EPA-660/3 74-004a
March 1974
Ecological Research Series
Turbulent D if fusion In
Liquid Jets: Part I
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
Hm Environmental Monitoring
5. socioeconomic Environmental Studies
This report has been assigned to the ECOLOGICAL
RESEARCH series. This series describes research
on the effects of pollution on humans, plant and
animal species, and materials. Problems are
assessed for their long- and short-term
influences. Investigations include formation,
transport, and pathway studies to determine the
fate of pollutants and their effects. This work
provides the technical basis for setting standards
to minimize undesirable changes in living
organisms in the aquatic, terrestrial and
atmospheric environments.
-------
EPA 660/3-74-004a
March 1974
TURBULENT DIFFUSION IN LIQUID JETS: PART I
Measurement of Particle Concentration
By a Light Scattering Probe
Charles H. Tinsley, Warren S. Stevenson and Victor W. Goldschmidt
Engineering Experiment Station
School of Mechanical Engineering
Purdue University
Lafayette, Indiana 47907
Project No. 16070 DEP
Program Element 1BA025
Project Officer
George R. Ditsworth
Pacific Northwest Environmental Research Laboratory
National Environmental Research Center
Corvallis, Oregon 97330
Prepared- for
OFFICE OF RESEARCHED DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by tbe Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.16
-------
EPA REVIEW NOTICE
This report has been reviewed by the Office of Research
and Monitoring; Environmental Protection Agency and approved
for publication. Approval does not signify that the con-
tents necessarily reflect the views and policies of The
Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
ii
-------
ABSTRACT
A technique for measuring particle concentrations in turbulent flows
was investigated. This technique is the measurement of the light
scattered from an incident beam by the solid contaminants present.
The results show that for moderate concentrations the scattering
system gives proportional increases in count to increases in particle
concentration. The limitations of this system are the signal to noise
ratio and the condition of singular scattering by the particles.
Suggestions on refinements on the correlation technique used are
made and observed phenomena which require further investigation
are discussed.
This report was submitted in (partial) fulfillment of Contract
16070 DEP under the (partial) sponsorship of the U.S. Environmental
Protection Agency.
iii
-------
CONTENTS
Section Page
I Conclusions 1
II Recommendations 3
III Introduction 5
IV Experimental Apparatus Description 21
V Measurements and Results 33
VI Discussion of Results 49
r
VII Discussion of Measurement Errors 51
VIII Acknowledgements 57
IX References 59
X Glossary of Symbols 63
XI Appendix 65
-------
FIGURES
No. Page
1. Optical System Used For Light Transmission 7
Measurements.
2. Probe With Miniature Photocell (Fiber Optics Probe). 8
3. Angular Distribution of Scattered Light For
Particles Used. 10
4. Optical Geometry Defining Scattering Volume. 12
5. Angular Scattering Distribution For Various Values
of Size Parameters. 13
6. Variation of Dimensionless Length With Solid Angle. 17
7. Effective Traversing Length. ' 19
8. Schematic Fluid Flow System. 22
9. Inlet Section and Test Region. " 23
10. Velocity Measurement Instrumentation. 25
11. Optical System. 26
12. Schematic of the Electronic System. 28
13. Electronics Used in Signal Analysis. 29
14. Particle Size and Number Measurement System. 31
15. Frequency vs. Diameter For Lycopodium Spores. 35
16. Frequency vs. Diameter for Microbeads. 36
17. Dimensionless Distance vs. Dimensionless Velocity. 37
18. Observed Count Per Second Versus Position For Dow
Beads. 40
19. Count Per Second Versus Position For Lycopodium
Spores. 41
20. Observed Count Per Second Versus Concentration For
Dow Beads. 42
vi
-------
21. Observed Count Per Second Versus Concentration For
Lycopodium Spores.
22. Observed Count Per Second Versus Sensitivity For
Dow Beads.
23. Observed Count Per Second Versus Sensitivity For
Lycopodium Spores.
24. Typical Coulter Counter Counts vs. Concentration
For Dow Beads.
25. Typical Coulter Counter Counts vs. Concentration
For Lycopodium.
26. Observed Count Per Second Versus Position For
Lycopodium Spores (Different Alignment).
27. Count Per Second Versus Position Upper and Lower
Limits.
28. Observed Count Per Second Versus Concentration
Upper and Lower Limits.
Vll
-------
TABLES
No. Page
1. Particle Data. 33
2. Particle Supplies and Properties. 34
viii
-------
SECTION I
CONCLUSIONS
It was found in this investigation that light scattering can
be used to measure concentration profiles in internal turbu-
lent flows in regions not adjacent to the wall.
Other important conclusions reached in this investigation are
presented below.
1. Changes in concentration levels resulted in
proportional changes in counts per second at
the sampler.
2. For low concentrations of solids with specific
gravity close to one, the concentration profiles
closely approximated that of the velocity. How-
ever, for higher concentrations the similarity
becomes increasing less pronounced.
3. For a given alignment the scattering system
seemed insensitive to changes of particle size
and real index of refraction.
4. The use of a Coulter counter was not feasible
as a check of concentration levels. A more
practical system would be to use a fiber optic
probe.
-------
SECTION II
RECOMMENDATIONS
This study was limited to the investigation of the possible
use of a light scattering probe as a particle concentration
sampler,
1. The use of alternate sampling methods might
be considered when applicable. For instance,
when sampling droplets or bubbles the use of
a modified hot wire or hot film anemometer
may well be more convenient and reliable.
2. The actual sampling volume of the light scat-
tering probe has to be selected to eliminate
multiple particle scattering signals, as well
as to permit a dependable calibration of the
samples.
3. Techniques for signal to noise optimization
should be studied in future tests if a device
of the type described is to be employed.
4. Additional studies are necessary to determine
the exact count per second and particle number
density relationships. These studies should
follow the direction taken by M.J. Fischer and
F.R. Kause (1967) or that suggested by H.L.
Morse et al. These authors used a two point
correlation method by either using two beams
or two detectors. This allows correlation to
help eliminate unwanted noise.
-------
SECTION III
INTRODUCTION
SCOPE OF PROJECT
The stated objectives of the work sponsored under Contract
#16070DEP, were twofold. One to establish an effective
method of tracking contaminants in a liquid jet, including
the use of a laser system. Two to determine the diffusion
due to turbulence of contaminants of different size and
density. The second objective was fully met. An earlier
report "Turbulent Diffusion in Liquid Jets" dealt directly
with it. On the other hand the first objective was only
partly met. It is now discussed in this report. The
second objective was met through the research of Dr. Strong
C. Chuang. His doctoral thesis was completed on August
1970. The research into the use of a laser as a sampling
device was conducted by Mr. Charles H, Tinsley as part of
his M. S. studies at Purdue University.
PROJECT BACKGROUND
Knowledge of the properties of two phase flow systems,
such as solid-gas, solid-liquid and liquid-gas, is impor-
tant in many aspects of engineering. In recent years ap-
plications in nuclear reactor development, chemical pro-
cesses, dispersion of contaminants in waterways, and the
dispersion of particulate matter into the atmosphere have
called for knowledge of the concentration and concentration
gradient of the dispersed phases.
Several techniques for the experimental evaluation of the
distributions of dispersed phases have been proposed. One
is the hot wire impact method developed by Goldschmidt and
Eskinazi (1966). This has been used to measure aerosol
particle concentrations (HDuseholder (1968)) and Lee (1970)
and to measure bubble concentrations (Chuang (1969)), all in
free turbulent jets. This method consists of inserting a
hot wire probe into the flow and measuring the signal re-
sulting from the impact of bubbles or liquid droplets on
the wire. Soo et al (1964) investigated several different
methods for measuring concentration distribution and mass
flow. The more successful of these was the light attenua-
tion method. In this method light is projected perpendicular
to the flow stream and the transmitted (unattenuated) light
is measured on the other side. The amount of light received
by the detector can be related to the size and the concentra-
tion of particles in the flowing medium. Phelps (1968)
-------
evaluated this dependence of light transmitted on the aver-
age concentration, particle diameter and particle properties.
The experimental arrangement is shown in Figure 1. Soo et al
(1964) used a modification of this method employing a fiber-
optic probe. This has the same dependency on particle size
and properties and particulate concentration as shown by Phelps
for the amount of light received. The primary difference bet-
ween these two methods is that the test section for the experi-
ments of Soo et al is defined by the geometry of the probe,
and the test section for Phelps is defined by the cross-section
of the flow. Consequently, Soo et al are able to obtain a
gradient and Phelps is only able to obtain an average concen-
tration. The disadvantage in the method of Soo et al is that
a probe must be inserted into the flow field. An example of
the type of probe used by Soo et al is shown in Figure 2.
This probe was used by Peskin and Baw to obtain concentration
profiles in pipe flow with solids and air as the two consti-
tuents .
The method used for the measurements reported herein was
light scattering. In this method no physical probe is in-
serted into the flow. The concentration is measured in a
finite volume which is determined by the geometry of the op-
tics. This method has been used previously by Rosensweig,
Hottel and Williams 1961, while studying the concentration
gradient in free turbulent flows of a gas-solid mixture. In
this study the phases are solid-liquid and the turbulent flow
under consideration is an internal flow. The object of this
study was to evaluate the feasibility of using a light scat-
tering system to determine the concentration and concentra-
tion gradient of the solid particles in a turbulent internal
flow.
-------
A. Spectra-Physics model 124
Stabilite gas laser X = 6328 A
L- Condensing lens
L_ Collimating lens
B. Test cell
C. Interference filter
D. RCA IP-28 photomultiplier tube
E. PPI Laboratory photometer
F. Leeds & Northr
strip recorder
C. D.C. power supply
E.
up Speedomax
b
ply
G
I 1
B
^^^
i
J
L
J
^
FIGURE 1 - OPTICAL SYSTEM USED FOR
LIGHT TRANSMISSION MEASUREMENTS (PHELPS (20))
-------
FLOW
00
I-GLASS ROD
PHOTOCELL -
TO VOLTMETER_
FIGURE 2 PROBE WITH MINIATURE PHOTOCELL
PESKIN AND BAW (18)
-------
THEORETICAL BACKGROUND
Scattering Parameters and System Response
Light that traverses a medium may be transmitted, absorbed
or scattered. There are certain parameters of the medium
and of the incident light that determine to what degree
transmission, absorption, and scattering occur. These
phenomena are related such that the amount of light not
transmitted (extinction) is equal to the sum of losses
due to scattering and absorption.
In equation form:
Extinction = Absorption + Scattering (1)
A brief summary of these parameters and their effect on
extinction, primarily taken from Green and Lane (1957),
is given here.
The relative index of refraction of a particle suspended in
a medium affects extinction. The relative index of refrac-
tion is given by:
n particle (2)
n _ __i
n fluid
where n = real part of the complex index of refraction (n - iko)
i = -1
k - a x
K° ~ 4rT
with a = absorption coefficient, and X = wavelength of the
incident beam
Scattering systems generally utilize particles which have a
zero absorption coefficient, thereby maximizing the amount
of the signal obtained by scattering alone. The effect of
the index of refraction on the angular distribution of scat-
tered light is shown in Figure 3 for the two types of parti-
cles studied.
Two other important parameters for scattering systems are the
angle 9 subtended between the beam and the detector, and the
solid angle, a viewed by the detector. These angles help
determine the length of the scattering volume. When the angle,
9, is zero the scattering system reduces to a transmission
system and the scattering volume is maximum. For angle, 9,
-------
6 = 30
(Microbeads)
(Dow)
10^ 10* IQ6 108
FIGURE 3 - ANGULAR DISTRIBUTION OF SCATTERED LIGHT FOR PARTICLES USED
-------
equal to ninety degrees the scattering volume is minimum.
These angles are shown in Figure 4.
Another important parameter in scattering studies is the
ratio of the incident light. This parameter is the size
parameter defined by:
q = 2rrrA (3)
where
r = radius of the particle
X = the wavelength of the incident light
Theoretical angular distributions of scattered light were
obtained for particles used in this study from an I.B.M.
computer program called Dame. This program has the capa-
bility of calculating scattering distributions for a size
parameter up to 5000. The distributions obtained are pre-
sented in Figure 5 as a function of size parameter.
There have been several theories proposed to explain the
relationship between the incident beam and losses associated
with the scattering and absorption. For the size range of
particles used the appropriate theory is that developed by
Mie. A detailed explanation of this theory is given by Van
de Hulst (1957). For the type of scattering experiments
conducted in this study a detailed examination of Mie's
theory is not needed. The underlying assumption of this
study is that the light scattered is directly proportional
to the number of scatterers in the control or scattering
volume. This is true as long as single or independent scat-
tering is the only type present. Sinclair (1950) has stated
that this assumption is valid as long as the distance between
particles is ten or preferably a hundred times the radius of
the particle.
The single scattering assumption has been used by many
investigators; Green and Lane (1957) give the following
simplified scattering expression:
= krPn
where
I = the scattered light intensity
r = the radius of the scatterers
P = parameter determined experimentally
n = number of scatterers
11
-------
Pinhole apertures (394 microns)
Photomultiplier Tube (detector)
- Bandpass Filter for 6328 A
Angle between detector and beam
(30°)
Solid angle formed by detector
(0.4° planar angle)
Diameter of incident beam (2mm)
Distance between apertures (2inches)
Length of scattering volume
Distance of the scattering volume
from origin of a (5 inches)
s
FIGURE 4 - OPTICAL GEOMETRY DEFINING THE SCATTERING VOLUME
-------
9 = 30'
5 = 150
u>
q
q
q
(Microbeads) n
9 = 150
FIGURE 5 - ANGULAR SCATTERING DISTRIBUTIONS FOR VARIOUS VALUES OF SIZE PARAMETER
-------
Rosensweig, Hottel and Williams (1961) extended this
simplified expression to the signal (current) produced
by the phototube which is given by:
i.-I /r* «. (5>
where
I is the phototube signal
S is the overall system sensitivity and is constant
for .a given optical arrangement and alignment
f* is given by lim and is defined as the average
vs * ° s
particle density contained within -the control volume.
N is the average number of particles contained in V
The desirable relation between I . S and f* is
5
Is = Sf* (6)
Rosensweig et al (1961) state that clearly V must be small.
However, decreasing V increases all noise levels of the sy-
stem. Therefore an optimization is needed. They also show
that if a point concentration is to be obtained from equa-
tion (5) the concentration profile must not have large curva-
ture. These conditions are controlled by the geometry of the
optics and by the velocity profile in the medium.
The relation between the count from an electronic counter,
monitoring the phototube output, and the number of parti-
cles per unit volume can now be obtained. Under the assump-
tion of homogenous flow, the total number of particles count-
ed is attributed to the passage of a frozen concentration
pattern through the scattering volume V . The pattern passes
at a velocity U - the local mean velocity of the flow field.
The average time for one particle to pass through the scat-
tering volume is given as:
(7)
where
b = diameter of the beam
U = stream velocity
14
-------
The length D/2 represents an average cross-sectional
length for the circular beam. The time it takes for
a length V /A to pass through the scattering volume
j _ SB
is:
V
S = At
us
where
V = scattering volume
S
A = cross-sectional area of the scattering
volume defined as L D, .
L = the length of the volume formed by the
interception of the solid angle and the
beam. (shown in Figure 4)
If the average number of particles per unit scattering
volume is given by:
r* = £ (9)
where
,%/
N = total number of particles contained
in V
s
V = scattering volume
S
and the total number of particles contained in the system
volume is N then the particle density is given by:
r = v <10>
and for homogenous flow
CV
I- I n-n
v "~ v * '
s
The number of counts per second is then given by the number
of particles per unit scattering volume divided by the time
for one scattering volume to be convected through the volume.
In equation form this is:
15
-------
6 . r* v5 (12)
0 At 5
where
6 = count/second
V = Scattering volume
s
or substituting for T * from equation 9 and for A t from
equation 8
« -I v
But |s - |SSo
6S = S ASU (13)
For a desired count the number of particles and ultimately
the weight of the particles were determined by the relation-
ship expressed by equation 13.
Resolution of the Scattering Volume
The length of the scattering volume may be determined dir-
ectly from the geometry of the optics. The angle as is de-
termined by the diameter of the pinhole apertures which
limit the field of view of the photomultiplier tube. The
angle 9 is the angle between the photomultiplier tube and
the incident beam. These variables are all shown in Figure
4. The equation that describes the dependence of the scat-
tering length upon the geometry is given by:
_ sin oos sin
-------
.20 -
.16 -
.12
.08 --
.04 -
L
s
-- versus ft
Ld
0 - 10
.< (SOLID ANCLE FORMED BY DETECTOR)
s
FIGURE 6 - VARIATION OF DIMENSIONLESS
LENGTH WITH SOLID ANCLE u
17
-------
The effective length for transversing the test section is
longer than Ls by a Length equal to AL^ and AL2- The lengths
AI« and AL2 are shown in Figure 7 and are given by:
AL, = b tan (9 + a) (15)
1 2
AL0 = °b tan (0 - a) (16)
2 2
This total length is the dimension which limits the ability
to make measurements close to the wall.
The conditions that are necessary in order that the volume
be small are:
a. large values of Q (Max. = 90° )
b. small diameter pinholes
c. large distance between apertures
d. small distance between the center of the
aperture and the beam
Equation 14 is derived neglecting the effects of refraction
at the walls of the test section. This effect is generally
negligible provided angle a is very small.
18
-------
INCIDENT BEAM
PINHOLE
APERTURE
FIGURE 7 - EFFECTIVE TRAVERSING LENGTH
19
-------
SECTION IV
EXPERIMENTAL APPARATUS DESCRIPTION
The experimental apparatus may be divided into various
subsystems - the fluid flow system, optical system, parti-
cle measuring system, and the electronic system.
Description of the Fluid Flow System
A schematic of the fluid system is shown in Figure 8.
The system consists of a plenum chamber, a reservoir, a
one-third horsepower centrifugal pump with a bronze head
and an Aqua-Pure water filter. All piping used was made
of Polyvinyl Chloride (plastic pipe) to insure that the
system remained as rust free as possible. The plenum
chamber was made of Plexiglass and had dimensions of I1
x 1' x 2.5'. The inlet section from the plenum to the test
section was constructed from a cylindrical Plexiglass piece
cut at 45° angles in such a manner that when it was reunit-
ed, it formed a bell shape inlet section for a square chan-
nel as shown in Figure 9. The plenum chamber also had an
overflow weir that consisted of a 3/4 inch diameter pipe.
The reservoir was constructed from a fifty-five gallon
drum that had been treated to prevent rust. The reser-
voir had four holes located in its bottom, leading to
valves 3, 4, 5, 6. Valve No. 5 is used1 to help control
the amount of flow going to the reservoir from the con-
stant speed pump. All four valves allowed the separation
of the reservoir from the rest of the system. Valve 4 is
between the reservoir and the drain. Valves 7, 8, 9, are
for further flow control. They allow either complete,
partial, or no flow through the filter. Valve 11 is to
control the weir draining rate and valve 12 is to control
the plenum draining rate. The remaining valves (1, 2)
were to allow collection of a flowing sample from the test
region.
The test section located immediately downstream of the in-
let nozzle was made of Plexiglass and designed in such a
manner that it could be detached from this position and
interchanged with another section of the channel. The
test section is shown in Figure 9.
The velocity profiles were measured with the device shown
in Figure 9. The measuring instrument was a Pace differential
21
-------
Sampling
bypass
(1)
J+
(6)
reservoir
(3)
plenum
filter
(7)
a>
(11)
orifice
meter
drain
FKJURE 8 - SCHEMATIC-FLUID FLOW SYSTEM
22
-------
(1) Test section
(2) Inlet nozzle
(3) Pilot tube with
traversing device
Photomultiplier tube
FIGURE 9 - INLET SECTION AND TKST
23
-------
transducer and the probe was made from 0.035" I.D. wire
tubing. The traversing mechanism used a micrometer to
position the probe. The electronics used for velocity
measurement are also shown in Figure 10. The electron-
ics consisted of a carrier demodulator and a VIDAR digital
voltmeter with external controls- allowing' timer averaging
of the velocity profile.
Optical System
The optical system consisted of a helium-neon laser, a
photomultiplier tube, a narrow band filter, and two pin-
hole apertures. The system is as shown in Figure 11.
This arrangement was chosen as a compromise between scat-
tering volume size and signal to noise ratio. In general
as the angle between the beam and the photomultiplier was
decreased, the scattering volume length increased and the
signal to noise ratio decreased reaching a minimum at the
zero degree angle. The prism shown in Figure 11 was used
just to change the direction of the beam and has no ef-
fect on the performance of the overall system. The wave-
length of the laser and the narrow band filter used was
6328 A. The apertures were approximately 310 microns in
diameter. '>'
To aid in understanding the operation of the system and
the purposes of each component a,detailed explanation is
presented. A uniform beam is incident upon the test sec-
tion normal to the flow direction. The normal condition
was chosen to minimize any aberrations due to refraction.
The test section was chosen to have flat walls to elimin-
ate the need for a compensator box to compensate for the
effect of curvature as was found necessary in work done
in internal flows with a Doppler,system by Denison (1969).
A system that involved the use of a lens that would focus
the collimated beam down to a particular point was tried.
It was discovered that with such a system, it would be
necessary to refdcus the system at each point in the flow
field as the concentration gradient was taken. This was
due to the change in optical path and consequently the
change in effective focal length of the lens.
If small particles are seeded into the flow system, they
will scatter radiation in all directions. The pinhole
apertures that are placed between the photomultiplier
tube and the flow field, together with the geometrical
relationship between the incident beam and the photomul-
tiplier tube, define the cone that intercepts the beam.
The common region between the cone and the incident beam
defines the scattering volume. The particles entering the
scattering volume will scatter light in all directions,
24
-------
(l) External timer (5) Carrier demodulator
(2) External start and : stop (6) Probe and transversing
(3) Vidar digital voltmeter mechahism
(£) Calibration mechanism (7) TransduceU
FIGUKE 10- VELOCITY MEASUREMENT INSTRUMENTATION
25
-------
(1) Laser
(?) Prism
(3) Apertures ani b
filter
CO Photomultiplier
(5) Laser power
11 - OPTICAL
26
-------
I
some of which is received by the photomultiplier tube.
The narrow band filter, which is placed directly in front
of the photomultiplier tube, limits the wavelength of
light that reaches the tube. Therefore, only radiation
that is the same wavelength as the filter bandpass ex-
cites the tube. The signal received by the tube is then
proportional to the scattered radiation from the parti-
cles, from scattering due to turbulence of the water
(Gurvich, (1968)) and from background radiation from
the room within the filter bandpass. Both of the latter
effects were found to be negligible in this study.
The entire optical system was mounted on a large alumin-
um plate which was in turn mounted on a lathe bed in such
a manner as to allow the scattering volume to be traversed
across the flow. Mounting the optics in this manner limit-
ed vibrations and allowed the whole optics to move normal
to the section where concentration gradients were taken.
Electronic System
The electronic system used in the analysis of the signal
from the photomultiplier tube is shown in Figures 12 and
13. This system consisted of an oscilloscope, bandpass
filter, an a.c. amplifier, an oscillator, a visicorder
and an electronic counter.
The signal from the photomultiplier tube was first chan-
nelled to the oscilloscope, and simultaneously to the a.c.
amplifier from which it entered the bandpass filter. The
signal was amplified to enable it to trigger the electron-
ic counter. The amplification factor was ten times the
input signal. The bandpass filter was used to improve the
signal to noise ratio. After passing through the bandpass
filter, the signal was simultaneously sent to the oscillo-
scope and to the electronic counter. Since the signal had
the appearance of shot noise, the filtered and unfiltered
signals were observed on the oscilloscope to insure that the
bas±c- form of the signals was not "changed.
The purpose of the oscillator was to serve as a standard
when in connection with the oscilloscope, by which the
electronic counter could be calibrated for sensitivity
versus accurate count setting.
The visicorder was used to double check the effect of the
bandpass filter, to compare the intensity of light received
from the different types of particles used, and also to
check on the count received from the counter at various
locations.
27
-------
photorcultiplii
tube
NJ
00
oscillator
filter
amplifier
O
oscilloscope
FIGURE 12 - SCHEMATIC OF THE ELECTRONIC SYSTEM USED
-------
(1) Vifccorder (5)
(2) Oscilloscope
(9) Narrow band filter (6)
CO Electronic counter (?)
Photomultiplier tube
voltage power supply
Oscillator
A.C. anplifier
KIIURE 13- ELECTONICS USED IN SIGNAL ANALYSIS
29
-------
Particle Measuring System
This system consisted of only a microscope and an analyti-
cal balance. The size distribution of the particles was
determined by measuring a sample of the particles used
under the microscope. The microscope at high magnifica-
tion will give the diameter to + 1 micron. It was origin-
ally proposed that size determination be done optically.
While the theory was sound, the system would involve an
extensive calibration. This was not considered feasible
for the scope of the present study. Separation of the
particles by size was also investigated. The results of
this investigation showed that only for gross differences
in diameter or properties is separation by size feasible.
Particles that were obtained from the manufacturers though
not uniform in diameter, represented the best practical
size separation that is presently available.
The analytical balance was used to measure the amount of
particles that were to be added to the fluid system. The
balance was accurate to 1/100 milligram. This system is
shown in Figure 14.
Average Concentration Measuring Device
In order that values obtained from the light scattering
experiments could be verified, a Coulter counter was used.
The Coulter counter works on the principle that the pre-
sence of particles passing through a small orifice changes
the electrical path through the orifice. This device is
thereby able to determine the particle size and concentra-
tion that is present in a small sampling vessel.
Concentration measurements from this device were determined
from small samples taken at each concentration level that
was used. The samples were taken with a bypass arrange-
ment from the^ reservoir that can be seen in Figure 8 and
was controlled by valves (1) and (2).
The results obtained from this instrument proved to be un-
reliable for the fluid (deionized water) and concentrations
used.
30
-------
(1) Microscope
(2) Analytical Balance
FIGURE Hf - PARTICLE SIZE AND NUMBER
MEASUREMENT SYSTEM
31
-------
SECTION V
MEASUREMENTS AND RESULTS
Particle Size Measurements
The particle size was determined by measurement with a
microscope. The standard deviation and average were
calculated from the following equations:
-
D = S f. x.
d=l
(17)
^ - D)2
(18)
where
N = Number of particles sampled
x.= Diameter of a given particle
f = Frequency of occurrence for that particle dia-
meter
T = Standard deviation of the particle diameter
5 = Average diameter
The results obtained are given in the following table:
Table I
TYPE OF PARTICLE
Lycopodium Spores
Microbeads
Dow Beads
D
31.66 p,
34.3 ji
45.4 |a
T
2.23 p
4.74
8.9
n
i Unavailable
1.51
1.587
33
-------
The frequency of occurrence of sizes for Lycopodium and
for microbeads can be found in Figures 15 and 16 » The
density and other properties of Lycopodium were taken
from Rouhiainen and Stachiewicz (1970) . The other par-
ticle properties were taken from manufacturer's data and
can be found in Table II.
TABLE II
PARTICLE SUPPLIERS AND PROPERTIES
MANUFACTURER
OR MATERIAL
SUPPLIER
INDEX
DIAMETER OF SPECIFIC
REFRACTION GRAVITY
Carolina
Biological
Supplies
Dow
Chemical
Microbeads
Minnesota
Mining
Reflective
Products
Lycopodium
Spores
Styrene
divinyl-
benze
Soda-lime
silca glass
Glass
32.2
34.3
1.587
1.51
1.5
0.621
1.05
1.5
1.5
Velocity Measurement
The measurement of the velocity was done prior to any of
the concentration runs. The nozzle and flow field is
shown in Figure 16. The mean velocity in the test sec-
tion was found to be 8 feet per second. The correspond-
ing Reynolds number based on the hydraulic diameter is 5.76
x 10^. Velocity profiles were found to be flat over most
of the test section in both horizontal and vertical direc-
tions. A typical profile is shown in Figure 17.
Concentration Measurements
The particles were injected into the plenum chamber and
34
-------
f versus x.
D - 31.66
T - 2.23 u
24 25 26 27 28 29 30 31 32 33 34 35
DIAMETER x , MICRONS
FIGURE 15 - FREQUENCY VERSUS DIAMETER FOR LYCOPODIUM SPORES
35
-------
100..
o
u
tn
O
w
S3
80
60.
40
20"
D - 34.3 p
T - 4.74p
f versus x
10
15
20
25
30
35
40
DIAMETER x , MICRONS
FIGURE 16 - FREQUENCY VERSUS DIAMETER FOR MICROBEADS
36
-------
1.0 ..
w
S
o
H
H
X
O
M
W
as
o
M
H
,8 --
,2 ..
y/h vs. U/U
X U - 7.87 ft/sec
vertical
0 U = 8.145 ft/sec
horizontal
i
°d
6
I
"O
b
i0
o
0.0
.2
,4
.6
1.0
II FT/SFC
--, RATIO OF VELOCITY TO AVERAGE VELOCITY,
FICURE 17 - DIMENSIONLESS DISTANCE VERSUS DIMENSIONLESS VELOCITY
37
-------
allowed to disperse for one hour to one and one-half
hours before any data was taken. This time was deter-
mined experimentally as the time delay necessary for a
consistent count. The procedure for data collection
was as follows:
1. Check to insure the following equipment
was turned on at least one-half hour be-
fore a run.
a. Oscilloscope
b. High voltage power supply
c. Visicorder oscillograph
d. Laser
e. Electronic Counter (set at maximum
sensitivity)
2. The inner wall of the test section was
located. (The location of the wall was
determined by the scattering pattern at
the interface between the wall and the
air.
3. Dial positioner was zeroed. (This dial
was used to determine step size of tra-
verses of the scattering volume in the
test section)
4. All room lights were turned off.
5. Readings of the count per second were
obtained at positions spaced 0.10 inch
apart.
6. Visicorder oscillograph and count per
second readings were obtained at posi-
tions 1, 4 and 7 which correspond to
the two wall regions and the center of
the test section.
7. Electronic counter sensitivity was varied
from a voltage threshold of 55 m.v. to
104 m.v.
8. The concentration level was changed.
(For lycopodium spores a wetting agent
must be used. For this investigation
Industroclean cleanser was added to the
system.)
38
-------
- 9. Electronic Counter was set back to maxi-
mum sensitivity (55 m.v.).
10. Steps 2-19 were repeated.
The data for the point concentration measurements were
taken as an arithmetic average of the values taken at
the point. Typical distributions for each type of par-
ticle run is presented in Figures 18-23 .
Coulter Counter Results
Typical results obtained for a few of the runs from the
Coulter counter are presented in Figures 24 and 25- T^6
Coulter counter proved to be a poor choice of a check for
the results obtained from the scattering system. The
primary reason is that the Coulter is extremely sensitive
to contaminants in the water used. The Coulter counter
would give high counts for clear water and give counts on
the -same order of magnitude for samples obtained from the
particulate system. In fact, in some cases higher counts
were obtained for clear water (uncontaminated deionized
water). The scattering system did not show this sensi-
tivity to clear water contamination; however, if it had
a zero level could have been set and exact counts could
have been obtained by substracting this known zero level
from obtained values.
39
-------
O - ix
D - 2X
A 3X
25.
20.
0
S
8 is.
U4
en
PS
w
Pi
g
8
5.
IX - 7.76 x 10* |S«i£ia
A
""A A A
>
n n n o n a
n n n __n_
O u w
i I I i II 1
123 45 67
POSITION
FIGURE 18 - OBSERVED COUNT PER SECOND
VERSUS POSITION FOR DOW BEADS
40
-------
O
35
30
25
20
8 15
w
w
PL,
8
10
a
TT
O
O
O
-D-
0-8X
a, 4x
A 2X
O IX
IX - 2.02 x 1,05 particles
-
CL
TT
A
O
i
12345
POSITION
(DISTANCE BETWEEN POINTS = 0.10")
FIGURE 19 - OBSERVED COUNT PER SECOND VERSUS
POSITION FOR LYCOI'ODIUM SPORES
41
-------
25.0
Q
S5
8
Ul
I
20.0
15.0
10.0
5.0
A
O
a
IX
position //7
position #4
position //I
1 milliter of
10% concentration
I
IX
2X
3X
CONCENTRATION
F1CURE 20 - OBSERVED COUNT PER SECOND
VERSUS CONCENTRATION FOR DOW BEADS
42
-------
o
u
w
to
PS
o
u
30
25
20
15
10
position //I
position //4
position //7
IX = .0161 gins
2X
4X
6X
8X
10X
CONCENTRATION
FIGURE 21 - OBSERVED COUNT PER SECOND
VERSUS CONCENTRATION FOR LYCOPODIUM SPORES
43
-------
o
u
w
tn
W
o.
o
u
10
5 __
40
60 80 100
COUNTER SENSITIVITY (mV)
120
FICURE 22 - OBSERVED COUNT PER SECOND
VERSUS SENSITIVITY FOR DOW BEADS
44
-------
O
J3
(O
fi
O
u
25
~~ O * point //I
*-J point #4
A - point #7
20
15
10
A
0
25
50
75
100 112.5
COUNTER SENSITIVITY, (raV)
FIGURE 23 - OBSERVED COUNT PER SECOND
VERSUS SENSITIVITY FOR LYCOPODIUM SPORES
45
-------
SO
I
o
o
o
M
g
O
Si
(X,
8
2.5
2.0
1.5
1.0
.5
.2
J_
I
IX - 10 ml of 1%
concentration
IX 2X 3X
CONCENTRATION
4X
5X
FIGURE 24 - TYPICAL COULTER COUNTER COUNTS
VERSUS CONCENTRATION FOR DOW BEADS
46
-------
xO
o
en
w
I
u
u
OS
w
eu
12 -
10
8
IX - 0.0161 gma
IX 2X 3X
CONCENTRATION
4X
5X
FICUKG 25 - TYPICAL COULTER COUNTER COUNTS
VERSUS CONCENTRATION FOR LYCOPODIUM
47
-------
SECTION VI
DISCUSSION OF RESULTS
The primary results obtained are presented in Figures
18 through 24; they illustrate typical data. Figures
20 and 21 show that a change in concentration results
in a corresponding change in counts per second and is
expressed by a linear relationship for moderate concen-
tration levels. It can be seen by comparing Figures 18
and 19 (which show the response of Dow beads and lycopo-
dium spores) that similar concentrations levels give
similar counts per second. Figures 18 and 19 also show
that the concentration profile, like the velocity pro-
file/ is flat across the test section for moderate con-
centrations. At high concentration levels a noticeable
assymetry in the profile was observed. The exact ex-
planation for this effect is not known. Figures 22
and 23 show the response of the system to varying de-
grees of sensitivity. These figures show that a corres-
ponding response to a concentration change was obtained
at all levels of sensitivity tested.
Figures 24 and 25 show the response of the Coulter count-
er to various concentrations. For lycopodium spores the
response is almost linear; however, for dow beads the
response was nonlinear. It was concluded from these
results and attempts at calibration, with this fluid and
these concentrations levels, that the Coulter's results
were not accurate.
49
-------
SECTION VII
DISCUSSION OF MEASUREMENT ERRORS
The data presented in the previous chapter illustrate
typical experimental results obtained with the system.
Due to the nature of the experiment rather large varia-
tions were observed in some of the measurements. The
source and magnitude of these variations is discussed
here.
Effect of Alignment
One of the disadvantages of this system is the sensi-
tivity to alignment. The amount of light received is
directly related to the angle between the detector and
the incident beam. One assumption that has been made
is that the detector and the beam are in the same plane
and that the pinholes and the beam are aligned. Devia-
tions from these conditions can greatly effect the re-
sults obtained at the counter. An example of this is
seen by comparing the results shown in Figures 18 and
26. For the same number concentration different counts
per second were obtained. Consequently, the actual
value of the count per second is not significant - only
the relative effect of changing the number concentra-
tion for a particular instrument alignment can be mea-
sured.
Concentration Measurement Error
The only error presented here is the error in consistent
counts. Figures 27 and 28 show the variation in concen-
tration profile and system response respectively. These
figures were obtained ,by taking the largest deviation
from the average. These curves show that the system re-
sponse was very good because the error in the count shown
could be expected just from the nature of turbulent flow.
Particle Suspension Considerations
To avoid the effect of particles separating from the flow
and becoming attached to some portion of the system, all
runs were made in one continuous sweep. The particles
were assumed to remain suspended throughout the run. The
effect of particles separating from the system was noted
when particles slightly heavier than water were used. For
51
-------
300.
250.
200.
8 150.
w
w
fx.
H
o
u
100.
50.
3 4
POSITION
IX - 2.02 x 10J particles
- 16X
- 8X
Clear water
FIGURE 26 - OBSERVED COUNT PER SECOND VERSUS
POSITION FOR LYCOPODIUM SPORES (DIFFERENT ALIGNMENT)
52
-------
= Lower limit
A - Upper limit
C.' » Average values
g
o
u
U4
U4
o
o
0
3 4
POSITION
FIGURE 27 - OBSERVED COUNT PER SECOND VERSUS
POSITION'UPPER AND LOWER LIMITS
53
-------
30
25
20
1 IS
a
8
10
A * Upper limit
O * Average Value
D * Lower limit
IX
3X
5X
7X
9X
CONCENTRATION
FIGURE 28 - OBSERVED COUNT PER SECOND VERSUS
CONCENTRATION UPPER AND LOWER LIMITS
54
-------
these particles the count first steadied and then rapidly
began to decrease. The lighter particles seemed to remain
suspended in the fluid rather well. This was determined
by allowing the system to run for several hours and then
rechecking the count/sec of particles.
55
-------
SECTION VIII
ACKNOWLEDGEMENTS
The reported work was the result of the Master of
Science research of Mr. Charles H. Tinsley, Jr.
Professors Victor W. Goldschmidt and Warren Stevenson
acted as supervisors of his work.
The support of the project by the Water Quality Office
Environmental Protection Agency, and the help provided
by Mr. George Ditsworth, Project Representative is
acknowledged.
The assistance of the staff of the School of Mechanical
Engineering and its Ray W. Herrick Laboratories during
the final stages of the work reported is also gratefully
acknowledged.
57
-------
SECTION IX
REFERENCES
Anderson, T.B. and R. Jackson. "A Fluid Mechanical
Description of Fluidized Beds." Industrial and Engi-
neering Chemistry Fundamentals. Vol. 8 pp. 137-144
February 1969.
Becker, H.A., H.A. Hottel and G. Williams. "On Light-
Scatter Technique for the Study of Turbulence and Mix-
ing." Journal Fluid Mechanics, Vol. 30, part 2, pp.
259-284. ^
Born, Max and Emil Wolf. Principles of Optics. Per-
gamon Press, 1959.
Chuang, Strong C. "Turbulence Diffusion of Small Gas
Bubbles in an Axi-Symmetric Water Jet." Ph.D. Thesis
School of Mechanical Engineering, Purdue University,
1969.
Denison, E.B. "Pulsating Laminar Flow Measurements
with a Directionally Sensitive Laser Velocimeter."
Ph.D. Thesis, Purdue University, 1969.
Fischer, M.J. and F.R. Kause. "The Crossed-beam
Correlation Technique." Journal of Fluid Mechanics,
Vol. 28, part 4, pp 705, 1967.
Goldschmidt, V.W. and S. Eskinazi. "TVo Phase Tur-
bulent Flow in a Plane Jet." Journal of Applied
Mechanics, Vol. 33 Series E, No. 4, December, 1966.
Green, H.L. and W.R. Lane. Particulate Clouds, Dust
Smokes and Mist. D. Van Nostrand Company, Inc. Chapter
4, 1957.
Gurvich, A.S. "The Determination of Turbulence Char-
acteristics from Light Scattering Experiments". At-
mospheric and Oceanic Physics, Vol.4, No. 2, 1968.
Hanesian, D. and A. Rankel. "Elutriation from a Mul-
tisize Particle Fluidized Bed." Industrial and Engi-
neering Chemistry Fundamentals, Vol. 7, No. 3, August,
1968.
Hino, M. "Turbulent Flow with Suspended Particles."
American Society of Civil Engineers Proceedings, Vol.
89, Journal of Hydraulics Division, Hy. 4, July 1963.
59
-------
Householder, Michael K. and V.W. Goldschmidt, "Tur-
bulent Diffusion on Small Particles in a Two-Dimension-
al Free Jet," Technical Report FMTR-68-3 School of Civil
Engineering, Purdue University, Sept. 1968.
Lee, J.B. "Turbulent Diffusion of Particles in Plane and
Circular Jets." M.S. Thesis, Purdue University, School of
Mechanical Engineering, January 1970.
McGill, D.R. "Scattering Experiments." Development of
Four Experiments for Applied Optics Laboratory. M.S.
Thesis, Mechanical Engineering, Purdue University, 1968.
Moder, J.J. and D.A. Dahlstrom. "Fine Size, Close-Specific-
Gravity Solid Separation with Liquid-Solid Cyclone." Chemi-
cal Engineering Progress, p. 75, 1952.
,-
Morse, H.L., B.J. Tullis, W.R. Babcock and H.S. Seifert.
"Development, Application and Design Specifications of a
Laser-Doppler Particle Sensor for the Measurement of
Velocities in two Phase Flow Rocket Exhaust." Volume I,
Sudaar, No. 356, p. 56.
Perry, R.L. and S.M. Henderson. Agriculture Process Engi-
neering, pp. 184-189, New York, Wiley, 1955.
Peskin, D.L. and P.S. Baw. "Solid Density Distribution in
Gas-Solid Pipe Flow." Technical Report No. 108-ME-F NYO
2930-6.
Petrie, J.C. and D.E. Black. "Improved Air Distributor
Caps for Fluidized Beds." Chemical Engineering Progress,
Vol. 62, No. 3, March, 1966 p. 82.
Phelps, Jean P. "Particle Size Determination Using A Laser
Light Transmission Technique." M.S. Thesis, U.S. Naval Post
Graduate School, March 1968.
Rosensweig, R.E., H.C. Hottel and G.C. Williams. "Smoke-
Scattered Light Measurement of Turbulent Concentrations
Fluctuations." Chemical Engineering Science, Vol. 15,
pp. 111-129, 1961.
Rouhiaineri, P.O. and J.W. Stachiewieg. "On the Deposi-
tion of Small Particles from Turbulent Streams." Journal
of Heat Transfer, Vol. 92, Series 1, February 1970.
Scholz, J.T. and D.R. Uhlmann and B. Chalmers. "Elutria-
tion Particle Separator." The Review of Scientific Instru-
ments, Vol. 36, Number 12, pp. 1813-1816, December 1965.
60
-------
Sinclair, D. Handbook on Aerosols. (Washington, B.C.:
U.S. Atomic Energy Commission) p. 84, 1950.
Soo, S.L. Fluid Dynamics of Multiphase System,. Blaisdell
Publishing Company, Chapters 1 & 4, 1966.
Soo, S.L., G.J. Trezek, R.C. Dimick and G.F. Hohnstreiter.
"Concentration and Mass Flow Distributions in a Gas-Solid
Suspension." Industrial and Engineering Chemistry Funda-
mentals, Vol. 3, p. 98, May 1964.
Van de Hulst. Light Scattering by Small Particles, Wiley
& Sons New York, N.Y., 1957.
Wallis, G.B. One Dimensional Two Phase Flow, Chapter 8,
p. 175, 1969.
Zenz, F.A. Fluidiza.ti_on of Particle Systems. Reinhold,
New York, N.Y., 1960.
61
-------
SECTION X
GLOSSARY OP SYMBOLS
A cross-sectional area of scattering volume defined as L a
s b
D average particle diameter
D. diameter of beam
f frequency of occurrence for that particle diameter
g 32.2 ft/sec/sec
I scattered light intensity
I phototube signal
S
k a 'X
O 4lT
L length of the volume by interception of solid angle a
S S
L stopping distance for impingement separators
S
L free fall travel distance in a cyclone separator
Vv
mV millivolts
N total number of particles added to water
j, average number of particles in Vg
NS number of times particle encircles cylindrical portion
of cyclone separator
N number of particles sampled
n real part of refractive index
n1 relative index of refraction
n number of scatterers
p parameter determined experimentally used to determine
light scattered
63
-------
2TTr
q size parameter jj
r radius of particle
S overall system sensitivity
U stream velocity
V total volume of fluid
V scattering volume
S
Vic inlet velocity to cyclone separator
V particle velocity
V.,V free fall or terminal velocity of one particle
*tz G
x. diameter of a given particle
a' absorption coefficient
a solid angle viewed by detector
F * average number of particles per unit scattering
volume
F average number of particles per unit volume
AL,&AL_ additive terms to L for traversing length
J- £t S
5 count/sec
9 angle subtended between the beam and the detector
X wavelength of the incident beam
p density of particle
p.. density of fluid
T standard deviation of particle size range
64
-------
SECTION XI
APPENDIX
In many experimental investigations, it is necessary
to have a supply of uniformly sized particles. Parti-
cles obtained from various manufacturers, in a range of
10 to 100 microns, do not adequately meet the uniform
size specifications. Examples of the types of variance
encountered are given in Table 1. The objectives of
this appendix are to present a survey of the various
techniques suitable for separation of micron range
particles by size and to examine the most practical
system.
Filtration Separators
Filtration separators are usually made of cloth or of a
porous element. This type of separator is commonly used
in conjunction with the removal of a particulate from a
gas stream. Filter efficiency is dependent on the length
of time in operation. Filters start at a low efficiency
but increase to values as high as 99.9%. After a short
period of operation, the pores become elogged from im-
pingement, adherance of fines or by blockage owing to the
larger-sized fractions in the dust being separated. The
filter will then build up a "precoat" layer on its up-
stream surface and thus the precoat actually serves as a
high efficiency filter bed.
Sieving is another form of filtration separation. Siev-
ing is a method used to separate the fine particles from
the course ones by means of nested screens. The sieving
screens are mounted on a metal rim to form cylindrical
pans. Such pan_s, each of different screen size, are
nested one atop another with the finest at the bottom and
the coarsest at the .top. The material is placed on the
top and then the screens are shaken. The difference be-
tween the pans of the nest represent a certain size range
of the material .separated.
Impingement Separator
Impingement separators depend on the inertia of the parti-
cle. The carrier fluid flows the streamline around an ob-
ject which the particle cannot immediately follow, due to
the particle momentum. The particle is, therefore, decel-
erated to zero velocity by striking the object. After
65
-------
striking the object, the particles are held into place
and collected - usually at shut down. Collection can be
made by counter current flow or by centrifugal force im-
parted on the carrier gas. Zenz (1960) gives a theoretical
explanation of this type of separator. The important para-
meter is the stopping distance (L_). The stopping distance
is a function of the particle velocity, free fall velocity,
and both particle and fluid densities. For a spherical parti-
cle the stopping distance is given by:
Ls = "P PP Ve (a)
where
v = particle velocity
v. = free fall velocity - v
T S
p = particle density
pf = fluid density
g =32.2 ft./sec/sec
Cyclone Separation
The cyclone or centrifugal separator is a device utiliz-
ing radial acceleration for separation of particles suspend-
ed in a gas stream. It consists of an outer cylindrical-
shell and a cone attachment, and is so arranged that the
dust-laded gas enters tangentially. This causes a vortex
of gas which ascends up the cone and finally ascends to an
outlet concentric with the outer cylinder. The particles,
however, are impinged against the curved wall and swirl in
a downward spiral path due to the effect of an outward
centrifugal force imparted by the momentum due to inlet
velocity and the force of gravity. The minimum diameter,
of the particles separated by this process, is given as a
function of fluid dynamic viscosity, free fall velocity,
fluid inlet velocity, and the difference in particle and
fluid densities.
In order to obtain the function which determines the dia-
meter of the particles to be separated, the distance traveled
by the particle laden gas must be determined. This distance
is simply 2rrr N of the cylindrical portion of the cyclone.
The time that it takes the gas to travel this distance is
t = 2rrr N_A, where V. is the inlet velocity to the
' ic
66
-------
cylindrical portion of the separation. This also is the
maximum time allowed for the particles to travel the dis-
tance LW. This is given by:
. , %
zc (b)
VV. = 2rrr N /V
t c s
The terminal velocity can be determined from an equation
given in a succeeding section if the drag force is com-
puted by Stoke 's law for a centrifugal field. The terminal
velocity is then given by:
Vt = 9Dp(Pp - pf)/18 nf (c)
Substituting and solving for D , the following is
obtained: "
Dp= 9
Gravity Separator
Gravity settling chambers are the simplest type of parti-
cle-collection equipment. The principle of operation is
that the fluid velocity is reduced below the free falling
velocity of the particle. This allows the particles to
settle from the carrier fluid by the influence of gravity.
The rate at which this occurs is dependent upon both fluid
and particle properties. For similar shapes, the settling
rate (free fall or terminal velocity) becomes dependent
only upon the characteristic length usually the hydraulic
diameter. In Wallis (1969) a derivation and also the
correlation are given for the terminal velocity.
Any expanded section in line can constitute a gravitational
settling chamber, however they are usually a long, empty,
horizontal vessel or a long vertical cylinder. For the
horizontal vessel, the basic idea is to allow the particle
to travel a vertical distance in free fall in less time
than it takes the carrier fluid to flow from inlet to outlet
of the settling chamber. For the case of equal densities but
different sizes the larger particles will fall out first and
therefore the length of the chamber determines which diameter
particles will settle. The cylinder settling chamber consists
of a stagnant or very slow moving fluid in which the particles
are dropped. The time at which a certain diameter particle
reaches a predetermined depth can be determined thereby al-
lowing a known range to be collected.
67
-------
Entrainment
Entrainment may be defined as the carry-over of the
particle by the fluid, from the particle bed to and
through the dispersed phase. The rate of entrainment
depends on factors that involve the apparatus, charac-
teristics of the solid and fluid, for all of these.
The entrainment process can be described in terms of the
free fall or terminal velocity of the particle. When
the fluid velocity exceeds the velocity of the free fall
of the particle, the particle is carried by the fluid
until either the particle is carried completely from
the column or the particle forces exceed the momentum
of the particle and it returns to the bed. Unfortunately,
the entrainment rate cannot be modeled by the terminal ve-
locity alone, other effects such as the throwing up of
the particles by the bursting action of the gas at the
particle bed and the concentration of the particles are
important. Since the entrainment rate is dependent upon
free fall velocity it also allows particle separation by
size.
Conclusion
The methods surveyed were:
(1) Filters
(2) Sieves
(3) Impingement Separators
(4) Cyclone Separators
(5) Gravity Chambers
(6) Entrainment Columns
Of these methods filters do not prove adequate because of
the nature of their operation. Filtration is essentially
a catch all operation. Sieving is used by most of the
manufacturers and if this method was chosen no finer dis-
tribution could be obtained. Impingement separation while
theoretically possible is experimentally infeasible because
it would require a test program for different types of
collector designs.
Gravity separators are useful only when all particles are
to be separated or when the densities or size are widely
68
-------
different. Even when the above conditions are satisfied,
collection of the particles by the different sizes is
difficult. Entrainment was chosen as the more practical
because it allows easy collection and because it can be
coupled with either cyclone, filter or sedimentation
chambers. The cyclone separator is thought to be the
best separator to couple with the entrainment bed because
it gives a second chance at size separation. Closer examin-
ation reveals that for particles with diameters around 100
microns free fall velocity is approximately 0.43 to 1.07
ft/sec. However, to separate particles with 4 (j, range
results in air velocity control as being less than .01
ft/second. It was concluded that for the particle size
separation needed, adequate controls were not available.
^U-S. GOVERNMENT PRINTING OFFICE: 1974 546-319/392 1-3 69
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
3. Accession No.
w
4. Title
TURBULENT DIFFUSION IN LIQUID JETS: PART I
7.
4. ^effotraltt!' ffrg&lsxaitat '
Reports? &,
Charles H. Tinsley, Warren S. Stevenson & Victor W. Goldschmi'dt
9. Organization
Engineering Experiment Station
School of Mechanical Engineering
Purdue University
.Lafayette. IN 49707
10. Project No.
16070 DEP
11. Contract/ Grant No.
16070 DEP
13. Type ol Report and
Period Coveted
is. .supplementaryNote* pARj j.IS SUBTITLED: Measurement of Particle Concentration by a
Light Scattering Probe. Environmental Protection Agency Report number 660/3-74-004«i,
March 1974
16. Abstract .
A technique for measuring particle concentrations in turbulent flows was investigated.
This technique is the measurement of the light scattered from an incident beam by the
solid contaminants present.
The results show that for moderate concentrations the scattering system gives
proportional increases in count to increases in particle concentration. The
limitations of this system are the signal to noise ratio and the condition of
singular scattering by the particles.
Suggestions on refinements on the correlation technique used are made and observed
phenomena which require further investigation are discussed.
17a. Descriptors
Equations, Instrumentation, *Particle Size, Suspension, Turbulent Flow
17b. Identifiers
Lasers, Light Scattering* * Particle Concentration, * Particle Size Measurement
17c. COWRR Field & Group 08 B
18. Availability
«. .
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 2O24O
Abstractor c D
| institution Environmental Prnt.pr.t1nn Aqpnry
WRS1C IO2 (REV. JUNE 1971)
------- |