—._—- -v __

_._'

 

\ '2C—‘“— —____

THE DESIGN OF A GUARDED
RlNG HOT PLAT-E FOR TESTING
T'Hi THERMAL CONDUCTIWTY
OF HOMOGWEOUS MATERIALS:

; than for the m of M. h
3 ? mama STATE com
. , Jam“ Trout Andaman
‘ 1943

11111

 

vu. . .n.‘ , . , r I
KYMkQ‘Et‘h‘...‘

E: I

.OJ

- . ..,.ln A.|\v.

0“ '1
1%; ’h¢.H 1' I Y a

e
.m
m
.m
m
.u
m
T

thesis entitled

he boom of n Wm not Pilate for
Testing moral Conductivity of W hteriale.

presented by

has been accepted towards fulfillment

of the requirements for

 

Major professor

DATE DUE

61

12397

m

m5

aw

mm

flu...

mm

mm

8

mm E

om U

mm D

mm E
T

.. ...

E

N

H

m

w

......

mm

m.
m
m...
m
0
Y
w
..W
t
u
0
a
e
.m
.5
.m
.v.
o
m
m
.0..
x
O
B
m
m
m

 

0*! fix.) .4 . 1. .-...i..- .. .I...‘ . . u}... 1 .1. ...... I... 1...... -..- ..1 .. - . . . . 1 1 . .. . .. . . . . . . . .
-0 It -‘c YT, .w -..-(.0.... ......r... ..a. ti. ’Junt... .... 0......1... n.r4‘ ...Ao..~vv1h . N‘1Thfuvqmir NIH. TIP-II...) . . ...A .. ... .. . . I . . . 1. . .LIZ .. I 19?... .0.$...I.l‘u. VIrx/VII .970. .. ... «J... n , . - V 1.
' V . . I. a . . . . .. I . . . \ .. .
...! dI. q... .. y. ‘ I..1‘ .... .v r.. . . 0 . . , .
| J...’ \— .9..I.N )I‘IMI‘IfD\ I130 l.“‘?... ’ a .w . v ... . . .. . \
A . .. \ 0. . . .. b n I}... .u .I. . ‘I.. ,.... . ,1.. . .. \1 D, \ a I
.p ..1 L‘. ..‘4 Q... t... . ..r . . L‘Mu’fl“ h.fi‘.. .rVaYU.J.L.‘. l}: \..) 341k)? 2 )fi-... '1. b .. KI”)! . - p. o ... . ...0. . ...l-
. . . v . Is .. . .. .. . . I. . .. .. 10‘. a . 1n. 0.. .. .1 1 . u L .01

......Q...I.....11I.Jhrfi...!vn., I .Fo AVAW 1.1.13 (....-. n 4W ..

 

(h‘. .“V Kuhn“. Fwy. .....JoI ......i...H.,‘N..JM~.I...4.w

 

_THE DESIGN OF A GUARDED RING HOT PLATE FOR TE STING THE
THERMAL CONDUCTIVITY OF HOMOGENEOUS MATERIALS

By
JAMES TREAT gymsom

A THESIS

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Mechanical Engineering
1948

THESIS

ACKNOWLEDGMENT

The author wishes to express his appreciation for
the advice and assistance of Professor Lorin G. Miller,
and for the help given by the other members of the
Department of Mechanical Engineering of Michigan State

College.

2021.45

TABLE OF CONTENTS

Introduction

Historical Background
Thermal Conductivity
Description of Apparatus
Guard Ring Principle
Standard Test Method Specifications
Heating Element

Power Measuring Meters
Power Supply

Guard Ring Power Supply
Vapor Proof Hood
Dehumidification Apparatus
Angle Iron Clamp Bracket
Selector Switch

Connection Board

Multiple Contact Connectors
Thermocouple Cold Junction
Coolant Supply

Coolant Pump

Galvanometer Shorting Switch
Cold Plate Construction
Sample of Test Data

Sample of Test Computations

Bibliography & References

11

Page

10
11
14
16
26
26
27
27
29
29
3o
30
32
32
34
34
35
35
36
37
38
40

TABLE OF ILLUSTRATIONS

Principal Forbes Bar Test

Preston's Guard Bung

Lees Disc Apparatus

Schematic Cross Section of Guarded Hot Plate

Enlarged Cross Sectional Views of Heat Flow
Lines in Edge of Test Section of Guarded
Hot Plate

Guarded Hot Plate Electrical Power Circuit

Schematic Diagram of Chemical Dehydration
System

Circuit Diagram For Common Cold Junction
Conductivity Plot

Temperature Plot Through Enlarged Cross
Section of Guarded Hot Plate

111

Page

12

15
28

31
33
39

4O

INTRODUCTION

Until the year 1945 all testing work on the thermal
conductivity of insulating materials was done in a some-
what haphazard manner, with each investigator deciding
for himself what the test conditions should be, and as a
result the information available in tabular form neces-
sarily had to be accompanied by explanatory notes as to
the conditions. In addition, it was left up to each
investigator to set for himself his own limits of accuracy.
All in all, this practice led to widely varying data for
materials seemingly the same.

However, in 1945, the American Society for Testing
Materials, in conjunction with the American Society of
Heating and Ventilating Engineers and the American Society
of Refrigeration Engineers, deve10ped and published a
standard method of testing, which should eliminate the
aforementioned difficulties. In addition, these organiza-
tions instituted a testing program for the purpose of re-
vising all available data. To insure uniform and accept-
able testing methods and equipment in the 000perating
laboratories, a standard sample of corkboard was sent to
each laboratory to be tested under the new method. The
results from this test along with the corkboard sample
were sent to the United States Bureau of Standards for

retesting and checking. Any laboratory whose test results

checked within three per cent of the Bureau of Standards
is to be accredited, and will proceed with testing for
data revision.

This standard methoa of test merely set up condi-
tions and limits for testing and did not specify in any
way how these conditions and limits were to be accomplished.
In this writing, it is hOped that a satisfactory method
for accomplishing the prescribed test is presented in such
a way that anyone wishing to set up similar equipment will
be able to do so without having to have a specialized
background in this type of work. As will be seen though,
a step by step rigid development is not used, but discus-
sion is employed, and latitude of method is left up to
the individual.

HISTORICAL BACKGROUND

Ever since Bernoulli1 conceived the molecular motion
theorem of heat in 1738, men have tried in various ways
to evaluate the amount of heat transmitted under given
conditions, by conduction. Among the first of these
experimenters was Count Rrumford,2 who, in 1804, set up
an iron rod with cepper cups soldered to the ends of it.
In the rod, at equal intervals, he drilled small holes
for the placing of thermometers, and in the COpper cups
he put crushed ice and boiling water, respectively. He
was trying to prove that conduction obeyed the same law
as radiation, and succeeded in merely finding that tem~
perature distribution in a uniform, homogeneous bar was
essentially uniform.

The next experimenter of note was Tyndall3 who set
up a relatively complicated apparatus for measuring rela-
tive conductivities. Roughly this consisted of an elec-
trically heated, mercury filled pad butting up against
one side of a small cubical sample, with another mercury
filled pad butting up against the Opposite side. The re-
ceiving pad had a thermocouple built in, connected to a

galvanometer. In Operation the heating pad was turned on

 

1. Tyndall, John, Heat As A Mode 0;;Mopggn, D.
Appleton & Co., 1890, p 119.

2. Thompson, Benjamin (Count anford), The WOrks_Q£
Rumford, American Academy of Arts & Sciences, Vol. II,
p 1 .

3. Tyndall, 0p. cit., p 245.

for exactly sixty seconds; then the galvanometer switch
on the thermocouple was closed and maximum deflection
noted. Of course, as Tyndall mentioned, this method did
not take into account the specific heat of the sample,
but it did have some value in determining relative con-
ductivities of wood across grain and with the grain.

The first really quantitative experimentation was
accomplished by Principal Forbes4 in 1850. Forbes setup
a long, homogeneous bar, exposed to the atmOSphere with
thermometer wells drilled at regular intervals. (See
Illustration Page 5.) One end of the bar was heated to
a constant temperature and steady state conditions main?
tained while temperature readings were taken. The same
bar was then heated its entire length to the initial tem-
perature, and time-temperature readings for the cooling
period, taken under the same atmospheric conditions as
the initial test, were made. Now with a knowledge of the
specific heat of the bar at all these temperatures, Forbes
computed the heat loss per unit length of the bar for any
temperature per unit length of the bar for any temperature
per unit of time. By taking any given cross section and
making a summation of the heat losses from the bar beyond
the cross section, he could say this summation must be
the heat flow through the cross section and establish the

thermal conductivity of the bar.

 

4. Forbes, Principal, Transactiong of the Royal
Society of Edinbur , 1861-1862.

K<m .rmuF 0 ....0m

hmm» mam mmmaoa ann_oz.en

o...u...m /

 

 

 

 

 

 

 

 

/ _ 1
a

 

 

 

 

 

 

 

_ \\
a X g \
-\

acupuaoaauz»

 

 

 

I...

In 1888, J. C. Maxwell5 claimed the obvious way to
measure thermal conductivity would be to form a test speci-
men into a uniformly thick plate, and bring one of its
surfaces to a known low temperature and the other surface
to a known high temperature and determine the quantity
of heat passing through in unit time. However, Maxwell
then gave this method up because of the seeming impossi-
bility of measuring the surface temperatures accurately.

6

Six years later, 1894, Thomas Preston described a
very crude forerunner of the present guarded hot plate,
seemingly based upon Maxwell's theorizing, and doing away
with the surface temperature difficulty. Preston started
with a large block sample enclosed on two Opposite sides

by containers filled with crushed ice and steam. (See
Illustration Page 7.) These containers had smaller con-
tainers within them, also facing upon the sample and also
filled with crushed ice in one and steam in the other.

A short distance in from these faces, two holes were drill-
ed for thermometers and the distance between the thermo-
meters was considered the thickness Of the sample for com-
putation purposes. The cross sectional area of the inner
containers was the area for computation and the heat sup-
plied was measured either by weighing the melted ice water,

or condensed steam per unit time.

 

5. Maxwell, J. C., Theory of Heat, Longmans, Green
& CO., 1888, p 268.

6. Preston, Thomas, Theory Of Heat, MacMillan & CO.,
1894: p 527.

Ox.m macaw m.zohmuad

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«IIIJ _
a. l. as a a... T.
A
1 . _l.
l xoch
, amok
g Q. .3. 3.
sawhm a<uhm .“ NT owzmomo ouxmaau
M; a;
-- . L - ----
.t l:
i. :.
:_ _
-F -..llu1lIIliL .F _ -1111 1‘...
m w
_ m
l l
_
;

mcupuzoaauxk

A more sensitive apparatus was devised by Lees7 in
1898. It had a series of three COpper plates with an
electrical heating coil sandwiched between plates one and
two, and the sample between plates two and three. (See
Illustration Page 9.) By measuring the power input to the-
coil and being careful about the air movement about the
pile, it was possible to calculate the surface coefficient
"h." Then by equating heat into the sample to heat out
of the sample, thermal conductivity was calculated.

The first experimental equipment built along the
lines Of present day equipment was described by Poensgen8
in 1912. It had a central heating plate, with guard ring,
flanked by cold plates supplied with cold water. The
whole was buried in a box of sawdust, and temperature
readings were taken by thermocouples.

In 1920, according to Robert Lander,9 professor at
the University of Minnesota, the generally accepted guard-
ed hot plate design was evolved and constructed. This hot
plate has proven successful and is the instrument used in

determining the thermal conductivities for insulation

given in references available at this date.

 

7. Poynting, J. H. and Thompson, J. J., Heat, A
Textbook of Physics, Charles Griffin & Co., 1904, p 102.

8. Saha, M. N. and Srivastava, B. N., A Textbogk
of Hggt, Indian Press Ltd., 1931, p 316.

9. Lander, Robert, Factgrs Affecting Thermal Condug-
tivit , University of Minnesota Technical Paper No. 49, 1944.

mDF¢m<ac< om.0 wuu4

I .< meF18U1§UF dekmzou

 

 

:zzrlts: :tii! {t ;J
on.o cuaaou

I'll

 

uaaaam pmu»

 

om.o cuaaoo

 

,4.oo u:.»<u: 4<omeou4u

 

 

 

 

"0."llul|-0llii'uo) - 4|

on.o cunaou

- II‘ 1 ... ..‘OIO.I‘ )i .10 -.‘l-ll.‘ .I- a I .01 .. A. Jul. { v-0.l. A: li'llllllllln ‘Inlll'lL

c.< maah<cunamh pz<rmzoo

 

THERMAL CONDUCTIVITY

Thermal conduction is defined as the transfer of heat
energy, in which the molecules of higher kinetic energy
transmit part of their energy to adjacent molecules of
lower kinetic energy by direct molecular action.10 Because
the kinetic energy of the molecule is directly prOportional
to its temperature the heat transferred will occur in the
direction of decreasing temperature. In the case of pure
conduction, the amount of heat transferred is affected in
linear preportion to the mean temperature of the material
in question. In other words a plot of thermal conducti-
vity versus mean temperatures would be linear. Present
day accepted methods of insulating is to trap small air
spaces in highly thermal reelstant material, giving a
cellular or granular insulation, which does not conform
to pure conduction transmission, but rather combines con-
duction with radiation across the minute air spaces.
Radiant energy transmission is prOportional to the fourth
power of the absolute temperature and therefore, when com-
bined with the linear characteristics of conduction, will
give a non linear variation with changes in mean tempera-
ture. Again referring to thermal conductivity versus mean
temperature it will be seen that now with increasing values

of mean temperature, thermal conductivity will increase at

 

10. American Society of Heating and Ventilating
Engineers, Guid , 1948, p 97.

10

a rate greater than linear. This effect becomes more
pronounced with the increased amount of entrapped air
space and the higher mean temperatures, and consequently
greater portion of radiant transfer. Due to this phe-
nomena it is not possible to simply determine two values
of thermal conductivity and extrapolate for desired values
beyond the temperature range covered by the two tests.

It is necessary to conduct a number of tests to definitely
establish the variance of conductivity with mean tempera-

ture over any given range of desired mean temperatures.
DESCRIPTION OF APPARATUS

The guarded ring hot plate consists of a central
electrically heated square flat plate with two heavy cOp-
per facings. This central plate has an electrically heat-
ed guard ring plate, in the same plane, surrounding it,
and separated by a small air gap from it. (See Illustra-
tion Page 12.) On the two faces, uniformly thick slabs
of homogeneous test material are placed, completely cover-
ing both central plate and guard ring. Now, two so called
cold plates are placed on the outside surfaces of the test
slabs; the whole is clamped together forming a large sand-
wich. These cold plates are liquid cooled copper plates
capable of conducting the heat away from the sample as
fast as it is supplied by the hot plate. This entire as-
sembly is surrounded by adequate insulation to keep heat

transfer not normal to the sample to a minimum.

11

 

 

 

 

 

 

 

 

Rune

 

 

when; an:

a a

_ _ l

_ g. .

_ ~ .

c M

R a

a

.

a

E E

E L E L E
T P T P T
A M A H A
L A L A L
P “a P S P
o .. T T o
L 3 S S L
0 E E E 0
C v: T T C

 

 

macaque as zo_»omw mtozu o.+«5m;om

 

fa]. o

. Val A“! bv‘\!3 .\

 

 

uraaa
kmuh
ho:

 

 

 

 

 

 

«2.x ox<aw

 

 

 

 

l2

The electrical circuits of the central plate and
guard ring are completely separate, the central plate
circuit being metered, and the guard ring being unmetered.
The object of the guard ring being to keep the heat flow
from the central plate perpendicular to the test slabs,
all that is necessary is that it be at the same tempera-
ture as the central plate. Electrical input to the plate
and ring is controlled by coarse and fine rheostats.

Temperature readings are all made by thermocouples
conveniently Joined to a selector switch. Temperatures
recorded are; side of test samples facing central plate,
side of samples facing cold plates, and differential tem-
perature between central plate and guard ring. 9

It can be seen from this description that all the
elements of the thermal conductivity equation are directly

measurable quantities:

k = 9L
tl-tg r
Where: k = thermal conductivity in b.t.u. inch per

sq. ft. deg. F. hr.

Q = b.t.u. input to central plate

thickness of test sample

h>
II

total area of both faces of central plate

t1 = average temperature of hot side of test
sample

t2 = average temperature of cold side of test
sample

r = time of test in hours

13

GUARD RING PRINCIPLE

Keeping in mind the principle that heat flow is com-
pletely dependent upon temperature difference, the reason
for keeping the guard ring at exactly the same temperature
as the test plate is readily apparent. All heat transfer
from the edge of the hot plate assembly is thus supplied
by the guard ring. However, the path of the heat flow
through the test Specimen is dependent upon three tempera-
tures, that of the test plate and guard ring, of the cold
plate, and of the atmoSphere surrounding the entire appara-
tus. The relatively large temperature differential between
the hot plate and cold plate will have the effect of caus-I
ing heat to follow a straight line between the plates,
while the surrounding temperature will have the effect of
diverting the heat flow lines either toward or away from
the edge depending upon the relative temperatures involved.
(See Illustration Page 15) This latter effect, of course,
is minimized by the insulation surrounding all edges of
the equipment, but must still be accounted for. Remember-
ing that for accurate results in testing for thermal con-
ductivity by this method, the heat flow must be as nearly
perpendicular to the plates as is practical, the ratio of
width of guard ring to thickness of test sample must be
considered. This ratio is set at one and one-half by the
standard test. A ratio of greater than one and one-half

means that a sample thinner than necessary is used and

14

ou>o5um zo.Ȣanmz. meow

 

 

 

 

 

 

 

 

uhqaa so: ououaaw.uo zo_huum emu» no usau z. muz_4 roam yam: uo mew.) Jazo.poum mmoco oueuaazu
rum» ucab<uuaauh tom emu» scarecusauk :u.z ,
. .uu o . .u .uuo cm. A
u uWeMa wanzam pmuk .lwasram hmuh .
. Ill-I...“ III... I! , All: will. [I
you» ahead 11
\\..\ ., ....33.;.!.l/~
..II\ \\| aIL . rill . III 1!
LPFPIPFFII. \\\\\ _ [IMMMWH411. 1!
\\ xx . I / / /JL
\\ I! /,/I
\\ \ifik QUUQ ON . TI / / l .0“ OUMO cm
on ouNO 00 1‘“ \ ..t'L UF¢JA . cu IOUO ON. / / A HP‘JQ .
\ \ 0.30. , rI/ / //.a pace.
«Ia \ «...: ,r/ /
2:25 \ \ .4 9.2.6 / / /
\ \ . I / / /
i1
\ \ \ \ / / / / a
\ \‘j I / / /
K x K x \ Jr 2, a A K

.u .uuo cs uguxumoapq . ,.u page or uzuzamoap<

15

consequently a greater percentage of error in measuring
the thickness of the sample is introduced. A ratio less
than one and one-half will not provide a wide enough ther-
mal protection border around the test section and thus

the temperature of the hot plate will be affected, putting
the test results in error.

At this point a knowledge of the contents of the
Standard Test Method adOpted in 1945, applying to the
guarded hot plate, would be helpful in understanding the
discussion that will follow. It is herewith reproduced

in its essentials.

"STANDARD METHOD OF TEST FOR

THERMAL CONDUCTIVITY OF MATERIALS BY MEANS
or THE GUARDED HOT PLATEl

A.S.T.M. Designation: C 177 - 45
AdOpted, 19452
***
"INTRODUCTION

1. (a) This method describes procedures to be used
in determination, by means of the guarded hot plate, of
the thermal conductivity of insulating, building, and
other materials, whose conductivities do not exceed the
maximum value for which these procedures are applicable
as specified in Section 2.

(b) Because of the requirements prescribed in this
method as to conditions under which conductivity tests
shall be made, it should be recognized that the conduc-
tivity coefficients obtained will not necessarily be the
values pertaining under all service conditions. As an
example, the method provides that the conductivity coeffi-
cients shall be obtained by test on dry specimens, while

 

1. Under the standarization procedure of the Society,
this method is under the jurisdiction of the A.S.T.M. Com-
mittee 0-16 on Thermal Insulating Materials.

2. Prior to adOption as standard, this method was
publiszed as tentative from 1942 to 1945, being revised

in 19 5.

in service such a condition will seldom be realized.

(c) The guarded hot plate is generally used for
determining the thermal conductivity of homogeneous mater-
ials in the form of flat slabs, and this method covers
the procedure for such tests. It is recognized, however,
that it is frequently desirable to determine the conduc-
tivity of certain materials used as pipe coverings, etc.,
and also materials constituting a wall construction or a
part thereof. For such purposes the guarded-end or cali-
brated-end pipe methods of test and the guarded hot box
method are recommended.

(d) For satisfactory results, the principles govern-
ing the size, construction, and use of apparatus for the
test described in this method should be followed. If the
'results are to be reported as having been obtained by
this method, then all of the requirements prescribed in
this method shall be met.

(e) It should be recognized that it is impossible
in a method of this type to establish details of construc-
tion and procedure covering all contingencies so that it
can be followed by a nontechnical person; and that, on
this account, technical knowledge on the part of those
using this method concerning the theory of heat flow, tem-
perature measurement, and general testing practices can-
not be dispensed with as a result of the standardization
of the method. It is further recognized that it would be
unwise, because of the standardization of this method, to
restrict in any way the activities of research workers in
the further develOpment of new and improved methods.

"SCOPE

2. (a) For practical purposes, this method of test
is limited to the determination of the thermal conductivity
of materials having conductivities not in excess of 5.0
Btu. in. per sq. ft. per hr. per deg. Fahr.

(b) The method is further limited in its application
to thermal conductivity tests between the extreme tempera-
tures of ~50 F. and 1400 F., or between mean temperatures
of approximately 0 F. and 1200 F.

"SYMBOLS AND DEFINITIONS * * *
"APPARATUS

4. (a) It is not intended in this method to include
detailed requirements for the construction or operation

of any particular guarded hot plate for determining con-
ductivity values. * * * Any plate conforming to the

17

limitations prescribed in Paragraphs (b) to (J) will be
satisfactory. * *

Note 2. -- The surfaces of the heater plates and
cooling plates should be finished to as nearly a true
plane as possible and should be checked periodically.
Variations over 90 per cent of the surface should be nil
with maximum variations on any portion of the surface
not exceeding 0.003 to 0.005 in.

(c) In the design of the guarded hot plate for
testing materials for use in any particular temperature
or conductiVity range, due consideration shall be given
to the materials used in the construction of the hot
plate with reapect to their performance at the tempera-
ture to which the plate will be subjected. Consideration
shall also be given to the rate at which heat must be
supplied and absorbed by the heater and the coolers in
designing the electrical resistance and current-carrying
capacity of the heating element. Otherwise, no particu-
lar variations need be made in the design of hot plates
for use in determining the conductivity of materials at
the widel different mean temperatures mentioned in Sec-
tion 2 (b .

(d) Heating units having metallic surfaces shall
have a definite separation or air gap not greater than
1/8 in. between the measuring area of the central surface.
plate and the guard surface plates. These plates shall
have highly emissive surfaces. The separation between
the main central section and the guard section of the
heating units shall not exceed 3/4 in., and this separa-
tion is allowable only if the spacing bars on either side
of the plate separation are of heavy COpper in order to
distribute the heat on the surface plates. In all other
cases the separation shall not exceed l/d in. The test
area shall be calculated from the center of one separa-
tion to the center of the othgr separation across the
central surface of the plate.

(6) The heater of the guarded hot plate shall be
provided with at least two thermocouples on each face of
the central surface plate, and at least two thermocouples,
on each face of the guard surface plate, located at cp-
posite edges. These thermocouples may be read either in-
dividually to indicate any temperature difference that
may exist between the central and guard surface plates or

 

4. In putting a new heating unit into Operation for
the first time, care should be taken to insure that the
two faces of the heating unit maintain essentially equal
temperature throughout the temperature range of the ap-
paratus.

18

they may be connected differentially, and thus indicate
such temperature difference directly. The differential
method of connecting the center-to-guard thermocouples is
the more sensitive, and is to be preferred. When the
thermocouples are to be read individually, they may be
peened, welded, or soldered to the surface plates. When
they are to be connected differentially, it is essential
that they be electrically insulated from each other.

This may be accomplished by installing the thermocouples
in shallow grooves in the surface plates with an electri-
cally insulating cement, in such a manner that the bead
of each thermocouple is in the plane of the surface of
the plate.

(f) The cooling units Shall have the same surface
dimensions as the heating unit. They may consist of
metallic plates cooled by a fluid, electrically heated
plates maintained at temperatures below that of the heat-
ing unit, or thermal insulation applied to the cool sur-
face of the test specimen, depending on the mean tempera-
ture desired.

(g) The edge insulation may be of any convenient
loose-fill or blanket-type insulating material to reduce
edge losses from the heater plate, test Specimens, and
cooling plates. It shall be of such thickness that the
resistance to edge losses shall be at least twice and pre-
ferably three or more times the thermal resistance of the
specimen in the direction of normal heat flow.

(h) The surface temperatures of the test specimen
may be determined either by means of thermocouples mounted
in the hot and cold plates or by thermocouples located in
the surface of the Specimen, depending on the nature of
the material to be tested. For nonrigid materials that
have a unit conductance of less than 1.0 Btu. per sq. ft.
per hr. per deg. Fahr. and that conform to the surfaces
of the plates, the surface temperatures of the Specimens
shall be taken as those indicated by the thermocouples
attached to the hot and cold surface plates. For this pur-
pose both the hot and cold plates Shall each be supplied
with at least two thermocouples mounted in the surfaces
of the plates as described in Paragraph (e). Preferably,
the thermocouples Should be insulated from the plates and
gread differentially between the hot and cold plates, the
surface temperatures of the Specimen being taken as those
of the plate surfaces in contact with it. For rigid
materials that fail to conform to the surfaces of the
plate, and for all materials having a unit conductance
higher than 1.0 Btu. per sq. ft. per hr. per deg. Fahr.,
separate surface thermocouples shall be used. These thermo-
couples shall be mounted on the surface of the Specimen in '
any convenient manner suitable for the purpose, such that

19

the thermocouple Junction is flush with the surface of
the Specimen. When thermocouples are used in the sur-
face of the Specimen, there Shall be placed between the
surfaces of the specimen and the hot and cold plates a
piece of blotting paper or asbestos paper, depending on
the temperatures encountered. For nonrigid materials of
low conductivity, the use of thermocouples in the plates
is preferred; while for rigid materials and those of
high conductivity, the use of surface thermocouples is
preferred. In the intermediate range the choice of
method is left to the Judgment of the Operator.

(1) The thermocouples mounted in the surfaces of
the plates shall be made of wire not larger than No. 23
A.w.g., while those used as surface thermocouples Shall
be made of wire not larger than No. 29 A.w.g.

(J) A potentiometer having a sensitivity of 5 micro-
volts or less shall be used for all measurements of electro-
motive force.

"SAMPLING AND PREPARAI‘ ION 0F SPECIl-ILENS

5. (a) When this method is used as a guide in deter-
mining the thermal conductivity of special samples for
control of manufacturing processes, for determining com-
pliance with purchase Specifications, and for other Similar
purposes, the selection and preparation of samples must
obviously be left to the discretion of the person desiring
the information. When, however, standard tests are to be
made for the purpose of reporting thermal conductivity of
a given material for consumer use, or for any other pur-
pose where a definite statement of the history and condi-
tion of the sample is not available, and in all cases
where the test is reported as having been performed in ac-
cordance with this method, the material shall be sampled
and the Specimens shall be prepared in accordance with

Paragraphs (b) to (h).

(b) The sample shall be selected so as to provide
two Specimens as nearly identical as possible and of such
size as to completely cover the heating unit. The Speci-
mens shall be of sufficient thickness to give a true aver-
age representation of the insulating material to be tested.
Since excessive thickness of the Specimens and consequent
'excessive dimensions of the area of the guarded hot plate
unnecessarily complicate the manipulation of the test pro-
cedure, the Specimens shall not be too thick. The thick—
neSS of the specimen shall be great enough to allow a suffi-
ciently accurate measurement of this dimension for calcu-
lating the thermal conductivity. The relationship between
the maximum thickness of the test Specimen used and the
minimum dimensions of the guarded hot plate Shall be as

follows:

20

Ninimum Linear Surface Dimensions of
Guarded Hot Plate (Square or Round), in.

Maximum
Thickness Central Section Guard Section
of Test or Test Area of or Guard Area of
Specimen, in. Heating Unit Heating Unit
100000 oooooo coo 1'" 1%
léooooooooooooo 8 2%
200.000.000.00012 3
4000.00.00.0000 12 6

(c) The sample shall be chosen as a fair represen-
tative of the material of the particular type on the mar-
ket, or that to be used by the consumer. It should, there-
fore, preferably be purchased on the Open market by an
unbiased person. Three separate samples, not expected to
have originated from the same day's output at the manu-
facturing plant, nor from a Single Shipment therefrom,
shall be thus obtained from three different sources. The
three or four tests prescribed in Section 6 (a) shall be
made on each of these three samples in determining the
conductivity of the given material.

(d) In testing all forms of homogeneous materials,
the surfaces of the test Specimens Shall be made as plane
as possible, by sandpapering or otherwise, in order that‘
intimate contact between the specimens and the plates or
the paper may be effected.

(e) The sample from which the test specimens are to
be taken Shall be weighed in the as-received conditions
and then dried at 215 F. until excess moisture is driven
off as indicated by a constant-weight determination. (If
the material is one that may be chemically affected by
heating to 215 F., the sample Shall be dried in a desic-
cator at from 120 to 140 F.) The aS-received weight, the
dry weight, and the necessary physical dimensions of the
sample for calculation of the density of the material as
tested shall be recorded. (For solid and blanket-type
materials the physical dimensions shall be determined
separately before and after drying. The density of loose-
'fill materials Shall be based on the volume occupied in
the guarded hot plate and the weight after drying.)

(f) Samples of homogeneous solid materials to be
used for test shall be dried in accordance with Paragraph
(6), and the Specimens cut to Size, or molded or pressed
into the prOper Size and shape, with due consideration
to the treatment of the material in question, weighed,

21

and placed in the guarded hot plate for measurement of
thickness, prior to testing, in accordance with Section
6 (c .

(g) Samples of blanket-type materials to be used
for test Shall, after being dried in accordance with
Paragraph (e), be measured for thickness in accordance
with the Standard Methods of Test for Thickness and
Density of Blanket Type Thermal Insulating Materials
(A.S.T.M. Designation: C 167) of the American Society
for Testing Materials.5 Two representative Specimens of
the material cut to prOper size shall be weighed, then
placed in the guarded hot plate and compressed to the
Specified thickness, and the thickness measured in ac-
cordance with Section 6 (c).

(h) Samples of loose-fill materials to be tested
at any prescribed density shall, after drying in accor-
dance with Paragraph (e), be confined in the Shape of a
square or circular slab of dimensions suitable for the
particular hot plate to be used and of a thickness not
less than five times the size of any particle composing
the test Specimen, provided that in no case shall the
thickness be less than 1 in. Two representative portions
of the sample, of such weight (determined to an accuracy
of plus or minus 0.5 per cent) as will give the prescribed
density when packed into Spaces of the required dimensions,
Shall be weighed out and fluffed up. The test specimens
Shall then be prepared in accordance with either method
1 or 2, as follows:

Method 1. -- The guarded hot plate shall be set up
with the required distances between the heating unit and
the cold plates. Each portion of the sample Shall be
divided into four equal parts. One part Shall be placed
in one side of the apparatus and vibrated or tamped until
it occupies Just one-quarter of the volume of that side.
The remaining parts shall be introduced in the same way,
one at a time, packing the material down by vibrating or
tamping until it occupies its apprOpriate volume. The
other sample shall then be placed in the other Side of
the apparatus in exactly the same manner.

Method 2. -- Two shallow square or circular boxes
whaving outside flat dimensions the same as those of the
guarded hot plate Shall be used. The edges shall be
made of wood strips % in. in thickness and of such width
as to make the depth of the box equal to the thickness of
the Specimen to be tested. The two square or circular
faces shall be made by gluing blotting paper or asbestos
paper to the edges of the strips. With one face in place
and the boxes lying horizontally, one portion of the

 

5. 1944 Book of A.S.T.M. Standards, Part II.

22

sample shall be placed in ach box, the material pressed
down, and the other paper face glued in place. These two
boxes containing the test Specimens shall then be placed
in the guarded hot plate. For materials the densities of
which cannot be altered at will, the container shall be
placed in a horizontal position with one side Open, an
excess of the dried sample weighed and poured in, the
container shaken, and the excess material leveled before
the upper paper face is glued in place. The weight of

the material used shall be obtained by weighing the excess
and subtracting it from the weight of the original material.

"PROCEDURE

6. (a) At least three and preferably four determina-
tions shall be made on each specimen at mean temperatures
which will cover the range of temperatures over which
the material is to be used. For any test, the tempera-
ture difference across the Specimen Shall be not less than
40 F. The above mean temperatures of the Specimen shall
differ from each other by at least 30 F.

(b) The atmosphere surrounding the test equipment
shall have a dew-point temperature not higher than the
coolest part of any surface or material in, or forming a
part of, the test apparatus.

(c) If the thermocouples mounted in the surfaces
of the plates are used to determine surface temperatures
(Section 4 (h)) the thickness of the test specimen shall
be taken as the distance between the surfaces of the hot
and cold plates when the Specimen is in place in the ap-
paratus. If separate surface thermocouples are used, the
thickness of the test Specimen shall be taken as the dis-
tance between the surfaces of the hot and cold plates
when the Specimen is in place in the apparatus, minus the
thickness of the two layers of blotting or asbestos paper
used.

(d) The heating element of the central heater shall
be supplied with electrical energy regulated to give the
desired temperature gradient through the test Specimen
and held constant within plus or minus 1 per cent, means
being provided to measure this energy. Automatic regula-
tion is recommended. Where automatic regulation is not
‘available, the energy input shall be regulated by means
of manual adjustment. The rate of electrical energy
input to the guard ring shall then be adjusted so that
the maximum temperature difference (Paragraph (e) between
the center and guard surface plates during the 5-hr. test
observation period shall be not greater than 0.75 per cent
of the average temperature drop through the two halves of

23

the Specimen, as determined by the differential thermo-
couples or the surface thermocouples. The average tem-
perature difference between these surface plates during
the test period shall be not greater than 0.2 per cent of
the temperature drOp through the Specimen.

(e) The cooling units shall be so adjusted that
the temperature drOps through the two test Specimens shall
not differ by more than 1 per cent.

(f) After steady state has been reached, the test
shall be continued with the necessary observations being
made to determine temperature difference, center-to-guard
thermal balance, and heat input until successive observa-
tions made at intervals of not greater than 1 hr., over
a period of 5 hr., give thermal conductivity values that
are constant to within 1 per cent.

(8) Upon completion of the test, the Specimen shall
be reweighed and the weight recorded.

"CALCULATIONS

7. (a) The density of the sample after drying, the
moisture in the sample as received, and the moisture re-
gain during test shall be calculated as follows:

 

D-.A_
B
MaC‘A
B
R=(G-F)x§x%

where:

D 2 density of the sample after drying (pounds per cubic
inch or grams per cubic centimeter),

M = moisture in the sample as received (pounds per cubic
inch or grams per cubic centimeter),

R = moisture regain during test ( ounds per cubic inch
or grams per cubic centimeter ,

A = weight of sample after drying prior to test (pounds
or grams),

B = volume of sample after drying (cubic inches or cubic
centimeters),

24

C = weight of sample as received (pounds or grams),

F = weight of Specimen prior to test (pounds or grams),
and

weight of specimen after test (pounds or grams).

G

(b) Thermal conductivity Shall be calculated by
means of Eq. 1 (Section 3 (b)).

"REPORT

8. (a) The report of the results of each test shall
include the following:

(1) Name and any other identification of the mater-
ial,

(2) Thickness of specimen tested.
(3) Dry weight of sample before test,
(4) Density of dried sample (before test),
(5) Mbisture in sample "as received,"
(6) Moisture regain during test,
(7) Temperature range of test,
(a) Hot-surface temperature,
(9) Cold-surface temperature,
(10) Mean temperature of test,
(11) Heat input in Btu. per sq. ft. per hr., and
(12) Thermal conductivity
(b) The conductivity-mean temperature relationship
for any material tested shall be obtained from a curve
resulting from plotting the thermal conductivity versus
mean temperature, and representing the average of the tests
on the three samples tested. The maximum deviation ob-

‘tained from this average curve, and the mean temperature
at which this maximum occurs shall be reported."

25

DISCUSSION

Basically the design of the National Research Council
on Heat Transmission, 1932, twelve inch Hot Plate, will
do as a starting point for the new design to conform to
the new specifications. (See blueprints in back cover
pocket.) Recommended changes, additions, or improvements

to this design are as follows:
HEATING ELEMENT

The existing heating element of nichrome wire in
conjunction with heavy COpper face plates is satisfactory
enough for equi-temperature distribution on the surface,
however, the possibility of substituting sheet rubber heat-
ing elements, (trade name Uskon,) should not be overlooked.
These elements would automatically take care of tempera-
ture distribution to the extent that metal foil faces
might be used in place of the heavy cOpper plates. Flex-
ible faces such as these would prove to be worth while in
the testing of materials which have irregular surfaces,
by better contact. At this date though, according to the
U. S. Rubber Co., manufacturers of Uskon, the gap neces-
sary, between the central plate and guard ring, to install

'conductors is greater than the one-eighth inch allowable.
POWER MEASURING METERS

In place of separate voltmeter and ammeter, a watt-

meter may be used, thus cutting down on reading time.

26

POWER SUPPLY

Specifications demand that the power input to the
hot plate Shall not vary more than plus or minus one
per cent for the duration of the test. This may be ac-
complished by using an electronic voltage regulator with
a direct current power supply. However, this is rather
burdensome, and an extremely easy solution to this pro-
blem can be found in using a constant voltage transformer
of the Solar Type. This transformer will regulate an
input from 95 to 125 volts, to 115 volts plus or minus

one-half per cent.
GUARD RING POWER SUPPLY

Adjusting the rheostats controlling power to the
guard ring can be accomplished by successive adjustment
followed by a waiting period to note changes on the
differential thermocouples between the central plate and
guard ring. This is a time consuming procedure and the
chance of over-correcting or under-correcting is high.
By arranging the guard ring power circuit so it may be
routed through the wattmeter by pushbutton will offer the
Opportunity of taking data on the guard ring power con-
sumption when prOperly balanced. (See Illustration Page
28.) This data may be used, on subsequent testing of
similar material, to initially set the rheostats to a
close approximation of the correct valve, thus saving

considerable time.

27

”-..-— ...-«---. -_ ,

. _ _. A ...__ .-7 A a, _.L_-m~ A, “*1.A_....

 

 

 

h.......«..,u amend. 4...“, a qu. an} A . t rendqmy
\ria. -7 - ,- .t 1
vi. ... Tutu. . .7 ... -. - ui. - 1 (Au .
"l A
l L\\»:: 4:,lliz.?;::e14filsisr {12:1
. f l,
- -- 1 1-7-1.111. (I ( X... ..\./\>\..\..,\>/\<>.\.<» ..fpilli- O 1 .i av o - M _
_
wakes... med 1i... a A ,,
Ammo» er W m H
‘ . w _ fl
is«.# unma-a W A _ a
m J .. a __ .. A m
r. J __
JV Col V 1 i 4:: 14,1159Iplrlv! .. 1) Jul: ,3: IIIIIIJ
\¥ «uq .\ ..
- ll).l.\1 1;. _ _ .
_ I I _ H
\Ytiw 10(li..-...l ( . . 7..

A
if. \..\<....\ .... . ,\.\ \./\ .- to-) i.\ ....w

mwkde or ..m

A._ M.-_.~_fl--- _ 4-

---“. -,_,,_..-__ r

p!

 

 

 

I'll,

quwuuxm
chalk ..waau:

pqrmomxa
UFQ JrL :UTma..UU:

A

. M

_ A

(>>8<()\</)\tiii4

#4.»..wbw12
wt...

|.|.I.-

A

 

. ...-.-._____.

Ir..- -..A -..—...“...A—J

?

l
_lnm

P)!!l.‘l’: in“)

_ m, _ -W~ -.~AI—.‘h.—--A§] ;..-_..

. 3 ,ilxill liIcp. J

wflfil‘,
“‘ ”I
l
I
1

J

-M—*—-— ...—*m-y ...—— . -7 ..

v-~..

 

uzaq

d.)

.3
C

o
_.
m

“AU.

 

 

28

VAPOR PROOF HOOD

To conform to the dew point specifications an ade-
quately vapor proofed hood must be supplied. For conven-
ience this hood should be large enough to allow all con-
nections, tubing, and wiring to be put through a station-
ary back section. The front, top, and sides could be
made with a ceiling pully and counterweighted. Construc-
tion would be two thicknesses of three ply plywood with
metal foil between them, the whole unit covered with

aluminum paint inside and out.
DEHUMIDIFICATION APPARATUS

To accomplish the lowering of the dew point of the
air surrounding the test equipment below the lowest tem-
perature in the equipment either mechanical or chemical
dehydration may be used. In case extremely low mean

11 are to be used in testing the mechanical

temperatures
refrigeration method has the advantage of lowering heat
input to the backsides of the cold plates from the atmos-
phere, Under ordinary Specified test conditions, (i.e.,
40 Deg. — 70 Deg. - 100 Deg. F. mean temperatures) this
effect is not so important, and the advantages of chemical

dehydration such as economy, ease of installation, and

convenience may be considered.

 

ll. Rowley, F. B., Jordan, R. C., and Lander, R. N.,
Low Tem er ture Thermal Conductivit Studies, University
of Minnesota Technical Paper No. 59, 1947.

29

The equipment may consist of a centrifugal forge
blower, a stack suitable for inserting chemicals, and
the necessary fittings. (See Illustration Page 31.)

The object is to "blow out" the hood surrounding the hot
plate with relatively dry air, then seal the hood and
utilize concentrated sulphuric acid to complete the dry-
ing process. The preliminary air was pre-dried by forc-
ing it over Open pans of calcium chloride and through a
ten inch deep bed of the same material. It was found
that starting with air containing 95 grains of moisture
per pound of dry air, this method would take out 60
grains per pound. The concentrated sulphuric acid was
found capable of keeping the dew point low enough to

employ 20 degree F. coolant in the cold plates.
ANGLE IRON CLAMP BRACKET

The solid angle iron bracket that holds the clamp-
ing screws quite often is a hindrance when charging or
working on the apparatus. Therefore, an improvement
would be to replace the fixed cross bar across the top of
the hot plate with a removable bar, which would need only
hooks at the ends to hold it in place similar to a fixed

C clamp.
V SELECTOR SWITCH

An improvement upon the thermocouple selector Switch‘
can be made by using a Silver contact, commercial, multi-

ple contact switch. The silver contacts will do away with

30

”Al I *
fl, .) ( I <lltl-.,l o
I... k
1: .-l
‘

a
.
A,
_
-

1
.W4

,P.«

a
a .

o>xmm see.:mru

1.15%.! 1 .fi 3..
A
A

-.— -r..-

\- ltllL 1).,b

.

i
F. ..l.

o.-a e «Areas» owwemyzmozno H ;

.1, u ,_

malemoxu 2,,OAqu w

4

l :

31

any oxidation that commonly affects the COpper contact
switch. This type of switch is also compact, does away
with possibilities of shorts, and uses soldered connec-
tions. All in all, it makes a more permanent, foolproof
setup. Switches of this type may be Obtained from the
Shallcross Manufacturing Company, Collingsdale, Pennsyl-

vania.
CONNECTION BOARD

Installation of a connection board or terminal board
on the back of the fixed side of the hot plate would facili-
tate the replacement and orderly arrangement of thermocouple
leads coming from the hot plate. Use of soldered connec-
tions is recommended for trouble-free Operation. This
terminal board would make the replacing of a defective or
broken thermocouple lead very easy. One word of caution,
make sure this board is at equi-temperature from one end
to the other. From the board to the selector switch ordi-

nary OOpper leads may be used. (See Illustration Page 33.)
MULTIPLE CONTACT CONNECTORS

For ease of Operation, removing the galvanometer,
potentiometer, and consequently the data station, a short
distance away from the hot plate is advisable. To keep
from a tangle of extension leads a multiple conductor

cable with suitable connectors may be used.

32

zo.»uzmp oaoo 201500 and :«ace.o p.2oc.o

 

 

 

um»w:o.»zwhoa A

AY.I.!:.I/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 1 I
. I My (\Mv zo.»ozaa
e pt: 4
M (c o ou
l 7
new cuegoo
meow . .
- . . aesz<emaou.luuiil
9 11?)... AAAAAAAAAAAA I V
I!!!
0, Av (\\\
F0 ; or . L9: .. lllllllllll / 23.812.
10....)m . I ..
he:
AIIIIIO cobouaum 0 Av ,L\\\
maeaoueacuz»
OW) louv'lAv ........ I
/’
Ow 7 Av (\w\

 

 

 

 

 

 

T5

THERMOCOUPLE COLD JUNCTION

An improvement may be made on the existing thermo-
‘ couple cold junction by replacing the cold junctions with

12 By prOper switching (See Illustra-

one cold junction.
tion Page 33) the cumbersome large cold junction now may

be compact and trouble free.
CONSTANT TEMPERATURE COOLANT SUPPLY

Up to the time of the 1945 standard test Specifica-
tions the mean temperature of test was allowed to fall
where it might, the cold plates being held at whatever
temperature the cold water supply was at. The new test
Specifications call for three mean temperatures, namely,
40 deg., 70 deg., and 100 deg. F., with a temperature
differential of at least 40 degrees F. across the sample.
This requirement forces the cold plates to be lower than
20 deg., 50 deg., 80 deg. F.,respectively, for the three
means. The 80 deg. F. temperature is rather easily ac-
quired by allowing a large reservoir of water to come to
room temperature. It will stay at this temperature, be-
cause the small amount Of heat being supplied from the
hot plate will be easily dissipated. To obtain a cOnstant
temperature below 50 deg. F. a commercial soft drink cool-
ing tank with a quarter horse power compressor, may be used.

Fill the cooler body with plain water and set for 28 or

 

12. American Institute of Physics, Tgmperatureg Itg
Measurement and Control_;n Science and Industr , Reinhold
Publishing Corp., 1941, p 201.

34

30 deg. F. This will cause a ring of ice to form on the
coils and remain there throughout the test, keeping the
temperature well below the 50 deg. F. To obtain a con-
stant temperature below 20 deg. F. adjust the compressor
expansion valve so that the pressure on the low Side is
low enough to insure below 20 deg. F. Operation, and
allow the compressor to run continuously with an alcohol
solution for coolant. A steady state condition will

finally result which will be constant enough for testing.
COOLANT PUMP

A laboratory centrifugal pump will prove most satis-
factory for circulating the coolant through the cold
plates. To insure the best conditions possible, arrange
a baffle in the coolant supply to make the distance traveled
by the coolant from pump outlet to pump inlet as long as
possible.

GALVANOMETER SHORTING SWITCH

A rather useful item in taking the numerous readings
required, by potentiometer and galvanometer, is a push
button type switch located near the potentiometer and
connected directly across the galvanometer. This saves

considerable time by cutting out the swinging periods.

35

COLD PLATE CONSTRUCTION

Formerly the cold plates were made with a double
Shell to provide an insulating air Space to cut down an
excessive condensation on the back sides. With a con-

trolled atmOSphere this is no longer necessary.

36

MATERIAL TESTED:

Time in
0
1
2
3
4

I:
z.

5 Hour Average

24.0
24.0

24.0

SAMPLE DATA

DATE:

Standard Corkboard From

July 28, 1947

 

Amneres

.149
.149
.150

.149

 

 

THERMDCOUPLE NUMBER

(Readings in.Mlllivolts)

 

 

 

 

Time in

EQM£§__ .A El 9. 2. El
0 1.080 .73 2.078 2.075 2.075
1 1.076 1.070 2.076 2.074 2.074
3 1.076 1.073 2.081 2.078 2.078
5 1.089 1.082 2.081 2.079 2.079
4 1.089 1.084 2.083 2.076 2.079
.§ 1.090 1.090 2.083 2.072 2.079
5 Hour 1.085 1.080 2.081 2.076 2.078

Average

37

E.
2.074
2.070
2.077
2.079
2.079
24mg.

2.075

[LOIS QI‘IOVOE.

 

Differential
ThermOOOuples
Averggg
.000
.000
-.002
.000
.000
~.__0_9_9_
-.0004
9 .7.
1-069 1.069
1-072 1.072
1-077 1.077
1.079 1.079
1.082 1.082
1-079 1.079

SAMPLE COMPUTATIONS

Average Cold Plate Temperature = 1.0807 M.v. = 81.03750 F.
Average Hot Plate Temperature = 2.0775 M.V. = 123.900 P.
Average Temperature Difference = 42.86250 F.

Mean Temperature = 102.470 F.

Power Input = 24.0 (.1492) (3.413) = 12.2212 Btu per Hour.
Total Area of Effective Test Area = .889 Sq. Ft.

Thickness of Test Samples = 1 inch.

Thermal Conductivity = K = ' L

T1 ‘ 2 P
12 2212 1 _
K '.‘66“°'9"T—‘L1'42.86 (1) 320781

38

 

1 1 1 1 U 1

1 1 1 1 1
1 A ._

.. ,1 1 1

1 fi __ fl

1 ll ,1 2 1 m 1

1 1 1 1 1

1 m .u autumn 2.1uu=»<cUg ub.n¢u:_ 1 1

1 . . , _ .. -- . - .1 . m 1

_ . 1

 

.0
?

 

 

 

1
1
.4 . . ,. .1
:. 1%. .1111.H .H. a @143. e , ”an.
1 1
1 1

— Lu it'll
a .1 1 1.., com.

1 1 , ._

M 1 1 1 1 1

1 1 . 1 h 1

1 1 1 1

1

 

,ooe.

 

 

 

.3».

I. A1Ifi119aouoo 1uu83u¢

“...—..-“... - -._.._...

 

 

 

wo- fip—w .....~—o_.—. _..-, -.

Elli:;-.ee.

...- .11 1111111111111”; 11.1 , ._._

 

 

 

 

uzah«u nau+ u<mm1em> >>.>

.onn.

 

 

 

 

1
1
. 1

 

 

oc¢o¢ucou haw» + . . ..,. .1
1.:
1

.0.. W ....h... __1_ — .J...__—-.__.._-._1._-.

 

 

 

 

 

H 1
_ _
lieil +i++é.!lii+%+.1-:iii++$iillzéil+; 1 1 :Iallllri

1 1 1 1 1 1 1 1
1 1 1 1 1 1 ,_ .
. 1 1 1 1
1 _ 1 11 1 . v v . 1 .
1 1 _ 1 1 1 1

ah '.'

mk¢4q

1
1

 

POT

ouacenu a»

u

“LE

./—
‘sok’r

/
—~

... .
r 1

r2. ,.

DAMDLE

.

Z ......Gzn

.-

....»

LiLam?

C

wmnranmefiuk

.
—
L 1 l
k C
r I: ..h.
2. .k

p. m11

I vi“

1.. (0.
—

.15.... a

hi‘
0

oh mwwacuo
w1eo»

meeeqeueeu»

4O

a”.
BIBLIOGRAPHY & REFERENCES

American Institute of Physics, Temperature, Its M asur -
ment and Control in Science and Indust , Reinhold
Publishing Corp., 1941, p 201.

American Society of Heating and Ventilating Engineers,
Guid , 1948, p 97.

Brown and Marco, Introduction to Heat Transfer, Chapter 2,
McGraw Hill. .

Forbes, Principal, Transactions of The Royal Society of
Edinburg, 1861-1862.

Lander, Robert M., Factors Affect1gg,Thermal Conductivit ,
University of Minnesota Technical Paper No. 49, 1944.

Maxwell, J. Clerk, Theory of Heat, Longmans, Green & Co.,
1888, p 268.

Poynting, J. H., and Thomson, J. J., Heat - A Textbook
of Ph sic , Charles Griffin & Co., 1904, p 102.

 

Preston, Thomas, Theory of Heat, MaoMillan & Co., 1894,
p 527.

Rowley, R. B., and Algren, A. B., Heat Transmission
Through Buildin Materials, University of Minnesota
Bulletin No. 8, 1932.

Rowley, R. B., Jordan, R. C., and Lander, R. N., Low
Temperature Thermal Conductivity Studies, University
of Minnesota Technical Paper No. 59, 1947.

Saha, M. N. and Srivastava, B. N., A Textbook of Heat,
Indian Press Ltd., 1931, p 316.

Thompson, Benjamin ( Count Rumford), The Whrks of Rumforg,
American Academy of Arts & Sciences, Vol. II, p 144.

Tyndall, John, Heat As a Mod of Motio , D. Appleton & Co.,
1890, pp 119-245.

41

 

n
O
8
r
e
d
n

 

 

 

HICHIGQN STQTE UNIV. LIBRRRIES

31293010939761