EPA-660/2-/3-012
October 1973
                         Environmental Protection Technology Series
  Negatively  Buoyant  Jets
  In  A Cross Flow
                                   Office of Research and Development

                                   U.S Environmental Protection Agency
                                   Washington, D.C. 20460

-------
                    RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series.  These five broad categories were established to
facilitate further development and application of environmental
technology.  Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface
in related fields.  The five series are:

     1.   Environmental Health Effects Research
     2.   Environmental Protection Technology
     3.   Ecological Research
     4.   Environmental Monitoring
     5.   Socioeconomic Environmental Studies

This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY STUDIES series.  This series describes research performed
to develop and demonstrate  instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and
non-point sources of pollution.  This work provides the new or
improved technology required for the control and treatment of
pollution sources to meet environmental quality standards.
                        EPA REVIEW NOTICE

This report has been reviewed by the Office of Research and
Development, U.S. Environmental Protection Agency, and approved
for publication.  Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commerical
products constitute endorsement or recommendation for use.

-------
                                                         EPA-660/2-73-012
                                                         October 1973
                    NEGATIVELY  BUOYANT JETS IN

                          A CROSS  FLOW
                                By

                        Jerry Lee Anderson
                         Frank L. Parker
                        Barry A. Benedict

                         Grant # R-800613
                         Project 16130  FDQ
                       Program Element  1BA032

                           Project Officer

                         Mr. Frank Rainwater
                  Pacific Northwest Water Laboratory
                National Environmental  Research Center
                      Corvallis, Oregon  97330
                           Prepared for
               OFFICE OF  RESEARCH AND DEVELOPMENT
             U. S. ENVIRONMENTAL PROTECTION  AGENCY
                     WASHINGTON, D.C.  20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $2.50

-------
                             ABSTRACT






     Negatively buoyant jets, or sinking jets, can be observed in many



problems of pollutant discharge.  Any chamical waste that is heavier than



the receiving water into which it is discharged may act as a negatively



buoyant jet.  In addition, when water is taken from the hypolimnion of



a deep lake or reservoir and used as cooling water, the temperature, and,



consequently, the  discharge may behave  like a negatively buoyant jet.



      Two  existing  jet  diffusion models  have been  utilized to predict



the  trajectory  and dilution  of a positively buoyant jet, or a rising jet,



and  have  been modified to  account for the  sinking effect.



      Twenty-four experimental  investigations were conducted involving



different combinations of densimetric Froude number  , velocity ratios,



and  initial  angle of discharge.   Salt was  used  as the tracer, yielding



a fluid that  was denser than the  ambient receiving water and facilitated



measuring concentration profiles  of the jet plume.  The coefficient of



entrainment,  the major mechanism  of dilution, was determined as a func-



tion  of the densimetric Froude number,  velocity ratio, and initial



angle of discharge.



      The redacted  drag coefficient  was  chosen as  zero for both models



since any  other value  would predict a trajectory  whose rise would be



less  than  experimentally observed.   For all angles of discharge the



entrainment coefficient increased with  a decrease in the velocity ratio



and with an increase in densimetric Froude number.  Additionally, there




was a marked decrease in the entrainment coefficient with a decrease in



the initial angle of discharge.
                                 11

-------
                        TABLE OF CONTENTS




                                                               Page




ABSTRACT	     ii




LIST OF FIGURES	      v




LIST OF TABLES	     xi




ACKNOWLEDGMENTS	    xii




CHAPTER SECTIONS




   I  CONCLUSIONS  	      1




  II  RECOMMENDATIONS	      3




  III  INTRODUCTION  	      6




  IV  REVIEW OF THE  LITERATURE	     10




   V  ANALYTICAL DEVELOPMENTS OF FAN'S AND ABRAHAM'S MODEL. .     36




  VI  METHODS AND  MATERIAL	     56




  VII  ANALYSIS FOR DATA AND  PRESENTATION OF RESULTS	     69




VIII  SUMMARY AND  CONCLUSIONS	    109




  IX  LIST OF REFERENCES	    121




   X  GLOSSARY - LIST  OF NOTATIONS	    125




  XI  APPENDICES




           A SALINITY-DENSITY  RELATIONSHIP   	   130




           B COMPUTER  PROGRAM  -  FAN'S MODEL  	   131




           C COMPUTER  PROGRAM  -  ABRAHAM'S MODEL  	   136




           D COMPUTER  PROGRAM  -  DRKGS	   141
                                 ill

-------
E  CALIBRATION OF o.5 gpm ROTAMETER	   145

F  CALIBRATION OF 60° V-NOTCH WEIR	   146

G  COMPUTER PROGRAM - ANALYSIS	   147

H  OBSERVED VALUES AND THEORETICAL CURVES PREDICTED
      BY FAN'S AND ABRAHAM'S MODEL	   152
                       Iv

-------
LIST OF FIGURES
Figure
1
2

3
4

5
6
7

8

9

10
11
12
13
14
15
16
17
18
19

A SIMPLE JET (AXISYMMETRIC CASE) 	
PROFILE VIEWS OF SIMPLE PLUMES IN UNIFORM AND STRATIFIED
ENVIRONMENTS 	
PROFILE VIEWS OF BUOYANT JETS [AFTER FAN (3)] 	
SCHEMATIC DIAGRAM OF A PROFILE VIEW OF A ROUND BUOYANT
JET IN A UNIFORM CROSS STREAM OF HONOGENEOUS DENSITY
LENGTH OF ZONE OF ESTABLISHMENT VERSUS 1/k 	
JET WITH NEGATIVE BUOYANCY 	
MAXIMUM HEIGHT OF NEGATIVELY BUOYANT JETS [AFTER
CEDERWALL (5)] 	
SCHEMATIC DIAGRAM FOR THE ANALYSIS OF A ROUND BUOYANT
JET IN A CROSS STREAM 	
SCHEMATIC RELATIONSHIP BETWEEN INITIAL DISCHARGE POINT
AND END OF ZONE OF FLOW ESTABLISHMENT 	
DETAILS OF PROBE CONSTRUCTION 	
PHOTOGRAPH OF CONDUCTIVITY PROBE 	
BASIC MEASURING CIRCUIT [AFTER CLEMENT (35)] 	
SCHEMATIC OF CONDUCTIVITY MONITOR [AFTER CLEMENTS (34)]
CONDUCTIVITY MONITOR 	
ESTERLINE ANGUS RECORDER. . . 	
COMBINATION OF CONDUCTIVITY MONITOR AND RECORDER 	
POLYETHYLENE BARREL WITH FLOW AND TEMPERATURE CONTROL . .
JET TEMPERATURE CONTROL 	
PARTIAL CUTAWAY VIEW OF RECIRCULATING FLUME 	
Page
11

12
13

17
20
26

31

37

51
58
58
59
59a
60
61
61
62
64
66

-------
Figure                                                             Page

  20     CONCENTRATION PROFILE AT s'/DQ = 3.54 FOR RUN NO. 34 ..    72

  21     REPRESENTATIVE PROFILE VIEW OF A JET'S TRAJECTORY. ...    77

  22     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 13	    85

  23     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 10	    86

  24     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 33	    87

  25     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 27	    88

  26     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 22	    89

  27     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 19	    90

  28     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  13	    92

  29     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  10	    93

  30     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  33	    94

  31     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  27	    95

  32     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  22	    96

  33     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S  MODEL - RUN NO.  19	    97

  34     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 13,
            F « 40, k  = 10, 0' = 90°	    98

  35     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 10,
            F = 10,  k = 5,   $'  = 90°	    99

  36     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 33,
            F  z 40,  k = 10,  g»  = 60°.	   100
                                 VI

-------
Figure                                                             Page

  37     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 27,
            F s 10, k * 5, B' = 60° ................   101

  38     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 22,
            F = 40, k « 10, g' = 45° ...............   102

  39     PHOTOGRAPH OF NEGATIVELY BUOYANT JET FOR RUN NO. 19,
            F = 10, k * 5, B1 = 45° ................   103

  40     INVERSE VELOCITY RATIO, 1/k, VERSUS BQ/3^ ........   115

  41     VALUES OF a FOR EXPERIMENTAL COMBINATIONS OF F, k, and
                                                                     117
  42     DENSITY OF A SALT WATER AS A FUNCTION OF SALT
            CONCENTRATION AND TEMPERATURE  (PERRY'S CHEMICAL
            ENGINEERS HANDBOOK, REF. 41) .............   130

  43     CALIBRATION OF  0.5 gpm ROTAMETER .............   145

  44     CALIBRATION OF  60° V-NOTCH WEIR .............   146

  45     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  18  ...............   152

  46     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  13  ...............   153

  47     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  12  .  .  .  .  ...........   154

  48     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  11  ...............   155

  49     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  16  ...............   156

  50     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  10  ...............   157

  51     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  9 ................   158

  52     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  15  ...............   159

  53     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL  - RUN NO.  34  ...............   160
                                    VII

-------
Figure                                                             Page

  54     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 33	161

  55     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 32	162

  56     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 28	163

  57     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 30	164

  58     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 27	165

  59     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 29	166

  60     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 31	167

  61     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 26	  168

  62     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 22	169

  63     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 21	170

  64     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 20	   171

  65     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 24	   172

  66     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 19	   173

  67     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 23	   174

  68     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            FAN'S MODEL - RUN NO. 25	   175

  69     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 18	   176
                                   Vlll

-------
Figure                                                             Page

  70     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 13	  177

  71     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 12	  178

  72     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 11	  179

  73     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 16	  180

  74     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 10	  181

  75     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 9	  182

  76     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 15	  183

  77     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 34	  184

  78     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 33	  185

  79     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 32	  186

  80     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 28	  187

  81     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 30	   188

  82     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 27	   189

  83     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 29	   190

  84     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 31	   191

  85     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 26	   192

  86     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 22	   193


                                    ix

-------
Figure                                                             page

  87     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 21	194

  88     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 20	195

  89     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 24	196

  90     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 19	197

  91     OBSERVED VALUES- AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 23	198

  92     OBSERVED VALUES AND THEORETICAL CURVES PREDICTED BY
            ABRAHAM'S MODEL - RUN NO. 25	199

-------
                             LIST OF TABLES
Table                                                               Page

  1     HEAVY VERTICAL JET EXPERIMENTS IN HOMOGENEOUS AMBIENT
           FLUID	   31

  2     COMPARISON OF JET TEMPERATURE WITH AMBIENT FLUID
           TEMPERATURE	   63

  3     COMBINATION OF DENSIMETRIC FROUDE NUMBER, VELOCITY
           RATIO, AND INITIAL ANGLE OF DISCHARGE ACCORDING TO
           FAN'S DEFINITIONS	   73

  4     COMBINATION OF DENSIMETRIC FROUDE NUMBER, VELOCITY
           RATIO, AND INITIAL ANGLE OF DISCHARGE ACCORDING TO
           ABRAHAM'S DEFINITIONS  	   74

  5     SUMMARY OF NEGATIVELY BUOYANT JET EXPERIMENTS IN A
           CROSS-FLOW FOR FAN'S MODEL	   80

  6     SUMMARY OF NEGATIVELY BUOYANT JET EXPERIMENTS IN A
           CROSS-FLOW FOR ABRAHAM'S MODEL	   82
                                    XI

-------
                         ACKNOWLEDGMENTS






     The successful completion  of this  investigation  owes a  great deal



to a great many - more  individuals  than can be mentioned here.  The



undertaking of the laboratory investigation would have been  virtually



impossible without the  cooperation  and  assistance of  the faculty, staff,



and students  of the Environmental and Water Resources Engineering Depart-



ment at Vanderbilt University.   Special thanks is due to several fellow



students who  not  only made helpful  suggestions for successful completion



of this investigation,  but aided in taking the laboratory data, especially



Ed Yandell,  Bob Reimers, Eung Bai Shin, Greg Waggener, and Aaron Parker.



Thanks is  also due Peggie Bush for  superb typing and  Larry Jones for his



drafting.   Above all, the author wishes to express his appreciation to



his wife,  Patsy,  for  her unwavering support, encouragement,  and financial



assistance.




     The investigations described herein were principally supported by



the National  Center for Research and Training in the  Hydrologic and



Hydraulic Aspects of  Water Pollution Control in the Department of



Environmental and Water Resources Engineering at Vanderbilt  University,



which was funded  by the Environmental Protection Agency, Contract Number



16130 FDQ. The senior author was also supported for one year by an Air



Pollution Traineeship and  for three years by a NASA Traineeship,  Grate-



ful acknowledgment is made for  the  financial support  from these groups.

-------
     The authors are also indebted to Mr. Frank Rainwater, Chief,




National Thermal Pollution Research Program and Dr.  Bruce Tichenor,




of the National Thermal Pollution Research Program for their helpful




discussions during the course of the work.



     The investigation also served as partial fulfillment of the Ph.D.




requirements of the senior author.
                                 Xlll

-------
                         I CONCLUSIONS






     A laboratory investigation of negatively buoyant jets using two




models originally derived for the prediction of characteristics of a




positively buoyant jet has been completed.  The results presented predict




trends rather than exact dilutions and jet trajectories.  The author  feels




that the utilization of Fan's model for the negatively buoyant jet is




theoretically more valid than the use of Abraham's model.  Abraham's




model  can also be used to predict the dilution of jet trajectory.




However, Abraham's model considers the direction of flow of the jet to




be  parallel  to the direction of flow of the ambient fluid at some




distance downstream from the discharge port.  This is not the case for




a negatively buoyant  jet, particularly one whose densimetric Froude




number is small,  i.e.,  one  in which the negative buoyancy term  is large.




In  the case of  a  jet  with a small  densimetric Froude number, the jet




will  deflect downward towards the  discharge  level  after reaching a




maximum height.   However,  Abraham's model is  advantageous  in that the




entrainment coefficients  are constant  and are not  restricted to a fixed




relationship with the densimetric  Froude  number, velocity,  ratio, and




 initial angle  of  discharge.



      Values of a, the entrainment  coefficient,  and C^,  the reduced drag




 coefficient, have been presented.   It was found from fitting  the pre-




dicted curves  to  the  experimental  data that  the best fit occurred when




 a value of  C,  equal  to zero was used.   Moreover,  a relationship which

-------
will predict the entrainment coefficient used in the modified Fan's



model for negatively buoyant jets as a function -^\ densimetric Froude



number, velocity ratio, and initial angle of discharge has been



presented.  Field studies of negatively buoyant jets in a cross-flow



are needed to verify appropriate values of a.  This study has increased



the understanding and  application of the integral theory of jet disper-



sion  to situations  other than  positively buoyant jets in cross-streams.

-------
                      II RECOMMENDATIONS






     The use of Fan's model fox prediction required a priori  knowledge




of both the length of the zone of flow establishment and the  reduced




angle of inclination at the end of the zone of flow establishment.  Much




more information detailing the effects of a cross-stream is needed  so




that one can adequately describe or predict the length of the zone  of




flow establishment as a function of k, the velocity ratio. Heretofore,




investigation on the length of the zone of flow establishment as a




function of k and 3' has been limited to discharges of 90° for submerged




jets.   Motz and Benedict  (25) investigated the effect of the initial




angle  of discharge and  velocity ratio on the length of the zone of flow




establishment for a heated surface jet  and found that the length of flow




establishment was strongly dependent upon the velocity ratio.  This was




the  approach taken by this author.  However, since data was only avail-




able for a discharge of 90°  for submerged jets, there should be further




investigation on the effect  on the length of the  zone of  flow establish-




ment caused by  a discharge angle  other  than 90°.



     The reduced angle  of inclination is  also an  important parameter for




the  utilization of Fan's model.   Scant  data exists  for  the effect of




the  velocity ratio on  the value of the  reduced  angle of inclination.




Present information  indicates  that the  ratio of the reduced  angle of




inclination to  the initial angle  of discharge should decrease with a




decrease in velocity ratio;  however;  the  data is  very scattered as

-------
evidenced by Figure 39.   Hence, no statement can be made regarding the



confidence with which one should use data now available.



     Fan's and Abraham's models predict the dilution and jet trajectory



information for a single port injection system only.  In practice, the



method of injection may vary from a single port system to a multiport



diffuser system.  Hence, additional work is necessary to delineate the



effect of interference as the  jet spreads and is  intersected by an



adjacent jet.



      Larsen and Hecker  (39)  investigated jet interaction for the case of



 submerged  diffusers with a  heated effluent.  They found that the surface



 dilution due to jet interference was  reduced for  the case of a multi-jet



 discharge.   However, for a  negatively buoyant jet, the  case of jet



 interference has not been investigated.  The reduced dilution will



 probably occur and give cause for some concern  since the waste will



 tend to sink back to the discharge level  and subject the benthic



 organisms to higher concentrations of the waste than were predicted by



 a single jet discharge.   Therefore,  further investigations are needed to



 characterize the extent of  dilution for a multiport discharge system.



      Another feature of the negatively buoyant  jet model that should



receive  additional attention is the maximum height of rise of the jet.



Holly and  Grace (40) present data for the maximum height of rise of a



negatively buoyant jet  in a flowing  stream.  However, no information is



available concerning the effects of  the initial angle of discharge



since the studies  conducted  by Holly  and Grace  involved only a discharge

-------
angle of 90°.  The equation developed by Holly and Grace utilized  an




ambient densimetric Froude number which could possibly be related  to  a



velocity ratio.




     The height of rise of the jet in a river or lake may be a limiting




criteria for the use of a particular initial angle of discharge.   As




noted by Holly and grace (40), a primary consideration in designing an




outfall system for dense waste is deciding whether the dense plume will




be allowed to reach the surface, or whether it will be controlled  so




that it remains submerged.  One would like to be able to maximize  the




dilution and still keep the jet submerged.  The jet may have to be dis-




charged at some angle other than 90° to keep it submerged.  Therefore, a




relationship difining the maximum height of rise of a negatively buoy-




ant jet as a function of the velocity ratio, jet densimetric Froude




number, and initial angle of discharge, is needed.

-------
                                CHAPTER I
                                  III
                               INTRODUCTION

     Jet dilution is  one of the techniques that may be used to meet stream
 quality standards in  order to disperse dense waste waters.   However, a
 basic  understanding of this flow  phenomenon of a negatively buoyant jet
 is needed to adequately control this type of pollution problem and meet
 the stream quality standards.  This thesis is expected to enhance the
 general understanding of the flow phenomenon of a negatively buoyant jet.
      Today, as never before, the  nation's attention is focused on ecology
 and the many parameters that affect an ecosystem.  This is true whether
 the ecosystem is found in a pond, lake, or in the very air we breathe.
 This is exemplified by the increase in time, effort, and money being
 utilized to alleviate some of the more pressing problems.  A challenge to
 all individuals was issued by President Johnson in 1968 in his assessment
 of the Nation's water resources under the Water Resources Planning Act of
 1965 when he said, "A nation that fails to plan intelligently for the
 development and protection of its precious water will be condemned to
 wither because of its shortsightedness" (1).
     The Water Resources Council  has reported that the conterminous
 United States  has a natural runoff averaging about 1,200 billion gallons
 per day (bgd).   Withdrawals have  been estimated at 270 bgd in 1965 and
 1,368  bgd in 2020.  The  large withdrawals estimated for 2020 in relation
to runoff indicate  that  even with increased inplant recycling, a large

-------
increase in reuse of water will be required.  Hence, there will be an




ever-increasing need for increased investment in water development,




water conditioning, waste treatment, and water management to meet the




estimated requirements (1).




     The ultimate disposal of man-made wastes is a major environmental




problem of today.  Pollution of the receiving waters has to be controlled



effectively and reduced to such a level as to preserve the delicate




balance of the natural biological processes.  This is true for all kinds




of waste for which an upper limit of pollutant can be defined.  Where a




limit on the concentration of a particular pollutant exists, the most




advantageous method of discharging the waste would be one in which com-




plete-mixing was accomplished instantaneously.  However, the method may




be limited due to the money available, hydraulic configuration of the




place of discharge, or other space restrictions.  Yet a waste cannot




simply be discharged into a receiving stream in a manner in which its




dilution would be inhibited.  Hence, a method of discharging a waste



must be found that lies between these two extremes, i.e., one that will




approach a completely-mixed concentration below the permitted upper




limit of the concentration of the particular pollutant within some




specified or known distance from the discharge point.




     The use of jet diffusion has received much attention in the past as




a means of disposing of domestic sewage in marine areas.  Brooks (2),




Fan and Brooks (3), and Cederwall (4,5) present solutions of a jet




diffusing into an ocean.  The degree of treatment and the choice of




outfall site, as well as the design of the outfall structure, must be




carefully considered so that water quality requirements of the receiving

-------
water can be met.  More and better information will be needed to



adequately design outfall  sites  and  describe  the dilution of waste water




discharges.



      In  the past, much work has  been conducted on  the trajectories and



dilution of positively buoyant jets  in a flowing stream  as exemplified



by the work of Fan  (3), Abraham  O), and Cederwall and Brooks  (7).



Cederwall  (5), Abraham  (8),  and  Turner (9) present solutions for the




maximum height of rise  of a vertical negatively buoyant  jet in  a stag-



nant environment and the concentration at its terminal height.  However,



no work has been conducted on negatively buoyant jets discharged into




 a flowing stream.



      Negatively buoyant jets, or sinking jets, can be observed  in many



problems of pollutant discharge.  Any chemical waste heavier than the



 receiving water may form a negatively buoyant jet.  In addition, the



negatively buoyant jet is found in releases of cold hypolimnic  waters to



the warmer receiving stream,  and the release of a  gaseous waste which



is heavier than the receiving ambient air.  It is  also expected that



this study on negatively buoyant jets will increase general under-



standing about mass and heat transfer across density gradients, a factor



of vital concern in many pollution problems.



      This  study will include  a laboratory investigation  of negatively



buoyant  jets  in a cross-flow. Various combinations  of Velocity ratios,



densimetric Froude  numbers, and  initial angles of  discharge will be



considered.  Models  for positively buoyant jets in a flowing stream



developed by Fan  (3) and Abraham (6) will be  modified and extended to

-------
include the negatively buoyant case.  Values of a,  the entrainment




coefficient, and, if applicable, C,, the drag coefficient,  will  be



determined.

-------
                              CHAPTER II


                                  IV


                       REVIEW OF THE LITERATURE





Jets in General



     In the past, more research attention has been devoted to the



problem of buoyant jets where the discharge fluid is lighter than the



receiving fluid and will  rise.  Fan  (3) presents a comprehensive review



of the research heretofore  accomplished.  His review encompasses the



field of simple jets  (Figures la and Ib), simple plumes  (Figures 2a and



2b), vertical  buoyant jets  and inclined or horizontal buoyant jets



 (Figures 3a and 3b).  The variables found on the figures are defined as:



     O1 = origin  of the coordinate system  (x',y'), point of jet



          discharge



     0  = origin  of the coordinate systems  (x,y), beginning of



          the  zone of established flow  (The zone of flow estab-



          lishment will be  discussed in detail in a later section.)



     0  = initial volume  flux at the nozzle
     xo


     D  = diameter of jet at orifice and orifice diameter



     r  = radial distance measured from the jet axis



     x1 = coordinate axis in horizontal direction on the same



          plane as jet axis with origin at 0'



     y1  = coordinate axis in vertical direction, with origin



          at 0'



     U   =  ambient uniform velocity
     £1
                                 10

-------
VIRTUAL
SOURCE
a. CONCEPT OF  A VIRTUAL  SOURCE
  0'
     ZONE  OF  FLOW
     ESTABLISHMENT
ZONE  OF
ESTABLISHED FLOW
b. ZONE OF FLOW ESTABLISHMENT AND ZONE OF ESTABLISHED
  FLOW
          FIGURE 1 - A SIMPLE JET  (AXISYMMETRIC CASE)
                          .

-------
                VIRTUAL SOURCE
                                             /J=CONST.
a  A SIMPLE  PLUME IN  A UNIFORM  ENVIRONMENT
                VIRTUAL SOURCE
                                  0'
b.  A SIMPLE  PLUME IN A  LINEARLY STRATIFIED
   ENVIRONMENT
      FIGURE     PROFILE VIEWS OF SIMPLE PLUMES IN UNIFORM
                AND STRATIFIED ENVIRONMENTS
                       12

-------
                                              Pa -• f (y)
a. AN INCLINED ROUND BUOYANT JET (FORCED  PLUME) IN A
  STAGNANT ENVIRONMENT  WITH  LINEAR DENSITY
  STRATIFICATION
           U,
                     yOa = constant
b  A VERTICAL ROUND BUOYANT JET IN A UNIFORM HORIZONTAL
   WIND (CROSS STREAM)
          FIGURE 3 - PROFILE VIEWS OF BUOYANT JETS

                     [AFTER FAN (3)]
                         13

-------
     U   =  initial  jet  discharge velocity



      u =  jet velocity at the centerline of the jet



     p   =  density  of the ambient fluid
      3.


     p   =  reference density taken as p (0)
      O                               3.


     PI =  initial  jet  density



       g =  gravitational acceleration



     A simple jet  (or ordinary momentum jet) is the turbulent  flow



 pattern generated by a continuous source of momentum.   Albertson, Dai,



 Jensen, and Rouse (10) have presented a model which describes  the



 behavior of a simple jet, whereas a simple plume is characterized by  a



 turbulent flow pattern generated by a continuous source of buoyancy.



 The more common case that has received most of the attention in the



 past is a steady release of heat.  The plume has no initial momentum



 flux.   Hence, the main direction of flow of the plume  is in the direction



 of the buoyancy force.  Due to continuous action of the buoyancy force,



 the momentum flux of the plume increases with increasing height in the



 case of a heated plume.  A buoyant jet (or forced plume) is characterized



 both by a steady release of mass, "momentum," and buoyancy from a source.



 The source may be situated in either a uniform environment or  stably



 stratified fluid.   Hence, the simple jet and simple plume are  limiting



 cases of buoyant jets.



     Morton (11) and Fan (3) have presented solutions  for buoyant jets.



 Fan presents a  solution for an inclined round buoyant  jet in a stagnant



 environment with linear density stratification and for a round buoyant



jet in  a uniform cross-stream of homogeneous density.   Development of



the round buoyant jet  in  a uniform cross-flow of homogeneous density
                                 14

-------
will be discussed at length in the following pages.




     Parker and Krenkel  (12) present a more recent review of analytical




models of jet studies.   In their report the integral approach of Morton,




Taylor, and Turner  (13)  is reviewed with Morton's  (14) basic assumptions:




     a.  The fluids are  incompressible



     b.  Flow is fully turbulent, implying no Reynolds number depen-




         dence and  that  molecular diffusion is negligible compared to




         turbulent  diffusion.




     c.  Longitudinal diffusion is negligible compared to lateral




         diffusion.



     d.  The  largest variation of fluid density through the flow field




         is small compared with the  reference density.  Hence, varia-




         tions of density can be neglected when considering inertial




         terms, but must be  included in gravity terms.  As Fan  (3)




         notes, this small  density variation  implies  that the conserva-




         tion of mass  flux  can be approximated by the conservation of




         volume flux.  The  assumption of  small density variation is




         commonly called the Boussinesq assumption.




     e.  Velocity profiles  are similar in consecutive transverse




         sections of the jet in the  zone  of established  flow.   Fan  (3)




         also assumed  a  similarity  for profiles of buoyancy and con-




         centration of tracer.  Buoyancy  and  concentration profiles




         are  given  by  the relation:



                                                -r2/b2
                  pa - p*(s,r,*)  = [Pa - PCs)]  e'*  /u                 (D
          and
                                  15

-------
              c*(s.r.*) = c(s) e'r                          (2)



where  p  = density of the ambient fluid
        3


       p* = local density within a jet



        s = distance along the jet axis from the zone of



            establishment



        r = radial distance measured  from the jet axis on A



        A = jet  cross-section normal  to the jet axis



          = angular  coordinate on  a cross-section normal to



            the  jet  axis



         b = local characteristic length of half-width of the



            jet



        c* = local concentration value



         c = concentration  at  the jet  axis



 Figure 4 shows a round buoyant jet discharging at a velocity



 u  into a uniform cross-stream of  velocity U  .  In addition to
  O                                         a


 the above variables, the other variables shown on Figure 4



 are defined below:



         6 = angle of inclination of the jet axis  (with respect



            to the x-axis)



 The x-axis is formed by passing a  plane through the origin, 0,



 in the same direction as the  flow  of  the ambient  stream.



        g1 = initial  angle  of  inclination



        3  = angle of inclination at the end of zone of flow



            establishment



        s' = length of  zone of flow establishment
        6
                        16

-------
                                    WAKE REGION
                            ZONE OF  FLOW ESTABLISHMENT
FIGURE 4 - SCHEMATIC DIAGRAM OF A PROFILE VIEW OF A ROUND BUOYANT JET
         IN A UNIFORM CROSS STREAM OF HOMOGENEOUS DENSITY

-------
          In addition  to  the above  assumption,  Fan made  two  additional



          assumptions  which are also appropriate in this study.  They



          are:



      f.   Within the range of variation, the density of  the  fluid  is



          assumed to be a linear function of either salt concentration



          or heat content above the reference level.  However, Morton



          (13) did consider the density difference expressed as  a  linear




          function of temperature difference and cubical expansion co-



          efficient.  This is a reasonable assumption as can be  seen in



          Appendix A.   It is appropriate in this case since  measurements



          of the conductivity of the fluid were used to  obtain the con-



          centration of salt along the jet axis.  The procedure  will be



          dealt with in a later chapter.



      g.  Curvature of the trajectory of the jet is small.   That is, the



          ratio of the local characteristic length to the radius of cur-



          vature is small.  Hence, the effect of curvature is neglected.



          In the case of the negatively buoyant jet, for small Froude



          numbers, the radius of curvature is small.  Hence,  the ratio



          of the local characteristic length to the radius of curvature



          may be large, but as a first approximation of  the  system, the




          effect of curvature is also neglected.






Zone  of Flow Establishment




      It  can be shown  that a zone of flow establishment  must exist



beyond the efflux section of either a two-dimensional jet or a  three-



dimensional submerged jet (10).   With reference to Figures  (Ib) and  (4),
                                 18

-------
the fluid discharged from the boundary opening may be assumed to be of




relatively constant velocity.  At the efflux section there will neces-




sarily be a pronounced velocity discontinuity between the jet and the




surrounding fluid.  The eddies generated in the region of high shear




will immediately result in a lateral mixing which progresses both in-




ward and outward with distance from the efflux section.  Fluid within




the jet is gradually decelerated; fluid from the surrounding region is




gradually accelerated and entrained into the jet.  The limit of the




zone of flow establishment is reached when the mixing region has




penetrated to the centerline of  the jet.  Albertson, et al.  (10),




reported that the zone of flow establishment for a three-dimensional




submerged jet in a  stagnant non-stratified environment is  X0/D0 = 6.2



where  x  = distance along the axis of the jet  from the efflux  (discharge)
       o


section to point of established  flow.   In the  case of a  jet  in a cross-




flow,  x  will be replaced by s^.



     Fan  (3) reported  values  obtained  from Gbrdier  (15),  Jordinson  (16),




and Keffer  and  Baines  (17)  for the  length of the  zone  of flow  estab-




lishment  for different velocity  ratios.   Fan developed  a plot  of s^/DQ




versus k  for k  values  of 4,  6, and  8 where



          s1  =  the distance  along the jet  axis  from the  discharge point

           e


               to the point  of established flow and




           k =  velocity ratio = U /U
                                 O  d



where



          U  =  jet discharge velocity and




          U  =  ambient  velocity.
           a
                                  19

-------
The equation of the data presented by Fan is
                          e  o

as reported by Parker and Krenkel  (12).  Note that s'/D  approaches
                                                                     (4)
6.2, which is the same  coefficient reported by Albertson, as 1/k

approaches zero.
     Since the  above  equation  was developed from using only 4 points,
it  was  felt necessary to seek  additional  sources of information.  Pratte
and Baines  (18)  have  presented results  as shown in Figure 5.
       10
        8
   0°
   cn
  i      i      i     i
                   0.05
0.10        0.15
      l/k
0.20
0.25
          FIGURE  5  -  LENGTH OF ZONE OF ESTABLISHMENT VERSUS l/k

      These authors felt that the different diameters used in their

studies  affected the length of the potential core (zone of flow

establishment).  However,  it is altogether possible that experimental

error could have just as easily been the villain, since the diameters

of the orifices  used were  all less than 1/2-inch. Hence, a least
                                 20

-------
squares fit of the data was performed, and as a result the equation




s'/D  = 5.91 e  *     was developed.  Hence, this lends credence to the
 e  o                           r



equation developed by Parker and Krenkel (12) and will be used in this



report.





Mechanisms of Entrainment in Turbulent Flow



     The mechanism of entrainment has received much attention over the



last 20 years in determining buoyant  jet behavior.  Tollmien  (19) and



Schmidt  (20) studied the problems of  turbulent non-buoyant and buoyant



jets,  issuing vertically upwards into a homogeneous fluid at rest.



It was noted that the similarity of velocity and concentration profiles



provided sufficient information to  solve for a jet's  trajectory with



either negligible or predominant influence  of buoyancy.  However, as



noted  by Abraham (21),  for jets which are characterized by a varying



influence  of buoyancy effects  within  the field of motion, the entrain-



ment principle  is necessary.



     Morton, Taylor, and Turner  (13)  first  proposed an entrainment



mechanism  for the dilution of  a maintained  plume.  They proposed  the



equation




                            -j^- = 2  irotub                               (5)
                            dx



where  -r*- represents the rate of change of volume  flux and a  is  the




coefficient of  entrainment.  This  equation  states that the rate of



entrainment at  the edge of the plume  is proportional  to some charac-



teristic velocity at that  point.  They noted that when a  stream is  in



contact with another stream the eddies which cause transfer  of  matter
                                  21

-------
between them are characterized by velocities proportional to the



relative velocity of the two streams.  They developed equations for



(a) a maintained plume in a uniform ambient stagnant fluid, (b) a point



source in a stratified fluid, and  (c) a uniformly stratified fluid.



The entrainment coefficient was considered constant at about 0.093.



     Morton  (11) states  that the structure of  the turbulence within a



plume  (jet in the  case of a forced plume  from  a point source of



momentum) and the  rate of entrainment at  its mean edge depends only



on the difference  in mean density  and mean velocity between the plume



axis and the ambient fluid.



      Fan (3) used  this  technique in  the  analysis of a turbulent round



buoyant jet  in a flowing stream.   The entrainment for a jet in a cross



 stream is assumed  to be  represented  by the equation




                          a£-21™*!^--uj                        (6)




where b is  again a characteristic  length defined by the assumed



velocity profile.   The  variables U.  and  U are defined by the following
                                   J       a



equations:



                          U.  = 1 (U  cos 8 + u)                        (7)
                           1      3-



                     IJ = t(U   cos  6) + 7(U   sin 0)                   (8)
                      3.      3.             3-



where  i  is a vector in the direction tangent to the jet axis and j is



a  vector perpendicular to the  jet  axis.   Hence,  JU. - U  | is the mag-
                                                  3    a


nitude of the vector difference in the two velocities.  Fan assumed a



to be constant in  the analysis.



     Abraham  (21)  investigated  the principle of entrainment and discussed
                                 22

-------
its restrictions in solving problems of jets.   It was  maintained that



the entrainment coefficient, as defined above,  was  not constant.



Abraham introduced a new constant, E, which relates the rate  at which



work is done by turbulent shear per unit time  in a  layer with some



thickness, dx, at some level, x, per rate of vertical  flow, Q .



     However, Fan and Brooks (22) in the discussion of Abraham (23),



considered the rate of entrainment to be proportional  to the  local



characteristic (or maximum) velocity and the local  characteristic



radius of the jet or plume.  They noted that the value of a for buoyant



plumes based on data of Rouse, Yih, and Humphreys (24) was 0.082, while



for momentum jets the value was 0.057 based on data by Albertson, Dai,



Jensen, and Rouse (10).  Fan and  Brooks recognized that the entrainment



coefficient could not be a universal constant, but varied as  the



relative buoyancy or local Froude number changed.  However, for  sim-



plicity's sake, it was assumed constant.  Fan  (3) considered  the  en-



trainment coefficient to be constant along the jet trajectory for a



particular set of values of velocity ratio and densimetric Froude number,



but was a variable dependent on each different set of values  of velocity



ratios and densimetric Froude number.  The densimetric Froude number  is



defined as:



                                     U

                             F =      °                               (9)
 /I AP

Vel^i
                                     p   o
where  U   = u  + U  cos B  = initial jet discharge velocity
        o    o    a      o



       U   = ambient uniform velocity
        a



       u   = jet discharge velocity at orifice
                                  23

-------
 also
                8   =  angle of inclination at the end of  the  zone of flow


                     establishment

                g =  gravitational constant
              |£!|  . Jj_L                                           do)
                a       a
                     U    u  + "U  cos g
                 k - — -  °    a
                 K -
                                 a
 For the case under study, i.e., negatively buoyant  jets, it  is advan-

 tageous to use the absolute value of the density difference.  A more

 explicit definition of the above variables is given in Figure 4.

      Fan (5) found that values for a for a round buoyant jet in a cross

 flow varied from 0.4 to 0.5 for a range of velocity ratios from 4 to

 16 and a range of densimetric Froude numbers from 10 to 80.  Benedict

 and Motz (25) reported values of the entrainment coefficient varying

 from 0.13 to 0.46 for heated surface jets discharged into flowing

 ambient streams.  Abraham (6) modified Fan's approach and considered

 the solution of a round buoyant jet with two distinct regions of en-

 trainment.   This modification will be discussed in more detail in

 Chapter III.


     Abraham  suggests  that the jet velocity, at a sufficiently great dis-

 tance downstream from  the nozzle of the jet fluid,  is about  equal to the

velocity of the  ambient  fluid.   Hence,  the entrainment may be described

as if the jet was a cylindrical thermal.  Richards (26) describes a
                                   24

-------
cylindrical thermal as a body of fluid of cylindrical shape with its




horizontal axis moving through a stagnant surrounding fluid due to a




density difference between the surrounding fluid and the particular




body of fluid under consideration.  Hence, Abraham describes the




volumetric flux and momentum flux of a round buoyant jet in terms of




the entrainment of a simple jet and the entrainment of a cylindrical




thermal.  The values of the entrainment coefficients used for the simple




jet and cylindrical thermal were 0.057 and 0.5, respectively.  These




values of the entrainment coefficient were considered constant for all




combinations of densimetric Froude numbers and velocity ratio.




     It can be seen that there are many different values for the en-




trainment coefficient, dependent upon the case under study.  However,




in most cases the value of the entrainment coefficient is unique for




a specific combination of densimetric Froude number and velocity ratio.




This approach is used in this study.






Negatively Buoyant Jets




     The  first attempt to deduce  the path followed by a jet of initial




density different  from its surroundings was reported by Groume-Grjimailo




 (27).  However, this  formula neglected all consideration of viscosity




or  entrainment and was really based on the parabolic path of a pro-




jectile.



      Bosanquet, et al.  (27), studied the effect of density difference




on  the paths of jets.  Equations were developed to predict the trajec-




tories.   However,  since these experimental tests were conducted in  a




transparent box with baffles, they were essentially conducted in a
                                 25

-------
stagnant environment.  Hence, the applicability to a negatively-buoyant

jet in a cross-flow is limited.

     Turner (9) studied jets and plumes with negative or reversing

buoyancy.  In an attempt to explain the oscillation occurring  at the

top of a cumulus cloud due to evaporation, Turner injected  salt water

vertically into a stagnant basin.  The most feasible explanation put

forth was that the evaporation produced a reversal of the buoyant force.

     In  these experiments the salt water was injected upwards.

Initially the pulse  of fluid looked like a buoyant plume with  a vortex-

like front and steady plume behind.  The velocity decreased more rapidly

with height and instead of rising indefinitely with constant shape, the

whole plume broadened, came to rest, and then started to fall  back.

The steady-state position was reached with the top at a lower  height

than that attained initially, with an upflow in the center, and a

downflow surrounding this.  Figure 6 is a sketch showing the relative

shape of a jet with  negative buoyancy.
                                UPFLOW

                       I      \   t  /      I
                  DOWNFLOW   I     /   DOWNFLDW
                FIGURE  6  -  JET WITH NEGATIVE BUOYANCY
                                26

-------
     The three phases distinguished were (a) a 'starting plume1,

advancing and growing by mixing over its top and sides, (b) the cap

stopped rising while it continued to grow;  negative buoyancy is being

accumulated here,  and (c) the cap collapsing.  Turner studied two

different cases,  a heavy jet injected upward and the plumes with re-

versing buoyancy.   Turner's results concerning the heavy jet were of

particular interest to this study.  For a heavy jet injected upwards,

two parameters, M (momentum flux) and F2 (buoyancy flux) were used to

define the flow from a small source (essentially a point source).  The

developed equation took the form

                            yt = CM°-75F-°-5                         (12)

where  C = a constant

      y  = the mean vertical height of rise  of the plume


       M = —2-2-                                                     (13)


              Pi - Pa  ^        D  ^_
                 Q                  4F

Evaluation of the experimental data shows that C = 1.85.   Hence,  sub-

stituting the value of C and Equations 13 and 14 into  Equation  12 and

dividing both sides of Equation  12 by DQf Equation 15  is  developed.


                              =p = 1.74  F                            (15)
                               o

     The main conclusion that can be drawn  from  the  above research  is

that jets in which the buoyancy  force always act downward,  and  which

must be driven upward by momentum at the source, reach a  steady height


                                  27

-------
and  fluctuate randomly about this height.



      Abraham (8)  has studied the problem of jets with  negative buoyancy



in homogeneous  fluids.  He made a distinction between  a zone with posi-



tive entrainment  near the orifice and a zone with negative  entrainment



near the ceiling level.  Abraham developed an expression for the ceiling



 level of a heavy jet injected upward, which is





                               ~ = 1.94 F                           (16)

                                o



 In the above Equations 15 and 16, the densimetric Froude number, F,  is



 calculated by dividing the density difference, Ap, by  the density of



 the  jet fluid, p , which is different from the definition of F in



 Equation 9.  However, the resulting value of F will not be  significantly



 changed if it is divided by either p  or p  since p = p .
                                     a     i        a    l



      Fan (28) used his model to theoretically predict  the dilution  and



 trajectory of waste gas discharges from campus buildings.  The mixing



 of waste gas discharged from a vertical furaehood exhaust in a wind  is



 basically a phenomenon of turbulent jet mixing in a crosswind.  The



waste gas jet bends over toward the downwind direction due  to the action



of ambient wind motion.  In the process, the jet entrains the ambient



air, growing in volume and width.



      Fan also considered a sinking jet in a calm atmosphere.  The



governing dimensionless parameters were velocity ratio and  jet densi-



metric  Froude number.   The cases for a negatively buoyant jet considered



were  for combinations of densimetric Froude numbers of 80,  40, and  20,



and velocity ratios  of 4,  8, 12, 20, 24, and 32.
                                  28

-------
     In a calm atmosphere (U  = 0 or k = °°) ,  a jet rises indefinitely
                            3.



if it has positive  buoyancy or F > 0, but a jet with negative buoyancy



would eventually sink back to the discharge level and spread over the



roof or ground after reaching a maximum height.  For a negatively



buoyant jet,  the initial upward momentum of the jet is gradually reduced



by the constant downward action of the gravitational force.  Using the



integral approach of Turner (9) , Fan made an estimate of the height of



rise of the jet as




                               -- =  1.9 F                          (17)
and the dilution ratio S  at y  = y  as
                        L-          U
                     s ~       F        (for F
                      L-    '  tJ
or
                         S  «0.25 F     (for a  =  0.082)             (19)
                          C-



     Morton (11) investigated the height of rise  of  a negatively



buoyant forced plume issuing vertically  into a  stagnant  environment.



He considered the effect of the discharge  of momentum from a virtual



source of buoyancy and presented a solution in  non-dimensional  parameters,



An equation was developed for the maximum  height  of  rise in terms  of



the non-dimensional parameters.  However,  as noted by Morton, Turner,



and Taylor  (13), a correction must be  made for  a  jet or  forced  plume



emitted from an orifice when a point  source  is  considered.


     Cederwall  (5) states that, in the case  of  a  jet issued vertically



upwards into a homogeneous,  lighter ambient  fluid, the initial  momentum
                                 29

-------
 and buoyancy force are opposing each other.  However,  the basic  con-



 siderations are essentially the same as for a positively buoyant jet,



 as long as the vertical momentum of the jet is sufficient to maintain



 positive entrainment.  Equations 20 and 21 were developed to describe



 the velocity and concentration profile for a negatively buoyant  jet.




                                  D                 -0.33
                    S~  (yr) = 6.2    (1 - 0.22---)                  (20)

                     o            y            DZF2
                                                o
 and
                               Do            v»  "°-33
                         = 5.6 -£  (1 - 0.22 -£— )                    (21)

                     o         y            D2F2
                                             o



      Equation 22 was also developed to predict the maximum height  of



 rise for a negatively buoyant jet.





                              £=2.9F°'67                         (22)

                               O



 However, Cederwall noted that for small values of F,  Equation 22 will



 not hold.  Figure 7 is a comparison of experimental data (Table 1)



 obtained by Cederwall and is compared to the previously developed



 Equations 15 and 16.  Also, Cederwall compared Equation 23, from the



 work of Priestley and Ball (29), in Figure 7 with data obtained at



 Chalmers Institute of Technology (5).





                               ~= 1.85 F                           (23)

                                o



The equations  that Cederwall  compared for the height of rise of a



negatively buoyant jet  in  a stagnant fluid are listed below:
                                 30

-------
    TABLE  1  -  HEAVY  VERTICAL JET EXPERIMENTS IN
             HOMOGENEOUS  AMBIENT FLUID
Run

1
2
3
4
5
6
F

9.3
10.5
20.4
28,8
31.8
20.6
D
o

14
17
30
32
43
26
D
o
Flowing Ambient
Fluid
--
--
--
54
58
38
FT
0

1.50
1.62
1.47
1.11
1.35
1.26
D F
0
Flowing Ambient
Fluid
--
--
--
1.88
1.82
1.84
3
2
LL.
\
cf
:£
~ 1

u(
,iiii
1 _-_ EQUATION 15
\ 	 EQUATION 16
\ __ EQUATION 23
^ O
• ^ 	 °o
0 STAGNANT AMBIENT STREAM
D FLOWING AMBIENT STREAM
, i i 	 1 	
) 10 20 30 40
DENSIMETRIC FROUDE NUMBER, F
^
m~
-
-

FIGURE 7 - MAXIMUM HEIGHT OF NEGATIVELY BUOYANT JETS
                [AFTER CEDERWALL (5)]
                        31

-------
                          1.   yt/DQ * 1.74 F                        (15)





                          2.   yt/D  = 1.94 F                        (16)






                          3.   yt/DQ = 2.9  F°'67                    (22)





                          4.   yt/D  = 1-85 F                        (23)





      Cederwall noted that the equations developed by Abraham  (Equation




 16) and Priestley and Ball (Equation 23) gave better prediction  of  the




 maximum height of rise of a negatively buoyant jet in a flowing  stream.




 However, the velocity of the flowing ambient stream [Cederwall  (5)]  is




 not given, nor is the value of the jet discharge velocity  given.  Hence,




 no statement can be made concerning the effect of the velocity ratio




 or the relative strength of the jet on the terminal height of rise  of




 a negatively buoyant jet in a flowing ambient stream.




      A review of the literature concerning negatively buoyant jets




 indicates a paucity of information concerning the experimental verifi-




 cation of any models which predict the dilution and trajectory of the




 jet.   However, Cederwall (5) noted that the theoretical considerations




 for a negatively buoyant jet are essentially the same as these for  a




 positively buoyant jet.   Cederwall noted that the only difference




 between a negatively buoyant jet and a positively buoyant  jet is that




 the initial  momentum and buoyancy force are opposing each  other.  Hence,




 the models  for a  positively buoyant jet, with an appropriate  modifi-




 cation  to the  buoyancy  force terms, should be able to predict the




dilution  and trajectory  of  a negatively buoyant jet.  Thus, the  inte-




gral approach of Morton,  Taylor,  and Turner (13) applied to a negatively
                                 32

-------
buoyant jet in the same manner as Fan (3) and Abraham (6) treated a



positively buoyant jet should yield a satisfactory and usable means of



evaluating the characteristics of a negatively buoyant jet.  These two



models are developed in more detail in Chapter III.
                                  33

-------
1.  yt/D  = 1.74 F                         (15)
2
                             .  y /D  =  1.94 F                         (16)





                            3.  yt/Do =  2.9 F0-67                      (22)




                            4.  yt/D  =  1.85 F                         (23)




     Cederwall noted that the  equations developed by Abraham (Equation



 16)  and Priestley and Ball (Equation 23) gave  better prediction of the



 maximum height  of rise of a negatively  buoyant jet  in a flowing stream.



 However, the velocity of the flowing ambient stream [Cederwall (5)] is



 not given, nor is the value of the jet  discharge velocity given.  Hence,



 no statement can be made concerning the effect of the velocity ratio or



 the relative strength of the jet on the terminal height of rise of a



 negatively buoyant jet in a flowing ambient  stream.



      Briggs (42) presents an excellent  review  of existing plume rise



 observations and formulas.  Nine formulas are  reviewed and compared with



 data from sixteen different sources.   Briggs chose  Equation 94 as the



 best predictive equation.  However, Equation 94 must be modified by



 assuming that a ceiling height is reached at a distance of ten stack



 heights downwind.  Other equations are  presented which are dependent upon



 different stability conditions.




                          Ah =  1.6 F U"1(X1)'667                       (94)
                                      cL



 in which Ah = plume rise above top of stack.



     A  review of the literature  concerning negatively buoyant  jets



indicates  a paucity of information concerning  the  experimental verifi-



cation of  any models  which predict the  dilution and trajectory of the
       34

-------
jet.  However,  Cederwall (5)  noted that the theoretical considerations




for a negatively buoyant jet  are essentially the same as these for a




positively buoyant jet.   Cederwall noted that the only difference




between a negatively buoyant  jet and a positively buoyant jet is that




the initial momentum and buoyancy force are opposing each other.  Hence,




the models for a positively buoyant jet, with an appropriate modification




to the buoyancy force terms,  should be able to predict the dilution and




trajectory of a negatively buoyant jet.  Thus, the integral approach of




Morton, Taylor, and Turner (13) applied to a negatively buoyant jet in




the same manner as Fan (3) and Abraham (6) treated a positively buoyant




jet should yield a satisfactory and usable means of evaluating the




characteristics of a negatively buoyant jet.  These two models are




developed in more detail in Chapter III.

-------
                                CHAPTER III


                                    V


            ANALYTICAL DEVELOPMENT OF FAN'S AND ABRAHAM'S MODEL





      In this chapter both Fan's and Abraham's models will be reviewed



 and normalized so that they can be used to describe negatively buoyant



 jets.





      Fan's Model for a Round Buoyant Jet in a Uniform Cross  Stream





      Fan  (3) applies the integral approach of Morton, Taylor,  and



 Turner  (13) to solve the problem of a round buoyant jet in a cross



 stream.   Figure 8 shows a round buoyant jet discharging at a velocity



 u  into a uniform cross stream of velocity U .   The densities  of the
  o                                       'a


 discharged fluid and the ambient fluid are p  and p , respectively.
                                             1      3-


 The flow  becomes fully developed at a short distance s' from the nozzle,
                                                       C


 O1 is taken at the beginning of the zone of flow establishment, and 0



 is taken  at the beginning of the zone of established flow.  9  is the



 angle of  the inclination of the trajectory with respect to the horizon-



 tal x-axis.



      For the case of a positively buoyant jet,  there are two reasons



 for the  deflection of the jet toward the downstream direction - the low



 pressure region established behind the jet and the entrainment of



 ambient horizontal momentum as the jet entrains the fluid of the cross



stream.  Yet, in the  case  of the negatively buoyant jet, the jet will



reach a maximum height and  then bend over toward the region  of y = 0
                                  36

-------
'
i
AMBIENT

   VELOCITY
                                                                          DRAG FORCE
                FIGURE 8 - SCHEMATIC DIAGRAM FOR THE ANALYSIS OF A ROUND BUOYANT JET

                                     IN A CROSS STREAM

-------
or towards the discharging  level or  lower.   The  jet will rise due to its



vertical momentum flux, but at  the same  time the negatively acting



buoyancy flux will tend to  reduce the  jet's  initial momentum in such a



manner that it bends over and returns  towards the discharge level.



     Fan found by dimensional analysis that  the  flow of a positive



buoyant jet was  characterized by the densimetric Froude number (Equation



9) and by  a velocity ratio  which represents  the  relative strength of the



jet  into a cross flow  as defined by  Equation 3.





Basic Assumptions



     The assumptions adopted by Fan  are  outlined below.



     1.  Velocity profiles  are  assumed to be similar and Gaussian above



         the  component of the ambient  velocity U cos 6




                  u*(s,r,) = U cos 9 + u(s) e"r /b                (24)
                                e*



         where u* = velocity at a local  point and b = local



         characteristic length.



     2.  The entrainment relation for  a  jet  in a cross stream is



         assumed to be represented by  the equation





                            - 2  TT ab |U. - Uj                      (6)




         where Q is the volumetric flux.



     3.   Buoyancy profiles  are  assumed to be Gaussian




                 p   - p*(s,r,<>) = [p   -  p(s)] e'
                  <&                  »



        where p* =  local density within a jet,



    4.  Concentration  profiles of a certain tracer are assumed to be
                                 38

-------
         Gaussian


                                          T2/h2
                       c*(r,s,4) = c(s) e"r /b



         where c* = local concentration value.



     5.   The effect of the presence of the pressure field can be lumped



         into a gross drag term proportional to the square of the veloc



         ity component of the ambient stream normal to the jet axis.



         The drag coefficient is assumed constant.




Peyel_opme_nt__of_ Equations



     Fan makes use of the equation of conservation of mass, the equa-


tion of conservation of momentum, and geometric equations to describe



the round buoyant jet in a cross stream.  The equations developed are



outlined below.



     1.   Conservation  equations - Equation 6 is integrated to attain



         an expression for the continuity of fluid.
                          ~^f  ~~ ^11 IAL/ 1 U • ~ U  1
                          ds        ' j    a1



         where Q is the volumetric flux.




                             Q = I   u* dA                           (25)
         where A is jet cross section normal to the jet axis.


     Fan defines the boundary of this jet as /2b.  Then, substituting


Equation 24 for u* into Equation 25 and integrating between the limits


of r = 0 and r = -/2b, Equation 25 becomes
                       r                   -r2/b2
                  Q =    2irr(U  cos 6 + ue      ) dr

                      J,.      a
                                 39

-------
After  the substitution of the limits
Q =
                                            -r2/b2
                                  cos 9 + ue   '  ) dr
 which can be approximated as
              Q   2TT
         r U   cos  9 dr +    ure        dr
                                                 -r2/b2
 After integration,
                         Q = irb2(2U  cos 6 + u)
                                   ol
 Hence,
                 §= |- E"b2(2U   cos  6 + u)] = 2irab|U. - "u,
                   QS        3.                     J    c
                                                    (26)
      Thus,  |U. - U  | is the vector  difference of the jet velocity  and
              J    a
 the ambient velocity.
                          U. =  i(U   cos  8 + u)
                           J      a
 and
                      U  = i U   cos  6  + j U   sin  6
                       a      a         J  a
 Therefore,
        JU.  - U |  = J(U  cos 9 + u  - U   cos  9)2  +  (0  - U  sin  6)2
               a    V  a              a                 a
Hence,
                    |U  - U" | = V (u2 + U  2  sin2  6)'
                      J    a,             a
Dividing Equation 26 by TT,
            ^-[b2(2U  cos 6 + u)] = 2ab J(U  2  sin2 6 +  u2)           (27)
            OS       di                   V  cL
                                 40

-------
     2.   The  momentum equations can be written by assuming a gross drag



         coefficient, C,,   In the x-direction, the rate of change of



         momentum flux is  equal to the rate of entrainment of ambient



         momentum flux plus the drag force acting on the jet.





        4- f   p*u*(u* cos  8)dA = 2™b p  U |U. - if  | + F  sin 6     (28)
        dsj.                          a  a  j    a     u




where Fn is the drag force per unit length assumed to be



                             p U2 sin2 6

                     FD *  Cd -^-~	 2 V^b                       (29)




     The left side of Equation 28, after substitution of Equation 24



and making use of Boussinesq's assumption that p*~g p&, becomes
                                         r2/b2  2
            §-  1    p  2irr(U  cos 6 + ue   '  )  cos  9  dr
            ds JQ    a      a




After integration,



                                                      ~l
                    p „ *  r b2     cos e + u)2 cos e               C30)
                    Ma  ds L 2    a                  -I


Substituting Equation 29 and 30 into Equation 28 and dividing by pair,


the x-momentum equation becomes
       _         cos
      ds L 2    a
                      0  +  u)2  cos  9 1=  2cxb U  J(U2 sin2 6
                                   J        a. t   A
                                         - U2b sin 3 6              (31)
                                         •TT    a
      3.   In the  y-direction,  the rate of change of the momentum flux



          is equal  to the gravity force acting on the jet cross-section



          minus the y-component of the drag force.  For a negatively



          buoyant jet, the buoyant force will be acting in a manner to
                                  41

-------
         reduce momentum in the y-direction.   Hence,  a negative sign

         correction to the buoyancy term will be made on Fan's equation

         for y-momentum so that the absolute value  of the Froude number

         can be used later.
 -
ds  A
   p*u*(u* sin e)dA =  -
'                       '
                       sin e)dA = -    g(p  - p*)dA - Fn cos 9       (32)
      The left-hand side of the equation is treated in the same manner

 as the left-hand  side of Equation 28, as shown by Equation 30.  The

 first term of the right-hand side of the equation is treated as below

 substituting Equation 1.
                  g(p   -  P*)dA = -  I  g(p  - p)e"r
                A   a
                  g(P   - P*)dA
                  -I
                  g(p  - p*)dA = - irb2g(p  - p)
               A                        a
Therefore,
                       COS
                   u)2 sin  6J = -nl

                                   p U2 sin2 6
                                                                    (33)
Dividing Equation 33 by p TT yields, for the y-momentum equation,
                         ct
                                 1      .  (pa "
                cos 9 + u)2 sin el  = -b2g •  a
                                              pa
                                      Cd-/2'
                                      - bU2 sin2 6 cos  6         (34)
                                42

-------
4.  The density deficiency flux induced is conserved since the



    ambient fluid is homogeneous, so





                   IF f  "*(Pa - P*)dA = 0                     (35)

                      •'A



    Substituting Equations 1 and 24 into Equation 35 yields
    d                       -r2/b2           -r2/h2
    ~     2Trr(U  cos 6 + ue r /D ) (p  - p)e r /D dr = 0
    ds          a                     a
d  ' b-U  cos Of        + — uf
        a cos  (.Pa   pj + ^  u^Pa ~ PJ
                                                   = 0
                        cos 6 •»- u) (p  - p)   =0              (36)
                                     d



    Therefore, Equation 36 yields




             4- [b2(2U  cos 6 + u) (p  - p)] = 0               (37)
             Uo       A              a.



5.  The flux of any specific tracer of concentration, c*, contained



    in the jet flow will be conserved in a fashion similar to the



    density deficiency, as shown below.





                      4-f  u* c* dA = 0                       (38)
                      ds J




    Substituting Equations 2 and 24 into Equation 38 and inte-



    grating yields




                 4~ [b2(2U  cos 6 + u)c] = 0                   (39)
                 dS       cl



6.  There exist two geometric relationships due to the jet



    trajectory which are utilized.




                        ^- x = cos 9                           (40)
                        ds



    and
                            43

-------
                                - y = sin 6                           (41)

         Hence, there are 7 unknowns - u, p  - p, c, b, 6, x, and y  -
                                           3.
         that can be solved for with seven simultaneous ordinary differ-
         ential equations - Equation Numbers (27),  (31), (34), (37),
          (39),  (40), and  (41).

 Initial  Condition  for Fan's Model
      Fan considers his model  applicable after the zone of flow estab-
 lishment.   The  initial conditions at the origin, 0, at the end of the
 zone  of flow establishment are given below.
                            u(0) = UQ

                            b(0) = bo

                            P(0) = PI

                            0(0) = eo = po                           (42)

                            c(0) = CQ

                            x(0) = 0
                            y(0) = 0
                            at s = 0
where  u  = initial jet discharge velocity at orifice

       b  = jet half-width at the end of the zone of flow establishment
            as defined in Equation 24
       Pt = jet density
       c   = initial concentration at the end of the zone of flow estab-
        o
            lishment
                                 44

-------
      x = horizontal coordinate measured  from the  end  of  the  zone of



          flow establishment



      y = vertical coordinate measured  from the  end  of the  zone  of  flow



          establishment



      s = parametric distance along  jet axis measured  from  the end  of



          the  zone of  flow  establishment



     By considering the conservation  of  the flux of density  deficiency



between the  two cross-sections at  point  0 and 0',
                   cos 6   * u)
                             0
                                 '2U  cos 6  + u
                                    a      o    o




            2U  cos 6  + u

where  k'  = —^	~	 = k + cos GO                             (43)

                    a
     . bo • Do -



     In addition, the initial angle, QQ, is the reduced angle of in-



clination.  It will always be less than the initial angle of discharge



due to deflection in the zone of flow establishment.  Fan developed a



functional relationship involving k to express the value of 9Q.  However,



for the negatively buoyant jet, no functional relationship exists.  So,



in each case, the jet trajectory is plotted, the  zone of establishment



is measured along the axis using values calculated from Equation 4, and



then  the angle at the end of the zone of  flow establishment is measured.



Hence, when Fan  uses 9  in his  derivation,  it is  equivalent to  3Q  in
                                  45

-------
 this report.  In addition, £' is the initial angle of inclination at the



 orifice.





 Dimensionless Parameters
      Fan's equations are normalized by defining dimensionless parameters



 as follows, using initial values.



 Volumetric flux parameter,



                              b2(2U  cos 9 + u)

                                                                     (45)
                              b2(2U  cos 8  + u )
                               o^  a      o    o'
 momentum flux parameter,
                    b2(2U  cos 6 + u)2

                m = - - -    (s-direction)             (46)


                    bo(2Ua COS 8o * V2



                h = m cos 6                 (x-direction)             (47)



                v = m sin 6                 (y-direction)             (48)



 velocity ratio,



                                u  + U  cos 8    U
                            v _
                            k
                                       a          a
 buoyancy parameter,
                        f = b0 g/^k.zuz)                  (49)

                                    a
                           k1 = k + cos 6                            (43)
                                         o                           ^   J
                                                                     (50)
                            i_ — —   .»      _i



where F = densimetric  Froude number,



coordinates,
                                 46

-------
                                      2c*
                              x  : n = r—  x
                                      b
                                       o

                                                                     (51)


                                    = —

                                       o



where a = coefficient of entrainment.





Normalization of Fan's Equations



     Fan's equations are normalized using the  dimensionless parameters



on pages 39 and 40, resulting in Equations 52-56,  60,  and 61.   Fan's



equations 27, 31, 34, 37, 39, 40 and 41:









        dri    1  u               Q


             k'2 v m



                      C1

     3. -77 = - f — - 	  —  sin2 6  cos 6                         (54)
     4. dJI= cos 9                                                   (55)
     5.  i= sin 9                                                   (56)
        d^



          r                             -i-5
where ^ = I  sin2 6 + (k' 2. . 2 cos  6)  2 J                            (57)




and  C' = C , -/2/a TT                                                  (58)
      d    a



     6.  The equations of  continuity  of  a  tracer,  38, and density



deficiency, 37, may be solved immediately.   The continuity of a tracer



is expressed as



                      b2(2U  cos  e  +  u)c = constant                 (59)
                           3.



Therefore,
                                 47

-------
                       r    b2(2U  cos 6  + u )    . .
                       c_ =  pi  a	o	oi_  = I                (6Q)


                        o     b2(2U  cos 9 + u)     y
                                   a



      7.  The conservation of density deficiency    .ntegrated as




                   b2(2U  cos 6 + u) (p  - p) = constant              (61)
                        a             a



      Hence, values of both c and (p  - p) can be determined easily from
                                     cl


 the solutions of the terms b2(2U  cos 6 + u).  This leaves  only five
                                 a


 unknowns contained in five simultaneous ordinary differential equations.



      The initial conditions for £ = 0 are



                              P(0) = 1



                              m(0) = 1
                                           o                         (62)



                              n(0) = 0



                                   = 0
              Abraham's Model for Round Buoyant Jets Issuing

                     Vertically into a Flowing Stream
      Abraham (6) has modified Fan's model by using entrainment  coeffi-



 cients that are constant, i.e., they are not functions of either F,  k,  or



 8'.  For Fan's model the value of the entrainment coefficient and the
  o


 drag coefficient is a function of the velocity ratio and densimetric



 Froude number.   Abraham recognizes that this may preclude the use of



 Fan's  model beyond the range of conditions covered by the experiments.





 Basic  Assumptions



     Abraham considers  two regions of the jet.   In the vicinity of the



nozzle the  trajectory of a round  buoyant jet in a cross flow may be
                                 48

-------
described as a vertical line, provided that the cross flow is weak and



that the initial momentum of the jet is strong.  The region near the



nozzle may be described as a jet primarily influenced by its initial



momentum, and thus the volumetric flux may be expressed as




                                - = a    2-nbu                        (63)
                                '    mom                                J



where a    = coefficient of entrainment for jets primarily influenced
       mom                                  j    r


             by initial momentum



      a    = 0.057 according to Albertson, et al.  (10)
       mom                 &              '



The variables u and b are the same as described earlier.



     The second region is a great distance downstream from the nozzle



where the velocity of the jet fluid is about equal to the velocity of



the ambient fluid.  Here, entrainment may be described by the relation-



ship for a cylindrical thermal  in a stagnant fluid as described by



Richards (26).  The entrainment satisfies the relationship,



                             dV

                             -5-9-= a ,  27T B_                        (64)
                             dy'    th     th



where  y' = vertical coordinate indicating position  of center of



            thermal jet



       V  = quantity of fluid contained in a thermal jet per unit of



            length



      B , = radius of thermal defined by  Equation  66
       th                                                            (65)



      a   = coefficient of entrainment for thermals  moving  through
       th


            stagnant ambient fluid  [0.5 according  to Richards  (26)]



     The entrainment of a cylindrical  thermal  may  be described  by



Equation 64 when  the concentration  of  a tracer carried by the thermal
                                 49

-------
 is expressed as





                    c*(s',r,«fr) = c(s') e     "'                       (66)



 This expression is equivalent to the one used by Fan shown in Equation



 2.  Abraham assumes the same velocity profile as described by Fan in



 Equation 24,



                                                  _T2/b2

                   u*(s',r,<|>) = U  cos 0 + u(s') e   '                (24)
                                 £L



      Abraham combines Equations 63 and 64 and describes the volumetric



 flux as


                   r

                     u*dA = 2irb(a    u + a.,  U  sin 0 cos 6)          (67)
              *—,   i  u \*rv ~~ f» ii u i \*    fc*  • **-_i_ **
              ds'   I.             mom      th  a
                   J A
                    r\



 where dy1 has been replaced by sin 6 ds1.



      The cosine function in Equation 67 was arbitrarily introduced in



 the term involving a ,  to avoid this term contributing to entrainment  in



 the initial  region near the orifice where the jet is primarily a momentum



 jet.  In addition, tMe prime notation  is used to indicate the coordinates



 of the jet from the discharge point.   Hence, comparing Fan's notation



 with Abraham's notation, the relationship for the coordinates is as



 follows:



                                s' = s1 + s
                                      e



                                x' = x1 + x                           (68)




                                y1 = y1 + y





where s', x1, and  y1  are the distances from the discharge point to the



end of the zone of flow establishment  [Figure (9)].
                                  50

-------
                                                  e
                jO I^VNN WW>.NNXXN"\\.\.X\xVvV\V
  FIGURE 9 - SCHEMATIC RELATIONSHIP BETWEEN  INITIAL DISCHARGE  POINT AND
                    END OF ZONE OF FLOW ESTABLISHMENT


Development of Equations


     The following equations were developed  by Abraham using Equation


67 to describe the volumetric and momentum flux equations .  Where the


equations are the same as developed by Fan,  they will receive  the same


number.   The equations are outlined below.


     1.   Continuity,
[b2(2Ua cos 6 + u)] =
  ds

2.   x-momentum,


    d   1 b2
                                    u
                                                 U&  sin  6  cos  6)     (69)
         ds
r MjL (2U  cos 6 + u)2 cos 8
  L ^     3-
                                               s\
         2bU (a    u + a ,  U  sin 9 cos 6) + —— U2 b  sin3  6        (70)
            a  mom      th  a                   n   a


     3.  y-momentum - The equation for y-momentum is essentially  the


         same as developed by Fan as shown by  Equation  34  with
                                 51

-------
         modification for the buoyancy force in the negative y-direction.
r- (2Ua cos e + u)2 sin e
                       P   -  P   C,j2
                                                 j =
                                  ,

                   -b2g -2 --- _ u2 b sin2 6 cos 6                (34)

                         pa       *   a
     4.   Continuity  of tracer,
                          [b2(2U  cos  9 + u)c] = 0                     (39)
                       ds        a
      5.   Density deficiency,
                    ^-r[b2(2U  cos  0 + u)(pa - P)] = 0                 (37)
                    ds       a            a



      6.   Geometric relationships,



                              3-T x1 = cos 8                           (40)
                              ds1



                              3JT /' = sin 9                           (41)



      It  is important to note that  the values of x1, y', and s? for



 Abraham's equation are measured from the orifice rather than the end  of



 the  zone of flow establishment and are defined as



           x1  = horizontal coordinate, measured from orifice



           y1  » vertical coordinate, measured from orifice



           s1  = distance along jet  axis, measured from orifice





 Initial  Conditions



     There are seven unknowns for  Abraham's model  (u, b, c, p  -p, 9,
                                                             3.


x1, and  yr) with  seven equations  (69, 70,  34,  39,  37, 40, and 41).  The



initial  conditions  are:



                              u(0)  = UQ




                              c(0)  = c
                                 52

-------
                             P(0)  = PI

                             b(0)  = DQ/2

                            x'(0)  = 0
                                                                    (71)
                            y'(0)  = 0

                             6(0)  = eo  = $•

                          at   s'  = 0

where c  = initial concentration,
       o

      u  = initial jet discharge velocity,

      B' = initial angle of inclination at discharge point, and

      D  = diameter of orifice.
       o
Dimensionless Parameters

     Abraham's equations are normalized in the same manner as Fan's

equations with the following exceptions:

     Buoyancy parameter:


                                           2U)                    (72)
                        f = (k/k')2/2F2                             (73)

     coordinates :

                            <; • r '  = — s'
                            S . (,    b
                                      O


                            x : n'  = ^-x'                          C74)
                            y  : V =f-y'
                                      o
                                 53

-------
Normalization of Abraham's Equation



     Abraham's Equations 69, 70, 34, 40, and 41 are normalized using the



dimensionless parameters listed on page 46, resulting in Equations 75-



79.




1.  ~- = a    Jm -- V— (2 a    - or,  sin 6)cos 9                  (75)
    d?'    mom      , , /—     mom    th
                    K'vm
  -  :rV = — — |2o     (— - 2 cos 0) + 2a .  sin e cos 9

      C          '2 ^
                — -   co        a .

VnT k'2    mom   y


                         3 e]                             (76)
                                sn
 3.  jV =  -f - - C,  —	^ sin2 6 cos e                           (77)






 4.  -p-f n1  = cos 9                                                  (78)





 5.  -4r V  = sin 9                                                  (79)
where C,  = -	                                                 (80)
       da      TT



     The density deficiency and continuity of tracer are solved in the



same manner as Fan's equation for density deficiency and continuity of



tracer.



     The initial conditions for ?' = 0 are:



                               v(0) = 1



                               m(0) = 1



                               6(0) - g^                            (81)



                              n'(0) = 0



                              C'(0) = 0
                                54

-------
                          Solution of Equations






     A system of simultaneous differential equations can be solved using




a Fourth-Order Runge Kutta technique.  Such a program is available at




the Vanderbilt University Computer Center in the Scientific Subroutine




library.   The listing of the computer program used to solve Fan's model




is included in Appendix B, and the computer program used to solve




Abraham's model is included in Appendix C.  The subroutine that contains




the Fourth-Order Runge Kutta solution to solve a system of simultaneous




ordinary differential equations is the same for both Fan's model and




Abraham's model and is included in Appendix D.  However, the main




programs, which contain the individual equations for the two different




models, are different.
                                  55

-------
                                CHAPTER IV
                                    VI
                           METHODS AND MATERIAL
 Introduction

      The objective of this research was to measure the vertical profile

 of a negatively buoyant jet at points downstream from the jet  discharge

 point in order to obtain dilution data, jet half-width data, and  jet

 trajectory data for a negatively buoyant jet.   Salt  solutions  of  dif-

 ferent concentrations were used to model the jet for two reasons.

 First, a salt solution is heavier than water and can easily be used to

 model a negatively buoyant jet.  Secondly, the vertical profiles  of the

 jet could be measured very easily by using a conductivity probe with

 appropriate monitoring devices.



 Conductivity Probe

      The conductivity probe was similar to the one used by Krenkel  (30)

 and Burdick (31) with a few modifications.  The electrode consisted of

 a 16-gage blunt stainless steel hypodermic needle as the outer electrode

 and a length of 0.01 in. diameter platinum wire as the inner electrode.

 The platinum wire was insulated from the hypodermic  needle by  a small

 capillary of plastic tubing into which it was threaded.  Both  ends of

 the  electrode  were sealed with a non-conducting epoxy cement.  The

bottom portion of the needle and the exposed portion of the platinum

were then plated  with platinum black to provide a large surface area.
                                  56

-------
However,  the hypodermic needle was first plated with copper as shown by




Krenkel (32) so that the platinum black would adhere more strongly.




Copper was plated on the stainless steel by using a length of cleaned




copper wire as one electrode and the needle as the other electrode.  A




5-volt power supply with variable output was used as the driving force




in the plating operation.  A voltage output of 1.5 volts was used.  This




circuit was completed by using 1 N copper sulfate as the electrolyte.




The copper wire was helically wound to increase the surface area in the




copper sulfate solution.  After the copper plating had been thoroughly




washed, the needle was placed in a 100-milliliter solution containing




3 grains of platinic chloride and 0.02 grams of lead acetate.  The needle




leads were then connected to about three volts direct current as shown




by Clements  (33).  The direction of current was reversed every  15




seconds until a uniform  coating was deposited on both electrodes.




(Two or three minutes is suggested.)  The electrodes were then  rinsed




in running tap water for half an hour and stored in distilled water




when not  in  use.   It was necessary to keep the probe wet at all times




in order  to  keep the platinum black coating in proper condition.




     The  probes vere finally covered with a non-conducting rubber  coating




to insure insulation and to give  the soldered  leads more strength.




Detailed  views of  the probe are  shown  in Figure  10  and  11.








Conductivity Monitor and Recorder



     The  conductivity monitor was designed  as  described by Clements




 (31).  The  basic premise of construction involved using a  series




resistance  arrangement  with an  a.c.  source  having a low internal
                                  57

-------
                                     i.
                                            I1
                                            16
PLASTIC
EPOXY RESIN



 PLATINUM
                 RUBBER  COATING

      DETAILS OF  PLATINIZED  TIP
  FIGURE 10 - DETAILS OF PROBE CONSTRUCTION
  FIGURE 11 - PHOTOGRAPH OF CONDUCTIVITY PROBE
                     58

-------
impedance as shown in Figure 12.   This type of circuit provided an  output
                                  CELL
                                          EC
                   FIGURE 12 - BASIC MEASURING CIRCUIT
                          [AFTER CLEMENTS  (35)]

voltage, E , which was nearly linearly proportional to the conductivity

G, of the test solution, where E. was a constant supply voltage of 5

volts and G was the reciprocal of probe resistance, R .  The linearity

may be seen by considering Equation 82.

                                     E.R
 If GR " .01, E  is directly proportional to G.  This type of monitor

 was chosen for the following characteristics.

     1.  Low cost - approximately $25.00 for parts.

     2.  Drift was low - about 0.12 percent per hour with normal

         ventilation.

     3.  Calibration was linear over a large range.

     4.  Polarization effects at the conductivity cell electrodes were

         negligible due to the 1 kHz excitation frequency.
                                  59

-------
01
vo
             IIOv.
             LINE
    47
.01  k

.01
                          NE-51
                                          100 ma.
                                          800 p.Lv.   470
                                      22            Iw.
                                  ——s/VA/^	J

                                  £
                                   250 v.
 20
350v.
,50k
 10 w.
                 S
                 r
                                          eASII FIL.
                             STANCOR
                              PS-8416
                            .001
                                        7hy.
                           36k   36 k    50 ma.
                           2w.    2w.
                                                        RESISTANCES  IN OHMS
                                                              k=IOOO

                                                   CAPACITANCES  IN  MICROFARADS
                                                                            47k <  = = .5
                                            D.C. OUTPUT

                                                  I


                                        INI9I   A.C. OUTPUT
                        .001     .001
                                                                                                COND. CELL
                  68 kS    68k>   68 k,
                              FIGURE  13 - SCHEMATIC OF CONDUCTIVITY MONITOR [AFTER CLEMENTS  (34)]

-------
     5.  Response was fast, with  an average  time  constant of 0.025




         second.




     The combination of  1-kHz  excitation frequency and the large surface




area exposed by the platinum black made the  polarization effects at the




conductivity probe neglible.   Figure 13 is a diagrammatic sketch of the




monitor and Figure 14 is a photograph of the conductivity monitor.
                     FIGURE  14  -  CONDUCTIVITY MONITOR




     The recorder chosen was an  Esterline-Angus Model T171B, Port-A-




Graph recorder.  The input  resistance  for  the recorder was 1.8 to 2




mega ohms.  The wide range  of  2,5,10,20,500,200 MV, 0.5, 1,2,5,10,20,50




V afforded the needed flexibility for  measuring various outputs with the




needed sensitivity.   Figures 15  and 16 show the Esterline-Angus Recorder




and the combination of conductivity monitor and recorder, respectively.
                                 60

-------
           FIGURE  15  - ESTERLINE-ANGUS RECORDER
FIGURE 16 - COMBINATION OF CONDUCTIVITY MONITOR AND RECORDER
                           61

-------
Injection and Flow Control

     A commercial grade salt was used as the trac   material.  The salt

was well-mixed using a 1500-gpm stirring apparatus in a 50-gal poly-

ethylene barrel.

     The barrel was equipped with  internal helically wound copper coils

to keep the tracer material at approximately the same temperature as the

ambient fluid  (Figure 17).  This was accomplished by passing water from

the same source as the ambient fluid through approximately 60 feet of

copper tubing.  This approach worked very well as may be seen in Table

2.
               FIGURE 17 - POLYETHYLENE BARREL WITH FLOW AND
                            TEMPERATURE CONTROL
                                  62

-------
TABLE 2 - COMPARISON OF JET TEMPERATURE WITH
          AMBIENT FLUID TEMPERATURE
Run No.
9
10
11
12
13
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Temperature
Ambient (°F)
64.6
66.3
66.3
66.8
67.6
68.7
68.5
70.5
73.3
73.3
73.2
73.2
75.5
76.0
76.0
72.2
76.2
76.2
77.6
77.6
77.6
76.1
76.1
75.5
Temperature
Jet (°F)
65.4
67.5
67.5
67.6
68.5
69.0
69.1
71.0
73.7
73.7
73.4
74.4
75.6
76.0
76.0
72.2
76.4
76.4
77.8
77.8
77.2
75.9
76.6
75.9
                   63

-------
     A student's t-test for the comparison of the means shows that there




is no significant difference in the two populatic--•-.  Hence, the conclu-




sion can be drawn that the data obtained came from the same population.




Figure 18 shows how temperature control was obtained.
                    FIGURE 18 - JET TEMPERATURE CONTROL




      The salt solution was then pumped through a PVC piping  system by  a




 12-gpm pump through a 0.5-gpm Fisher-Price rotameter to measure the  flow




 rate.   The excess was recirculated back into the mixing chamber.  The




 rotameter calibration curve is presented in Appendix E.  The desired




 flow-rates were obtained by adjustment of valves located in  the piping




 network.  The piping network was then connected to the  injection  port




 located in the bottom of the laboratory flume.
                                   64

-------
Laboratory Flume



     The flume used is located in the hydraulics laboratory of the



Department of Environmental and Water Resources Engineering at Vanderbilt



University.  The flume is 60 feet long, 1.0 foot deep, and 2.0 feet wide.



The channel walls consist of glass panels, and the steel bottom was



painted with an epoxy-based paint to prevent chemical corrosion.  The



entire flume system is supported at two points, one of which has a



mechanical screw jack mechanism for changing the channel slope.  However,



all experiments were made with a horizontal slope.  A variable-speed



recirculating pump is located at the lower end of the flume by which



water may be recirculated if it is necessary.  At times it was necessary



to recirculate to obtain the desired ambient flow rate in the flume at



the required depth.  The maximum flow rate with recirculation was 0.48



cfs.  The maximum flow rate available for a single pass through the



flume was 0.15 cfs.  The flow rate was measured by a 60° V-notch weir



installed at the upstream end of the flume.  A calibration curve of



flows over the weir was used which appears in Appendix F and is very



close to the theoretical equation of



                                         2.5
                           0  = 1.4076 H
                           xa           w
where Q  = flow,  cfs
      H  = head of water above apex of notch, feet
       w
     A bank of ripple siding was installed near the upstream end to



aid in straightening the flow.  A point gauge mounted on a traversing



mechanism was used to measure the flume depth and local positions of



the jet with an accuracy of 0.001 foot.  The required depth of flow was
                                 65

-------
 controlled by means of a perforated baffle installed  at the downstream

 end of the flume.  A diagrammatic sketch of the flume and appurtenances

 may be seen  in Figure 19.
                                  60 FEET
            A. PROPELLER PUMP
            B. PERFORATED BAFFLE
            C. INJECTION PORT
            D. PIVOT  FOR SLOPE
              ADJUSTMENT
E.  POINT  GAGE
F.  BAFFLE
G.  SHARP-CRESTED
   WEIR
          FIGURE  19  - PARTIAL CUTAWAY VIEW OF RECIRCULATING FLUME


      The  injection  of the salt solution was obtained either by using a


 length  of standard  3/8-inch copper tubing as the port or injecting

 vertically from  a port in the bottom of the flume.  For the 90° jet,


 the salt  water was  injected vertically from the bottom of the flume using


 a previously installed withdrawal tap.  The diameter of the jet at this


point is  0.80 centimeter.  The 45° jet was a standard 45°-3/8-inch copper

ell inserted into the withdrawal port.  This essentially raised the


elevation of the jet 0.106 feet above the bottom of the flume.
                                 66

-------
The diameter of the 45° jet was 0.95 cm.  The 60° jet was constructed




by bending a length of 3/8-inch copper tubing to an angle of 60° from




the horizontal.  The jet was then inserted into the withdrawal port,




thus raising the elevation of the jet 0.081 feet above the bottom of the




flume.   The diameter of the 60° jet was 0.72 cm.  The location of the




withdrawal tap is in the center of the flume 1.0 feet from each wall.








Procedure for Obtaining Correct Salt Concentration




     The limiting factors for determining the various parameters during




the experiments were the ambient flow rate and the depth of the water




in the  flume.  The usual depth of water in the flume was between 10 and




12 inches.  A depth of 10 inches and a maximum flow rate for a single




pass of 0.15 cfs was used for preliminary calculations.  This set the




preliminary ambient velocity.  Then, using the required velocity ratio,




the preliminary jet velocity was determined.  Knowing the required jet




velocity and densimetric Froude number, the preliminary density differ-




ence was calculated for a particular jet diameter.  The salt was added




to water and sufficient time was allowed for complete mixing.  The




density of the solution was then determined using previously weighed




specific gravity bottles.  After the density difference was accurately




determined, the jet velocity was again determined using the required




densimetric Froude number.  Then, using the required velocity ratio,




the ambient velocity was calculated.  Using the maximum flow rate of




0.15 cfs, the depth of flow in the flume was calculated.




     If recirculation was required to obtain the desired combination of




parameters, a salt solution was initially mixed and then the density
                                 67

-------
difference was determined as for a single pass.  Then, the required jet




velocity and ambient velocity were calculated.  Using a depth of 12



inches, the ambient flow rate was calculated.  The required flow rate



was then obtained using the recirculating pump.
                                 68

-------
                                CHAPTER V
                                  VII
              ANALYSIS OF DATA AND PRESENTATION OF RESULTS
     The objective of this research was to ascertain the applicability

of two different jet models which describe the dilution, jet half-width,

and jet trajectory of positively buoyant jets in flowing streams  to  the

case of a negatively buoyant jet in a flowing stream.   The specific

objectives are outlined below:

     1.  Determine if Fan's model could be used to predict the dilution,

         jet half-width, and jet trajectory for a negatively buoyant

         jet.

     2.  Determine if Abraham's model could be used to predict the

         dilution, jet half-width, and jet trajectory of a negatively

         buoyant jet.

     3.  Determine the range of values of the coefficient of entrain-

         ment, a, and drag coefficient, C^.

     4.  Seek the functional relationship of a and Cd as a function  of

         the velocity ratio, k, and the densimetric Froude number, F,

         and the initial angle of discharge, p^.

     5.  If both models can be used, determine the best model.

     A chronological schedule for obtaining and evaluating the data  is

outlined below:

     1.  Determine the actual conditions for a particular laboratory

         investigation.

         a.  jet velocity


                                 69

-------
        b.   ambient velocity



        c.   density of salt water



     2.  Obtain probe  calibration curve of millivolt output versus




         concentration of salt in jet.



     3.   Obtain concentration profile data measured in millivolt  output




         from laboratory investigation.



     4.   Reduce millivolt output data describing the profile  to actual



         concentration data using the calibration curve  found in  step 2




         above.



     5.   Determine the centerline concentration, the vertical position,



         y,  and the standard deviation, o, , for a particular  profile



         which is located at a known distance downstream,  x,  from the




         discharge point.



     6.  Measure the  length of the zone of flow establishment along the



         axis  of the  jet trajectory using Equation 4.



     7.   Determine 6   from  plotted jet trajectory data obtained using




         the profile  data.



     8.   Obtain theoretical dilution,  jet half-width,  and jet trajectory




         for initial  guesses of a and  C,.



     9.   Fit theoretical data to experimental data to obtain the best




         values of a  and C,.



     A laboratory investigation was used to  answer the specific objec-



tives of this research.  Experiments were  conducted for various combina-



tions of k,  F, and 6  , with values of  k ranging from 5 to 20, F ranging



from 10 to 40, and initial  angles of discharge  of 90°, 60°,  and 45°.  A



laboratory run consisted of an experimental  investigation of one
                                 70

-------
particular combination of the above parameters,  e.g.,  k  ~  5, F  ~ 10, and
g1 = 90°.  The exact combinations are found in Table  3 for Fan's defini-
 o
tions and in Table 4 for Abraham's definitions.   The  laboratory conditions
were set up as discussed in Chapter IV.

Procedure for Obtaining Salt Concentration Profiles
     For each laboratory run, a calibration curve of  millivolts output
versus salt concentration was made for various dilutions.   Hence,  there
was no need for a temperature correction, since  the probe  was calibrated
each time a run was made.  Concentration profile data was  initially
taken starting at the downstream end of the jet.  This procedure was
followed so that any disturbance created by the  probe would be  felt only
downstream of the profile under investigation.  Figure 20  is a  typical
example of the data taken over one cross-section.
     The strip charts of millivolt output at each sampling point were
analyzed for each individual run.  Values of the average millivolt output
at a point in a cross-section for various time increments, At,  were
estimated by eye from the trace on the strip chart.  These were then
averaged according to Equation 83 to obtain the time-averaged millivolt
output for that point .

                                iV-i.j.k""!'
                         •    ' ^— -                   t83)
                                      I  At.
where   m. ,  = time-averaged millivolt output at location j in cross-
         J »k
               section, k
      m.  . ,  = millivolt output at location (j,k) over time increment
                                  71

-------
                                                                             S'/D
                                                                             x'/D
                                                                              c/c
3.54
2.00
0.48
0.55
i >
                        FIGURE  20  -  CONCENTRATION  PROFILE  AT  sVDQ  =  3.54  FOR  RUN  NO.  34

-------
TABLE 3 - COMBINATION OF DENSIMETRIC FROUDE NUMBER,
  VELOCITY RATIO, AND INITIAL ANGLE OF DISCHARGE
          ACCORDING TO FAN'S DEFINITIONS
Run
No.
9
10
11
12
13
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Froude
No., F
10.3
10.9
20.4
21.1
40.7
10.1
20.2
44.2
11.7
21.6
23.5
43.2
10.9
20.8
10.4
46.9
11.3
21.3
10.6
20.6
10.3
22.5
42.7
45.4
Velocity
Ratio
k = U /U
o a
10.3
5.5
10.2
5.3
10.2
20.1
20.2
5.5
5.9
10.8
5.9
10.8
10.9
20.8
20.9
5.9
5.7
10.6
10.6
20.6
20.6
5.6
10.6
5.7
Initial Angle
of
Discharge
*o
90
90
90
90
90
90
90
90
45
45
45
45
45
45
45
45
60
60
60
60
60
60
60
60
Angle at End of
Zone of Flow
Establishment
3o
74
62.5
78
74.8
76.3
81.8
78.3
58.3
31.8
36.5
28.0
36.3
28.0
34.5
29.3
30.0
48.8
51.3
52.0
53.0
53.8
52.0
53.8
47.8
Reynolds
No.
R
e
1650
1190
2200
1050
2030
1110
2240
4700
1730
3200
1600
3140
2540
4980
1590
4580
1180
2220
1870
3636
1035
1080
2040
2940
                       73

-------
TABLE 4 - COMBINATION OF DENSIMETRIC FROUDE NUMBER,
  VELOCITY RATIO, AND INITIAL ANGLE OF DISCHARGE
         ACCORDING TO ABRAHAM'S DEFINITIONS
Run No.
9
10
11
12
13
15
16
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Froude No.
F
10.0
10.0
20.0
20.0
40.0
10.0
20.0
40.0
11.4
21.4
22.8
42.8
10.7
20.7
10.4
45.7
11.0
21.0
10.5
20.5
10.3
22.0
42.0
44.0
Velocity
Ratio
k - U /U
o a
10.0
5.0
10.0
5.0
10.0
20.0
20.0
5.0
5.7
10.7
5.7
10.7
10.7
20.7
20.7
5.7
5.5
10.5
10.5
20.5
20.5
5.5
10.5
5.5
Initial Angle
of
Discharge
B;
90
90
90
90
90
90
90
90
45
45
45
45
45
45
45
45
60
60
60
60
60
60
60
60
Reynolds
Number
R
e
1580
1090
2180
1900
2000
1110
2220
4250
1700
3170
1555
3110
2440
4930
1580
4320
1150
2200
1840
3610
1000
1060
2030
2850
                      74

-------
           i = index number of time  period


           j = index number denoting a given location in a given  cross-



               section, k


           k - index number denoting a given cross-section



           n = number of time increments used



         At. = time period used in time-averaging




     The time-averaged millivolt outputs were then used as input  data



for a computer program called Analysis.  A listing of the computer



program, Analysis, is included in Appendix G.  This program used  the



calibration curve to convert the time-averaged millivolt output,  uij^*



to a time-averaged concentration, c. , .   Knowing c. ,, which was  the
                                    j ,K            j j^


concentration at location  (j,k), yR, the  y-location of the mean value



of the concentration profile,  ck, the mean values of  the  concentration



profile, and the standard  deviation, a^,  were calculated  using



Equations  84, 85, and  86,  respectively.
                                 N
                            ,  ,
                           y  ,

                           y
                            k

                                          k
                                         '
                                                                    (84)
                           c  =
                            k
                                   N



                                                                    (85)
                                 75

-------
                                                       2
                                       N

                                       f   r     • v1
                                           j,k   yj,k

                            (y-  J2
                                            E   c. k


        ok-|/	l±L_i!	           (86)



                                  E   c. .

                                 *,i   J'k
where      N = number of j  location in a  given cross-section, k


        y!  .  = the vertical distance from the  discharge point
         J >K


         CI = Ay, the incremental  distance between each y. ,
      It  is important to note that  c. , and c, are equivalent to
                                   J »K       K


 c*(s,r,)  and c(s),  respectively,  where c*(s,r,4>) is a local con-



 centration value for any given y-distance from the center line and



 for  any  given cross -sect ion, k, and  c(s) is  the center line con-



 centration for any given cross-section, k.   This is equivalent to



 the  nomenclature used in Equation  2  for Fan's model and Equation



 66 for Abraham's model.   The values  of c,, y/, and a,  are then



used as  input  to the  computer programs for comparison with the



theoretical predictions.  Using Equation 24, the relationship between



o,  and b, the  jet half -width,  is defined as
                               b =  /2 ak                           (87)
                                   76

-------
     Hence, the centerline concentration, c, , was calculated.   For




each c,,  there was both an x'-coordinate and a y'-coordinate.   The




x'-coordinate represented the horizontal distance from the discharge




point, whereas the y'-coordinate was the vertical location of the




centerline concentration measured from a horizontal plane passed through




the discharge point.  Figure 21 gives a more explicit description of




the location of a jet's trajectory.
      FIGURE  21 - REPRESENTATIVE PROFILE VIEW OF  A JET'S  TRAJECTORY






 Jet  Trajectories



      Using  the x'-  and  y'-coordinates,  the  jet  trajectory of each run




 was  plotted.  Twenty-four runs  were made.   Table  3 gives  the various




 combinations  of densimetric Froude number,  velocity  ratio, and initial




 angle of discharge  according to Fan's definitions.   The angle at the




 end  of  the  zone of  flow establishment and  the jet Reynolds number are




 also included, where the jet Reynolds number is defined as
                                   77

-------
                                    U  D

                               R  = -2—2.                           (88)
                                e     v
where R  = jet Reynolds number
       C



       v = kinematic viscosity of salt water



     Table 4 gives the various combinations of densimetric Froude number,



velocity ratio, and initial angle of discharge according to Abraham's



definitions.  Table 4 also includes the jet Reynolds number.



     The difference between the two tables arises from the different



definitions of the initial jet velocity.  Fan defines UQ, the initial



jet velocity at the end of the zone of  flow establishment, as



                           U  = u  -i- U  cos 0                        (89)
                            o    o    a     o



whereas Abraham defines U  as




                           U  = u  + U  cos B1                       (90)
                            o    o    a     o



For both Fan's model  and Abraham's model, the value of UQ is the same,



but the addition  of the vector component of the  ambient velocity with



two different initial angles give rise  to two values for the component



vector.  The process  of obtaining the value of  BQ will be discussed  in



detail  later, whereas g' is simply the  horizontal angle that the discharge



orifice makes with the ambient flow.  By reviewing  the initial  conditions,



Equations  42  and  52,  this  difference  in the initial angle  is illustrated.



      Thus, using  Equation  4, and knowing the value  of the  velocity ratio,



k,  the  length of  the  zone  of  flow establishment, s^, can be determined.



This  length is  then measured  along the  jet  axis. Once the  end  of  the



 zone  of flow establishment is  determined, the x^ and y^  coordinates  are



 determined.  This is  only  necessary  for the use of  Fan's model, since
                                  78

-------
Abraham does not consider the zone of flow establishment.  Once the end


of the zone of flow establishment is determined, the initial angle of


inclination, B , of the jet is measured.  Hence, all of the unknown


variables are known and the theoretical solution using Fan's and


Abraham's model can be sought.


     Only six runs will be discussed in the main text.  However, the


plots of the dilution, jet half-width, and trajectory for all 24 runs


and for both models can be found in Appendix H.  Tables 5 and 6 give


the pertinent location, dilution, and jet half-width data for the six


combinations discussed using Fan's model and Abraham's model, respectively,




Fitting of Data to Theoretical Curves


     Trial computer solutions were made for the fitting of the data to


Fan's model.  Initial guesses of a and C, were made based on experience.


However, it was soon found that, for any value of Cd, the jet trajectory


was under-predicted.  The theoretical considerations concerning the


behavior of C. in the development of a positively buoyant jet apparently
             d

are not applicable when applied to the case of a negatively buoyant jet.


The buoyancy forces that tend to retard the vertical momentum are


apparently  greater than the drag forces.  Hence, in an effort to force


the model to fit the data, the value of Cd = 0.00 was used  for all runs.


Hence, only an initial guess  for a was needed.


     Several trial computer solutions were made using different values


of a.  Theoretical values of  the jet  dilution  and jet half-width were


plotted on  the same graph as  the experimental  data.  Experimental values


of jet dilution and jet half-width were both used to  select an appro-


priate value of a which predicted the best fit.  In most  cases, a best



                                  79

-------
                                  TABLE 5 - SUMMARY OF NEGATIVELY BUOYANT JET EXPERIMENTS
                                               IN A CROSS-FLOW FOR FAN'S MODEL
Run
No.
13







10







33







8
po
90°







90°







60°







F

40.7







10.9







42.7







k

10.2







5.5







10.6







gi
po
76.3







62.5







53.8







s1
e
D
4.45







3.19







4.45







X1
e
D
0.75







1.10







2.43







"i
D
4.38







3.07







3.75







s
D
9.75
15.75
21.20
31.56
56.81
112.50
156.94
207.02
5.81
11.38
16.39
21.52
36.42
41.62
51.69
76.88
4.06
10.42
20.90
31.22
41.39
51.39
101.53
201.58
X
D
4.25
9.25
14.25
24.25
49.25
104.88
149.25
199.25
3.90
8.90
13.90
18.90
33.90
38.90
48.90
73.90
2.57
7.57
17.57
27.57
37.57
47.57
97.57
197.57
£
D
8.71
12.05
14.20
16.72
19.45
19.86
19.12
18.09
4.28
6.60
7.52
5.98
5.90
4.56
3.84
1.96
3.14
6.97
10.01
11.94
13.14
12.33
14.83
11.57
c
o
c
9.13
14.60
17.80
30.97
36.86
62.34
87.71
87.04
9.99
11.59
14.33
17.56
20.25
35.37
36.74
40.70
4.01
7.43
16.76
25.46
31.97
35.13
60.06
77.11
b_
D
o
4.47
4.15
5.19
6.42
6.81
9.61
11.03
11.68
2.14
3.24
3.56
2.95
4.97
5.35
5.06
4.11
2.20
3.58
4.91
4.96
6.44
7.23
9.12
13.50
00
o

-------
                                                 TABLE  5  -  Continued
Run
No.
27






22






19






6°
60°






45°






45°






F
11.3






43.2






11.7







5.7






10.8






5.9





Bl
0
48.8






36.3






31.8





s;
D
3.19






4.45






3.19





x'
e
D
1.94






3.45






2.68





yl
~D
2.57






2.93






1.81





s
D"
3.97
9.44
19.58
29.75
39.86
50.00
75.49
2.00
7.68
18.26
28.58
38.84
48.86
99.10
2.58
8.00
18.05
38.10
48.21
68.36
x
D"
3.06
8.06
18.06
28.06
38.06
48.06
73.06
1.54
6.55
16.55
26.55
36.55
46.55
96.55
2.32
7.32
17.32
27.32
47.32
67.32
y
D
2.43
4.75
5.93
7.04
6.00
4.78
0.00
1.18
3.95
7.54
10.08
11.99
13.41
17.52
1.18
3.06
4.08
4.15
3.69
0.96
c
o
c
4.34
7.44
11.65
18.62
21.28
23.66
45.05
2.56
4.11
7.44
12.25
14.77
19.58
29.82
3.67
5.56
10.12
17.82
25.08
31.70
b
D
o
1.68
2.61
3.24
3.90
4.48
4.50
5.26
1.74
2.57
3.39
4.52
5.18
5.63
6.62
1.72
2.20
3.15
4.30
5.40
5.42
00

-------
                              TABLE 6 - SUMMARY OF NEGATIVELY BUOYANT JET EXPERIMENTS
                                        IN A CROSS-FLOW FOR ABRAHAM'S MODEL
00
NJ
Run
No.
13







33







22






B
po
90°







60°







45°






F

40







42







42.8






k

10







10.5







10.7






sf
D
14.20
20.20
25.65
36.01
61.26
116.95
161.39
211.48
8.50
14.86
25.35
35.67
45.83
55.83
105.97
206.03
6.45
12.14
22.72
33.03
43.29
53.32
103.56
xf
D
5.00
10.00
15.00
25.00
50.00
105.62
150.00
200.00
5.00
10.00
20.00
30.00
40.00
50.00
100.00
200.00
4.99
10.00
20.00
30.00
40.00
50.00
100.00
zl
D
13.15
16.49
18.64
21.16
23.89
24.30
23.56
22.52
6.89
10.72
13.76
15.69
16.89
16.08
18.58
15.32
4.10
6.87
10.46
13.01
14.92
16.34
20.45
c
o
c
9.13
14.60
17.86
30.97
36.86
62.54
87.71
87.04
4.01
7.03
16.76
25.46
31.97
35.13
60.06
77.11
2.56
4.11
7.44
12.25
14.77
19.58
29.82
b
D
o
4.47
5.15
5.19
6.42
6.81
9.61
11.04
11.63
2.20
3.58
4.91
4.96
6.44
7.23
9.13
13.50
1.70
2.57
3.39
4.56
5.18
5.63
6.63

-------
                                                 TABLE 6 - Continued
00
Run
No.
10







27






19






B°
90°







60°






45°







10







11






11.4







5







5.5






5.7





s'
D
9.00
14.56
19.58
24.71
39.61
44.81
54.88
80.06
7.17
12.64
22.78
32.94
43.06
53.19
78.68
5.77
11.19
21.24
41.29
51.40
71.55
X1
D
5.00
10.00
15.00
20.00
35.00
40.00
50.00
75.00
5.00
10.00
20.00
30.00
40.00
50.00
75.00
5.00
10.00
20.00
40.00
50.00
70.00
y1
~D
7.34
9.66
10.59
9.04
8.96
7.62
6.90
5.01
5.00
7.32
8.50
9.61
8.97
7.35
2.57
2.99
4.87
5.89
6.06
5.50
2.77
c
o
c
9.99
11.59
14.35
17.56
20.25
35.37
36.74
40.70
4.34
7.44
11.65
18.62
21.28
23.68
45.05
3.67
5.36
10.12
17.82
25.08
31.70
b
D
0
2.15
3.24
3.36
2.95
4.97
5.35
5.05
4.11
1.68
2.61
3.24
3.90
4.48
4.50
5.26
1.72
2.20
3.15
4.30
5.40
4.42

-------
fit curve pertaining to a particular value of a was obvious.  However,




in several cases two or more values of a could fit the data, i.e., one



value of a would cause the theoretical value of jet dilution to best



fit the experimental data of dilution, and a different value of a would



cause the theoretical value of jet half-width to best fit the experi-



mental data  of jet half-width.   Since the equations of jet dilution and



jet half-width are coupled, a  technique using the squared difference



between the  theoretical  and experimental data was used to select the



best  fit value of a.  The squared difference between a theoretical



point and an experimental point was calculated both for the dilution



and jet half-width.  Then all  the squared differences for each value of



a were summed.  The  value of a giving the smallest value of squared



difference was used.



      Figures 22, 23, 24, 25, 26, and 27 show the theoretical fit of



Fan's model  and the  experimental results for six of the twenty-four



combinations of k, F, and 3' in question.  Lines on the figures represent



the theoretical values of jet  dilution, jet half-width, and jet



trajectory.   A circle, o, represents an experimental value of dilution,



and a triangle, A, represents  an experimental value of jet half-width



at a  location s/D on the jet axis.  On the trajectory curve, a circle,



o, represents an experimental  location  (x/D, y/D) of its respective



centerline dilution.



     The fitting of the theoretical curves to experimental data for



Abraham's model was much easier than the fitting of the data to Fan's



model, since  the entrainment coefficient for Abraham's model is



assumed constant, i.e., it is not a function of k, F, and B'.  Again,
                                 84

-------
60



40-
                                             I	1	T
20
   -o
        J	L
                                       J	,	L
  0
 IOO
40
80          120
      x/D
160
200
                                                            KXX>
                                s/D
        FIGURE 22 - OBSERVED VALUES AND THEORETICAL CURVES

              PREDICTED BY FAN'S MODEL - RUN NO. 13
                             85

-------
30
                         4O
                        50    60
                        x/D
70    80    90    100
  100
   IO
F  * 10.9
K  =5.5
  is90°
  * = 62.5°
a" « 0.7
      . DILUTION
          b/D
                 ' '
                    ' 11
                     -\—i—i   i 111
                                   Oo
                                     X
                                   i  I  i i i i
                                         i   i  i	I  I i I I
                       10
                                s/D
                                  100
                IOOO
          FIGURE 23 -  OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY FAN'S MODEL - RUN NO. 10
                               86

-------
60
       F   «42.7
     - K   « 10.6
                                 s/D
                                                             1000
         FIGURE  24  - OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY FAN'S MODEL - RUN NO.  33
                               87

-------
30
                                      60   70   80    90
K>    20    30    40
100
 lOOc
                                                            1000
           FIGURE 25 - OBSERVED VALUES AND THEORETICAL CURVE
                PREDICTED BY FAN'S MODEL - RUN NO. 27
                               88

-------
30-
  0     K>    20    3O   40    50    60    70   80   90    100
                                                            1000
                                s/D
         FIGURE 26 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL -  RUN NO. 22
                              89

-------
30-
20
 K)
        10    20    30    4O    50    60    70   80   9O    100
 100
       £0«3I.8°
       a   « 0.3
                                8/0
                                                           1000
         FIGURE 27 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL - RUN NO. 19
                             90

-------
the value of the drag coefficient,  C,, was assumed equal to zero.   For


each run, only the values of the densimetric Froude number and velocity


ratio are needed to obtain the theoretical results as outlined in Table


4.


     Figures 28, 29, 30, 31, 32, and 33 show the theoretical values from


Abraham's model and the experimental results for the six combinations


under consideration.  In addition,  lines on the figures represent the


theoretical values of jet dilution, jet half-width, and jet trajectory


predicted using Abraham's model.  A circle, o, represents an experi-


mental value of dilution, and a triangle, A, represents an experimental


value of jet half-width at some location s'/D on the jet axis.  On the


trajectory curve, a circle, o, represents an experimental location


 (x'/D, y'/D) of its respective centerline dilution.  In addition,


Figures  34, 35, 36, 37, 38, and 39 are photographs of the jet in actual


 laboratory conditions.



Discussion of  Results - Fan's Model


     For all angles of  discharge (90°, 60°, and 45°), the best  fit


 occurred at values  of high  densimetric Froude numbers and  low velocity


 ratios,  e.g.,  F = 46.9, k = 5.9, and  B^  =  90°.  For  the sake of dis-


 cussion  and comparison  of the "goodness-of-fit" of the  data to  the


 theoretical model  for the different combinations  of F,  k,  and  B^,  the


 value  of U  will  be considered  constant  in determining  the values  of F
          o

 and k  as defined  by Equation  9  and 11,  respectively.  At a high


 densimetric Froude  number,  e.g., F =  46.9, the difference between  the


 initial  density of the  jet  and  the ambient density is very small.
                                  91

-------
  60
  40
§
  20
                4
8C
120
160
200
  100
                                                              1000
                                  s'/D
            FIGURE 28 - OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY ABRAHAM'S MODEL - RUN NO. 13
                                 92

-------
30
        10    20    30    40    50    60    70    80    90    100
       F  =10.0
       K  =5.0
          DILUTION
                                          100
1000
                               S'/D
       FIGURE 29  - OBSERVED VALUES AND THEORETICAL CURVES
            PREDICTED BY ABRAHAM'S MODEL - RUN NO. 10
                            93

-------
  60
  40
§
  20
                40^       ^0          120          160         20O
                                  X'/D
                                 s'/D
                                            100
1000
           FIGURE 30 - OBSERVED  VALUES AND THEORETICAL CURVES
               PREDICTED BY ABRAHAM'S MODEL - RUN NO. 33
                                94

-------
15
10
              10

                                X-/D
40
50
                                                             IOOO
          FIGURE 31 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY ABRAHAM'S MODEL -  RUN NO. 27
                                95

-------
 30





 20
I
•»
s


 10
         10     20    30    40    50    60   70   80   9O    100

                                 x'/D
  100
                                                             1000
                                S'/D
        FIGURE 32 - OBSERVED VALUES AND THEORETICAL CURVES

             PREDICTED BY ABRAHAM'S MODEL - RUN NO. 22
                              96

-------
  30
I
  20
          10    20   30   40   50    60    70    80    90    100
   lOOr
         F  =11.4
      ILK  =5.7
                                            IOO
1000
                                 s'/D
           FIGURE 33 - OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY ABRAHAM'S MODEL - RUN NO. 19
                                97

-------
FIGURE 34 - PHOTOGRAPH OF NEGATIVELY BUOYANT JET
    FOR RUN NO. 13, F * 40,  k = 10, B1  = 90°
                     98

-------
FIGURE 35 - PHOTOGRAPH OF NEGATIVELY BUOYANT JET
     FOR RUN NO. 10, F * 10, k * 5, B  = 90°
                      99

-------
FIGURE 36 -  PHOTOGRAPH OF NEGATIVELY BUOYANT JET
    RUN NO.  33,  F = 40, k * 10, 6  = 60°
                    100

-------
 4IO    31
                      I/,  e
FIGURE  37 - PHOTOGRAPH OF NEGATIVELY BUOYANT JET
     FOR RUN NO.  27,  F * 10, k ~  5, 3' = 60°
                    101

-------
                      I
FIGURE 38 - PHOTOGRAPH OF NEGATIVELY BUOYANT JET
    FOR RUN NO.  22,  F * 40,  k = 10, 6'  = 45°
                     102

-------



FIGURE 39 - PHOTOGRAPH OF NEGATIVELY BUOYANT JET
     FOR RUN NO. 19, F = 10, k - 5, 3  = 45°
                     103

-------
Consequently, the effect of the negatively buoyant force which tends




to retard the vertical momentum is diminished.  The relationship of




the buoyancy parameter, f, in Equation 54 as a function of the




densimetric Froude number is defined in Equation 50.  In addition to




the small density difference for high densimetric Froude numbers, at




low velocity ratios, i.e., k = 5.9, the strength or relative velocity




of the  ambient fluid is greater.  Considering the same value of U , at




low velocity ratios the ambient velocity, U , is greater than at
                                           3.



higher  velocity ratios.  Hence, the component of the jet velocity due




to the  ambient velocity will be a larger percentage of the jet velocity.




Also, the normal component of the ambient velocity, U  sin 6, will be
                                                     cl


larger.  Referring to  Equation 6 and substituting Equations 7 and 8,




Equation 91 can be developed.







                       ^= 2irab(U2 sin2 6 + u2)0'5                   (91)
                       Q-S         <£




Hence,  the rate of dilution of the pollutant, salt in this case, will




be greater than at higher velocity ratios.  This can be envisioned as




a larger normal component of the ambient velocity acting to entrain




more of the ambient fluid into the jet to increase its dilution, thus




reducing the effect of the negatively buoyant force which is acting to




retard  the vertical momentum of the jet.  This larger normal component




will necessarily increase the vorticity in the region of the jet, as




outlined by Flatten and Keffer (38); thus, it will increase the rate of




dilution or mixing.  The cross-flow component, U  sin 8, decreases
                                                3.



slowly  relative to the growth in cross-sectional area.  Hence, the




cross-flow shearing force acts over an increasing jet circumference and
                                 104

-------
the rate of change of vortex inflow velocity would increase with




distance as discussed by Flatten and Keffer (38).




     Also, the case of a high densimetric Froude number and low




velocity ratio more nearly approximates Fan's basic assumptions as




outlined in Chapter II.  The radius of curvature for this combination




of densimetric Froude number and velocity ratio is large.  Hence, there




will be no overlapping or interacting of the concentration profiles.




     In addition, the cases of low densimetric Froude number, i.e.,




the density difference between the initial jet density and ambient




density is large, and low velocity ratios, e.g., F = 10.9, k = 5.5,




$' = 90°, can be fitted using Fan's model.  However, the fit is not as




good as for the previously discussed case of a high densimetric Froude




number and low velocity ratio.  Yet the ability to fit the data with




the model for low densimetric Froude numbers and low velocity ratios




is attributable to the greater relative strength of the  ambient velocity




where the normal component of the ambient velocity is  large.  Hence,




the dilution of the pollutant occurs at a faster rate  for  the same




reasons advanced earlier.  Consequently, the effect of the negatively




buoyant force is diminished.



     Yet  for the cases of  low densimetric Froude numbers and high




velocity  ratios, the  ability to predict the  jet dilution,  jet half-




width,  and jet trajectory  is not very  good.  The  cases of a  low



densimetric Froude number  are the  ones with  a  larger  density difference.




Hence,  the negatively buoyant force tends  to retard the  vertical




momentum  and  the  advancement  of the y-component of the jet axis to  a




 greater degree than  in the case of a high densimetric Froude number.
                                 105

-------
Once more, for high velocity ratios, the normal component of ambient




velocity will be small compared to the jet velocity.  Hence, the dilu-




tion capacity of the ambient stream is lessened.  Consequently, the




effect of the negatively buoyant force is not readily diminished.




Also, the radius of curvature for this case is very small with the




result being an overlapping and interacting of profiles which will




inhibit the ability to adequately describe the concentration profiles.






Discussion of Results - Abraham's Model
     The fitting of Abraham's model to the experimental results was much




easier than for Fan's model, for reasons outlined earlier.  The same




type of situation was found fox Abraham's model as for Fan's model.  The




cases of high densimetric Froude number and low velocity ratios, e.g.,




F = 40.0, k = 10.0, and &' = 90°, were fitted with the best results.




All cases of high densimetric Froude number would fit the experimental




data for all angles of discharge.  This is due, in part, to the small




density differences.  The radius of curvature is very large, hence the




dilution due to the action of the cylindrical thermal is more closely




approximated.  This is true for this case because the jet axis is nearly




parallel to the direction of the ambient velocity.  In addition, for low




velocity ratios, the strength of the ambient velocity is greater than at




high velocity ratios.   Actually, for a high densimetric Froude number,




the jet may be treated as a simple jet consisting of only momentum forces,




The first term on the right-hand side of Equation 70 accounts for the




region of the jet which is affected by its initial momentum.




     In the cases of low densimetric Froude numbers and low velocity
                                 106

-------
ratios, e.g., F = 10.0, k = 5.0, and g' = 90°, Abraham's model begins to




deviate from the experimental data after the jet reaches its maximum




height.  Apparently, in the initial reaches where the jet behaves like




a momentum jet, the model will predict the jet dilution and jet tra-




jectory satisfactorily.  However, after the jet reaches its maximum




height and is deflected towards the discharge level, the predicted data




begin  to deviate from the experimental data.  Similarly, as discussed




above, the assumption was made that the direction of flow is parallel




to the jet axis at distances far downstream.  This assumption is violated.




However, at a distance far downstream, the confidence with which the




centerline dilutions are measured and reported is open to question, due




to techniques of measuring the concentration profiles.  In addition, the




jet dilution is under-predicted.  As a result, the model will under-




predict the jet trajectory due to the  larger negatively buoyant  force




which  will hinder the  jet from increasing  its y-component.




     In the cases of low densimetric Froude numbers  and high velocity




ratios, e.g., F = 10.0, k =  20.0, and  6^ = 90°,  the  fit of Abraham's




model  to the experimental data  is poor.  Due  to  the  large density




difference,  the vertical momentum is retarded to such  an extent  that  the




jet trajectory  is grossly under-predicted.  The  assumption  that  the  jet




centerline  is parallel to the  direction  of flow  at  distances  far down-




stream is  flagrantly violated  as indicated in the example  cited.



      Hence,  when  the density differences are small  and the  strength of




the ambient velocity is  large  with  respect to the jet  velocity,




Abraham's  model  can be used to predict the jet dilution,  jet  half-width,
                                  107

-------
and jet trajectory.   For the cases of a low densimetric Froude number




and high velocity ratio, Abraham's model should not be used.
                                108

-------
                               CHAPTER VI
                                 VIII
                         SUMMARY AND CONCLUSIONS

Introduction
     Two existing jet models have been analyzed to determine the
feasibility of their use to predict the dilution, jet half-width, and
jet trajectory of a negatively buoyant jet.  An equation has been
developed to predict the coefficient of entrainment, a, used in Fan's
model as a function of the jet densimetric Froude number, velocity
ratio, and initial angle of discharge.  The limits of application,
sources of errors, and recommendations for further research are found
in the discussion that follows.

Fan's Model
     Fan's model (3) was developed to predict the dilution and buoyant
jet trajectory in a cross-stream with a uniform velocity.  An integral
approach was used to develop equations of  (a) continuity, (b) momentum,
(c) conservation of tracer and density deficiency, and  (d) two geometric
equations.  The same approach could be modified to predict the behavior
of a negatively buoyant jet in a cross-stream.  The main difference
would be in the derivation of the y-momentum equation where the direction
of the buoyancy term is simply reversed  (i.e., given a negative sign) in
order for the analogy of a negatively buoyant jet to be appropriate.
     Twenty-four experiments were completed in the laboratory flume at
Vanderbilt University.  Eight combinations of densimetric Froude number

                                109

-------
and velocity ratio were utilized for each angle of discharge,  i.e.,  90°,




60°, and 45°.  The data was analyzed using the seven equations




developed in the integral approach.  Values wer-     ruined for a,  the




coefficient of entrainment, and C,, the reduced drag coefficient, by




trial and error fitting of the experimental data to the theoretical




solutions.  The value of the drag coefficient, C,, was assumed to be




zero because, for any value of C,, the jet trajectory was under-pre-




dicted.  Fan (3) and Motz and Benedict (25) stated that the value of




C, was predominately used in determining the best fit for the jet




trajectory and played an insignificant role in its effect on the dilu-




tion of the  salt water tracer along the jet axis.  Hence, only values




of a were obtained from the fitting of the experimental dilution and




jet half-width data along the jet axis to theoretical dilution and  jet




half-width data.



     In all cases (90°, 60°, and 45°), the best fit of experimental




and predicted data occurred at values of high densimetric Froude numbers




and low velocity ratios.  At high densimetric Froude numbers, the effect




of the density difference was small.  In addition, at low velocity




ratios, the momentum and velocity of the ambient fluid is greater.




Hence, dilution of the salt water is attained at a faster rate.  Con-




sequently, the sinking of the jet after attaining its maximum height is




much less for jets with low velocity ratios compared to jets with a




high velocity ratio.  In addition, for jets with small densimetric




Froude numbers  (large density differences) and small velocity ratios,




the prediction for the dilution and the jet trajectory was also better




than for the same densimetric Froude number and high velocity ratios.
                                110

-------
This is due to the higher relative effect of the ambient velocity.




However, due to the higher density differences for low values of den-




simetric Froude number, sinking of the jet after the jet reaches its




maximum height is more pronounced than for jets investigated with high




values of densimetric Froude number.  For jets with small densimetric




Froude numbers, there is an inflection in both the theoretical dilution




and jet half-width curves.  By comparing the jet trajectory curve and




the dilution curve, this inflection occurs at the point where the jet




has reached its maximum height.  Consequently, it is at the same point




along  the  jet  axis that the angle,  9, changes from a positive value  to




a negative value.



      In most cases, Fan's model can be used  to predict  the dilution




and jet trajectory.   One of the basic assumptions for  the development




of  the equations  was  that the  radius of  curvature was  large.  However,




for a negatively  buoyant jet with  a low  densimetric  Froude number,  this




is  not the case.   The radius of  curvature of the jet axis  is  small  for




 low values of densimetric  Froude number  due to  the  greater  density




 differences after the maximum  height has been reached, which causes a




 more  rapid downward deflection and sinking of the jet axis  than for




 jets  with high values of densimetric Froude number.   In addition,




 accuracy in measuring the real profile may be limited due to the



 interacting or overlapping of sections of profiles along the jet axis.




 Nevertheless, if a value of a is known for some combination of densi-




 metric Froude number, velocity ratio, and initial angle of discharge,




 Fan's model can be used to predict the dilution and jet trajectory.  A




 mathematical  expression for obtaining a  as  a function of F, k, and  3;






                                 111

-------
will be evaluated later.






Abraham's Model




     Abraham's model is a modification of Fan's model.  Abraham con-




siders the coefficient of entrainment constant, not a function of the




densimetric Froude number, velocity ratio, or initial angle of dis-




charge.  Abraham considered the equation of continuity to consist of




two terms, a term related to a simple jet in a region near the discharge




point and a term related, to a cylindrical thermal for a region at some




distance downstream from the discharge point.  In addition, the equa-




tion governing the momentum in the x-direction was different from that




used in Fan's model due to the above consideration.  Moreover, Abraham




does not consider a zone of flow establishment in the development of




his model.




     The use of Abraham's model has one advantage over Fan's model in




that the coefficients of entrainment are set a priori.  Hence, the




prediction of the dilution, jet half-width, and jet trajectory curves




rely only on a knowledge of the densimetric Froude number, velocity




ratio, and initial angle of discharge.  In this study, Abraham's model




was used to predict dilution, jet half-width, and trajectory curves




for the experimental conditions.  For values of high densimetric Froude




number and low velocity ratios, the fitting of the jet trajectory




curves was excellent.



     Albertson, et al. (10), developed two separate equations to predict




the dilution of a simple momentum jet.  One predicted the dilution




(volumetric flux ratio) in the zone of flow establishment while the
                                112

-------
other predicted the dilution in sections beyond the zone of flow estab-




lishment.  Abraham used a relationship developed by Albertson to express




the dilution of the jet tracer near the discharge when, in reality,




the equation can actually be used only to describe the dilution in the




region of established flow.  Hence, the prediction of the dilution in




the region of zone of flow establishment is not valid, but this does




not preclude the use of the model to predict dilution at cross-sections




beyond the zone of flow establishment.




     In the case of a high densimetric Froude number and low velocity




ratio, the jet is approaching the case of a simple jet.  The case of a




high densimetric Froude number and low velocity ratio approaches the




case for which Abraham's assumptions are valid.  Basically, the region




near the nozzle is a momentum (simple) jet, and in the region where the




cylindrical thermal approximation is valid, the predominant direction




of flow of the jet is parallel to the direction of flow of the flowing




ambient stream.  Hence, for these cases, Abraham's model fit the




experimental data exceptionally well.  However, for the cases with




low values of densimetric Froude number and high velocity ratios, the




fit of the trajectory curve was not as good as discussed above.  In all




cases the predicted height of rise of the jet was underestimated with




respect to the actual height of rise observed in the experiments.




The assumption that at some distance downstream from the discharge point




the direction of flow is parallel to the direction of flow of the ambient




stream is violated.  Hence, the theoretical prediction of the jet tra-




jectory after the deflection in a downward direction is not valid.
                                 113

-------
Angle at the End of the Zone of Flow Establishment

     One of the more important parameters in the utilization of Fan's

model was the reduced angle of inclination or angle at the end of flow

establishment, 8 •  There was no data existing for a priori prediction

of B  for a negatively buoyant jet.  However, in the region in which

this initial deflection occurs, Abraham considered the jet to act as a

simple jet since the initial momentum forces are greater than the

buoyancy forces.  With this in mind, data for the reduced angle of in-

clination of a simple jet in a cross-flow does exist.  Fan (3) pre-

sented results from Gordier  (15), Jordinson  (16), and Keffer and Baines

(17) in his study of a round, positively-buoyant jet in a cross-flow.

For the above studies the angle of discharge was 90°.  Fan used Equa-

tion 92 to predict the reduced angle of inclination at the end of the

zone of flow establishment as

                            8n
                            g£- = 90° - 110/k                        (92)
                            po

     Motz and Benedict (25) presented results obtained for 8/6* in a

study of a heated surface jet into a cross-flow.  In the present study,

the reduced angle of inclination was obtained as discussed earlier.

Figure 40 represents the data found in Fan's (3), Motz and Benedict's

(25) and the present study of the angle at the end of the zone of flow

establishment versus the inverse of the velocity ratio, 1/k.   The

value of 8 /81  seems to decrease with an increase in 1/k.  With an in-
          o  o

crease in 1/k,  the relative value of the ambient velocity is higher.

Consequently, the deflection of the jet will be more due to the in-

creased amount of entrained ambient momentum in the x-direction.
                                114

-------
r o
     1.0
     0.9
     0.8
     0.7
     0.6
     0.5
        0.0
0.1
                           O - 90°
                           A - 45°
                           O- 60°
                           • - 90° ~ MOTZ - BENEDICT (25)
                           A - 45° '
                           • -60°-   "
                           • - 90° ' FAN (3)
0.2               0.3
       l/k
0.4
                        FIGURE  40 - INVERSE VELOCITY RATIO,  l/k, VERSUS

-------
However, there is quite a bit of scatter in the data, particularly for



the 45° jet.  For the 60° and 90° jets at low 1/k values, the values of



8 /81 are clustered closely around values of 8 /B' = .89 for 1/k - 20
 o  o                                         o  o


and 8/6' =  .845 for 1/k ~ 10.  However, the rest of the data exhibits



a high degree of scatter.  Hence, no statement can be made for a general



predictive equation relating the value of the angle at the end of zone



of flow establishment as a function of the inverse of the velocity



ratio, 1/k.





Coefficient  of Entrainment



     The most important parameter for the prediction of the dilution,



jet half-width, and jet trajectory curves is the coefficient of en-



trainment, regardless of whether it is a positively or negatively



buoyant jet.  In Fan's model the coefficient of entrainment is constant



for a given  combination of initial angle of discharge, densimetric



Froude number, and velocity ratio.



     Fan  (3) presents data for the prediction of the coefficient of



entrainment, a, for a positively buoyant jet.  The present study presents



data for an  a priori prediction of a for a negatively buoyant jet.  The



values of a  were obtained for the 24 combinations of densimetric Froude



number, velocity ratio, and initial angle of discharge.  Figure 41 shows



the relationship between the values of a for the various combinations.



A multiple linear regression analysis was performed on the data.  The



basic equation related a as a function of Log F, Log (k), and sin (8^)•



Equation 93 was developed and the correlation coefficient was 0.91.



       a = -0.107 + 0.104 Log(F) - 0.553 Log(k) + 1.05 sin (6')      (93)
                                116

-------
         DENSIMETRIC  FROUDE  NUMBER, F
FIGURE 41 - VALUES OF a FOR EXPERIMENTAL COMBINATIONS OF F, k, and 3'

-------
A test of significance of the correlation coefficient shows that, with



the null hypothesis, there is no relationship between the variables



which can be rejected at the 1% level.  Hence, Equation 92 will give a



value of a for any combination of densimetric Froude number, velocity



ratio, and initial angle of discharge within the ranges of these



parameters used in this study.



     The values of a ranged from 0.15 to 0.9.  The value of a = 0.15



was  found for the combination of F - 10, k - 20, and 3' = 45°, and the



value of a = 0.9 was found for the combination of F - 40, k ~ 5, and



3' = 90°.  The value of ot decreases as the case of a coflowing stream



is approached.



     Due to geometric  limitation, the degree that ambient fluid is



entrained is  less for  discharge of 45° than  for a discharge of 90°.



Hence, as 01  is decreased the volume of ambient fluid  available for



in-flow and dilution of the jet on the side  of the jet near the boundary



is decreased.  Motz and Benedict  (25) obtained similar results for a



heated surface jet.  When the discharge angle was changed from 90° to



45°, the value of the entrainment coefficient was changed from 0.4 to



0.2.  Equation 92 will predict a decrease in ot for a decrease in the



initial angle of discharge.  Also, the value of a tends to decrease as



the  velocity ratio increases.  As the velocity ratio increases, the



effect of the ambient current decreases (i.e., the dilution caused by



the  ambient current is less).  Consequently, the values of the entrain-



ment coefficient are less.  One would expect lower values of a for this



situation, since the jet is approaching the  case of a  simple jet issuing




into a stagnant environment with a = 0.057.
                                 118

-------
Jet Reynolds Number
     Values of the jet Reynolds number as defined by Equation 87 for the
cases presently under study are listed in Tables 3 and 4.  The values
range from 1000 to 5000.  Rawn, Bowerman, and Brooks (35) stated that,
for R  > 2,000, the jet flow will usually be turbulent, but turbulence
     6
is probably not fully developed until R  reaches 10,000 or 20,000.
                                       G
Frankel and Cummings  (36) stated that, for a fully turbulent jet, the
Reynolds number is sufficiently large to have no appreciable influence
on the dilution.  Rouse  (37) noted that, for low Reynolds numbers, the
expansion in a lateral direction will be due solely to mixing caused by
molecular diffusion, i.e., viscous shear.  However, at high Reynolds
numbers, eddies will be generated in the zone of maximum instability,
and the expansion process will occur much more rapidly.  The eddies
forming in this region will not only produce the deceleration and
acceleration of the respective zones, but will themselves penetrate
farther and farther into each zone.  Burdick  (31) concluded that, as
long as the jet could be considered turbulent, the Reynolds number
could be neglected as a  significant parameter.
     The runs with jet Reynolds numbers  greater than 2000 had the best
agreement  of predicted data to experimental data.  For  experimental runs
with jet Reynolds numbers  less than 2000, the  rate of decay of  the
potential  core  will be  less than  for the fully turbulent  case.   Hence,
the  length of  the potential core  will be longer.   Consideration of the
longer  length  of  the  potential core would cause the  experimental values
of the  y-coordinate to  be  decreased.  This  would  imply  a  better fit of
the  jet trajectory  since the  jet  trajectory was under-predicted using
                                 119

-------
the previously determined values of the y-coordinate.   Yet  this  effect



may be offset by the possibility of a smaller reduced  angle of inclina-



tion, B  , for a longer potential core.   A smaller BQ would  imply a




smaller  initial vertical momentum component which would tend to reduce



the height of rise of the jet and to again cause under-pediction of



the jet  axis.  In addition, there will be less dilution in  the cases of



jet Reynolds numbers less than 2000 due to the decreased turbulence.




Consequently, values of a for the experimental runs with Rg > 2000 may



be too  low  in considering application to a fully turbulent  jet.   Hence,



o should be higher to reflect the increased dilution due to turbulent



flow  conditions.  A higher a would also imply a better fit  of the



trajectory  data.  Hence, the difficulty of fitting the trajectory data



could possibly have been due to violation of the model assumption of




fully turbulent flow.



      The reasons for the low jet Reynolds number are twofold.  The



maximum flow in the laboratory flume was restricted and the size of the



flume was also restricted.  Hence, in some cases the jet velocity had



to be so low to meet the above restrictions that the criterion that the




jet Reynolds number be greater than 2,000 had to be violated.
                                 120

-------
                                  IX

                          LIST OF REFERENCES


1.   Water Resources Council, "The Nation's Water Resources," Washington,
    D.  C., 1968.

2.   Brooks, N., and Koh, R., "Discharge of Sewage Effluents from a Line
    Source into a Stratified Ocean," International Association for
    Hydraulic Research, Xlth Congress, Leningrad, September, 1965, pp.
    1-8.

3.   Fan, L. N., "Turbulent Buoyant Jets into Stratified or Flowing
    Ambient Fluids," Technical Report No. KH-R-15, W. M. Keck Laboratory
    of Hydraulics and Water Resources, California Institute of Tech-
    nology, Pasadena, California, June, 1967.

4.   Cederwall, K., "Jet Diffusion:  Review of Model Testing and Com-
    parison with Theory," Hydraulics Division, Chalmers Institute of
    Technology, Gb'teborg, Sweden, February, 1967.

5.   Cederwall, K., "Hydraulics of Marine Waste Water Disposal," Report
    No. 42, Hydraulic Division, Chalmers Institute of Technology,
    Gb'teborg, Sweden, January, 1968.

6.  Abraham,  G., "The Flow of Round Jets Issuing Vertically into  Ambient
    Fluid  Flowing in a Horizontal Direction," Proceedings  of the  5th
    International Water Pollution Research Conference,  San Francisco,
    July-August, 1970, pp.  Ill 15/1 -  III 15/7.

7.  Cederwall, K., and Brooks, N.,  "A  Buoyant Slot Jet  into a  Stagnant
    or  Flowing Environment," Report No.  KH-R-25, W. M.  Keck Laboratory
    of  Hydraulics and Water Resources,  California  Institute of Tech-
    nology,  Pasadena,  California, March,  1971.

8.  Abraham,  G., "Jets with Negative  Buoyancy in Homogeneous  Fluid,"
    Journal  of Hydraulic  Research,  Vol.  5,  No.  4,  1967, pp.  235-248.

9.  Turner,  J. S., "Jets  and  Plumes with Negative  or  Reversing
    Buoyancy," Journal  of Fluid  Mechanics,  Vol.  26,  Part  4,  1966,
    pp. 779-792~T~

10.  Albertson, M.  L.,  Dai, Y.  B.,  Jensen,  R.  A., and Rouse  H
    "Diffusion of Submerged Jets," Transactions^ American Society o±
    Civil Engineers,  Vol. 115,  1950,  pp. 639-697.

11.  Morton,  B.  R.,  "Forced Plumes," Journal of Fluid Mechanics, Vol.  5,
     Part I,  1959,  pp.  151-163.

12    Parker  F  L.   and Krenkel,  P. A., "Thermal Pollution:  Status of
     the Si "Report No.  3, Naiional Center for Research and Training
     £ t£ Hydrotgic and Hydraulic Aspects of Water Pollution Control,
     Vanderbilt University, Nashville, Tennessee, December, 1969.

                                 121

-------
13.   Morton, B.  R.,  Taylor, G.  I.,  and Turner, J.  S.,  "Turbulent
     Gravitational  Convection from  Maintained and  Instantaneous Sources,"
     Proceedings,  Royal Society of  London, Vol. 234A,  No.  1196, January,
     1956, pp.  1-23.

14.   Morton, B.  R.,  "On a Momentum-Mass Flux Diagram for Turbulent Jets,
     Plumes, and Wakes," Journal of Fluid Mechanics, Vol.  10,  1961, pp.
     101-112.

15.   Gordier, R. L.,  "Studies on Fluid Jets Discharging Normally Into
     Moving Liquid," Technical  Report No. 28, Series B, St.  Anthony
     Falls Hydraulic Laboratory, University of Minnesota,  Minneapolis,
     Minnesota,  August, 1959.

16.   Jordinson,  R., "Flow in a Jet  Directed Normal to the Wind," Aero.
     Research Council, Report and Memo No. 3074, 1956.

17.   Keffer, J.  F., and Baines, W.  D., "The Round Turbulent  Jet in a
     Cross-Wind," Journal of Fluid  Mechanics, Vol. 15, 1963, pp. 481-
     496.

18.   Pratte, B.  D., and Baines, W.  D., "Profiles of the Round Turbulent
     Jet in  a Cross Flow," Journal  of the Hydraulics Division, American
     Society of Civil Engineers, Vol. 92, No. HY6, Proc. Paper 5556,
     November,  1967, pp. 53-64.

19.   Tollmien, W., "Strahlverbreiterung," Zeitschr. Angew. Math, und
     Mech.,  1926, pp. 468-478.

20.   Schmidt, W., "Turbulente Ausbreitung eines Stromes Erhitzter Luft,"
     Zeitschr. Angew. Math, und Mech., Vol. 21, 1941, pp. 265-278,
     pp. 351-363.

21.   Abraham, G., "Entrainment Principle and Its Restriction to Solve
     Problems of Jets," Journal of  Hydraulic Research, Vol.  3, No. 2,
     1965, pp. 1-23.

22.   Fan, L. N., and Brooks, N. H., Discussion of "Horizontal Jets in a
     Stratified Fluid of Other Density," by G. Abraham, Journal of the
     Hydraulics  Division, American  Society of Civil Engineers, HY2,
     March, 1966, pp. 423-429.

23.   Abraham, G., "Horizontal Jets  in a Stagnant Fluid of Other Density,"
     Journal of the Hydraulics Division, American Society of Civil
     Engineers,  HY4, July, 1965, pp. 139-154.

24.   Rouse, H.,  Yih, C. S., and Humphreys, H. W.,  "Gravitational Con-
     vection from a Boundary Source," Tellus, Vol. 4, 1952, pp. 201-210.
                                 122

-------
25.   Motz,  L.  H.,  and Benedict, B.  A., "Heated Surface Jet Discharged
     into a Flowing Ambient Stream," Report No.  4, National Center  for
     Research  and  Training in the Hydrologic and Hydraulic Aspects  of
     Water Pollution Control, Vanderbilt University, August, 1970.

26.   Richards, J.  M., "Experiments on the Motion of Isolated Cylindrical
     Thermals  Through Unstratified Surroundings," International Journal
     of Air and Water Pollution, Vol. 7, 1963, pp. 17-34.

27.   Bosanquet, C. H., Horn, G., and Thring, M.  W., "The Effect of
     Density Differences on the Path of Jets," Journal of Fluid Mechanics,
     Vol. 5, Part  I, January, 1959, pp. 340-352.

28.   Fan, L. N., and Brooks, N. H., "Dilution of Waste Gas Discharge
     from Campus Buildings," Technical Memorandum 68-1, W. M. Keck  Lab-
     oratory of Hydraulics and Water Resources,  California Institute of
     Technology, Pasadena, California, January,  1968.

29.   Priestley, C. H. B., and Ball, F. K., "Continuous Convection from
     an Isolated Source of Heat," Quarterly Journal of the Royal
     Meteorological Society of London, England, Vol. 81, 1955.

30.   Krenkel,  P. A., and Orlob, G. T., "Turbulent Diffusion and the
     Reaeration Coefficient," Journal of the  Sanitary Engineering
     Division, American Society of Civil Engineers, Vol. 88, SA2,
     March, 1962.

31.   Burdick,  J. C., and Krenkel, P. A., "Jet Diffusion Under Stratified
     Flow Conditions," Technical Report No. 11, Environmental and Water
     Resources Engineering, Vanderbilt University,  1967.

32.   Krenkel,  P. A., "Turbulent Diffusion  and Kinectics of Oxygen
     Absorption," Ph.D. Dissertation, University  of California, Berkeley,
     California,  1960.

33.   Clements, W. C., "Pulse Testing  for Dynamic  Analysis.  Part I.
     An  Investigation of Computational Methods  and  Difficulties in
     Pulse Testing.  Part  II.  Application of Pulse Testing to Flow and
     Extraction Dynamics,"  Ph.D. Dissertation,  Vanderbilt University,
     Nashville, Tennessee,  June, 1963, pp.  91-93.

34.  Clements, W. C., "Electrical  Conductivity  Monitor,"  Instruments
     and Control  Systems,  Vol. 41, November,  1968,  pp.  97-98.

35.  Rawn, A. M., Bowerman,  R.  F., and  Brooks,  N. H.,  "Diffusers for
     Disposal of  Sewage  in Sea Water,"  Journal  of the  Sanitary Engineering
     Division, American  Society  of Civil Engineers, Vol.  86, March, 1960.

36.  Frankel,  R.  M.,  and Gumming,  J.  D., "Turbulent Mixing  Phenomena of
     Ocean Outfalls," Journal  of the Sanitary Engineering  Division,
     American Society of Civil Engineers,  Vol.  91,  April,  1965.
                                 123

-------
37.  Rouse, H., Engineering Hydraulics, Wiley and Sons, Inc., New York,
     1950.

38.  Flatten, J. L., and Keffer, J. F., "Entrainment in Deflected
     Axisymmetric Jets at Various Angles to the Stream," Technical
     Publication No. 6808, Department of Mechanical Engineering,
     University of Toronto, June, 1968.

39.  Larsen, J., and Hecker, G. E., "Design of Submerged Diffusers and
     Jet Interaction," Preprint No. 1614, presented at the American
     Society of Civil Engineers National Water Resources Engineering
     Meeting, Atlanta, Georgia, January, 1972.

40.  Holly, F. M., and Grace, J. L., "Model Study of a Dense Fluid in a
     Flowing Fluid," Preprint No. 1587, presented at the American Society
     of Civil Engineers National Water Resources Engineering Meeting,
     Atlanta, Georgia, January, 1972.

41.  Perry's Chemical Engineers' Handbook, Fourth Edition, McGraw-Hill,
     New York, New York,  1969.

42.  Briggs, G. A., "Plume Rise," AEC Critical Review Series, No. 26,
     1969.
                                124

-------
    GLOSSARY






LIST OF NOTATIONS

-------
                                 X

                         LIST OF  NOTATIONS



  A - Jet cross-section normal to jet axis,  L2


  b - Local characteristic length of half-width of the jet,  L


 b  - Jet half-width at the end of the zone of flow establishment,  L
  o


B ,  - Radius of thermal, L
 th


  c - Concentration at the jet axis


 c* - Local concentration value


 c  - Initial concentration at the end of the zone of flow establish-
  o

      ment, i.e., at 0


 CI - Ay,  the incremental distance between each y' ^,L



 C, - Drag coefficient for Fan's model



 C',  - Reduced drag  coefficient for Fan's Model
   d


 C    - Drag coefficient  for Abraham's  model
  da


c.    - Local concentration at  location (j,k)



  c  - Mean value of the concentration profile



  D  - Diameter  of jet at orifice and orifice diameter,  L
   o


  E.  - Constant  supply voltage


  E  - Internal  voltage of conductivity monitor
   o


   f - Buoyancy parameter


   F - Densimetric jet Froude number


  F  - Drag force per unit length, L T-




                               125

-------
     g  -  Gravitational  acceleration,  LT  2



     G  -  Conductivity of salt  solution



     h  -  Dimensionless  horizontal  or  x-momentum flax parameter



    H  -  Head of water  above apex  of  notch,  L
     W



     i  -  Vector in the  direction tangent to  the jet  axis  or  index number



         of time period



     j  -  Vector perpendicular  to the  jet axis  or  index number denoting



         a given location in a cross-section,  k



     k -  Velocity ratio or index number denoting  a given  cross-section



    k'  -  k + cos
                  o


     m - Momentum flux parameter
  m.
m.  . ,
 1,3,k
    - Time-averaged millivolt output at location j  in cross-section,  k
J >*


      Millivolt output at location (j,k) over time increment A^




  n - Number of time increments


  N - Number of j location in a given cross-section, k



  0 - Origin of the coordinate system (x,y), beginning of the zone of



      flow establishment



 0' - Origin of the coordinate system (x',y?), beginning at the point



      of jet discharge



  Q - Volumetric flux, L3!"1



 Q  - Flow in laboratory flume, L3T-1
  3.



 Q  - Initial volume flux at  the nozzle, L T




  r - Radial distance measured from the jet axis, L



 R  - Initial jet Reynolds number
                                 126

-------
  s  -  Distance  along  the  jet  axis  from  the  zone  of  flow  establishment,


      L


 s'  -  Distance  along  the  jet  axis  from  the  jet discharge point, L


 s1  -  Jet axis  length of  the  zone  of flow establishment, L
  6


 S  -  Dilution  ratio  at terminal height of  rise, y



At.  -  Time period used in time-averaging, T



  u -  Jet velocity at the centerline of the jet, LT 1


 u  -  Jet discharge velocity at orifice, LT'1



 u* -  Jet velocity at a local point, LT 1


 U  -  Ambient uniform velocity, LT
  a
                                 "1
                                       '1
U  - Initial jet discharge velocity, LT



 v - Dimensionless vertical or y-momentum flux parameter


V  - Quantity of fluid contained in a thermal jet per unit of



     length, L2


 x - Horizontal coordinate measured from the end of the zone of


     flow establishment, L


x1 - Coordinate axis in horizontal direction on the same plane as


     jet axis with origin at 0', L


x1 - Horizontal distance from jet discharge point to the end of the
 e

     zone of flow establishment, L


 y - Vertical coordinate measured from the end of the  zone of flow



     establishment, L


y'  - Coordinate axis in vertical direction, with origin at 0', L


y'  - Vertical distance from the jet discharge point to the end of
Je

     the  zone of  flow establishment,  L



                              127

-------
  y£ - y-location of the mean value of the concentration profile, L




y! ,  - Vertical distance of local point at location (j ,k) from jet
 J >K


       discharge point, L



  y  - Mean vertical height of rise of the plume, L




   a - Coefficient of entrainment



a    - Coefficient of entrainment for jets primarily influenced by



       initial momentum



 a ,  - Coefficient of entrainment for thermal moving through a



       stagnant ambient fluid



   £ - Dimensionless vertical distance (y)



  £' - Dimensionless vertical distance (y1)



   n - Dimensionless horizontal distance  (x)



  r\' - Dimensionless horizontal distance  (x1)



   C - Dimensionless distance along s-axis (s)



  C1 - Dimensionless distance along s-axis (s')



  8  - Initial angle of inclination at the end of the zone of flow



       establishment, degrees



  g' - Initial angle of inclination at jet discharge point, degrees




   9 - Angle of inclination of the jet axis (with respect to the



       x-axis), degrees



  Q  - Initial angle of inclination (with respect to x-axis) (6  =




       6  for Fan's model and 6  = g'  for Abraham's model), degrees
        o                      o    o



   M - Volumetric flux parameter



   v - Kinematic viscosity, L2?"1
                               128

-------
p  - Density of the ambient fluid, FT2L k
 3.



p  - Reference density taken as p (0), FT2L~k
 O                               3.



p  - Initial jet density, FT2L~1+




p* - Local density within a jet, FT2L ^



  - Angular coordinate on a cross-section normal to jet axis,



     degrees



a, - Standard deviation at cross-section k,  L
 k



Ah - Plume rise above top of stack
                              129

-------
            XI
         APPENDIX A

SALINITY-DENSITY RELATIONSHIP

-------
o
            10
            >-
                                                            TEMPERATURES IN °C
                             0.5      1.0       1.5     2.0      2.5      3.0      3.5

                               SALT  CONCENTRATION  IN  WEIGHT  PER CENT
              FIGURE 42 - DENSITY OF SALT WATER AS A FUNCTION OF SALT CONCENTRATION AND TEMPERATURE
                                (PERRY'S CHEMICAL ENGINEERS HANDBOOK, REF.  41)

-------
          APPENDIX B
COMPUTER PROGRAM - FAN'S MODEL

-------
C****THI S PROGRAM  CONTAINS  THE MAIN PROGRAM* AMD
C    TWO  SUBROUTINES NEEDED TO PROVIDE  INPUT AMD
C    OUTPUT CONTROL  TO SOLVE THE FIVE EQUATIONS IN
C    FANS MODEL.   THE MAIN  PROGRAM CONTAINS THE
C    NECESSARY INPUT INFORMATION WHILE  SUBROUTINE  FCT HAS
C    THE  EQUATIONS FOR INTEGRATION AND  SUBROUTINE  OUTP
C    PROVIDES THE  NECESSARY OUTPUT CONTROL.  THE OUTPUT
C    VARIOUS DISTANCES ARE  IN TERMS OF  DIAMETERS.   THE
C    OTHER OUTPUT  VALUES ARE SELF-EXPLANATORY.
C    IN  ADDITION*  SUBROUTINE DRKGS MUST ALSO BE USED AS
C    OUTLINED IN THE MAIN PROGRAM.
       IMPLICIT REAL*3C A-£)
       INTEGER NDIM* IHLF* I*NDATA*MI
       REAL OUTP* FCT
       REAL*3 AUX<<3* 5)*Y( 5)*DERYC 5)*PRMTC5)* SDATAC25)*
      ABDATAC25)*XDATA( 25) * Y DAT AC 25) * CONC 1 ( 25) * VARC 90)
       COMMON /INPUT/ K, BETA1* BOUPAR, CD, KK* BO* DI AM* ALPHA
       EXTERNAL FCT*  OUTP
       DATA DERY/5*.2DO/
       DO  973 MI  =  1*10
C******INPUT THE VALUE OF INDEX***********************
C******INDEX = 1   ALLOWS ONE TO INPUT  EXPERIMENTAL  DATA
C******AND THEN  MAKE AN INITIAL GUESS  AT  A VALUE  OF
C****** ALPHA.  IF  ANOTHER VALUE OF ALPHA  IS DESIRED
C****** AFTER THE INITIAL GUESS SIMPLY  FOLLOW THE
C******INI TI AL DATA  FOR INDEX=1 WITH A CARD WITH  INDEX=2
C******  AND THEN  ANOTHER CARD WITH THE SECOND GUESS AT
C******ALPHA.  THIS  MAY BE DONE FOR AS MANY TIMES AS
C******DESIRED UP  TO 10-  IF YOU DESIRE THE CALCULATION
C******TO STOP.   INSERT A CARD WITH INDEX  = 3«
C
C
       READC5*974) INDEX
       I FC INDEX .EG).  2) GO TO  975
       I FC INDEX .EG.  3) GO TO  976
C**TITLE OF PROJECT  OR RUN UNDER INVESTIGATION**********
       READC 5*804)< VAR( I )* 1= 1*20)
C******INPUT INITIAL JET LOCATION* JET TERMINAL  DISTANCE*
O******INCREMENTAL  DISTANCE*  AND ALLOWABLE ERROR
       READ(5*800)(PRMTCI )* 1 = 1*4)
C******PARAMETERS  NECESSARY  FOR OPERATION
C******
C       1   FROUDE  NUMBER
C       2  ALPHA-COEFFICI ENT  OF ENTRAINMENT
C       3  K- VELOCITY RATIO
C       4  BETAI -INITIAL ANGLE OF DISCHARGE-DEGREES
C       5  CD-DRAG  COEFFICIENT
       READC 5* 8 79) FROUDE* ALPHA* X* BETAI * CD
C****INPUT INITIAL CONDITIONS****************************
                              131

-------
C      1.  YC 15 = 1.0
C      2.  YC2)=1.0
C      3.  YC3) = ANYTHING—NEVER USED
C      4.  YC4)=0»0
C      5.  YC5)=0.0
       READC5*801)CYCI5*1=1,5)
       READC5>802) NDIM
C****INPUT REDUCED ANGLE OF INCLINATION  AT END OF ZONE
C***        OF FLOW ESTABLISHMENT
       READC5*973)BETA1
       BETA1 = BETA1/57.29578
C
C*******************************************************
C
C      INPUT THE NUMBER  OF CROSS-SECTIONS ALONG THE
C      JET AXIS CNDATA),  THE DIAMETER  OF  THE JET
C      CDIAM) FOR THIS PARTICULAR RUN.
C      INPUT THE X-DISTANCE*Y-DI STANCE*  S-DI STANCE*
C      CONCENTRATION  RATIO*  AND JET HALF-WIDTH DATA FOR  A
C      PARTICULAR CROSS-SECTION DOWNSTREAM.
C
C
C******************************************************
       READC 5*9 705NDATA* DIAM*(XDATACI) , YDATAC I )* SDATAC I ) ,
      4CONC1CI)*BDATACI)>I=1*NDATA)
       GO  TO 953
  975  READC5*973)ALPHA
  953  CONTINUE
       CORR = YC2)
       Y(2)=CORR*DCOS(BETA15
       YC 35 = CORR*DSIN(BETA15
       KK=K+DCOSCBETA1>
       BOUPAR=-C((K/KK>**2.5)/((DSQRTC2.DO))*ALPHA))/FROUDE**
     $2
C****WRITE THE WORKING PARAMETERS************
       WRI TEC 6*905)
       WRITEC6*504)(VARCI)*1=1*20)
       WRITEC 6* 900)FROUDE* K* ALPHA* BOUPAR* CD* BETAI
       IFCINDEX .GT.  15GO  TO  977
       BETA5=BETA1* 5 7.29 5 78
       WRITEC 6*910)BETAS
       WRITEC6*911)DIAM
       BO  =  DIAM*CDSQRTCK/(2.DO*KK))5
       WRITEC6*969)K*KK*BO
       WRITEC6*905)
       WRITEC 6*968)
  968  FORMATC1X* 'EXPERIMENTAL f* Tl6* 'EXPERIMENTAL'* T32*
     & 'EXPERIMENTAL'* T45* 'EXPERIMENTAL '* T59* 'EXPERI
     «MENTALf/T6* 'JET'*T16* 'CONCENTRATION ', T36* 'JETST46
     ** 'X-DISTANCE'*T60> 'Y-DI STANCE VT5* 'AXIS'*T33* 'HA
                            132

-------
&LF- WIDTH '/T6* 'S
& '//)
 DO 20 I=1*NDATA
                              'B/D'* T49 , 'X/D1* 162, 'Y/D
    SDATACI) =  SDATACI )/DI AM
    XDATACI) =  XDATACI )/DI AM
    YDATACI) =  YDATACI )/DI AM
    BDATACI) =  BDATACI ) *DSQRTC 2 . DO 3 /DI A>M
    COMCKI) =  l.DO/CCONCK I )*.01DO)
    WRITEC6*966) SDATAC I )>COMCI< I )*BDATA< I )* XDATACI ) , YD AT AC
   $1 )
 20 COMTIMUE
977 WRI TEC 6*904)
    CALL DRKGSCPRMT*Y^DERY*NDIM* IHLF* FCT^OUTP^ AUX)
    YC 1)=1.0
    YC2)=1-0
    YC 55 = 0-0
97B CONTINUE
    IFCIHLF.EO. 1 1) GO  TO 1 1
    IFCIHLF.EO. 1?)GO  TO 1 2
    WRITEC6j901) IHLF
  11 WRI TEC 6*902) IHLF
  12 WRITEC6..903) IKLF
    STOP
800 FORMATC3F10. 1,D10.4>
801 FORMATC5F5-1)
802 FORMATCI 1)
804 FORMATC20A4)
879 FORMATCF10.5*F5-2* 3F5. 1)
900 FORMATC///T88* 'FAN ' '5  MODEL ' // ' DEN SI METR1 C FROUDE  MO.'
    &, F6.3/ 'VELOCITY  RATI 0 ' * T33* f= ' * T39* F4. I/ "ENTRAIMMEMT C
    SOEFFI CIENT C
    &ALPHA) % T33> '= '* T38* F5. 3/ 'BUOYANCY PARAMETER *> T33* '= '*
    $T35*FS-5/ 'DR
    &AG COEFFICIENT', T33> '= ' , T3B , F5 . P/ ' I MI TI AL AMGLEOF  T)l S
    SCHARGEST33*
    *'= ST39,F3.0//)
901  FORMATC/ '****ERROR --- SI GNC PRMTC 3) ) -ME. SI GNCPRMTC
    &2)-PRMTC 1) ) . '/ 'IKLF =  t*I2*')V)
902  FORMATC/ '****ERROR --- IN3ITIAL  INCREMENT  BISECTED
    &MORE THAN  10 TIMES1/ ' C IHLF  =  SI2j')f/)
903 FORMATC/ «****ERROR --- PRMTC 3)   =  O'/'CIHLF  =  SIS-1)1/)
904  FORMATCT?> MET%T18* ' VOLtMETRI C '> T 32, 'CONCENTRATI OM '-
    &T49> 'MOMENTUM %T65* ' VELOCI TY f > T3 3-» 'JETST99* 'XST1 14*
    &'Y'*T126> 'BETAVT7* 'AXISf*T20> »FLUX'>T51, 'FLUXf,T66*
    A'RATIO'*T79> 'K ALF- VI DTHf • * T9 5, 'DI STAMCE ' , T 1 1 0,  'DI STA
    ftNCEV/TS* 'S/D',T65* 'UJ/UO STS3* 'B/D'*T98* 'X/DST1 13*
    &'Y/D'>T124* 'DEGREES'//)
905 FORMATC 1H1)
                            133

-------
  910 FORMATC/'BETAS  = '^FS.2/5
  911 FOR^JATC/'DI AM = **F5.2>

  969 FORM ATC T21, 'Kt*T26, '= ' * T30» F8 . 5/T2 1 * 'K * • STS6* '= '» T
  970 FORMATCI2-. F3.2/C5Fi0.5»
  973 FORWATCF10.5)
  974 FORMATC 111)
  976 END
C
C
C*****SUBROUTI!ME  FCTCX* Y» DERY)
C     COMTAIMS  THE DIFFERENTI Al, EQUATIONS THAT ARE
C     TO  BE INTEGRATED AMD  ARE LISTED  AS  DERYC I ) .
C
C
      SUBROUTINE  FCTCX, Y* DERY)
      IMPLICIT  REAL*3
      REAL*B    YC5)^ DERYC5)
      COM-MOM /INPUT/ K>BETA1*BOUPAR*CD*KK,BO,DI AM, ALPHA
      M=DSCRT(Y(2)**2-»-Y( 3>**2>
      BETA=DATAMCYC3)/Y(2) )
      IFCBETA) 1>2*2
     1 IFCCD)2*2*3
     3 CD = -CD
      PRINT 901^  CD
     2 COMTIMUE
      SS=DSIM(BETA>**2
      SSS=DSIM( BETA)**3
      CS=DCOSCEETA)
       TJSW=YC D/DSGRTCM)
       MU=M/Y( 1>
       UM=Y( 1>/M
       CHI=DSORT( SS+CKK*MU-2.DO*CS>**2>
       DERYC 1)=(USM*CHI )/KK
       DERY(2)=(USM/KKK)*(2.DO*CHH-CD*SSS)
       DERYC 3) = BOUPAR*UM-( CD*USM*SS*CS/KKK)
       DERYC4)=CS
       DEPYC 5)=DSIM(BETA)
  901  FORMAT( lX*F5-25
       RETURN
       END
C
C
O****SUBROUTINE  OUTPCX* Y* DERY, IHLF*MDIM, PRMT)
C
C
       SUBROUTIME  OUTP(X>Y> DERY*
       IMPLICIT REAL*8CA-Z>
                              134

-------
    INTEGER  *JDIM, IHLF
    REAL*3     YC5)*  DERYC5)*  PRMTC5>
    COMMON /INPUT/ K*BETA 1,BOUPAR,CD,KK,BO,DIAM, ALPHA
    IFCDABSC(X*!O.DO)-IDIMTC
    BETA =  CDATANCYC 3)/YCS» )
    R=R*80/DIAM
    URATIO  = CYC1)*KK*BQ**S/R**2>-DCOSCBETA>
    BETA =  BET A* 5 7. 29 5 78
    XX=X*BO/CDIAM*2.DO*ALPHA)
         YC^)*BO/C DIAM*2.DO*ALPHA)
         Y( 5)*BO/CDI AM*2.DO*ALPrIA)
    WRITE(6*901)XXJY< 1 ) > COMC, M^ URATI 0* R, XKKX* YYYY* BF.TA
901 FORMATC
    RETURM
                            135

-------
            APPENDIX C
COMPUTER PROGRAM - ABRAHAM'S MODEL

-------
C****THIS PROGRAM  CONTAINS  THE MAIN  PROGRAM,  AMD
C     TWO  SUBROUTINES MEEDED TO PROVIDE INPUT  AND
C     OUTPUT CONTROL TO SOLVE THE FIVE EQUATIONS IN ABRA-
C     HAMS MODEL.   THE MAIM  PROGRAM  CONTAIN THE
C     NECESSARY  INPUT INFORMATION WHILE S',DERY<5>,?RMTC5),SDATAC25),
      &BDATAC 25),XDATAC25>,YDATAC25), CONCH 2 5), VARCRO)
       COMMON  /INPUT/ K,KK,BOUPAR,CD, AMOM, ATHERM,BO,DI AM
       EXTERNAL  FCT, OUTP
       DATA DERY/5*.2DO/
C** TITLE OF PROJECT OR RUN UNDER  INVESTIGATION**********
       RF.AD(5,S04)C VARCI ), 1 = 1,20)
C******r\jprjT  INITIAL JET LOCATION,  JET TERMINAL DISTANCE,
C*******INCREMENTAL DISTANCE* AND ALLOWABLE  ERROR
       READC 5,300>(PRMT(I ), 1= 1,4)
C******PARAMETEES NECESSARY  FOR OPERATION
f ******
C       1  FROUDE  NUMBER
C       ?.  K- VELOCITY RATIO
C       3  BETAI-INI TI AL  ANGLE OF  DI SCHARGE- DEGREES
       READC 5,£?03)FFOUDE, K, BETAI
C***   THF. VALUES OF THE  EMTRAINMENT COEFFICIENT
C***   A\in T.P.^-r-  COEFFICIENT ARE CONSTANT FOR  ABRAHAM'S
C***   MODEL.  HENCE THE  VALUES CAN  BE IDENTIFIED
C***   FOR ALL COMBINATIONS WITHIN THE
       COMPUTER  PROGRAM.
       £MOM=0.057DO
              0« 5DO
       NP'JT INITIAL CONDITIONS****************************
C      I .  Y ( 1 ) = 1 . 0
C      ?.  YC?)=1.0
f      3.  Y(3)=  ANYTHING- -NEVER USED
C      4.  YC4)=0-0
C      5-  V(5) = 0.0
       PEAD< 5,^01)C Y< I) , 1=1, 5)
       READ(5,^Of5) NLIM
       BETAI =  BFT^I/57.2957R
C***
                              136

-------
c***
C***  IT  IS NECESSARY  TO INPUT  THE VALUES OF  THE
C***  RESPECTIVE LENGTHS FOR THE  ZONE OF
C***  FLOW ESTABLI SHMENT* I .E. XESTAB= THE X
C***  DISTANCE WITHIN  THE ZONE  OF LOW ESTABLISHMENT
      YESTAB= THE Y  DISTANCE WI THIN THE ZONE  OF FLOW
      ESTABLI SHM EN T*  AND SESTAB=  THE S DI STANCE
      WITHIN THE ZONE  OF FLOW ESTABLISHMENT.   THE
C***  VALUES OF XDATACI),  YDATAC I ) , AND SDATAC I ) THAT
C***  FOLLOW MUST HAVE THESE VALUES OF XESTAB* YESTABj
C***  AND SESTAB ADDED SO THAT  THE DISTANCES  WILL BE WITH
C***  RESPECT TO THE DISCHARGE  POINT-
C***
      READC 5,8031XESTAB, YESTAB, SESTAB
C
c* ******************************************************
c
C     INPUT THE NUMBER OF CROSS- SECTI ONS  ALOWG THE
C     JET AXIS CNDATA), THE DIAMETER OF THE  JET
C     CDIAM) FOR  THIS PARTICULAR  RUN.
C     INPUT THE X-DI STANCE* Y- DISTANCE* S-DI STANCE*
C     CONCENTRATION  RATIO* AND  JET HALF- WIDTH DATA FOR  A
C     PARTI CULAR  CRO SS- SECTION  DOWNSTREAM.
C
C
c************************* *****************************
      READC 5*970)!MDATA., DI AM* (XDATAC I >*YDATAC I ) * SDATAC I ) >
      &CONC1C I )*BDATACI )*!= 1*NDATA)
      CORR = YC2)
      YC2)=CORR*DCOSCBETAI )
      YC 3)=CORR*DSINCBETAI )
      KK=K+DCOSCBETAI)
      BQUPAR=-K**2/C 2. DO*KK**2*FROUDE**2)
      PETAI=BETAI*57-2957S
      VRITEC 6*905)
C****WRITE T4E VOFKING PARAMETERS************
      VRITEC 6*804) C VAR< I )* 1= 1* 20)
      VRI TEC 6* 900) FROUDE* '«* AMOM* A THERM* BOUPAR* CD* BETAI
      CD=CD*DSQRTC 2.DO)/PI
      WRITEC 6*91 1 )DI AM
      BO  = DI A>1/ 2. DO
      VRI TEC 6*9£9)K*KX*BO
       DO 20 I=1*NDATA
       SDATAC I) =C SDATAC I )+ SESTAB)
       KDATAC I ) =CXDATAC I )+XESTAB)/DI AM
       YDATAC I ) = C YDATAC D+YESTAe)/DI AM
       BDATAC I ) =  BDATAC I )*DSQRTC2.DO)/DI
       CONS1CI) =  l.DO/CCONCKI>*.01DO>
                              137

-------
       WRITEC 6*966)SDATACI> , CONC 1 C I ) * BOATAC I ) * XDATAC I ) * YDATAC
      SI)
   20 CONTINUE
       WRI TEC 6*904)
       CALL DRKGSC PRMT* Y* DERY* NDIM*IHLF*FCT* OUTP* AUX)
       IFCIHLF.EQ.11)GO TO 11
       IFCIHLF.EQ.12)GO TO 12
       WRI TEC 6*90 DIHLF
   11 WRITEC 6*902)IHLF
   12 WRITEC 6*903)IHLF
       STOP
  800 FORMATC3F10.1*D10.4)
  801 FORMATC5F5.1)
  802 FORMATCI1)
  803 FORMATC3F10.5)
  804 FORMATC20A4)
  900 FORMATC///T26* 'ABRAHAM "S MODEL '// 'DENSIMETRI C '*
      &' FROUDE NO. f*T33* '=f*T37* F6.3/'VELOCITY  RATIO'
      &*T33*'='*T39*F4.I/'MOMENTUM COEFFICIENT'/
      *'OF
      &' THERMAL'/'COEFFICIENT OF
      4T3S*F5.3/ 'BOUYANCY PARAMETER'* T33* '= '*T35*F3.5/
      A'DRAG COEFFICIENT'*T33* '='*T38*F5.2/'I NITIAL
      &' OF DISCHARGE'*T33* '= ' * T40* F3-0)
  901 FORMATC/'****ERROR	SIGNCPRMTC3)).NE.SIGNCPRMTC
      &2)-PRMTC 1)). VIHLF  = '*I2>')'/)
  902 FORMATC/'****ERROR	INITIAL INCREMENT BISECTED
      &MORE THAN  10 TIMES V'CIHLF =  '*I2*')'/>
  903 FORMATC/ »****ERROR	PRMTC3) =  O'/'CIHLF  =  '*I2*f)V
  904 FORMATC///'  JET  VOLUMETRIC CONCEN- MOMENTUM'*
      &'    JET'* T46* fX'*T56* 'Yf* T66* 'BETA'/' AXIS'*
      «T10* 'FLUX'*T 18*  'TRATION'*T28* 'FLUX'* T36* 'HALF'*
      ST43* 'DISTANCE   DISTANCE    DEGREES VT36* ' WI DTH '
      &/ '   S/D'*T1 1* *U'*T29* 'M'*T37* 'B/D'*T45* 'X/D'*
      &T56*'Y/D'///)
  905 FORMATC1H1)
  911  FORMATC/'DIAM  =   ',F5»2>
  966 FORMATC T5* F8 . 4* T20* FS • 4* T34* F8 - 4, T47* F8 . 4* T60* F8 • 4 )
  968  FORMATCIX*'EXPERIMENTAL'*T16*'EXPERIMENTAL'*T32*
      A'EXPERIMENTAL',T45*'EXPERIMENTAL'*T59* 'EXPERIMENTAL'
      &/T6* 'JET'*T16* 'CONCENTRATION'*T36* 'JET'*T46*
      &'X-DI STANCE'* T60* 'Y-DI STANCEVT5* 'AXIS'* T33*
      &'HALF-WIDTH'/T6* 'S/D'* T36* 'B/D'*T49* 'X/D'*T68*
      4*Y/DV/)
  969 FORMATCT21* 'K'*T26* '= '* T30* F8 . 5/T2 1* 'K" '*
      *T26* '= '*T30*F8.5/T21* 'BO **T26* '= '* T30* F8 . 5)
  970 FORMATCI2*F3.2/C5F10.5))
       END
C
C
                              138

-------
C***** SUBROUTINE  FCTCX* Y* DERY)
C      CONTAINS  THE DIFFERENTIAL EQUATIONS THAT  ARE
C      TO  BE INTEGRATED AND ARE LISTED AS DERYCI).
C
C
       SUBROUTINE  FCTC X* Y* DERY)
       IMPLICIT  REAL*8
       REAL*5    Y<5)* DERYC5)
       COMMON /INPUT/ K* KK* BOUPAR* CD* AMOM* AT4ERM* BO* DI AM
       M=DSQRTCYC2)**2+YC 3)** 2)
       BETA=DATANCYC 3)/YC2) )
       IFCBETA) 1*2*2
     1  IFC ATHERM)2*2* 3
     3  CD  = -CD
       ATHERM=-ATHERM
       WRITEC6*901) CD*ATHERM
     2  CONTINUE
       S=DSINCBETA)
       SS=DSINCBETA)**2
       SSS=DSIN C BETA) ** 3
       C5=DCOS(BETA)
       USM=YC 13/D50RTCM)
       MU=M/YC 1)
       UM=YC 1)/M
       DERYC 1)=4. DO*C AtMOM*DSQRTCM)-C USM/KK) * ( 2. DO* AMQM*CS- AT4
      SERM*S*CS) )
       DEPYC 2> = C 4. DO*USM/XKK5*C 2. DO*AMOM*C KX*MU-2. DO*CS)
      &+2.DO*ATHERM*S*CS+CD*SSS>
       DERYC 3> = 4.DO*CBOUFAR*UM-CCD*USM*SS*CS/K.KK) )
       DERY(45=CS
       DERYC 5) = S
  901  FORMATC 1X*F5«2* 7X* F5.2)
       RETURN
       END
C
C
C***** SUBROUTINE  OUTPCX* Y* DERY* IHLF*MDI M* PFMT)
C
C
       SUBROUTINE  OUTPCX,Y, DERY* I !-rLF*NDIM* PRMT)
       IMPLICIT  REAL*3CA-?:>
       INTEGER NDIMj I 4LF
       REAL*3     YC5)->  DERYC 5)* PRMTC 5>
       COMMON"/ I MPUT/ X*KK* BOUPAR, CD* AMOM* A THERM* BO* DI AM
       IFCDABSC (X*10.DO)-IDINTCCX+ lD-6>* 10. D0» >  • GT.
      & .001 DO) RET URN
       IFCX  .LT.  l.ODO)  GO  TO  10
       IFCDABSCX-IDINTCX+ ID- 6) > - GT. . 00 1 DO) RETURN
       IFCX  .LT.  40- 01) GO TO  10
                               139

-------
    I FCDAPSC (XX 10.DCD-IDIMTCCX+lD-6)/10.DO> ) .  GT,
   &  .001)RETURN
 10 CONG =   l.BO/YC1)
    M=DSCRT(YC 2
    BETA = CDATAM
    IvRI TE(6*901)XX>YC 1 > , COMC^M, R, XXXX, YYYY* BETA
901 FORMATC IX* F6. 2.. T8* F7. 8* T 18* F7. 5^ T27, F7. 2* T36*
    &F6.P»T44,F7.S,T52* FB . 2* T65* F6- S)
    RETURN
    EMD
                            140

-------
       APPENDIX D
COMPUTER PROGRAM - DRKGS

-------
c
c
C   THIS ISA  SCIENTIFIC LIBRARY PROGRAM TO INTEGRATE A
C   SYSTEM OF  SIMULTANEOUS  ORDINARY DIFFERENTIAL
C   EQUATIONS  USING A FOURTH-ORDER RUNGE-KUTTA TECHNIQUE.
C   THIS PROGRAM IS AVAILABLE A MOST  COMPUTER CENTERS-
G   THE INPUT  VARIABLES ARE EXPLAINED IN THE PROGRAMS
C   FAN AND/OR ABRAHAM.

C
C
C
      SUBROUTINE DRKGSC PRMT, Y* DERY^NDIM, IHLF* FCT* OUTPj AUX)
      DIMENSION YC 1)>DERYC 1>* AUXC8> 1 ) > AC 4) , BC 4) * CC4>*PRMTC 1)
      DOUBLE  PRECISION PRMT> Y,» DERY* AUX.. A., B> C^X^XEND* H, AJ, BJ,
     1DELT
      DO  1  I = 1*NDIM
    1 AUXC8* I ) = . 066666666666666667DO*DERY( I )
      X=PRMTC 1)
      XEND=PRMTC2)
      H=PRMT(3)
      PRMT(5)=O.DO
      CALL  FCTCX*Y*DERY)
C
C     ERROR TEST
      I FCH*CXEND-X> ) 38* 37* 2
C
C     PREPARATIONS  FOR RUNGE-KUTTA METHOD
    2 A( 1>».5DO
      A(2)=. 2928932188 1345248 DO
      A< 3)= 1-707 10678 1 1865475DO
      A<4>=. 16666666666666667DO
      B< 1)=2.DO
      B< 3) = 1
      B(4>=2.DO
      C( 1) = .5DO
      C(2)=. 2928932188 1345248 DO
      CC 35=1.707106781186547500
      C(4)=.5DO
C
C     PREPARATIONS  OF FIRST RUNGE-KUTTA STEP
      DO  3 I = 1*NDIM
      AUXC 1*I)=YCD
      AUXC2, I)=DERYCI )
       AUX(6*I)=0»DO
       IREC=0
       IHLF=-1
                              141

-------
      ISTEP=0
      IEND=0
C
C
C     START OF A RUMGE-KUTTA STEP
    4 IF< CX+H-XEND)*H> 7*6* 5
    5 H=XEND-X
    6 IEND=1
C
C     RECORDING OF  INITIAL VALUES OF THIS  STEP
    7 CALL OUTP
      BJ=BCJ>
       CJ=CCJ)
       DO  11 I=1*NDIM
      R1=H*DERYCI)
       R2=AJ*I > + R2-CJ*Rl
       IFCJ-4)12*15*15
    12 J=J+1
       IFCJ-3)13*14*13
    13 X=X+.5DO*H
    14 CALL FCTCX*Y*DERY)
       GOTO 10
C     END OF INNERMOST RUNGE-KUTTA LOOP
C
C     TEST OF ACCURACY
    15 IFC I TEST) 16* 16*20
C
C     IN CASE ITEST=0 THERE  IS NO POSSIBILITY FOR  TESTING OF
C     ACCURACY
    16 DO 17 I=1*NDIM
    17 AUX(4*I)=Y(I )
      ITEST= 1
      I STEP= I STEP+1 STEP-2
    18 IHLF=IHLF+1
      X=X-H
      H=.5DO*H
      DO 19 I=1*NDIM
      Y(I)=AUX(1*1)
      DERYCI)=AUX(2*I)
                              142

-------
   19 AUXC6* I ) = AUXC3* I )
      GOTO  9
C
C     IN  CASE ITEST=1  TESTING  OF ACCURACY IS POSSIBLE
   20 IMOD=ISTEP/2
      IF(ISTEP-IMOD-IMOD)81*S3*81
   21 CALL  FCTCX*Y*DEKY)
      DO  22 I=1*NDIM
      AUX<5*I)=YU)
   22 AUXC7*I)=DERY(I)
      GOTO  9
C
C     COMPUTATION  OF  TEST  VALUE Dh.LT
   23 DELT=O.DO
      DO  24 I=1*NDIM
   24 DELT=DELT+AUXC8*I)*DARSCAUXC4*I)-Y
      IFCDELT-PRMTC4))28*28* 25
C
C     ERROR IS TOO GREAT
   25 IFCIHLF-10)26*36, 36
   26 DO  27 I=1*NDIM
   27 AUXC4, I) = AUX( 5* I )
      ISTEP=ISTEP+I STEP-4
      X = X-H
      IEND= 0
      GOTO 18
 C
 C     RESULT VALUES ARE GOOD
   28  CALL FCTCX*Y*DERY)
       DO  29  I=1*NDIM
       AUXC1*I)=YCI)
       AUXC2*I)=DERY(I)
       AUXC3*I)=AUX(6*I)
      Y
       IF(PRMTC 5))40* 30* 40
    30 DO 31  I = 1*NDIM
       YCI)=AUX(1*I)
    31  DERY(I) = A.UX(2* I )
       IREC=IKLF
       I FCI END) 32* 32* 39
 C
 C     INCREMENT GETS  DOUBLED
    32 IHLF=IHLF-1
       ISTEP=ISTEP/2
       H=H+H
       IFCIHLF)4* 33* 33
    33 I MOD=I STEP/2
       IFCISTEP-I MOD-I MOD)4* 34*4
                               143

-------
   34 IFCDELT-.02DO*PRMT<4»35> 35>4
   35 IHLF=IHLF-1
      ISTEP=ISTEP/2
      H=H+H
      GOTO 4
C
C
C     RETURNS TO CALLING PROGRAM
   36 IHLF=11
      CALL FCT(X*Y*DERY)
      GOTO  39
   37 IHLF=12
      GOTO  39
   38 IHLF=13
   39 CALL  OUTPCX*Y.»DERY*IHLF>NDIM.»PRMT>
   40 RETURN
      END
                              144

-------
          APPENDIX E
CALIBRATION OF 0.5 gpm ROTAMETER

-------
                  100
C/1
              5
              UJ
                              O.I
    0.2        0.3        0.4        0.5
 FLOWRATE,  GALLONS  PER  MINUTE
FIGURE 43  - CALIBRATION OF 0.5  gpm ROTAMETER
0.6

-------
          APPENDIX F
CALIBRATION OF 60° V-NOTCH WEIR

-------
   0.7


   0.6



1  0.5
a:
    0.4
8  0.3
oc
u

|  0.2



fe



I
CD
x
    O.I
                                       i     i     i   i   i   i
      0.01
                     0.02     0.03   0.040.05             O.I


                           FLOWRATE, CUBIC FEET  PER SECOND
0.2       0.3
                          FIGURE 44 - CALIBRATION OF  60° V-NOTCH WEIR

-------
        APPENDIX G
COMPUTER PROGRAM - ANALYSIS

-------
c
C COMPUTER PROGRAM  TO CONVERT RAW LABORATORY DATA
C TO JET  AXIS DATA.
C
C
      REAL MC100)
      DIMENSION CC100)>SC100)*DC100)*X3C100)*CXDC100)*CXDSQC
     $100),C1C100
     &)*CALC2C100)*XDISTC25)*DIF2C100)*YMEAN1C25)*CONMAXC25)
     f *STDEV1C25)
C     YCEPT AND SLOPE ARE  THE PERTINENT PARAMETERS TO
C     DEFINE THE CALIBRATION  CURVE.   YCEPT AND SLOPE ARE
C     DIFFERENT FOR EACH RUN.  THE  CALIBRATION CURVE ISA
C     LINEAR EXPRESSION WHICH RELATES THE MILLIVOLT OUTPUT
C     TO  THE CONCENTRATION  OF SALT.
C
      READC 5*950)YCEPT*SLOPE
C     IRUNNO = THE  EXPERIMENTAL RUN NO.
C     PGTV   = IS THE RELATIVE DEPTH OF THE MEASURING  DEVICE
C               TO TOP OF THE  WATER
C.    PGBV   = IS THE RELATIVE DEPTH OF THE MEASURING
C               DEVICE TO THE  BOTTOM OF THE WATER OR THE
C               BOTTOM OF  THE  LABORATORY FLUME
C     PROBTW = IS THE RELATIVE LOCATION OF THF, CONDUCTIVITY
C               PROBE WITH  RESPECT  TO THE MEASURING DEVICE
C               AND  TOP OF  THE WATER
C     ELEJET = THE  HEIGHT  OF  THE JET DISCHARGE POINT FROM
C               BOTTOM OF  THE  FLUME
C     THE ABOVE VARIABLES  ARE NEEDED TO TRANSFORM THE
C     LABORATORY DATA OF  THE  MEASURED DEPTHS TO ACTUAL
C     DEPTHS WITH RESPECT  TO  THE JET DISCHARGE POINT.
      READC 5*903)1RUNNO* PGTW*PGBW*PROBTW* ELEJET
      READC 5*920)FROUDE
C     FROUDE = DENSIMETRIC FROUDE  NUMBER FOR THIS EX PERI -
C               MENTAL RUN
      WRITEC6*802)IRUNNO
      DEPTH =PGTW-PGBW
      DEPINC = PGTW-PROBTW
      ELEINC = ELEJET - PGBW
      WEI TEC 6* 907)PGTW* PGBW* PPOBTW* DEPTH* DEPINC* ELEINC
C     N = THE NUMBER OF CROSS-SECTIONS ALONG THE JET AXIS
C     Xl= THE RELATIVE LOCATION OF  THE JET DISCHARGE POINT
C          IN THE X-DIRECTION.  THIS VALUE MAYBE ZERO OR
C          SOME OTHER RELATIVE X VALUE
C     DIAM = THE  JET DIAMETER
      READC5*900)N*X1*DIAM
      READC5*912)AK*BETA
C     AK = THE VELOCITY RATIO FOR  THIS EXPERIMENTAL RUN
C     BETA = THE  INITIAL  ANGLE OF  DISCHARGE
      WRITEC6*801)N*X1*DI AM
                              147

-------
      DO  10  J=1,N
C     Nl  = NUMBER OF Y LOCATIONS ON THE  CROSS-SECTION OF  THE
C         JET AXIS THAT  WERE MEASURED
C     X = THE RELATIVE X LOCATION OF THE CROSS-SECTION  WITH
C         RESPECT TO THE JET DISCHARGE POINT.
C     CI  = THE DELTA Y VALUE OR THE INCREMENTAL Y DISTANCE
C          ALONG THE CROSS-SECTION
C     MCI) « THE AVERAGE MILLIVOLT OUPUT AT THIS PARTICULAR
C             Y LOCATION
C     DC I) = THE RELATIVE  Y  LOCATION AT  WHICH THE MILLIVOLT
C         READINGS WERE TAKEN
      READC5,910)N1*X,CI,CMCI),DCI ),I=1*N1)
      CI=CI*30.48
      S02PI  a 2.506628274631
      DIF22  = 0.0
      CNSUM1  =0.0
      CNSUM2  * 0-0
      FREO a 0.0
      X2=CX-X1)/DI AM
      XDI STCJ)=X2*DI AM
      WRI TEC 6* 901) J> X, X2
      WRITEC6,902)
C
C
C    CALCULATES THE MEAN  Y-DISTANCE*  THE  STANDARD DEVI A-
C    TION* AND VARIANCE OF  THE PARTICULAR CROSS-SECTION
C
C
      DO  20  I=1*N1
      DCI)=DCI)-PGBW+DEPINC-ELEINC
      DO) = DCI)*30.48
      C(I)=YCEPT  +  SLOPE*MCI)
      IFCCCI) -GT. 0.000000)00 TO  21
      CCI) = 0-00000000
      SCI) = 10000000.00
      GO  TO  22
   21 S(I)=1./CCI)
   22 CKI)  = CCI)*100.00
      FREO = FREQ +  CKI)
      CXDCI ) = CKI) * DCI)
      CXDSQCI) - CXDCI)  *  DCI)
      CNSUM1  = CNSUM1   +  CXDCI)
      CNSUM2   = CNSUM2  + CXDSQCI)
   20 CONTINUE
      YN1EAN  = CN SUM 1/FREO
      STD =  SQRTCCCNSUM2  -  CCCNSUM1**2)/FREQ))/FREQ)
      YMEAN1CJ)=YMEAN
      STDEV1CJ)=STD
C
C

                             148

-------
C   CALCULATES THE  CONCENTRATION AT THE MEAN Y-DI STANCE
C
C
      CONMAXC J>=CFREQ*CI >/CSTD*SQ2PI )
      VAR = STD**2
      DO  40 I = 1,N1
      WRI TEC 6,8 00) MCI), SC I ) , DC I > , CC I ) , C 1 C I )
   40 CONTINUE
      WRI TEC 6*81 1)FREQ
      WRITEC6,812)YMEAN, STD, VAR
      WRITEC6>814>
C
C
C   CALCULATES THE  GAUSSIAN DISTRIBUTIONS USING  THE ABOVE
C   INFORMATION  AND THE COMPARES THE CALCULATED  VALUE
C   WITH  THE OBSERVED VALUE AND CALCULATES THE
C   SQUARED DIFFERENCE.
      DO  60 I =  1,N1
      X3CI) = DCI>  -YMEAN
      CALC2CI) = C(FREQ*CI >/ C STD+SQ2PI » *EXPC -CX3C I ) **2> /C 2.
      $*STD**2) )
      DIF2CI) =  **2
      DIF22 = DIF22 + DIF2CI)
      VRI TE( 6, 8 1 3) DC I ) ,X3C I ) > C 1 C I ) , C ALC2C I ) , DI F2C I )
   60 CONTINUE
      VRITEC6*815)DIF22
      NPTS  = Nl
      XMAX  = 60-00
 C
 £
 C     XMAX  IS A  SCALING FACTOR  FOR THE PLOT  IN  SUB-
 C      ROUTINE PLOT2D.IT SHOULD  HOWEVER BE  SOME
 C      INTEGER VALUE OF  12  SINCE 120 COLUMNS  ARE
 C      USED  IN THE OUTPUT
       WRITEC6*957MRUNNO,FROUDEJ AK*XDISTC J)
       CALL  FLO T2D< NPTS* XMAX, D*Cl*CfiLC2)
    10 CONTINUE
       WRITEC6,958)                                ^Amj
 C     WRITES OUT  THE PERTINENT INFORMATION  FOR  EACH
 C     CROSS- SECTION, X- DISTANCE,  MEAN Y- DISTANCE, MAXI-
 C     MUM  CONCENTRATION,  AND THE STANDARD DEVIATION

       WRI TEC 6,959 )XDI STC J3 , YMEAN 1 C J5 , CONMAXC J> , STDEV1 C J)


                                                        , I 2/T 1 1,  0
   801 FORMATCT11,'NO.  OF OBSERVATIONS '

                        , F5-2,'  CENTIMETERS VT11, -DIAMETER OF
      SETS T33, '= '>
      &T36,F5-2,  ' CENTIMETERS1///)
                                149

-------
802 FORMATC1H1/////T33*'RUN NO.  «*I2///)
811 FORMATCT58*'	'/T46*'SUMMATION =  SF1I.7)
812 FORMATC T26* 'Y-MEANST48* '= »*T52>F8-4/T26* 'STANDARD DEV
   $IATION'*T48
   ** '='*T54*F8.6/T26* 'VARIANCE'*T48* •« '*T54*F8«6)
8 13 FORMATC T9*F6. 3* T23*F6. 3*T35*F1 1. 1, T49*F1 1. 7* T73* Fl 1 . 7)
814 FORMATC1H3T9*'OBSERVED'*T22*'DISTANCE'*T37*'OBSERVED',
   ST51*'CALCULA
   ATEDST72* ' DI FFERENCE VT1 1* 'DEPTH '*T24* 'FROM'*
   4T34, fCONCENTRATION',T53* •SQUARED VT11> 'CCM) S T24*
   A'MEAN',T38* 'X100'*T54* fX100'jT72* 'FOR NORMAL '/
   &T24*•CCM>'*T51-'NORMAL CURVE*,T75*'CURVE')
815 FORMATCT73*'	VT60* 'SUMMATION  =  '*
   *T76*F1 1-7)
900 FORMATCI5*F5.2»F5.3)
901 FORMATC1H1/////T26*'CROSS-SECTION NO.  '*I2//T4,'DISTAN
   $CE FROM DISC
   4HARGE POINT '>T45*'='*T48*F6.2*'  CENTIMETERS'/T4,'X/D
   S(NO.  OF JET
   ADI AMETERS  DOWNSTREAM) '*T45> '= ST48*F6-2>
902 FORMATCT27* *Y- DISTANCE VT28> 'FROM JET V IX* 'M1LLI VOLT',
   $T14>'DILUTIO
   *N'*T28* 'DISCHARGE'*T41* 'CONCENTRATI ON'*T5S* 'CONCENTHAT
   JION*
   S/T30* 'POINT'* T62* 'X100VT30* '(CM) ')
903 FORMAT(I2*4F5.3)
907 FORMATCT7* 'POINT GAGE TOP  OF WATER'*T35* '='*T39*F5.3*T
   $47* 'FEETVT7
   ft*'POINT GAGE BOTTOM OF WATER'*T35*'='*T39*F5.3*T47*'FE
   $ET'/T7*'PROB
   &E TOP OF WATER'*T35* '=**T39*F5.3* T47,'FEET VT7* 'DEPTH
   JOF WATER**T3
   A5* '='*T39*F5.3*T47* 'FEETVT7, 'DEPTH INCREMENT'* T35* '= '
   $*T39*F5-3*T4
   A7^ 'FEET VT7j. 'LOCATION OF JET ABOVE VT7* 'BOTTOM  OF FLUM
   $E',T35* '='*T
   A39*F5.3*T47*'FEET'///>
910 FORMATCI5*F5.2*F5-3/C6F10.2))
912 FORMATC2F5.2)
920 FORMATCF5.2)
950 FORMATC2F15.9)
957 FORMATC1HIT53*'RUN  NO. =  '*T64*I2//T50*'FROUDE NO. =  '
   $*T64*F5.2//T
   A50*'VELOCITY'/T52*'RATIOCK)  =  '*T64*F5.2//T50* 'X-DISTA
   $NCE  ='»T63*F
   A8.4*T72*•CENTIMETERS'////)
958 FORMATC 1H1T3* 'X-DI STANCE'* T20* 'Y-DI STANCE'* T41 * 'MAXIMU
   $M'*T54* 'STAN
   ADARDVT39* 'CALCULATED'* T54* 'DEVIATION'/
   AT38* 'CONCENTRATIONVT2* '< CENTIMETERS) '*T21* 'CMEAN DI ST
   $.) V
                            150

-------
     &T20*'(CENTIMETERS)'//)
  959  FORMATCT5* F8 • 4, T22., F 7. 3.. T4 1 * F9 . 5* T56-. F 7. 4)
       END
C
c
C
C   PLOTS THE OBSERVED AMD  CALCULATED VALUES ON THE  SAME
C   PLOT
C
C
       SUBROUTINE PLOT2DCNPTS* KM AX.. D* Cl* CALC2)
       INTEGER BLANK/'  '/*O/ '0 ' /*G/*G'/
       INTEGER BAR/"C•/>MINUS/'-'/
       DIMENSION  D< 100)*C1C100)>CALC2(100)*LINE(121)*XTEMPC13
     $)
       DELTY = D(1>-DC2>
       DO 100 I=1^NPTS
       DO 101 J=1 * 121
   101  LINE(J)=BLANK
       IFCABS(D(I)>   .GT.  DELTY/2.5GO  TO 200
       DO 102 K=l,121
   102  LINECK)=MINUS
   200  DO 104 J=1>I2l> 10
   104  LINECJ)=BAR
       N=«CALC2CI )*120«>/XMAX)+ 1.5
       LINECN)=G
       N=CCC1CI )*120.)/XMAX)+l-5
       LINECN>=0
   100  WRITE(6^900)DCI )*LINE
       X = 0-0
       DELTX=XMAX/12.
       DO 103  J=l*13
       XTEMP(J)=X
   103  X=X+DELTX
       WRITEC 6^901)XTEMP
   900  FORMATCIX*F9.4^2X>121A15
   901  FORMATC 12X,13C 'C %9X)/Tl1, 13(F6.2
       RETURN
       END
                               151

-------
              APPENDIX H
OBSERVED VALUES AND THEORETICAL CURVES
PREDICTED BY FAN'S AND ABRAHAM'S MODEL

-------
30
20
        K>    20    30    40
60    70    80    90    IOO
                                                            1000
                                s/D
         FIGURE 45  - OBSERVED VALUES  AND THEORETICAL  CURVES
                PREDICTED BY FAN'S MODEL - RUN NO.  18
                            152

-------
60



40
20
   -o
  0
40
80          120
      x/D
160
                                 8/D
200
         FIGURE 46 - OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED BY FAN'S MODEL - RUN NO.  13
                             153

-------
60
40
20
              40
80          120
      x/D
160
200
       F   =21.1
           =5.3
       ff. * 90
                                                            1000
        FIGURE  47 - OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED BY FAN'S MODEL - RUN NO.  12
                            154

-------
60



40
                                      T~   "I	1	1	r
20
                                            J	L
  0
 100
40

                                 ,n
                               x/D
160
200
                   1111
       F  =20.4
       K. * 10.2
       ft s 78.0
       a  «0.5
  10
         DILUTION
                             I -- 1 - 1 — [MM
                    0 0
                                    1
                                  i   I   I   1 11 r

                                           y
                                               ,    ,  ,   ,  , M
                      10                 100
                               8/D
                                             1000
        FIGURE 48 - OBSERVED VALUES AND THEORETICAL CURVES

              PREDICTED  BY FAN'S MODEL - RUN NO.  11
                            155

-------
30
  TO     K>    20
30    40   50   60
           x/D
70    80   90   IOO
  100
       F  *20.2
       K  -20.2
                                                          BOO
                               3/D
         FIGURE 49  - OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED BY FAN'S MODEL - RUN NO. 16
                            156

-------
30
        K>    20
30    40    5O    60
            x/D
70    80    90    100
  100
       F  « 10.9
     UK  -5.5
             *>•
                                                            1000
           FIGURE  50  - OBSERVED VALUES AND THEORETICAL CURVES
                  PREDICTED BY FAN'S MODEL - RUN NO.  10
                               157

-------
100
10
1
• T 1 1 | 1 1 1 1
F «I0.3
LK «io.3
. 0;«90°
£*«74.0°
a «0.4
: o
— 1 	 1 	 1 [Mil
y
DILUTION"-/
o0/
o 7
•—Tar-"** A
A
—I— TT
i i i 1 i i ij^
10 100 KX
X)
                          s/D
FIGURE 51 - OBSERVED VALUES AND THEORETICAL  CURVES
       PREDICTED BY FAN'S MODEL -  RUN NO.  9
                      158

-------
                                x/D
100
 10
               ~~  T
   -  F  « 10.1
   Z- K  * 20.1
       0* 81.8°
      a  «0.2
         1 _ 1 _ 1  1 III!
    1 -- 1 - 1 |  1 1 1 1
DILUTION-;
                      10
                                                              50
                                                1   1   1  | 1 1 1 L
                        i   i   i   i i i i
                               s/D
100
                                   1000
       FIGURE 52 - OBSERVED VALUES AND THEORETICAL  CURVES
              PREDICTED BY FAN'S MODEL - RUN NO.  15
                           159

-------
30
  0     10    20    30    40    50    60    70    80    90    100
 100
                                                           KXX>
         FIGURE  53 - OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY FAN'S MODEL - RUN NO.  34
                             160

-------
                         80          120
                               x/0
KX>
200
      F  «42.7
      K« 10.6
         • 60°
           53.8°
      a  -0.3
                               3/D
                                                           KXX>
         FIGURE 54 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL -  RUN NO. 33
                            161

-------
30
                    1	r
                                      —   —I	1	r
        10   20   30   4O   50   60   70    8O    9O    IOO
                               X/D
       F  «22.5
       K  « 5.6
          •60°
       A, « 52.0°
       a  «0.4
                               s/D
                                         100
KXX>
         FIGURE 55 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL -  RUN NO. 32
                             162

-------
60
                                       T~   	1	1	T
40
2O
              o  o
 100
                                                   160
ZOO
       F   =21.3
       K= 10.6
                                                             1000
                                 s/D
         FIGURE  56  - OBSERVED VALUES AND THEORETICAL CURVES
                PREDICTED BY FAN'S MODEL -  RUN NO. 28
                              163

-------
30
                                                        90    100
  100
    [—

 F  «20.6
. K  • 20.6
           . 53 0«
        a  «0.2
                                 s/D
           FIGURE 57 - OBSERVED VALUES AND THEORETICAL CURVES
                 PREDICTED BY FAN'S MODEL - RUN NO.  30
                               164

-------
  30
                      T	1	1	T
Q
>,
20



 10
                                        60    70    80    90    100
                                  s/D
                                                              1000
           FIGURE  58  - OBSERVED VALUES AND THEORETICAL CURVES

                  PREDICTED BY FAN'S MODEL - RUN NO.  27
                               165

-------
K>
 100
  10
I -- 1 — I
       F  «I0.6
     IK  «I0.6
            520-
            0.3
-i—i—i  i  i 111
                DILUTION 7
                           b/De
                                     1_L_L
T -- 1
                       10
                                3/D
                                100
                                                   U
                    I    «  I  I I  I 11
                                1000
         FIGURE 59 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL - RUN NO. 29
                              166

-------
10

 5
              K>
100
20
                                  ,r.
                                x/D
40
50
                                                             1000
         FIGURE 60 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL -  RUN NO.  31
                             167

-------
30
20
 10
        -i	1	1	1	1	1      i	r
               j - 1
                                 -I	L
"90    IOC
         _
        K>    20    3O    40
50
x/D
60    70    80
 lOO
       F  " 46.9
     IK  «5.9
     - A-30LO-
       a  «0.3
                                3/D
        FIGURE 61 - OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED  BY FAN'S MODEL -  RUN NO. 26
                              168

-------
K>    2O    30    40    50   60   70   80   90   TOO
                         s/D
  FIGURE  62  - OBSERVED VALUES  AND THEORETICAL CURVES
         PREDICTED BY FAN'S MODEL - RUN NO.  22

-------
60


40
        -i	r
                                                          -T—   —r
20
                                                               o  .
       F  = 23.5
       K  =5.9
                                 8/D
                                                             KXX)
        FIGURE 63 - OBSERVED  VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL - RUN NO. 21
                             170

-------
60
        T	T
                                 I	T
40
20
80          120
      x/D
                                                   160
200
 lOOc
       F   =21.6
      . K   = 10.8
                                 3/D
          FIGURE 64  - OBSERVED VALUES AND THEORETICAL CURVES

                 PREDICTED BY FAN'S MODEL -  RUN NO. 20
                              171

-------
30
                                      60    70   80    90   100
K)    20    30    40
  100
                                                            KXX>
                                 3/D
        FIGURE 65  - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL - RUN NO.  24
                             172

-------
30
20
 10
        K>    20    30    40   50    60    70   80   90    100
  100
                                 3/D
          FIGURE 66  -  OBSERVED VALUES AND  THEORETICAL CURVES
                 PREDICTED BY FAN'S MODEL  - RUN NO. 19
                              173

-------
10
              K)           20          30          40          50
                                x/D
                                                           KXX)
       FIGURE 67 - OBSERVED VALUES AND THEORETICAL  CURVES

              PREDICTED BY FAN'S MODEL - RUN NO.  23
                            174

-------
15
K>
              10
20    x/D   3°
        40
50
100
  10
             1  1    1
      F   * 10.4
      K   =20.9
          = 29.3°
      a   « 0.15
                             J	I	L
                      10
                               ./o
                                                1    1  1   1  1 1 1
100
                                  IOOO
        FIGURE  68 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY FAN'S MODEL - RUN NO. 25
                            175

-------
30
        10    20   30   40    50    60   70   80   90   100
 too
       F  *40.0
       K  -5.0
  K>
DILUTION
                      o


                    #
                     A

                   ^/
               ,
                                  I
                                               T   I  I
                      10                 100
                              8'/0
                                                 1000
        FIGURE 69 - OBSERVED VALUES AND THEORETICAL CURVES
             PREDICTED BY ABRAHAM'S MODEL -  RUN NO. 18
                            176

-------
60
40
20
                                       ~i	r
                 •o-
                                   o
              45-
                                x'/D
120
160
200
100
                               s'/D
                                          IOO
                     1000
         FIGURE 70 - OBSERVED VALUES  AND THEORETICAL CURVES
              PREDICTED BY ABRAHAM'S  MODEL - RUN NO. 13
                             177

-------
 6O-





 40

>
h
S

 20
                                              T	1	1	r
               o  o
                                 x'/D
                                        20
160         200
 100
       F  «20.0
       K  -5.0
                                                            1000
       FIGURE 71 -• OBSERVED VALUES AND THEORETICAL CURVES

            PREDICTED BY ABRAHAM'S MODEL - RUN NO.  12
                            178

-------
  60
                                               T	1	1	r
  40
§
  20
                           ~SO~        120         160         200
                                  x'/D
        F  « 20.0
        K , - 10.0
          ' * 90°
                                s'/D
                                           100
1000
         FIGURE  72 - OBSERVED VALUES AND THEORETICAL CURVES
              PREDICTED BY ABRAHAM'S MODEL - RUN NO. 11
                              179

-------
30
        10    20   30   40   50   60    70    80    90    100
100
      F
      K
  10
^  "~
20.0
20.0
90°
                           DILUTION
                                     I I I I
                                               I	I
                     10                 100
                              s'/D
                                                         ITT
                                               1000
        FIGURE 73 - OBSERVED VALUES AND THEORETICAL CURVES

            PREDICTED BY ABRAHAM'S MODEL - RUN NO. 16
                           180

-------
30
        10    20    30    40    50   6O   70   80    90    100
           DILUTION
                                                            1000
                                s'/D
        FIGURE  74 - OBSERVED VALUES AND THEORETICAL CURVES
             PREDICTED BY ABRAHAM'S MODEL - RUN NO. 10
                            181

-------
15
              10
20
                                x'/D
30
100
4O
50
                                                            1000
                               S'/D
           FIGURE 75 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY ABRAHAM'S MODEL  - RUN NO. 9
                               182

-------
15

                               KVD
100
                  1  1 II
      F  * 10.0
      K  =20.0

         • 90°
  10
                            I -- 1 - 1   MM
                        DILUTION ?

                         /a
                                                 40
                                                             50
                                                           1000
        FIGURE  76 - OBSERVED VALUES AND THEORETICAL CURVES

             PREDICTED BY ABRAHAM'S MODEL - RUN NO. 15
                             183

-------
30
20
 10
                                 i      i
o     o
        K>    20    30   4O   50    60    70    80    90    IOO
                                X'/O
       F  «44.0
       K  -5.5
          " 60°
                               s'/D
       FIGURE 77 - OBSERVED VALUES AND THEORETICAL CURVES
            PREDICTED BY ABRAHAM'S MODEL  - RUN NO. 34
                            184

-------
  60
  40
§
                                               T	T

                                  ^^ * L^
                                                               :
                                                  o
        F  -42.0
        K  -10.5
                                 s'/D
                                           IOO
1000
          FIGURE 78 - OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED BY ABRAHAM'S MODEL - RUN NO. 33
                              185

-------
  30
I
  20
  10
          10    2O    3O    4O
5O
x'/O
60    70    80    90    100
                                  s'/D
                                            100
                             1000
         FIGURE 79 - OBSERVED VALUES AND THEORETICAL CURVES

              PREDICTED BY  ABRAHAM'S MODEL - RUN NO. 32
                              186

-------
60
                                 x'/D
                                                                200
                                                             1000
         FIGURE 80 - OBSERVED VALUES AND THEORETICAL CURVES
             PREDICTED BY ABRAHAM'S MODEL - RUN NO. 28
                              187

-------
30
        10    20   30   40   50    60    70    80    90    100
             20.5
             20.5
             60°
          DILUTION
                               s'/D
                                         IOO
IOOO
      FIGURE 81  - OBSERVED VALUES AND THEORETICAL CURVES

           PREDICTED BY ABRAHAM'S MODEL -  RUN NO. 30
                          188

-------
15-
              10
       20    ,^  3°
             x'/D
100
 10
MM
      F  « 11.0
    1 K  « 5.5
      DILUTION
                             1 -- 1 - 1   MM
    10
                             i   i   i i  i i i i
         40
50
                             I   I' T  { 1  1 1 I.
                               s'/D
100
                                                           1000
          FIGURE 82 -  OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED  BY ABRAHAM'S MODEL -  RUN NO. 27
                             189

-------
                                x-/D   *>
50
lOOr
                               S'/D
         FIGURE 83 - OBSERVED VALUES AND THEORETICAL CURVES
              PREDICTED BY ABRAHAM'S MODEL  - RUN NO. 29
                              190

-------
15


10
             10
100
  10

      F   «I0.3
      K   -20.5
          - 60*
        DILUTION
20
                               x'/D
30
                                  I  I I  I 11
                      10
                              s'/D
               100
40
50
                     1000
        FIGURE 84 - OBSERVED VALUES AND THEORETICAL CURVES
            PREDICTED BY ABRAHAM'S MODEL - RUN NO. 31
                            191

-------
 O    10    20    30   40   50    60    70    80    90    100
                               X'/D
100
                               87D
        FIGURE 85  -  OBSERVED VALUES AND  THEORETICAL CURVES
             PREDICTED  BY ABRAHAM'S MODEL  - RUN NO. 26
                            192

-------
30
20
 10
  "0 ~   10    20   3O    40    50    60    70    80    9O    100
                                 X'/D
                                                             1000
         FIGURE 86 - OBSERVED VALUES AND THEORETICAL CURVES
             PREDICTED BY ABRAHAM'S MODEL - RUN NO. 22
                            193

-------
  60
                                          T— 	T
  4O
§
>»
  20
                              1_	l_

                                                      160
200
                                                              1000
           FIGURE 87 - OBSERVED VALUES AND THEORETICAL CURVES

                PREDICTED BY ABRAHAM'S MODEL - RUN NO. 21
                                194

-------
60

40
              40
                                x-/o
160
200
     -   DILUTION
                               S'/D
                                         100
         1000
        FIGURE 88 - OBSERVED VALUES AND THEORETICAL CURVES
            PREDICTED BY ABRAHAM'S MODEL - RUN NO. 20
                            195

-------
 30




 20
I
*
•»

 10
  10     10    20    30    40    50    60    70    80    9O    100

                                 x'/D
  100
                                                             1000
          FIGURE 89  -  OBSERVED VALUES AND THEORETICAL CURVES

               PREDICTED BY ABRAHAM'S MODEL  - RUN NO. 24
                              196

-------
3O
20
 10
                                                               1000
          FIGURE 90 - OBSERVED VALUES AND THEORETICAL  CURVES
               PREDICTED BY ABRAHAM'S MODEL - RUN NO.  19
                                197

-------
15
10
             10
                               x'/D
                               A f IS
100

4O
50
      F  -I0.7
      K  -I0.7
         ' • 45°
                                                           1000
                               f'/D
          FIGURE 91 - OBSERVED VALUES AND THEORETICAL CURVES
               PREDICTED BY ABRAHAM'S MODEL -  RUN NO. 23
                              198

-------
15
100
  10
T	1—r~ I  i i 11
       F   -10.4
     _K   -20.7
    -  DILUTION
          i    '	L
                                      /
                                                                   50
                                                          I  [MIL
                        10
                                 s'/D
                                   100
1000
          FIGURE  92  -  OBSERVED VALUES AND THEORETICAL  CURVES
                PREDICTED BY ABRAHAM'S MODEL  -  RUN NO.  25
 . GOVERNMENT PBINTING OfFICE:1974 546-319/385
                                 199

-------
SELECTED WATER
RESOURCES ABSTRACTS

INPUT TRANSACTION FORM
                                             Kt-
                                                               w
           Negatively  Buoyant  Jets  in  a Cross  Flow
     Anderson, J.L.,  Parker,  F.L., &  Benedict,  B.A.
                                                                ".  P-  'iroth  "~JZ"C •
                                                                  A  ~-rtK,..
                                                                   16130 FDQ
     Vanderbilt University
     Environmental & Water Resources Engineering
     Nashville, Tennessee
i?  ;v°'--"-W3 r-MI*  ti     Environmental Protection Agency
     Environmental Protection Agency No.  EPA-660/2-73-012,  October 1973,
                                                                              nd
            200 words

Modification of Fan's and Abrahams jet diffusion models were  used  to  predict  the
trajectory and dilution of a negatively buoyant jet.  Such jets can occur  when  a
chemical waste is discharged into a less dense ambient water  or when  cool, hypo-
limnetic water is used for condenser cooling water and discharged  into  less dense
surface waters, then a sinking jet would result.  Experimental investigations
were conducted involving different combinations of densimetric Froude number, velocity
ratio, and initial angle of discharge.  Salt was used as  the  tracer,  yielding a fluid
that was denser than the ambient receiving water and facilitated measuring concentratioi
profiles of the jet plume.  The experimental data was then fitted  with  predicted jet
dilution, trajectory, and cross sectional values for each model.   The values of the
entrainment coefficient were chosen as the one which best fit the  experimental
data for the particular combination of densimetric Froude number,  velocity ratio,
and initial angle of discharge.  The value of the drag coefficient was  chosen as
zero for both models since any other value would predict a trajectory whose rise
would be less than experimentally observed.  Typically, for all angles  of  discharge
the value of entrainment increased with a decrease in the velocity ratio and with an
increase in densimetric Froude number.  Additionally, there was a  marked decrease
in the entrainment coefficient with a decrease in the initial angle of  discharge.
17a. Descriptors

Thermal Pollution, Thermal Power Plants, Density Currents, Entrainment
17h. Identifiers
Jet Discharge, Near Field, Jet Trajectory, Negative Buoyant Jet, Waste Dilution,
Densimetric Froude Number
                      05B
                        9,  K   '.city   -ss..
                       20.  Sccur  y Cfass.
                            '
                                        ?.}.
                                              . of
                                                    Send To:
                                                   WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                                   U.S. DEPARTMENT OF THE INTERIOR
                                                   WASHINGTON. D. C. 2O24O
          Dr.  Frank L. Parker
                                               Vanderbilt University, Nashville, Tenn.

-------