AC-AIOS 095 ARIZONAUNly TUCSON ENGINEERING STATION

EXPERIM4ENT F/6 20/13

CONVECTIVEHEATTRANSFER CU)
FOR SHEIPPROPULSION.
APR 81 N A MADlS. 0 M NCELIGOT MNOOOIVTS-C-0694
UNCLASSIFIED 1248-8 IA.
 S1.0 ,
I- 8 j12.5
IIII
1112 111112-2

.I25 I IIIIf.

MICROCOPY RESOLUTION TESTCHART

NATIONAL!SLIRKAUOf TANDARP 1q
 Sixth Annual Summary ReDort r
CON Contract 'o. 000l4-75-C-,94 LEVEL
C Contract Authority NR-097-395

CONVECTIVE HEAT TRANSFER FOR SHIP PROPULSION

Prepared for

Office of Naval Research

Code 473
Arlington, Virginia

Prepared by

M. A. Habib and D. M. McEligot D TIC

ELECT
1 April 1981 DEC 3 1981

SDiSTIBUlION STATEM:NT A I
Approved for public reloase D
Distribution Unlimited

ENGINEERING EXPERIMENT STATION

COLLEGE OF ENGINEERING
THE UNIVERSITY OF ARIZONA
TUCSON, ARIZONA 85721
 11nrvtq
Ia Hfi id
SECURITY CLASSIFICATION OF THIS PAGE (WhmenDoe Entered)
INSTRUCTIONS
READCOMPLETING
REPORT DOCUMENTATION PAGE BEFORE FORM

IIA sea . OVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

148-8__ _ _ _ _ _ _ _ _ _ _ _ _

" 4. TITLE (and'Subtite) S. TYPE OF RILJORT 6 PelIOO COVERED

Convective Heat Transfer for Ship Propulsion. Annual gummary Report /

(Sixth Annual Sumary Report) //115 Jan 180-30 Mard=981

I. AUTNO (a) £. CONTRACTOR GMANTNUMBER(s)

, M. A..Habib amd D. M.iMcEligot , ,O 4-75-C -694

S. PERFORMING ORGANIZATION NAME ANO ADDRESS IW. PROGRAM ELEMENT. PROJECT. TASK
NRW-_? RK UNIT NUMBERS
Engineering Experiment Station
University of Arizona -
Tucson, Arizona 85721 I.:--lJ
II. CONTROLLING OFFICE NAME AND ADDRESS -M R"6T OATE
1 /
Office of Naval Research / Aprit-0198
Code 473 3 BER OF PAGES

Arlington, Virginia 22217 100

14. MONITORING AGENCY NAME & AOORESS(If different from Controlling Office) IS. SECURITY CLASS. (of this report)
Office of Naval Research Unclassified
Bandolier Hall West - Rm. 204
University of New Mexico 1S. OECLASSIFICATION/OOWNGRAOING
SCHEDULE
Albuquerque, N. M. 87131
16. DISTRIBUTION STATEMENT (of this Report)

Unlimited
DISTRIBUTION STATEMENT A
Approved for public relesei
Distribution UnlHnited

I?. OISTRIBUTION ST. 4ENT (of abstract entered In Block 20. If different from Report)

Unlimited

,,. SUPP ENTARY rEs

19 KEY WORDS (Co.ntinue on reverse side ineceaene, and Identify by block number)
Heat transfer Turbulence model
Turbulent flow Forced convection
Swirl Elliptic partial differential equations
Sudden expansion Inlets

20. \AGSTRACT (Continue on reverse side It necessary and Identify by block number)
"A common geometry recurring in compact Naval propulsion plants is a change in
duct size in the primary fluid loop. As a consequence of the upstream plumbing,
the fluid is often swirling about the axis in the piping. Heat losses from the
primary fluid and thermal stresses in the component depend on the convective heat
transfer from the fluid to the component as it undergoes this geometrical transi-
tion. The idealized problem is a study of heat transfer in a sudden expansion
with swirl flow, the subject of the present report. The same situation also oc-
curs at the entrance to some heat exchanger tubes. - .-.
DO AN3 1473 Unclassified
SECURITY CLASSIFICATION OF THIS PAGE (When Dete grntered)

.A A
SE RtY C1..A5 IICACPON Or TMIS PAGg£(ha D.e8 £,.C,, Unclassified

lCalculations of the flow and heat transfer of a swirling turbulent flow after a
sudden pipe expansion are reported. The calculations were obtained by the numeri-
cal solution of the time-averaged forms of the continuity, momentum and thermal en-
ergy equations together with equations for the turbulence kinetic energy and its
rate of dissipation. For a sudden expansion without swirl, the predictions pro-
duced satisfactory agreement with available data for the Nusselt number. In the
swirling case, there were no heat transfer measurements available for comparison.
However, existing measurements of a swirl flow at an expansion without heat trans-
fer were used to test the validity of the flow field calculations.
The predicted effects of varying the swirl number from zero to 1.0 on the heat
transfer behavior are presented for a range of Prandtl numbers from 0.7 to 10. The
expected effects of the swirl on the velocity and temperature fields are also re-
ported. The results predict, for example, that the maximum Nusselt number is in-
creased as the swirl number increases and its position moves towards the inlet sec-
tion. They also show an increase in the Nusselt number as the Prandtl number in-
creases.

,II.

Unclassified
SCUMIlY 1Z.ASSIPICAIO 4 OP TWIS PAG~tMlh@ DO* IrCOVE)
 Sixth Annual Summary Report

CONVECTIVE HEAT TRANSFER FOR SHIP PROPULSION

By

M. A. Habib and D. M. McEligot

Aerospace and Mechanical Engineering Department
University of Arizona
Tucson, Arizona 85721

Research Sponsored by

I Office of Naval Research

ONR Contract Number N00014-75-C-0694
ONR Contract Authority NR-097-395

Atoession For
.1 NTIS -GRA&I

DTIC TAB ]
]
DTIC
Unannounced
Justtficatio
Dt
tviuton ___1 April 1981 E ECTE
Availability Codes E
- Avail and/or -- DEC3 1981
~ j Dist SpecialD
' D-
!D

Approved for public release; distribution unlimited. Reproduction in whole

or in part is permitted for any purpose of the United States Government.

._ _ _ _ _ _
 I CONVECTIVE HEAT TRANSFER FOR SHIP PROPULSION

M. A. Habib* and D. M. McEligot

I Aerospace and Mechanical Engineering Department
5 University of Arizona, Tucson, Ariz. 85721 USA

j Abstract

I Calculations of the flow and heat transfer of a swirling turbulent

flow after a sudden pipe expansion are reported. The calculations were

I obtained by the numerical solution of the time-averaged forms of the

continuity, momentum and thermal energy equations together with equations

for the turbulence kinetic energy and its rate of dissipation. For a

sudden expansion without swirl, the predictions produced satisfactory

agreement with available data for the Nusselt number. In the swirling

I case, there were no heat transfer measurements available for comparison.

However, existing measurements of a swirl flow at an expansion without

heat transfer were used to test the validity of the flow field calculations.

1 The predicted effects of varying the swirl number from zero to 1.0
on the heat transfer behavior are presented for a range of Prandtl numbers

I from 0.7 to 10. The expected effects of the swirl on the velocity and

temperature fields are also reported. The results predict, for example,

that the maximum Nusselt number is increased as the swirl number increases

I and its position moves towards the inlet section. They also show an in-
crease in the Nusselt number as the Prandtl number increases.

I *Presently at the Mechanical Engineering Department,

Cairo, Egypt.*
Cairo University,

1 Now
 TABLE OF CONTENTS
Page

ABSTRACT. .. .................. .. .. .. . ..

TABLE OF CONTENTS....................... .. .. .. . . ..

LIST OF ILLUSTRATIONS .. ................... iv

NOMENCLATUE.. .. ....................... v

1. INTRODUCTION .. .. ....................... I

1.1 Background. ....... .. .. .. .. .. .. .... 1

1.2 Previous w .. .. .. .. .............. 4
1.3 Present work .. ..................... 5

2. ANALYSIS .. .......................... 7

3. PREDICTIONS .. ................. ....... 13

3.1 Zemanick and Dougall - heat transfer without swirl . 13

3.2 Beltagui and MacCallum - swirl without heat transfer .15

3.3 Heat transfer with swirl at a sudden expansion. .. ... 20

4.* CONCLUSIONS .. ................ ........ 32

APPENDIX. Computer program TEACH-Y. .. .......... A-1

REFERENCES .. ........................ R-1

DISTRIBUTION .. ....................... D-1

I
I
I LIST OF ILLUSTRATIONS

i Figure Page

I 1 Idealized geometry ...................... 3

2 Comparison for heat transfer at a sudden expansion

12
without swirl ............................
3(a) Mean velocity distributions for swirl flow without
heat transfer ..... ...................
.... 16

3(b) Mean velocity distributions for swirl flow without

heat transfer (cont'd) ............... 17

S4 Development of mean velocity and temperature pro-

files for different swirl numbers .... ......... 21

1 5 Development of wall shear stress for swirl flow . 23

6 Effect of swirl on development of local Nusselt

I numbers ...... ...................... .... 25

7 Effect of swirl on overall heating .... ......... 27

1 8 Effect of Prandtl number on temperature distribu-

tions .. ....................... .. 29

9 Effect of Prandtl number on local Nusselt number.. 31

I
I

I
I
j iv

IAL
 I
I
I NOMENCLATURE

I Symbol Definition

1 CR Constant in turbulence model, for correction for stream-

line curvature, equation (11)

D Diameter of pipe downstream of the expansion - 2Ro

d Diameter of the pipe upstream of the expansion 2r

G Generation of turbulence kinetic energy K

H Step height = (D - d)/2

H Enthalpy
1 au U
K Turbulence kinetic energy i i

Nu Nusselt number

Nufd Nusselt number of fully developed pipe flow

Pr,z Molecular Prandtl number

Pr,t Turbulent Prandtl number

qw Wall heat flux

R Radius of downstream pipe

r Radial coordinate

Re Reynolds number; Red, based on diameter upstream; ReD,

based on diameter downstream

Rf Flux Richardson number

S Swirl number, IorO U)rr2dr/r ro 2rdr

0 0

I Tb Bulk temperature, ftTrdr

fUrdr

Tin Inlet temperature

Iv
Li.~
 I
I vi

I NOMENCLATURE--Continued

Symbol Definition

T Wall temperature

IAxial mean Velocity

UUav Bulk velocity = fpUrdr/fprdr

u Fluctuations of axial velocity

V Radial mean velocity

v Fluctuations of radial velocity

SW Tangential mean velocity

I w Fluctuations of tangential velocity

x Axial coordinate

X Axial distance downstream from the expansion

y Transverse coordinate measured from the wall

I Rate of dissipation of turbulence kinetic energy

i :w von Karman constant for turbulence model, equation (10)

Tw Wall shear stress in the axial direction, equation (9)

I Subscripts

I o Wall

p Node point nearest the wall

I
I
I
I
K , ii
 CHAPTER I

INTRODUCTI ON

1.1 Background

Current Naval propulsion plants are powered by variations of the

Rankine cycle (steam) or the open gas turbine cycle (air and combustion

products) plus some diesel engines in small ships. Alternative power

systems suggested include the closed gas turbine cycle and cycles involving

dissociation of the working fluid in either a Rankine or a gas cycle.

These latter two are believed to of fer the potential of substantial im-

provement in the power-to-weight ratio of the propulsion plant. The present

and proposed studies consider basic problems in convective heat transfer

j and flow friction that are important in all of the above.

Convective heat transfer provides the dominant thermal resistance in

several components of conventional steam power plants as well as in all

heat transfer components in gaseous cycles. For example, heat transfer

from the condensing steam to the cooling water in the main condenser of

a Naval ship is via the condensation heat transfer coefficient which is

high, then conduction conductance through thin tubing of high conductivity,

I and finally via the moderate heat transfer coefficient of the cooling

I water flowing through smooth circular tubes. Thus, the overall thermal
resistance is dominated by the convective thermal resistance of the cool-

ing water inside the tubes. One can expect significant reductions in tube
 2

length and, therefore, size and weight of the condenser and overall plant

if thie convective heat transfer coefficient of the cooling water is in-

creased appreciably.

A common geometry recurring in compact Naval propulsion plants is

a change in duct size in the primary fluid loop. As a consequence of the

upstream plumbing, the fluid is often swirling about the axis in the

piping. Heat losses from the primary fluid and thermal stresses in the

component depend on the convective heat transfer from the fluid to the

component as it undergoes this geometrical transition. The idealized

problem is a study of heat transfer in a sudden expansion with swirl flow,

the subject of the present report. The same situation also occurs at the

entrance to some heat exchanger tubes.

The flow downstream of a sudden expansion consists primarily of a

central jet, with or without rotation due to swirl, surrounded by a re-

circulating donut-shaped vortex induced by the separation as the jet

leaves the central tube. This jet gradually spreads as the flow proceeds

downstream until it reattaches to the downstream wall and fills the larger

j tube. Eventually a fully developed velocity profile can be expected to be

established if the tube length is sufficiently long. The centrifugal

body forces induced by swirl modify the recirculating vortex, the develop-

j ment of the jet and the reattachment location.

Significant enhancement of heat transfer rates generally occurs

when fluids are transported through sudden enlargements in pipes (Fig. 1)

or duct cross-sections or when they are subjected to angular tangential

 I

I I 3
I
I
Tin

Ii n
Red

x D-2R

I
I
I
I
I
I
!4
momentum due to twisted tapes or the application of vane-generated

swirl. In both cases, the levels of turbulence kinetic energy are

increased and the heat transfer rate generally increases as well.

High shearing rates created by both separation and swirling effects

are associated with high generation rates for the kinetic energy

of turbulence; in addition the rate of dissipation of turbulence kinetic

energy decreases due to an increase in the length scales. The two

Ieffects lead to elevated magnitudes of the kinetic energy of turbulence.

g All these effects are expected to reduce the thickness of the viscous

sublayer through which heat must pass largely by molecular diffusion

and so will increase the heat transfer rate.

1.2 Previous work

The problem of heat transfer to separated and reattached sub-

I sonic turbulent flows downstream of abrupt pipe expansions has been

studied experimentally by a number of investigators such as Ede, Hislop

and Morris [1956], Krall and Sparrow [1966] and Zemanick and Dougall

I [1970]. Reviews of the heat transfer literature for separated and

reattached flows are reported by Eckert et al. [1971] and by Fletcher,

Briggs and Page [1974]. The experimental data of Krall and Sparrow

[1965] for water, and of Zemanick and Dougal [1970] for air are for
typical geometries representing axisymmetric abrupt enlargement of

tubes. Both teams found a maximum in the local heat transfer coeffici-

ent at the assumed point of flow reattachment. Adequate descriptions

of wall heat transfer at solid boundaries in regions of separated and

1
I
'I
5

reattached flows have been limited by the lack of a complete solution

for the fluid flow field. Spalding [1967] presented a power law re-

lation between the Stanton and the Reynolds numbers for heat transfer

from a wall adjacent to a region of separated flow; he employed a one-

dimensional model of the flow near a wall. Recently, Chieng and Launder

[1980], in calculations of the turbulent heat transfer downstream of

an abrupt pipe expansion, found that the Nusselt number is overpredicted

by 20% in flows with high Reynolds numbers and that its maximum value

occurs about one step height upstream of the calculated reattachment

point, relative to the data of Zemanick and Dougall.

For swirling flow, a number of studies have been done to investi-

gate the effect of swirl on heat transfer parameters. Experiments

performed by Thomas [Eckert et al., 1971], with swirling air flow

in a tube indicated that, for turbulent flow and Reynolds numbers less
5
than 10 , the swirl flow transfers heat more efficiently than a non-

swirling flow. Zaherdadeh and Jagadish [1975] concluded that an aug-

mentation of up to 80% was obtained, with a constant wall temperature

boundary condition, with the use of tangential vane type swirl generators.

1.3 Present work

Calculations of the effects of separation and swirl on heat transfer

parameters have rarely been reported and the combined effects of both

apparently have not been studied. Thus, we attempt the prediction of

such cases in contrast to the "Dostdiction" usually employed in numerical

studies. In the present paper, a calculation method is provided and

__
__
__-
 6

is applied for the prediction of flow and temperature fields in the

swirling, separated and reattached flow in the vicinity of a sudden

expansion of a pipe.

An idealized version of the geometry is shown as Figure 1. The

expansion ratio is 2 and the Reynolds number upstream of the expansion

is taken as 5 x 104 for the calculations. The effects of swirl on the

heat transfer parameters are studied in the range of swirl number

from 0.0 to 1.0 and of Prandtl number from 0.7 to 10. For confidence

testing, the calculations are compared with the measurements of Zemanick

and Dougall [19701 for zero swirl number and those with swirl, but without

heat transfer, are compared to the data of Beltagui and MacCallum [1976].

The governing equations and boundary conditions are described

in the following section which is ended by a description of the solution

procedure. The results are then presented and discussed. The paper

ends with a summary of conclusions.

I

CHAPTER 2

ANALYSIS

Equations representing conservation of mass, energy, turbulence

kinetic energy and its dissipation rate plus the momentum equations

were solved in axisymmetric coordinates with an extension of the

TEACH computer program developed originally by Gosman and Pun [1974].

In the present study, the version of Habib and Whitelaw [1980], which

treats double coaxial jet flows, was modified to give predictions of

swirl flow downstream of an abrupt pipe expansion and to yield the

temperature field and heat transfer parameters by adding the solution

Iof the thermal energy equation.

I The equations modeled and the turbulence model are as follows:

continuity, -

I oU0 (1)

momentum,
'- (2)

j 3
thermal energy,

-_14 (3)
turbulence kinetic energy,

I
-1( (4
I
8

and dissipation of turbulence kinetic energy

w i t h ( 6)

and

where 6ij - 0 if i J and 6ij 1 if i j. (7)

Fluid properties were idealized as constant except in some cases with

air where the density was allowed to vary due to pressure variation.
w In addition to the set of equations listed above, the following

i inlet and boundary conditions are considered. In the case of non-

swirling flows, data for fully developed pipe flow were taken as initial

] values for mean velocity and turbulence kinetic energy. The inlet
Fluidspropertis wer idaie as0cosatecprnoecsswt
3/2 its model,
dissipation rate was then determinedK from

where Z is the length scale that characterizes the energy containing

o
ai For swirl, the initial axial and tangential mean velocities were

interpolated from the data of Habib 1980].

-- The turbulence kinetic (8)
 9

energy was taken as a function of the total mean kinetic energy and the

dissipation rate was determined via equation (8) with Z assumed to be

given as r0 (0.03 + 0.07S).

All axial gradients, ;/ x, were presumed zero at the exit plane

of the tube well downstream. Axial symmetry was specified so V- 0 and the

radial gradients of all other quantities were zero at the centerline. No

slip and impermeable wall conditions were specified on all solid surfaces.

Wall functions [Launder and Spalding, 1972] corresponding to the

dependent variables were taken as

Uy
w y (9)

/
+
where UP 1 £n E
-- p y
Ey+p and C1 /4 KP1 /2p /v

- H) y+
i (H
and - p p (10)
y U- + P
p p f

where P -C (F__ _ (1! \ /

The value of the kinetic energy of turbulence near the wall, K is

calculated from the transport equation for K with the flux of energy

to the solid wall set to zero. The corresponding value of e was cal-

culated from equation (8) with Z C1 / 4

K
YP"
 10

The constants used in the above equations are in accord with those

of Khalil and Whitelaw (1974] and Pope and Whitelaw [1976]. They are

summarized as follows:

0
C 0.09, aK = 1.0, cc =1.22, aH .9 , CI 1.45,
(10)

C = 1.9, Cf = 9.24, = 0.42, E = 9.8

Modification of C to account for streamline curvature is described in

Section 3.2 later.

The solution procedure described by Patankar and Spalding [1972]

was used to solve the above equations. From trial values of the pressure

distribution, the momentum equations can be solved to obtain an esti-

mate of the velocity field at the first iteration. However, there is

no guarantee that the resultant velocity field will satisfy the continu-

ity equation, therefore, after each calculation over the solution

domain, adjustments are made to the pressure and velocity fields to

satisfy continuity. The procedure is repeated until convergence is

obtained. The energy equation is then solved and the temperature field

is presented.

Most of the computations were done with a grid of 25 x 18 nodes

with more points concentrated near the walls and regions of separation

and high velocity gradients. The computer time (per iteration per

g grid node) using the CDC CYBER 175 was 9.5 x 10 -4 Sec to solve the velocity

field and 1.2 x 10

-4 Sec to obtain the temperature
field. (These time

estimates typically corresponded to $24/run and $6/run, respectively,

at University rates). The storage requirements of the program were

37,000 octal words.

____
 11

A grid independence test was conducted during the comparisons to

the data of Zemanick and Dougall [1970]. It showed that an increase

in the number of nodes from 20 x 17 to 27 x 21 affects the predicted

Nusselt number only in the region of its maximum (see Figure 2) and

then less than three percent. The effect was negligible in the down-

stream region. For most of the predictions presented in the next section

a grid of 25 x 18 nodes was employed so they are expected to have better

numerical accuracy than the former case.

I ,
 I
i 12

I
11
I

1 8 0 eangic
° and Dougall [1970]

I Chieng and Launder [1980]

Nu Re =47,640

Nu i {:j0 0 d/D = 0.43

04 0 0

I 27 X210
I 0 0

10
0 5 10 15 20
I X/H

K' Figure 2. Comparison for heat transfer at a sudden expansion without swirl.

H'
I
."
 I CHAPTER 3

PREDICT IONS

In this section results from the computer program are first compared

to the available data to test its validity. Then, the predictions of

the separated flow with both swirl and heat transfer are presented.

3.1 Zemanick and Dougall - heat transfer without swirl

The results are first compared with the experiment of Zemanick

and Dougall (1970] for zero swirl. Since inlet velocity profiles were

not presented, the initial conditions were taken as those of fully

developed pipe flow in the smaller tube. In the run chosen for compari-

son, the expansion ratio was 0.43 and the Reynolds number calculated

downstream from the expansion was 47,800. In the experiment resistive

heating was employed so the appropriate thermal boundary condition would

be a specified wall heat flux. To correspond with the procedure of the

present program, the wall temperature distribution which was measured

I was entered as the thermal boundary condition and the wall heat flux

was calculated via equation (10).

As noted by Zemanick and Dougall [1970] there were some difficulties

j with their experiment so it is not ideal for a close comparison, but

it is one of the few with data readily available. Axial conduction

I in
the tube was significant and the thickness of the copper electrode

at the expansion extended about a step height in the downstream direc-

tion. In addition it may be noticed that their correlation of local

13
 14

Nusselt numbers in the fully developed region downstream is about 15%

higher than commonly accepted values [McEligot, Magee and Leppert,

1965]. Unfortunately, quantitative estimates of the experimental

uncertainties [Kline and McClintock, 1953] were not presented. If

one assumes an uncertainty of 1*C in measuring the wall temperature,

typical of premium grade thermocouple wire, it can be shown that this

effect alone would lead to an uncertainty of about twenty percent in

the Nusselt number near the entry and approximately five percent down-

stream; these estimates are shown by brackets on Figure 2. Thus, it is

best to restrict the comparison to trends rather than absolute magnitudes.

Figure 2 shows a comparison between the calculated and measured

distributions of Nusselt numbers; the curves represent the numerical

predictions. It is shown that the agreement is generally good in the

upstream and downstream locations with the present calculations under-

predicting the maximum Nusselt number by about 20%. The location of

the maximum Nusselt number is found to be one step height upstream

from the corresponding peak in the experiments. By examination of

the predicted velocity distributions it may also be found that the

calculated maximum Nusselt number occurs two step heights upstream

from the calculated reattachment point. This direction was also indi-

cated by Chieng and Launder [1980]; however, they indicate the

difference in location to be one step height only.

The discrepancy between the calculations and the experiment may

be attributed partly to the numerical accuracy in the calculation

 15

procedure. Increasing the number of grid points to reduce effects of

the numerical diffusion leads to a slight increase in Nusselt number

in the entry region as shown on Figure 2. However, as noted earlier,

the increase in the number of nodes from 20 x 17 to 27 x 21 gives a

change of less than 3% which is essentially negligible for the present

purposes.

3.2 Beltagui and MacCallum - adiabatic flow with swirl

For swirl, data with heat transfer are not available to compare

to predictions; however, the adiabatic measurements of Beltagui and

MacCallum [1976] were considered to test the flow field calculations

and to provide guidance. It is expected that when the model provides

predictions which are satisfactory for the flow field calculations, then

the heat transfer calculations which depend on them will consequently

be reasonable too.

Predictions were calculated for the conditions of the experiment

by Beltagui and MacCallum [1976] with Re - 2 x 10 5, S - 0.67 and

d/D = 0.413 as expansion ratio. Their measurements do not include

inlet velocity profiles; therefore initial values for the present cal-

culations were interpolated to S = 0.67 from the data of Habib [1980].

Results based on the standard K-e model of Section 2 are shown

as curves with centerline markings (dash-dot) in Figure 3(a) for axial

mean velocity profiles and Figure 3(b) for circumferential mean velocity

distributions. Dashed lines represent the experimental observations.

It should be noticed that the ordinate scale changes as effects decay

downstream.
 9 16

-- -11

00

m4- )

CL

0 L

-- 0 0

4-) L

0-A

c-o

0 0CO

o - 2

Q~II
 17

X/D 0.06 [X/D 3.72

6.0 -3.0

W~

/C R
2.0 1.0 / /R0

0 0/. I
0.4 0.8 1.0 0.4 0.8 1.0
r/ R r/R

1Figure 3(b). Mean velocity distributions for swirl flow without heat transfer (cont'd).
 18

One effect of this relatively high swirl is that the central jet

rapidly takes the form of a circular sheet which moves radially outward

and attaches to the wall of the outer tube in a short distance. Conse-

quently, the surrounding vortex becomes significantly smaller in size

than without swirl and is seen as a small reversed flow region at only

the first axial location. Along the centerline of the larger tube

a reversed flow also appears showing the existance of another recir-

culating vortex of size comparable to the diameter of the smaller tube.

The qualitative trends are in agreement with the data. The axial vel-

ocity profiles agree well except near the centerline where the vortex

is predicted to be shorter than measured (e.g., compare centerline

velocities at X/D = 2.63). Also the tangential velocity in the central

region is predicted to decay more quickly than observed.

The discrepancies are partially explained by Khalil and Whitelaw

[19741 and Habib and Whitelaw [1980], who showed that the standard
K-e model often yields underpredicted lengths for the central recircu-

lation zones of swirling flows. The work of Pope and Whitelaw [1976]

for flows behind a baffle and of Habib and Whitelaw [1981] for confined

flows with a sudden expansion indicates that the dissipation equation

is the weakest of the governing equations since its modelling is only

based on assumptions. As shown by Bradshaw [1973], body forces due to

streamline curvature can have a strong effect on turbulence and the

equations of turbulence models should account for these effects. Re-

cently, Rodi [19791 and Morse [1980] found that it is essential to

.I
 19

correct one of the empirical constants in the dissipation equation as a

function of Richardson numbers to achieve agreement with experiments.

To improve the present calculations, we replaced C the empirical

constant in the dissipation equation, by CI(l + CR Rf) where Rf is the

flux Richardson number [Bradshaw, 1973]

Rf r~~\
Dr~ (11)

The use of CR= 0.2 was found to increase the length of the recircu-

lation zone from about 1.7D to 3.6D. This value was chosen to pro-

vide the best agreement with the measured size of the central recircu-

lation zone.

Preictonswit CR = .2 are presen..d as solid curves in Figure 3.

The axial velocity profiles show considerably better agreement with the

data in the central recirculating region, as would be expected from the

manner of adjustment, and slightly better agreement near the wall.

The predicted circumferential velocity distribution is also improved

near the wall, the region with the greatest effect on the heat transfer,

but as the centerline is approached the values appear to be under-

predicted. Comparison of the predicted velocities with CR=0 and

CR= 0.2 shows that modifying C 1with this functional dependence on

Rf tends to decrease the rate of diffusion in the downstream sections

and to suppress the development of the jet.

 20

3.3 Heat transfer with swirl at a sudden expansion

Calculations of heat transfer to a swirling flow were made after

an abrupt enlargement of D/d = 2 for an inlet Reynolds number (Re d)

of 50,000 and with constant wall temperature imposed after the expansion.

It was idealized that there was no heat transfer through the annular step

between the inner and outer tubes. For flows at the different swirl

numbers studied the inlet distributions of velocities and kinetic energy

of turbulence were interpolated from the data of Habib [1980). The

effects of varying swirl number and Prandtl number were examined sep-

arately.

Effects of swirl were studied with the Prandtl number taken as

0.7 as for common gases. Predictions are shown in Figures 4 through 7.

Figure 4, presenting the development of velocity and temperature dis-

tributions, provides information which helps understanding the later

figures. On these figures the solid curves represent the reference

case: zero swirl.

In Figure 4 the velocity profiles show that at S = 0.1 the flow

is only slightly modified, at S - 0.4 there are significant differences

and for S = 1 the general pattern of the flow changes. These runs

might be considered as showing low, medium and high swirl, respectively.

As the swirl (or angular momentum) increases, the reattachment point

of the separated region in the corner moves successively closer to

the entrance and the central jet spreads more rapidly towards the wall.

j After a short distance the velocity profile for S =0.4 resembles a

 T-Tin21
GOV D/d=2, Red: 5 0 ,OOO Tw -Tin 0.8 Pr 0.7
0.6

2 0.4
0.2
00.0 I

0.8

s~o 0.6

2O.i4 X/H=IO0.
Y. _..~ _.-- 0.2

0 -0 0.0 -

-1 0.8 -

6 -0.6-

2 - X /H 5 0. - 0
0.04
0 WO
-2--' 0.8

6- 0.6-

X/H I
2 //0.2
0 0.01-o I
C 0.2 0.4 0.6 0.8 1.0
I i
C~0. 2 0.4
g -2
0.6
I Ir/R
0.8 1.0

g . r/R

Fiqure A. Develoornent of m'ean velocity and ternrerature orofiles for

1 difFerent swirl numbers.

- -IL -. i4
 22

plug flow. The velocity near the heated wall increases with S in this

entry region.

At S - 1 the flow pattern resembles that of Beltagui and MacCallum

[1976] shown earlier. For this high swirl number the flow exhibits

a long central recirculation zone while the outer recirculation zone is de-

creased to 0.13H. Flow development is delayed as the central jet be-

comes a wall jet with the reversed flow at the centerline persisting to

beyond XIH -17. These features are also demonstrated by Habib 119801.

The corresponding variation of the non-dimensional wall shear

stress or friction factor is shown on Figure 5 (note the change of

scale for S -1). Negative values occur due to the reversed flow near

the wall. The value of zero indicates the location of the reattachment

point or position where the dividing mean streamline meets the wall.

As S increases this point progresses upstream as noted from examination

of the velocity distribution; the points of maximum shear stress like-

wise move upstream. The magnitudes of the extremes also increase with

swirl; an effect of the centrifugal forces apparently is to drive the

higher velocity central flow towards the wall, thus steepening the

velocity gradient. For fully developed, turbulent flow in a tube at

Re D= 25,000 the friction factor is expected to be about 0.006 [Kays,

1966] so all the flows must go considerably further downstream before

becoming fully developed.

The temperature profiles in Figure 4 show approximately isothermal

zones across the recirculating regions, thereby demonstrating high transnort

----
--
 23

0.03 0.15
Re d 50, 000

0.02 S /: 0.10

0.011 . * -- 0.05

,TW 2 I.
TP b _ __
0
_ 0_ _ _ _ _ _ _

-0.01 -0.05

-0.021 -0.10

0 5 10 15 20
X/H

Figure 5. Development of wall shear stress for swirl flow.

I
24

rates due to the turbulent mixing in these regions. For the range

S < 0.4 there is only a slight change in temperature distributions as

the swirl is increased.

The interesting case is for high swirl flow. Within 5 step heights

(or 1.25 D) the central recirculating flow is approximately isothermal

By X/H = I this reversed flow has carried thermal energy forward from further

downstream and, thereby, has raised the centerline temperature sub-

stantially. The small recirculating region near the wall has done the

same for that region, giving the appearance of a cooled wake at an

intermediate radius. However, before X/H = 5 the central region has

expanded and has eliminated the wake appearance in the temperature

distribution.

Predicted local Nusselt numbers are shown in Figure 6 normalized by

0 8 0 4
Nufd = 0.020 Re . Pr .

a correlation for fully developed turbulent flow with a constant wall

temperature [Kays, 1966]. As for the wall shear stress, increasing

swirl increases the maximum value and moves it upstream towards the

entry. The higher velocity gradient corresponds to flow at a higher

Reynolds number so the viscous layer, which dominates the thermal

resistance, is thinner and the Nusselt number is increased. Comparison

to Figure 5 shows that for each swirl number the maximum Nusselt number

occurs near the point of reattachment, which was shown earlier to move

upstream as swirl increases,

-.
 25

Red= 50,000

4 lS=I D/d 2, Pr 0.7

0 5 20

II
 26

For the high swirl case, the increase in Numxdue to the combined

effects of separation at the annular step and swirl is four times while

for separation alone the improvement is 2.5, relative to fully developed

flow. Thus, the major increase is due to the separation. Further

downstream, the low and moderate swirl cases show a slight decrease

in Nusselt number as swirl number increases. This result may be the

consequence of swirl reducing the length of the outer separated zone

so that the development of the normal tube flow has progressed further

at a given location, i.e., for given X, X - Xreattachment increases

with S. For S - 1 the strong wall jet phenomenon continues to dominate

the velocity profile and, apparently, maintains a slightly higher

Nusselt number than for zero swirl.

For applications such as heat exchangers, the designer is interested

in the resulting bulk temperature variation of the fluid as it is heated

(or cooled). Figure 7 presents this information and it is seen that

an effect of swirl is to increase the bulk temperature and its rate

of increase slightly. After viewing the temperature profiles, it may

seem surprising that the increase is not greater for high swirl. The

difficulty is that, though the temperature is approximately uniform, the

velocity changes sign in the large reversed flow in the central region.

In calculating the bulk temperature via its definition (JtTrdr/UUrdr)

the enthalpy flow rate there is subtracted from the enthalpy flow rate

in the positive direction giving a net value which is smaller than

expected. Therefore, while local heat transfer rates are increased

substantially by swirl the overall improvement is not proportionally

as large.

-1
 27

0.30
Red= 50, 000
0.25 - D/d = 2, Pr =0.7 .

0.20
Tb -Tin 01

0.10 0'

0
0 5 10 15 20
X/H

Figure 7. Effect of swirl on overall heating.

 28

Predictions of the effects of Prandtl number on heat transfer were

calculated for S = 0.7, again at Red 000adDd 2 xmnto

of the expected velocity profiles in Figure 8 shows this situation to

be a high swirl case with a substantial recirculating vortex in the

central region. In comparison to the profile for S = 1 at X/H = 1

in Figure 4, one finds the same shape except the maximum does not fall

as close to the wall in this case with lower swirl. Lower circumfer-

ential momentum at the entry causes the main streamlines of the annular

jet to have less rapid divergence from the axis.

Temperature profiles were calculated for Pr = 0.7, 3 and 10, corres-

ponding to common gases and liquid water at about 60*C and 10*C.

respectively. Results are also shown on Figure 8. The general behavior

is in agreement with the earlier predictions for S =1.

At each axial location the temperature profiles have the same

general shape for all Prandtl numbers. Only the magnitudes change

significantly. Thus, the total thermal energy transferred to the flow

is greater as the Prandtl number decreases. The similarity of the

shapes may be taken as an indication that convection dominates the .A

thermal energy transfer processes in these regions despite the in-

creased thermal conductivity implied by lower Prandtl numbers. The

levels of the curves then depend on the heat fluxes at the wall where

the thermal conductivity does dominate the thermal resistance. Thus,

accurate prediction of heat transfer parameters requires accurate

knowledge of the behavior of the viscous layer in this flow as well as

in simpler ones.
 29

1.0 3
X/H=5
0.8 -2

T-i*0.6 -u I X/H 20 low

Tw-Tin Pr z0.7______

0.2 -

L.0.4 0.8 1.0 0.4 0.8 1.0

r/R r/R
Red=50,OOO D/d=2 S=0.7

1.0 1.0
X/H~I X/H=20

0.8 -0.8

TTn 0.6 - -i 0.6 -P=.

T -T Ti 04-1
w'n 0.4 -Pr 0.7 /04-
3
0.2 /0.2
10 ->10 - .-

*10.4 r/R
0.8 1.0 0.4
r/R
0.8 1.0

Figure 8. Effect of Prandtl number on temperature distributions.

 30

The axial distribution of local Nusselt numbers is presented as

Figure 9; the curve for Pr - 0.7 is consistent with the trends of

Figure 6 earlier. Again, as with the temperature profiles, the shape

of the Nusselt number distribution is primarily dependent on the flow

parameters--Re, S and D/d--and is relatively independent of the Prandtl

number. However, the level or magnitude of the results increases as

the Prandtl number increases so, as a general conclusion, one may

note that the local Nusselt numbers increase with the Prandtl number

as in most flows.

I
i-
kL~ .i.L
 31

Red= 50, 000

D/d =2, S=0.7

NU 4
Nuf d

0
0 5 10 15 30 40 50
X/H l--Change of Scale

i Figure 9. Effect of Prandtl number on local Nusselt nunber.

 CHAPTER 4

CONCLUSIONS

The following important conclusions can be extracted from the

preceding results:

1) For heat transfer without swirl, generally good agreement

is obtained between measured and calculated Nusselt numbers.

2) For an adiabatic flow with swirl, predictions could be ob-

tained in agreement with available data by modifying the

turbulence model to account for effects of streamline cur-

vature.

3) For swirl flow with heat transfer, (a) the maximum local

Nusselt number is predicted to increase and its location

moves towards the entry as the swirl number increases, and

(b) the local Nusselt number increases as the Prandtl number

increases.

3
32
I
.. - " "-
 APPENDIX. COMPUTER PROGRAM

1. Comments concerning parameters used in Block data of Teach T.CDC

IGET: This parameter indicates continuation of calculations

from previous output saved on magnetic tape

=l Does not read or write any output

>1 Writes data on tapeS8

>2 Reads from Tape 7 and writes on Tape 8

IHT Parameter controls heat transfer calculations

ElThere is a heat transfer calculation (the thermal

energy equation is solved)

EQ No heat transfer calculations

IC? Parameter used to restrict the program to calculations for a

constant property fluid
=Q Flow field calculations only

=l Energy equation is solved alone provided that the flow

is known from a previous set of iterations. In this
case (i.e., IC? = 1) IHT should be set to 1 also.
The program may consider changes in density when the
fluid is air. For this case set IHT = 1 and IC? 0.

IAIBN Defines North wall (i.e., larger tube) as adiabatic

EQ There is heat transfer through the North wall

El There is no heat transfer through the North wall

IABW Sidewall (end wall) boundary situation

=0 There is heat transfer through the side wall.

(This option was not tested.)

-- There is no heat transfer through the side wall

*Questions concerning the details of the program should be directed to

Dr. M. A. Habib, Power Section, Mechanical Engineering Department, Cairo
University, Cairo, Egypt.
A-1
 I --

A-2

CRF Constant C used with equation (11) for streamline curva-

ture correction. (It is only useful in the swirl cases.)

=0 There is no correction

=0.2 There is a correction (with C R = 0.2)

MAXIT Maximum iterations allowed. (It will write the results

if IGET>2)

INDPRI Defines the number of iterations after which the program

shows a sample of the calculations
2
PO Atmospheric pressure, N/m

CPl Specific heat, J/Kg C*

GASCON Universal gas constant = 8314 J/Kgmole C*

WFLUID Molecular weight, Kg/Kgmole

COND Thermal conductivity, Watt/m C*

NOTAIR

=0 The fluid is air or a comparable gas

=1 The fluid is not air and the properties for the fluid
must be inserted. These are identified as CPFL,
CONDFL, VISFL, DENFL)

PRLAM, PRH Laminar and turbulent Prandtl numbers

DENFL

CPFL The properties of the fluid to be used if it is not con-

sidered to be air. (NOTAIR should be equal to 1 to con-
VISFL sider these.)

CONDFL

ICHF Indicates whether there is a constant heat flux boundary

condition along the larger tube.

-0 Not constant heat flux; the program needs a specified

wall temperature distribution

=1 Constant heat flux boundary condition

 A- 3

QADD Wall heat flux in the case of a constant heat flux

boundary condition, J/m2 s. (To be used only if
ICHF = 1)

TIN Inlet temperature, K0

SWNO Swirl number, defined as the axial flow of tangential

momentum divided by the axial momentum rate,

f --r2 dr/R0
2
rdr, where R is the nozzle radius.

=0 No swirl

=any value Magnitude of swirl number

UFUIN The maximum inlet velocity (on the centerline);

important only if SWNO = 0. The program then
treats the inlet as fully developed pipe flow and
UFUIN is the value at the centerline.

RLARGE Geometry
& +
DNOZLE/2 RLARGE
DNOZLE

IDATA Parameters specify if the input velocities, turbu-

lence kinetic energy and dissipation will be sup-
plied via READ statements or will be calculated
by the program as fully developed pipe flow.

=0 For swirl number = 0 (SWNO = 0)

=1 For others. In this case, at the University of

Arizona Computer Center, the program Teach.Vel
should be run first; its output should be saved
on Tape 4; then the presentprogram is run and
it will automatically read U, W, K, E from Tape 4.

NI, NJ Number of nodes in x and y directions, respectively

JNOZLE Number of radial nodes in inlet tube. /

JNOZLE --------- _)

X Locations of nodes in the x-direction

Y Location of nodes in the radial direction (starts

from centerline)

I
 A-4

TWALN North wall temperature, K

TWALW Side wall temperature, K

2. Program listing (following pages)

 A- 5

Ic

Nt -- o

0-N0

-* --

;" 0 m ft -
- m
ft- cm m. 6m Z x 4
ZZz 66 WP-

N ~aNamm:4
NN P..
A6n
t.y ~a 0Z 0 ~
12.3L:
j~a
L&5 aC
WL
w... -
a.* ;2La~'

" aJ a 5 ZN vcu a." 3 =;322

- '
~'5-.50
aP P
A4J
A. a 5 LL
05L- 5 5
a~,
Zi
7 . .
;~
^4 ..
4. S- -
:
-

~ ~
-- ~ 2 ra&- ~ ~ I'a .=Z=-~ ~ ~~ ~~ - C

- ~
-4 -c a-u. 4-7 N *. a cZ-O.. A.V. *. 5

a.. ~ ~ ~
s ~ 'N a - *5-*
 A-6

I'In

.f7

Co 2 mm

*l - c-1

* . )r4 2.
.7Olnc.

- .7 I-
 A- 7

*1 wMj4 lMI ----

.0 ftW.~IC
0.~' .0 C'-f~ r~i"~P- Ma 1- Z ~ -t 400~CC.. ~ ~

oj- 4- ac*

.0 NZ "AO

*. A* %Z ; tm
z M* ft =(" U
* w D- C -"J0
47" * - Lj 0 A

t. r'. :- - - -- =
, * *0 3 :f ; *N

*m
4
4j
a *
*k
-Pjr
-
%~ -ft
~ ft f
t
C..
i
91L,* - .
*l N .O~0
c cc x
-. *-. - - a*J 3*m4- L 0. -a~.N ~ 0
* f * 4
pM * A... 00 - N I~l0.~* 4 4 0 j

~. . o. * - r - I~ - - -cd I..d~A-r

- *4 - L
ft - Q . N N 03-4j 4
ca** I 4 ~ *0. 0L. - X Z.

4L .£ 4 N. M0 N - -'N~f .4 3

0.*36~. 2 4j 1..~AN N N 04 ~ C . Z *-

J.d f N I 4 -
c
I-Ix
* 0
.
.VI* 4
f.-
~N *0 0 ~ U.Z f-
 A- 8

.0Ua C ^'w'

Z ZZt:C9- !Z!Z!.2U E7!!:C zzi4.Izz.x.I- .Z .zC C = = z

c Ty w z
9-.D I7 Z.4
Z . a.r..,, ,t gr u u. C'.C.''C.fI7~N.

N6

-I. I

_JU . 2.O a.

.3u .4 . .

* a.

0 N- .a. e-4 ,J

N ~ -4 .4-.ja
U..
 A-9

r~ ~~~~~~ - - . . -- ----- -- -- -- - - - --

- ~ ~
Tr71r=rzT 7Z==ZZ rZZZTZ -TZ
ZZrZr rETZ2 zrrT TT K

4&
o4
*
*
~~ ~ 44
s
44
4b4
.*
a.
44 4 4
4
4

44
44
4
440
44 4 4.44 4 44 440
44 44 44 * 4444 4
44 ~~~44 44 4 4 44
44 4 4444 4 44 440

C14 444 o. 44w4 4 44

'. cc 4h 44. 4 4 44
o4

04 44b4 444 44

2. 44 44 4 44 444 44
- 44 44w 44 o4 4 4 44

w 44 4494
p4 4 444

4-4 4 4 4 4 4 4 44

-4 4444 4 44 4 44 4444

4 444 - . : 4 4 =x4
Z 44c 6 .v

4 w4 48. 44 j 44 44 4: 44.

- ~ ~ ~ ~ Z. c4 fm4 4 444 44

44 -o 44C 0. 44 A40 4 44

Z% 4-,4j .J.~ 4 44 . 4 44_ uU/ 4 4

 A-10

r-.f CC-r 0. .'* Co e

4 ,If le.. 0' If a-. c :;^j0'7 fl a- a C,
en.~- In' :4',N
o :M m~ wr we r

.** ~Z zzXZ Z Z Z Z Z Z Z Z Z Z Z Z Z Z ~ 7 Z Z ~
c. Nc-z _- -- -

- 44 4 44 44 4 44 44 4 44 4 44 44 4 44 44 4 44 444L4 c

x, = 41

-4r

-- j .1*= ~ r i

-o
-- ~ .24
o n~
 A-11

~~ -'N4. ~~ ~ ~ - 00-K.~pj~ C-

o
*
e~ooo
A~NA
ooe
ooeect-
N N f*NA.
,
A N
.-.-.-..-
A t. Ai
E..I.
AN.N AJIN NAJI
.~.
A N.
.
NA
~
~NN m-i-
'&A.A
t--
N. NNAf0
4r4
AI4A

-f-,

* 4*

Z* t2

*l 4
I A-i12

e'l wcazt -' C wa0 ; Mrn rf '.Fa0 ; -v' ,.r fj q

-~ 441.~4 4 xx

% ~ ~~rr~T~w~rTma

i., 494

f4.
Z' CL-~ 4 4.4

-.- C xf - a 4

11 Q
 A-13

* S' 'otfccWo~~
0C ~m w r a
~ ~ I
*~~-- a C.
a~CN'AIAr-,Cf-

- U3

- 444
44 4444 444 444 4444 444 444 4444 444 444

%~c rTr~~r7~~c LUZ~YT~r;1r

TrrrTrT

. . . ...

- ------- - - -.. -
z -

C. Z- vC .M ,0 L L L

I-I-
 A-1 4

44-.

A4

* Ir

C. C
-AO- I at * z: =

.7 d -7r ZP.0 Z o

.-- . a k - X-z,7I

z_ N- -,4 z

C.Qt AI CA W

X%-, -~FC5 Z- 0 ~
At~1 j -V--~
 A-15

4.

- uz2:vr'TTr,-wrgrTr SICI

NJ

:w
00e

. C
"

,. .'. " -'

OMQ. AL

. I-

Itmot"I

b..........

I. _N
 A-1

N 0

-y N

CL - a .- ~Z -

N0 - . a ,.-

* 2 - ww=.i
O- .f. 13 Z' W-11! 49 'r

3t ~ ~ 4 C

-NC . .~-- =PJl:t a Z=~N

C. I'M
- ~ N J .. If' ,,

C .N ~ ~ Z ~ - q) C6

4-- - . ~4J . t --

01r 2 -- ~ .. *

O ~ a ~ U.. -I
4 U 'M..). ~ 2' C U..
 A-1 7

4- C

CL o.

W,
C2 a;

-. t-.

- a aa - 3 -- ~ ,X4

- -, -- - - . - I
I
I A-i8

I '7

I o

N
~

I - -
~
- - ~ -

I N

I N

I z
~ '4
N

I
I
-h

.4
.4

I *1

ON

~~
t.-

- 0
0

. 'I
0~.4O~0~V~O0

N (NJN

I N

I 4'
.~ 4, *,~ 0
- N N ~ '7

1
1 I. ' . t.
 A- 19

o CVCICNS S'NSE~~ E"'

4
'4? ~

1~~~ zVVVE"~AE 'El 1'4'~~U

o-% -1.

LA W. w m -L

N
As JaZ
4- aEJ -L

:z.(S. 13 Z-- 0. '45 'L'O

z N" 442 J-
aW==- Q0 - 4.- f4

"3~CI
at Oa S%~-
Z--2~ '.

o a a CW5 p- J0. &1 a

Z Z -

r~~~~~~ N- 0 -3 - - -a~
-~g -4 - -34s
CIL 9(* EJU ~ ~ a- iU

Si
0. N 3
~~ ~ ~ O
.J
jEE
JL
Ek
VE."44aN
0
N.
4
-
'aa
40.0
N4"0
4-. E~a
S 5Ja*44O~S
%Itt.-U.
--
-
zZ4.

Ch -' fO -a. )'5.. aA t .. ,\ -

1
I A-20

I
N %J~J('~~4%JA~J0~~ E~~9E~' ~ E'~ ,t.t ~r

I ~ V V V%~ (4fVW'U'U
I -. ~ a.~&a.a~a-~.a.a.
I ~ ce~Dccccr~'r~~
- ~
% ~ ~.&.

4.~J
I.-'

U. U,
0
I-,

- 0
~ -J
* *0
- U,

o -U~
U. - 0

K ~ C
* I
~*J . -
%J* - ~
0* - -
* -- . -

* C- -, 7

9-. - - )-~~ C
-, CL
C - V~U. -

9- '.n--4-C
.5 --
* -- 0 .~.,uZ.
T- TT~..- ~
90 ~VU-J

I
,ZT

~ -~r a .~
- 'V.J '~Je.JU,-0C~~.4~9. ,,
* w- 4 M-~ U-EU.G-
.7 JJ9)4 *U.~ ~
~ I~I
C )0~ C a I a- ~
-or ---

I ~

~Co-~
~Z~Z

CC~-
~VQV

I to
9--

I '~ -V
a) C ~- f

I
I
I
 A-2 1

4 LC
*
Crwtv
4
IJL.~~~z
r4z
4
T ~ (J
gr ~r z FTr2 31ft
-1 _.
~ r
-1L(I.ILI
Z.-~aJJ.J

o, N N
N;:7
,
-a4 -.
': It
.- M4 "0
*

Ne - 7' -A.J r 4-

O
7is N N Z Ire t U.. L-- _

- .f 1 t 4- , 4fln

N L -,, 40- b:2 03-

4.43_0 . -el a L g - 0.

-w-C NM. '' - - -d-= 4% * c-*

ol X.3 a M- ~ 0- M M -- a. MA .

a .~lU .U .. 8 ac-

tu3 J N30~C31-
dU
NI 401
4 0 LI - 0-.. - .. -

t-k1 z- Z U. . *.- L Li

- '4~N
- On J - ' ~ja- N00000.0-
 A-2 2

4.7

AM
o NNE
'7~...t~77~t
~ 7
C.7~~CE
17.7.7.7f~
r.a O . ~ U P 07-7
~tC~.OC.Cfl..7 O-tE7tt
-77 7

-- j at00

% LJ~
L)~W4J
6JL~L ~J.J4L'LJ~ ~ ~U1 M...M.J~~~~CL
4~~ .J..J ~ ~ 1.JJJ ~ *.J 0h.J L ~ J .. J j J~ ccJ~

9L I

-az

iC%. -a
-* 77 "L I-l44 00
C. . JC
. xx
-a-- 3 N
00

-- ~~0 040- CLq~

-~L - t-I
ji

ac . - -

33 JD 9~ 4 -
I ~

1
I A-2 3

.3
s NrcPo..fl% W'.s flOP-a e
o r-NP-E~earrrctr

ft.

'% LJLJ UL IJLJJtJIJE.JLAJLJ

~j~e jLu~Ja~J ~jlJ
-o C.EC4444CC4 4.44
-. LhIJ(JLLDLHJIJLJIJLJtflJ
4'

'U

.3
*0
N
C. .3
.3

.3

.3
'4.
.3
z
C
-i .4

-
-
a
44,
a
4. Lu
- La
-, r
* LU
- a
- Lu
a. .~
.4 U.

4,

t~ r

~I? Z
C~ 4L1L C
- U.
'% r~.. - -
-D tN
'~ C 2 -
~*a r

- Ctfl~ 4-~j
- .JtUIQ * 3
.7 Li6 *S4..iJ ZV~
r- a--an .3 -4
1'flZZ ~.jZ
a,. ..-- 0 a.
Lu~ 4 s-'Z

£aZ~~ C4uZ
a

4, aQU. C
Lu I CC a C
C I MP' C .3
'-I I -
U 4,4,1...

C
a

0
- N

~aafl~L*
 A-2 4

-.. C., C'.eJ fs C' - CWC. C'r -S0'x7 f--

00

el ft N 0% zm XZ

a
Z C - x
4l =U iA7 Ma -~ t
* 0 a>,M 0,

*
.7
0'S-
4
; aL
~N p
N-
-
*
- 7-
.73
W
-

~
NA,
0'r-zn ~ ~ ~ w zoa J gL- 4..
a-66
.a, 9S ;WeiS . L&Ck I-- C. -- 4 .-

,,i
0
tM4 0i
~~ ~ 1'a ~~a 7 44
.
C^ .
~ j
a 5
aa.-

44 -~~ - 4)5),. - AZ ) N I'' 4-

C.
N C*- N a 4..~ .)-2m a - -a-u 44 U6...C
4 aa4N-- O 0-4Sc-v. --

9Z A N. OZ a .~a-n. CL

-~~W
a05 Na 0 0' a 4 . .. a 4 .. 4.Z ... i
a- - - - e" 4
en a w44 -,' J0, ifZ~
 A-2 5

f% kel %fIU l .lF

~S'l'
C'P-~PC~ ~ LP~. C~-~ oJ~ .C-~C ~N0lL ,~p.* .- e0 ' -'

0~AINp.
e l~t. N ? ~ ~rt.C Lt-,~r Jjc CC ~C IL Cr-#L"-P-

Wa

anv 1

a3 rC~.7c Zu

4 N4

-- N TI I

4 I I 1-JZ N ..

-- C-C-CLIz - x- 7 7

CL 0

-- 0
 A-2 6

M, ! C

1 -0=
LO

CL

ow-

I~ I.i 4~ -
*~Lick-

-I C -- *
 A-2 7

o~~~~~~~~ NA - r~E'
---. -- - -C-t- -I-

'S ~~ ~ ~ ~ ~ ~-~44. ~ ~ LU tJ
4-~LI ~ L~JJ1jL~~L

Mo C

L14 -- '

9-4
e M 0.C
-9
o 4 N mm*L U 4 4.

*
'I
~ N-. 4~
~ -~-
N
4y
a. a.
==C1
3 --
.
--
*
- zw zt U N-
0 ~~ ~ ~ ~ ~ ~' ~I4~.
fl Z. .4

- t.NC 0 - - 4!.-- *

U. 3 0N N N 0 ja. aU. 3 N * - .

a 1%) N ~fl'I
3 - .4 4 ~-~ Z - - -

CN-----'L----.U ~a.- ~ ~ -* ~ 0
-- 3 . N3 N3 0-.~ C - - -,,
 A-28

utok.itill %Ij-4 jA f..4 " 4"Jj#IJ.$P-.L4j 4fJ.WL

Itjj-j Oa.'mLjv j . j jta'j- -A- O
J-Jjrt . - -j -j -jj- t OA
o . Pi AIt-- .W
%j I $ M E 7.PIt4~,t,~~tI~tC ~ ~ -

~L&&
d&L-aa..
La~& LL &*. & A.O & O-& IL
t~*4JJ~l
I ~ JL$J~ .Lr4

IJJU~IJ~tWI.~LJ4
.II.4JI ~ JLJI MJLI ~ .J~j~M~&jj~ JU~uJL~~zLLJ4.I~.Jq
C"

If4
Au I i V

41 t

ck n
Ir
IC!
e c cU Cl
V.
f K '
I-1 olo a.

WIr,~ j CK r 'A -,l

ft 71-t I, .t 1

d-4 0 %D a 4&tu[ ILIL

,--- V%-- I - -,z

.4 ;= = O5 _- IJ Zj
- -WUU4-. a *.a -
ralm
-0,
--
"I~
Ooaa
;
I4
41
- *4
~ l'
Z
C .0 .h -V4
,b-
-?C ' X-- 4 c. " 'u"
* ~ ~ ~ ~~ - z 2 VII.U.=J *U- N ~
CO *w2
M * -K

(010~ 3 a. a.~.
0IJ~
4coo-01 40-0 new 1 %r 44 '"m -aC4Ua

ii 0 ~~c
0
~
. I
u 3
('0 £ 0 A
 A-29

AL
U
tto= %,LAIL,
- - .
tj~j~f-UIv- %Kuj u~4

JIuiJlJIJ.U
V.J

G.O

a0 fil VIa.'l.

a[ 'GA-C U. 0-0CI

w.
CL

Cal 00I
J V
 A- 30

Co o= O C l4..9CW W 4 --

333In

N~~ eq-l

-C E -- . - CIL

LIN 0-- A - Z .. iiJ -

002 1. .
3 4 )- * -4
O-c:m=;W;
=4 Z I -. :7 Z- - -
rmt. - el XZ. M i- ill i.
- *4 -J~ ZL - 51 - - *
.a k.ju w Z .'- J' m- -

-CL Z=9 -q- Z tv .

M.t-n -

0-1 C-i - 9L X N
:A- .a Z .. A5~' .- -
-- N N 0. yv.'---t Z 4..J ! .1 -.- = - - -
NA!e~
A!= 051, * "'e . . q Z a- .. .1 4 .

3m0Ucx -
-l
U! ~*45Nv ~ 0NV~ s ~~ ;,a:;<=a52
C
'
~ ~0 L~I4
U',
40

p-
a.-"44Z
Z
,1..3
3
~L-'*M,4 1 4

~5'..N4 ~ili. ~ - N 0.0 -24~.

a. . -~0-i i ~iw a. N~ - 44. C~i. It

 A- 31

o ~ ~ ~ N('. ~ ~ ~ ~ ~ ;P-wrr
7'?4M. r~~eN
I44is4 W4~ c4~~c
'r7. Wj~j 4,ri-C~i P-

Nw j- i.

UUUL q-jujC1t .JL .?j-q 4jjjkq LIL4

4UK LJU
k.A U ~J L4

P4 Ci u

N 4

* II

0 .co . a, M- o
-
-7 * ~06U C-
ta *
 A- 32

'N, f mf0"M m ww

AC

-Z z "

-C
;N 0 -
a:: A- e
-A~N, --
-, ,~ M' 4- -- -.

N (^Pf -4 .i S.

N .. . - N *
3 N
* N~ ..
16
X^4
=
~ 54.

Nl Ns I= 0 4-
WC s4
* 3 2 4 =.-a--.
AJ-L;= N 13 - - - 4--

r% .7 71 CL4. N ~ ~
Z~~~ N 4 CSS - Vqi - .q..

c UL:9-! CLr-
aN 6- IZL.)e 0:~

AJN Mg.g IA .4 Z I 3)( - ZI...5. -

A - ---
'%4 -.- -- - -. -. ' U
.- UAaJ3 ~ ~ .. C ~ - .3..... A Ca

A. NI4A-A . 2
 A- 33

o tlN -llj ll-. &

A 'UWU~.~UUU~UW n.wU ~. LUU 4.U LU .U 4.UUUU4UU~.W~.UU ~

* 4..

Aj*, -4

4 A -4 41 R

=C- - - -
- -* " A l ._>.

I C. 1 .1 - 4 - -. - ---

% a,- -- 1) 4 -
 A- 34

II
.z M

w' ~ '.~ a c~ .

aao

MA c

ofm-0 -... aU-=

'

49 - n - f . w

- -I-
 A- 35

IC

- oao cacoaccc
%~ ~~~~ -
-i-L-L- - - ---- - - - %lLLJt.,Lit4IiiLLLL

:
= ~~ ~ ~ ~~t zitJ'L~i ~ ~ ~~ Aj
i L~L iiLLiti2L iiiJi'L i) L iii

o= .Z uw *

NN Ni M. C6= M - e-t -w,

7.51. - 0. z, .J r-

&w .'.u . %- C 1
*L tc d6i - - 0.
* 30 -= tj -U._-: a

N-. 4 .
a ~ A- ~
N
*- -9M
-7
- - 0.
=
N A 7.

-=C nr~ - -~-. -4~-

NN lt.k~ V. r a. . a -

0 *N - .a .A c ' .. - U N7.1a

xN . Q64LKZ .

. - U 3 Adw N I. .C- - ' .~

 A- 36

.74

1% O~t~~CP
C.6,.7 ~r~ .?- 3.-f0 7gCp-1 OO~%I.CM
o Na~~a~NN .7,.7rlr 1~C.COU4M
WOAN

- C03033003 COCC 3033 00003330cc 00cooceo oc

~~.m
-.
a~.ML.
at of..~tI
~q.J4.JL4..JLLtJJI4~L
q.t.4.L.
i
z..U.U
~4~f64LI26 ~J~k~~L A.UU

4 4

b- t~J

CL- - - 7 -

; < 4

ZA .. ,wd x - L~qQ
" :; 101 - 1-:;

4,4 .a1L
 A- 37

'F

~C- ~
N

- coccca
T

~
- .g4444'7
- ULJL3LLJL7

0
'F

'a.
N 'E'
-'7

~ '1*
'S.
A
U. C
- -
'7

(.5z

n0

7.5 -
'7 ~
C 'A-

-S. C
~ 'F e- Li,
C. ~

'F ~f'

- 'F '~I
'F Z*~~
I~ 'F -
-2.-I

o '~

'F .t..i~-z
C
,j - 0
- ~. 0
'7 'F
'I
7.57-I

F
-

7'

I.
0

-
-

vi -~
 J A-
38

%A 14;L U-1 LA EUEUK CEs_ 4_& WLCA L) W f , %_j~fc

' L--t

acc

- I e

CC,

m m, I. " t -1 -0,

- - NN= " -=- -

*~-- 2 -

4 N 4" ;13p Ct .4 2'

Z -n - 4 C%
tA of-1 P .. U. -

m .
=Nm e . 4&A C.I.*Z-

I) Z, NN3
M~ 4- 74 t!7- -t,~- -
lIU W*=g I of
2 1 - 4%1. -a-W

r-.w ~ -= ~
" . 1 .. S-C E

s~~ - ~~
.N~~~ N. . 0. w0 L_~4=-. 44 --

Nl0U-ta
'.N4U~ 2 ~- ~ ;z.

.f~. N
- ~ ~ ~~
-AM04Ac.Z
L. ~- I
C ~ %j
4 ~ 1'-
~*N
J ~ ~C~5..CI
~ ~ - ~ -- 3/0 15 4,

- .. ~N..... . .AN .fl

=~ OX
-: 44 wz,%"Q I
.lm LJ .f-S
.x4-- zz M . N .J.-

j LV ~ ~ L4L LI~a.~L ~ I..JE..s... I

rO I4 4

L 6ANI.NlE-NE7.W -
 A- 39

"II

f '~~ ~..qJJI~. L ~.L~..~4 4JJ AjLL

~ - .. J...J
*.. J..JJ J.J.J.J
.J. .J .JA.J.J . .J... .. ~. ..... J .. j.J.....I.J cc .. ~J
.4 4444 4 444
444~d44 4 4 4444 444 444 444 4444 444

- qLAJ.
i~L~J ~.JLLL
L'LL' L3~t~JLL(~ L..J. .L)3LL. .X

=ZzNc

* -
U- Mw .0

0OW4 4. 10:d -0.o ;,

0 4. - "
4. - - J 1

4L0

ca4N4ac4 L
 A-40

°-
I'
1C
 A-4 1

IC ~ -ONJ"F4~N O 'F~7~,C8 C j~C F(.- C~..NF.*' c.= c

3
-- -3-3 - -- -- ~ 3

0 C.

J -1 I -tc ~ -

Ni ~ cla Ni N .j '- .

NJ- ~ ~ ~ A4^
-_j - -- 4

*~~~-. 6. a * CM~a

- - A - C - -0. J'I.) 4.

-~e 'Y" - - - -JNJ. -a- - -

a CSJ -
~
O.C4. ~~ i
a
-'
~Z .s
~ C
-
'4 N~
O..33
LC
~ Jaa
( '
~ -
m M.s

-- N N C a- Cadt 3 ck -jc --

''N z.i. ' W a' ' J F .. -

~
o ~ ~ ~ ~
F'a.> -4.S z - NIA
NN3~C. ~ A
~ N
F
--
4a..4
ZZ
a
-
s~~
-2 7

-NZ 0a ~ a C-=3 a-
 A-4 2

-- -- -- -- ---- 1 --- - - - - -

x- z

Ie 9L, 4 L

- I1

I I C
-t:;

AlI
 A- 43

ccoo
* it
00tr
L
my
-- ----- ~ -XXx m ------
i~

c
mt80.80, c cc
n~,iiJJi 00CL ItJ .J J J, .x.0 00
L"S%-C .~00,~000

o Sj

-ji :elli~i~IKj
oL CY 24 -

ml-

0 r4 4~ 5N
- - N

lin i .- ~

.w -. LlJ. w0 wb -Z-' -a. mi4

r= J4 f2 0h.~.
C64I 4~.. .

- ~-Wl
(%j
IJ41. & -S 4.80
66 21.
LZi
W
0Z6
*
-~
'45
IAM.- CMN f - 7

9- ~ ~
NS NN m t0L. OL-

8.
9L~ IC. N UJ7S - "j 24LU. -j2 ~ ~ - . N

NJ4"C 0~CU'
u E 79'4'4fr0 4-A - 6. ;:t 44~'.

CI 279...-.-- -- C

-P 'n C, -, ZZ.it2C3
-"C'
, ,
ZZ J8..
~.r2*- .IJ *4** 5= .. C
N.- =,A~02...

0'Ul9'~~~~~~~NIX
2 uLu Ti -.-- N2.2 aO40
o-
11,11 111 -- n ZC.... CP--- N

'7- e 0L SU.~J . ~ ~ ~ 2i
 A-4

-- - - -- -W a&' Mae~ aCO-N U. a. ft w or or m t&=C9-w=eW.

0 a. 6J~a~I' t~d~t .~'W t~rI~s- ~ a.~~a

17

Cp *. I~

I~r ~ z
IA -W. IMW.
-
;4 Z 6, -6- aC
-o ZOL
Z'-
*, , - -M ----

S -- c -z
.*., ,-

* 2r. 1 * 1.

a' ~ - 744* a ZamzZzw,7'-.

J1 01 .O a
 A-4 5

II
I In

-10

3k

'1 IL -. 1 *

e-I I 'I

a.L

*1- 7.
=( WU- m 6 .

-111 - .4
===UJ-1c

-
S- - z. .~ - = -
-A M4- 3 ..a . l
Z-0 Za- =- le Ia lo . --

*~~~l
-. =. Ia
O 4 00 .a -
"Lt -. 3 .~ (-I a
IN

-l-

U L OUu.W C8 ' E .C- t0j

.8.

-4j.

4 6- 4Aa [* -

0- in bA CA-C% C
-. 40
0. .0

_j z

r. MC.,

01 Q M- t- 1 w IM -

CL - J
4 42.P-
 A-4 7

U 0;eCIL-cc
.~jd~wv C .CD-f t ".1 ,.c~P-&C'.A F..r
a a QC COTC co-.
'O~ cc~c
o
C~C-OZO' ~
in'.cfr a a-%f.ra f
* ~
- ~ .~N%. N4fl #:ftp.ft tk
-==
kI tf 0
=Nr...,-
u% Nftf
t I.~&%ejt~rf
46%, .esjI."s- ft N4f,
j
.- mmIIni fl rr* w rI.g4
rMf^. Wjp4tfip 4 ~-e jc%fuf\i~%.

On0

39.4

an

ar -0 t0
-3c

ZV 7. 4 fl
m- -t-2 4.,a VI CS.(O I

Li L

.n M
 A-4 8

A~~~ 6%A O-O-

?-'C " .l C *- 0CN'** 0co- -'4 ... o

- coc cc-c cCccccc c cc a ccr--cccr-cc r-c~crcrc cc c c cc c~crcccc

c CCCcCCCcc CCCC CCCCCCCCCCCICle ctcCCCC-CCC

ZZ. 39- I

-0 - N 4. -jC

4. %w Cm.4-mc

-LU -- 0

oJ' -' -- t '-Zi r:3T

- ',r 5

=o- -i on 1 J- z

-4 ~a-
I. -Do
 A-49

W
.c
c-raC
7-rlrr
c
fiar-'W' ~lr - -Wl~f r cc N~vlfc
llllll l~lww4II

x
ce, xx xx x uxz XX X Z'~ a4~ 77.-. X7x
37... a' M:xr IC~r

-
&'
CCn C I
~&MOMCC LC n3C n c cc
=rCC cccac
CC CC C --'CC C ~CaC=.

- 1rZ.E

le w ,

- - U~b3.~a 3=ak&
- a.&&~
a~~a~~a~aa.&
ao~ ade~a a.a~a8.O~.C

44
 A-50

'.7

d~ .~NW
Ff'.~fl?- C-A.~4e
6OJ,.77 N~.7~~t T f~*.?* c-s o-e~.i.j .PC6

o 39

CL 7...

- wr

z (W.

aa

Qo- 0 --

3 4
-)C-

-~~r 4-. "4

 A-5 1

o ~ ~N ?nclIa7,~
,. cc-PIPt t 7 J

e% r

41"

ola

7:

It
 A-5 2

a-r,.C{.P. ~l.. jlC- ~ ' .

-
C L i ? . , ar e e. Nt N Ntw rr t ;wt
- -r.T

w* .

!
! -

-- .4.'-o 0.°,
 A-5 3

:,

*~~~~~~
0 00 0 0
0 0 .... . .0.0.0.0.0...

........
4... ... .* * ........... *.... .. O.

* ~P4f.0~IsJ~.7..........44..4-
NJAJJf~fJIJ~~s...-.....-..4.4

oca000000000000000000100 coo 000000000000000000

.............. 4 4... 44 4 * ** ** 4*

I U U WWU4U
JAJ"L W WULLUUJ WUW' 0 L L~~tUWAWUWJL uJAWW4A6JL WU WJUWLW4-.J LU W
LL, UN~ , m ~ ~ f? ,'

- . .... * **
......... * ** 4 4*4 *** ** *

*~~ ~~~~~~ E* . .N.. . . . .a .??-mI4~m~i

r
.~$
* r
 I i A-54

*... ........

*,
ujujWWL=
WW W Wu uJ9

.r. PaInAJ4IuN... W% NIJoa

a 4.

.. * . *------------
---

.................................................
,3M4omq I~~ . . . . . . . . .OO
>e,* 4 e~
21-u WaA WWWI) U300 W WW

. . . . . ....... . . . . . . .

a. ."a a .1 01 . ..

-. - -~-. -- ----
-*- -

1 00 0 0
 A-55

It%--4...-0000000O000mm00

.................................

mooooeo0000

W%%n~
+* * ** ...... .

WNtl

00000000000000000000

.A WWWJJU~nonl~ IU~UU.,UNW
n nl^I nr Mm l~l

... ..................

0000000-000e00"0000'

N
NM

.... a........... olf

I- M~W UuWWj~fA~~~~ o-UJm

CJNNNI "

ooaoooooooooacoooooo-
AD-AIOS 095 ARIZONAUN1V TUCSON ENGINEERING EXPERIMENTSTATION F/0 20/13
CONVECTIVEHEATTRANSFER FOR SHIP PROPULSION.
LUI
APR B1 N A HAIG, 0 N N4CELIGOT N00011?5-C0095
UNCLASSIFIED 124" -S

E
I fllllllffffff
2
 1362 12.2

I II' III
1111112

MICROCOPY RESOLUTION TEST CHART

NATONALRUtAU I1%
OF STANDARD, A

• . .... AMR
 A-5 6

~0e000000 orev0co 0000000000000000000

-= C- 00000'c") ,00 ~ 00

I %rcnc.cc0000000 0o0 ocl oaf

*000000000 0. 00000000000000000000b
I ~~IJWWSU&JU.MJUJ~UJW.&UJUWM
uA J 4LUWL hI4L4I J UJU.t&ujuiWW1juj4uj~&u A

* ~ dY .d.44.

i .
.. . . . . ... . . .. ... . . . . . . . .

I .~ Ylmm44',~I.~i .... 4m
m ,94~I..I m...............4mff~

I Coo I

.~.4"~4000004200-

ej.411J AA f14legf mm%"""Mm ftl41#"Mo

~~???~~4 00000000000000000000oo

4 4.O4.
**.** **+ +***.

AJE%~~JA~t*E~J
a~tNft
. 0000.000.. . .. 00 .. . . .
fiaj.A~a~4fAII1.wut a44~ .. .. .1... ..

44*44444*449444*40
*4444 0 O ~
0 0 00 0 0 0 0 0
04~4444444*44444 O0 0 0 0 0
II I
0 0 0 0 0 00%0
IIi a I
0 0 0
99gSI 9 9 9
 * r A-5 7
00000000000000000000c~ 00000000000000

cc000000000000000cc 000000(I~0 00

I 000000000000000000-0 000000000000000~0000
W,".ujuiWW
M ~~ ~
AWWUJLULALSWJWWW(
W1'UI ~ ~ UJUUIW&m. IUWJWWULWUJUJUJUUI%
Iv @not,'waw W'ter ~W %WflUI

--.---. 0004000004000000( O00 j

00C000O0000OOC0caoo000C~00

000000000t ~O000dI CC~0000O0000110000Q

&O.wWWw Uj~ujuujuujwijLjwWUISuw IL UJ&UMIwwwwUJSwwwwwUUUJUW

0 0 000000-00000111201-00
00000
0000C ~C000C p~oO 00000000000010000-000

0000000001000000 0 00~10o0o0 0000oo0oo Oooooooooooe

U..I ****4 *** 4 *. ........

CIO ccoooaooo' 00000000000 0000000000

.*..* C................*

. .............. r ............

.. . . . C
.... .

Wc000,0000000000000-4 OC00000000000000 o
00000C 001000000000,

.1 ... 4......4 . ...... ........

.00&'r 1
I ,000I000~0O00000 o~00000000 0000O00.0000M
 A-5 8

0 el *a I

Ii a* to 00 coc

* P** a 4
00

@
~~~~~~~
SM~ MW M S
*
I. w
*
0"

*I
a

M m
.0
ii

M
M SM 51m
4
4SM
A
S
I~

M
44
hi ~~~~
NMM
.....
----------- 0f~
* 4%.~ ~At
[ A-59

03 f

11 9' 0s ft ,~ -F ow .4 fl - ' % W"-7 10 fn n 0W An W4%I 9% 1% w irN'e

o, t. 4% !N - 'V
0% & :w -e -r @% NI bs ft s Cd - heS - -w t

I .
e In a 10 .0
o 0 0 0 0 0 0 a 00 a a N 0 No

=
L .. 1 U. SM
U W S& W ' %L
. I"M 6W
.

a I Z
S IL

f
l

#
W...LL SM MS

' I
LL. 1W ... J LL W W

0'j1
611 ef
0
~ ~on i.st0 ~ 0 ~ 0 ~m0 140 ~f"
~ la0 4 0"'
a
in
00
-
0
" f'fC'
to 0 10 0
t I"
10
C0
0C' i" C't
fA
a a
9 4 .*
'# '0 * + .4 # 4 4 4 * 4o * I+ 4 * I* 4
#4 * 4* .4 4
A" 6w 0"ft 4y ru
M Uf"C
U L" M n|SM. UI UIka " u Sm
p" L us, WSM
&u o m f"w w %w f-wi uud
wS SM f 4
f" As. 4 9
& 00 '
01 PI. 4n
a, F W T-% F %f"' 4y No 1" 4% '.. 4 1
l-P V. Cl Y" 4) 1£ 0 fm i f- 4 A. p. 6r C4 A£ cl F'0 : f" *.NY -

J:~ * . lty *. . . . . . . .4Ow * '4 .4 -. . . . . . .

NO 1
~~00'0 00 C0 00
I~0 10 100 10 I 0 0

- m 0. . 4 9- Q%III -4 4%l W 0- IV- 4 NY ,41 W . m0 WT

a o- C *, 0-v W% %. 0 P'- In. W"s P. ft 1.
"0 m c 0
s ^I In w t
a

F" 'A of n h Db Y r W%. I

'' 44 IN N I'm 44 I f
 I A-60

f.c .I ! .

ONf" f" f " f. o

- co0.0.
a0 a, 0 0 aa000- 0 00

~
$A~ IDa4 ~ 40 " 4 vi f-
.4 0.ONed VI ey ft fOOO ft
I

1 REFERENCES

Beltagui, S. A. and N. R. L. MacCallum, 1976. Aerodynamics of vane-

swirled flames in furnaces, J. Inst. Fuel, 49, p. 183.

Bradshaw, P., 1973. Effects of streamline curvature on turbulent

flows, AGARDograph No. 169.

Chieng, C. C. and B. E. Launder, 1980. On the calculation of turbulent

heat transport downstream from an abrupt pipe expansion, Numerical
Heat Transfer, 3, p. 189.

IEckert, E. R. G., E. M. Sparrow, R. J. Goldstein, C. J. Scott, W. E.

Ibele and E. Pfender, 1971. Heat transfer - A review of 1970
literature, Int. J. Heat Mass Transfer, 14, p. 1883.

Ede, A. J., C. E. Hislop and R. Morris, 1956. Effect on the local

heat transfer coefficients in a pipe of an abrupt disturbance
of the fluid flow: Abrupt convergence and divergence of diameter
ratio 2/1, Proc. Inst. Mech. Eng., London, 170, p. 1113.

Fletcher, L. S.,D. G. Briggs and R. H. Page, 1974. Heat transfer

in separated and reattached flows: an annotated review, Israel
Journal of Technology, 12, p. 236.

IGosman, A. D. and W. M. Pun, 1974. Lecture notes for course entitled

"Calculation of recirculating flows," Imperial College, Mech.
Eng. Dept. Report HTS/74/2.

I Habib, M. A., 1980. Confined flows with and without swirl, Ph.D.
Thesis, Imperial College, London.

Habib, M. A. and J. H. Whitelaw, 1980. Velocity characteristics of

a confined coaxial jet with and without swirl, ASME J. Fluids
imln. 102, p. 47.

Habib, M. A. and J. H. Whitelaw, 1981. Calculations of confined

coaxial-jet flows, Third Symposium on Turbulent Shear Flows,
University of Calif., Davis, Sept. 9-11, 1981.

Kays, W. M., 1966. Convective Heat and Mass Transfer, New York:
McGraw-Hill.

I
I R-1

~ ~*0w
!
R-2

Khalil, E. E. and J. H. Whitelaw, 1974. The calculation of local-

flow properties in two-dimensional furnaces, Imperial College,
Mech. Eng. Dept. Rpt. HTS/74/38.

Kline, S. J. and F. A. McClintock, 1953. Describing uncertainties

in single-sample experiments, Mech. Eng., 75, pp. 3-8.

Krall, K. M. and E. M. Sparrow, 1966. Turbulent heat transfer in

the separated, reattached and redevelopment regions of a circular
tube, ASME J. Heat Transfer, 88, p. 131.

Launder, B. E. and D. B. Spalding, 1974. The numerical computations

of turbulent flows, Computer Methods in Applied Mechanics, 3,
p. 269.

McEligot, D. M., P. M. Magee and G. Leppert, 1965. Effect of large

temperature gradients on convective heat transfer: the downstream
region, J. Heat Transfer, 87, pp. 67-76.

Morse, A. 1980. Axisymmetric turbulent shear flows with and without

swirl, Ph.D. Thesis, University of London, England.

Patankar, S. V. and D. B. Spalding, 1972. A calculation procedure

for heat, mass and momentum transfer in three-dimensional
parabolic flows, Int. J. Heat Mass Transfer, 15, p. 1787.

Pope, S. B. and J. H. Whitelaw, 1976. The calculation of near-wake

flows, J. Fluid Mech., 73, p. 9.

Rodi, W., 1979. Influence of buoyancy and rotation on equations for

the turbulent length scale, 2nd Symposium on Turbulent Shear Flows,
July 2-4, Imperial College, London.

Spalding, D. B., 1967. Heat transfer from turbulent separated flows,

J. Fluid Mech., 27, p. 97.

Zaherdadeh, N. H. and B. S. Jagdish, 1975. Heat transfer in decaying

swirl flow, Int. J..Heat Mass Transfer, 18, p. 941.

Zemanick, P. P. and R. S. Dougall, 1970. Local heat transfer down-

stream of abrupt circular channel expansion, ASME J. Heat
Transfer, 92, p. 53.
U WNW - -

DISTRIBUTION LIST

HEAT TRANSFER

One copy except

as noted

-AMr. M. Keith Ellingsworth 5

Materials and Mechanics Proqrams
-T Office of Naval Research
800 N. Quincy Street
Arlington, VA 22203
-Defense Documentation Center 12
Building 5, Cameron Station
-. Alexandria, VA 22314
Technical Information Division 6
- Naval Research Laboratory
4555 Overlook Avenue SW
- Washington, DC 20375
Professor Paul Marto
Department of Mechanical Engineering
US Naval Post Graduate School
I Monterey, CA 93940
Professor Bruce Rankin
- Naval.-Systems Engineering
I US Naval Academy
Annapolis, MD 21402
I Office of Naval Research Eastern/
I _ Central Regional Office
Bldg 114, Section D
I -- 666 Sumter Street
I Boston, Massachusetts 02210
t - Office of Naval Research Branch Office
1 536 South Clark Street
Chicago, 111. 60605

I Office of Naval Research

Western Regional Office
1030 East Green Street
I Pasadena, CA 91106
Mr. Charles Miller, Code 05R13
Crystal Plaza #6
Naval Sea Systems Commnand
Washington DC 20362
I D-1
 D-2
Steam Generators Branch, Code 5222
National Center 14
Naval Sea Systems Command
Washington, OC 20362

Heat Exchanger Branch, Code 5223

- National
Naval Sea Center
Systems#3Command
Washington, DC 20362

- Mr. Ed Ruggiero, NAVSEA 08

National Center #2
- Washington, OC 20362

" Dr. Earl Quandt Jr., Code 272

. David Taylor Ship R&D Center
Annapolis, MD 21402

Mr. Wayne Adamson, Code 2722

David Taylor Ship R&D Center
- Annapolis, MD 21402

" Dr. Win Aung

Heat Transfer Program
National Science Foundation
- Washington, DC 20550
- Mr. Michael Perlsweig
- Department of Energy
Mail Station E-178
- Washington, OC 20545

Dr. W.H. Thielbahr

- Chief, Energy Conservation Branch
Dept. of Energy, Idaho Operations Office
- 550 Second Street
Idaho Falls, Idaho 83401
. Professor Ephriam M. Sparrow
Department of Mechanical Engineering
- University of Minnesota
Minneapolis, Minnesota 55455
- Professor J.A.C. Humphrey
Department of Mechanical Engineering
- University of California, Berkeley
Berkeley, California 94720
Professor Brian Launder
Thermodynamics and Fluid Mechanics Division
- University of Manchester
Institute of Science & Technology
- P088 Sackville Stteet
Manchester M601QD England
IProfessor Shi-Chune Yao
Department of Mechanical Engineering
j3S Carnegie-Mellon University
Pittsburgh, PA 15213
 Professor Charles B. Watkins D-3
Chairman, Mechanical Engineering Department
Howard University
Washington, OC 20059
"
-Professor Adrian Bejan
- Department of Mechanical Engineering
University of Colorado
Boulder, Colorado 80309
Professor Donald M. McEligot
I - Department of Aerospace and Mechanical Engineering
Engineering Experiment Station
University of Arizona 85721
I Professor Paul A. Libby
Department of Applied Mechanics and Engineering Sciences
University of California San Diego
I -Post Office Box 109
La Jolla, CA 92037
-Professor C. Forbes Dewey Jr.
Fluid Mechanics.Laboratory
Massachusetts Institute of Technology
-Ca Cambridge, Massachusetts 02139
- Professor William G. Characklis
m _ Dept. of Civil Engineering and Engineering Mechanics
Montana State University
- Bozeman, Montana 59717
Professor Ralph Webb
- Department of Mechanical Engineering
Pennsylvania State University
S-
_
208 Mechanical Engineering Bldg.
University Park, PA 16802
S- Professor Warren Rohsenow
Mechanical Engineering Department
- Massachusetts Institute of Technology
_
-77 Massachusetts Avenue
Cambridge, Massachusetts 02139
l_ Professor A. Louis London
Mechanical Engineering Department
Bldg. 500, Room SOIB
Stanford University
- Stanford, CA 94305
Professor James G. Knudsen
I- Associate Dean, School of Engineering
Oregon State University
" 219 Covell Hall
Corvallis, Oregon 97331
 D-4
Professor Arthur E. Bergles
Mechanical Engineering Department
Iowa State University
Ames, Iowa 50011

I Professor Kenneth J. Bell

School of Chemical Engineering
Oklahoma State University
Stillwater, Oklahoma 74074

* -Dr. James Lorenz

Component Technology Division
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois 60439

Dr. David M. Eissenberg

j - Oak Ridge National Laboratory
_ P.O. Box Y, Bldg. 9204-1, MS-O
Oak Ridge, Tennessee 37830

I -Dr. Jerry Taborek

Technical Director
- _Heat Transfer Research Institute
1000 South Fremont Avenue
- Alhambra, CA 91802
J- Dr. Simion Kuo
- Chief, Energy Systems
Energy Research Laboratory
- United Technology Research Center
East Hartford, Connecticut 06108

i - Mr. Jack YampoIsky

a
General 81608 Company
P.O. Box Atomic
I _ S San Diego, CA 92138

- Mr. Ted Carnavos

I- Noranda Metal Industries, Inc.
Prospect Drive
Newtown, Connecticut 06470

Im- Dr. Ramesh K. Shah

Harrison Radiator Division
General Motors Corporation
- Lockport, New York 14094
* Dr. Ravi K. Sakhuja
_-Manager, Advanced Programs
Thermo Electron Corporation
- 101 First Avenue
Waltham, Massachusetts 02154

i -.. .

... ,.*".m - • . . . . . . .. . . .. , .. . . . . .. ±. . .. . .. -. .' .

 D-5
Mr. Robert W. Perkins
Turbotec Products, Inc.
533 Downey Drive
New Britain, Connecticut 06051

Dr. Keith E. Starner

York Division, Borg-Warner Corp.
P.O. Box 1592
York, PA 17405
Mr. Peter Wishart
C-E Power Systems
Combustion Engineering, Inc.
Windsor, Connecticut 06095
Mr. Henry W. Braum
Manager, Condenser Engineering Department
Delaval
Front Street
Florence, New Jersey 08518

Or. Thomas Rabas

Steam Turbine-Generator Technical Operations Division
Westinghouse Electric Corporation
- Lester Branch
P.O. Box 9175 N2
Philadelphia, PA 19113

Dr. Al Wood
Director, Mechanics Program
Office of 'laval Research
800 '. Quincy Street
Arlington, VA 22203

- Mr. Walter Ritz

Code 033C
1laval Ships Systems Engineering Station
Philadelphia, Pa 19112
Mr. Richard F. Wyvill
Code 5232
NC #d
'laval Sea Systems Command
Washington, DC 20362

Mr. Doug Marron

Code 5231
.11C #4
Naval Sea Systems Command
Washington, DC 20362

S. *
 D-6

Mr. T. M. Herder
Bldg. 46462
General Electric
1100 Western Co.
Avenue
Lynn, MA 01910

Mr. Ed Strain
AlResearch of Arizona
Dept. 76, Mail Stop 301-2
P. O. Box 5217
Phoenix, AZ 85010
(Tel. 602-267-2797)

Mr. Norm Mclntire

Solar Turbines International
2200 Pacific Highway
San Diego, CA 92101
Prof. Daryl Metzger
Chairman, Mechanical and Energy
Systems Engineering
Arizona State University
Tempe, AZ 85281