TRW REPORT N0. 14103-60d3rRO-<
FINAL REPORT
DEVELOPMENT OF LASER INSTRUMENTATION
FOR
PARTICLE MEASUREMENT
JUNE 1971
prepared by
B. J. MATTHEWS and R. F. KEMP
CONTINUATION OF EPA CONTRACT CPA 70-4
TRW
SYSTEMS GROUP
ONE SPACE PARK • H 6 D O N D O BEACH • CALIFORNIA
-------
OFFICIAL USE ONLY
Final, interim and monthly reports submitted under this contract
con:tain j nformation and statements which are prel iminary' and represent
only the sta~e of the .i~formation developed as of the reporting date.
. This information ;s also subject to review and critique by EPA
personnel. before. re') easeoutsi de of Ef'A. . To preventinappropri ate
. dissemination of information which could be misinterpreted and/or mis-
leading, you are asked to regard these. reports strictly as internal
.workingdoc:umeritsand treat them accordingly. You are requested to
'observe/the following guidelines: " . .
(2)
Use the reports for information and coordi nati on purposes
only.
00 not discuss the reports or the information contained
in them with persons outsid~ of EPA.
(1)
(3)
Refer all inquiries relating to the reports to the Project
Offi cer. . .
(4)
Provide comments to the Project Offi cer for hi s use in
managing the contract.
-------
TRW REPORT NO. 14103-6003- RO-OO
FINAL REPORT
DEVELOPMENT OF LASER INSTRUMENTATION
FOR
PARTICLE MEASUREMENT
JUNE 1971
prepared by
B. J. MATTHEWS and R. F. KEMP
CONTINUATION OF EPA CONTRACT CPA 70-4
TRW
SYSTEMS GROUP
ONE SPACE PARK' REDONDO BEACH' CALIFORNIA
-------
FOREWORD
This report summarizes analytical and experimental work accomp-
lished under Modification No.7 to the Environmental Protection Agency's
Contract CPA 70-4, dated 20 April 1970. Submittal of this report is
in accordance with provisions of the referenced contract.
Technical direction and administration of the program by the
Environmental Protection Agency was provided initially by
Dr. Frederic C. Jaye, and later in the program, by Mr. Robert M. Statnick
The TRW Project Manager was Mr. B. J. Matthews. TRW staff engineer,
Mr. R. F. Kemp, was responsible for developing and demonstrating
the laser-illuminated synchronous detection system used at the Shawnee
Power Plant during field tests. Mathematical modeling of laser beam
scattering together with basic particle scattering measurements was
accomplished by Dr. L. O. Heflinger of TRW's Systems Group Research
Staff. Dr. R. F. Wuerker of the Systems Group Research Staff developed
the holographic methods for recording both side and back scattering of
the pulsed ruby laser beam. The authors are also indebted to
Drs. Wuerker and Heflinger for the use of the TRW Chlorophyll dye cell
used in the successful demonstration of holography of side and back
scatter over extended distances.
i; ;
-------
CONTENTS
1.
INTRODUCTION AND SUMMARY........... . ... . . . . . .. .. . . .. . . . . . .
1 . 1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Summary of Results ...................................
2.
ANALYSIS. ... .. .. . . .. . . .. . . . . ... .. .. . . .. . . .. .. . . . . . . . . ... . .
2.1 Scattering Relationships .............................
2.2 Supporting Laboratory Tests ... ................. ......
2.3 Particle Microscopy..................................
3.
EXPERIMENTAL EQUIPMENT ............... ............. ........
3. 1 Rub y Las e r Ho 1 0 came ra ................................
3.2 Synchronous Detector.................................
4.
RES UL TS [[[
4.1 Unit 10 Transmission Measurements ... .... ..... ..... ...
4.2 Unit 10 Light Scattering Experiments .................
5.
6.
CONCLUSIONS.. . . .. . . ... . . . . . .. .. . . . . .. . . .. .. .. . ... .. . . .. . . .
RECOMMENDATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
Page
-------
16.
17.
ILLUSTRATIONS
1.
Schematic diagram of probing beam and scattered light
geomet ry ..................................................
2.
3.
Detail of beam intersection ..... ...... ..... ... .... ........
Plot of the concentration function, y = xe-x ....... .......
4.
Schematic diagram of apparatus used for scattering and
transmission measurements.................................
5.
Transmission of laser beam as a function of concentration
for Unit 10 flyash and flyash-limestone mixtures .. ........
6.
Ninety degree scattering of laser beam as a function of
concentration for Unit 10 flyash and flyash-limestone
mi xtures ..................................................
7.
Straight path transmission versus 900 scattering for
mechanically and electrostatically precipitated flyash
samples with and without limestone ........................
8. Polar plot
s ta t i cally
9. Polar plot
statically
of scattering intensity for Unit 10 electro-
precipitated flyash-limestone mixture .. ........
of scattering intensity for Unit 10 electro-
precipitated flyash dust ....... ......... .......
10.
Polar plot of scattering intensity for Unit 10 mechanical-
ly collected flyash dust .......... .................. ......
Scanning electron microscope (SEM) photograph of flyash-
1 imestone mixture.........................................
11.
12.
13.
SEM micrograph (300X) of broken flyash sphere ...... .......
SEM micrograph of broken flyash sphere of 80 micron broken
sphere. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14.
15.
Extreme enlargement of SEM micrograph of hollow sphere ....
SEM micrograph of fractured flyash sphere ..... ..... .......
SEM micrograph of flyash particle with hole melted in
surface. .. .. . .. . . . .. . . . . .. . .. .. .. . . . . . . ... .. . . . . . . . .. . . . . .
SEM micrograph of second flyash sphere with hole melted
ins urface ................................................
vii
Page
6
7
8
10
11
13
14
15
16
17
20
21
22
23
24
26
27
-------
31.
32.
ILLUSTRATIONS (Continued)
18.
Transmission micrograph (400X) of 80 micron diameter
parti cl e ..................................................
Six micrographs of 150 micron particle showing progressive
.disintegration of structure...............................
Microqraph (160X) of small 50 micron particle imbedded in
larger 140 micron flyash particle........... ........ ......
19.
20.
21.
Schematic diagram of hologrqphic arrangement for recordinq
ruby 1 aser back scatteri ng .~............................:.
Schematic diagram of holographic arrangement to record side
scattering of ruby laser beam... ... """"" ... ..... "'"
22.
23.
Schematic diagram of laser illuminated synchronous detector
24.
25.
Photomultiplier tube and focusing telescope assembly. ....
Schematic diagram of setup for transmission measurements
at elevation 392 of the Unit 10 superheater region .......
26.
Schematic diagram of setup for scattering measurements
at elevation 376 of the Unit 10 boiler... .......... .....
27.
Relative locations from which scattered light measure-
ments were made. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . .
28.
Strip chart record of laser light transmission across Unit
10 superheater at elevation 392, on 16 March 1971 ........
Strip chart record of 90 degree scattering at elevation
376 on 18 March 1971......................................
29.
30.
Strip chart record of 125 degree scattering at elevation
376 on 18 March 1971...............""""'''''''''''",
Strip chart record of 115 degree scattering at elevation
376 on 18 March 1971 .....................................
Schematic diagram of proposed three-beam holocamera design
for use at the Shawnee Unit 10 boiler. ...,.. .... "" "'"
viii
Page
29
30
31
33
35
37
38
40
43
46
52
54
55
56
59
-------
I.
II.
I I I.
I V.
V.
VI.
LIST OF TABLES
Transmission measurements, Unit 10 superheater,
elevation 392, 16 March 1971 ... ...... ..... ...........
Values of nand mfp for transmission data, Unit 10
superheater, elevation 392,16 March 1971............
Tabulation of results of "single point" method using
Eq. (6) to solve for transmittance with and without
limestone, and a list of the ratios of transmittance
with limestone to transmittance without limestone "'.
Calculation of the scattering fraction B, assuming
mfp = 66 inches.....................................
Calculation of the scattering fraction ~, assuminq
mfp = 55 inches... .................................
Calculations of the scattering fraction ~ using the
II two- po i nt II method..................................
ix
Page
41
42
47
49
49
51
-------
1.
INTRODUCTION AND SUMMARY
1.1
BACKGROUND
This report summarizes recent developments in laser-illuminated
instrumentation for measurina particulate density distributions in an
operatinq coal-fired steam boiler. The development work, sponsored by
the Environmental Protection Agency (EPA), is part of an ongoing program
at TRW Systems Group to characterize limestone particle distributions
injected into a boiler to remove sulfur oxide from the combustion gases.
Field tests in this program were conducted at the Tennessee Valley
Authority's (TVA) Shawnee Power Plant where the full scale dry limestone
program is being evaluated on one of ten identical 140 megawatt boilers.
This report is the third in a series by TRW covering development
of laser instrumentation for particulate visualization and measurement.*
The previous studies are reported in References 1 and 2; however, a
brief review of the highlights of this past work is useful as an intro-
duction to the present investigation.
Initial work under this contract was concerned with applying single
beam (Gabor) pulsed ruby laser holography to the recording of particulate
over extreme distances. Although this technique was not ultimately
used at the boiler, Gabor holography was extended into an area never
before explored; namely, recording distributed phenomena across distances
up to 50 feet.
Two-beam ruby laser holography was successfully utilized at the
Shawnee Unit 10 boiler to record particulate clouds and obtain spatial
information as well as qualitative details of the limestone clouds in
the presences of flyash backgrounds.l This was by the fO~/ard scattering
method. Subsequently, a three-beam scattered liqht holography technique
was devised to acquire quantitative particle number density data
directly from the reconstructed holographic areal real image.2 Means
were also developed to holographically record ruby laser back scatter by
particulate thus adding more flexibility to the basic technique. The
latter method, first described in Reference 2, was accomplished in a
preliminary manner and served as a basis for the present work.
Earlier work involving transmission holography of low angle forward
scattering by particulate in the Unit 10 boiler necessitated using two
directly opposed ports in the boiler walls. The absence of opposed ports
in the reqion of interest (elevation 376) precluded transmission studies.
It was desired, therefore, to investigate the feasibility of recordinq
laser back scatter through a single port, or recording side scatter by
*
EPA Contract CPA 70-4
-------
the use of a combination of two of the several existinq ports at this
elevation. The objective of the present work then, was to determine
if limestone dust in the presence of a flyash bac~ground could be
recorded holographically via back or side scatterlng of a ruby laser
beam.
1.2 SUMMARY OF PRESENT WORK
To investigate the feasibility of making holograms of side
(approximately 90 degree) or back scattering of a ruby las:r beam~ a
mathematical model was first derived. This model (Eq. 6, ln Sectlon 2.1)
relates the following factors which describe the "transmittance" of
scattered light: (1) geometric relationships of the incident and.
scattered (received) beams; (2) scatterinq efficiency of the partlculate;
(3) angular dependence of the scattering;' and, (4) the particulate
concentration dependence. The relationship of these four factors gives
the transmittance along the scatterinq path from the laser to a
detector (hologram, photomultiplier tube, etc.), and allows analysis of
experimental measurements.
Development of a mathematical model of the scattering was
augmented by laboratory tests to: (1) assess variations in scatterinq
efficiency between Shawnee Unit 10 flyash and flyash-limestone mixtures;
and, (2) determine angular dependence of the scattering intensity for
both types of particulate. The tests showed that there was less than a
factor of two difference in the scattering efficiency between flyash and
mixtures of flyash and limestone dust. Intensity of the scattering was
predominantly in the low angle forward direction with a smaller lobe
in the back scatter direction. Side scattering (~40 to 160 degrees)
was the 1 eas t intense. It was a factor of rv 103 1 ess than the i ntens ity
of forward scatter, and between 10 and 20 times less than back scatter
intensity.
In addition to the scattering measurements, particle microscopy
studies were extended beyond those reported in Reference 2, with the
aid of a scanning electron microscope (SEM). This work revealed that
in some instances at least, smaller particles are encapsulated in larger
flyash spheres during the boiler combustion process. Determination
of the mass fraction of encapsulated particles was not attempted.
Early in the program, it was decided to change the scope of work to
include preliminary scattering measurements at the Shawnee Unit 10
boiler.* Original plans called only for straight path transmission
measurements at elevation 392 of the Unit 10 superheater. Ultimately,
scattering holograms were contemplated for elevation 376 (Plane A-A) and
* Modification No.7 to Contract CPA 70-4.
2
-------
there were no opposed ports at this elevation with which to measure
transmittance. To assess the feasibility of recording back and side
scattered light holograms, it was also necessary to make preliminary
scattering measurements at this lower elevation. Subsequently,
transmission measurements were made across the superheater (elevation 392)
and scattering measurements were made at Plane A-A during the period 16 to
18 March 1971.
The transmission measurements were made in the absence of any
limestone injection and indicated that averaqe flyash particle densities
varied between 4000 and 4600 particles/in3.
Existing port locations were used at elevation 376 to make
scattering measurements both with and without limestone injection into
the boiler, on 17 and 18 March 1971. Measurements were made at scatterinq
angles of 90, 115 and 125 degrees. The measuring apparatus consisted of
a continuous wave helium-neon gas laser, a chopper to convert the incident
laser beam into a 750 Hz a.c. signal, a telescope and photomultiplier
assembly to observe and measure the intensity of the scattering signal,
and a lock-in amplifier. The output signals from the photomultiplier
and the lock-in amplifier were recorded on a two-channel strip chart
recorder. This assembly of components was termed a synchronous detection
system.
A factor of between 1.3 and 2.4 increase in the amount of scattered
light was measured during periods of limestone injection. The traces
exhibited a high degree of noise. Analysis of the measured data using
the mathematical model led to inconclusive results with regard to
determining particle concentrations. Differences in particle concen-
tration between conditions with and without limestone injection could
not be ascertained. They were obscurred by signal noise and
uncertainties in geometric and scattering factors in the mathematical
model.
Calculations based upon the measured scattering intensities at
elevation 376 indicate that pulsed ruby laser holography of particle
back scatter may be feasible. The depth of field of the holograms will
probably be limited to a distance of about six feet inside the boiler.
The scattering measurements at elevation 376 were also analyzed to de-
termine particle flow field dynamic characteristics. This analysis
indicated that limestone distribution at this elevation does not appear
spatially or temporally uniform. Random variations in signal intensity
and frequency were noted during periods of limestone injection.
3
-------
2.
ANALYSIS
This section is concerned with development of a model for scatte~ing
by flyash particles of a light from a probing laser beam. The analyslS
was done by Dr. L. O. Heflinger of the TRW Systems Group Research Staff
(SGRS). The work was in preparation for a test program at the. Shawnee
Unit 10 boiler (Section 4.2). In addition, laboratory scatterl~g
measurements were made by Dr. Hefliner in support of the.analysls .and
subsequent Unit 10 data reduction and evaluation. Also lncluded ln
this section are the results of supporting particle microscopy work.
This is an extension of previous microscopic studies of flyash.2
2.1
SCATTERING RELATIONSHIPS
The followin9 is a brief summary of scattering relationships
between the probing laser light and Unit 10 flyash particles. The
scattering relationships led to a mathematical formula which gives the
intensity of scattered light in terms of particle concentration and the
geometry of the optical configuration. In the following discussion, let
n = the number of particles per unit volume.
o = the geometric cross section area of a particle.
For a spherical particle, 0 = TI 02/4, where 0 is the diameter of the
particle. In the scattering formulas, it is the product no which enters
the formulas. From TVA data on Shawnee Unit 10 particle size
distribution, it has been determined by integration that the effective
particle diameter is 40~ (40 x 10-6 m), giving an effective 0 of:
0eff = TI/4 (40 x 10-6)2
-9 2
0eff = 1.26 x 10 m.
(1 )
If this effective value is used, the value of n obtained from the
formulas will be the total particle concentration, counting all sizes of
particles.
Note that 4 no is the total surface area of the particles per
unit volume, for spherical particles. Thus, the light scattering
experiments can be interpreted as measuring the particle surface area
per unit volume directly. It can be shown that the value of no
determined by experiment gives the total particle surface area per unit
volume essentially independent of the particle size distribution.
Transmission Measurements
The number of particles per unit of volume may be obtained by
measuring the transmitted light through a scattering medium. In this
4
-------
case~ let
L = length of path through the scatterin9 medium.
T = transmittance of this path
= intensity of output beam
intensity of input beam
Then for a uniform scattering region
T = e-2ncrL
(2)
This can be rewritten
or ~
ncr = ~L (-2,n T)
1 1
ncr = 2[ In f
(3)
If the scattering is not uniform throughout the region~ the value of
ncr obtained is the average over the path L. The coefficient of L in
Equation (2) is the extinction coefficient and has the dimensions of an
inverse distance. Hence~ the distance 1/2ncr is characteristic of the
scattering medium~ and is the distance over which the intensity of
transmitted light is attenuated by the factor e = 2.718... Hence~
a given photon has about a 37 percent probability of traversing this
distance within the medium before encountering a scattering particle.
We take the liberty of calling this distance "mfp" and note that it
has nothing to do with the mean-free-path of kinetic theory. This
distance may be calculated from a transmission measurement by rewriting
Equation (2) again:
1 L
"mfp" =-=-
2ncr -wT
(4)
Scattered Light Measurements
At most planes of interest within the Unit 10 boiler~ directly
opposed ports for transmission measurements do not exist. Therefore~
one would like to estimate local particle concentrations based upon low
angle back scatter measurements where only a single viewing port is
required. An alternative is the measurement of side scatter intensities
provided two ports are available which allow intersecting optical paths.
For the measurement of direct back scatter or of side scatter~ a
number of geometric factors must be determined. The relationship
of the probing and scattered beams is shown in the Figure 1 sketch.
In this sketch~ 8 = angle of observation (8 = 0 is looking into the
beam) .
5
-------
PROBING
BEAM 1 N
..
1 1
FURNACE
SCATTERED
BEAM OUT
Figure 1.
Schematic diagram of probing beam and
scattered light beam geometry.
The small volume element from which liqht is scattered into the
receiving optical system is defined by the intersection of two beams.
One is the laser beam and the other may be called the viewinq beam. A
small aperture (approximately 1 mm diameter) is placed at the focal
point of the telescope objective lens. Light passing through this
aperture falls on the photomultiplier tube. Consequently, the image
of this aperture within the scatterina volume determines the axial
position and diameter of the viewing beam. It is convenient, though
not exact, to approximate the resulting scattering volume element as
the intersection of two cylinders having intersectina axes at an
angle 8. If the diameter of the viewing beam w at the scattering
point is larger than the laser beam, the situation is particularly
simple: If 5 equals the cross-section area of the laser beam, then
the fraction of light incident on the scattering volume element which
each particle removes is 0/5. The total number of particles N in the
volume element is n times the volume of the element or
N = nw5/sin 8
and the total fraction of light removed from the beam within the volume
element is Na/5 = now/sin 8. Of this light, a fraction, B, is scattered
6
-------
(we assume here) uniformly in all direction.
is th us ~
The scattered light fraction
Pout - w
Pin - sin e B na
(s teradi an - 1 )
Figure 2 shows a detailed view o~ the intersection of the laser beam and
the viewing beam in the case where the telescope aperture is a slit
and the image of this slit has width wand thickness less than the
diameter of the laser beam. In this, and other similar cases, the
fraction a may be used with w to give the effective length of the laser
beam in the scattering volume element. The fraction a is chosen so
that waS/sin e is equal to the actual volume seen by the telescope.
The solid angle subtended2at2the scattering volume element by the
receiver optics if r = nr /R (steradian) where r is the radius of the
entrance pupil and R is the distance from the scattering volume to the
telescope. Finally~ the total path length for absorption is L = £1 + £2'
so that the total transmittance for the system is given by
T = r (1) Wa B -2naL
4; sin e na e
IN
..
W
INTERSECTION
VOLUME
THIS IS a
OF THE
FUL L BEAM
PERSPECTIVE VIEW
TOP VIEW
~
Figure 2.
Detail of beam intersection
7
-------
This can be rewritten as
T - r Wa S (2naL)e-2naL
- 8TIL sin e
(5 )
in order to obtain a factor involving the product na which has the form
-x
y = xe
x = 2naL
and which may be called the concentration function. This function is
plotted in Figure 3, and may be used to predict the behavior of the
transmittance of the system for a given path length L as the (uniform)
number density is increased from zero. For low values of particulate
loading, the received intensity is nearly proportional to number
density. However, as n increases to the point where 1/2na = L, the
intensity reaches a maximum. Beyond this point, extinction proceeds
more rapidly than scattering fraction, and the received intensity
decreases as the number density increases. However, one should not
overlook the fact that if the particulate loading is highly non-uniform,
scattering of light is proportional to the number density within the
scattering volume element, while attenuation is related to average
number density over the path length L. In general~ the scattering of
light is not uniform, so that the angular dependence is not simply
1/4TI as aSJsumed above, but is given by a function f(e) such that over
a sphere, f(e)de = 1.
On the basis of these parameters and considerations, the following
formula gives the transmittance via the scattering path from the laser
to the detector, and will serve as a mathematical model for analysis
.5
.4
.3
Y
.2
. 1
0
0 2 3 4 5 6 7
X=2naL
F i CJ u re 3.
Plot of the concentration function, y = xe-x.
8
-------
of experimental results.
rWa
T = 8TIL . B .
--- -.-
/
/
Geometric
Factors
fee)
sin.e
(2no- L)e-2ncrL
(6)
I
I
Particle
Scattering
Efficiency
'\.
"
Angular
Dependence
~
Concentration
Dependence
The following section describes some laboratory experiments in
which measurements were made of the angular dependence of the scattering
fraction. In as much as this turned out to be a slowly varying function
for the range of angles used in scattering measurements within the
Unit 10 boiler at the Shawnee Power Plant, Equation 6, above, was used
in the data analysis.
2.2 SUPPORTING LABORATORY TESTS
Laboratory measurements were made to assess particle scattering
efficiency and angular dependence terms in Eq. (6) for flyash and
mixtures of flyash and limestone. Samples were obtained from both the
Unit 10 mechanical collector and electrostatic precipitator.
A schematic diagram of the test apparatus is shown in Pigure 4. A
beam from a 1.6 milliwatt helium-neon gas laser is directed through a
scattering cell. The light scattering cell contains two flat windows
through which the laser beam passes. The rear portion of the cell
is frosted to prevent reflections. The output of the laser and the
intensity of the light after transmission through the scattering cell
are monitored with silicon detectors as shown in the schematic. The
laser silicon detectors and scattering cell are mounted on a heavy
duty rotating azimuth head. The telescope and photomultiplier assembly
used to measure the scattered light remain stationary.
Alignment of the telescope onto the laser beam passing through the
scattering cell is accomplished using a mirror with an aperture located
at the focal point of the 2 inch diameter, 8 inch focal length objective
of the telescope. An eyepiece mounted normal to the telescope axis
allows the observer to accurately align the laser beam scatter in the
telescope aperture. The photomultiplier then receives the maximum
scattering signal.
Measurements of the light transmission and scattering intensities
of flyash and of flyash-limestone mixtures were made by suspending the
appropriate particulate in water. A mechanical agitator was used to
keep the particles in suspension. Measurements were made for different
particulate concentrations as well as for changes in the scattering
9
-------
LASER
INTENSITY
MONITOR -
SILICON
~ DETECTOR
~
- D t::::=
I
. FLAT
I WINDOW
I \
He Ne ~ T ~
1--- ---- - - - --~ - ---- SILICON
1.6mw '~---
I :' " ~ DETECTOR
BEAM / ' /
SPLITTER //~ e./ FLAT
,/" / WINDOW
" I
" MEASURED
WITH TELESCOPE
AZ IMUTH HEAD
LIGHT
SCATTERING
CELL
TRANSMITTED
/ INTENSITY
/" (microammeter)
LASE R
-'
o
MIRROR WITH ACCURATE
APERTURE TO
SELECT MEASURED
PORTION OF BEAM
SCATTERED INTENSITY
(DC VOLTMETER)
2" D1AMETER,
8" FOCAL LENGTH
OBJECT IVE
CELL FROSTED ON
REAR SIDE
TO PREVENT
REFLECTIONS
""'~
,-
',- ' ".1 L EYE FOR ALIGNMENT
, OF APERTURE ON
" " LASER BEAM
Figure 4.
Schematic diagram of apparatus used for scattering and
transmission measurements.
-------
angle, 8. The angle e was varied within the limits of the apparatus
setup resultina in measurements from 8 = 6 degrees to 8 = 174 degrees
(8 = 180 degrees represents back scatter while 8 = 0 degrees is
transmitted light).
Transmission Measurements
Transmission data for various concentrations of both electro-
statically precipitated and mechanically separated Unit 10 flyash and
flyash-limestone mixtures are plotted in Figure 5. The transmission
path was 35.5 mm. The data cover a range of volumetric concentrations
of from 0 to 6. The unit of volumetric measure is arbitrary and termed
a "mill ispoon" (ms). For the mechanically separated dust, one ms
contains 6.4 x 10-j grams (dry). One ms of electrostatically
precipitated dust contains 5.5 x 10-3 grams. Thus at a concentration
of 4 with mechanically separated flyash, the density would be
(6.4 x 10-3) glms x (4 x 10-2) mslcc = 2.56 x 10-4 glee. Similarly, at
the same concentration the electrostatically precipitated flyash
density would be 2.20 x 10-4 glee, etc.
The curves of Figure 5 show that at hiqher concentrations of
particulate, there is an increase in the attenuation of the transmitted
laser beam. Further, for equivalent concentrations, the light attenuation
is greater for suspensions of electrostatically precipitated dust than
for mechanically separated particulate. This is due to the
smaller size distribution of the electrostatically precipitated flyash.
For example, in one analysis, approximately 55 percent of the dust
I
~
E
E
"'
:;]
~
o
z
Q
~ 0.1
Z
~
~
I
~
:'!.
~
I
()
~
t;
1.0
---t- ----
- Li.....! --- --
I I
.- - - L--L
I i
i---t-- -i-- ,-
NOMENCLATURE
M MECHANICALLY SEPARATED FLYASH
M, MECHANICALLY SEPARATED FLYASH-
LIME ST 0 NE MIX TURE
E ELECTROSTATICALLY PRECIPITATED FLYASH
EI ELECTROSTATICALLY PRECIPITATED FLYASH-
. LIMESTONE MIXTURE
SUBSCRIPT NUMBERS REFER TO DIFFERENT SAMPLES
OF SAME MA TERIAL
I
I
1---
I
0.01
o
2 3 4 5
CONCENTRATIONS, MILLISPOON/iOO CM3
Figure 5.
Transmission of He-Ne gas laser light beam as a function of
concentration for Unit 10 flyash and flyash-limestone mixtures.
11
-------
collected from the Unit 10 electrostatic precipitator was found to be
less than 10 microns in size.3 Conversely, only 7-10 percent of the
dust collected by the mechanical separator on Unit 10 was less than
10 microns.3 The significant increase in fines in the electrostatic
precipitator flyash results in a corresponding increase in the total
particulate surface area intercepted by the transmitted laser beam.
The expression for light transmission through a scattering medium is
T = exp (-na£) where na is the product of the number of particles
and the average particle scattering cross-section over the path length £.
The product na is also proportional to the total surface area of the
particles. Therefore, as the total particulate surface area increases,
more light in the beam is intercepted and the beam attenuation is
increased.
Of greater significance to the present work, however, is that there
generally appears to be little difference in the light attenuation
characteristics between flyash samples containing limestone dust and
those without limestone. This is seen in Figure 5. Some deviation
does occur between the samples with and without limestone for the
electrostatically precipitated dust, but only at very high concentrations.
In other words. one could not expect to readily differentiate between
flyash and flyash-limestone mixtures by straight path transmission
measurement based upon the data obtained during the laboratory tests.
Scattering Measurements
Light intensity measurements as a function of the scattering
angle 8 were made for different concentrations of both flyash and
flyash-limestone mixtures. Of particular interest was the 90 degree
scattering condition. These data are plotted in Figure 6. The measure
of concentration is the same as that described for Figure 5. As with
the transmission data, it is apparent that the presence of limestone
has virtually no affect on side (90 degree) scatter intensity. Indeed,
there is less than a factor of two difference between the mechanically
and electrostatically precipitated dust samples.
The data of Figures 5 and 6 are cross-ploted in figure 7, where
straight path transmission versus 90 degree scattering curves are
presented for mechanically and electrostatically precipitated flyash
both with and without limestone. The slope of each curve at the
origin is proportional to the scattering efficiency or S f(90o) term
in equation (6). These values are tabulated on the next page. It can
be seen that the presence of limestone in the f1yash has a minimal
affect on the scattering efficiency of the dust. There is less than
a factor of two difference in the spread of values measured, and no
consistant trend.
The polar plots in figures 8 through 10 were obtained using the
scattering cell apparatus described previously. !hey show the angular
variation in scattering intensity for e1ectrostat1ca11y precipitated
f1yash both with and without limestone, and for mechanically trapped
f1yash without limestone. On each plot there are two curves. The
right-hand curve is the apparent scattering from ~ beam. The left-hand
curve represents single particle scattering, and 1S derived from the other
12
-------
-
\I")
I--
...J
o 12
G
I--
I
o
...J 1 0
CI
w
c.::
W
I--
I--
<{
~ 8
o
o
0-.
20
E£l
18
16
M
14
6
4
NOMENCLATURE
M = MECHANICALLY SEPARATED FLYASH
M£ = MECHANICALLY SEPARATED FLYASH -
LIMESTONE MIXTURE
E = ELECTROSTATICALLY PRECIPITATED FLYASH
El = ELECTROSTATICALLY PRECIPITATED FLYASH -
LIMESTONE MIXTURE
SUBSCRIPT NUMBERS REFER TO DIFFERENT SAMPLES
OF SAME MATERIAL
2
o
o
2 3 4 5
CONCENTRATION, MILLISPOONS/100 CM3
6
7
Figure 6.
Ninety deqree scattering of He-Ne gas laser light
beam as a function of concentration for Unit 10
flyash and flyash-limestone mixtures.
13
-------
Z 0.8
o
V)
V)
-+::>
~
V)
Z 0.6
«
~
I-
:r:
I-
«
CI.. 0 4
I- .
:r:
o
~
I- E
V) 0.2 E£
1.0
NOMENCLATURE
M = MECHANICALLY SEPARATED FLYASH
M£ = MECHANICALLY SEPARATED FLYASH -
LIMESTONE MIXTURE
= ELECTROSTATICALLY PRECIPITATED FLYASH
= ELECTROSTATICALLY PRECIPITATED FLYASH -
LIMESTONE MIXTURE
SUBSCRIPT NUMBERS REFER TO DIFFERENT SAMPLES
o OF SAME MATERIAL
o 2 4 6 8 10 12
90° SIDE SCATTERED LIGHT (VOLTS)
Ell
--
.
E 12
14 16 18 20
Figure 7.
Straight path transmission versus 90° scattered light
intensity for mechanically and electrostatically
precipitated flyash samples with and without limestone.
-------
1800
90°
1200
120°
90°
600
o
Fi gure 8.
Polar plot of scattering intensity distribution for Unit 10
electrostatically precipitated f1yash-1imestone mixture.
15
-------
1800
900
1200
1200
900
600
10000
Figure 9.
o
Polar plot of scattering intensity distribution for Unit 10
electrostatically precipitated flyash dust.
16
-------
1800
1200
900
o
Fi gure 10.
Polar plot of scattering intensity distribution for Unit 10
mechanically collected flyash dust.
17
-------
Type S f(900) Value
ESP flyash wlo limestone .087
ESP flyash wi 1 imestone, #1 sample .104
ESP flyash wi limestone, #2 sample .121
Mech. flyash wlo limestone .138
Mech. flyash wi limestone, #1 sample .121
Mech. flyash wi limestone, #2 sample .12
Notes: ESP = Electrostatically precipitated flyash
MECH = Mechanically separated flyash
curve as follows: The intensity measurements of the laser beam scatter vary
as beam intersection volume changes with the angle of measurement; i.e.,
the scattering volume = llsin e x constant, where e is the angle between
the probing laser beam and the beam of scattered light received by the
observing telescope (see Figure 1). When e = 0 degrees (straight path
transmission), the scattering volume is infinite. The observed scattering
volume is a minimum value at e = 90 degrees. To obtain single particle
scattering, it is necessary to remove the geometric factor from the measure-
ments by applying a sin e factor. This results in the left-hand curve on
each of the three polar plots. The data are normalized to transmission
measurements as a function of particle concentration and thus all curves
may be directly compared in the three plots. The curves are open at both
the forward and back scatter (e = 0 degrees and 180 degrees) regions due to
limitations of the measuring apparatus setup.
An evaluation of these curves shows that only two significant
scattering lobes exist with the predominant lobe in the direction of
forward scatter (e = 0 degrees). The scattering intensity is near a
minimum at 90 degrees. The intensity of the forward scatter is estimated
to be between 50 and 100 times greater than for the back scatter condition.
In the scattering tests performed at the Shawnee Unit 10 boiler (described
subsequently in Section 4.2), measurements were made at values of e= 90,
115 and 125 degrees. It can be seen from the polar plots in Figures 8
through 10 that the scattering intensities are minimal in this region.
A comparison of the polar plots of Figures 8 and 9. scattering
intensity for flyash only and flyash with limestone, shows that there
is little difference in the shape of the two curves or in the absolute
values at any given angle. There is less than a factor of two
difference in the scattering intensity between the two types of dust
for values of e equal to those for which measurements were made at
the Unit 10 boiler. This fact inhibits the detection or identification
18
-------
of limestone dust in the presence of a background of flyash particulate.
In short, the scattering characteristics oT the two different types of
dust proved to be remarkably similar.
2.3 PARTICLE MICROSCOPY
Determi nati on of s catteri ng characteri s ti cs of flyash by TRW Sys tems
led to the discovery that some larger spherical particles in the samples
were hollow and apparently filled with numerous smaller particles. This
was revealed during a series of scanning electron microscope studies of
flyash samples from the electrostatic precipitator and mechanical separator
of the Unit 10 boiler. The work was a continuation of microscopy studies
reported in Reference 2.
One of the first examples of a scanning electron micrograph which sug-
gested that large hollow spheres contained smaller particles is seen in
Figure 11. This is a ~lOOOX photomicrograph of a mixture of limestone and
flyash obtained from mechanical separator hoppers of Unit 10. The arrow
in the upper right-hand part of the photograph shows a ~30 micron sphere
with a portion of its outer surface broken. It appears that smaller
particles rest inside this shell.
Interest in this broken flyash particle resulted in a search for other
such examples. To obtain some degree of particle separation, acetone was
used to float the dust onto a pedestal before examination. Greater
dispersion was achieved thus aiding the examination of individual particles.
A second particle with a fractured shell was located and photographed using
the scanning electron microscope. Three micrographs of one particle made
at successively larger magnifications (~300X, 1000X and 3000X) are shown in
Figures 12 through 14. Figure 12 is an overall view of the sample area
scanned with the electron microscope. The particle of interest is ~80
mi crons in di ameter. Surface detail s of the fracture and the internal
ensemble of particles are more clearly seen in Figure 13. A more extreme
enlargement (~3000X) of this particle is shown in Figure 14 where two
distinct types of internal particles are clearly observed; namely, spherical
and irregularly shaped particles in the two to 20 micron size range. The
physical appearance of the two different types suggests that the spheres
are probably smaller flyash parti cl es res ulti ng from the combus ti on of
pulverized coal in Unit 10. The angular particles which look like "rock
candy" strongly resemble photomicrographs of limestone calcined at ~1200°C
as reported by McCl ell an in Reference 4.
Review of these rather dramatic photomicrographs raised the question
of whether small particulate inside the larger flyash shell entered after
the shell surface was fractured, or had somehow been encapsulated during
the combustion process. In an attempt to answer this question, additional
micrographic studies were made. The first two tests utilized the TRW
scanning electron microscope. A small sample of dust from the Unit 10
mechanical separator was examined and then rolled with a metal rod to
fracture larger particles. A photomicrograph of a deliberately fractured
particle is shown in Figure 15. The results were considered indicative of
19
-------
Fi gure 11.
Scanning electron microscope (SEM) photograph of a fly'-
ash-limestone mixture in which a broken spherical particle
(arrow) was observed to contain still smaller particles
on the inside. The flyash sample was obtained from the
hoppers of the mechanical separator on the Shawnee Unit 10
coal-fired steam boiler~ Limestone dust was being injected
into the furnace to remove SOx from the flue gases at the
time this sample was obtained Micrographs in all suc-
ceeding illustrations were made from Unit 10 flyash-
1 imes tone mi xtures.
20
-------
Figure 12.
SEM micrograph (-300X) of broken f1yash sphere showing
internal particles. This micrograph was made in a
deliberate attempt to find large broken particles
containing smaller dust on the inside. Acetone was
used to "float" the dust sample onto the SEM pedestal
to achieve a greater degree of particle separation.
The next two figures are photographs of the same
broken particle but taken at magnifications of lOOOX
and 3000X, respectively.
21
-------
Figure 13.
Higher magnification SEM micrograph of broken particl~
seen in Figure 12. This illustration clearly shows the
contents of this hollow flyash sphere. Note the
exceedingly thin wall of the hollow sphere. Particle
diameter is about 80 microns. When this micrograph
was made, it was not known if the internal particles
had fallen in after the shell was broken or had been
encapsulated during the combustion process.
22
-------
J...----- -
Figure 14.
Extreme enlargement of Figure 12 showing contents of
hollow flyash sphere. Note the two distinct types
of particles. The spherical particles are thought
to be small flyash dust while the angular "rock
candy" particles are believed to be calcined lime-
stone. Microprobe analysis is needed to identify
the particle chemistry.
23
-------
Figure15.
SEM micrograph of flyash particle broken by rolling a
metal bar over a dust sample. This was an early attempt
to determine if the smaller particles were originally
encapsulated in the larger sphere or had fallen into
the void after the shell was broken. Although indica-
tive that the internal dust was trapped during the
combustion process, this test did not fully confirm
the fact.
24
-------
an encapsulation process; however, the micrograph could not be considered
as conclusive evidence since the rolling method of fracture could also
have introduced smaller particulate into the hollow sphere.
Another and more unique method was tried. The micrographs of Figures
16 and 17 show the results of the new method. In each instance, a large
unbroken flyash particle was located in the field of view of the scanning
electron microscope. The electron beam was concentrated at the surface of
the flyash particle. Localization of the beam energy caused the surface
material to melt away. Evidence of the molten state of the flyash shell
is easily seen in the two figures. The porous nature of the surface in the
region of the molten particle wall suggests that outgassing occurred
during the melting process.
In Figure 16, the opening in the flyash particle is approximately 45
microns maximum width. The opening in the companion micrograph is about
80 microns. In each instance, the internal volume of the flyash sphere
is full of smaller particles. All examples in the preceding illustrations
were taken from flyash-limestone mixtures obtained from Unit 10 at Shawnee.
The encapsulated particulate in all cases seemed to consist of both
smaller flyash spheres and the irregularly shaped glossy particles thought
to be calcined limestone.
In addition to the scanning electron microscopy studies under the
present contract (CPA 70-4), further microscopic studies were done under
TRW sponsorship using a conventional light microscope. Flyash is typically
a fused silicious residue with a glassy transparent appearance~ This
allows one to conveniently observe particle structure and characteristics
using transmitted light and a high quality light microscope. The following
micrographs were made in this manner using an Izumi Polarization Microscope
(SIN 19637) with a Polaroid film back adaptor and various objective lenses.
Selected particles from samples of Shawnee Unit 10 flyash (mechanical
collector) were crushed while under observation so that the fragmentation
process could be observed. A small quantity of dust was scattered on a
microscope slide. From this ensemble, a single particle of spherical
shape and white color was removed with the aid of a 12X eye loupe and a
teasing wire. The selected particle was then immersed in a drop of
dioctyl phthalate (OOP) oil on another microscope slide, and covered
with two 0.020 inch thick cover glasses. The thickness of the double
cover glasses very nearly matched the distance between the particle and
the barrel of the microscope objective lens at its focal distance. It was
possible, therefore, to use the microscope focusing mechanism to apply
pressure to the particle with the lens barrel while the particle was under
observation.
Micrographs of individual particles were recorded. The microscope
objective lens was then depressed until the particle was observed to
fracture. Further disintegration was observed as additional force was
exerted. Micrographs were recorded at various stages of this process.
25
-------
Figure 16.
SEM micrograph of f1yash particle with hole melted
in previously unbroken surface by concentrating the
energy of the electron beam at the surface. This
and subsequent micrograph using this technique re-
vealed numerous small particles inside the large
particle thus tending to confirm that encapsulation
occurred during combustion process.
26
-------
Figurel~
SEM micrograph of a second flyash particle in which the
surface was melted using the concentrated energy of the
electron beam. Note the voids in the edge of the
encapsulating shell suggesting that entrapped or dis-
solved gases were liberated as the result of the
material being heated by the electron beam.
27
-------
The two 400X micrographs of Figure 18 illustrate an 80 micron
diameter spherical flyash particle photographed before and after fracture.
The upper photograph shows the particle intact and clearly demonstrates
that smaller particles are encapsulated in this larger sphere. The
other photograph was ma~after the particle had been broken by the
method described. A portion of the larger particle has drifted to the
left. The pieces of this particle are still immersed in the oil film
between the two glass slides which serves to hold the fragments within the
field of view.
The series of six micrographs in Figure 19 show several progressive
stages of particle disintegration (from left to right) as a result of
repeatedly squeezing the entrapped large particle. In its original form,
the flyash particle was approximately 150 microns in diameter (Figure 19a).
Definite major fracture lines are seen in Figure 19b. The existence of
one interesting 15 micron particle is followed (arrow) in each of the
micrographs. In the last micrograph (19f) this smaller particle had also
been fractured.
The last illustration (Figure 20) is a lower power (160X) micrograph
of a small particle (arrow) partially assimilated into a larger 140 micron
diameter particle. The smaller particle is perhaps 50 microns in diameter.
Several particles of -10 microns diameter are also seen attached around
the outside of the large particle.*
As mentioned, the particle microscopy reported here is a continuation
of work begun under an earlier portion of the basic contract and reported
in Reference 2. The microscopy was initiated as an aid to understanding
the scattering characteristics of the Unit 10 flyash and limestone-flyash
mixtures. In the process of examining various samples of ash, the contain-
ment of very small particles by larger spheres was noted and investigated.
This work continued to a point where it could be confirmed that the small
particles were indeed encapsulated durinq the combustion process and did
not, as first suspected, become entrapped in the cavities of previously
broken large spheres.
It is suggested that the encapsulating phenomena is important from
two considerations. If a mechanism by which the encapsulation takes place
can be understood, it may be possible to enhance or promote the process
thus reducing the number of fines in the stack gas. Conversely, if
particle collection methods tend to fracture the thin shells of larger
spheres, the possibility exists that additional microscopic dust may be
introduced into the stack gases. The investigation of flyash formation
and encapsulation processes was not within the scope of the present work;
however, additional work and understanding in this area would seem Quite
worth while. .
*
The dark curving line in the upper portion of this micrograph is the
edge of the DOP oil droplet in which the flyash particle was immersed.
28
-------
Fi gure 18.
t":
--1 SOfL
(a)
--- ----
vi
.
@
!II
-1 }}
, .,
t: 4#"'
..to ~-
,.~
~J' l w" \.
.~. .
-------
.50~ t--
it.
a
w
a
d
I -
b
'.7
\. \ e)' .
,.~.
.. ......;~ \ - .. "'~- ~ - '
., \. '~..~ [\~ _4"'.....
t'!.' . ..,',."'K -.
. - -. : ~T I ~ ..
~~- .)t.~ ,,;t~:_,
'" ~. . ~ ~ .,~ ..~.(."-';'~-~'.
~ r' , . ", ~ \ ~' . 4 .~ : ?
~~ ~." ~. IS, :ij;., t-........
. . . . ."k ~. . ~ . , . . "
--c" '\ . . ~. ' ~\'!I ." tJ;'
~ \, .. . ,. ,
'~.~. \:'~'-'-l..i~" ',\ .
. '- ~-r):.~-V'~~ ,,1
or. -- .
..-t~ i:1iM . .". I
--- ------ .-.I
e
"
c
\.
.~'~: '. '
.. ._~ ..
r'-.. . .~..)
., , y ~... ().;.. f
,. =i .~ \.,"",-- ..". " '."
;"'. . ,. . ~. . . / ~.
. .~:\,!., ~,_'>.,\t. "~'*~'. ~1{~. ".,~,,:\.
,.. .r," .~. t:., " . , Al. ~,~....
,... .' ...' . . .... -, .,
# ~ ,... ~ ..,~.. - , 't"-.... .., . ~.. - "..P\
.. J " ~....,~ - ..,.'
:..': -;:. ~.... \ '~. - .~
. (~."'~'~I' "..""1: '.) _1',
,~'~--:', ;. .t"'. ~:it'L. ~.-. ,
,". II '.~ ~., .; . ~... .,
~ ,,, J ';'(4 .. . I \
(:!I, \ ~1X'" .J'. ,,~ '
1.!-~.;~':.. ,--",,) -. _..l~ 01",.-
f
Six micrographs of a large 150 micron particle showing the progressive disintegration of
the structure under increasing pressure from the light microscope objective lens. Several
small internal particles can be traced in the disintegration process. Observation of
particle breakup was continued until one of the internal particles (arrow) was eventually
broken. At th;s po;nt. most of the res;due was ~lO microns.
Figure 19.
-------
-'1100}1- ~
~
Figure 20.
Micrograph (160X) of a small 50 micron
particle partially imbedded in a larger
140 micron flyash particle.
31
-------
3.
EXPERIMENTAL EQUIPMENT
The apparatus used durin~ the experimental effort on this program is
conveniently divided into two categories. The ~irst concerns the ~olographic
tests conducted in a TRW Systems laboratory, whlle the second conslsts of
equipment designed and built for scattering measurements at the Shawnee
Unit 10 boil er.
3.1
RUBY LASER HOLOCAMERA
The TRW pulsed ruby laser illuminator described,in Reference 2 was
aqain used for the present work. The Kerr cell Q sWltch was removed and
replaced with a Chlorophyll d dye cell to achieve increased laser coherence
necessary for the present back and side scatter holography application.
The Chlorophyll dye cell, developed at TRW by Dr. Ralph Wuerker, is
reported in detail in Reference 6. With the dye cell positioned in the
ruby oscillator cavity, laser coherence was increased from a few centimeters
to approximately 3 meters. The laser pulse duration increased from
approximately 50 nanoseconds to about 100 nanoseconds. Energy content of
the laser pulse was between 1 and 1.5 joules.
Demonstration of both back and side scatter pulsed ruby laser holo-
graphy necessitated construction of different holocamera configurations.
All of the holography during the present work was accomplished in a large
laboratory using breadboard configurations. The schematic diagram of
Figure 21 shows the arrangement used for holographically recording ruby
laser back scatter. The arrangement is a three beam system similar to that
used in the earlier transmission holography.2 The ruby laser output beam
is divided into two components using a wedge beam splitter. Approximately
90 percent of the ruby light passes through the beam splitter (W-l) to
form the scene beam. A right angle prism diverts the incident scene beam
into the scattering medium of the scene. A reflection off the first surface
of the scene beam prism is incident on a photodiode used to monitor the
ruby laser output. That portion of the ruby laser light reflected by the
wedge beam splitter is expanded and collimated into a 5-inch diameter ref-
erence beam using a telescope. The collimated light of the reference beam
is further divided into two beams with the aid of a second wedge (W-2).
Reflected light from the second wedge beam splitter is directed onto the
hologram using a front surface mirror (M-l). A second front surface mirror
(M-2) reflects the remaining reference illumination transmitted through
the second beam splitter (W-2) onto the hologram. The path lengths of the
two reference beams are matched.
The hologram is located adjacent to the scene beam such that low angle
~5 degrees) back scatter may be recorded. In tests made to date, back
scatter of pulsed ruby laser light over distances of up to 15 feet have
been successfully recorded. Test aerosols have included chalk dust and
dioctyl phthalate (DOP). These tests were essentially the same as earlier
work2 except that a second reference beam was added as in the previous
transmission scattered light holograms.
32
-------
PHOTO DIODE
W-l
RUBY LASER
TELESCOPE
M-l
W-2
5-INCH DIAMETER
REFERENCE BEAM
.....1
0:: SCATTERING
o 0 0:'. SC ENE
0000 VOLUME
SCENE BEAM
Fi gure 21.
Schematic diagram of laboratory holographic arrangement
used to record back scatter of a ruby 1 aser scene beam.
33
-------
Port geometry and availability at elevation 376 at Shawnee Unit 10
suggested an alternate holographic arrangement, namely the recording of
side scattering of the incident scene beam. The holography of side
scattering (90 degree scattering) offered a potential for covering more
lateral area in the plane of interest at this boiler elevation.* For this
reason, the feasibility of recording side scatter of a pulsed ruby laser
beam was investigated in the laboratory. A breadboard holocamera was
setup to demonstrate feasibility of the techn~que. A schematic diagram
of this holographic arrangement is shown in Flgure 22. For the sake of
simplicity, the second reference beam was omitted in this setup. A two
beam holocamera was adequate for demons trati ng the feas i bil i ty of
holographically recording side scatter from the ruby laser scene beam.
From Figure 22, it will be seen that the output of the ruby laser
was divided into scene and reference components using a wedge beam
splitter (W-l). The reference beam was expanded and collimated into a
5-inch diameter beam using a telescope. Two front surface mirrors
(M-l and M-2) directed the reference beam onto the holographic plate.
Light transmitted through the wedge beam splitter (W-l) formed the scene
beam. Two 90 degree prisms (P-l and P-2) were used to direct the scene
beam in front of and parallel to the holographic plate. The distance
between the scene beam and film plate was approximately 6 feet. For some
tests, a cylinder lens was placed in the path of the laser beam reflected
by prism P-2 to form a fan-shaped scene beam. With the configuration
shown in Figure 22, a nominal scene width of 4.5 feet could be obtained
establishing a viewing angle of about 37 degrees. When the cylinder
lens was used, the maximum width of the fanned beam was about 1 foot.
The scene width was 4.5 feet.
With the holographic arrangement described, both static scenes and
aerosols were recorded. The purpose of these tests was to show the basic
feasibility of holographically recording side scatter of a probing ruby
laser scene beam. Holograms of static objects were first recorded and
success fully recons tructed. Tenuous aerosols of chalk dus t and DOP were
subsequently recorded. Reconstructions of side scattering by aerosols
were, in general, not as satisfying as similar recordings of forward and
back scattering of a ruby laser beam. The side scatter holograms were
not as bright. This result was not unexpected based upon the angular
scattering intensity distribution measurements made earlier in the
program (see Section 2.2). Quantitative measurements of the aerosol
number density from variations in scattering intensity recorded by the
hologram were not attempted during these feasibility experiments.
*
A holocamera for recording side scattering at Unit 10 was designed during
the course of the present work. It is presented and discussed in Section 4.
34
-------
RUBY LASER
PHOTODIODE
P-l
TELESCOPE
SCENE BEAM
5-INCH DIAMETER
REFERENCE BEAM
41-6"
~
SCATTERING SCENE
VOLUME
61 - 81
P-2
CYLINDER LENS
,
,
"
.,
HOLOGRAM
Fi gure 22.
Schematic diagram of holographic arrangement used to record
side scattering of ruby laser scene beam during laboratory
feas i bil i ty tes ts.
35
-------
3.2 SYNCHRONOUS DETECTOR
Although back and side scatter pulsed ruby laser holograms of tenuous
aerosols could be made under laboratory conditions~ the utilization of
this technique at the Shawnee Unit 10 ~oiler re~ain~d t~ be ~etermined.
It was desired to study limestone partlculate dlstrlbutlons ln the presence
of flyash, at elevation 376 of the boiler. A series of transmission
measurements at this elevation under various boiler operatinq conditions
would enable one to determine the potential and limits of the holographic
technique at this elevation. Unfortunately, opposed ~orts d~d not exist
at the 376 elevation and hence, no measure of the optlcal thlckness of
the scene volume could be easily obtained. As a result, it was decided
to build and test a helium-neon gas laser illuminated synchronous detection
system for measuring the power of the scattered laser light under various
geometry and boiler operating conditions.
The diagram in Figure 23 shows schematically the components of the
laser illuminated synchronous detection system which was constructed and
tested at elevation 376 of the Shawnee Unit 10 boiler. The illuminator
is a continuous wave Spectra Physics Model 124A helium-neon (He-Ne) gas
laser which emits approximately 18 milliwatts of visible red light at a
wavelength of 0.6328 micron. The light is collimated in a 1 .1mm diameter
beam. The d.c. light signal produced by the laser is converted to an a.c.
output with the aid of a chopper. The light chopper consists of an 1800
rpm synchronous motor drivinq a 25 tooth gear which intersects the laser
beam. The result is a 750 Hz square wave light signal. Light pulses
passinq between the gear teeth are directed into a scattering medium of
interest (i.e., particulate in the boiler combustion gases). Laser light
reflected by the gear teeth is incident on a photocell which generates a
750 Hz electrical reference signal for the lock-in amplifier (Figure 23).
Background flamelight in the boiler was sufficiently intense that
light scattered from the 18 mil1iwatt probing laser beam by particulate
entrained in the combustion gases could not be observed with the unaided
eye. To make scattered light measurements. a very large improvement in
signal-to-noise ratio was required. This was accomplished by several
means. The principal improvement came through use of the synchronous
detector system consisting of the light chopper and lock-in amplifier.
The lock-in amplifier, an EMC, Inc. Model RJB, comprised a resonant
signal amplifier, inverter, gating circuit and averaging circuit. The
six db/octave time constant T of the averaging circuit defines both
the rise time of the measurement and the equivalent bandwidth f = l/(TIT)
of the system, centered at the reference frequency of 750 Hz. A reference
frequency of 750 Hz was selected because this frequency is neither a
h~rmonic.of 60 Hz nor of 120 Hz and, therefore, should give good discrimina-
tlon agalnst these power-related noise frequencies.
The chopped laser beam scattering intensity is observed and measured
using a focusing telescope and photomultiplier tube (PMT). The telescope
and PMT assembly, mounted on a rotating azimuth head is shown in the
36
-------
LIGHT CHOPPER
(1800 RPM SYN. MOTOR
WITH 25 TOOTH GEAR)
W
'-J
POWE R
SUPPL Y
HARRISON 6515A
0- 1 000 V
SCENE
(SCATTERING) /-~....
_/ "-
---- ,
/ \
// \
/ ~ - I
I ./ I
1,__./ -_/
/ /
/1 /
\ I
/" '------/
~=-H£AT ABSORBING
, ~ ~~ (I.R.) GLASS FILTER
/ \:J NARROW BAND (.63281' )
y-" GLASS FILTER
0/ ~ 1 MM2 PINHOLE IN MIRROR FOR FOCU5"ING
EYEPIECE OF TELESCOPE
C-W He-Ne LASER
SILICON LIGHT
DETECTO R
(REFERENCE SIGNAL
GENERATOR)
PHOTOMUL TlPLIER
TUBE (l P2l)
A.C. COMPONENT OF
OPTICAL BACKGROUND
NOISE
CA
PLUG-IN
PRE - AMP
TEKTRONIX
535
OSc.
EMC, INC. MOD. RJB
LOCK-IN
AMPLIFIER
COHERENT
LIGHT
SIGNAL
BRUSH TWO -
CHANNEL
CHART RECORDER
REF. SIGNAL INPUT
Figure 23. Schematic diagram of laser illuminated synchronous detection system.
-------
Figure 24.
Photomultiplier tube and focusing telescope assembly mounted
on rotating azimuth table.
photograph of Figure 24. The device is capable of fine adjustment in
pointing, elevation and focus so that light scattered from a small volume
at considerable distance can be made to impinge on the active surface of
the lP21 phototube. The liqht gathering telescope consists of a single
achromat lens of 1-7/8 inch diameter and approximately 8 inch focal length.
The observer's eyepiece is a 1 inch focal length lens mounted to intersect
the telescope optical axis at the focal point of the objective lens. A
one mm area of the mirror coating was removed at the intersection of the
optical axis to form an aperture through which focused scattered light can
proceed to the PMT. With this arrangement, it is thus easy to observe
through the eyepiece when the detector is pointed at the desired target
region of the probing gas laser beam.
From the photograph of Figure 24, it will be seen that the telescope
and PMT housing are mounted in one assembly with a focusing tube arranged
to provide focus range from about 4 feet to infinity. The configuration
is essentially a "breadboard" design, but it proved entirely adequate for
field use at the Unit 10 boiler.
Referring again to Figure 23, the output of the PMT is amplified
initially with a Tektronix Model 535 oscilloscope and CA Plug-in pre-
amplifier unit. From the oscilloscope, the signal is fed into the lock-
in amplifier which selects and amplifies that component of the input
38
-------
*
signal which is in phase with the 750 Hz reference signal. A heat-
absorbing glass filter and a narrow-band dielectric interference filter
placed before the telescope objective lens provide some attenuation of the
background flamelight to enhance discrimination between the 0.6328 micron
laser scatter and blackbody radiation of the flame. Variation in scattering
intensity as a function of time is obtained by recording the coherent
light signal output of the lock-in amplifier with a Brush strip chart
recorder. Laboratory tests in the absence of background flamelight indicated
that thr full range of sensitivity for the synchronous detector system is
3 x 10- to 5 x 10-13 watts of chopped laser light arriving at the photo-
detector. The calibration procedure is described in Section 4.
* The output of the PMT could also be fed.directly into the ~ock-in
amplifier. Loss of oscilloscope power durlng tests at the b~lle~
necessitated this mode of operation with an attendant reduc~lon ln
signal amplification by a factor of ~10. The system was stlll adequate
for the Unit 10 measurements.
39
-------
4.
RES UL TS
The experimental portion of the program was divided into two dis-
tinct parts; namely, transmission and scattering measurements at the
Shawnee Unit 10 boiler, and several holographic tests designed to con-
firm the feasibility of recording pulsed ruby laser back scatter
and side (90 degree) scatter. This section of the report describes the
results of both efforts. The test setup for both the transmission and
scattering measurements at the Unit 10 boiler are shown schematically
for reference purposes. The Unit 10 measurements and the laboratory
holographic tests led to design of a three-beam holocamera for use at
elevation 376 of the boiler. This design is presented and discussed
as a part of the results.
4.1
UNIT 10 TRANSMISSION MEASUREMENTS
Laser light transmission measurements were made across the 50 foot
wide Unit 10 superheater region at elevation 392. A schematic of the
test setup at this elevation is shown in Figure 25. The apparatus used
for transmission measurements was essentially the same as that described
in Section 3.3 for the Synchronous Detector with but one exception.
The focusing telescope and photomultiplier tube assembly was not used;
instead, the laser beam was made to impinge directly on a silicon diode
detector. A one cm2 silicon diode was mounted at the focal point of a
4 inch diameter, 6 inch focal length lens. The lens and diode together
with a heat absorbing (infra-red) glass filter and Wratten No. 70 red
filter were assembled in an aluminum housing mounted on a tripod. The
signal amplifying equipment and the chopped laser light illuminator
were the same as described previously in Section 3.3.
CHOPPER
UNIT 10 SUPERHEATER
ELEV. 392
He-Ne LASER
..
..
..
REF. SIGNAL
48 FT
L.I.A.
CHART
RECORDER
Figure 25.
Schematic diagram of setup for transmission measurements
at elevation 392 of the Unit 10 superheater section.
40
-------
All of the transmission measurements were made in the absence of
any pulverized limestone injection into the boiler. Unit 10 was operating
at an electrical output of 127 megawatts during the transmission test
period. At this output, the average coal feed rate was 109.6 x 103 lb-
coal/hr with a corresponding air flow of 1 x 109 lb/hr.
Four data points were reduced from the Unit 10 superheater strip
chart records. These data points were selected to obtain values of
minimum and maximum light transmittance. The data are tabulated in
the following table.
Table I. Transmission Measurements
Unit 10 Superheater, Elevation 392 Ft.
16 March 1971
Data Power Received Power Output Fraction Recei ved
Point P in watts P in watts P/P
0 0
Mi nimum 2. 1 x 10-6 .023 -5
9.13 x 10
Maximum 6.8 x 10-6 .023 -4
2.96 x 10
Minimum 2.2 x 10-6 .023 9.56 x 10-5
Maximum 6.0 x 10-6 .023 2.61 x 10-4
From these data, it will be seen that the He-Ne laser li~ht beam
attenuation varied between approximately 1 x 10-4 and 3 x 10-4 during
the measurement period. The fraction P/Po' which is the ratio of laser light
received across the superheater, to the incident laser light, may be used
to calculate values of the average particle number density n for each of
the data points in the preceeding table. This is accomplished using
Eq . ( 2 ) .
whe re ,
T
e
n
0"
L
T = e-2nO"L
is the transmittance = P/P
o
is the base of the natural system of logarithms
is the average number density of particles
is the average physical cross-section
(1.26 x 10-5 cm)
is the optical path length (48 ft)
of the particles
41
-------
One may also calculate values of "mfp" from the transmission measure-
ments in the superheater. These are useful for comparison with similar
calculations made from the scattering data obtained at elevation 376,
where straight path transmission measurem~nts co~ld not be made. .In
other words no direct measure of the optlcal thlckness of the bOller
gases could'be determined at the lower eval~ation thus ~ntroducing
uncertainties into the particle number denslty calculatlons.
Following is a tabulation of average particle densities
and mfp lengths derived from values of PIPo in Table I using Eq. (2) and
mfp = 1/2na. The average of the four measurements gives mfp = 66.1 inches
Table II. Values of nand mfp for
Transmission Data, Unit 10 Superheater
Elevation 392 ft., 16 March 1971
Data PI Po Partic1e Number Dens ity mfp
Point mlcm n/in3 Feet Inches
Minimum 9.13xlO-5 281 4612 5.16 61. 9
Maximum 2.96 x 10-4 246 4029 5.91 70.9
Mi nimum -5 280 4589 5.19 62.2
9.56 x 10
Maximum -4 250 4091 5.82 69.8
2.61 x 10
4.2
LIGHT SCATTERING EXPERIMENTS AT UNIT 10
In Figure 26 is a schematic diagram of the light scattering experi-
ments performed at Unit 10 of Shawnee Power Plant, 17 and 18 March 1971.
The sketch shows interconnections of equipment and the relative positions
of viewing ports in the boiler wall at Plane A-A (Elevation 376). The
chopped beam from the helium-neon laser was made parallel to the boiler
wall at the same elevation as the centers of the viewing ports. It was
then directed into one or another of the ports by a pentaprism (which
has the property of maintaining a precise 90-degree angle of reflection
regardless of small errors in prism rotation). The scattered light
signal was measured usinq the synchronous detector described in Section
3.3. Two recording channels were used for recording of data. One
channel was used to record an indication of flame light (total light)
and one was used for the coherent detector output signal.
The first attempts to record scattered laser light were made by
viewing along the inside wall of the furnace. These were unsuccessful,
until it was realized that the fresh-air draft drawn into the open port
was blowing all the scattering centers out of the field of view.
Accordingly, a port cover was made with a 5/16" diameter hole near the
center for the input laser beam to pass through. This was of thin sheet
42
-------
metal, and was customarily held in place by the differential pressure
between the inside and outside of the furnace. Inasmuch as the laser
beam w?s not visible to the ey~ in the presence of flamelight, it was
occasslonally advantageous to lnsert a 1/4" diameter stainless steel
tube through the hole in the port cover for preliminary aiming of the
telescope.
N+
UNIT 10
ELEV 376
~ LASER INPUT BEAM! VIEWING PORT
: TELESCOPE LINE OF SIGH~
1-- -- --
L////d v///d!r////j L/////A
PORTS - No.2 No.3: No.4 No.5
---P-HOTO-OIO-DE------- --- - - - - --jt790° PRISM LOCK-IN AMPLIFIER
IN
REF. SIG.
OUT
PMT
POWER SUPPLY
TELESCOPE
BRUSH
STRIP CHART
RECORDER
Figure 26.
Schematic diagram of setup for scattering measurements
at elevation 376 of Unit 10 boiler.
The calibration procedure used for this setup is described in the
section immediately following. Measured results are reported and
discussed in Section 4.2.
Calibration of Scattered Light Detector
The reference standard for power in the helium-neon laser beam is
a type 401 Power Meter*, which has a useful range of 0.1 to 100. milli-
watts. This standard is applied to the synchronous detector system as
follows: The laser beam is made to fall on the surface of a magnesium
carbonate block placed at some convenient distance, say 20 feet, from
the viewing telescope. The intensity of the beam is measured with the
* Spectra-Physics, Inc.
43
-------
power meter at the location of the block with the chopper turned off. In-
asmuch as the diffuse reflectivity of magnesium carbonate is about 0.99
for light of this wavelength, it is assumed that the incident laser beam
power is diffusely reflected with a consine-normal distribution, i.e.,
I ( e) = £ cos e
TI
(watts per steradian)
(7)
where p is the power incident on the block and e is the polar angle from
the block normal. Now, if the angle between the telescope axis and the
block normal is small, power into the telescope entrance pupil is
Pt =
2 2
pTId - ~
TI4R2 - R2
(watts)
(8)
where d = 2r is the diameter of the entrance pupil and R is the distance
from the reflective block to the telescope.
With about 20 milliwatts of laser light incident on the magnesium
carbonate block at about 20 feet from the telescope, a light signal
is received which is several orders of magnitude stronger than scattered
light signals from the experiment, but is well within the linear range
of the photomultiplier tube. (This was checked over 4 orders of mag-
nitude using neutral-density filters). Accordingly the very accurate
attenuators in the pre-amplifier and the lock-in amplifier can be used
to obtain a reading which corresponds to a known power arriving at the
telescope.
The calibration reading thus obtained defines a calibration con-
stant for the system which may be used during subsequent data reduction
for determining the scattering ratio (more specifically, the trans-
mittance via a particular scattering path) for a particular measurement.
This is done as follows: Given a reading p from the lock-in amplifier
(p goes from 0 to 1; 1 = full-scale on the meter) and an attenuation
ratio (attenuator setting) a (1 ~ a ~ 10°), the calibration constant is
Pt
c =-
pa
and represents the power incident on the telescope represented by a
full-scale reading with maximum sensitivity. Note that the calibration
power level is calculated from the de power of the laser, while the
calibra~ion constant c is ,derived from the chopped light signal. The
conve~slon f~o~ a ~rapezold~l wave to de is thus accounted for, and the
lock-ln ampl~fler lS effectlvely a "dc" power meter having a range of
about 1 x 10 13 to 5 x 10-6 watts. A measurement pi is an indication
of laser light power scattered into the telescope during the half-
c~cle of the chopper period when the light is on, and is numerically
glven by
(watts)
(9)
pi = pac
(watts)
(10)
44
-------
The calibration constant thus obtained is valid for the system as
~ong as optics remain fairly clean and the voltage on the photomultiplier
1S held constant. Values of c measured at Shawnee were:
=
1.145 x 10-12 watts
7.7 x 10-11 watts
(with pre-amp, Vpmt = 600 volts)
(without pre-amp, V t = 700 volts)
pm
cl
c2 =
Scattered Light Measurements
Measurements were made of laser light scattered from 9 locations
within the Unit 10 boiler at Pl~ne A-A lelevation 376). The locations
are shown in Figure 27. Of the 9 measurement locations, scattering from
flyash only was measured at 5 points. Of these 5 points, 4 were also
measured with limestone injection. In addition, measurements were made
at 4 other points during limestone injection.
The 9 points are identified by numbers in Figure 27, and are also
characterized by the intersections of laser beams injected through
ports 2, 3, 4 and 5 with viewing telescope axis poisitions at scattering
angles of 90, 115 and 125 degrees as shown.
Associated with each of the 9 scattering positions are two attenu-
ation distances, £ and £ , together with a geometric distance R. The
distance £, is th~ dista~ce inside the boiler to the scattering volume
element. FOr the case of 90 degree scatter, £ is essentially zero
since the probing gas laser beam was parallel io the inside surface
of the boiler tube wall and there was a separation of only a few inches
between the wall and the beam. The distance £2 is that distance from
the scattering volume element to the inside edge of the exit (viewing)
port of the boiler. The geometric distance R is the total distance
from the scattering volume to the telescope objective lens located
outside the boiler. It is used in calculating the steradian correction
to the measurements.* Values for tl' £2 and R are given in Table III.
The fourth column in Table III is the factor Wa in Equation 6, and
is the effective length of laser beam in the element of scattering
volume from which light is received. The values of transmittance
listed in columns 5 and 6 were calculated using Equation 6 and represent
the transmittance with flyash only, and with flyash-limes~one mixt~res.
respectivelyJ In addition, values of the ratio of.transm1ttance w1th
limestone to transmittance without limestone are llsted for the four
points where scattering measurements were obtained for both conditions.
* The diameter of the telescope entrance pupil is 1.75 inches.
l' Transmittances shown in Table III are in some cases, mean values
of several data points from the strip chart records.
45
-------
N~
KEY
. 1 POINT NO LIMESTONE
. 4 POINTS WITH LIMESTONE
. 4 POINTS WITH AND WITHOUT LIMESTONE
TOTAL = 9 POINTS OF MEASUREMENT
4
3711-\ TELESCOPE
ENTRANCE
I PUPIL
8 = 900
UNIT 10
HEV 376
.p.
m
3
2
r-,."""""",,'\) I r-.."""""""""""",' I r-,."""""""""",,'\) I 1.""""""""""","1 I ~
PORT 2 PORT 3 PORT 4 PORT 5
~39" .~39" .~39"-+19.5"
8 = 1150
8 = 1250
Fi gure 27.
Relative locations from which scattered light measurements were made.
-------
Table HI.
Tabulation of results of solving Equation 6 for transmittance
with and without 1 imestone injection
point 9,1 ~2 R Wa Transmittance Transmittance Transmittance
No. Inches Inches lnches lnches Plyash only Flyash + limestone Ratio
O. 19.5 56.5 .0799 1.30 x 10-8 1.96 x 10-8 1.3
2 O. 58.5 95.5 .2598 6.0 x 10-9 9.89 x 10-9 1. 65
3 O. 97.5 134.5 .3858 3.0 x 10-9
4 O. 136.5 173.5 .5039 1.0 x 10..9
5 9.09 21 .52 62.34 .1096 1 .0 x 10-8
+::> 5.64 x 10-9
........ 6 27.28 64.55 105.37 .3228
7 45.47 107.58 148.4 .474 1.52 x 10-9
68.97 .1432 -9 1 .2 x 10-8
8 13.65 23.81 6.0 x 10 2.0
116.59 .3975 -9 7.2 x 10-9 2.4
9 40.96 71.42 3.0 x 10
-------
A review of the Table III data shows that for the four points from
which measurements were made both with and without limestone injection,
more light was received from the limestone-flyash mixture than from fly-
ash alone. There was, however, a significant increase in fluct~ations
of light intensity during the limestone injection measurements.
An attempt to solve Equation 6 directly for particle number density
ran into difficulties due in part to the uncertainty of the value of the
scattering fraction S, and to the fact that a given value of the concen-
tration function C does not yield a unique value of na. Accordinaly,
a "mean free path" (mfp) value of 66 inches based upon transmi ss i on
measurements across the Unit 10 superheater was used with the experi-
mental data obtained at Elevation 376 to solve for S. The results of
these calculations are presented in Table IV.
Calculated values for the scatterinq fraction S in Table IV are
given for measurements with flyash only (Sf) and for those taken while
limestone dust was.being injected into the furnace (Sfl). The values
of Sf are not unreasonable, and tend to indicate some consistency in
the aata. The mean value is 0.358 with an uncertainty of about 30 per-
cent. For the limestone numbers, the mean (Sfl) is 0.571 with an un-
certainty of about 60 percent, indicating either that the limestone has
a different scattering fraction from the flyash, or that the assumed
"mfp" was wrong.
Inasmuch as the limestone measurements were very noisy, a good
probability is that one value of mfp is not correct for the entire set
of measurements. For example, Table V shows the result of repeating
the calculation with an input value of mfp = 55 inches. In this case,
the uncertainty in Sf values is about the same, while the value of Sfi
for point 1 is improving, it is still unreasonably high. Further, the
value of Sfl for point 9 says that more light is being scattered from
the intersection volume than is beinq received from the laser. Hence,
the results are still unsatisfactory.
Another attempt was made to derive the value of na directly from
the scattering measurements. For each set of measurements (with and
without limestone), comparisons were made between pairs of measurements.
In the ratio of two given transmission measurements, some of the factors
in the basic equation cancel out (in particular, the quantity S).
That is,
T = r 1 w 1 dl Sf(900) f(e) 2nae-2naLl
1 8n f(900)sine
T = r 2w2d2 Sf(900) f(e) 2nae-2naL2
2 8n f(900)sine
* s~~ discussion following.
48
-------
Table IV
Calculations of the Scattering Fraction ~
Assuming mfp - 66 inches
Point /3f ;Jf
No.
1 .481 .725
2 .352 .581
3 0.000 .425
4 .326 0.000
5 0.000 .389
6 0.000 .538
7 0.000 .495
8 .242 .485
9 .388 .932
Table V
Calculations of the Scattering Fraction ~
Assuming mfp = 55 inches
Point jJf ,/if
No.
1 .425 .641
2 .351 .578
3 0.000 .476
4 .411 0.000
5 0.000 .356
6 0.000 .592
7 0.000 .656
8 .226 .453
9 .455 1.092
49
-------
whence
1 -
mfp = 2nq -
1 n
Ll L2
( T 2/ A2)
T 1/ Al
(11 )
where
r'w. QI.
A = 111
i 8TI
The results of this calculation are shown in Table VI. Here, it seems
that there are a number of results which tend to cluster about some
value, and a few wild results. This can be accounted for partly as
fo 11 ows : Equa ti on 11 ass umes that the values of n () for the two meas ure-
ments were the same, and therefore cancel. If this is not the case, a
"wild" result can occur. If one arbitrarily ignores resultant values
above 100 inches in maqni tude, the two sets of res ul ts i ndi cate a "mean
free path" of about 54:6 and 54.2 inches respectively, with an uncer-
tainty of about 33 percent. If anythin9, this tends to indicate that
the total particulate loading (in terms of extinction coefficient per
unit length) is not strongly changed during the addition of limestone
dust; hence, the difference between "measured" values of 13, with and
withoud limestone (Tables IV and V), may indeed be due to differences
between scattering properties of the flyash and limestone particles.
One must regard such conclusions as tentative, in view of the sparseness
of data and the simplified character of the present model.
Dynamic Behavior of Transmitted and Scattered Light Data
It was not an original objective of the measurements at the Shawnee
Power Plant to perform a frequency-spectrum analysis of the number
densities in the boiler. Nevertheless, some comments may be made on
the dynamic behavior of the data, and four figures are included here
as basis for such remarks. The first sample, Figure 28 is a record of
the transmission measurement (no limestone injection) in the superheater
region made on 16 March 1971. It is a recording of the output of the
lock-in amplifier. At the left of the figure, the transmittance is
fairly steady for a minute or so, then there are four cyclic variations
during which the transmittance varies by almost a factor of two. The
period of these variations is 30 to 35 seconds. During the course of
the measurements, this phenomenon was seen to repeat itself at intervals
of a few minutes. Similar cyclic bursts were later noticed on chart
records in the boiler control room, suggesting that these slow oscil-
lations may be related to the operation of some automatic control system
within the boiler. For this record, the lock-in amplifier time constant
was set at 0.003 second, giving a IIhalf-powerll upper cut-off frequency
of about 50 Hz. Frequency components covering the range of 0.1 to
about 10 Hz can be observed in the trace.
50
-------
Table VI
Calculations of the Scattering Fr~ction
Using the "Two-Point" Method
Fl y as h Only Flyash and Limestone
Point mfp Point mfp
Numbers Inches Numbers Inches
1, 2 43.223 1, 2 47.963
1, 4 54. 1 06 1, 3 45.440
1, 8 18.758 1, 5 14.027
1, 9 57.270 1, 6 51.843
2, 4 61.899 1, 7 55.512
2, 8 - 381. 707 1, 8 26.611
2, 9 74.888 1, 9 80.284
4, 8 82.192 2, 3 43. 168
4, 9 44.615 2, 5 1320.483
8, 9 112.775 2, 6 57.262
2, 7 59.367
2, 8 152. 185
2, 9 156.740
3, 5 72 .349
3, 6 17.650
3, 7 80.604
3, 8 57.637
3, 9 -26.581
5, 6 101.501
5. 7 75.874
5, 8 -58.478
5, 9 224.110
6, 7 60.579
6, 8 75.483
* Equati on 11 6, 9 -86.253
7, 8 66. 781
7, 9 32.564
8, 9 155.434
51
-------
L.I.A. TIME CON STANT = .003
<.Tl
N
Figure 28.
Strip chart record of laser light transmission across Unit 10
superheater at elevation 392, on 16 March 1971. Measurements
were made in the absence of limestone injection.
-------
An interesting record was made at the end of the experimental period
(18 Ma~ch 1971)~ as. the limestone injection was being shut off. This
trace.ls shown ln Flgure 29. The upper trace is lock-in amplifier out-
put wlth zero at 2 major divisions from the bottom of the record. The
lower trace is proportional to total light received by the phototube,
z~ro at. the top of. the record. Chart speed is the same as for the pre-
v~ous flgure. It lS easy to notice, at least qualitatively, a significant
dlfference between traces with and without limestone injection, both in
amplitude and "noisiness" of the total-light and scattered-light signals.
Figure 30 is a trace made with laser light scattered from point 9
(See Figure 27) with limestone injection. For this trace, the laser
beam was shuttered off, allowing the lock-in amplifier response to decay
to zero. After 20 to 30 seconds it was turned on again. This was done
with three different settings of the lock-in amplifier time constant in
order to demonstrate the effect of this parameter on the character of the
traces. With a 3 second time constant, the trace is reasonably smooth,
but the decay time prevents one from obtaining a reading until about 20
seconds after a steo change in input. With a time constant of 0.3 second,
the circuit responds more rapidly, but noise in the trace leads to an
uncertainty in the magnitude of the signal which can only be removed by
taking an "eyeball average" of the signal over the subsequent period of
about 20 seconds. An intermediate time constant of 0.1 second is also
shown, with a similar result.
The final sample of chart records, Figure 31, was made during
recording of scattered light from point 6 with limestone injection. For
this trace, the recorder was allowed to run at maximum chart speed. The
lock-in amplifier time constant was 0.3 second (upper cut-off about 0.5
Hz)~ nevertheless, frequency components up to about 70 Hz or so can be
observed in the trace. Two other frequency components seem to predom-
inate; one at about 8 - 9 Hz, and another between 16 - 20 Hz. This was
about the noisiest trace recorded during the limestone injection.
No attempts were made to correlate these traces with any aspects
of the limestone injection process. It seems likely, however, that the
large amount of flickering of light intensity is an indication that the
injection process does not result in anything like a uniform limestone
density distribution. Nor does it result in a stationary distribution
which could be described in a meaningful way by measurements made at one
instant. This fact alone suggests that considerable improvement in the
limestone process could be achieved through improvement in injector
nozzle design.
53
-------
~-L-
_i --+--t-:_~~~_Lj .. :, . --:=::- -:-~- . r-~!,--r+ -- r -r \' l
'. \' '\., \ --~_r -"~ \ . \', "rn\-\_\_, '1 '
\ --~- \=:;'. TL_\\.--,~_\-- \ ,i_I \- 'L_L_L--\, u'L
- ~ . -~__~~~E ":Z ,LT~r'il'-!~Oi;).=:i X":?;f~/-~
!
LIMESTONE OFF
3:02 P.M.
3-18-71
WATTS X 10-12
----,-- r-, --r-;--------r--T-,-j----r-T-T--.' .
i . ::, /!:!! / I !
. ,r t- 300);,;. T' fn-~- Ti-r--j- 1-- J
I iI"I.J i I 1 . I I
~.,' , I:
- I ; - ~'ooi ! I 1 ; i ; I : i
T-t l'OO~I' 'i: i :----
, ~1 I Ii: \ I i
-t-till! -\ -- \ - _n~
-++-+ 0 -r-n- -
-1- j j \ \--~\- + - Lt--\--~_. \
\ \ ' . , \ \ \ \ \
- - \ WATT S .l----L.....L ---L- L -
--;------r- 01/-7------; I ' J ---:Ti - ,
I , , i ' ; i I / I ' '
-- -'I---f --j-- -I-- +-r--r--T--r--:--+--L. 7
, 1 001- i I i I I ! i I '- .
--:--tlli!/lli;
I ; , , I ' , I, I
-,--+' - 200L_-L--...l, - -- '--.:"--+---1
~. ~ ": ;
,
\
-----------
-~--------'---- -
L.I.A. TIME CONSTANT = 1.0
1 MINUTE
-- . --- ~ ---,--
, ' ,
-- T ;
"~';"--~I,!~-"--~~~
- _J - i. ~ - ~. - . I ' !
; . , - -------r-------- - -
, ! :
..:. ;..J.-
. \ l ,
, . . . .
~ -~---1---~-- - -~--- - -:--- --;- --;----:T--- ~ - - ~--- T-----;-- -~-
j l \ "
-+- \ " ---c " ,,'
'I \ \ \ \ ; \ \ \, \ ',t \ \ " .
-!--_\---t_t~\=~I~ t~----~--'C\L~_\:-\'_:I". ,
Figure 29.
Strip chart record of 90 degree laser light scatter using port #5 at elevation
376, on 18 March 1971. Lower trace is proportional to total light received by
PMT and thus related to combustion flamelight. Upper trace represents coherent
signal from lock-in amplifier due to particulate scatter of the chopped laser
input beam. The traces show the reduction in both total light and scattered
light intensity as the limestone injection system was turned off.
-------
WATTS X 10-10
L. I. A.
TIME CONSTANT
= 3.0
L.I.A.
TIME CONSTANT
= 0.3
L. I. A. TIME CONSTANT = 1.0
Ul
Ul
Figure 30.
Strip chart record of 125 degree laser light scatter using port #4 at elevation
376, on 18 March 1971. The strip chart trace is coherent scattering signal from
point 9 (see Figure 27) during limestone injection and shows the effect of
different time constant settings on the lock-in amplifier.
-------
L.I.A.
TIME CONSTANT = 0.3
, '
H~: ,~m,m "'."
I 1 .' ! ' ,
~ ~,'~ ':~r T -in -:~;"
1 SEC
-.1
--:~+-~~--"
-< ,\ L c;.=: -:
~ ,
, ~
I
!
,
J
I
l
I
I
!
I
I
J
i
I
I
!
I
'] , I I ~. T f' I I I '(--TIMET' :..--r-T~T-r'~r-~i7-r'--i-T-fTi-
----,.!-.- ~"'--~IL__+,"'~-TI! - --~I' ~--,l'-'-"i---- -~I: _-""""1"",I~--J--11l_'~--IJ---,-,!~--~I, "IL--1(L-+I'~--J,1
,~ -,Il'~-",,! I~,F-~,'",L~~"j--,f ~,',/ -~"t-:-"/-=-,,,L~,',,/-~L-,~'r,L~,(=_.;p'r, , I
," .. . - -. : .,,1 ,r / ! " / ~ / . / / . ! I ,I I J ,
; J : ' i : I I I "" I:' ~----~t - --_.J-, '-~,' 'I _._~-, ---T-,--~~-----4--------L_-,......-~--t---,~ -~,: --7,'-
I ' , ,i " : I : ,,/
I " : ; : : [
I '--r-------;--~-+----o- : I Ii,: ;:!,;
, , T\ -.t ,+--~u-i-i- "-1--: u'l~ ~ . "I-~-~~ 'T-t-~;' - \--t--t-.;_.\--t'i-t--r-t
\ \ '} 1 I \' " \ \ \ \\' \ '\ \ ~ ': 'I '~-T--+
\ i \ \\\ " \ " \ \ \~- : \ , \ \ \ \. \ \ \ \ \ \ \ \
' i \ t-+-\ ---\-.-t-~--\-~- \-uT' ~\,. L._+.--:. - ~- "," "\"-'~'\'\' ",\"_I'L' ~, .',',"'" ' ;,:,-",T,\-n"\',_\,-,.','r\.,.~ ,\~'I,--1,\-, \,,'-" "-,,. ',',~,\, -- t"" "~\,~-l.-,,-,\~., -,.-",',',"..\
_Ll_L--Ll__l.l.~j-_\_j--\ L--\.:L.\. ~ . ~'- L L ,L n J. ~ ,~," -,- , .
U1
(j)
\
\
J
I :
-~
1 \ ;
-l-.\-. \
, ' \
,~ 1
Fi gure 31.
Strip chart record of 115 degree laser light scatter using port #3 at elevation
376, on 18 March 1971. In this example, the strip chart speed was a maximum at
125 mmjsec, with lock-in amplifier time constant at 0.3. Upper trace is coherent
scattering signal. Lower trace is proportional to total light received by the PMT.
-------
Feasibility of Holography of Light Scattered at Large Angles
~ principal objective of the experiments reported here was to
~etermlne whether meaningful holograms could be made at the same plane
ln the Shawnee Unit 10 boiler as was used in the scattered light measure-
ments. Two important aspects of this question are considered here - the
amount of light available in the holographic scene beam, and the possible
effects of particle motion. These considerations indicate that holo-
graphy may be possible at scene depths comparable to those of the present
measurements. A tentative design for a holocamera setup is presented
and discussed briefly. However, it is not considered possible to make
a hologram which will cover the entire distance from the viewing port
to a far wall.
The first aspect of the holographic process which can be examined
using the scattered light data is the amount of light available in the
scene beam. With Agfa 8E75 plates, a reference beam intensity of about
20 microjoules per square centimeter is typically used in order to
achieve near-optimum photographic density. The required scene beam
intensity may be stated in terms of a minimum scene-to-reference beam
intensity ratio. A series of holograms was recently made at TRW to
determine the limiting value of this quantity.7 Two reference beams
of equal intensity, and a scene of adjustable intensity were made to
fall on the holographic plates. The beams were separated by convenient
angles, and, from the viewpoint of the hologram, were derived from point
sources of light. Scene to reference beam intensity ratios ranged from
10-4 to 10-7. At 10-6, the faint point of light was barely visible to
an observer viewing the reconstructed image. The point at 10-7 intensity
ratio was not visible to the eye, but could be detected with the aid of
a telescope (using the light gathering ability of the larger lens).
Data from point 7 (see Table III and Figure 27) can be used as
follows: The telescope entrance pupil has an area of about 15.5 cm2,
and the transmittance from point 7 was 1.52 x 10-9 with limestone.
Accordingly, if the holographic 1as~r pulse em~t~ ~ 4ou1e of energy into
the scattering volume, and one conslders the V1Slbl11ty of the same
scattering volume ele~ent as .in.the present experiment, the scene-to-
reference beam intenslty ratlo lS
-9
1. x 1.52(10)
-6
15.5 x 20(10)
=
-6
4.5 x 10
which is near the lower limit of the range.
the visibility of one small volume element;
show scattered light from a great many such
feet of the incident laser beam.
This estimate considers only
hopefully, the hologram will
elements covering several
57
-------
The other aspect of the problem is the possible effects of particle
velocity on visibility in the holographic image. A commonly used
criterion says that to make a good hologram, the object should not move
during the laser pulse such as to cause a path length change.of m~re ~han
1/10 wavelength in the scene beam. Slight departure from thlS crlterlon
will not necessarily cause complete failure of the holographic process,
although unpredictable changes in visibility may result. Without g~ing
through all the details of the analysis. one may observe that the dlrec-
tion of particle velocity is of crucial importance, and that a velocity
component greater than about 3 meters per second in an adverse direction
may make a scattering particle invisible in the hologram. However, if
the velocity components greater than this magnitude are in a direction
perpendicular to the plane containing both the incident and received
beams, particle visibility may not be impaired. We have the favorable
result, also, that scattered liqht holograms have been successfully made
in the Unit 10 boiler. However, we note that plane A-A has a narrower
cross section than the plane at which holograms were previously made,
and that curvatures in the flow (due to flow around the "nose" of the
superheater) and turbulence could cause erroneous variations in image
intensity.
The schematic diagram of Figure 32 shows a proposed holocamera
design for recording side scatter of a ruby laser beam by particulate
in the Shawnee Unit 10 boiler at elevation 376. Arrangement of the
holocamera optics makes use of existing boiler ports at this elevation.
The incident ruby scene beam enters the boiler through one of the
existing four ports on the west side of the boiler. The incident beam
is shown as an unexpanded collimated beam; however, the addition of a
cylinder lens at the entrance to the boiler port would permit the
fanning of the incident beam within the geometric limitations of the
port.* Scattering of the incident beam is received by the hologram
through the large port on the southwest corner of the boiler.
The reference beam is formed using a wedge and telescope arrange-
ment. The reference beam is expanded and collimated to a 5-inch dia-
meter, and directed parallel to the outside of the west wall of the
boiler. A large wedge divides the primary reference beam into two
components. Front surface mirrors are used to recombine the two
reference beam components at the plane of the holographic plate. The
result is a three-beam holocamera similar to that demonstrated in the
laboratory.2 The addition of the second reference beam enables the
reconstruction process to be linearized and provides the sensitivity
necessary to record very weak scene intensities.
* Refer to the results of the low angle forward scattering holograms
recorded previously at elevation 365 for examples of recordings with
fanned scene beams (Reference 1).
58
-------
SCATTERING FLYASH
.6943/1
50 X 10-9 SEC
~
MIRROR
N+
RUBY LASER INPUT BEAM
UNIT 10
ELEV 376
U1
<.0
TYPICAL BO ILER PORT "'-
/ / / / / / / / / / / / 1
VIEWI NG PORT-----.......
HOLOGRAM
MIRROR
WE DGE
MIRROR
RUBY LASER
Fi gure 32.
Schematic diagram of proposed three-beam ruby laser ho1ocamera for use at the
Shawnee Unit 10 boiler to record side scattering. With modification~ same
apparatus could be adapted to record back scatter.
-------
5.
CONCLUSIONS
From the scattering analysis, laboratory tests and measurements made at
the Shawnee Unit 10 boiler, the following conclusions have been reached
concerning each of the following topics:
Holography of Scattered Light
.
The holography of back scatter and side scatter is feasible.
Recordings were made in the laboratory of static scenes and
aerosols to demonstrate this feasibility.
The holography of side scatter and back scatter by particu-
lates in the boiler is rossible. This is based upon scat-
tering measurements made with the synchronous detector at
elevation 376 in the boiler. Demonstration of feasibility
remains to be verified by test at the boiler. Uncertainties
with respect to particle velocity components and actual
scene-reference ratios prevent a definite conclusion by
analysis only.
.
~
Of the two alternative methods considered durinq the present
work, side scatter holography and back scatter holography,
the latter appears to offer the most promise. This is based
on the fact that laboratory tests indicated a factor of ~20
increase in the amount of liqht scattered in the backward
direction, over that scattered in the 90 degree to 125 degree
side scatterinq directions.
.
Both the side scattering and back scattering techniques suffer
limitations with regard to particle velocity components due
to turbulence, etc. Velocity limitations for both techniques
are about the same (around 1 to 3 meters/see).
.
Based upon all of the past and present scattered light holo-
graphy studies, the forward scattering method is by far the
optimum method for visualizing particle distributions and
acquiring particle density data. In addition to being pre-
viously demonstrated at the boiler, the holography of low
angle forward scatter has two distinct advantages; namely,
it is relatively insensitive to particle velocities, and
more favorable scene-to-reference beam ratios result since
particle scatterinq intensity is predominantly in the forward
direction.
Acquisition of Particle Density from Scattering Measurements
.
Straight path transmission measurements at elevation 392
of the superheater indicated that particle densities ranged
between 4000 and 4600 flyash particles/in~ in the absence
of limestone dust injection into the boiler.
60
-------
.
Comparative tests at elevation 376 showed that there was an
increase in scattered light intensity of between 1.3 and 2.4
during periods of limestone dust injection. Strip chart
traces exhibited considerable amounts of "noise".
.
Direct solution of the particle scattering model proved
impractical. This was due to uncertainties concerning the
scattering fraction S, and the fact that qiven values
of the concentrati on functi on (2ncr Le-2ncr L) di d not res ul t
in unique values of no.
.
The alternate "two point" method of derivinq no from the
scattering measurements did not demonstrate a sensitivity
to increases in particle density due to the presence of
limestone dust. Tentatively, it is concluded that the
attenuation coefficient per unit length (2no) is not signifi-
cantly changed at Plane A-A. Stated anouther way, changes
in the attenuation coefficient are within the "noise level"
of the scattering intensity measurements.
.
A partial reason for the apparent lack of change in the
attenuation coefficient may be due to the particle en-
capsulation phenomenon noted durinq the scanninq electron
beam microscopy studies. - -
The concentration function y = xe-x places a limit on
the distance into the boiler particle flow field which
may be interrogated via back scattering measurements.
At Plane A-A, the characteristic distance was approximately
6 feet inside the boiler volume.
.
Particle Flow Field Dynamic Characteristics.
Limestone particle distribution at Plane A-A does not
appear spatially uniform nor is the distribution constant
or repetitive with time, based upon the apparently random
intensity and frequency variations of the strip chart
records.
.
.
The frequency characteristics of the transmitted and scat-
tered light strip chart traces vary randomly and range
from about 0.1 to 10 Hz in the absence of limestone in-
jection.
During limestone injection, the degree of "noise" in-
creased. Frequency components up to about 70,Hz were
observed. Two frequencies appeared to predomlnate; one
at 8 - 9 Hz, and another at between 16 and 20 Hz.
.
61
-------
.
Occasionally, definite repetitive cyclical variations did
occur. In one example, the period of these variations
was approximately 30 to 35 seconds. It was tentatively
concluded that these slow oscillations correlate with
some automatic control system associated with the operation
of the boiler.
62
-------
6.
RECOMMENDATIONS
It is recommended that the method of low angle forward scatter
holography be utilized to investigate and optimize limestone dust
distribution in the Shawnee Unit 10 boiler. A laser-illuminated
synchronous detector would be used in conjunction with holography to
provide a continuous monitor of localized dust concentrations. The
synchronous detector would be a part of the holocamera setup and
provide complimentary data to the instantaneous holographic records.
The output of the detector should also be integrated into the overall
data acquisition system for the dry limestone injection apparatus
and boiler operation.
It is recognized that the utilization of forward scattering
(transmission) holography will necessitate cutting additional ports
in the boiler; however, the capability of making in situ detailed
measurements (instantaneous and continuous real time) of particle
number density distributions as a function of limestone injection
variables would aid in the development of the dry limestone system.
63
-------
REFERENCES
1.
B. J. Matthews and R. F. Kemp, "Holographic Determination of
Injected Limestone Distribution in Unit 10 of the Shawnee Power
Plant," TRW Report No. 14103-6001-RO-00, June 1970.
B. J. Matthews and R. F. Kemp, "Investigation of Scattered Light
Holography of Aerosols and Data Reduction Techniques," TRW Report
No. 14103-6002-RO-00, November 1970
2.
3.
Private communication from D. T. Clay, 17 March 1970. See
Tennessee Valley Authority Report No. 54 covering particle
analysis of limestone and flyash during period of 9 July -
of 1969.
a 1 so
size
6 Augus t
4.
G. H. McClellan, "Physical Characteristics of Calcined and Sul-
fated Limestones," Paper presented at the NAPCA Symposium on
the Dry Limestone Injection Process, Gilbertsville, Kentucky,
June 1970.
5.
Walter C. McCrone in: Air Pollution, Vol. II, by Arthur C.
Stern (Academic Press, New York) 1968.
6.
R. F. Wuerker and L. O. Heflinger, "Laser Holography Study,"
Interim Report, Air Force Avionics Laboratory Report No. AFAL-
TR-70-178, January 1971.
7.
L. O. Hefl inger and R. E. Brooks, "Holographic Instrumentation
Studies," TRW Systems Report No. 12122-6007-RO-00, prepared for
NASA Ames Research Center under contract NAS2-4992, December 1970.
64
------- |