yt WATER POLLUTION CONTROL RESEARCH SERIES • 16130 DJU 02/71
               AN ANALYTICAL
      AND EXPERIMENTAL INVESTIGATION
  OF SURFACE DISCHARGES OF HEATED WATER
ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE

-------
          WATER POLLUTION CONTROL RESEARCH SERIES

The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollu-
tion of our Nation's waters.  They provide a central source
of information on the research, development, and demon-
stration activities of the Water Quality Office, Environ-
mental Protection Agency, through inhouse research and grants
and contracts with Federal, State, and local agencies, re-
search institutions, and industrial organizations.

Inquiries pertaining to the Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Development, Water Quality
Office, Environmental Protection Agency, Washington, D.C. 20242,

-------
AN ANALYTICAL AND EXPERIMENTAL INVESTIGATION

   OF SURFACE DISCHARGES OF  HEATED WATER
                     by



            Keith D. Stolzenbach


                    and


            Donald R. F. Harleman
         RALPH M PARSONS LABORATORY

   FOR WATER RESOURCES AND HYDRODYNAMICS

      Department of Civil  Engineering

   Massachusetts Institute of Technology

      Cambridge, Massachusetts  02139
                   for  the


            WATER QUALITY OFFICE

       ENVIRONMENTAL PROTECTION AGENCY


         Research Grant No.  16130 DJU


               February, 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office
            Washington, D.C., 20402 - Price $1.75

-------
                 EPA Review Notice
This report has been reviewed by the Water Quality Office,
EPA, and approved for publication.  Approval does not signi-
fy that the contents necessarily reflect the views and poli-
cies of the Environmental Protection Agency, nor does mention
of trade names or commercial products constitute endorsement
or recommendation for use.

-------
                                ABSTRACT
         The temperature distribution, induced in an ambient body of water by a
surface discharge of heated condenser cooling water must be determined for eval-
uation of thermal effects upon the natural environment, for prevention of recircu-
lation of the heated discharge into the cooling water intake, for improved design
of laboratory scale models and for insuring that discharge configurations meet
legal temperature regulations.

         An analytical and experimental study of discharges of heated water is
conducted.  The discharge is a horizontal, rectangular open channel at the sur-
face of a large ambient body of water which may have a bottom slope or a cross
flow at right angles to the discharge.  Of interest is the dependence of the
temperature distribution in the receiving water as a function of the initial
temperature difference between the heated discharge and the ambient water, the
initial discharge velocity, the geometry of  the discharge channel, the bottom
slope, the  ambient  cross flow, and the  transfer of heat to the atmosphere  through
the water surface.

          The  theoretical development  assumes  that  the  discharge  is a three-
dimensional turbulent jet with an unsheared  initial core and a turbulent region
in which  the  velocity and  temperature distributions are related  to centerline
values by similarity functions.   Horizontal  and vertical entrainment of  ambient
water  into  the  jet  is proportional to the  jet centerline velocity by an  entrain-
ment  coefficient.   The  vertical  entrainment  is a  function  of the local vertical
stability of the jet and  thci  buoyancy of  the discharge increases lateral spreading.
A cross  flow deflects  the  jet by entrainment of  lateral momentum and a bottom slope
 inhibits  vertical entrainment and buoyant  lateral spreading.

          Experiments are performed  in a laboratory basin  in  which  all of the  rele-
vant  parameters, including the cross flow and the bottom  slope,  are varied and
 three-dimensional temperature measurements are taken  in the  heated discharge.

          The experiments verify that the  theoretical  model predicts  the  behavior
 of heated discharges.   The theory contains no undetermined parameters and  the
 comparison of the experimental and theoretical results does  not  involve  any  fitt-
 ing of the theory to data.   The rate of temperature decrease in  the jet  and  the
 vertical and lateral spreading are controlled by the  initial densimetric Froude
 number,  the ratio of channel depth to width, and the  bottom slope.  A cross  flow
 deflects the jet but does not greatly affect the temperature distribution.  Heat
 loss does not significantly affect the temperature distribution  in the  heated
 discharge within the region treated  by the theory.

          Application of the theory to prediction of temperatures in an  actual
 heated discharge is possible if the  temperatures, velocities,  and the  geometry
 of the discharge may be schematized by representative steady state temperatures
                                   -2-

-------
and velocities and by an equivalent rectangular channel.  If a model study is
necessary the theory indicates that temperature similarity requires an undistor-
ted model.

         The theoretical model developed in this study may be extended by treating
a stratified ambient condition, by considering recirculation of the heated jet in
a finite enclosure and by development of a theory for the transition of the heated
discharge jet into a buoyant plume.

         This report was submitted in fulfillment of Research Grant No. 16130 DJU
between the Water Quality Office, Environmental Protection Agency and the Massachusetts
Institute of Technology.  .
                                                            /
Key words:  thermal pollution; turbulent, buoyant jets; stratified flow; temperature
            prediction; heated surface discharges.
                                   -3-

-------
                            ACKNOWLEDGEMENT







        This investigation was supported by the Water Quality Office,




Environmental Protection Agency, under Research Grant No. 16130 DJU as part




of a research program entitled "Prediction and Control of Thermal Pollution".




The project officer was Mr. Frank Rainwater, Chief, National Thermal Pollution




Research Program, Pacific Northwest Water Laboratory, at Corvallis, Oregon.




The cooperation of Mr. Rainwater is gratefully acknowledged.




        Stone and Webster Engineering Corporation of Boston, Massachusetts, sup-




ported the purchase of experimental equipment in conjunction with another research




project.




        The research program was administered at M.I.T. under DSR 71335 and 72324.




The staff and students at the Ralph M. Parsons Laboratory for Water Resources and




Hydrodynamics gave valuable suggestions throughout the study.  The experimental




work could not have been performed without the particular help of Mr. Umit Unluata




Miss Alician Quinlan, Mr. Edward McCaffrey and Mr. Roy Milley.  Typing was done




by Miss Kathleen Emperor.  All computer work was done at the M.I.T. Information




Processing Center.




        The material contained in this report was submitted by Mr. Stolzenbach




in partial fulfillment of the requirements for the degree of Doctor of Philosophy




at M.I.T.
                                 -4-

-------
                        TABLE OF CONTENTS






                                                                  Page




ABSTRACT                                                            2




ACKNOWLEDGEMENT                                                     4




TABLE OF CONTENTS                                                   5




I.   INTRODUCTION                                                   9




     1.1  Why Heated Discharges?                                    9




     1.2  Effects of Heated Discharges                             10




     1.3  Temperature Prediction                                   12




     1.4  Objectives of this Study                                 14




     1.5  Summary of the Study                                     17




II.  REVIEW OF PREVIOUS INVESTIGATIONS                             18




     2.1  Scope of the Review                                      18




     2.2  Turbulent Jets                                           18




          2.2.1  Physical Properties                               18




          2.2.2  Analytical Solutions                              22




          2.2.3  Deflected Jets                                    27




          2.2.4  Confined Jets                                     27




     2.3  Turbulent Flows with Density Gradients                   28




          2.3.1  Reduction of Turbulence                           28




          2.3.2  Submerged Buoyant Jets                            32




          2.3.3  Two Layer Flow                                    32




     2.4  Surface Heat Exchange                                    33




     2.5  Surface Discharge of Heated Water                        35
                             -5-

-------
                                                                    Page



III.  THEORETICAL CONSIDERATIONS                                     A2




     3.1  Statement of the Problem                                   42




     3.2  The Governing Equations                                    44




          3.2.1  Basic Equations                                     44




          3.2.2  Steady, Turbulent Flow Equations                    45




          3.2.3  Hydrostatic Pressure                                47




          3.2.4  Scale Variables                                     4g




          3.2.5  Large Reynolds Numbers                              40




          3.2.6  Small Density Differences                           5Q




          3.2.7  Basic Scale Relationships                           5^




     3.3  Non-Buoyant Surface Jets                                   52




          3.3.1  Governing Equations                                 52




          3.3.2  Structure of the Jet                                54




          3.3.3  Boundary Conditions                                 58




          3.3.4  Integration of the Equations                        ^




          3.3.5  The Solution for Non-Buoyant Jets                   55




     3.4  Buoyant Jets                                               7^




          3.4.1  Governing Equations                                 71




          3.4.2  Structure of the Jet                                73




          3.4.3  The y Momentum Equation                             75




          3.4.4  Vertical Entrainment                                73




          3.4.5  Integration of the Equations                        80




     3.5   The Deflected Jet                                          80




          3.5.1  The Governing Equations                             80




     3.6   Bottom  Slopes                                              90




     3.7   Solution  Methods                                           96



                              -6-

-------
                                                                   Page
IV.  EXPERIMENTAL EQUIPMENT AND PROCEDURE                           97
     4.1  Previous Experiments                                      97
     4.2  Experimental Set-up                                      101
     4.3  Experimental Procedure                                   105
     4.4  Data Analysis                                            106
V.  DISCUSSION OF RESULTS                                          112
     5.1  The Controlling Parameters                               112
     5.2  An Example of a Theoretical Computation                  114
     5.3  Comparison of Theory and Experiment                      118
     5.4  Comparison of Theory with Experimental Data              135
          by other Investigators
     5.5  Summary of Jet Properties                                143
VI.  APPLYING THE THEORY                                           152
     6.1  Prototype Temperature Prediction                         152
     6.2  Modeling Heated Discharges                               158
     6.3  A Scale Model Case Study                                 162
VII. SUMMARY AND CONCLUSIONS                                       169
     7.1  Objectives                                               169
     7.2  Theory and Experiments                                   169
     7.3  Results                                                  171
     7.4  Future Work                                              174
BIBLIOGRAPHY                                                       176
LIST OF FIGURES AND TABLES                                         179
LIST OF SYMBOLS                                                     182
                               -7-

-------
                                                                   Page
APl'KNULX I.   The Computer Program for Theoretical                 187
              Calculation of Heated Discharge Behavior
APPENDIX II.  Deflected Jet in a Cross Flow:  Integration          208
              of the x and y Momentum Equations Over the
              Entire Jet
                            -8-

-------
I.   Introduction




1.1  Why Heated Discharges?




        The discharge of heated water into rivers, lakes and coastal




regions has its roots in the growth of demand for electric power and




in the economics of power production.  The demand for power generating




capacity in the United States is expected to continue doubling each




decade.  As hydro power plant sites are developed to their full capa-




city, the demand will be met increasingly by fossil fuel and nuclear




steam plants.  Regional economics and economics of scale result in an




increase in plant size; often fuel costs make nuclear plants more econ-




omic than coal or oil facilities (6 )•




        The heat rejected by a steam power plant is related to the




thermodynamic efficiency of the power production process.  Present lev-




els of efficiency are about 40% for coal and oil plants and about 30%




for nuclear plants.  The nuclear plants reject significantly larger




amounts of heat:  10,000 BTU/kwh vs. 6,000 BTU/kwh (10).  Present lev-




els of efficiency are determined by the economics of power plant design




and are mostly a function of the cost of capital items such as turbines




and condensers.  Water is the only economic cooling fluid for large




power plants and at present the costs of its supply and of the heat dis-




posal system do not greatly influence the choice of economic efficiency.




Significant technological improvements in the efficiency of convention-




al steam cycles or the development of more efficient thermal power gen-




erating methods are not foreseen in the near future  (6 ).




        The total volume of cooling water used for all industries  in
                              -9-

-------
the United States is fifty percent of all water use; the power industry




accounts for all but ten of that fifty percent  0.9).  The large volumes




are dictated by the economic design of the cooling systems which require




low operating temperatures and high flows.




        Thus the amount of heat rejected, the volume of water required,




and the temperature rise in the cooling system necessary to satisfy the




demand for power are determined largely by in-plant technological and




economic considerations not related to the ultimate heat disposal pro-




cess or its effect upon the environment.




        The choice of heat disposal systems for a given power plant site




reflects the relatively fixed constraints on temperatures and flows.




Economic re-use of waste heat for domestic or industrial purposes is




inhibited by the low temperature uses and the large flows dictated by




the plant design.  The rejected heat must in most cases be returned to




the natural environment.  The present methods for heat disposal are




cooling towers, cooling ponds, and once-through cooling.  For a 1000 MW




power plant unit the cost of cooling towers may be $10,000,000 and the




cost cf purchasing land for a cooling pond as high as $5,000,000.  The




once-through cooling method, in which water is withdrawn from an adja-




cent waterway and returned after being heated, is the most economical




assuming the availability of the natural water (6 ).




1.2  Effects of Heated Discharges




        Balancing the economic benefits of once-through cooling is the




impact which the discharge "has upon the natural environment and how




this in turn affects  the power plant directly and the users of the
                             -10-

-------
power plant indirectly.




        The most immediate undesirable effect of the heated discharge




upon the power plant is the recirculation of the heated water into the




intake, creating a feedback of heat which significantly raises the temp-




erature of the cooling water.  The resulting reduction in plant effic-




iency is considered as a cost to the power company.  Prevention of re-




circulation was the major concern in the design of a discharge system




before the era of large nuclear plants and public concern over thermal




effects.  Separation of the intake and discharge structures or the de-




sign of the intake as a selective withdrawal device are the most common




techniques of preventing recirculation.  A knowledge of the flow and




temperature distribution in the receiving waters near the heated dis-




charge is necessary for proper location and design of the intake for




the prevention of recirculation.




        The indirect effects of heated discharges are related to the




cuanges in physical and chemical properties of water resulting from a




temperature rise. As the temperature of water increases the density,




viscosity and ability  to dissolve gases decrease while the vapor pres-




sure of water increases.  The effects of these changes may be striking:




density differences, although small, may produce stable stratified flows




which differ markedly  from natural flow patterns; and evaporative water




losses are increased as vapor pressure rises.  Dissolved oxygen, a prime




indicator of water quality, may decrease.  Biochemical reactions which




consume oxygen and produce undesirable tastes and odors are promoted by




increases in water temperature.  Higher level species such as fish or
                             -11-

-------
 shellfish may not  survive  exposure  to high  temperatures  or may have




 vital  life  cycles  inhibited  by  temperature  rises  (19).




        The  above effects have been  lumped together  under the  title of




 "thermal pollution"  since  the undesirable changes in  the environment




 are  the result of  discharges of heat.   Prediction of  the response of




 the  environment to increases in temperature is  still  at  a very basic




 level;  a fundamental requirement for such an evaluation  is always a




 knowledge of  the temperature distribution induced in  the receiving




 water  by the heated  discharge.




 1.3  Temperature Prediction




        Once-through  cooling  systems are characterized by an intake and




 a  discharge structure.  The  water velocity  at the intake, for a given




 pumping rate, is a function  of  intake design since  converging flow




 structures  are not complicated  by extreme flow  separation.  The minimum




 possible discharge velocity, however, is usually closely related to the




 in-plant velocity  since special  efforts must  be made  to  de-accelerate




 a water flow.  Economic power plant design  velocities are typically of




 the  order of 5 to  10  ft. per second.  Thus  even for a designed area




 increase of ten times the discharge velocity  is near  1 ft. per second.




 This is in  general higher than many ambient flow velocities in rivers




 and  coastal areas.   At the point  of discharge the momentum of the dis-




 charge  is significantly greater  than that of  the receiving fluid and




jet-like mixing of the heated and ambient water occurs.  Ambient water




 is entrained into  the discharge jet and the average temperature and




velocity of the heated flow are reduced.  Further from the discharge
                                 -12-

-------
point the jet velocity is reduced to the magnitude of the ambient cur-




rents.  In this region, entrainment does not occur and the temperature




distribution is determined by natural turbulent diffusion and convec-




tion.  Ultimately, all of the rejected heat is passed to the atmosphere




through the water surface, a process driven by the  elevation of the




water surface temperature.  The density difference between the heated




discharge and the ambient water may inhibit mixing in the initial jet




region and may promote spreading of the discharge as a heated layer.




       The temperature distribution induced by a heated discharge is




thus determined near the outlet by jet-like diffusion and in an outer




region by natural processes.  Surface heat loss and density effects




occur throughout the flow.  The interrelationship  of these processes




is very complex.  Some basic understanding of each phenomena alone




has been achieved:  turbulent jet theory and solutions to convective




diffusion equations are well established and predictions of surface




heat loss rates are now possible.  The effects of stratification are




the least understood.  Analytical predictions of temperatures in the




vicinity of heated discharges are possible under the assumption that




the flow is either a pure jet, a plume, or a stagnant cooling region.




       Often it is not possible to isolate the dominant physical pro-




cess and the flow may be significantly affected by density gradients




or the presence of solid boundaries.  It is customary in such cases  to




construct a physical model of the discharge and ambient waterway.   Un-




fortunately model scaling laws must be based on an understanding of




the very process which the model is supposed to reproduce.   It  is
                               -13-

-------
often impossible, at the present level of knowledge about heated dis-




charges, to build a model which accurately replicates all of the import-




ant features of the temperature distribution.




       For a given cooling water flow the performance of the discharge




structure is a function of its location, size and shape.  Submerged




jets achieve greater mixing than surface discharges.  The presence of




an ambient cross current increases natural heat dispersion while solid




boundaries inhibit mixing.  Several small orifices of high velocity




produce more mixing than a single large one.  The cost of the discharge




structure varies widely from the case of a simple pipe ending at the




water surface to a deeply submerged multiple port diffuser.  Economical




design of the discharge requires the least expensive configuration which




will meet the restrictions imposed on the temperature distribution for




prevention of recirculation and thermal pollution.  In particular fed-




eral and state standards are presently expressed in terms of the per-




missible area in which a certain temperature rise, usually from 1.5 to




5°F, is permitted.  The power company must prove before construction




permits are granted that the design of the discharge structure will




meet the legal standard.  The pressing demand for new power facilities




requires that accurate methods for relating discharge design to _the




temperature distribution induced by the heated discharge be developed.




1.4  Objectives of this Study




       This investigation is motivated by the need, as discussed in




previous sections, for accurate temperature prediction techniques which




consider the effect of power plant condenser water discharge design




upon the temperature distribution in the ambient v;ater.   Prevention




                              -14-

-------
of recirculation, evaluation of thermal effects, construction of reli-




able scale models, and economical design of discharge structures re-




quire an understanding of the physics of heated discharges.




       The following general problem is considered:  A horizontal sur-




face discharge of heated water from a rectangular open channel of arbi-




trary width and depth into a large unstratified body of water which




may be stagnant or flowing with an arbitrary velocity distribution at




right angles to the direction of discharge and which may be infinitely




deep or may have a sloping bottom.  The temperature and velocity dis-




tjribution in the discharge are considered as functions of the discharge




channel geometry, the initial velocity and temperature rise, the mag-




nitude of the ambient cross jlow and bottom slope, and the surfa.ce




heat transfer rate.  The discharge is considered only to the point




where jet-like behavior ceases and natural turbulence and convection




dominate the temperature and velocity distributions.




       Figure 1-1 shows a schematic of the discharge scheme and the




relevant parameters.  This problem is chosen because:




       1) Often the allowable mixing zone, as defined by temperature




          standards, is within the initial jet  region.




       2) Turbulent diffusion, density gradients, and surface




          heat loss may be studied within the defined region.




       3) The number of controlling factors is  relatively  few.




       4) The discharge may be studied on a laboratory scale,




          thereby allowing accurate measurements and close




          control of all parameters.
                                  -15-

-------
                                             Surface Heat  Loss
                    Heated Discharge
Side
  View
                          Turbulent
                        Entrainment
       Discharge
       Channel
Ambient Cross Flow
                         Turbulent
                        Entrainment
                      Heated Discharge
 Top
View
                 Turbulent
                Entrainment
                                                                     Bottom Slope
                  Figure 1-1.  Schematic of Heated Discharge

-------
       The region dominated by natural processes,  commonly called the




"far field", is excluded because of the infinite number of possible




conditions and because of the difficulty in modeling natural phenomena




on a laboratory scale.




1.5  Summary of the Study




       A guiding principle of this investigation is that an increased




understanding of the problem as defined is to be found in a synthesis




of existing knowledge of the physical phenomena of turbulent jets,




stratified flows, and surface heat loss.  A review of the scientific




literature in these fields provides the framework for a three-dimen-




sional theory based on a turbulent jet model with modifications intro-




duced to include density gradients and surface heat loss.  The theory




is developed carefully from the basic governing equations to a set of




integrated equations which may be solved numerically.




       The discharge is modeled in a laboratory basin.  All of the




parameters of interest are varied and three-dimensional temperature




measurements are taken.  Comparisons of theory and experiment verify




the validity of the theoretical model.




       The application of the theory to temperature prediction in actual




power plant discharges is discussed.  The implications of the theoreti-




cal results for the design of scale models are illustrated by a case




study.  Finally, some suggestions for future work are made.
                              -17-

-------
11.  Re\djew__p_f__P_re^/io us  Inves t i ga t ions_




2.1  Scope of the Review




       The literature  in the  fields of turbulence  and  flows  of non-




homogeneous fluids  is  vast.   This  review  discusses those works which




are the major contributors  to the  theoretical  analysis  of  this study.




Analytical and experimental investigations  of  turbulent jets, flows




with stable density gradients, and heat transfer in  turbulent fluids




are examined.   Whenever  possible reference  is  made to  the  best avail-




able summary rather than to original publications.




       Several recent  investigations treat  horizontal surface discharg-




es of heated water.  These efforts are examined.




2.2  Turbulent Jets




       2.2.1  Physical Properties




       Abramovich (l ) and Townsend (34)  discuss the general character-




istics of various kinds  of submerged turbulent jets:   (see Figure 2-1)




       1) A turbulent shear flow region which  increases in size away




          from the jet origin.  There  is  a  distinct  boundary between




          the  turbulent  flow and a potential flow  outside  the jet.




          The  jet width, b, is defined as the mean distance from




          the  jet centerline to this boundary.  In practice, however,




          a half width b, ,„ is often defined as the  point  where the




          mean x velocity, u,  is one half the  centerline mean velo-




          city, u .  For isolated submerged jets both b and b.. ,„ are




          observed to grow linearly with x.  In plane and  axisyrmr.etric




          jets, b = .22  x.
                               -18-

-------
I
l-
                                                  Turbulent Region
                           Figure 2-1.  Structure of a Submerged Turbulent Jet

-------
       2) An initial core  region of unsheared fluid which is gradually




          engulfed by  the  spread of the  turbulent region.




       3) Entrainment  of ambient fluid into the  turbulent region.  The




          entrainment  process is characterized by small scale eddies




          near the jet boundary under the overall control of larger




          scale features such as the mean shear.  The entrained fluid




          velocity normal  to the jet is  v .




       4) Heat, mass, momentum and kinetic energy are transported




          within the jet by convective and diffusive processes.   The




          nature of the transport is different for each quantity.




       5) Similarity of lateral profiles of velocity in the turbulent




          region.  The velocity may be expressed by:








          u  =  uc f(y/b)                                        (2.1)









          where u  is the  centerline velocity and f is a similarity




          function.




          In heated jets,  the lateral temperature profiles are similar




          to each other but have a shape different from the velocity




          profiles.
          AT = AT  t(y/b)                                        (2.2)
          where AT  is the centerline temperature difference and t is




          a similarity function.




       Pearce (26) investigates the dependence of the jet characteris-




tics on the Reynolds number of the jet.   For circular jets the Reynolds



                               -20-

-------
number is
             u d

             -2-°                                                (2.3)
where  u  = initial velocity



       d  = initial diameter
        o


       v  = kinematic viscosity




Pearce summarizes the changes in the structure of a circular water jet



as Uincreases:



       IR < 500 - A long laminar cylindrical column, expanding



       slightly.  Some small stable wave instability but no



       turbulence for x/d  <100.
                         o


       500
-------
           the same magnitude in all directions.

        2)  The magnitudes of the turbulent fluctuations are small

           compared to the maximum mean velocity at any section.

        2.2.2   Analytical Solutions

        The equations governing free turbulent jets are derived by

Schlichting (29).   The assumptions common to all the cases considered

 are:

        1)  Molecular transport terms are negligible.

        2)  Longitudinal diffusion is negligible with respect  to

           longitudinal advection.

        3)  Constant  pressure throughout the jet.

 The equations  expressing conservation of mass, momentum,  and heat  ener-

 gy for  a steady state, two-dimensional flow are:
                        -> ______
               3uv   _   5u'v'
                                                                  .
                                                                  (2<5)
        3Tu      3Tv   _   3v'T'
       "~  +        =  ~                                       <2'6)
where the quantities u'v1  and v 'T' are  the  time  averaged fluxes of mom-

entum and heat in  the y direction due to  turbulent motions.  A similar

set of equations may be formulated for  axisymmetric  flow.

       The analytical solution  of Equations  2.4  - 2.6 proceeds as

follows:
                              -22-

-------
       1)  The velocity  and temperature distributions  are assumed to

          be given by similarity functions f  and t as in Equations

          2.1 and 2.2.   The functions  f and t are not specified, but

          boundary conditions  are stated.   The most common boundary

          conditions are that  the jet  is symmetrical about the center-

          line axis, that the  velocity is zero, the temperature con-

          stant,  and all turbulent transfer is zero at the jet boun-

          dary.

       2)  The turbulent terms  are related to the mean variables.

          Abramovich (  1 ), Pai (24 ), Schlichting (29 ) , and Hinze (17 )

          discuss the well—known assumptions:
          Prandtl's mixing length
          u'v'
         2 9u
           17
3u
ay
u'T' = £  —
          ay
3u
ay
                                             (2.7)
          Taylor's vorticity transfer
 —
u'v
_3_u
ay
                          jhi
                          ay
 u'T1 =  2iT
                                    3y
                       3T
                       ay
                  (2.8)
Reichardt's inductive theory

TTTTr.   , 3(u2)
           ay
          Prandtl's new theory
          u'v'
            9u     -t^rr        3T
                   UT   =   e
                                                                 (2.9)
                                        (2.10)
In the above relationships the quantities X,, e  , and e  are functions
                              -23-

-------
   of x only.  The x dependence of these parameters is related


   to the jet structure on dimensional grounds:





   £  -  b
   ev * GT - Ucb
   Careful experimental investigation has shown that none of


   the local relationships between turbulent transfer and


   mean gradients (Equations 2.7 - 2.10) are valid (34).   For


   many cases,  however, the solutions based on these assumptions


   do predict the gross behavior of the jet reasonably well.


3) The similarity functions (2.1 & 2.2) and the turbulent


   hypotheses (one of Equations 2.7 - 2.10) are substituted


   into the equations of motion (2.4 - 2.6).  For jets, the


   sufficient criteria that unique equations for the similarity


   functions f  and t exist, is  found to be (34):
    L  =  const.   =  e                                    (2.12)
   dx
   where e  is determined later  by comparison of the solution


   with experimental  data.   The functions f and t  are determined


   subject  to the stated boundary conditions.   A particular


   failing  of the mixing length hypothesis is  that it results


   in the same similarity function for both velocity and  temp-


   erature.   Taylor's theory predicts  that heat is diffused


                             -24-

-------
          laterally faster than momentum and that;







          t  =  Jl                                               (2.13)







          Prandtl's new theory gives the same result if e  = 2e  (see




          Figure 2.2) (29)-




       4) The behavior of u  and AT  is determined  by integrating the




          equations of motion (2.4 - 2.6) over the  jet cross section




          to yield expressions of constant momentum and heat flux in




          the x direction.  This completes the analytical solution.




          K property of the solutions for plane and axisymmetric jets




          is that






          ve =  auc                                              (2.14)







          i.e. the entrainment velocity is proportional to the center-




          line velocity with the coefficient of proportionality called




          an entrainment coefficient.




       Analytical solutions which describe the jet behavior as accurat-




ely as those derived by the above process may be obtained by a consid-




erably less complicated procedure (24):




       1) A specific form for the similarity functions, f and t, is




          assumed.




       2) Either the jet is assumed to spread linearly  (Equation 2.12)




          or the entrainment relationship  (2.14) is assumed.  The




          appropriate constant, eor a, is determined empirically.




       3) The integrated equations of motion are used  as  in  (4)  above.




                              -25-

-------
                                            1.0
                                          OCtO-Or
                                        .75 --
                                        .50
                                        .25 --
-2.0
-1.5
-1.0      -.5
.5
1.0
1.5
                                                                                  y/b
2.0
l/2
Figure 2-2.  Velocity and Temperature Distributions in Submerged Jets
     Reichardt's Data:    Velocity    Q
                          Temperature Q
                                                           u/uc = (1 - .29 ;3/2)2
                                                          AT/AT  = 1 - .29

-------
       2.2.3  Deflected Jets



       The case of turbulent jets discharged into a cross flow is dis-



cussed by Abramovich.  The literature on the subject is well reviewed



by Fan (11).  In all the cases considered the jet is deflected by two



processes:



       1) Entrainment of the lateral cross flow momentum.  The result-



          ing force per unit length on the jet in the direction of



          the cross flow is:
                                                                 (2.15)
          where V is the cross flow velocity and q  is the total en-



          trainment per unit length of jet.



       2) A net pressure force caused by eddying of the ambient fluid



          in the lee of the jet and by the distortion of the jet boun-



          daries.  This force per unit length is:




                    V2
          FD = CD p r.  D                                        (2.16)






          where C  is a drag coefficient and D is the jet diameter.



       2.2.4  Confined Jets



       The problem of jets discharging over a solid boundary is re-



lated to diffuser flow as discussed by Schlichting (29).  Abramovich



(1) also discusses the confined half jet.  The effects of the boun-



dary are:




       1) Pressure gradients in the lateral and longitudinal direc-




          tion.  Adverse longitudinal pressure gradients result in
                             -27-

-------
          separation and eddying.  For turbulent flows the point of



          separation is determined only empirically.



       2) A viscous boundary layer near the solid boundary.  Longi-



          tudinal momentum is decreased by wall shear.



2.3  Turbulent Flows with Density Gradients



       2.3.1  Redaction of Turbulence



       A discussion of turbulence in the presence of stable density



gradients is given by Phillips (27).  Deissler (7 ) formulates a sim-



plified model of initially isotropic turbulence in the presence of a



vertical density gradient.  Experimental work with turbulent, density-



stratified flow is reported by Webster (36).  In all discussions the



parameter controlling the effect of the density gradient upon the flow



is the Richardson number given by:



              £ 3p
                                                                 (2.17)

               3ul
As R. increases from 0 to 1 the vertical turbulent fluctuations dimin-
    i


ish; vertical transfer of mass, heat and momentum decrease, mass and



heat transfer more so than momentum transfer.  The region of maximum



density gradient becomes an "interface" between two layers of differ-



ent densities.  Kato and Phillips (22) investigate the entrainment of



mass across such an interface, using a surface shear to generate tur-



bulence.  They find the entrainment to be inversely proportional to the



Richardson number.  Ippen and Harleiaan (20) report that viscous under-



flows are stable for IF<1.0 where IF is the densimetric Froude number
                              -2S-

-------
given by






       IF =  — -                                               (2.18)
V is the average flow velocity, h is the flow depth and Ap is the den-



sity difference across the interface.  Underflows of moderate turbulen-



ce are experimentally evaluated by Lofquist  (23).  Experiments over a



large range of Richardson numbers for fully  turbulent shear flows are



treated by Ellison and Turner  ( 9 ) .  Their determination of entrainment


                                                                 2
as a function of a gross Richardson number expressed as R. = I/IF is



given in Figure 2.3.  The entrainment coefficient a is defined as the



ratio of the velocity across the interface, v , to the average velo-



city parallel to the interface, u.  For R. = 0 the entrainment coeffi-



cient is equal to a , the value observed in  homogeneous fluids.



       Phillips (27) proposes a model for vertical transport of momen-



tum in a shear flow, u(z).  The model is based upon a wind wave genera-



tion theory; after considerable manipulation the eddy viscosity, e,



defined by





       T  =pe|^                                               (2.19)





is given by
       e ~  w'2 G                                                 (2.20)
                 o
where

              00


       G
        °     o        w1
                                            R(OdS                (2.21)
                              -29-

-------
I

-------
is the integral time scale of the turbulent vertical velocity fluctua-



tions.  The more persistent the vertical turbulent fluctuations, the



greater the vertical momentum transfer.  Phillips suggests that in a



stratified flow the vertical velocity spectrum reflects the frequency



of the natural oscillation associated with the density gradient, the



Brunt-Vaisala frequency, N:
       N
                       1/2
                P
                                                          (2.22)
and that in the stratified flow.
G
              oo


              f  R(O cos N WC  < G                              (2.23)

              •                   "  °
Assuming that the Richardson number is not so large  that vertical  tur-



bulence has been completely damped i.e. that w1 is still the same  mag-



nitude as the other components of the turbulent velocity fluctuations,



then the ratio of stratified shear to unstratified shear.





       T      e      G

       —  =  -5-  =  -2-                                           (2.24)
       T      e      G
        ooo




The vertical shear is related to the vertical entrainment by





       T ~ u v  - au                                               (2. 25)
             e




Thus it must be that for a given shear flow u,


              oo

              f R(C) cos N£dC


       —  =  - -  < 1                                (2.26)
                  j

                  •'
                              -31-

-------
providing a relationship between  the reduction in entrainment and the




vertical density gradient.




       2.3.2  Submerged Buoyant Jets




       A number of analytical and experimental investigations have been




done for the case of submerged jets in which buoyancy acts on the jet




as a whole, producing a vertical acceleration but in which the turbu-




lent structure of the jet is not significantly altered by the density




gradients.  Fan (ll) gives a review of the investigations which include




stratified and non—stratified environments, cross flows, and plane or




axisyrametric jets.




       A major difference between the analysis of the buoyant and the




non-buoyant jets is that the linear spread of the turbulent region is




no longer generally valid because of the addition of buoyant forces.




In place of a linear spread assumption it is usually assumed that the




entrainment coefficient is constant along the jet.  This assumption is




correct for non-buoyant jets as previously discussed, and gives gener-




ally good results for the buoyant jets considered.  The value of the




entrainment coefficient is determined by comparison with experiment.




       2.3.3  Two Layer Flow




       If the local Richardson number is high or the overall densimetric




Froude number low, two layer flows of different densities may exist.




Harleman (12)  discusses the equations of motion governing two layer




flow in open channels.   Of particular interest is the case of a flow




issuing from a channel into a large body of denser fluid.  If the den-




simetric Froude number based on the upstream depth,  h ,  and velocity,
                             -32-

-------
u , is less than one, a "stagnant wedge" of dense water extends upstream



from the end of the channel beneath the flow of lighter fluid (see Fig.



2-4).  The length of the wedge is given by:
                l  _!_  _2 + 3 F2/3-f IF4/3


           fi  I   E
(2.27)
             u

where IF =  	    and f. is a friction factor for interfacial shear
             	         i
            P    °

stress,   averaged over the length L. Harleman and Stolzenbach (13) es-



timate the friction factor by using smooth turbulent friction laws in



which f. is a function only of the Reynolds number of the flow.  At



the junction of the channel and the reservoir the interface rises



rapidly.  The upper flow densimetric Froude number, based on the act-



ual layer depth and velocity, is unity at this point.



2.4  Surface Heat Exchange



       A comprehensive review of surface heat exchange processes and



predictive equations is given by Brady (3 ).  The prediction of natur-



al water temperatures requires a determination of all the natural heat



inputs to the water body.  The steady state natural water temperature



for fixed heat inputs is referred  to as the equilibrium  temperature,



T .  In a large body of water the  response time  for temperature changes



is longer than the characteristic  oscillations in heat inputs.  The



actual water temperature is never  exactly equal  to T  for this reason.



In rivers, lakes, and reservoirs  the difference  between  the actual and



equilibrium temperatures is usually less than 5°F.  In the ocean it



may be as high as 10° - 20°F.





                               -33-

-------
                                                              1.0
Figure 2-4.  Two Layer Flow in the Discharge Channel

-------
       The "excess" temperature rise induced by the addition of power



plant heat rejection may be calculated by assuming that the "excess"



heat flux through the surface, q , is proportional to the difference
                                s


between the heated water surface temperature, T  and the equilibrium
                                               S


temperature, T .
              pcp K (Tg - Te)                                    (2.28)
where the coefficient of proportionality is a function of the water



surface temperature and meteorological conditions.  Brady gives form-



ulas and graphs for determining the value of K given wind speed  and



dewpoint temperature.



2.5  Surface Discharge of Heated Water



        Experimental studies of three-dimensional surface discharges



of heated water have been done by Hayashi (16) , Wiegel (37), Jen (21),



Stefan (31) and Tamai (32).  Discussion of the experiments is presented



in Chapter 4 and some of the results given in Chapter 5.  Previous



attempts to treat the problem analytically are presented in this sec-



tion.  Figure 2-5 shows the physical problem considered by each study.



        Edinger and Polk present analytical solutions which assume that



the heated discharge is a point source of heat at the water surface on



the boundary of a uniform stream which is infinitely wide and deep.(8)



No lateral or vertical convection occurs, the temperature distribution



being determined entirely by vertical and lateral eddy diffusivities



and by convection in the direction of the uniform stream.  Buoyant
                              -35-

-------
        Uniform Cross Lurrenl
        HUM;
                                                  L'n i I orm c ross Current
Heat Source
   /  /
                     Isotherms
                Infinitely Wide
                and Deep Basin
       Edinger and Polk
Constant Depth,
Infinitely Wide Basin
                                                                                            and i ross i urrent
                         No Vertical Entrainment
 No Vert it a 1  Lnt rainment
 Constant Depth Jet into
 Infinitely Wide and Deep
 Basin

HDopes
                                                                                lun ninensional thannel
          Hayashl
               rigure  2-5.   Schematics of  Previous Theoretical  Investigations

-------
effects are not considered.  The results are fitted to field data by




adjusting the value of the vertical and horizontal diffusivities.  No




correlation between the fitted values of the diffusivities and values




estimated a priori is found.  This treatment is clearly valid only for




the region dominated by ambient turbulence as discussed in Chapter 1.




The difficulty in determining natural turbulent diffusivities is a sev-




ere limitation on the practical use of Edinger and Folk's theory.




        Carter treats the case of a heated discharge from a channel into




an infinitely long and wide basin whose depth is uniformly the same as




that of the discharge channel and in which a uniform cross current is




flowing at right angles to the discharge.  The jet velocity and tempera-




ture are assumed to be uniform across the width and depth of the jet,




varying only along the centerline coordinate.  The jet is deflected by




pressure drag and entrainment of lateral momentum as expressed in inte-




gral momentum equations.  Laboratory experiments are performed in which




the temperature field is measured.  The entrainment into the jet is




determined from the experimental data and is input to the theory to ob-




tain predictions of the centerline trajectory which are compared with




the data.  The scatter of  the data is large but the points are near




the theoretical curves on  the average.  The neglect of buoyancy and




surface heat loss seems justified for this physical situation although




the spreading of the jet is not measured and may be dependent on buoy-




ant forces.  Extrapolation of Carter's results to a deep receiving basin




are not possible since his measured entrainment curves would not be




valid.  (5)







                              -37-

-------
        Hoopes treats a physical problem very similar to Carter's.  He




assumes that the effect of buoyancy is to make the vertical entrainment




zero and that the jet is vertically uniform and of constant depth even




though the receiving basin is infinitely deep.  Hoopes assumes Gaussian




similarity functions for velocity and temperature and accounts for lat-




eral entrainment by defining a lateral entrainment coefficient.  Inte-




grated continuity, momentum, and heat equations yield expressions for




jet deflection, width, and centerline temperature and velocity.  The




theoretical model is compared with field measurements.  Uncertainty in




the values of the ambient cross flows makes comparison of the theoreti-




cal and actual jet trajectories difficult.   Also most of the temperature




data is taken far from the discharge where the temperature is nearly




the ambient value.  This study illustrates the extreme difficulty in




using field data to check a theoretical model. (18)




        Hayashi proposes a theory for a discharge from a channel of




finite width into a large, stagnant receiving body of water.  He assumes




that the buoyant stability is so great that vertical entrainment is zero




and that all of the convective terms in the governing equations are neg-




ligible.  The equations are integrated over the jet depth using a simi-




larity function to describe vertical temperature and velocity distribu-




tions.  Analytical solutions for temperature and velocity are obtained




by assuming that horizontal eddy diffusivities are not functions of




space and taking into account heat loss from the water surface.  Experi-




ments done by Hayashi do not agree with the theory and show a large




amount of scatter.  The discrepancy between data and theory is a result





                                 -38-

-------
ot initial mixing near the discharge point which the theory does not




account for.  As discussed in Chapter 1, initial mixing is important




in most heated discharges.  Also, the formation of stagnant cold water




wedges in discharge channels of low densimetric Froude number, IF, pre-




vents the local value of IF at the discharge point from being less than




unity.  Thus Hayashi's theory, which is valid for 3F= 0, will not apply




to actual heated discharges.  (15)




        Wada considers a two-dimensional flow (vertical and longitud-




inal) of heated water and treats the full equations of motion in which




turbulent terms are represented by eddy coefficients.  The vertical




eddy coefficients for heat and momentum have the form
                                                                  (2.29)
where e , e.., and c are constants which are different  for heat and




momentum and R  is the local Richardson number. (35)




        The equations of motion are transformed into a vorticity equa-




tion and solved numerically.  The results are physically reasonable,




indicating an initial entrainment near the discharge and a  stable two




layer system farther from the discharge.  Wada states  that  the theory




is valid for discharges in which the ambient turbulence controls the




diffusion processes, making e  and e, relatively  constant in  space.




        The details of Wada's numerical scheme are  not given  in his




papers and a calculation for only one case is shown.   It is doubtful




that many field conditions correspond to Wada's problem.  If  the dis-
                             -39-

-------
charge channel is wide, the jet may behave initially like a two-dimen-




sional flow with little lateral variations in temperature and velocity,




but the turbulent structure close to the discharge origin is jet-like




and is not constant in space as Wada assumes.




        Table 2-1 summarizes the studies from the point of view of the




processes affecting the temperature distribution and the number of phys-




ical dimensions considered.  Each of the authors presents a satisfact-




ory theory based on the physical situation he assumes.   None of the




theories consider all of the factors discussed in Chapter 1 as being




important for the prediction of temperatures in the vicinity of a heat-




ed discharge.
                                -40-

-------
             Direction of
             Temperature
              Variation
                                Factors Affecting Temperature Distribution
                                              Discharge
                          Initial             Channel    Bottom    Surface    Ambient      Ambient
Investigator   x   y  z   Mixing   Buoyancy   Geometry   Slope    Heat Loss  Cross Flow   Turbulence
 Edinger     yes yes yes   no        no
 (8)
                                    no       no
                                                        no
                                                                   yes
                                             yes
 Carter      yes yes  no   yes
 (5)
                        no
                                   yes
                                             no
                                                        no
                                yes
                                                                                 no
 Hoopes
 (18)
yes yes  no   yes
                        no
yes       no
                                                        no
            yes
                                                                                 no
 Hayashi     yes yes yes    no      yes
 (15)
                                   yes
                                             no
                    yes
                                                                    no
                                                                                 no
 Wada
 (35)
yes  no yes   yes      yes
yes      yes
                                                        no
                                                                    no
                         yes
 This
 Study
yes yes yes   yes      yes
yes      yes
yes
yes
                                                                                 no
                         Table 2-1.   Summary of Previous  Theoretical Investigations

-------
III.  Theoretical  Ccmsideratiions



3.1  Statement of  the  Problem



       The preceding chapter  reviews  earlier  theoretical investigations



of heated surface  discharges.   Each is  severely  limited:  only Hayashi



considers surface  heat loss.   There is  no  attempt  to include buoyant



convective motions and none of the theories reproduce the initial core



region in a turbulent  jet.  There is  a  need for  a  three-dimensional



theoretical development which  considers all these  factors.



       This study  treats a horizontal discharge  of heated water at the



surface of a large body of ambient water.  Figure  3-1 shows the rele-



vant geometrical characteristics of the discharge  and the definition



of a Cartesian coordinate system.  A  theory for  predicting the three-



dimensional velocity and temperature distribution within the discharge



jet is developed in the following sections.  The initial temperature



difference between discharge and ambient water,  the discharge velocity,



the channel geometry,  surface  heat transfer, the ambient cross flow,



and the bottom slope will determine the temperature and velocities in



the jet.



       The theory  is limited to the region of flow where jet turbulence



dominates ambient  turbulence.  The ambient cross flow is not necessar-



ily uniform but far from the jet the free surface  is uniformly at z = Q



and the temperature is  T .
                        cl


       The basic hydrodynamic  and thermodynamic  equations governing the



flow are  first transformed into usable form for  the case of a non-buoy-



ant discharge into deep, stagnant ambient water.  The effects of an





                                 -42-

-------
Initial Velocity » UQ

Initial Temperature = TQ
                                t
                                                                 Bottom Slope - S
                                                                                            Front View
         u ,T
          o' o
r
      Top View
                                   Ambient Temperature = T
                  Figure 3-1.  Characteristics  of the  Discharge  Channel

-------
initial temperature difference  are  then  included  in  the  theoretical


model.  Finally, a buoyant jet  discharged  over  a  bottom  slope  or  into


a cross flow is considered.

3.2  The Governing Equations

       3.2.1  Basic Equations


       A fluid flow is characterized by  its velocity, density, pres-


sure, and temperature, all of which may  be a  function of space and


time.  These four variables are related  by equations expressing con-


servation of mass, momentum, and heat energy  and  an equation of state


which relates density, pressure, and temperature.  In Cartesian coor-


dinates the complete equations for water flow, incorporating the Bouss-

inesq approximation,  are:

                                 =
                                                                  J"L
       3t     3x
                     ay
3u     3u
at     ax
              3uv      3uw _
                 -
                                                 2
                                                3_u
                                                      3z
                                                                 (3.2)
jiw
at
at
       3uw
       ax
              3v
                             p  3y
3vw
3y
3y
3w _ 1 3p
az p az
1 awT t
I • — K
[32I
U2
                                           fa2*
                                         32T

                                         3y2
                                                U
                                                 ,
                                                 3y
A
. 2
                                                      3z
32wl
                                                                  (3.3)
                                                                  (3.4)
                                                                  (3.5)
       3jp_
       3T
               a  =  0
                                                                 (3.6)
                              -44-

-------
where  u,v,w  =  velocity components in the x,y,z direction




           p  =  density




           p  =  pressure




           T  =  temperature




           v  =  viscosity




           g  =  gravitational acceleration




           k  =  thermal conductivity




           a  =  thermal expansion coefficient




             =  viscous generation of heat





       In conjunction with  initial and boundary  conditions  the  fluid




flow is determined by these equations.  There is no  closed  form solu-




tion to the governing equations.  Approximations must  be  introduced




which simplify the equations  to  the point where  solutions may be ob-




tained.  The  transformation of the governing equations will consist of




dropping certain negligible terms, of assuming the  form of  certain




variables, and of integrating the equations over a  region in space.




       3.2.2  Steady, Turbulent  Flow  Equations




       In accordance with standard practice when dealing  with  turbulent




flows the velocity, pressure, density,  and  temperature are  separated




into mean and fluctuating parts:






       u - u  + u'




       v = v  + v1




       w = w  + w'                                                 (3.7)




       P = P  + P1




       P • P  + P'




       T = f  + T1              __

-------
        In  general the mean values may be  functions of time when con-



sidered over periods longer than the periods  of the turbulent fluctua-



tions.   This study considers only steady  mean flows.



        The result of substituting Equations 3.7 into  Equations 3.1 -



3.6 and taking  a time average of each t_en.:  .; the well known set of



turbulent  mass,  momentum, and htac. equations.
3y
                         3x
                                                       3y
                                                               3z
                                                                 (3.8)
3u    3uv    3uw
      3uv  .  3uw      1  3p    , _ „ .
^   •+• -z— + -r— =  -  —r-*- + v I—^ +
3x    3y     az       p  3x    I „ 2
                                     A     O
          u .  3*u  .  3"u|    3u'2    au'v1
                                                   3x
                                                                 (3.9)
8uv   3v     9vw
—-   -j- r-T-. ^  ^. „
                     1  3p  ,   I 3 v
                =	-^ + V 	r-
                     P  °y     I ~ 2.
                        1     LdX
               9—    9 —
        -«-    *.£•    ./
        3 v   a v   3 v
                                2   32   3«2
                                    ']    ,3u'v'   3v'     3v'w'
                                                                 (3.10)
3uw   3vw   3w2

3x    3y    3T

                                                      3u'w'    3v'w'

                                                        3x  '  8y
                                                                 (3.11)
3uT .  3vT .  3wT
          r -^ —
3x  '  9y
)2T   82T
"" 7^~ '    f\"
                                         3u'T'    3v'T'    3w'T
                                                    3y
                                                                 (3.12)
                                                                 (3.13)
                               -46-

-------
Henceforth the mean variables are written without the bar.  Fluctuating


qualities are still designated by a prime.


       3.2.3  Hydrostatic Pressure


       The mean pressure is considered to be the sum of two parts:
       P = Ph + Pd                                               (3.14)
where the hydrostatic pressure,? , satisfies the equation

                3ph
                    -g                                           (3.15)
The remaining component, p , is a dynamic pressure resulting from fluid


acceleration.



       Equation 3.15 may be integrated:




       P   = - (   Pgdz  +  p                                     (3.16)
where z  is a point at which p  = p .  In this case  the elevation of



the water surface, n, is a point where p  = 0.  Thus
       Ph  = -  (  Pgdz                                           (3.17)

                'n



       The density is separated into components:
             pa - Ap                                              (3.18)
              d
where p  is the density of the ambient water and  is not  a function of
       ci


space.  The deviation from the ambient density  is -Ap  and in  a  heated



                              -47-

-------
jet  results  from  a  temperature  difference  as  defined  by  the  equation




of state.  If  T   is the  ambient temperature corresponding  to p  , then
               a                                              a
               -f
                  'T
and
P  - Pa - I    adT                                        (3.19)

      3   'l
            a
              (T

       Ap  =  \     adT                                             (3.20)

              'T
If Equation  3.18 is substituted into the pressure Equation 3.17:
       Ph  =  -  I  Pagdz  +  I  Apgdz                             (3.21)

                 'n           'n




The pressure terms in the momentum equations are then:
         1^  5p _ _   3n _ 	3	  f   3Ap  ,      1   "''d
                         rz  .,..         i    3P,
                                 dz -
  p  3x       3x   p -Ap  J   3x
                    a      n
         1  IE  -   „ 3H     g    fZ  3Ap dz     1   3Pd          ,.
       •H —  — ..T ..  s: » o •'• - ^ — •- • •!*-• — •'  I     »  vi^i ^  •• • • •'  • ' — —          i j y 4 j

         p   3y      s 3y   Pa~AP  )   3y       pa~AP Sy             '   J
       _ I  IE  =  „ - -1—
         p  3z     g   p -Ap
                        a
       3.2.4  Scale Variables



       The dominant terms in each equation are identified by estimating



the scale of each variable and the magnitude of spatial changes in each



variable.  For this purpose the following "characteristic" or "scale"
                                -48-

-------
variables are defined:




       x*,y*,z*  =  scale of changes in the x,y,z directions


             n*  =  scale of changes in the water surface elevation


       u*,v*,w*  =  scale of the mean velocity components


            p *  =  scale of hydrostatic pressure changes


            p *  =  scale of dynamic pressure changes


            AT*  =  scale of mean temperature changes


            Ap*  =  scale of mean density changes


    u'*,v'*,w'*  =  scale of the velocity fluctuations u1 ,v' ,w'


           AT1*  =  scale of the temperature fluctuations


           Ap'*  =  scale of the density fluctuations


             0*  =  scale of the viscous heat production




       In discussing relative magnitudes it is useful to  define a quan


tity 6 such that the following sequence of terms:



       ,21      1
       6  ,  6,  1,  j   ,  ~2
                          o



is increasing and each  term is an order of magnitude different from


adjacent  terms.  The symbols « and  » are stronger statements of rela


tive magnitude.


       3.2.5  Large Reynolds Numbers


       The ratio of any of  the momentum terms  to  the viscous terms is :
        (u*,v*.w*)(x*,y*.z*)

                 v
                               -49-

-------
where  the  appropriate  choice  of  length  and  velocity  scale  is  made.



This ratio  is  the  Reynolds  number,  R.  A condition for  fully  developed



turbulence  is  that  the Reynolds  number  be very  large i.e.  H » 1.   The



viscous  terms  are  then negligible with  respect  to the convective  terms.



Since  k  is  the same order as  v,  the molecular heat transfer terms are



also negligible.




       In  the heat energy equation, the ratio of viscous heat  produc-



tion to  heat convection is
                                (u*2,v*2w*2)

                                                                      '
       Heat Convection      R     c  AT*
                                   P


                                                         2   ?   ?
                                                       (u*  v*  w*  )
where c  is the specific heat of water.  The quantity - - '    * - -


                                                         °P

is at most of order 1 in the jet.  The viscous production of heat  is




therefore neglected.



       3.2.6  Small Density Differences



       The deviations from  the ambient temperature, T  , under consider-
                                                     C*


ation are of the order 0 -  30°F.  The coefficient of thermal expansion



is considered a constant over such a range and Equation 3.20 becomes
       Ap  =  a (T - T )  =  aAT                                 (3.27)
                      3.
For T  from 40°F to 100°F the value of - varies from 0 to ,0002°F~1.
     a                                 p


Thus the maximum value of Ap is .006 for AT = 30°F and T  = 100°F.  It
                          *                              3.

                          Pa

may be said that
           « 1                                                  (3.28)

       pa



                              -50-

-------
       The pressure Equations 3.22 - 3.24 are simplified by noting that
       —±—  :   -±-                                              (3.29)
       Pa-Ap      pa
       Since AT'* can be no greater than AT*, it must be that Ap'* is



less than or equal to Ap*.  The mass conservation equation is simplif-



ied by dropping all terms scaled by Ap* or Ap1* since these  are negli-



gible compared to terms scaled by p .  The mass conservation then be-
                                   3.


comes the familiar continuity equation for incompressible fluids.
       3.2.7  jSasic Scale Relationships



       The continuity Equation  3.30  implies  that:






       v* -  ^-r  u*  and w*  -   -—- u*                             (3.31)







       The scale for free surface changes,  r\*,  may be determined by



considering the momentum equations at  very  large depths i.e. as z -v -».



The effect of  the  jet is negligible  and  there is no motion at such



depths.  Equations 3.22 and  3.23 become:




                3       0  /-°° a.        -,   3p,

       0 = - g M   _  _S_  I   JZ^fi. dx  -  —  —- at z = -oo         (3.32)

                      'a  ' n             a
                        a   n
                                -51-

-------
It must be that
                        _                                        (3.34)

                         pag
       In a fully turbulent shear flow the turbulence is locally iso-



tropic.  The scale of the fluctuating components, (u'*,v'*,w'*), are



the same order of magnitude.  Furthermore the mean velocity in the dir-



ection of the flow is an order of magnitude greater than the fluctuat-



ing velocities:
                    '*
3.3  Non-Buoyant Surface Jets



       3.3.1  Governing Equations



       This section considers the case of a discharge of water horizon-



tally from a rectangular open channel into a large body of water at the



same temperature as the discharge.  There is no bottom slope and no



ambient cross flow.  With these conditions the equation of mass conser-



vation 3.30 is exact and the heat and state equations are not needed



since p and T are not variables.



       The momentum equations with all negligible terras dropped are


                                                   u*2
shown below.  The scale of each term is divided by —r—  for convenience
                                                   X


in comparison of terms:




                              -52-

-------
  o                                          o

3u      3uv    3uw _     jh} _  _1_   pd _  3u'	9u'v'  _ 3u'w'
3x
1
3x
3y 3z 8 3x p 3x 3x 3y Sz
Si
* * (3-36)
n U* o U* ^
"a * a
2 •}_ o
+ 3v + 3vw 3n 1 Pd 3u'v' 3v' 3v'w'
3y 3z 8 3y p 3y 3x 3y 3z
a J
(3.37)
  JL      jt      JL      * J    "Jt    P J    A                «L       JL.


 x*     x*     x*     ~T2 y*    72  y*"      &     6 y*    6 "z*"
                      pu*       pu*
                       a         a





3uw +  3vw +  3w   =  J^_  ^P_d_  _ au'w'   _  ly'w1   _  3w'2

3x     3y     3z      p   3z  ~   3x        3y       3z

                       S                                           (3.38)


 j^    j*     z*     ^d   x*^               x*        x*

 x*    x*     x*        2 z~*      6       6 7*      6 7*
                    pau*                    y         2
       If the turbulent  terms  in the x equation are to be the same



order as the convective  terms  it must be that:







       £  ~ 6  and   £ -  6                                       (3.39)







The first turbulent  term is then negligible in all the momentum  equa-



tions.



       In the y and  z  equations the turbulent terms are of order 1  if



3.39 is true.  The pressure term must also be of order 1 since the  con-



vective terms are of order   6:
                               -53-

-------
        Pd*
       n U*
       pa
                                                                  (3.40)
The pressure term in the x equation is then negligible.  The approxi-


mate equations become:
         2
       3u      3uv   ,  3uw  _    3u'v*     ju'w'
                                                ~
       0 . _ g in
             g
               3y    p   3y      3y
                      a

       3.3.2  Structure of the Jet


       The set of equations is a well known result for turbulent shear


flows.  Further assumptions about the nature of the jet are taker, from


established experimental and theoretical results in Chapter 2.


       The gross features of the jet are defined in Figure 3-2.  The


core region has a half-width s and depth r.  The turbulent shear region


expanding horizontally has width b;  the vertical shear region is of


depth h.   Section A-A in Figure 3-2  shows four regions in the jet cross
                              -54-

-------
     Side View
                                                  Jet Boundary
//////
Turbulent
 Region
  fop View
                                         'Core Region
                                                      Regionfeeglon,
                                                     1 3   | 1   1
                                                                                  (RegionReglonl

                                                                                  1  *   1  2   1
                                                                                   b     s
                                                                                        Section A-A
                Figure 3-2.   Geometrical Characteristics of the Jet

-------
section:  an unsheared  core  region (//I) :  a vertically sheared region



(//2) : a horizontally  sheared region (#3);  and a region sheared in both



directions  (#4).  The jet  is assumed  to be symmetric about the plane



y =  0.  Thus only half  the width  of the jet will be considered in the



theoretical development.



       The core regions are  eventually engulfed by  the spreading tur-



bulent zones.  The distance  at which  r and s become zero  are  x  and
                                                               r


x  respectively.  Ifx   
-------
 I
Ul
                                           Centerline Velocity


                                               i—-Center-line Temperature Difference
                                                   •Temperature Distribution


                                                •Velocity Distribution
                                                                                                       I _Reglpn_l	
                                                                                                        t 1 t T tl t T~t t 11 7
                                                                                                          Kegion 2
                                                                                                        M t M Ml T t f I.
                                                                                                                      I
                           Figure  3-3.   Velocity and Temperature Characteristics  of  the Jet

-------
The  form of  the  similarity function, f, is a basic assumption of this



theory.  A function  introduced by Abramovich (1 ) for the main region



of plane and axisymmetric jets is  used






       f  =  (1 -  ;3/2)                                          (3-47)






Abramovich uses a different function for the core region.  The reason



for using 3.47 throughout the jet in this theory is discussed in a



later section.



       3.3.3  Boundary Conditions



       The conditions at x = 0 are assumed to be:
       1  "  ho



       s  =  b

              °                                                  (3.48)


       b  =  h  =  0



       u   =  u
        c      o
These conditions imply that there is no significant turbulent boundary



layer in the discharge channel.



       The theoretical development treats each of the four jet regions,



Thus boundary conditions must be specified on the boundaries between



regions.  In the following development ose will be made of the kine-



matic boundary condition which states that particles of fluid on an



impermeable flow boundary remain at the boundary.



z = n:  The kinematic boundary condition holds on the free surface:
                             -58-

-------
                              at z = n                           (3.49)
There is no turbulent momentum transfer across the free surface
 w'2 = u'w1  =  v'w'  =0    at z = n                           (3.50)



y = 0:  The jet is symmetrical about this plane.  There is no net flux

of mass or momentum across y = 0.
         2
    v  = v'  =  u'v'  =  v'w'  =0    at y = 0                     (3.51)
z = -r:  The boundary of the vertical shear region is not a rigid sur-


face but there is no momentum transfer across this boundary:




  w?2= u'w'  -  v'w'  =0     at z - -r                          (3.52)



It is assumed that  the vertical velocity at z = -r is given by:
       w  =  w              0
-------
  = s:  This boundary  is  analogous  to  z  =  -r.   The  conditions are:
                          t 2

       u'v*  =  v'w'  =  v   = 0
       v=v     -r
-------
y a s + b;  This surface is analogous to z = -r-h.  The conditions are:
       v = v
       v - v f( C )
            e    z
                          -r < z < n
                  -h  <  z  < -r
                                  at  y  =  s + b
                                                          (3.57)
       v - 0               z < -h

 v«2 = u'v'  =  v'w'  =  0


       3.3.4  Integration of the Equations

       To solve the equation set (3.41 - 3.49) it is necessary to inte-

grate the equations over certain portions of the yz plane such that the

contributions of the turbulent transfer terms, about which no assump-

tion has been made, are zero.

       If the y and z equations  (3.43 and 3.44) are integrated over

the entire yz plane of the half jet the result is:


                                                 p
                                              -  —       dz      (3.58)
                                                    y=0
0 -
g(r+h)
S
=b~
V
0
)
-c
p J
Q
(r+h) pa
y
=b
Pd
pa
       ,s+b   p,
0  = - \      -^
       Jo     pa
                         z=-n
                                  dy
                                -(r+h)
 Using  Equation  3.40  it must  be  that
                                                                  (3.59)
n*    u*
Z*    gz*
                                                                  (2.60)
        It  is  assumed that the discharge channel flow is subcritical so

 that  the water level at  x = 0 is controlled by the level of the ambient

                               -61-

-------
                                                          u

 fluid.   This is true if t'ae channel Froude number, IFO =  	 ,  is



 less  than unity.   It then must be that
       i*   <  «                                                  (3.61)





 and water  surface  elevations  are negligible in comparison to  the  scale


 of vertical changes in the jet.   Integration over the jet in  the  ver-


 tical  direction may begin at  z  = 0 without  significant error.


       Equations 3.58  and 3.59  imply that  the total  pressure within


 the jet  is constant from the  interior to the boundary.  This pressure


 may be estimated by Equation  3,40.
       Ph*  +   Pd*  "  &  pu*                                        (3.62)





Consider  the flow of  fluid  outside  the jet.   Entrained  fluid is trans-



ported from rest far  from the jet,  where  the  pressure difference  in-



duced by  the jet is zero, to the boundary of  the jet where it has



velocity w  or  v .  The  flow outside the  jet  is nearly  irrotational and



Bernoulli's theorem may  be  applied  along  a streamline.





       t*\           _1_
            £    G       d


                                                                  (3.63)


       0  =-^pw    +  p,
            2    e       d






or using 3.62
                             -62-

-------
       v  - 6u*
        e
       w  - 6u*
        e
                                                                 (3.64)
       The entrainment velocities are proportional to the local longi-



tudinal velocity scale and are an order of magnitude less.  This rela-



tionship is formalized by introducing entrainment coefficients:
       v   =  a u
        e      y c
       w   =  a u
        e      z c
                                                                  (3.65)
where a  and a  are variables to be determined by  the  theory.



       The x momentum and the continuity equations are integrated over



each region of the jet using (3.41 and 3.42)  the similarity  forms for



velocities, the boundary conditions, and the  entrainment  definition:
Region //I  -  continuity

          du

       rs - — + rv  - sw  =0
          dx      s     r
Region # 2  -  continuity
            du h


            -dx- +   SUc  d^ + Vbhll  + S
Region  //  3   -   continuity




            du b       .
             C.         Q.S
        I,r  — j — + ru   -j—   -  w, bl,  - r [v  -au]  =  0
        1   dx      c  dx       hi        s    y cj
                              -63-

-------
 Region //A  -  continuity




   2  du hb                  .

 *i    ~~j	+u *i  [ b -;	l-h-r-]+ [w -a u ]I,b - (v,-a u )I,h = 0
  1     dx      clL   dx     dxj     hzcl      bycl
 Region //I  -  momentum



            du

       2 rs -~-  +  r  v  - s w   =0
            dx         s      r




 Region it 2  -  momentum



             du 2h        2

        s I, —1=—  +  s u   -r^  +  u v, hi.  +  u sw  =0
           2dx        cdx      cb2      cr




 Region // 3  -  momentum                                          (3.66)

                 •)                                                cont'd


              duc b        2  ds
        r I9  —,	  +  r u    -7^- - u w, bI0 - u rv  =0
           2dx        cdx    ch2    cs
Region  // 4  -  momentum



         2 du  hb

        12  -ir-  + \\  ^ If + b & + «ci2  (whb
  r1                 r1
= (    f(c) dc   I, = \

  '0             *   JQ
                                      2
where  I   =     f(c) dc   I  =  \   f
       These eight equations are not^ sufficient  to  determine the eleven



variables.  The following experimentally  observed relationships  are



assumed (1 ):
                               -64-

-------
       S  -  E
       Abramovich (1 ) notes that the value of e is different in the



core regions from that in the outer jet.  He also uses a different



similarity function for the core regions and matches the inner and



outer regions by allowing h and b to change over a transition region.



This difficulty is bypassed by allowing  e and f to be the same every-



where.  The results for the outer region are the same as in Abramovich's



work and the differences in the core region are small.  It is assumed



that e = .22 for all values of x.



       3.3.5  The Solution for Non-Buoyant Jets



       The solution of the Equations 3.66 - 3.68 is:





       For all x,
       v   =  v,  =  w   =  w,  =   0
        s       b      r       h
              h  =  b  =   ex
                               ho             bo
        Initial  Region     x <  ——  ,    x <   —
        u   =   u
         c       o
        r  =  h  - el_x
               o     2
        s   =  b  - eI0x
               o     2
       -ay =  az  =  eC^ - I,,)
                              -65-

-------
For h  <  b
     o    o
                           el,
            el.
       u   =  u
        c      o
       r  =  0
       s  =  b  - eI0x
              o     2
For b   <  h
     o     o
el.
     < x  <
el.
       u   =  u
        c      o
       r  =
                                                                  (3.69)
                   -
For x  >bQ/eI2 ,  x  >
                 Jh b
                 • o o
       u   =  u	
        C      O  I „ '-X
                                -66-

-------
             s  =  0
Defining the dimensionless variables:
       x  =
            /h b
              o o
            fh b
              o o
                                                                 (3.70)
             fh b
              o o
             f h b
              o o
             'h b
              o o
and the constants:
                     1

                    el.
      G>   _i_
s    >/h    eI0
                                                                  (3.71)
       A   =   r—
                                -67-

-------
the solution for velocity and jet dimensions is:
       b  =  h  =  ex       x>0
       u  -  1.0
>    ;<
            Sk
                                      x   and  x
      s  =  0
                                 X   < X  <  X
                                  s        r
      r  =  0
            /A
                                 X  < X < X
                                  r        s
                                                                 (3.72)
      u
      r  =  s  =  0
                                 x > x   and  x
                                      s        r
                             -68-

-------
For the assumed similarity function:
              .450


                                                                 (3.73)



              .316
and thus since e « .22;
       x   -  14.A ft.


                                                                  (3.74)



              14.4
       x   =  	
       A plot of these solutions is shown in Figure 3-4.  The data



shown is from experiments by Yevdjevich (38) for turbulent air slot



jets with values of A from 1.0 to 94.0.  Scaling x by /h b  allows all



of Yevdjevich's data to be plotted on a single line, where he presents



separate curves for each value of h /b  and scales x by b .
                                   o  o                  o


       This section has considered a non-buoyant horizontal discharge.



The solution is obtained by proposing similarity forms for velocity



and temperature and by assuming linear jet spreading as discussed in



Chapter 2.  The definition and determination of entrainment coefficients



are not necessary for the solution of the equations.  However, in the



following section, which treats buoyant discharges, the linear spread



assumption is not valid and the entrainment coefficients determined



for non-buoyant jets will be used to give information about entrainment
                                -69-

-------
c
I
                                                                                         1000    2000
                             Figure  3-4.   Non-Buoyant  Jet  Solutions: Data from Yevdjevich:   1<  A< 94

-------
in the buoyant case.



3.4  Buoyant Jets



       3.4.1  Governing Equations



       This section considers the horizontal surface discharge  of water



from a rectangular open channel into a large body  of water.   The temp-



erature of the discharge is greater than that of the ambient  water.



There is no bottom slope and no ambient cross current,  these  effects



will be included later.  The heat energy and state equations  are coup-



led with the mass and momentum conservation equations.



       The scaled momentum and heat equations  (3.9 - 3.12)  are:
3u  + auv + auw _    in _ JL ( °°3Ao dz _ _i

3x    3y    3z       ax   p  )  3x       p
                           a.  T\           o.
               Pa"
                             gz*
                           u*
                                                 ax
ay
                                                                   (3.75)
                                                 <5    6 „*     6 T*
                                                                   x*

                                                                   z*
3uv H
ax
y*
av
ay
, 3vw an g 1
az ay p
a
">»*"
. p.*
y* d
9AD dz 1
) 9y p« 9i
n "
^8Z* x*
d 3u v
i ax
r
                                                           '
                                                          X*

                                                          y*
                                                                   (3.76)
auv/
ax
z*
X*

3vw 3w 1 d au w
+ 	
ay
z*
X*

T —
az
z*
X*


Pa 9Z 8X
n *
Pd x*
*2 z*
pu*
a
_ av w
ay
A X*
S*

r\\J
az
A x*
0 £

                              -71-
                                                                    (3.77)

-------
 3x      3y      3z          3x          3y         32

                                                                   (3.78)



 111          6        67$       <5 77T
       If, in the x equation  (3.75),  the pressure  term related  to  the


density difference is to be significant, it must be  that:
Ap*


     - 1                                                (3.79)
        g Pa
         u»
The y equation (3.76) indicates that if y*/x*  -  6, there is no  term of


sufficient magnitude to balance the y pressure terms which would be of


order 1/6 .  The conclusion is that






       y*/x*  - 1                                                  (3.80)





and as in the non-buoyant jet, for some turbulent term to be important


in the x equation,





       z*/x*  -  6                                                 (3.81)





       The z equation (3.77) implies that:
        P/
       -S-  -  6                                                  (3.82)
       pu*
                               -72-

-------
       The equations may now be written without the negligible terms.



The progression from the non-buoyant jet to the buoyant case is of in-



terest.  For this reason all the terms appearing in the non-buoyant


equations are retained.  In addition the free surface variation terms



are eliminated by using Equations 3.22 and 3.23.  The density differ-


ence is replaced by the temperature difference using Equation 3.27.
        H.  +              =  0                                     (3>83)
       ax     9y     3z
3uw _ ag f

32  ~ P., )„
                                ML dz  . inbLl _ inl^l            (3.84)
        3x     9y     3z     p       9x          9y       9z
                          3.   Z
        9uv    9v2    3uv _  ag  (~°° 3AT          9P       '
        3x     3y     3z     p   J    3y        pa  3y      3y      3z
                          £1   Z             3
                 3z        3y       3z
                                                                   (3>86)
                 3vAT     3wAT  =  _ 3v'AT'   _

                                ~  ~         ~
         3x       3y       3z          3y         3z





        3.4.2  Structure of the Jet



        A comparison of the non-buoyant governing equations (3.41 - 3.44)



 and those derived for buoyant jets (3.83 - 3.87) indicates the effects



 of adding an initial temperature difference to the non-buoyant case:
                               -73-

-------
       1) Buoyant forces  in  the x  direction  tending  to accelerate


          the jet.


       2) Buoyant forces  in  the y  direction which increase the lateral


          convection.




These effects are incorporated into the jet structure as defined for the


non-buoyant jet.  All the definitions of initial channel geometry, jet


regions, velocity distributions, and velocity boundary conditions are


assumed to be the same as those defined for the non-buoyant jet as in


Sections 3.3.2 and 3.3.3.


       The temperature is assumed  to be distributed as follows:
       AT  =  AT  T (y) T (2)                                     (3.88)
                i_  y     L
where AT  is the centerline temperature difference.  The similarity


function for temperature is related to the similarity function for velo-


city by Equation 2.13:
T   =  1.0                   0 < y < s
 y


T   =  t(£ )                 s
-------
where





       t  = /f  =  1 -C                                           (3.90)






       The boundary conditions on the turbulent transfer of heat are



as follows:
       w'T'  =0                      z  =  -r -h

                                                                   (3.91)



       v'T'  =  0                      y  =  0 and y  =  s + b






i.e. no transfer of heat is allowed from the jet into the ambient  fluid.
       w'T'  =  K(T - T  )  at z = n                                (3.92)
                       3.
where K is  the surface heat  transfer coefficient as  defined  in  Equation



2.28.  It is assumed  that  the ambient  temperature T   is  nearly  the  eq-
                                                   Q.


uilibrium temperature T  .



       3.4.3  The y Momentum Equation



       The  y momentum equation  (3.43)  formulated for a non-buoyant  dis-



charge expresses a balance between  turbulent momentum transfer  and  the



dynamic pressure gradient.   The integrated  form of Equation  (3.43)  im-



plies a relationship  between the centerline jet velocity and the en-



trainment velocity which is  formalized by defining a lateral entrain-



ment  coefficient, a  , as in  Equation 3.65.  The y  equation in the case



of  a  buoyant discharge  (3.85) contains lateral convective terms and a
                                 -75-

-------
 pressure  gradient  resulting  from  a  lateral  temperature  gradient.   It  is

 assumed that  the non-buoyant balance  between lateral  turbulent  entrain-

 ment  and  the  dynamic  pressure as  expressed  in Equation  3.34 holds  for

 the buoyant jets and  that  the discussion  in Section 3.3.4  is valid for

 horizontal entrainment:
                                                                   (3.93)
where a  is assumed  to be  the  same for  each jet  region as determined

for the non-buoyant  jet  (Equations 3.69).

       Subtracting Equation  (3.43) from Equation (3.85) yields a rela-

tionship batween  the lateral temperature gradient  and the lateral con-

vective terms
                               p         3y
                                3   Z

       This equation may be integrated over the entire yz plane of the

half jet:
           b+Sro                                     T
                                                           dy
   (b+Sro                  /s+b fo      (
£        uv dzdy  -  *&  \
   'o '-h-r           pa  > o   y-(r+h);
where ^ Is a dummy variable for z in the integration.  In  (3.94)  a

lateral velocity, v, appears for which there is no previously assumed

distribution which will make the integration possible.  This lateral
                                -76-

-------
velocity is a consequence of lateral pressure gradients  and  causes the

lateral jet spread to be greater than the non-buoyant  linear spread,

         It is assumed that the lateral spreading velocity is related
to the difference in the buoyant and non-buoyant  spreading.
             I db      i*-*1-11  i                                         /-» r»/• \
             IS  -   feU"                                       0.96)
Since the spreading is  induced by  the  lateral  temperature difference in

the jet, the distribution  of v is  assumed to be  proportional to the tem-

perature gradient  in  the y direction which is:
        If =   0               0   (1 - Cy   >  Cy /  s < y < b
         v  •  0                  b < y                             (3.98)
                                  -77-

-------
        3.4.4  Vertical Entralument:




        In Section 3.3.5  the solution of  the non-buoyant jet equations




included vertical and horizontal entrainment coefficients, a  and a .
                                                            y      2


In the previous Section 3.4.3 it is assumed that for buoyant jets the



lateral entrainment is not affect--" by the ;   is the  same as derived for the non-



buoyant jet.   As discussed in Chapter 2,  vertical entrainment is re-



duced in the presence of  a vertical density gradient.  In this section



a relationship expressing the reduction of entrainment as a function of



the local Richardson number  (2.17) or alternately the local densimetric



Froude number (2.18) is formulated.  It is assumed that the entrainment



for zero density gradient is, a , the value derived for non-buoyant
                               ti


jets.



        The theory by Phillips presented  in Chapter 2 is used to relate



the vertical entrainment  to the local temperature gradient.  Hinze indi-



cates that a reasonable form for double correlation functions such as



R in Equation 2.24 is:
                  2 2

             - e~5 C                                             (3.99)




Then it is true that:





        G  -  |r                                                 (3.100)
         o    *•*-




and using (2.26) with a  - a :
                                                                 (3.101>




                             -78-

-------
where a   is the reduced vertical entrainment coefficient.
       sz


        Hinze points out that for flow advected in the x direction with



velocity u, the integral time scale G is G = A/u where X is the inte-



gral space scale of the turbulence.  Furthermore in free turbulent jets



A  is proportional to the width of the turbulent region  (34).  It must



then be true that for vertical turbulent motions:
        G   ~ -  - h/u
         o    c       c
       2
since N   (see Equation 2.22) may be approximated by
(3.102)
        N
_£.

P_
(3.103)
 the  entrainment  relationship becomes:
          sz
                                                                  (3.104)
where  IF  is  a  local  densimetric  Froude  number which  corresponds  very
        LJ


closely  to  that  defined  by  Ellison and  Turner.   Their data,  see  Figure




2-2, indicates that  6 ;  5.0.   Thus it is assumed that in the buoyant




jet:
          sz
                 a  exp
                  z    f
                         -5.0
                               P u
                                a c
(3.105)
                               -79-

-------
         3.4.5   Integration  of  the  Equations


         The momentum  and  continuity  equations are integrated separately


 over  the four jet  regions shown  in Figure 3-2 as in the case of the


 non-buoyant jet.   The heat  equation  is  integrated over the entire jet.


 The free surface elevation  is  estimated as
                        « 1                                     (3.106)
                   '"a



and may be neglected in the integration.  The complete set of equations


are given in Table 3-1.


        The equations governing the non-buoyant jet are a system of


non-linear first order differential equations.  There is in general


more than one solution curve, the correct choice being determined by


the conditions at x = 0.  Physically a jump from one solution to another


is possible as long as momentum, mass, and heat are conserved without


a gain in mechanical energy.   The system of equations may also have


singular points where some or all of the derivitives are infinite.


This study will not treat the possibility of jumps in the jet solution


since downstream conditions must be specified.


3.5  The Deflected Jet


        3.5.1  The Governing Equations


        This section treats a jet discharging into an ambient cross


flow.   Figure 3-5 defines a coordinate system (x, y, z) which is orien-


ted along the centerline of the jet.  The angle between the x and y


axis is  8,  the jet being discharged with an initial angle 9 .  The


                              -80-

-------
Region //I  —  continuity




            du
              Q

        rs —7—  +   rv  -  sw    =0
            dx        s     r





Region # 2  -  continuity
             duch          dr

             -dx~  +   SUc  dJ   + vbhll + S Iwr -"sz
Region // 3  -   continuity




            ducb
        I, r -,C   +  ru -p-  -  w.bl,  -r  [v  -au]  =  0
         Idx        cdx      hi         s    y c





Region // 4  -   continuity





     du hb

I,   —§—  + u I, [b  %L + h 5s-|  [w -a u ]I.b -  (v -a u  )I1h -  0
 1     dx       c  1 L  dx      dxj  l h  szcj 1      b  y c' 1






Region // 1 - momentum


      j                               9
                            pa
     du  2h        n J                          r   dAT h2
                               r H    r           d      1
U |2rs -~ + r  v  - s w 1 + ags -r~AT -=- +   I,,r — AT h   =  0
 cL   dx        s       rj  —~—Idx  c ^        J  dx   c J





Region  //  2 - momentum





     du  h       _ ,                          r    QAI n            ,  i

s  I~ —§— + s u    -r^- + u v, hi. + u sw  + ags  I.  —-7=— + I,AT h -~\m
   2dx      cdx    cb2    cr   —'°— L 4   dx      3  c  dxj

                                              a




Region  //  3 - momentum





     duc2b       2 ds                      r13 dATcbr2     2  dATcbh
r  I0 —3— + r u    -7— - u w, bl,, - u rv +aa 1	+1  r	—
   2dx      cdx    ch2    c   s —fi |_ 2    dx       3     dx




            Table 3-1.  Integrated  Equations for  Buoyant Jets



                               -81-

-------
 Region //  4  -  momentum


        2                                                           2
   2  du hb      o      j       H                           r     dAT h b

 X    --+  u     h    + b    ] +       (    - vh) +
                                                        ,  -  .   dx

                                                      a
            ,  AT h2 4s" +  I,2  AT hb ~ |  =  0
            4   c   dx     3
                      Chbfl  .
                      c    dxj
y-momentum
__

dx
heat:
~
where
        ^f
               'n
                      hI7) (s + bI7)J  -f KATc (s
                   = (  (1 -C3/2) d£ =  .4500
                          - f

                            fn
      f1  2         ^       3/2 4
I2  = \  f (^)dc  = \  (1 - C '  ) dC  =  .3160

      'o
    = \  t(5)d£  = f

      »n           'n
                              (1 - C)d;  =  -6000
           Table 3-1.   Integrated Equations for Buoyant Jets (.cont'd)





                             -32-

-------
      (  [  t(c)  dcdc  = [   I
      Jo >c              'o /
I   =       t(c) dcdc  =    I  (1 - C)0

        (I  - I )e     r  ?0

                      r = 0                        pauc
                                                       hn
                                                       1
 e =  spreading rate of a free turbulent region =  .22
   Table  3-1.   Integrated Equations  for  Buoyant  Jets  (cont'd)


                       -83-

-------
db

dx
u
AT
    o



=   b  =  0







=   e





=   u
    o



=   AT
Table 3-1.  Integrated Equations for Buoyant Jets  (cont'd)
                     -84-

-------
                                                            Cross  Flow = V -  V(x)
I
00
             Top

            View
                                Figure 3-5.  Characteristics of a Deflected Jet

-------
cross  flow velocity, V, may be a function of x but not y.



         It is  assumed that the deflection of the jet is sufficiently



gradual  that locally a cylindrical coordinate system nay be defined in


                                      ae ~1
which  rde  = dx,  dr = -dy, and r = -  ~



         The governing equations for buoyant jets,  (3.92 - 3.96)  then



contain  extra  terms which account for the jet curvature:
                3v  +       +  v      =0                         (3.107)
                By     3z        3x                                      '
  2                              /•—00
t -Tt    f\ _-w_?t-   r\ ^.Ti       *> rt     — -_ f    £\ AT1

                                 f


                                                                   (3.108)
jluy  , _3v^    jyw    -2 _3_6    ~2 ^6 = a£ ( °° 3AT   .   J^ _.^d.   3v]2

r\ T  ""•*" n. "—   "•" ^T^-   ~"~U  «*•  i  V  n."'      I   'i-,   U ti    -  !5^r      ^iv
                       3x       3x   P  ) ~  3y        P   3y      3y
                       UA             "z            a
             8v>w'   + a'2  li  - v'2 |^                            (3.109)
               3z           3x        ox
                                                                     .
        3uAT  A  3vAT    ,  3wAT  _      Sv'AT*      3w' AT

        --  +  ~~"     "~  "  "   "
                                                                   (
         In  the  following treatment the effect of  the  cross  flow on the



jet will be expressed in terms of integrations of  the equations of






                                     -86-

-------
motion over the entire jet.  These results will be incorporated into
the structure of the theory for the undeflected buoyant jet.  A major
assumption of the following development is that the cross flow velo-

cities are small compared with the jet centerline velocities i.e. defin-
ing a cross flow scale velocity y*:
        ^* -  5                                                 (3-112)
        If the x and y equations  (3.108 and 3.109) are integrated over
the entire jet  (see Appendix II),
o       s+b
                        ' dvdz
                          y
            fd      fS+b     _            fo       fs+b    f- a-
        -1  \       I       u  ^~z =M  (                 ^
        dx  J-(r+h) /-(s+b)           pa  J-(r+h)  •'-(s+b)/z dx
                          +  q  V  cos  6                           (3.113)
         ,    /o       /"S+b
         a   (        f
             •'-(r+h)  >-(s
where  q   is  the  total  entrainment into the jet per unit distance along
the  x  axis.   The derivation in AppendixII assumes 1)  symmetry of the
jet  about the y  = 0  plane and  2)  a potential flow outside the jet which
is equivalent to assuming that there is no pressure drag force on the

jet.
         Equations 3.113 and 3.114 are valid for the whole jet.  The

                             -87-

-------
 effect  of the  cross  flow  is  1)  to  introduce an additional  term in the



 ic momentum equation  to  account  for entrainment of cross flow x momen-



 tum  and  2)  a bending  of the  jet  so as  to balance the entrainment of



 cross flow y momentum.  The  cross  flow has the result of making the



 boundary  conditions on velocity  unsymmetrical around the jet boundary



 so that strictly speaking Equations 3.113 and 3.114 do not hold for



 any  separate region of  the jet.  The jet will no longer be symmetrical



 and  the analysis of the preceding  sections on buoyant and non-buoyant



 jets cannot be applied.



        It has been assumed, however, that for small cross flow velo-



 cities relative to the jet velocity the distortion of the jet is small.



 To apply  the governing equations to the four separate jet regions the



 following  is assumed also:



        1) Equation 3.113 which holds for the entire jet is assumed



           to hold for each of the four jet regions separately,



           i.e. the effect of the  cross flow on the integrated x



           momentum equation (3.108) is to add a. term which is



           the product of  Vcos6   and the entrainment per unit length



           into the jet region being considered.



        2) The similarity functions for velocity and temperature



           within the jet are all  the same as for the undeflected



           jet except for u which  is given by:
        u  =  u F F   +  V cos 6
               c y z
where   F  and F  are as defined in (3.46),
         y      z
                                -88-

-------
        3)  The boundary conditions are the same as for the undeflected



           jet except the conditions on the velocity at the jet boun-



           dary which are:
        w = w  - V cos 6 -rr        0 < y < s
             e           dx            J
        w = w f(£ ) -V cos 6 -pr    s < y < b
        v = v  + V cos 6 -rr        -r < z <  n
             e           dx
        v-vf(z;)+V cos 6-7^    -h < z < -r
             e   z           dx
                                                z  = -r-h
                                                                 (3.116)
                                                y  =  s + b
where w  and v  are related to the centerline velocity, u  , by the



entrainment coefficients a  and a  as defined for the undeflected jet.
                          sz     y


        4) Equation 3.114 which holds for the entire jet is assumed



           to hold for each half of the jet on either side of the



           surface y = 0.  If the y momentum equation  (3.109) is then



           integrated over the half jet as in Section  3.4.3, and 3.114



           is used, the result is a lateral spreading  equation  iden-



           tical with  (3.95)  in the deflected coordinate system.  The



           lateral spreading  velocity  is  assumed  to be distributed



           as in Section 3.4.3 taking  into account the  different dis-



           tribution for u  i.e. for s
-------
                                                                 (3.117)
and v = 0 elsewhere.


                                                            fab'
        For deflected jets, the non-buoyant spreading rate,



is no longer a constant but is defined by the assumption that the en-



trainment coefficients a  and a  are the same, in the deflected non-
                        y      sz


buoyant jet as in the non-deflected, non-buoyant jet.  It is also as-
sumed that  -r  Fl    is negligible i.e. that a non-buoyant, deflected
            dx  (dxJNB


jet boundary will not have a large curvature.



        5) In the continuity Equation (3.107), the curvature term



           is negligible with respect to the other terms.  Equation



           3.114 implies that:
           d 6 *

        ^-  - *  S                                         (3-118>
Thus this assumption is consistent with the basic premise (3.112).



        With the above assumptions, the integrated equations governing



a deflected, buoyant jet may be obtained with (3.114) determining the



new variable 9.  The coordinate system x, y, z may also be related to



the x,y,z system by geometrical considerations.  The complete equation



set is given in Table 3-2.



3.6  Bottom Slopes



        The presence of a solid boundary complicates the derivation of



                              -90-

-------
   Region 1:  continuity   rs -r=-   [u  + Vcosfl] + rv  - sw   =0
                               dx     c              s     r






   Region 2:  continuity   s [—  [h(u I  + Vcos6)] +   (u  + Vcos6) -^  + w  - a  uj  + v hi  = 0
                               QX      C J.                C          QX     TL    S Z Cr     D  X





   Region 3:  continuity   r [—  [b(u I  + VcosS)]  +  (u  + VcosG) -p - v  +au]  - w bl  = 0
                               QX      C J,                C          CtX    S    y C      JT  J_





   Region 4:  continuity   -pr  fhb(u I,  + VcosO)]   + (u I,  + VcosO)   |b ~ + h ~\  + (w,  - a  u )I,b
     0                  J    dx  L     c 1         /J       c 1            L  dx     dxj    x h    sz c' 1

T                                   - (v.  - a u ) I h = 0
                                        b    y c   1




                                             d                                     d       r2     d
   Region 1:   x momentum  (u + Vcos6)[2rs ~ [u  + Vcos9] + rv  - sw  ] + —& s  [-p- (AT   —)+I r— (AT h) J

                                                                             a
                                                                                                      0
                               [jo                   99                     9   H T
                              -pr  [h (u  I. + 2Vcos6u I, + V cos 6)]  +   [u  +  VcosG]    -£
                              dx       c  2          c 1                   c             dx




                                               dAT h2               -

   +  w  [u  + Vcos9] - a  u Vcos6   + ^*-  [I,  —T:	  + I0AT h -—•  ]| + v,h  (u 3
       re             szc         p     4dx       3  c  dx  j    b    c
                  Table 3-2.   Integrated Equations for Deflected  Buoyant  Jets

-------
Region 3:  x momentum    r  -r=-   [b  (u   !„  4- 2Vcos6u I  + V cos  0)]  + [u  + Vcose]  -jl-  -
                            LQ.A       C   fc           C X                   C           Q.3C



                                                .,                                r I    dAT br2

                     v  [u  + VcosQ] + a  u  Vcose  - w.b  [u  I0 +  VcoselJ + ^   ~  —7^	
                      sc              yc    J     hc2          1    p   U 2     dx
                                                                              SL



                                      dAT bh             2
                              .  _. £       ^«     |   . fji   t      1_  "f
                           j         92             222            2                  99
 Region A:  x momentum    -r=-  Ihb(u  I0   + 2Vcos6u I,  -f V  cos  9)]   +  [u  I. +  2Vcos0u I,+V cos  6]
   °                      dx       c  L            cl                   Ic2           Cl





                              [  dr  ,   .  i~.
                              Jb  -r^  +  " "3T
                              L  dx       dx_




                                               dAT h'b
                                                           .
                                     p  L 3 4    dx        Ac    dx    3    c    dx
                                      cl
Jet  y momentum           ~  fe  - B|]  ] Cu/bl^r + hI2) + 2Vcos6ucbI5(r  + hlj_) + V2cos20b(r+h)]

                               •          nJK •
                               - ^ AT  ( \  +

                                 Po   C    2
                     Table  3-2.   Integrated Equations  for Deflected Buoyant Jets  (cont'd)

-------
        Jet heat
df  f
     ucATc(s + bV  (r + hI7)  + VcoseATc 
                                            (s + bI3)  =  0
        Jet bending
      (s + bI2)  (r + hI2) +  2Vcos6uc  (s + bl^  (r
 I

?
        Jet x position
                                    V2cos20(s-l-b)  (r + h)J ^| - ucVsin0[-as2(s + bl^  + a  (r + h^
        Jet y position
  •
dx
    -  cose  -  0
       where:
                      ' I7 are as defined in Tatle 3-1
                        Table 3-2.  Integrated Equations for Deflected Buoyant Jets  (cont'd)

-------
                                db
                                —    =  non-buoyant spreading  rate  calculated from above equations
                                dx
                                   NB
                                         with AT = 0
                                              s  > 0
                                               s  = 0
i
VD
4>
                       a
                                     - I)e
                                              r = 0
                                     a   = a  exp 1 -i.O
                                      sz        *
                     [-'•'
                                                                                       a c
                        e  =   spreading rate of a free turbulent region = .22
                        r  =  h
                             db

                             dx

                             u
                                                                 u
                        X
 o

y  =  h
AT  = AT

 e  =  e
                 I
                                                                            J
                                                                                    x  =  0
                      Table 3-2.   Integrated Equations for Deflected Buoyant Jets (cont'd)

-------
the governing equations since a number of the basic assumptions made



for no slopes are no longer valid as discussed in Chapter 2.  The jet



will remain attached to the bottom until adverse pressure gradients or



buoyancy effects cause separation, at which time vertical entrainment,



until then inhibited, will begin.  This study will not attempt a theory



which predicts the point of separation due to adverse pressure grad-



ients.  It is assumed that the jet will not separate from the bottom,



over the entire width of the jet, until the buoyant effects of lateral



spreading cause the slope of the bottom of the jet to be less than the



bottom slope i.e.
        ^   +   ^  <  S                                        (3.119)
        dx       dx      x
Before  this point, vertical entrainment, w  , is zero and the jet is



attached  to the bottom:
        •     +    ^=5                                        (3.120)
        dx       dx       x
 In  addition,  the presence  of  the bottom  invalidates  the  expressions  for



 pressure  gradients  since the  pressure  gradient  at  z  = -°° is  no  longer



 known.  It  is assumed  that until buoyant separation  occurs that the



 buoyant terms have  no  effect  on  the  jet.



        The formulation of Equation  3.120 in the pre-separation zone,



 requires  the relaxation of one of  the  other governing equations.   It



 is  assumed  that frictional losses  are  negligible and that the x momen-



 tum equations are valid.   Since  the  non-buoyant y-equations result in




                                 -95-

-------
  •^    =  T~ which  is  known  to be  true only for jets free from solid
  dx      dx

boundaries,  the  y-equation  is dropped.


        The  calculation of  a jet over a bottom slope proceeds by testing


at  each x step whether, if  the bottom slope were not present, the ineq-


uality  3.119 would hold.  If the jet is spreading less than the bottom


slope,  calculation proceeds as if the bottom were not there.   This


assumes that once separation occurs, the losses in the entrained flow


between the jet and the bottom are small.   If the jet is spreading


more rapidly than the bottom allows, Equation 3.120 replaces  the y-


equation and the calculation proceeds with that set of equations.  This


formulation is valid with and without cross flows.


3.7  Solution Method


        The solution of the governing equations is done using a fourth


order Runge-Kutta integration technique supplied in the IBM Fortran


Scientific Subroutine Package.  The routine chooses a step size to meet


a specified error bound for each step.  The calling program keeps track


of which region the calculation is in i.e. whether r or s are zero and


supplies the appropriate equations.  The actual IBM routine was modified


slightly to prevent negative values of the physically positive vari-


ables.   A listing of the program and an input format is given in Appen-



dix I.


        The results of the calculations are output in dimensionless


form with u , AT  and /h b   being the normalizing velocity,  tempera-


ture and length respectively.
                                  -96-

-------
IV.  Experimental Equipment and Procedure




4.1  Previous Experiments




        As mentioned in Chapter 2, experimental studies of heated sur-




face discharges have been done by Wiegel, Jen, Stefan, Tamai, and




Hayashi.  Some of the data from these experiments are compared with the




theory of this study in the next chapter.  An examination of the experi-




mental techniques of the previous investigators is useful in designing




improved experiments and in avoiding duplication of effort.  Table 4-1




gives a summary of the characteristics of each previous experimental




study:




        1) Interest has clearly focused on low densimetric Froude




           number jets.  Only Stefan considers a case where




           IF < 1.0 and a wedge of ambient water intrudes into the




           discharge channel.  Good control of discharge flow and




           temperature is necessary if IF is to be held constant




           during the course of an experiment.  Wiegel1s and Jen's




           experiments have a large variation in F for each  case.




        2) The magnitudes of the  Reynolds numbers  in  Table 4-1




           reflect  the difficulty of achieving a large Reynolds




           number in a laboratory experiment, especially when low




           Froude numbers are required.  Wiegel, Stefan and  Hayashi




           report values of  Hclose  to or below 3000,  the  minimum




           value for fully developed turbulence in circular  jets




           as  discussed  in Chapter  2.




         3) The dimensions  of  the  receiving  basins  are typical of







                               -97-

-------
    Investigator   IF
                                              Boundary
                                               „,         Maximum
                                               Flow
                           Receiving  Basin   Control     x/A b
                                                             o o
                                                           Measurements
     Jen  (21)    18-180     8300-21000
                                                                        Thermocouples
                            26' x  15' x  1.5'     none   .010'
-------
   university laboratory facilities except for Tamai who




   uses a narrow channel.   It is likely that the side walls




   of Tamai's channel severely limit lateral mixing and




   spreading, thus distorting the jet behavior.  For this




   reason, Tamai1s data is not discussed in Chapter 5.  None




   of the experiments considered cross currents in the re-




   ceiving basin.




4) Only Stefan provides for a disposal of the heated discharge




   at the boundary of the basin.  Such control is necessary




   if steady state conditions are to be achieved during an




   experimental run.  Boundary flow control may also reduce




   eddying of the ambient water in the far corners of the




   receiving basins.




5) Low densimetric Froude number discharges require large




   channel dimensions if Reynolds numbers are  to remain




   high.  The available scaled distance,  x//h b   , decreases




   as IF decreases.  Many of  Stefan's results are for  x//h b




   less than 50.  In such a  limited distance the differences




   in jet behavior between separate runs  are small  and it




   is difficult to determine the effects  of  the controlling




   parameters.




6) As noted  by Wiegel,  buoyant  surface jets  have a  tendency




   to meander in  the horizontal plane  about  a  mean  centerline.




   A fast  scanning multi-probe  temperature  measurement system





                        -99-

-------
            is best  for  obtaining  an  accurate  three-dinensional




            determination  of  temperature.  Measurements with a sin-




            gle movable  probe such as  Stefan uses must account for




            the unsteady nature  of the flow.




         7)  The velocity of  the  discharge in the laboratory is




            usually  less than 1  ft/sec and decreases rapidly away




            from the jet origin.   A three-dimensional velocity deter-




            mination requires  a  probe  capable of measuring velocities




            less than  .1 ft/sec.   Only  Stefan uses such a probe during




            the experimental  runs.




         8)  None of the  previous experiments involve control of the




            surface heat exchange  from  the water in the receiving




            basin.   This requires  a room in which temperature and




            humidity are regulated.  This is usually difficult to




           achieve in a room large enough to hold the experimental




           basin necessary for jet studies.





        In light of the above discussion of previous studies, the exper-




iments in this investigation are  designed with the following objectives:




        1) Close control of discharge  flow and temperature.




        2) Reynolds numbers greater than 10,000.




        3) Boundary flow control which allows operation of the




           experiment without a significant change in the steady




           state behavior.
        4) Values of x//h b  greater than 50.
                                   -100-

-------
        5) An accurately positioned,  fast scanning temperature




           measurement system.




        6) The capability for generating cross flows in the




           receiving basin.




        7) Addition of a sloping bottom in the receiving basin.





The measurement of velocity and the complete control of heat transfer




are considered beyond the resources of this study.  However, the labora-




tory is air conditioned so the ambient temperature remains reasonably




constant.




4.2  Experimental Setup




        Figure 4-1 shows the arrangement of the experimental equipment.




The receiving basin is 46' x 28' x 1.51.  All the walls within the




basin, including the discharge channel, are constructed of marine ply-




wood.  The adjustable sluice walls along the sides are held by c-clamps.




The probe platform rests on four leveling jacks which may be reached




from the sides of the basin.  Point guages at each corner of the plat-




form assure accurate control of the platform level.  The sloping bottom




is a lightweight aluminum frame over which industrial plastic  sheeting




is stretched.  The water level in  the basin is monitored by a  fixed




point guage.




        The heated discharge is supplied from a steam heat  exchanger




capable of delivering up to 60 gallons per minute of water  at  a cons-




tant temperature up to 150°F.  The flow  is introduced into  the discharge




channel  through a gravel filter to insure a uniform velocity distribu-




tion in  the channel.



                              -101-

-------
Ba
1
O
NJ
1
t>" Pipe.

9


Sluice Wall


Side Channel







Hose End
Support
Frame
8' Manii Id 0 LI iy***^
1 — * u Pip| D'D "











1=^. To Pumping System
1
/i
B
B
• • : • 4- - :
' PVl Pipe
-•:••-?-• V\V« 	 	 	 	 	 : 	 __
tintlflrmmimyr " - " -^
U-*D ^-Hose Manifold
Wooden Probe Frame
y Sluice Wall
v ^*\








1
A
»— •
jBl""








Leveling Jack
^Aluminum

anne ^



c
	 Discharge Channel
travel Filter
C
j
anger
hannel Bottom Channel
> » ,M^£^ *i-
'
1
Adjustable Wai 1
/^xFixed Wall
HP-R B u
l...ftr
-rrrrrl^r
c-c
1
c
f-" Pipe
^
" L
^^ I.1 Pumping Svstem
•> It.
                                                    scale
                                    B-B
Figure 4-1.   Experimental  Set-Up

-------
        Boundary flow control is established at the sides and at the




rear of the basin.  Water may be pumped into or out of each side chan-




nel to create a cross flow in the receiving basin.  At the rear of the




basin there is a manifold consisting of 40 lengths of 3/4" hose with




one end mounted on the basin and the other end tapped into a 30' x 8"




PVC pipe outside the wall.  The position and orientation of the hose




ends in the basin are adjustable along a frame suspended parallel to




the back wall.




        A five horsepower centrifugal pump is used to provide two modes




of boundary control.  Water may be withdrawn from the rear of the basin




through the. hose manifold and returned to the side channels or a cross




flow may be established in the basin by pumping from one side channel




to the other.  A drain line and a cold water inflow line are connected




to the pumping system to permit replacement of heated water withdrawn




from the basin by new colder water.




        All flows are monitored by Brooks rotameter flowmeters except




for the drain and cold water input flows which are measured with sim-




ple Venturi meters.  Most of the pipe fittings and valves  are schedule




80 PVC plastic.  Water hose is used in place of pipe whenever possible.




        The temperature measurement system consists of  67  Yellow Springs




Instrument #401  thermistor probes, 59 of which are mounted on the basin




platform  (see Figure 4-2) and  the  remainder  in the various inflow and




outflow pipes.   The  probes are  connected  to  a Digitec  (United Systems)




scanning  and  printing  system capable  of  a  scanning rate of 1 probe  per




second.   The  temperature  in  degrees Fahrenheit is printed  on a  paper

-------

Figure 4-2.  Mounted Thermistor Temperature Probes
                    (scale in inches)
                    -104-

-------
tape.   The probes have an accuracy of about -f . 25°F with an interchange-




ability of + .05°F.   The response time constant  of the probes is 7 sec-




onds which filters out turbulent fluctuations.   The probes are leveled




using the water surface as a reference plane.




        The laboratory building is air conditioned and the temperature




is relatively constant.  This has the advantage  that surface heat loss




rates do not vary greatly.  The wet and dry bulb air temperatures are




measured using two thermistor probes, one of which is covered by a




water saturated gauze strip, mounted in an air flow provided by a small




blower.  The wet bulb and dry bulb temperatures are constant at about




b6°F and 73°F from day to day.




4.3  Experimental Procedures




        An entire experimental run is completed in about one hour.  Sur-




face temperatures are scanned initially to insure that no ambient temp-




erature gradients are present.  The basin boundary flows are then es-




tablished.  In the case of no cross flow, the back wall manifold  is




primed and an equal  flow  to  each  side wall of 50  gpm  is established in




all cases.  The  back manifold does not function well  as a selective




withdrawal device and cannot withdraw from a surface  layer without suc-




tion of air.  It is  used  solely  as a means for maintaining  the  basin




level  and  for preventing  large  eddies in  the ambient  fluid.  The  ver-




tical  position of the  hose  ends  is maintained at  about  .5  ft.  from  the




basin  bottom, pointing  toward  the discharge.  A very  slow ambient cur-




rent  from  the  side walls  to the  back  is  set  up which  reduces eddying




but does  not  distort the  jet structure.   The capability of  the pumping
                                -105-

-------
system  to replace the manifold withdrawal with cold water is not used




at all  except  to maintain  the basin level by discharging into the drain




line a  flow equal to the channel discharge.




        The channel flow is established by first turning on a bypass




flow from the  heat exchanger to a drain and adjusting the temperature




of this flow to the desired value.  The flow to the discharge channel




is then gradually increased to the proper rate.




        Table  4-2 gives a  summary of the parameters of each run.  The




Reynolds number of all runs is near 10,000 and the density differences




are all about  .004, corresponding to a temperature difference of about




25°F.   In Runs 1-10 only the densimetric Froude number and the channel




geometry are varied with no bottom slope or cross flow.  In Runs 11-19




the bottom slope is placed in the basin and varied from 1/100 to 1/25




for several values of IF.  The last runs, 20-25, investigate the case




of an ambient  cross flow, without bottom slope, over the entire width




of the  basin.  To avoid difficulty in comparing the results of differ-




ent cross flow runs,  the same cross flow was maintained continuously




during Runs 20-25.   The distribution of cross flow velocity, measured




by timing submerged floats, is shown in Figure 4-3.




4.4  Data Analysis




        The data from each experiment consists of flow rates, tempera-




tures in the jet,  and temperatures of the various inflows and outflows




from the basin.  The ambient temperature is defined as the initial basin




temperature.   A variation in the surface temperature at the extremities




of the basin is determined to be about 0.5°F per hour but the sub-sur-





                               -106-

-------
Physical Variables
Run

No.
1

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
h (ft)
0

.290
(.417)
.258
.164
.452
.355
.079
.088
.081
.164
.083
.246
.246
.246
.355
.355
.355
.079
.079
.079
.164
.164
.164
.164
.164
.164
b (ft)
0

.380

.315
.188
.104
.063
.229
.104
.063
.042
.042
.313
.313
.313
.063
.063
.063
.043
.043
.043
.188
.188
.188
.042
.044
.042
ft
u
o sec

.24
(.14)
.19
.32
.35
.50
.48
.58
.72
.95
1.06
.20
.20
.20
.60
.60
.60
1.25
1.25
1.25
.33
.22
.21
.74
.98
.74
T °F
a

71.5

69.9
72.1
71.2
71.8
73.3
72.3
72.3
72.6
71.8
73.7
73.3
72.8
73.2
74.0
74.2
70.4
70.8
71.2
70.9
73.1
72.8
72.1
71.5
72.6
AT °F
o

25.3

25.6
24.3
25.8
24.1
27.4
26.1
27.0
23.7
27.6
23.6
23.9
24.3
26.3
25.6
25.1
28.0
27.6
27.2
32.2
36.1
15.4
34.1
31.2
15.5
Ap/p xlO3
3.

4.10

4.05
3.95
4.19
3.89
4.65
4.31
4.50
3.86
4.60
3.91
3.94
3.99
4.42
4.33
4.24
4.55
4.50
4.45
5.45
6.50
2.35
6.14
5.29
2.35
V ft

max sec

0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.085
.085
.085
.085
.085
.085
Dimensionless
S
X

00

00
CO
OO
OO
00
00
OO
OO
OO
.01
.02
.04
.01
.02
.04
.01
.02
.04
OO
OO
CO
OO
OO
OO
IRxlO
o

1.5
(1.1)
1.1
1.1
1.2
1.1
1.1
1.1
1.0
1.3
1.2
1.1
1.1
1.1
1.3
1.3
1.3
1.4
1.4
1.4
1.1
.8
.7
.9
1.3
.9
4 F
o

1.00
(.58)
1.03
1.83
1.42
2.35
4.40
5.22
6.53
6.60
9.55
1.13
1.13
1.13
2.70
2.70
2.70
11.60
11.60
11.60
1.16
1.20
1.88
4.10
5.85
6.60
Parameters
h /b
o o

.76
(1.10)
.82
.87
4.30
5.70
.35
.84
1.30
3.95
2.00
.78
.78
.78
5.70
5.70
5.70
1.90
1.90
1.90
.87
.87
.87
3.95
3.95
3.95
V
max

u
o
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.26
.39
.40
.12
.09
.12
K n.5
— xlO
u
o
8.5

10.5
6.2
5.8
4.0
4.2
3.5
2.8
2.1
1.9
10.0
10.0
10.0
3.3
3.3
3.3
1.6
1.6
1.6
6.2
9.0
9.5
2.8
2.0
2.8
Note:  a) All cross flows as in Figure 4-3.                     _5
       b) Surface heat loss coefficient, K, assumed to be 2 x 10   ft/sec in all cases  (see Chapter 5)
       c) For Run 1 numbers in parenthesis indicate values upstream of cold water wedge.
d) Reynolds number, JR, defined as 4u
where
hydraulic radius
                                                                                b h /(h + b ).
                               Table 4-2.  Summary of Experimental Runs

-------
   14
 oo
•H
 M
O


 a
 o

-------
face ambient temperature is constant over the time of the experiment.



Figures 4-4 and 4-5 are a sample of the three-dimensional temperature



distribution measured by the movable probes for Run No. 10 and Run No.



22.  The data anlaysis resulted in centerline temperature rises and jet



widths and depths as follows:



        1) The lateral profiles of surface temperature at all x are



           plotted i.e. T vs. y at all x.  Centerline temperature



           rises above the ambient, AT  , are immediately available



           as in the position, y , of  the centerline in the case



           of deflected jets.



        2) The half width of  the jet,  b^, is  the distance between



           the centerline and the point where the surface temperature



           rise is half the centerline temperature rise.  Actually



           the average of the left and right half widths  is deter-



           mined in the data  analysis.



        3) At each value of x centerline  temperatures  at  each depth



           are determined and plotted  i.e.  AT   vs. z at all x.   The



           half depth, h... ,„,  is  the  depth at which the centerline



           temperature rise  is half  the  surface centerline  temperature



           rise.




 The temperature  rises  and  distances  obtained by these  processes are nor-



 malized by the initial   temperature  rise, T  -  T ,  and the length,
                                            O    <1


 /h b ,  respectively.
   o o
                              -109-

-------
                                              70.8   • 74.2
       • -"7.4   • Bj!4   • 81.0   • 82.7  • 82.6  • 83.3   • 82.2
                                Surface Temperature Contours (z - 0)
                                  .ertical Section at y = .25 ft.
Vertical
 Scale
                                          .? «71.7» 71.7» 71.6   «71.5
                                         Section at x • 5.4 ft.
     Figure  4-4.   Temperature Distribution  Measured  for Run No.  3

                                   Probe Position  •
                                                                                                •76. V
                 70.6  •  70.9   • 74.0  •  77.2  •  77.7   W77
          71.S   • 71.7/» 81.8   • 80.4  • 80.5  • 80.5   • 81.8
                                     »83.3    «82.5   «82.0
         t> 0   «92 4/  • 87.5^/084.7   • 84.7  «83.5    «84.7
         S5T5
                                                                                                 •BQ.6
                                                                                                 •80.3
                                  . 3  •  80.6 • 80.5   • 81.0
               ^71.7   • 71.7   • 71.6  • 71.6  »71.7   • 71.
                                                                                                 •71.7
                                        -110-

-------
                77.6
 72
                                           78.7  »78.6/  • 77.8
                                                            • 77.7
                               ft.
                                                                     76.7
                                                                     77.1
Figure 4-5.   Surface (z = 0) Temperature Distribution Measured
              for Run 22 - With Cross Flow
                      Probe  Position  •
                         -111-

-------
V.  Discussion of Results


5.1  The Controlling Parameters


        This chapter is a comparison of the results of the laboratory



experiments, in this study and others, with the calculations of the theor-


etical model developed in Chapter 3.  The response of the characteris-


tics of the heated discharge to the dimensionless parameters describing


the discharge and the ambient conditions is examined.


        The physical variables characterizing each discharge may be com-


bined into the following dimensionless parameters (see Table 4-2):
                u              u

        IF =  	  =   	~	  =  densimetric Froude number
          o

        A = h /b   =  discharge channel aspect ratio
             o  o
        —  *  surface heat loss parameter
        u
         o
        S   =  bottom slope
         x
        —  =  cross flow parameter
        The centerline velocity,  centerline temperature and the shape of


the jet are functions of the above parameters and of the distance from



the origin:

                                 -112-

-------
   tu    /\T      ,        u
    c     c     b       h       r          ,

   u~ »  ~AT~~ '  ,-"'--" '  r~—  ' /     >  ;••-•'-'  =   function
    o     o   Jh b    yh b    /h b   yh. b
              voo     oo     oo     oo
                           }-
               A   JL    s    _L      _*	
            9   "• 1     )  ^ 9      )

                    o          o     v h b -
                                       o o
        The internal velocities v . w , v,  and w. are not determined ex
                                 s   r   b       h


plicitly by the theory nor are they measured in  the experiments.  The



values of u and AT away from the centerline are  calculated using the



similarity functions f and t as in Equations 3.46 and 3.89.



        Other interesting dependent variables which are also  functions



of x may be defined.  In the buoyant jet the vertical entrainment is a



function of a local densimetric Froude number:
        IF  =   	  =  local densimetric Froude number

                 'agAT h
where u  ,  AT  , and h are local values  at a given  distance  from the  ori-



gin.  The  total  flow in the jet may be determined by  integrating the x



velocity over the jet  cross section.   The ratio of the flow at a given



x to the initial flow, u h b  , is  called the  jet  dilution:
                        o o o
                -s+b

                              dzdy
/-Sf D  fO



\           U
/o    y-h-r
                   u h  b                      u h b
                    o o o                      o o o
                                -113-
                                       uc(r + IjhHs + Ijb)

                                                     	   =  dilution

-------
 where the similarity function for u CEquation 3.451  is used  in the inte-


 gration.   The effect of surface heat loss upon the heated discharge may


 be evaluated by calculating the excess heat flow in  the jet  where  excess


 heat is defined as  heat which raises the water temperature above ambient


 A parameter, analogous  to dilution,  may be defined as  the ratio of  the


 local excess heat flow to the initial  excess heat flow, pc AT u h b •
                                                          p  o  o o  o"
  H
         ss+bro

     PC            uATdzdy              _h)(s+I_b)
         yo   ^-h-r          _    c   c     7       7
       PCP AToUohobo              AToUohobo            initial excess heat
 5.2  An  Example  of a Theoretical  Computation


         Only  b,  h, and  AT  are  determined in the experiments.  Thus


 values of UG, r,  s, IF   D and H cannot be compared with the theory.


 Figures  5-1 and  5-2 show the calculated results for Run No. 6  (see Table


 4-2) which includes all the variables and which illustrates the follow-


 ing general structure of heated surface discharges.


         1) A core region in which the centerline velocity and  the


 centerline temperature rise decrease very slightly.  The lateral spread


 of the jet is very large, requiring an initial rise of the jet bottom


 boundary to satisfy mass conservation.  The dilution, D, and the local


 densimetric Froude number,  F  , do not vary greatly in this region.   The


magnitude of E is much larger than F because the initial depth of the


 turbulent region, h,  is zero.   There is no significant surface heat loss


 in the core region.


        2) An entrainment region in which the centerline velocity and



                              -114-

-------
           Core Ke&iun
                              Entrapment Region      Region    Regie
               Cor*
             ^, Region	Entninaent Region
Vamt Lois
 ••lion
    -trr
Figure  5-1.   Theoretical Calculation  of Jet Structure for Run 6.

                IF •»  4.4    K/u  -  4.2 x  10~5   V/u  - 0
                  O              O                      O

                A -  .35     S   - 0
                               -115-

-------
                                    90    100   0    10    20    30   40   SO   40
Figure 5-2.   Calculated  Isotherms of AT /AT   for  Run 6.
  0                                      co
              IF  - 4.4
               o


              A - .35
K/u  - 4.2  x  10
   o


S  - 0
                                               -5
V/u  - 0.0
                          -116-

-------
temperature drop sharply, approximately as 1/-X, as in the non-buoyant




jet.  The jet spreads downward by turbulent processes at a rate of the




order dh/dx r .22.  The lateral growth is dominated by gravitational




spreading at a much greater rate than the vertical turbulent spread.




Because of this large ratio of lateral to vertical spread, the jet reach-




es a maximum depth beyond which the bottom boundary rises to maintain




mass conservation.  Local densimetric Froude numbers in this region de-




crease rapidly and the dilution rises sharply as a result of entrain-




ment.  Surface heat loss is still negligible in this region i.e. H =




1.0.




        3) A stable region in which vertical entrainment is inhibited




by vertical stability as indicated by the  local densimetric Froude num-




ber which is of order one or less.  The jet depth continues to decrease




because of lateral spreading.  The small jet depths reduce  the lateral




entrainment, the  dilution and centerline temperature remaining relatively




constant in this  region.  The centerline velocity, however, drops sharply




as a consequence  of the  large lateral spread.  The surface  temperature




pattern, as shown in Figure 5-2,  is dominated by  the wide,  constant  temp-




erature stable  region.




        4) A heat loss region marking the  end  of  the stable region.




The  lateral spread is sufficiently large  to allow significant  surface




heat transfer and the temperature begins  to fall  again.   Once  surface




heat loss has begun  to be  important,  the  rate  of  temperature  decrease




±3  very  rapid and in conjunction with the low  centerline velocity the




discharge may no  longer  be considered as  a jet.







                                -117-

-------
         5)  A far field region dominated by ambient  conyectiye and diffus-


 ive  processes which are beyond the scope cf this  study.  If ambient flow


 velocities  are large,  the far field region may begin before the stable


 region has  formed.   The excess discharge heat is  ultimately lost to the


 atmosphere  in the  far  field.


 5.3  Comparison of  Theory and Experiment


         In  the previous section the general behavior of heated discharges


 is discussed.   Many of the characteristics of the discharges cannot be


 generalized and must be determined by  examining each case.  This sec-


 tion, in addition  to comparing the data with the  theory, examines the


 behavior of the jet as a function  of the controlling parameters discus-


 sed  in Section 5.1.


         Chapter 4 includes a  description of the determination of AT ,


 h, ,? and b.. ,„  from  the experimental  data.   It should be noted that the


 half width,  b1/2> and  nalf depth,  h-,-,  are defined in terms of the temp-


 erature  distribution rather than by  the velocity distribution as dis-


 cussed in Chapter 2.   The  determination of the half width, b  ,   is


 difficult because of a)  distortion of  the surface temperature distribu-


 tion in  the  lateral  portions  of the  jet  by recirculation of heated water


within the  experimental  basin and  b) the small temperature differences


 in the jet  far  from  the  discharge  point.


         In  the  theoretical computations  for  all the laboratory cases,


 the surface  heat loss  coefficient  is assumed to be 2 x 10   ft/sec.

                                                   2
 corresponding  to a  heat  transfer rate  of  110 BTU/ft -day-F°. This is in


 the low  range  cf values  predicted  from  standard formulas (3 )  and is
                               -118-

-------
close to Hayashi's (.15) estimate for laboratory environments, 1.8 x 10



ft/sec.



        In Runs 1-10 the densimetric Froude number, IF  , and the aspect



ratio, A - h /b , are varied.  There is no bottom slope or cross flow.



Figures 5-3 to 5-12 show the experimental data and the theoretical com-



putations.  For low values of F the jet spreading is too large to per-



mit determination of the half width, b.. ,_.



        Runs 1 and 2 illustrate the case of F -1.0 which is the lowest
                                              o


possible value of IF .  In Run No. 1 the channel densimetric Froude



number is actually less than one and a  cold water wedge is observed in



the channel.  As discussed in Chapter 2, the wedge will force the heated



flow  to obtain f = 1.0 at the  discharge point.  The value of u  and h
                 O                    of                     O      O


to be used in the theoretical calculation is  determined by this condi-



tion.  There is  little difference between the temperature distribution



for the wedge and non-wedge  runs.  In Run 2 the "toe" of the wedge is



positioned at the channel expansion and the vertical structure of the



jet is approximately the same as if the "toe" had  intruded into the



channel.  For such low IF discharges the entrainment region  is very



small, the temperature of the stable region is high, and the heat loss



is promoted by large lateral spreading.



        Runs 3 and 4 show the effect of the discharge channel aspect



ratio, A, on low IF  discharges  in which horizontal entrainment dominates



vertical  entrainment.  Large aspect ratios provide larger  available  area



for  lateral entrainment  and the temperature drops  more  quickly.   Sur-



face  heat loss behaves in an opposite  manner, being promoted by the
                               -119-

-------
    1.0




    !i

    wo



     .5
              CmurliM T«p*ntur« U««: 4T lit
J	1      I  I  I I 11
                                                    I   I  I  I I  I I
                                                              100
                                                             1—r
                                                                                Too
       -J
                    -5
                                           100
                                               »«H-D.pth; -t»
                                                  Theory   -



                                               Tot»l D«j>thi -(r
                                                  Thaory
                                                                        200
Figure 5-3.   Theoretical  Calculations  and Experimental  Data  for Run  1.



               IF -  1.00  (wedge present)     K/u   •= 8.5 x  10       V/u   « Q
               A «•  .76
                                 S   » «
                                  x
                                     -120-

-------
   iT
    c

   4T
         Centerline Temperature Rl»e:





             Experiment  •
               J	L
                                till
                                             J	I	I	I	I  I I  I I
                                                                            J	I	L
                                                              150
          -2
          -3
Half-Depth:  -h. ,,/
        	LI t.  o o



    Exp«ri**nt ^



    Theory   —



Total Depth:  -(r -f h)//h~b~
_. „ _ _             o o
                                                      Theory
Figure  5-4.   Theoretical Calculations and Experimental Data for Run 2.



                IF  - 1.03      K/u   - 10.5 x  10~5      V/u  -  0
                  o                  o                            o


                A  - .82        S   -  »
                                   -121-

-------
tTc
"o
.5

.1





^^oX.
^"Vnrv
Ccnterlin* Tecnpcrftturc Si»t: fil /4T
Experiment •
Theory . •
i 11,111 i iii
1 10
0
Ch.urxl
lotto»
-' 1
/TT
0 0

-2
-•)
100
«0
60
40
20
>
0 ,5^1
o ^ so 100
S i i
'V0^x'x
Half Dtpth:
-
^
-
1 1 1 1 1 1 II
100 '•'^\^ soo
ISO «/-'\^ 200
1
-h1/2//sx
ExptrlMnt •
Theory
Total Depth:
Theory
1 1
0 SO 100
/ I Half width:
.
o o

1
ISO n/-^^ JOO
1
"i/2//E?:
/
f j Exp«rl»*f)t 0
/ / • Theory .
~ I / Total Width
// •• Theory
Channal
Sid*
: (• + k)/r1i^

1






re 5-5. Theoretical Calculations and Experimental Data for Run 3


F - 1.83 K/u - 6.2 x 10~5
0 O
A = .87 S » »
X
-122-
V/u - 0
o


-------
          Experiment



          Theory  ••
                                                                 100
                                                                                      500
       -3
              Channel

              Bottom
                                             100
                                                             150     x//h b    200
                                                                      o o
                                                      H.lf Depth:  - h1/2'''h
-------
          C«nt
-------
1.0


IT
  c

ST
           Experiment



           Theory
                                I I  I
                            50

                           —r
                                            100
                                                            150
                                                  X//5TT"     200
                                                     o o
       Channel

       Bottom

           77
                                                     Half Depth:  -h, ,,/A b
                                                               II L  o O



                                                        Experlnent  Q



                                                        Theory      i



                                                     Total Depth:  -(r + h)/A b
                                                     	o o
                                                        Theory
       80
                                             100

                                            T-
                                            150

                                           —r
                                                                              200
                                                      Half Width: b. .,/|h b
                                                      	lit  o o



                                                          Experiment   0






                                                      Total Width:  <» + V>lJ\C\T
                                                      	o o
                                                          Theory
Figure 5-8.
Theoretical  Calculations  and Experimental Data  for Run 6.


                                      -5
                 IF  -  4.4
                   o


                 A -  .35
                  K/u   -  4.2  x  10
                   S  - »
                    x

                  -125-
V/u   - 0

-------
                                                                                  500
        Channel
        Bottom
       -1
       -2
       -3
                             50
                  \
                  Experiment  £)
                  Theory   _^^«__
               Total Deptu: -(r + h)/A b
                                                   Theory
                                                                         200
Figure 5-9.   Theoretical Calculations and Experimental Data  for Run 7
                                                  -5
               P  = 5.22
                o
               A = .84
K/u   = 3.5  x 10
   o
V/u
                                 -126-

-------
                                                                  200
                                               Half Width:  b. .,//h b
                                                        i/z  o o
                                                  Experiment
         Channel

          Side
Figure 5-10.   Theoretical Calculations  and Experimental Data for Run 8.
                IF  - 6.53
                 o


                A - 1.30
  K/u  -  2.8 x 10





  S  • oo

   X


-127-
                                                 -5
V/uo  - 0

-------
         C«nt«rlln* Tenpcracarc Rise: AT /AT
       -3
                                                  lUlfUldU,:  b  /.VT
          ltd*
Figure 5-11.   Theoretical Calculations and  Experimental Data  for Run  9,


                                                 ,-5
                F  - 6.60
                 o


                A » 3.95
K/u   - 2.1 x  10
   o
                                  X


                                -128-
V/u   = 0
   o

-------
         Centerline Temperature Rise:  AT /AT
      0 0


      60
        Channel

         Side
                                       100
                                                Half Width: b
                                                         . ,
                                                         iff  0 O
                                                   Experiment



                                                   Theory
Figure 5-12.
Theoretical Calculations  and Experimental  Data  for Run 10,


                                   -5
                IF  - 9.55
                  o


                A - 2.00
                  K/u   = 1.9  x 10
                     o


                  S  -  »
                   x


                 -129-
V/u  - 0
    o

-------
 large horizontal surface areas in low aspect ratio discharges.   The




 theory does not reproduce the centerline temperature drop  for large as-




 pect  ratios as  well as  for smaller values.




         The same principle is illustrated in Runs  5 and  6.  An  increase




 in aspect  ratio may outweigh an increase in  IF  as  shown  in Figures  5-7




 and 5-8.   In Runs 7-10  (Figures 5-9 to 5-12)  the effect  of aspect ratio




 and densimetric Froude  number is diminished  as  IF  increases.  The jet




 depths  are much larger  and heat loss  is not  as  significant as in the




 discharges of lower ]F  .




         Runs 1-10 indicate that the theory agrees well with the data




 and is  capable  of predicting the structure of heated  discharges as  a




 function of F  and A.   The experimental error is greatest  in the deter-




 mination of b, ,-  an(* least for  AT .   Since the  theory agrees best with




 the data for ATc  and worst for  b, ,2 some  of the discrepancy between the




 data and the theory, especially for the half widths, can be attributed




 to experimental error.




         In  the  runs involving bottom slopes (Runs 11-19)  (Figures 5-13




 to 5-15) only surface temperatures are measured because of the diffi-




 culty of positioning probes near  the solid bottom.   The bottom slope




has two counteracting effects upon the jet:




        1)  Vertical entrainment is inhibited by the presence  of  the




           solid boundary.




        2)  The vertical core region of depth r persists as  long  as




           the jet is attached to the bottom.   This promotes  lateral




           entrainment over the full depth of the jet.




                                -130-

-------
1.0



ATc

AT"
  o



 .5




         Centerline Temperature Rise:   AT /ATQ
Experiment  Run 11: •



           Run 12:•



           Run 13: 4



Theory




   I      I    	I
                                     - 1/100



                                     - 1/50



                                     - 1/25
                                   (value of S  shown)
                                    i  i  i i
                              i   i   i  i  i  i
                                          10
                                                                                100
                                                                                                           500
        Figure 5-13.   Theoretical Calculations  and Experimental Data for Runs  11, 12 and 13.
                       F -  1.13
                         O


                       A - .78
K/u   - 10.0 x  10
   O


V/u   - 0
   o
                                                         "5
                                                     Run  11:   S  - 1/100
                                                                X


                                                     Run  12:   S  - 1/50
                                                                x


                                                     Run  13:   S  - 1/25

-------
1.0
                         sx - 1/100



                         Sf • 1/50



                         S, - 1/25



                         (value of S shown)
                                10
                                          100
                                 150


                                T-
                                                               x//iTT"
                                                                  o o
                                                200
                                     Half Wldch: b,,-/Sh~b~
                                     _ _ if &  O O
               Experiment   Run 14:



                        Run 15:



                        Run 16:



               Theory  — — ^_ —
                                                         S  - 1/100



                                                         S  - 1/50



                                                         S^ - 1/2S



                                                         (value of S shown)
        Figure  5-14.
        IF  - 2.7
          o

        A  - 5.7
Theoretical Calculations and Experimental

Data for Runs  14,  15 and 16.
K/u  -  3.3 x 10
    o

V/u  -  0
    o
                                              -5
                             Run  14:   S  -  1/100
                                         X,

                             Run  15:   S  -  1/50


                             Run  16:   S  -  1/25
                                  -132-

-------
.5
       Centsrline Temperature Rise:  AT /AT
       __	 c  o
          Experiment  Run 17:  0 S - 1/100




                   Run IS:  • Sx - 1/50




                   Run 1»:  * Sx - 1/25




          Theory  ___^^_   (vslue of S shown)







         __l	1	L
                                          J	1	I	I	I  I  I  I
                                                                         I      I'
                                                               100
                            50
                                                           ISO
                                                                            200
      80
      60
             Hslf Wtdth:  b
                       . ,,
                       1/2  o o
Experiment  Run 17: ^  S - 1/100




         Run IB: B  S^ - I/SO



         Run 19: A  S^ - 1/25



Theory  — — _   (value of S, «horo)
       Figure 5-15:
        F  -  11.6
         o

        A  -  1.9
        Theoretical Calculations and Experimental

        Data  for Runs  17,  18 and 19.



           K/u  - 1.6 x 10~5      Run 17:   S   - 1/K
               o                                  x
               u


           V/u  -  0
               o
Run 17:   S   - 1/100


Run 18:   S   - 1/50
            x

Run 19:   S   - 1/25
                                     -133-

-------
 The resultant effect of these two phenomena is dependent upon the values




 of ]P  and A.   Runs 11-13 indicate that for low IF  the jet detaches from




 the bottom almost  immediately and that the centerline temperature rise




 is nearly the same for all the slopes considered and is  near  the  value




 computed  for  no bottom slope.   Runs 14-16 (Figure 5-14)  show  that for




 large  aspect  ratios and low IF  the bottom slope actually promotes higher




 entrainment than without a slope.  As with Runs 11-13 the data  shows no




 dependency upon the value of the bottom slope,  indicating early separa-




 tion of the jet from the bottom.   In Runs 17-19 (Figure  5-15) the reduc-




 tion of vertical entrainment increases as the  slope  decreases.  This be-




 havior is a consequence of the  large densimetric Froude  number  and mod-




 erate  aspect  ratio for these cases.   The jet remains  attached to  the




 bottom and the dilution is dependent  upon the value  of the slope.




        The theory reproduces  the trends  in the data  as  discussed above.




 Th'e low IFo computations (Runs 11-16)  show more  dependency on the  value




 of Sx  than do  the  data,  indicating  that  the theory is not predicting sep-




 aration of the jet  sufficiently close  to  the jet origin.  For the large




 IF  runs (17-19)  the  theory underestimates  the effect of  the slope on the




 reduction  of entrainment.  The theory  fails to  reproduce  the lateral




 spreading  of the jet which is seen  to be  close  to the calculated curve




 for no slope rather  than for the  very  small spread the theory predicts




when a slope is present.  The assumption put forth in Section 3.6 that




 the jet does not exhibit buoyant  spreading  is clearly not valid.  The




jet must be spreading laterally at  the edges and remaining somewhat




 attached to the bottom near  the centerline, distorting the jet beyond
                                  -134-

-------
the capability of the theory to describe it with simple similarity func-



tions .



        In Cases 20-25 a cross flow deflected the jet.   Determination of



the jet centerline position from the temperature data is very difficult



since the total lateral temperature variation within the experimental



measurement area and far from the discharge origin is only a few degrees.



This is particularly true of the runs with low values of 3F .  The agree-



ment of the experimental and calculated jet trajectories are scattered



but good on the average (see Figures 5-16 to 5-21).  The measured center-



line temperature values are also close to the theoretical  curves, being



not much different than the no-cross flow cases with the same IF and A.
                                                                o


As expected from the formulation of the equations governing deflected


                                                     y
jets,  the jet deflection increases with IF , A, and  — since the only

                                                     o

force  turning the jet  is entrainment of lateral momentum.  The  theoreti-



cal calculations are terminated when u  = Vcos6 since  the  basic assump-



tion of small cross flow velocities does not hold beyond  that point.



The theoretical  calculations  used  an analytical expression for  the  cross



flow velocity as a function of x  (see Figure 4-3).



5.4  Comparison  of Theory with Experimental Data by  Other Investigators



        As mentioned  in Chapters  2 and  4  a number  of other studies



collected  field  and  laboratory data for three-dimensional heated surface



discharges.   These  results  are compared with the  theory in this section.



Data from  Stefan (31), Hoopes (18) and  Tamai (32)  are not used.  Stefan's



data points  are too  close to the jet origin in most cases to permit exam-



ination of trends in the data; Hoopes1  field temperature measurements
                               -135-

-------
  1.0



  AT
  	c

  AT
         Center line Temperature Rift:  ATc/
            Experiment



            Theory   —
             J	L
                            '  '  ' '
                                                J	I	I	I  I  I I I
                                          1	T
      100
                            50


                            —T
        100

       —r~
                       150
                                                   Foiitton of Centerllne



                                                       Experiment  0
          200
Figure 5-16.   Theoretical Calculations  and  Experimental Data for Run  20.
                 F  -  1.16
                  o


                 S  -  .87
                  x
K/u  - 6.2 x  10"
    o


S   - <*>
 x
Vmax/uo  - '26
                                   •J36-

-------
4T
	£

4T
           Experiment



           Theory
                                               I    1   1  I  1  I  I 1
      80
      A\ b
       o o
        Channel

         Side
                           50
                                           100
                                                          150    X/.VTT    200
                                                 Position of Centcrline



                                                     Experioeot  9



                                                     Theory  ———
  Figure 5-17.   Theoretical Calculations and  Experimental Data for Run 21.



                  _  » j.. ^u        K./U  •  y.u x  J.T
                    o


                  A - .87
K/UQ - 9.0 x  10"
V     - u  -  .39
 max     o
                                       x
                                    -137-

-------
    1.0

    n
    .5
           Centerllne Temper Cure MM:




              Experiment  «
               J	L
                                                        I  I I  I  I I
                                                         J	1   I.
                                   10
                                                                100
                                                                                    300
        100
         10
        to
        *0
                              50

                             ~r
                                            150

                                           T~
                                               Position of C«nt«rllD«
                                                  Theory
Figure 5-18.  Theoretical  Calculations  and  Experimental  Data  for Run 22.
IF  - 1.88
  o


A - .87
                                    K/u  -  9.5 x  10
                                        o


                                    S   - »
                                     x
                                                      -5
Vmax/uo * •*
                                   -138-

-------
         Centerllne T«np*rature Rlae: AT /AT
                                 10
                                                                                  500
        100
        SO
        60
                  Po«ltlon of C«nt«rllg«
                     Theory
                                                                          200
          ChttMl

           S14*
Figure 5-19.  Theoretical Calculations and Experimental Data for Run  23.
                V  - 4.10
                  o


                A - 3.95
K/u  - 2.8 x 10
    o
                                                     -5
V    /u  - .12
 max  o
                                    -139-

-------
                                                                                SOO
         100
        /h b
          o o


         60
         20
                             50


                            T
Poiition of Ccnt«rlint



   Ixpcriiwnt  %


   Theory  ^—_
          Chaaiul

          Slit
Figure 5-20.   Theoretical  Calculations and Experimental Data for  Run 24.
                F  -  5.85
                 o


                A - 3.95
                K/u   - 2.0 x 10
                   o


                S  -  »
                 x
                                                   -5
Vmax/uo -  -09
                                 -140-

-------
  1.0
   .5
         Cmttrllm Ttap«ntur> Rlit: AT /AT,
        20
Figure  5-21.  Theoretical Calculations and Experimental Data for Run  25.
               TF -  6.60
                 o


               A  - 3.95
K/u  -  2.8 x 10
   o


S  « co

 X
                                                 -5
V    /u  - .12
 max  o
                                   •441-

-------
 are  too  close  to ambient values  to  be  useful;  and Taniai's  data is  sever-



 ely  affected by the  fact that  the experimental basin is  no more than 40



 times as wide  as the discharge orifice.   This  inhibits the lateral en-



 trainment which is so  important  in  three-dimensional heated discharges.



        Jen's  (21) experiments are  >'one  wit  circular orifices and values



 of IF from 18  to 180.  He  found  that the centerline  temperature drop,



 beyond the core region, is  independent of IF and  given empirically by
         AT       14R
where R  is the initial jet radius.  For IF =  °° ,  the temperature  rise
       o                                   o


beyond the core region predicted by  the theory  of  this study  is:
         AT        10.7,/h b

              -   —^                                       «•«
which is valid for rectangular discharge channels.  The experimental



data for Jen indicates that the theoretical expressions for  I" = »



are approximately valid for IF  > 15.  The quantity v/hb   is  the square



root of half the discharge channel area.  If it is assumed that for a



fully submerged circular discharge pipe the proper scaling is also the



square root of half the discharge area, then the quantity >/h^b~ is re-



placed by R  /•£  = 1.25 R .   Then the expression for the centerline temp-
r       J  o V2          o                                             r


erature rise beyond the core region (IF = <=°) and a fully submerged circu-



lar discharge pipe becomes,  according to the theory of this  study:
                               -142-

-------
        AT       13.4 R

        ^.  -   	2.    (approximately valid for I" ?> 15)     C5.3)
        AJ-          X                                 O
which is almost identical with Jen's empirical result (5.1).  (see Fig-


ure 5-22)  For circular discharges the aspect ratio should be taken as


2.0.


        Wiegel (37) studied rectangular orifices with aspect ratios of


1/2 and 1/5 and W from 20 to 30.  Figure 5-23 shows that his data, when


normalized by /h b   falls close to the line given by Equation 5.2.


Figure 5-23 also shows Wiegel's data for a discharge of F  from 20 to 30,


aspect ratio of 1/5, and bottom slopes from 1/200 to 1/50.  The beginn-


ing of the entrainment region is delayed by the  inhibition  of vertical


entrainment.  Later dilution  is seen to be a  function of  the slope value.


The theory of this study satisfactorily reproduces  the  delay in  initial


entrainment but not the bahavior farther from the origin.


         Hayashi  (16) studied  heated discharges from rectangular  orifices


into  stagnant receiving water with no  cross  flow.   His  data and  the


corresponding theoretical  calculations are shown in Figure  5-24.   The


data  scatter  is  large  and  the agreement mostly qualitative. The results


illustrate  the property  of heated  discharges  that  a lower densimetric


Froude number produces less dilution  in  the  entrainment region but higher


eventual heat  loss far from the discharge  point.


5.5  Summary  of  Jet Properties


         The previous  sections have considered the general structure of
                               -143-

-------
fe
r
       1.0



       AT
       AT   -
        .5  _
                                                                       I    I   I  I  I  I  I
                                                                                                 1           I
                Centerline Temperature Rise:  AT /AT
                                                            I       I    I    I   I   I
                                               10
                                   100   x/R /»/2
                                           o
                                                                                                               500
                Figure  5-22.   Theoretical Calculations and Experimental Data for  Circular Orifices  -

                               Jen  (21)


                               IF - 18 - 180      K/u  =* .5 x 10~5
                                o


                               A - 2.0
                      V/u  - 0
                          o
S  -

 X

-------
fe
r
          — Centerline Temperature Rise:   AT /AT
                            (values of A and S  shown)
       .1
                                                                                     100    x//h b
                                                                                              o o
                                                                                  500
              Figure 5-23.   Theoretical Calculations  and Experimental Data for Rectangular Orifices -

                             Wiegel  (36)


                                              K/u  -  1.0 x 10~5        V/u  = 0
F  - 20-30
 o
                                                • o

-------
1.0




AT


AT"
  o




 .5
           Theory



              This Study



              Hayashl





                I      I
      Centerllne Temperature Rise:   AT /AT
           Experiment   O  IF * 1.
                          F - 2.6
                           o
    (value of ff  shown)
             o
I   I   I  I  I  I
     I	I
         F » 1.4

I    I   I  •"!  I '
                                                         I  •  2.6
                                                          o
I     I
                                          10
                                                                               100
                                                                h b
                                                                 o o
                                                                                                          500
       Figure 5-24.   Theoretical Calculations and Experimental Data for Rectangular Orifices -

                       Hayashi (16)
                          - 1.4 and 2.6
                        o


                       A - 2.0
K/u  - 10.& x 10
   o


S  « eo
                                                                   -5
                                             V/u   - 0
                                                o

-------
heated discharges, the dependence of the detailed structure of the dis-



charge on the dimensionless parameters Tf  , A, K/u , S  and 7/u , and
                                        O         O   JC        O


the ability of the theory of this study to reproduce experimental res-



ults.  The theory successfully describes the jet behavior except for



lateral spreading in the presence of a bottom slope.  In this section



the theoretical calculations are used to quantify some of the general



statements in the previous sections.



        The dilution and corresponding centerline temperature in the



stable region are of interest since the stable region is a significant



portion of the surface temperature distribution as previously discussed.



The centerline temperature rise may be related to the dilution by assum-



ing that no heat has been lost i.e. H = 1.0  and  considering a position



beyond the core region  i.e.  r = s - 0.  The  definitions  of D and H may



then  be combined  to yield:
         AT       I -i i    i      -,  c
          c   -     1     !   -   i-5                                  /c ,,
                  I~J    n   ~   IT                                  (5.*)
         Figure 5-25 is a plot of the dilution in the stable region vs.



 T  with values of A indicated.   The points are the theoretical calcula-
  o


 tions for Runs 1-10 and for several of the previous investigations, all



 with no bottom slope or cross flow.  Some tentative lines indicating



 the curves for constant A are drawn but many more calculations are



 needed to improve the reliability of the plot.  Nevertheless it is clear



 that the ultimate stable dilution in a buoyant jet depends primarily



 upon W and to a lesser extent upon A.



                                -147-

-------
u
 Stable
            15
            10
                                        1.3
                                          84
                                     35
                                                  A-1.0
                                                           10
           Figure 5-25.  Dilution in the Stable Region of Buoyant Discharges
                        (Points are calculated by the theoretical model -
                         Values of A indicated)
                              -148-

-------
                                                 h

        The maximum depth of the discharge jet,   -5si. j  ia aiso o/f inter-

                                                /h b
                                                  o o

est and may be determined from the theoretical calculations.   Figure



5-26 indicates that the maximum vertical penetration of the jet is a



function of ]F with a very small dependence upon A.  The results are



approximated by the following relationship:
                  .5 P                                            (5.5)
                      o
        ftv b
         o o
        Other interesting features of the heated discharges  treated in



 this study are the values of x at which  the various jet regions begin.



 Unfortunately the boundaries of  the jet  regions are not well defined,



 transition zones being present between the regions.   The  lengths  of the



 core regions, x  and  x  , are not the  same as  for a non-buoyant  jet, in
               S      r


 which  they are functions only of A, but  are rather a  function of  both



 IF  and A in  the buoyant  discharge.  In the absence of a  great many cal-
  o


 culations from which  -to  make  general  plots,  the  detailed structure of



 each specific discharge  is  best  determined by a  separate theoretical



 calculation  for  each  case.



         In the previous  section, little  was  said about the effect of the



 heat loss parameter,  K/u ,  upon the jet  behavior.   For all but the low-



 est values of 3F  and A,  the theoretical  calculations indicate that heat



 loss has no significant effect until after a stable  region  has formed and



 the jet has undergone considerable spreading.  This  is also indicated by



 the experimental data.  Thus the value of K/u  will  affect  only  the rate





                                -149-

-------
 max


'h  b
 o o
                                                          10
                                            IF
         Figure 5-26.  Maximum Jet Depth Vs. F as Calculated by Theory
                            -150-

-------
at which the stable region loses heat.   Since jet-like behavior ceases




at this point, the value of K/u  may be said to have no effect upon the




region of the discharge considered in this study.  Surface heat loss may,




however, be important in the far field region as discussed in the follow-




ing chapter.
                               -151-

-------
 VI.   Applying the Theory




         This  chapter considers the use of the theoretical .model for sur-




 face  discharges  of heated water in meeting the objectives outlined in




 Chapter 1  with respect to prototype temperature prediction and the de-




 sign  of scale models of heated discharges.




 6.1   Prototype Temperature Prediction




         Prediction of temperatures in the vicinity of a specific dis-




 charge  configuration requires  the  parameters  IF  , A, S , K/u , and V/u
                                               o      x     o         Q


 be specified  in  addition to AT , T ,  and  the  discharge channel area.
                               O   £L


 The geometry  of  actual field configurations are  rarely as simple as




 those assumed  in the theoretical development  and judgement must be exer-




 cised in the  schematization of the discharge  to a less complex form




 for which  the  dimensionless parameters may be  given.  This section is




 a discussion of  the  data requirements  for  treatment of a heated surface




 discharge  by  the  theory  of  this study  and of  some schematization tech-




 niques.




        The following  are  the  physical data needed as input to the




 theoretical calculations:




        The ambient  temperature, T ,  is assumed in the theory to be




constant in space  and  time.  Actual ambient receiving water temperatures




are often  stratified vertically or horizontally and may be unsteady due




 to wind, tidal action  or diurnal variations in solar heating.   The nat-




ural stratification of the  ambient water may be increased by accumula-




tion of heat at  the water surface  if  the discharge is in a semi-enclosed




region.   The value of  the ambient  temperature, for a given initial temp-
                              -152-

-------
erature rise, &T , determines the initial density difference between the




discharged and ambient water.  Also tbe effectiveness of the entrainment




of ambient water in reducing the discharge temperatures is a function




of the ambient stratification.




        If the temperature differences resulting from ambient stratifica-




tion in the vicinity of the discharge are the same magnitude as the init-




ial discharge temperature rise, the theoretical model of this study




should not be applied without further development to account for the




stratification.  The theory  is valid if an ambient temperature, T  , may
                                                                 cl


be chosen which  is  representative  of the receiving water temperatures,




that is, if  the  temporal or  spatial variations in ambient  temperature




do not differ from  T  by more than a few degrees Fahrenheit.
                    &



        The  initial discharge temperature rise,  AT  ,  is determined from




the  discharge temperature, T ,  and the  chosen ambient temperature,  T  .




The  value of AT   will be equal  to  the  temperature  rise through the con-




densers only if  the intake  temperature  is equal  to  T .  Actual discharge




temperatures should be  relatively  steady unless  the power plant has sev-




eral different  condenser  designs in which case  the discharge temperature




will vary with  the  plant  load.   Use of the  theory requires that a con-




stant  value of  ATQ  be specified, the  choice being based on the likely




 steady value of the discharge temperature.




         The discharge channel geometry is important since the theory




 uses one half the square root of the discharge channel flow area as a




 scaling length.  Calculation of the discharge densimetric Froude number,




 JF  , requires specification of the initial depth, h  ,  and the aspect
                                -153-

-------
 ratio,  A,  requires both, h  and the initial  width,  b  .  Rectangular and



 circular discharge geometries are discussed in  Chapter 5.  The following



 procedure  is  suggested  for  any channel  shape:



        a)  Let  h   be  the actual maximum discharge  channel depth



            so that calculation of F  is not affected by the schema-



            tization.



        b)  Let  b   be  such that the correct  discharge channel area



            is preserved:
        ,      channel area

        bo "  	2h	
                    o
        Then the aspect ratio is given by:




            h            2hQ2


        A "  ~
            b       channel area
             o
        The discharge channel geometry may vary with time if the eleva-



tion of the receiving water changes due to tidal motion or other causes.



In this case separate calculations for each elevation of the receiving



water must be made, assuming that the steady state theory predicts the



instantaneous temperature distribution.



        The initial condenser water discharge velocity, u  is a function



of the power plant condenser water pumping rate and the channel area.



        The bottoir. slope, S  is usually not uniform as the theory assum-



es.   The average value used as input  to the theory should be close to



the  value of the slope near the discharge point since it is the initial





                               -154-

-------
portion of the discharge jet which is most affected by the bottom slope.




        The surface heat loss coefficient, K, may be determined from




meteorological data as discussed by Brady (3) who gives several useful




diagrams for estimation of K.




        The cross flow velocity, V, in the receiving water may be meas-




ured by drogue surveys or estimated from current records or flow measure-




ments  in the case of a river.  The theory accepts as input  (see Appen-




dix  I) any analytical function which gives V as a function of x.




        The dimensionless parameters IF   , A,  S  , K/u  , and V/u  are  func-
        	O  *   '  X*    O	 (3


tions  of the above physical  variables.   The  initial density difference




Ap   does not appear explicitly  in  the theory but is represented by a T  .




In calculating  the initial densimetric Froude number,  F  ,  the actual




value  of  Ap   should be  used  to  minimize  the error  in  the assumption  that




the  coefficient  of thermal expansion, a,  is constant.   This choice of




 Ap   is equivalent  to  choosing an average value  of  a  over the  temperature




range  T  to  T   + AT  .
        3.      ai     O



        When the discharge  channel densimetric  Froude number, IF  , is




 less than unity, a wedge of  ambient water intrudes into the discharge




 channel as discussed  in Chapter 2 and the heated flow will be forced to




 obtain a  densimetric  Froude number of unity at the discharge point  (see




 Figure 2-4).  The depth of the heated flow  in the presence of a wedge,




 h *, is given by:   (12)
  o





         h *
                                -155-

-------
where  h  and  IF  are based on the channel dimensions.   A  flow  area may  he
        o       o                                                     v


calculated  based on the depth h * and  the aspect  ratio calculated as des-



cribed previously in this section.   The  actual  discharge may  be directed



parallel  to a wall which will effectively cut off lateral  entraintnent



from one  side of the jet.   In this  case  the  discharge  may  be  schematized


by  assuming that the solid boundary is the centerline  of a jet whose



discharge channel is twice the width of  the  actual discharge  (see Figure



6-1).   The  width b  is  then twice  the width  calculated by  the procedure


discussed previously.



        More  complicated arrangements of the discharge channel relative


to  the boundaries  of the  ambient  receiving waters  are  beyond  the capab-



ilities of  the  theory of  this  study.  If the discharge channel is nearly


at  right  angles  to the  solid boundaries,  the theory may  be used as dev-


eloped, or  if  the  discharge is  directed  parallel  to a  straight boundary,


the schematization described above may be  used  (see Figure 6-1).  How-



ever,  if  it is not  clear whether the jet will entrain water from one or


both sides or if  irregularly shaped  solid  boundaries will  deflect the


jet or distort it  from  the  form assumed  in the  theory, a meaningful



schematization is  not possible.



        Once the schematization of the discharge configuration is achiev-


ed, the theoretical  calculation is performed by the computer program des-



cribed in Appendix I.   The inputs to the  program  for  each calculation


are F   , A, S , K/u  and V/u   (as a function of x/Vh b ).  The program



outputs the values  of ATc/ATQ, UC/UQ, b/J^Tb^, h/y^Tb^,  r/y^hlT, s/v^TT",


6, and the position  of  the jet centerline as a function  of the distance



along  the centerline.

                               -156-

-------
7  7  III
                    /  r  i f  i
                 90°
   ~  near 90°
   o
                 Discharges with Entrainment from  Both Sides
J
1
b
, 0
T f 1 t Tl I i I i • i i i i
              Discharges with Entrainment  from One Side
1 r





' ' ' i

1

f
i
\ /





f ' ' ' '
f

f

f
                                                    ™
        Schematization Possible
  Schematization not  Possible

Because of Irregular  Geometry
                         Discharges with Obstructions
           Figure 6-1.   Discharge  Channel  Schematization
                                        -157-

-------
 6.2  Modeling Heated  Discharges




         Laboratory scale models  are  used when the  determination  of  temp-




 eratures in  the vicinity of  a  prototype  heated discharge  is beyond  the




 capability of analytical temperature prediction techniques.  The theory




 of  buoyant discharges  developed  in this  study makes  the construction of




 a model  unnecessary if the region of interest is within the scope of the




 theory and if the  prototype  may  successfully  be schematized as discussed




 in  the previous section.




         Many  prototype situations may be divided into a near field  and




 a far field  region.   In the  near field region the  temperature distribu-




 tion  is  controlled by  jet entrainment, buoyant  spreading and the local




 ambient  cross flow.  The  theory of this  study is a treatment of  the




 near  field.   In the far field  region ambient  turbulent dispersion,  con-




 vection, and  surface heat loss determine  the  temperature distribution.




 The far  field  and near  field are not independent of each other;  the




 near  field local ambient conditions  are  a function of the far field




 temperature and flow distribution and the input of heat to the far  field




 is controlled  by the near field characteristics.




         Information may be needed about  the near field or the far field




or about both.  Thermal pollution regulations may be such that maximum




temperature levels must be met within the near  fiej.d.  On the other hand




an evaluation  of the effecLs of the heated discharge upon the natural




environment requires information about the far field temperature dis-




tribution.  Recirculation studies may involve either of the regions




depending upon the relative location of the intake and discharge.






                                -158-

-------
        The theory of this study may he used to analyze the prototype




temperature distribution if the near field temperatures are of interest.




Schematization of the prototype near field is possible within the limita-




tions discussed in the previous section.  In general the theory may be




applied if the far field ambient temperature is independent of the near




field structure and if the geometry in the vicinity of the discharge is




relatively simple.




        In the case that the prototype does not yield to schematization




or the far field conditions may not be estimated and may be a desired




result of the analysis, a model study is required.  If the far field




temperature distribution  is of  interest the model must reproduce not




only  the  far field physical phenomena but also  the manner  in which  the




near  field inputs heat  to  the  far  field.   In a heated discharge,  this




input is  characterized  by  the dilution  and  corresponding  temperature




of  the stable region.   A  model  which  is to  reproduce  the  prototype far




field temperature distribution  must as  a  minimum  requirement  have  a




near  field stable region  with  the  same  temperature  rise  and total  flow




at  the same  location as in the  prototype.




         The  physical dimensions and fluid flows in a model of a heated




discharge must  be scaled  from  the  prototype in a  manner which insures




similarity of the temperature  distribution between model and prototype.




Analytical statements of  the proper scaling are commonly called "model




laws" and will be different  depending upon which physical process dom-




 inates the distribution of heat.




         The characteristics of the near field behavior of buoyant jets







                                 -159-

-------
 are discussed in Chapter 5.  The diraensionless centerline controlling



 parameters are IF  ,  A, S  and y/u f surface heat loss being unimportant



 in the near field.  For similitude of model and prototype near field



 temperature distribution it is sufficient that IF  ,  A, S  and V/u  be
                                                 O	X        o


 the same in model and prototype.   If the following scale ratios are de-



 fined:



         z  • ratio of vertical lengths in model and  prototype



         L  « ratio of horizontal  lengths in model  and prototype



         u  - ratio of velocities  in model and prototype



        AT  • ratio of temperature rises in model and  prototype



        Ap  * ratio of density differences in model and prototype




 then the similarity  requirements  for the near field  become:
                                   1/2
                1.0  or  ur  =  (Aprzr)                                (6.4)
        Ar " Sxr = 1<0  or  Lr =  Zr
        Thus the model must be undistorted  (Equation 6.5) with the velo-



city ratio related to the length scale ratio and  the density ratio by



the well-known densimetric Froude model law (Equation 6.4).  In most



cases Ap  is not much different from unity  because of the problems of



working with very large or very small temperature differences in the



laboratory.



        The temperature and dilution in the stable region of a heated



discharge are functions of IF and A (see Figure 5-25).  Thus even if




                               -160-

-------
the detailed structure of the near field region is not  of  interest,  sim-



ilarity of near field heat input to the far field requires an undistorted,



densimetric Froude model.



        Distorted models have some distinct advantages  over undistorted



models especially in regard to laboratory space requirements.  For a



model to reproduce a large horizontal area of the prototype, making



z   > L  results in more workable water depths and higher model Reynolds



numbers than if  z  = L .  Distortion may be required by model laws for



similarity of heat loss  (14) or frictional effects (13) in the far



field.  If a distorted model is constructed, the aspect ratio and bottom



slope will be different  in the model than  in the prototype even if IF



is  the same.  If this were the only effect of  distortion, one might



attempt to adjust the model values of  IF to compensate  for  the distor-



tion of S  and A.  However, the ratios  of  dimensionless horizontal and
         X


vertical distances in model and prototype  are:
           0  O
        i ,	          	i       .1 „                            (6.6)
         y£ir
                                                                  (6.7)
 The distortion thus tends to make the dimensionless x and y values of any



 physical position less in the model and the dimensionless z distances



 more.  The model temperature distribution would be wider, longer, and





                              -161-

-------
 shallower than the prototype  by  a  factor  of Vz^/L  .  Varying IF in  the
 model arbitrarily may  produce the  sane  stable  temperature or ,flow as the
 prototype but will not counteract  the vertical  and horizontal distortion
 of  the  temperature field.  To compensate  for a  larger A in the model the
 value of F  must be lowered in the model, which will produce an even
 wider and shallower model temperature distribution.  It is concluded that
 it  is impossible to build a distorted scale model which reproduces cor-
 rectly  the minimum required characteristics of  the near field region of
 a heated surface discharge.
        This result has been  generally  realized in the past.  In certain
 cases a separate undistorted  near  field model has been constructed and
 the results used to adjust the discharge  configuration in a distorted
 far field model to reproduce  the stable region  as observed in the near
 field model.  The theory of this study  should make it possible to cir-
 cumvent the construction of near field models except in cases where
 geometrical  complications are  extreme and the theory cannot be used.
 6.3  A Scale Model Case Study
        A laboratory study of  a heated discharge was described by
 Harleraan and Stolzenbach (14)   in 1968.  The power plant is located on
 the coast of Massachusetts near Plymouth  (see Figure 6.2).  The primary
discharge configuration under  consideration was a surface discharge from
a trapezoidal open channel into a  coastal region having a relatively uni-
form bottom  slope.   The ambient cross current in the coastal zone was
primarily due to wind-driven circulation.   The  intake was directly down-
wind of the  discharge and was protected by two breakwaters (see Figure
6-2).
                              -162-

-------
 N
       1/2 MILES
CAPE COD BAY
SCALE
                                         -30

                                 Dischorge Channel
                     Pilgrim
                     Station Nuclear
                     Power Plant Site
                                     MANOMET
          FIGURE 6.2  LOCATION  OF POWER  PLANT

-------
        The  purpose of  the model  study was  twofold;   to determine  the



 extent  of  far field recirculation of  the heated water into  the  intake



 and  to  evaluate  the near  field  temperature  distribution.  Thus  both far



 and  near field regions  were of  interest.  Any analytical approach  to



 predicting the near field temperatures would have been frustrated  by the



 impossibility of determining whether  the discharge jet entrainment would



 be cut  off laterally by the breakwaters protecting the intake.  This



 was  a particularly strong possibility because the prevailing current



 would deflect the discharge toward the breakwaters.



        In the model study (AT  )  was determined from the arbitrarily



 chosen  condition (Ap)   = 1.0.   The horizontal length scale was L  = 1/240



 to allow reproduction of a sufficient length of coastline in the experi-



 mental  basin.  A vertical scale ratio was chosen as z  = 1/40 on the



 basis of far field surface heat loss similarity (see Ref.  14 ).  Harleman



 and  Stolzenbach justified the 6 to 1 distortion in the model on the



 basis of Wiegel's data  for bottom slopes with A = 1/5 (see Figure 5-23).



 This data indicated that the effects of the bottom slope and aspect



 ratio distortions tended to counteract each other with respect  to AT .
                                                                    c


 Figure  6-3 shows the temperature distribution measured in the model for



a typical tide and current condition.  The temperature distribution and



observed current patterns in the model indicated that the discharge jet



does entrain water from both sides.



        With the knowledge that the entrainraent in the jet is not later-



ally inhibited and using the theoretical results of the present study,



the model and prototype may be schematized as discussed in the previous
                               -164-

-------
  z = 0 ft.
 Scheme T
 QQ"  650 cfs
Surface Current
.6 knots
  z - -1.7 ft.
  z - -3.3 ft.
L
    Figure  6-3.  Temperature Rise Contours Measured
                 in  the Pilgrim Plant Model  Study

-------
 section.   The relevant parameters are:
  a

 AT
 u
 c


 V



 K
 IF
  o


 A



 K/u



 V/u
 Prototype



  65°F



28.7°F



 720 cfs



 6.7 ft



26.0 ft



 2.0 ft/sec



 1.0 ft/sec



 3.6 x 10~5 ft/sec

(200 BTU/ft2-day°F)



1/60



 2.0



 ,26



 1.8 x 10~5



 .5
  Modal



  77°F



  25"F



 5.2 gpm



 .17 ft



 .11 ft



 .32 ft/sec



 .16 ft/sec



 2 x 10~5  ft/sec

(110 BTU/ft2-daycF)



 1/10



 2.0



 1.5



 6 x 10~5



 .5
        The centerline surface  temperatures  observed  in the model are




shown in Figure 6-4.  The model data  is plotted  in prototype distances




(ft.) by scaling the model distances  with  the horizontal scale ratio



L  = 1/240.  A theoretical calculation using the  theory of this study




with the tabulated model parameters is also shown.  In the theoretical




calculation V/u  = 0 and S  = « since the  cross flow  in the model was
               O          J»



not significant in the near field region and the  large value of S (1/10)
                                                                 A


in the distorted model had little effect on such  a lev; densiraetric
                               -166-

-------
1.0
                                                            i    i    i   i   I  \  i
                                                                  500
                                                                             1OOO
                                                                                                       5000
                                                                  x - Dlcttnce Along Jet Centerline
                                                                        In Prototype (ft.)

        Figure 6-4.  Theoretical Calculations and Model Data  for Pilgrim Station

                                  Power  Plant Model  Study

-------
Froude number discharge.  The agreement between  the model data and the




theoretical calculation plotted in prototype  distances  Oft.) is good.




It should be kept in mind that the scheraatization of  the model for in-




put  to the theory is based upon the observed  two-sided  entrainment in




the  model.  Figure 6-4 also shows the  theoretical calculation for the




prototype parameters.  The model and prototype curves are close together




up to about 500 ft. from the discharge, but the  curves  diverge from that




point on.  The argument based on Wiegel's data is obviously limited by




the  fact that the F  in Wiegel's runs was 25  rather than 2.  The model




case shows the formation of a stable region which is apparently not




present in the prototype although the limitations of the theory with




respect to discharges with bottom slopes must be kept in mind.   The




model temperature distribution extends farther from the discharge than




that of the prototype by roughly a factor of  2 to 3 which is what Equa-




tion 6.6 predicts for z  = 6L  as was the case for this model.




        This case study indicates that the distorted model does not re-




produce the prototype temperature distribution although the results are




conservative since the model showed higher temperature  rises at corres-




ponding distances from the discharge channel.  Also the agreement of




model results and theoretical calculations is a verification that the




theory could be substituted for the construction and operation of the




model as far as near field temperatures are concerned.
                                -168-

-------
VII.  Summary and Conclusions




7.1  Objectives




        The use of natural waterways for disposal of power plant rejected




heat necessitates prediction of the induced temperature distribution in




the receiving water for:




        a) Evaluation of thermal effects upon the natural environment.




        b) Prevention of recirculation.




        c) Design of discharge structures to meet temperature standards.




        d) Scale model design.




        This study treats a three-dimensional horizontal surface dis-




charge of heated water from a rectangular open channel at  the surface




of  a large ambient body of water which  may have  a bottom  slope  and  a




cross flow at  right angles to  the  discharge.  Of interest is  the de-




pendence  of  the  temperature distribution in  the  vicinity  of the dis-




charge point upon  the  initial  temperature difference between the ambient




and discharged water,  the initial  discharge  velocity, the discharge




channel  dimensions,  the bottom slope,  the ambient cross  flow, and  sur-




face heat loss to  the  atmosphere.   The primary  objective is the predic-




tion of  the  temperature distribution within the near field region  domin-




ated by  turbulence generated by the discharge.   Temperature distributions




 in the far field region where ambient processes dominate are beyond the




 scope of the study.




 7.2  Theory and Experiments




         Previous knowledge related to heated discharges consists of




 established theories for isothermal turbulent jets, measurements of  the
                              -169-

-------
physical properties of turbulent jets, experimental work on stratified




turbulent flows,  and a sraall number of laboratory experiments on heated




surface jets.   Previous attempts by others to develop a theory for heat-




ed three-dimensional surface discharges have been limited by failure to




consider all the  factors affecting the temperature distribution in the




discharge.




        The theory of this study has the following structure:




        a) The discharge is basically a three-dimensional turbulent




           jet with a well defined turbulent region in which velocity




           and temperature are related to centerline values by simil-




           arity  functions.   Entrainment of ambient fluid into the




           turbulent region is proportional to the centerline velo-




           city by an entrainment coefficient determined from non-




           buoyant jet theory.   Unsheared core regions are accounted




           for in the theory.




        b) The effect of buoyancy on the discharge is to reduce ver-




           tical  entrainment and to promote lateral spreading of the




           of  the jet as a density current.   A major contribution




           of  this study is  the incorporation of  these buoyancy




           effects into the  turbulent jet theory.




        c) Surface heat loss is assumed to be proportional to the




           surface temperature rise above ambient  in the discharge




           and is accounted  for throughout the entire discharge




           length.




        d) The jet is deflected in the presence of an ambient
                             -170-

-------
           cross flow by entrainment  of  lateral momentum.




        e) A bottom slope is assumed  to  inhibit vertical  entrainment




           and buoyant lateral spreading as long  as  the discharge




           jet remains attached to the slope.




        The theoretical development synthesizes  the  results of previous




investigations of turbulent jets and stratified  flows. 'No new empirical




constants are introduced in the theory and the comparison of the theory




with experiments does not involve any "fitting"  of the theory to the




data.




        Experiments are performed in which all the controlling parameters




are varied.  Care is taken  to insure establishment of  fully turbulent




steady state conditions in  the discharge and accurate  three-dimensional




temperature measurements are  taken.  The data reduction  results  in cen-




terline temperatures, discharge depths  and widths, and centerline posi-




tion in the case of  an  ambient cross current.




7.3  Results




        A comparison of experimental data  from this  study and others




with the  theoretical calculations indicates  that the theory reproduces




the  centerline temperature rise,  the  jet width and  depth, and the  cen-




terline position.   When a bottom slope  is  present good agreement is




achieved  for centerline temperature  rises  although  the theory does not




predict  correctly  the lateral spreading of the  heated discharge.




         Particular theoretical results  are:




         a) For nonbuoyant jets (F  = °°) the discharge spreads linearly




            with distance from the origin.   The centerline temperature






                                  -171-

-------
   beyond the core region is giyen by;
   AT                                             1/2
   	£  _     ,    (1/2 x Area of Discharge Channel)^

   AT   ~  1J<*                 x
   This expression is valid for discharges in which IF > 15.



b) Buoyant jets (1 < IF  < 10) consist of a core region, an



   entrainment region, a stable region, and a heat loss



   region which marks the end of the jet and the beginning



   of the far field which is dominated by ambient processes.



c) Entrainment increases with increasing initial densimetric



   Froude number, IF  , and with discharge channel aspect



   ratio, A « h /b .   The temperature and diluted flow in
               o  o


   the jet in the stable region are functions of F  and A



   (see Figure 5-25).



d) The vertical spread of the jet is inhibited and the lateral



   spread increased by the buoyancy of the discharge.  Spread-



   ing increases with decreasing IF and the maximum vertical



   penetration of the jet is a function of IF  (see Figure 5-26).



e) Surface heat loss  has no significant effect on the behavior



   of the discharge in the core region, the entrainraent region



   or the stable region.



f) Ambient cross flows whose velocities are small relative to



   the initial discharge velocity deflect the discharge center-



   line but do not greatly affect the temperature distribution.
                        -172-

-------
       g) Bottom slopes tend to inhibit vertical entrainment for
          heated discharges in which IF is greater than about 5.
          Discharges with smaller values of F tend to separate
          from the bottom close to  the discharge point and show
          little dependence on S  ,  the value  of the slope.  There
                                X
          is some indication  that  in the region where  the jet is
          attached to  the bottom,  the  increased lateral area actually
          promotes increased  entrainment in  low 3F  discharges.
                                                  o
        This study of heated  surface discharges satisfies  its broad ob-
jectives to  the following extent:
        a) Evaluation of thermal  effects:   The theoretical model pre-
           dicts the  limits of the near field temperature distribution
           in which temperature rises are significant.  The diversion
           of natural flows by jet entrainment is also a result of
           the  theoretical calculations.
        b) Recirculation:  The dependence of  the maximum vertical
           discharge depth on ]F  (Figure 5-26) enables design of
           skimmer wall structures which require a knowledge of the
           depth of the heated layer at the intake.  The calculated
            dimensions  and position  of  the discharge jet aids in the
            location of an intake  for prevention of  recirculation.
         c)  Discharge design;  The  temperature and  shape of  the heated
            discharge is directly  related to  the densimetrie  Froude
            number,  IF  , and  the  aspect ratio, A,  of the  discharge
            channel by  the theory.   Theoretical calculations may be

                              -173-

-------
           used to evaluate the success of discharge configurations




           in meeting temperature standards.




        d) Scale models;  The theory indicates that near field temp-




           erature distributions are not reproduced in distorted models.




           If the prototype discharge configuration may be schematized




           as input to the theory, the theoretical calculations will




           make construction of an undistorted near field model unneces-




           sary.




7.4  Future Work




        The immediate need is for incorporation into the theory of some




of the more complex aspects of heated discharges:




        a) Stratified receiving water:   This requires more complicated




           assumptions for heat transfer within the jet and for buoy-




           ant spreading rates.




        b) Accumulation of heat  in the  far field:  If the far field is




           of finite size, the rejected heat will cause the far field




           surface temperature to rise  until equilibrium with surface




           heat loss is achieved.   The  resulting stratification must




           be described and its  interaction with the discharge entrain-




           ment and buoyant spreading accounted for as in (a) above.




        c) A far field model:   Buoyant  plume theories (4 )  (28) have




           already been postulated.   A  complete picture of the temp-




           erature distribution  in the  near and far field regions awaits




           the development of  a  theory  for the transition region between




           the near and far fields.






                             -174-

-------
        Much, of the field data on temperatures in the. vicinity of actual




prototype discharges is either unavailable to researchers or is of such




poor quality that it is unusable.  Improvements in the theory of this




study and the development of extensions as described above depend cruc-




ially upon the availability of accurate, complete field data for power




plant heated discharges.
                                -175-

-------
                          BIBLIOGRAPHY
 1.  Abramovich, G. N. , The Theory of Turbulent Jets, TheM.I.T.  Press,
     M.I.T. Cambridge, Massachusetts, 1963.

 2.  Batchelor, G. K., An Introduction to Fluid Dynamics, Cambridge
     University Press, 1967.

 3.  Brady, D. K. and others, "Surface Heat Exchange at Power Plant
     Cooling Lakes", Edison Electric Institute, Publication No.  69-901,
     November 1969.

 4.  Brooks, N. H., "Diffusion of Sewage Effluent in an Ocean Current",
     in Waste Disposal in the Marine Environment, Pergamon Press, 1960.

 5.  Carter, H. H., "A Preliminary Report on The Characteristics  of a
     Heated Jet Discharged Horizontally into a Traverse Current Part I
     - Constant Depth", Technical Report 61, Chesapeake Bay Institute,
     Johns Hopkins University, November 1969.

 6.  Cootner, P. H. and G. 0. G. Lof, Water Demand for Steam Electric
     Generation, Resources for the Future, Inc. , The Johns Hopkins
     Press, Baltimore, Maryland, 1965.

 7.  Deissler, R., "Turbulence in the Presence of a Vertical Body Force
     and Temperature Gradient", Journal of Geophysical Research,  Vol.
     67, No. 8, July 1962.

 8.  Edinger, J. E. and E. M. Polk, Jr., "Initial Mixing of Thermal
     Discharges into a Uniform Current", Report No. 1, Department of
     Environmental and Water Resources Engineering, Vanderbilt Univer-
     sity, October 1969.

 9.  Ellison, T. H. and J. S. Turner, "Turbulent Entrainment in
     Stratified Flows", Journal of Fluid Mechanics, Vol. 6, Part  3,
     October 1959.

10.  "Environmental Effects of Producing Electric Power", Hearings before
     the Joint Committee on Atomic Energy, Congress of the United States,
     U. S. Government Printing Office, 1969.

11.  Fan, Loh-Kien,"Turbulent Buoyant Jets into Stratified or Flowing
     Ambient Fluids", W. M.  Keck Laboratory of Hydraulics and Water
     Resources, California Institute of Technology, Report No. KH-R-15,
     June 1967.

12.  Harleman, D. R. F., "Stratified Flow", in Handbook of Fluid  Dynamics.
     V. L. Streeter, Ed., McGraw Hill Book Company, Inc., 1961.
                              -176-

-------
13.  Harleman, D.  R.  F.  and K.  D.  Stolzenhach,  "A Model  Study of Thermal
     Stratification Produced by Condenser Water Discharge", M.I.T.  Hydro-
     dynamics Laboratory Technical Report No. '107,  October 1967.

14.  Harleman, D.  R.  F.  and K.  D.  Stolzenbach,  "A Model  Study of Proposed
     Condenser Water Discharge Configurations  for the Pilgrim Nuclear
     Power Station at Plymouth, Massachusetts", M.I.T. Hydrodynamics
     Laboratory Technical Report No.  113, May  1969.

15.  Hayashi, T. and N.  Shuto,  "Diffusion of Warm Water  Jets  Discharged
     Horizontally at the Water Surface", IAHR  Proceedings, Ft. Collins,
     Colorado, September 1967.

16.  Hayashi, T. and N.  Shuto, "Diffusion of Warm Cooling Water Dis-
     charged from a Power Plant", Proceedings  of the llth Conference
     on Coastal Engineering, London,  1968.

17.  Hinze, J. 0., Turbulence. McGraw Hill Book Company, Inc.,  1959.

18.  Hoopes, J. A. and others, "Heat Dissipation and  Induced Circula-
     tions from Condenser Cooling Water Discharges into Lake Monona",
     Report No. 35, Engineering Experiment Station, University  of
     Wisconsin, February 1968.

19.  Industrial Waste Guide  on Thermal  Pollution.  U.  S. Department of
     the  Interior, Federal Water Pollution Control Administration,
     Corvallis, Oregon, September  1968.

20.  Ippen, A.  T.  and D. R.  F. Harleman,  "Steady-State  Characteristics
     of Subsurface Flow", U. S. Department  of  Commerce, NBS  Circular
     521,  "Gravity Waves",  1952.

21.  Jen,  Y.  and  others, "Surface  Discharge of Horizontal Warm Water
     Jet", Technical Report HEL-3-3, University of California,  December
     1964.

22.  Kato, H.  and 0. M. Phillips,  "On  the Penetration of  a Turbulent
     Layer into Stratified Fluid", Journal  of  Fluid  Mechanics,  Vol. 37,
     Part 4,  July 1969.

 23.  Lofquist, K., "Flow and Stress  Near an Interface Between Stratified
     Liquids", The Physics of Fluids,  Vol.  3,  No.  2, March-April 1969.

 24.  Pai, Shih-I, Fluid Dynamics of  Jets, D.  Van Nostrand Company, Inc.,
      1954.

 25.   Parker, F. L. and P.  A. Krenkel,  Engineering Aspects of Thermal
      Pollution, Vanderbilt University Press,  1969.
                                -177-

-------
 26.  Pearce, A. F., "Critical Reynolds Number  for Fully-Developed Turbu-
     lence  in Circular  Submerged Water Jets",  National Mechanical Engin-
     eering Research  Institute, Council for  Scientific and Industrial
     Research, CSIR Report MEG 475, Pretoria,  South Africa, August 1966.

 27.  Phillips, 0. M., The Dynamics of the Upper Ocean, Cambridge Univer-
     sity Press, Cambridge, England, 1966.

 28.  Prych, E. A., "Effects of Density Differences on Lateral Mixing
     in Open-Channel Flows", W. M. Keck Laboratory of Hydraulics and
     Water Resources, California Institute of  Technology, Report No.
     KH-R-21, May 1970.

 29.  Schlichting, H. , Boundary-Layer Theory, Translated by J. Kesten,
     Sixth Edition, McGraw Hill Book Company,  Inc., 1968.

 30.  Stefan, H.  and F. R. Schiebe, "Experimental Study of Warm Water
     Flow Part I", Project Report No.  101, St. Anthony Falls Hydraulic
     Laboratory,  University of Minnesota, December 1968.

 31.  Stefan, H.  and F. R. Schiebe, "Experimental Study of Warm Water
     Flow into Impoundment? Part III",  Project Report No. 103, St.
     Anthony Falls Hydraulic Laboratory,  University of Minnesota,
     December 1968.

 32.  Tamai,  N.  and others,  "Horizontal  Surface Discharge  of Warm Water
     Jets",  ASCE  Journal of the Power  Division, No.  6847, PO 2,
     October 1969.

 33.  Townsend,  A.  A.,  "The  Mechanism of  Entrainment in Free Turbulent
     Flow",  Journal of Fluid Mechanics,  Vol.  26,  Part 4,  1966.

 34.  Townsend,  A.  A.,  The Structure of  Turbulent Shear Flow^  Cambridge
     University Press, Cambridge,  England, 1956.

 35.  Wada, A.,  "Studies  of  Prediction  of  Recirculation of Cooling  Water
     in a Bay", Proceedings of the llth  Conference on Coastal Engineering,
     London, 1968.

36.  Webster,  A.  G.,  "Experimental Study  of Turbulence in a Density-
     Stratified Shear Flow", Journal of  Fluid Mechanics,  Vol.  19,  Part
     2, June 1964.

37.  Wiegel, R. L.  and others, "Discharge of  Warm Water Jet over Sloping
     Bottom",  Technical  Report HEL-3-4, University of California,
     November 1964.

38.  Yevdjevich,  Vujica  M.,  "Diffusion of Slot Jets with  Finite Length-
     Width Ratios",  Hydraulic  Papers, Colorado State University, Ft.
     Collins,  Colorado,  No.  2, 1966.
                              -178-

-------
                     LIST OF FIGURES AND TABLES



Figure                                                              Page

 1-1       Schematic of Heated Discharge                             16

 2-1       Structure of a Submerged Turbulent Jet                    19

 2-2       Velocity and Temperature Distributions                    26
           in Submerged Jets

 2-3       Reduction of Entrainment as a Function of                 30
           Richardson Number

 2-4       Two Layer Flow in the Discharge Channel                   34

 2-5       Schematics of Previous Theoretical Investigations         36

 3-1       Characteristics of the Discharge Channel                  43

 3-2       Geometrical Characteristics of the Jet                    55

 3-3       Velocity and Temperature Characteristics of  the Jet       57

 3-4       Non-Buoyant Jet Solutions                                 70

 3-5       Characteristics of a Deflected Jet                        85

 4-1       Experimental Set-Up                                       102

 4-2       Mounted Temperature Probe                                 104

 4-3       Cross  Flow Velocity in  Experimental  Basin                 108

 4-4       Temperature Distribution Measured  for Run  3

 4-5       Surface Temperature Distribution Measured  for
           Run  22

 5-1       Theoretical  Calculations of Jet  Structure  for            115
           Run  6

 5-2       Calculated  Isotherms  of AT /AT  for Run 6                 11 A
                                      c   o                           •LJ-°

  5-3       Theoretical  Calculations and Experimental  Data           120
            for  Run 1

  5-4        Theoretical  Calculations and Experimental  Data
            for  Run 2

                               -179-

-------
Figure                                                            Page

 5-5       Theoretical Calculations and Experimental              122
           Data for Run 3

 5-6       Theoretical Calculations and Experimental              123
           Data for Run 4

 5-7       Theoretical Calculations and Experimental              124
           Data for Run 5

 5-8       Theoretical Calculations and Experimental              125
           Data for Run 6

 5-9       Theoretical Calculations and Experimental              126
           Data for Run 7

 5-10      Theoretical Calculations and Experimental              127
           Data for Run 8

 5-11      Theoretical Calculations and Experimental              128
           Data for Run 9

 5-12      Theoretical Calculations and Experimental              129
           Data for Run 10

 5-13      Theoretical Calculations and Experimental              131
           Data for Runs 11,  12 and 13

 5-14      Theoretical Calculations and Experimental              132
           Data for Runs 14,  15 and 16

 5-15      Theoretical Calculations and Experimental              133
           Data for Runs 17,  18 and 19

 5-16      Theoretical Calculations and Experimental              136
           Data for Run 20

 5-17      Theoretical Calculations and Experimental              137
           Data for Run 21

 5-18      Theoretical Calculations and Experimental              138
           Data for Run 22

 5-19      Theoretical Calculations and Experimental              139
           Data for Run 23

 5-20      Theoretical Calculations and Experimental              140
           Data for Run 24
                             -180-

-------
Figure                                                            Page

 5-21      Theoretical Calculations and Experimental
           Data for Run 25

 5-22      Theoretical Calculations and Experimental
           Data for Circular Orifices - Jen (21)

 5-23      Theoretical Calculations and Experimental
           Data for Rectangular Orifices - Wiegel (37)

 5-24      Theoretical Calculations and Experimental
           Data for Rectangular Orifices - Hayashi (15)

 5-25      Dilution in the Stable Region of Buoyant
           Discharges

 5-26      Maximum Jet Depth Vs. IF as Calculated by
           the Theory              °

 6-1       Discharge Channel Schematization

 6-2       Location of Power Plant

 6-3       Temperature Rise Contours Measured  in                  165
           the Pilgrim Station Model

 6-4       Theoretical Calculations and Model  Data                 167
           for the Pilgrim Station Power  Plant

 II-l      Generalized Cross  Section of a Deflected  Jet           21°
 Table                                                             Page

  2-1       Summary of Previous Theoretical Investigations          ^1

  3-1       Integrated Equations for Buoyant Jets                   81

  3-2       Integrated Equations for Deflected, Buoyant Jets        91

  4-1       Summary of Previous Experiments                         98

  4-2       Summary of Experimental Runs                           107
                              -181-

-------
                         LIST  OF  SYMBOLS
A       —   the discharge, channel  aspect  ratio  =  h  /b
                                                 o o

a       -   the coefficient  of  thermal  expansion  =  - ~p
                                                    31
b       -   horizontal distance from  the  jet centerline to
            the jet boundary

b       -   one half the width  of  a rectangular discharge channel

b. ,„    -   horizontal distance from  the  jet centerline to  the point
            where the temperature  or  velocity,  as defined,  is one
            half the centerline value

D       -   the ratio of flow in the  jet  to the initial discharge
            flow = dilution

D   ,.  -   the dilution in  the stable  region of a  heated discharge


 db
 -j—•     -   the spread of a  non-buoyant jet
    NB
F      -  the densimetric Froude number =
                                              P
IF      -  the densimetric Froude number of the discharge channel
                 u
               Ep
               ™ *   gn
               D     O
                a
                                                               u
It'      -  the local densimetric Froude number in the jet =     -c-
                                                           [3/2T
                                                    1 -  (j;)   J
G       -  the turbulent integral time scale in a homogeneous flow

G       -  the turbulent integral time scale in a stratified flow
 s

H       -  the ratio of heat flow in the jet to the initial discharge
           heat flov;

h       -  vertical  distance from the jet centerline to the jet
           bound :irv
                             -182-

-------
h       -  depth, of the discharge channel

h *     -  depth of the heated flow at  the point of discharge if a
           cold  water wedge is present

h^ 12    -  vertical distance from the jet  centerline to the point
           where the temperature or velocity, as defined, is one
           half  the centerline value

h       -  the maximum value of h obtained in a heated discharge
          ,1
                    = .4500
       -I
       -t
       -c
       -tt
                     =  .3160
                      .6000
I,      -       t(Od?d?  =  .2143
                            .2222
I6      -  I  f  UK"" - .1333
          C
          £
I       -  \ f(C)t(OdC  - .3680
 K      -  the surface heat  loss coefficient

 k      -  the thermal conductivity of water

 L      -  the scale ratio for horizontal lengths between
           model and prototype
                                        r    3 i  1/2
 N      -  the Brunt-Vaisala frequency - \-~  -^H

 p      -  the pressure

 p,      -  the dynamic pressure

 P_      -  the hydrostatic pressure
  d.

                            -183-

-------
Q       -   the discharge  channel  flow

q       -   the entrainment  flow into the jet per unit length of jet

R       -   the Eulerian time coefficient of a turbulent flow
                                      £_ 5p
R.      -   the Richardson number  =   p §z
R,.      -  the hydraulic radius of the discharge channel flow

R       -  the radius of a circular discharge pipe

 °                                    AuoRH
H      -  the jet Reynolds number =  	


r       -  the vertical distance from the jet centerline to the
           boundary of the core region

S       -  the bottom slope
 x

s       -  the horizontal distance from the jet centerline to the
           boundary of the core region

T       -  the temperature

T       -  the ambient temperature
 3

T       -  the equilibrium temperature

T       -  the jet surface centerline temperature

T       -  the temperature of the heated flow in the discharge
           channel

T       -  the temperature at the water surface
 S

AT      -  temperature rise above ambient in the jet - T - T
                                                            £1

AT      -  the surface temperature rise above ambient at the jet
           centerline T  - T
                       c    a

AT      -  the temperature difference between the discharge and
           the ambient water = T  - T
                                o    a

AT      -  the scale length for temperature rises between model
           and prototype

                                                          3/2
t        -  the similarity function for temperature = 1 - £

                             -184-

-------
u,v,w   -  the velocity components in the fixed coordinate system

UjV^w   -  the velocity components in the coordinate system relative
           to the centerline of a deflected jet

u       -  the velocity in the discharge channel

u *     -  velocity of the heated flow at the point of discharge
           if a cold water wedge is present

u       -  the surface centerline jet velocity

u       -  the scale ratio for velocities between model and prototype

V       -  the ambient cross flow velocity

v       -  the lateral velocity of the entrained flow at the jet
 e         boundary

v,      -  the lateral velocity in the jet at    y » s and -h  s and -r < 2 < n
 S

w,      -  the vertical velocity in the jet at z » -r and s < y < b

w       -  the vertical velocity of the entrained flow at the jet
           boundary

w       -  the vertical velocity in the jet at z * -r and 0 < y < s

x,y,z   -  the fixed  coordinate directions

x,y,z   -  the coordinate directions  relative  to the centerline
           of a  deflected jet

x       -  the distance  from the jet  origin to the point where s  »  0
  S
x       -  the distance  from the jet  origin to the point where r  «  0

z       -  the scale  ratio for vertical  lengths  between model  and
           prototype

a      -  the entrainment coefficient

a      -   the entrainment coefficient in a homogeneous  flow

 a       -   the  entrainment coefficient in a stratified flow
  S

 a       -   the  lateral entrainment coefficient in non-buoyant and
           buoyant jets
                             -185-

-------
a       -  the vertical entrainment coefficient in a non-buoyant jet


a       -  the vertical entrainment coefficient in a buoyant jet
 sz

6       -  a scaling quantity one order of magnitude less than unity

e       -  the spread, -:— , of the turbulent region in an undeflected

           non-buoyant jet

n       -  the water surface elevation

9       -  the angle between the jet centerline (x axis) and the y axis

6       -  the angle between the discharge channel centerline and the
           y axis

p       -  the density of water

p       -  the density of the ambient water
 3i

p       -  the density of the heated discharge

Ap      -  the difference between the density and the ambient water
           density = p - p
                          £L

Ap      -  the difference between the density of the heated flow in
           the discharge channel and the ambient density = p  - p
                                                            o    a
Ap      -  the scale length for density differences between model
           and prototype

$       -  viscous generation of heat

v       -  the kinematic viscosity of water

Superscript *   -   indicates a characteristic magnitude of the
                    variable

Superscript     -   indicates a turbulent fluctuating quantity
                             -186-

-------
            Appendix I.   The Computer Program for Theoretical



                         Calculations of Heated Discharge Behavior






       The computer program is written in Fortran IV,  G level,  Mod 3




and the calculations for this study are done on an IBM 360-65 computer.




The program consists of a main program and five subroutines:




MAIN;  This program reads the input data for the theoretical  calcula-




tions, sets up the initial conditions for each calculation, and calls




the subroutine SRKGS which performs the actual calculations.






              Input Data                                   Format




            Number of Calculations                           13
One Set  [IF , S , A, K/u , Q   *   ERR, b.                 8F10.5
         I   o    x        O   O   r        i



for Eachj I]L, 1^ T.y 1^, 1^, Ig, I?,  e                     8F10.5




Calcula-[vi, V2, V3> V4, V5§ Vfi, V , Vg                    8F10.5


  tion
                                                           u

       where:  IF = initial densimetric Froude number «  	—
                 o
I
               S  = bottom slope                         Jp   8 o



               A  * channel aspect ratio » h  /b
                                            o o


             K/u  = surface heat  loss coefficient parameter




               6  = initial angle of the discharge with  the



                    boundary  of the ambient region




               XF * The  value of  x/Vh b  at which the  program should




                    terminate for this  calculation.   In  this study,




                    x., varies from 50  to  200.
                      i
                              -187-

-------
             ERR = Maximum allowable numerical  truncation error  in



                   each  dependent variable, at each  time step.  In this



                   study a value of ERR  =  .01 is used.



              b. • Initial value of h/yh b   and b/>^i b   to be  used
               i                         o  o         o o


                   in  this calculation.  The mathematical formulation



                   of  the theory sets h=b=0atx=0 but in  the



                   numerical scheme a finite value  must be given.  A


                                   —8
                   value of b. » 10   is used in this study.  The



                   calculations are found  to be independent of the



                   choice of b. as long  as  it is small.



         I  - I  = Integrations of similarity functions as given



                   in  Table 3-1.



               e = Spreading rate of the turbulent  region of a non-



                   buoyant jet.



         Vn - V0 = Input constants describing the cross flow - (see
          J.    o


                   description of subroutine CROSS).




SRKGS;  This subroutine  is a modified version of the IBM Scientific



Subroutine Package routine DRKGS.  It is a  fourth order Runge-Kutta



solution to a system of  first order differential equations.



       The form of the equation is;
               dx~  '    ci
where a   is a coefficient matrix, y.  is a vector of the independent
       -LJ                           J


variables and c. is a vector of the constants  in the i equations.





                             -188-

-------
        The program MAIN provides the initial values y.  and the sub-



 routine SRKGS advances the solution by calling subroutine FCT (described

                                                     dy

 below) which calculates the values of a..> c., and  -r-*-  at each step.



 The solution is terminated when x «= x,.  The results of each step are



 output by subroutine OUTP.



 FCT;  This routine takes the current values of x and y. and computes
                                                       J


 the values a., and c  according to the theory of this  study, the system



 of equations in (I--1) are then solved by an IBM Scientific Subroutine



 DGELG which returns the values of 9y. /3x to FCT which  returns them to



 SRKGS.



 OUTP:  The independent variables y.  and u/u  , AT /AT  , h//h b , b//h b  ,
 	                              J        OCO      OO      OO
  r//h b  » s//h bQ, db/dx, x//h b  , y//h b  , and e  •  The values  of  these



  variables are output after each  step of the  calculation.   The value  of



  the dilution D and the heat flow H  (see Chapter  5)  are  also  calculated



  and printed.  Each page of output displays the input  variables  for that



  particular  calculation on the page  heading.



CROSS_  This  subroutine is called  whenever  the velocity of  the cross flow



  is required.  The subroutine uses the values of  V-  -  Vg to compute the



  cross flow  as a  function of x//h b  .  For each particular application
                                  o o


  of the  theory, the  subroutine may be  different.   The  only requirements



  are as  follows :



         a) The  subroutine  returns the  value of V/u  at the given x/ >h b .
                                                      o
         b)  V8 is reserved for return of  ^-j-       ,^-j to the main

                                             °     "    o o

            program.
                               -189-

-------
       In this study the following algorithm is used;
                                                                  (1-2)
                                       00
                                                                 (1-3)
                             o o
Note;  The program may generate numerical underflows in certain calcu-
lations and corresponding error messages may be printed out.  The re-
sult of an underflow calculation is set to zero automatically.
                           -190-

-------
C     MAIN PROGRAM
      REAL II, 12, IT, 15, 16
      DOUBLE  PRECISION PRMT ,Y , AUX , DERY, TH, XP, YP, 8, U , G, H, BX, S ,R, X, XL IM
      INTEGER  RR
      EXTERNAL  FCT,OUTP
      EQUI VALE MCE  ( Y( 1 ) , TH) , ( Y(2 ) ,XP) , ( Y( 3 ) ,YP ), (Y (41 , B) , (Y(5),U)
      FSinVALEf^CF  (Y(6),G),(Y(7),H),
-------
  DERY(N)=..l
2 CONTINUE

  CALL  SRKSS(PRMT,Y,DERYfNDIM,IHL F, FCT, CUJTP, AUX)
1 CONTINUE
  CALL  EXIT
  FND

-------
 c
 c
 c
 c
u>
c.
c
  SUBROUTINE  SRKGS(PRMT,Y,DERY,NDIM,THLF,FCT,OUTP,AUXl
  DOUBLE PRECISION PRMT,YfDERY,AUX,A,R,C,X,XEND,H, AJ,BJ,CJ, Rl ,R2,
 1DELT,DABS
  DIMENSION Y(!4),OCRY(14),AUX(8tl^)»A(4),B(4)fC(4),PRMT(5)
  DP 1 1=1 ,NDIM
1 AUX( 8»I )=.'^6666666666666666700* DERY( I)
  X=PRMT(1)
  H=PRMT( 3)
  PPMT (5) = "«.D'i
  CALL FCT(X,Y,DERY)

  ERROR TEST
  IFf.781186547«5Dn
  A{4) = . 1666f>666666666667D^
  B(l ) = 2.D'1
  B(2)=1.D"
  3(3 )=1,D.°
  B(4)=2,D~
  C( 1 ) = .5DA
  C(2 )=. 292893 2 1881 34574800
  C (3) = 1,7' 71 "^6781186547500
 PREPARATIONS OF FIRST RUNGE-KUTTA STFP
 DO  3  I=1,NDIM
 AUXd ,!) = YU)
 AUX(2tI )=
                                                                            DRKS1130
                                                                            DRKG1140
                                                                            DRKG115P
                                                                            DRKG1160
                                                                            DRKG1170
                                                                            DRKG1180
                                                                            DRKG1190
DRKG1210
DRKG1220
DRKG1230
DRKG1240
DR
-------
c
c
c
c
c
c
c
      H=H+H
      IHLF=-1
      ISTEP=0
A IF( ( X+H-XEN[5)*H)7,6,5
  START OF  A  1UNGE-KUTTA STEP
5 H=XENO-X
6 IEND=1

  RECORDING OF  INITIAL VALUES OF  THIS STEP
7 CALL OUTP(X,Y,DERY,IREC,NDIM,PRMT)
  IF( PRHT(5) )A"!t 8, AO
8 ITEST=0
9 ISTEP=ISTEP+!
   START OF  INNERMOST RUNGE-KUTTA  LOOP
   J = l
   DO 91 I=1TM3IM
   AUX( 3, 11=^,0^
91 AUX{ 6, I )=0,r)^
10 AJ=A(J)
   BJ=B(J)
   CJ=C( J)
   00 11 I=1»NOIM
   R1=H =DERY(I )
   R2=AJ*'(R1-BJ'-AUX( 6, I ) )
   Y( I )=Y(I
                                                                        DRKG14-60
   11  AUX(6, I) = AUX(6, I)+R2-CJ*Rl
       IF
-------
 c
 c
 c
 c
 c
 c
 VO
 Ln
C
c
c
c
14 CALL  FCT(X,Y»DERY)
   GOTO  1<>
   END 3F INNERMOST  RUNGE-KUTTA LOOP
    TEST OF ACCURACY
 15  IFUTESTI16tl6t2i

    IN  CASE ITEST="> THERE  IS  NO  POSSIBILITY FOR TESTING OF  ACCURACY
 16  DO  17 1=1 ,NOIM
 17  AUX(4,I»=Y(I)
    ITEST=1
    ISTEP=ISTEP+ISTEP-2
 18  IHLF=IHLF-H
    X=X-H
    H=.500*H
    DO  19 I=1,NDIM
    Y(I)*AUX(lfII
    DERY( I)=AIIX(?tI)
 19  AUX(6,I)=AUX(3tI)
    GOTO 9

    IN  CASE ITEST=1 TESTING OF ACCURACY  IS  POSSIBLE
 2r  IMOD=ISTEP/2
    IF(ISTEP-IMOD-IMOD)21,23,21
 21  CALL  FCT(X,Y,DERY)
    DO  22 I=1,NOIM
    AUX(5,I)=Y(I)
 22  AUX(7,I)=DERY(I)
    GOTO  9

    COMPUTATION OF TEST VALUE DELT

    DO  24 I=!,NDIM
?4  DELT=OELT*AUX(8, I)*DABS(AUX( 4,1 )-YU ) )
    IF(OELT-PRMT(4) )?8,28,25
DRKG1780
DRKG1790
DRKG1800
DRKG1810
DRKG1820
DRKG1839
DRK31840
DRKG1850
DRKG1853
DRKG1870
DRKG1880
ORKG1890
DRKG1900
DRKG1913
DRKS1920
DRKG1930
DRKG1940
                                                                       DRKG1960
                                                                       DRKG1970
                                                                       DR
                                                                       DRKG2123
                                                                       DRKG2130

-------
C                                                                         DRKG214D
C     ERROR  IS  TOO  GREAT                                                 DRKG215O
   ?5 GO  TO  26
   ?6 00  27  1=1, NO I M                                                     DRKG2170
   27 AUX(4,I)=AUX(5 ,1)                                                   DR
-------
   35 IHLF = IHLF-1
      ISTEP=ISTFP/?                                                       DRKG2490
      H=H+H                                                               DRK3250O
      GOTO 4                                                              DRKG7510
C                                                                         DRKG2520
C                                                                         DRKG2530
C     RETURNS TO CALLING  PROGRAM                                         DRKG2540
   36 THLF=ll                                                             DRKG2550
      CALL FCT(X,Y,DERY)                                                  DPXG2560
      GOTO 39                                                             ORKG257?
   37 IHLF=12                                                             DRKG2580
      GOTO 39                                                             DRKG2590
   38 IHLF = 13                                                             ORKG26CO
   39 CALL OUTP
-------
    SUBROUTINE  FCTr H=l.^D-7
7^1 CONTINUE
    SLPPE=1
    FLAG=1
    CAL L  CRnSS(XP,V)
    IF(V)  7=1,75,76
 75 EPSB=EPS
    FLAG=2
 76 IF  (R-.OTO'U) 8^,80,81
 RP R=«~'.
    WINF=Il-"U*EPS/2
    GO  TO 811
 81 WINF=(I1-I2) *.EPS<:U
811 IF  (S-.000^1) 82,82,83
 82 S=r%
    VINF=-IlfcUJ EPS/2

-------
     GO  TO 831
  83 VINF=-U1-I2)*EPS*U
 831 VCT=OCOS(TH)*V
     WINFS = WINF
     FIS=FI
     B2=S+6*T2
     83=8*16
     B4=B*I5
     BT=S+B*IT
     H? = R-«-H*I2
     HGG=R*R  3 5+Tl*HfeR+T2*H*H
     HT=R-l-H*IT
VO
VO
     UM1= U* 12+2.  VCT*I1
 888 DO 1 N=l,14
     DP 1 NN=lf14
1 CONTINIJF
  GO TO
                       FLAG
     GO TO 6
 301  ARG= 53':G*H/(U*U*FR*FR)
 312  ARG=0.
 3^3  IF  (ARG-1H, )
     GO TO 6
 3C  WINF=WINF*EXP(-ARG)
  5  A(l,l ) = U*U
     A (2, 2 I *1,1
     A(3,3)=l,n
                                                VCT*VCT*(S+B)*{R+H)

-------
    A<4,4) = 1,0
    OERY<4)=  8X
    A(5,1)=-H*B*VST
    A (5, A )= Il*U*H*Il + H
    ACS, 5 )=I1*H*B*I1
    A( 5,7)=I1*U*B*I1*B*VCT
    A(5,9)=UtH*IH-H*VCT
    At 5,10) = U*8*IH-B*VCT
    A(5,11)=-H*I1
    A(5,12)=8*I1
    A(6,1)=-2.*H-B*(U*I1*II+VCT)*VST
    A(6,4 ) = U" H^UM2 + TlvT2* H^H*G1' F
    A(6, 5)=?
    A(6, 7) =
    A(6,9I=
    A(6,10)
    A(6,ll >=-H*UMl
    A(6,12)«B*UMl
    A(7,l )=-G*HG BG'VST
    A ( 7f 4)=G* ( U*I T*HT+T1*VCT*HG
    A(7,5) = GtHT^BT
    A (7, 7) = GMU*
    A { 7 ,9 ) =G** I U* HT+VCT*HG )
    A(7,1C)=G*( U«-BT*VCT*BG)
    GO TO (3*4,3^5),  FLAG
304 A(8,8)=l,0
    OERY(4)=0.
    GO TO 3f>6
305 GO TO (3^51,3052), SLOPE
    A( 8,1)=-(BX-EPSBJ*VST*2. *( U*H1* B4+VCT* (R+H)*B)
    A(8,4)= (BX-EPSB)*(U*U*I6*H2+2.*VCT*U*H1*I5+VCT*VCT*' 
-------
             (BX-EPSB)MU*UM2*B3+2.*VCT'B4*I1*U+VCT*VCT*B)
                          . *VCT*B4*H1+VCT*VCT* -  U^R-^R
     Adi, 12)= -I1*B
     A(ll,14)= -R
     A(12tl)*-2. «B*R*(U*Il-»-VCT)*VST
     AC 12,4) = R MJ*U*!H-R*VCT*VCT+T1*G*FI*< . 5*R*R4-R*H*T1I
     A(12t6)=  T1«B*<, «5*R^R+R*H-*T1 )*FI
     A(12,7)=T1"T1*B*R*G*FI
     A ( 1 2 ,9 ) =R* ( U -U*2, *U«-VCT-K R*. 5+H* Tl ) *G*F I ) *R*VCT* VCT
     A(12,lr>)=A(12,7)/Tl
     A(12,14)=  -R (U*VCT)
  6' IF  (S»  61,61,62
  61 A(9,9)*l,'>

-------
     A(14,14)=l
     GO TO 70
  62  A<9, 1 )=-
     A(9,7) = I!*S*U+S*VCT
     A(9,10)=S*U+S*VCT
     A(9,11)=I1*H
     A( 9,13)=S
     A(11,1}=-2,*S*H*
-------
         ( 1?)=
     OERY(13)=
     IF  (V)  7?, 73, 71
  71 VR=VCT-VEL(8)/VST
     00 72 1=2,14
     DEPY( I^DE^Yd
  72 CONTINUE
  73 CONTINUF
     on 2^ 1=1 ,14
     no 21 j=i,i*
     OA=OABS(A( I, J) )
     IF 1DA-AMAX)  21,21,22
  22 AMAX=DA
  21 CONTINUE
     IF (AMAX)  2% 2*^,24
^ 24 DP 23 J=l,14
g    A( I,J)=A( I, JJ/AMAX
1  23 CONTINUE
     OERY(I)=DERY(I)/AHAX
  2r CONTINUE
     EPS = .10-16
     CALL OGELGCOHRY, A, 14, 1 , ERS ,1 ER)
     IF (IER)  9,8,7
   9 WRITE (P,5'^>  IER, FLAG, SLOPE
     WRITE (P,5H1)  < (A(I,J) ,J=l,14),OERYm,I=l,14)
     CALL EXIT
   7 WRITE (P^Oi)  IER.FLAG, SLOPE
     FORMAT (]X,18HERRnR  IN  MATRIX = ,12,14,14)
     FORMAT (15(08.2))
     GO TO (32%321 ),  FLAG
     FLAG=2
     FPSB=DERV(4)
     FI=FIS
     GO TO 888

-------
 3?1  GO TO (322, 323), SLOPE
 322 SJ=DERY(7) + DERY(ir\)
     SXC=SX"DSIN( TH)
     IF (SXC-SJ) 3221*3221,323
3221 SLOPE=2
     WINF=0.
     GO TO 88B
 323 FI = FIS
     VEL<8)=0,
     RETURN
     END
O
•o

-------
      SUBROUTINE OUTP(X,Y,OERY , I HLF,NDIM, PRMT )
      INTEGER P
      DOUBLF PRECISION DERY, Y, PRMT, X, TH,XP , YP ,U ,B ,G ,H,B X, S, R, XUM
      REAL IItI?»TTf15,16,M
      DIMENSION DERY(14),PRMT(5) ,Y(14) ,VEL( 8)
      COMMON NPAGE,NLINE, FR,FI,AS,E,I1, 12,1T,T1,T2,EPS,I 5,16,SX,THI,VEL
      COMMON FRR
      TH=Y(1)
      XP=Y(2)
      YP = Y{ 3»
      
      RX=Y(8)
r!,     S=Y(9)
g     R=Y(10)
1     P = 6
      CALL CROSS«P,V)
      IF (NLINE-«5-U ^,11,11
  11  NPAGE  = NPAGE +1
      NLINE  = "
      WRITE  2) NPAGE
 23?  FORMAT  (1H1,24HBUOYANT JET  C ALCULAT IONS, ir»X, 5HPAGE ,12)
      WRITE  
-------
     FC=ABS( FC)
     FC=FC*':.5
     FC=FR/FC
     Q=U*( S+BM1)1 <*+H*Il)-i-V*DCOSf TH)*(S-t-B)*  FORMAT  ( 3 < IX ,08 .3 ), IX, 09. 3, I'M 1 Xt 08. 3) » IX, I 3)
     IF (U-V-DCOS(TH) ) 500,500,501
 500  PRHT(5)=1
 5D1  RETURN
     END
o
T

-------
SUBROUTINE CROSS(XP,V)
REAL  lit 12,IT,15,16
DOUBLE PRECISION XP
DIMENSION VELC8)
COMMON NPAGE,NLINE,FR,FI,AS, E, II, 12, IT,T1,T2 ,EPS, 15,1 6, SX ,THI ,VEL
A=XP*VEL(4I
A= A-VELC1)
VEL(8) = -A*VEL (2 )'2.*' VEL ( A)
A=A*A
A=VEL (2)*A
V=VEL(3)*EXP(-A)
RETURN
END

-------
            Appendix II. '"'of! crteri Jet  TO  a  Cross  Flow:




               Integration. of_ the x and y'Jj orient urn



                  Equations over the Entire  Jet
        This appendix considers the x and y momentum equations (3.108



and 3.109) for a deflected jet as given in Section  3.5.1.   These equa-



tions will be integrated over the entire jet  to  give equations 3.122



and 3.123 which are used in Section 3.5.1 to  develop the  governing equa-



tions for deflected jets.  The basic assumptions  of  the derivation are:



        1) The velocity distributions within  the  jet are  symmetrical



           about the y = 0 plane i.e. the cross  flow does  not  greatly



           distort the jet from the structure defined for  the  undeflec-



           ted jet in Chapter 3.



        2) The flow outside the jet is irrotational.



Both of the above assumptions require that the cross flow  velocity is



small with respect to the initial jet velocity as discussed in Section



3.5.1.



        The x and y equations (3.108 and 3.109)  are:
 _2    „_,     „.,__                              	     	.
                     /i ~ •. (? t)   3.2 \   o & 1  * ~    gll  v     o u W     «•* ** i — t  do
                     2uv —- = —a \   —z— az	=— -  •	vn'-u'  -i"i
3x     3y      3z        3x   p  ),  3x          3y       32
                               3  Z
                                                                  (II-l)
3uv   3v    3vw _ 38   2 _ ~2. _ ag_ (    8AT  ,-__!_

Sx    3y    3z    9x             p  / -  39        p
                                  a  z            a
                  v'2)                                            (II-2)
                              -208-

-------
        Figure II-l shows a local section of the jet in which the shape

of the jet cross section, A, is arbitrary except for symmetry about the

y = 0 plane and in which the distribution of entrainment flow around the

jet section circumference, C, is also arbitrary.

        Equations II-l and II-2 are integrated over the entire jet area,

A, taking advantage of basic scale relationships introduced in Chapter

3, the symmetry of the jet, and the boundary conditions at the jet boun-

dary as given in Chapter 3.
        -pr $ u2 dydz = -^$1        dz'dydz +    u(wdy-vdz)
           A             a A ' 2                  C
          de  Lf  ,2  ,,,,      jL pd
              A               C  pa      C
                                            v(wdy-vdz)
         It  is  assumed  that  everywhere  on  the  jet boundary  (see Figure

 II-l):
         u
                V cos 9                                            (II-5)
 Noticing that the total entrainment per unit length of jet, q ,  is given
 by:
         q   =   9  (wdy - vdz)                                   (II-6)
 then the integrated ic equation (II-3) becomes:
                              -209-

-------
VcosG /   \Vsin6
     Cross Flow
                    Top View of Deflected Jet
                                    1
     VsinQ
     Cross  Flow
                           Cross  Section of Jet
                    Total  Entrainment/Unit Length = q
              Figure  II-l.  Generalized Cross Section of a Deflected Jet
                               -210-

-------
-£  <$  u2 dydz  -  Sfi.  ^  j    ^|i-  dz'dydz +  q£ V cos 0       (II-7)
                    pa   A  ' z
The above equation, when the integrations are written in terms of the

jet geometry defined in Chapter 3, is the desired result i.e. Equation

3.113.

        The y momentum Equation (11-4) is not so easily dealt with since

the velocities on the boundary are not known without determining the

flow  of the ambient fluid around  the jet.  The  assumption  of irrotation-

al  flow makes it possible to use  several properties of potential flows

to  evaluate the terms in II-4.

        A. Bernoulli equation is written from  a  point  far from  the jet

to  the jet boundary:
         pv2
          a
+
      1     r.2 .  -2 ,  -2,
Pd +  2  pa [U  + V  + W ]    r
                           on C
                                                                  (11-8)
 where p,  is the dynamic pressure far from the jet.  This may be

 written:
         P   =  H -    p   [v2 + w2]   on C                       (II-9)
                 ir                  O   O
 where   H  -   £^-  +  p      -  p V cos 6                         (11-10)
 is constant and will not contribute to the pressure integral  over  the
                               -211-

-------
circumference of the jet.

        Then Equation II-4 becomes:
          <**  A              C    2
Defining a complex coordinate system in the y-z plane


        C  =  y + i z                                            (II-12)


and a complex potential, Q, for which the velocities v and & are given

by
             »  v - i w                                          (11-13)
Equation 11-11 becomes
                 u2 dyd2  -  Real   ±
Batchelor (2) shows that fot an arbitrary shaped cylinder in the pres-

ence of a cross flow, V sin 6, and having a total mass flux per unit

length into the cylinder q  the above integral may be easily evaluated

and the integrated y equation becomes simply:
               H2 dydZ = - q  V sin 0                             (11-15)
            A

which is the desired result.

                              -212-

-------