a gas fill should be good for approximately 1,000,000 shots or 2
weeks.

A Scientech Vector series laser power/energy meter system was used to measure
the pulse energy during the calibration. The system consists of the models 8200 single
channel power and energy indicator with digital readout, P25 25 mm pyroelectric detector,
and VSZSUV quartz beamsplitter with a 3.35/82.5 R/T ratio. The beamsplitter must be
used because the detector's coating cannot handle the full 38 MW/sq. cm peak power
density of the excimer laser. The beamsplitter reflects only 3.35 percent of the beam into
the detector and transmits 82.5 percent. The R/T ratio can be entered into the indicator so
that it will display the energy or power of the transmitted beam. The beamsplitter was
calibrated with S polarized light and therefore must be positioned such that the reflected
beam is directed vertically when used with the excimer laser. The indicator displays
energy, average energy or power. It also collects and calculates statistics on a user-

defined sample size of pulses.

3.1.4 Intensified CCD Detector

Fluorescent images were taken with a Princeton Instruments intensified CCD
(Charge Coupled Device) detector system and stored on the Dell computer running
Princeton Instruments CSMA (CCD Spectrometric Multichannel Analyzer) imaging and
spectroscopy software. A model ICCD-576 G/RB detector was controlled by a model
ST-130 controller. A model F G-100 pulse generator provided external high voltage gate
pulses to the intensifier. The intensifier acts as the shutter and exposes the 576 x 384
pixel array when it receives a gate pulse. The time between array readouts is specified by
the exposure time set in the software. The intensifier is controlled separately from

readout, thereby permitting multiple array exposures per readout in External Sync timing

l3

mode. This results in on-chip averaging which improves signal-to-noise over in-memory
averaging.

The detector only sees light intensity, not wavelength. The images are assigned
colors based on intensity values by a look-up table in the software when displayed on the
monitor. Several look-up tables are included in the software, or the user can create
customized tables. User defined look-up tables can be used to change the resolution of an
image and to specify preferred shades of colors. The CSMA software also has image
processing capabilities such as adding, subtracting or rotating images, and defining a

rectangular region of interest on which to do statistics or zoom in.

3.1.5 Isolated Droplet Generator

Liquid droplets of a known size were required for the liquid phase calibration. An
isolated droplet generator designed and constructed in the Department of Chemistry at
Michigan State University was used for this purpose (1 l,12). The droplet generator is
based on the principles of induced breakup of a liquid jet discovered by Rayleigh. Its
electronic control module can be operated in a stand-alone mode or, for greater versatility,
in a computer-controlled mode.

Figure 6 shows the basic components of the droplet formation module. Liquid is
forced through a glass capillary to form a liquid jet. A rectangular piezoelectric crystal
called a bimorph is attached to the capillary. When an oscillating voltage is applied to the
bimorph, it deflects, causing the capillary to vibrate at the oscillation frequency. These
vibrations are transmitted through the capillary to the surface of the liquid jet. Normally,
the jet would travel in a cylindrical mass for some distance before it breaks up under the
action of surface tension into a series of randomly sized drops. However, if the capillary is
vibrated and the wavelength of the vibrations is greater than the circumference of the jet,
the break up occurs in a uniform manner and produces droplets of uniform size and

spacing.

l4

bimorph
driver signal

(Pl—U0

/ high voltage
O

 

 

 

 

 

 

 

 

 

bimorph deflection
charging plates
electrode 0 O
z r 0000000 000 00
teflon capillary
tubing ___

 

 

FIGURE 6: Schematic of the droplet formation module of the isolated droplet generator.

This droplet generator can theoretically produce droplets at rates of 1000 to
50,000 per second (13). Because the production rates are greater than desired for some
experiments, this droplet generator has the ability to charge and deflect droplets out of the
main stream so that they can be trapped. Selected droplets are charged as they pass
through a cylindrical electrode as shown in Figure 6. They are then deflected as they pass
between the high voltage plates and trapped, thereby decreasing the delivery rate in the
main stream. The ratio of charged to uncharged droplets can be varied.

Droplet size and production rate are controlled by capillary size, bulk flow rate
through the capillary, and applied bimorph fiequency. Capillaries range in size from 10 to
65 microns inside diameter. Liquid is supplied by a constant pressure delivery vessel.
Regulated pressure fiom a nitrogen gas cylinder is used to control the flow rate. Flow

rate versus pressure has to be calibrated for each capillary. The liquid delivery system

15

contains a 2 micron in-line filter to keep the capillaries from getting clogged. Droplet
volume can be calculated by
Vd = F / f (3-1)

where F is the bulk flow rate and f is the production frequency. By substituting the
formula for the volume of a sphere, droplet radius can be expressed as

rd = (3F / 41:01/3 (3-2).
The overall error in the determination of the dr0plet radius by this method is reported to
be on the order of 0.33 percent (11).

For this work, the isolated droplet generator was operated using 15 to 45 micron
diameter capillaries, 20 to 60 psi liquid delivery pressures, and 7000 to 28,000 Hz applied
bimorph frequencies. At production rates up to 28,000 Hz, droplet charging and
deflection was not required as the droplets were visibly separated in the fluorescent

images.

3.2 Procedure
The following explains the procedure used to calibrate an exciplex system. A

separate calibration is required for each of the two phases.

3.2.1 Vapor Phase

The vapor phase calibration was carried out in two steps using a sample of TMPD,
the vapor phase marker. The first step was to determine the vapor concentration versus
temperature for the sample. The second step was to determine fluorescence intensity
versus vapor concentration using the concentration/temperature relationship from the first

step to establish known concentrations.

16

3.2.1.1 Sample Preparation

For the vapor phase calibration, a small sample of TMPD was prepared using a
standard quartz fluorimeter cell with a stem and 10 mm light path length fiom NSG
Precision Cells. Before placing the sample in the cuvet, an absorbance measurement was
made on the empty cuvet with the spectrophotometer. This is the blank signal and was
used to subtract the effect of the cuvet walls from the subsequent sample measurements.

TMPD was dissolved in spectrophotometric grade hexane to make a solution with
a concentration of approximately 1.0 gram TMPD/liter. Approximately 100 microliters of
solution was then transferred to the cuvet using a microliter syringe fitted with a 10 inch
hypodermic needle so that the sample could be placed on the bottom of the cuvet, keeping
the walls optically clear. The hexane was evaporated at room temperature. Because of
the narrow opening in the cuvet stem, the hexane evaporated extremely slowly. The
process was greatly speeded by attaching the hypodermic needle to a supply line and
blowing dry nitrogen gas across the surface of the solution. The sample cuvet was then
attached to a vacuum line, fiozen with liquid nitrogen, evacuated, and sealed under
vacuum.

Two samples were prepared. Sample 1 contained 100 microliters of solution and
was evacuated to 10'3 torr. Sample 2 contained 125 microliters of solution and was
evacuated to 10‘4 torr. Sample preparation was done using the facilities of the

Department of Chemistry.

3.2.1.2 Determination of Vapor Concentration Versus Temperature

The first step was to determine the vapor concentration versus temperature from
absorbance measurements with the spectrOphotometer. Increasing the temperature of the
sample increases the vapor concentration by increasing the saturation vapor pressure. The
temperature-controlled sample heating system was mounted in the sample compartment of

the spectrophotometer for this purpose as shown in Figure 7. Measurements were made

l7

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

lighlt trap
F II
f
l J
temperature heater
controller element sample
‘ I] chamber .2
U0 L sample
0 o ”EH-m °§mpmem
+ II IL . __________________________________________
[__— T spectrophotometer
flexible ball
shaft coupler bearings

FIGURE 7: Schematic of setup for determining vapor concentration.

at 20 degree Celsius intervals from room temperature to 180 degrees.

The spectrophotometer was allowed to warm up approximately one-half hour and
zeroed with the sample chamber windows of the heating system in the sample beam. This
removes the effect of the windows from the measurements. The reference beam was
empty, i.e. air was the reference. The sample was then placed in the sample chamber of
the heater and absorbance measurements taken after the temperature equilibrated at each
interval. The spectrophotometer settings used are in Table 1. After subtracting the blank
signal, concentration was calculated using Beer's Law

A = e b c (3-1)
where A is the absorbance, e is the molar absorptivity (1.74 x 104 L/mol- cm at 266 nm),

b is the path length, and c is the unknown concentration.

18

TABLE 1: Spectrophotometer settings used for making absorbance measurements.

 

 

Parameter Settigg
photometric mode absorbance
abscissa mode sample number
wavelength 266 nm
spectral bandwidth 1.0 nm FWI-IM
source lamp UV

sample averaging time 1.0 sec

3.2.1.3 Determination of Fluorescence Intensity Versus Vapor Concentration

The second step in the vapor phase calibration was to determine fluorescence
intensity versus vapor concentration using the vacuum sealed TMPD sample. Again,
temperature was used to vary the vapor concentration of TMPD, except that the
concentration at each temperature was known from the first step. The setup for this step
utilizes the sample heating system, excimer laser, laser energy meter, light sheet forming
optics, and intensified CCD detector as shown in Figure 8.

The excimer laser was used as the excitation source, operating at a wavelength of
351 nm. The laser beam was passed through a custom-made variable beamsplitter with
removable quartz discs for adjusting the pulse energy to the desired level, and a field stop
to cut off the fringes. It then passed through the beamsplitter for the energy meter, a
cylindrical focusing lens with a 300 mm focal length, a cylindrical recollimating lens with a
-300 mm focal length to form a 550 nm wide vertical sheet, and the sample in the heating
system.

Images of the fluorescence were taken perpendicular to the laser sheet at
approximately 20 degree Celsius increments with the intensified CCD detector. The
system was synchronized by externally triggering the CCD detector from the excimer

laser. Several images were taken at each concentration level using 10 exposures per

l9

 

sample
CFC heating system bandpass filter
0 O

 

 

 

 

 

 

 

 

 

 

i
ICCD detector
EU I:

\ recollimating lens

 

é focusing lens

___—J
energy meter detector ___—- [3:3

 

—— field stop

 

 

 

variable beamsplitter

 

excimer laser

FIGURE 8: Schematic of setup for determining fluorescence intensity of vapor phase.

readout with background subtraction, and stored on the computer's hard disk. The
average pulse energy of the 10 pulses most nearly associated with each image was
recorded for image processing later. The system components and user-defined parameter
values for this step of the calibration are given in Table 2. The detector hardware
parameters are adjusted by ten-tum potentiometers but are not linear with the turn units
on the dial face. Calibrations for these potentiometers along with other operational

information to assist the user are contained in the Appendix.

20

 

TABLE 2: System components and parameter values for vapor phase calibration.

Component Parameter Value
Laser rep. rate 5 Hz
charging voltage 22 - 24 kV
Detector optics extension tube 18 mm
zoom 50 mm
aperture f/16
apparent focal distance 4.0 it
bandpass filter 400nm; 40nm FWHM
Detector hardware gain variable
delay 640
gate pulse width 250 ns

Detector software

timing mode

exposure time

image size

background subtraction

ext. sync. normal
2.0 sec

bin by 2

on

Fluorescence intensity data was extracted from the images using the CSMA

software. A rectangular region of interest containing the cuvet image (except the edges)
was defined, and the average intensity per pixel was calculated. The region of interest
remained the same for all images. The average intensities were normalized by the pulse
energy, corrected for detector gain, and corrected for the attenuation effects of the quartz
windows and cuvet walls. Finally, the intensities were plotted resulting in a calibration
curve of fluorescence intensity versus mass concentration of TMPD vapor. The
concentration of fuel vapor is then inferred based on the mixture ratio of TMPD to fuel.
Figure 9 shows the temperature-controlled sample heating system mounted in the
sample compartment of the spectrophotometer for the first step of the vapor phase

calibration. Figure 10 shows the quartz windows in the sample chamber of the heating

21

system with the insulation removed. Figure 11 shows the heating system, CCD detector,

and sheet forming optics setup for the second step of the vapor phase calibration.

 

FIGURE 9: Setup for determining vapor concentration.

22

 

FIGURE 11: Setup for determining fluorescence intensity of vapor phase.

23

3.2.2 Liquid Phase

The liquid phase calibration was carried out in a single step by taking fluorescent
images of known sized droplets of exciplex-doped fiiel with the CCD detector. The total
fluorescence intensity of the droplets was then calculated from the images and plotted
versus droplet mass. A schematic of the laser, droplet generator, and detector

configuration is shown in Figure 12.

isolated droplet generator

\ bandpass filter

 

 

©l::l®

.. l n I:

ICCD detector

 

 

 

 

 

 

 

 

 

 

 

 

 

\ recollimating lens
11:: focusing lens
energy meter detector —— Db.
——; — -—field stop
E:— variable beamsplitter
a s

 

excimer laser

FIGURE 12: Schematic of setup for determining fluorescence intensity of liquid phase.

24

Again, the excimer laser provided excitation for the fluorescence and images were
taken with the CCD detector. The sheet forming optics and energy meter configuration
were the same as in the vapor phase, but the bandpass filter in fi'ont of the detector was
now centered at 500 nm with a 40 nm FWHM bandwidth. The stream of droplets
produced by the isolated droplet generator was directed downward through the vertical
laser light sheet. Droplets were generated in the 75 to 200 micron diameter range by
varying the flow rate through 20, 30 and 45 micron capillaries, and by varying the droplet
production frequency from 7000 to 28,000 drops per second.

Fluorescence intensity data was extracted from the images by zooming in on
individual droplets and calculating the total intensity by summing the intensity values of
the pixels containing the droplet. Since this calibration could be done in the open
atmosphere and no quartz windows or cuvets were in the optical path, the intensities only
needed to be corrected for detector gain. The intensities were again normalized by the
pulse energy and the result was a calibration curve of fluorescence intensity versus mass of
liquid fuel.

The isolated droplet generator, sheet forming optics, and CCD detector setup for
the liquid phase calibration are shown in Figure 13. Figure 14 is the droplet forming
module showing the glass capillary and bimorph. Figure 15 is a fluorescent image of
droplets taken during the liquid phase calibration at a droplet production rate of 28,000
per second. Figure 16 is a zoom of some droplets from Figure 15 using the image

processing software. Each square is a pixel in the detector's array.

25

 

.. ‘- v'. -‘

FIGURE 131 Setup for determining fluorescence intensity of liquid phase.

 

FIGURE 14: Droplet formation module of the isolated droplet generator.

26

 

FIGURE 16: A zoom of some individual droplets from Figure 15.

CHAPTER 4

RESULTS AND DISCUSSION

4.1 Vapor Phase

The TMPD vapor concentration versus temperature for the first step of the vapor
phase calibration is shown in Figure 17. The upper plateau is the result of the entire
sample being vaporized. The absorbance measurements were corrected for the effect of
the cuvet walls using the blank signal before being converted to concentration. As shown
by the four runs, the measurements were quite repeatable.

Sample 1 was used primarily for testing and adjustment of the equipment, while
the data being presented here were taken with sample 2. The maximum concentration of
12.26 micrograms/cc from Figure 17 times the volume of the cuvet results in a total
sample size of 84.6 micrograms. The amount of TMPD placed in the cuvet during sample
preparation was approximately 125 micrograms (125 microliters of 1.0 gram TMPD/liter
solution). The difference is believed to be due to sublimation of TMPD during evacuation
of the cuvet. Therefore, it is important that the sample only be evacuated long enough to
remove the air fi'om the cuvet.

After the TMPD sample has been heated and driven into the vapor phase, it
recrystalizes on the surface of the cuvet upon cooling. It is desired that it do this on the
bottom and not the walls where it would interfere with optical access. Therefore, the
bottom of the cuvet should be cooled first. This feature happened to be inherent in the
design of the heating system. The bracket from the external walls of the sample chamber
attaches to the bottom of the cuvet holder. Thus, when the temperature is reduced, heat is

conducted to the external walls from the bottom of the cuvet first.

27

28

 

 

v.II6III
m.II6III

N.IIAYII

 

 

oer

decent—mo 893 .593 me new Em .8 0.338an8 mama? 5:828:00 .093 DE U C mMDOE

ADV 9:552.th

 

 

cow ovw om? cor on 00 ov ON 0
p P b - p b p n p n . P n p - p p o
— - J u u - 1 W
r
..N
m
a m
1: V w
I a
u
.. n
. w
tom.
.. u
. m
11 w 3
1 X
I
.- 0
V
.. 9
it cw
fl
#9
H

 

9.5.299th .m> 55.22850 .59; 952,—.

29

The normalized fluorescence intensity versus vapor concentration data, as shown
in Figure 18, is nearly linear, as expected, since fluorescence intensity is directly
proportional to concentration. The calibration constant resulting from a linear regression
curve fit is 8.19 counts/(nicrogram/cm3). The scatter at the higher concentrations may be
due to temperature effects and should be investigated fiirther. The fluorescence intensities
were corrected for detector gain and for the attenuation of the window and cuvet wall
using the transmittance of these components. The transmittances were measured with the
spectrophotometer at 380 nm, the peak of the vapor phase emission. The transmittances
of the window and cuvet wall were also measured at 351 nm and used to correct for
attenuation of the incident laser pulse energy.

The fluorescence intensities were normalized by the incident laser pulse energy.
Incident laser pulse energy determines how many molecules are put in the excited state,
and the number of excited molecules determines fluorescence intensity. In other words,
fluorescence intensity is proportional to pulse energy as well as concentration, but
normalized fluorescence intensity is proportional to concentration.

The excimer laser emits a 5 mm wide by 20 mm high rectangular beam with a 2
mm diameter dead spot in the center. The dead spot is caused by a spot of reflective
coating in the center of the laser's output coupler. When the beam was focused in the
horizontal direction to form a vertical sheet, the intensity profile was not uniform. This
could be seen as a dark band passing horizontally through the center of a fluorescent
image taken perpendicular to the sheet. Since this would affect the fluorescence intensity
measurements, only the portion of the image above the dark band was used for extracting
data. Other ways of dealing with the dead spot would be to use only the portion of the
beam between the spot and side of the beam or between the spot and top of the beam, and
use optics to form the desired size sheet.

Melton et al. observed the development of a brown discoloration on the inside of

the cuvet walls where the laser enters and exits after continual irradiation. In this work,

30

dots—88:8 83> mg «35> E885 858863 Bum—«E62 HM: $5...va

33: 983 __ezabeooeev
VF N— or w o v N o

T
-
1..
-_
..
_-
r

 

 

 

 

 

 

5:55:8an .893 9:5. .m> mam—.8:— ooaoomoueim “5&3...an

ON

0%

cm

on

§

ON?

Kirsuaiul oouoosaaonu pazuutuao N

31

the walls of sample 1 gradually became clouded around the entrance and exit points of the
laser, but sample 2 did not. Also, the TMPD in sample 1 turned brown, but that in sample
2 did not. The difference is believed to be due to the degree of evacuation. Sample 1 was
evacuated to 10'3 torr, while sample 2 was evacuated to 10"4 torr. Therefore, more
complete removal of air from the cuvet is important to reducing the photoreaction caused
by the laser.

Oxygen is an efficient quencher of TMPD fluorescence, but nitrogen is not (6). As
a result, fuel distribution studies in engines must be done with nitrogen under motored
conditions. Exciplex systems that are not quenched by oxygen need to be developed so

that studies can also be done in a firing engine.

4.2 Liquid Phase

A plot of the normalized fluorescence intensity versus droplet mass for exciplex-
doped decane is shown in Figure 19. The fluorescence intensities were again corrected for
the detector gain and normalized by the incident laser pulse energy.

Droplet sizes were not predetermined because not all frequency and flow rate
combinations of the droplet generator could produce stable droplets. As a result, droplet
sizes were completely random, but there were more than enough combinations that could
produce stable droplets. An attempt was made to produce smaller droplets using a 10
micron capillary. This capillary would produce a stable liquid jet with straight decane, but
the jet would exit in random directions with exciplex-doped decane. Apparently,
dissolving naphthalene in the decane made it too viscous.

The data in Figure 19 showed more scatter than expected. The scatter is believed
to be due to hot spots in the laser beam. The intensity profile of the beam was not
measured and quantified, but a profile of the beam using thermal paper showed dark spots
scattered throughout. (The darker the paper, the greater the intensity.)

The calibration of fluorescence intensity versus droplet mass shown in Figure 19 is

32

.mmmE 832w 38? 33:85 85880:: macaw—«E62 ”2 9505

 

GE: 3 mas—z 3E2:
wan n ad ad YN NN N m... 9. Yr Né _. ad 0.0 to N6 o
.1 “ u w w u u u u u w “ w u u n
. w... ..
I
I I I II’ F-I .
I I I - fi
- I I u. I. l
I I II
I I I
I ll
. . .
I I I 1...
I I 1.
I l'
I

2:89: comet
any—ESQ .5.“ was: 8E8: .m> >323:— 35898=E EEEESZ

 

oc _.
com
com
oov
8m
80
con
com
com

.§_

8:
oom—

Kusuoiul oouaosaaonu pazuaruroN

33

not linear. This nonlinearity is due to the larger droplets being optically thick, i.e. having
optical densities greater than 0.2 (6). An optical density of 0.2 corresponds to a droplet
diameter of approximately 80 microns (13). In an optically thick droplet, the laser
intensity decreases as it passes through the droplets, causing the total fluorescence
intensity to be decreased.

For optically thin droplets, those with optical densities up to 0.2, the calibration
curve should be linear. So, instead of using single droplets that were not optically thin at
the larger masses, a new method was developed. This method adds up the masses and
fluorescence intensities of several small droplets, which are individually optically thin, to
generate the calibration at the larger masses. The result was a linear calibration curve of
fluorescence intensity versus mass of liquid fuel as shown in Figure 20. The calibration
constant from a linear regression curve fit is 398.81 counts/microgram. Since the droplets
fi'om a high pressure diesel (a decane based firel) injector are primarily in the optically thin

regime, a calibration developed with optically thin droplets would be desired.

34

:23888 838: 29:38 mam: 3:8 2:3. 3E3 36:25 85880:: ecu—2:02 Hom 950E

61:5 as: 235
Nd m mN oN vN NN N w; w; v; N... P md ed «.0 Nd o

 

 

 

T u u a n i u l i u u l l i i u a
-- co.
toom N

0
:8». m
m.
:83. m.
D.
:8... u
m
4: com M
n
:2: m
D
a
In con m...
m.
:8: m
tooeh
-- cop:
.r oom:

55:55.5 3395
295:2 win: 3:2 2:34 .m> >223:— oosomoaeam moi—«8.82

CHAPTER 5

SUMMARY AND CONCLUSIONS

. Exciplex liquid/vapor visualization would be a useful tool for quantitatively studying
fuel distribution in internal combustion engines and combustors.

. Multiple droplet summation of optically thin droplets should be used to generate liquid
phase calibration curves for use in studies with high pressure injectors.

. The effects of optical access windows in a setup can be removed by measuring their
transmittance at the appropriate wavelength. This allows a single calibration to be
used for setups with different elements in the optical path.

. Since a calibration is setup-dependent, preliminary fluorescent images should be taken
in the setup before calibrating to determine the required ranges of the calibrations.

. The cuvet should be cooled from the bottom after taking data so that the TMPD
recrystallizes on the bottom and doesn't interfere with optical access by recrystallizing
on the walls.

. Fluorescence intensity is proportional to pulse energy and should be normalized.

. The dead spot in the center of the excimer laser's beam must be dealt with when
focusing the beam into a sheet.

. Significantly less clouding and discoloration of the cuvet walls was observed in a
sample evacuated to 10'4 torr when compared to one evacuated to 10'3 torr. Also,
the sample should be evacuated only long enough to remove the air, in order to
prevent significant loss of TMPD. Therefore, a high vacuum line should be used to

evacuate the sample for a short period of time.

35

36

9. The vapor phase fluorescence from the naphthalene/TMPD system is strongly
quenched by oxygen. Fuel distribution studies in engines using this system can only be
done with nitrogen under motored conditions.

10. The isolated droplet generator produced uniformly sized and spaced droplets of

known size and was very well suited for this application.

PW.“

.V'

CHAPTER 6

RECOMMENDATIONS

The effects of high temperatures on fluorescence intensity should be investigated.
The development of exciplex systems for other fuels needs to be continued.
Exciplex systems which will work in an oxygen environment need to be developed.
Methods should be investigated for making the laser beam intensity profile more
uniform.

The use of fiber optic image carriers for exciplex imaging should be studied.

The percent of TMPD evaporated versus the percent of fuel evaporated from a
droplet, at the temperature at which fuel distribution studies will be done, should be

investigated to determine how accurately the TMPD tracks the vapor.

37

LIST OF REFERENCES

10.

39

LIST OF REFERENCES

Melton, L. A., "Spectrally Separated Fluorescence Emissions for Diesel Fuel
Droplets and Vapor”, Applied Optics, Vol. 22, No. 14, pp. 2224-2226, 1983.

Melton, L. A. and Verdieck, J. F., "Vapor/Liquid Visualization in Fuel Sprays",
Twentieth Symposium (International) on Combustion, (The Combustion Institute,
1984), pp. 1283-1290.

Bardsley, M. E. A., Felton, P. G. and Bracco, F. V., "2-D Visualization of Liquid
and Vapor Fuel in an I. C. Engine", SAE Paper 880521, 1988.

Bardsley, M. E. A., Felton, P. G. and Bracco, F. V., "2-D Visualization of a Hollow-
Cone Spray in a Cup-in-Head, Ported, I. C. Engine", SAE Paper 890315, 1989.

Melton, L. A., "Quantitative Use of Exciplex-Based Vapor/Liquid Visualization
Systems ( A Users Manual)", Final Report, Army Research Office Contract
DAAL03-86-K-0082, 1988, (NTIS order # ADA191649).

Rotunno, A. A., Winter, M., Dobbs, G. M. and Melton, L. A., ”Direct Calibration
Procedures for Exciplex-Based Vapor/Liquid Visualization of Fuel Sprays",
Combust. Sci. and Tech., Vol. 71, pp. 247-261, 1990.

Melton, L. A., "Exciplex-Based Vapor/Liquid Visualization Systems Appropriate for
Automotive Gasolines", Applied Spectroscopy, Vol. 47, No. 6, pp. 782-786, 1993.

Shimizu, K, Matumoto, S., Furuno, S., Murayama, M. and Kojima, 8.,
"Measurement of Air-Fuel Mixture Distribution on a Gasoline Engine Using LIEF
Technique”, SAE Paper 922356, 1992.

Murray, A. M. and Melton, L. A., "Fluorescence Methods for Determination of
Temperature in Fuel Sprays", Applied Optics, Vol. 24, No. 17, pp. 2783-2787,
1985.

Ingle, J. D. and Crouch, S. R., Spectrochemical Analysis, Prentice-Hall, Englewood
Cliffs, New Jersey, 1988.

11.

12.

13.

14.

40

Weigand, P. M. and Crouch, S. R., "Evaluation of a Programmable, Hardware-
Driven Isolated Droplet Generator", Applied Spectroscopy, Vol. 42, No. 4, pp. 567-
571, 1988.

Wiegand, P. M., Ph.D. Dissertation, Michigan State University, East Lansing, 1985.
Archontaki, H. A. and Crouch, S. R., "Evaluation of an Isolated Droplet Sample
Introduction System for Laser-Induced Breakdown Spectroscopy", Applied
Spectroscopy, Vol. 42, No. 5, pp. 741-746, 1988.

Felton, P. G., Bracco, F. V. and Bardsley, M. E. A., "On the Quantitative
Application of Exciplex Fluorescence to Engine Sprays", SAE Paper 930870, 1993.

APPENDIX

41

APPENDIX

The Appendix contains information to assist the user in the operation of the

equipment in this setup and is meant as a supplement to the equipment manuals.

A.l Equipment Connections for Laser Induced Fluorescence Imaging

The connections for the ICCD detector system and excimer laser for the laser

induced fluorescence portion of the project are shown in Figure 21. The connections

 

, Dell Computer

 

 

J3

 

 

 

ST 130 Controller trigger out 1:13-100

, Pulse Generator
ext. sync. r enable in

notscan

 

 

output

 

 

 

 

 

 

Excimer
Laser

 

detector variable

ate ,
‘ ICCD
Detector

FIGURE 21: Connections for laser induced fluorescence equipment.

 

 

 

 

 

 

 

sync.
out

 

42

shown in Figure 21 are a combination of Case 1 and Case 2 in section 6.3.1, ICCD Gating
Experiments Examples, of the Princeton Instruments manual. The timing diagram for this

configuration is shown in Figure 22. The excimer laser triggers the pulse

Laser
Trigger

3
:1
P
j
j
:1

Laser
Emission

Gate
Pulse

_——————-p———————-——
—
—
_
—
_
_

 

__l.

 

 

 

— dela
settiiig l- l— -200 V
readout
ICCD -_I data aquisition [

 

Notscan __i ......................................................................................... 1

FIGURE 22: Timing diagram for ICCD detector system configuration.

generator which sends a high voltage gate pulse to the intensifier in the ICCD detector.
The delay between the laser trigger and gate pulse is adjustable so that the detector can be
gated after the laser emission occurs. The 'enable in' to 'not scan' connection prevents the
array from being exposed during readout by blocking the gate pulses. The 'trigger out' to
'ext. sync.‘ connection initiates readout when the system is being externally triggered and
the controller is being operated in the external synchronization mode. The time during
which data is acquired is determined by the 'exposure time' set in the software. The
number of exposures accumulated by the detector array per readout is determined by the
exposure time and the laser repetition rate. For one exposure per readout, the exposure

time can be set to zero or the laser repetition rate can be adjusted accordingly.

43

A.2 Operational Information for the Laser Induced Fluorescence Imaging System
The gain of the ICCD detector and the pulse width of the gate pulse generator are
controlled by ten-tum potentiometers, but are not linear with the potentiometer setting.
Experiments were done to determine the gain and pulse width versus potentiometer
settings.
I The ICCD detector gain is variable from 1 to 100 times. Table 3 summarizes the
average intensity gain versus potentiometer setting for the detector. Potentiometer

settings are read from 0 to 1000.

TABLE 3: Average intensity gain versus potentiometer setting.

Potentiometer setting Intensity gain
1000 100.0
990 85.3
980 73.7
960 54.6
940 39.8
920 29.0
900 20.8
850 9.01
800 3.86
750 1.91
700 1.26
650 1.07

The pulse width of the FG—lOO pulse generator is variable from 18 to 650 ns in
range 1 of the variable width mode. A Tektronix 11302 oscillosc0pe was used to measure

the pulse width. Table 4 summarizes the pulse width versus potentiometer setting.

44

TABLE 4: Gate pulse width versus potentiometer setting.

Potentiometer setting Pulse width (ns)
1000 600 to 650
998 550
997 500
996 450
995 400
993 350
990 300
986 250
980 200
970 150
945 100
0 18

The potentiometer collar with the setting indicator mark on it must be turned clockwise
when setting the position. There is a small amount of play in the collar which will affect
the pulse width by as much as 50 ns. If the potentiometer is turned gently to 1000, the
pulse width will be 600 ns. If light torque is applied to the knob, the pulse width will
increase to 650 ns. All other positions were repeatable.

Another experiment was conducted to determine the gate pulse delay setting
required to initiate detector gating immediately after the laser emission pulse, yet avoid
gating before laser emission had decayed. This was necessary to ensure that the image
intensities in the calibration were due only to fluorescence emission and not direct laser
emission. It was carried out by focusing the detector on a piece of aluminum which
scattered the laser pulse while emitting negligible fluorescence that would obscure the
decay of the laser emission pulse. The laser pulse width is approximately 18 ns. The
shortest detector gate pulse width, also of approximately 18 ns, was used. The maximum
intensity of the images was then monitored while varying the delay setting. The peak

image intensities occurred at a delay potentiometer setting of 624. Increasing or

45

decreasing the delay fiom 624 resulted in a decrease in the intensities. Therefore, the laser
emission pulse and the detector gate pulse were completely overlapped at 624.

The absolute delay time could not be measured. However, the change in delay
time for a given change in potentiometer setting could be measured with the oscilloscope.
From this it was determined that the delay time increases 1.6 ns per unit change of the
potentiometer setting.

Delaying the gate pulse an additional 20 ns should place the leading edge of the
gate pulse 2 ns after the trailing edge of the laser emission pulse. This translated into a
delay setting of 637. At this setting, the image intensities indicated that the trailing edge
of some of the laser pulses was being caught by the gate pulse. This was probably due to
trigger jitter in the excimer laser. A delay setting of 640 cleared this up and was used for
the calibration.

Changing the gate pulse width does not affect the delay setting required. The
trailing edge of the gate pulse moves when the pulse width is changed. The leading edge

remains fixed relative to the trigger.

"iliiiliiIIII