Model For Horizontal Tube Settlers

Iowa State University Capstones, Theses and
Retrospective Theses and Dissertations

Dissertations
1985
Model for horizontal tube settlers

Ahmed Ashry Fadel
Iowa State University
Follow this and additional works at: https://lib.dr.iastate.edu/rtd

Part of the Civil Engineering Commons
Recommended Citation
Fadel, Ahmed Ashry, "Model for horizontal tube settlers " (1985). Retrospective Theses and Dissertations. 12059.
https://lib.dr.iastate.edu/rtd/12059
This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University
Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University
Digital Repository. For more information, please contact digirep@iastate.edu.
INFORMATION TO USERS
This reproduction was made from a copy of a document sent to us for microfilming.
While the most advanced technology has been used to photograph and reproduce
this document, the quality of the reproduction is heavily dependent upon the
quality of the material submitted.
The following explanation of techniques is provided to help clarify markings or

notations which may appear on this reproduction.
1.The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the missing
page(s) or section, they are spliced into the film along with adjacent pages. This
may have necessitated cutting througii an image and duplicating adjacent pages
to assure complete continuity.
2. When an image on the film is obliterated with a round black mark, it is an

indication of either blurred copy because of movement during exposure,
duplicate copy, or copyrighted materials that should not have been filmed. For
blurred pages, a good image of the page can be found in the adjacent frame. If
copyrighted materials were deleted, a target note will appear listing the pages in
the adjacent frame.
3. When a map, drawing or chart, etc., is part of the material being photographed,
a definite method of "sectioning" the material has been followed. It is
customary to begin filming at the upper left hand comer of a large sheet and to
continue from left to right in equal sections with small overlaps. If necessary,
sectioning is continued again-beginning below the first row and continuing on
until complete.
4. For illustrations that cannot be satisfactorily reproduced by xerographic

means, photographic prints can be purchased at additional cost and inserted
into your xerographic copy. These prints are available upon request from the
Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases the best
available copy has been filmed.
Universi^
Microilms
International
300 N. Zeeb Road
Ann Arbor, Ml 48106
8524649
Fadel, Ahmed Ashry
MODEL FOR HORIZONTAL TUBE SETTLERS
Iowa State University PH.D. 1985
University
Microfilms
Intornâtio nâf 300 N. Zeeb Road, Ann Arbor, Ml 48106
PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy.
Problems encountered with this document have been identified here with a check mark V .
1. Glossy photographs or pages
2. Colored illustrations, paper or print
3. Photographs with dark background
4. Illustrations are poor copy
5. Pages with black marks, not original copy
6. Print shows through as there is text on both sides of page
7. Indistinct, broken or small print on several pages
8. Print exceeds margin requirements
9. Tightly bound copy with print lost in spine
10. Computer printout pages with indistinct print
11. Page(s) lacking when material received, and not available from school or
author.
12. Page(s) seem to be missing in numbering only as text follows.
13. Two pages numbered . Text follows.
14. Curling and wrinkled pages
15. Dissertation contains pages with print at a slant, filmed as received
16. Other
University
Microfilms
International
Model for horizontal tube settlers
by
Ahmed Ashry Fadel
A Dissertation Submitted to the
Graduate Faculty in Partial Fulfillment of
Requirements for the Degree of
DOCTOR OF PHILOSOPHY
Department: Civil Engineering

Major: Sanitary Engineering
Approved :
Signature was redacted for privacy.
For pfie Major Department
For the Gr.
Iowa State University

Ames, Iowa
1985
ii
TABLE OF CONTENTS
Page
DEDICATION xi
INTRODUCTION 1
REVIEW OF LITERATURE 3
Discrete Particle Settling 3
Flocculant Settling 6
Flow Regimes 7
Sedimentation Theory 8
Development of Tube Settlers 10
Tube Settlers 15
Design and Experimental Approaches • 20
Gulp et a l . 20
Comments 2f
Yao 29
Comments 38
Willis 40
Comments ' 42
Hernandez and Wright 43
Comments 45
Van Vliet 47
Comments 50
El-Baroudi and Fuller 50
Comments 54
Factors Affecting the Performance of Tube Settlers 54
Summary 56
UNSETTLED QUESTIONS 58
Actual Particle Trajectory 58
The Fadel model 63
Essentially-Horizontal vs Steeply-Inclined Tube 66
Inlet and Outlet Arrangement 68

ill
THEORETICAL STUDY 72
Factors Affecting Tube Performance Using Fadel's Model 80
Differences Between the Fadel Model and Yao and Gulp Models 87
Method of Using the Model 94
Essentially Horizontal Tubes 102
EXPERIMENTAL STUDY 113
Equipment and Materials 113
Equipment 113
Miscellaneous equipment 120
Materials 120
Suspension preparation 120
Disadvantages of the elutriation process 123
Experimental Procedure 124
RESULTS AND DISCUSSION 130
Part 1 130
Run #1. (2/18/85) 130

Run #2. (2/20/85) 132
Run #3. (2/22/85) 133
Run #4. (2/25/85) 133
Run #5. (2/27/85) 134
Run #6. (3/5/85-3/6/85) 135
General Comments, Runs #1 - #6 135
Run #7. (3/17/85) 136

Run #8. (3/19/85) 138
Run #9. (4/20/85) 139
Run #10. (4/5/85) 140
Run #11. (4/6/85) 141
Run #16. (4/11/85 to 4/13/85) 142
Run #21. (4/25/85) ' 144
General Comments on Runs #7, #11, #16, and #21 147
Part 2 148
Set #1 • 149
iv
Set #2 151
Set #3 151
General Discussion on the Run Sets Conducted in Part 2 154
CONCLUSIONS 157
From the Theoretical Study 157
From the Experimental Study 158
RECOMMENDATIONS 160
REFERENCES 161
ACKNOWLEDGEMENTS 163
APPENDIX A. COMPUTER PROGRAMS 165
APPENDIX B. NUMERICAL SOLUTION OF FADEL'S MODEL 179

V
LIST OF TABLES
Page
Table 1. Maximum flow rates in GPM/ft for steeply inclined

tube (9 = 60°) 45
2
Table 2. Maximum flow rates in GPM/ft for essentially
horizontal tubes (6 = 5° M) 46
2
Table 3. Maximum flow rates in GP/ft for settling
activated sludge mixed liquors (6 = 60°) 46
Table 4. Langhaar constants 59
Table 5. The effect of sludge accumulation on the required

tube length for a 3" tube, settling velocity =0.1
fpm, and degree of inclination = 5.0° 109
Table 6. The effect of sludge accumulation on the

required tube length for the 4" tube, settling
velocity = 0 . 1 fpm, and degree of inclination =
5.0° 109
Table 7- Results and computer-predicted results in Run #7 137
Table 8. A comparison between the different models using Run

#7 data 138
Table 9. Results and computer-predicted results in Run #9 140
Table 14. Diameter of glass beads collected in the cups in

Run set #1 149

Run set #2 151

Run set #3 154
vi
LIST OF FIGURES
Page
Figure 1. Drag coefficients of spheres, disks and cylinders 5
Figure 2. Settling column and settling curves for flocculant

particles 5
Figure 3. Settling paths of discrete particles in an ideal

basin 9
Figure 4. Different settling velocities in an ideal basin 9
Figure 5. Flocculated particle trajectory in an ideal basin 11
Figure 6. The effect of reducing the depth of a settling

basin receiving flocculant particles 11
Figure 7. Tray in tank provides added floor area and

increases solids removal 12
Figure 8. Possible tube cross-seciton designs 17
Figure 9. Basic configuration for tube settlers 17
Figure 10. Self-supporting beam module of steeply inclined

tubes 18
Figure 11. Basic arrangement of horizontal tube alignment 18
Figure 12. Basic arrangement of inclined tube alignment 19
Figure 13. Downflow and upflow tilted plate separator 19
Figure 14. Normal and backwash operations of the Aqua-Reuse

plant 21
Figure 15. Suggested arrangement for steeply inclined tube

settlers by Willis 21
Figure 16. Straight line trajectory for the particles in tube

settlers suggested by Gulp et a l . 23
Figure 17. Schematic diagram for the experimental test

equipment used by Gulp et a l . 26
Figure 18. The effect of tube diameter, tube length, flow rate
flocculation on the percent removal of turbidity 27
vil
Figure 19. The effect of the tube settler's degree of

inclination on the percent removal of turbidity 28
Figure 20. Coordinate system for the design model by Yao 31
Figure 21. Relative settler length vs performance at ©• = 0 31
Figure 22. Relative settler length vs performance a t 9 = 20°,

40° 33
Figure 23. Angle of inclination vs performance using Yao's

approach 33
Figure 24. Turbidity lemoval efficiency vs flow rate using

Chen's data 36
Figure 25. Inlet water turbidity vs efficiency using Chen's

data 36
Figure 26. Flow velocity vs efficiency using Chen's data 37
Figure 27. Tube's relative length vs efficiency using Chen's

data 37
Figure 28. Performance of upflow vs downflow tube settlers

after Yao 39
Figure 29. Suggested single tube dimensions by Willis 39
Figure 30. Schematic diagram of modules in primary clarifier

for the experiment conducted by Van Vliet 49
Figure 31. General performance of the clarifiers and the

modules in Van Vliet experiment 49
Figure 32. Sedimentation volume classification according to

El-Baroudi and Fuller 52
Figure 33. Schematic diagram for the model tank used in El-
Baroudi and Fuller experiment 53
Figure 34, Langhaar calculated velocity profiles 61
Figure 35. Schematic diagram illustrates the particle path

inside the tube using Fadel's model 64
viii
Figure 36. The upper surface of a steeply inclined tube bundle

in a sedimentation tank. 70
Figure 37. Suggested inlet-outlet arrangement for an

essentially horizontal tube bundle in a
sedimentation tank 71
Figure 38. Particle trajectory for Example 1 81
Figure 42. The effect of increasing tube diameter on the

required tube length using Fadel's model 85
Figure 43. The effect of the degree of inclination on the

Figure 44. The effect of increasing the flow velocity on the

Figure 45. The effect of increasing the particle settling

velocity on the required tube length using Fadel's
model 89
Figure 46. The effect of temperature increase on the required

tube length using Fadel's model 90
Figure 47. Yao's shape factor equal the area under the
parabolic velocity profile divided by the area
under the uniform velocity profile 92
Figure 48. The ratio of the required length by Yao to Gulp's

calculated length for 60° tubes 93
Figure 49. The design chart for 2" diameter tubes 95

IX
Figure 54. versus Vav/Vsc for 2" diameter tube 103
Figure 59. Temperature effect on the "n" value 108
Figure 60. Sludge accumulation effects on the required length

for an essentially horizontal tube settler with a
3-inch diameter 111
Figure 61. Sludge accumulation effects on the required length

for essentially horizontal tube settler with a 4-
inch diameter 112
Figure 62. Schematic diagram for the experimental equipment 114
Figure 63. Schematic program for the modified experimental

equipment 116
Figure 64. Schematic diagram for the three tubes used in the
experiment 117
Figure 65. Schematic diagram presents the elutriation process

equipment used in the experiment 122
Figure 66. Glass beads (bulk) used in the study 143
Figure 67. Glass beads settled at the tube end, cup 1^1,
Run #16 145
Figure 68. Glass beads settled at the tube end, cup /'3,
Run #16 145
Figure 69. Glass beads settled at the tube end, cup #5,
Run 1^16 146
Figure 70. Glass beads settled at the tube end, cup #7,
Run #16 146
Figure 71. The relationship between the depth reduction ratio

(ds/D) and particle diameter in Set //I 150
X

(ds/D) and particle diameter in Set #2 152

(ds/D) and particle diameter in Set #3 153
xi
DEDICATION
To my wife, Rawia
1
INTRODUCTION
Separation of suspended solids from water is one of the oldest and
most important problems in water and wastewater treatment. Sedimen
tation i s the most widely used solids-removal process; approximately
one-third of the total capital expense of an entire water treatment
plant i s devoted to i t s sedimentation tanks. Several attempts have been
made to reduce sedimentation costs. These efforts range from attempts
to increase sedimentation basin efficiency to elimination of
sedimentation tanks altogether by using direct filtration processes
instead. However, direct filtration processes are limited in
application because they cannot be used to treat high turbidity water (>
200 NTU), nor are they applicable to waste water treatment.
Certain technical advances can reduce the size and cost of
sedimentation units. One such advance i s the use of high rate
sedimentation, which usually reduces the liquid retention time in the
settling device by reducing the distance necessary for the particles to
settle in effecting their removal. These devices are normally tubes;
parallel plates or other shapes are also considered as being tubes placed
either horizontally or inclined at some angle to the horizontal.
Hazen (1904) suggested the idea of shallow-depth settling, Camp
(1946) [4] explored i t , and Hansen and Culp (1967) [5] demonstrated i t s
practical application. Sedimentation-tanks incorporating tube with
detention times of 15 minutes or less settlers can now achieve settling
efficiencies comparable to or better than those normally obtained in
conventional rectangular settling tanks generally having detention times

2
of two hours or more.
There are only two basic configurations of tube settlers, the
"essentially horizontal" and the "steeply inclined." Essentially
horizontal tubes have an angle of inclination, 0, less than 7.5°, while
steeply inclined tubes have 9, as any angle up to 60° (8 is the angle to
the horizontal). Several theoretical and experimental studies evaluating
the performance of tube settlers under different conditions of flow rate,
suspended solids constituents, tube lengths, tube diameters, and angle of
inclination have been conducted. These studies indicate that the tube
settler concept needs further exploration to provide operating
information and to verify design criteria, especially in the "essentially
horizontal" tubes.
3
REVIEW OF LITERATURE
Sedimentation may be defined as the separation of suspended
particles by gravitational settling. When solids are heavier than water,
four types of particle settling phenomena, or sedimentation, may occur.
These types and their definitions follow:
a. Discrete particle sedimentation: Particles settle
individually, having no significant interaction with neighboring
particles.
b. Flocculant particle sedimentation: Particles flocculate or
increase in mass and size during settling, thus they settle at a faster
rate.
c. Hindered sedimentation: Particles are affected by interpartide
forces which hinder their settling, thus causing them to remain in fixed
positions relative to each other and causing a liquid-solid interface to
develop at the top of the settling mass.
d. Compressive sedimentation: Particles are much more highly
concentrated than in hindered settling, thus forming a higher particle
concentration "zone" or structure in which further settling may occur
only by compression of the structure.
Neither hindered nor compressive settling have any significant
effects on the design of tube settlers. Therefore, the present study i s
limited to analysis of discrete and flocculant particle sedimentation.
Discrete Particle Settling
Newton's and Stoke's laws are used to analyze discrete particle
settling. Newton's law allows for calculation of the particle's terminal

4
velocity by equating the gravitational forces on the particle to i t s
frictional resistance, or drag. For spherical particles, the law yields:
i g(P. -P)d ^
' c - J C^P »>
where :
= particle settling velocity
g = gravitational acceleration
Pg, P = the density of the particle and the density of water
d = diameter of the particle
Cg = drag coefficient.
The value of the drag coefficient, C^, depends on whether the flow
regime surrounding the particle i s laminar or turbulent. Figure 1, after
Metcalf and Eddy-[17], shows the drag coefficients of three particle
shapes plotted against the flow regimes represented by their Reynolds
^cd
numbers (R^ = ^ ). Stoke's law may be applied if the Reynolds number i s
less than 0.3. The law i s stated as:
gCp -P)d^
'c'-lsTT— «'
Sand i s a good example of a discrete particle.

5
100 000
"vV Disks
10.000
1.000
too Cylinders
0 0001 0.01 10 100 10.000 1.000.000
RevnolcJS number.
Figure 1. Drag coefficients of spheres, disks and cylinders
^ Sampling
i ports
Figure 2. Settling column and settling curves for flocculant particles

6
Flocculant Settling
Particles may coalesce during sedimentation. When this occurs,
particle mass increases and particles settle faster. Flocculation
depends mainly on the individual particle's opportunity for contact with
other individual particles. These opportunities vary with different
sedimentation processes, flow rate conditions, basin depths, system
velocity gradients, and particle concentrations and size ranges.
A settling column provides the simplest test method to determine
this type of sedimentation requiring only a column having a 150 mm
diameter and a height of 3 m for satisfactory performance. The test i s
easy to make and the results are readily available. Column sampling
ports should be spaced at 0.6 m intervals, with samples collected at
varying intervals of time.
Uniformity of temperature and particle size distribution from the
top of the column to the bottom should be carefully monitored and
controlled. After analysis, the percent of suspended solids removed may
be computed and plotted against time and depth. The resulting plotted
points are then drawn as curves showing the percent of suspended solids
removal. A settling column and settling curves for flocculant particles
are shown in Figure 2.
The percent of suspended solids removal may be calculated as :
5 hi
Percent Removal = £ —r—
\ ^i+1
(3)
i=I "5
However, the column test provides only an approximation of overall
particle settling velocity and removal efficiency.

7
Flow Regimes
Flow regimes are usually classified according to their Reynolds
number, RN, which expresses the relative magnitude of the accelerative
(inertial) and viscous forces. In parametric form, this relationship i s :
RN = — (4)
V
where :
RN = Reynolds number
V = a relative velocity between fluid and boundary;
L = a dimension (such as sphere diameter, depth of flow, etc.);
V =— the kinematic viscosity of the f l u i d .

P
This relationship clearly indicates that the influence of viscosity
is greater when Reynolds numbers are smaller, and less when Reynolds
numbers are larger. The inclusion of viscosity allows for two vastly
different flow conditions, laminar and turbulent. Laminar flow is
characterized by fluid movement in parallel layers without crosscurrents,
whereas turbulent flow i s characterized by pulsatory crosscurrents.
Laminar flow conditions exist only at low Reynolds numbers so, as the
Reynolds number increases, a stage eventually must be reached at which
the flow becomes unstable (or turbulent); disturbances of a particular
magnitude will then result.
Reynolds number for laminar flow conditions occur within a
given range of low numbers (RN < 2100) and turbulent conditions occur
within a higher range (RN > 4000). Between these values, either laminar
or turbulent flow may occur (depending on whether the starting flow

8
conditions were laminar or turbulent). Thus, laminar flow conditions may
exist up to RN = 4000, the flow velocity increases gradually. This
region has been termed the "transition zone."
Sedimentation Theory
Camp (1946) [4] defined an "ideal basin".as a settling tank in which
settling takes place exactly as i t would in a quiescent settling column
of equal depth. Such an ideal basin would be rectangular with continuous
flow, and would have the following characteristics:
(1) Horizontal flow, with uniform velocity in all parts of the
settling zone.
(2) Uniform vertical distribution of suspended particles
at the inlet zone.
(3) Removal of particles from suspension Waen they reached the
bottom of the settling zone.
The settling paths of discrete particles in the settling zone of an
ideal basin are illustrated in Figure 3, after Camp. A discrete
particle will have two velocity components in the ideal basin, and V.
The downward settling velocity of the particle itself is represented by
, and the velocity of the moving fluid by V, making the particle's
actual velocity the vector sum of these two components. Both the
trajectory of a discrete particle having a settling velocity slow enough
to be barely sufficient for particle removal and the different velocity
components in the basin are illustrated in Figure 4. This resultant
settling velocity may be termed the "critical settling velocity," or
All particles having a settling velocity greater than or equal to

Figure 3. Settling paths of discrete particles in an ideal basin [2]
DISCRETE
PARTICLE
Figure 4. Different settling velocities in an ideal basin

10
will reach the bottom of the tank and be removed; however, particles
with settling
° velocities lower than Vsc will not reach the bottom of the
tank unless they enter some distance below the top. Different settling
velocities in an ideal basin are illustrated in Figure A.
Dick (1969) [7] suggested a trajectory for a flocculant particle
illustrated under ideal basin condition, in Figure 5. By definition, the
flocculant particle must increase in mass due to agglomeration with other
particles along i t s settling path, thus becoming heavier and larger with
time and gaining more drag through the increased area. Therefore, i t
settles faster as i t progresses through the tank. Figure 6, after Dick
[7], shows the effect of reducing the depth of a settling basin receiving
flocculant particles. The flocculant particles will not be removed.
Development of Tube Settlers
The development of tube settlers resulted from experimentation in
settling tanks and basins. Camp (1953) [3] reported a 1952 modification
to sedimentation basins in Cambridge, Massachusetts, in which trays were'
added, increasing the settling capacity to nearly three times that
available in the old basins. Figure 7, he concluded that adding trays
increased the surface area, and significantly decreased surface loading.
Addition of trays also proved that sedimentation i s independent of depth:
between two sequential trays, each space worked as a separate
sedimentation tank having the same surface area and surface loading as
the tank before modification.
Fischerstrom [10], a later researcher, thought tray spacing was
limited by the solids removal problem. Suggesting that the space between
11
Figure 5. Flocculated particle trajectory in an ideal basin[7]
-r LUCuLLcN
JC RTCLE
_lpj NOT
ARTICLE ReMO^-ED
Figure 6. The effect of reducing the depth of a settling basin

receiving flocculant particles [ 7 ]
12
p w 'tirtn
,LU L'J..LL-1. -- —rff
Ploeeuiotion Chamber# S# fiUmg 8c9in 0#3'*m dcpifi

(o) Tn r 0 u 9 h-f I o«r type of basin
Filf#r ootrorirtf
qell try
Pip# a0li try
To fiftitt
C l> • m < c 0 1
0 » t » # CH
rr«ccw(cn*ff S« ftliMf SOIlA
(b) Rtturn-fio« fyp# of bosin
Figure 7. Tray in tank provides added floor area and increases solids
removal [3]
13
two trays be large enough to allow for sludge removal equipment,
Fischerstrora reported that one should also add vertical baffles to reduce
the values. Both vertical baffles and horizontal trays increase the
wetted perimeter, reduce the hydraulic radius, and accordingly, reduce
the Reynolds number.
Joining these three factors (wetted perimeter, surface loading, or
overflow rate and tank depth) to problems experienced in adding trays to
shallow sedimentation tanks. Gulp and Hansen [5] proposed the idea of
tube settlers. When stacked in layers, tube settlers increase the
sedimentation surface area by a number equal to the number of layers
stacked, so that a stack of three tube settlers has three times the
surface area of one, a stack of six has six times the surface area, and
so on. The wetted perimeter also increases by an equivalent number over
that of an ordinary settling basin having the same volume as that of the
tubes. According to the above discussion, there are two phenomena that
have contributed to the development of tube settlers.
Phenomenon 1 : The settling of discrete particles i s
independent of the sedimentation tank depth. This can be
proved as follows:
The velocity of the settling particle in the direction of flow,
V, is
where Q is the flow rate through the basin, w is the width of
the basin, and d i s the depth of the basin. From similar
triangles, one may derive

14
- d
V L
where = particle settling velocity due to gravity, and
= velocity i s the velocity of the smallest completely
removed particle.
L = the length of the basin.
Accordingly,
V (6)
sc L dwL A ^ /
where A = wL = basin surface area.
Equation (6), ^ i s called the "overflow rate," "surface
settling rate," "surface loading." The surface loading i s the
basic design parameter in water clarification. Equation (6)
implies that removal of free settling particles in an ideal
settling basin depends on surface loading but not on volume,
indicating that removal i s entirely independent of the basin's
depth.
Phenomenon 2: Laiainar flow conditions yield better settling
performance. This flow regime depends on the boundary
conditions of the settling tank. Gulp et a l . [6] reported that
most tanks at that time were operated at RN values of 1000 to
25,000. Fischerstrom [10] reported that sedimentation tanks
should be operated in the laminar region if good sedimentation
were to be obtained. The R value for laminar flow in wide,

15
open-channel flows has been found to be 500 where RN = ——,

V
and d' is the depth of the channel. Since the laminar region
for wide channels i s considered to be less than 500, Gulp et
a l . data indicate that most tanks were being operated at flow
rates several times the optimum sedimentation rate.
Good, economical sedimentation processes will be obtained by incorpor
ating these phenomena in the design operations of sedimentation
facilities.
Tube Settlers
A typical tube settler consists of a modular unit of thin-walled
tubes having a small cross-sectional area. The units are relatively
lightweight, since they are usually made of PVC plastic or similar
materials. The cross-sections of the individual tubes may be circular,
square, rectangular, or even hexagonal. .Six possible tube designs are
illustrated in Figure 8. Tube dimensions range from 2 to 3 inches wide,
with practical lengths ranging from 2 to 8 feet.
The two basic tube configurations shown in Figure 9 were described
by Gulp e t a l . [6] as:
(a) Essentially horizontal, in which tubes are inclined to the
horizontal at an angle ranging from 0 to 7.5°; and
(b) Steeply inclined, in which the degree of inclination to the
horizontal is approximately 60°.
Gulp et a l . also showed that tube settlers in the steeply inclined
position can be made into a "self-supporting beam" by alternating the
direction of tube inclinations within the module. Such alteration

prevents short circuiting of flow by limiting current velocity. Short-
circuiting of flow may result if a l l tubes are inclined in the same
direction. An example of a self-supporting beam module i s shown in
Figure 10.
Yao [28] reported that tube settlers can be classified according to
the direction of flow as upflow and downflow high-rate settlers. He
reported that an angle of inclination of 30° may be used for the downflow
tube settlers, and that this angle of inclination i s sufficient to obtain
self-cleaning action.
Several methods of arrangement have been proposed, and put into
practice, for tube settler location in sedimentation basins. Two
treatment units manufactured by Neptune MicroFloc, Inc. (USA) are shown
in Figures 11 and 12 (after Gulp [5]). One uses a nest of horizontal
tube settlers in which the tubes are backflushed by the backwashing water
of the filters. The unit must be dewatered before backflushing if this
treatment method i s to be used. The other treatment method uses a
steeply-inclined, self-cleaning tube having an inclination of 60° to the
horizontal.
Another high rate settler configuration i s the "Tilted Plate
Separator," reported by Yao [28]. Used fairly widely in Europe and
listed by Chemical Engineering magazine as being among the top ten most
popular new equipment items introduced in the year ending July, 1969 , the
device consists of closely spaced inclined parallel plates. A downflow
and upflow Tilted Plate Separater i s illustrated in Figure 13. Yao
mentioned that an upflow parallel plate separator unit was successfully
used in an extended aeration system called Aqua-Reuse; the plant had a

17
Figure 8. Possible tube cross-section designs
COAGUUNTS PORTION OF 0W
FOR TUBE WASHING
FLOC- JUBE SETTLER

CULAT- FILTER
OR
WASHWATER DRAIN
ESSENTIALLY HORIZONTAL TU9E SETTLER
BW TO
COAGULANTS WASTE
RAW <
WATER
TUBE
•SETTLER
FILTER
SLUDGE
ORAWOFF
STEEPLY INCLiNEO TUBE SETTLER
Figure 9. Basic configuration for tube settlers

18
Figure 10. Self-supporting beam module of steeply inclined tubes
Calctte -Settling • Backwash Clearwell

Column -Chamber" Storage
Raw Water
Use
Piping
Filter Cycle
——— Backwash
Figure 11. Basic arrangement of horizontal tube alignment [6]

19
EJîtuint Liurdef
Settling Tjr.h
Flcccwijtor Tanfc
(2 tl 0 <n X4 tt 6 in.%6 tt 0 m.)
i2S0 Cil)
Sluûge O'io Oit^
Figure 12. Basic arrangement of inclined tube alignment [6]
OIL SEPARATION SOUO-LIQUlO SEPARATION
Figure 13. Dovmflow and upflow tilted plate separator [28]

20
backwash arrangement to keep the unit clean. The plates in this unit are
0.375 inch apart and inclined a t 60° to the horizontal. Suspended solids
removal efficiency was found to be 95%. Backwashing water was supplied
by an air l i f t pump operated by a timing device and using the same air
source as aeration. Both normal operation and backwashing for two
different Aqua-Reuse Plant configurations are shown in Figure 14.
Willis [25] suggested the arrangement for steeply inclined tube
settlers shown in Figure 15. Proposing a rectangular sedimentation tank
with steeply inclined nested tube settlers, he recommended that
distribution headers similar to the underdrain of rapid sand filters be
placed below the inlet face to the tube bundle. On top of the tube
bundle, Willis suggested another submerged discharge bundle, or launders,
with adjustable weirs to control flow distribution through the tubes.
Several arrangements for tube settlers in industrial and waste water
treatment plants can be found in the literature.
Design and Experimental Approaches
Unfortunately, l i t t l e has been done to explore tube settler
concepts. Out of eleven published papers, only five have design
approaches; of these five, only three include data from experimental
investigations. These papers are summarized below. Comments follow each
summary.
Gulp et a l .
Gulp et a l . [5,6] considered a straight line trajectory for the
particles in the tube settlers shown in Figure 16. If one knows Vq
(average flow velocity in the tube) and the minimum settling velocity
21
IN VENT
VENT r-AIR
_fL
—PLATE
SETTLER
^AIR DIFFUSER
NORMAL OPERATION BACKWASH
Figure 14. Normal and backwash operations of the Aqua-Reuse plant [28]
Co'ieci'om Wt<«
4 T'owgn or
J / /
- Tuoe *nçie 9 W CfÇ

Z.iifewiiom Htaoff ~
^ ,/ e e
./
o o #— o o
\.
o o\
— P'lm* ol BfCe'Timg A'fi

Ol TuBf Gunce A|«< 0'
P«»n» ij me Piee A'h *«
Figure 15. Suggested arrangement for steeply inclined tube settlers by

Willis [25]
22
for the particles to be removed, V , one can then assume a reasonable

sc
tube diameter of 2 to 8 inches; the required length i s then obtained by
triangular similarity, according to the equation below.
d(V„ - V sin 9)
'•s • V e—
sc
where: = the required settling length
d = tube diameter
0- = degree of inclination of the tube from the horizontal
V q = the average velocity across the tube.
Gulp et a l . stated that at a certain V- and V , where V. > V , the

• 0 sc 0 sc
required length of the settling surface decreases as the angle increases
from zero to a certain 6; this 0 value depends on the ratio of to
V , e . g . , for V„ = 2.5 V , 0 will be between 25-30 degrees. After

sc 0 sc
this angle i s reached, the required length increases. Accordingly, as
the angle of inclination (6) i s increased to 90°, the length required
approaches infinity.
Gulp et a l . conducted two sets of experiments on circular high-rate
settlers. In the f i r s t set, four tubes having diameters of 1/2 i n . , 1
i n . , 2 in., and 4 i n . were used. Three different tube lengths of 2 ft.,
4 f t . , and 8 f t . , based on the tube cross-sectional area, were used.

2
Flow rates of 2, 5, and 8 GPM/ft were used. Tlie raw water turbidities
used were 150 or 450 Jackson units. Some of the runs were conducted
using a polyelectrolyte dose of 0.2 and 0.5 mg/l. The second set of
experiments was designed to check the effects of the degree of
inclination on the tube's performance by testing five tubes each 4 feet

23
Direction
of flowy
Discrete
particie
y, (resultant)
'sc
Figure 16. Straight line trajectory for the particles in tube settlers
suggested by Gulp et a l . [6]
24
long and having a 1- inch diameter. The five tubes were re-tested
individually and inclined at angles of 0, 5, 20, 40, and 90°. Figure 17
illustrates a schematic diagram of the equipment used in his two
experiments.
From the f i r s t series of experiments. Gulp et a l . concluded:
(1) Smaller tube diameters enhanced effluent turbidity reduction at
the same flow rate and the same inlet turbidity. Figure 18 shows the
effects of tube diameter on the turbidity percent removal at different
conditions of flow rate and tube length.
(2) Longer tubes performed better. Indicating that smaller
diameter long tubes would perform better than large, short ones. Figure
18 shows both effects on the removal efficiency of the tubes.
(3) Lower flow rates are recommended in order to increase the
settling efficiency; Figure 18 indicates that a low flow rate through a
small, long tube yields the best performance.
(4) Polyelectrolyte coagulation ahead of the tube bundle
significantly increased removal efficiency.
(5) Inclining the tubes slightly in the direction of flow permits
sludge removal by gravity drainage and eliminates the need for mechanical
sludge removal equipment.
(6) Horizontal tube settlers provide at least 24 hours of sludge
storage.
In the second experiment. Gulp et a l . observed that :
(1) Tube performance decreases slightly when the degree of
inclination increasnd from zero to 10 - 20 degrees.
(2) Increasing inclination from 10 - 20 degrees to 40 - 45

25
degrees increases efficiency. However, past 45 degrees, efficiency
decreases again until i t reaches i t s lowest value with the tubes inclined
at 90 degrees.
Figure 19 illustrates Gulp et a l . observations; unfortunately, Gulp
et a l . did not explain the difference between the two curves shown in
this figure.
Comments
Gulp et a l . assumption that the particle's settling path would be a
straight line was based erroneously on an assumed uniform flow velocity
through the tube. In fact, for laminar flow the tube's actual velocity
profile varies from uniform flow at the entrance to a fully developed
laminar velocity profile after a certain flow length. This variation
increases the required length for particle settlement. Therefore,
assuming a straight-line trajectory led Gulp et a l . to underestimate the
required tube length.
Gulp et a l . experimental method was poor, in that he controlled the
flow through the tubes by using a pinch clamp at the outlet end of each
tube and measured turbidity using a light transmittance colorimeter.
This i s a poor turbidity measuring device; there were better devices
available at the time.
Gulp et a l . mentioned the non-seIf-cleaning tubes had an advantage
in that the build-up of solids in the tubes would cause a self-orificing
of the flow leading to even distribution using the Neptune micro-floc
unit. Equal flow distribution through the run i s essential for
successful operation. This implies that the inlet for the system
26
Flocculated Water
• Inlet Plenum
o
CC
jO~O^—
Cn —
CT
• Flow Control Valve
[-in. Oiamx4«tt Long Tubes
To Drain
Flexible Plastic Tubing

for Connection
2 Vertical
Entry Ports
Inlet Plenum
3 Horizontal Entry Ports
Mag-Mixer UseO to Prevent
Flocculated Water Deoosiiion in Inlet Plenum
Elevation
Figure 17. Schematic diagram for the experimental test equipment used by
Gulp et a l . [ 6 ]
27
1-in. Diam. Tube

2-in. Diam. Tube
4.in. Oiam. Tube
Flow rate was 5 gpm/sqft; ahtm dosage

Flow rate was 2 gpm/sqft; alum dosage,
100 mg/l; polydectrolyte dosage, 0.5
100 mg/l; flocailation time, 20 min; and
mg/l; jlncculaiion time 20 min; and
average influent turbidity, 450 Jackson average influent turbidity, 450 Jackson
units. There was no polyelectrolyte
units.
dosage.
Figure 18. The effect of tube diameter, tube length, flow rate and
flocculation on the percent removal of turbidity [ 6 ]
28
90
80
0 10 20 30 40 50 60 70 30 50
Tube Inclination'deg
Figure 19. The effect of the tube settler's degree of inclination on the
percent removal of turbidity [6]
(Figure 11) was improperly designed.
Yao
Yao [27] assumed that flow conditions in a high rate settler were
laminar and particles we ce discrete. Essentially, Yao separated the
velocity components in a manner similar to Gulp et a l . (Figure 20). The
velocity u i s the local fluid velocity. Using the angle of inclination,
0, and the following velocity components in the x and y directions, Yao
defined :
u = u - V sin 0 (3)
X sc
u = -V cos 9 (9)
y sc
By definition, u^ = dx/dt and u^ = dy/dt. Dividing u^ and leads
to a trajectory equation resulting from the combined effects of fluid
drag and gravitational settling, as follows:
= -V cos 9/u - V sin 0 (10)

dx sc sc
The integration of Equation (10) leads to the following equation:
udy - V y sin 9 + V x cos 9 = c (11)

sc sc
Dividing a l l terms in Equation (11) by , the average flow velocity
(which i s equal to the flow, Q, divided by the area normal to the tubes)
and d (the depth of the flow normal to u) leads to the general equation
for particle trajectories shown below:

30
u/V dy - V Y sin 0 + V /V X sin e = c, (12)

o sc — sc o 1
where :
= an adjusted integration constant
= y/d, and
X = x/d
According to Camp's analysis of settling velocities in an ideal basin,
particles following trajectory (Figure 20) have the lowest settling
velocity; thus, they represent the "critical settling velocity" used
in the above derivation.
Evaluating different tube settler shapes by using the general
equation for particle trajectories and the relationship between u and V^.
Yao concluded that, for the limiting trajectory, , there are two
boundary conditions which may be used to arrive at a design equation for
each specific tube shape. These boundary conditions are as follows:
X = L at ^ = 0 and X = 0 a t Y^ = 1.
L = the relative length.
= 1/d (the length of the settler divided by the distance
normal to the fluid velocity i . e . , the tube diameter.)
The integration constant, c^, in Equation (12) evaluated by substituting
the f i r s t boundary conditions into the trajectory equation was evaluated
for the specific settler shape, as follows:
V /V (sin 0 + L cos 0) = S (13)

sc o c
Yao found the value of to be: 4/3 for circular tube settlers, 1 for
31
TRAJECTORY
e. ANGLE OF INCLINATION
OF HIGH-RATE SETTLER
Figure 20. Coordinate system for the design model by Yao
0 = 0=
.6
'sc
.4
,*v^ClRCULAR TUBE
-PARALLEL PLATES
80 100
Figurii 21. Relative settler length vs performance at 9 - 0

32
parallel plates, 11/8 for square conduits, 1 for shallow open trays, and
1 for systems with uniform velocity distribution.
The relationship for the ratio (u/V^) used in deriving the
values i s a result of the fluid's viscosity. Viscosity introduces a
resistance to motion, creating a non-linear velocity profile
distribution. Evaluation of (u/V^) for circular tubes i n laminar flow

2
results in a parabolic shape, described by the equation 8(y - y ) ,
From the settling characteristics of the solid the settler
shape (S^), the angle of inclination (0), the spacing or height of the
settler (d), and the flow rate, one can determine the required tube
length using Yao's Equation (13).
For a fixed V , Yao showed that the value of V decreases as L

0 sc
increases (when L is relatively small). The rate of decrease in
drops appreciably after L reaches 20 and becomes nearly f l a t after L
reaches 40, as illustrated in Figure 21. The angle of inclination chosen
for this graphical representation was 0 = 0 ° . Since this angle of
inclination i s impractical, Yao presented another drawing. Figure 22, for
0 = 20 and 40°, i n which he obtained the same pattern as that for 0 = 0 ° .
The curves flatten, after L reaches 40, so increasing L reduces the V

sc
value very l i t t l e ; hence, L should be kept below 40 and probably closer
to 20 for optimal economical design.
Yao found that the optimum angle for a minimum value to be:
0 = tan ^(1/L) (14)
For L equal to 20, the optimum inclination angle becomes 2°52'.
Performance for various angles of inclination are shown in Figures (23);

33
.8
PARALLEL PLATES
.6
.4
.2
0 =40
0
0 20 40 60 80 100
L
Figure 22. Relative settler length vs performance at 9 = 20°, 40°

[27]
L»
(00
,30
0* 90'
e
Figure 23, Angle of inclination vs performance using Yao's approach

note that performance decreases rapidly beyond a value of about 40°.
In the same study, Yao recommended that designers consider the
entrance length required to establish a laminar flow and stated that for
practical installations, the settlers will probably be connected to an
inlet chamber having a relatively large sectional area. At the entrance
to a settler, a transition region exists in which uniform flow gradually
changes into fully developed laminar flow, due to the influence of the
solid boundary and fluid viscosity. In circular tubes, the relative
length, L' (relative length equal to the transition length divided by
tube diameter), for this transition region may be estimated using
Langahaar's equation:
V d
L' = 0.058 (15)
This length could be added to the design length of the tube settlers. If
the required entrance length i s longer than the design length, Yao stated
that the total should be taken as twice the design length. It i s
noteworthy that this equation i s based on the fact that the flow has a
uniform velocity value at the entrance face of the tube.
Yao [28] later reviewed Chen's thesis data (National Taiwan
University). Chen conducted an extensive experimental study of circular
upflow high-rate settlers similar to those described by Gulp et a l . [6] ,
using four tube sizes (0.5, 1, 2, and 3 inches), three tube lengths (2,
3, and 5 feet), and four flow velocities. Raw water was taken from an
irrigation canal and a koalin suspension was added to adjust the
suspended solids level to the desired level. Chen conducted three sets
of runs, the f i r s t set of which used a constant angle of inclination of

60° at ten different raw water suspended solids concentrations, ranging
from 15 mg/l to 90 mg/l. The second set held raw water suspended solids
at a constant 50 mg/l, with six different angles of inclination: 0, 15,
30, 45, 60, and 75°. In the third set, Chen lowered his raw water solids
level to 50 mg/l; other experimental conditions duplicated those of the
second set. An average alum dose of 20 mg/l was used with flocculation;
Reynolds numbers ranged from 15 to 370 and relative lengths ranged from 8
to 120.
Using a digital computer, to analyze Chen's data, Yao illustrated:
(a) Turbidity removal efficiently decreased as the flow rate
increased. Figure 24.
(b) Higher raw water turbidities yielded better efficiency. Figure
25.
(c) Better flocculation before settling and particle aggregation
during settling could improve efficiency.
(d) Higher flow velocities yielded lower removal efficiency. Figure
26.
(e) There was no definite trend indicating that settler efficiency
would be adversely affected if the angle of inclination exceeded a
certain limit within the range studied, Figure 24.
(f) Bemoval efficiency was not affected by the wider range of
relative lengths used (8-120), Figure 27.
Yao concluded that circular settlers with flow velocities less than
0.54 fpm tend to perform better than those with flow velocities of higher
than 0.8 fpm. In the same paper, Yao explored the idea of using a down-
flow settler and compared the theoretical efficiency of the downflow

36
OVERFLOW RATE,CMO/SM
^0 20 10 60 80 too
* *0. 0® '
o JO*
0 feS®
A 7 5'
L• B TO 120
' a
A A
0
6S •
3 Oo 0 • ^ o
•
•
0 5 00 1000 1500 2000 2500
OVERFLOW RATE,USGPO/SF
Figure 24. Turbidity removal efficiency vs flow rate using Chen's data
[28]
OVERFLOW RA7E,CM0/SM
0 20 UP 60 SO 100
" 1
RAW WATER
TURBIDITY
mg/l j
- Q 90
0 5 0
^ 3 0
SQI 1^ < I !
0 500 1000 1500 2000 2500
OVERFLOW RATE,USGPO/SF
Figure 25. Inlet water turbidity vs efficiency using Chen's data

[28]
37
OVERFLOW RATE, CMO/SM
0 20 iO 60 50 100
' FLOW'VELOCITY '
GPM/SF FPM CM/MIN
0 2 0.27 e.i
•' fc 0.51 16.3
CJ 0 6 0.80 2k.t
A A 8 1.07 32.4
e - 0* TO 75'
\\
\K. y\
\A
O
-3 D— =•
^^0 500 1000 1500 2000 2S00
OVERFLOW RATE . USOPO/SF
Figure 26. Flow velocity vs efficiency using Chen's data

[28]
OVERFLOW RATE, CMO/SM
0 20 10 60 SO 100
3S| 1 : n
a L- a
A « 12
0 2(,
A . 120 ^
9 - 0 TO 75®
—S.-i
(
' " c0
° .
Ù 0 • X
)
30 0 0
'^0 500 1000 1500 2000 2500
OVERFLOW RATE, USGPO/SF
Figure 27. Tube's relative length vs efficiency using Chen's data

[28]
settler versus the upflow s e t t l e r . His study illustrates that down-flow
settlers tend to give better performance if the settling system must rely
on a self-cleaning action for sludge removal. Figure 28, after Yao,
shows that a lower V can be achieved for a down-flow settler inclined

sc
to the horizontal at 30° than for upflow settlers inclined at 60° to the
horizontal when L i s above 4 .
Comments
Yao assumed fully developed laminar flow exists in the tubes, a
condition which exists only after a transition length; in shorter tubes
the flow will be a mixture of uniform and laminar flow. Since the
performance of a high-rate tube settling system with uniform flow i s
either comparable to or better than that of a similar system with laminar
flow, as indicated by the values, this suggests that if the transition
length i s long enough ( i . e . , L' 2 L) the required particle may settle
before i t reaches the region in which laminar flow i s fully developed.
Accordingly, the laminar flow trajectory Yao proposed may not exist.
In a second paper [28], Yao's conclusions contradicted those in his
f i r s t [27]; according to his theoretical studies, the tube's performance
should deteriorate rapidly a t angles of inclination higher than 40°,
Figure 23. In the second paper, he asserted that for angles of
inclination up to 75°, there was no major trend observed which altered
the tube's performance. Another contradiction exists in the observation
that there was no major difference in tube performance at L ranges
between 8 and 1 . According to his theoretical studies, tube performance
should improve rapidly at L ranges between 4 and 20, and remain constant
39
"ael-ao')
"sececfi
0 ZO aO 60 80 100
L - l/d
Figure 28. Performance of upflow vs downflow tube settlers after Yao

[28]
Singif Tu8«
Too View Cantigu'inon Cross Section
Figure 29. Suggested single tube dimensions by Willis

40
at L values higher than 40. The only reasonable explanation for the
contradiction i s that Chen used flocculation, which might produce a much
heavier particle. Such particles settle much faster than discrete
unflocculated particles, regardless of the length and degree of
inclination of the tubes.
Willis
Willis [25] stated that three basic requirements are essential to
successful tube settler performance. These are:
1. Laminar (or viscous) flow conditions must exist within the
tubes at the maximum flow rate required. Laminar flow i s
- essential so that each slowly-settling floe particle within a
tube maintains a steady descent to the collecting surface of
the tube and i s not intermittently swept upward by turbulent
currents within the tube.
2. The residence time within each tube must be of sufficient
duration that a floe particle entering at the extreme upper
edge of the tube will have sufficient time to settle to the
collecting surface a vertical distance below. (Once the
particle reaches the collecting surface, the coalescing
tendency between particles creates a steady sludge formation.)
3. The velocity of flow through the tubes must not exceed a
critical maximum that would cause the settled sludge to lose
stability and be swept out of the tube in the direction of
normal flow. As a corollary, the volume of the tube must be
ample to allow either sludge accumulation or a continual

41
discharge backward of a l l sludge, without critically changing
the normal flow rate through the tube.
Willis' mathematical analysis i s based on the fact that residence time
for the flow within the tube settler, t^, should be equal to or greater
than the time required for the smallest particle which i s to be removed
to settle the longest vertical distance, t^, i . e . , >_ t^. Figure 29.
Applying this concept, he derived the following equation:
1 A, (sin 6)
>_ 3 m/cos ©• (16)
KQ
where :
1 = tube length
A^ = the face area of the tube
K = i s a factor allowing for the tubes wall thickness of the tubes
plus any dead spaces around the tubes,or is
Total area of the tubes inlet face

Open area in the tubes
9- = angle of the tubes to the plane of the face area
g = a constant related to the settling rate and the units of
measure
m = the settler's diameter .
Willis proposed that a Reynolds number of 400 or less should be
considered in tube settler design. Also, he suggested a flow rate of .25

2
GPM/ft , based on the tube cross-sectional area, in order to eliminate
or reduce sweep-out of settled particles in both the horizontal and

42
inclined tubes.
His design equation led Willis to state that essentially horizontal
tubes with a degree of inclination of 7.5° have the advantage of reducing
the maximum drop distance by almost half, when compared with steeply
inclined tubes (60°) of the same size and shape. However, he warned that
the 7.5° tubes are much more susceptible to sweep-out, and that sludge
will accumulate in the tube bottom, thus reducing the tub.e's available
cross-sectional area and, consequently, increasing the flow velocity and
the sweep-out tendency. He also pointed out that 60° tubes are self-
cleaning, eliminating the problems associated with sludge accumulation in
horizontal tubes.
Comments
Willis uses a "theoretical"'approach to derive design equations for
settler length. This approach is incorrect for the following reasons:
1. Willis neglects the negative component associated with the
settling velocity vector.
2. Willis assumes, as Gulp et al. did, a uniform velocity flow
distribution; i t is actually a variable velocity profile which i s uniform
at the inlet and parabolic after the transition length.
3. Willis' design approach was based on the assumption that the
velocity of flow through the tubes must not exceed a critical maximum
velocity that causes the settled particles to sweep out. If a particle
could overcome the high velocity region in the central part of the tube
cross-sectional area and settle, i t would be impossible to sweep i t out
from the tube, especially at the bottom of the tube where the velocity
43
would approach or equal zero.
However, by incorporating Yao's values and the utilization of the
negative velocity component, manipulation of Willis' equations could
produce equations giving the same results as Yao's for any of the settler
shapes.
Hernandez and Wright
Hernandez and Wright [12] used data from several laboratories and
treatment plants. The data covered a wide range of tube variables but
only two angles of inclination, 5° and 60°. Lengths of 2, 4, 6, and 8
feet were used; tube diameters measured 0.5, 1, 2, and 4 inches. Raw
water turbidity ranged from 15 to 450 TU; waste water solids
concentrations (mixed liquor) ranged from 1126 to 2900 mg/1. Hernandez
and Wright also presented data covering the effect of adding
polyelectrolytes on tube settler efficiency. They concluded that tube
settler performance was related to the generalized design criteria
expressed in Eq. (17):
where :
V = velocity of flow in ft/sec, computed from the face velocity

2
of the tube in GPM/ft of the tube entrance area (the area is
perpendicular to the tube).
R = hydraulic radius of the tube in f t .
L = tube length in f t .
Using this factor, they observed the following effects for the different
44
variables on the tube's performance:
(a) Degree of inclination
Hernandez and Wright stated that removal efficiencies appear to
differ for tube angles of 5° and 60°, with the 60° tube nest producing a
v2R
higher quality effluent at the same value of —However, only nine
data points were available for the 60° tubes, vs 145 for the 5° tubes.
Hernandez and Wright observed that the performance of steeply inclined
tube settlers is not influenced as greatly by the physical and chemical
nature of the suspended particles as i s the performance of essentially
horizontal units. Also, they noted that good settling results are
V^R
obtained in the 60° tube nests at much higher values of —^ than are
possible with 5° tube nests.
(b) Flow rate, tube length, and tube diameter :
Hernandez and Wright stated that tube performance deteriorates with

V^R
increasing values of — . Also, small diameter tubes perform better
than large ones such that the hydraulic radius of the tube should be kept
as small as practical, regardless of the nature of the raw water supply.
Accordingly, flow rate (or velocity) becomes the critical factor. Low
flow rate i s essential. This, also, led to the observation that long
tubes perform better than short ones.
(c) Polyelectrolytes:
Use of appropriate polyelectrolytic doses had the same approximate
effect on turbidity removal as a 50 percent reduction in the flow rate;
the influence of —— i s diminished by the general floe characteristics.

V^R
Hernandez and Wright presented some maximum values of —^ under different
treatment conditions which yield good performance depending upon the

45
angle of inclination, and the flocculant nature and density of the solids
being removed. These values are on the order of 4 x 10 ^ for 5° units,
40 X 10 ^ for 60° units. Hernandez and Wright suggested the following
three tables to meet these criteria.
2
Table 1. Maximum flow rates in GPM/ft for steeply inclined tube (9 =
60°) [12]
L, ft
D, in 2 4 6 8
1 9 13^ 16^ 18^
2 6 9 12^ • 13^
3 5 8 9 10^
4 4 6 8 9
Standard Coagulants Used

No polyelectrolytes added
a 2
Flow rates above 9 GPM/ft are not recommended.
Comments
2
Hernandez and Wright's proposed design parameter ('^^) > is of
questionable validity. The authors should have set limits for the
V^R
variables, V, R, and L; they did not. For the same value of — , one
may choose high R and low V and rationally expect the same tube
performance as if one choose high V and low R. By similar reasoning, one

46
2
Table 2. Maximum flow rates in GPM/ft for essentially horizontal tubes
(a = 5 ' ) [12]
L, ft
D, in. 2 4 6 8
1 3 4 5 5.6
2 2 3 3.6 4
3 1.7 2.4 3 3.5
4 1.4 2 2.5 3
Standard Coagulants Used

No polyelectrolytes added
2
Table 3. Maximum flow rates in GPM/ft for settling activated sludge
mixed liquors (0- = 60°) [12]
L, ft
D, in. 2 4 6 8
1 2 3 3.6 4
2 1.4 2 2.5 3
3 1.2 1.7 2 2.4
4 1 1.4 1.8 2
should expect the same performance from tubes having large L and high R
values as one would expect from a short tube with a small diameter; the
47
2
V R
two tubes having the same —values. However, experimental results
showed that these conditions do not exist. Long tubes with large
diameter perform less well than short tubes of small diameter. Also,
2
Hernandez and Wright [12] set values for -j— for alum-treated water and
activated sludge mixed-liquor which simply do not apply to other types of
treatment, such as iron-treated water and/or polymer-treated water and
lime-softened water.
Van Vllet
Van Vliet [22] installed both a tube and plate modules in a primary
clarifier, Figure 30. The floe blanket in the primary clarifier was
operated just below the module's intake. The tube unit had a module-end
2
cross-sectional area of 0.93 m ; the vertical height of the tube bundle
was 0.5 m. The tubes were inclined at 30° to the vertical (i.e., 60° to
the horizontal), and had a square cross-section with 50 mm wall-to-wall

2
spacing. The plate unit had a module-end cross-sectional of 0.29 m
consisting of corrugated fiberglass plates, spaced 20 mm apart and
inclined at 30° to the vertical. The water velocity through the two
modules ranged from 3 m/hr to 12 m/hr.
Van Vliet tested the tube settlers and inclined plates using water
treated with four different flocculants:

3+
(a) Ferric chloride dosage of 5 mg/l (as Fe ) plus an anionic
polyelectrolyte dosage of 1 mg/l.
(b) Ferric chloride, as a concentration of 5 mg/l only.
(c) Anionic polyelectrolyte, at a concentration of 1 mg/l only.
(d) No flocculant.
48
The flocculant was applied ahead of the primary clarifier. Figure 31
illustrates the experimental results. Van Vliet reported that the tube
or plates performed better than the clarifier, and removed an additional
60 percent of the turbidity from the normal clarifer effluent.
Performance of the plate module was marginally better than that of the
tube module.
Van Vliet proposed the following model:
TM = aexp O/T^) TCyFR^ + 6 (18)
where :
TM = fitted turbidity of the overflow from tube or plate
module, FTU.
TA = absolute temperature, K° (288 < TA < 298).
TC = turbidity of primary clarifier overflow, FTU 1 < TC < 6•
FR = ratio of module hydraulic loading to primary clarifier
hydraulic loading (1.5 < FR < 6.0).
The symbols a, 3, y, 6, and 0 represent model coefficients arising
from the modeling procedure and assuming a specific set of values for
the tube and plate module at each flocculant dose. The model was
developed from data fitting analysis, using an IBM digital computer. The
model indicates that turbidity removal efficiency should increase with
increases in the clarifier's effluent turbidity. The effect of the
flocculant doses was described by Van Vliet as showing:
"There appeared to be very l i t t l e difference in module
performance for the two low dosing conditions, i . e . ,
in which only the ferric chloride was used. These

49
Inclined plates Inclined tubes
•Jkj Primory
clonfier over
flow Tube module
Monopump
'overflow
Plate module overflow Monopumo
Floe blanket
Figure 30. Schematic diagram of modules in primary clarifier for the

experiment conducted by Van Vliet [22]
Floccuignf dosing conditions ^

Day I througn 4 . PolyelectrolyJe only
OoySttiroughS • Ferric chloride plus polyelectrolyte
Hydfoulic looding conditions:(modules)
Day 1 ond 5 •' 3 mti" '
Day 2 ond 6 : 6 mti" '
Primory clarifier overflow. —'
Day 3 ond 7 • 9 mh" '
Tube module overflow : —
Day 4 and 8 : 12mh''
Plote module overflow •
t «'*
Figure 31. General performance of the clarifiers and the modules in Van
Vliet experiment [22]
50
two conditions were therefore considered to be
equivalent in respect of clarification within the
modules. Although the performance of the modules
deteriorated as flocculant dosage decreased, they
nevertheless produced consistently better overflow
clarities than the primary clarifier, even at zero
flocculant dosage."
Comments
Van Vliet's approach applies only to the condition and configuration
in which he used the tube settlers. His model's equation was restricted
nonhomogenous. One cannot use or apply this equation to design another
system with different water treatments. His conclusion that the plate
module performed better than the tube module i s misleading, since the
plate spacing was less than half the tube diameter, helping to improve
plate efficiency.
El—Baroudl and Fuller
El-Baroudi and Fuller [8] used tracer dispersion techniques to study
the changes in hydraulic characteristics brought about by selected high
rate settler a configurations as compared to those of the standard or
conventional counterparts. They found that increasing surface area will
increase removal efficiency and that introducing tubes and plates will
induce optimum hydraulic characteristics, thus enhancing sedimentation.
They guided the flow to reduce water depth and Reynolds numbers, thereby
creating better laminar flow conditions with better velocity
distribution.
El-Baroudi and Fuller divided the high rate settling tank under
ideal hydraulic conditions into three zones, Figure 32. Tliese were:
(a) The inlet zone, , in which a complete mixing condition was
encountered wherein a l l properties were the same at any given instant,
and equal to the inlet zone's effluent.
(b) The horizontal flow in volume 2, (Vg); flow is laminar and
velocity is uniform along the water depth, so no turbulent mixing takes
place. This horizontal velocity is reduced linearly along the length of
this zone because of the uniform diversion of flow to volume 3.
(c) The upward flow in volume 3, ( ) , in which the inclined tubes
or plates are placed, i s laminar with a uniform velocity.
The experiment was conducted using a model settling tank divided
into two units, as shown in Figure 33. One unit worked as a standard
sedimentation tank. I t was placed in parallel with the second unit,
which represented a high rate settler. The "tank i s described in more
detail in the reference. El-Baroudi and Fuller presented a mathematical
and graphical solution in which various values for were assumed
Vi
and (t/T) - V„ were calculated for various ratios of — ^. Also,
m 3 ?! + *2
values of (C/c)^ were calculated and plotted for assumed + Vg
fractions and — — ratios, where:

1 + 2
C = the time elapsed after injection of tracer weight divided
by (V^ + Vg + Vg).
c = the desired efficiency concentration.
m = maximum tracer concentration values in effluent.

52
INLET OUTLET
''/'/y/
INFLUENT
EFFLUENT
0
V
Figure 32. Sedimentation volume classification according to El—Baroudi

and Fuller [8]
53
INLET SETTLING OUTLET

15' 033
TRACER
INJECTION
METER' TUBE MODULE 12.0
EFFLUENT
PUMP.
[r—^rn
CONSTANT
MEAD
STORAGE 3.83
METER
STANDARD
SAMPLING
PROSES
Figure 33, Schematic diagram for the model tank used in El-Baroudi and
Fuller experiment [ 8 ]
54
T = tank detention time.
El-Baroudi and Fuller found the high rate model to demonstrate rather
poor detention characteristics and concluded that the tapered flow volume
(Vg) has an effect on tracer dispersion which is equivalent to that of a
completely mixed volume. The experiments emphasized the inçortance of
having an adequate volume (Vg) beneath the tube bundle and demonstrated
that this volume should be as large as 70% of the total tank volume to
increase the tube bundle's efficiency. The large volume i s essential to
reduce the flow velocity in this part of the tank. Headloss for the flow
in this volume will be minor; hence, differences in flow across the tubes
at the first portion of the bundle will not be significantly lower than
at the far end of the bundle.
Comments
El-Baroudi and Fuller did not consider that a fourth volume i s
required at the top of the tube's effluent. Their work i s in error
because this volume i s essential to control the effluent flow for the
tube in the model tank. This i s why they found the high rate model to
demonstrate relatively poor detention characteristics. This volume
should be inserted in the model if the model is to be accurate in
reflecting tube settler performance.
Factors Affecting the Performance of Tube Settlers
Several factors noted in research studies discussed [5,6,12,21,22,
25,27,28] affect the performance of tube settlers. These are:
(a) Flow velocity and flow regime ; All researchers agree that the
flow regime should be laminar. Some [6] have set a Reynolds number
55
limit (based on hydraulic radius) of 500; others have set limits on flow
velocity (0.6 fpm by Yao; 0.33 fpm by Willis, and 1.2 fpm by Hernandez
and Wright).
(b) Tube diameter and tube length : All the researchers agree that
smaller tube diameters and longer tube lengths yield better performance.
Gulp et a l . suggested that tube diameters should be in the range of 1 to
4 inches and tube lengths should be in the range of 2 to 4 feet. Willis
stated that a tube diameter of two inches is considered economical and
reasonable. Hernandez and Wright suggested tube diameters of 1 to 4
inches and lengths of 2 to 8 feet.
(c) Tube cross-sectional shapes ; Tube cross-sectional shape
affects performance. Yao found that parallel plates perform better than
circular or square conduits. Willis found that wide horizontal plates
are hydraulically unstable.
(d) Inlet and outlet arrangement ; The design of a proper inlet and
outlet arrangement is as inçortant as the design of the tube settler
itself. To utilize the tube's capacities to remove suspended solids, one
must ensure equal distribution of the flow through the tube bundle. Both
Willis and El-Baroudi showed the importance of having a good design for
the inlet and outlet for the tubes.
(e) Degree of inclination : Only the essentially horizontal and
steeply inclined tube configurations are defined by the degree of
inclination; these are at angles of (& = 0 -7..5° and & = 45 - 60°,
respectively). Theory states that essentially horizontal tubes yield
better performance than steeply inclined tubes (Willis, [25]); however,
steeply inclined tubes are more self-cleaning. Hernandez and Wright

56
V^R
reported that for the same values of -j—, the steeply inclined tubes
yield better performance than essentially horizontal tubes. Yao [28],
however, concluded, based on an analysis of Chen's data, that the degree
of inclination has l i t t l e effect on tube performance in the range of 0°
to 75°.
(f) Entrance length: In the entrance region, a nearly inviscid
upstream flow converges and enters the tube. The initially rectangular
velocity distribution in this region i s gradually transformed into a
parabolic distribution by the action of viscous forces further
downstream. This transformation takes place in the transition relative

V d
length which Yao considers equal to 0.058 . The transition, or entrance,
length i s mentioned only by Yao, who believes that little settling occurs
in this portion of the tube. He recommends that the required tube length
calculated from his model be increased by the transition length if the
transition length is less than the actual length calculated from the
model. On the other hand, if the transmition length is greater than the
calculated tube length, the tube length provided in practice should equal
two times the calculated tube length. All researchers agree that the use
of alum and/or polymers increases removal efficiency. They found that
the amount of flocculant required depends on the type and concentration
of the suspended solids. Flocculation increases particle settling
velocity by increasing particle size and mass.
Summary
Previous work in tube settler design has characteristically shown
poor experimental design in conducting tests to evaluate design

models. In some cases ( i . e . , Yao), no experiments were in fact
performed. However, all researchers have demonstrated that long, small
diameter tubes yield better performance than short, large-diameter
tubes. While there i s substantial variance in practical use of tube
diameters and lengths, a tube diameter of two inches and a tube length
of four feet or more are commonly held desirable. The actual variations
in tube length and diameter may be accounted for by differences in
experimental method and purpose.
Tube settlers have the advantage of increasing the settlement
capacity of a sedimentation basin while making no increase in the
physical size of the basin. This was first demonstrated by Camp and
confirmed by Gulp. Since their early work, the essentially horizontal
and steeply inclined tube configurations have emerged as basic patterns
of design. While both patterns offer advantages for specific uses, the
steeply inclined form i s self-cleaning while the essentially horizontal
i s not.
The research summarized herein presents four design models having
varying degrees of efficiency.

58
UNSETTLED QUESTIONS
The literature review suggested three important questions requiring
further investigation. These are: a) What i s the actual particle
trajectory in the tube?, b) Will the design for the essentially
horizontal tubes be the same as that for the steeply inclined tubes?, and
c) What is the best inlet and outlet arrangement for the essentially
horizontal tubes configuration? This portion of the research will answer
to some extent these unsettled questions.
Actual Particle Trajectory
Studying velocity profiles in circular tubes, Langhaar [15] focused
on the change of the velocity profiles from a uniform velocity
distribution at the entrance to a fully developed laminar parabolic
velocity profiles at some distance along the tube. By using an equation
developed by Langhaar", derived from Bessel functions, Table 4, and the
hyperbolic Bessel functions, the velocity profiles for laminar flows in a
tube can be calculated. Langhaar's equation states that:
\ = [igCY) - ig/m)] / igCy) (19)
where A. = w/vav = dimensionless axial velocity
IgCy) ~ The hyperbolic Bessel function of 0-order for y value
I Q CYQ ) = The hyperbolic Bessel function of 0-order for yq value
I ^(y ) = The hyperbolic Bessel function of 2nd-order for y value
w = components of the fluid velocity
vav = average velocity in the tube
y = 3a = dimensionless parameter
59
Table 4. Langhaar constants [15]
Y 20.00 10.00 9.00 8.00 7.00 6.50
0.00082 0.00418 0.00541 0.00722 0.00997 0.01188
Y 6.00 5.50 5.00 4.50 4.00 3.50
0- 0.0143 0.0174 0.0214 0.0267 0.0335 0.0426
Y 3.25 3.00 2.75 2.50 2.25 2.00
a 0.0483 0.0549 0.0625 0.0715 0.0821 0.0947
Y 1.85 1.70 1.55 1.40 1.30 1.20
cr 0.1034 0.1132 0.1241 0.1365 0.1459 0.1560
Y 1.10 1.00 0.90 0.80 0.70 0.60
0- 0.1671 0.1795 0.1934 0.2091 0.2270 0.2479

60
q = r/a = dimensional radial co-ordinate
a = tube radius
g = certain parameter function of z alone
CT = z/aR = dimension less axial co-ordinates z, r = cylindrical

co-ordinates of the point z, x
RN = Vav a/v = Reynolds number (according to Langhaar)
V = kinematic coefficient of viscosity
The distance between two velocity profiles can be obtained using the
following equation:
L = (a^ - a RN (20)
where i = number of the velocity profile section.
Figure 34, after Langhaar, shows some of the calculated velocity
profiles. Langhaar found that the transition length required for
developing a complete laminar parabolic velocity profiles can be obtained
from the following equation:
L = 0.058 a RN . (21)
The parabolic velocity profiles will remain constant after the
transition length until the end of the tube i s reached. Experimental
work confirms the validity of these equations (White, [24]). The works
of Langhaar and White proves that the assumption of Gulp and Hansen [5]
and Willis [25] i s not valid; i . e . , flow through the tube i s not
uniform, and therefore the particle trajectory proposed by them is
incorrect.
The particle will begin to settle as soon as i t enters the tube

c.5
2 0.00082 0.CC72! \0.02 CD

a
Figure 34. Langhaar calculated velocity profiles [iS]

because the flow in the transition length is laminar. One discovers that
Yao's particle trajectory i s not applicable by following the particle
path across the tube and taking into consideration the variation in
velocity profile through the tube and i t s changes along the transition
zone.
As previously mentioned, Yao assumed a particle trajectory in the
fully laminar parabolic velocity profile part of the tube. However, a
particle may have settled part of the way through the tube diameter or i t
may be completely settled before i t reaches the end of the transition
length. The position of the particle in the tube with regard to the
transition length will depend on the average flow velocity and on the
particle settling velocity. If the average flow velocity is low and the
particle's settling velocity i s high, one can predict that the particle
will settle before the end of the transition length. On the other hand,
if the tube flow velocity i s high and the particle's settling velocity is
low the particle may settle in the region of the fully developed laminar
velocity profile, i . e . , after the transition length. Therefore, if one
adds the transition length to the design length using Yao's equation, as
recommended by Yao, one will overestimate the required tube length.
Therefore, the question of how the particle trajectory may be calculated,
and hence, how the particle settling length may be derived, must be
answered. The author has developed a model which determines both the
particle trajectory and the required tube settling length, and which
considers the velocity profile variation through the tube calculated
using the equations developed by Langhaar.

63
The Fadel model
In order to use the author's model, one must calculate the different
velocity profiles and the distance between these profiles along the tube
length by using Table 4 and Eq. 19. These calculations are shown in
sample form in Appendix B. Basically, if the local velocities through
the tube are known, the particle trajectory and the tube length may be
calculated by the following steps;
1) Assume that the particle which i t i s desired to remove
completely will enter the tube at the upper most point "A" in section I,
Figure 35.
2) Assume that the particle will reach point "B" in section II by
assuming settlement of distance in the process.
3) Assume the distance equal to 0.05D (D i s now the tube
diameter).
4) From the calculated velocity profiles find the four local
velocities VI, VI', VII, VII'.
5) By assuming that the particle will move from point "A" to "B"
with a velocity equal to the average of the local velocities, i . e . , Vav =
(VI + VI' + VII + VII')/4, the time required for the particle transfer
from "A" to "B" can be calculated :
t = L/Vav
6) By knowing the particle settling velocity, Vsc, the actual
settling depth can be calculated:
X = Vsc X t
a
7) If x^ > or < x^-assumed, then assume a new x equal to the depth
x^ calculated from step 6., and repeat steps 2 through 6 until the _
VII
Voy
vu'
Figure 35. Schematic diagram illustrates the particle path inside the
tube using Fadel's model
65
assumed.
8) When _ x^-assumed, the particle will be considered moved to
the calculated new point "B".
9) The velocity VII' will be the velocity VI for the part of the
tube between section II and III.
10) Repeat step 2 through 8 for the particle movement from point "B"
to "C," and so on, until the particle settles to the bottom of the tube
at some distance along the tube.
11) The summation of the L's until settling occurs will equal the
required tube settling length.
If the particle does not reach the tube bottom prior to reaching the
end of the transition length, the following equation should be used to
determine, mathematically, the remaining length required for the particle
to reach the tube bottom:
= d[8.0(Vav/Vsc.cose)(Yr^2.0 - Yr^/3) - YrtanS] (22)
Where Yr = ^
d
x^ = the particle's settled depth from the tube's top at
the end of the transition length.
This equation was developed by Yao as a general equation for the
trajectories of all the particles entering the tube and assumes a fully
developed flow velocity profile.
The author's model i s easy to follow and use, but the calculation Is
time consuming. Accordingly, a computer program i s used to solve the
model. Appendix A. Samples of the calculated trajectories and tube

lengths obtained using this model can be found in the chapter entitled
"Theoretical Study." Users of the computer program should enter the
desired flow velocity, particle settling velocity, water temperature,
water viscosity, and the required tube diameter to obtain the particle
trajectory and tube length from the computer program.
The model may be applied to any tube's cross sectional shape. The
only changes required are the velocity profiles for the new cross-section
and the "ff" values to be used. This can be achieved by adding new data
files to the computer, similar to the FOR010.DAT, and FOR01.DAT. The
F0R010.DAT file, shows the ratio of the local velocities for the circular
tube to the average velocity; the FOR011.DAT file shows the a values.
Essentially-Horizontal vs Steeply-Inclined Tube
All previous researchers [5,6,12,25,27,28] used the same model for
designing both essentially-horizontal and steeply-inclined tube
configurations, mentioning that the essentially horizontal tubes
theoretically yield better performance. All researchers agreed that
practically the steeply inclined tube removal efficiency was higher than
the essentially horizontal tube's, however, and stressed the importance
of the big advantage of the steeply inclined tube's self-cleaning
characteristics, noting that the horizontal tubes requires a periodic
flushing.
Of course, the two design configurations have significant
differences. The steeply inclined tubes are nearly always in a steady
state operating condition, i . e . , changes in the flow velocity and the
effective tube diameter as a function of time are negligible because the

steeply inclined tubes are self-cleaning. In self-cleaning tubes, the
particle settling on the tube bottom will immediately slide down and exit
the tubes. On the other hand, the essentially horizontal tube stores the
settled particles until the tube i s backflushed to wash the settled
particles out of the tube. These stored particles occupy part of the
tube volume, which decreases with time, reducing the available flow depth
and increasing the flow velocity. This means that the essentially
horizontal tubes are not in a steady state condition, i . e . , both the
velocity and the effective settling depth change with time.
When a designer uses the same model for designing both tube
configurations, he underestimates the tube diameter and tube length
required for the essentially horizontal tubes. The tube diameter
calculated by the designer will be for a clean tube; this diameter will
change with time. As more particles are stored at the tube bottom, the
velocity of the flow increases and particles begin to sweep out due to
increasing the shear on the stored particles surface. This i s why
horizontal tubes seem less efficient than steeply inclined tubes. If the
design procedure included the required storage volume in the essentially
horizontal tubes, the tubes would yield performance equivalent to that of
steeply inclined tubes.
Another significant difference between the two configurations i s in
their application in the field. The steeply inclined tubes are used
primarily as a polishing step in water treatment; most large, heavy
particles settle in the large inlet volume provided for this
configuration to ensure adequate flow distribution. The low settling
velocity particles will settle in the tubes. Essentially horizontal

68
tubes remove all the suspended solids in water, permitting little or
nothing to settle in the inlet zone ahead of the tube modules.
From an economic point of view, the essentially horizontal tubes
,offers an advantage in that the tank required will be much smaller than
that required for the steeply inclined tubes. The difference in the
volume i s significant, because: a) The steeply-inclined tubes require a
larger inlet and outlet volume, and b) The required cross sectional area
for steeply inclined tubes i s larger than that for the essentially
horizontal tubes for the same ratio of average flow velocity to particle
settling velocity. However, an important point which must be considered
when comparing the two configurations is the fact that both labor and
backflushing water are required to backflush or clean the essentially
horizontal tubes.
Inlet and Outlet Arrangement
As-discussed previously, the proper design of the inlet and outlet
arrangement is as important as the design of the tubes themselves.
Therefore, Willis [25] proposed an inlet and outlet arrangement for the
steeply-inclined tubes consisting of an inlet piping system similar to
the under drain of a rapid sand filter and an outlet having troughs with
an adjustable weir. Figure 15. His proposed inlet/outlet flow control
system makes good hydraulic sense and will improve the performance of
steeply-inclined tubes. The inlet piping system will reduce the large
inlet volume required by use of the inclined tubes, improving the flow
distribution through the tube bundle; the outlet system will reduce flow
short-circuiting in the tank. Figure 36 shows the upper surface of an

inclined tube bundle in a sedimentation tank; Figure 36a shows the
surface condition at the beginning of service, and Figure 36b shows the
surface condition after one month of service. A big portion of the
available surface area of the tube was plugged, indicating that the flow
was unevenly distributed.
Unfortunately, l i t t l e information is available in the literature
about the operation problems and performance of essentially horizontal
tube sedimentation tanks in practice. The only configuration found in
the literature is the system designed by Neptune-Micro floe and shown in
Figure 11. Gulp et al. mentioned that an uneven flow distribution was
obtained in self-orificlng tubes; hydraulically, an even flow
distribution should be obtained using this system. The explanation for
the observed distribution may be that large heavy particles entered the
lower portion of the tube bundle, plugging i t faster than the upper
portion where smaller particles predominated.
The author suggests the inlet-outlet arrangement shown in Figure 37.
The staggered baffles force the flow to be evenly distributed through the
bundle. The inlet zone should be as short as possible, and the fluid in
this zone should be completely mixed to keep the particles in suspension.
The staggered outlet baffle will prevent short-circuiting from taking
place in the tank. In order to prove the adequacy of this arrangement as
well as Willis arrangement for steeply inclined tubes, field application
i s required.
70
«snscww» rmR!PBTn?2y5STV5S5?BS XBZB3S$8C8F2S5!5f?51
a) At the beginning of service
«mm
b) After one month of service
Figure 36. The upper surface of a steeply inclined tube

bundle in a sedimentation tank
71
Inlet
Staggered
Saffle Outlet
Sludge
Outlet
a) Cross-sectional elevation
Sludge
Ouclec
Inlet
Sludge
Ou t l e c
b) plan
Figure 37. Suggested inlet-outlet arrangement for an essentially

horizontal tube bundle in a sedimentation tank
72
THEORETICAL STUDY
In order to evaluate the effects of significant variables on the
design of tube s e t t l e r s , six design examples are presented in this
section. The Fadel model was used and the calculations were made using
the author's computer program described in Appendix A. The computer
output predicts the particle location inside the tube a t any time, i . e . ,
the depth the particle settles versus the distance along the tube length
that i t moves.
Example 1 :
Conditions :
Tube diameter = 3 . 0 inches

Average flow velocity = 1 . 5 fpm
Particle settling velocity = 0.8 fpm

Water temp. = 90°F
Degree of inclination = 5 deg.

Transition length = 8.87 f t
S e t t l . depth ( f t ) Detention time (min) S e t t l . length ( f t )
0.021 0.026 0.031
0.081 0.101 0.156
0.100 0.126 0.202
0.129 0.161 0.268
0.170 0.213 0.371
0.200 0.251 0.443
0.250 0.314 0.500

73
Example 2:
Conditions ;
Tube diameter = 3 inches

Particle settling velocity = 0 . 4 fpm

Water temp. = 90 F.
Degree of inclination = 5 . 0 deg.

0.012 0.031 0.031
0.049 0.122 0.160
0.059 0.148 0.206
0.073 0.184 0.275
0.094 0.237 0.380
0.109 0.272 0.453
0 .126 0.316 0.545
0.148 0.371 0.664
0.175 0.441 0.817
0.221 0.554 1.019
0.250 0,627 1.069
Example 3:
Conditions ;

Average flow velocity = 1.5 fpm

Water temp. = 90 F.
74

Transition length = 8.87 f t .
0.007 0.035 0.032
0.030 0.152 0.161
0.036 0.181 0.209
0.044 0.220 0.278
0.056 0.278 0.385
0.063 0.316 0.459
0.072 0.362 0.552
0.084 0.419 0.671
0.098 0.489 0.826
0.115 0.577 1.031
0.137 0.686 1.293
0.165 0.831 1.645
0.165 0.927 1.865
0.211 1.060 2.120
0.250 1.254 2.278
Example 4 :
Conditions :

Average flow velocity = 1.5 fpm

Water temp. = 90 F.

75
0.004 0.038 0.032
0.019 0.194 0.162
0.022 0.229 0.210
0.028 0.278 0.280
0.034 0.344 0.387
0.038 0.388 0.461
0.044 0.442 0.555
0.050 0 .506 0.676
0.058 0.586 0.831
0.068 0.685 1.036
0.080 0.803 1.301
0.095 0.952 1.654
0.104 1 .040 1.876
0.114 1.140 2.132 .
0.125 1.251 2.427
0 .138 1.381 2.777
0.153 1.534 3.188
0.172 1.725 4.016
0.186 1.868 4.396
0.204 2.053 4.820
0.204 2.407 4.820
0.250 2.510 4.844

76
Example 5:
Conditions :


Water temp. = 90 F.

Settl. depth ( f t ) Detention time (min) Settl. length ( f t )
0.002 0.040 0.032
0.012 0.249 0.163
0.015 0.293 0.211
0.018 0.354 0.280
0.022 0.441 0.388
0.025 0.495 0.463
0.028 0.561 0.557
0.032 0.641 0.678
0.037 0.738 0.833
0.043 0.859 1.040
0.050 1.002 1.305
0.059 1.179 1.659
0.064 1.284 1.881
0.070 1.399 2.140
0.076 1.527 2.434
0.083 1.670 2.785
0.091 1.832 3.198

77
0.100 2.016 3.689
0.107 2.140 4.030
0.113 2.276 4.409
0.121 2.425 4.833
0.129 2.593 5.317
0.135 2.719 5.683
0.142 2.856 6.076
0.150 3.008 6.508
0.159 3.181 6.991
0.169 3.384 7.533
0.181 3.630 8.144
0.197 3.948 8.841
0.250 5.019 10.014
Example 6:
Conditions :


Water temp. = 90 F.

S e t t l . depth ( f t ) Detention time (min) Settl. length ( f t )
0.001 0.040 0.020
0.009 0.352 0.104

78
0.010 0.413 0.135
0.012 0.492 0.179
0.015 0.603 0 .250
0.017 0.674 0.296
0.019 0.760 0.356
0.022 0.864 0.434
0.025 0.988 0.533
0.029 1.141 0.665
0.033 1.323 0.3 35
0.039 1.547 1.061
0.043 1.677 1.203
0.046 1.821 1.368
0.049 1.979 1.557
0.054 2.156 1.781
0.059 2.354 2.045
0.064 2.576 2.359
0.068 2.722 2.576
0.072 2.882 2.820
0.076 3.053 3.09 2
0.081 3.241 3 .400
0.085 3.380 3.635
0.088 3.525 3.887
0.092 3.682 4.163
0.096 3.853 4.472
0.101 4.042 4.818
0.106 4.250 5.210

0.112 4.484 5.656
0.250 10.000 13.113
The examples were selected to present different particle trajectory
possibilities. The f i r s t example i l l u s t r a t e s the effect that a low
average flow velocity to particle settling velocity ratio has on Che
particle trajectory. Only 7 velocity profiles were used in calculating
the required settling length because the particle settles a short
distance from the tube entrance. Examples 2, 3 , and 4 illustrate the
effect on the required settling length of increasing the ratio of the
average flow velocity, Vav, to the particle settling velocity Vsc. As
the ratio Vav to Vsc increases, the number of velocity profiles also
increases so that, in case of example 5 , the required settling length i s
larger than the transition length of the tube. For this reason, the
l a s t integration step in the computer output was obtained by calculation
only (Eq. 22).
The transition length for the 3 inches tube with an average flow
2
velocity of 1.5 fpm and water viscosity of 0.00001 (90F) ft/sec i s 8.87
ft. The remaining length, 1.173 f t , represent the distance where the
fully developed parabolic velocity profile influences the particle
traveling in the tube. In example 6 , the degree of inclination was
changed from 5 to 60 degrees.
The particle's trajectory can be obtained by drawing the
relationship between the particle settling depth versus the distance i t
moves in the tube. Figures 38, 39, 40, and 41 show the particle
trajectories developed in examples 1 through 4.

80
Particle trajectories shown in the l a s t four figures illustrate the
effect that variation in the flow velocity profiles has on the particle
path inside the tube. At the tube entrance, where the velocity
distribution i s close to uniform across the tube, the particle trajectory
i s close to a straight l i n e . In the second part of the tube, the
particle trajectory departs from a straight line when the velocity
distribution profile i s closer to a parabola than i t i s to the uniform
velocity distribution a t the entrance.
Factors Affecting Tube Performance Using Fadel's Model
Five factors affecting tube performance may be predicted from the
computer solutions of the model. These are:
1) Tube diameter
Figure 42 presents the effect of increasing the tube diameter on
the required tube length. The conditions of flow velocity, particle
settling velocity, degree of inclination, and temperature remain
constant. As expected, large tube diameters require longer tubes; small
diameter tubes yield better performance.
2) Degree of inclination
Figure 43 shows the effect of the degree of inclination on the
required tube length. Angles in the range of 5 to 20 degrees from the
horizontal have l i t t l e influence on the required tube length. However,
the required tube length increases significantly at angles of 20 degrees
or more; a t 60 degrees, the required length i s almost twice that of tube
on an inclination of 5 degrees.
.00 .05
D 26-1
V
E .24-
.22-
P .20-
T .18-
H .16-
.14-
I .12-
N .10- 00
N)
.08-
1
.06- 7"^
F .04-11
T .02 A,
.00 A
.0 .1 .2 .3 .4 ,5 ,6 .7 .8 .9 1.0 1.1
SETTLING LENGTH IN FT.

Figure 39. Particle trajectory for Example 2
00
J OJ
.0 .2 .4 .6 .8 1.0 1.2 1.4 1.6 1.0 2.0 2.2 2.4

Figure AO. Particle trajectory for Example 3
D .26-1•w
E .24-
.22-
P .20- X
T .18-
H .16-
.14-
I .12-
N . 10-
.08-
.06-
F .04-1
T .02- \
.00-
.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Figure 41. Particle trajectory for Example 4
y
y
y
y .
12- y
Vav = 1.0 f pm y
11- Vsc = 0.1 f pm y
y
y
10-
9-
<a ^
y
8
y
7- y
y
6- y
y < —
5- y 00
Ln
y
4 y
3
2
y •
1
0
.Dd
1
.25 .50
1 1
.75
t
1.00 1.25
1 1.50
1 1.75
1 2.00
1 2.25
1 1
2.50
1
2.75 3.00
1 3.25
1 3.50
1 3.75
1 4.00
Diameter in inches
Figure 42. The effect of increasing tube diameter on the required tube
length using Fadel's model
111.11
Vau = 1.5 fpm

9. C- Vsc = O.I fpm
3 inches
8.5-
0.0-
7.5-
7.0-
G.O-
Degree of inclination
Figure 43. The effect of the degree of inclination on the required tube
87
3) Average flow velocity
Higher flow velocities require longer tubes. Figure 44 illustrates
this effect.
4) Particle settling velocity
Lower particle settling velocities also require longer tubes.
Figure 45 shows the effect of lowering the particle settling velocity on
the required tube length.
5) Temperature
Temperature affects both fluid viscosity and the particle settling
velocity. Changes in water temperature changes the distance between the
velocity profiles. Figure 46 i l l u s t r a t e s this effect on the required
length, which i s insignificant. The temperature effect on the particle
velocity i s the responsibility of the designer. In general, the lower
the water temperature, the longer the required tube length.
Differences Between the Fadel Model and Yao and Gulp Models
In order to explain the differences between the author's model and
the other two models, the difference between the Yao and Gulp et a l . ' s
models should be f i r s t discussed.
Yao's Equation (13) may be rearranged as:
Vsc (sin 9- + 1/D cos 0) = 1.33 Vav .
From this, i t follows that:
Vsc sin 0- + 1 (Vsc cos 9)/D = 1.33 Vav
and
1.33 Vav - Vsc sin 9- = 1/d Vsc cos 0- . (23)

7-
rH
EH
Diam = 3 inches
Vsc = 0.1 fpm
.0 .2 .4 .6 .0 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.0 3.0
Average flow velocity, fpra
Figure A4. The effect of increasing the flow velocity on the required
tube length using Fadel's model
4.5-
4.0-
3.5-
Vav 0.8 fpm

3 inches
60°
.c
w) 2.0-
1.5-
03
.5-
0 1 .2 .3 .4 .5 6 .7 .0 .9 1.0
Particle settling velocity, fpm
Figure 45. The effect of increasing the particle settling velocity on

the required tube length using Fadel's model
13-
12-
11-
TEH'P. 5UF
10-
9-
8-
TEMP. 0OF
7-
B-
5-
4-
Vav/Vsc » 10
3-
0
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.[
Diameter in inches
Figure 46. The effect of temperature increase on the required tube

91
When this i s done, the following equation is obtained:
(1.33 Vav - Vsc sin 0-)/l = (Vsc cos 0-)/d . (24)
When one compares this equation with that equation obtained by Gulp et
al.,
(Vav - Vsc sin 9-)/l = Vsc cos 9-/d (25)
i t becomes apparent that the only difference between these two models i s
the constant 1.33 which i s multiplied by Vav. This difference results
from Yao's assumption of a parabolic velocity profile. The constant,
1.33, can be obtained simply by integrating the area under the parabolic
velocity profile divided by the area under the uniform velocity profile,
Figure 47. At 0- equal zero, the difference between the length required
by Yao's model i s 1.33 times the length required by Gulp e t a l . model.
This i s also true a t 5 degrees of inclination. At angles greater than 5
degrees, the effect of Vsc sin 9- will be significant a t low Vav to Vsc
ratios, but insignificant at higher Vav to Vsc ratios. Figure 48
compares Yao's r a t i o of required length to Gulp et a l . calculated length
for 60 degree tubes.
The model proposed herein can be expressed in the same way. The
equation would be written as follows :
(n Vav - Vsc sin &)/l = (Vsc cos 9-)/d (26)
where "n" i s a variable whose value depends on the number of velocity
profiles used to calculate the required tube length. The value of "n"
represents the average of the areas under these velocity profiles divided
92
Vav
Vmax
Figure 47. Yao's shape factor equals the area under the parabolic
velocity profile divided by the area under the uniform
velocity profile
2.0-
0 2 4 G 8 10 12 14 16 10 20
Vav/Vsc
Figure 48. The ratio of the required length by Yao to Gulp's calculated
length for 60° tubes
94
by the area under the uniform velocity. The "n" value can be calculated
simply by assuming an angle of inclination equal to zero, which yields
the following :
n Vav/1 = Vsc/D (27)
and
L (suggested) = L' = nVav D/Vsc . (28)
Dividing L' by the length calculated by Gulp e t a l . (L = Vav D/Vsc), the
value of "n" will equal to L'/L, i . e . , n = L'/L.
Method of Using the Model
Three methods have been devised to use this model.
a) Computer program
To use the Fadel model most effectively, tube designers should have
a copy of the author's computer program. While this method would provide
the most accurate tube design results, the program i s not yet available
commercially.
b) Design charts
Design charts have been developed from successive computer solutions
which allow the designer to obtain the required length for different tube
diameters, average flow velocities, particle settling velocities, degrees
of inclination and water temperatures. Figures 49, 50, 51, 52, and 53,
are design charts for the most practical tube diameters, tube diameters
of 2, 3 , 4, 5 , and 6 inches. Tlie abscissa of the charts represents the
ratio of the average flow velocity to the design particle settling
velocity, Vav to Vsc. As long as the ratio remains constant, the same
u
bû
0 2.5 5,0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5
Vav/Vsc
Figure 49. The design chart for 2" diameter tubes

-
X x-
/
^ 0 tj
/
' 1
,tao//
-
y
/ v1
/ r ;>
X
/ -
/
y
/
/
%
/ / J
/
/
/ "
-
/
/
.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5
Vav/Vsc

0 2
VO
00
10 12 M
Vav/Vsc

18
Vav/Vsc

100
tube lengths will be obtained regardless of their numerical values. The
ratio of Vav to Vsc is limited by a maximum value of Vav assigned on the
basis of the maximum Reynolds numbers associated with laminar flow
conditions in the tubes. Based on the tube hydraulic radius, the maximum
Reynolds number i s 500, as recommended in the literature.
The designer using these charts should first establish a value for
the following parameters to be evaluated in tube design:
1) Angle of inclination,
2) Tube diameter,
3) Average flow velocity,
4) Minimum expected particle settling velocity, and
5) Water temperature range.
When these parameter values are known, the designer selects the
chart for the selected tube diameter and calculates the ratio of Vav to
Vsc, plugging these values and the degree of tube inclination into the
design chart to obtain the required tube length. Particles having a
settling velocity higher than the chosen values will settle in lengths
shorter than the required length indicated on the chart.
3) Empirical equation
The only unknown in the suggested Eq. 26 is the value of n, which
varies from 1 at very Vav/Vsc low ratios to 1.4 at high ratios. The
value of n can be determined from the following empirical equation:
n • 0.2 log ('=" : + 1.05 • (29)

° Vsc cos 6
Knowing n, the required length can be determined from Eq. 26.
The empirical equation was found by:

101
1) Assuming 0- equal to zero; in this case the required length for
the Vav/Vsc ratio range of 1 to 25 at 80°F temperature was obtained by
using the computer program for diameters of 2 to 6 inches.
2) Dividing the designing length by the calculated length from Gulp
et a l . equation; the corresponding n values are then obtained.
3) Drawing the relationship between n and the ratio Vav/Vsc; in
using the curve fitting technique for this step, the following equations
were obtained for diameters of 2 to 6 inches:
For: diameter = 2 inches
n - 0.085 In - Vsc sine ^ , Q55

Vsc cos6
diameter = 3 inches
n = 0.082 In (Vav - Vsc sinO^ + i,o89

Vsc cos6
diameter = 4 inches
n . 0.092 In + 1.039
Vsc cos6
diameter = 5 inches
n . 0.091 In (Vav - Vsc slnfl ^

Vsc cos6
diameter = 6 inches
n . 0.086 In f + 1.024
Vsc COS©
4) By averaging the constants in the five equations above, the
following equation was obtained:

102
» - 0.087 In (V»v - Vsc + j .046 .

Vsc cos6
5) By changing the In term to a log term, Eq. 29 was obtained.
The equation gives a value of "n" with an accuracy of + or - 0.04 of
the "n" value calculated by the computer program. Figures 54, 55, 56,
57, and 58 illustrate the above procedure. Figure 59 shows the effect
of temperature on the value of "n" which is shown to be in the range of
0.03 at the maximum practical Vav/Vsc ratio of 25.
As a factor of safety, the following modified equation is desirable:
n . 0.2 log f + 1.10 . (30)

Vsc cosQ
Essentially Horizontal Tubes
To simplify the study of the effect of suspended solids accumulation
on the bottom of an essentially horizontal tube, the following
assumptions are necessary:
1) Solids accumulation on the inside bottom of the tube is
uniformly distributed.
2) The tube cross-sectional area remaining after the area occupied
by the solids i s subtracted will be treated a circular tube having a new
diameter equal to its free depth.
3) The velocity of flow in the solids-free area (the free area is
the area calculated by subtracting the area of the sector occupied by
the solids from the tube total cross section area) will equal the
discharge divide by the solids-free area.
Tables 5 and 6 represent variations in the required tube length,

Diameter = 2 inches
Vav/Vsc
Figure 54. "n" versus Vav/Vsc for 2" diameter tube

Diameter = 3 inches
1.100'-
.0 2.5 5.0 7.5 .0 12.5 15.0 17.5 20.0 22.5
Vav/Vsc
Figure 55. versus Vav/Vsc for 3" diameter tube

1.34-:-
1.32-
1.30-
1.28-
1.26-
1.24-
n 1.22-
1.20-
1.18-
1.16-
Diameter = A inches
1.14-
1.12-
1.10--
j I 1 I I
.0 10.0 12.5 15.0 17.5 20.0
versus Vav/Vsc for 4" diameter tube

.28-
1.22-
o
o\
Diameter = 5 inches
Vav/Vsc

1.26-i
n 1.20-
1.18-
1.16-' Di ameter 6 inches
1.14-
r
.0 2.5 5.0 7.1" 10.0 12.5 15.0 17.5 20.0
Vav/Vsc

1.375
1.350- TEMP. 50F
1.325-
1.300-
1.275- TEMP. 80F
1.250-
n 1 .225-
1.200-
1.175-
1. 150- Diameter 4 inches
1.125-
1.100-
1.075
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Vav/Vsc
Figure 59. Temperature effect on the "n" value

109
Table 5. The effect of sludge accumulation on the required tube length

for a 3" tube, settling velocity = 0.1 fpm, and degree of
inclination = 5.0°
Solids—free depth Flow velocity Settl. length

(in) (fpm) (ft)
3.00 1.00 3.223
2.25 1.24 3.055
2.00 1.36 2.984
1.50 2.00 3.334
1.00 3.40 3.831
0.75 5.00 4.240
Table 6. The effect of sludge accumulation on the required tube length

for the 4" tube settling velocity = 0.1 fpm, and degree of
inclination = 5.0°
Solids-free depth Flow velocity Settl. length

(in) (fpm) (ft)
4.00 1.00 4.165
3.50 1.08 3.990
3.00 1.24 4.007
2.50 1.52 4.116
2.00 2.00 4.395
1.50 2 .92 4.875
1.00 5.11 5.7 59

110
based on the above assumptions for two tubes with diameters of 3 and 4
inches, respectively.
Figures 60 and 61 illustrate the sludge accumulation effects
presented in the above two tables, in which the required tube length
decreases with solids accumulation until a certain depth is reached at
which the length then increases with solids accumulation. The
explanation for this behavior can be related to the geometrical
distribution of the circular sections. When the tube fills with solids,
residual tube depth diminishes more rapidly than the tube solids-free
area, meaning that the depth through which the particle settles
decreases faster than the flow velocity increases. This condition
favors decreasing the required settling length until the point is
reached when small changes in diameter lead to large decreases in the
solids-free area and significant increases in flow velocity. This
condition i s approached when the solids deposited approach close to the
center of the circular tube. This leads to an increase in the required
settling length.
If experimental work is consistent with this situation, the designer
may use any of the above methods to design an essentially horizontal
tube, with the provision that periodic backflushing must be performed
when the tube is approximately 40% full. If the above situation i s
inconsistent with the experimental work, the recommended design should be
established according to the results of the experimental work.

2.25
2.00
1.75-
1.50
A 1.00 Di ameter 3 inches

T3
Vav/Vsc 10
3.2 3.3 3.5 3.9
Settling length, f t .
Figure 60. Sludge accumulation effects on the required length for an

essentially horizontal tube settler with a 3-inch diameter
4.5
4.0-
3.0
E. 2.0
Diameter = 4 Inches
Vav/Vsc
CO
3.0 4.0 4.2 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0
Settling length, f t .
Figure 61. Sludge accumulation effects on the required length for

essentially horizontal tube settler with a 4-inch diameter
113
EXPERIMENTAL STUDY
Previous research left two questions unsettled at the design and
practical levels: "What is the actual particle trajectory in the tube?"
and, "Will the design for the essentially horizontal tube be the same as
that for the steeply inclined tubes?" Theoretical answers had been
provided earlier by the author. The purpose of the experimental study
was to investigate the validity of the theoretical answers. The
experimental study was divided into two parts: first, a study was
conducted to demonstrate the trajectory of the particles as predicted
using the Fadel model; second, a study was conducted to evaluate the
affect of sludge accumulation on the bottom of the tube on the
essentially horizontal tubes performance. The.experimental study was
conducted in the sanitary engineering laboratories at Iowa State
University.
Equipment and Materials
Equipment
As shown in Figure 62, the experimental equipment consisted of four
main parts: the tube's water supply and inlet tank, the tube settler
itself, the tube's outlet, and the equipment for feeding particulates at
the top of the inlet to the tube.
The water supply used in the study was obtained from the building
water supply and delivered to a constant head tank made from plexiglass
and mounted on a steel platform. The tank overflow was wasted. The
constant head tank was connected to a filter nozzle by a 3/4-inch garden

Constant Head Tank
Mixer
18.5"
Outlet Tank
M a s t e r f l e x Pump IB.5
Inlet Tank 9.4
0=5
Figure 62. Schematic diagram for the experimental equipment

115
hose. The filter nozzle was made of plastic and had 0.04 inch openings;
i t was supplied by General Filter Company, Ames, Iowa. The nozzle was
placed in a small box (5x5x5 inches) made of aluminum screen having
openings of 0.4 x 0.4 inch. The box was filled with plastic beads
(Cullsan P) provided by Culligan U.S.A., Northbrook, Illinois. The
beads were cylindrical in shape and had a diameter of 0.06 inch and a
length of 0.08-0.12 inch. The box was wrapped in cloth to keep the
beads in place (box openings were larger than the bead dimensions) and
to work as a filter for the tap water. The box was placed in a
plexiglass tank with dimensions of 18.5 inches x 18.5 inches x 18.5
inches. The modification shown in Figure 63 was made after Run #15.
The modification included: a) redesigning the inlet box to have
dimension of 4x4x4 inches (box opening was 0.06 inch in diameter and
made of stainless steel; the same beads were used), b) Hooking the
inlet box on the wall of the plexiglass inlet tank facing the tube
opening, as shown in the figure, c) Connecting the inlet hose from the
constant head tank to the inlet box outside the plexiglass inlet tank.
The tube settler was mounted through a circular opening of 2.75
inches located 4 inches from the inlet tank bottom. The opening was
provided with a 1/4—inch rubber "0" ring to seal off water around the
tube settler.
Three plexiglass tubes were used at various times as tube settlers.
Figure 64. Tube dimensions were: inner diameter of 2 inches, outer
diameter of 2.5 inches, and length of 71 inches. Each tube was provided
with a 1/2-inch circular opening at one end for the suspension inlet.
Constant llend Tai\k
McTgne t 1 c
Mixer
M a s t e r f l e x Pump
Outlet Tank
Inlet Tank
Figure 63. Schematic program for the modified experimental equipment

F i r s t Tube
Tkr-^jiUP^nLlJ—IMJ—It!]—ikj—iLlr=iUr
.34
Second Tube
I-'
M
Ï U | — — i U r = n L r T ' —~
11 —V- 4 —\ J '—
t —o
Third Tube
.26-
-1.99-
7.24
-1.96"-
1=7.24"
-1.80'i
The Three Plate's Dimensions
Enlarged cross-section of the third tube
Figure 64. Schematic diagram for the three tubes used in the experiment
118
One of the tubes was provided with eight copper cups (3/4-inch) located
at the tube bottom (considering the suspension inlet opening to be
located at the tube top). The cups were spaced at 4-inch intervals with
the first centered 34 inches from the tube inlet. The third tube was
designed differently than the other two. A 0.32-inch section along its
bottom was removed for the full tube length and replaced by a flat
plate, except for 6 inches at each end of the tube. A flat plate was
placed inside the tube to provide the same reduction of the total depth
as that for the removed part. The third tube then can be considered as
an open circular tube cross-section having a simulated particle depth
equal to 0.32 inch. This design represents an essentially horizontal
tube with 0.32 inch of i t s depth occupied by accumulated sludge. Eleven
cups, described above, were provided along the flat bottom of this tube
starting 24 inches from the tube inlet and spaced at 4-inch intervals.
Three plates were used to represent added depth of sludge
accumulation on the bottom of the third tube. These plates have the
same length as that of the tube; they were provided with 3/4-inch
circular halls in the same locations as those in the third tube,
permitting the settling experimental particles to be collected in the
cups. The dimensions of the plates were as follows: The first plate
has a width of 1.8 inches and a thickness of 0.24 inch; the second plate
has a width of 1.96 inches and a thickness of 0.24 inch; and the third
plate has a width of 1.99 inches and a thickness of 0.26 inch.
The tubes 5 when installed to the inlet and the outlet tanks, had 4
inches of their length at each end inside the tanks.

119
The outlet tank had the same dimensions as the inlet tank, with
only one difference: the outlet tank was provided with an outlet
overflow tube 1/2-inch diameter from which the water flowed to waste.
The outlet tube was located on the opposite side the outlet tank, i . e . ,
facing the 2.75 inch opening, and 9.5 inches from the tank bottom.
The suspension feeding system consisted of an 8-liter plastic
bucket, a 1/30-horsepower mixer provided by Talpays Engineering
Corporation, Emerson, New Jersey; an adjustable speed control for the
mixer; and a Masterflex pump. Model No. 7565, provided by Cole-Parmer
Instrument Co., Chicago, Illinois. The pump had several heads which
could be mounted separately or in groups of 1, 2, 3, or 4 units. The
head used in this study was Model No. 7014, which pumps water at
adjusted rates up to 100 ml/min. The flow rate required can be adjusted
by an adjustable control. The inlet to the pump was a 0.04 inch Tygon
tube having i t s inlet side submerged in the suspension bucket. The
pump's outlet was the same as its inlet except that a 2-inch long steel
tube having the same diameter was attached to the outlet end. The steel
tube was bent and placed in the top opening of the tube settler. The
bent end was directed to the flow direction. A later modification
placed the suspension bucket just above the Masterflex pump and used a
magnetic stirrer instead of the mixer. Accordingly, a shorter inlet
tube to the Masterflex pump was used. This tube was kept as nearly
vertical as possible; the reasons for these modifications will be
explained when the results are discussed.

120
Miscellaneous equipment
A microscope manufactured by the American Optical Scientific
Instrument Division, Buffalo, New York, was used to magnify the settled
glass beads in order to measure their diameters. A Polaroid camera was
used to photograph the glass beads. A nest of sieves manufactured by
the Allan Bradley Company, Milwaukee, Wisconsin, was used to sieve glass
beads to obtain the range of desired diameters. The sieve openings used
were 30, 45, 53, and 74 microns, with factory specification stating that
the actual openings are within + 2 microns of the stated openings.
Materials
Two materials were used to provide particulates as suspension in
the study: diatomite and glass beads. The dlatomlte was provided by
Johns Manville Product Corporation, Lompac, California. The diatomite
grade used, Cell te 535, has a median particle size of 25 microns and a
size range of 5 to 175 microns. The glass beads were provided by the
Fine Industrial Supply Corporation, Baldwin, New York. Eighty percent
of the glass beads purchased were in the size range of 44 to 95 microns.
They were spherical in shape and contained no more than 10% irregularly
shaped particles. They were reasonably free of sharp angular particles,
as well as particles showing mllkiness, surface scoring, or foreign
matter.
Suspension preparation
Two methods of preparing particles having a narrow size range were
practiced. These are water elutriation and sieving analysis. The water
121
elutriation process was used for preparing the diatomite samples; the
sieving process was used for preparing the glass bead samples.
The elutriation process was used in an attempt to prepare a
suspension with a single settling velocity. The apparatus consisted of
three glass tubes, each with a different diameter arranged vertically,
as in Figure 65. The three tubes had a different lengths and diameters
as shown in the figure. The lower tube had a diameter of 1.5 inches,
the middle one had a diameter of 3 inches, and the top tube had a
diameter of 6 inches. The inlet to the elutriation tubes was provided
through a flow ratemeter and a needle valve, which controlled the flow
rate through the tubes. Water flowed from the flow ratemeter into the
bottom of the lower tube and up through the second and the third tubes,
where i t was discharged from the top of the third tube.
The elutriation process may be summarized as follows: If a sample
of diatomite is placed in the lower tube, the particles with settling
velocities equal to or greater than the water flow velocity in the
bottom tube will remain in i t , and the particles with settling velocity
less than the flow velocity will escape to the second tube. Since the
second tube has a larger tube diameter than the first, the fluid
velocity becomes proportionally smaller. Thus, particles with settling
velocities matching the flow velocity will stay in this tube, while the
particles with lower settling velocities will escape to the third tube,
and so on. Because diatomite particles have an irregular shape, the
suspension collected in each tube will have a very wide size range.
The elutriation process time was continued until the particles

122
Outlet
5- ID
RUBBER •1/4" ID
STOPPER
11 FLOW 1-1/2
XCONTROL
VALVE
Flow Rate Meter
*1-1/2" ID
Inlet
1/4" ID
Figure 65. Schematic diagram presents the elutriation process

equipment used in the experiment
123
become stationary and discrete, i . e . , the time when the particles were
no longer moving up or down. The time required to terminate the
elutriation process varies with the flow velocity, longer times being
required at low flow rates than high flow rates.
At the end of the elutriation process, the stopcocks were closed in
the middle, then the top tube, and finally the lower tube. The middle
tube was then separated from the system. A volume of water equal to the
cone volume of the middle tube was then drained into a graduated
cylinder to obtain the exact volume. The remaining water then was
collected in a beaker. This water contained the particles used in the
tube settler experiment.
The operation was repeated several times, until a considerable
weight of particles was collected.
Disadvantages of the elutriation process
Several disadvantages to the elutriation process were discovered:
1. The three columns must be nearly vertical; since tube
inclination disturbed the flow pattern and flow rotation
occurred.
2. The stopcocks must be essentially vertical when opened; any
minor tilting disturbed the flow pattern.
3. The cone in the column contains particles with a wide range of
flow velocity, starting from the average flow velocity from
the previous tube to the average flow velocity of the tube
that includes the cone. However, this problem was reduced by
wasting a water volume (and the particles contained therein).

4. The most critical disadvantage of this method was that the
flow regime across the tubes is laminar; therefore, velocity
profiles within the tube will change from uniform to
parabolic. These velocity profiles will collect suspension
having particles with wide range of settling velocities. This
problem becomes more critical when the transition length
(between uniform velocity profile and laminar flow velocity)
i s shorter than the tube length than that when the transition
length is longer than the tube length. The velocity at which
the transition length equals the tube length was found to be
0.13 ft/min.
Experimental Procedure
Two experimental procedures were used in this study. The first was
designed to demonstrate the validity of the Fadel model as an accurate
predicter of where particles would settle out. The second procedure was
designed to examine the suggested behavior of the essentially horizontal
tubes taking into consideration the effect of the settled solids
accumulated on the tube bottom.
The following steps were assigned in the first procedure depending
on the material of the suspension:
Using diatomite:
1. The diatomite particulate was prepared by the elutriation
process.
2. After collecting a reasonable amount of the suspension, the
suspension was placed in the plastic bucket and continuously

125
mixed.
3. The suspension was then fed to the inlet end of the tube
settler at the top, using the Masterflex pump at a rate equal
to the average flow velocity across the tube settler.
4. The uniform flow rate across the entrance to the tube settler
was measured by recording the average time required to f i l l a
2-liter beaker. At least five replicates were timed and
averaged.
5. When the average flow velocity across the tube settler and the
flow velocity of the second elutriator tube are both known,
the following quantities may be calculated :
a) Transition length in the elutriation tube. If this length is
less than the 9-inch elutriation tube length, the average settling
velocity for the particles will equal 1.33 times the flow velocity, and
the maximum settling velocity will equal twice the flow velocity. The
minimum settling velocity of the particles will be less than or equal to
the flow velocity. If the transition length i s larger than the column
length, the "n" value will be less than 1.33. The "n" value then can be
calculated and the average particle settling velocity will equal "n"
times the flow velocity. The maximum settling velocity can be
calculated using the Langhaar equation.
b) In each of the above cases, the particles settling velocity
range can be determined and, the required settling length can be
calculated using the Fadel computer program, and a settling range
determined.
126
6. When the settling range is known, i t can be compared with the
actual location of the particles in the tubes at the end of
the run. The validity of the model may then be determined by
comparing the theoretical with the actual location of the
particles.
Using glass beads ;
The procedure described below was used in the remaining runs except
Runs #5 and #6. These two runs were conducted using the same procedure
used for the diatomite.
1. The suspension was prepared by mechanically sieving the glass
beads. The glass bead size range used in the experimental
work was 20 to 74 microns.
2. A small amount of monobasic sodium phosphate, which acted as a
dispersant, was added to the glass beads suspension.
3. As with the diatomite, the suspension was fed into the tube
using the Masterflex pump. The second tube having eight cups
was used in these runs.
4. The flow rate was measured using a one-liter graduated •
cylinder and a stopwatch.
5. The run was terminated after a reasonable amount of glass
beads lay on the bottom of the tube. Usually, a run lasted 16
to 24 hours at low flow velocities, and 6 to 9 hours in high
flow velocities runs.
6. At the end of each run, the tube was tilted 20 to 40 degrees
around the axis of the flow direction, and the two tanks were
drained by simultaneous syphoning with 1/2 inch tygone tubes
(one for each tank). The tube is tilted in order to move the
glass beads accumulated in the tube to a new position far from
the cups. Thus, when the water backs up while the tanks are
drained, the glass beads will not be carried to the cups.
After draining the two tanks, the cups were removed and a
sample of the glass beads settled in four cups chosen at
different locations was examined under the microscope.
The lOOOx magnification lens was used to measure the diameter
of unbroken spherical particles. Two or three pictures were
taken at two to three locations in the same sample. Again,
the locations of the photographs made were chosen to present
the maximum possible number of unbroken spherical glass beads.
If variation occurred in the measured diameters, the average
diameters of those photographed was taken. The variation in
the measured diameters was in the range of 1 to 3 microns,
when particles having diameters of 22 to 46 microns were
measured.
When both the diameter and the specific gravity of the glass
beads are known, the particle settling velocities may be
Stoke's settling velocities of the particles and the average
flow velocity across the tube are known, the theoretical
length at which the particle is predicted to settle may be
found using the computer program.
The validity of Fadel's model then may be determined by

comparing the theoretical length with the actual length at
which the glass bead particles were found.
The following operating steps were used in the second procedure in
which the glass beads were the only suspension used:
1. For a specified flow, the previously noted steps one through
nine were repeated to evaluate where the glass beads would
settle in the tube with no sludge accumulation in the tube.
2. Using the same flow and the third tube, a complete run was
then conducted in which the glass beads settled in tubes with
a simulated sludge accumulation. Pictures were taken of the
glass beads that settled in the same places as those in the
first step.
3. Then, the first plate was placed on the flat bottom of the
third tube. This added plate represents more sludge
accumulation on the bottom of the tube. Again using the same
flow, a complete run was conducted.
4. Step two was then repeated after adding the second and then
the third plates. Again, a set of pictures was taken after
each plate was added to represent additional sludge
accumulation. The glass beads were collected in the same
places as those collected in the previous runs.
5. A relationship between the ratio of storage depth to the tube
diameter and the particles diameters at the same spots along
the tube length was drawn.

129
Steps one through five were then repeated for different flow
values.
130
RESULTS AND DISCUSSION
Twenty-five runs were conducted in the experimental study.
Thirteen runs were conducted to demonstrate the validity of the Fadel
model. Of these, four runs were conducted (Runs #1 to #4) using
diatomite and nine runs (Runs #5 to #11 and Runs #16 and #21) were
conducted using glass beads. In the second part of the study, twelve
runs were conducted to determine the effect of sludge accumulation on
the performance of essentially horizontal tubes.
The results of the diatomite runs and the first two glass bead runs
were expressed in terms of the range of the distance of the settled
diatomite or glass beads from the tube inlet. The later glass bead runs
were used to determine the mean diameter of the settled glass beads, and
their locations along the tube. The following discussion concerns
results from the runs conducted to demonstrate the validity of the Fadel
model.
Part 1
Run £l_^ (2/18/85)
This run was designated as the first run; it was conducted after 3
to 4 weeks of experience in running the system to determine final
equipment and test precedure operating conditions. For example, the
inlet box was first filled with gravel having a range of 1/2 to 3/4 inch
diameter. When dye was used to study the flow regime, it was discovered
that the flow distribution was not as uniform as it should be. It was
then decided to change this media to the Cullsan P. In another early

131
trial, the suspension was fed by gravity through a very narrow tube
placed in the 1/2-inch hole provided at the tube settler top. This
method was troublesome. Every 1/2 hour, the feeding tiibe clogged with
diatomite and cleaning was required. It then was decided to use the
Masterflex pump, which eliminated the clogging problem.
In Run #1, the average flow velocity, Vav, was 3.54 fpm. The
diatomite was elutriated with an average upward velocity of 0.319 fpm in
the second tube, and 1.58 fpm in the first tube.
Using the computer program, the theoretical location for the
settled particles indicated that, for the 1.48 fpm elutriation tube
settling velocity and the 3.54 fpm tube settler Vav, the settling length
should be 0.43 ft, and for the 0.319 fpm elutriation tube velocity, the
settling length should be 2.214 ft. The observed range was 0.43 to
3.0 ft, with most of the suspension in the range of 0.9 ft to 2 ft.
These results were quite encouraging, but confusing. However, the fact
that we are dealing with a circular tube (the second tube) in the
elutriation process reduces this confusion. The second tube will have a
velocity profile and a transition length similar to that found in the
tube settler. Accordingly, there will be a range of settling velocities
rather than a single value of settling velocity particles. Knowing that
the maximum average velocity for a transition length of 9 inches is
0.13 fpm, one realizes that lower velocities yield wider settling length
ranges and higher velocities yield narrower ranges.
A complete laminar flow velocity profile will be established inside
the elutriation tube at velocities less than 0.13 fpm. The maximum
132
upflow elutriation tube velocity will equal twice the average velocity;
i.e., the particles will have settling velocities ranging from the
average velocity to twice that value. On the other hand, for higher
velocities, the maximum velocity will be less than twice the average
velocity. Accordingly, the particles elutriated will have settling
velocities ranging from average velocity in the elutriation tube to less
than twice that value.
In this run, the average settling velocity used in the elutriation
process was higher than 0.13 fpm. Accordingly, the maximum velocity
will be less than twice the 0.319 fpm used. However, some of the
particles used in this run had settling velocities less than the average
flow velocity in the second tube. Accordingly, the range over which
the particles settled in the tube settler can be considered applicable .
and promising but not confirming.
(2/20/85)
A lower average tube settler flow velocity of 2.94 fpm was used in
this run, and particles with a lower settling velocity were prepared
using the elutriation process. The average velocity in the second tube
was 0.244 fpm. The cone in the second tube was drained at the end of
each elutriation process to remove particles having velocities higher
than those found in the second tube and lower than those in the first
tube (volume was 235 ml).
The particles settled were located at a distance of 1.4 to 2.77 ft
from the tube settler inlet. According to the computer solution,
particles with a settling velocity of 0.244 fpm settled in a

133
2-inch tube settler with an average settler tube velocity of 2.94 should
settle at a distance of 2.5 ft from the tube settler inlet. Again, as
in the first run, we are dealing with a range of .settling velocities and
not a single value, so the observed range of tube length where settling
occurred is not surprising. Using the Langhaar Equation (19), the
maximum velocity inside the second column can be calculated. The
maximum velocity was found to equal 1.95 times the average velocity,
i.e., 0.476 fpm. This settling velocity will give a settling length of
1.22 ft. Again, the range is considered reasonable but the Fadel model
is unconfirmed.
Run£^ (2/22/85)
In this run, the flow average velocity through the tube settler was
2 fpm and the average upward velocity used in the elutriation process
was 0.14 fpm. The maximum settling velocity was twice 0.14 fpm; i.e.,
0.28 fpm. The settling range where the particles were found was 1.5 to
3.5 ft. from the tube inlet. The computer solution showed that the
expected range should be 3.139 ft for the 0.14 fpm (Vsc) particles to
1 .487 ft for the 1.28 fpm (Vsc) particles.
Run (2/25/85)
A lower particle settling velocity was used in this run, but the
tube settler velocity was kept the same. The average particle settling
velocity was 0.1 fpm ( <.13 fpm) and the maximum was 0.2 fpm.
The experimental results showed that the particles settled over a
range of 2.00 to 4.2 ft from the inlet; the computer-projected range was
134
2.16 to 4.44 ft. Again, there is a good agreement between the observed
and the computed location of the settled particles.
^ (2/27/85)
In this run, glass beads in the size range of 45 to 53 microns were
used instead of diatomite. The tube settler flow velocity used, Vav,
was 4.20 fpm. Using Stoke's Law, the glass beads with a diameter of 45
microns will have a settling velocity of
Vsc = 980 cm/sec^(2.5 - I)(0.0045 cm)^

18 X 0.012 cm^/sec
= 0 .138 cm/sec
= 0.272 fpm (45 micron glass bead).
For the 53-micron beads, Vsc is equal to 0.377 fpm. The range over
which the particles settled in the tube settler found in the tube bottom
was 2.0 to 3.4 ft, while the computer range was of 2.28 to 3.23 ft. It
was noticed that the observed range was a little wider than computed
because the actual sieve openings were +2 micron. Accordingly, the
actual range to consider is 43 to 55 microns. This glass head size
range gave a settling velocity range of 0.248 to 0.406 fpm with a
computer-predicted settling range of 2.1 to 3.56 ft.
The experimental results are in good agreement with the predicted
results. However, a thin line of settled glass beads lay on the tube
bottom from the observed range to both ends of the tube. The explana
tion for the presence of this line is that particles smaller than the
range specified were sieved with the sample. They could be stuck with
135
other sieved particles. These particles had lower Vsc than
the settling velocity of the 43-micron particles. Also, irregular
particle shapes such as half spheres, rods, etc., could be sieved
within the range of 43 to 55 microns. These particles would have
settling velocities higher or lower than the spheres. For example,
cylinder-like particles have a lower drag coefficient than spherical
particles. Figure 1. Accordingly, the cylindrical particles will have a
higher settling velocity than the spherical particles, and will settle
earlier in the tube.
Run (3/5/85-3/6/85)
The tube settler flow velocity was set to equal 5.08 fpm. This flow
velocity gave a Reynolds number of 1055 based on thé tube diameter. The
range of glass bead sizes was from 51 to 76 microns (actual range). The
glass bead settling velocities were then from 0.35 to 0.76 fpm. At the
end of the run, most of the particles were laying from of 1.1 to 2.80 ft
from the tube inlet. The computer results indicate that the beads were
predicted to lay from is 1.25 to 2.98 ft from the tube inlet.
General Comments, Runs #1 - //6
The results obtained from Runs #1 to #6 indicate that a reasonable
agreement exists between the observed and the calculated location of the
particles in the tube settler. Although these runs were nearly
successful, neither elutriation nor the sieving processes were as
helpful as expected. The results are not a confirmation of the Fadel
model. A confirming result would be one demonstrating that the settled

136
particle can be picked up and its actual settling velocity calculated.
Using this particle settling velocity, the computer-projected solution
should provide a calculated settling length equal to the actual settling
length observed. A modified procedure based on this premise may be
accomplished by using the second tube, where the eight cups work as
traps for the settled particles. Accordingly, at this stage of the
study, it was decided to conduct a few runs using the second tube in
order to measure the diameters of particles collected in the eight cups.
Run (3/17/85)
The second eight-cup tube was used in this run. The average tube
settler flow velocity was 3.5 fpm. Glass beads in the size range 35 to
47 microns were fed through a tube having an inner diameter of 1/16-inch
and outer diameter of 1/8-inch. The suspension feeder was connected to
the top of the tube settler wall.
After 12 hours the run was terminated and microphotographs were
taken of the glass beads trapped in the first, fourth, and seventh cups.
Table 7 shows the cup number, the distance between the inlet and the
specified cup (equal to the settling length), the glass bead particle
diameter found in the specified cup, the settling velocity, Vsc, for the
particles, and the computed predicted distance along the tube settler
where the particle should settle.
The computer-predicted length = (the length computed according to
the run conditions of Vav, Vsc, and diameter) - (the effect of letting
the particles enter the tube 1/8 of an inch lower than the tube top
(i.e., the computer output gives the settling length for particles
137
Table 7. Results and computer-predicted results in Run #7
Cup# Distance^ Particle Diameter Vsc Computer-predicted

ft microns fpm location, ft
1 2.85 44.00 0.250 2.83
4 3,.85 38.75 0.192 3.84
7 4.85 33.80 0.152 4.94
^Distance from the inlet.
entering the tube from the uppermost point until settlement). The
correction can be found by using the first option in the computer
program (Appendix A), which allows us to trace the particle trajectory
across the tube. This option was used in presenting Examples 1 through
6 (under the theoretical study). Knowing the outer diameter of the
suspension feeder tube, which equals 0.011 ft, the distance the particle
will travel from the tube inlet until it reaches this depth can be found
using the above option. The distance was found to equal 0.07 ft.
Comparisons between settling lengths predicted by the different
models discussed in the literature review and the Fadel model are
presented in Table 8.
Another observation should be mentioned here. Even though the
sieving range was from 35 to 47 microns, particles with diameters of
less than 35 microns were found in cup #7.

138
Table 8. A comparison between the different models using Run #7 data
Yao
Cup # Willis Gulp Fadel actual
model recommended length, ft
design length
1 3.11 6.22 2.34 2.33 2.83 2.85
4 4.04 8.09 3.04 3.03 3.84 3.85
7 5.12 10.73 • 3.85 3.84 4.94 4.85
Run (3/19/85)
After the very successful Run #7, it was decided to reduce the
average flow velocity in the tube settler to 1.85 fpm in Run #8. During
Run #8, the author observed that a large number of the particles that
entered the tube settler were large-sized and settled rapidly. Before
they reached the middle of the tube, however, the agglomerated particles
began to disperse into a large number of much smaller particles. These
smaller particles settled much more slowly when dispersed than when they
were flocculated together. In short, small particles which flocculate
prior to entering the tube settler form larger particles which settle
faster until they hit a region of high local velocity close to the tube
center where they start to separate into the original small particles.
Such a situation leads to faulty results. Another problem was also
observed associated with the suspension feeding tube. The flow
velocity (at the minimum flow of the Masterflex pump) exiting this tube
was two times the average flow velocity of the tube settler. Both of
these conditions will lead to shorter observed settling lengths than
those predicted based on the settling of discrete particles. Before the
termination of the run, a sample of the glass bead laid at the end of
the tube was pulled out by a means of a pipette. The diameter of these
glass beads was found to equal 20 microns. This diameter of particle
has a settling velocity of 0.051 fpm. For this settling velocity and at
the average flow velocity in the tube settler, the required settling
length projected by computer is 8.08 ft. However, these particles were
located only 5.9 ft from the tube inlet. The effect of these two
conditions led to an error of 2.16 ft.
(4/20/85)
The results of Run #8 caused the author to increase the mean tube
settler velocity to overcome these effects. The tube flow velocity was
increased to 2.9 fpm. Better results were obtained than in Run #8, but
they were not as good as those obtained in Run #7. Table 9 presents the
results and the computer-predicted location of the particles collected
in each cup.
The tube carrying the particles from the suspension bucket to the
Masterflex pump was almost 3/4 full of particles; these rolled in very
slow motion. Also, at the outlet feeding tube, the particles exited in
large numbers at one time, resembling hindered rather than discrete
settling. Again, the first effect mentioned in Run #8 still existed,
but the particles separated faster than those in Run #8.

140
Cup # Distance,^ Particle diameter, Vsc, Computer-Predicted,

ft microns fpm location, ft
1 2.85 38.00 0.194 3.03
5 4.19 32.00 0.138 4.38
7 4.85 29.00 0.113 5 .35
^Distance from inlet.
Run #10. (4/5/85)
In this run, suspension feeding tube was replaced by a larger
diameter tube in order to match the flow of suspension exit from the
feeder with that of the tube settler. The inner diameter of the new
feeder tube was 1/8 inch, the outer was 3/16 inch. A dispersant was
added to the suspension (monobasic sodium phosphate) to keep the
particles dispersed. The mean flow rate in the tube settler in this run
was 2.25 fpm. Table 10 presents the results.
While the results were not as expected, they were better than those
obtained in Run . The researcher noticed, after the run started, that
the delivery pipe was carrying too many particles—even with presence of
the dispersant material—and the particles exiting the tube were
clustered together. This may be explained since (1) the particles
clustered together. This may be explained since (1) the particles came
very close to each other in the feeding system such that the dispersion
effect was not apparent, or (2) the amount of dispersant was

141
Table 10. Results and computer-predicted results in Run //lO
Cup # Distance,^ Particles diameter, Vsc, Computer-predicted

ft microns fpm Location , ft
1 2.85 32.00 0.138 3.38
3 3.52 29.00 0.113 4.18
6 4.52 25.10 0.085 5.70
insufficient. Also, an air bubble was found in front of the feeding
tube outlet which drove the particles downward when they left the
feeder. This air bubble was formed by the action of the mixer inside
the suspension tank when the tank became half full.
Run #11. (4/6/85)
Another modification, consisting of cutting the suction tube length
to a minimum by placing the suspension tank on the top of the pump was
made. Accordingly, a magnetic stirrer replaced the mixer and a two-
liter plexiglass tank having the suspension outlet at its bottom
replaced the plastic bracket. The suspension inlet inside the tank was
kept vertical to prevent air bubbles from entering the suction tube.
The suction tube itself was also kept vertical as shown in Figure 64, to
reduce the particle's retention time in the suction tube. Whenever the
particle enters the suction tube, it will fall by gravity as well as by
the suction action from the pump to the pump head.
In Run #11, the average tube settler flow velocity was 3.00 fpm.
142
The results showed a significant improvement. Table 11 shows these
results.
Cup # Distance^ Particles diameter, Vsc, Computed-Predicted

ft microns fpm Location, ft
1 2.85 37.60 0.190 3.15
3 3.52 34.50 0.160 3.82
5 4.19 32.70 0.144 4.45
7 4.85 30.00 0.121 5.22
^Distance from inlet.
The observed error in this experiment may be due to the high
suspension concentration used. Since the number of particles leaving
the feeding tube was still high, it was decided to use a lower
concentration suspension (3 grams of glass beads in the two-liter tank).
Runs #12 to #15 were used to study the effects of sludge
accumulation on the performance of the essentially horizontal tubes
where a flow of 0.065 cubic ft/min, giving a 3 fpm Vav for full diameter
tube, was monitored as a continuation for Run #11.
Run #16. (4/11/85 to 4/13/85)
In this run, a very low suspension concentration and a tube settler
flow velocity of 2fpm were used. The run lasted 48 hours, until a
considerable amount of glass beads lay in the tube bottom. The results
were very encouraging, as shown in Table 12. Figure 66 shows a

143
Figure 66. Glass beads (bulk) used in the study

144
microphotograph of the bulk glasls beads used in the study. Figures 67
to 70 are microphotographs of the glass beads settled in cups 1, 3, 5,
and 7 in this run.
Table 12. Results and computed-predicted results in Run #16
Cup # Distance Particles diameter, Vsc, Computed-Predicted

ft microns, fpm Location, ft
1 2.85 32.20 0.140 3.06
3 3.52 30.00 0.121 3.55
5 4.19 28.20 0.104 4.18
7 4.85 26.00 0.090 4.87
Run #21. (4/25/85)
In this run, the tube settler flow velocity was increased to 4.0
fpm. The suspension concentration was higher than that used in Run #15
because of the higher tube velocity. The run was successful and the
observed lengths almost matched the computer-predicted location of the
particles, as shown in Table 13.
The difference of 4.4 inches between the actual and computer-
predicted location of the 34.50 micron beads found in the seventh cup
could result from: (a) assuming a straight-line trajectory between two
consecutive velocity profiles (the error generated from this assumption
will be higher when higher tube settler velocities are encountered); (b)
145
Figure 67. Glass beads settled at the tube end, cup #1, Run #16
Figure 68. Glass beads settled at the tube end, cup #3, Run #16
146
Figure 70. Glass beads settled at the tube end, cup #7, Run
147
Cup # Distance, Particles diameter. Vsc, Computed-predicted

ft microns fpm Location, ft
1 2.85 45.65 0.280 2.85
3 3.52 41.00 0.226 3.54
5 4.19 37.50 0.189 4.37
7 4.85 34.50 0.160 5.22
the allowable 1% error of the computer program gives a 0.6 inch error at
cup in ; and, (c) the interval depths in calculating the required
settling length in the computer program were chosen to equal 0.05 of the
tube diameter. A smaller value could lead to more accurate results but
would be a more time consuming program to run.
The accumulation of these effects on the generated calculation of
settling length using the computer program increases with distance as
the particle travels through the tube. This can be detected by
comparing differences found in cup #5 with that of cup #7. In cup //5
the error was 2.16 inches while in cup §7 the error was 4.4 inches.
These differences were also observed in Runs #7 and #11.
General Comments on Runs #7, #11, #16, and #21
The trend of the results showed that certain precautions should be
taken to obtain a successful run:
1. The suspension concentration should be as low as possible.

148
2. The velocity of the suspension exiting the feeder should match
the tube average flow velocity.
3. The suspension feeding should be initiated only after
establishing a uniform flow pattern in the tube settler. This can be
detected by feeding a dye solution through the suspension feeder.
4. The suction pipe from the suspension tank to the pump's head
should be as short as possible to prevent the accumulation of glass
beads in the suction pipe.
The results of Runs #7, #9, #11, #16, and #21 illustrate the
applicability of Fadel's model.
Part 2
In this phase of the study using the same sample of glass beads as
in previous ones , the variables used were as follows:
Tube Settler Average Sludge
Set £ Run £ Flow Velocity, fpm Accumulation, in
11 3.00 0.00
12 3.35 0.32
1 13 3 .89 0.56
14 7.79 0.80
15 6.49 1.06
16 2.00 0.00
17 2.23 0.32
2 18 2.59 0.56
19 3.19 0.80
20 4.33 1.06
21 4.00 0.00
22 4.46 0.32
3 23 5.19 0.56
24 6.38 0.80
25 8.66 1.06
149
Set #1
The first run in this set was Run #11. The remaining four runs in
this set were conducted under the same conditions of suspension
concentration and feeding tube outlet diameter. As mentioned earlier,
the third tube was used. A complete run, as described in Run #11, was
conducted and the the glass beads settled in cups 1, 3, 5, and 7 were
collected and photographed. Subsequent runs in the set were made after
adding the first, the second, and the third plate to simulate the
accumulation of sludge. The results of these runs are listed in Table
14.
Figure 71 presents the relationship between the depth reduction
ratio, ds/D, and particle diameter for the four locations along the tube
settler. The ds/D ratio represents the relative depth occupied by the
Table 14. Diameter of glass beads collected in the cups in Run set #1
Glass bead diameter, microns
Cup # Distance, full 0.32" 0.56" 0.80" 1.06"^

ft tube sludge sludge sludge sludge
1 2.85 37.60 36.50 32.00 34.20 46.40
3 3.52 34.50 33 .47 29.00 31.20 33.50
5 4.19 32.20 30.43 26.35 28.15 30.00
7 4.85 30.00 28 .15 24.35 25.80 27.40
^Flat plate thickness to simulate sludge accumulation.

<1
Particle Diameter, microns
Figure 71. The relationship between the depth reduction ratio (ds/D) and
particle diameter in Set //I
151
settled sludge in the bottom of the tube settler.
Set £2
The procedure used in the above set was also followed with an
average tube velocity of 2 fpm, assuming a full-tube diameter. The
results are illustrated in Table 15 and shown in Figure 7 2.
Set #3
The runs conducted in this set followed the operating procedures of
Set #1 and Set #2. The full-diameter average flow velocity in the tube
settler was 4 fpm. Table 16 and Figure 73 shows the results obtained.
Table 15. Diameter of glass beads collected in the-cups in Run set #2
Cup # Distance, full 0.32" 0.56" 0.80" 1.06"

ft tube sludge sludge sludge ' sludge
1 2.85 32.20 29.67 20.60 31.00 32.00
3 3.52 30.00 28.91 28.10 28.80 29.60
5 4.19 28.20 26.63 25.67 26.60 26.45
7 4 .52 26.00 24.35 23.36 24.00 25.00

<1
2-
20 22 24 26 2B 30 32 34 36 38 40 42 44 46 48 50
particle diameter in Set //2
M
Lfl
W
30 32 34 36 38 40
particle diameter in Set #3
154
Table 16. Diameter of glass beads collected in the cups in Run set #3
Cup # Distance,. full 0.32" 0.56" 0.80" 1.06"

ft tube sludge sludge sludge sludge
1 2.85 45.65 42.00 38.80 38.84 42.60
3 3.52 41.00 36.50 32.00 33.38 37.20
5 4.19 37.50 34.00 29.67 32.72 34.20
7 4.85 34.50 29.00 27 .38 30.00 32.50
General Discussion on the Run Sets Conducted in Part 2
The three sets of runs showed the same pattern : The diameter of
the particles collected in the same cup decreased until the reduction of
the tube diameter reached 0.56 inch, then the particle diameter
increased with increasing sludge depth. In Figures 71 through 73, the
pattern was consistent in the three sets.
One can easily see the difference between the theoretical effect of
sludge accumulation in tubes (Figures 60 and 61) and the the actual
experimental results (Figures 71 and 73). Figures 60 and 61 suggest
that whenever ds/D reaches 0.4, the settling length (or the settled
particle diameter) will begin to exceed that for the empty tube.
Figures 71 through 73, based on experimental results, suggest that even

when ds/D equals 0.53, the required settling length in a tube with
substantial sludge accumulation is still less than that calculated for
an empty tube using the Fadel model. The difference between the two
patterns result from the assumptions made in the theoretical study,
which did not consider the effect of having flat bottomed velocity
profiles inside the tube. Also, the assumptions neglected the effect of
the wide horizontal velocity profile on the narrow vertical profile we
are dealing with. However, the experimental pattern resulting in
diameter decrease and then increase matches the pattern suggested by
theory.
Actually, there is little or no information available in the
literature which helps to evaluate or calculate velocity profiles under
operation conditions in which sludge has accumulated in the tubes.
The question which must now be answered is not how to develop a
model to handle the "sludge accumulation (which is beyond the scope of
this study), but to suggest how the designer can gain from these
results. The shape of the sludge accumulation inside the tube will
depend on particles concentration, particles size distribution, and
particles inlet distribution to the tube. A survey study of existing
treatment units is required to determine these shapes. The actual
configuration may be considered to behave like the theoretical
configuration assumed in the model development. However, as a factor of
safety, a 40% reduction in the depth could be considered as the maximum
allowable sludge accumulation, meaning that the designer should set the
backflush timing to begin when sludge accumulation reaches 0.4 of the

156
tube depth.
The above discussion illustrates that the designer can use the
Fadel model for designing both the essentially horizontal and steeply
inclined tube configurations, except that one must monitor the
backflushing process in the essentially horizontal tubes. Simple
calculations will be required to determine the time interval between
backflushing. For example, for a treatment plant which treats water
carrying 100 ppm of suspended solid in a tube settler having an 80%
removal efficiency, 80 ppm of solids will be collected in the tubes.
The average flow velocity inside the tubes is assumed to be 2 fpm and the
diameter to be 3 inches and its length to be 6 ft. The storage volume

3
allowed per foot of length at 0.4 ds/D will equal 0.0151 ft . This
volume will fill up with settled sludge after 6.5 hours (assuming 90% of
the settled sludge volume is water). Thus, the whole tube length would
fill to that level in 38 hours, i.e., back flushing must take place at
least once in 38 hours.

157
CONCLUSIONS
In recent years, many systems have been proposed to improve the
performance or increase the capacities of existing water and wastewater
treatment facilities. Shallow-depth clarification through the use of
the tube settlers is one of these systems. The study described herein
was devoted to clarifying some of the unsettled questions about tube
settlers. The following conclusions were reached.
From the Theoretical Study
1. A theoretical model for predicting the performance of circular
(or hexagonal) tube settlers taking into consideration the
effect of the velocity profile variations from uniform at the
tube entrance to a fully developed laminar flow profile at the
end of the transition length was established. This model is
called the Fadel model.
2. The model can be used in three different ways:
a) The model has been written in Fortran language for solution
on a VAXA computer system. (See Appendix A).
b) Model output converted into design charts. Five charts
were constructed for predicting the performance of 2-, 3-, 4-,
5-, and 6-inch diameter tubes. The designer must first
establish an average tube flow velocity, the minimum settling
velocity of the particle to be removed ,the degree of tube
inclination and the tube diameter. He may then obtain the
required tube settler length from the chart.

158
c) An equation (30) which approximates the results obtained
with the Fadel model may be used; however, the designer must
establish the same five values as in b) to use the equation.
An assumption was made to predict the performance of an
essentially horizontal tube under conditions of sludge build up
on the tube bottom. A 40% storage depth for sludge
accumulation was set as the maximum safe limit based on the
assumpti on.
An inlet and outlet arrangement for the essentially horizontal
tube sedimentation tank was proposed.
From the Experimental Study
The experimental work demonstrated the validity of the Fadel
model as a method of determining the settling length required
in circular tube settlers. This is the first time
experimental work has succeeded in verifying the performance of
a tube settler using a design model.
The experimental work showed that the limit for the allowable
maximum flow velocity through tube settler is much higher than
the limits recommended in the literature. An average flow
velocity of up to 5 fpm was used with no observed sweep out of
settled particles.
Uniformly sized particles were collected in the cups of the
laboratory tube settlers. If all particles have the same
shape, this procedure may be used to obtain uniform particle
size diameters under the specified conditions. The

159
experimental equipment is economical to build and easy and fast
to operate.
4. Both the essentially-horizontal and the steeply inclined tubes
can be designed using the Fadel model.
5. For circular tubes, a maximum allowable storage depth of 40% of
the tube diameter is recommended. The experimental work
indicated that a sludge depth up to 53% of the tube diameter
can be attained, without resulting in failure of tube
performance. However, in design, tube settlers should be
cleaned when the sludge accumulation reaches 40% of the tube
diameter.
160
RECOMMENDATIONS
The following recommendations for further work were suggested in
the course of both the experimental and the theoretical studies
conducted for evaluating tube settler performance:
1. Regularly shaped tube settlers, i.e., square, rectangular,
etc., need to be theoretically and experimentally studied using
the procedures described in this dissertation.
2. Irregular shapes (like the sixth shape in Figure 8, and the
shape proposed by Willis, Figure 29) have no theoretically
studied velocity profiles, and will need to be investigated
using the same experimental method.
3. A pilot plant should be established to determine the best inlet
and outlet arrangements for both tube settler configurations.
For example, such a plant could employ the arrangement proposed
for essentially horizontal tubes by Fadel, and proposed for
steeply inclined tubes by Willis.
4. It is recommended that the experimental method be examined as a
device for separating uniform-size particles using other
particle shapes. Also, this experimental method may be used to
determine the shape factor for different particles shapes by
comparing their settling velocities with those obtained from
the spherical glass beads.

161
REFERENCES
1. Beyer, William H. ICRC Standard Mathematical Tables. 26th

edition. Boca Raton, Florida, CRC Press, Inc., 1981.
2. Camp, Thomas R. A Study of the Rational Design of Settling Tanks,

Journal of Sewage Works 8:742-758, June 1936.
3. Camp, Thomas R. Studies of Sedimentation Basin Design. Journal

of Sewage and Industrial Wastes 25:1-12 , 1953.
4. Camp, Thomas R. Sedimentation and the Design of Settling Tanks.

Transaction, American Society of Civil Engineering, Vol. 3,
pages 895-936, April, 1945.
5. Culp, G. L. and S. P. Hansen. Applying Shallow Depth Sedimentation

Theory. Journal of the American Water Works Association 59:1134-
1148, Sept., 1967.
6. Culp, G. L., S. Hansen, and G. Richardson. High Rate

Sedimentation in Water Treatment Works. Journal of the American
Water Works Association 60:681-698, June, 1968.
7. Dick, R. I. Fundamental Aspects of Sedimentation. Journal of

Water and Wastes Engineering 6:45-50, February, 1969.
8. El-Baroudi, H. M. and D. R. Fuller. Tracer Dispersion of High Rate

Settling Tanks. Journal of American Society of Civil Engineering,
99 (EE3):347-369, June, 1973.
9. Fair, G. M., J. C. Gleyer, and D. A. Okum. Water and Wastewater

Engineering. Volume 2. Water Purification and Waste Water
Treatment and Disposal. New York: Wiley and Sons, 1968.
10. Fischers trom, C. N. H. Sedimentation in Rectangular Basins.

Proceedings of the American Society of Civil Engineering, Journal of
Sanitary Engineering Division, pages 325-339, May, 1955.
11. Hanke, E. F. and F. Lssch. Tables of Higher Functions. 6th

edition. New York: McGraw Hill, 1960.
12. Hernandez, John W. and John R. Wright. Design Parameters for Tube-
Settlers. Purdue Proceedings of the 25th Industrial Waste
Conference, May, 1970. Engineering Extension Series No. 137, Part 2,
pages 805-829.
13. King, Williams H. and Ernest F. Brater. Handbook of Hydraulics

for the Solution of Hydraulic Engineering Problems. New York:
McGraw-Hill, 1976.
162
14. Kreyszig, E. Advanced Engineering Mathematics. Third edition,

New York: Wiley, 1972.
15. Langhaar, H. L. Steady Flow in the Transition Length of a Straight

Tube. Journal of Applied Mechanics, A:55-58, June, 1942.
16. Langhaar, H. L. Dimensional Analysis and Theory of Models. New

York: John Wiley and Sons, 1960.
17. Metcalf and Eddy. Wastewater Engineering: Treatment Disposal

Reuse. New York: McGraw Hill, 1979.
18. Rouse, Hunter. Engineering Hydraulics. New York: John Wiley &
Sons, Inc., 1950.
19. Schlichting, H. Boundary-layer Theory. Sixth edition. New

York: McGraw Hill, 1968.
20. Slechta, A. F., and Walter R. Conley. Recent Experience in Plant-

Scale Application of the Settling Tube Concept. Journal of Water
Pollution Control Federation 43(8):1724-1738, December, 1971.
21. Szymanski, V. Sur L'ecoulement Non Permanent Du Fluide Visqueux

Dans Le Tuyau. Proceedings of the International Congress of
Applied Mechanics, Stockholm 1:249-254, April, 1930.
22. Van Vliet, B. M. The Efficiency of Inclined Tube and Plate Modules
in a High Line Clarification Process. Journal of Water Research,
11:783-788, 1977.
23. Verhoff, F. H. Optimal Design of High Rate Sedimentation Devices.

J. ASCE, 105(EE2):199-215, April, 1979.
24. White, M. F. Fluid Mechanics. New York: McGraw Hill. 1979.
25. Willis, R. M. Tubular Settlers - A Technical Review. Journal of

the American Water Works Association 59:331-335, June, 1978.
26. Yao, Kuan M. Column Settling Test and Tube Settling. Journal of
American Water Work Association, 71:109-112, February, 1979.
27. Yao, Kuan M. Theoretical Study of High Rate Sedimentation.

Journal of the Water Pollution Control Federation 42(2, Part 1):218-
228, February, 1970.
28. Yao, Kuan M. Design of High Rate Settlers, Journal of the

Environmental Engineering Division, ASCE, 99(EE5):621-637, February,
1973.
163
ACKNOWLEDGEMENTS
The studies described in this dissertation were funded in part by
the Engineering Research Institute, which the author was funded by and
also an Egyptian Mission Fellowship.
The author wishes to express his appreciation to his many
associates who contributed to the author's work.
• To Dr. E. Robert Baumann, who served as my major professor and
provided research guidance and technical advice as well as moral support.
• To Dr. C. S. Ouiman for his guidance, encouragement, and very
helpful suggestions during the course of study,
• To Dr. T. A. As tin for his encouragement and moral support.
• To Drs. Dennis Johnson and Homer Walker for their help as
members on my committee,
• To ny fellow graduate students, Roger V. Stephenson and Adrian
Hansen for their good and hard times,
• To Rebecca Shiwers , not only for her fast, efficient typing,
but also for her helpful suggestions in editing the disserta
tion.
Also, I want to express a very special appreciation to members of
my family without whose support I would not have found this study so
satisfying to ue.
• To my wife, Rawia, and my sons, Ahmed, Maysara, and Hatem, who
were understanding and patient with me in spite of the long
periods I had to be away from them while doing my research,
studying, and writing this dissertation.

164
• To my mother, my sisters, Rawia and Rabha, and my brother,
Maysara, for their continued support and encouragement during my
graduate studies.
Finally, I want to thank all the Egyptian employees in the Egyptian
Mission Bureau in Washington, D.C. and Cairo, Egypt for their faithful
cooperation and fast efficient work.

165
APPENDIX A. COMPUTER PROGRAM
INTEGER Z,Y,N,M
CHARACTER*! RESP
CHARACTER*! CHOICE
CHARACTER»80 TITLE
CHARACTER*50 GTITLE
REAL A(20),THETA(20),VAV(20),VSC(20),NU(20),LAMBDA(21,31 )
REAL V(21,30),SIGMAC31),D,LM1,LM2,SH,YH,DEG,VISC
REAL XT0T,TT0T,DCT0T,X(30),DA,V-1(30),V2(30),V3(30),Vlt( 30)
REAL V2P(31 ),V3P(31),V4P(31),V2DP(31),V3DP(31),V4DP(31)
REAL VP,T(30),DC(30),VSCV(20),L,LR,LS,LC,LW,VSCH(20),TL,LM
LOGICAL SAMINC , AGRPH
PI=3.14159
C
C NAME OUTPUT FILE: TUBEMD.OUT
C
OPEN (UNIT=99, FILE='TUBEMD.OUT',STATUS='NEW')
C ENTER PROBLEM TITLE
C
WRITE(6,*)(' DO YOU WISH TO HAVE A DATA FILE PREPARED FOR
+ LATER USE OF AGRAPH ? ( YES=Y, NO=ANY KEY )')
READ(*,500,ERR=800,END=900) RESP
IF (RESP .EQ. 'Y' .OR. RESP .EQ. 'y')THEN
WRITE(15,*)(' -1000 1')
AGRPH = .TRUE.
ELSE
AGRPH = .FALSE.
END IF
C
C READ DEPTH V S . LAMBDA TABLE FROM INPUT FILE=F0R01O.DAT
C
DO 10 1=1,21
READ(10,»,ERR=SOO,END=900)(UMBDA(I,K),K=1,7)
READ(10,*,ERR=800,END=900)(LAMBDA(I,K),K=8,13)
READ(10,*,ERR=800,END=900)(LAMBDA(I,K),K=26,31 )
10 CONTINUE
C
C READ SIGMA TABLE FROM INPUT FILE: F0R011.DAT
C
READ(n,*,ERH=800,END=900 )(SIGMA( J),J=1,30)
C
C READ INPUT DATA
C
WRITEC6,119)
119 FORMATCl '/
+ ' ************************************************7
+ ' * » '/
166
* CHOOSE ONE OF THE FOLLOWING OPTIONS ; *7

* * ' /
* 1- SOLVE FOR DEPTH VS. SETTLING LENGTH (LS) *7

* 2- SOLVE FOR DIAMETER VS. LS »7
* 3- SOLVE FOR THETA VS. LS . *7
+ * 4- SOLVE FOR AVERAGE VELOCITY (VAV) VS. LS »7
+ ' * 5- SOLVE FOR SETTLING VELOCITY (VSC) VS LS *7
* 6--SOLVE FOR KIN. VISCOSITY (NU) VS. LS »7 .
* *7
************************************************ 7/)
R£AD(*,600,ERR=800,END=900) CHOICE
IF ( CHOICE .NE. '1' ) THEN
WRITE (6,120)
120 FORMAT(/' HOW MANY VALUES OF ( DIA,THETA,VAV,VSC,OR NU)7
+ ' DO YOU WANT TO SOLVE FOR ? ( ENTER A NUMBER BETWEEN
+ 2 AND 20 )7)
READ(*,502,ERR=300,END=900) NR
602 FORMAT(12)
END IF
IF(CHOICE.EQ.'DTHEN
WRITE(6,*)(' ENTER DIAMETER OF PIPE IN INCHES :')
READ(*,*,ERR=800,END=900)A(1)
A(1)=A(1)/24.
WRITE(6,*)(' EÎJTER THETA IN DEGREES ;')
READ(*,ERR=800,END=900)THETA( 1)
THETA(1)=THETA(1)*3.1U159/180.
WRITEC6,»)('.ENTER VAV IN FEET PER MINUTE :')
READ(»,*,ERRr800,ENDr900)VAV( 1)
WRITE(6,*)(' ENTER VSC IN FEET PER MINUTE ;') .
READ(*,*,ERR=800,END=900)VSC(1)
WRITE(6,»)(' ENTER NU IN FEET SQUARE PER SEC :')
READ(*,»,ERR=800,END=900)NU(1)
NU(1)=NU(1)»60.0
DO 151 K=1,NR
A(K) =A(1)
THETA(K)=THETA(1)
VAV(K) =VAV(1)
VSC(K) =VSC(1)
NU(K) =NU(1)
151 CONTINUE
VSCV(K)=(VSC(K))*COS(THETA(K))
VSCH(K)=(VSC(K))*SIN(THETA(K))
VSCV(K)=VSCV(1)
VSCH(K)=VSCH(1)
ELSE IF (CHOICE .EQ. '2') THEN
WRITE(6,*)(' ENTER VALUES OF DIAMETER IN INCHES :')
READ(*,*,ERR=800,END=900) ( A(K), K=1,NR )
DO 235 K=1,NR
167
A(K)=A(K)/24.0
235 CONTINUE
WRITE(6,*)(' ENTER THETA IN DEGREES ;')
READ(»,*,ERR=800,END=900)THETA(1)
THETA(1)=THETA(1)*3.14159/180.
WRITE(6,*)(' ENTER VAV IN FEET PER MINUTE ;')
READ(*,*,ERR=800,END=900)VAV(1)
WRITE(&,*)(' ENTER VSC IN FEET PER MINUTE ;')
READ(»,*,ERR=800,END=900)VSC(1)
WRITE(6,*)(' ENTER NU IN FEET SQUARE PER SEC :
READ(*,*,ERR=800,END=900)NU(1)
NU(1)=NU(1)»50.0
DO 152 K=1,NR
THETA(K)=THETA(1)
VAV(K) =VAV(1)
VSCCK) =VSC(1)
NU(K) =NU(1)
152 CONTINUE
DO 161 K=1,NR
VSCV(K) =(VSC(K))»COS(THETA(K))
VSCH(K) =(VSC(K))»SIN(THETA(K))
161 CONTINUE
ELSE IF(CH0ICS.EQ.'3') THEN
WRITE(6,»)(' ENTER DIAMETER OF PIPE IN INCHES
READ(»,*,ERH=800,END=900)A(1)
A(1)=A(1)/21|.
WRITE(6,*)(' ENTER VALUES OF THETA IN DEGREES
READ(*,*,ERR=800,END=900) ( THETA(K),K=1,NR)
DO 236 K=1,NH
THETA(K)=THETA(K)*3.1416/180.0
236 CONTINUE
WRITE(6,*)(' ENTER VAV IN FEET PER MINUTE :')
wRITE(6,*)(' ENTER VSC IN FEET PER MINUTE :')
wRITE(6,»)(' ENTER NU IN FEET SQUARE PER SEC :
READ(*,«,ERRr 800,END=900)NU(1)
NU(1)=NU(1)»60.0
DO 153 K=1,NR
A(K) =A(1)
VAV(K) =VAVC1)
VSC(K) =VSC(1)
NU(K) =NU(1)
153 CONTINUE
DO 162 K=1,NR
VSCV(K) =(VSC(K))*COS(THETA(K))
VSCH(K) =(VSC(K))*SIN(THETA(K))
162 CONTINUE
168
ELSE IF(CHOICE.EQ.'4') THEN

WRITE(6,*)(' ENTER DIAMETER OF PIPE IN INCHES :
READ(*,*,ERRr800,END=900)A(1)
A(1)=A(l)/24.
WRITE(6,*)(' ENTER THETA IN DEGREES :')
READ(*,*,ERRr800,END=900)THETA(1)
THETA(1)=THETA(1)*3.14159/180.
WRITE(6,*)C' ENTER VALUES OF VAV IN FEET PER MINUTE
READ(*,*,ERR=800,END=900) (VAV(K),K=1,NR)
WRITE(6,*)(' ENTER VSC IN FEET PER MINUTE :')
WRITE(6,*)(' ENTER NU IN FEET SQUARE PER SEC :')
READ(*,*,ERR=800,END=900)NU(1)
NU(1)=NU(1)*60.0
DO 154 K=1,NR
A(K) =A(1)
THETA(K)=THETA(1)
VSC(K) =VSC(1)
NU(K) =N0(1)
154 CONTINUE
DO 163 K=1,NR
VSCV(K) =(VSCCK))»C0S(THETA(K))
VSCH(K) =(VSCCK))*SIN(THETA(K))
163 CONTINUE
ELSE IF(CHOICE.EQ.'5')THEN
READ(*,*,ERRrSOO,END=900)A(1)
A(1)=A(1)/24.
WRITE(6,*)(' ENTER THETA" IN IE GREES :')
READ(»,ERR=800,END=900)THETA(1)
• TKETA(1)=THETAC1)*3.14159/180,
WRITE(6,*)(' ENTER VALUES OF VSC IN FEET PER MINUTE
READ(*,*,ERRrSOO,END=900) (VSC(K),K=1,NR)
WRITE(6,*)(' ENTER NU IN FEET SQUARE PER SEC ;')
READ(*,»,ERR=800,END=900)NU(1)
NU(1)=NU(1)*60.0
DO 155 K=1,NR
A(K) =A(1)
THETA(K)=THETA(1)
VAV(K) =VAV(1)
NU(K) =NU(1)
155 CONTINUE
DO 164 K=1,NR
164 CONTINUE
169
ELSE IF(CHOICE.EQ.'6') THEN

READ(*,*,ERR=800,END=900)A(1)
A(1)=A(1)/24.
WRITE(6,*)(' ENTER THETA IN DEGREES ;')
READ(»,»,ERRrBOO,END=900)THETA(1)
THETA(1)rTHETA(1)»3.14159/180.
READ(*,»,ERR=800,END=900)VAV(1)
WRITE(6,*)(' ENTER VSC IN FEET PER MINUTE :')
READ(*,»,ERR=800,END=900)VSC(1)
WRITE(6,*)(' ENTER VALUES OF NU IN FEET SQUARE PER SEC:')
READ(»,*,ERR=800,END=900)(NU(K),K=1,NR)
DO 239 K=1,NR
NU(K)=NU(K)*60.0
239 CONTINUE
DO 156 K=1,NR
A(K) =A(1)
THETA(K)=THETA(1)
VAV(K) =VAV(1)
VSC(K) rVSCd)
156 CONTINUE
DO 165 K=1,NR
165 CONTINUE
END IF
WRITE(5,»)(' ENTER WATER TEMP IN DEC F :')
HEAD(*,*,ERR=800,END=900)TEMP
IFCCHOICE.EQ.'1') NR=1
C INITIALIZE DISTANCE SUM, TIME SUM, CALCULATED DEPTH SUM, & LOGICAL
C
r
C HEADING OF OUTPUT TABLE
C
IFCCHOICE.EQ.'l')THEN
PRINT *, 'DIAMETER = ',24*A(L),' IN.'
736 F0RMAT(F3.1)
PRINT *,' THETA =',THETA(1),' DEG.'
737 F0RMAT(F4.1)
PRINT », ' AVERAGE VELOCITY = ',VAV(1),' FPM'
738 FORMAT(F5.3)
PRINT *,' SETTLING VELOCITY = ',VSC(1), 'FPS'
739 FORMAT(F5.3)
PRINT KINEMATIC VISCOSITY = ',NU(1),' SQ.FT/S'
7U0 FORMAT(F10.8)
WRITE(6,200)
170
WRITE(6,650)
WRITE(99,200)
WRITE(99,650)
ELSE
WRITE (99,741)
741 FORMAT('1',5X,'DIAM.',4X,'VAV',5X,'VSC',4X,'TEMP.',3X,
+ 'K. VISCOSITY',3X,'THETA',3X,'SET. LENGTH')
WRITE(99,742)
742 FORMAT(6X,'
WRITE(99,743)
743 F0RMAT(6X,'IN.',5X,'FPM',7X,'FPM',5X,'F',7X,'SQ.FT/SEC',5X,
+ 'DEG.',5X, 'FEET')
WRITE(99,742)
WRITE(99,745)
WRITE(6,741)
WRITE(6,742)
WRITE(6,743)
WRITE(6,745)
END IF
C SOLVE FOR THE TIME, DISTANCE, AND DEPTH SUMS OF PARTICLE AT EACH
C SECTION
DO 100 L=1,NR ,
XTOT=0
TTOT=0
DCT0T=0
SAMINC =.FALSE.
DO. 101 1=1,21
LAMBDA(I,1)=LAMBDA(I,1)*A(L)
101 CONTINUE
DO 30 1=1,21
DO 20 K=2,31
V(I,(K-1))=VAV(L)»(UMBDA(I,K))-VSCH(L)
20 CONTINUE
30 CONTINUE
DO 50 J=1,29
C FIND WIDTH OF EACH SECTION
X(J)=((SIGMA(J+1)-SIGMA(J))*(VAV(L)-VSCH(L))*A(L)*»2)/NU(L)
C ASSUME DEPTH OF THE PARTICLE FOR SECTION
DA=0.10»A(L)
C SOLUTION
400 IF(J.EQ.1)THEN
V1(J)=VAV(L)
V2(J)=0
V3(J)=VAV(L)
V4(J)=DA*V(2,2)/(0.10»A(L))
ELSE
171
V1(J)=VU(J-1)
Z =INT(DCT0T/(0.10*A(L)))
Y =Z+1
N =INT((DCTOT+DA)/(0.10*A(L)))
M =N+1
IF(DCTOT.GE.(N»0.1»A(L)).AND.DCTOT.LE.(M*0.1*A(L)).AND.(DCTOT+DA)
.GE.(N*0.1*A(L)).AND.(DCTOT+DA).LE,(M*0.1*A(L)))THEN
SAMINC=.TRUE.
ELSE
SAMINCr.FALSE.
END IF
IF( SAMINC .EQ. .TRUE. .AND. N .EQ. 0)THEN
V2DP(J)=V(2,(J+1))
V3DP(J)=V(2,J)
V4DP(J)=V(2,(J+1))
V2(J) =V2DP(J)«DCT0T/(0.10*A(L))
V3(J) =V3DP(J)*(DCT0T+DA)/(0.10*A(L))
V4(J) =V4DP(J)*(DCT0T+DA)/(0.10*A(L))
ELSE
Z=Z+1
Y=Y+1
N=N+1
M=M+1
V2P(J) =V(Z,(J+1))
V2DP(J)=V(Y,(J+1))
V3P(J) =V(N,J)
V3DP(J)=V(M,J)
V4P(J) =V(N,(J+1))
V4DP(J)=V(M,(J+1))
V2(J)=V2P(J)+(DCTOT-(0.10»A(L)*(Z-1)))»(V2DP(J)-V2P(J))/
(0.10»A(L))
V3(J)=V3P(J)+((DCTOT+DA)-(0.10»A(L)»(N-1)))»(V3DP(J)-V3P(J))/
(0.1»A(L))
VU(J)=V4P(J)+((DCTOT+DA)-(0.10*A(L)*(N-1)))*(V4DP(J)-V4P(J))/
(0.1*A(L))
END IF
END IF
VP =(V1(J)+V2(J)+V3(J)+V4(J))/4,0
T(J) =X(J)/VP
DC(J) =T(J)*VSCV(L)
XTOT =XTOT+X(J)
TTOT =TTOT+T(J)
DCTOT =DCTOT+DC(J)
IF (DCTOT .EQ. 2*A(L) .AND. CHOICE .EQ. '1' ) THEN
IF (AGRPH)CALL AGROUT(DCTOT,TTOT, XTOT,A)
CALL OUT (DCTOT,TTOT,XTOT,AGRPH)
GOTO 4000
END IF
172
IF (DCTOT .EQ. 2*A(L) .AND. CHOICE .NE.T) GOTO 520

IF(DCTOT.GT.(2«A(L)))THEN
TTOT =TTOT-T(J)
XTOT =XTOT-X(J)
TL =(2*A(L)-(DCT0T-DC(J)))/VSCV(L)
VP =(V1(J)+V2(J))/4.0
DCTOT = 2*A(L)
TTOT =TTOT+TL
XTOT =XTOT+(TL*VP)
IF (CHOICE .NE. '1') GOTO 520
IF(CHOICE.EQ.'r) THEN
IF(AGRPH)CALL AGROUT(DCTOT,TTOT,XTOT,A)
CALL OUT(DCTOT,TTOT,XTOT,AGRPH)
GOTO 4000
END IF
ELSE
IF((ABS(DC(J)-DA)/DA).LE. O.ODTHEN
IF(CHOICE.EQ.'r) THEN
IF(AGRPH)CALL AGROUT(DCTOT,TTOT,XTOT,A)
CALL OUT(DCTOT,TTOT,XTOT,AGRPH)
END IF
SAMINCz.FALSE.
GOTO 50
ELSE
DCT0T=DCT0T-DC(J)
XTOTz XTOT-X(J)
TT0T=TT0T-T(J)
DA=DC(J)
GOTO 400
END IF
END IF
CONTINUE
IF(DCT0T.LT.(2*A(L)).AND.CHOICE.EQ.'1')THEN
S=2*A(1)-DCTOT
YH=S/(2*A(1))
LM=8.0»(VAV(1)/VSCV(1))*(YH**2/2.0-YH**3/3.0)-YH*TAN(THETA(1))
LM1=2*A(1)»LM
XTOT=XTOT+LM1
DCT0T=2«A(1)
TT0T=TT0T+S/VSCV(1)
IF(AGRPH) CALL AGROUT (DCTOT,TTOT,XTOT,A)
CALL OUT (DCTOT,TTOT,XTOT,AGRPH)
GOTO 4000
ELSE IF(DCTOT.LT.(2»A(L)).AND.CHOICE.NE.'1')THEN
S=2.0*A(L)-DCTOT
YH=S/(2*A(L))
LM=8.0*(VAV(L)/VSCV(L))»(YH**2/2.0-YH**3/3.0)-(YH*TAN(THETA(L)))
LM1=2»A(L)»LM
173
XT0T=XT0T+LM1
DCT0T=2*A(L)
TTOT=TTOT+S/VSCV(L)
GOTO 520
END IF
520 D=24»A(L)
DEG=l80*THETA(L)/3.14159
VISC=NU(L)/60
WRITE(6,7«5)D,VAV(L),VSC(L),TEMP,VISC,DEG,XTOT
WRITE(99,7U5)D,VAV(L),VSC(L),TEMP,VISC,DEG,XTOT
745 F0RMATC6X,F3.1,4X,F5.3,5X,F5.3,3X,F4.1,5X,F10.8,4X,F4.1,2X,
+ F5.3)
100 CONTINUE
C PROVIDE OPTION FOR SOLVING THE PROBLEM BY OTHER
C METHODS (YAO,GULP,WILLIS)
4000 IFCCHOICE.EQ.'DTHEN
PRINT *,'D0 YOU WISH TO SOLVE THE SAME PROBLEM BY OTHER
+ METHODS ( ENTER Y OR N )'
READ(*,600,ERR=800,END=900)RESP
IF(RESP .EQ. 'Y' .OR. RESP .EQ. 'y') THEN
L R=(0.227»VAV(1)*A(1)**2)/NU(1)
L =( 1.333*VAV(1)/VSC(1)-SIN(THETA(1)))*2*A(1)/COS(THETA(1))
IFCLR.GE.DTHEN
LS=2*L
ELSE
LS=L+LR
END IF
WRITE(6,*)(' YAO METHOD :')
WRITE(99,*)C' YAO METHOD :')
WRITE(99,»)(' ************')
WRITE(6,*)(' ************')
WRITE(99,*)(' TRANSITION LENGTH =',LR,' FT')
WRITE(99,*)(' MODEL LENGTH =',L,' FT')
WRITE(99,*)(' SUGGESTED TUBE LENGTH =',LS,' FT')
WRITE(6,*)(' TRANSITION LENGTH =',LR,' FT')
WRITE(6,*)(' MODEL LENGTH =',L,' FT')
WRITE(6,»)(' SUGGESTED TUBE LENGTH =',LS,' FT')
LC=2»A(1)*(VAV(1)-VSCH(1))/VSCV(1)
WRITE(6,»)('GULP METHOD: 'SUGGESTED TUBE LENGTH: ',LC,' FT')
WRITE(99,»)('GULP METHOD: ','SUGGESTED TUBE LENGTH: ',LC,' FT')
LW=2*A(1)»VAV(1)/VSCV(1)
WRITE(6,*)('WILLIS METHOD; ','SUGGESTED TUBE LENGTH: ',LW,' FT')
WRITE(99,*)('WILLIS METHOD: ','SUGGESTED TUBE LENGTH: ',LW,'FT')
END IF
END IF
GOTO 3000
500 FORMAT(ABO)
600 FORMAT(AD
174
800 PRINT ERROR ENCOUNTERED IN ENTERING DATA - START AGAIN'

GOTO 3000
900 PRINT END OF FILE ENCOUNTERED IN READING INPUT DATA'
GOTO 3000
200 FORMATCl SETTLING DEPTH(FT) DETENTION TIME(MIN) SETTL
+ING LENGTH(FT)')
650 FORMAT('
201 F0RMAT(9X,F6.4,12X,F7.4,12X,F9.4)
3000 CALL EXIT
END
SUBROUTINE OUT(DCTOT,TTOT, XTOT,AGRPH)
REAL DCTOT,TTOT,XTOT
LOGICAL AGRPH
WRITE(5,201)DCTOT,TTOT,XTOT
WRITE(99,201)DCTOT,TTOT,XTOT
201 F0RMAT(9X,F6.i|,12X,F7.4,12X,F9.4)
RETURN
END
SUBROUTINE AGROUT(DCTOT,TTOT,XTOT,A)
REAL Y,DCTOT,TTOT,XTOT,A
CHARACTER*40 GTITLE
Y = 2*A(L) - DCTOT
WRITE(15,*) XTOT,Y
CALL OUT(DCTOT,TTOT,XTOT,AGRAPH)
RETURN
800 PRINT ERROR ENCOUNTERED IN ENTERING DATA - START AGAIN'
GOTO 3333
900 PRINT END OF FILE ENCOUNTERED IN READING INPUT DATA'
175
File: FOR010.DAT
0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.1 0.8 0.7459 0.7219 0.6859 0.6462 0.6250
0.6033 0.5807 0.5575 0.5337 0.5098 0.4859
0.4628 0.4744 0.4519 0.4410 0.4313 0.4215
0.4159 0.4103 0.4057 0.4010 0.3981 0.3954
0.3913 0.3888 0.3863 0.3841 0.3841 0.3821
0.2 1.1079 1.0467 1.0303 1.0093 0.9826 0.9671
0.9498 0.9310 0.9104 0.8883 0.8649 0.8406
0.8283 0.8160 0.8036 0.7912 0.7803 0.7693
0.7629 0.7564 0.7509 0.7454 0.7423 0.7389
0.7365 0.7340 0.7308 0.7275 0.7258 0.7229
0.3 1.1079 1.1603 1.1626 1.1626 1.1586 1.1548
1.1497 1.1430 1.1346 1.1244 1.1125 1.0989
1.0917 1.0845 1.0767 1.0689 1.0618 1.0547
1.0502 1.0456 1.420 1.0383 1.0353 1.0336
1.0319 1.302 1.0276 1.0249 1.0237 1.0219
0.4 1.1079 1.2047 1.2198 1.2357 1.2511 1.2548
1.2654 1.2718 1.2772 1.2817 1.2850 1.2871
1.2874 1.2876 1.2874 1.2873 1.2869 1.2865
1.2857 1.2848 1.2843 1.2838 1.2835 1.2833
1.2823 1.2827 1.2818 1.2808 1.2805 1.2802
0.5 1.1079 1.2222 1.2447 1.2708 1.3 1.3159
1.3327 1.3502 1.3681 1.3865 1.4049 1.4229
1.4317 1.4405 1.4436 1.4566 1.4642 1.4718
1.4752 1.4788 1.4821 1.4856 1.4870 1.4891
1.4911 1.4930 1.4936 1.4941 1.4956 1.4978
0.6 1.1079 1.2292 1.2557 1.2877 1.3260 1.3479
1.3719 1.3979 1.4258 1.4558 1.4873 1.52
1.5366 1.5532 1.5694 1.5855 1.6009 1.6163
1.6242 1.6320 1.6393 1.6466 1.6506 1.6550
1.6591 1.6632 1.6652 1.6671 1.6712 1.6756
0.7 1.1079 1.2320 1.2605 1.2944 1.3398 1.3675
1.3945 1.4266 1.4620 1.5008 1.5428 1.5874
1.6106 1.6337 1.6568 1.6789 1.7022 1.7245
1.7361 1.7476 1.7587 1.7697 1.7759 1.7818
1.7881 1.7944 1.7977 1.8009 1.8060 1.8130
0.8 1.1079 1.2331 1.2627 1.3 1.3470 1.3752
1.4072 1.4452 1.4833 1.5288 1.5782 1.6315
1.6596 1.6876 1.7159 1.7442 1.7718 1.7994
1.8076 1.8158 1.8361 1.8563 1.8641 1.8717
1.8795 1.8873 1.8921 1.8969 1.9032 1.9116
0.9 1.1079 1.2336 1.2636 1.3017 1.3504 1.38
1.4137 1.4520 1.4953 1.5440 1.5978 1.6564
1.6874 1.7183 1.7501 1.7816 1.8126 1.3435
176
1.8601 1.8767 1.8923 1.9078 1.9255 1.9265

1.9340 1.9425 1.9482 1.9538 1.9680 1.9745
1.0 1.1079 1.2337 1. 2 6 3 8 1.3022 1.3514 1.3814
1.4157 1.4547 1.4989 1.5488 1.6041 1.6645
1.6965 1.7286 1.7613 1.7939 1.8260 1.858
1.8753 1.8923 1.9000 1.9249 1.9342 1.9432
1.9521 1.9609 1.9668 1.9727 1.98 1.99
1.1 1.1079 1.2336 1.2636 1.3017 1.3504 1.38
1.4137 1.4520 1.4953 1.5440 1.5978 1.6564
1.6874 1.7183 1.7501 1.7816 1.8126 1.8435
1.8601 1.8767 1.8923 1.9078 1.9255 1.9265
1.9340 1.9425 1.9482 1.9538 1.9680 1.9745
1.2 1.1079 1.2331 1.2627 1.3 1.3470 1.3752
1.4072 1.4452 1.4833 1.5288 1.5782 1.6315
1.6596 1.6876 1.7159 1.7442 1.7718 1.7994
1.8076 1.8158 1.8361 1.8563 1.8641 1.8717
1.8795 1.8873 1.8921 1.8969 1.9032 1.9116
1.3 1.1079 1.2320 1.2605 1.2944 1.3398 1.3675
1.3945 1.4266 1.4620 1.5008 1.5428 1.5874
1.6106 1.6337 1.6568 1.6789 1.7022 1.7245
1.7361 1.7476 1.7587 1.7697 1.7759 1.7818
1.7881 1.79U4 1.7977 1.8009 1.8060 1.8130
1.4 1.1079 1.2292 1.2557 1.2877 1.3260 1.3479
1.3719 1.3979 1.4258 1.4558 1.4873 1.52
1.5366 1.5532 1.5694 1.5855 1.6009 1.6163
1.6242 1.6320 1.6393 1.6466 1.6506 1.6550
1.6591 1.6632 1.6652 1.6671 1.6712 1.6756
1.5 1.1079 1.2222 1.2447 1.2708 v.3 1.3159
1.3327 1.3502 1.3681 1.3865 1.4049 1.4229
1.4317 1.4405 1.4486 1.4566 1.4642 1.4718
1.4752 1.4788 1.4821 1.4856 1.4870 1.4891
1.4911 1.4930 1.4936 1.4941 1.4956 1.4978
1.5 1.1079 1.2047 1.2198 1.2357 1.2511 1.2548
1.2654 1.2718 1.2772 1.2817 1.2850 1.2871
1.2874 1.2876 1.2874 1.2873 1.2869 1.2865
1.2857 1.2848 1.2843 1.2838 1.2835 1.2833
1.2823 1.2827 1.2818 1.2808 1.2805 1.2802
1.7 1.1079 1.1603 1.1626 1.1626 1.1586 1.1548
1.1497 1.1430 1.1346 1.1244 1.1125 1.0989
1.0917 1.0845 1.0767 1.0689 1.0618 1.0547
1.0502 1.0456 1.420 1.0383 1.0353 1.0336
1.0319 1.302 1.0276 1.0249 1.0237 1.0219
1.3 1.1079 1.0467 1.0303 1.0093 0.9826 0.9671
0.9498 0.9310 0.9104 0.8883 0.8649 0.8406
0.8283 0.8160 0.8036 0.7912 0.7803 0.7693
0.7629- 0.7564 0.7509 0.7454 0.7423 0.7389
0.7365 0.7340 0.7308 0.7275 0.7258 0.7229
1.9 0.8 0.7459 0.7219 0.6859 0.6462 0.6250
177
0.6033 0.5807 0.5575 0.5337 0.5098 0.4859

0.4628 0.4744 0.4519 0.4410 0.4313 0.4215
0.4159 0.4103 0.4057 0.4010 0.3981 0.3954
0.3913 0.3888 0.3863 0.3841 0.3841 0.3821
2.0 0.0 0.0 0.0 0.0 0.0 0.0
O.O 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
File: FOROU.DAT
0.0
0.00082
0.00418
0.00541
0.0072
0.00997
0.01188
0.0143
0.0174
0.0214
0.0267
0.0335
0.0426
0.0483
0.0549
0.0625
0.0715
0.0821
0.0947
0.1034
0.1132
0.1241
0.1365
0.1459
0.1560
0.1671
0.1795
0.193^
0.2091
0.2270
0.2430
179
APPENDIX B: NUMERICAL SOLUTION OF FADEL'S MODEL
The required steps for the model are calculated below:
1. Assume:
1- Diameter = 2"
2- Average flow velocity, Vav = 2.88 fpm
3- Particle settling velocity, Vsc = 0.0742 fpm
2
4- Kinematic viscosity, = 1 x 10 ft /sec
5- Tube length = unlimited
Reynolds number = Vav x D/4
_ 2.88 fpm X 2" x (1 ft/12 in) ^ < 500

4 X 1x10' ft"/sec X (60 sec/min)
laminar flow condition.
2. Calculate velocity profiles:
Sample section velocity profiles were calculated using Eq.
(19) and Table 4.
Profile of section I:
From Table 4: CT = 0.00082, Y = 20:
The hyperbolic Bessel function of 0. order and 2nd order for
values higher than 10 are not available in the literature. Accordingly,
the velocity profile for this section was established by direct
measuring from Figure 34.
r
q = — IA.
^ a
0 1.1079
0.1 1.1079
0.2 1.1079
180
0.3 1 .1079
0.4 1 .1079
0.5 1 .1079
0.6 1 .1079
0.7 1 .1079
0.8 1 .1079
0.9 0.8000
Profile of section II:
Table 4:
a =: 1D.00418 Y = 10
Reference [11] IgCy) = 2816 1,(10) = 2281.75
q = f q IgCyq) À*
0.0 0 1 1.2337
0.1 1 1.2661 1.2336
0.2 2 2.2800 1 .2331
0.3 3 4.8810 1 .2320
0.4 4 11.30 20 1 .2292
0.5 5 27.2400 1.2222
0.6 6 67.2300 1.2047
0.7 7 168.5900 1.1603
0.8 8 427.6000 1.0467
0.9 9 1093.6000 0.7549
A. was calculated using Eq . (9).

181
Profile of section III:
From Table 4:
a = 0.00541 y = 9
From Reference |ll] Iq( y) = 1093.6 ^2^^^ ~ 864.535
9 = f q . igCyq) x'
0.0 0.0 1 1.2638
0.1 0.9 1.2130 1 .2636
0.2 1.8 1.9900 1.2627
0.3 2.7 3.8420 1.2607
0.4 3.6 8.0280 1.2557
0.5 4.5 17.4800 1.2447
0.6 5.4 39.0100 1.2198
0.7 6.3 88.4600 1.1626
0.8 7.2 202.9000 1 .0303
0.9 8.1 469.5000 0.7219
The remaining velocity profiles can be found in Appendix A, file
FOR010.DAT.
3. The distance between two adjacent profiles is calculated using
(20) as
L : ^ Vav/4
Thus, the distance between section I and II is:
= (0.0008 2 - 0) (1/12)2 X 2.88 fpm/4 x 0.00060 ft^/min
= (0.00082 - 0) X 33 .33 = 0.027 ft ,
and the distance between section II and III is:
= (0.00722 - 0.00082) x 33.33 = 0.213 ft,

182
and so on. The remaining distances between the other sections can be
calculated in the same manner.
4. To follow the particle path through the tube, using the ten steps of
the model (page 63).
a) From tube entrance to section II.
First trial:
Assume X = 0.05 D = 0.05 x 2"/12 = 0.00833 ft
VI = Vav = 2.88 fpm
VII = 0 X 2.88 = 0 fpm
VI' = Vav = 2.88 fpm
VII' = X Vav = 0.9 X 2.88 = 2.592 fpm
Vav = (2.88 + 0 + 2.88 + 2.592)/4 = 2.088 fpm
The time required for the particle to travel from the entrance to
section I is calculated as:;
t = 0.00833 ft/2.088 fpm = 0.0129 min.
Actual settling depth:
X = 0.0129 min x 0.0742 fpm = 0.000959 ft
0.05 D » 0.000959, a second trial is required.
Second trial:
Reassume X = 0.000595 ft = X from the first trial

a
VI = 2.88 fpm
VII = 0 fpm
VI' = 2.88 fpm
VII' = 0.9 X 0.000959 ft x 2.88 fpm/0.00083 ft
= 0.3 fpm
Vav = (2.88 + 0 + 2.88 + 0.3)/4 = 1.51 fpm

183
t= 0.00833 ft/1.51 fpra = 0.0177 min
X = 0.0177 min x 0.0742 fpm = 0.00132 ft

a
X assumed « X^ calculated, a third trial is required.
Third trial:
Reassume X = 0.00132 ft = X from the second trial

a
VI = 2.88 fpm
VII = 0 fpm
VI' = 2.88 fpm
VII' = 0.9 X 0.00132 ft X 2.88 fpm x 0.00132 ft
= 0.41 fpm
Vav = (2.88 + 0 + 2.88 + 0.41)/4 = 1.54 fpm
t = 0.00833 ft/1.54 fpm = 0.0175 min
X = 0.0175 min x 1.54 fpm •- 0.00130 ft

a
Because X assumed is approximately equal to the value
calculated (1 + 0.05) X^, then the particle will travel from section II
to section III starting at a depth equal to 0.0013 ft from the tube
invert; the distance from the tube entrance is 0.00833 ft.
b) From section II to section III
First trial:
Assume X = 0.05 D = 0.00832 ft
VI = VII' in the previous trial
= 0.41 fpm
VII = 1.872 X 0.0013/0.00833 = 0.293 fpm
VI' = 2.592 + (3.18 - 2.59) x 0.0013/0.00833
= 2.68 fpm
184
VII' = 1.872 + (2.88 - 1 .87) x 0.0013/0.00833
= 2 .027 fpm
Vav = (0.41 + 0.30 + 2.027 + 2.69)/4
= 1 .36 fpm
t = 0.212/1.36 = 0.156 min
X = 0.156 X 0.0742 = 0.0116 ft

a
X actual » X assumed
a
Second trial:
Assume X = 0.0116 ft
VI = 0.41 fpm
VII = 0.30 fpm
VI' = 2.592 + (3.188 - 2.592) x (0.0116
+ 0.0013 - 0.0083)/0.0083 = 2.92 fpm
VII' = 1.872 + (2.880 - 1.872) x (0.0116
+ 0.0013 - 0.0083)/0.0083 = 2.43 fpm
Vav = (0.41 + 0.30 + 2.92 + 2.43)/4
= 1.51 fpm
t = 0.213/1.51 = 0.14 min
X = 0.14 X 0.0742 = 0.0104 ft

a
. Accuracy = (0.0116 - 0.0104) x 100/0.0116 = 10% .
Since the model requires a 95 percent confidence level
(+5%), these results are insufficient; a third trial is required.
Third trial:
Assume X = 0.0104 ft
VI = 0.41 fpm
185
VII =0.30 fpm
VI' = 2.592 + (3.188 - 2 .592) x (0.0104
+ 0.0013 - 0.0083)70.0083 •= 2.83 fpm
VII' = 1.872 + (2.880 - 1.872) x (0.0104
+ 0.0013 - 0.0083)70.0083 = 2.28 fpm
Vav = (0.41 + 0.30 + 2.83 + 2.28)74
= 1.43 fpm
t = 0.21371.43 = 0.146 min
X = 0.146 X 0.0742 = 0.0108 ft

a
Accuracy = (0.0104 - 0.0108) x 10070.0104
= 4% < 5% .
Therefore, we may confidently state that after a specified time (0.0175
+ 0.14, 0.1575 min) and at a specified distance from the entrance (0.027
+ 0.213 = 0.24 ft) the particle will settle to a specific depth (equal to
0.0013 + 0.0104 = 0.0134 ft).
The same calculation was repeated until the particle reached the
bottom of the tube. The required length was found to be 8.67 ft, and the
detention time to be 2.25 min.

Model For Horizontal Tube Settlers

Uploaded by

Copyright:

Available Formats

Model For Horizontal Tube Settlers

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Model For Horizontal Tube Settlers

Uploaded by

Copyright:

Available Formats

Iowa State University Capstones, Theses and

Retrospective Theses and Dissertations

Model for horizontal tube settlers

Follow this and additional works at: https://lib.dr.iastate.edu/rtd

The following explanation of techniques is provided to help clarify markings or

2. When an image on the film is obliterated with a round black mark, it is an

4. For illustrations that cannot be satisfactorily reproduced by xerographic

Fadel, Ahmed Ashry

MODEL FOR HORIZONTAL TUBE SETTLERS

Iowa State University PH.D. 1985

1. Glossy photographs or pages

2. Colored illustrations, paper or print

3. Photographs with dark background

4. Illustrations are poor copy

5. Pages with black marks, not original copy

6. Print shows through as there is text on both sides of page

7. Indistinct, broken or small print on several pages

8. Print exceeds margin requirements

9. Tightly bound copy with print lost in spine

10. Computer printout pages with indistinct print

12. Page(s) seem to be missing in numbering only as text follows.

13. Two pages numbered . Text follows.

14. Curling and wrinkled pages

15. Dissertation contains pages with print at a slant, filmed as received

Ahmed Ashry Fadel

A Dissertation Submitted to the

Graduate Faculty in Partial Fulfillment of

Requirements for the Degree of

Department: Civil Engineering

Signature was redacted for privacy.

Signature was redacted for privacy.

For pfie Major Department

Signature was redacted for privacy.

For the Gr.

Iowa State University

Discrete Particle Settling 3

Development of Tube Settlers 10

Design and Experimental Approaches • 20

Factors Affecting the Performance of Tube Settlers 54

Actual Particle Trajectory 58

The Fadel model 63

Essentially-Horizontal vs Steeply-Inclined Tube 66

Inlet and Outlet Arrangement 68

Factors Affecting Tube Performance Using Fadel's Model 80

Method of Using the Model 94

Essentially Horizontal Tubes 102

EXPERIMENTAL STUDY 113

Equipment and Materials 113

Experimental Procedure 124

RESULTS AND DISCUSSION 130

Run #1. (2/18/85) 130

General Comments, Runs #1 - #6 135

Run #7. (3/17/85) 136

General Comments on Runs #7, #11, #16, and #21 147

General Discussion on the Run Sets Conducted in Part 2 154

From the Theoretical Study 157

From the Experimental Study 158

APPENDIX A. COMPUTER PROGRAMS 165

APPENDIX B. NUMERICAL SOLUTION OF FADEL'S MODEL 179