THE FREE ENEKGY OF FORMATION AND THE INTERFACIAL ENTHALPY

IN PEARLITE*

J. J. KRAMER,t G. M. POUND; and R. F. MEHL:

-1 constant heat-flow calorimeter was used t,o measure the ent~halpy of the pearlite-austenite trans-
form&ion as a function of pearlite spacing in the iron-carbon and copper-aluminum eut,ectoids, and
the interfacial enthalpies of the peariitic phases were found t,o be 1400 -‘_ 300 ergs/cm” and 1560 zz 606
ergs/cm2, respect~ively. The eorrespox~ding interfacial free energies xvere estimated to be 700 ergs/cm2
and 1000 ergs/cm2. Thus ihe interfacial free energy between ferrite and cementite is less t,han one half
and probably about one fourth ofthat required by the Zener theory for pearlite growth. This means that
only somethmg between one fourth and one eighth of the free energy available for the t,ransforrnation
is conserved as pearlitic int,erfaciwl free energy, instead of one half as required by the theory. Free
energies of formation of bulk pearlite are presented for both eutertoid systems.

L’ENERGIE LIBRE DE FORMATION ET L‘ENTHALPIE D’INTERFBCE DE LA PERLITE

Un calorimetre adiabatique a et6 utilise pour mesurer l’enthalpie de la transformation austenite-
perlite, en fonction de I’&wtnment des lamelles de la perlite dans les rutectoides de Fe-C et Cu-Al.
Les auteurs ont trouve que les cnthalpies d‘interface des phases perlitiques sont respectivement de 1400
-+ 306 ergs/cm2 et 1500 * 600 ergs/cm *, tandis que l’estimation des energies libres correspondantes
conduit aux valeurs de 700 ergs/cm2 et 1000 ergs/cm2. 11 en r&x&e que les energies libres d’interface
entre la ferrite et la cementite sent infirieures k la moitie, et probablement. au quart, de relies prPvues
par la theorie de Zener pour la, croissance de la perfite. Cefa signifle qu’une partie comprise entxe l/4
et, l/S de l’energie Iibre dis~)nible pour la t~ransformat,ion se trouve sons la formr d’energie libre de
l’int.erface de la perlite, tandis que la theorie exigeait que I‘Bnergie conservee corresponde it la moitie de
l’energie disponiblc. Les nuteurs donnent egalement des valcurs des energies libres de formation de IA
perlite massive pour les deux eutertoides.

DIE FREIE BILDUSGSENERGIE UND DIE GRENZFLACHENENTHALPIE

BE1 DER PERLITBILDUNC
Die Enthalpie der Perlit-Auntenit-Umwandlung wurde kalorimetriseh als Funktion des Perlitabstandes in
eutektoiden Eisen-Kohlenstoff- und Kupfer-Xluminium-Geftigen gemessen. Dsraus ergaben sich fur die
Grenzfliichenenthalpien der perlitischen Phasen Werte von 1406 + 300 erg/cm2 bzw. 1500 f 600
erp/em2. Die entsprechenden freien Energien der Grenzfliichen warden zu 700 erg/cm2 und 1006 erg/cm2
geschiitzt. Demnach betrapt die freie Brenzfliichenenergie zwischen Ferrit und Zementit, weniger als die
Halfte und wahrsrheinlich nur etwa ein Viertel des in der Zenerschen Theorie des Perlitwachstrums
geforderten Wertes. Dies bedeutet, dass nur ein Viertel bis ein Arhtel der gesamten freien Energie der
~rn~~andlung als Energie der Perlitgrenzfl~~hen vorliegt, anstntt der Hiilft,e. wie van der Theorie
gefordert. Fiir die beiden untersucht~en eutektoiden Systeme werden ferner Wet% fiir die freie Bildungs-
energie des Perlits angegeben.

INTRODUCTION appeared as a recent, contribution from this labora-

The rate of nucleation and the rate of growt’h of tory.@) The present paper is an extension of thir;
pearlite in the austenite-pearlite reaction depend earlier work with an aim to greater calorimetric
upon, among other things, the free energy of formation precision.
of pearlite and the interfacial free energy between It has long been realised t,hab a relat,ionship must
ferrite and cementite. This subject has been reviewed exist between the rate of growth and the interlamellar
in detail.(l) A preliminary study of the subject spacing in pearlit8e.(i~3) Zener(4) has shown that t,he
edgewise growth rate of a peaditic structure should
* This paper is from part of a thesis presented by J. J. be maximized if one-half of the total Gibbs free energy
Kramer to the Committee on Graduate Studies of the
Carnegie Institute of Technology, Pitt.sburgh, Pennsylvania, available for the reaction is consumed to drive the
in partial fulfillment of the requirements for the degree of diffusional processes, while the other half is used to
Doctor of Philosophy in Metallurgical Engineering, September
1956. The work was supported by the U.S. Army Office of form the interfaces between the two phases. This
Ordnance Research under Contracts No. DA-36.061.ORD-205 treatment is subject to the condition that diffusion is
and -350. Received March 6, 1958.
j’ Westinghouse Electric Corporation, Pittsburgh, Penn- the rate controlling step in the reaction, i.e. interface
sylvania. reactions are sufficiently rapid to require negligible
$ Department of Metallurgical Engineering, Carnegie
Institute of Technology, Pittsburgh, Pennsylvania. driving force. The total free energy change available
ACTA -METALLURGICA, VOL. 6, DECEMBER 19.58 763
704 ACTA METALLURGICA, VOL. 6, 1958

for the reaction affects both the growth rate and the evaluated a number of measurements of the copper
interlamellar spacing of the pearhte. Zener’s value of grain boundary free energy and give the most reason-
the interfacial free energy between the ferrite and able value for this energy as 550 ergs/cm2 between
cementite in the iron-carbon eutectoid, calculated 950 and 1050°C.
from available free energy and pearlite spacing data,
is 2800 ergs cm2. EXPERIMENTAL PROCEDURE
Thompson’5) measured the interfacial free energy
between ferrite and spherodized cementite by means a. Principle of calorimeter
of the Gibbs-Thomson effect and found it to be 1350 The constant heat-flow calorimeter used for these
ergs/cm2. From the relative interfacial free energies measurements was a modification of the type originally
measured in the iron-carbon system by Smithc@ and developed by Smith. (la) A specimen on which thermal
the measurements by Van Vlackc7) of the absolute measurements are to be made is positioned within a
grain boundary tension * of gamma iron, the ferrite- small refractory container in a tubular furnace (Fig. 1).
spherodized cementite interfacial free energy at Power to the furnace is controlled to maintain a
1100°C was computed to be 710 ergs/cm2. The constant temperature difference between the outside
discrepancies between the interfacial free energy and inside surfaces of the container. Under these
required by Zener’s theory and these measured values conditions, the heating or cooling rate of the specimen
constitute a difficulty in understanding the mechanism is a function of its enthalpy, the thermal conductivity
of pearlite growth.7 However, there is some un- and heat capacity of the container, and any extraneous
certainty in the reported experimental values of the losses.
specific interfacial free energy. Thus, it seemed If no transformation requiring the absorption of
desirable to estimate this value by calorimetric means, heat occurs within a given temperature range, the
and this was one of the objects of the present work in specific heat of a specimen can be calculated within
which the interfacial enthalpy was measured and the this range from the heating rate curve of the container
interfacial free energy computed by assuming a value without any specimen, the curve with a specimen of
for the entropy of the interface. The interfacial known heat capacity, and the curve for the specimen
enthalpy was determined by measuring the enthalpy
of the pearlite-austenite transformation as a function
of pearlite spacing. Also, using the data obtained in
the course of these measurements, the free energy of
formation of bulk
pearlite (infinite interlamellar
spacing) was computed. From this value and the
specific interfacial free energy, the fraction of the
available free energy conserved as pearlitic interfacial
free energy was determined in order to check the
prediction that it should be one-half.
As a subsidiary investigation, the specific interfacial
enthalpy between the pearlitic phases in the copper-
aluminum system was measured and the interfacial
free energy estimated. The interfacial free energy
between the pearlitic phases (a and y2) in the copper-
aluminum system has not been calculated on the basis
of Zener’s analysis because of the lack of thermo-
dynamic data. Smith’s measurements(@ of the relative
interfacial free energy between phases in alloys
containing copper indicate that these interfaces
Support
usually have smaller free energies than the large-angle
copper grain boundary. Fisher and Dunntg) have
Imtlotion Brick
*
Although, in the case of solids, there is a small difference
between specific interfacial free energy and interfacial tension, Nlchrome Winding
these terms will be used interchangeably.
t Another difficulty of the theory is that it predicts an
inverse linear relationship between edgewise growth rate and
pearlite spacing which is not borne out by experiment.‘8’ FIG. 1. Schematic diagram of calorimeter.
 KRAMER, POUND AND MEHL: IXTERFACIAL ENTHALPY IN PEARLITE 765

of unknown thermal properties. The expression for into this well, specimens could be replaced in the
specific heat is furnace. Small guides in the crucible positioned the
specimen reproducibly t,o within l/32 in.
Control of the temperature difference through the
container was effected by an electronio controller
which was activated by a differential thermopile.
where the subscripts s, k, c, refer to the test specimen,
The thermopile (of No. 2X gage platinum, pIatinum
the specimen of known thermal properties. and the
104: rhodium) consisted of twelve alternate beads on
container, respectively; W is the weight of the
the outside and inside surfaces of the container. With
specimen, Cps its specific heat, and At/AT the slope
of the inverse heating rate curve. special preparation, this wire was found to give very
When a transformation occurs, the specimen reproducible readings for long periods of operation of
absorbs or evolves heat as determined by the require- the calorimeter, and was sufficiently fine that thermal

ments of the transformation. For transformations leakage from the outside to the inside surface of the
which are sufficiently rapid, no temperature change container was not a problem. Control of the tempera-
occurs throughout the process of the reaction, and the ture difference was to within jO.l”C for a difference
ent~halpy of t$he transformation can be expressed as of 50°C. The specimen temperature was measured
with a calibrated Chromel-Alumel thermocouple
At,, which was replaced after each measurement cycle.
AH, = (~~~~~)~~. ~~--- -- (3
i (At/AT), --.^(At/AT), The potentiometer circuit consisted of an electronic
recorder and the temperatures were accurate to t”C.
where At, is the time required for the transformation. Any stray electrical pickup in the thermocouple
Sluggish transformations occur over a temperature circuit, as e.g. from the furnace winding, was elimi-
range, and part of the enthalpy of transformation is nated by encasing the container in a thin nickel sheath
manifested as an apparent specific heat. Evaluation which was grounded.
of the ent’halpy of such a reaction requires knowledge The container and all insulation were of Armstrong
of t,he t.rue specific heat and an integration of the A-20 brick which was found to be chemically stable
difference between true and apparent specific heats and free of any phase t,ransforlnation within the
over the t~n~perat,ure range of the reaction. temperature range of calorimeter operation. The low
thermal conductivity and low temperature coefficient
of conductivity of this brick made it particularly
desirable. However, it is somewhat hygroscopic, and,
In order to measure the small variation in enthalpy
accordingly, the furnace was evacuated and baked out
of transformation as a function of pearlite spacing,
at 400°C to remove any adsorbed or chemically
the calorimeter was designed for maximum repro-
combined moisture. Good reproducibility was ac-
ducibility. This was accomplished by the control of
hieved when the calorimeter was used in measurements
four major variables.
on a heat,& cycle, and was found to be best when the
1. The reproducibility in position of all components controller was set to give a flow of heat through the
of the calorimeter was carefully maintained. container walls of approximately 1 calisec. Under
2. Thermocouples were selected which were chemi- t,hese conditions, heating curves on specimens of
cally and structurally stable for long periods of identical heat capacity are reproducible to within
t,ime when used within the temperature range 10 set over a 3 hr period of operation, an overall
0-850”c. variation of 50.1 per cent. (A more complete
3. Stray electrical pickup in the thermocouple discussion of experimental error will be given in the
circuits was eliminated. Discussion of Results.)

4. Heat transfer was made reproducible by using a

chemically stable refractory specimen container c. Specimen preparation
and eliminating moisture from the furnace.
The plain-carbon eutectoid steel was obtained* in
Test specimens were replaced in the crucible by the form of annealed 1 in. round bars. Two heats of
removing the thermocouple and the assembly con- steel were used in the experiments. Precautions were
taining t,he plug, the container cap, and the thermo- taken in the preparation of the alloys to prevent
couple guide tube. The t,hermocouple well of each
specimen was thread-tapped, and by screwing a rod
766 ACTA METALLURGICA, VOL. 6, 1958

decarburization and to minimize alloy banding. The

compositions were as follows :

Alloy No 1 ’ 0.81
Alloy No 2 0.80
-.____

Specimens wa re cut from the 1 in. bars and austeni-

tized in a well-deoxidized lead pot at 900 & 2°C for
3 hr. They were then isothermally reacted at various
temperatures below 727’C to form pearlite of the
required spacing. A $ in. rim was machined from the
surface of all specimens to remove any possible trace
of decarburization, and a small sample was removed
from each specimen and used for spacing determina-
tion.(ll) All pearlite formed was found to be uniform
with little degeneracy (Fig. 2).
The non-ferrous copper-aluminum cutectoid* was
received in the form of 2 in. extruded rods. This alloy FIQ. 2. Fe-C fine pearlite, Picral etch. x 2000.

analyzed in weight 88.07 o/o copper, 11.89% aluminum, apparent spacing ranges much larger than the true
0.01 y. manganese, O.O2o/o iron, and the balance, trace spacing, this method of determination can be used
impurities. The same austenitizing schedule used for for true spacings as narrow as 2000 A ahhough the
the Fe-C eutectoid was employed in the preparation of limit of resolution of the light microscope is 5000 A.
this alloy. Further protection from oxidation was Spacings ranged between 2000 and 8000 A.: For
provided by encasing each sample in a graphite
capsule. Isothermal transformations were conducted
at temperatures below the eutectoid temperature of
565’C. This pearlite was also found to be uniform and
free from degeneracy (Fig. 3). The necessary trans-
formation times at various temperatures were
obtained from appropriate isothermal transformation
curves.(12~13) The reaction temperature necessary to
obtain pearlite of any given spacing was determined
from the experiments of Pellissier et uZ.oi) on the iron-
carbon system and Haynes(14) on the copper-aluminum
system.

d. Spacing determination
Pearlite spacings were determined by the method
developed by Pellissier et aZ.(ll) which consisted of
matching the experimentally observed distribution of
apparent spacings? on a plane of polish to the
theoretical distribution of apparent spacings. Two
hundred spacing counts at Xl000 were used to obtain
FIG. 3. &-Al fine pearlite, Ferric chloride etch. x 2000.
the experimental distribution of apparent spacings.
With this number of counts, the accuracy in spacing
2 Roberts and Mehl’ls’ found that pearlite formed by a
or interfacial area is within 5 to 10 per cent. Since the normalizing treatment (Fe-C pearlite of spacing ~3500 A)
transforms 99.5% to inhomogeneous austenite within 90 set
* Prepared by the Ampco Metal Company. at 732°C.
7 The apparent spacings are those observed on a plane of Darken and G~rryc’~’ determined the heat of solution of
polish, and represent a range of spacings from the true cementite in austenite and found that, to a first approximation,
perpendicular spacing between lamellae to observed larger it is independent of carbon concentration in ihe austenite.
spacings which result from the inclination of the lamellae Therefore, no heat effect is expected from homogenization of
to the plane of polish. the carbon in the austenite.
 KRAMER, POUND AND MEHL: INTERFACIAL EKTHALPY IX PEARLITE 567

coarser pearlites, diffusion distances are too great and of the container with and without a specimen of known
t’he rate of transformation to austenite becomes too specific heat after every 3 or 4 measurements on the
sluggish. Therefore, very accurate determination of pearlite. This permitted a check of the stability of the
the heat of the t’ransformation of coarse pearlites is calorimeter. Copper was selected for the calibration
not possible. specimen because its specific heat’ is well-known and
does not differ greatly from the specific heat of steel.
e. Operatim of the calorimeter and results
The data of Kelly(r’) were taken as the best values for
A sufficient number of samples of Fe-C and Cu-Al the specific heat of copper.
pearlites were investigated to allow statistical analyses The reproducibility of both the pearlitic and
of the data. Measurements were made on pearlites of austenitic specific heats in successive measurement8s
medium and fine spacing in a random sequence in on an iron-carbon alloy is shown in Fig. -1. The curve
order to minimize the effect of any systematic change of specific heat for the copper-aluminum eutjectoid
in the thermocouples or calorimeter. The calorimeter (Fig. 5) shows an abrupt change in specific heat of the
n-as calibrated by measuring the heating rate curve austenitic phase close to the eutect’oid trmperat’ure.

0.320
t

Y
4280 -
2
! Q260-

2 Q240-
P
s
4. Q220-
0

$ Cl200
ii 0.160
u
b 0.160

!#
0.140

OJOOI I I

440 460 520 560 600 640 660 720 760 b

TEMPERATURE , “C

FIG. 4. Scatter in specific heat data from successive runs on plain carbon eutectoid steer.

0.240 -

a040 - /
I’
0.020 I I/ I I I 1 L I
440 460 520 560 600 640 680 720 760 8
TEMPERATURE , %

FIG. Ti. Mean value of specific heat ns a function of temperature for Cu-Al eutectoid.
768 ACTA METALLURGICA, VOL. 6, 1958

a
Sf
L . 16.60 -

x\\x o
ak! x
x
ez x
uS’~-
l-w
30
5:
LC
I760 - +*O\
r;:
L

$30
2 4 6 6 IO I2 I6
8-w 0 _:’
Zt FERRITE-CEMENTVIE INTERFACIAL AREA, CM2x IO / GRAM

Fro. 6. Variation of enthalpy with interfacial area for plain carbon steel.
x-alloy no. 1. o-alloy no. 2. +-least square points.

AHB=6.9

\+,

x0
o
+\
Y
x
Y

t 6.Od ,
0 2 4 6 6 l0 I2 I4 I6
INTERFACIAL AREA , CMf I8 / GRAM

FIG. 7. Variation of enthalpy with interfaoial area for Cu-Al eutectoid.

+-least square points.

This makes determination of an enthalpy or free pearlite and 565°C for Cu-Al bulk pearlite.* The
energy change-temperature relationship difficult exact slopes of the curves were determined from
because of uncertainty in extrapolating the austenitic a least squares analysis of the data. Interfacialenthal-
specific heat curve to lower temperatures. pies of the pearlite were determined from the enthalpy
For the iron-carbon alloys, complete reaction variation by the relationship
occurred within a temperature range of 3-4°C. For (8AH/&4), = iYs. (3)
the copper-aluminum alloys, however, this range was
For the Fe-C pearlitic interface this value is, from
found to be as great as 10°C. The enthalpy of the
Fig. 6, 1400 & 300 ergs/cm2 at 727°C while for the
pearlite to austenite reaction at the equilibrium
Cu-Al pearlite (Fig. 7) it is 1500 f 600 ergs/cm2 at
eutectoid temperature for both iron-carbon and
565°C. By extrapolation of the enthalpy-surface area
copper-aluminum alloys as a function of the inter-
relationships, the enthalpy for the reaction of the bulk
facial area is given in Figs. 6 and 7. The equilibrium
eutectoid temperature is taken as 727°C for Fe-C bulk * Bulk pearlite refers to pearlite of negligible interfacial area.
 KRAMER, POUND AND MEHL: INTERFACIAL ENTHALPY IN PEARLITE 569

constituents transforming to austenite is 18.4 (727°C) difference between the specific heats of the pearlitic
Cal/g for the Fe-C alloy and 8.9 (565°C) Cal/g for the and austenitic phases. The results are presented in
Cu-Al alloy. Using these enthalpy and specific heat Figs. 8 and 9.
data, the free energies of the reactions were calculated
DISCUSSION OF RESULTS
from the thermodynamic relationship,
(a) Precision
hF T;,_l’ = -$Y[ AHTO --JTY CAPI The reproducibility of measurements on the specific
heat of the plain-carbon eutectoid steel was &l per
-Cc,(y or B)dT
1 d(l/T) (4) cent, and the same reproducibility
Cu-Al eutectoid at the lower temperatures.
was obtained for the
However,
where AHTO is the enthalpy of the austenite to near the Cu-Al transformation temperature (566”C),
pearlite reaction at the equilibrium transformation the reproducibility of specific heat for this alloy was
temperature TOoK, AF, is the free energy change at only &4 per cent. This is thought to be due to the
any temperature, and C,(P) - C,,(y or p) is the presence of an ordering reaction in the CI (low

I I I 1 I I I I
440 480 520 560 600 640 680 720 750 800
TEMPERATURE , “C

Fm. 8. Free energy change vs. temperature for au&mite to pearlite reaction in eutectoid plain carbon steel (bulk alloy).

~4 , , , , , , , , ]
Y 400 440 460 520 660 600 640 660 720 760
I.,

E TEMPERATURE ,“C

FIG. 9. Free energy change of /3 to pearl& reaction vs. temperature for G-Al eutectoid (bulk alloy).
770 ACTA METALLURGICA,
temperature) phase and the slowness of the austeni-
tization process in this substitutional alloy. The
reproducibility of the enth~l~y of tra~form~tion was
10.5 per cent for the eutectoid steel and f4 per cent
for the C&Al eutectoid.
From an a.nalysis of the experimental method it is
estimated that the precision measure (supposed
absolute accuracy) of the specific heats is -&3 per cent
for the Fe-C alloy. However, the error is undoub~~y
greater (perhaps &IO per cent) in the ease of the Cu-Al
eutectoid for the reasons given above. The precision
measure of the enthalpy of transformation is estimated
to be about A5 per cent. But again, the error for the
&-Al eutectoid is probably about *lO per cent.
The same precision measures are assumed to apply to
the enthalpies of transformation of the pearlites of
infinite interlamellar spacing, and similar errors should
apply to the free energies of transformation of these
“bulk” pearlites.
The present specific heat data on steel are in
reasonable agreement with the results of Awbery and
Snow,(i9) Esser and Baerlecken,@O) and Hagel et o1.Q)
Also, the present data on Cu-Al eutectoid specific heat
are in reasonable agreement with the results of
Masumoto et uZ.(~*) Nevertheless, the present value for
enthalpy of transfo~?tion of eutectoid steel of
infinite interlamellar spacing (18.4 Cal/g at 727°C) is
not in good agreement with the values reported by
Esser and Grasst2n (19.3 Cal/g) and Hagel et CZZ.(~)
(19.6 Cal/g). Neither does it agree well with the value
of 20.0 cal,‘g calculated indirectly from the data of
Darken and co-workers.f1~2sf
However, the precision measure of the interfacial
enthalpy between the eutectoid phases of the plain
carbon steel or the Cu-Al eutectoid is not limited by
the absolut,e accuracy of the enthalpies of transforma-
tion. Rather it is de~rmined only by the probable
error in the slope of Fig. 6 or Fig. 7, respectively, due
to scatter (reproducibility) of the data. Thus the
precision measure of the interfacial enthalpy is
estimated from the data to be 520 per cent for the
plain-carbon eutectoid steel and &4O per cent for the
&-Al eutectoid. Ac~or~gly, the ~te~acial enthalpy
between the eutectoid phases is given as 1400 -& 300
ergs/cm2 for the steel and 1500 & 600 ergs/cm2 for the
CU-Al eutectoid.
In order to estimate specific in~rfacial free energy
between the eutectoid phases, one must assume a value
for interfacial entropy. Not many data are available
to provide a basis for accurate selection of an inter-
facial entropy value. The entropies of solid gold(23)
VOL. 6, 1958
and silver(24) surfaces have been determined from
surface tension measurements to be 1 erg/cm2/“K at
13OO’K and 0.5 erg/cm2j”K at lfOO°K, respectively.
For the purposes of the present computation, S, was
assumed to be 6 erg/cm2 “K. Accordingly, the specific
interfacial free energy between cementite and ferrite
was calculated from
Fg = H, - TS,, (5)
to be 700 ergs/cm2 at 727’C. The corresponding figure
for the Cu-Al eutectoid is 1000 ergs/ems.
It is evident from the results, that the specific intor-
faoial free energy between ferrite and cementite is less
than one half and probably about one fourth of that
required by the Zenerf4) theory for pearlite growth
(2800 ergs/cm2).* This means that only something
between one fourth and one eighth of the free energy
available for the transformation is conserved as
pearlitic interfacial free energy, instead of one half as
required by the theory. It has been suggested that
there may be at least one other source of irreversi-
bility in pearlite growth besides that of carbon
diffusion.(25,26) For example, there may be appreciable
irreversibility in one of the interface reactions.‘26)
CONCLUSIONS
The enthalpy of the interface between the ferrite
and cementite phases of plain-carbon eutectoid
steel is 1400 & 300 ergs/cm2. The enthalpy of
the interface between the eutectoid phases of the
cop~r-alum~um system is 1509 f 600 ergs~cm2.
Assuming an interfacial entropy of 8 ergfcma/“K
for both systems, the interfacial free energies for the
eutectoid phases of the iron-carbon and copper-
aluminum systems are 700 f 300 ergs/cm2 and
1000 * 600 ergs/om2, respectively.
The specific interfacial free energy between ferrite
and eementite is less than one half and probably
about one fourth of that required by the Zenerc4)
theory for pearlite growth (2800 ergs/cm2). This
means that only something between one fourth and
one eighth of the free energy available for the
transformation is conserved as pearlitic interfacial
free energy, instead of one half as required by the
theory.
The enthalpy of copper-aluminum eutectoid trans-
formation is -8.9 & 0.9 Cal/g at the eutectoid
temperature of 565%
* The upper limit of one half arises from the extreme
assumption of an interfacial entropy 8~ equal to zero, and
thus uncertainty in s$ does not affect the validity of this
oonolusion.
 KRAMER, POUSD AND MEHL: INTERFACIAL ENTHALPY IN PEAHLITE 771

ACKNOWLEDGMENT 12. W. C. HAGEL, Thesis, Metallurgy Department, Carnegie

Institute of Technology (May 1953).
Thanks are due t.o Dr. Q. A. Roberts of the 13. D. J. MACK, Trans. Amer. Inst. Xin. (IWetnZZ.) Engru.
~anad~uu~ Alloy Steel Compan~l, who supplied the 175,240 (1948).
14. R. HAYNES, J. Xwt. Met. 83, 105 (1954).
plain-carbon eutectoid steels. 15. G. A. ROBERTS and R. F. MEHL, Tra.ns. Am. &‘oc. Mekxls
31, 613 (1943).
REFERENCES 16. L. S. DARKEN and R. W. GURRY, Physical Chemistry of
1. R. I?. MEHL and W. C. HAGEL, “The Austenite: Pearlite Metals. McGraw-Hill, New York (1953).
Reaction” in Progress in Metal Physics Vol. 6. Pergamon 17. K. K. KELLEY, Bureau of Mines Bulletin 476 (1946).
Press, New York (1956). 18. Ii. KISUMOTO, H. SAITO, and M. TAFCAHASHI, Sei. Reps.
2. W. C. HACEL. G. M. Pooxo. and R. F. MEHL. Acfa Met. Re.~~ch Inst. Tohoku U&u. 7, 5, 465 (1955).
4, 37 (1956). 19. J. AWBERY and A. SNOW, Spec. Report No. 24, Second
3. K. F. MEKL, “The Physics of Hardenability“ in Harden- Report of the Alloy Steek Research, Compnitiee, p. 216.
nbility of lllloy Steek. Am. 8oc. Metals, Cleveland (1939). Iron and Steel Inst., London (1939).
4. C. ZENER. TWW. Amer. Inst. Min. (Metall.) En,ars. ” 167. 20. TX. ESSER and H. BAERLECKEN, Arch. Eisenhuttenw. 14.
550 (1046). 617 (1941).
.5. F. C. THOMPSOX, Trans. Pw&ay Sot. 17, 391 (1922). 21. II. ESSER and H. GRASS, Stahl u. E&en 53, 92 (1933).
6. C. R. S~~ITH, “Interphase Interfaces” in Imperfections in 22. I,. S. DARKEN and It. P. SMITH, I&. Emg. Chem. 43, 1815
Seurly Perfect Crystnls. John Wiley, New York (1952). (1951).
‘i. L. H. Vau VLA~K, Trans. dmer. fnst.&fin. (Met&.) Engrs. 23. F. H. BUTTNER, H. UDIN. and J. WULFF, Tmns. Amer.
lS1, 251 (1951). Inst. Min. (Me~~Z.) Eng?x. 191, 1209 (1951).
R. F. MENL~ J. Iron St. Inst. 159, 113 (1948). 24. E. R. FUNK, H. UDIN, and J. WOLFF, Trans. Amer. Irtst.
:: J. C. F~SXF,R and C. G. DUNN, “Surface and Interfacial Min. (Metall.) Engra. 191, 1206 (1951).
Tensions of Single Phase Solids” in Iwqxvfections in 25. J. C. FISHER, “Eutectoid Decompositions” in Thevmo-
Ye&u Perfect Crust&. John Wilev. New York (19521. dynamics in Physical Meta.llurgy. Am. Sot. Metals,
10. C. S.-RMI<H, T&u. Amer. In&“&in. (M~tnli.) Er&ws. Cleveland (1949).
137, 236 (1940). 26. E. S. MACHLIN, T?-ans. Inst. Xin. (Met&.) Engrv. 5, 437
11. G. E. PELUSSIER, M. F. HAWICES, W. A. JOHNSON, and (1953).
R,. F. MEHL, Trrtns. A,mer. Sot. Met& 30, 1049 (1942).