Louisiana State University

LSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1953

The Effect of Composition on Heat Transfer

Coefficients for Condensing Binary Vapors With a
Single Liquid Phase.
James Berwick Todd
Louisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

Part of the Chemical Engineering Commons

Recommended Citation
Todd, James Berwick, "The Effect of Composition on Heat Transfer Coefficients for Condensing Binary Vapors With a Single Liquid
Phase." (1953). LSU Historical Dissertations and Theses. 8038.
https://digitalcommons.lsu.edu/gradschool_disstheses/8038

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in
LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact
gradetd@lsu.edu.
T H E E F F E C T OF C O M P O S ITIO N ON H E A T TR A N S FE R
C O E F F IC IE N T S FO R CONDENSING B IN A R Y VAPORS
W IT H A SIN G LE L IQ U ID PHASE

A D issertatio n

Submitted to the Graduate Faculty of the

Louisiana State U n iv ersity and
A g ric u ltu ra l and M echanical College
in p a rtia l fu lfillm e n t of the
requirem ents fo r the degree of
D octor of Philosophy

in

The D epartm ent of C hem ical Engineering

by
James B . Todd, J r.
B .S ., Louisiana State U n iversity, 1948
M .S ., Louisiana State U n iversity, 1950
June, 1953
 UMI Number: DP69416

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

Dissertation Publishing

UMI DP69416
Published by ProQuest LLC (2015). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code

ProQuest
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346
 MANUSCRIPT THESES

Unpublished theses submitted for the master’s and doctor’s

degrees and deposited in the Louisiana State University Library

are available for inspection* Use of any thesis is limited by the

rights of the author. Bibliographical references may be noted, but

passages may not be copied unless the author has given permission*

Credit must be given in subsequent written or published work*

A library which borrows this thesis for use by its clientele

is expected to make sure that the borrower is aware of the above

restrictions.

LOUISIANA STATE UNIVERSITY LIBRARY

 : -0
A C K N O W LED G M EN TS

The author is v e ry deeply indebted to D r , Jesse Coates,

d ire c to r of this p ro je c t, fo r his advice and suggestions*

G rate fu l acknowledgment is also made:

To D r. P . M . H orton fo r his many helpful suggestions.

T o M r . £ . E* Snyder fo r building the equipment and

hie advice on ce rtain m echanical features

T o his fellow students fo r th e ir constructive c ritic is m

z 7/ . '/£

L .9 3 o d u
/< ? S 3
 T A B L E O F C O N TE N TS

CHAPTER PAGE

I. IN T R O D U C T IO N 1

It . T H E O R E T IC A L CONSIDERATIONS 5

A. The Nugselt Equation 5

B. Other Considerations 19

H I. PR EVIO U S E X P E R IM E N T A L WORK 31

IV . A P P A R A TU S 34

V. E X P E R IM E N T A L W ORK 55

V I. R ESULTS 61

V II. CONCLUSIONS 105

Vm. R E C O M M E N D A TIO N S 107

S E L E C T E D B IB L IO G R A P H Y 109

NOM ENCLATURE 112

C A L C U L A T IO N OF C O N FID E N C E L IM IT S 115

A U TO B IO G R A P H Y 122

ill
 L IS T OF T A B L E S

TABLE PAGE

X* F IL M C O E F F IC IE N T S F O E CONDENSING VAPORS
(Single Components) 63

XL F I L M C O E F F IC IE N T S F O E B IN A R Y VAPORS 6?

m. F IL M C O E F F IC IE N T S A T A CONSTANT T E M P E R A ­
T U R E DROP SO

iv
 U S T OF FIG U R ES

1 T e m p e ra tu re and Concentration Gradients fo r

a Condensing M ixed V apor 12

2 T h e rm a l Conductivity V ersus Composition 23

3 D ensity versus Composition 25

4 V isc o sity versus Composition 26

A 3 p i. 1/4
5 The Group ( —. ) versus Composition 27

6 F ilm Coefficients of H eat T ra n s fe r fo r Conden~

sation of E th an o l-W ater Vapors at Constant
T e m p e ra tu re D ifferen ce Based on the B oiling Pt.
(W allace & Davison) versus Composition 30

7 Therm ocouple In stallatio n 36

8 The B o ile r 38

9 Cooling W ater System 40

10 The Vapor Jacket 42

II V iew s of the Apparatus 49

12 A V ie w of the Apparatus B efore Insulating 50

13 F ilm Coefficients of Heat T ra n s fe r of Single

Components versus T e m p eratu re D rop A cross
the F ilm 66

14 F ilm C oefficient of Heat T ra n s fe r at Constant

Com position versus T e m p eratu re Drop A cross
the F ilm , 17. 2 M o l % Acetone in Methanol

15 F ilm Coefficients of Heat T ra n s fe r at Constant

Composition versus T e m p eratu re Drop A cross
the F ilm , 42, 3 M o l % Acetone in Methanol

v
16 F ilm C oefficients of Heat T ra n s fe r a t Constant
Com position versus T e m p e ra tu re D rop A cross
the F ilm * 5 7 .3 M ol % Acetone in M ethanol 89

17 F ilm C oefficients of H ea t T ra n s fe r at Constant

Com position versus T e m p eratu re D rop A cross
the F ilm # 80 M o l % A ceto ne in M e th a n o l
(Azeotrope) 90

18 F ilm C oefficients of H eat T ra n s fe r at Constant

Com position versus T e m p eratu re Drop A cross
the F ilm ! 4 6 .5 M o l % M ethanol in Benzene 91

19 F ilm Coefficients of H eat T ra n s fe r a t Constant

Com position versus T e m p e ra tu re Drop A cross
the F ilm , 6 1 .2 M o l %, M ethanol in Benzene
(A zeotrope) 92

20 F ilm Coefficients of H eat T ra n s fe r a t Constant

Com position versus T e m p eratu re Drop A cross
the F ilm , 83 M o l % Methanol in Benzene 93

21 Ring Appearances in F ilm of Condensate at low

concentration of Methanol in Benzene and Inter**
m ediate Rates of Condensation 95

22 Ir r e g u la r P a tte rn of Condensate a t Low Concen-

tra tio n of M ethanol in Benzene at Low Concen­
tra tio n Rates 96

23 F ilm Coefficients of Heat T ra n s fe r at Constant

T e m p e ra tu re Drop versus Com position (A cetone-
M ethanol System) 97

24 F ilm Coefficients of H eat T ra n s fe r at Constant

T e m p e ra tu re Drop versus Composition (Methanol**
Benzene System) 99

25 F ilm Coefficients of H eat T ra n s fe r a t Constant

T e m p e ra tu re Drop versus Com position (M ethanol -
Isopropanol System) 100

26 F ilm Coefficients of H eat T ra n s fe r at Constant

T e m p e ra tu re Drop Based on the Boiling Point
versus W eight P e r Cent 102

vi
27 Enthalpy-C oncentr ation D ia g ra m (Acetone -
M ethanol) 117

28 Enthalpy ~€ one ent ra t ion D iag ra m (M e th a n o l-

Benzene) 118

29 B oiling P oin t D iag ra m (A cetone-M ethanol) 119

30 B oiling Point D iag ra m (M ethanol-B enzene) 120

31 B oiling P oin t D iag ra m (M ethanol-Isopropanol) 121

v ii
 'ABSTRACT

M uch w o rk has been done on the determ ination of film c o effi­

cients of heat tra n s fe r fo r pur© components, but fo r b in ary systems

little w ork has been done. D ata fo r the ethanol*w ater system only,

and that over a n a rro w range of composition, w ere found in a search

of the lite ra tu r e . C onsiderable disagreem ent between these e x p e ri-

m ental resu lts and predicted values fro m the N usselt Equation was

noted.

A study of the th eo retical background of the problem indicates

that the N usselt Equation should apply in the prediction of film co effi­

cients of heat tra n s fe r fo r b in ary vapors that fo rm a single liquid

phase if the operating conditions approxim ate those fo r which the

equation was developed, and if special care is taken in evaluating

the te m p eratu re drop across the film of condensate. H ow ever, the

p ra c tic a l use of the equation is n u llifie d by the lack of necessary

physical data fo r solutions. T h is fact emphasizes the d e s ira b ility

of being able to p red ict the value of the film coefficient fo r the con­

densation of b in ary vapors fro m a knowledge of the values fo r the

single components, which a re ea sily obtainable. A study of fr a g ­

m en tary data fro m the lite ra tu re indicates that it m ay be possible to

p re d ic t coefficients fo r id eal solutions fro m the sim ple weighted

v iii
average of the p ro p erties of the single components but not necessarily

fo r no n-ideal solutions*

A n apparatus previou sly designed by the author was adapted

to the conditions m et in m easurem ent of film coefficients of heat

tra n s fe r of b in ary solutions. The noteworthy features of this apparatus

a re the a b ility to m aintain iso th erm al conditions the en tire length of

the condensing surface and the technique of mounting the surface

therm ocouples to give c o rre c t tem p eratu re readings* The apparatus

was designed in such a fashion as to allow a stream of vapor of con­

stant composition to be fed to the condensing surface. Other design

featu res insured the maintenance of steady state conditions.

Data fo r th re e b in ary systems a re included here as w ell as

th at fo r the single components: the acetone-m ethanol and benzene-

methanol data collected by the author and the iso propanol-m ethanol

data collected subsequently by Trachtenberg on the same apparatus.

The resu lts indicate that a w orkable apparatus has been de­

signed which w ill give consistent re s u lts . F o r bin ary systems with

a single liquid phase that approach id eality, the experim ental results

indicate that the film coefficient fo r condensation of the vapor is a

lin e a r function of the composition at a constant tem p eratu re drop

across the film * F o r b in ary systems with a single liquid phase which

do not approach id eality, no such simple relationship is apparent.

Under the conditions which N useelt's Equation is applicable, it appears

ix
that the c o rre c t tem p eratu re drop is the difference between the boiling

point of the condensate and the surface te m p eratu re .

x
 CHAPTER X

IN T R O D U C T IO N

The p ro p er design of heat tra n s fe r equipment necessitates

the accurate evaluation of the o v e ra ll coefficient of heat tra n s fe r fo r

use in Newton1s Equation*

q s U A At (1)

q * rate of heat tra n s fe r, B T U /H r.

U 5 o v e ra ll coefficient of heat tra n s fe r,
B T U /h r ,-ft.2 -°F .
A = a re a of heat tra n s fe r, ft. ^
A t * tem p eratu re difference, ° F

The conventional approach to the evaluation of the o v e ra ll heat tra n s -

fe r coefficient is to m ake use of the resistance co n cep t--th at the total

resistan ce to the flow o f heat is the sum of the various resistances in

s e rie s .

In the case o f tra n s fe r of heat fro m a vapor to a cooling

m edium through a containing w all w herein the vapor is condensed,

the to tal resistance, Rfc, is made up of the resistance of the film of

condensate, R ^ the resistance of any d irt o r scale covering the

vapor side of the containing w a ll, R^v, the resistance of the con­

taining w a ll, R w, the resistance of any d irt o r scale on the opposite

of the containing w all, R ^ , and the resistance of a film of the cool­

ing m edium , Rj^*

• RV + Rdv + Rw * a dL + RL (2)
1
 2

Now the to tal resistance is defined as the re c ip ro c a l of the o v e ra ll

co efficient and the individual resistances as the recip ro ca l of the in ­

dividual coefficients* In the specific case of the condensation of a

vapor on the outside o f a pipe, the o v e ra ll coefficient of heat tra n s fe r

m ay be represented by the following equation w herein the te rm s fo r

d iam e ter rep resent the a re a corrections:

J L i JL * J L + L U D 0 <3)
*>0 hv **dv k Davg ^ d L 1^

Of these coefficients, perhaps the most uncertainty exists with

respect to the evaluation of the individual coefficient of the condensing

vapor. The coefficients fo r the d irt and scale form ations depend on

the tim e elapsed since cleaning the surface and the nature o f the

vapor and the cooling m edium and must be determ ined ex perim en tally

o r estim ated fro m experience* The resistance of the w a ll is ea sily

determ ined since its thickness and th erm a l conductivity in most cases

is re a d ily available. The coefficient fo r the cooling m edium m ay be

d eterm ined fro m sem i ^e m p irical relationships which have been sub-
( 12)
stantiated by many investigators.

A th e o re tic a l treatm ent of the problem of predicting the value

of the film coefficient fo r the condensation of vapor on a colder s u r-

(13)
face was published by Nusselt in 1916. This treatm ent is

dependent on several assumptions which w ill be discussed in a la te r

section. This treatm ent applies only fo r a condensate form ing a

 3

continuous, homogeneous film on the condensing surface. F o r the

cases of Mtwo phase" and "d ro p w is e ” condensation, no com pletely

th e o re tic a l relationships have been found.

O perating under the conditions assumed by Nusseit in the

development of his equation, m any investigators have found good

agreem ent between the predicted values and the experim ental values

of film coefficients fo r pure compounds. In m any cases, the m ag­

nitude of the predicted and the experim ental coefficients w ere some­

what d ifferen t; but in m ost cases the relatio n of physical properties

as p red icted by Nusseit has been substantiated. Several investigators

have studied the condensation of b in ary vapors form ing a two-phase

condensate, but no general theory has been proposed. It has been

said that the Nusseit Equation should apply in the condensation of

b in a ry vapors form ing a single liq u id p h a s e ^ ^ .

A search of the lite ra tu re revealed that only one system

form ing a single liq u id phase has been studied previous to this

p ro ject and that fo r a lim ite d range of compositions, ^ ^ )(9 ) -phis

m eagre data shows considerable deviation between the experim ental

values and th eo retical e x p e c t a t i o n s ^ T h e present w o rk was under­

taken to co llect additional data on other systems in the hope of aiding

in the establishm ent of some general system of behavior. As a

n ecessary adjunct, a previously designed apparatus was adapted to

the conditions encountered in dealing with b in ary vapors and was

tested to determ ine its re lia b ility .
 C H A P T E R IX

T H E O R E T IC A L CO NSIDERATIONS

The N usseit Equation

The conditions encountered in the study of film coefficients

of heat tra n s fe r fo r b ilia ry vapors condensing to fo rm a single liquid

phase a re in m any ways s im ila r to those m et in a lik e study of pure

component vapor condensation. Since an excellent analysis and solu­

tion of the problem fo r single components has been achieved by

N usseit, it is necessary that his basic w ork be included here.

Nusseit found it necessary to impose the following condi­

tions in o rd e r to solve the problem of calculating a film coefficient

o f heat tra n s fe r fo r the condensation of a vapor on a colder surface.

1) The tem p eratu re of the liquid must be the same as the w all

te m p eratu re at the w all surface and the tem p eratu re at the surface

in contact with the vapor the same as the vapor tem p eratu re. 2) The

velocity o f the film at the w all surface must be eero and the film

m ust cling to the m e ta llic w all; and the surface of the d im in contact

w ith the vapor m ust be under the same pressure as the vapor* 3) The

heat passing through the film is that of the latent heat of condensation.

4) The velo city of the vapor p a ra lle l to the surface of condensation

is less than one m e te r p e r second. 5) The cu rvatu re of the film and

5
the velo city component perpendicular to the w a ll can be neglected.

6) The tem p eratu re d istrib u tio n through the film can be represented

by a straight line relationship. 7) The tem perature drop across the film

is constant. 8) The physical constants to be used in the equation may

be evaluated at average film te m p eratu re .

L e t u be the downward velocity component of the film a distance

a fro m the w a ll a n d ^ ^ th e coefficient of viscosity. W ith a velocity

change du/da at the position a, there resu lts a shearing

stress S;

( 1)

A t a distance da fro m this position*

<21

is the change in shearing stress. T h is shearing stress change is

equal to the downward component of the liquid due to the force of

g rav ity. If the surface is at an angle with the horizontal plane,

and if designates the weight per unit volume of condensate, then

(3)

or

<4)

Upon integration
If the liq uid clings to the w a ll, then a » G, u = 0, and € 2 = 0* then

it - n a - P S/At ( P gZ
a ‘ c 'a - T ^ r w

The value C j depends on the flow conditions of the vapor, fo r if the

vapor s trik es the film ’ s shearing stress w ill be set up on the film .

Since the velo city of the vapor with respect to the film has been

assumed so sm all as to be negligible, this stress was neglected.

If aQ is the thickness of the film at a spot and correspondingly

du/da = 0, then

t * /° S lN 0? £
(7)
* M ?

and

a - $ Bpa. - /0^~. ^ 3 2 (8)

. (z * .a -a ‘) (9)

ums ~ a [(L & M z fd a . (io>

s 0 J0 KM ?

u„ - £ & £ /& * . £ ) , ? S „ tp ,
* -* ■ , I t -5^ - a * ,u l

In a tube length of unity and an arc elem ent d (p

if,
 8

Based on the equations

(13)

w ■■ ( iv tJ - x ) (14,
A- cLo

where q = rate of heat tra n s fe r

k « th e rm a l conductivity

t v a t e m perature of the vapor side of the film

tw * tem p eratu re of the surface side of the film

w = the mass rate of tra n s fe r

JL s the latent heat of vaporization

F ro m Equation (11)

<15)

and if 3 « 3^ f f * * -,£ * )

F ro m (12) and (15)

£ c L (p = 8.0 eL(3-J S in Ip <1 3 . V 3* <Los(?<J$

(16)
X *
Now le t 2, ( - t y 1* - •g ' (1^)

Then fro m (16)

5/4 S \N ($ — * + CoS - O
cL
Equation (17) is a lin e a r d iffe re n tia l equation whose solution is

Z , S s Z * 5 p L z f $ ,N i~ ^ ) (19)

I f no liquid drops on top of the tube, when » 0, the film thickness

 9

ac is fin ite and hence z j is fin ite . Then the constant C must be zero,

and

z ' m ( if , S n/ (20)

G raph ical integ ration of the in te g ra l $ between

lim its ( ( m 0 and gives the value 3 .4 2 $ .

(Recent evaluation ^ * Vby gamma functions has shown the value to be

3 .4 4 9 5 - 0 .0 0 0 2 , but this m akes only a slight difference in the r e ­

sulting equation giving a constant of 0. 728 as compared to Nusseit*s

0. 7 2 5 .)

F ro m (17)
a.» ( 21)

The lo cal coefficient at angle is

k -. -- J jjL - (221
* 3.0 fry - jf « ) (22)

B y integrating fro m (Qt to the mean coefficient is obtained.

(23)
L W V * w c & F v iJ ~ « r
\ r
Substituting the value fo r \P fro m (21) and integrating graphically

P T A. H i* <i-
krn ' \ / - * X (24)
' sUf 7> ( i y - i y , )

The physical prop erties are the average properties fo r the film and

should be evaluated at the film tem p eratu re, t|.

if* - 34 (25)
as shown by D re w , ty is the saturated vapor tem perature and A t

is the tem p eratu re drop across the film . It is to be noted that this

equation takes into consideration only the heat of condensation, ig n o r­

ing that heat which may be given up by the cooling of the film of

condensate or any superheat of the vapor. A co rrection for this fault

m ay be made, but in most cases the correction is of such sm all

magnitude as to be negligible. Another point to be kept in mind is

that the film of condensate m ust flow downward in stream line motion

if the equation is to be s tric tly applicable. The above equations can

ur
be put into dim ensionless fo rm including the Reynolds Num ber ( - j r )

where P * ; w being one half the mass rate of flow and L the

length of the horizontal tube where the equivalent d iam eter is equal

to four tim es the film thickness. This equation is

Each of the conditions imposed by Nusseit in his derivation

m ust now be studied w ith reference to a b inary system to see whether

his equation can be expected to hold true fo r calculations of film

coefficients of heat tra n s fe r fo r b in ary systems form ing a single

liquid phase. Condition Num ber 1 ‘ that the tem perature of the liquid

must be the same as the w all tem perature at the w all surface and the

tem p eratu re at the surface in contact w ith the vapor the same as the

vapor tem p eratu re" req u ires ca re fu l study. This assumption was
 11

made sim ply to obtain the te m p eratu re drop across the film - The

d erivation does not depend on the identity of the vapor side te m p era­

tu re of the film w ith the vapor tem p eratu re . Th is assumption m e re ly

served to establish the vaporside film tem p eratu re easily fo r pure

compounds. Hence the assumption m ay be m odified to ’’the te m p era­

tu re drop across the film m ust be d e t e r m in a b le .C e r t a in ly the

te m p eratu re on the w a ll side of the film is very near that of the outer

surface of the w a ll. H ow ever, the evaluation of the tem perature on

the w a ll side of the film is much m ore com plicated in the case of
(41
b inary vapors* Colburn and B re w ' have studied the problem, by

re la tin g mass and heat tra n s fe r. T h e ir w ork is sum m arized in the

following paragraphs w ith the nom enclature changed to the generally

accepted symbols as appear in m ore recent lite ra tu re . The

theory in fe rs that between the interface and the m ain vapor stream

th ere exists a la y e r of vapor that is stagnant or in stream line flow .

(F ig . No. 1). A t points w ithin the film or lay er of vapor, the m olal

ra te of tra n s fe r of the m o re vo latile component toward the interface

in the case of equim olai counter diffusion is given per unit a re a as

Na * -D .,dcA . -D v dp a (27)
ds R T da
c 3 concentration, (m o ls /ft.
Dv = m olecular d iffusivity in gas (ft. /h r)
= ra te of diffusion (m o ls /h r)
p * p a rtia l pressure of m ore vo latile component (atm )
R 2 Gas Constant (ft. ^ )(a tm )/(lb m o l)(°R )
T * Absolute tem p eratu re, (°R )
z s distance in d irectio n of diffusion, ft
 12

FIGURE I
TEMPERATURE AND CONCENTRATION
GRADIENTS FOR A CONDENSING
MIXED VAPOR

CD
 13

Equation (27) m ay be integrated to give

Nn *
(28)
•R-r*

Th is equation expresses the resistance to mass tra n s fe r due to

m olecular diffusion alone* H ow ever, another fa c to r, eddy diffusivity,

is involved which is also proportional to the p a rtia l pressure gradient.

Since the effect of m olecular diffusion and eddy diffusion cannot be

separated at the present tim e , a coefficient lumping them together is

commonly used as defined by the following equation*

n a : kg<Pi-P8> (29)

where is called the m ass tra n s fe r coefficient. In the case where

eddy d iffu s ivity is v e ry la rg e as compared to the m olecular d iffusivity

then kg_ ie seen to approach the value D„ / ftfa..

In the general case of condensation of b inary vapors the mass

tra n s fe r to the in terface between vapor and liquid is brought about by

diffusion through the "so called" stagnant film and bodily tran sp o rt due

to condensation. The above m ay be expressed m ath em atically by the

approxim ate equation

(30)
NA * ”k* * 2 ? + <Na + NB )yA
W here the f ir s t te rm on the rig h t represents the ra te of diffusion

and the second the bodily tran sp o rt. The fact that th e rm a l diffusion

is neglected as w e ll as the effect of component B on the mass diffusion

is a lim ita tio n of the above equation.

 14

Equation (30) can be integrated between the lim its a • O

and a- B , y>jf • J*\i and * Jzi to give

{H n + * b ) ■- \ f In Z ^ S L l^ i | (3i)
' /V*f A/® /

A t the eondeneate vapor in terfac e, and N g m oles a re at a te m p e r­

atu re tj^ Assum ing that the enthalpies of the pure components in both

the liq uid and vapor phases a re additive, the latent heat of the m ixtu re

} k N» < ^ V ( \e - kj t J
(32)

where ^ is the heat erf condensation and h^ is the enthalpy of the

saturated vapor and hj is the enthalpy of saturated liq uid . The

m o ia l enthalpies a re evaluated at Colburn and D rew neglected

the heat of m ixing fo r non-ideal solutions to sim plify the equation.

If it is fu rth e r approxim ated fo r sim plication that the m o lal latent

heats of the components a re equal then q ^ * (N a 4 N b ) A.m where

is the m ean m o la l latent heat of the components.

The heat flu x which a r r iv e s at the in terface as sensible heat

is the re s u lt of the cooling of the m ain stream of vapor and the cooling

of the 4 N3 moles per hour which a re condensing fro m the

tem p eratu re of the vapor stream to the tem p eratu re at the in terfac e.

Th e sum of these two is expressed in te rm s of the heat tra n s fe r co­

efficien t h. Since the flow of m atte r across the film w ill convey heat

bodily, it m ust be considered in a rriv in g at an equation to express the

 15

amount of sensible heat a rriv in g a t the in terface* T h e re fo re

P ' ' kx6 ^ f { ^ Cr « + C? * ) ( * ' t i )

In the above equation it is assumed that no condensation takes place

w ithin the f ilm . The f ir s t te rm fakes info consideration the heat

flux across the vapor la y e r a ris in g fro m the lo cal tem p eratu re

gradient* The second te r m in the equation takes into consideration

the heat c a rrie d by m ass tra n s fe r. Of the quantities expressed in

Equation (35) above, only £ v a rie s w ith £ and integration results

in

(34)

w here C0 - 1 |K C N0 C ^ and CpB a re m o lal heat

capacities.

F r o m Equations (31), (3 2), and (36)

N*+N&

w here ^ is the to ta l heat effect.

The values h and kg m ay be evaluated fro m the following

relations

(36)

(37)
 16

w here G and G _ a re the m ass ra te of flow and m o lal rate of flow

m

re s p e c tiv e ly and f, the dim ensionless fric tio n fa c to r. The j factors

a re found to ap proxim ate f /2 e x p e rim e n ta lly fo r c e rta in sim ple

s tre a m lin e shapes,

Johnston and F ig fo rd ^ ^ re p o rt the follow ing equation to

give excellen t resu lts when w orking w ith b in a ry m ixtu res in a wetted

w a ll colum n*

JD S 0. 033 Re ”° * 23 (38)

Since the m o la r rate of condensation is usu ally known, the

value o f N a m ay be determ ined by solving Equations (31) and

na +nb

(3 7 ), I f we le t X a * Na and express Equation (31) in exponen­

na+ n b
tia l fo rm we have

c ~^*6 « l 0 3 P ^ B (39)
' *

°r x . <40>
/-£■
The value of yv and y^ lie s between zero and u n ity. Since (N ^ + N g )

is po sitive i f condensation is taking place, E^ m ust also have a value

between zero and unity*

I f the value o f E is one as if is when no condensation takes

place, then Equation (39) shows that y^ would equal yv and if th ere

w ere any condensate, it would be in static eq u ilib riu m w ith the m ain

vapor s tre a m . This condition would be approached if only a ve ry

 17

sm all am ount of the to ta l flo w of vapor was allow ed to condense. At

the other extrem e w h ere E is ae ro or approaches zero due eith er to

the la rg e am ount of condensate o r to the operating conditions which

m ake the fa c to r kg v e ry sm all* Equation (40) shows th at would

equal yyt o r in sh ort, bulk flo w would explain the m ass tra n s fe r.

In th is w o rk, the re a l question is the establishm ent of the

te m p eratu re a t the in te rfa c e . In the case w here E has a value of

one, then the condensate is in static e q u ilib riu m w ith the vapor

and the te m p eratu re of the in te rfa c e m ust approach or equal that of

the m ain vapor stream provided th e re is no superheat. In the other

case, w here E is equal to or approaches zero and the com position of

the condensate approaches th at of the m ain vapor stream , the

te m p eratu re a t the in te rfa c e would be that a t which the solution of

th at com position would ex ist at the given p res su re o r, in short, its

n o rm a l b o ilin g point* The lim its of the te m p eratu re of the in terfac e

w ould then be the b o ilin g point of a solution and the dew point of the

vapor of the sam e com position depending upon the m ethod of op era­

tio n . B etw een the lim its described above, the tem p eratu re of the

in te rfa c e could be established m aking use of m ass tra n s fe r data.

W ith the te m p eratu re a t the in te rfa c e known, the tem p eratu re drop

across the film of condensate could be established fo r use in

N u s s e lt’ s equation.

The second condition set fo rth by N usselt "T h at the velo city
 13

of tho film a t the w a ll surface m ust be ze ro , and the film m ust cling

to the w a ll; and the surface of the film in contact w ith the vapor m ust

he under the same p res su re as the vapor** c e rta in ly applies as w e ll

to b in ary liq u id solutions as to pure liq u id s.

Th e th ird condition "that the heat passing through the film

is th at of the laten t heat of condensation*1 ap plies as w e ll to b in ary

solutions as to pu re liquids provided th ere is no appreciable super­

heat and th at the laten t heat of condensation is evaluated fro m an

en thalpy-concentration d iag ram . Th is d iag ram shows the enthalpy

of a m ix tu re as a function of te m p eratu re and com position. (See

F ig . N os. 27 and 28 in A ppendix). Thus, if we know the com position

of a liq u id and the in itia l conditions of tem p eratu re and phase at a

constant p ressu re and the fin a l conditions of te m p eratu re , the change

in enthalpy can be read e a s ily fro m this d iag ram . The accurate

construction of this diagram calls fo r data which is not too p revalen t,

but approxim ations can be m ade to a lle v ia te this shortage of data.

T h is change in enthalpy m ay be thought of as a ”laten t heat of con­

densation*' although it is not ac tu a lly such. .As fa r as the N usselt

Equation is concerned, the use of laten t heats is to sim p lify the

determ in atio n of the to ta l heat tra n s fe rre d and the s tric t usage cf

the la te n t heat is not a lim ita tio n so long as a c o rre c t value fo r the

to ta l heat tra n s fe rre d ie used.

I f the operating conditions a re such that the ve lo c ity of the

 19

vapor past the film of condensate is w ith in the bounds established by

N usselt and the ra te of condensation is such that stre am lin e m otion

p re v a ils » then the re m a in d e r of N u s s e lt*& stipulations should apply

equally w e ll to b in a ry solutions as to pure compounds.

It appears then th at N u s self's Equation should apply in the

p red ictio n of heat tra n s fe r co efficients fo r b in a ry vapors th at fo rm

a single liq u id phase since the soundness of the equation fo r pure

compounds has been proven by m any in ves tig a to rs.

O ther C onsiderations

I t has been shown th at the N usselt Equation should be usable

in the p red ictio n of the film coefficients of heat tra n s fe r fo r the con*

densation of b in a ry vapors th at fo rm a single liq u id phase. How­

ever the obstacle to its use is the absence of the necessary physical

data needed fo r its solution. T h e re is a shortage of the necessary

physical data fo r solutions, but th e re is adequate data av aila b le fo r

the pure components of m ost solutions. The obvious hope then is

to apply some weighted average to the p ro p erties of the pure compo­

nents.

Using the best data a v a ila b le , F lit c r a f t ^ and Rhodes and

Y o u n g er^ ^ found that the group

independent of te m p eratu re over short te m p eratu re ranges which

tends to im prove the hope of determ in in g the film co efficien t of heat

tra n s fe r fo r condensing b in ary vapors that fo rm a single liq uid

 20

phase since th e re is u sually a w ide v a ria tio n of te m p e ra tu re w ith

com position of a b in a ry system . Laten t heat of condensation need

not be considered since it is p ra c tic a lly a lin e a r function of com posi­

tio n .

The question now a ris e s ais to what can be done to determ ine

an average value fo r the aforem entioned group. Obviously,, the

prob lem can be separated into two p a rts , p ro p e rtie s fo r id eal solu­

tions and p ro p e rtie s fo r n o n -id ea l solutions.

T h e re is a lim ite d am ount of data on th e rm a l conductivity

as a function of com position of b in a ry solutions. I t m ust be pointed

out th at m ost of th is m eagre data is of doubtful accuracy as pointed

out by Sakiadis and C o a te s ^ *^ . H ow ever, it m ay serve to illu s tra te

the problem s a t hand.

B efo re going fu rth e r, a thorough under standing of th e rm a l

conductivity is necessary since it plays such an im po rtant p a rt in

d eterm inin g the value fo r co efficients of heat tra n s fe r as shown in

N usselt*s E quation. T h e rm a l conductivity m ay be thought of as one

of the tra n s p o rt p ro p e rtie s of a substance. These p ro p erties include

d iffu s iv ity and v is c o s ity . T h e rm a l conductivity is a p ro p e rty w hich

is te m p eratu re dependent. Num erous attem pts have been made to

re la te th e rm a l conductivity to the m ore com m only used and e a s ily

obtained physical p ro p e rtie s . T h is approach has not been e n tire ly

sa tis fa c to ry up to the present tim e fo r the equations w hich have been

 21

developed though rigo ro us e re not useful to r they are so com plicated.

(XI)
H ow ever, P a lm e r has im proved a sem i -e m p iric a l re la tio n to tie

in th e rm a l conductivity to other physical p ro p erties*

•d- O.OS+ 7 rfm (Vr) (41

Cp * specific heat (g. c a i. /g ra ra -°C )
k * th e rm a l co adu ctivityfC al. / sec. -cm ? - ° C /c m .)
m » m o lecu lar weight
T * bo ilin g point (°K )
A. 9 laten t heat of vap o riaatio n (C al. /m o l)
/* 9 density (g /c c )

T h is equation was found to hold fo r 4$ liquids w ithin £ 3. 8%.

The la s t p a rt of the equation ( A ./T ) w arrants discussion here since

it was introduced by P a lm e r to show the effect of hydrogen banding

on th e rm a l conductivity. He noted that fo r liquids that w ere h yd ro ­

gen bonded, such as w ater and m ethyl alcohol, the th e rm a l conduc­

tiv ity was ab n o rm ally high com pared to that expected on com parison

w ith substances of lik e m o lecu lar weight. The entropy of vaporisation

was chosen to show the effect of hydrogen bonding. The v a lid ity

of this b e lie f was enhanced by application of this te rm fo r it im proved

the accuracy of the equation fro m 1 23% to the present £ 3. 8% fo r

the 48 substances fo r which exp erim en tal data w ere av aila b le .

P a lm e r went fu rth e r in the study of the effec t of hydrogen bonding

on th e rm a l conductivity and by studying a homologous se rie s proposed

that fo r w ater 80% of the heat conducted was due to the presence of
 22

hydrogen bonds and lo r m eth yl alcohol 32%.

P erhaps* the reasoning in the foregoing paragraphs m ay be

ap plied to the case of b in a ry solutions. I t would seem reasonable

to expect th at on m ixin g tw o substances which fo rm an id eal solution*

th at each would contribute an ad ditive p a rt to the th e rm a l conductivity

of the solution. F u rth e rm o re if appears that a p lo t of th e rm a l con­

d u ctivity versus com position gives m ore n e a rly lin e a r relation ship s

if the com position is handled on a w eight p er cent basis ra th e r than

on a m ole p er cent b asis. D ata fo r the th re e system s studied in th is

w o rk a re not av ailab le* but data fo r s im ila r system s a re av ailab le

over a lim ite d te m p e ra tu re rang e. A lso* it m ust be strongly em pha­

sized th at the values fo r th e rm a l conductivities h e re in used a re

h ighly questionable but it is hoped that the re la tiv e effec t of com posi­

tio n is not too g re a tly in e r r o r .

^ F ro m the foregoing discussion* it m ight be expected that 'f

th e rm a l conductivity fo r an id e a l b in a ry solution m ay be determ ined

fro m averaging values fo r the pure components on a w eight b asis.

In o rd e r to check th is assum ption* data fo r a system such as benzene-

toluene should be used. U nfortunately* th at data is not a t present

exlstan t, but th e re is some data fo r the ethanol-m ethanol system*

w hich is a v e ry n e a rly id eal system . (F o r a discussion of the

reasons fo r n o n -id e a lity in solutions see G au treau x1 S'A T h e rm o ­

dynam ic A n alysis of C o rre la tio n s fo r V a p o r-L iq u id E q u ilib ria in

 FH1
13
S3
jtt:
rS
m
ST
L-3-ITU
TFf.::T
 24

N o n -Id e a l S o lu tio n s ^ . A p lot of th is data a t 6 8 °F shows only a v e ry

slig h t d eviatio n fro m lin e a rity (F ig . N o. 2 ), H ow ever, the p lo t fo r

e th an o l-w ate r shows a g re a te r d ep artu re fro m lin e a rity , and w ith

b en zen e-acetic acid , the d ep artu re is e x tre m e ly g rea t, and fu rth e r­

m o re a m in im u m is reached a t ap p ro xim ately 40 w eight p er cent

ac etic a c id . The p lo t fo r a c etone - w ater shows the same general

tendency of d ep artu re fro m lin e a rity w ith the indication th at a t high

w eight concentrations of acetone, the effec t of the acetone on the

th e rm a l conductivity tends to lessen. It seems then that the th e rm a l

conductivity of a solution is dependent on the degree of id e a lity of

the solution. Since it has been proposed that th e rm a l conductivity is

to some extent dependent on hydrogen bonding and since hydrogen

bonding has been found to be a v e ry g re a t fa cto r in the behaviour of

solutions as fa r as id e a lity is concerned, it is not su rp risin g that the

above noted behavior of th e rm a l conductivity in solutions is the case.

Q u a lita tiv e ly , th is behavior m ight be explained as fo llo w s . The

th e rm a l conductivity of associated liquids is found to be ab n o rm ally

high. Now if some substance is introduced which causes a decrease

in the association of the m olecules of the substance, then not only

should the o v e ra ll th e rm a l conductivity fo r the solution be changed

due to the presence of the fo re ig n m olecules as such but the additional

effe c t of th is dissociation should be apparent. C e rta in ly th is m ight

explain the case fo r the system including w ater discussed above; and
25
 26

m M .
RO UP

wr

|
10
 28

by analogy i t should c e rta in ly hold fo r the system s used in this w ork

substituting m eth yl alcoh ol, an associated liq u id , though not to such

a high degree as w a te r, fo r w a te r. S im ila r plots fo r densities and

vis co sitie s of the above solutions {F ig s . 3 and 4) show the same

general tendency as th at noted fo r the th e rm a l conductivities.

F o r the th re e system s fo r which data a re a v a ila b le , a plot

was m ade of ^ ^ I versus Com position (F ig . N o. 5).

7
T h is p lo t shows th at the group
f
^
*
^
V fc] *
J is v e ry n e a rly

U near fo r the system s which do not contain a highly asso ciated

liq u id , but fo r the w a te r-e th a n o l system , a g reat deviation fro m

lin e a rity is noted. T o be sure th is com putation was c a rrie d out at

a low te m p e ra tu re but la c k of su fficien t data was the lim ita tio n .

T h e system s used in th is w o rk a re isopropanol-m ethanol,

rep o rted by T ra c h te n b e rg ^ ^ , acetone-m ethanol, and benzene-

m ethanol. Th e choice of these components aUows the effect of

vario u s classes of substances on an associated substance to be

studied as fa r as heat tra n s fe r coefficients a re concerned. The class

(61
of the substance re fe rre d to is th at given by E w e ll et a l ' in his

q u alitativ e discussion of id e a lity of solutions.

C lass X - Liquids capable of fo rm in g 3-d im en sio n al netw orks

of strong hydrogen bonds e. g. w a te r, g lyco l.

C lass XX - Liquids composed of m olecules containing both

ac tive hydrogen atom s and donor atom s (O , N , F I)
e . g .alco h o ls, acid s.
 C lass U I - L iquid s composed of m olecules containing donor
atom s but no ac tiv e hydrogen atom s e. g. eth er,
ketones, aldehydes,

C lass IV - Liquids composed of m olecules containing ac tive

hydrogen atom s but no donor atom s e. g. C H C L ,
c h 2c i 2 .

C lass V * A ll other liq uid s having no hydrogen bond fo rm in g

c a p a b ilitie s .

A ccording to th is grouping, m ethanol is a C lass 11 compound, acetone

a C lass 111 compound, benzene a C lass V compound, and isopropanol

a C lass XX compound. The significance of th is c la s s ific a tio n is to de­

te rm in e the effec t upon m ixin g . W hen benzene and m ethanol a re

m ixed , according to E w e ll, the predom inant effect is to b reak hydro­

gen bonds of the associated liq u id and it would seem to decrease the

value of the th e rm a l conductivity. When acetone and m ethanol a re

m ixed, hydrogen bonds of the associated liq u id a re broken, but

ad ditional hydrogen bonds between the unlike m olecules a re fo rm ed .

The outcome is not p re d ic ta b le . It depends upon w hich effect is p re ­

dom inant. W hen the two alcohols, m ethanol and isopropanol, a re

m ixed the outcome is again unpredictable.

 30

e:a t TUAN
t Ha 'n c

FjHT

rtt
i
lii

IS

Fx;

E!

13
rttJ

11

95
 C H A P T E R IH

P R E V IO U S E X P E R IM E N T A L W ORK

A search o l the lite ra tu re has reve ale d th at v e ry little w ork

has been published on heat tra n s fe r coefficients of b in ary vapors

fo rm in g a single liq u id phase* Evidence of only two investigations

was found, th at of W allace and D a v ie o n ^ ^ and Hagenbuch ^ who

have rep o rted a scanty am ount of data on the eth an o l-w ater system .

The fo rm e r's data was collected fo r condensation on a h o rizo n tal

c y lin d ric a l su rface, and Hagenbuch1s data is fo r condensation on a

v e rtic a l c y lin d ric a l su rface. The data of these two in vestigators

has been consolidated by McAdams^ and com pared to a s e m i-

th e o re tic a l p lot based on the weighted average of the film c o e ffi­

cients fo r the pure components. The agreem ent is fa r fro m good.

A ctu al com putation of the th e o re tic a l values fo r the film c o e ffi­

cients cannot be m ade since the data fo r the vario us physical

p ro p erties as a function of com position and te m p eratu re is not

a v a ila b le . H ow ever, data a t one low tem p eratu re (2 5 °C ) has been

found in the lite ra tu re and the v a ria tio n of the group

as a function of com position has been made and discussed (See Theory).

The g en eral shape of this curve is ju s t about the same as that ob­

tained when W allace and D avison1s data is plotted a t constant

31
te m p e ra tu re drop across the film . (F ig . N o. 6) U nfortunately,

the data is fo r too n a rro w a concentration range to draw any sound

conclusions* Hagenbuch's data is even m o re scanty than W allace and

Davison*s* but the two a re in gen eral ag reem ent when converted to a

common b a s is . W allace and D avison 's data indicated to them a

general ag reem ent w ith the conclusions reached by C olburn and D rew in

th e ir th e o re tic a l study of the p ro b lem .

Th e condensing su rface fo r W allace and D avidson's w ork

consisted of a b rass tube 36 inches long having an outside diam eter

of 0 .3 4 inches. The tube was placed inside a steel ja c k e t 3 inches in

d ia m e te r. Therm ocouples w ere im bedded in the surface of the tube

and th erm o m e ters w ere placed in the vapor stre a m . The vapor was

introduced a t one end of the ja c k e t and w ithdraw n a t the opposite end.

Cooling w ater flow ed inside the tube and th erm o m eters m easured the

in le t and ou tlet w a te r te m p e ra tu re . The liq u id was heated and

vaporised in a b o ile r which was a steel drum provided w ith a steam

c o il. P ro visio n s w ere made to co llec t the condensate fro m the tube*

A heat balance was m ade based on the tem p eratu re ris e and the

am ount of w ater flow ing through the tube.

In calculating the film co efficien t, it was assum ed that the

com position of the condensate was equal to the com position of the

liq u id a t the v a p o r-liq u id in te rfa c e . By re fe rrin g to the boiling point

diag ram fo r the eth an o l-w ate r system , the tem p eratu re a t the in te r -
 33

face was assum ed to be equal to the b o ilin g point of the com position of

the condensate. B y evaluating the te m p e ra tu re a t the in te rfa c e , if was

possible to determ in e the film co efficie n t o f the m ix tu re .

In itia lly , the com position of the condensate was kept constant

but la te r w o rk was p e rfo rm e d allow ing the com position of the conden-

sate to v a ry , keeping the cooling w ater ra te constant* The com post*

tlo n of the en terin g vapor ap proxim ated the com position of the
i

condensate w ith the v a ria tio n seldom exceeding 2% and in m ost

instances less than 1%.

"A p paren t” and " T ru e " film co efficients w ere calculated.

The fo rm e r w ere based on the te m p eratu re d ifferen ce between the

vapor and the surface w h ile the la tte r w ere based on the tem p eratu re

d ifferen ce between the estim ated in te rfa c e te m p eratu re and the s u r­

face te m p e ra tu re .
 C H A P T E R IV

A P P A R A TU S

In previous w o rk , the author1 designed an apparatus fo r

the d e term in atio n of fiim co efficien ts of heat tra n s fe r. Since p a rt

of the p resen t w o rk was to com plete the testin g and to adapt the

apparatus to the d e term in atio n of the film co efficients of heat tran s -

fe r fo r condensing b in a ry vapors, a short discussion of the equip­

m ent w ill be given h e re . D e ta ils of the apparatus a re shown in

F ig u re s 7, 8, 9. and 10. F ig u re s 11 and 12 a re photographs of the

ap paratus.

B rie fly , the apparatus consisted of a steam heated b o ile r

connected to a vapor ja c k e t surrounding a 3 .9 0 8 " copper b ar 36"

long w hich acted as the condensing su rface. The vapor ja c k e t was

connected to a condenser to take care of any vapor not condensed

on the p rim a ry condensing su rface. Cooling w ater was introduced

into both ends of the copper b a r through sp ecially d rille d holes and

rem oved a t the opposite ends. A system was provided fo r the c o lle c ­

tio n of condensate fro m the condensing su rface. The d etails w ill be

discussed in the fo llow ing section.

A s it has been pointed out p revio u sly, any apparatus fo r

•the d eterm in atio n of film co efficients of heat tra n s fe r m ust provide

an accu rate means fo r evaluating the am ount of heat tra n s fe rre d ,

 35

and the te m p e ra tu re drop acro ss the film of condensate, A t, fo r a

given heat tra n s fe r a re a . O f these te rm s , the m ost d iffic u lt to e v a l­

uate a c c u ra te ly is the te m p e ra tu re drop acro ss the film of condensate,

fo r th is te rm depends on the accu rate d e term in atio n of the te m p e ra tu re

of the condensing su rface . F o r pure liq u id s, the te m p eratu re drop

across the film is sim p ly the te m p e ra tu re of the saturated vapor

m inus the te m p e ra tu re of the condensing su rface, but w ith b in a ry m ix ­

tu re s it is the d ifferen ce between the e q u ilib riu m te m p eratu re of the

condensate and the surface te m p e ra tu re . A fte r a c a re fu l study of the

m eans a v a ila b le fo r m easuring surface te m p e ra tu re ^ ^ , it was de­

cided to use therm ocouples.

In deciding on the exact technique of in s ta lla tio n of the s u r­

face therm ocouples fo r th is p a rtic u la r p ro je c t, the follow ing co n sid er­

ations w ere kept in m ind .

1. The junction should be as sm all as possible.

2. The junction should be as near the surface as possible.

3. Th e condensing surface a t the ju nction should be as

smooth as possible.

4. The leads in the im m ediate v ic in ity of the junction

should p a sb through an iso th erm al zone to prevent

the conduction of heat to or fro m the junction.

5. The in su latio n of the leads m ust be of a m a te ria l

th at re s is ts d e te rio ra tio n due to the action of the

m a te ria l being condensed.

 36

( I ) Denotes Location O f
Row 9 'i t -i - > K ^ow 9 Va" >?<- ■9^' -7h - Junction
i
i
Row 2 ft & " ■ ---- 9% — ->J< -75*- 8 tz . -

i
Milled S lot 7 '/z . -Hi* 3 '4 “
Row 3

END V IE W OF BAR LOCATION OF T H E R M O C O U P L E S

ON C O P P E R BAR

36"

JSLOT
</f

1/16’ Undercut To R e c e iv e
P r o te c tin g P la te
5 /8 " Holes D rilled The Length O f
The B ar And Tapped To Receive
S E C TIO N A -A / 1 /2 " S S P ip e - H o le s D rille d On
z" D ia m e te r

T 7* I k
Groove Cut
By R o lle r

Groove Cut
By R o lle r

M ille d S lot To R e c e iv e
T h e rm o c o u p le W ire

1/8 Hole D rilled

To In tersect S lo t LO C A TIO N OF W ATER DUCTS
{F u ll S c a le )
CROSS S E C T IO N AT T H E R M O C O U P L E
L O C A T IO N
(F u ll S c a le )

FIGURE 7
THERMOCOUPLE INSTALLATION DEPARTM ENT OF CHEMICAL
EN G IN E ER IN G

LOUISIANA ST A T E U N IV E R S IT Y

APPARATUS FOR MEASU RING

F IL M C O E F F IC IE N T S
Drown B y. J B TODD 2 0 JAN 1 9 5 0
S cale: 1/4" To l" Drowmo hfo. 5
 37

In o rd e r to obtain a v e ry sm all bead, 30 Gauge (B & S)

constantan and copper w ire s w ere selected. The insulation was of

woven fib e r glass w ith silicone varnish* T h is ©m all w ire allow s the

fo rm a tio n of a v e ry sm all bead and © till gives su fficien t strength.

E-ach therm ocouple was then c a lib ra te d , determ inin g the electro**

m otive fo rc e corresponding to a known te m p e ra tu re . The value of

the te m p e ra tu re was established by the b o ilin g points of pure ethyl

alcohol, d is tille d w a te r, and pure hexane.

A s m entioned p revio u sly, the condensing surface is a copper

b a r 3 .9 0 8 " in d ia m e te r and 36" long mounted h o rizo n ta lly . A t the

low est point on the circu m feren c e, a slot 1 /4 " w ide and 3 /8 " deep

was m ille d lo n g itu d in ally along the surface and te rm in a te d 2" fro m

eith er end of the b a r. Using a blunt pipe cutting w heel in a lathe,

a groove was ro lle d 0 .1 4 2 " on e ith e r side of the slot its fu ll length* -

Th is r o lle r pushed up the m eta l a t its edge© and le ft the m etal 1 /3 2 ”

higher than its o rig in a l position. Then the m e ta l between the groove

and the slot w a ll was m ille d out 1 /8 " deep to fo rm a shoulder to serve

a© a base fo r the cover p la te . {See F ig . 7 ). A t one end, a hole 5 /1 6 "

in d iam e ter was d rille d through the rem aining 2" of m eta l between the

end of the slot and the end of the bar a t such a depth that the bottom of

the slot coincided w ith the edge of the hole nearest the center of the

b a r. Thus the surface of the b a r was unm olested between the end of

the slot and the end of the b a r. F ro m each point w here it was d esired
 38
D E P A R T M E N T OF CHEMICAL
Holes To R e c e ive E N G IN E E R IN G
Steom Pipes
LO U ISIA NA S T A T E UNIVERSITY
35 APPARATUS FOR MEASURING
Bolt C ir c le - 15-1/16 F IL M C O E F F IC IE N T S
Drawn By: $ B Tgdd I M ay 1949
Scale '• 1/4 To I Drawing No. 2
I SS Pipe And F ittin g s
Holes To R eceive
1/2" B o lts
/'&
/6

- 22-

R e tu rn Bends

S team Coils — ---------- 20 V ------------------------

Welded To P la te ,lCl
P L A T E “C 1 T O P VIE W O F S T E A M C O IL S
P L A T E "O'
1/2 S t e e l P la te
(In n e r S u rfa c e M o c h in e d )

Holes To R e c e ive
1/2" Rolfs
IZ%- Ends To Be W elded To Vapor
Jacket

I 8 1 /2 " S ta n d a rd Br a s s
G a te V a lv e s

1/2" Bolts Holes Tapped To

3 / 4 " S t e e l Plate
Machined To Smoothness After Welding Receive 3 /8 " Gauge
1 /2 S ta n d a rd S c re w e d
Cocks
1 1/2 S S Pipe U n io n s
R im "B " W elded Flush zas
W ith Edge O f 3 6
Section Of 12" S S Pipe 1
3 /8 " Sight Gouge
Cocks(Screwed)

R eturn Bend
Steam Inlet

To Steam Trap

P la te "A" W e ld e d To PLATE A ”
P la te C H o le T ap p ed To Receive \
End O f S e c tio n O f 1 /2 " S te e l P ltfe
A s b e s to s 1 /2*' S S P ip e And D rd in C o ck
G osket
FIGURE 8 Hole T ap p e d To Receive
1 /2 " Feed Pipe
THE BOILER
 39

to lo cate a therm ocouple, a groove was ro lle d in the surface of the b ar

w ith the blunt cutting w heel along the circu m feren ce to a point 1 /2 ”
y*,
i
*Vfrom the s lo t. T h e depth of th is groove was 1 / 16“ w ith a ridg e of

m eta l pushed up to e ith e r side of the w heel 3 /6 4 " high. The w idth of

the bottom of the groove was 0. 040” . A t the point w here the ju nction

was to be located, the groove was allow ed to becom e m ore shallow

"■V
u n til the l/$ 2 " bulb b a re ly rem ained below the su rface. A hole 5 /6 4 “

in d ia m e te r was d rille d through the rem aining 1 /2 “ of m eta l between

the end of the groove and the slot in such a m anner that the v e ry

bottom of the groove was connected to the v e ry bottom of the s lo t.

The next op eratio n was to im bed the therm ocouples.

The therm ocouples w ere then laced through the slot into the

groove. A thoroughly cleaned piece of copper w ire was then la id

over the therm ocouple w ire and the m eta l fo rc ed up by the w heel was

peened over the therm ocouple ju n ctio n and lead w ire . The additional

copper w ire served as a f ille r . The surface was somewhat roughened

by the peening but the therm ocouple w ire and the additional copper

w ire le ft the peened surface higher than the unm olested surface a llo w ­

ing th is roughness to be smoothed w ith a file and fin e em ery p ap er.

This operation was repeated w ith each therm ocouple. When the

smoothing op eratio n was com pleted, each therm ocouple could be b a re ­

ly detected in the copper surface due to the contrast in appearance of

constantan and copper. Thus the therm ocouples w ere located at or

 40

FIGURE 9
CO O LING WATER S Y S T E M

1/2 S S Pipe Screwed

Fittings O f Copper Tubing
Into H oles 2, 4 . 8 6 1 /2" Copper T ubing Soldered
O f P la te " F ' Screw ed Into H o les 1 , 3 , 8 5
To Inside O f R in g l(AMAs O f P la te "F"
Shown In Fig A 1/2" SS Pipe W elded As S h o w n ,
__________ 6 " Long_____________
W ELL B
1/2" S S P ipe Welded To
End O f W e ir'B "

© 3 SS Pipe
Copper Bor 1
w ater Ducts
Fig. A

P la te F P la te F

2 Sections O f 3 And 4
2 Section O f 3 SS S S Pipe Mounted Concen­
Pipe Sealed At Each tric a lly W ith The Ends
End W ith 1/8" S t e e l P late/ S e a le d By 1/8" S te e l P la te
Welded On

Cover Plate

3 / 4 " Spa c e r(3 "x 8 ")

3 /8 " Pyrex P la te Glass

To Condenser

G askets

Plate W elded To V ap o r
Jacket
\ V a p o r Line

End View O f V a p o r Jacket

V - Groove For
S h ow in g Locatio -n O f Copper B ar
S e p a ra tin g C ondensate
S ig h t G la s se s

C o oling W ater S tream

End View Showing Position

Vapor Line C ollecting Trough
O f Trough And B a r
From Boiler
D E P A R T M E N T O F C H E M IC A L
Condensate O utlet E N G IN E E R IN G

L O U IS IA N A ST A T E U N IV E R S IT Y

A PPAR ATU S FOR M E A S U R IN G

F IL M C O E F F IC IE N T S
D raw n B y - J B TOD D 2 0 JAN 1 9 5 0
S c a le ; |/4 " T o l" Draw ing No 5 A
 41

near the su rface, the leads departed through an is o th e rm a l &one, and

the su rface a t the in s ta lla tio n was smooth and v is ib ly undisturbed. It

should be m entioned h ere th at in the m akeup of the therm ocouples care

was taken to in su re th at the in su latio n on the copper and the constantan

w ire s was in ta c t rig h t up to the 1 /3 2 ” bulb fo rm in g the ju n ctio n . T h is

insulation was fib e r glass covered w ith silicon e v a rn is h . Thus the

te m p e ra tu re of the bulb is the te m p e ra tu re m easured as the surface

te m p e ra tu re .

In o rd e r to com plete the is o la tio n of the therm ocouple leads

fro m the vap o r, the slot was sealed. T h is was accom plished by fittin g

a 1/8'* copper s trip snugly over the slot on the bearing surface p re ­

vio u sly p re p a re d . T h is op eratio n was com pleted by peening down the

m e ta l pushed up by the cutting w heel in p rep arin g the slot.

T o te s t the effectiveness of the seal obtained, the b a r was

subm erged in w ater w h ile m aintaining the slot under a ir pressu re of

4 P S IG . Some bubbles appeared but w ere re a d ily stopped by fu rth e r

peening. It can be said then that the seal was effe c tiv e .

The w ire em erging fro m the b ar was le ft fre e fo r connection

to a 10 point sw itch and thence to a vacuum bottle to contain an ice

bath fo r the cold ju n ctio n s. The exposed portion of the w ire was

p ro tected by encasing it in p ro tective tubing leading to the te rm in a l

block of the m u lti-p o in t sw itch. The cold junctions w ere insulated

fro m each other by covering w ith fib e r glass tubing and then w ere
 42
DEPARTMENT OF CHEMICAL
ENGINEERING_________
FIGURE 10
LOUISIANA STATE UNIVERSITY
THE VAPOR JACKET APPARATUS FOR MEASURING
FILM COEFFICIENTS
Drawn By: ,J B Todd 8 May 1949
S c a le -. 1/4* To I Drawing No.3A

JACKET TO BE M AD E FROM A 3 6 "

S E C T IO N OF 6 " S S P IP E

Vapor L in e To
5/1 6 Holes Condenser

SS P ip e And F it t i n g s

3 / 8 " Holes

PLATE "F"
1/4" S teel Plate I S ta n d a rd B rass
G ate V a lv e s

<3 — ©
3 /8 Holes

5 /1 6 " Thickness
R ig h t Head To B e C o n s tru c te d
\ O f Jacket
P L A T E "E" Id e n tic a l To L e f t H e a d
3 / 4 " S te e l Plate I 1/2" Vapor Ducts F ro m Boiler
PLATE F PLATE E
Welded To Edge O f 6 " S S Pipe W elded To J a c k e t As Shown
And M achined Sm ooth G A SK E T

Indicates Holes Tapped

To Receive 5/16" Bolts

Ind icates 1 1 /3 2 " Holes

Slots To Be Cut From
Wall As Indicated

35%" ' 1 71%

1/2" S te e l P la te To B e W elded '
Over S lo ts In J a c k e t As Shown ^

&k- y ^, !0‘
S te e l P la te s Welded
Flush W ith S lo ts

L o c a tio n O f Slots For Sight G lasses

Vapor Duct From
\ B o iler
Plates To Be Welded On
As Shown Above
 43

w rapped together w ith rubb er tape as a w aterp ro o f protection. The

cold junctions could then be introduced into the ice bath.

In o rd e r to obtain the vapor te m p eratu re , a therm ocouple

in s ta lla tio n was made in the b o ile r. A therm ocouple w e ll was made

by b ra zin g a 9 H sealed length of 3 /1 6 " copper tubing to a 1i Z u pip©

plug which was in tu rn screwed into the vapor space of the b o ile r.

Th e tubing was bent downward so that o il m ight be retained in the w e ll

although i t was mounted h o rizo n ta lly . The length of the w e ll was such

that it is fe lt that the conduction of heat fro m the location Of the June*

tio n to the b o ile r w a ll is n eglig ible. Th e therm ocouple was connected

to the m u lti-p o in t switch and its cold junction assem bled w ith those

of the surface therm ocouples.

A m e r c u r y -fille d th erm o m eter was in stalled in the vapor

ja c k e t as an aid in m aintaining a constant composition when w orking

w ith b in a ry system s, fo r a constant vapor te m p eratu re indicates a

m ix tu re of constant composition, pressure rem aining constant.

Tw o means a re availab le fo r determ ining the amount of heat

tra n s fe rre d fro m the vapor to the cooling flu id . One is to base the

d eterm in atio n on the change in tem p eratu re of the cooling m edium

a t a known ra te of flo w if specific heat data a re availab le, and the

other is to base the determ ination on the amount of condensate co lle c ­

ted fro m a given heat tra n s fe r a re a if latent heat data a re a v aila b le .

In the fo rm e r method, w ith low flu id rates and sm all tem p eratu re
 44

changes, v e ry p re c is e m easurem ents of te m p e ra tu re a re necessary if

any degree of accuracy is to be obtained. F o r exam ple, fo r a change

of te m p e ra tu re of 1 0 °F , fo r accuracy of 1%, readings acc^Sate to

0 f 1 °F m ust be m ade. T h is in its e lf is not an obstacle, but experience

w ith th is p a rtic u la r apparatus shows that even w ith the care taken to

m ain tain a constant ra te of flow of cooling flu id , variatio n s of 0 . 4 °F

w e re noted w ith in a perio d of approxim ately a m inute. H ow ever, by

m easuring the weight of condensate collected fro m a given a re a and

m aking use of latent heat data v e ry accurate resu lts can be obtained.

The condensate can be weighed w ithin 1 gram in 1000 gram s, which

is an e r r o r of 0 .1 % . The problem then is to collect the condensate

a c c u ra te ly fro m a.given a re a . Due to the fa c t that the copper bar

m ust be suspended fro m eith er end th ere is the p o ssib ility that heat

m ay be conducted to the surroundings and that condensate fro m the

suspension system m ay m ingle w ith that fro m the desired a re a . To

preven t this undesirable effect, it was decided to collect the conden­

sate fro m the ce n tral 27 1 5 /1 6 ” of the 36” b ar allow ing the condensate

fr o m the 4 ” end sections to d rip back to the b o ile r. T o achieve this

re s u lt, special rings w e re constructed to separate the a re a s . These

rings w ere made fro m 1 /4 ” brass stock. The outer surface of the

rin g was grooved to a depth of 3 /1 6 ” in a ”V ” shape. T h is rin g was

placed on the b a r w ith the center of the rin g over the point 4 ” fro m

the end of the b a r. A trough constructed of th in copper sheets was

 45

placed in the bottom of the vapor ja c k e t so that the ends of the trough

w e re located ju s t below the deepest po rtion of the f,V H~cut of the rin g .

Thus a s p lit of the condensate was obtained between the end sections

and the c e n tra l section. A visual check through the sight glasses of

the vapor ja c k e t shows that the sp lit does occur sharply and that a ll

the condensate fro m the desired section is collected in the trough.

T h e condensate was seen to follow the surface of the b a r to the lowest

section of the b a r and not to drop off before this point was reach ed .

A piping system allow s this condensate to be collected and weighed

or to re c y c le to the b o ile r. Speculation that the vapor m ight be of

a q u ality less than 100% was discounted by checking the heat tra n s fe r

by m easuring the condensate against that of the cooling w ater and

finding agreem ent w ith in ap p roxim ately 2% w ith that of the w ater being

the h ig h e r. T o p ro tect the b a r against im pingem ent of the vapor fro m

the b o ile r and th erew ith any droplets of liquid* baffles w e re placed

over the vapor in le t ports fo r the end ducts. The ce n tral duct was un­

im peded since it was fe lt that the presence of the condensate trough

served the same purpose fo r that p a rtic u la r duct. It m ight be noted

that the placem ent of the end ducts was such that the a re a contacted

by the d ire c t stream of vapor fro m the b o ile r was that a re a excluded

fro m collection of condensate. S im ila rly fo r the ducts to the conden­

s e r, the placem ent was such that any condensate dripping back u^ould

f a ll on the a re a excluded fro m collection of condensate in the case of

the end ducts. However* fo r the m iddle duct# a copper b affle 3" wide

supported h a lf way between the top of the vapor jacket and the b a r on

a brass w ir e fra m e was constructed. T h is shield extended downward

fa r enough to in sure that any d rip fe ll fre e of the copper b a r. As fo r

condensate on the inside of the vapor jacket* no d iffic u lty was en­

countered. In w a rm -u p periods* the condensate followed the surface

of the ja c k e t to its lowest point before dropping off. A fte r insulation

was applied to the jacket* v e ry little condensate was seen to collect

on the inside of the vapor ja c k e t w a ll.

When the condensate was rem oved fo r weighing it had to

be cooled to room te m p eratu re to prevent evaporation losses.

T h is cooling was accom plished by use of a sm all heat exchanger con­

structed of 1 /2 " copper tubing w ith w ater as the cooling m edium .

T h e condensate was collected fo r tim ed periods in h a lf-g allo n jugs.

T h e outlet fro m the cooler was connected to a loosely fittin g 3 /8 ” tee.

To each end of th is tee, about 2" above the mouths of the jugs# a 6"

nipple was screwed. Thus by pushing downward on the nipple, the

flo w of condensate could be changed instantly to either of the jugs.

Th e jugs w e re fitte d w ith loosely fittin g copper covers to prevent ex­

cessive evaporation losses. Thus an accurate means fo r collecting

the condensate was provided.

The b o ile r was constructed fro m a 36" section of 12" stand­

a r d steel pipe. One end was sealed com pletely by welding on a 3 /4 "

steel plate. The other end was provided w ith a flange and a rem ovable
 47

head, V ap o r ducts w e re IhofaHed a t shown (F ig , 8), The heating

elem ent was o rig in a lly composed of th ro e sections of 1” standard

ft
ste e l pipe fa b ric a te d as shown. H ow ever, using the liquids w ith low

laten t heats of vaporisatio n, the m inim um ra te of heat tra n s fe r to

the liq u id in the b o ile r proved too g reat, and one of the sections had

to be rem oved. Even this did not obtain the d esired resu lts, and

the system of m aintaining a pressure slightly above atm ospheric

p res su re by means of a p res su re reducing valve and tra p had to be

abandoned. Instead, the tra p was taken off, and a globe valve put in

the lin e ju s t downstream fro m the p ressu re reducing v a lv e . The out­

le t fro m the coils was put under a liquid seal provided by the con­

denser w ater and by adjusting the valve, the proper amount of steam

was bled in to give the d e sired heat tra n s fe r. The use of the pressure

reducing valve ju s t ahead of the adjusting valve proved to be a v e ry

desirab le fe a tu re in m aintaining a steady ra te of heat tra n s fe r.

F o r this p a rtic u la r w o rk, I t is m ost im portant that, steady

state conditions be m aintained. C ertain featu res w ere incorporated

into th is apparatus to a tta in this desired end. F ir s t , of course, a

constant flo w of vapor to the condensing surface is necessary. As

m entioned above, this constant flow of vapor was attained by

m aintaining a constant heat input to the b o ile r. O thm er and Whit*^

devised a scheme fo r obtaining is o th erm al conditions over the length

of a c y lin d ric a l condensing surface by using the counter fla w of a

coolant through the c y lin d e r. F e a tu re s fo r obtaining this counter flow

a re evident in the accompanying drawings {F ig . No. 9 ) and fu rth e r

discussion m ay be found in a previous w o rk of the a u th o r ^ ^ . Not

only m ust th ere be a counterflow of the coolant, but the flo w in each

leg of the system m ust be u n ifo rm and equal. T h is constant flow was

obtained by use of a constant head box to overcom e the fluctuations

in the p re s s u re of the c ity m ain s . In order to be able to check this

equality of flo w , m anom eters w e re in stalled in each leg of the system*

The use of the c ity w ater supply w ith a constant head box obviously

lim its to a ra th e r n a rro w degree the v a ria tio n in the quantity of heat

which m ay be rem oved. T h is lim ita tio n was undesirable when w o rk ­

ing w ith organic compounds of w id ely d iffe re n t latent heats since

ap p ro xim ately equal te m p e ra tu re drops across the film of condensate

w ere desired and this could only be controlled by v a ria tio n of the heat

rem oved by the coolant.

T h e re fo re it was necessary to v a ry the tem p eratu re of the

coolant. A means was provided fo r both heating and cooling the w ater

before it was introduced to the constant head box. The w ater was

heated and m aintained a t the d esired tem p eratu re by use of a Sarco

" B le n d e r.” T h is "B le n d e r”, a co m m ercial device, allow ed the

te m p e ra tu re to be controlled w ithin 1 1°F and m ostly even better

than th at. F o r obtaining tem p eratu res low er than those p revailin g

in the city m ains, a re frig e ra tio n unit was used. T h is unit was not
 49

FIGURE 11

VIEWS OF THE APPARATUS

Above: The Com plete Apparatus

The Condensate C o llecting S ystem

 FIGURE 12

A VIEW OF THE APPARATUS BEFORE INSULATING

 51

quite la rg e enough to m aintain a constant te m p eratu re of the coolant

during an e n tire run, but the d r ift upward in tem p eratu re wag only

about 4 ° F .

T o show that the desirab le design featu res were attained

the follow ing ex perim en tal evidence is offered . F ir s t , the reasons

fo r using the standard T e m p eratu re * E lectro m o tive F o rc e re la tio n ­

ships fo r copper-constantan thermocouples as published by Leeds

and N orth ru p w ill be discussed. A H the w ire fro m the th erm o ­

couples used was cut fro m the same spool which should make fo r

u n ifo rm ity among therm ocouples. Six of the thermocouples w ere

tested in cooperation w ith Orr^ according to his procedure using

n o rm al hexane, ethyl alcohol, and d is tille d w ater to establish the

p r im a r y te m p e ra tu re . The average deviation was -0 , 3 °F fo r the

six therm ocouples. I t m ust be pointed out that the insulating w e ll

used in this testing was a th ick w alled glass tube which could have a

b earin g on th is low te m p eratu re finding although if would not in the,

le a s t affec t O r r ^ uses. The rem aining thermocouples w ere tested

against one of these six and the deviation was found to be zero when

averaging the values fo r a ll the therm ocouples. A ll of this testing

was done p r io r to the in stallatio n of the thermocouples on the con­

densing surface. To be accurate, the therm ocouples, of course,

should be calib rated in place, but no means w ere available to c a rry

out this c a lib ra tio n . Since the results of this q u asi-ca lib ratio n
 52

showed re s u lts only slightly lo w er than that expected in tem p eratu re

determ ination s (about 0 . 1 °F ) and since th ere was a po ssib ility of a

slight change la c h a ra c te ris tic s of the therm ocouples upon Installation

and since the te m p eratu re s obtained by the therm ocouples would be

avers ged to obtain the surface tem p eratu re, it was decided to use the

standard T e m p e ra tu re -E le c tro m o tiv e F o rc e relationship ra th e r than

a c a lib ra tio n curve of doubtful value* In o rd e r to check the te m p e ra ­

ture m easured by the various thermocouples one against the other, the

following scheme was used. W ith the "blender", the tem p eratu re of

the coolant could be made the same as that of the vapor surrounding

the condensing surface. W ith the apparatus in operation, the tem p er -

itu re of the coolant was raise d grad ually until the last signs of con-

iens&tion w ere observed. The tem p eratu re of the coolant was then

.owered slig h tly u n til a v e ry thin film of condensate appeared. This

’ilm was so th in that only a v e ry few drops of condensate w ere seen

o f a ll fro m the surface. The tem p eratu re m easured by each couple

ras then recorded a fte r allow ing a reasonable tim e fo r eq uilib rium

o be established. The resu lts a re here recorded.

 53

Therm ocouple No. E lec tro m o tiv e F a rc e T e m p eratu re ( ° F )

0 2 .3 7 0 136.0
1 2 .3 7 3 136.1
2 2 .3 7 4 136. 1
3 2.3 6 9 136.0
4 2 .3 7 6 136.2
5 2.3 71 136.0
6 2 .3 7 0 136.0
7 2 .3 7 6 136.2
8 2.3 7 7 136. 2
9 2 .3 7 2 136.1
10 (V ap o r Space) 2. 374 136.2
(F o r Therm ocouple locations see F ig u re No. 7}

The consistency of the readings fu rth e r v e rifie s the v a lid ity of using

the standard T e m p e ra tu re -E M F relatio n s.

T o show that steady •state conditions w ere obtained, a

ty p ica l "ru n ” is cited. T h is Hru n M was c a rrie d out over a period of

25 m inutes.

Therm ocouple N o. T em p eratu re (° F )

A t s ta rt of operation A t end of operation

0 154.8 155.8
1 156.8 157.0
2 157.6 158.1
3 158. 1 157.6
4 158. 1 158.4
5 159.2 159.2
6 159.6 159.9
7 158.2 158.7
8 159.0 159.5
9 159.0 159.4
10 (B o ile r T e m p .) 176.7 -
s^e Surface T e m p e ra tu re 158.0 158.4

T h is constancy of surface te m p eratu re indicates the u n ifo rm ity of

 54

operation over the tim e d p e rio d . As described so fa r the equipment

was fo r use w ith pure compounds. The use of m ixed vapors, how­

ever* introduced one other p rob lem .

Th is p rob lem came about as a re s u lt of the fa c t that as

a solution is boiled, the composition is changed (azeotrope excepted,

of course). T h is change in composition brings about a change in

condensing te m p eratu re which is undesirable in this w ork. So, the

follow ing adaptation of the equipment was m ade. It was thought that

steady-state conditions could be attained if a constant holdup of con­

densate was made and as much solution fed back to the b o iler as was

rem oved fro m the b o ile r. T h is called fo r a m eterin g device fo r

establishing the feed ra te through t r ia l and e r r o r . F o r this purpose

a ro ta m e te r was placed in the feed lin e . The use of this ro tam eter

along w ith the proper technique of operation to be described la te r

brought about satisfactory re s u lts .

 CHAPTER V

E X P E R IM E N T A L WORK

Th e choice of the systems to b© used in this w ork was d ic ­

tated by two p r im a r y considerations. F ir s t , the components of the

system had to have w id ely d iffe re n t heat tra n s fe r coefficient®.

Secondly, systems w ere chosen fo r which v a p o r-liq u id eq uilib rium

data w e re av aila b le , fo r obvious reasons. A contributing fa cto r to

the choice was the d e s ira b ility of having re fra c tiv e indices w idely

d iffe re n t fo r the pure components of a system in order to use the r e ­

fra c tiv e index of a solution fo r establishing its composition. F u rth e r ­

m o re, the systems chosen should have no adverse effects on the

experim en tal equipment and should cause a m inim um perm anent

deposit on the condensing surface. O ther considerations discussed in

the T h e o ry Section along w ith the above point® led to the choice of

m ethanol-benzene and m ethanol-acetone solutions.

Reagents

Obviously, the purest reagents obtainable should be used in

any experim en tal w o rk . H ow ever, due to the larg e quantity necessary

to operate this apparatus successfully, 10 gallons, it was decided to

use c o m m e rc ia l grade substance®. The com parison of the various

re fra c tiv e indices w ith the lite r a tu r e values fo r pure compounds as

55
 56

shown in the follow ing tables indicates that the difference is only slight,

Substance R e fra c tiv e Index ^

L ite ra tu re E xperim ental

Methanol 1.3267 1.3268

Benzene 1.4967 1* 4942
Acetone 1.3566 1.3555
Isopropanol 1.3758 1*3710

Operating P ro ced ure

B efo re any series of determ inations was c a rrie d out, the

apparatus was thoroughly cleaned. T h is cleaning included the con*

densing surface. I t was necessary to dism antle one end of the appa­

ratus to clean the " b a r11 or condensing surface. T h is cleaning was

accom plished through the use of crocus cloth and soft rags. I t was

not necessary to rem ove the "b ar" fro m the vapor jac ke t. T h is

cleaning was repeated when the "b ar" appeared to be tarnished. Only

when using the acetone-m ethanol system did appreciable deposits

occur. T h is deposit was a fin e, w hitish powder, and it was rem oved

as soon as it became v is ib le , about every three days. The rem ainder

of the cleaning was accom plished by rinsing w ith sm all amounts of the

substance to be used and purging w ith a ir .

The substance to be used then was poured into the b o ile r in the

amount of ap p ro xim ately 6 gallons. T h is amount was enough to

cover the steam coils during operation, which could be checked by the
 57

sight glass on the b o ile r* It is im po rtant that the coils be covered to

prevent any superheating of the vapor.

T o begin the determ ination, the flow of w ater through each

branch of the cooling system of the condensing surface was equalised

a t the d esired ra te to obtain any p a rtic u la r rate of condensation. If

a te m p eratu re of the coolant higher than that of the city m ains was

d esired , the "blender" was adjusted p ro p erly. If colder w ater was

desired , the re frig e ra tio n system was adjusted to the proper le v e l.

A ls o , the flow of w ater through the a u x ilia ry condenser and the con­

densate cooler was regulated,

A l l valves in the vapor stream fro m the b o iler to the

a u x ilia ry condenser w ere opened and p a rtic u la rly the vent in the

a u x ilia ry condenser. The valves in the condensate line w ere adjusted

to a llo w recyclin g to the b o ile r. Steam was then adm itted to the

heating coils to brin g about boiling. During this in itia l warm up

p erio d, considerable vapor was fed to the a u x ilia ry condenser to

fo rc e out any a ir in the system . The steam flow then was adjusted

m anually to give only a slight amount of condensate fro m the a u x ilia ry

condenser. Th e amount of condensate was checked v is u a lly during

w arm up and was adjusted to about 50 m l. p e r 20 m inute period.

T h is ra te was m aintained a t such a low value to assure alm ost to tal

condensation in the vapor space and yet to m aintain the system fre e

fro m a i r . R ig id control of the steam flow was necessary to m aintain

 58

a constant flo w ©£ vapor to the condensing surface* T h is control of

the steam was d iffic u lt and m odifications w ere necessary to get

s a tisfacto ry resu lts*

It was necessary to prep are tine ice bath fo r the th e rm o ­

couples a t le a s t one hour before a "run" was m ade. A fte r about an

h our's "w a rm up*’ p erio d, checks of the surface te m p eratu re showed

that steady*state conditions had been reached* Runs of 20-25 minutes

then w e re made* F o r the pure components, the to tal amount of con­

densate fo r the tim ed perio d was collected and then weighed* How­

ev er, fo r the b in ary m ixtu res of non~a»eotropic composition, a

d iffe re n t technique was used due to the change in composition of the

\
solution during prolonged boiling. In this case, during both the w a rm - A

up and during the re g u la r run about 1/2 gallon of the solution was

m aintained as holdup. B y m aintaining this hold-up constant and feed­

ing hack to the b o ile r the same amount of liquid as was boiled off per

unit tim e , the com position of the vapors entering the vapor jacket was

m aintained constant. T h is procedure called fo r continual collecting,

weighing, and dumping of the condensate, The feed-back ra te could

be m easured by the ro ta m e te r in the line fro m the feed tank to the

b o ile r. The ra te was a rriv e d at through t r ia l and e r r o r un til the

condensate collection was ju st equal to the feed-back. The success of

this technique was proven by the analysis of sm all amounts of con­

densate collected p e rio d ic a lly . The v a ria tio n shown by analysis was
 59

©f the o rd e r of 0 .5 m ol p e r cent* T h is technique req u ire s considerable

e ffo rt on the p a rt of the o p erato r. The condensate then could be

weighed over any tim e d period*

D uring the tim e d run, the electrom otive fo rc e generated

by each therm ocouple was re a d at least tw ice by use of the Leed© and

N o rth ro p Type 1C P o te n tio m eter. T h re e samples of the condensate,

a co llective sample of the condensate fro m the a u x ilia ry condenser,

and a sam ple fr o m the b o ile r w ere taken during the run. A nalysis

was c a rrie d out by use of the Abbe R efracto m eter making use of a

p rev io u sly p rep ared re fra c tiv e index-com position curve. A nalysis

fo r com position in this m anner gives a high degree of accuracy as

shown by G autreaux

The te m p eratu re indicated by each thermocouple was ob­

tained fro m a plot of electrom otive force versus tem p eratu re. These

te m p eratu re s fo r the ten surface thermocouples w ere then averaged

a lg e b ra ic a lly to determ ine the average surface tem p eratu re. During

the same tim e , p erio dic determ inations of the vapor tem perature

w ere made through use of the therm ocouple mounted In the bo iler*

Thus the basic data w ere obtained*

In o rd e r to determ ine the amount of heat tra n s fe rre d per

unit tim e , the amount of condensate was weighed. T h is weight of

condensate then was converted to heat units by use of the latent heat

of condensation. F o r the pure components, this heat of condensation

 60

was s im p ly the la te n t heat of v a p o riz a tio n fr o m the lite r a tu r e . For

the s o lu tio n s , h o w e ve r, the la te n t heat o f con densation was d e te rm in e d

by use of an e n th a lp y -c o n c e n tra tio n d ia g ra m . These d ia g ra m s w ere

constructed according to the m ethod o u tlin e d by Dodge

(51 using data

fro m the In tern atio n al C r itic a l Tables^ and the Table of Phyeico-
{211
C hem ical Constants1 . Th ere is a lack of specific heat data fo r both

solutions, consequently an average value was calculated using m ol

percentage as the basis. In getting this specific heat, the effect of

tem p eratu re on the specific heat of the pure liquid was taken into

consideration. The heat of condensation then was read fro m the d ia *

g ram using a line of constant composition. This point is fu rth er

considered in the Results Section. Thus the necessary bits of in fo rm a «

tion w ere determ ined fo r solution of Newton's Equation.

 CHAPTER V I

R E SU LTS

I t m ust be rem em b ered that the object of this w o rk is to study

the effect of com position on the heat tra n s fe r coefficients fo r the conden­

sation of b in a ry vapors fo rm in g a single liquid phase. The problem

was attacked in the follow ing m anner. The f ilm coefficients of heat

tra n s fe r fo r the pure components w ere f ir s t determ ined. Then, the

film coefficients of heat tra n s fe r fo r various condensate compositions

w ere m easured, A series of determ inations was c a rrie d out under

such conditions as to m aintain a constant condensate composition and

alm ost to ta l condensation. V ario u s series w ere then run fo r each

system , m ethanol-benzene and acetone -m ethanol. The data of

(23V
Trachtenberg fo r isopropanol and methanol is included in this d is ­

cussion fo r it was determ ined on the same apparatus and under the

same conditions a t a subsequent date. The data determ ined fo r the

pure components w ill be discussed fir s t .

T ab le No. 1 gives the experim ental results fo r methanol,

acetone, and benzene. The film coefficients of heat tra n s fe r w ere

determ ined in the follow ing m anner. The tem p eratu re drop across

the film was determ ined by subtracting the average surface te m p e ra ­

tu re fro m the vapor te m p eratu re as m easured. The heat tra n s fe rre d

2
through a given a re a (2 . 38 f t ) was m easured by collecting the
 62

condensate and m aking use of heat of vaporization data. The known

values w ere then substituted in Newton* s Equation to determ ine h,

* / / ? u y /v
the heat tra n s fe r coefficient. Since the group — — *j has been

found to be alm ost constant as a function of tem perature* a plot of h

versus a t, the tem p eratu re drop across the film should give a straight

lin e w ith a slope of *0 , 25 on logging paper. This plot was made fo r each

pure component (F ig , No. 13), The lines shown a re the resu lt of the

calcu latio n to determ ine the best straight lin e to represent the exp er­

im en tal data by the method of least squares. In addition* the s ig n ifi­

cance of the data was tested by statis tica l methods as outlined by

B ro w n lee' The confidence lim its show that 95% of the determ inations

should fa ll w ith in these bounds o r in short we have the consistency of

the data. This does not establish absolute accuracy but only the con­

sistency of the determ inations w ith this equipment. The confidence

-j.
lim its show that the consistency of the data is approxim ately - 4. 5%,

This value would seem to im p ly that the equipment gives reproducible

results and hence accurate results based upon the findings discussed

in the Apparatus Section,

Since N usselt’ s equation has been found generally to be in

agreem ent w ith experim en tal results determ ined by many investigators,

a com parison of ex perim en tal data here determ ined and theoretical

results is in o rd e r. Lines representing th eoretical film coefficients

are included on F ig . No. 13, The v e ry good agreem ent between the
 63

TABLE I

E X P E R IM E N T A L RESU LTS

F I L M C O E F F IC IE N T S F O R CONDENSING VAPORS

(Single Components)

A. Benzene

Run No. V ap o r T e m p , A vg. W all Condensate Tem p Drop F ilm C oef-

(°F ) Tem p* (OF) (# /h r ) A cross F ilm fic ie n t h,
it(° F ) (B T tJ /(h r)
(£r)(°F)
1 176.9 146.9 9 3 ,2 3 0 .0 221

2 176.9 153.6 7 3 ,5 2 3 ,3 225

3 176.9 158.5 6 0 ,2 18.4 233

4 176.9 159.0 6 0 ,4 17.9 239

5 176.5 155.0 6 9 .3 2 1 .5 229

6 176.8 151,9 8 2 .2 2 4 .9 234

7 176.8 155.7 6 8 ,4 21. 1 232

8 176.5 159.9 5 7 .5 16.6 248

9 176.9 168.1 3 5 .5 8 .8 287

10 176.9 165,1 4 4 ,9 11.8 270

11 176.9 162.4 51.9 14.5 256

12 175.9 159.7 58.1 16.2 258

13 175.8 152.9 6 3 .1 17.9 251

14 176.3 157.3 6 4 .5 19,0 241

 64

T A B L E I (C o n t.)

B. M ethanol

Eon H o. V apo * T e m p . Avg. W a ll Condensate T e m p D ro p F ilm C o effi-

(® F ) T e m p , (O f) (#/h ir) A cross F ilm cient h» (B T tJ /
(° F ) (h r)(ft2M °F )
at
1 147.8 135.6 2 9 .7 12.2 487

2 147.9 135.1 3 0 .0 12.8 467

3 147.5 136. 8 2 6 .3 10.7 491

4 147.3 136.9 2 7 .1 10.4 516

5 14 7.4 136,9 2 7 ,4 10.5 517

6 147.7 133,1 3 3 .6 14.6 457

7 147.6 133.0 3 3 .5 14.6 456

8 147.6 130.5 36 .3 17.1 421

9 14 7.4 130.1 35 .6 17.3 411

10 147.3 128.9 38.9 18.4 421

11 1 4 7 .4 129.3 39 .8 18.1 437

12 147.4 128.6 39 .0 18.8 414

13 147.7 139.4 2 1 .0 8 .3 505

14 147.6 127.3 4 1 .5 2 0 .4 404

15 147.3 127.2 4 0 .4 20. 1 400

16 147.4 127,4 44. 1 2 0 .0 438

17 147.7 126,7 45. 1 2 1 .0 427

18 147.7 126.7 4 5 ,7 2 1 .0 433

19 148.2 118.4 5 6 .7 2 9 .8 378

20 148.0 116.6 5 7 .5 3 1 .4 364

 65

T A B L E I (C o a t.)

C. Acetone

tn N o. V ap o r T e m p . Avg. W all Condensate Tem p Drop F ilm C o e ffi­

(°F ) T e m p .(° F ) A cross F ilm cient h, (B T U /
(° F ) (h r)(ft2)(°F )
at
1 133.0 116.3 52.1 16.7 293

2 13 2 .4 118.9 4 5 .5 13.5 318

5 13 2.6 114.6 58 .0 18.0 304

4 13 2.6 112.8 59 .9 19.8 284

5 132.7 120.8 4 2 .3 11.9 334

6 132.7 121.7 3 9 .7 11.0 339

7 133. 1 123.4 37 .5 9 .7 364

8 132.9 110. 1 7 1 .7 2 2 .8 296

9 132.7 106.7 8 1 .5 2 6 .0 295

10 132.6 107.4 7 5 .2 2 5 .2 281

11 132.7 115. 1 55 .3 17,6 296

 a 9 , 10

O F SI
v e rs u s
CROSS. X H E 1 F I U M

MiSeHWfKH

Experimental DOto I
—- ^ t-T h e d r e tiC o i Prediction
9 5 % Confidence! Lim its

TEMPERATURE DROP p F l
O'
O'
 TABLE II
EXPERIMENTAL RESULTS
FILM COEFFICIENTS O F HEAT TRANSFER OF BINARY VAPORS

A. 17.2 Mol % A cetone in M ethanol

T z V apor T em p eratu re = 143*4 ®F(Avg) AT V * Tv - Tw

xV

f « B oiling Point = 138.4 ° F AT ^ s T - T

BP " BP BP w
Dew Point - 142. 9 ° F 6 T d p 9 XDP * T w
DP "

No. Avg. W a ll Tem p. Condensate Tem p eratu re Drop ( ° F ) F ilm C oefficients

<©F) (# /h r ) ATv 6T gp 6 T Dp ( B T U /h r - f t 2 ~ ° F |
hv hB P h DP
i 122. 2 34. 1 2 1 .6 16.2 2 0 .7 292 389 304

2 120.9 35 .4 2 2 .7 17.5 2 2 .0 287 372 296

3 117.5 39.5 2 5 .8 2 8 .9 2 5 ,4 283 349 287

4 125.4 2 9 .2 18.0 13.0 17*5 299 414 308

5 1 2 7 .1 26. 1 16 .4 11 .3 15 .8 294 426 305

6 129. 1 22. 1 14.5 9, 3 13 ,8 282 439 296

7 112.6 42, 1 3 1 .0 2 5 .8 30 .3 251 302 257

8 129*7 2 2 .2 13 .7 8 .7 13.2 302 475 313

9 115.0 4 1 .7 2 8 ,7 2 3 .4 2 7 ,9 266 329 274

 TABLE II (C o n t.)

B. 42. 3 Mol % A cetone In Methanol

Ty s 137. 1 °F (A V G ) T D p = 135. 7 ° F

T B p = 132. 6° F

Run No. Avg. W a ll Condensate T em p eratu re D rop ( F ) F ilm C oefficients

Tem p ( ° F ) (# /H r .) ATy ( B T U /h r - f t - ° F )
atbp atdp
hv hB P hD F
i 119.8 2 9 .6 17. 1 12.8 15. 9 280 374 301

z 117. 1 3 4 .9 19.8 15,5 18 .6 285 365 304

3 112.2 2 9 .0 24. 7 2 0 .4 2 3 .5 256 310 269

4 120.9 30. 1 16.6 11 .7 14,8 293 415 329

5 108.2 4 6 .3 2 9 .0 2 4 .4 2 7 .5 258 307 273

6 108.4 4 6 .5 2 8 .7 2 4 .2 27. 3 261 310 275

7 114.5 38. 4 2 2 .7 18* 1 2 1 .2 275 344 294

8 123.3 2 6 .4 14.4 9 .3 12.4 283 438 328

9 121.3 2 9 .8 16* 4 11.3 14.4 292 424 333

10 116.4 34. 3 2 1 .2 16.2 19*3 263 344 289

 T A BLE II (C o n M

C. 57. 3 Mol % A cetone in M ethanol

T v “ 13 J. S °F (Avg) T B p = 131.6 ° F 1 0 °F
:D P 5 132-
Run No. A vg. W all Condensate T em perature Drop ( ° F ) F ilm C oefficients
T e m p < °F ) <# / h r) ATv a T BP 4 T DP < B T U /< h rH ft2H ° F )
hv hB P hD P
1 118.6 32. 8 14 .6 13 .0 1 3 .4 322 362 351

2 115.7 3 9 .0 17.5 15.9 16. 3 319 351 342

3 113.4 45. 1 20. 6 18.2 18.6 312 353 345

4 113. 2 4 5 .6 2 0 .8 1 8 .4 18.8 304 344 336

5 119.2 3 2 .6 14 .6 1 2 .4 12 .8 319 376 365

6 108.5 5 4 .9 2 5 .0 2 3 .1 2 3 .5 313 337 332

7 123.0 2 7 .0 10.9 8 .6 9 .0 354 449 428

8 122.5 2 6 .8 11.1 9 .1 9 .5 344 420 402

9 124.0 2 4 .9 9 .8 7 .6 8 .0 363 468 445

10 123.9 2 4 .8 9 .7 7 .7 8. 1 367 461 438

11 115.7 3 9 .0 17.7 15.9 16.3 315 351 342

12 107.4 5 3 .2 2 6 .2 2 4 .2 2 4 .6 291 315 309

 TABLE n (Coat* )
D. BO M ol % A cetone in M ethanol (A zeotrope)
TV *
t bp s t dp ' 131. 3 °F (A vg) Nr = N a p = ^D P
Ron No. Avg* W a ll Condensate T e m p e ra tu re D rop F ilm C oefficients,
T emp ( ° F ) (# / h r) « °F ), 2 Tv

1 115.3 4 5 .3 16.0 335

2 118.6 4 0 .8 12.7 379
3 115.7 4 5 .9 15.6 348
4 118. 7 3 9 .6 12.6 371
5 111.3 5 1 .9 19.8 308
6 112.9 5 5 .8 1 8 .4 349
7 107.6 6 4 .6 2 3 .7 320
8 107.2 6 5 .3 2 3 .9 323
9 112.7 5 1 .5 18.6 324
10 106.9 6 3 .9 2 4 .4 308
11 122.9 3 1 .0 8 .7 422
12 122.6 3 0 .0 9 .0 396
13 12 2.4 3 0 .5 9 .1 386
14 11 2 .4 5 3 .6 18.9 335

-a
o
 T A B L E I I (C o n t.)

E. 17. 7 M o l % Methanol in Benzene

T v = 1 6 5 .8 ° F (A v g .)
T B p » 137. 1 °F (A v g .)
T d p = 165. 2 ° F (Avg. )

Run No. Avg. W all Condensate T e m p eratu re D rop F ilm C oefficients

Tem p ( ° F ) ( # /h r ) (°F ) (B T U /(h r )(ft2 )(° F )
AT h
^B P DP hv hB P DP

1 114.9 6 0 .5 5 1 .2 2 2 .2 5 0 .2 118 273 120

2 116.7 5 4 .6 4 8 .5 2 0 .4 48. 0 114 269 115
3 117.0 5 6 .8 4 3 .2 2 0 .3 48. 6 117 279 116
4 123.5 4 4 .3 4 1 .7 1 3 .7 4 1 .8 107 324 107
 TABLE II (C e n t.)

F. 46. 5 M ol % M ethanol in B enzene

Tv * 1 4 3 .7°F (Avg) T b p • 136. 0°F (Avg) TDP * 1 4 5 .9 °F (Avg)

Run No. A vg. Wall Condensate T em perature Drop F ilm C o efficien ts

Tem p (°F ) (# / hr) A TV A I J j A TD p (B T U /(h r ) (ft2 ) (° F )
K hB P hDP
1 116.8 4 2 .5 2 6 .7 19.2 2 8 .7 220 306 205

2 117,6 4 1 .5 2 5 .9 18.4 2 8 .2 221 309 203

3 115.6 4 6 .1 2 7 .9 2 0 .4 2 9 .8 227 310 213

4 11 3.4 48, 1 30. 1 2 2 .6 3 2 .6 219 291 202

5 120.7 3 8 .4 2 2 .8 15.3 25, 5 231 342 205

6 127. 1 30. G 1 7 .4 9 .0 2 0 .0 234 451 204

7 12 5.4 3 1 .6 1 8 .6 10,7 2 1 .7 231 400 198

8 121.9 3 4 .8 22. 1 14.2 2 5 .2 213 331 185

9 124.9 3 1 .9 18. 1 11.2 2 1 .6 240 388 201

10 129* 1 2 6 .5 1 4 .4 6 .9 16 .7 259 537 223

11 130. 1 2 5 .0 13 .4 5 .9 16. 1 255 577 211

-«4
8\>
 T A B L E n (C o n t.)
G. 6 1 .2 M o l % M ethanol in Benzene (A zeo trop e)
T y * T g p » T g p = 135* 7 ° F (Avg) hy - h g p * h p p

Run N o. A vg. W a ll Condensate T e m p e ra tu re Drop ( ° F ) F i lm C o efficien ts, hy,

Tem p ( ° F ) ( # /h r ) ATy < B T U /(h rH ft M °F )

1 117.7 17.9 317

2 119.2 1 6 .4 333

3 121.2 14.2 345

4 123.5 12.0 362

5 12 5.4 10.0 405

6 123.3 12 .5 348

T 120.4 15.4 335

8 116. 4 1 9 .4 321

9 114.8 2 1 .0 313

10 108. 2 2 7 .6 283

11 109.2 2 6 .6 278

12 126.6 9 .3 379

13 125.5 1 0 .4 365

14 108.2 2 7 .7 295
 T A B L E II (C o n t.)

H. 83 M o l % Methanol in Bensene
Tv : 1 4 2 .4 ° F (Avg) TB p : 1 3 6 .9 ° F TDp s 1 4 2 .4 ° F

Run No. Avg. W all Condensate T e m p e ra tu re Drop ( ° F ) F ilm C o e ffic i| p is

Tem p ( ° F ) <# / h r) AT A Tn n AT (B T 0 /(h r )< F t4') ( ° F )
v BP DP
\ feB P hB P
I <126. 3 2 4 .3 16 .0 10.6 16. 1 277 418 276

2 125.8 24. 1 16.5 11. 1 16 .6 266 395 264

3 119.8 3 4 .9 2 2 .5 17. I 2 2 .6 284 373 283

4 12 1.7 3 2 .4 2 0 .5 15. 2 2 0 .7 275 371 273

5 121.8 3 1 .9 2 0 .2 15.1 2 0 .6 288 385 282

6 119.4 3 7 .5 2 3 .2 17.5 2 3 .0 295 391 297

7 118.4 3 9 .0 2 4 .2 18.5 2 4 .0 293 383 295

8 114. 3 46. 1 2 8 .3 2 2 .6 28. I 301 377 303

9 114.5 4 4 .6 2 8 .3 2 2 .4 2 7 .9 288 364 293

10 113.4 4 3 .4 2 8 .9 2 3 .5 2 9 .0 275 338 274

11 129.9 18.5 12 .9 7 .0 12.5 261 482 269

•4

12 129. 1 18.6 13 .7 7 .8 13. 3 248 436 255

 75

th e o re tic a l and exp erim en tal values fo r methanol is ve ry evident^

H ow ever, the values fo r acetone and benzene are g re a tly in variance,

although the slopes of the lines a re in good agreem ent with theory.

These slopes fro m experim en tal data a re -0 . 27, -0 ,2 3 , and -0# 22 fo r

m ethanol, acetone, and benzene resp ectively. The fact that the

m ethanol data is so good discounts the effect of a faulty apparatus.

The only reason that is evident fo r the discrepancy noted in the ben*

aene and acetone data is the im p u rity of the reagents as pointed out

p reviou sly. This p u rity , along with questionable data fo r the physical

p ro p e rtie s , could cause this disagreem ent. It is believed, on the

strength of the above discussion, that the data presented here

represents accurate determ inations fo r the substances used in this

w ork.

Since a calculation of the operating conditions necessary to

give a Reynolds Num ber o f 2100 showed that a ll experim ental d e te r-

j
m inations w ere w ell into the stream lin e flow region, turbulence in

the film could not appreciably affect these resu lts.

Surface tem peratures m easured in this w ork show agreem ent

with previous w ork as fa r as va ria tio n of the tem perature around the

circu m ference of the cylin d er is concerned. The highest tem perature

occurred at the top of the cylind er w ith the low est tem p eratu re indicated

by the therm ocouple on the lowest portion of the cylinder. This v a ria ­

tion amounted to approxim ately 2. 5 ° F on the average determ ination.

 76

It is thought that the placem ent of the thermocouples on the condensing

surface allow ed alg eb raic averaging of the surface tem peratures to

determ in e a usable surface tem perature*

A t a ll tim e s , w ith the pure components, the d im of condensate

appeared to be u n ifo rm and to flow evenly over the surface. V e ry

little tarnishing of the surface was noted. The drops fe ll fro m the

condensing surface at u n ifo rm in te rv a ls showing that the b a r was in

a p e rfe c tly horison tal position.

E x tre m e ca re had to be exercised in preventing an excessive

amount of vapor to flow to the a u x ilia ry condenser. The effects of

this vapor flow on the coefficients was evident from >several runs

made w ith e x tre m e ly adverse conditions p rev ailin g . The coefficients

so determ ined w ere abno rm ally high. This velocity effect was

probably accentuated by the design of the jacket surrounding the con*

densing surface, an annular space of slightly less than an inch.

This featu re was incorporated into the design fo r possible future use

in the study of velocity effects on heat tra n s fe r coefficients*

E x p e rim e n ta l data fo r the benzene-m ethanol and acetone -

methanol systems a re presented in Table No, G, Operating conditions

w ere controlled in such a m anner as to give alm ost total condensation

of vapors and a condensate of constant composition as previously

d escrib ed . The condensate was weighed fo r a tim ed period and the

heat tra n s fe rre d was calculated using an enthalpy-concentration

 77

d iag ram (See Appendix), The value of the tem perature drop across

the film how ever could he reckoned in several possible w ays.

The th e o re tic a l discussion has shown the lim itin g values

o f the com position at the in terfac e depending on the operating condi-

tions. A calcu latio n was made to determ ine the th eo retical re la tio n ­

ship between the com position of the vapor and the condensate* Using

Equations {37 & 38), k was calculated fo r a typical run*

s
In o r d e r to make this calculation, some necessary approxim a­

tions w e re m ade. Since the total flow past the surface was alm ost

sere, values fo r and the Reynolds Num ber w ere obtained by

dividing the to tal rate into the vapor jacket by 4 since the flow was

sp lit equally on each side of the b a r and the ra te at the top of the bar

was p ra c tic a lly ze ro . F ro m Equation (38), j was calculated to be

£3
2 /3
ap p ro xim ately 0. 008. A value fo r ( ) % . 86 was taken at an
^ Dv
average of those repo rted by Johnston and P igford fo r bin ary m ixtu res.

The value of was assumed as one atm osphere. The value fo r

2
k_
g
was then found to be _
0. 009
_
# m o ls /(h r)(ft )(atm os). The value of

E was then calculated fro m Equation 39 to be 1 /e which is p ra c tic a lly

zero .

This finding substantiates th eo retically the fact that the com ­

position of the condensate should be p ra c tic a lly that of the m ain vapor

stream as shown by Equation (40).

* i = yv - E yj
1- E
 78

The exp erim en tal findings w ere in close agreem ent w ith this

conclusion* fo r analysis of the sample fro m the a u x ilia ry condenser

showed i t to be only slightly ric h e r in the m ore vo latile component

than the condensate. T h is enrichm ent m ight w e ll have been increased

by p a r tia l condensation in the pipe leading to the a u x ilia ry condenser.

I t seems reasonable to expect that the tem perature of the

liq u id at the in te rfa c e should be the n o rm al boiling point of the solu­

tio n a t the given pressu re fo r these operating conditions. How*

e v e r the ex p erim en tal resu lts a re presented with the heat tra n s fe r

coefficient based upon a tem p eratu re drop across the film of

condensate using the boiling point corresponding to the condensate

composition* the dew point* and the tem p eratu re of the entering

vapor* Th e resu lts a re lis te d in T able No. H I,

T h e b o ile r te m p eratu re was determ ined d ire c tly fro m the

therm ocouple located in the b o iler* The dew points and boiling points

w ere obtained fro m v a p o r-liq u id eq u ilib riu m diagram s (F ig Nos. 29,

30, and 31). T h e re was some d iffic u lty in choosing the proper set of

v a p o r-liq u id e q u ilib riu m data fo r the a c et one - methanol system fro m

the fo ur seta availa b le . The therm odynam ic consistency of each set

of data was tested and that of Q th m e r^ *^ was chosen based on the con­

sistency test and the fa c t that this was the only data that gave the te m ­

p e ra tu re of the saturated vapor equal to or low er than the tem p eratu re

 79

m easured in the vapor space in this work* This data, though not

e n tire ly consistent therm odynam ically, was smoothed according to the

Gibbs - Duhem Equation, This smoothing caused only slight tem p era­

tu re changes. The data of Chu^ ^ and G autreaux^^ fo r the benzene -

m ethanol w ere in fa ir ly good agreem ent. The data determ ined by

G autreaux w ere e x tre m e ly accurate and w ere taken on ve ry pure sub­

stances, It has a lrea d y been pointed out that the p u rity of the sub­

stances used in the present experim ental w ork is not too great* At

the aseotropic composition, the tem peratures m easured in the b o iler

w ere in good agreem ent w ith Chu’ s aseotropic tem perature but some­

what lo w e r (1 ° F ) than that of Gautreaux, Also between the azeotropic

com position fo r the system and that of pure methanol, the tem perature

of the saturated vapor was higher fo r G autreaux4 data than that m easured

in the b o ile r, a case which seems to be unreasonable. Chu's data, how­

e v e r, gives the expected agreem ent* This is not a c ritic is m of

G autreaux1 data but it is believed the difference noted is due to the

p u rity of the substances used, that of Chu’ s being closer to that used

h ere. F o r this reason Ghu's vapor-liqrud eq u ilib riu m data was used.

F ro m the discrepancies noted here it would seem that the va p o r-liquid

e q u ilib riu m data should be obtained using the very sam e quality sub­

stances as used in the heat tra n s fe r determ inations, o r ex trem ely pure

substances should be used in the heat tra n s fe r determ inations.

A plot of the heat tra n s fe r coefficient versus the tem perature

 T A B L E IU

F IL M C O E F F IC IE N T S A T A C O N S T A N T T E M P E R A T U R E DRO P

A, F ilm C oefficients of Single Components as a Function of T e m p e ra tu re Drop A cross the F ilm

(T - T )
v w

T v * Vapor T e m p e ra tu re , ° F T w - Surface T e m p e ra tu re , ° F

T em p eratu re Drop F ilm C oefficient ( B T U /h r - f t ^ - ° F )

Ojp Benzene Acetone Methanol Isopropanol

10 278 343 502 285

15 252 315 450 242

20 238 290 418 218

25 227 278 392 198

30 218 265 375 185

GO
O
 T A B L E IB (C o n t.)

B. F ilm C oefficients of the A cetone-M ethanol System at a Constant Tem perature D rop
A cross the F ilm as a Function of Com position*

T g p • B oiling Point o f Condensate (° F ); T w = Surface T e m p e ra tu re ( ° F )

Tem perature Drop F ilm C oefficient* h g p {B T U /h r-ffc ^ -°F }

°F 1 7 .2 M o l % Acetone 42* 3 M o l % 57. 3 M ol % 80 M o l %

10 435 415 397 385

15 385 367 363 350

20 357 340 338 327

25 335 320 315 305

30 320 302 302 288

Co
 T A B L E IH (Cont. )

C. F ilm C oefficients of the A cetone-M ethanol System at a Constant T e m p e ra tu re D rop ( T j j p - T ^ )

as a Function of Com position

*^I>F * ^ o in t Condensate (° F ); T w * Surface T e m p e ra tu re (°F J

Tem perature D rop ( ° F ) F ilm C o efficien t, hj^p, (B T U /h r*

17. 2 M o l % Acetone 42. 3 M o l % 57. 3 M o l % 80 M ol

10 310 352 395 385

15 310 317 357 350

20 305 295 332 327

25 292 280 317 305

30 270 267 302 2 88

GO
 T A B L E U I <Cont. )

D. F ilm C oefficients of the A cetone-M ethanol System at a Constant T e m p e ra tu re D rop ( T y - T ^ )

as a Function of Com position

T v * V apor T e m p eratu re ° F T w * Surface T e m p e ra tu re ° F

Tem perature D rop ( ° F ) F ilm C o efficien t, hy# ( B T U / k r - f t ^ - ° F )

17. 2 M o l % Acetone 42. 3 M o l % 57. 3 M o l % 30 M o l %

io 250 255 355 385

15 290 265 330 350

20 300 278 310 327

25 275 295 298 305

30 260 320 287 233

00
w
 TABIJE m (C on t. )

E. F ilm C oefficients of the M ethanol-B enzene System at a Cons tart T e m p e ra tu re D rop A cross the
F ilm ( T g p * T ) As a Function of Com position

^B F s B oiling Point of Condensate (° F ); T w - Surface T e m p e ra tu re { ° F }

Tem perature Drop (° F ) F ilm C oefficients, h g p , ( B T U /h r ~ ft ^ -° F )

17. 7 M o l % Methanol 46. 5 M o l % 61. 2 M o l % 83 M o l %

10 345 415 375 415

15 304 335 335 380

20 280 300 310 360

25 263 280 292 345

30 250 265 279 330

00
I*.
 T A B L E I I I (Cont. )

F* F ilm C oefficients of the M ethanol-B enaene System at a Constant T e m p e ra tu re D ro p ( T j^ p - T w )

as a Function of Com position

T£)p r Dew Point of Condensate (° F ) s Surface T e m p e ra tu re , ° F

Tem perature Drop F ilm C oefficients, h r ,p , (B T U /h r -£ tZ- ° F )

17. 7 m ol % Methanol 4 6 .5 M ol % 6 1 .2 M o l % 83 M o l %

10 223 375 272

15 217 335 280

20 210 310 286

25 207 292 290

30 203 279 294

CO
tn
 T A B L E I I I {C o n t.)

G, F ilm C oefficients of the M ethanol-B enzene System at a Constant T e m p e ra tu re D rop (T V* T W)

as a Function of Com position

T y = V apor T e m p e ra tu re ( ° F ) T * Surface T e m p e ra tu re ( ° F )

Tem perature D rop { ° F ) F ilm C o efficient, (B T U /h r -ft^ -° F )

17. 7 M o l % M ethanol 46. 5 M o l % 6 1 .2 M o l % 83 M o l %

10 — 220 375 258

15 - - 225 335 269

20 231 310 278

25 -- 240 292 283

30 * » 252 279 290

CO
O '
 1 2 3 -4 5 6 7 3 9 1 2 3 4 5 6 7 3 9 10
10

COEFFICIENTS DF
A T ' C O N ST A N T C

LIRE ACROSS TFE FILM

% ACETONE IN W

not

TT

U_

t t ^ M V V
• — Singly Coinportetr

10 20 30
TEMPERATURE DROP ( ° F )
oo
^4
 O
l 01 si 0 0

F l'L M C O E F F IC IE N T .

(O

; ;£ <7>rn
■ ■■■! s » r l v
: HH ! ^ H
T i l -S '-^ r
:? ? ! !? =o
5? I#:'. ■*!

^: S T-n-

TEMPERATURE
ro
D R O P (°F )
o

oi
o
Ol
 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 0 0 10
10

FILM COEFFICIENTS OF HEAT

8
IT ; ATI CONSTANT COMPOSI
7 :___ . versus _ _____
u. TEMPERATURE OROF ACROSS
6 57 . ,A£ET&NEr4N
(M
5

X
4

ul o
2 ■4T* . V b f Tw ) : —
f T = (T dp- r w) :
•«T= ( T y- T,w ) i :
4 S in g le C o itn p o n e n ls

10 20 30
TEMPERATURE DROP ( ° F )

00
vO
TEM PERATURE DROP ( * F )
 O
D '0
flL M COEFFICIENT, h,| (B T U /H r - Ft -°F

TEMPERATURE
DROP
(°F )

16
TEMPERATURE DROP (° F)
 1 2 3 4 - 5 6 7 3 3 1 2 3 4 5 6 7 B 3 1 0
1

F i m COEFFICIENTS OF HL.
a h AT • CONSTANT-COMPOS'
- ■ t - - versus : ' ' :* !
TEMPERATURE DROP ACROS
l | 3 M O iJ% M ETH A N O U IN B
C O E F F IC IE N T ,M B T y /H r-

m
m. —

er
'FILM

t * . JT v - t w )
Single; Components

10 20 30
TEMPERATURE DROP ( ° F )
vO
oo
drop across the film based on the th ree tem p eratu res m entioned above

was made fo r each series of runs at a constant com position (See Fig«

Nos* 14-20)* A gain it seems that the plot should be a straig h t lin e

w ith ap p roxim ately the th e o re tic a l slope of *0* 25 on lo g -lo g paper

( ■4 13f \
f- J is believed to be alm ost independent

of tem p eratu re just as w ith the pure substances as discussed in the

theory section* Th is w asfound to be g e n erally the case fo r the co­

efficients based on the boiling point* The best lin e representing the

experim en tal data was then plotted* F ro m this line* the heat tra n s fe r

coefficients fo r a tem p eratu re drop of 10°, 15°# 20°* 2 5 °, and 3 0 °F .

w ere read and recorded in Table No* 111*

D iffic u lty was encountered at low concentrations of m ethanol

in benxene in getting a smooth film o f condensate* A t a concentration

of 44* 5 mol% m ethanol in benzene a p e cu liar type of condensation

was noted (F ig u re 21)* A t low te m p e ra tu re drops across the film , a

rin g appearance was noted in the film o f condensate* These rings

o f condensate w ere u n ifo rm ly spaced as shown. They seemed to move

gradually toward the center o f the condensing surface* As the ra te of

condensation was fu rth e r reduced, the u n ifo rm ring fo rm atio n was d is ­

rupted and a random uneven p attern was noted in the film (Fig* No. 22).

It is believed that this n o n-u niform film at the lo w er rates o f condensa­

tion accounts fo r the curve upward in the determ ination o f the heat

tra n s fe r coefficients fo r 4 6 .5 m ol p e r cent methanol in benzene* For

 I

F IG U R E 21

R IN G A P P E A R A N C E IN F I L M O F C O N D E N S A T E A T LO W C O N C E N T R A T IO N
O F M E T H A N O L IN B E N Z E N E A N D IN T E R M E D IA T E R A T E S O F C O N D E N S A T IO N

<_n
 FIGURE 22

IR R E G U L A R P A T T E R N OF CONDENSATE A T LO W C O N C E N T R A T IO N
OF M E T H A N O L IN B E N Z E N E A T LOW CO NDENSATIO N R A TE S
 97

high rates of condensation, the u n ifo rm f ilm again appeared. F o r 17, 2%

methanol in benzene, the condition was even worse, A uniform film

could be obtained only at v e ry high rates of condensation, consequently

determ inations at tem p eratu re drops less than 2 0 °F w ere not made.

This same so rt of film appearance has been recently reported by

B r o m le y ^ fo r pure w ater and fo r pure acetone, not a m ix tu re j and

this observation was noted on an apparatus which was constructed fro m

stainless steel.

Plots w ere then made of the film coefficient of heat tra n s fe r

at a constant te m p eratu re drop as a function of m ol frac tio n as shown

in F ig s, 23, 24, and 25. The data of Trachtenberg is here Included

fo r com parison, A com parison of the three sets of curves fo r each

system im m e d ia tely brings out the n o n -s im ila rity . F o r the acetone-

methanol system and the benzene -m ethanol system , there are three

points in common, the te rm in a ls and the azeotropic composition points.

F o r the isopropanol-m ethanol systems only the te rm in a l points are in

common.

In each case the u n ifo rm ity of the curves w ith the coefficient

based on the boiling point is s trik in g . In each case, the s im ila rity

of the curves based on the dew point and the vapor te m p eratu re is evident

and such should be the case fo r the two tem peratures are not g re a tly

d iffe re n t. In a ll three system s, a m in im u m is reached in the c u rv e s

based on the vapor and dew point tem p eratu res. In the case of the
 ENGRAVING -334-3, 10 X 10 TO THE HALP INCH.
WHEN C
’ ftlrtKtNG STATE (-’ Oi.OP, orAW'MC lb OAI-.-D

FILM COEFFICIENT, h, ( B TU /H r- Ft2 -<>F)

04 6l
CJ1 tn

cr
■rSS-
o '1®
2</>
H I "H
- ’_ 0 -

CO
'H1 m
cr'
m .&
H o

■t

0,

Tl
■ 00
 FILM COEFFICIENT, h, ( B T U /H r - F t 2 - ° F )

ot 01
m cn

Q JO1

t o fO O
3 --- 1
m
o T|
k l
m
m z
05
'N — i\r~
n> -35H-----
CL
-ir 05
aj >
~pp=r

jsr
~n -n
- 3 -

OJ
KEUFFEL
& ESSER
CO,
 100

Fi£DBE._25j
F.JLM-. COEFFIGIF W S .... OF
XEMPFtRAT.UH£ o ro r
versus
C O M P O S IT IO N
[M e th a n o l- Isopropanol System )

COEFFICIENT I ON THE c o e f f ic ie n t Ea s e d o n t h e COjEFflcjENfr ^ASEtj)

BOILING POINT d£w ; p o in t VAPOR TEMPERATURE
( Tv- t J
dp ;v
500
CO EFFIC IEN T, h, (B T U /H r-F t - ° F )

450

400

350
FILM

300

250

200

ISO
40 60 100
KtUFFHL s. E S S E R CO N Y
MOL % M ETHANOL
isopropanol-m ethanol sysfcern3 the curve i® un ifo rm and approaches

the m in im u m gradually* A check of the tem peratu re-co m p osition

plot fo r isopropanol shows that there is a re la tiv e ly s m a ll change in

te m p eratu re between the saturated vapor and liquid compositions*

^he higher saturated vapor tem p eratu re accounts fo r the dip in the

curves plotted* The ac©tone-methanol tempo ratu re-co m p o sitio n plot

shows a w id er v a ria tio n in the tem p eratu res of the saturated vapor and

liquid at a given com position except near the azeotropic point* T h ere is

a pronounced m in im u m a t ap p ro xim ately 80 m ol per cent m ethanol

but none at compositions h igher than the azeotropic composition* The

la tte r is due to the fa c t that the d ifferen ce between the saturated vapor

and liquid tem p eratu res between zero and twenty m o l p e r cent m ethanol

is nil* A glance at the plots fo r the m ethanol-benzene system shows

m ore pronounced m inim um s on e ith e r side of the azeotropic com position

fo r coefficients based on the dew points and boiling tem p eratu res*

Again* these m in im a occur at points of g reatest te m p e ra tu re differen ce

between liquid and saturated vapor*

In o rd e r to determ ine the significance of these curves* it

must be re c a lle d that a plot of the group on

of composition was made fo r the ethanol-w ater* acetic acid-benzene,

and e t h a n o l -m ethanol system s(Fig* No. i>). These lines* though not

a ll approaching lin e a rity * did not show in any case m o re than one point

of inflection* It seems that we could expect s im ila rly shaped curves

 102

FILM COE^FI (j)F HE<Kt TRANSFER

AT CONSTANT RAfrURE i DROP
BASED INC POINT
versus ,

ACETONE■jMEjrHANOL BENZfNE- ^ E JHANOL IsOPROPANfrL- METHANOL

COEFFICIENT, h, (BTU/Hr-Ft -° F )

I T =ip ° r
FILM

20 40 60 80
K E U F F E L & E S S E R CC M V.
WEIGHT PER CENT METHANOL
 103

fo r the heat tra n s fe r coefficients as a function of composition at a con­

stant tem p eratu re drop across the film since only the tem perature drop

across the f ilm would be appreciably affecting the values. If only the

te rm in a l points and the aseotropic composition are considered, such

a curve does result* C e rta in ly , the curve based on the boiling point

approaches these expected results closer than those obtained using the

dew point o r vapor te m p eratu re as a basis fo r calculating the heat

tra n s fe r coefficients.

Since the plots of th e rm a l conductivity, density, and viscosity

seem ed to approach lin e a r ity when plotted as a function of weight per

cent ra th e r than a function of m ol p e r cent, the same scheme was

c a rrie d out using the film coefficients of heat tra n s fe r based on the

boiling point {Fig. No, 26), This change did serve to smooth out the

dip somewhat except in the case of the acetone-methanol system. In

the case of the isopropanol-m ethanol system we have lines which ap­

proach lin e a rity closely, p a rtic u la rly at the higher rates of heat tra n s ­

fe r . This system is one that approaches id eality.

I t must be pointed out that a sm all e r r o r in the vap o r-liq u id

equilibrium diagram w ill cause m ore Ir re g u la r ity in the lines re p re ­

senting heat tra n s fe r coefficients fo r sm all tem perature drops than fo r

the la rg e r ones fo r obvious reasons. Thus the effect of composition

on the coefficients should be b e tte r shown fo r the constant tem perature

drop of 30 F ,
 1

F o r th© benzene-m ethanol plot, a sm all upsurge in the lin e is

noted at ap p roxim ately 20 weight % methanol* This probably stems

fro m the peculiar condensation noted previously w herein the film was

not of u n ifo rm smoothness resulting in higher values fo r the coefficients*

^he acetone- m ethanol plot has the same appearance fo r low

m ethanol concentrations and in addition, a downward bend is noted in

the neighborhood of 70 weight p e r cent. The reason fo r this is not

known, but it m ust be rem em b ered that the v a p o r-liq u id eq u ilib riu m

data was somewhat of questionable value* How ever, these variatio ns

are not v e ry fa r fro m a straig h t lin e connecting the values of the co­

efficients of the single components.

 C H A P T E R V II

CONCLUSIONS

It seems c le a r fro m the w ork done on pure compounds and

the s ta tis tic a l analysis of the data, that a w orkable apparatus has

been designed. The d esirab le features of iso th erm al conditions along

the length of the condensing surface and an im proved technique of

therm ocouple in s tallatio n have been successfully incorporated into the

apparatus.

The w ork w ith b in a ry solutions has evolved a workable p ro ­

cedure to be followed in determ ining film coefficients of heat tra n s fe r

as a function of composition* It appears that fo r id eal systems, the

coefficients at any com position m ay be determ ined by plotting that

fo r the pure components at a given tem p eratu re drop at aero and 100

weight p e r cent and connecting the two points by a straight lin e . The

coefficient at a given weight p e r cent then can be read sim ply fro m the

plot. This conclusion is not fin a l, ra th e r fu rth e r evidence fro m s im ­

ila r investigations on solutions approaching id e a lity is needed* F o r non­

id eal solutions* tru e values fo r the film coefficients do not seem to be

so e a s ily obtained* True* in this case the v a ria tio n fro m lin e a r r e la ­

tionship is sm all; and in the absence of experim ental data a procedure

m ight be followed s im ila r to that fo r ideal solutions without too much

e rro r* H ow ever, it is believed that solutions containing w ater as one

105
 106

component should not be handled in this m anner, since the v a ria tio n of
, ( a s p f\y +
tne group ^ '“^ C f J composition calculated fro m existing data

though m eag er was seen to v a ry non^uniform ly w ith composition* This

seems to be borne out by the findings of W allace and Davison, scant

though they are* W ater is singled out here since it is such a highly

associated substance a® pointed out by E w e l l ^ .

F u rth e rm o re * it appears that the tem perature drop across the

film of condensate is the differen ce between the surface tem perature

and the boiling point of the condensate fo r these p a rtic u la r operating

conditions. It m ust be rem em bered that this is a lim itin g condition

as opposed to a case in which only a sm all frac tio n of the entering

vapor is condensed* F o r the la tte r case, the use of the boiling point

as the in te rfa c ia l tem p eratu re w ill not necessarily be c o rre c t since

other resistances to tra n s fe r enter into the p ictu re.

It seems reasonable to expect that the film coefficients of

heat tra n s fe r could be predicted fo r b in ary solutions that fo rm a single

liq uid phase by the N usselt Equation when the necessary physical data

become available*
 C H A P T E R V III

R EC O M M E N D A TIO N S

F u rth e r w ork seems to be in order on both ideal and non-ideal

solutions, w ith w ater as one component p a rtic u la rly of im portance*

Th is w ork has served to point out the d e s ira b ility of having

values fo r density, visco sity, and p a rtic u la rly th e rm a l conductivity

as a function of composition and tem perature* It is believed that this

equipment m ight be used to study the value of the vapor side film

te m p e ra tu re as a function of the degree of p a rtia l condensation when

this data becomes availab le.

It is believed that v a p o r-liq u id eq u ilib riu m data should be de­

term in ed fo r the reagents actually used in the heat tra n s fe r w ork, as sm all

te m p eratu re e r r o r s (even 1 o r 2 ° F ) are im po rtant p a rtic u la rly at the

lo w er te m p eratu re drops across the film .

Slight changes in the apparatus should be made to ease the

operating technique as w e ll as to im prove the results obtainable. The

I
vapor jacket probably should be enlarged to m inim ise adverse vapor

velocity effects. Steps fchould be taken to provide a b e tte r control of

the heat input. Cleaning of the condensing surface should be made ea sie r,

o r the condensing surface should be made less lia b le to tarnishing p e r ­

haps through plating on a thin film of no n -co rro sive m eta l. An ea sie r

107
 108

m ethod of m easuring the condensate should be devised, perhaps a vo l­

u m e tric system which would be alm ost automatic* The condensate

re tu rn lin e should be brought to the center of the b o ile r to prevent a

concentration gradient across the b o ile r fro m one end to the other*

P erhap s, the te m p eratu re m easurem ents m ay be speeded by connecting

a ll the surface therm ocouples in series and reading a total voltage

generated, then dividing by the number of thermocouples to get an

average value. This would allow in stallatio n of thermocouples at other

strateg ic points, since the lim ita tio n has been the tim e necessary to take

the readings* Separate readings fo r each surface thermocouple do not

seem to be so im p o rtan t fo r subsequent w ork since the va ria tio n

around the c ircu m feren c e has been found to be ra th e r sm all*

 SELECTED
B IB L IO G R A P H Y

(1) B ro m le y , L . A * , j i t *L* " H e a t T ra n s fe r in Condensation,11

In d u s trial and Engineering C h e m is try , 44 (1952) 2962*
ZfSFf

(2) B row nlee, K , A . , In d u s tria l E xp erim en tatio n , L>ondon: His

M a je s ty 's S tation ery O ffice, 1947, C hapter IX ,

(3) Chu, J , , D is tilla tio n E q u ilib riu m Data, Reinhold Publishing

Company, New Y o rk , 1950,

(4) Colburn, A , P , * and D rew , T . B . , " T h e Condensation of M ixed

V a p o rs ," Transactions A m e rican Institute of C hem ical
E n g in ee rs, 33 (1937), 197*213,

(5) Dodge, B . F . , C h em ical Engineering Therm odynam ics, M cG raw -

H ill Book Company, In c ,, New Y ork, 1944,

(6) E w e ll, R. H , , _et a l, "A aeo tro p ic D is tilla tio n ,>r In dustrial and
E ngineering C h e m is try 36 (1944) 871.

(7) F litc r a ft, R* K , , M . S. thesis, Washington U niversity, St* Louis,

M is s o u ri, 1948,

(8) G autreaux, M . F , , " A Therm odynam ic Analysis of C orrelations

fo r V a p o r-L iq u id E q u ilib ria in N on-Ideal S o lu tio n s," Un­
published Thesis in C hem ical Engineering, Louisiana State
U n iv ersity , June, 1951,

(9) Hagenbuch, W, H * , " H e a t T ra n s fe r fo r P a rtia l Condensation of

Steam and Ethanol M ix tu r e s ," Unpublished Thesis in C hem ­
ic a l Engineering P ra c tic e , Massachusetts Institute of T ech ­
nology*

(10) International C r itic a l Tables, New Y ork, M c G ra w -H ill Book

Company, In c ., 1933.

(11) Johnstone, H. F * , et a l, " H e a t T ra n s fe r to Clouds of F allin g

P a rtic le s , " Transactions of A m erican Institute of C hem ical
E ngineers, 37 (1941), 95.

(12) M cAdam s, W, H . , Heat Transm ission, New Yorks M c G ra w -H ill

Book Company, Inc. , 1942. Chapters V II and IX ,

109
 110

(13) N usselt, W *, " D ie Oberflachenkondensation des W asserdam pfes, M

Z e its c h rlft des V ereines Deutsche r Ingenieure, 60 (1916)
569*575*

(14) O rr, V ,, Vapor * Liquid E q u ilib riu m of N o n -Id e al Solutions, "

Unpublished D is s e rta tio n in C h em ical Engineering! Louisiana
State U n iv e rs ity , 1950*

(15) Othmifer, • D . F . , ’'C om position of Vapors fro m Boiling B in a ry

Solutions, " In d u s tria l and Engineering C h e m is try, 20
(1928), 743*

(16) O th m er, D * F * , W hite, R* E , , "Condensation of Vapors*Apparatus

and F ilm C oefficients fo r Lo w er A lc o h o ls ,11 Transactions
of A m e ric a n In stitute of C hem ical Engineers, 37, (1941), 135*

(17) P a lm e r, G . , " T h e r m a l C onductivity of Liquids, " In d u strial and

E ngineering C h e m is try , 40 (1948), 89-92*

(18) Rhodes, F * H * , and Younger, K* R * , "R a te of Heat T ra n s fe r be­

tween Condensing Organic Vapors and a M etal T u b e ,"
fejfaStrtS* a° d.. Engineering C h e m is try, 27 (1935) 957.

(19) Sakiadis, B. Sa, and Coates, J *, A L ite ra tu re Survey of the T h e r­

m a l C onductivity of Liquids, Engineering E xperim ent Station
B u lle tin Ho. 34, Louisiana State U niversity, Baton Rouge,
Louisiana, 1952.

(20) Sherwood, T . K . , and P igford, R* L . , Absorption and E xtraction.

Second Edition* New Y ork: M c G ra w -H ill Book Company, In c .,
1952. pp. 1-97.

(21) T im m e rm a n s , J , , P hysico-C hem ical Constants of P ure Organic

Compounds* New Y o rk: E ls w ie r Publishing Company, 1950.

{22} Todd, J. B . , J r . , " A n Apparatus fo r M easuring F ilm Coefficients

of Heat T ra n s fe r of Condensing V apo r® ," an unpublished Thesis
in C hem ical Engineering, Louisiana State U niversity, 1950.

(23) Trachtenberg, J* J . , " H e a t T ra n s fe r Coefficients fo r Condensing

Vapors of M ethanol-Isopropanol M ix tu r e s ," Unpublished
Thesis in C hem ical Engineering, Louisiana State U n iversity,
June, 1952*

(24) W allace, J. L , , and Davison, A. W . , "Condensation of M ixed

V apors, " In dustrial and Engineering C h e m is try, 30 (1938),
948-953.
A P P E N D IX
 112

N O M EN C L A T URE

distance fro m w a ll out into film of condensate (ft)

A a re a of heat tra n s fe r (ft2)

B * assumed a constant
B * 3 s C tf ^Vy-^w )

c * concentration (m els /ft**)

C * constant o f integration

CpA’ CpB ~ m ^l&l specific heat of components A and B respectively

(B T U )/(# m o l)(° F )

D ftVg = average d iam e ter of pipe (ft)

Di = inside d iam e ter of pipe (ft)

= outside d ia m e te r of pipe (ft)

Dv 9 m o lecu lar d iffu s ivity in gas (ft /h r )

g I
% = ac celeratio n due to g rav ity (4.17 sc 10 f t /h r )
2
G s m ass rate of flo w ( # /(h r )( ft )

Gm = m o la l rate of flow (# m o l/(h r)(ft2)

h * lo c a l coefficient of heat tra n s fe r (B T U /(h r )(ft2)(° F )

hag, hfcg = enthalpy of saturated gas of components A and B respectively

(B T U /# )

hdL * coefficient of heat tra n s fe r of d irt on liquid side of the pipe

(B T U /(h r )(ft2M0r )

- heat tra n s fe r coefficient of d ir t on the vapor side of the w all

(B T U /(h rM ft2)(° F )
 113

k L» * co efficient of heat tra n s fe r of liquid

( B T U /h r - f t 2 - ° F )

hlA* hlB = enthalpy of saturated liquid of components A and B

resp ectively

hm s m ean coefficient of heat tra n s fe r ( B T U /h r -ft^ -° F )

hv * vapor side film coefficient of heat tra n s fe r, (B T U /h r - f t ^ - ° F )

s dim ensionless facto rs fo r heat and mass tran sfer respectively

k = th e rm a l conductivity ( B T U / h r - f t ^ - ° F /f t )

= m ass tra n s fe r coefficient (# m o ls )/(h r-ft^ -a tm )

k8
L a thickness of the pipe w a ll (ft)

M = m o lecu lar weight

S ra te o f m ass tra n s fe r of m ore volatile component (# m o ls /h r)

na

P s p a rtia l pressure (atm )

P s to ta l p ressure (atm )

= log m ean pressure of less vo latile vapor (atm )

pB M

Pi = p a rtia l pressu re at in terface (atm )

Ps
s p a rtia l pressure in m ain stream (atm )

q - rate of heat tra n s fe r (B T U /h r )

R a gas constant ( ft^ -a tm )/ (#m o l~°R )

r = radius of pipe (ft)

*t
3 to ta l resistance

S s shearing stress (# f o r c e /ft2)

T 3 absolute tem perature ( ° K)

3 film tem perature * t y - 3 /4 £ t , where At is tem perature

H
 114

drop across the film ( ° F )

tj » te m p e ra tu re at in terface (° F )

tv a vapor te m p e ra tu re ( ° F )

tw * te m p e ra tu re of condensing surface (° F )

u = downward ve lo c ity of the film (( f t ) /h r )

U0 * o v e ra ll coefficient of heat tra n s fe r based on outside are a of

pipe ( B T t l/( h r ) ( f t 2)( ° F )

U * o v e ra ll coefficient of heat tra n s fe r (B T U /(h r )(ft^ )(° F )

w : m ass ra te of tra n s fe r (# /(h r )( ft2)

xa s m o l fra c tio n of m ore v o la tile component in liquid phase

y s m o l fra c tio n of vapor, yv in m ain vapor stream at interface

Y - f ilm thickness at a given point (ft)

z = distance in d irec tio n of diffusion (ft)

z i s a function of film thickness

p* - mass ra te of flow over a surface per unit length ( lb /h r - f t )

di s te m p eratu re d ifferen ce ( ° F )

A * latent heat of vap o rizatio n (B T U /# )

Am * average xnolai latent heat fo r solution (B T tf/# m o l)

A /L = coefficient of visco sity (M a s s /u n it length - unit tim e ) « A if

s viscosity in fo rce units (F o rc e *• T im e /L e n g th )

s density (# * s /ft3) or consistent units

(p z angle of surface w ith reference to h o rizontal plane (rad ius)

\p ~ a function of film thickness in substitution

f s re fe rs to average film properties

 115

C alculation of Confidence L im its According to Brownlee^

T h e re is reason to expect that heat tra n s fe r coefficients can

be c o rre la te d fo r a given substance or a given composition of a m ix ­

tu re by the follow ing relatio n :

log h = log k - m log AT

where h is the heat tra n s fe r coefficient, k and m are constants, and

AT is the tem p eratu re drop across the film .

To test the significance of this lin e a r relationship, the c o rr e ­

lation coefficient, r c m ust be calculated as follows

£ fAT-AfKh-M
C \j £(* T - ar )* -£ ( l i - T i ) *
where 21 denotes summation and

- / ,-r l^ ( . i . h ' T ) *'

2(AT-AT) * 2 (AT )- —

n = num ber of observations.

Z(h-K)*- t ( h l ) -

- / 1 - T ~\ (h)
2(AT-AT)(h-K) • Z [ h * r ) ^

By use of Table IV p. 150, the p rob ab ility of getting such a value is

determ ined. If the prob ab ility is high then there is good reason to

suspect that the lin e a r relationship exists. F o r the determ ination

on m ethanol, the co rrelatio n coefficient is 0. 9499. The prob ab ility

 116

read fro m the t&ble of g e ttin g this value in the absence of the c o r r e la ­

tio n is n il* consequently the use of the relationship is sound.

T h e equation fo r the lin e is calculated according to the method

of le a s t squares. The calculation fo r methanol gives an equation

log h * 2.994566 «* .27235 log \T .

T h e standard deviation, Cr m ay he calculated as follows

T h is standard deviation m ay then he used to draw the lim its in which

a given percentage of the data should fa ll. F o r lim its in which 95%

of the data w ill fa ll, use T a b le 1, p. 144, to determ ine a value fo r jt.

F o r m ethanol the standard deviation was found to be 0.00993134.

The value of t is 2 .1 to give a deviation fro m the line of - . 020856

lo g arith m ic units, thus establishing the confidence lim its .

 CO.
&ESSER
KEUFFEL
O r < jlM m m

ig lH A

vl± +i hr-;

Saturated

t -ju IL li: .'i.i.i-j- x

, :n '

PER CENT ACETONE

‘ uajtfd s«( fioot
... ] J * ; w i«u Ml s a w
'M 3<tV d n H T D V y i HO ~ S W jH lr tiq 'a Q lQ O J i l V i S 9 H IW C U 0 NSW
•HoNf ?nyH'.3Hi p i Oi’ X..OJ 'e ^ s e $Niwyt>N3
 ENTHALPY (B T U /P O U N D )

£
O
II ^
51 in
If 33
n , o
g.i<£ m

- ;i 3
■ in *
m
P H
11 X
>
2
11 O

•fmH

K.EUFFEL & E S S £R CO. N. Y.

. 811
 fH.i

iiiis

HE

M
fritra:
 120

■r:

K:

rtn

Hti
ir:

;:ti;n

m
 121

U N e ; PQtw
opqno
rr t

¥¥

3iii

tiHi

ri:
l -rm n i

44

METH^NOg £
 A U TO B IO G RA PH Y

James B . Todd* J r . , was born January 17* 1924, in

F ra n k lin , Louisiana. He attended public schools in E ast Baton

Rouge P a ris h and graduated fro m Baton Rouge High School in 1941.

H e was en ro lled at Louisiana State U n iversity, Baton Rouge, Louis­

iana, fro m June, 1941 to June, 1943 at which date he entered the

A rm y of the United States. Upon discharge fro m the A rm y in Sep­

te m b e r, 1946, he again enrolled at Louisiana State U n iv ersity and

was graduated in June, 1948, with the Degree of Bachelor of

Science in C hem ical E ngineering, He received the M a s te rfs Degree

in C hem ical Engineering in 1950 fro m Louisiana State U niversity.

H e continued graduate studies at the same institution until August,

1951, at w hich tim e he re -e n te re d the A rm y to serve as a Gunnery

In s tru c to r in the A rm y F ie ld A r t ille r y School until Novem ber, 1952.

He then re -e n te re d the U n iv e rs ity and at present is a candidate fo r

the Degree of Doctor of Philosophy,

122
 EXAMINATION AND THESIS REPORT

Candidate: James B. Todd, Jr.

Major Field: Chemical Engineering

Title of Thesis:
The Effect of Composition on Heat Transfer Coefficients for
Condensing Binary Vapors with a Single Liquid Phase.
Approved:

Major Professor and Chairman

uate School

EXAMINING COMMITTEE:

Q i,

Date o f E x a m in a tio n :