ASM International, Doru M Stefanescu (editor) - ASM Handbook, Volume 1A_ Cast Iron Science and Technology-ASM International (2017)

Prévia do material em texto

ASM HandbookW
Volume 1A
Cast Iron Science and Technology
Prepared under the direction of the
ASM International Handbook Committee
Volume Editor
Doru M. Stefanescu, FASM, The Ohio State University and The University of Alabama
Division Editors
Steve Dawson, SinterCast Ltd.
Hasse Fredriksson, KTH Stockholm
Wilson Guesser, TUPY
Richard Gundlach, Element Materials Technology
Harry Tian, GIW Industries
ASM International Staff
Victoria Burt, Content Developer
Steve Lampman, Content Developer
Amy Nolan, Content Developer
Susan Sellers, Content Development and Business Coordinator
Madrid Tramble, Manager, Production
Kelly Sukol, Production Coordinator
Patty Conti, Production Coordinator
Diane Whitelaw, Production Coordinator
Karen Marken, Senior Managing Editor
Scott D. Henry, Senior Manager, Content Development
Editorial Assistance
Warren Haws
Elizabeth Marquard
Jo Hannah Leyda
Lilla Ryan
ASM InternationalW
Materials Park, Ohio 44073 0002
www.asminternational.org
Copyright # 2017
by
ASM InternationalW
All rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, 
photocopying, recording, or otherwise, without the written permission of the copyright owner.
First printing, September 2017
This Volume is a collective effort involving hundreds of technical specialists. It brings together a wealth of information from worldwide 
sources to help scientists, engineers, and technicians solve current and long range problems.
Great care is taken in the compilation and production of this Volume, but it should be made clear that NO WARRANTIES, EXPRESS OR 
IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PUR
POSE, ARE GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM can
not guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by persons having 
technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of ASM’s control, ASM assumes no 
liability or obligation in connection with any use of this information. No claim of any kind, whether as to products or information in this publi
cation, and whether or not based on negligence, shall be greater in amount than the purchase price of this product or publication in respect of 
which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE EXCLUSIVE AND SOLE REMEDY OF BUYER, 
AND IN NO EVENT SHALL EITHER PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER 
OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of the material 
under end use conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended.
Nothing contained in this Volume shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any 
method, process, apparatus, product, composition, or system, whether or not covered by letters patent, copyright, or trademark, and nothing 
contained in this Volume shall be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or as a 
defense against liability for such infringement.
Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International.
Library of Congress Cataloging in Publication Data
ASM International
ASM Handbook
Includes bibliographical references and indexes
Contents: v.1. Properties and selection irons, steels, and high performance alloys v.2. Properties and selection nonferrous alloys and 
special purpose materials [etc.] v.23. Materials for medical devices
1. Metals Handbooks, manuals, etc. 2. Metal work Handbooks, manuals, etc. I. ASM International. Handbook Committee. II.
Metals Handbook.
TA459.M43 1990 620.1’6 90 115
SAN: 204 7586
EISBN: 978-1-62708-134-4 
ISBN 13: 978 1 62708 133 7
ISBN 10: 1 62708 133 X
ASM InternationalW
Materials Park, OH 44073 0002
www.asminternational.org
Printed in the United States of America
Foreword
In this year of renewal at ASM International, it is especially fitting to release Cast Iron Science and
Technology, Volume 1A of the ASM Handbook series. Its focus on improving materials performance is
a key value that ASM International strives to offer its members and those who research, develop, process,
make, and buy cast irons. Volume 1A covers the processing and applications of cast irons, which differ
entiates it from, and supplements, Properties and Selection: Irons, Steels, and High Performance Alloys,
Volume 1, ASM Handbook.
Coverage in this Volume includes fundamentals, primary processing, fabrication, effects of processing
on properties, process and product design, and the engineering properties of specific grades, types, and
product forms of iron castings.
ASM International is grateful for the work and dedication of volunteer editors, authors, and reviewers.
They devoted their time and expertise to develop a reference work that reflects the continuing commit
ment of ASM International to present a publication of the highest technical and editorial quality. The
result is a comprehensive body of knowledge from the world’s leading innovators, researchers, and prac
titioners in the cast iron field that gives readers the tools to solve problems. ASM International is indebted
to Volume Editor Doru M. Stefanescu, a world renowned expert who worked tirelessly to oversee this
undertaking.
William E. Frazier
President
ASM International
William T. Mahoney
Managing Director
ASM International
iii
Preface
“Research isn’t practical.” Neither are babies. They are costly, dirty and
have no practical value. They net no return on the investment for
20 years, and even then they may be a liability rather than an asset.
There are many reasons for not having babies and for not doing
research. The result of yielding to those superficial reasons is the same
in both cases a dim and declining future climaxed by extinction.
H.W. Lownie (foundryman), 1961
Cast iron is probably the most complex alloy used by human civiliza
tion. It includes in its chemical composition more elements than super
alloys, that is, base elements (C, Si, Mn, P, S), alloying elements
(Cu, Sn, Ni, Cr, Mo, V, Al), and minor elements (As, B, Bi, Cd, Pb,
Sb, Se, Te, Ti, Zr). Depending on composition and cooling rate, it soli
difies with either stable or metastable eutectic and with the carbon rich
phase, graphite, in a variety of morphologies, from flake/lamellar to
nodular/spheroidal. Cast iron is the first man made metal matrix com
posite, combining crystalline iron and crystalline graphite. It has a wide
range of properties, including higher specific properties (property/den
sity) than many of its competing materials. For example, cast iron has
higher specific fatigue strength and higher specific tensile strength at
temperatures above 100 �C than aluminum, and all this at a much lower
price. This explains why iron castings represent approximately 70% of
the total tonnage of castings worldwide. Thus, collecting the available
information on the history, science, and technology of cast iron in a sin
gle volume is a worthwhile endeavor. As, at the beginning of human
civilization, iron processing was considered magic, which then evolved
into an art, then technology, and finally science, culminating today with
virtual cast iron, this endeavor is not just worthwhile but also
challenging.
Yet, this book intends to be more than a technical compendium. It
aspires to also acknowledge the history of cast iron, an important
attribute if we care to consider the fast pace of knowledge development.
The Renaissance genius Sir Francistimes.
While the meteorite iron/sky/divine connec
tion is undoubtedly flattering to the metallur
gist, another theory of the initial advent of
iron mastery in human accomplishments was
promulgated. Some iron examples uncovered
in Anatolia suggest that iron was a by product
obtained during smelting of iron containing
copper ores (Ref 6). However, the beginning
of iron metallurgy on an industrial scale was
not possible until the secret of smelting magne
tite or hematite was discovered, followed by the
art of hardening the metal through quenching
(approximately 1200 to 1000 B.C.) in the
mountains of Armenia (Ref 7).
Other sources (Ref 6) place the beginning of
large scale production of iron with the Renn
kilns in eastern Anatolia, at approximately
2000 to 1000 B.C. While iron was still a precious
metal, as attested by the iron artifacts found in the
royal tombs ofAlacahoyuk,Anatolia, and by cune
iform tablets in Assyrian which state that iron was
more valuable than gold, it was increasingly used
to make weapons and tools in addition to luxury
and art objects. The principle of the Renn kiln
involves reduction of the iron ore with charcoal to
obtain sponge iron (loupe or luppe), which is a
mixture of slag, charcoal, pure iron, and unreduced
iron ore. The sponge iron is then forged and
cleaned of residuals to produce a malleable iron.
Early Cast Iron in Mesopotamia and
China
The earliest successful iron founding is gen
erally credited to the ancient Mesopotamian
civilizations (Babylonians, Assyrians, and Chal
deans) many centuries before Christ (Ref 9).
Although the Greeks and Romans understood
the art of casting iron, their early applications
did not compare with the extensive development
of cast iron in China. There is ample evidence
that the Chinese capitalized on the early
evolutionary work, probably passed along to
them by migrating Mesopotamian craftsmen.
The Chinese became the first people to produce
iron castings successfully and regularly as early
as 800 to 700 B.C., with the earliest sand mold
being traced to 645 B.C. (Ref 1). One ancient
document (513 B.C.) refers to a requisition for
272 kg (600 lb) of iron for casting a tripod on
which the criminal code was to be inscribed.
Cast iron plowshares were recorded in 233 B.C.
The oldest cast iron objects found to date
were cast during the Han dynasty (206 B.C.
to 220 A.D.) and include a stove (Fig. 2), an
ink pallet, a vase, a pan, and various fittings.
Cast iron became so popular in China that it
was used not only for home implements but
also for art (Fig. 3), worship objects such as
incense burners and statues, pagoda roof tiles,
and even true cast iron pagodas, such as the
iron pagoda of Yuquan Temple (Fig. 4).
One of the major surviving masterpieces is
the iron lion of Cangzhou, cast in a single mold
(Fig. 5). The technique, also used in ancient
Chinese bronze casting, starts with a clay model
of the sculpture, which is covered with a new
layer of clay after drying. This outer layer of
clay is then cut into pieces and removed before
it dries completely. In the next step, material is
taken off the surface of the inner clay model to
provide room for pouring the iron between the
outer and inner mold. Because casting pro
ceeded in several stages, fault lines were intro
duced into the cast at regular intervals, which
mark the filling height of the mold at successive
casting stages. These fault lines were bridged
with wrought iron rods that were plunged into
the solidifying surface of the iron from the pre
vious pour and then covered in the next pour.
The amazing progress in cast iron technology
that occurred in China is attributed to the devel
opment of melting equipment capable of pro
ducing greater air draft (the box bellows
furnace) and to the abundant supply of the nec
essary raw materials. Evidence suggests that
blast furnaces that convert raw iron ore into
pig iron, which can be remelted in a cupola
Fig. 2 Oldest known cast iron stove, from the Han
dynasty
Fig. 3 Recumbent cast iron lion, 502 A.D.
Fig. 4 Iron pagoda in front of Yuquan Temple in
Dangyang. Built in 1061, it incorporates
38,300 kg (84,400 lb) of cast iron and stands 17.9 m
(58.7) tall.
Fig. 5 The iron lion of Cangzhou, cast in 953 A.D., is
the largest (40 to 50 tons) known old surviving
iron cast artwork in China.
4 / Introduction
 
 
 
furnace to produce cast iron, were operational
in China by 722 to 481 B.C. (Ref 10).
A second reason for the shift from bronze to
iron in China seems to be the understanding of
a process consisting of holding an iron ore/car
bon mixture at low temperature to produce a soft
mass of pure iron (melting point 1530 �C, or
2785 �F), followed by holding this iron at high
temperature in the presence of carbon, to pro
duce a carbon rich iron with a melting point of
1170 �C (2140 �F). In addition, because the iron
ore was rich in phosphorus, and high phosphorus
coal was added during melting, the resulting iron
contained 6 to 7% P, which allowed pouring of
this iron at 980 �C (1795 �F) 100 �C (180 �F)
below the melting point of copper.
Cast Iron in Europe in the
Medieval Ages
While metal casting was known to both the
ancient Greeks and Romans, little evidence of
cast iron was found from that period in Europe.
After the Roman legions departed the island,
iron was smelted in Britain by Anglo Saxon
monks, as attested by a small cast statuette
dated 170 A.D. found in Sussex. In continental
Europe, as Europe descended into the Dark
Ages, the metal casting art was preserved dur
ing the Merovingian dynasty by the Gauls,
famous for their metalworking talent during
the Roman period (Ref 1). Knowledge was kept
secret and transferred almost solely by word of
mouth. It was not until 1122 A.D. that the monk
Theophilus, in his manuscript On Divers Arts,
included some description of foundry practice.
The cast iron of those days was an inferior
material termed “corrupt metal” even as late
as the 15th century, because it was believed that
the melting of iron ruined its properties. Not
surprising, because the iron had very high car
bon content (since charcoal was used as fuel)
and little silicon and thus was very brittle.
The beginning of the progress of blast fur
naces in Europe has been traced to the char
coal fueled Catalan forge developed by the
Moors in the 8th century A.D. The product
was sponge iron (loupe), which was further pro
cessed by forging. This furnace was followed
by improved models in Switzerland, Germany,
and Sweden. The Swedish model used manu
ally operated leather bellows. In 1325, the
water driven bellows was introduced, marking
the beginning of modern iron foundry practice.
The temperature in the furnace was high
enough to allow removal of the slag and
tapping the molten iron into a large basin and
then into smaller and smaller molds, resembling
a sow with suckling pigs, which is probably the
origin of the term pig iron (Ref 1).
Large scale introduction of cast iron in Europe
did not occur until approximately 1200 to 1450
A.D. For more than 400 years, foundry processes
and materials often relied on the methods
described by Biringuccio (Fig. 6), an Italian met
allurgist and author of De la pirotechnia, a
manual on metalworking that was published
posthumously in 1540. This book is credited
with starting the tradition of scientific and
technical literature. It preceded by 14 years the
printing of De re metallica by Georgius Agri
cola. Biringuccio, who is considered the father
of the foundry industry, recommended using
the dregs of beer vats and human urine as bin
ders for molding sand, both of which were in
use well into the 20th century. Development by
Biringuccio of a standard bell scale is one of
the earliest instances on record of the metal
caster and the engineer combining their skills
for the production of perfect castings.
An important cast iron success was the intro
duction of cast iron water pipes in the 15th cen
tury. Apparently, the first one was installed at
the Dillenburg Castle in Germany in 1455,
althoughearlier installations are mentioned.
Early Modern Period (16th
to Mid-18th Century)
A partial chronological list of the advance
ment in cast iron technology and science
achieved after 1500 A.D. is presented in
Table 2.
By the early 17th century, cast iron reached
America. The Virginia Company of London
established Falling Creek Ironworks in 1619,
the first iron production facility in North Amer
ica, which was short lived due to an attack by
Native Americans three years later. However,
cast iron developments continued, as attested
by the Saugus pot shown in Fig. 7, the first
Fig. 6 Vannoccio Biringuccio, as depicted in the
Specola Museum in Florence
Table 2 Chronological list of developments and use of cast iron during the modern period
Date
Development
LocationEarly modern period (16th to mid-18th century)
1619 North America’s first iron furnace is built at Falling Creek, VA, on a branch of the
James River, 100 km (62 miles) from the Jamestown colony.
United States
1642 The first American casting: iron pot made at the Saugus Iron Works in Massachusetts,
America’s first iron metal-casting facility (and second industrial plant)
United States
1664 Flanged cast iron pipes laid at Versailles France
1709 Cast iron produced with coke as fuel, Coalbrookdale England
1715 Boring mill of cannon developed Switzerland
1722 de Reaumur develops whiteheart malleable iron France
Late modern period
1776 Metalcasters Charles Carroll, James Smith, George Taylor, James Wilson, George Ross,
Philip Livingston, and Stephen Hopkins sign the American Declaration of
Independence.
United States
1779 Cast iron used as architectural material: Iron Bridge over the Severn River England
1794 John Wilkinson invents the first metalclad cupola furnace, using a steam engine to provide
the air blast.
England
1809 Centrifugal casting is developed by Eckhardt. England
1825 Aluminum, the most abundant metal in the Earth’s crust, is isolated from aluminum
chloride by Hans Oerstad.
Denmark
1863 Henry Sorby develops metallography after the invention of the microscope in1860. England
1886 Electrolytic refining of aluminum (the Hall-Héroult process) is invented independently
by Charles Hall and Paul Héroult.
United States, France
1908 First attempts at liquid treatment of cast iron with FeSi, Ca, and V by Geilenkirchen Germany
1928 First specification (DIN 1691) for cast iron; classes 140–280 MPa (20–41 ksi) Germany
1931 Augustus Meehan obtains a U.S. patent for the addition of calcium silicide. United States
1935 First scanning electron microscope image by Max Knoll England
1940 Chvorinov develops the relationship between solidification time and casting geometry. Germany
1942 Piwowarsky in Aachen publishes Hochwertiges Gusseisen, the first cast iron “bible.” Germany
1943 Keith Millis discovers that magnesium addition to molten iron produces a spheroidal
graphite structure.
United States
1938–
1949
Patent rights for the production of cast iron with spheroidal graphite granted to Adey
(1938) to Millis, Gagnebin, and Pilling (1949), and to Morrogh (1949)
Germany, United
States, England
1948 Industry’s first ductile iron pipe is cast at Lynchburg Foundry, Lynchburg, VA. United States
1951 Ford Motor Co. in Dearborn converts 100% of its crankshaft production to ductile iron. United States
1956 Formulation of the constitutional undercooling criterion by Chalmers opens the road for
applications of solidification science to metal casting.
Sweden
1965 First scanning electron microscope marketed by the Cambridge Scientific Instrument Co. England
1966 Mathematical theory of eutectic solidification by Jackson and Hunt England
1966 First computer model for the solidification of alloys (cast iron) by W. Oldfield England
1969 Patent rights for the production of cast iron with at least 50% vermicular graphite granted
to Schelleng
United States
1972 Commercialization of austempered ductile iron: a 0.5 kg (1 lb) crankshaft for a refrigerator
compressor produced at Wagner
United States
1976 Foote Mineral Co. and the British Cast Iron Research Association develop compacted
graphite iron.
United States, England
A History of Cast Iron / 5
 
 
 
surviving cast iron artifact produced in Amer
ica. In France, cast iron pipes were installed at
the palace of Versailles by order of King Louis
XIV (Fig. 8); some of these pipes are still being
used today (2016).
An important development occurred in 1709,
when Abraham Darby from Coalbrookdale,
England, initiated the use of coke as a furnace
fuel for iron production. In 1715, Johann Mar
itz, Master Founder at Burgdorf, Switzerland,
developed the procedure of casting cannon
solid and then machining the bore, a technology
further developed by the French. Another sig
nificant French contribution to cast iron during
this period was the development of whiteheart
malleable iron by de Reaumur, which dispelled
the notion that cast iron is an inherently brittle
material and opened the way to the many dis
coveries that the understanding of metallurgy
bestowed on cast iron.
Late Modern Period
The late modern period in human civilization
begins at approximately 1760, when the
Industrial Revolution started in England. It ush
ered the change from muscle power (hand
production methods) to water power and then
steam power (steam engine). New chemical
manufacturing and iron production processes,
the development of machine tools, and the rise
of the factory system were also the hallmarks
of this first Industrial Revolution. Cast iron
tram road rails produced in Coalbrookdale in
1756 replaced wooden rails, and the famous
Iron Bridge was built in 1779 (Fig. 9).
Before the invention of the microscope, only
two types of iron were known, and they were
classified based on the appearance of their frac
ture: white and gray. The strength was limited
to 80 to 100 MPa (12 to 15 ksi). In 1863, Sorby
used a microscope to study polished samples,
enabling metalcasters to microscopically exam
ine metal surfaces and understand the constitu
ents of alloys. Still, cast iron was slow to
develop to the modern, high properties, widely
used material that we know today (2016). The
limited knowledge of the subject is summarized
in the first paper on cast iron to be published in
the newly created Journal of the American
Foundrymen’s Association in 1896 (Ref 11),
which stated, “The physical properties of cast
iron are shrinkage, strength, deflection, set,
chill, grain and hardness. Tensile test should
not be used for cast iron, but should be confined
to steel and other ductile materials. Compres
sion test should be made, but is generally
neglected, from the common erroneous impres
sion that the resistance of a small cube or cylin
der, which is enormous, is always in excess of
loads which can be applied.”
Fig. 7 The Saugus pot (1642), the first casting made in
the Americas
Fig. 8 Sewer pipes in Versailles (1664). The initials
“LF” stand for Louis of France. Source: Ref 1
Fig. 9 The cast Iron Bridge over the Severn River near Coalbrookdale, England (1779). (a) General view. (b) Detailed
view showing surface defects of the castings poured in open molds. Photos taken by the author in 2012
6 / Introduction
 
 
 
The march of cast iron toward higher
mechanical properties achieved a turning point
during the late 1920s and early 1930s, when
the Ross Meehan foundry in Chattanooga, Ten
nessee, discovered the advantages of inoculat
ing iron with controlled additions of calcium
silicide. The initial patent on the process was
issued to Augustus Meehan in 1931. The pro
cess allowed the production of gray iron with
tensile strength up to 500 MPa (72 ksi).
Significant progress was also achieved in
Germany, where, beginning in 1930, Piwo
warsky performed systematic studies of the
use of sodium, calcium, lithium, magnesium,
cerium, strontium, and barium for inoculation
of gray iron. By 1936, Adey was preoccupied
in obtaining spheroidal graphite. Quoting from
the famous book by Piwowarsky (Ref 12),
whose first editionwas published in 1942, Adey
obtained a patent in 1938 for a “process for pro
duction of cast iron of higher strength, charac
terized by a eutectic or hypereutectic cast iron
free of slag inclusions with a minimum content
of 1% Si in which, after fast solidification, the
graphite is whole or in part of spheroidal form
in the metallic matrix.” As can be inferred from
Fig. 10, it appears that the material was malleable
iron with spheroidal graphite obtained through
heat treatment (“thermische vergütung”).
Yet, the quest for an ideal as cast iron with
properties equal or superior to malleable iron
continued. At the 1943 Convention of the
American Foundrymen’s Society, one of the
speakers, J.W. Bolton, addressed the following
question to the audience: “Your indulgence is
requested to permit the posing of one question.
Will real control of graphite shape be realized
in gray iron? Visualize a material, possessing
(as cast) graphite flakes or groupings resem
bling those of malleable iron instead of
elongated flakes.” A few weeks later, in the
International Nickel Company Research Labo
ratory, Keith D. Millis made a ladle addition
of magnesium (as a copper magnesium alloy)
to cast iron and produced spheroidal graphite,
discovering ductile iron, whose expansion in
industry in the following years was explosive
(Ref 5). At the American Foundryman’s Society
annual meeting on May 7, 1948, in Philadelphia,
Millis announced this achievement during a
brief discussion period after a technical presen
tation by H. Morrogh, who independently con
ducted work in England on spheroidizing the
graphite through additions of cerium. This led
to patents by Millis (U.S. Patent 2,485,760 in
1949) and Morrogh (U.S. Patent 2,488,511 in
1949). The major discoveries related to graphite
shape control ended in 1969 with the recogni
tion of compacted graphite iron as a grade in
its own merit through a patent for “cast iron
with at least 50% of the graphite in vermicular
form” granted to R.D. Schelleng. Finally, with
the commercialization of austempered ductile
iron, the strength of cast iron rivaled that of
many steels, as shown in Fig. 11, which sum
marizes the increase in strength of cast iron
over the years.
In approximately 1950, the second Industrial
Revolution started with the advent of transis
tors, computers, and microchips, which helped
to replace and enhance mental effort, made pos
sible the invention of robots to perform danger
ous or boring jobs, triggered major productivity
increases, and decreased demand on natural
resources. The second Industrial Revolution
helped propel cast iron in the body of advanced
materials following the birth and growth of
solidification science and computational model
ing. The formulation of the mathematical corre
lation between casting volume/surface ratio and
solidification time by Chvorinov (Ref 13)
in 1940 had a major impact. Then, Chalmers
(Ref 14) transformed solidification science
from a purely physics discipline into an engi
neering science with his formulation of the con
stitutional undercooling criterion, which opened
the road to the understanding of the effects
of cooling rate on the microstructure of cast
alloys. Two of the most significant advances
in the mathematics of solidification, with major
effect on the engineering science of cast iron,
occurred in 1966 with the publication of two
papers. The first one is the classic paper on
eutectic alloys by Jackson and Hunt (Ref 15),
a rigorous analytical analysis of regular lamel
lar eutectic growth that established the correla
tion between the processing parameters and
microstructure for eutectic alloys, including
cast iron.
The age of virtual cast iron (computational
modeling of microstructure, properties, and
soundness of cast iron) was started by the bril
liancy of scientist W. Oldfield (Ref 16), who
developed a computer model that could calcu
late the cooling curves of gray iron. His seminal
paper was the first attempt to predict solidifica
tion microstructure through computational
modeling and the first attempt to validate such
a model against cooling curves. Nobody ever
remembers the first one to be second in any
human endeavor. Yet, the author of this article
will have to take credit for this position, since
in 1973 he was the first one to continue Old
field’s work (Ref 17). By 1985, solidification
modeling of cast iron became an area of inten
sive research (Ref 18). Simulation of cast iron
microstructure and properties has made gigantic
strides. Today (2016), computer software com
panies offer complete packages that include
integrated simulation of the entire process
(mold filling, solidification, and cooling) using
a micromodeling approach to investigate final
structures and properties of iron casting. Some
models predict graphite morphology (lamellar,
nodular), carbide formation, and microstructure
length scale (eutectic grain size, type and aver
age size of lamellae, or number of nodules).
Fig. 10 Page from the laboratory notebook of C. Adey from 1936, showing malleable iron with spheroidal graphite
Fig. 11 Temporal evolution of the tensile strength of
cast iron. ADI, austempered ductile iron; DI,
ductile iron; CGI, compacted (vermicular) graphite iron;
LG, lamellar graphite
A History of Cast Iron / 7
 
 
 
They can calculate the eutectoid transformation
and thus the final structure and predict proper
ties such as hardness, yield and tensile strength,
and fracture elongation.
Cast Iron—A High-Tech,
Economical, Modern Material
A recent commercial produced by Cleveland
Golf that introduced a new line of golf wedges
stated, “The CG10 wedge is made from a pro
prietary material called carbon metal matrix.
This material, while not a composite, is infused
with 17 times more carbon than traditional car
bon steels. The carbon is infused into micro
scopic spheres suspended within the molecular
structure, creating a matrix that is 10% less
dense and 15% softer than steel. The density
relieving spheres damp vibrations. . .”. The
reader may have guessed, and the published
microstructure confirms, that the material is
nothing else but spheroidal graphite iron, which
is indeed a graphite iron composite, the first
man made composite. This is further confirma
tion that cast iron has achieved the status of a
high tech material. There are more compelling
examples of high performance cast iron parts,
such as large ductile iron castings for the wind
mill industry (e.g., the hub in Fig. 12, frame,
and gearboxes) or ductile iron bodies for naval
engines (Fig. 13).
The application of cast iron in works of art is
as old as cast iron itself. More modern art appli
cations are in architecture. Cast iron architec
ture became a prominent style in the Industrial
Revolution era, when cast iron was relatively
cheap and modern steel had not yet been devel
oped. Ditherington Flax Mill in England, built
in 1796, is the oldest iron framed building in
the world. As such, it is seen as the world’s first
skyscraper and is described as “the grandfather
of skyscrapers.” A famous example is the
Bulgarian Iron Church in Istanbul. The richly
ornamented church is a three domed, cross
shaped basilica with a 40 m (131 ft) high bell
tower (Fig. 14). It was completed in 1898.
The main skeleton of the church was made of
steel and covered by prefabricated cast iron
boards weighing 500 tons that were produced
in Vienna.
Many other examples of cast iron architec
ture survived in London, New York, Boston
(Fig. 15), and many other cities.
Another exciting application of cast iron is in
the art of cooking. Cast iron distributes heat
evenly, favoring the development of the Mail
lard reaction (Ref 19) during cooking, which
is a chemical reaction between amino acids
and reducing sugars that gives browned food
its desirable flavor. Thus, it is one of the best
media for cooking now advertised by such tele
vision celebrities as Alton Brown. The author
of this article is himself a big fan of cast iron
cookware (Fig. 16).
The markets for iron castings includecon
struction, motor vehicles, farm equipment,
mining machinery, engines, valves, pumps,
home appliances, ware, and oil and natural gas
pumping and processing equipment. The reader
is referred to the paper by Prucha et al. (Ref 5)
for a more complete list. These examples
should be convincing, but a more rigorous anal
ysis may be used to fully establish cast iron
credentials.
Fig. 12 Ductile iron hub for large windmill
Fig. 13 Ductile iron cylinder head for a naval engine weighing 83 tons
Fig. 14 Bulgarian St. Stephen Iron Church in Istanbul. Photo taken by the author in 2004
8 / Introduction
 
 
 
Over recent years, aluminum has been the
material of choice for a large number of auto
motive components because of its low density
and lower energy requirements during use and
postuse, compared with ferrous materials.
Automotive aluminum use has grown steadily
for 40 years. A survey of North American auto
makers found that automakers will increase
their use of aluminum from 148 kg (327 lb) in
2009 to 250 kg (550 lb) in 2025, doubling the
aluminum percent of vehicle curb weight from
8 to 16%. Yet, when conducting optimization
analysis on the two competing materials, alumi
num and cast iron, an interesting picture emerges
(Ref 20). The objective of optimization when
selecting a material for a particular application
is to optimize a number of performance metrics
(P) in a particular product. Typical metrics for
the problem of interest are cost, mass, fatigue
resistance, strength, stiffness, and so on. A first
approach to optimization is to directly compare
selected properties of the competing materials
or, when the weight is important, as in the case
of automotive parts, the specific property of the
material (property/density ratio).
For example, fatigue strength can be used as
an optimization parameter. The ability of a
material to withstand long term cyclic stress is
typically described by the stress (S)/number of
cycles (N) curve. As shown in Fig. 17(a),
Fig. 15 Cast iron façade on a building in Boston. Photo taken by the author in 2002
Fig. 16 Cast iron cookware, produced by Lodge Manufacturing, in the author’s kitchen
Fig. 17 Optimization through direct comparison of
properties. (a) Typical applied stress (S)/
cycles to failure (N) curves for cast aluminum alloys and
ductile iron (DI). (b) Typical specific stress/cycles to
failure curves for cast aluminum alloys and DI. r,
density. Source: Ref 20
A History of Cast Iron / 9
 
 
 
aluminum alloys exhibit a lower S N curve than
ductile iron (DI). In addition, cast iron exhibits
a fatigue limit (stress under which failure does
not occur, regardless of the number of cycles),
while aluminum does not. More importantly,
the specific stress of DI is superior to that of
aluminum alloys when the number of cycles
exceeds 107 (Fig. 17b).
The more detailed analysis presented in
Fig. 18 shows that die cast alloys have similar
fatigue resistance to ferritic ductile irons, but
even the premium A357 die cast alloy cannot
compete with pearlitic iron. The fatigue
strength of tempered and austempered DI
exceeds several times that of solution treated
as cast aluminum alloys.
Another property of particular interest for
automotive parts is the strength at elevated tem
peratures. As shown in Fig. 19, at temperatures
above 200 �C (390 �F), the specific strength of
ductile iron rapidly overtakes that of aluminum
Fig. 18 Specific fatigue strength of selected solution-
treated cast aluminum alloys and ductile
iron. r, density; SC, sand cast; DC, die cast; DI, ductile
iron; F, ferritic; FP, ferritic-pearlitic; P, pearlitic; T,
tempered; AUST, austempered. 355 = Al7Si; 356 =
Al7Si0.4Cu; 357 = Al7Si0.8Cu. Source: Ref 20
Fig. 19 Influence of temperature on the specific tensile
strength of aluminum alloys and ductile iron
(DI). UTS, ultimate tensile strength; r, density. Source: Ref 20
AUS CG
Gray
AI DIΔ
0.30.10
0
0.2
0.2
0.4
0.4
C
os
t ×
 ρ
/E
ρ/E
DDDDDD
1
1
0.5
0.5
0
0
1.5
1.5
C
os
t ×
 ρ
/σ
y1/
2
ρ/σy
1/2
2.50.50
0
1.5
2
3.5
4
C
os
t ×
 ρ
/E
1/
3
ρ/E1/3
DD
1 2 3 4
D DDD
0.2 0.4 0.6
0.4
0
0
0.8
0.8
C
os
t ×
 ρ
/σ
y2/
3
ρ/σy
2/3
(a)
(c) (d)
(b)
Fig. 20 Comparison between cast iron (AUS, austempered; DI, ductile iron; CG, compacted graphite iron) and
aluminum alloys for multiobjective optimization using mass-cast as performance metrics. (a) Tie, stiffness
prescribed. (b) Panel, stiffness prescribed. (c) Panel, strength prescribed. (d) Beam, strength prescribed. The cost is in
$/kg; density (r) is in Mg/m3; Young’s modulus (E) is in GPa; and yield strength (sy) is in MPa. Source: Ref 20
Table 3 Materials indices for different
applications
Function Example Objective Constrain Index(a)
Tie Cable support Minimum
weight
Stiffness r/E
Beam Aircraft wing Minimum
weight
Stiffness r/E1/2
Panel Automobile
door
Minimum
weight
Stiffness r/E1/3
Beam Auto
suspension
arm
Minimum
weight
Strength r/sy
2/3
Panel Table top Minimum
weight
Strength r/sy
1/2
(a) r, density; E, Young’s modulus; sy, yield strength. Source: Ref 21
1.20×108
8.00×107
6.00×107
4.00×107
P
ro
du
ct
io
n,
 m
et
ric
 to
ns
2.00×107
1.00×108
0
2006 2008 2010 2012 2014 2016
Magnesium,
0.2%
Aluminum,
15.5% Other nonferrous, 2.8%
Cast iron,
70.9%
Steel,
10.8%
Total
Cast iron
Steel
Aluminium
Year(a) (b)
Fig. 21 Worldwide cast iron production. (a) Evolution of tonnage of various casting alloys between 2007 and 2014. (b) Share of total production of various casting alloys in 2014
10 / Introduction
 
 
 
alloys. Thus, for high temperature applications
(e.g., engine parts), ductile iron is a better
choice than aluminum.
A more detailed optimization analysis must
include the particular function of the product.
Then, the metrics depend on the geometry of
the product and the constraints imposed on it.
More complicated equations that define a mate
rial index are developed (Ref 21), as exempli
fied in Table 3. The values of the performance
metric for competing materials scale with the
material index. By using this concept, selection
of a material becomes a simple case of choos
ing materials with the smallest index character
izing the performance metrics. For example,
examining the data in Fig. 18, the best material
is austempered DI having a density/fatigue
strength of 0.14 to 0.18, while sand cast alumi
num alloys are in the range of 0.41 to 0.5.
When there are two or more optimization
objectives, solutions rarely exist that optimize
all at once. One way of optimizing several objec
tives is to compare the materials in a P1 P2
graph, where P1 and P2 are the metrics of the
two objectives. An example is provided in
Fig. 20 for mass cost optimization for four dif
ferent applications. The slopes of the parallel
lines on the graphs are drawn such that a unit
increase in P1 corresponds to a unit increase in
P2. For all applications in this example, cast iron
is either clearly superior or slightly superior to
aluminum alloys, because the values for cast iron
are closer to the origin of the graph show lower
cost for higher indexes.
This analysis demonstrates that in applica
tions where mass and cost are the objective of
optimization, cast iron should be selected over
aluminum alloys. The main reason why alumi
num is replacing cast iron in automotive appli
cations seems to be the inability or lack of
interest of iron foundries to produce lightweight
iron castings, that is, iron castings with thin
walls, despite the significant advances made in
this direction (see the articles “Thin Wall Gray
Iron Castings” and “Thin Wall Ductile Iron
Castings” in this Volume).
To conclude this section, it is useful to pro
vide an analysis of current trends in the world
wide casting production. As shown in Fig. 21
(a), the tonnage of all casting alloys has
increased by almost 11% between 2007 and
2014. While the percentage of cast iron from
the total tonnage has slightly decreased in
2014 compared with 2007, it is stillat more
than 70%, by far the highest in the competition
of casting alloys (Fig. 21b).The share of alumi
num over the same time period has increased
from 13.4 to 15.5%, while that of magnesium
has decreased from 0.3 to 0.2%.
Today (2016), cast iron remains the most
important casting material. The main reasons
for cast iron longevity are the wide range of
mechanical and physical properties associated
with its competitive price. If all ferrous alloys
are considered, their share of the world casting
production is above 81%. Thus, as far as
this author is concerned, we are still in the
Iron Age.
REFERENCES
1. B.L. Simpson, History of the Metal Casting
Industry, 2nd ed., American Foundry
men’s Society, Des Plaines, IL, 1997
2. Ö. Bilgi, H. Özbal, U. Yalçin, Castings of
Copper bronze, in Anatolia, cradle of cast
ings, ed. Ö. Bilgi, Graphis Matbba
3. M. Goodway, History of Casting, Casting,
Vol15,MetalsHandbook, 9thed.,D.M.Stefa
nescu, Ed., ASM International, 1988, p 15 23
4. Timeline of Casting Technology, Mod.
Cast., Cast Expo Issue, May 2005
5. T.E. Prucha, D. Twarog, and R.W. Monroe,
History and Trends of Metal Casting, Cast
ing, Vol 15, ASM Handbook, ASM Interna
tional, 2008, p 3 154
6. U. Yalçin, Iron Technology in Antiquity,
Anatolia, Cradle of Castings, Ö. Bilgi, Ed.,
Graphis Matbba, Istanbul, 2004, p 221 224
7. M. Eliade, The Forge and the Crucible,
The University of Chicago Press, 1978
8. T.C. Mitchell, Tubal cain, New Bible Dic
tionary, London, IVF, 1962, p 1302
9. C.F. Walton, The Gray Iron Castings
Handbook, Gray Iron Founders Society,
Cleveland, OH, 1958
10. D.B. Wagner, The State and the Iron Indus
try in Han China, Copenhagen: Nordic
Institute of Asian Studies Publishing, ISBN
87 87062 83 6, 2001
11. J. Am. Foundrymen’s Assoc., Vol 1, 1896
12. E. Piwowarsky, Hochwertiges Gusseisen,
Springer Verlag, 1951
13. N. Chvorinov, Theory of the Solidification of
Castings,Giesserei, Vol 27, 1940, p 177 186
14. B. Chalmers, Trans. AIME, Vol 200, 1956,
p 519
15. K.A. Jackson and J.D. Hunt, Trans. Metall.
Soc., Vol 236, 1966, p 1129
16. W. Oldfield, ASM Trans., Vol 59, 1966,
p 945
17. D.M. Stefanescu, Ph.D. Dissertation, Poli
tehnica University of Bucharest, Romania,
1973
18. D.M. Stefanescu, Metall. Mater. Trans. A,
Vol 38, 2007, p 1433 1447
19. L.C. Maillard, Formation of Melanoidins in
a Methodical Way, Compt. Rend., Vol 154,
1912, p 66
20. D.M. Stefanescu and R. Ruxanda, Light
weight Iron Castings Can They Replace
Aluminum Castings? Proceedings of the
65th World Foundry Congress, C.P.
Hong et al., Ed., The Korean Foundry
men’s Society, Seoul, Korea, 2002,
p 71 77
21. M.F. Ashby, Multi Objective Optimization
in Material Design and Selection, Acta
Mater., Vol 48, 2000, p 359 369
A History of Cast Iron / 11
 
 
 
Classification and Basic Types of
Cast Iron*
Revised and updated by Doru M. Stefanescu, The Ohio State University and The University of Alabama
THE TERM CAST IRON, like the term steel,
identifies a large family of ferrous alloys. Cast
irons are multicomponent ferrous alloys. They
contain major (iron, carbon, silicon), minor
(0.1%) ele
ments. Cast iron has higher carbon and silicon
contents than steel. Because of the higher car
bon content, it solidifies with a eutectic. The
structure of cast iron, as opposed to that of
steel, exhibits a carbon rich phase. Depending
primarily on composition, cooling rate, and
melt treatment, the carbon rich phase may be
graphite (Gr) or iron carbide (cementite,
Fe3C). Cast iron may solidify according to
the thermodynamically metastable iron iron
carbide (Fe Fe3C) system or the stable iron
graphite (Fe Gr) system. Referring strictly to
the binary Fe Fe3C or Fe Gr system, cast iron
can be defined as an iron carbon alloy with more
than 2% C. However, because silicon and other
alloying elements considerably change the maxi
mum solubility of carbon in austenite (g) and in
the eutectic, a more general definition of cast iron
is that it is an iron carbon base alloy that solidi
fies with eutectic. Alloys with less than 2% C
can exhibit a eutectic structure and still belong
to the family of cast iron.
The presence of higher amounts of silicon in
cast iron as compared to steel produces signifi
cant differences in the solidification and cool
ing to room temperature of the two classes of
alloys. As shown in Fig. 1, silicon changes all
the characteristic compositions and tempera
tures (eutectic, eutectoid, maximum solubility
in austenite). The eutectic and eutectoid tem
peratures change from a fixed value to a range.
This significantly affects microstructure evolu
tion during solidification and subsequent
cooling. A detailed analysis of the effect of Si
and other elements on the Fe C diagram is
provided in the article “Thermodynamics
Principles as Applied to Cast Iron” in this
Volume.
ASM Handbook, Volume 1A, Cast Iron Science and Technology
D.M. Stefanescu, editor
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
* Revised from D.M. Stefanescu, Classification and Basic Metallurgy of Cast Iron, Properties and Selection: Irons, Steels, and High-Performance Alloys, Vol 1, ASM Handbook, ASM
International, 1990, p 3–11.
0 2.5 5.0 7.5 10.0
Melt + δ – S.S.
1498 °C
1152 °C
1145 °C
E�
E�
C´
D
?
F�
K�
K
F
D
S�P�
S
0
P
C
G 910 °C
738 °C ± 3°
723 °C ± 2°
760 °C
M
N 1400 °C
7539°C
1600
1300
1100
800
600
Melt + 
δ – solid solution
δ - solid
α + δ – S.S.
δ + δ – S.S.
δ – S.S.
α – S.S.
Melt + Fe3C 
& graphite
Melt
A H
B
J
%C=1.30+2.57·10–3+°C
Carbon content, at.%
12.5 15.0 17.5 20.0 22.5 25.0
1600
1500
1400
1300
1200
1100
1000
900
800
700
600
0 0.5
0
(a)
(b)
25
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
, °
C
50 75 100
1 1.5 2 2.5 3
Carbon content, wt%
Iron carbide content, wt%
α – solid solution + Fe3C & graphite
δ – solid solution + Fe3C & graphite
3.5 4 4.5 5 5.5 6 6.5 7
1500 2.4% Si
Te
m
pe
ra
tu
re
, °
C
Te
m
pe
ra
tu
re
, °
F
Carbon content, %
1200
900
600
2730
2190
1650
1110
0 0.5 1.0 1.5 2.0
δ+γ+L
δ+L
γ+L
γ +Cγ
γ+L+C
α+γ+C+γ
α
α
α+C
L+C
Lδ+γ
2.5 3.0 3.5 4.0
δ
Fig. 1 Effect of silicon on Fe-Fe3C equilibrium diagram. (a) Fe-Fe3C equilibrium diagram. Source: Ref 1 (b) Isoplet
section of ternary Fe-Fe3C-Si diagram at 2% Si. Source: Ref 2
DOI: 10.31399/asm.hb.v01a.a0006294
 
 
 
The formation of stable or metastable eutec
tic is a function of many factors, including the
nucleation potential of the liquid, its chemical
composition, and the cooling rate. The first
two factors determine the graphitization poten
tial of the iron. A high graphitization potential
will result in irons with graphite as the car
bon rich phase, while a low graphitization
potential will result in irons with iron carbide.
Schematic representations of the structure of
the common types of commercial cast irons,
as well as the processing required to obtain
them, are shown in Fig. 2 and Fig 3 (after
Ref 3).
The two basic types of eutectics, the stable
austenite graphite or the metastable austenite
iron carbide, have wide differences in their
mechanical properties, such as strength, hard
ness, toughness, and ductility. Therefore, the
basic scope of the metallurgical processing of
cast iron is to manipulate the type, amount,
and morphology of the eutectic to achieve the
desired mechanical properties.
Classification
A number of criteria can be used for the
classification of cast iron:
� Fracture aspect
� Graphite shape
� Microstructure of the matrix
� Commercial designation
� Mechanical properties
Classification by Fracture
Historically, the first classification of cast
iron was based on its fracture. Two types of
iron were initially recognized:
� White iron: exhibits a white, crystalline frac
ture surface because fracture occurs along
the iron carbide plates; it is the result of
metastable solidification (Fe3C eutectic)
� Gray iron: exhibits a gray fracture surface
because fracture occurs along the graphiteplates (lamellae, flakes); it is the result of
stable solidification (Gr eutectic)
Classification by Graphite Shape and
Microstructure of the Matrix
With the advent of metallography, and as the
body of knowledge pertinent to cast iron
increased, other classifications based on micro
structural features became possible:
� Graphite shape: lamellar (flake) graphite
(LG), spheroidal (nodular) graphite (SG),
compacted (vermicular) graphite (CG), and
temper graphite (TG); temper graphite results
from a solid state reaction (malleabilization).
Additional graphite shapes include coral
graphite and the newly introduced superfine
interdendritic graphite (Ref 4). Some typical
examples of graphite shapes are given in
Fig. 4. Note that, when examined in three
dimensions, LG is actually interconnected
graphite plates within a spherical grain
(Ref 5).
Solid-state transformation
(cooling through
eutectoid interval)
Graphite shape depends
on minor elements
Spheroidal
Slow
Gray
cast
iron
Pearlite + graphite
(αFe + Fe3C)
Ferrite + graphite
(αFe)
Fast
Compacted
γ + graphite
γ + graphite
γ + Fe3C + graphite
Pearlite + Fe3Cγ + Fe3C
γ + Fe3Cγ + Fe3C
Mottled cast iron
White iron
Flake
Solidification Medium
Graphitization
potential
Liquid
cast iron
(iron-carbon-
alloy)
High
Low
Solid-state transformation
(cooling through
eutectoid interval) Reheat above
eutectoid interval
Hold above
eutectoid
interval
Cool
through
eutectoid
interval
Fast
Pearlite + temper graphite Ferrite + temper graphite
Malleable iron
Slow
Fig. 2 Basic microstructures and processing for obtaining common commercial cast irons. Source: Ref 3
Fig. 3 Basic microstructures and processing of special cast iron (B: Bainite, M: Martensite, ADI: austempered ductile
iron). Source: Ref 3
Classification and Basic Types of Cast Iron / 13
 
 
 
� Matrix: ferritic, pearlitic, austenitic, marten
sitic, bainitic (austempered)
Depending on the chemical composition, a
variety of graphite shapes, substantially differ
ent than those introduced in Fig. 4, may be
found, as illustrated in Fig. 5. The correspon
dence between the ASTM International and
International Organization for Standardization
(ISO) graphite shapes is given in Table 1.
Lamellar graphite is further subdivided into
five categories, as shown in Fig. 6. Type A
graphite occurs in well inoculated irons. Type
B graphite appears at moderate rates of cooling
and may indicate marginal inoculation. Type C
graphite occurs in hypereutectic irons. Type D
graphite is normally associated with high cool
ing rates in thin sections. Type E graphite is
normally seen in strongly hypoeutectic irons.
Graphite shape is the single most important
factor affecting the mechanical properties of
cast iron, as shown in Fig. 7, which compares
the tensile strength of irons with different
graphite shapes.
Classification by Commercial
Designation
The classification based on graphite shape
and/or matrix is seldom used by the floor foun
dryman. The most widely used terminology is
the commercial one. A first division can be
made into two categories:
� Common cast irons: for general purpose
applications; they are unalloyed or low alloyed
� Special cast irons: for special applications,
generally high alloy
The correspondence between commercial
and microstructural classification, as well as
the final processing stage in obtaining common
cast irons, is given in Table 2.
Gray Cast Iron (Lamellar Graphite Iron,
or LGI). These irons have the carbon rich
phase in the form of lamellar graphite. The
graphite lamellae are interconnected within the
eutectic grain. Gray iron has good machinabil
ity because the graphite helps break the turning
chip. It also has good wear resistance and vibra
tion damping ability, high thermal conductiv
ity, and, as graphite, it is a lubricant and can
retain lubricants.
Ductile Cast Iron (DI) (Spheroidal Graph
ite Iron, or SGI). Ductile iron, which is also
known as nodular iron, is produced from the
same types of raw material as gray iron but
usually requires slightly higher purity (in par
ticular, low sulfur). To produce spheroidal
graphite, small amounts of magnesium (e.g.,
0.04% Mg) and/or cerium are added to the
melt during the liquid treatment of the iron.
The main advantage of ductile iron over gray
iron is its combination of high strength and
ductility. Ferritic SGI may have elongations
10 μm10 μm 50 μm
(a)
(b)
(c)
(d)
Fig. 4 Graphite shapes in cast iron. Left column: optical microscopy, unetched; right column: scanning electron
microscopy, deep etched. (a) Lamellar (flake) graphite. Source: Ref 5. (b) Superfine interdendritic graphite.
Source: Ref 4. (c) Compacted graphite. (d) Spheroidal graphite
14 / Introduction
 
 
 
of approximately 20% combined with tensile
strength of 415 MPa (60 ksi), as compared to
only approximatley 0.6% elongation for a gray
iron of comparable strength. Martensitic ductile
irons with tensile strengths of approximately
830 MPa (120 ksi) exhibit at least 2% elonga
tion, and the newer austempered ductile irons
exhibit in excess of 5% elongation at even
higher tensile strengths (1000 MPa, or 145 ksi).
Compacted graphite iron (CGI) is also
known as vermicular graphite iron. It is charac
terized by graphite interconnected within the
eutectic cell, similar to lamellar graphite in gray
iron. However, CG is coarser and has rounded
tips when viewed on a metallographic sample
(ASTM type IV). Both the structure and the
properties can be considered roughly intermedi
ate between those of gray iron and ductile iron.
The combination of higher mechanical proper
ties than gray iron with higher thermal con
ductivity than ductile iron makes it preferable
to either gray or ductile iron in applications
such as disc brake rotors and diesel engine
heads and motor blocks. The CGI is produced
through liquid treatment similar to that of SGI,
but with lower magnesium content (e.g.,
0.02% Mg). Undertreatment may result in gray
iron, while overtreatment may produce high
nodularity. Thus, because of the narrow win
dow for magnesium, the process is more diffi
cult to control.
Malleable iron is produced by the heat
treatment of white cast iron. During this pro
cess, the iron carbide (cementite) of white
iron decomposes in austenite and temper
graphite. Subsequent slow cooling transforms
the austenite into ferrite or pearlite, depending
on the cooling rate. The ductility and toughness
of malleable iron is close to that of ductile iron.
Because of heat treatment constraints, mallea
ble iron is limited to section sizes up to approx
imately 100 mm (4 in.) thick. In recent years,
malleable irons have been replaced by the more
economically processed ductile irons for many
applications.
White Cast Iron. White iron solidifies when
the carbon in solution in the molten iron does
not precipitate as graphite upon solidification
but remains combined with the iron as iron car
bides. White irons are hard and brittle. They
have high compressive strength and good
strength and hardness at elevated temperature.
The high amount of carbides provides excellent
resistance to wear and abrasion.
Special cast irons differ from the common
cast irons mainly in the higher content of
alloying elements (>3%), which promote micro
structures having special properties for elevated
temperature applications, corrosion resistance,
and wear resistance. A classification of the main
types of special cast irons and their main proper
ties is shown in Fig. 8.
The United States specifications for iron cast
ings are summarized in Table 3.
Principles of the Metallurgy
of Cast Iron
The goal of the metallurgist is to design a
process that will produce a sound casting with
a structure that will yield the expected
Fig. 5 Typical graphite shapes after ASTM A247. I, spheroidal graphite; II, imperfect spheroidal graphite; III, temper
graphite; IV, compacted graphite; V, crab graphite; VI, exploded graphite; VII, flake graphite
Fig. 6 Typical flake (lamellar) graphite shapes
specifiedin ISO 945-1 (equivalent to ASTM
A247). (a), uniform distribution, random orientation; (b),
rosette groupings; (c), primary graphite, also called kish
graphite (superimposed flake sizes, random orientation);
(d), undercooled graphite (interdendritic segregation with
random orientation); (e), interdendritic segregation with
preferred orientation
Table 1 ASTM International and equivalent
International Organization for
Standardization (ISO) classification of
graphite shapes in cast iron
ASTM
A247
ISO/R 945-
1969 (E) Description
I VI Nodular (spheroidal) graphite
II VI Nodular (spheroidal) graphite,
imperfectly formed
III IV Aggregate, or temper carbon
IV III Quasi-flake graphite
V II Crab-form graphite
VI V Irregular or “open”-type nodules
VII I Flake graphite
Classification and Basic Types of Cast Iron / 15
 
 
 
mechanical properties. The two basic types of
eutectics in common cast irons the stable aus
tenite Gr or the metastable austenite Fe3C
have wide differences in their mechanical prop
erties, such as strength, hardness, toughness,
and ductility. Therefore, the basic scope of the
metallurgical processing of cast iron is to
manipulate the type, amount, and morphology
of the eutectic to achieve the desired mechani
cal properties. This requires knowledge of the
structure properties correlation for the alloy
under consideration, as well as the factors affect
ing the structure. When discussing the metal
lurgy of cast iron, the main factors of influence
on the structure that must be addressed are:
� Chemical composition
� Liquid (molten metal) treatment
� Cooling rate
� Heat treatment
Chemical Composition
All the elements present in the chemistry of
an iron affect its graphitization potential
(whether the iron solidifies with a Gr eutectic
or a Fe3C eutectic) and the room temperature
matrix. The effect of various elements on the
graphitization potential depends on whether
they increase carbon solubility in the melt
(carbide stabilizers) or decrease it (graphite
stabilizers). It can be estimated thermodynam
ically from the effect of the element (X) on the
solubility of carbon in the molten ternary
Fe C X alloy (see details in the article “Ther
modynamics Principles as Applied to Cast
Iron” in this Volume). A high negative solubil
ity factor (the solubility factor is the ratio
between the change in solubility of carbon
upon addition of a third element, and the
amount of element added) implies a high
graphitization potential (Gr forming ten
dency), while a high positive factor indicates
a low graphitization potential (Fe3C forming
tendency). Table 4 from Ref 7 presents some
of these values for a number of elements com
mon in cast iron.
Although listed as a graphitizer (which is
true thermodynamically), phosphorus also acts
as a matrix hardener. Above its solubility level
in austenite (~0.08%), phosphorus forms a very
hard ternary eutectic. While manganese is a
carbide promoter, it can combine with sulfur.
The resultant manganese sulfides act as nuclei
for lamellar graphite. In industrial processes,
nucleation phenomena may sometimes override
solubility considerations.
For common cast iron, the main elements of
the chemical composition are carbon and sili
con. Figure 9 from Ref 8 shows the range of
carbon and silicon for common cast irons as
compared with steel. It is apparent that irons
have carbon in excess of the maximum solubil
ity of carbon in austenite, which is shown by
the lower dashed line. High carbon content
increases the amount of graphite or Fe3C. High
carbon and silicon contents increase the graphi
tization potential of the iron as well as its
castability.
The combined influence of carbon and sili
con on the structure is usually taken into
account by the carbon equivalent (CE) calcu
lated using the solubility factors in Table 4:
CE %Cþ 0:31 %Siþ 0:33 %P 0:029 %Mn
þ 0:41 %S (Eq 1)
where the percent symbol signifies mass% of
the element. Additional information on carbon
equivalent is available in the article “Thermo
dynamics Principles as Applied to Cast Iron”
in this Volume.
In foundry practice, an additional carbon
equivalent, the carbon equivalent liquidus
(CEL), is used to estimate the composition of
the iron through thermal analysis. The CEL is
calculated as:
CEL %Cþ 0:25 %Siþ 0:5 %P (Eq 2)
Liquid Treatment
After melting, it is common practice to add
specially formulated alloys to the molten metal
in the furnace, in the pouring ladle, or in the
mold. This operation is called liquid treatment.
Fig. 8 Classification of special high-alloy cast irons. Source: Ref 6
Table 2 Classification of cast iron by commercial designation, microstructure, and fracture
Commercial designation Carbon-rich phase Matrix(a) Fracture Final structure after
Gray iron Lamellar graphite P Gray Solidification
Ductile iron Spheroidal graphite F, P, A Silver-gray Solidification or heat treatment
Compacted graphite iron Compacted vermicular graphite F, P Gray Solidification
White iron Fe3C P, M White Solidification and heat treatment(b)
Mottled iron Lamellar Gr + Fe3C P Mottled Solidification
Malleable iron Temper graphite F, P Silver-gray Heat treatment
Austempered ductile iron Spheroidal graphite At Silver-gray Heat treatment
(a) P, pearlite; F, ferrite; A, austenite; M, martensite; At, austempered (bainite). (b) White irons are not usually heat treated, except for stress relief
and to continue austenite transformation.
Fig. 7 Influence of graphite morphology on the stress-
strain curve of several cast irons
16 / Introduction
 
 
 
It is of paramount importance in the processing
of these alloys, because it can dramatically
change the nucleation and growth conditions
during solidification. As a result, graphite mor
phology, and therefore properties, can be signif
icantly affected.
There are two types of liquid treatments:
� Inoculation: Its goal is to increase the num
ber of nuclei during solidification.
� Modification: Its main purpose is to change
the morphology of the eutectic either by
changing the graphite shape (e.g., from LG
to SG) or by promoting Gr austenite eutectic
solidification over the Fe3C austenite eutec
tic (increase graphitization potential).
Typical alloys used for inoculation of both LG
and SG irons are based on ferrosilicon that con
tains any number of other elements, such as cal
cium, aluminum, barium, strontium, cerium,
and so on. The main results of a good inocula
tion are decreased chill and higher number of
eutectic grains or graphite nodules. Typical
additions consist of 0.15% of the weight of
the melt for high efficiency inoculants, to
0.4% for standard inoculants.
It has been demonstrated that lamellar graph
ite nucleates on MnS or complex (MnX)S com
pounds that have low crystallographic misfit
with graphite (Ref 9 12).
Because inoculation is based on the crea
tion of regions of chemical nonhomogeneities
in the melt, and because these nonhomogene
ities are unstable, the effect of inoculation
disappears in time (fading of inoculation).
Figure 10 shows the effect of time before
pouring on the number of eutectic grains
(cells) as well as the efficiency of various
inoculants.
Modification of the eutectic morphology is
achieved by addition of some minor elements.
The most widely used element for the
production of spheroidal graphite is magne
sium. Because the melting point of magnesium
(649 �C, or 1200 �F) is much lower than that
of cast iron, magnesium vaporizes on contact
with the liquid iron and bubbles rapidly to
the surface of the melt. The burning of magne
sium in contact with the atmosphere results in
a violent reaction that produces fumes
and light.
The generic influence of various elements
on graphite shape is given in Table 5. The ele
ments in the first group, the spheroidizing ele
ments, can change graphite shape from flake
through compacted to spheroidal. The anti
spheroidizing elements will revert the process,
degenerating the graphite shape for spheroidal
to some less compact shape.
The most accepted theory on nucleationof
ductile iron stipulates that SG nuclei are
Fig. 9 Carbon and silicon composition ranges of
common cast irons and steel. Source: Ref 8
Table 4 Solubility factors of various third
elements for carbon saturated Fe C X melts
Graphite stabilizer Carbide stabilizer
Element Solubility factor Element Solubility factor
B 0.54 Ti +0.159
C 0.61 V +0.105
Al 0.22 Cr +0.064
Si 0.31 Mn +0.029
P 0.33 Nb +0.058
S 0.41 Mo +0.014
Ni 0.051
Cu 0.076
Sn 0.110
Source: Ref 7
0
0
2
4
6
8
10
12
4 8 12 16 20
Time after inoculation, min
E
ut
ec
tic
 c
el
ls
 m
m
–2
FeSiBa
FeSi
FeSiCe
FeSiSr
Fig. 10 Fading of inoculation
Table 3 Standard specifications for iron castings
Material Standard Characteristic
Gray iron ASTM A48 Gray iron castings
ASTM A74 Cast iron soil pipe and fittings
ASTM A126 Gray iron castings for valves, flanges, and pipe fittings
ASTM A159, SAE J431 Automotive gray iron castings
ASTM A278, ASME SA278 Gray iron castings for pressure-containing parts for temperatures up to 345 �C (650 �F)
ASTM A319 Gray iron castings for elevated temperatures for non-pressure-containing parts
ASTM A823 Statically cast permanent mold castings
ASTM A834 Common requirements for iron castings for general industrial use
High alloy gray and white iron ASTM A436 Austenitic gray iron castings
ASTM A518 Corrosion-resistant high-silicon iron castings
ASTM A532 Abrasion-resistant white iron castings
Compacted graphite iron ASTM A842 Compacted graphite iron castings
Malleable iron ASTM A47, ASME SA47 Ferritic malleable iron castings
ASTM A197 Cupola malleable iron
ASTM A220 Pearlitic malleable iron
ASTM A338 Malleable iron flanges, pipe fittings, and valve parts for railroad, marine, and other heavy-duty service up to 345 �C (650 �F)
ASTM A602, SAE J158 Automotive malleable iron castings
Ductile iron ASTM A395, ASME SA395 Ferritic ductile iron pressure-retaining castings for use at elevated temperatures
ASTM A439 Austenitic ductile iron castings
ASTM A476, ASME SA476 Ductile iron castings for paper mill dryer rolls
ASTM A536, SAE J434 Ductile iron castings
ASTM A571, ASME SA571 Austenitic ductile iron castings for pressure-containing parts suitable for low-temperature service
ASTM A874 Ferritic ductile iron castings suitable for low-temperature service
ASTM A897 Austempered ductile iron castings
Classification and Basic Types of Cast Iron / 17
 
 
 
sulfides (MgS, CaS) covered by magnesium
silicates (e.g., MgO�SiO2) or oxides that have
low potency (large disregistry). After inocula
tion with FeSi that contains elements such as
aluminum, calcium, strontium, or barium, hex
agonal silicates (MeO�SiO2 orMeO�Al2O3�2SiO2)
form at the surface of the oxides, with coherent/
semicoherent low energy interfaces between sub
strate and graphite (Ref 10). However, recent
research shows that many other types of inclusions
can serve as nuclei for SG (see the article “Micro
structure Evolution during the Liquid/Solid Trans
formation in Cast Iron” in this Volume).
Cooling Rate
The cooling rate (section size of the casting)
has a major influence on the microstructure and
thus on the mechanical properties. A high cool
ing rate refines the structure (finer dendrites,
higher number of eutectic grains or graphite
nodules) but also promotes higher carbide for
mation (chilling tendency). The effect of cool
ing rate is specific to the type of cast iron
(SG, CG, LG, or temper carbon).
Heat Treatment
Heat treatment can significantly alter the solid
ification microstructure, with corresponding
changes in mechanical properties. Specific heat
treatments are used for the various classes of cast
irons and are discussed in greater detail in articles
in the Section “Heat Treatment” in this Volume.
Process Control
The most important part of process control is
melt control, because this guarantees the deliv
ery of the desired solidification microstructure.
Typical methods of melt control include chem
ical analysis of melt samples (liquid or solid),
thermal analysis (TA), linear displacement
analysis (LDA), and evaluation of macro/
microstructure (wedge test for chill in cast iron,
nodularity examination in a polished cross sec
tion of a cylindrical bar for ductile iron, rapid
automated metallographic examination).
Thermal analysis that consists of recording
and analyzing the cooling curve of the alloy
of interest has developed into a widespread
on line method for melt control. Originally,
TA was used only for evaluating the chemical
composition of the iron. Using the correlation
between the liquidus temperature (the tempera
ture at which the austenite begins to solidify
from the melt) and the CEL in a standard sand
cup, one can calculate CEL (Ref 13):
TLA 1623:6 112:36 CEL (Eq 3)
Then, using Eq 2 and an additional equation,
the carbon and silicon content can be obtained.
However, the cooling curve contains much
more information than the composition of the
melt. Because the cooling curve shape is
affected by the heat transport from the casting
to surroundings and by phase transformations
during cooling, the cooling curve includes the
genetic algorithm of the solidifying metal. An
example of such a cooling curve and some of
the parameters of interest is presented in
Fig. 11. Through the use of computer analysis
of the derivatives of the cooling curve, it
became possible to use TA for the prediction/
calculation of graphite morphology, latent heat
of solidification, evolution of fraction solid,
amount of phases, dendrite coherency, and den
drite arm spacing (Ref 14 17). Other techniques
include two thermocouples in the same cup (the
Sintercast method for producing CG iron is
based on such an approach) or two or more
cups (Ref 18).
In the LDA method, the linear displacement
occurring during the solidification of the iron
is measured through quartz rods introduced
directly into the liquid metal and connected to
transducers. The concomitant TA and LDA
enables the direct correlation between expan
sion/contraction and the temperature change
during solidification events such as graphite
formation, and thus the understanding of the
kinetics of graphite expansion (Ref 19). An
example of TA and LDA curves for a gray iron
is shown in Fig. 12. The characteristic para
meters obtained through the two methods are
shown on the curves. Note the discrepancy
between the beginning and end of solidification
determined by the two methods. As explained
in the article “Principles of Thermal Analysis”
in this Volume, this is because it is incorrect
to use the minimum on the cooling rate to deter
mine the beginning and end of solidification.
Such information is of paramount importance
in evaluating graphite shape and avoiding micro
shrinkage, in particular in ductile iron and CG
iron castings. An example of the early work on
the use of combined TA and LDA to control
graphite shape in compacted graphite iron is pre
sented in Fig. 13, after Ref 20 and 21.
The following sections in this article discuss
some of the basic principles of cast iron metal
lurgy. More detailed descriptions of the metal
lurgy of cast irons are available in separate
articles in this Volume that describe the various
types of cast irons.
Gray Iron (Flake or Lamellar
Graphite Iron)
Composition and Classes of Gray Iron
The composition of gray iron must be
selected in such a way as to satisfy three basic
structural requirements:
� The required graphite shape and distribution
� A carbide free (chill free) structure
� The required matrix
The range of composition for typical unal
loyed common cast irons is given in Table 6.
The classes of gray iron according to ASTM
A48 94a are listed in Table 7. Note that as the
carbon equivalent increases, the strength and
the hardness decrease. Superfine interdendritic
graphite irons have typical strength in the range
of 300 to 350 MPa (44 to 51 ksi) with low hard
ness of 185 to 200 HB.
Increasing the carbon and silicon contents
improves the graphitization potential and there
fore decreases the chilling tendency. However,
the strengthis adversely affected (Fig. 14) because
of ferrite promotion and the coarsening of pearlite.
The manganese content varies as a function
of the desired matrix. Typically, it can be as
low as 0.1% for ferritic irons and as high as
1.2% for pearlitic irons, because manganese is
a strong pearlite promoter.
From the minor elements, phosphorus and
sulfur are the most common. They can be as
high as 0.15% for low quality iron and are con
siderably less for high quality iron, such as
ductile iron or compacted graphite iron. The
manganese/sulfur ratio is very important because
it directly affects nucleation and formation of
undesired iron sulfide (FeS) at grain boundaries.
The stoichiometric ratio is manganese/sulfur =
1.7. Typically, to obtain lamellar graphite that
solidifies mostly on MnS inclusions, the manga
nese content must exceed this ratio, as shown in
the following equation:
%Mn 1:7 %Sþ 0:15 (Eq 4)
However, if the %S is too low, and therefore
insufficient for the formation of MnS, graphite
nucleation occurs at the austenite/liquid interface,
Time, s
100 200 3000
1000
(1830)
1100
(2010)
1200
(2190)
Te
m
pe
ra
tu
re
, °
C
 (
°F
) 1300
(2370)
1400
(2550) Pouring temperature
dT/dt: Cooling rate
Superheat
ΔTmin
TL
TETER
TLA
TEU
Local solidification time
Total solidification time
ΔTmax ΔT
Fig. 11 Cooling curve and characteristic parameters.
Source: Ref 17
Table 5 Influence of minor elements
on graphite shape
Element category Element
Spheroidizing or compacting Mg, Ca, lanthanides (Ce, La,
Y, etc.)
Neutral Fe, C, alloying elements
Antispheroidizing or
anticompacting
Al, As, Bi, Te, Ti, Pb, S, Sb
18 / Introduction
 
 
 
and interdendritic graphite will result (Ref 22).
This type of graphite is typically associated with
a ferritic matrix.
More recently, Gundlach (Ref 23) argued
that the concept of excess manganese required
to tie all the sulfur is invalid, because the
reaction does not go to completion at the solid
ification temperature of cast iron. Indeed, ther
modynamic calculations indicate that there is
sufficient sulfur in solution at the eutectic tem
perature. The equilibrium constant of the reac
tion Mn + S = MnS is:
K
aMnS
aMn aS
� 1
%Mn %S
(Eq 5)
where a stands for the activity of the compound
or of the pure element. For a eutectic tempera
ture of 1160 �C (2120 �F), it was calculated that
K0 = 1/K = %Mn � %S = 0.03. The value of K0 is
temperature dependent, as shown in Fig. 15.
Below the equilibrium lines, excess sulfur will
exist, while above the lines, manganese will
be in excess. At low manganese/sulfur ratios,
MnS forms at high temperature and can float
to the surface of the casting, producing blow
hole defects.
When plotting tensile strength data obtained
from Ref 24 as a function of the %Mn � %S
product, Gundlach noted that the maximum
strength was obtained at %Mn � %S = 0.03, as
shown in Fig. 16. It was also suggested that at
%Mn � %S > 0.03, MnS are the favored inclu
sions that act as heterogeneous nuclei. Con
versely, at values smaller than 0.03, other
inclusions act as nuclei.
Other minor elements, such as aluminum,
antimony, arsenic, bismuth, lead, magnesium,
cerium, and calcium, can significantly alter
both the graphite morphology and the micro
structure of the matrix.
Both major and minor elements have a
direct influence on the morphology of lamel
lar graphite. The typical lamellar graphite
shapes are shown in Fig. 6. Type A graphite
is found in inoculated irons cooled with
moderate rates. In general, it is associated
with the best mechanical properties. Cast
irons with this type of graphite exhibit mod
erate undercooling during solidification
(Fig. 17). Type B graphite is found in irons
of near eutectic composition, solidifying on
a limited number of nuclei. Large eutectic
grain (cell) size and low undercooling are
common in cast irons exhibiting this type
of graphite. Type C graphite occurs in hyper
eutectic irons as a result of solidification
with minimum undercooling. Type D graph
ite is found in hypoeutectic or eutectic irons
solidified at rather high cooling rates, while
type E graphite is characteristic for strongly
hypoeutectic irons. Types D and E are both
associated with interdendritic distribution
and high undercooling during solidification.
Not only graphite shape but also graphite
size is important, because it is directly
related to strength, as shown in Fig. 18 from
0
1000
1050
1100
1150
1200
1250
1300
1350
2
Time, min
Heat 812.1
4 6 8
Ts
TE
TL
10
–3.0
–2.5
–2.0
–1.5
–1.0
0.5
0.0
–0.5
Te
m
pe
ra
tu
re
, °
C
 
Temperature, °C 
dT/dt, °C/s
d
T
/d
t, 
°C
/s
 
Te
m
pe
ra
tu
re
, °
C
 
D
is
pl
ac
em
en
t, 
m
m
0
1000
1050
1100
1150
1200
1250
1300
1350
2
Time, min
Heat 812.1
4 6 8
TSε
εS 
εγ
εGr
Tγshr
TGrexp
10
–0.4
–0.3
–0.2
–0.1
0.0
0.3
0.2
0.1
Temperature, °C
Displacement
Fig. 12 Cooling curves, cooling rates, and linear displacement for gray iron. Source: Ref 19
Nodular
+ compacted graphite
Compacted
+ lamellar graphite
Liquid cast iron
Compacting treatment
Fe-Si-Ca-Mg-Ca-Ti
Compacted graphite cast iron
Fe-Si-Mg
postinoculation
Fe-Si-Ti
postinoculation
FeSi 75
postinoculation
Compacted graphite
1130 °C 1130 °C 1130 °C
Fig. 13 Use of combined thermal analysis/linear displacement analysis to evaluate and correct graphite shape in
compacted graphite iron. If the nodularity is too high, ferrotitanium is added to degenerate the graphite. If
nodularity is too low a magnesium-containing ferrosilicon is used as a postinoculant to improve graphite
compactness. Source: Ref 20, 21
Classification and Basic Types of Cast Iron / 19
 
 
 
Ref 25. ASTM A247 provides a standard for
the evaluation of the size of the graphite
flakes.
Alloying elements can be added in common
cast iron to enhance some mechanical proper
ties. They influence both the graphitization
potential and the structure and properties of
the matrix. In general, alloying elements can
be classified into three categories, discussed in
the following paragraphs.
Silicon and aluminum increase the graphiti
zation potential for both the eutectic and eutec
toid transformations and increase the number of
graphite particles. They form solid solutions in
the matrix. Because they increase the ferrite/
pearlite ratio, they lower strength and hardness.
Nickel, copper, and tin increase the graphi
tization potential during the eutectic transfor
mation but decrease it during the eutectoid
transformation, thus raising the pearlite/ferrite
ratio. These elements form solid solution in
the matrix. Because they increase the amount
of pearlite, they raise strength and hardness.
Chromium, molybdenum, tungsten, and
vanadium decrease the graphitization potential
at both stages of transformation. Consequently,
they increase the amount of carbides and pearl
ite. They concentrate principally in the car
bides, forming (FeX)nC type carbides, but also
alloy the ferrite (aFe) solid solution. As long
as carbide formation does not occur, these ele
ments increase strength and hardness. Above a
certain level, any of these elements will deter
mine the solidification of a structure with both
Gr and Fe3C (mottled structure), which will
have lower strength but higher hardness.
In alloyed gray iron, the typical ranges for the
elements discussed previously are as follows:
Element Composition, mass%
Chromium 0.2–0.6
Molybdenum 0.2–1
Vanadium 0.1–0.2
Nickel 0.6–1
Copper 0.5–1.5
Tin 0.04–0.08
The influence of composition and cooling rate
on tensile strength (TS) can be estimated using
the following equation (Ref 25):
TS 162:37þ 16:61=D 21:78 %C 61:29 %Si
10:59ð%Mn 1:7 %SÞ þ 13:8 %Cr
þ 2:05 %Niþ 30:66 %Cuþ 39:75 %Mo
þ 14:16 %Sið Þ2 26:25 %Cuð Þ2
23:83 %Moð Þ2 (Eq 6)
where D is the bar diameter (in inches). This
equation is valid for bar diameters of 20 to
Table 6 Range of compositions for typical unalloyed common cast irons
Type of iron(a)
Composition, mass%
C Si Mn P S
White 1.8–3.6 0.5–1.9 0.25–0.8 0.06–0.20.06–0.2
Gray (LG) 2.5–4.2 1.0–3.0 0.15–1.0 0.02–1.0 0.02–0.25
Compacted graphite (CG) 2.5–4.0 1.5–3.0 0.2–1.0 0.01–0.1 0.01–0.03
Ductile (SG) 3.0–4.0 1.8–4.5 0.1–1.0 0.01–0.1 0.01–0.03
Malleable (TG) 2.2–2.9 0.9–1.9 0.15–1.2 0.02–0.2 0.02–0.2
(a) LG, lamellar graphite; CG, compacted graphite; SG, spheroidal graphite; TG, temper graphite
Table 7 Compositions (mass%) and mechanical properties of various classes of gray irons
according to ASTM A48 94a
Class Carbon, % Silicon, % Carbon equivalent
Tensile strength
Hardness, HBMPa ksi
20 3.40–3.60 2.30–2.50 4.30 152 22 156
25 . . . . . . . . . 179 26 174
30 3.10–3.30 2.10–2.30 3.88 214 31 210
35 . . . . . . . . . 252 36.5 212
40 2.95–3.15 1.70–2.00 3.67 293 42.5 235
50 2.70–3.00 1.70–2.00 3.47 362 52.5 262
60 2.50–2.85 1.90–2.10 3.34 431 62.5 302
Fig. 14 General influence of carbon equivalent on
the tensile strength of gray iron. Source: Ref 8
0.2
0.05
0.1S
ul
fu
r, 
%
0.15
0.2
0.4 0.6 0.8 1.0 1.2 1.4
Manganese, %
1200 °C
1280 °C
1350 °C
0.15% S – 0.8% Mn
0.05% S – 0.8% Mn
Fig. 15 Equilibrium %Mn � %S = 0.03 lines for
various temperatures. Source: Ref 23
MnS nucleates graphite
0.00
20
25
30
40
35
45
50
55
U
lti
m
at
e 
te
ns
ile
 s
tr
en
gt
h,
 k
si
U
lti
m
at
e 
te
ns
ile
 s
tr
en
gt
h,
 M
P
a
0.02 0.04
(%Mn)·(%S)
0.06 0.08 0.10
138
172
207
241
276
310
345
379
Other inclusions
nucleate graphite
Fig. 16 Correlation between tensile strength and the
%Mn � %S product for data from Ref 24.
Source: Ref 23
Fig. 17 Characteristic cooling curves associated with
different flake graphite shapes. TE, equilibrium
eutectic temperature
Fig. 18 Effect of maximum graphite flake length on the
tensile strength of gray iron. Source: Ref 25
20 / Introduction
 
 
 
50 mm (0.8 to 2 in.) and compositions within
the following ranges: 3.04 to 3.29% C, 0.1 to
0.55% Cr, 0.03 to 0.78% Mo, 1.6 to 2.46%
Si, 0.07 to 1.62% Ni, 0.089 to 0.106% S,
0.39 to 0.98% Mn, and 0.07 to 0.85% Cu.
Cooling Rate
The cooling rate, like the chemical composi
tion, can significantly influence the as cast
structure and therefore the mechanical proper
ties. The cooling rate of a casting is primarily
a function of its section size. The dependence
of structure and properties on section size is
termed section sensitivity. Increasing the cool
ing rate will:
� Refine both graphite size and matrix struc
ture; this will result in increased strength
and hardness
� Increase the chilling tendency; this may
result in higher hardness but will decrease
the strength
Consequently, composition must be tailored in
such a way as to provide the correct graphitiza
tion potential for a given cooling rate. For a
given chemical composition and as the section
thickness increases, the graphite becomes
coarser, and the pearlite/ferrite ratio decreases,
which results in lower strength and hardness
(Fig. 19). Higher carbon equivalent has similar
effects.
Liquid Treatment
In gray iron practice, the liquid treatment
used is termed inoculation and consists of min
ute additions of minor elements before pouring.
Typically, ferrosilicon containing aluminum
and calcium, or proprietary alloys are used as
inoculants. The main effects of inoculation in
gray iron are:
� Increased graphitization potential because of
decreased undercooling during solidifica
tion; as a result, the chilling tendency is
diminished, and graphite shape changes
from type D or E to type A
� Finer structure, that is, higher number of
eutectic cells (grains), with a subsequent
increase in strength
As shown in Fig. 20, after Ref 26, inocula
tion improves tensile strength. This influence
is more pronounced for low CE cast irons.
Heat Treatment
Heat treatment can considerably alter the
matrix structure, although graphite shape and
size remain basically unaffected. A rather low
proportion of the total gray iron produced is
heat treated. Common heat treatment may con
sist of stress relieving or annealing to decrease
hardness. More information is available in the
article “Heat Treatment of Gray Irons” in this
Volume.
Ductile Iron (Spheroidal Graphite
Iron)
Composition and Classes
of Ductile Iron
The main effects of the chemical composi
tion are similar to those described for gray iron,
with quantitative differences in the extent of
these effects. The carbon equivalent has only
a mild influence on the properties and structure
of ductile iron, because it affects graphite shape
considerably less than in the case of gray iron.
Nevertheless, to prevent excessive shrinkage,
high chilling tendency, graphite flotation, or a
high impact transition temperature, optimum
amounts of carbon and silicon must be selected,
as suggested in Fig. 21.
As mentioned previously, minor elements
can significantly alter the structure in terms of
graphite morphology, chilling tendency, and
matrix structure. Minor elements can promote
the spheroidization of graphite or can have an
adverse effect on graphite shape. The minor
elements that adversely affect graphite shape
are said to degenerate graphite shape. The
generic influence of various elements on graph
ite shape is given in Table 5. The effect of mag
nesium, the most widely used element for the
production of spheroidal graphite iron, on
graphite shape is illustrated in Fig. 22, cited in
Ref 27. The amount of residual magnesium in
the iron, Mgresid, required to produce spheroidal
graphite is generally 0.03 to 0.05%. The precise
level depends on the cooling rate. A higher
cooling rate requires less magnesium. The
amount of magnesium to be added in the iron
is a function of the initial sulfur level, Sin, and
the recovery of magnesium, Z, in the particular
process used:
Mgadded
0:75 Sin þMgresid
Z
(Eq 7)
A residual magnesium level that is too low
results in insufficient nodularity (that is, a low
ratio between the spheroidal graphite and the
Fig. 19 Influence of section thickness of the casting
on (a) tensile strength and (b) hardness for a
series of gray irons classified by their strength as-cast in
30 mm (1.2 in.) diameter bars. Source: Ref 8
Fig. 20 Influence of inoculation on tensile strength as
a function of carbon equivalent for 30 mm
(1.2 in.) diameter bars. Source: Ref 26
Fig. 21 Typical range for total carbon (TC) and silicon
contents in good-quality ductile iron. Source:
Ref 8
Classification and Basic Types of Cast Iron / 21
 
 
 
total amount of graphite in the structure). This
in turn results in a deterioration of the mechan
ical properties of the iron, as illustrated in
Fig. 23. If the magnesium content is too high,
carbides are promoted.
The presence of antispheroidizing (deleteri
ous) minor elements may result in graphite
shape deterioration, up to complete graphite
degeneration. Therefore, upper limits are set
on the amount of deleterious elements to be
accepted in the composition of cast iron. Typi
cal upper limits are as follows, after Ref 28:
Element Upper limit, %
Aluminum 0.1
Arsenic 0.02
Bismuth 0.002
Cadmium 0.01
Lead 0.002
Antimony 0.002
Selenium 0.03
Tellurium 0.02
Titanium 0.1
Zirconium 0.1
These values can be influenced by the combina
tion of various elements and by the presence of
lanthanides (rare earths) in the composition.
Furthermore, some of these elements can be
deliberately added during liquid processing to
increase nodule count.
In principle, alloying elements have the same
influence on structure and properties as for gray
iron. Because better graphite morphology
allows more efficient use of the mechanical
properties of the matrix, alloying is more com
mon in ductile iron than in gray iron.
ASTM specification A536 lists five classes of
ductile iron based on their minimum tensile prop
erties (Table 8). Some of the grades can be pro
duced as cast while others through heat treatment.
Cooling Rate
When changing the cooling rate, effects similar
to those discussed for gray iron also occur in duc
tile iron (DI), but the section sensitivity of ductile
iron is significantly lower. Indeed, as shown in
Fig. 24, propertiesBacon was thought to be the last per
son who knew everything a person could know, at least in a European
nation, in 1600. Until 1900, human knowledge doubled every century.
By 1945, knowledge was doubling every 25 years. Today, different types
of knowledge have different rates of growth: Nanotechnology knowledge
doubles every 2 years, but clinical knowledge doubles every 18 months.
On average, human knowledge doubles every 13 months. According to
IBM, the building of the “internet of things” will lead to the doubling
of knowledge every 12 hours.
But what is knowledge? According to Aristotle (384 322 B.C.E.),
considered to be the father of science and the scientific method and
the inventor of the language of science, knowledge includes theoretical
(episteme knowing and understanding), practical (praxis doing), and
technical (techne making, production). It is one of the ambitions of
this Volume to include aspects of all these types of knowledge on cast
iron. And, as Plato (428 347 B.C.E.), considered to be the founder of
Western spirituality, stated, “where there is number there is order;
where there is no number there is nothing but disorder,” this Volume
also stresses the mathematical, quantitative aspects of the science of
cast iron. This is a logical objective, as many of the processes used in
iron casting are still empirical in nature, but many others are deeply
rooted in mathematics. The knowledge ladder includes generation of
knowledge, transfer of knowledge, and implementation of knowledge.
Thus, knowledge is not merely the possession of information but rather
its implementation and use,which brings us to the main goal of this
Volume to package and transfer knowledge in a form that can facili
tate its implementation in praxis. Because this is a monumental, almost
impossible task, its completion necessitated the involvement of the top
iron casting engineers and scientists in the international community.
Their collective effort was successful in assembling what I believe to
be the most complete text on cast iron available in the English language
today.
This Volume is structured in eleven sections, starting with an intro
duction that covers the history of cast iron and a detailed classification
and discussion of the basic types of cast iron. The following section is a
rather academic treatment of the fundamentals of the metallurgy of cast
irons, including thermodynamics principles specific to cast iron, micro
structure evolution and volumetric changes during solidification and
solid state transformation, and prediction of solidification microstruc
ture through computational modeling, which was dubbed earlier in this
preface as “virtual cast iron.” Next, an extensive discussion of the many
facets of the science and engineering of processing of cast ironis
provided, with particular emphasis on liquid metal preparation, casting
processes, and heat treatment. The section on secondary processing
addresses issues such as machining, inspection, and quality control.
The properties of various types of iron and the effects of processing
are treated in a section that concludes with another “virtual cast iron”
subject, computer aided prediction of mechanical properties. The speci
fications, selection criteria, microstructure, and production particulari
ties of the main classes of cast iron gray iron, ductile iron,
compacted graphite iron, high alloy iron, and malleable iron are then
Professor Doru Michael Stefanescu
Volume Editor
iv
discussed in great detail in separate sections. Attention is given to more
recent developments, such as thin wall iron and heavy section ductile
iron castings. Most articles include a large number of references that
serve a dual purpose: to give credit where credit is due, and to direct
the reader to additional information on the subject, if the reader is
interested.
This Volume is the product of the combined efforts of an interna
tional team of top scientists and metal casting specialists from no less
than 12 countries (Argentina, Brazil, China, Denmark, France, Nor
way, Poland, Romania, Spain, Sweden, the United Kingdom, and the
United States of America) and of the outstanding diligence of the
ASM International technical and support personnel, to whom the Edi
tor is deeply grateful. The Editor would like to extend his personal
appreciation to the leaders of the ASM International team, Mr. Steve
Lampman, Senior Content Developer, and Ms. Vicki Burt, Content
Developer, for their remarkable efforts in coordinating this gargantuan
task and their personal contributions to the text. It required many,
many days. We the authors, the ASM International team, and the
Editor do hope that the readers will find in this Volume answers to
most of the questions that they may have on cast iron for many years
to come.
v
Policy on Units of Measure
By a resolution of its Board of Trustees, ASM International has
adopted the practice of publishing data in both metric and customary
U.S. units of measure. In preparing this Handbook, the editors have
attempted to present data in metric units based primarily on Système
International d’Unités (SI), with secondary mention of the corresponding
values in customary U.S. units. The decision to use SI as the primary sys
tem of units was based on the aforementioned resolution of the Board of
Trustees and the widespread use of metric units throughout the world.
For the most part, numerical engineering data in the text and in tables
are presented in SI based units with the customary U.S. equivalents in
parentheses (text) or adjoining columns (tables). For example, pressure,
stress, and strength are shown both in SI units, which are pascals (Pa)
with a suitable prefix, and in customary U.S. units, which are pounds
per square inch (psi). To save space, large values of psi have been con
verted to kips per square inch (ksi), where 1 ksi = 1000 psi. The metric
tonne (kg � 103) has sometimes been shown in megagrams (Mg). Some
strictly scientific data are presented in SI units only.
To clarify some illustrations, only one set of units is presented on art
work. References in the accompanying text to data in the illustrations are
presented in both SI based and customary U.S. units. On graphs and
charts, grids corresponding to SI based units usually appear along the left
and bottom edges. Where appropriate, corresponding customary U.S.
units appear along the top and right edges.
Data pertaining to a specification published by a specification writing
group may be given in only the units used in that specification or in dual
units, depending on the nature of the data. For example, the typical yield
strength of steel sheet made to a specification written in customary U.S.
units would be presented in dual units, but the sheet thickness specified
in that specification might be presented only in inches.
Data obtained according to standardized test methods for which the
standard recommends a particular system of units are presented in the
units of that system. Wherever feasible, equivalent units are also pre
sented. Some statistical data may also be presented in only the original
units used in the analysis.
Conversions and rounding have been done in accordance with IEEE/
ASTM SI 10, with attention given to the number of significant digits in
the original data. For example, an annealing temperature of 1570 �F con
tains three significant digits. In this case, the equivalent temperature
would be given as 855 �C; the exact conversion to 854.44 �C would
not be appropriate. For an invariant physical phenomenon that occurs at
a precise temperature (such as the melting of pure silver), it would be
appropriate to report the temperature as 961.93 �C or 1763.5 �F. In some
instances (especially in tables and data compilations), temperature values
in �C and �F are alternatives rather than conversions.
The policy of units of measure in this Handbook contains several
exceptions to strict conformance to IEEE/ASTM SI 10; in each instance,
the exception has been made inof thin walled DI castings
(2.5 to 4 mm, or 0.10 to 0.16 in., thickness) fall
in the range of the general properties of DI. This
is because spheroidal graphite is less affected by
cooling rate than flake graphite.
Liquid Treatment
The liquid treatment of ductile iron is more
complex than that of gray iron, because it typi
cally requires two stages:
� Modification, which consists of magnesium
or magnesium alloy treatment of the melt,
with the purpose of changing graphite shape
from lamellar to spheroidal
� Inoculation (normally after the magnesium
treatment postinoculation) to increase the
nodule count. Increasing the nodule count
is an important goal, because a higher nod
ule count is associated with less chilling ten
dency (Fig. 25) and a higher as cast ferrite/
pearlite ratio.
Heat Treatment
Heat treatment is extensively used in theproces
sing of ductile iron, because better advantage can
be obtained from thematrix structure than for gray
iron. This is discussed further in the article “Heat
Treatment of Ductile Iron” in this Volume. The
heat treatments usually applied are as follows:
� Stress relieving
� Annealing to produce a ferritic matrix
� Normalizing to produce a pearlitic matrix
� Hardening to produce tempered structures
(bainite, martensite)
� Austempering to produce a ferritic bainite
A typical temperature time diagram for the
austempering process is presented in Fig. 26.
Higher austempering temperatures produce
coarser structures associated with good ductility
and dynamic properties. Lower austempering
temperatures generate finer structures that have
higher strength and wear resistance. The
difference between ausferrite and bainite is fur
ther discussed in “The Austenite to Ausferrite
Transformation” article in this Volume.
The advantage of austempering is that it results
in ductile irons with twice the tensile strength for
the same toughness. A comparison between some
mechanical properties of heat treated ductile
irons is shown in Fig. 27. ASTM A897 lists five
classes of austempered ductile iron based on their
minimum tensile properties (Table 9).
Compacted (Vermicular) Graphite
Irons
Compacted graphite (CG) irons have a
graphite shape intermediate between spheroidal
and lamellar. Typically, compacted graphite
looks like type IV graphite (Fig. 5). Conse
quently, most of the properties of CG irons lie
in between those of gray and ductile iron.
Composition and Classes of Compacted
Graphite Iron
The chemical composition effects are similar
to those described for ductile iron. Carbon equiv
alent influences strength less obviously than for
the case of gray iron but more than for ductile
iron, as shown in Fig. 28. ASM specification
A842 lists five grades of CG irons based on the
minimum tensile strength (Table 10). The graph
ite shape is controlled, as in the case of ductile
iron, through the content of minor elements.
Fig. 23 Influence of (a) residual magnesium and
(b) nodularity on some mechanical properties
of ductile iron. Source: Ref 29, 30
Table 8 Ductile iron grades in ASTM A536
Grade
Minimum tensile strength Minimum yield strength
Minimum elongation, % Brinell hardness Matrix microstructureMPa ksi MPa ksi
60-40-18 414 60 276 40 18 149–187 Ferrite
65-45-12 448 65 310 45 12 170–207 Ferrite/pearlite
80-55-06 552 80 379 55 6 187–255 Pearlite/ferrite
100-70-03 689 100 483 70 3 217–269 Pearlite
120-90-02 828 120 621 90 2 240–300 Tempered martensite
Fig. 22 Influence of residual magnesium on graphite
shape. Source: Ref 27
22 / Introduction
 
 
 
Cooling Rate
The cooling rate affects properties less than
for gray iron but more than it does for ductile
iron (Fig. 29). In other words, CG iron is less
section sensitive than gray iron. However, high
cooling rates are to be avoided because of the
high propensity of CG iron for chilling and high
nodularity in thin sections.
Liquid Treatment
There are four common methods to produce
CG iron:
� Controlled undertreatment with magnesium
containing alloys
� Treatment with alloys containing both com
pacting (magnesium, cerium, lanthanum,
calcium) and anticompacting (titanium, alu
minum) elements
� Treatment with lanthanides base alloy or
magnesium lanthanides alloys
965
896
827
758
689
620
552
483
414
345
276
207
138
69
0
120 160 200
HB
240 280
10
20
30
40
50
60
70
80
90
100
110
S
tr
es
s,
 k
si
S
tr
es
s,
 M
P
a
e f
, %
ef
YS0.2
UTS
120
130
140
ef
YS0.2
UTS
Fig. 24 Static mechanical properties of ductile iron. Solid lines delimit the typical range of properties for ductile
iron. Source: Ref 8. Dashed lines are summary data (regression). Source: Ref 31. Symbols are data from
Ref 32 on thin-walled ductile iron castings. UTS, ultimate tensile strength; YS, yield strength
Fig. 25 Influence of the amount of ferrosilicon (75%
Si) added as a postinoculant on the nodule
count and chill depth of 3 mm (0.12 in.) plates. Source:
Ref 33
9801800
1600
1400
1200
1000
800
600
400
200
0
0
0
0.
5
1.
0
1.
5
2.
0
2.
5/
0 30 60 90 12
0
15
0
18
0
21
0
24
0
24
00
24
,0
00
24
0,
00
0
870
760
650
540
430
315
205
95
A
B
D
E
F
C
Ms
Mf
Pearlite
Bainite
Austenitizing, h
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
, °
C
Austempering, min
Ausferrite
Austenite
Q
uench
Fig. 26 Austempering process for cast iron. Source: Ref 34
Fig. 27 Properties of some standard and austempered
ductile irons. Source: Ref 35
Classification and Basic Types of Cast Iron / 23
 
 
 
� Treatment of a base iron containing high
amounts of anticompacting elements (sulfur,
aluminum) with alloys containing compact
ing elements (magnesium, cerium)
From the standpoint of controlling the structure,
it is easier to combine compacting and anticom
pacting elements. However, most compacted
graphite iron today (2016) is produced through
undertreatment of the melt with magnesium
(~0.02% Mg), and process control through ther
mal analysis.
Figure 30 shows how increased amounts of
magnesium will change graphite shape from
lamellar to compacted and then to spheroidal.
It is also seen that the range for CG is very nar
row, 0.016 to 0.019% Mg.
Liquid treatment may include two stages, as
for ductile iron. However, postinoculation must
be maintained at a low level to avoid excessive
nodularity.
Heat Treatment
Heat treatment is not common for CG irons.
Malleable Irons
Malleable cast irons differ from the types of
irons previously discussed in that they have an
initial as cast white structure, that is, a structure
consisting of iron carbides in a pearlitic matrix.
This white structure is then heat treated (anneal
ing at 800 to 970 �C, or 1470 to 1780 �F), which
results in the decomposition of Fe3C and the for
mation of austenite (g) and temper graphite. The
basic solid state reaction is:
Fe3C ! gþ Gr (Eq 8)
Most of the malleable iron is produced by this
technique and is called blackheart malleable
iron. The final microstructure consists of graph
ite in a matrix of pearlite, pearlite and ferrite,
or ferrite. The structure of the matrix is a func
tion of the cooling rate after annealing. Some
malleable iron is produced in Europe by decar
burization of the white as cast iron, and it is
called whiteheart malleable iron.
Composition and Classes of
Malleable Iron
The composition of malleable irons must be
selected in such a way as to produce a white
as cast structure and to allow for fast anneal
ing times. Some typical compositions are
given in Table 6. Although higher carbon and
silicon reduce the heat treatment time, they
must be limited to ensure a graphite free struc
ture upon solidification. Both tensile strength
and elongation decrease with higher carbon
equivalent. Nevertheless, it is not enough to
control the carbon equivalent. The annealing
time depends on the number of graphite nuclei
available for graphitization, which in turn
depends on the carbon/silicon ratio, among
other factors. As shown in Fig. 31, a lower
carbon/silicon ratio (that is, a higher silicon
contentfor a constant carbon equivalent)
results in a higher temper graphite count
(Ref 37). This in turn translates into shorter
annealing times.
The manganese content and the manganese/
sulfur ratio must be closely controlled. In gen
eral, lower manganese content is used when fer
ritic rather than pearlitic structures are desired.
The correct manganese/sulfur ratio can be
calculated with Eq 4, which is plotted in
Fig. 32. Under the line described by Eq 4, all
sulfur is stoichiometrically tied to manganese
as MnS. The excess manganese is dissolved in
the ferrite. In the range delimited by the lines
given by Eq 4 and the line Mn/S = 1, a mixed
sulfide, (Mn,Fe)S, is formed. For manganese/
sulfur ratios smaller than 1, pure FeS is also
formed. It is assumed that the degree of com
pacting of temper graphite depends on the type
of sulfides occurring in the iron (Ref 38). When
FeS is predominant, very compacted, nodular
temper graphite forms, but some undissolved
Fe3C may persist in the structure, resulting in
lower elongations. When MnS is predominant,
although the graphite is less compacted, elonga
tion is higher because of the completely Fe3C
free structure. The manganese/sulfur ratio also
influences the number of temper graphite parti
cles. From this standpoint, the optimum manga
nese/sulfur ratio is approximately 2 to 4
(Fig. 33 from Ref 39). Alloying elements can
be used in some grades of pearlitic malleable
irons. The manganese content can be increased
to 1.2%, or copper, nickel, and/or molybdenum
can be added. Chromium must be avoided
because it produces stable carbides, which are
difficult to decompose during annealing.
ASTM specification A220 lists eight classes
of malleable iron based on their yield strength
and elongation (Table 11).
Cooling Rate
Like all other irons, malleable irons are sen
sitive to cooling rate. Nevertheless, because
the final structure is the result of a solid state
reaction, they are the least section sensitive
irons. Typical correlations between tensile
strength, elongation, and section thickness are
shown in Fig. 34.
Liquid Treatment
The liquid treatment of malleable iron
increases the number of nuclei available for
Fig. 28 Effect of carbon equivalent on the tensile
strength of flake, compacted, and spheroidal
graphite irons cast in 30 mm (1.2 in.) diameter bars.
Source: Ref 36
Table 9 Austempered ductile iron grades in ASTM A897
Grade
Minimum tensile strength Minimum yield strength
Minimum elongation, % Brinell hardness
Impact energy
MPa ksi MPa ksi J ft�lbf
1 850 125 550 80 10 269–321 100 75
2 1050 150 700 100 7 302–363 80 60
3 1200 175 850 125 4 341–444 60 45
4 1400 200 1100 155 1 388–477 35 25
5 1600 230 1300 185 . . . 444–555 . . . . . .
Table 10 Compacted graphite iron grades
in ASTM A842
Grade
Minimum
tensile
strength
Minimum
yield
strength
Minimum
elongation, %
Brinell
hardnessMPa ksi MPa ksi
250 250 36 175 25 3.0 179 max
300 300 44 210 30 1.5 143–207
350 350 51 245 36 1.0 163–229
400 400 58 280 41 1.0 197–255
450 450 65 315 46 1.0 207–269
Fig. 29 Influence of section thickness on the tensile
strength of compacted graphite irons
24 / Introduction
 
 
 
the solid state graphitization reaction. This can
be achieved in two different ways:
� By adding elements that increase undercool
ing during solidification. Typical elements in
this category are magnesium, cerium, bis
muth, and tellurium. Higher undercooling
results in finer structure, which in turn means
more austenite Fe3C interface. Because
graphite nucleates at the austenite Fe3C inter
face, this means more nucleation sites for
graphite. Higher undercooling during solidifi
cation also prevents the formation of
unwanted eutectic graphite.
� By adding nitride forming elements to the
melt. Typical elements in this category are
aluminum, boron, titanium, and zirconium.
Heat Treatment
The heat treatment of malleable iron deter
mines the final structure of this iron. It has
two basic stages. In the first stage, the iron car
bide is decomposed in austenite and graphite
(Eq 8). In the second stage, the austenite is
transformed into pearlite, ferrite, or a mixture
of the two. Although there are some composi
tional differences between ferritic and pearlitic
irons, the main difference is in the heat treat
ment cycle. When ferritic structures are to be
produced, cooling rates in the range of 3 to
10 �C/h (5 to 18 �F/h) are required through
the eutectoid transformation in the second
stage. This is necessary to allow for a complete
austenite to ferrite reaction. A typical annealing
cycle for ferritic malleable iron is shown in
Fig. 35. When pearlitic irons are to be pro
duced, different schemes can be used, as shown
in Fig. 36. The goal of the treatment is to
achieve a eutectoid transformation according
to the austenite to pearlite reaction. In some
Type 1 – MgL.R. De and Y.J. Xiang, Trans. AFS, Vol
99, 1991, p 707 712
10. T. Skaland, F. Grong, and T. Grong, Metall.
Trans. A, Vol 24, 1993, p 2321, 2347
11. M. Chisamera, I. Riposan, and M. Barstow,
Paper 3, AFS International Inoculation
Conference (Rosemont, IL), 1998
12. E. Moumeni, D.M. Stefanescu, N.S. Tiedje,
P. Larrañaga, and J.H. Hattel,Metall. Mater.
Trans. A, Vol 44 (No. 11), 2013, p 5134 5146
13. J.G. Humphreys, BCIRA J., Vol 9, 1961,
p 609 621
14. D. Rabus and S. Polten, Giesserei Rund
shau, No. 9, 1972, p 1 8
15. P. Strizik, Giesserei, Vol 61, 1974,
p 615 618
16. L. Bäckerud, K. Nilsson, and H. Steen,
in The Metallurgy of Cast Iron, B. Lux,
I. Minkoff, and F. Mollard, Ed., Georgi
Publishing, Switzerland, 1975, p 625
17. I.G. Chen and D.M. Stefanescu, Trans.
AFS, Vol 92, 1984, p 947
Table 11 Malleable iron grades according
to ASTM A220
Grade
Minimum
tensile
strength
Minimum
yield strength
Minimum elongation, %MPa ksi MPa ksi
40010 414 60 276 40 10
45008 448 65 310 45 8
45008 448 65 310 45 6
50005 483 70 345 50 6
60004 552 80 414 60 4
70003 586 85 483 70 3
80002 655 95 552 80 2
90001 724 105 621 90 1
Fig. 35 Heat treatment cycle for ferritic blackheart
malleable iron. Source: Ref 6
(a) (b)
Fig. 34 Influence of bar diameter on the (a) tensile strength and (b) elongation of blackheart malleable iron. Source:
Ref 40
Fig. 36 Heat treatment cycles for pearlitic blackheart
malleable irons
26 / Introduction
 
 
 
18. T. Kanno, I. Kang, Y. Fukuda, M. Morinaka,
andH.Nakae,Paper06 083,AFSTrans.,2006
19. D.M. Stefanescu, M. Moran, S. Boonmee,
and W.L. Guesser, Trans. AFS, Vol 120,
2012, p 365 374
20. D.M. Stefanescu, L. Dinescu, S. Craciun,
and M. Popescu, Paper 37, Proceedings of
the 46th International Foundry Congress
(Madrid, Spain), CIATF, 1979
21. D.M. Stefanescu and C.R. Loper, Gies
serei Prax., Vol 5, 1981, p 73 96
22. D.M. Stefanescu, G. Alonso, P. Larrañaga,
and R. Suarez, Acta Mater., Vol 103,
2016, p 103 114
23. R. Gundlach, Influence of Mn and S on
Mechanical Properties of Gray Cast Iron,
Part I: Historical Perspective, Paper 14
079, AFS Proc., 2014
24. K.M. Muzumdar and J.F. Wallace, Effect of
Sulfur in Cast Iron, AFS Trans., 1973, p 412
25. C.E. Bates, AFS Trans., Vol 94, 1986, p 889
26. T.E. Barlow and C.H. Lorig, Trans. AFS,
Vol 54, 1946, p 545
27. E. Nechtelberger, H. Puhr, J.B. von Nessel
rode, and A. Nakayasu, Cast Iron with Ver
micular/Compacted Graphite State of the
Art Development, Production, Properties,
Applications, Proceedings of the 47th
International Foundry Congress, April
1982 (Chicago, IL), CIATF
28. H.Morrogh, Trans. AFS, Vol 60, 1952, p 439
29. R. Barton, BCIRA J., No. 5, 1961, p 668
30. R.W. Lindsay and A. Shames, Trans. AFS,
Vol 60, 1952, p 650
31. L.J. Basaj, T.A. Dorn, M.D. Rothwell, B.D.
Johnson, and R.W. Heine, Trans. AFS, Vol
107, 1999
32. L.P. Dix, R. Ruxanda, J. Torrance, M.
Fukumoto, and D.M. Stefanescu, Trans.
AFS, Vol 111, 2003, p 1149 1164
33. D.M. Stefanescu, AFS Int. Cast Met. J.,
June 1981, p 23
34. G.M. Goodrich, Tech. Ed., Iron Castings
Engineering Handbook, American Foundry
Society, 2003
35. J.F. Janowak and R.B. Gundlach, Trans.
AFS, Vol 91, 1983, p 377
36. G.F. Sergeant and E.R. Evans, Br. Foun
dryman, May 1978, p 115
37. D.M. Stefanescu, Metalurgia (Romania),
No. 7, 1967, p 368
38. K. Roesch, Stahl Eisen, No. 24, 1957, p 1747
39. R.P. Todorov, in Proceedings of the 32nd
International Foundry Congress (Warsaw,
Poland), International Committee of
Foundry Technical Associations
40. K.M. Ankab, O.E. Shulte, and P.N. Bidu
lia, Isvestia Vishih Utchebnik Zavedenia
Tchornaia Metallurghia, No. 5, 1966,
p 168 (in Russian)
Classification and Basic Types of Cast Iron / 27
 
 
 
 
 
 
Thermodynamics Principles as
Applied to Cast Iron
Doru M. Stefanescu, The Ohio State University and The University of Alabama
Jacques Lacaze, Université de Toulouse
THE FINAL MICROSTRUCTURE of cast
parts is the result of phase transformations occur
ring during cooling from liquid state to room tem
perature. These transformations are the liquid
solid transformation (solidification), which occurs
when the liquid cools under the liquidus or the
eutectic temperature, and the solid solid transfor
mation, which occurs when austenite cools under
the eutectoid temperature. In many instances, in
order to improve certain properties, the iron cast
ings may be submitted to further solid solid trans
formations as they undergo heat treatment.
The control of the solidification process of cast
iron requires understanding and control of the
thermodynamics of the liquid and solid phases
and of the kinetics of their solidification, including
nucleation and growth.While the information that
thermodynamics can provide covers a rather
wide range including processing of minerals,
controlling metal slag interactions and behavior
of linings, gas metal interactions, endogenous
precipitation, solidification and solid state trans
formations, the issues discussed in this article
include: (a) the influence of temperature and com
position on solubility of various elements in iron
base alloys; (b) calculation of solubility lines, rel
evant to the construction of phase diagrams;
(c) calculation of activity of various components,
which allows for determination of probability of
formation and relative stability of various phases.
The role of alloying elements then is discussed in
terms of their influence on the activity of carbon,
which provides information on the stability of
themain carbon rich phases of iron carbon alloys,
that is, graphite and cementite.
Thermodynamics of Binary
Fe-X Systems
The Fe C, Fe Si, and Fe S systems, and the sol
ubility of gases in iron are discussed in this section.
The Fe-C System
There are two crystalline modifications of
iron: body centered cubic (bcc) for a iron
and d iron (ferrite), and face centered cubic
(fcc) for g iron (austenite) with the phase
transformations occurring at the following
temperatures:
aFe !910 �C gFe !1392 �C dFe !1537 �C liquid Fe
The change of the lattice constant of the two
crystalline forms of iron are presented in
Fig. 1 (Ref 1). The interatomic distance as a
function of the lattice parameter a is as follows:
for bcc, d ¼ a 3
p
=2; for fcc, d ¼ a= 2
p
. This
gives the following values at the a ! g trans
formation temperature: aa = 2.898 � 10�10 m,
da = 2.51 � 10�10 m, ag = 3.639 � 10�10 m, and
dg = 2.573 � 10�10 m.
The elements forming interstitial solutions
(e.g., hydrogen, boron, carbon, nitrogen, oxy
gen) have larger solubility in austenite than in
ferrite because ag > aa. The elements with fcc
structure (e.g., nickel, cobalt) dissolved in iron
as solid solution extend the temperature stability
range of austenite (g promoters), while the bcc
elements (e.g., silicon, chromium, vanadium)
promote the ferrite phase.
The equilibrium phase diagram of the binary
Fe C system includes the stable (Fe graphite)
and metastable (Fe Fe3C) equilibria. It is pre
sented in Fig. 2 after Okamoto (Ref 2). The tem
perature and composition of the characteristic
points are presented in Table 1 after Okamoto,
Neumann (Ref 3), and Gustafsson (Ref 4). Note
the ASM compilation from Ref 2 is slightly dif
ferent from the known thermodynamic assess
ment by Gustafsson.
For the stable system, that is, with graphite as
the equilibrium high carbon phase, the follow
ing equations are used to describe the maxi
mum solubility of carbon (mass%) in various
phases as a function of temperature (in �C
unless otherwise specified):
� For the liquid in the interval eutectic temper
ature, 1600 �C:
%CL=Gr
max 2:11þ 1:213 10�6 T þ 5:197 10�7 T2
(Eq 1a)
%CL=Gr
max 1:3þ 2:57 10�3 T (Eq 1b, Ref 3)
where T is the temperature in �C. The % nota
tion refers to mass% here and later in this text,
unless otherwise specified. Using thermody
namic quantities, Eq 1(b) can be written as:
logN
L=Gr
Cmax
12:728=Tð�KÞ þ 0:727 logTð�KÞ
3:049 (Eq 1c, Ref 3)
where N
L=Gr
Cmax is the maximum solubility of car
bon in molefraction.
� For the austenite in equilibrium with graph
ite between the eutectoid and the eutectic
temperatures:
%Cg=Gr
max 1:948þ 3:51 10�3 T (Eq 2a)
%Cg=Gr
max 0:16 1:898 10�4 T þ 1:8 10�6 T2
(Eq 2b)
ASM Handbook, Volume 1A, Cast Iron Science and Technology 
D.M. Stefanescu, editor
DOI: 10.31399/asm.hb.v01a.a0006295
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
Face-centered
cube
2.8
3.2
La
tti
ce
 c
on
st
an
t a
, Å
0
3.6
4.0
400 800
Temperature, °C
1200 1600
32 750 1470 2190 2910
γFe
Body-centered
cube
αFe δFe
γFe αFe, δFe
Fig. 1 Crystal lattice of iron as a function of temperature.
Source: Ref 1
 
 
 
For the metastable system, which has Fe3C
as the high carbon phase, the following equa
tions apply:
� For the liquid:
%CL=Fe3C
max 195:38 0:33979 T þ 1:51 10�4 T2
(Eq 3)
� For the austenite in equilibrium with
cementite:
%Cg=Fe3C
max 1:6287þ 3:29 10�3 T (Eq 4a)
%Cg=Fe3C
max 0:49 1:1487 10�3 T þ 2:18 10�6 T2
(Eq 4b)
Equations 1(a), 2(a), 3, and 4(a) were calcu
lated based on the data in Fig. 2. Equations 2
(b) and 4(b) were calculated with data obtained
from Ref 3.
Prediction of phase stability relies on knowl
edge of activity of various elements. Activity is
a function of concentration, but usually not a
simple one. In the case of Fe C alloys, an
increase in the activity of carbon in the liquid
parallels a greater ability of carbon to separate
as graphite; that is, an increased activity of car
bon is a measure of an increased tendency of
the alloy to solidify as a stable system. On the
contrary, a decreased activity of carbon reflects
carbide promoting behavior of the system, that
is, metastable solidification tendency.
For gases or ideal solutions, the activities, ai,
are the same as the mole fraction, Ni , (Raoult’s
law), that is ai = Ni. However, most metallurgi
cal solutions (alloys or slags) are strongly non
ideal. The departure from equilibrium (depar
ture from linearity in the activity concentration
relationship) of an element i is measured through
the activity coefficient, which is defined using
either mole fraction or mass% as:
gi
ai
Ni
or fi
ai
%i (Eq 5)
In the first equation, ai is the Raoultian activity
(used in conjunction with the pure component),
while in the second it represents the Henryan
activity (used with the 1 mass% standard state,
i.e., at 25 �C and 1 atm).
For solute concentration in terms of mass%,
up to several percent of solute content, in a first
approximation it can be assumed that log fi
increases or decreases linearly with higher sol
ute concentration:
log fi ei %i (Eq 6)
where ei is the solute interaction coefficient. For
liquid iron at 1600 �C, eC = 0.18.
However, a more accurate description of the
thermodynamic properties of Fe C Si phases
at high carbon and silicon contents requires
the use of second terms. The activity coefficient
of carbon in iron as a function of carbon con
centration and temperature is given in Fig. 3
after Elliott et al. (Ref 5).
The general equation used to describe the
activity of a component 2 in a component 1
according to the Darken formalism (Ref 6) is:
log g2=g
o
2 a12ðN2
2 2N2Þ (Eq 7)
where a12 is the function of temperature, usu
ally of the shape A/T + B, where A and B are
constant.
For the activity of carbon in an iron melt,
Eq 7 is valid for component 1 (iron) and com
ponent 2 (carbon) with (Ref 7):
aFeC ð1270=T þ 1:74Þ (Eq 8)
log goC 1180=T 0:87 (Eq 9)
For the Fe C system, using the natural loga
rithm (base e) instead of the common logarithm
(base 10), Eq 7 can be written as:
ln gC 2714=T 2þ ð2920=T þ 4:01Þð2NC N2
CÞ
(Eq 10)
This equation is in good agreement with exper
imental data, as shown in Fig. 4(a).
Since Kaufmann’s pioneering work in the
1970s, the thermodynamic description of the
1600
0
0.09
0.53
0.16
2.1
2.14
0.65
0.021
0.022 0.76
1153°C
1147°C
4.2
L
4.3
5 10 15 20 25
1400
1200
1000
912°C
800
600
400
Fe
0 1 2 3
C, wt%
Te
m
pe
ra
tu
re
, °
C
C, at.%
4 5 6
1538°C
1394°C
(δFe)
(γFe)
(αFe)
740°C
727°C
L + C (graphite)
1252°C
F
e 3
C
1493°C
Fig. 2 Phase diagram of the binary Fe-C system: The stable system (Fe-graphite) is shown with full lines; the
metastable system (Fe-Fe3C) is shown with dotted lines. Source: Ref 2
Table 1 Temperatures and carbon compositions (mass%) of selected characteristic points
from the Fe C phase diagram from various sources
Point
Okamoto (Ref 2) Neumann (Ref 3) Gustafsson (Ref 4)
Temperature
% C
Temperature
% C
Temperature
% C�C �F �C �F �C �F
Melting point of iron . . . 1538 2800 0 1539 2800 0 1538 2800 0
L + d – L + g – L trijunction . . . 1493 2720 0.53 1499 2730 0.53 1495 2723 0.53
Max solubility of C in d ferrite . . . 1493 2720 0.09 1499 2730 0.08 1495 2723 0.09
Invariant peritectic point . . . 1493 2720 0.16 1499 2730 0.18 1495 2723 0.18
Invariant eutectic point Metastable 1147 2100 4.30 1145 2090 4.30 1148 2100 4.38
Stable 1153 2110 4.20 1152 2110 4.26 1154 2110 4.34
Melting point of Fe3C . . . 1252 2290 6.67 1545 ? 2810 ? 6.687 1225 2240 6.687
Max solubility of C in austenite Metastable 1147 2100 2.14 1145 2090 2.03 1148 2100 2.05
Stable 1153 2110 2.10 1152 2110 2.01 1154 2110 2.03
Invariant eutectoid point Metastable 727 1340 0.76 723 1330 0.80 727 1340 0.76
Stable 740 1370 0.65 738 1360 0.68 738 1360 0.68
Max solubility of C in a ferrite Metastable 727 1340 0.022 723 1330 0.025 727 1340 0.019
Stable 740 1370 0.021 738 1360 0.023 738 1360 0.018
32 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
various phases that exist in metallic liquids and
solutions, compounds, and slags has led to the
development of several models (Ref 8, 9) and
dedicated software. While the so called
Darken Wagner approach or its improvements
(Ref 10) still are in use, in particular for very
dilute solutions, much of the effort in this area
since the 1980s is on the development of com
puterized databases. The thermodynamic prop
erties of most binary systems, many ternary
systems, and higher order systems have been
critically assessed. Figure 4(b) shows the same
calculation as in Fig. 4(a) extended to a range
of temperatures.
The Fe-Si System
With respect to the effect on the phases in the
Fe X systems, the elements can be classified as
a or g promoters. Silicon is a strong a pro
moter. It strongly extends the ferrite field in
the binary Fe X system as seen from the Fe Si
equilibrium diagram in Fig. 5 from Ref 11.
Equation 7 can be used to calculate the activ
ity of silicon in the Fe Si system, with the fol
lowing remarks (Ref 7): NSi � 0.2; component 1
is iron; component 2 is silicon:
aFesi ð153=T þ 3:02Þ (Eq 11)
log goSi 0:914 6863=T (Eq 12)
Introducing these two equations in Eq 7 and
using a natural logarithm rather than a common
logarithm results at 1600 �C in:
ln gSi 6:33þ 7:14ð2NSi N2
SiÞ (Eq 13)
As shown in Fig. 6(a), calculations with this
equation are in good agreement with experi
mental data. Fig. 6(b) shows the same graph
(though in normal logarithm) from an assess
ment of the whole Fe Si system (Ref 12).
The Fe-S System
Sulfur plays a particularly important role in the
solidification of cast iron, because it determines to
a large extent graphite shape. The Fe S diagram is
shown in Fig. 7. There is a peritectic invariant at
1365 �C (2490 �F) where the equilibrium phases
are gFe with 0.05% S, dFe with 0.18% S, and liq
uid with 12% S. At mucsh lower temperature,
there is a eutectic at 30 mass% S and 988 �C
(1810 �F) between liquid, g, and FeS compound.
Solubility of Gases in Iron (Fe-Gas
Systems)
The gases that influence the properties and
structure of ferrous alloys can be divided into
four categories:
� Monoatomic gases (inert gases): Ar
� Elemental diatomic gases: H2, N2, O2
0.04
Reference state:
fC = 1 when %C→0
0
0.2
0.4
lo
g 
X
C
0.6
0.8
1.0
0
0 1 2 3
C, mass%
4 5 6
0.08 0.12
Graphite saturation
1460 °C
(2660 °F)
1560 °C
(2840 °F)
1660 °C
(3020 °F)
1760 °C
(3200 °F)
1260 °C
(2300 °F)
Austenite saturation1360 °C
(2480 °F)
Eutectic
0.16 0.20
XC
Fig. 3 Activity coefficient (fC) of carbon on liquid iron. Source: Ref 5
2.0
1.5
1.0
Eq 8
Chipman (1970)
0.5
–0.5
0A
ct
iv
ity
 c
oe
ffi
ci
en
t o
f c
ar
bo
n,
 In
 γ
C
0 0.05
(a) (b)
0.1
Mole fraction of carbon, XC
0.2 0.250.15
Sanbongi/Ohtani
Richardson/Dennis
Rist/Chipman
Hsu/Poljakov/Samarin
C, wt%
A
ct
iv
ity
 C
0
0 1 2 3 4 5 6 7 8 9 10
0.1
0.2
0.3
0.4
0.5
Liquid
1733
0.6
0.7
0.8
0.9
1.0
Sanbongi et al. 1713 K 1953
Sanbongi et al. 1825 K
Czan–czi et al. 1823 K 1969
Czan–czi et al. 1873 K
Yavayskiy et al. 1908 K 1972
Richardson et al. 1833 K 1953
Marshall et al. 1813 K 1942
’’
’’
1723 K
1763+/ –10 K
3
3
0 3
1833
1933
2033
Fig. 4 Dependence of the activity coefficient of carbon on the mole fraction of carbon in liquid iron. (a) At 1550 �C (2820 �F). Source: Ref 7. (b) At various temperatures from 1460
to 1760 �C (2660 to 3200 �F). Source: Ref 4
Thermodynamics Principles as Applied to Cast Iron / 33
 
 
 
� Compound diatomic gases: CO
� Triatomic gases: SO2, H2O vapor, CO2
In the solid Fe C alloys, gases may be pres
ent in one of these forms:
� Chemically combined: oxide, nitride
� Physically dissolved
� Molecular: cavities of varying sizes, from
defects in the metal lattice to large gas pre
cipitation in the form of blowholes
Monoatomic gases do not dissolve in liquid
iron or Fe C alloys and could be used to flush
out other unwanted gases. The others dissolve
in steel and cast iron and affect their structure
and properties.
Elemental diatomic gases are soluble in both
liquid and solid iron and Fe C alloys. Accord
ing to Sievert’s Law, when a diatomic gas
reacts with a metal it dissolves in the atomic
form:
1
2
i2g $ ½i� (Eq 14)
where i is the elemental gas, g stands for gas
eous state, and the brackets indicate that the
element is dissolved in the liquid or solid phase.
At a given temperature and pressure, once
the solution reaction has proceeded to equilib
rium, there is a constant ratio (equilibrium con
stant) between the active masses of products
and reactants. Assuming that the gas behaves
ideally, so that the activity is equal to its partial
pressure, pi2, then:
K
ai
a
1=2
i2
ai
pi2
p (Eq 15)
where ai is the activity of the dissolved gas,
and K is the equilibrium constant. Assuming
that the activity of gas dissolved in the liquid
follows Raoult’s Law, that is, fi = 1: ai = fi �
[%i] = [%i] , then Eq 15 becomes:
½%i� K pi2
p
(Eq 16)
Thus, at a given temperature, the amount of
gas dissolved in the metal depends on the par
tial pressure of the gas on the liquid metal.
The solubility of gases in liquid metals also
depends on temperature. Indeed, the standard
free energy of solution of a gas in a liquid
metal, DGo, is given by:
�Go RT lnK �Ho T�S (Eq 17)
where R is the gas constant, T is the absolute
temperature, DHo is the standard heat of solu
tion of the diatomic gas, and DSo is the standard
entropy of solution. Then, assuming that DHo
and DSo are independent of temperature, the
van’t Hoff equation gives:
lnK
�Ho
RT
þ�So
R
(Eq 18)
For a partial pressure of gas of 1 atm, com
bining Eq 16 and 18 gives:
ln ½%i�1atm
A
T
þ B (Eq 19)
where A and B are constants. This equation
indicates that solubility increases generally with
temperature in a given phase. However, the
Si, wt%
Te
m
pe
ra
tu
re
, °
C
 
Si, at.%
0
1700
1600
1500
1400
1300
1200
1100
1000
900
800
770°C
912°C
1394°C
1538°C
1410°C
L
17.6
1.9 α2
α1
16.5
10.9
(γFe)
(αFe) F
e 5
S
i 3
)
F
eS
i
αF
eS
i 2
F
e 2
S
i
βF
eS
i 2
13.4
1203°C 1212°C 1220°C 1207°C
1414°C
937°C
982°C 54
54.6
58.25134.2
(Si)
Magn. Trans.
Fe Si
1060°C
700
600
500
10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
825°C
965°C
1212°C
Fig. 5 Phase diagram of the binary Fe-Si system: a1 and a2 are ordered forms of ferrite (aFe). Source: Ref 11
Matoba/Gunji/Kuwana
Hsu/Poljakov/Samarin
Turkdogan/Grieveson/Beisler
Schwerdueger/Engell
Mole fraction of silicon, Xsi
0
8
6
4
2
0
0.1 0.2 0.3 0.4 0.5
A
ct
iv
ity
 c
oe
ffi
ci
en
t o
f s
ili
co
n,
 In
 γ
S
i
In γSi = –6.33 + 7.14 (2XSi – X2
Si)
(a)
0
–3.0
–2.5
–2.0
–1.5
–1.0
–0.5
0
0.5
0.2 0.4 0.6 0.8 1.0
Si, mole fraction
lo
g 
γ S
i
(b)
Fig. 6 Dependence of the activity coefficient of silicon on the mole fraction of silicon in liquid iron at 1600 �C (2910 �F). (a) Source: Ref 7. (b) Source: Ref 12
34 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
solubility is much lower in solid phases than in
liquid, so that during solidification the solubil
ity of gases decreases abruptly. This may result
in gas porosity in the casting.
Hydrogen. The effect of temperature on
hydrogen solubility in pure iron is presented
in Fig. 8. Note the abrupt decrease in solubility
at the melting point of iron. The temperature
dependence of K (with pH2
in atm) for various
states of iron is as follows (Ref 1):
logKaFe; dFe 1418=Tð�KÞ þ 1:628 (Eq 20)
logKgFe 1182=Tð�KÞ þ 1:628 (Eq 21)
logKliqFe 1900=Tð�KÞ þ 2:423 (Eq 22)
Nitrogen. Nitrogen dissolves in iron base
alloys by a reaction similar to Eq 14 with the
equilibrium constant given by Eq 15. The solubil
ity of nitrogen in pure iron is shown in Fig. 8.
According to Warda and Pehlke (Ref 13), the
equilibrium solubility in the liquid phase at
1 atm N2 is described by:
log ½%N�sat 247=Tð�KÞ 1:22 (Eq 23)
The temperature dependence of the equilibrium
constant for various states of iron is (Ref 1):
logKaFe;dFe 1570=Tð�KÞ þ 2:98 (Eq 24)
logKgFe 450=Tð�KÞ þ 2:05 (Eq 25)
logKliqFe 188=Tð�KÞ þ 2:76 (Eq 26)
Oxygen. Oxygen is dissolved in iron base
alloys during melting from the atmosphere or
from oxidized charging materials. The solubil
ity of oxygen in pure iron (see Fig. 9) is much
higher than that in iron base alloys, where the
oxygen level in the melt is controlled by several
oxidation deoxidation reactions.
The solubility of oxygen in pure liquid iron is
given by (Ref 1):
log ½%O�sat 6380=Tð�KÞ þ 2:765 (Eq 27)
Other Gases. Other gases are soluble in liq
uid iron, including CO, H2O, SO2, and H2S (see
Ref 1).
Thermodynamics of Ternary
Fe-C-X Systems
Activity coefficient and solute concentrations
are related for a ternary system by:
log fi eii½%i� þ eij½%j� (Eq 28)
For multicomponent systems, the following
summation is used:
log fi eii½%i� þ �eji ½%j� (Eq 29)
As pointed out previously, first order interactions
may not be sufficient for describing Fe C Si alloys
with compositions relating to cast irons. Further
examples of strong interactions are presented sub
sequently in the case of Fe C S and Fe Si S alloys,
as well as in the case of Fe X N alloys.
Data on interaction coefficients in liquid iron
for carbon, hydrogen, nitrogen, oxygen, and
sulfur are given in Table 2. For metallurgical
reactions involving dilute solutions, it often is
necessary to change the standard state from
pure component to that of 1 mass% in solution.
The free energy change is given by:
�Gs RT ln
0:5585 goi
Mi
(Eq 30)
Te
m
pe
ra
tu
re
, °
C
1600
1500
1400
1300
1200
1100
1000
900
0 0.05 0.10
γFe
Liquid
Sulphur, wt.%
δFe
1365°
0.18% S
988°
913°
0.050% S
0.15 0.20
γFe + Liquid
γFe + FeS
δFe + Liquid
Fig. 7 The iron-sulfur diagram. Source: Ref 1
600
0
10
N
N
γFe δFeαFe
N
H
H
H
20
30
40
50
800 1000 1200 1400 1600 1800 2000
Temperature, °C
1110 1470 1830 2190 2550 2910 3270 3630
Temperature, °F
N
itr
og
en
, m
as
s%
·1
03
H
yd
ro
ge
n,
 p
pm
 (
m
as
s)
Fig. 8 Solubility of hydrogen and nitrogen in iron at a pressure of 1 atm
0.20
28202730 2910
0.16
0.10
Solid Liquid
0.008
1500 16001550
Temperature, °C
O
xy
ge
n,
 m
as
s%
Fig. 9 Solubility of oxygen in pure iron as a function
of temperature. Based on data from Ref 14
Thermodynamics Principles as Applied to Cast Iron / 35
 
 
 
where Mi is the atomic mass (g) of the solute.
For the mass concentration of carbon and sil
iconabove 1%, the values in Table 3 should be
used for the activity coefficient of sulfur. Free
energies of solution, DGs, of selected elements
in liquid iron are given in Table 4.
Detailed equilibrium diagrams of many ter
nary systems are available in Ref 11. Only
selected systems are discussed in this article.
Solubility of Gases in Fe-X Alloys
Nitrogen. Third element additions to the
Fe N system significantly alter nitrogen solubil
ity. For an Fe 2%Si alloy, the nitrogen solubil
ity was established to be (Ref 17):
logð%NÞ 1:178 572=Tð�KÞ (Eq 31)
For the solubility of nitrogen in the Fe C Si
system, Uda and Pehlke (Ref 17) established
the following relationships:
log fN ð%CÞð280=T 0:055Þ
þ ð%SiÞð171=T 0:031Þ 0:005ð%CÞ2
þ 0:0037ð%CÞð%SiÞ (Eq 32)
logð%NÞ ð 306=T 1:201Þ þ 0:5 log pN2
½ð%CÞð280=T 0:055Þ þ%Sið171=T
0:031Þ þ 0:005ð%CÞ2
þ 0:0037ð%CÞð%SiÞ� (Eq 33)
The solution of nitrogen in pure iron base
alloys is diffusion controlled. However, accord
ing to Pehlke and Elliott (Ref 18), in the pres
ence of surface active elements such as oxygen
and sulfur, absorption is controlled by surface
reactions and the rate of solution is reduced,
although the solubility limit is unchanged.
Some typical nitrogen contents in cast iron and
the effect of the amount of steel in the charge are
summarized in Table 5 after Elliott (Ref 19).
Magnesium Vapor. The measured solubility
of magnesium in liquid Fe C and Fe Si alloys is
presented in Fig. 10. From the slopes of the
lines, the following interaction coefficients are
obtained (Ref 1, 22):
eCMg 0:15 and eSiMg 0:046
The intercepts of the lines with the ordinate
axis of %C and %Si = 0 give the equilibrium
constant KMg = [%Mg]fMg/PMg for pure liquid
iron. From the same figure, its temperature
dependence is:
logKMg 4110=Tð�KÞ 3:698 (Eq 34)
which gives the following free energy equation
for the solution of magnesium vapor in liquid
iron:
�GS 78;690þ 70:8T in J mole�1 (Eq 35)
For the quaternary melts Fe C Si Mg, the
activity coefficients of magnesium may be
approximated by the product:
fMg f CMg f SiMg (Eq 36)
In a multicomponent Fe C alloy, there are
extensive chemical reactions between the vari
ous impurities. Magnesium reacts with both
oxygen and sulfur, as can be inferred from
Fig. 11(a), which shows that at the same mag
nesium level in the melt the oxygen activity is
higher at higher sulfur content because some
of the magnesium is used to produce MgS.
The relation for oxygen activity in the presence
of magnesium is given by:
log ao 25751
1
T
þ 6:28 log aMg þ log aMgo
(Eq 37)
where T is in degrees K, and a represents the
activities of Mg and MgO (Ref 23). This rela
tionship and Fig. 11(a) and (b) show that oxy
gen activity deceases with higher magnesium
and lower temperature. The correlation between
the oxygen and nodularity, summarized by
Fig. 11(c), indicates that decreasing oxygen
favors the lamellar graphite to spheroidal
graphite transition, but also that oxygen is not
the only factor affecting it (see the large data
spread).
The Fe-C-Si System
The most important ternary phase diagram
for cast iron is the Fe C Si diagram. The addi
tion of silicon to a binary iron carbon alloy
decreases the stability of Fe3C, which already
Table 2 Interaction coefficients in dilute
solutions of ternary iron base alloys for
carbon, hydrogen, nitrogen, oxygen, and
sulfur at 1600 �C (2910 �F)
Element j e j
C eH
j
e j
N e j
O e j
S
Aluminum 0.043 0.013 –0.028 –3.9 0.035
Boron 0.24 0.05 0.094 –2.6 0.13
Carbon 0.14 0.06 0.13 –0.13 0.11
Cobalt 0.008 0.002 0.011 0.008 0.003
Chromium –0.024 –0.002 –0.047 –0.04 –0.011
Copper 0.016The fourth section at
1373 K (1100 �C) has been calculated by sus
pending all phases but austenite and liquid in
order to illustrate the equilibrium between these
latter phases when the eutectic transformation
proceeds with some undercooling. It is seen that
the silicon partition coefficient is higher than 1
for alloys containing up to 6 wt% Si and does
decrease below 1 only at values far away from
the domain related to cast irons.
The Fe-C-P System
Because the partition coefficient of phospho
rus between solid and liquid iron is low,
0
0
0.5
B
1.0
1.5
2.0
2.5
3.0
4.0
3.5
4.5
C
E1
5.0
2 4
Si, wt%
C
, w
t%
Ferrite α
Ferrite
Hilliard and Owen
Patterson et al.
Ferrite
Austenite γ
Graphite G
Austenite
Austenite
6 8 10
Fig. 12 Projection of the liquidus surface of the Fe-C-Si phase diagram in the iron-rich corner for the stable system.
Source: Ref 12
1.2
0.8
0.4
0
0
O
xy
ge
n 
ac
tiv
ity
, p
pm
0.02 0.04
Mg, %
0.06
S, % no.
0.005 156
0.006 627
0.007 540
0.008 664
0.015 24
0.025 860
(a)
theor.
O
xy
ge
n 
ac
tiv
ity
, p
pm
6.4
0.001
(b)
0.01
0.1
1
10
1300 1400
Temperature, °C 
Mg 0
Mg 0.037–0.055%
1500
6.2 6.0
1/T (× 104)
5.8 5.6 5.4
Oxygen activity at 1420 °C (2590 °F), ppb
0
0
10
20
30
40
50
60
70
80
90
100
100 200 300 400 500 600
(c)
N
od
ul
ar
ity
 (
L/
T
1166 1117:2 48:8 oC
Some typical experimental results for silicon,
chromium, and vanadium are shown in Fig. 20
after Oldfield (Ref 31). These results are in
agreement with calculations.
Partition of a Third Element between
Various Phases in the Fe-C-X System
Calculated and experimental data for parti
tion coefficients of various elements between
austenite and liquid, k
g=L
X , and cementite and liq
uid, k
Fe3C=L
X , are given in Table 7 after Kagawa
and Okamoto (Ref 30). They are valid on a
0.20
L/Fe3P
L/Fe 3
C
L/Gr
L/γ
1100
1100
1100
1050
1050
1000
980
960
1000
1050 1000
980
963 °C
1050
11
00
0.15
0.10
P
ho
sp
ho
ru
s 
co
nt
en
t, 
y P
Carbon content, yC
0.05
Eutectic temperatures
L/γ + Fe3P + Gr: 954 °C
L/γ + Fe3P + Fe3C: 948 °C
0.05 0.10 0.15 0.20
0
0
Fig. 16 Calculated projection of the liquidus surfaces for the stable and metastable Fe-C-P phase diagrams. Source:
Ref 27
1800
1600
1400
1200
1000
800
600
1800
1600
1400
1200
1000
800
600
0.0 0.5 1.0 1.5 2.0
Carbon content, mass% Carbon content, mass%
Te
m
pe
ra
tu
re
, °
C
Te
m
pe
ra
tu
re
, °
C
2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
 α + liquid α + liquid
γ + liquid γ + liquid
γ + α γ + α
γ + α + graphite γ + α + cementite
Liquid
Liquid
G
+
liquid
Austenite Austenite
Austenite + graphite
Cementite
+ liquid
Austenite + cementite
Ferrite + graphite Ferrite + cementite
(a) (b)
Fig. 17 Isopleth Fe-C sections of the Fe-C-Si-Mn system at 3 mass% Si and 1 mass% Mn. (a) Stable system. (b) Metastable system. Calculations performed with Thermocalc
software and the TCFE8 database.
Thermodynamics Principles as Applied to Cast Iron / 39
 
 
 
mole fraction basis. Similar calculations on a
mass fraction basis are reported that were per
formed with TCFE8 for Fe C Si X alloys with
2.5 mass% Si.
Influence of a Third Element on Carbon
Solubility in the Fe-C-X System
According to the compilation by Neumann
(Ref 3), the influence of a third element on the
solubility of carbon in liquid iron can be
expressed by:
�NX
C;liq NX
Cmax
NCmax
(Eq 38a)
�%CX
liq %CX
max %Cmax (Eq 38b)
where �NX
C;liq and �%CX
liq represent the
increase or decrease of carbon solubility in mole
fraction or mass percent, respectively; NCmax
and %Cmax represent the saturation concentra
tion in the iron carbon system calculated from
Eq 1(b) or (c); and NX
Cmax
and %CX
max are the car
bon saturation concentrations in the Fe C X
system.
The changes in carbon solubility resulting
from additions of third elements, X, are shown
in Fig. 21. These data are valid in the range
from 1200 to 1700 �C (2190 to 3090 �F), with
the exception of silicon and sulfur at high con
centrations. In the region of low concentration
of the third element (%X = 0 to 5%), the
change in the solubility of carbon can be
represented by the temperature independent
linear equation:
�NX
C;liq m NX (Eq 39a)
Table 6 Influence of a third element on the temperature change of critical points on the
iron carbon diagram
Element
Max solubility of
C in g, �C/mass% X
Eutectoid,
�C/mass% X
Eutectic,
�C/mass% X
Eutectic,
�C/at.% X
Eutectic,�C/mass%
X with TCFE8
Metast. Stable Metast. Stable Metast. Stable Metast. Stable Metast. Stable
Silicon –10 to –15 +2.5 8 0–30 –10 to –20 +4 –3.25 10.91 –12.1 +5.4
Copper –2 5.2 . . . –10 –2.3 5 –3.32 8.41 –2.9 +6.4
Aluminum –14 8 10 10 –15 8 –7.74 0.91 +4.8 +44.4
Nickel –4.8 4 –20 –30 –6 4 –1.33 6.5 –2.6 +1.9
Cobalt . . . . . . . . . . . . . . . . . . –2.23 1.48 –2.7 +0.1
Chromium 7.3 . . . 15 8 7 . . . 4.97 –10.23 +11.4 –1.9
Manganese 3.2 –2 –9.5 –3.5 3 –2 –2.33 –7.15 –3.6 –4.5
Molybdenum . . . . . . + + . . . . . . –8.9 –12.36 –5.3 –5.4
Tungsten . . . . . . + + . . . . . . –12.13 –15.55 –2.1 –2.6
Boron . . . . . . . . . . . . . . . . . . –15.47 –18.53 –55 –58
Nickel . . . . . . . . . . . . . . . . . . 19.67 16.87 +35 +77
Titanium . . . . . . . . . . . . . . . . . . –16.91 –18.91 –13 –10
Vanadium +6–8 . . . 15 + 6–8 . . . . . . . . . –1.7 –10.4
Phosphorus –180 –180 + 6 –37 –30 16.05 –16.98 –24.2 –18.7
Sulfur . . . . . . . . . . . . . . . . . . . . . –18.53 –20 –20
Metast., metastable; +, increase; , decrease. Source: Ref 28 except for last two columns showing calculations with TCFE8 for the eutectic
Fe
600 1100
1290
1470
1650
1830
2010
2190
2370
2550
2730
2910
700
800
900
1000
1100
0.03
0.04wt% S
0.02
0.01
1200
1300
1400
1500
1600
0.5 1.0 1.5 2.0
γ + Fe3C
γ + L1γ 
δ
γ + α
0
C, wt%
Te
m
pe
ra
tu
re
, °
C
Te
m
pe
ra
tu
re
, °
F
γ + L2 (or FeS)
Fig. 18 Effect of the sulfur content on the boundaries
of the austenite phase field. Source: Ref 28
Fig. 19 Classification of the influence of a third element in the Fe-C-X system on the graphite- or carbide-promoting tendency in cast iron, based on their influence on the eutectic
stable and metastable temperatures. (a) Strong graphite stabilizers Si, Al, Ni, and Cu. (b) Weak graphite stabilizers P and As. (c) Strong carbide stabilizers Cr, V, and Mn.
(d) Weak carbide stabilizers Mo and W
Te
m
pe
ra
tu
re
, °
F
Chromium content, %(a)
1160
Te
m
pe
ra
tu
re
, °
C
1140
1120
1100
 γ + carbide eutectic
2120
2085
2050
2010
0 0.2 0.4 0.6 0.8 1.0 1.2
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
, °
C
Silicon content, %(b)
1160
1150
1140
1130
1120
1100
 γ + graphite eutectic
 γ + carbide eutectic
2120
2100
2085
2065
2030
2050
0.5 1.0 1.5 2.0 2.5
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
, °
C
Vanadium content, %(c)
1160
 γ + graphite eutectic
 γ + carbide eutectic
1140
1120
1100
2120
2085
2050
2010
0 0.2 0.4 0.6 0.8 1.0
 γ + graphite eutectic
Fig. 20 Influence of (a) chromium, (b) silicon, and (c) vanadium on the equilibrium eutectic temperatures of cast irons. Source: Ref 31
40 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
or
�%CX
liq m0 %X (Eq 39b)
where m and m0 are solubility factors. In Eq 39
(a) and (b), m and m0 as well as �NX
C;liq and
�%CX
liq are positive for an increase in the solu
bility of carbon and negative in the other case.
It also is generally accepted that carbide
promoting elements increase the solubility of
carbon (activity coefficient of carbon in solu
tion is decreased), while graphite promoting
elements decrease it (activity coefficient of
carbon in solution is increased). Therefore,
third elements that have a negative solubility
factor promote graphitization, while elements
that have a positive solubility factor promote car
bide formation. The value of the solubility factor
is proportional to their effect. It must be noted,
however, that the accuracy of the prediction of
the behavior of elements based on solubility fac
tor is questioned by Kagawa and Okamoto
(Ref 30) for elements that have a strong influence
on the austenite liquidus.
Furthermore, kinetic effects may supersede
thermodynamic behavior. Experimental work
has shown that elements that have a strong
affinity for nitrogen (such as silicon, aluminum,
and titanium) may exhibit a higher than
expected graphitizing influence. Indeed, it is
accepted that the nitrides may act as nuclei
for the stable eutectic, promoting gray
solidification. This is why experiments show
that titanium, for example, can act either as a
carbide or graphite promoter, depending on
the nitrogen content (Ref 30).
As indicated in experimental values are close
to the theoretical values for many elements.
Equations similar to Eq 39(a) and (b) can be
used to calculate the change in the maximum
solubility of carbon in austenite upon the addi
tion of a third element:
�%CX
g m0g %X (Eq 40)
Data for m0g also are given in Table 8. In
addition, values of m0 and m0g calculated with
TCFE8 are listed.
Thermodynamics of Multicomponent
Iron-Carbon Systems
Carbon solubility in multicomponent systems
is discussed in this section, along with satura
tion degree and carbon equivalent.
Carbon Solubility in Multicomponent
Systems
It can be assumed, at least in the region of
lowconcentrations, that the effect of alloying
elements on the solubility of carbon can be
considered to be additive:
�%CSi;Mn;P;S;...
max �%CSi þ�%CMn þ�%CP
þ�%CS þ . . . (Eq 41)
Assuming further that the values in Table 8
for m and m0, which were determined for the
Table 7 Equilibrium partition coefficients of a third element, X, in the Fe C X system, and
CALPHAD calculations with TCFE8 (last two columns)
Element
k
g=L
X kX
Fe3C=L kX
g=L kX
Fe3C=L
Calc. Exp. Calc. Exp. Calc. TCFE8 Calc. TCFE8
Silicon 1.71 1.72 0 0.05 1.14 0
Copper 1.57 1.62 0.12 0.08 1.6 0
Aluminum . . . 1.15 . . . 0.03 2.3 0.005
Nickel 1.46 1.61 0.43 0.32 1.18 0.38
Cobalt 1.18 1.13 0.59 0.60 1.03 0.54
Chromium 0.53 0.55 1.96 1.95 0.82 3.08
Manganese 0.7 0.75 1.03 1.21 0.67 0.78
Molybdenum 0.41 0.38 0.6 0.84 0.2 0.57
Tungsten 0.23 0.42 0.42 0.88 0.26 0.97
Boron 0.06 . . . 0.22 . . . 0.0015 0.46
Nitrogen 2.04 2.04 2.12 . . . 2.57 0.02
Titanium 0.04 . . . 0.09 0.27 0.29 0.43
Vanadium . . . . . . . . . . . . 0.14 1.86
Phosphorus 0.15 . . . 0.08 0.09 0.07 0
Sulfur 0.06 . . . . . . . . . 0.004 0
The first four columns are on a mole fraction basis, the last two on a mass% basis. Calculations with TCFE8 have been made for an alloy with
2.5 mass% Si at the stable (kg/LX ) and metastable (kX
Fe3C/L) eutectic composition and temperature. 1 mass% was added to the alloy in most calculations,
except for B, N, and Ti where it was 0.1 mass%, and S at 0.01 mass%. Calc., calculated; Exp., experimental. Source: Ref 30
SiC-
saturation
Aluminum
Antimony
Nickel
Manganese
2.0
1.0
–1.0
–2.0
–3.0
–4.0
–5.0
–6.0
0 5 10 15 20 25
0
Molybdenum
Vanadium
(b)
Chromium
CopperTin
Phosphorus
C
ha
ng
e 
in
 th
e 
ca
rb
on
 s
ol
ub
ili
ty
, Δ
%
C
x
Alloying element X, wt%
S
ul
fu
r 
(1
35
0 
°C
)
1290 °C
1490 °C
1690 °C
SiliconSiC-saturation
Aluminum
Antimony
Nickel
Manganese
Molybdenum
Vanadium
0.05
–0.05
0
–0.10
–0.15
(a)
–0.20
0 0.1 0.2 0.3 0.4
Chromium
Copper
Tin
Phosphorus
Mole fraction of the alloying element, X
x
C
ha
ng
e 
in
 th
e 
so
lu
bi
lit
y 
of
 th
e 
ca
rb
on
. Δ
X
x C
S
ul
fu
r 
(1
35
0 
°C
)
1290 °C
1490 °C
1690 °C
Silicon
Fig. 21 Influence of third elements, X, on the solubility of carbon in molten iron. (a) In mole fraction. (b) In mass percent. Source: Ref 3
Thermodynamics Principles as Applied to Cast Iron / 41
 
 
 
ternary Fe C X system, can be extended to mul
ticomponent systems, the saturation concentra
tion of carbon in multicomponent systems can
be calculated as (Ref 3):
%CSi;Mn;P;S;...
max %Cmax þ
X
�%CX
max (Eq 42)
which is Eq 37(b) rewritten for multicomponent
systems. Then, using for example, Eq 1(b), and
data from Table 8, it can be written that:
%CSi;Mn;P;S...
max 1:3þ 2:57 10�3 Tð�CÞ
0:31 %Siþ 0:027 %Mn
0:33 %P 0:4 %S� . . .
(Eq 43)
Saturation Degree and Carbon
Equivalent
It is well known that mechanical properties of
cast iron strongly depend on the amount and shape
of graphite. For hypoeutectic gray irons, the
amount of graphite depends only on the amount
of eutectic. The amount of eutectic can be calcu
lated with the lever rule. For hypoeutectic irons:
Sr
% eutectic
% eutecticþ%austenite
%Canal %Cg
%Canal %Cg þ%Ceut %Canal
%Canal %Cg
%Ceut %Cg
(Eq 44)
where Sr is the rectified saturation degree,
%Canal is the analyzed carbon content of cast
iron, and %Ceut and %Cg represent the carbon
content of the eutectic and of the austenite,
respectively, at the eutectic temperature in the
multicomponent system. Assuming the solubility
factors can be extrapolated to the hypoeutectic
region, %Ceut can be calculated with Eq 43. A
similar approach can be used to calculate %Cg
using data in Table 8 for m0g:
%Cg 2:11 0:11 %Si 0:35 %Pþ 0:006 %Mn
0:08 %S (Eq 45)
In foundry practice, the following simplified
form of Eq 44 is used to define the saturation
degree:
SC
%Canal
%Ceut
%Canal
4:26 0:31 �%Siþ 0:027 �%Mn 0:33 �%P� . . .
(Eq 46)
Further simplification gives:
SC
%Canal
4:26 0:31 %Si 0:33 %P
(Eq 47)
where SC 1 for hypereutectic
irons.
Saturation degree is used in most of the
European literature. In the Anglo Saxon
literature and foundry practice, carbon equiva
lent rather than saturation degree is used. Car
bon equivalent (CE) is equal to the carbon
content plus the amount of carbon equivalent
from the added elements, and can be calculated
as:
CE %Canal �%CSi �%CMn �%CP
�%CS . . . %Canal þ 0:31 %Si
0:027 %Mnþ 0:33 %Pþ 0:4 %S� . . .
(Eq 48)
The thermodynamic basis for this relationship
is that the carbon equivalent for the multicom
ponent solution has the same carbon activity
as the equivalent amount of carbon in the
binary solution (Ref 19). Thermodynamic cal
culations for a 3.5% C iron using available
interaction coefficients for a liquid iron temper
ature of 1500 �C (2730 �F) produced (Ref 32):
CE %Canal þ 0:32 %Siþ 0:33 %P (Eq 49)
These coefficients as well as the ones in
Eq 48 are in good agreement with the experi
mental ones evaluated in Ref 33.
Eutectic iron has CE = 4.26%. It must be
noted that while Sr gives directly the amount
of eutectic in the structure (e.g., Sr = 0.9 means
90% eutectic), SC and CE, although easier to
calculate, do not allow for direct estimation of
the amount of eutectic.
Castro et al. (Ref 34) used a different
approach for the calculation of the change in
the solubility of carbon by addition of a third
element. Assuming that the liquidus tempera
tures of austenite and graphite can be expressed
by linear relations of composition, they esti
mated the slopes of these lines from phase dia
gram information, and then calculated %Ceut as
a function of these slopes and the amount of
third element. The proposed equation for car
bon equivalent is:
CE %Canal þ 0:28 %Siþ 0:007 %Mnþ 0:303
%Pþ 0:033 %Crþ 0:092 %Cuþ 0:011
%Moþ 0:054 %Ni (Eq 50)
Note that the effect of all elements appearing
in Eq 50 is of the same sign that is in disagree
ment with data in Table 8 for chromium, man
ganese, and molybdenum.
Composition Control of
Iron-Carbon Melts
Composition determines the microstructure,
the properties, and the soundness of the casting.
Composition control is achieved during melting
and subsequent melt treatment. Certain ele
ments such as carbon, silicon, and alloying ele
ments must be kept under prescribed limits,
while others, such as sulfur, oxygen, and gases,
must be reduced. While deoxidation a very
important purification process in steel is not
Table 8 Experimental and calculated solubility factors of various third elements for
carbon saturated Fe C X melts and calculations made with TCFE8 in the stable system(a)
Third element X
DNX
liq m � NX D%CX
liq m0 �%X D%CX
Y m0g �%X TCFE8
m exp. m calc. Validity m0 exp. m0 calc. Validity m0g Validity m0 m0g
B –0.575 –0.51–0.11 –0.104an effort to improve the clarity of the
Handbook. The most notable exception is the use of g/cm3 rather than
kg/m3 as the unit of measure for density (mass per unit volume).
SI practice requires that only one virgule (diagonal) appear in units
formed by combination of several basic units. Therefore, all of the units
preceding the virgule are in the numerator and all units following the
virgule are in the denominator of the expression; no parentheses are
required to prevent ambiguity.
vi
List of Contributors and Reviewers
Aquil Ahmad
Eaton Corp., retired
Tito Andriollo
Technical University of Denmark
Juan Asensio-Lozano
University of Oviedo
Brian Bendig
Penticton Foundry Ltd.
Roberto E. Boeri
National University of Mar del Plata
A.A. Burbelko
AGH University of Science and
Technology
Pierre-Marie Cabanne
Rio Tinto Iron and Titanium
John Campbell
University of Birmingham
Manuel Castro
Cinvestav, Unidad Saltillo
A.V. Catlina
Caterpillar Inc.
Sidney Clouser
SIFCO ASC
Steve Dawson
SinterCast Limited
Lucian Vasile Diaconu
Jönköping University
Attila Diószegi
Jönköping University
J.L. Dossett
Consultant
Diego O. Fernandino
National University of Mar del Plata
Michael E. Finn
Finn Metalworking and Cutting Solutions
Hasse Fredriksson
KTH Stockholm
Martin Gagné
Rio Tinto Iron and Titanium
Marcin Górny
AGH University of Science and
Technology
Serge Grenier
Rio Tinto Iron and Titanium
John A. Griffin
The University of Alabama
Wilson Guesser
Tupy S.A. and State University of Santa
Catarina
Richard Gundlach
Element Materials Technology
Dika Handayani
The Pennsylvania State University
Jesper Hattel
Technical University of Denmark
K. Hayrynen
Applied Process Inc.
Jayson L. Helsel
KTA Tator, Inc.
Amjad Javaid
Natural Resources Canada Canmet
MATERIALS
W. Kapturkiewicz
AGH University of Science and
Technology
J.R. Keough
Applied Process Inc.
Chantal Labrecque
Rio Tinto Iron and Titanium
Jacques Lacaze
Université de Toulouse
Pello Larranaga
Maristas Azterlan Engineering
Simon N. Lekakh
Missouri University of Science and
Technology
Laurentiu Nastac
The University of Alabama
Jakob Olofsson
Jönköping University
Tom Prucha
American Foundry Society
Jingjing Qing
Missouri University of Science and
Technology
Janina M. Radzikowska
The Foundry Research Institute, retired
Von Richards
Missouri University of Science and
Technology
Iulian Riposan
Politehnica University of Bucharest
Valery Rudnev
Inductoheat Inc.
Roxana Ruxanda
Emerson Climate Technologies
Kumar Sadayappan
Natural Resources Canada Canmet
MATERIALS
Angella Sell
Applied Process Inc.
Torbjorn Skaland
Elkem Foundry Products
Doru M. Stefanescu
The Ohio State University and
The University of Alabama
Ramon Suarez
Maristas Azterlan Engineering
K.V. Sudhakar
Montana Tech of the University of
Montana
Bo Sundman
CEA Saclay
Ingvar L. Svensson
Jönköping University
Peter Svidró
Jönköping University
József Tamás Svidró
Jönköping University
Judit Svidró
Jönköping University
Kenneth B. Tator
KTA Tator, Inc.
Jesper Thorborg
Technical University of Denmark
Harry Tian
GIW Industries
vii
Niels Skat Tiedje
Technical University of Denmark
A. Udroiu
Maristas Azterlan Engineering
George F. Vander Voort
Struers Inc., consultant
Nikolaj Vedel-Smith
Technical University of Denmark
Robert Voigt
The Pennsylvania State University
Ron Walling
Cummins Inc.
Menk Werner
Georg Fischer AG
Charles V. White
Kettering University
Douglas White
Elkem Materials Inc.
Franco Zanardi
Zanardi Foundry
M. Zhu
Southeast University, China
viii
Officers and Trustees of ASM International (2016–2017)
William E. Frazier
President
Naval Air Systems Command
Frederick E. Schmidt
Vice President
Advanced Applied Services
Jon D. Tirpak
Immediate Past President
ATI
William T. Mahoney
Managing Director
ASM International
Craig D. Clauser
Treasurer
CCECI
Ellen Cerreta
Los Alamos National Laboratory
Kathryn Dannemann
Southwest Research Institute
Ryan M. Deacon
United Technologies Research Center
Larry D. Hanke
Materials Evaluation and Engineering
Roger A. Jones
Solar Atmospheres Inc.
Sudipta Seal
University of Central Florida
T.S. Sudarshan
Materials Modification Inc.
David B. Williams
The Ohio State University
John D. Wolodko
University of Alberta
Student Board Members
Swetha Barkam
University of Central Florida
Allison E. Fraser
Lehigh University
Rachel Stewart
Colorado School of Mines
Members of the ASM Handbook Committee (2016–2017)
Alan P. Druschitz, Chair
Virginia Tech
Craig J. Schroeder, Vice Chair
Element
George Vander Voort, Immediate Past Chair
Vander Voort Consulting, LLC
Craig D. Clauser, Board Liaison
Craig Clauser Engineering Consulting
John D. Wolodko, Board Liaison
University of Alberta
Sabit Ali
National Bronze and Metals Inc.
Kevin R. Anderson
Mercury Marine
Scot Beckwith
Sampe
Narendra B. Dahotre
University of North Texas
Volker Heuer
ALD Vacuum Technologies GmbH
Martin Jones
Ford Motor Company
Dana Medlin
SEAL Laboratories
Brett A. Miller
MR Metallurgical Services
Erik M. Mueller
National Transportation Safety Board
Scot M. Olig
U.S. Naval Research Laboratory
Valery Rudnev
Inductoheat Incorporated
Satyam Suraj Sahay
John Deere Technology Center India
Jeffery S. Smith
Material Processing Technology, LLC
Jaimie S. Tiley
U.S. Air Force Research Laboratory
George E. Totten
G.E. Totten & Associates, LLC
Dustin A. Turnquist
Spectrum Forensics, LLC
Junsheng Wang
Kaiser Aluminum Trentwood
Charles V. White
Kettering University
Dehua Yang
Ebatco
Joseph Newkirk
ex-officio member
Missouri University of Science
and Technology
Chairs of the ASM Handbook Committee
J.F. Harper
(1923 1926) (Member 1923 1926)
W.J. Merten
(1927 1930) (Member 1923 1933)
L.B. Case
(1931 1933) (Member 1927 1933)
C.H. Herty, Jr.
(1934 1936) (Member 1930 1936)
J.P. Gill
(1937) (Member 1934 1937)
R.L. Dowdell
(1938 1939) (Member 1935 1939)
G.V. Luerssen
(1943 1947) (Member 1942 1947)
J.B. Johnson
(1948 1951) (Member 1944 1951)
E.O. Dixon
(1952 1954) (Member 1947 1955)
N.E. Promisel
(1955 1961) (Member 1954 1963)
R.W.E. Leiter
(1962 1963) (Member 1955 1958, 1960 1964)
D.J. Wright
(1964 1965) (Member 1959 1967)
J.D. Grahaui
(1966 1968) (Member 1961 1970)
W.A. Stadtler
(1969 1972) (Member 1962 1972)
G.J. Shubat
(1973 1975) (Member 1966 1975)
R. Ward
(1976 1978) (Member 1972 1978)
G.N. Maniar
(1979 1980) (Member 1974 1980)
M.G.H. Wells
(1981) (Member 1976 1981)
J.L. McCall
(1982) (Member 1977 1982)
L.J. Korb
(1983) (Member 1978 1983)
T.D. Cooper
(1984 1986) (Member 1981 1986)
D.D. Huffman
(1986 1990) (Member 1982 2005)
D.L. Olson
(1990 1992) (Member 1982 1988, 1989 1992)
R.J. Austin
(1992 1994) (Member 1984 1985)
W.L. Mankins
(1994 1997) (Member 1989 )
M.M. Gauthier
(1997 1998) (Member 1990 2000)
C.V. Darragh
(1999 2002) (Member 1989 )
Henry E. Fairman
(2002 2004) (Member 1993 2005)
Jeffrey A. Hawk
(2004 2006) (Member 1997 )
Larry D. Hanke
(2006 2008) (Member 1994 )
Kent L. Johnson
(2008 2010) (Member 1999 )
Craig D. Clauser
(2010 2012) ( Member 2005 )
Joseph W. Newkirk
(2012 ) (Member 2005 )
ix
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
A History of Cast Iron
Doru M. Stefanescu, The Ohio State University and The University
of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
The Beginnings of Metal Casting and of the Iron Age . . . . . . . 3
Early Cast Iron in Mesopotamia and China . . . . . . . . . . . . . . . 4
Cast Iron in Europe in the Medieval Ages . . . . . . . . . . . . . . . . 5
Early Modern Period (16th to Mid 18th Century) . . . . . . . . . . . 5
Late Modern Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Cast Iron A High Tech, Economical, Modern
Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Classification and Basic Types of Cast Iron
Revised and updated by Doru M. Stefanescu, The Ohio State
University1. E.T. Turkdogan, Fundamentals of Steel
making, The Institute of Materials, London,
1996
2. H. Okamoto, Phase Diagrams of Binary
Iron Alloys, ASM International, 1992
3. F. Neumann, The Influence of Additional
Elements on the Physico Chemical Behavior
of Carbon in Carbon Saturated Molten Iron,
Recent Research on Cast Iron, H.D. Mer
chant, Ed., Gordon and Breach, 1968, p 659
4. P. Gustafsson, A Thermodynamic Evalua
tion of the Fe C System, Scand. J. Metall.,
Vol 14, 1985, p 259 267
5. J.F. Elliott, M. Gleisler, and V. Ramak
rishna, Thermochemistry for Steel Making,
Vol 2, Addison Wesley Publ. Co., Reading,
MA, 1963
6. L.S. Darken, The Thermodynamics of Ter
nary Metallic Solutions, Trans. TMS, Vol
239, 1967, p 80 90
7. F. Neumann and E. Dötsch, Thermodynam
ics of Fe C Si Melts with Particular Empha
sis on the Oxidation Behavior of Carbon
and Silicon, The Metallurgy of Cast Iron, B.
Lux et al., Ed., Georgi Publishing, 1975, p 31
8. N. Saundersand and A.P. Miodownik,
CALPHAD Calculation of Phase Dia
grams, A Comprehensive Guide, Pergamon
Materials Series, 1998
9. H.L. Lukas, S.G. Fries, and B. Sundman,
Computational Thermodynamics, the
CALPHAD Method, Cambridge University
Press, 2007
10. C.H.P. Lupis, Chemical Thermodynamics
of Materials, North Holland, 1983
11. O. Kubaschewski, Alloy Phase Diagrams,
Vol 3, ASM Handbook, ASM International,
1992, p 2.203
12. J. Lacaze, and B. Sundman, An Assessment
of the Fe C Si System, Metall. Trans. A,
Vol 22A, 1991, p 2211 2223
13. H. Warda and R.C. Pehlke, Nitrogen Solu
tion and Titanium Nitride Precipitation in
Liquid Fe Cr Ni Alloys, Metall. Trans. B,
1977, p 441
14. H.A Wriedt, Alloy Phase Diagrams, Vol 3,
ASM Handbook, ASM International, 1992,
p 2.199
15. E.T. Turkdogan, Physical Chemistry of
High Temperature Technology, Academic
Press, 1980
16. G.K. Sigworth and J.F. Elliott, The Ther
modynamics of Liquid Dilute Iron Alloys,
Met. Sci., Vol 8, 1974, p 298 331
17. M. Uda and R.D. Pehlke, Cast Metals Res.
J., Vol 10, 1974, p 30
18. R.D. Pehlke and J.F. Elliott, Solubility of
Nitrogen in Liquid IronAlloys: 1. Thermody
namics, Trans. Met. Soc. AIME, 1960, p 1088
19. R. Elliott, Cast Iron Technology, Butter
worths, London, 1988
20. P.K. Trojan and R.A. Flinn, A New Method
for Determination of Liquid Equilibria as
Applied to the Fe C Si Mg System, Trans.
ASM, Vol 54 (No. 3), 1961, p 549 566
21. P.J. Guichelaar, P.K. Trojan, T. Cluhan,
and R.A. Flinn, A New Technique for
Vapor Pressure Measurement Applied to
the Fe Si Mg System, Metall. Trans. B,
1971, p 3305 3313
22. E.T. Turkdogan, Foundry Processes, Their
Chemistry and Physics, S. Katz and C.F.
Landfeld, Ed., Plenum Press, New York,
1988, p 53 97
23. F. Mampaey, D. Habets, J. Plessers, and F.
Seutens, The Use of Oxygen Activity Mea
surement to Determine Optimal Properties
of Ductile Iron during Production, Giesser
eiforschung (Int. Foundry Res.), Vol 60
(No. 1), 2008, p 2 19
24. F. Mampaey and K. Beghyn, Oxygen
Activity in Cast Iron Measured in Induc
tion Furnace at Variable Temperature,
Trans. AFS, Vol 114, 2006, paper 06 11
25. J.O. Andersson, T. Helander, L. Höglund,
P.F. Shi, and B. Sundman, Thermo Calc
and DICTRA, Computational Tools for
Materials Science, CALPHAD, Vol 26,
2002, p 273 312
26. Thermo Calc Software TCFE8 Steels/Fe
alloys database version 8, www.thermo
calc.com/media/10864/tcfe8 flyer format
ted bh.pdf (accessed October 6, 2016)
27. M. Hillert and P.O. Söderholm, White
and Gray Solidification of the Fe C P Eutec
tic, The Metallurgy of Cast Iron, B. Lux
et al., Ed., Georgi Publishing, 1975, p 197
28. H. Ohtani and T. Nishizawa, Calculation of
Fe C S Ternary Phase Diagram, Trans.
ISIJ, Iron and Steel Institute of Japan, Vol
26, 1986, p 655 663
29. N.G. Girsovitch, Ed., Spravotchnik po
tchugunomu litja (Cast Iron Handbook),
Mashinostrojenie, 1978
30. A. Kagawa and T. Okamoto, Partition of
Alloying Elements on Eutectic Solidifica
tion of Cast Iron, The Physical Metallurgy
of Cast Iron, H. Fredriksson and M. Hillert,
Ed., North Holland, 1985, p 201
31. W. Oldfield, BCIRA Journal, Vol 9, 1961,
p 506 518
32. R.C. Creese and G.W. Healy, Metallurgical
Thermodynamics and the Carbon Equiva
lent Equation, Met. Trans. B, Vol 16B,
1985, p 169
33. R.W. Heine, The Fe C Si Solidification
Diagram for Cast Irons, Trans. AFS, Vol
94, 1986, p 391
34. M. Castro, M. Herrera, M.M. Cisneros,
G. Lesoult, and J. Lacaze, Simulation of
Thermal Analysis Applied to the Descrip
tion of the Solidification of Hypereutectic
SG Irons, Int. J. Cast Metal. Res., Vol 11,
1999, p 369 374
35. D.M. Stefanescu and S. Katz, Thermody
namic Properties of Iron Base Alloys,
Casting, Vol 15, ASM Handbook, ASM
International, 2008, p 41 55
36. W. Weis, The Importance of Deoxidation
in the Crystallization of Cast Iron, The
Metallurgy of Cast Iron, B. Lux et al.,
Ed., Georgi Publishing, 1975, p 69
16
1470 1830 2190 2550 2910
8
0
–8
m
n
Fe-Mn-O
Fe-S-O
γ +
Oxi.+
FeS
j
q
p
–16
–20
800 1000 1200 1400
Temperature, °C
1600
1
M
n,
 w
t%
 fo
r 
cu
rv
e 
j
R
T
 In
 %
 M
n,
 k
ca
l
Temperature, °F
0.001
0.01
0.1
10
100
Mn-S-O
γ + Oxi. + ‘Mns’ γ + Oxi. + l1
γ + Oxi. + l2
Fig. 25 Univariant equilibria in the Fe-Mn-S-O
system in the presence of g-Fe and (Mn,Fe)O
phases. Source: Ref 43
44 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
37. R.J. Fruehan, Ladle Metallurgy: Principles
and Practices, Iron and Steel Society of
AIME, 1985, p 7 8
38. S. Katz and C.F. Landefeld, Cupola Desul
furization, Cupola Handbook, American
Foundrymen’s Society, 1984, p 351 363
39. S. Asai and I. Muchi, Fluid Flow and Mass
Transfer in Gas Stirred Ladles, Foundry
Processes: Their Chemistry and Physics,
S. Katz and C.F. Landefeld, Ed., Plenum
Press, 1988, p 261 329
40. S. H. Kim and R.J. Fruehan, Physical Mod
elling of Liquid/Liquid Mass Transfer in
Gas Stirred Ladles, Metall. Trans. B, Vol
18, 1987, p 381 390
41. J. Ishida, K. Yamaguchi, S. Sugiura, N.
Demukai, and A. Notoh, Denki Seiko, Mar
uzen Yushodo Co Ltd, Vol 52, 1981, p 2 8
42. C.F. Landefeld and S. Katz, Kinetics of
Iron Desulfurization by CaO CaF2,
Proceedings of the Fifth International
Iron and Steel Congress, Vol 6, Iron
and Steel Society of AIME, 1986,
p 429 439
43. E.T. Turkdogan and G.J.W. Kor, Metall.
Trans., Vol 2, 1971, p 1561
Thermodynamics Principles as Applied to Cast Iron / 45
 
 
 
The Liquid State and Principles of
Solidification of Cast Iron
Doru M. Stefanescu, The Ohio State University and The University of Alabama
Roxana Ruxanda, Emerson Climate Technologies
UPONMELTINGOFGRAPHITIC IRON, the
graphite will dissolve if enough time at the super
heating temperature is allowed. Thus, the struc
ture of liquid iron is a function of chemical
analysis, temperature, and holding time in the liq
uid state. X ray analysis on liquid cast iron
demonstrated that, for a Fe 4.1%C 1%Si alloy,
the size of undissolved graphite immediately after
melting was 36 to 38 nm (Ref 1). It decreased by
half after 5 to 6 h holding at 1220 �C (2230 �F).
The graphite completely dissolved after approxi
mately 11 h. For a low silicon alloy, Fe 4%
C 0.02%Si, the size of the graphite particles after
melting was approximately 17 nm, and the graph
ite dissolved completely in 3 to 5 h.
Iron carbon alloys with low carbon content
(steels) in liquid state are condensed phases with
compact distribution of atoms in short range
order. X ray and neutron wide angle diffraction
performed by Steeb and Maier (Ref 2) on molten
iron carbon alloys with up to 5.5 wt% C in the
temperature range of 1150 to 1600 �C (2100 to
2910 �F) found that, for pure iron, the number
of nearest neighbors (number of atoms in the first
coordination sphere) is NI = 9, and the nearest
neighbor distance is rI = 2.6 � 10�10 m. Up to
1%C, the packing density is increased as the dis
tance increases to 2.67 � 10�10 m, and the number
of neighbors increases to 10.4 (Fig. 1). Between
1.8 and 3% C, the nearest neighbor distance
remains constant, but the number of neighbors
increasesto 11.2 atoms, which means that the
packing density is further increased. Amaximum
packing density is reached at 3% C, and it
remains constant at higher carbon concentra
tions. At 3.5% C, the authors concluded that
short range ordered regions rich in carbon exist
in the melt, but they were unable to establish
their structure. Indeed, viscosity measurements
summarized in Fig. 2 (Ref 3) show a correlation
between viscosity and percentage of carbon.
The melts containing short range ordered
regions show high viscosity values.
Thus, liquid iron carbon alloys with low
carbon content (3.5% C, i.e., cast irons rich in carbon)
are colloidal dispersed systems with carbon
clusters in liquid solution. The nature of the car
bon clusters is not clear. There are two hypoth
eses regarding their structure: they are Fe3C
molecules, or they are Cn molecules.
From thermodynamic considerations, Darken
(Ref 4) concluded that the existence of Fe3C
molecules in iron carbon melts is possible.
Activity measurements also support short range
order similar to Fe3C (Ref 5). Because the
nucleation energy for Fe3C is smaller than that
for graphite, it is thermodynamically possible
for the carbon rich regions to exist as Fe3C
clusters.
Other investigators consider the carbon clus
ters to be stable in iron carbon melts with more
than 2% C (Ref 6, 7). Their size is supposed to
be in the range of 1 to 20 nm, and it increases
with the carbon equivalent, lower silicon con
tent, and lower holding time and temperature.
According to Ref 2, these carbon clusters contain
approximately 15 atoms (C15) with a stability
time interval of approximately 10�10 s. It is to
be expected that the carbon rich clusters existing
in molten iron carbon alloys are in dynamic
equilibrium and that they diffuse within the melt.
Fundamentals of Solidification of
Cast Iron
Solidification processing is one of the oldest
manufacturing processes, because it is the prin
cipal component of metal casting processing.
While solidification science evolved from the
need to better understand and further develop
casting processes, solidification science today
(2106) is at the base of many new develop
ments that fall out of the realm of traditional
metal casting.
Solidification is, strictly speaking, the trans
formation of liquid matter into solid matter.
The microstructure that results from solidifica
tion may be the final one, in which case it
directly affects the mechanical properties of
the product. In other cases, heat treatment or
other processes may be used after solidification
to further modify the solidification microstruc
ture. However, the outcome of this additional
processing will be greatly affected by the solid
ification microstructure.
Length Scale of Solidification
Structures
The effect of solidification on the morphol
ogy of the matrix can be deconstructed at four
different length scales (Ref 8) (Fig. 3):
ASM Handbook, Volume 1A, Cast Iron Science and Technology 
D.M. Stefanescu, editor
DOI: 10.31399/asm.hb.v01a.a0006311
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
Neutron
X-ray
Iron, mass%
N
um
be
r 
of
 a
to
m
s
949596979899
9
10
11
12
100
Fig. 1 Number of atoms in the first coordination
sphere obtained by neutron diffraction and
x-ray. Source: Ref 2
rI
rI
NI
NI
const.
rI, NI const.
Short range ordered
regions
4.54.03.53.02.52.01.51.00.5
4
0
5
6
7
8
9
10
4
5
6
7
8
9
10
Carbon concentration, wt% V
is
co
si
ty
 o
f i
ro
n-
ca
rb
on
 a
llo
ys
, c
P
V
is
co
si
ty
 o
f i
ro
n-
ca
rb
on
 a
llo
ys
, P
a 
· s
 ×
 1
0–3
�, �
Fig. 2 Viscosity of iron-carbon alloys as a function of
carbon concentration. Source: Ref 3
 
 
 
� The macroscale (macrostructure) is of the
order of 10�3 to 1 m. Elements of the macro
scale include shrinkage cavity, macrosegre
gation, cracks, surface roughness (finish),
and casting dimensions. A typical example
of a solidification macrostructure is given in
Fig. 4, after Boeri and Sikora (Ref 9), which
illustrates columnar grains growing inward
into the cast iron rod.
� The mesoscale is of the order of 10�4 m. It
allows description of the microstructure fea
tures at the grain level, without resolving the
intricacies of the grain structure. As seen in
Fig. 3, the solid/liquid (S/L) interface is not
sharp. Three regions can be observed: liquid,
mushy (containing both liquid and solid
grains), and solid. Mechanical properties
are affected by the solidification structure
at the mesoscale level, which is described
by features such as grain size and type
(columnar or equiaxed), the type and con
centration of chemical microsegregation,
and the amount of microshrinkage, porosity,
and inclusions. The term mesoscale has been
introduced in solidification science to more
accurately describe the results of computer
models. An example of a solidification
mesoscale structure is given in Fig. 5, after
Moore (Ref 10).
� The microscale (microstructure) is of the
order of 10�6 to 10�5 m. The microscale
describes the complex morphology of the
solidification grain. In a sound casting,
mechanical properties depend on the solidifi
cation structure at the microscale level. To
evaluate the influence of solidification on
the properties of the castings, it is necessary
to know the as cast grain morphology
(i.e., size and type, columnar or equiaxed)
and the length scale of the microstructure
(interphase spacing, e.g., dendrite arm spacing
and eutectic lamellar spacing). The term
microstructure is the classic term used in met
allography to describe features observed under
the microscope, as seen in the micrograph
from Fig. 6, which shows graphite and pearlite
in a gray iron.
� The nanoscale (atomic scale) is of the order
of 10�9 m (nanometers) and describes the
atomic morphology of the S/L interface. At
this scale, nucleation and growth kinetics
of solidification are discussed in terms of
the transfer of individual atoms from the liq
uid to the solid state. Features such as dislo
cations, atomic layers, and even individual
atoms are observed with electronic micro
scopes. An example of graphite layers in a
spheroidal graphite aggregate seen at nano
scale magnification is given in Fig. 7, after
Purdy and Audier (Ref 11).
As discussed in some detail in the following
sections, two basic phenomena must take place
in the liquid for solidification to occur: under
cooling and nucleation. If these conditions are
met, the nuclei can grow into the new solid
grains.
Undercooling
Global equilibrium phase diagrams are
frequently used to understand alloy behavior
when the alloy is cooled from the liquid state
to room temperature. Global equilibrium requires
uniform chemical potentials and temperature
across the system. Under such conditions, no
changes occur with time. When global equilib
rium exists, the fraction of phases can be calcu
lated with the lever rule, and the phase diagram
gives the uniform composition of the liquid and
solid phases. Such conditions exist only when
the solidification velocity is much smaller than
the diffusion velocity. Uniform chemical poten
tials and temperature may truly appear only when
solidification takes place over geological times.
Macro Meso
Solid Mush Liquid
Micro Nano
Fig. 3 Solidification length scale. Source: Ref 8
Fig. 4 Macrostructure of 30 mm (1.2 in.) diameter bars
showing columnar grains (primary austenite
dendrites). Source: Ref 9
Fig. 5 Room-temperature eutectic grain structure in
lamellar graphite iron. Original magnification:
14�. Source: Ref 10
Fig. 6 Pearlitic gray iron showing type A graphite and
fine pearlite
The Liquid State and Principles of Solidification of Cast Iron / 47
 
 
 
Solidification as encountered in common
processes does not occur at equilibrium, because
during solidification of most castings, both tem
perature and composition gradients exist acrossthe casting. Elementary thermodynamics demon
strates that a liquid cannot solidify unless some
undercooling below the equilibrium (melting)
temperature, Te, occurs. Five types of solidifica
tion undercooling have been identified: kinetic
undercooling, thermal undercooling, constitu
tional (solutal) undercooling, curvature under
cooling, and pressure undercooling.
Nevertheless, in most cases, the overall solid
ification kinetics can be described with suffi
cient accuracy by using the local equilibrium
condition, that is, by using the mass, energy,
and species transport equations to express the
temperature and composition variation within
each phase and by using equilibrium phase dia
grams to evaluate the temperature and composi
tion of phase boundaries, such as the S/L
interface (corrections must be made for inter
face curvature). Most phase transformations,
with the exception of massive (partitionless)
and martensitic transformations, can be
described with the local equilibrium condition.
When the stable phase cannot nucleate or grow
sufficiently fast (e.g., gray to white transition in
cast iron), metastable local equilibrium can
occur. For both stable and metastable local
equilibria, the chemical potentials of the com
ponents across the interface must be equal for
the liquid and for the solid.
However, at large undercooling, the solidifi
cation velocity exceeds the diffusive speed of
solute atoms in the liquid phase (rapid solidifi
cation). The solute is trapped into the solid at
levels exceeding the equilibrium solubility.
Typically, for solute trapping, the solidification
velocity must exceed 5 m/s (16 ft/s).
Kinetic Undercooling. When a number of
simplifying assumptions are introduced (pure
metal, constant pressure, no thermal gradient in
the liquid, and flat S/L interface that is, the
radius of curvature of the interface is r = 1),
the only undercooling driving the S/L interface
is the kinetic undercooling. It is a nanoscale
length effect, resulting from the net difference
in atoms transported from L to S and from S
to L. Typically for metals, the kinetic undercool
ing is very small, of the order of 0.01 to 0.05 K.
When the simplifying assumptions are
relaxed to reflect typical solidification scenar
ios, the free energy of the liquid solid system
upon the solidification of a discrete volume of
liquid, DFv, will increase by:
�Fv �GT þ�Gc þ�Gr þ�FP (Eq 1)
where F and G are the Helmholtz and Gibbs
free energy, respectively. The four right hand
terms are the change in free energy because of
temperature, composition, curvature, and pres
sure variation, respectively. Solidification can
not occur unless each of these energies is
balanced by a corresponding undercooling of
the system, as discussed in this section.
Thermal Undercooling. If nucleation does
not occur, a pure metal can undercool under
the equilibrium temperature because of heat
extraction. The liquid is said to be thermally
undercooled by a quantity:
�TT Te T� (Eq 2)
where DTT is the thermal undercooling, Te is the
equilibrium (melting) temperature of the inter
face, and T* is the S/L interface temperature.
Constitutional (Solutal) Undercooling.
During alloy solidification, solute is rejected
by the solid. This can be understood from the
phase diagram in Fig. 8. For a given tempera
ture, T*, the composition of the solid, CS, is
smaller than that of the liquid, CL, in equilib
rium with the solid. The ratio k = CS/CL is
called the partition coefficient. For the case in
the figure (where the equilibrium temperatures
decrease with increased alloy composition),
knucleus interface energy.
The value of this energy depends on the crystal
structure of the two phases. The interface
between two crystals can be coherent, semico
herent, or incoherent.
Coherent interfaces may have slight devia
tions in the interatomic spacing, which causes
lattice deformation and induces a strain in
the lattice (Fig. 10). If the deviation in spacing
is too large to be accommodated by strain, dis
locations may form in distorted areas. The
interface is said to be semicoherent. If there
is no crystallographic matching between the
two lattices, the structure changes abruptly
from one crystal to the other; the interface is
incoherent.
An efficient heterogeneous nucleant (inocu
lant) should satisfy the following requirements:
� The substrate must be solid in the melt; its
melting point must be higher than the melt
temperature, and it must not dissolve in the
melt.
� There must be a low contact angle between
the metal and nucleant particles or a high
surface energy between the liquid and the
nucleant.
� The nucleant must expose a large area to the
liquid; this can be achieved by producing a
fine dispersion of nucleant or by using a
nucleant with a rough surface geometry.
� Because the atoms are attaching to the solid
lattice of the substrate, the closer the sub
strate lattice resembles that of the solid
phase, the easier nucleation will be. This
means that, ideally, the crystal structure of
the substrate and the solid phase should be
the same, and that their lattice parameters
should be similar (isomorphism). They
should have at least analogous crystalline
planes (epitaxy). Because the crystal struc
tures of the solidifying alloy and the sub
strate may be different, the substrate must
have one or more planes with atomic
spacing and distribution close to that of one
of the planes of the solid to be nucleated
(coherent or semicoherent interface), that
is, have a low linear disregistry, d (Ref 13):
d ðan aSÞ=aS (Eq 6)
where an and aS are the interatomic spacing
along shared low index crystal directions in
the nucleant and the solid nucleus, respectively.
� Low symmetry lattice (complex lattice) is
desirable. While it is impossible to assign
numbers to lattice symmetry, to some extent
the entropy of fusion can be used as a mea
sure of lattice symmetry. In general, less
symmetrical lattices have higher entropies
of fusion.
� It should have the ability to nucleate at very
low undercooling.
Inoculation and Grain Refining. The nucle
ation concepts introduced in the preceding
paragraphs are helpful in the understanding of
the widely used inoculation processes of cast
iron. Inoculation is often used in cast iron pro
cessing to control the grain and graphite size
and, to a lesser extent, graphite morphology.
Typical inoculants for cast iron are based on
ferrosilicon or calcium silicide. Inoculation
must not be confused with modification. Modi
fication, typically obtained through magnesium
additions to the melt, is a process related mostly
to graphite growth and morphology. The main
purpose of inoculation is to promote grain
refinement and avoid metastable solidification
(chill), while modification is used to change
the morphology of the eutectic aggregates.
Bramfitt (Ref 14) argued that the Turnbull/
Vonnegut equation for linear disregistry
(Eq 6) cannot be applied to crystallographic
combinations of two phases with planes of dif
fering atomic arrangements (e.g., cubic iron and
hexagonal tungsten carbide). He modified the
equation in terms of angular difference between
the crystallographic directions within the plane
to produce the planar disregistry equation:
d hklð ÞS
hklð Þn
X3
i 1
d
uvw½ �i
S
cosy
� �
�d
uvw½ �in
���
���
d
uvw½ �in
3
100 (Eq 7)
where (hkl)S is a low index plane of the sub
strate, [uvw]S is a low index direction in (hkl)S,
(hkl)n is a low index plane in the nucleated solid,Fig. 10 Coherent and semicoherent interfaces. Source: Ref 8
The Liquid State and Principles of Solidification of Cast Iron / 49
 
 
 
[uvw]n is a low index direction in (hkl)n, is the
interatomic spacing along [uvw]n, d½uvw�
S
is
the interatomic spacing along [uvw]S, and y
is the angle between [uvw]S and [uvw]n. The
effect of selected carbide and nitride additions
to pure iron (99.95%) were then evaluated.
Their effectiveness as nucleants was estimated
based on the effect of the solidification under
cooling. A good nucleant produced a lower
undercooling. The main results are listed in
Table 1 together with the planar disregistry
between the nucleant and iron. It is observed
that the highly effective inoculants have low
disregistry (continue to grow. On the contrary, if GT > GL,
the interface will remain planar (Fig. 13a).
For small constitutional undercooling, the
instabilities will only grow in the solidification
direction (the x direction), and a cellular inter
face will result (Fig. 13b, c). This is shown in
Fig. 14. The planar to cellular transition occurs
at a gradient Gp/c. As the constitutional under
cooling increases because of the lower thermal
gradient, the spacing between the cells
increases, and constitutional undercooling may
also occur perpendicular to the growth direction
(in the y direction). Instabilities will develop on
the sides of the cells, resulting in the formation
of dendrites (Fig. 13d). This is the cellular to
dendrite transition. It takes place at a tempera
ture gradient Gc/d. Both cellular and dendritic
growth occurring from the wall in the direction
opposite to the heat transport can be described
as columnar growth.
If constitutional undercooling is greater,
equiaxed grains can be nucleated in the liquid
away from the interface. The dendritic to
equiaxed transition occurs at Gd/e. If the ther
mal gradient is almost flat, that is, GT = 0, the
driving force for the columnar front will be
extremely small. A complete equiaxed structure
is expected.
Table 2 Typical compositions of inoculants
Inoculant
Composition, mass%
Si Al Ca Ba Sr Zr Mn Others
Standard FeSi 75–80 1.2–2 0.3–1.2 . . . . . . . . . . . . . . .
FeSi-Mn-Zr 60–65 1.2 1–3 . . . . . . 5.6 5.6 . . .
FeSi-Ba 60–65 0.5–1.7 1.0 9–11 . . . . . . . . . . . .
FeSi-Ba 60–65 1.5 2.0 5–6 . . . . . . 9–10 . . .
FeSi-Zr 80 1.5–2.5 2.5 . . . . . . 1.5 . . . . . .
FeSi-Sr 75 CS). Conse
quences of this phenomenon are the occurrence
of constitutional undercooling and segregation.
Constitutional undercooling is instrumental in
destabilizing the S/L interface and promoting
interface morphologies different than planar. As
inferred by Eq 8, there is a critical solute content
(Co) of the alloy for a given GT/V ratio combina
tion, at which the interface becomes unstable.
This can be presented graphically as shown in
Fig. 16, where the line for Eq 8 indicates the pla
nar to cellular transition. As the GT/V ratio con
tinues to decrease (or Co to increase), the S/L
interface becomes increasingly unstable with
successive formation of a columnar dendritic
and then equiaxed dendritic structure.
Fig. 12 Nucleation and coalescence of eutectic grains in cast iron. (a) Early solidification. (b) Late solidification. (c) After solidification (room temperature). Original magnification:
20�. Source: Ref 18
Fig. 13 Change of morphology of the solid/liquid (S/L) interface as a function of growth velocity (V) in a transparent organic system (pivalic acid, 0.076% ethanol) directionally
solidified under a thermal gradient of 2.98 K/mm. (a) Planar interface, V = 0.2 mm/s. (b) Cellular interface, V = 1.0 mm/s. (c) Cellular interface, V = 3.0 mm/s. (d) Dendritic
interface, V = 7 mm/s. Same scale for all images. Source: Ref 17
The Liquid State and Principles of Solidification of Cast Iron / 51
 
 
 
The formation of the equiaxed dendritic struc
ture requires bulk nucleation. In the absence of
bulk nucleation, the columnar front will continue
to grow.
Planar Interfaces. Planar growth of alloys
can usually be achieved only in crystal growth
furnaces at high temperature gradients and low
solidification velocities. For example, for planar
solidification of an alloy with DT = 5 K and
GT = 100 K/cm, the maximum allowable solid
ification velocity calculated with Eq 8 is 2 mm/s.
However, most commercial cast irons solidify
with nonplanar interfaces, because the solidifica
tion velocity is much higher.
Cellular Structures. When constitutional
undercooling occurs, the S/L interface morphol
ogy becomes cellular or dendritic. For condi
tions of growth where the GT/V ratio is only
slightly smaller than the ratio DT/DL, the
interface is cellular, as shown in Fig. 17(a) for
a hypoeutectic iron, after Tian and Stefanescu
(Ref 20).
Dendritic Structures. The dendritic mor
phology is the most commonly observed solidi
fication structure of solid solutions, including
austenite in steel and cast iron. Examples of
dendrites observed in directionally solidified
cast iron are presented in Fig. 17(b, c).
Effect of Crystallographic Orientation. Den
drites are single grains that have preferred
growth directions. The morphology of a colum
nar dendrite is influenced by the orientation of
the grain with respect to that of heat extraction,
as shown in Fig. 18, where the heat extraction
direction is upward (Ref 21).
Influence of the Type of Phase Diagram. The
nature of the material as represented by the type
of phase diagram will also influence the den
dritic structures. If the phase diagram shows
complete solid solubility (Fig. 15a), the struc
ture will be single phase, containing only den
drites. If, as is the case for cast iron, the phase
diagram contains a eutectic (Fig. 15b), the
interdendritic regions will be composed of the
two phase eutectic. Figure 19, from Aguado
et al. (Ref 22), presents a low magnification
microstructure of a hypoeutectic gray iron.
The microstructure exhibits a large number of
austenite dendrites with interdendritic austen
ite graphite eutectic.
Effect of Constitutional Undercooling. As
shown in Fig. 16, as the amount of solute
increases, or as the GT/V ratio decreases, a cel
lular to dendritic solidification occurs. This is
because the constitutional undercooling is
large. Such a transition is not common in cast
iron, because the solidification conditions are
conducive to mostly dendritic structures.
Figure 14 indicates that for rather steep thermal gradients, columnar dendrites will form,
while for shallow gradients, equiaxed dendrite
will solidify. In a continuously cooled casting,
the decrease in the GT/V ratio may produce a
columnar to equiaxed transition, as seen in
Fig. 4 for a gray iron bar.
Effect of Solidification Velocity. As empha
sized previously, solidification velocity is,
together with the temperature gradient, the most
important variable affecting microstructure
transitions. The change in solidification veloc
ity may determine a planar S/L interface to
become cellular and then dendritic. In addition,
the morphology of the equiaxed dendrites
(branching and tip radius) depends significantly
on the cooling rate and/or undercooling. The
effect of solidification velocity over a wide
range of velocities can be understood from
Fig. 20. At very small velocities, the dendrite
tip radius is very large, even infinity, in which
case a planar interface is obtained. As the
velocity increases, the radius decreases, and
the morphology changes from planar to globu
lar/cellular, then to regular equiaxed dendritic.
Further increase in solidification velocity in
the range of rapid solidification determines a
transition from fully branched to globular/cellu
lar dendrites and finally again to planar inter
face (absolute stability). A typical example
illustrating the influence of cooling rate on
the morphology of equiaxed dendrites of an
Al 7Si alloy is given in Fig. 21 (Ref 23).
Planar
X
T Gp/c
Gc/d Gd/e
TL
Cellular
Dendritic
Equiaxed
Fig. 14 Correlation between the thermal gradient at
the interface and the interface morphology.
Source: Ref 8
Fig. 15 Binary phase diagrams. (a) Complete solid solubility. (b) Partial solid solubility with eutectic reaction. L,
liquid solution; a and b, solid solutions
Fig. 16 Transition to different interface morphologies
as a function of the temperature gradient/
solidification velocity ratio (GT/V) and solute
concentration (Co)
52 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
The solidification time scale also influences
the secondary dendrite arm spacing (SDAS).
The SDAS is the distance between adjacent
branches growing from the main dendritic
arm. It is directly related to certain mechanical
properties. It is generally accepted that the
SDAS is a function of the local solidification
time, tf, described by:
SDAS mo t
1=3
f (Eq 9)
where mo is a material specific constant (coars
ening constant). Extensive experimental data
on secondary arm spacing have also been
reported to fit a SDAS cooling rate equation
(Ref 24):
SDAS m1 ð _TÞ�0:34	0:02
(Eq 10)
where m1 is a material specific constant, and _T
is the cooling rate.
Solute Redistribution and Microsegrega
tion in Dendritic Solidification. Rejection of
solute from the solid during solidification that is
responsible for the formation of the solutal
boundary layer (Fig. 9) produces compositional
nonuniformity across the dendrite during solidi
fication, called microsegregation. To understand
the mechanism of formation of microsegrega
tion, consider the volume element extending
from the axis of the dendrite arm to the edge of
the final dendrite (at the end of solidification)
shown in Fig. 22. The thick line in the lower part
of the figure represents the composition change
in the solid during solidification. At the begin
ning of solidification, when there is no solid
formed, the fraction solid is fS = 0. The first solid
to form will have the composition kCoat temperature TE and composition CE.
At this point, two solid phases, a and b, solidify
simultaneously from the liquid, L. The eutectic
reaction can be written as: L ! a + b. As many
as four phases have been observed to grow
simultaneously from the melt. However, most
technologically useful eutectic alloys consist
of two phases. The particular morphology of
the eutectic is a function of processing condi
tions and of the nature of the two phases.
Classification of Eutectics. Many eutectic
classifications have been proposed, based on
different criteria. A first classification of eutec
tics based on their growth mechanism is:
� Cooperative growth: The two phases of the
eutectic grow together as a diffusion couple.
� Divorced growth: The two phases of the
eutectic grow separately; there is no direct
exchange of solute between the two solid
phases and no trijunction.
Cooperative eutectics can be further classi
fied based on the ratio between the fractions
of the two phases of the eutectic, fa and fb,
and on the morphology of the S/L interface
(Ref 27), as shown in Fig. 23. The nondimen
sional entropy of fusion, DSf/R, where R is the
gas constant, is used to distinguish between fac
eted and nonfaceted morphologies.
Alloys such as lead tin and Al Al2Cu, where
there are approximately equal volume fractions
of nonfaceted phases, solidify as regular, lamel
lar eutectics. If one of the phases is nonfaceted,
the morphology becomes irregular, because the
faceted phase grows preferentially in a direc
tion determined by specific atomic planes. This
is the case of lamellar graphite iron, where aus
tenite is nonfaceted and graphite is faceted. In
this case, one solid phase may project into the
liquid far in advance of the other solid phase.
When the volume fraction of one phase is
significantly lower than that of the other (typi
callyamounts, crystallo
graphic factors, interfacial energies, impurity
content, and alloy composition. The lamellar
spacing, l, and the solidification velocity are
related by the simple equation l2V = constant.
The effect of solidification velocity is illustrated
in Fig. 25. It is seen that the spacing of irregular
eutectics is significantly larger than that of regu
lar eutectics.
The adjustment in the eutectic spacing during
growth occurs through faults. Two types of faults
are shown in Fig. 26. Figure 26(a) shows a no net
fault in which the number of lamellae on both
sides of the fault is the same. Figure 26(b) shows
a net fault in which one side of the fault has one
more lamellae than the other side. This fault is
analogous to an extended dislocation in that the
number of lamellae above and below the fault
differ by one (Ref 28).
For equiaxed eutectics, the length scale
may include grain size in addition to lamellar
spacing. Metallographic identification of the
grain size is alloy specific.
Solidification Structures of
Peritectics
Peritectic solidification is very common in the
solidification of metallic alloys. Many technically
important alloy systems, such as steels, copper
alloys, and rare earth permanent magnets, display
peritectic reactions in the regions of their phase
diagrams where phase and microstructure selec
tion play an important role for the processing
and the properties of the material. Basically,
peritectic solidificationmeans that at the peritectic
temperature, TP, a solid phase g of peritectic com
position, CP, solidifies from a mixture of liquid,
L, and solid phase d. The peritectic solidification
can be written as L + d ! g. A phase diagram
with peritectic solidification is presented in
Fig. 27. The different reactions occurring along
the solidus lines, corresponding to various compo
sitions, produce three structural regions: d + g, g,
and L + g.
Two different mechanisms are involved in
peritectic solidification: peritectic reaction and
peritectic transformation. These mechanisms
are shown in Fig. 28. In a peritectic reaction,
all three phases (d, g, and liquid) are in contact
with each other. In the peritectic transforma
tion, the liquid and the primary d phase are
isolated by the g phase. The transformation
takes place by long range diffusion through
the secondary g phase. A variety of microstruc
tures can result from peritectic solidification,
mostly depending on the GT/V ratio and nucle
ation conditions. The possible structures
include cellular, plane front, bands, and eutec
tic like structures.
Simultaneous growth of two phases in the
form of oriented fibers and lamellae has been
0.1
1
2
5
10
20
50
100
4 40
40
400
Eutectoids
Regular eutectics
Irregular eutectics 400
4000
4000
4 × 104
1
Solidification velocity (V ), mm/s
S
pa
ci
ng
 (
l)
, m
m
 
S
pa
ci
ng
 (
l)
, m
in
. 
Solidification velocity (V ), µin./s
10 102 103
Fig. 25 Comparisonof thelamellar spacing/solidification
velocity correlation for eutectics and eutectoids.
Source: Ref 8
Fig. 26 Cross sections of a directionally solidified lead-cadmium eutectic showing the presence of faults in the
lamellae. (a) No-net fault. (b) Net fault. Etchant not reported. Source: Ref 28
Two phase
0.08
1400
2802
2728
2550
1498
LL + δ
L + γ
γ
γ
TMγ
TM
δ = 1539
Te
am
pe
ra
tu
re
, °
C
Te
am
pe
ra
tu
re
, °
F
0.16 0.53
Carbon, %
Single phase
δ + γ
δ
Fig. 27 Schematic phase diagram of the peritectic region of carbon steel. Source: Ref 8
56 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
observed in some peritectic alloys when the
composition was on the tie line of the two solid
phases and the GT/V ratio was close to the limit
of constitutional undercooling for the stable
phase having the smaller distribution coefficient
(Ref 29). Figure 29 shows such a structure for
an iron nickel alloy.
Fluid flow can further complicate the possi
ble microstructures. While peritectic reactions
are typical for cast steel, they do not occur in
cast iron, because the carbon content is always
above the higher limit (0.53%) of the peritectic
solidus.
ACKNOWLEDGMENT
This article was adapted from Doru M. Stefa
nescu and Roxana Ruxanda, Fundamentals of
Solidification, Metallography and Microstruc
tures, Volume 9, ASM Handbook, American
Society for Metals, 1985, p 71 92.
REFERENCES
1. M.V. Volostchenko, On the State of Graph
ite in Liquid Iron, Liteinoe Proizvod.,
No. 2, 1976, p 5 7
2. S. Steeb and U. Maier, in The Metallurgy
of Cast Iron, B. Lux, I. Minkoff, and F.
Mollard, Ed., Georgi Publishing, St.
Saphorin, Switzerland, 1974, p 1 11
3. W. Krieger and H. Trenkler, Arch. Eisenh€ut
tenwes., Vol 42 (No. 3), 1971, p 175
4. L.S. Darken, “Equilibria in Liquid Iron
with Carbon and Silicon,” Tech. Pub.
1163, AIME Metals Technology, 1940, p 1
5. E. Schürmann, private communication
quoted in Ref 1
6. A.A. Vertman and A.M. Samarin, Dokl.
Akad. Nauk SSSR, Vol 134 (No. 3), 1960,
p 629
7. A.A. Vertman and A.M. Samarin, Liteinoe
Proizvod., No. 10, 1964
8. D.M. Stefanescu, Science and Engineering
of Casting Solidification, 3rd ed., Springer,
2015
9. R.E. Boeri and J.A. Sikora, Int. J. Cast
Met. Res., Vol 13 (No. 5), 2001, p 307
313
10. J.C. Moore, in Metallography, Structures
and Phase Diagrams, Vol 8, Metals Hand
book, 8th ed., American Society for Metals,
Metals Park, OH, 1973, p 93
11. G.R. Purdy and M. Audier, Electron Micro
scopical Observations of Graphite in Cast
Irons, The Physical Metallurgy of Cast
Iron, H. Fredriksson and M. Hillert, Ed.,
Materials Research Society Symposia Proc.
(Stockholm), North Holland, New York,
1985, p 13 23
12. D.M. Stefanescu, G. Alonso, P. Larrañaga,
and R. Suarez, On the Stable Eutectic Solidi
fication of Iron Carbon Silicon Alloys, Acta
Mater., Vol 103, 2016, p 103 114
13. D. Turnbull and R. Vonnegut, Ind. Eng.
Chem., Vol 44, 1952, p 1292
14. B. Bramfitt, Metall. Trans., Vol 1, 1970,
p 1987 1995
15. A.L. Greer, A.M. Bunn, A. Tronche, P.V.
Evans, and D.J. Bristow, Acta Mater.,
Vol 48, 2000, p 2823
16. R. Elliott, Cast Iron Technology, Butter
worths, London, 1988
17. R. Trivedi and W. Kurz, Solidification of
Single Phase Alloys, Casting, Vol 15,
ASM Handbook, D.M. Stefanescu, Ed.,
ASM International, 1988, p 114
18. H. Tian, Ph.D. dissertation, University of
Alabama, Tuscaloosa, 1992
19. H. Tian and D.M. Stefanescu, Experimen
tal Evaluation of Some Solidification
Kinetics Related Material Parameters
Required in Modeling of Solidification of
Fe C Si Alloys, Modeling of Casting,
Welding and Advanced Solidification Pro
cesses VI, T.S. Piwonka, V. Voller, and L.
Katgerman, Ed., TMS, Warrendale, PA,
1993, p 639
20. H. Tian and D.M. Stefanescu, Dendritic
Growth during Directional Solidification
of Hypoeutectic Fe C Si Alloys, Metall.
Trans. A, Vol 23, 1992, p 681 687
21. S. Akamatsu, G. Faivre, and T. Ihle, Phys.
Rev. E, Vol 51, 1995, p 4751 4773
22. E. Aguado, D.M. Stefanescu, J. Sertucha,
P. Larrañaga, and R. Suárez, Effect of
Carbon Equivalent and Alloying Elements
on the Tensile Properties of Superfine
Interdendritic Graphite Irons, Trans. AFS,
Vol 122, 2014, p 249 258
23. C.P. Hong and M.F. Zhu, in The Science of
Casting and Solidification, D.M. Stefanescu,
Fig. 29 Quenched solid/liquid interface of simultaneous two-phase growth in peritectic iron-nickel alloy. Source:
Ref 29
Fig. 28 Mechanisms of peritectic solidification. Source: Ref 8
The Liquid State and Principles of Solidification of Cast Iron / 57
 
 
 
R. Ruxanda, M. Tierean, and C. Serban,
Ed., Editura Lux Libris, Brasov, Romania,
2001, p 110 118
24. D. Bouchard and J.S. Kirkaldy, Metall.
Mater. Trans. B, Vol 28, 1997, p 651
25. W.J. Boettinger, Metall. Mater. Trans.,
Vol 5, 1974, p 2026
26. J.H. Perepezko and J.J. Paike, in Rapidly
Solidified Amorphous and Crystalline
Alloys, B.H. Kear, B.C. Giessen, and M.
Cohen, Ed., North Holland, 1982, p 49
27. W. Kurz and D.J. Fisher, Fundamentals of
Solidification, 3rd ed., Trans Tech Publica
tions, Switzerland, 198928. R. Trivedi, J.T. Mason, J.D. Verhoeven,
and W. Kurz, Metall. Mater. Trans. A,
Vol 22, 1991, p 2523 2533
29. M. Vandyoussefi, H.W. Kerr, and W.
Kurz, Acta Mater., Vol 48, 2000,
p 2297 2306
58 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
Microstructure Evolution during the
Liquid/Solid Transformation in Cast Iron
Doru M. Stefanescu, The Ohio State University and The University of Alabama
CAST IRON is a binary iron carbon or a
multicomponent Fe C X alloy that is rich in
carbon and exhibits a considerable amount of
eutectic in the solid state. Two such possible
eutectics may result, as follows:
� If solidification occurs according to the meta
stable diagram, Fe Fe3C, the white eutectic or
austenite iron carbide (Fe3C), forms.
� If solidification follows the stable diagram,
iron graphite, the gray eutectic or austenite
graphite, results.
Depending on composition, cooling rate, and
liquid treatment, it also is possible to produce a
mixed white gray structure called mottled
structure. The two basic types of eutectic are
very different, with mechanical properties such
as strength, ductility, and hardness varying over
very large intervals as a function of the type
and the amount of eutectic formed.
Thesolidificationofhypoeutecticcast ironstarts
with the nucleation and growth of austenite den
drites, while that of hypereutectic iron starts with
the crystallization of primary graphite in the stable
system or cementite in the metastable system.
Nucleation and Growth
of Austenite Dendrites
While the austenite (g) dendrites are the most
important solidification phase in the development
of mechanical properties (the strength of graphite
is, for practical purposes, nonexistent), most of
the structure mechanical properties correlation
developed for cast iron are based on the shape
and distribution of the graphite. This anomaly is
the consequence of research being biased toward
the study of graphite because of the difficulty in
outlining the austenite on the microstructure, as
the austenite undergoes a solid state transforma
tion during cooling, resulting in recrystallization
into pearlite and/or ferrite.
A clear understanding of solidification of
austenite in cast iron became possible only after
2001, when Boeri and Sikora (Ref 1) developed
the macroetching technique dubbed direct aus
tempering after solidification that allows visual
ization of the solidification austenite dendrite
grains at room temperature. As shown in
Fig. 1, columnar and equiaxed regions, and
even columnar to equiaxed transitions, can be
observed in both gray and ductile iron. In
lamellar graphite iron, as the carbon equivalent
increases, the amount of equiaxed grains in the
structure also increases. As discussed later,
these grains include not only the primary aus
tenite but also, in most instances, the eutectic
austenite.
Nucleation in cast alloys is heterogeneous,
that is, on nuclei having a different chemical
composition than the solidifying iron carbon
alloy. Predictions with the classic theory of het
erogeneous nucleation, which is an extension of
the steady state homogeneous nucleation the
ory, fail to match experimental data because
the mechanisms of the two types of nucleation
are different (Ref 2). Homogeneous nucleation
results from the stabilization of a transient
grouping of atoms, so that a nucleus consisting
of many atoms is formed all at once. In hetero
geneous nucleation, the atoms of the metal to
be nucleated attach themselves to the best loca
tions on the nucleant, and the nucleus grows
atom by atom.
Ideally, the crystal structure of the nucleus
and the solid phase should be the same, and
their lattice parameters should be similar (iso
morphism). At a minimum, they should have
analogous crystalline planes (epitaxy). The
atomic spacing and distribution in these planes
should be close to that of one of the planes of
the solid to be nucleated (coherent or semico
herent interface), that is, should have low lattice
disregistry. A linear disregistry can be calcu
lated as d = (an as)/as, where as and an are
the interatomic spacing along shared low index
crystal directions in the nucleating solid and the
nucleant, respectively. Bramfitt (Ref 3) modi
fied this equation to calculate a planar
disregistry:
d ðd1 þ d2 þ d3Þ=3 (Eq 1)
where d1, d2, and d3 are the linear disregistry
calculated along the three lowest index
ASM Handbook, Volume 1A, Cast Iron Science and Technology 
D.M. Stefanescu, editor
DOI: 10.31399/asm.hb.v01a.a0006304
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
Fig. 1 Macrostructure of cast iron bars showing primary austenite dendrites. Etched with direct austempering after solidification + 5% picral. CE, carbon equivalent.
(a) Hypoeutectic lamellar graphite (LG) iron (3.94 CE), 20 mm (0.8 in.) diam bar. (b) Eutectic LG iron (4.27 CE), 20 mm diam bar. (c) Hypereutectic LG iron (4.64 CE),
30 mm (1.2 in.) diam bar. (d) Spheroidal graphite iron, 30 mm diam bar. Courtesy of R. Boeri and G. Rivera
 
 
 
directions within a 90� quadrant of the planes
of the nucleus and the nucleated solid. From
experimental data on pure iron, it was con
cluded that a disregistry conducive to nucle
ation should be lower than approximately 10%.
Another critical parameter for heterogeneous
nucleation is the wettability of the nucleant.
The wetting problem can be solved practically
by the formation of an intermediate phase that
wets the nucleant. For a detailed discussion on
the subject of heterogeneous nucleation, the
reader is directed to Ref 4 and 5.
Nucleation of austenite in cast iron is still in
need of research. Titanium is well known to be
a deoxidizer and structure refiner in steel. The
literature data on the effect of titanium in cast
iron are rather inconsistent. It is claimed that tita
nium additions nucleate dendrites, favoring the
formation of small, equiaxed dendrites (Ref 6).
Alternatively, the number of austenite dendrites
can be increased by reducing the carbon equiva
lent, adding elements such as titanium and
boron, or by adding materials that serve as sub
strates for austenite nucleation (nitrides, carboni
trides, and carbides of various elements such as
titanium and vanadium) (Ref 7). Yet, TiC did
not appear to be a nucleation site for the primary
austenite in low sulfur irons, because it was
found mostly at the periphery of the secondary
arms of the austenite, in the last region to solid
ify (Ref 8). It also was stated that titanium addi
tions refine the secondary arm spacing in both
gray and ductile iron (Ref 9, 10).
Addition of titanium to cast iron melts
produces a number of titanium compounds,
including TiN (35 at.% N), (MnTi)S, and
TiC (Ref 11). Several theories try to explain
the increased number of austenite dendrites
achieved through titanium. Ruff and Wallace
(Ref 7) postulate that the effect of titanium is
related to the increased undercooling resulting
from reduced nucleation potential for graphite
or from restricted growth of the eutectic grain.
Okada and Miyake (Ref 12) suggest that
because titanium combines with carbon in the
melt to produce TiC, low carbon regions are
produced at the solid/liquid interface, favoring
formation of type D graphite. Wilford and
Wilson (Ref 13) stated that in irons with up to
0.4% Ti, first, titanium will react with nitrogen,
producing TiN or Ti(CN). The excess titanium
then will react with sulfur. Formation of TiS
decreases the available sulfur for MnS forma
tion and increases undercooling, which is
responsible for type D graphite formation.
Directional solidification experiments com
bined with a liquid metal decanting technique
revealed the evolution of dendrite tip radius
and spacing (Ref 14). The typical morphology
of dendrites in an Fe 3.08%C 2.01%Si alloy
presented in the scanning electron microscopy
(SEM) micrographs in Fig. 2 shows the parabo
loid shaped tip of the dendrite, which is consis
tent with observations in other systems. At a
growth velocity of 0.65 mm/s, a cellular to dendritic transition occurred.
Nucleation of Graphite
Experimental evidence indicates that various
types of nuclei become effective at various tem
peratures. Indeed, if only one type of nuclei
would be active in a cast iron melt, an undercool
ing superheating curve will show a single arrest
for the temperature region over which the nuclei
become effective. However, as shown in Fig. 3,
a number of steps are observed in the relationship,
suggesting that various foreign nuclei become
effective as superheating is increased (Ref 15).
Increasing the superheating apparently destroys
the effective nuclei. Consequently, the number
of eutectic grains (eutectic cells) decreases as
the superheating increases, and, as a result, under
cooling increases. However, when the undercool
ing is increased at constant superheat by
increasing the cooling rate or the growth velocity,
the eutectic cell count will increase, as shown in
Fig. 4 (Ref 16).
The analysis of the vast literature on this sub
ject reveals that lamellar graphite (LG) and
spheroidal graphite (SG) nucleate on a variety
of substrates, and that the chemistry of these
substrates is quite different for LG as compared
with that of SG iron. Thus, nucleation of graph
ite is discussed separately for the two irons.
The heterogeneous nucleation theory devel
oped over the last 30 years is focused on the non
metallic inclusions present in all commercial cast
irons, such as oxides, nitrides, sulfides, and sili
cates, to list a few. To act as possible nucleation
sites, the inclusions must satisfy some specific
conditions, such as good crystallographic com
patibility, low lattice disregistry or mismatch, fine
dispersion in the melt (1 to 3 mm), and high sta
bility at elevated temperatures (Ref 17). Theories
advocating one stage, two stage, or multistage
nucleation have been offered.
Nucleation of Lamellar Graphite
Nucleation of LG can occur on carbon rich
regions in the liquid, such as carbon molecules
or undissolved graphite. Indeed, a large body
of inoculation experiments indicates that graph
ite is a potent nucleant in LG iron. Direct
experimental evidence is missing because, even
with electron microscopy, it is not easy to dis
tinguish the graphite nucleus from the graphite
that has grown on it. While these nuclei are
not homogeneous nuclei in the classical sense
of the term because they are postulated to pre
exist in the melt, they are of the same nature
with the graphite phase growing on them.
Fig. 2 Interface morphology at decanted solid/liquid interface in an Fe-3.08%C-2.01%Si alloy (G = 50 K/cm).
(a) Paraboloid-shaped austenite dendrite tip. (b) Austenite cell. Source: Ref 14
2001000
0
18
36
54
72
90
108
0
10
20
30
40
50
60
32 212 392 572 752 932
300 400 500
Superheating, °C
Superheating, °F
U
nd
er
co
ol
in
g 
(∆
T
),
 °
C
U
nd
er
co
ol
in
g 
(∆
T
),
 °
F
Fig. 3 Relationship between superheating and
maximum undercooling in lamellar graphite
cast iron. Source: Ref 15
Undercooled by
rapid freezing
Undercooled by
superheating
Undercooling (∆T ), °C
Undercooling (∆T ), °F
0
0
0 18 36 54 72
0.4
0.8
1.2
1.6
2.0
0
10
20
30
40
50
10 20 30 40 50
90
E
ut
ec
tic
 c
el
l c
ou
nt
 p
er
 m
m
E
ut
ec
tic
 c
el
l c
ou
nt
 p
er
 in
.
Fig. 4 Influence of undercooling on the number of
eutectic grains (cell count) in lamellar graphite
cast iron. Source: Ref 16
60 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
X ray and neutron wide angle diffraction on
molten iron carbon alloys brought supporting
evidence of short range order regions in melts
with higher than 3.5% C (Ref 18). The authors
argued that they are carbon clusters containing
approximately 15 atoms. At carbon contents
consistent with the short range order, the melt
exhibited increased viscosity (Ref 19). These
and other experiments (Ref 20 22) indicate that
either Cn or (Fe3C)n clusters exist in dynamic
equilibrium in molten iron carbon alloys. These
carbon rich clusters (or molecules) may serve
as homogeneous nuclei for graphite. Small
sized crystalline graphite already present in
the melt (undissolved graphite) also could serve
as nuclei for graphite (Ref 23).
Eash (Ref 24) presented the idea of silicon
rich regions around the dissolving graphite par
ticles when the melt was treated with silicon
base inoculants, which could promote the pre
cipitation of graphite. However, Feest et al.
(Ref 25) argued against this supposition,
because the dissolution time of ferrosilicon in
liquid iron is only a few seconds, and graphite
forms at the interface between dissolving parti
cles and liquid iron. They suggested that the
seed crystals can be preserved in the melt only
if barium or strontium is present in sufficient
amounts to prevent the graphite from dissolving
back into the melt.
Following the dissolution of ferrosilicon in
liquid cast iron, Fredriksson and coworkers
(Ref 26, 27) observed that SiC crystals and
graphite particles were formed in the melt
close to the dissolving ferrosilicon particles.
Assuming that the local supersaturation of
carbon and silicon in the melt, subsequent to
the SiC dissolution, provides the necessary
driving force for homogeneous nucleation of
graphite, a theory was developed and calcula
tions were performed to explain the nucle
ation of graphite and the fading mechanism
of these particles. The fading effect was
explained by the homogenization of carbon
and silicon in the melt through convection
and diffusion.
Recently, Stefanescu et al. (Ref 28) demon
strated that, in low sulfur gray irons, graphite
nucleates at the austenite/liquid interface with
out the presence of any foreign inclusions. This
supports the nucleation on carbon rich clusters
theory, which is a one stage nucleation model.
Another example of the one stage nucleation
model is the saltlike carbides nucleation theory
advanced by Lux (Ref 29), based on experi
ments that found that pure metals such as lith
ium, calcium, barium (Ref 30), strontium, and
sodium (Ref 31) promote graphite nucleation
in LG iron. Lux suggested that these and all ele
ments from groups I, II, and III from the peri
odic table, when introduced in molten iron,
form saltlike carbides that develop epitaxial
planes with the graphite and thus constitute
nuclei for graphite. Yet, because all these
metals are strong oxide and sulfide formers,
development of carbides rather than of the more
stable oxides or sulfides in the melt is
questionable.
However, other carbides, such as TiC or Ti
(CN), have been demonstrated to act as nuclei
for LG (Ref 28, 32), although in a rather limited
manner. An SEM micrograph illustrating this
nucleation is provided in Fig. 5.
The first multistage nucleation mechanism
seems to have been proposed by Weis (Ref 33),
who argued that nucleation of LG occurs on
SiO2 oxides formed by heterogeneous catalysis
on CaO, Al2O3, and oxides of other alkaline
metals. Thermodynamic calculations (Ref 34)
lead to the conclusion that while homogeneous
nucleation of silica is improbable, the cristoba
lite variety (tetragonal) lends itself to heteroge
neous growth of graphite, because there is only
3% incoherency between the longer tetragonal
axis of cristobalite (0.69 nm) and the c axis of
the graphite (0.67 nm). Homogeneous nucle
ation of CaO, Al2O3, and ZrO2 is highly proba
ble. Silicates, and in particular the hexagonal
2CaO�SiO2 with a lattice size of 0.72 nm, offer
better sites for graphite nucleation. It was con
cluded that the most effective common inocu
lating agent for LG is calcium, which operates
primarily via oxide embryo formation.
After Wallace (Ref 10) revealed the role of
MnS in graphite nucleation, a consensus was
reached that graphite lamellae nucleate on
MnS or complex (MnX)S compounds that
have low crystallographic misfit with the
graphite (Ref 17, 35, 36). Based on these and
similar findings, Riposan et al. (Ref 37) sug
gested that LG nucleation occurs on complex
(MnX)S sulfides, which in turn grow on com
plex oxides of aluminum, silicon, zirconium,and The University of Alabama . . . . . . . . . . . . . . . 12
Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Principles of the Metallurgy of Cast Iron. . . . . . . . . . . . . . . . 15
Gray Iron (Flake or Lamellar Graphite Iron) . . . . . . . . . . . . . 18
Ductile Iron (Spheroidal Graphite Iron) . . . . . . . . . . . . . . . . . 21
Compacted (Vermicular) Graphite Irons . . . . . . . . . . . . . . . . 22
Malleable Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Special Cast Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Fundamentals of the Metallurgy of Cast Iron. . . . . . . . . . . . . . . 29
Thermodynamics Principles as Applied to Cast Iron
Doru M. Stefanescu, The Ohio State University and
The University of Alabama
Jacques Lacaze, Université de Toulouse . . . . . . . . . . . . . . . . . . 31
Thermodynamics of Binary Fe X Systems . . . . . . . . . . . . . . . 31
Thermodynamics of Ternary Fe C X Systems . . . . . . . . . . . . 35
Thermodynamics of Multicomponent
Iron Carbon Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Composition Control of Iron Carbon Melts . . . . . . . . . . . . . . 42
The Liquid/Solid Transformation (Solidification)
The Liquid State and Principles of Solidification of Cast Iron
Doru M. Stefanescu, The Ohio State University and
The University of Alabama
Roxana Ruxanda, Emerson Climate Technologies. . . . . . . . . . . . 46
Fundamentals of Solidification of Cast Iron . . . . . . . . . . . . . . 46
Length Scale of Solidification Structures . . . . . . . . . . . . . . . . 46
Undercooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Nucleation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Growth and Interface Stability . . . . . . . . . . . . . . . . . . . . . . . 50
Solidification Structures of Solid Solutions . . . . . . . . . . . . . . 51
Solidification Structures of Eutectics . . . . . . . . . . . . . . . . . . . 54
Solidification Structures of Peritectics . . . . . . . . . . . . . . . . . . 56
Microstructure Evolution during the Liquid/Solid
Transformation in Cast Iron
Doru M. Stefanescu, The Ohio State University and The
University of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Nucleation and Growth of Austenite Dendrites. . . . . . . . . . . . 59
Nucleation of Graphite . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Nucleation of Austenite Iron Carbide Eutectic . . . . . . . . . . . . 63
Growth of Graphite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Eutectic Solidification of Cast Iron . . . . . . . . . . . . . . . . . . . . 70
Principles of Thermal Analysis
Hasse, Fredriksson, KTH Stockholm
Doru M. Stefanescu, The Ohio State University and
The University of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Basics of Cooling Curves . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Solidification Temperature and Chemical Composition . . . . . . 82
The Gray to White Transition . . . . . . . . . . . . . . . . . . . . . . . 83
Cooling Curves and Graphite Shape . . . . . . . . . . . . . . . . . . . 85
Nonequilibrium Solidification. . . . . . . . . . . . . . . . . . . . . . . . 86
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Volumetric Changes during the Solidification of Cast Iron
Attila Diószegi and Peter Svidró, J€onk€oping
University, Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Methods to Measure Volume Changes . . . . . . . . . . . . . . . . . 88
Direct Measurements of Volume Changes . . . . . . . . . . . . . . . 88
Indirect Measurement of Volume Changes . . . . . . . . . . . . . . 89
Dilatometer Measurements. . . . . . . . . . . . . . . . . . . . . . . . . . 89
Problems Associated with Volume Change
Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Anisotropic Displacement . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Kinetic of Graphite Expansion . . . . . . . . . . . . . . . . . . . . . . . 92
Computational Models for Prediction of Solidification Microstructure
A.V. Catalina, Caterpillar Inc., USA
A.A. Burbelko and W. Kapturkiewicz, AGH University of Science
and Technology, Krakow, Poland
M. Zhu, School of Materials Science and Engineering, Southeast
University, China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Macroscopic Transport Equations . . . . . . . . . . . . . . . . . . . . . 94
Analytical Microscopic Models for Solidification . . . . . . . . . . 95
Macro Microscopic Modeling of Cast Iron Solidification
Microstructure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Cellular Automaton Modeling . . . . . . . . . . . . . . . . . . . . . . . 101
The Solid/Solid Transformation
The Austenite to Pearlite/Ferrite Transformation
Jacques Lacaze, Université de Toulouse . . . . . . . . . . . . . . . . . 106
Stable and Metastable Three Phase Fields . . . . . . . . . . . . . . . 106
The Eutectoid Austenite Decomposition . . . . . . . . . . . . . . . . 107
Austenite Decomposition to Ferrite and Pearlite in
Spheroidal Graphite Irons. . . . . . . . . . . . . . . . . . . . . . . . . 108
Austenite Decomposition to Ferrite and Pearlite in Lamellar
and Compact Graphite Irons . . . . . . . . . . . . . . . . . . . . . . . 110
Modelling Austenite Decomposition to Ferrite
and Pearlite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
The Austenite to Ausferrite Transformation
Robert Boeri, UNMdP INTEMA . . . . . . . . . . . . . . . . . . . . . . . 114
General Features of the Decomposition of
Austenite into Bainite . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Heat Treatment Cycle and Microstructure . . . . . . . . . . . . . . . 116
Factors Affecting the Transformation of Austenite during
Austempering of Free Graphite Cast Irons . . . . . . . . . . . . . 117
ASM Handbook, Volume 1A, Cast Iron Science and Technology
D.M. Stefanescu, editor
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
xi
Processing of Cast Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Liquid Metal Preparation
Cast Iron Melting Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Cupola Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Refractory Linings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Water Cooled Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Emission Control Systems . . . . . . . . . . . . . . . . . . . . . . . . 125
Cupola Control Principles . . . . . . . . . . . . . . . . . . . . . . . . 125
Specialized Cupolas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Electric Melting Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Holding Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Electric Arc Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Induction Crucible Furnaces. . . . . . . . . . . . . . . . . . . . . . . . . 131
Coil and Transformer Yokes. . . . . . . . . . . . . . . . . . . . . . . 133
Refractory Linings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Refractory Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Refractory Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Crucible Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Duplex Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Pouring Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Pressure Actuated Pouring Furnaces . . . . . . . . . . . . . . . . . 140
Pouring Magnesium Treated Melts . . . . . . . . . . . . . . . . . . 142
Other Pouring Technologies . . . . . . . . . . . . . . . . . . . . . . . 143
Cast Iron Melt Quality Control. . . . .magnesium, and titanium. An illustration of
this mechanism from more recent work is
provided in Fig. 6.
Fig. 5 Lamellar graphite (Gr) growing on a cuboidal
TiC. Reprinted with permission from Elsevier.
Source: Ref 28
Fig. 6 Graphite nucleated on a MnS sulfide, which in turn nucleated on an aluminum, magnesium, silicon, calcium oxide. Source: Ref 8
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 61
 
 
 
Nucleation of Spheroidal Graphite
Nucleation of SG is even more complex than
that of LG. Review papers have been produced
periodically (Ref 38, 39). An early theory was
the gas bubbles theory, which postulates that
the tiny bubbles in the liquid metal, created by
gas evolution, are ideal sites for nuclei that
will give rise to growth of graphite nodules
(Ref 40). The graphite grows radially from the
outside into the bubble, as seen in Fig. 7. This
theory relies on the presence of carbon monox
ide bubbles. However, in industrial ductile iron
heats, the addition of magnesium and lantha
nides removes the oxygen, whether dissolved
or as CO gas. Furthermore, it is highly unlikely
that a complete graphite nodule will extend into
the entire volume of a gas bubble, because this
eventually would have to involve carbon diffu
sion through a graphite shell.
As early as 1966, Warrick (Ref 41) suggested
that nuclei for LG and SG are composed
of complex oxides and sulfides. Several investi
gators (Ref 32, 42 44) have suggested that the
sulfides, oxides, or nitrides, which are formed
after the addition of inoculants in Fe C Si
alloys, can act as nucleation sites during the
solidification of graphite. They concluded that
the majority of graphite spheroids are asso
ciated with nonmetallic inclusions, mainly,
magnesium calcium sulfides. Also, Skaland
(Ref 45) reported that oxysulfide particles will
have at least one lattice spacing that could
match the graphite lattice spacing and creates
the possibility of a favored substrate for graph
ite growth.
A two stage nucleation theory of double lay
ered (cored) nucleation was proposed by Jacobs
et al. (Ref 46) for SG in 1974. Using the results
of SEM analysis, they contended that SG nucle
ates on duplex sulfide oxide inclusions (1mm
diam); the core is made of calcium magnesium
or Ca Mg Sr sulfides, while the outer shell is
made of complex Mg Al Si Ti oxides. The ori
entation relationships were established as fol
lows. For the nucleus core/nucleus shell:
ð110Þsulfidejjð111Þoxide and ½1010�sulfidejj
ð211Þoxide. For the nucleus shell/graphite:
ð111Þoxidejjð0001Þgraphite and 111d eoxidejj
1010d egraphite. The x ray diffraction data
showed that the first few graphite layers, adja
cent to the (111) oxide, had a dilated lattice
(0.264 nm, instead of 0.246 nm). It was sug
gested that the spacing within the graphite
layers decreases away from the oxide until
unconstrained spacing is reached. Dislocations
were observed frequently in the matrix, and it
was suggested that these were generated to
relieve some of the elastic strain in the graphite
layers adjacent to the oxide.
This idea was further developed by Skaland
et al. (Ref 17), who argued that SG nuclei are
sulfides (MgS, CaS) covered by magnesium
silicates (e.g., MgO�SiO2) or oxides that have
low potency (large disregistry). After inocula
tion with FeSi that contains another metal
(Me) such as aluminum, calcium, strontium, or
barium, hexagonal silicates (MeO�SiO2 or
MeO�Al2O3�2SiO2) form at the surface of the
oxides, with coherent/semicoherent low energy
interfaces between substrate and graphite.
Igarashi et al. (Ref 47) also found examples
of two stage nucleation: a core of CaO, MgO,
Al2O3, or MgO/MgS, enveloped by a complex
nitride (MgSiAl)N, surrounded by a graphite
spheroid.
Secondary ion mass spectroscopy analysis on
duplex graphite nodules yielded interesting
results for both nucleation and growth of graph
ite (Ref 48). A duplex nodule (Fig. 8a) is made
of a core resulting from graphite precipitation
from the liquid, surrounded by a graphite shell
produced by carbon diffusion through a solid
austenite envelope. For an iron treated with
cerium mischmetal, a high level of lanthanum
and cerium was detected in the middle in the
core part of the duplex nodule (Fig. 8b). Their
rather uniform distribution across the core
implies no nucleation effect but continuous
incorporation in the graphite growing in the liq
uid. Calcium appears segregated at the periph
ery of the graphite core.
In the iron treated with magnesium and tita
nium, which exhibited 40% compacted graph
ite, titanium concentration is seen in the
position of the nucleus, suggesting a titanium
base inclusion. Cerium again was found uni
formly distributed throughout the core, but not
in the graphite shell, and again distributed in
the matrix. Magnesium, calcium, and titanium
exhibited a peak outside of the nodule, indicat
ing the presence of an inclusion. Other than
that, magnesium was uniformly distributed
throughout the graphite and the matrix.
A
B
C
D
Fig. 7 Karsay’s gas bubble theory. A, gas bubble; B,
graphite; C, melt; D, austenite: Source: Ref 40
Duplex
Inner part
Ca
C
ou
nt
s
C
ou
nt
s
Ca
30 mm
30 mm
Ce
Fe
La
C
nodule
(a)
(b)
(c)
Growth through
the g shell 
Duplex
Inner part
Ca
Ca
Ce
C
Mg
Mg
Fe
Ti
Ti
Ti
nodule
Nucleus
Growth in
the melt
Fig. 8 Secondary ion mass spectroscopy step-scans
across duplex graphite nodules. (a) Duplex
graphite nodule. Reprinted with permission from
Cambridge University Press. (b) 80% nodularity iron
produced through the addition of 0.19% Ce-mischmetal
(0.081% Ce). (c) 60% nodularity iron produced with
0.13% Mg and 0.1% Ti additions (0.021% Mg, 0.11%
Ti). Source: Ref 48
62 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
Recently, Alonso et al. (Ref 49) investigated
the chemistry of nuclei in hypo and hyper
eutectic SG iron samples obtained through
interrupted solidification. The irons were posti
noculated with an FeSi base inoculant contain
ing approximately1% Al, 1.8% Ca, 6% Mn,
0.13% Ti, and 6.8% Zr. SEM generated spec
tra, mapping, and line scans were used to detect
and analyze the nuclei. The x ray composition
maps in Fig. 9 show two different inclusions
in the graphite core: a complex sulfide (quasi
homogeneous distribution of sulfur and magne
sium and, to a lesser extent, of calcium) and a
carbonitride (similar distribution of nitrogen,
titanium, and zirconium). Previous investiga
tions have shown that titanium has a very high
affinity to oxygen and sulfur, but in the samples
analyzed in this work, it appeared to be present
much more as a carbide than a sulfide or an
oxide.
In many instances, the nucleus was made of
two or three different compounds, and all of
them were in contact with the graphite, as seen
in the examples in Fig. 10. The MgS and TiC
compounds were the major nucleation sites for
SG. Theories inferring that graphite spheroids
nucleate on nonmetallic inclusions that contain
a MgS core surrounded by an oxide shell, or a
shell of complex magnesium silicates, did not
explain many of the findings in their work.
Free energy of formation of the various
inclusions is of paramount importance in
establishing their probability of formation.
Selected free energies of formation of the sig
nificant compounds, calculated with the ther
modynamics software FactSage, are presented
in Table 1. Titanium nitrides and carbonitrides
or titanium, zirconium carbonitrides were not
found in the database of the software. It is seen
that the most probable is the complex magne
sium, calcium, aluminum oxide. The silicates
and double oxides also have high probability
of formation.
Nucleation of Austenite-Iron
Carbide Eutectic
The iron carbide (cementite) is based on an
orthorhombic unit cell with 12 iron atoms and 4
carbon atoms per cell and therefore has a carbon
content of 6.7 mass%. Its density is 7600 kg/m3.
Very little information is available on the
nature of the nuclei of the white eutectic. Nev
ertheless, it is accepted thatthe nucleation of
Mg
CaC
Ti
N
S Zr
Fig. 9 Energy-dispersive x-ray analysis composition maps in a graphite spheroid.
Source: Ref 49 Fig. 10 Inclusions found in the center of graphite (Gr) spheroids. Source: Ref 49
Table 1 Free energy of formation of possible compounds in spheroidal graphite nuclei
Compound DG, J/mol Compound DG, J/mol Compound DG, J/mol
Complex oxides Oxides Nitrides
2MgO�2CaO�14Al2O3 0.62 � 107 Al2O3 0.21 � 106 AlN 0.38 � 105
5CaO�4TiO2 0.47 � 106 Ti2O3 0.1 2 � 106 Ca3N2 0.39 � 105
3CaO�Al2O3�3SiO2 0.85 � 106 Fe3O4 0.49 � 105 Mg3N2 0.18 � 105
2MgO�CaO�2SiO2 0.98 � 106 SiO2 0.49 � 105
2CaO�FeO�SiO2 0.70 � 106 MgO 0.85 � 105
Double oxides CaO 0.78 � 105
3CaO�2SiO2 0.96 � 106 Sulfides Carbides
Al2O3�SiO2 0.86 � 106 Fe9S10 0.06 � 105 TiC 0.63 � 105
2MgO�SiO2 0.57 � 106 Ti2S3 0.38 � 105 Al4C3 0.23 � 105
2FeO�SiO2 0.01 � 106 ZrS3 0.11 � 105
5CaO�4TiO2 0.47 � 106 CaS 0.25 � 105 Carbonitrides
MgO�2TiO2 0.82 � 106 MgS 0.94 � 105 CaCN2 0.39 � 105
FeO�2TiO2 0.55 � 106 FeS 0.07 � 105
Al2O3�SiO2 0.86 � 106
2TiO2�ZrO2 0.19 � 106
Source: Ref 49
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 63
 
 
 
Fe3C occurs at lower undercooling than that of
graphite. Cooling curve data show that the
solidification of white iron begins at lower tem
perature than that of gray iron. There is some
evidence that silicon dioxide and aluminum
oxide can serve as substrates for the growth of
the Fe Fe3C eutectic, and that the nature of
the substrate can influence the morphology of
the eutectic (Ref 50).
Growth of Graphite
A large number of reviews on the subject are
available. This article notes the ones in Ref 5,
38, and 51 to 53.
Graphite Morphologies in Cast Iron
The three main graphite morphologies crystal
lizing from the iron melts during solidification
are lamellar (LG), compacted or vermicular
(CG), and spheroidal (SG). Examples are
provided in Fig. 11, after Ref 54 and 55. The
internal structure of LG shown in Fig. 12 exhi
bits parallel graphite layers and a large number
of defects between the graphite sheets. The inter
nal structure of SG is quite different. As shown
in Fig. 13(a), it exhibits conical sectors of paral
lel graphite planes extending radially from the
center. In imperfect SG shapes, such as exploded
graphite, the sectors are partially broken
(Fig. 13b). In some cases, the annular rings show
zigzag steps of the (0001) planes, suggesting
columnar crystals of graphite with different
orientations (Ref 56).
Transmission electron microscopy (TEM)
observations reveal that a graphite lamella is
made of a large number of thin plates and
incorporates numerous defects (Fig. 12). Higher
magnification unveils that the lamellae are
imperfectly crystalline on a local scale and
may contain amorphous regions (Ref 58).
Recent TEM work on graphite spheroids
(Ref 59) found a microcrystalline structure at
the center of the spheroid (small areas with
different orientations), while another TEM
report (Ref 60) identified an amorphous central
region surrounded by annular rings of a layered
intermediate region and then an outer region
made up of large polygonal crystalline platelets
in a mosaic like structure.
Many times, but not always, the graphite
spheroids may exhibit radial sectors joining in
the center (Fig. 13a). The angle between the
[0001] directions of graphite in the two adjacent
radial sectors has been found to have a range of
values (Fig. 14). The misfit at the joining of radial
Fig. 11 Typical graphite shapes obtained from commercial cast iron through deep etching and extraction. (a)
Lamellar graphite, lettuce growth. Source: Ref 54. (b) Compacted graphite, cauliflower growth. Source:
Ref 55. (c) Spheroidal graphite, cabbage growth
Fig. 12 Transmission electron micrograph of a graphite lamella exhibiting the pattern of a layered crystal with iron
entrapped between the layers. Source: Ref 54
Fig. 13 Scanning electron microscopy images of spheroidal graphite showing conical sectors and graphite nuclei. (a) Well-formed graphite spheroid. Reprinted with permission
from The American Foundry Society. Source: Ref 57. (b) Graphite spheroid with separated sectors. Reprinted with permission from The American Foundry Society. Source:
Ref 57. (c) Conical sectors with zigzag steps. Original magnification: 2500�. Reprinted with permission from The Japan Institute of Metals and Materials. Source: Ref 56
64 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
sectors precludes the continuity of graphite plates
across the joining (Ref 61).
A number of intermediate graphite shapes exist
between the standard coarse lamellar (type A) or
interdendritic lamellar (types D or E) and sphe
roidal graphite. They include superfine inter
dendritic graphite (SIG), coral graphite,
compacted graphite, and a number of irregular
graphite spheroids (chunky, exploded). The
SIG is short (10 to 20 mm) and stubby,
with round ends, and is obtained in low sulfur
and moderate titanium content irons (e.g.,
preferred
growth habit for graphite is in the a direction,
producing a sheet. Complete destruction of the
graphite structure occurs only at approximately
4000 �C (7230 �F). This explains the presence
of some graphite aggregates in molten iron
even at temperatures considerably higher than
the liquidus temperature.
The growth of a graphite crystal starts with
the formation of 2 D graphene sheets that can
grow in the a direction. To produce a graphite
platelet (a multilayer sheet that is the building
block of the graphite aggregate), some growth
in the c direction is required. It generally is
agreed that thickening of the platelets occurs
through spiral growth at screw dislocation steps
or by 2 D nucleation of the sheet in the c direc
tion. There is microscopy evidence for both.
The graphite plates exhibit two types of defects:
twin boundaries, which tilt the flake out of the
basal plane, and twist boundaries (stacking faults)
that lie on the basal planes (Fig. 17a, b). Twin
boundary defects may produce graphite branching
through splitting along its basal plane while grow
ing in the a direction. Twist boundaries cause a
rotation about the c axis of the graphite. Thus,
graphite lamellae are composed of layers of fault
free crystal, some 10�4 mm (4� 10�6 in.) thick.
Effect of Impurities in the Melt
The numerous impurities affecting graphite
growth can be divided into two categories:
� Reactive impurities favoring the transition
from LG to SG, such as magnesium,
calcium, yttrium, and lanthanides (cerium,
lanthanum). They typically are termed
compacting or spheroidizing.
� Surface active impurities favoring the transi
tion from SG to LG, such as sulfur, oxygen,
aluminum, titanium, arsenic, bismuth, tellu
rium, lead, and antimony. They are known
as anticompacting or antispheroidizing.
The solidification graphite shape is affected
by the reactive and surface active impurities
on the defect growth mechanism of graphite
and by the cooling rate of the alloy. Both influ
ence the constitutional undercooling of the
melt. Higher cooling rates favor the LG to
CG to SG transition.
All elements decrease the surface energy
of liquid iron carbon alloys. However, while
nickel, copper, and silicon slightly reduce the
surface energy, calcium, magnesium, cerium,
sulfur, selenium, and tellurium have a much
stronger effect.
There are extensive chemical reactions
between the impurities in the two categories.
Magnesium reacts with both oxygen and sulfur
as described to a certain extent in the article
“Thermodynamics Principles as Applied to
Cast Iron” in this Volume. Oxygen activity
decreases with higher magnesium and lower
temperature. Decreasing oxygen favors the
LG to SG transition.
Reactive impurities, such as magnesium and
cerium, remove surface active impurities gener
ating high surface energy in the melt. The effect
of magnesium on the surface tension of iron is
presented in Fig. 18, after Ref 74. It is seen that
the maximum surface energy is reached at a
magnesium level lower than that required for
SG formation, and that the effect of magnesium
decreases over time (fading).
According to McSwain and Bates (Ref 75),
there is a difference between the surface
energy of the melt and the liquid/graphite
interface energy. The data summarized in
Table 2 show that in iron magnesium alloys
the iron/graphiteprism interface energy is
higher than the iron/graphitebasal energy. Con
sequently, it was concluded that graphite
grows from the melt normal to the plane with
the lowest interfacial energy, which is the
c direction for a Fe C Mg alloy and the a direc
tion for a Fe C S alloy.
Auger analyses of sulfur containing irons
show concentrations of oxygen and sulfur in
iron, adjacent to the iron/LG interfaces, but not
in the graphite (Ref 76). Type A lamellae were
covered with a monolayer of sulfur with patches
of oxygen in the form of iron oxide having a
thickness of approximately 2 nm (Ref 77).
In magnesium treated iron, neither magne
sium nor oxygen or sulfur were detected on
the graphite surface but appeared isolated in
combined form as Mg S P compounds. This
seems to imply that magnesium does not act
directly on the graphite, but rather that it acts
as a scavenger of the impurities that stabilize
LG. However, because good SG cannot be
obtained by simply reducing the sulfur and oxy
gen content to nil, and because magnesium
containing irons produce good graphite
0.00 0.01 0.02 0.03 0.04
1300
1350
1400
1450
1500
T~1390 °C
Maganesium content, %
S
ur
fa
ce
 te
ns
io
n,
 d
yn
e/
cm
S
ur
fa
ce
 te
ns
io
n,
 d
yn
e/
cm
(a)
4 8 12 16 20 24 28
900
1000
1100
1200
1300
Holding time, min
(b)
Fig. 18 Effect of (a) magnesium content and (b) holding time on the surface tension of cast iron. Source: Ref 74
Fig. 17 Scanning electron micrographs of defects in graphite. (a) Twinning of plates. (b) Twist boundaries. Source:
Ref 73
Table 2 Surface properties of Fe C Si alloys
on graphite in the absence of oxygen
Alloy Graphite
Surface
tension of
iron, J/m
Iron/graphite
interfacial
energy(a), J/m
Fe-3.7C-
2.8Si-
0.037Mg
Basal 1.128 1.460
Polycrystalline 1.167 1.621
Prism 1.147 1.721
Fe-3.7C-
2.4Si-
0.05S
Basal 1.057 1.270
Polycrystalline 1.017 951
Prism 1.153 846
(a) Calculated from contact angle, surface energy of graphite, and
surface tension of iron. Source: Ref 75
66 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
spheroids while cerium or calcium containing
irons produce only quasi SG, it is reasonable
to conclude that the reactive impurities also
play a direct role on the graphite habitus.
Impurities in the melt also will affect the
growth habitus of the graphite crystal. Recent
research by Muhmond and Fredriksson
(Ref 78), who used simulations with a molecule
editor program and visualizer, established that
trace elements in the melt can attach to the
basal plane of a graphene layer, and that pen
tagonal, hexagonal, and high order carbon rings
can be present as defects in the basal plane. In
the absence of defects, graphite crystals grow
mainly in the a direction. However, the pres
ence of some trace elements, vacancies, and
carbon ring defects creates situations for
growth along the c direction and/or curvature
in the basal plane, as exemplified in Fig. 19
for oxygen. Nitrogen behaves similarly to oxy
gen. Other elements, such as sulfur, selenium,
and boron, attach to the basal plane and stabi
lize lamellar growth. It may be assumed that
magnesium also produces bending of the gra
phene sheet, and that the effect is very strong
once the other surface active elements are
eliminated as oxides or sulfides.
In metal casting practice, the LG to SG tran
sition is triggered through the addition of small
amounts of magnesium or lanthanides to a low
sulfur iron. High cooling rates decrease the
magnesium addition needed for the transition
to occur. Some typical limits of selected impu
rities associated with the various graphite
shapes are summarized in Table 3. They should
be considered only as guidelines, because cool
ing rate and other impurities will affect the
ranges. In principle, the LG to SG transition is
favored by higher cooling rate, decreasing
amounts of anticompacting impurities, and
increasing amounts of compacting impurities.
Crystallization of Graphite
from the Liquid
In industrial cast iron alloys, graphite can be
produced through solidification or through
solid state transformation (heat treatment). The
room temperature morphology of graphite pro
duced through the solidification route is the
result of a four stage growth process:
1. Crystallization from the liquid
2. Cooperative or divorced graphite/g growth
during the eutectic transformation
3. Thickening during cooling to the eutectoid
temperature
4. Thickening during the eutectoid transformation
This section is concerned only with graphite
crystallized directly from the liquid or in con
tact with the liquid during the eutectic reaction.
Some early concepts of the mechanism of
solidification of various graphiteshapes are
summarized in Fig. 20, after Ref 57, 79, and
80. It purports to show that the LG/g eutectic
grain grows in a radial manner as graphite
sheets bend, twist, and branch while growing
in the a direction (Fig. 20a). For CG, bending
and stacking of graphite plates along the c axis
is suggested (Fig. 20b d). Chunky graphite
appears to be made of conical sectors of plates
along the c axis (Fig. 20e). The graphite spher
oids are made of conical sectors growing from
the same nucleus (Fig. 20f, g).
The artistic rendition of the mechanism is
correct. The models assume that in CG and
SG iron the graphite crystals are growing pre
dominantly in the c direction, which is not sup
ported by experimental facts. As discussed later
in this section, significant growth along the
c axis is not probable for the graphite crystal,
but the graphite aggregate can grow predomi
nantly in the c direction.
The significant change in surface energy caused
by the addition of reactive elements prompted a
number of investigators (Ref 81 84) to conclude
that the higher surface energy promotes SG as
the system attempts to decrease its energy. There
is a critical graphite/liquid (Gr/L) interface energy
above which polycrystalline SG is favored over
single crystal LG. Yet, there are many arguments
against this theory, as summarized in Ref 5.
Maybe the most significant one is that the maxi
mum surface energy is reached at approximately
0.018% Mg (Fig. 18a), while at least 0.03% Mg
is needed for well formed SG.
The effect of the reactive and surface active
elements in modifying the graphite shape then
was attributed to their role in changing the ratio
between the growth velocity on the prism
½1010� face (a direction) and that on the basal
[0001] face of graphite (c direction) (Ref 85).
As the growth direction changes from a to c,
the graphite shape changes from lamellar to
spheroidal. (See Ref 5 and 53 for in depth anal
ysis of this theory.) While it is clear that graph
ite aggregates, such as chunky graphite, can
grow in the c direction, there is little evidence,
if any, of graphite crystals growing significantly
in the c direction.
Another theory considering the differences in
growth velocity on the graphite surfaces,
although not based on surface energy argu
ments but on kinetic ones, is the defect growth
of graphite theory (Ref 86). Three growth
mechanisms are considered: 2 D nucleation,
step of a defect (twisted) boundary, and screw
dislocation. The first two mechanisms are gov
erned by exponential laws and apply to the
½1010� surface, but the third is governed by a
parabolic law and applies to the (0001) surface
of the graphite crystal. When weak, reactive
impurities such as sulfur are present in the melt
(contaminated environment), the edge energy
of steps changes, resulting in relative position
change of the growth rates involved, as shown
in Fig. 21(a). The curve for growth on the step
of a defect boundary, Vstep, is at a lower under
cooling than those for growth by 2 D nucle
ation, V2-D, or by screw dislocation, Vscrew. In
a pure environment such as an Fe C Si alloy
with no sulfur, the growth rate curves are dis
placed to higher undercooling (Fig. 21b). In a
melt of sufficient purity, or when increasing
cooling rate, the higher degree of undercooling
may allow growth through screw dislocations,
so that graphite spheroids can form. The LG
to SG transition was obtained experimentally
for pure nickel carbon alloys by increasing the
cooling rate of the melt, or for ultrapure iron
carbon alloys by cooling slowly in a vacuum.
In an environment with reactive impurities
(e.g., magnesium), the impurity will react with
the surface, and the growth at a step of a twist
boundary will be neutralized. Only the curves
for V2-D and Vscrew are left, and they are dis
placed to greater undercooling (Fig. 21c).
More recently, the role of the growth velocity
ratio in the a and c directions in determining
the graphite shape also has been advocated by
Fig. 19 Growth of graphene in the c-direction, caused by the attachment of oxygen out of the basal plane, and
carbon-ring defects (Avogadro software). Pentagonal rings create curvature in the basal plane, which can
cause conical or spheroidal growth of graphite. Source: Ref 78
Table 3 Graphite shape as a function
of typical impurity levels for small
and medium sized castings
Graphite shape Sulfur, %
Oxygen at
1420 �C
(2590 �F), ppm Magnesium, %
Lamellar type A
(LGA)
>0.03 >0.75 0.012 >0.75 0.035
Source: Ref 5
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 67
 
 
 
Amini and Abbaschian (Ref 87). They argued
that a roughening transition from faceted to dif
fuse Gr/L interfaces produced by supersatura
tion is responsible for the LG to SG transition.
They argued that the growth in length (a direc
tion) of LG is diffusion controlled, while the
thickening (c direction) is surface controlled
through 2 D polynucleation growth. Further,
they assumed that at small solidification rates,
the graphite crystal basal and prismatic planes
are faceted. As the interface velocity increases,
the supersaturation increases, and the faceted
interface becomes gradually rough.
Based on TEM observations, Theuwissen
et al. (Ref 88) argued that graphite aggregates
consist of growth blocks stacked upon each
other, and that graphite crystals develop mainly
by a 2 D nucleation and growth mechanism.
Yet, unlike the layer by layer growth, in which
each new layer corresponds to a graphene sheet
advocated by Amini and Abbaschian, they
argued that there should be a critical block
thickness required for further growth of graph
ite aggregates instead of atomic layers.
According to Double and Hellawell (Ref 89),
the ability of a graphene sheet to bend in steps
of 20�450 about three 1100h i axes mutually
inclined at 120� makes it possible for a lamellar
crystal to roll upon itself as conical helices,
which may grow into a spheroid (Fig. 22a).
This model explains well the conical sectors
observed in TEM images of SG. There is
SEM evidence for helical growth (Fig. 22b),
although not for complete conical helices.
In the growth model proposed by Sadocha
and Gruzleski (Ref 70), a graphite spheroid
may result from repeated bending of the graph
ite sheets. It was postulated that a large number
of steps on the surface of the spheroid grow in
the a direction by curved crystal growth, with
the low energy basal plane of graphite exposed
to the liquid. In the presence of surface active
impurities that decrease the surface tension
(sulfur or oxygen), the spherical shape is dete
riorated into a lamellar one. While there is
(a) (b) (c) (d)
(e) (f) (g)
a 
a 
a 
a 
a 
a 
Lamellar graphite
sheets
Austenite
-axis
-axis -axis
Fig. 20 Schematic representation of models for growth mechanisms of various graphite morphologies. (a) Lamellar graphite (LG)/austenite eutectic grain. Source: Ref 79.
(b) Compacted graphite (CG)/austenite eutectic grain. Source: Ref 79. (c) CG developing out of LG. Source: Ref 80. (d) CG developing out of spheroidal graphite.
Source: Ref 80. (e) Chunky graphite. Source: Ref 57. (f) Irregular graphite nodule. Source: Ref 57. (g) Graphite spheroid. Source: Ref 57. Reprinted with permission from The
American Foundry Society
Undercooling, ΔT Undercooling, ΔT Undercooling, ΔT
G
ro
w
th
 v
el
oc
ity
, V
 
G
ro
w
th
 v
el
oc
ity
, V
G
ro
w
th
 v
el
oc
ity
, V
A AB BBC C C
(a) (b) (c)
Fig. 21 SuggestedDT V correlation for ð1010Þ and (0001) crystal faces of graphite growing in various environments. (a) Contaminated environment (e.g., sulfur-containing Fe-C-Si
alloy). (b) Pure environment (e.g., pure Fe-C-Si alloy). (c) Environment with reactive impurities (e.g., magnesium-containing Fe-C-Si alloy). Three growth mechanismsare
discussed: A, on the step of the defect boundary; B, two-dimensional nucleation; and C, screw dislocation. Source: Ref 86
68 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
metallographic evidence of curved growth of
graphite, this model cannot explain the occur
rence of the radial sectors. A modification of
the model assumes that, while growing in the
a direction, the graphite sheets tilt through twin
ning to minimize surface energy (Fig. 22c). The
formation of neighboring detached radial sectors,
as seen in Fig. 13(b), remains unexplained. Note
that the models in Fig. 22 are postulating the
a direction as the main growth direction, albeit
through different mechanisms.
A completely different line of thinking was
advanced by Saratovkin (Ref 90) as early as
1959. Based on observations on the growth of
cadmium iodide, he developed the concept of
foliated crystals, which are assemblies of thin
plates separated by solvent impurity layers,
and foliated dendrites. Foliated dendrites occur
because of solute accumulation on the basal
planes. Protuberances originating in defects
such as spiral dislocations will grow in the
c direction and resume anisotropy controlled
growth once they reach liquid poorer in solute
(Fig. 23). This concept then was used to explain
graphite growth in cast iron and the entrapment
of iron between the foliated graphite plates.
Growth of iron carbide in cast iron also was
considered to be a case of foliated crystals.
Dendritic graphite aggregates were observed
both in SG and CG irons (Ref 56, 91, 92), as
shown in Fig. 24. While some researchers
(Ref 92) argued that every branch of the den
drite is an independent columnar crystal grown
from their own nucleus situated along the prin
cipal trunk of the dendrite, others (Ref 56)
could not confirm whether the dendritic pattern
consists of a single crystal or many columnar
crystals radiating from nuclei scattered along
the principal axis of the dendrite.
Saratovkin’s theory was revived by Rovi
glione and Hermida (Ref 93), who advanced
the idea that the constitutive elements of CG
and SG are clusters of randomly distributed
and heavily distorted small, faceted crystals,
with basal planes forming major surfaces, and
prismatic planes forming minor ones. They rea
soned that these are foliated dendrites, and that
addition of reactive elements produces compac
tion forces on the graphite by the austenite shell
and the melt and cause increased twinning of
the foliated dendrites, resulting in the crystalli
zation of CG or SG.
The existence of thin graphite platelets with
nanometer height and micrometer width as the
building blocks of both SG and CG was recently
confirmed (Ref 68, 91, 94). The platelets are par
ticularly visible in samples obtained through
interrupted solidification, as shown in Fig. 25.
Furthermore, it was established that in themagne
sium free irons, graphite platelets grow into foli
ated crystals that assemble in a tiled roof
configuration (Fig. 23a), forming graphite plates
that grow in the general a direction (Fig. 26).
(a) (b) (c)
(q)
[0001]
(a) (b) (c)
(q)
[0001] a
Fig. 22 Growth models of graphite crystals. (a) Growth by helical bending. Conical helices radiating from a common center. Source: Ref 89. (b) Scanning electron microscopy-
based drawing of helical growth of a graphite crystal in a chunky graphite aggregate. Reprinted with permission from Springer. Source: Ref 5. (c) Circumferential growth
with boundary tilt through twinning. Reprinted with permission from Springer. Source: Ref 5
Fig. 24 Optical micrographs showing dendritic graphite aggregates in cast iron. (a) Graphite spheroid with dendritic
outgrowth. Original magnification: 250�. Reprinted with permission from The Japan Institute of Metals and
Materials. Source: Ref 56. (b) Compacted graphite dendrites. Source: Ref 91
(a) (b)
Fig. 23 Schematic representation of the growth mechanism of foliated dendrites. (a) Growth of graphite platelets as
foliated dendrite organized in a tiled-roof configuration. (b) Growth of graphite platelets as disorganized
foliated dendrite. Source: Ref 91
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 69
 
 
 
In the magnesium modified melts, the
graphite platelets become disorganized
(Fig. 23b, 27a) and stack along the c axis. A
gradual consolidation of the platelets into
clusters (blocky graphite) with random orienta
tion (Fig. 27a) and then reorganization of the
clusters into columns (Fig. 27b) follow at higher
magnesium content or higher cooling rate. As
solidification advances, thickening of the plate
lets occurs through growth of additional graphene
layers nucleated at the ledges of the graphite
prism and through recrystallization of amorphous
carbon diffused from the liquid through the
austenite shell. Thickening through 2 D nucle
ation or screw dislocation nucleation also is prob
able. At a further increase in themagnesium level
during early solidification, the graphite platelets
lose some of their hexagonal shape and begin
organizing into conical sectors or tiled roof con
figuration (Fig. 25b, c). As solidification
advances, the platelets assemble into clusters,
and the clusters assemble into conical sectors
growing from the same nucleus. The conical sec
tors may partially occupy the volume of the
sphere, forming chunky graphite (Fig. 27b), or
they may fill the whole volume to produce a
graphite spheroid. The large number of cavities
(defects) observed between the platelets in all
graphite morphologies is consistent with growth
of foliated dendrites.
A summary of the foliated dendrite mechan
isms working in the crystallization of various
forms of graphite is presented in Fig. 28. It
illustrates the gradual change in the stacking
of the foliated platelets from along the a axis
to along the c axis.
Compacted and spheroidal graphite grow
initially in the liquid, but significant growth
occurs after encapsulation into an austenite
shell, through carbon diffusion from the liq
uid through the solid shell. Transmission
electron microscopy evidence (Ref 95, 96)
shows that amorphous carbon is deposited
on the graphite surface and then recrystallizes
to produce the layered structure seen in SG
(Fig. 29).
It is most certain that in strongly hypereutec
tic irons, both LG and SG grow initially in con
tact with the liquid. For hypoeutectic irons,
solidification starts by the formation of austen
ite dendrites. The graphite then forms in the
interdendritic liquid by a eutectic reaction, with
major differences between the growth of LG
and SG.
Eutectic Solidification of Cast Iron
As discussed in detail in Ref 5, the basic
parameters affecting the morphology of the
eutectic are the temperature gradient (G)/
growth velocity (V) ratio and the composition.
Two different solidification processes must be
considered:
� Continuous cooling solidification, where the
controlling factor is the G � V product (the
cooling rate)
� Directional solidification, where microstruc
ture formation is controlled by the G/V ratio
Fig. 25 Scanning electron micrographs on samples obtained through interrupted solidification, demonstrating that the graphite aggregates are composed of a multitude of graphite
platelets of various orientations. (a) Disorganized growth of graphite platelets in a foliated dendrite configuration in compacted graphite. Source: Ref 91. (b) Platelets with
conical sector orientation in spheroidal graphite. Reprinted with permission from Elsevier. Source: Ref 68. (c) Platelets with tiled-roof orientation in spheroidal graphite. Reprinted with
permission from Elsevier. Source: Ref 68
Fig. 26 Scanning electron micrographs of sand-cast lamellar graphite (LG) irons at room temperature. (a) Parallel
graphite (Gr) platelets at the g/liquid interface in low-sulfur LG iron. Reprinted with permission from
Elsevier. Source: Ref 68. (b) Fracture surface showing tiled-roof configuration of graphite platelets in a graphite
lamella. Courtesy of W.L. Guesser and the Tupy/SENAIproject
70 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
The discussion also must include the two equi
libria: stable austenite graphite eutectic (gray)
and metastable austenite iron carbide eutectic
(white).
Coupled Zone in Cast Iron
The degree and type of eutectic growth that
occurs in cast iron can be determined by using
tools such as growth velocity curves to locate
coupled zone regions, isothermal time tempera
ture diagrams to gage susceptibility to carbide
formation, and growth velocity/composition
plots to ascertain parameters that affect both
directional and multidirectional solidification.
The coupled growth region of the eutectic in
cast iron is asymmetric. It is possible to con
struct a theoretical coupled zone for gray iron
from the condition of equal growth rate of the
austenite and graphite phases (Ref 97). See also
the discussion on this subject in Ref 53.
The transition from a fully eutectic to a eutectic
+ dendrite structure in pure iron graphite alloys of
eutectic composition and the calculated g/iron and
graphite/iron eutectic boundaries by Jones and
Kurz (Ref 98) are shown in Fig. 30. The diagrams
suggest that alloys of eutectic composition can
exhibit primary austenite in the microstructure
when solidifying at high undercooling resulting
from relatively high cooling rate.
Stable Solidification of Austenite-
Graphite Eutectic—Continuous Cooling
In hypoeutectic LG iron, austenite graphite
eutectic grains can nucleate at the austenite/liquid
interface or in the bulk of the liquid, depending on
the sulfur andmanganese content and on the cool
ing rate. When nucleation occurs on the primary
austenite, several eutectic grains can nucleate
and grow on the same austenite dendrite
(Fig. 31a). The eutectic austenite then grows on
the primary austenite and has the same crystallo
graphic orientation. Thus, a final austenite grain
may include several eutectic grains (Ref 28).
In eutectic irons, the quasi spherical eutectic
grains nucleate and grow in the liquid
(Fig. 31b). Low nodularity (ratio between the radii of the
g shell and the graphite spheroid (rg/rGr = 2.3)
throughout the microstructure evolution.
The primary austenite growing into the liquid
will tend to grow anisotropically in its preferred
crystallographic orientation (Fig. 35a). However,
isotropic diffusion growth will impose an
increased isotropy on the system. Consequently,
the dendritic shape of the austenitewill be altered,
and the g/L interface will exhibit only small pro
tuberances instead of clear secondary arms
(Fig. 35b). This process is dominant toward the
end of solidification. The result is large austenite
dendrites (primary and eutectic) that incorporate
numerous graphite spheroids. A eutectic grain
cannot be defined for the austenite/SG eutectic,
because it is not possible to separate the primary
austenite from the eutectic one.
There has been some debate over the
uninodular (Ref 106, 107) or multinodular
morphology of the eutectic grain in SG iron.
Rivera et al. (Ref 108) used color etching
metallography to demonstrate that several
graphite spheroids typically are surrounded
by a highly segregated last to freeze region.
This aggregate was defined as a multinodular
eutectic grain. Further research by the same
scientists (Ref 1) revealed that the solidification
microstructure of SG iron includes large austenite
grains with numerous graphite nodules.
As discussed, the solidification mechanisms of
LG and SG cast iron are quite different. An
important practical consequence of this difference
is that LG iron solidifies with skin formation,
while SG iron is characterized by mushy solidifi
cation (Fig. 36).
Stable Solidification of Austenite-
Graphite Eutectic—Directional
Solidification
Because the growth velocity and the temper
ature gradient can be controlled independently,
the information obtainable through directional
solidification (DS) experiments is extremely
valuable in understanding the intricacies of
solidification of cast iron.
(a)
(b)
(d)
(c)
Fig. 32 Solidification of the eutectic in lamellar graphite iron during continuous cooling (different gray shades
indicate different crystallographic orientations). (a) Eutectic iron, early solidification. (b) Eutectic iron, late
solidification. (c) Hypoeutectic iron or eutectic iron at high cooling rate, early solidification. (d) Hypoeutectic iron or
eutectic iron at high cooling rate, end of solidification. Reprinted with permission from Elsevier. Source: Ref 28
(b)
(a)
Fig. 33 Models for the solidification morphology of
near-eutectic-composition spheroidal graphite
iron, mushy-type solidification. (a) After Engler and
Ellerbrok. Adapted from Ref 101. (b) Adapted from Ref 102
γm
L
L
γ + L
G/γ
Gr/γ
γ /L
γ /L
L/γ
L/γ
γ
γ
T
Te
m
pe
ra
tu
re
, °
C
Gr
Gr
rGr rγ Radius
γm
L
L+G
γ + Lγ
L
Gr/γ//
γ /Lγ
L/γ//
γ
Gr
Gr
Composi ion, %
C
om
po
si
tio
n,
 %
Fig. 34 Isothermal growth of a graphite spheroid
within an austenite shell. Source: Drawing in
Ref 5 after Ref 103
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 73
 
 
 
As shown in Fig. 37, it is possible to achieve
a variety of structures in cast iron when varying
the G/V ratio and/or the level of impurities
(e.g., cerium). As the G/V ratio decreases or
the composition Co (e.g., magnesium or
cerium) increases, the solid/liquid interface
changes from planar to cellular and then to
equiaxed, while graphite remains lamellar.
The austenite and graphite grow cooperatively.
Further decrease of G/V or increase of Co
brings about formation of an irregular interface,
with austenite dendrites protruding in the liq
uid. Graphite becomes compacted and then
spheroidal. Eutectic growth is divorced.
Interrupted DS experiments summarized in
Fig. 38 confirmed the sequence proposed in
Fig. 37. It is seen that the solid/liquid interfaces
of SG and CG irons are coarse, while that of LG
iron is closer to planar. During solidification of
LG iron, the graphite and the austenite grow coop
eratively. Graphite is the leading phase. The CG
iron shows an austenite/cellular interface that
includes the CG. The SG iron solidifies with an
austenite/dendritic interface that incorporates
graphite nodules. In these last two cases, the lead
ing phase during solidification is the austenite. In
the case of SG iron, even for hypereutectic irons,
the graphite spheroids are associated with austen
ite dendrites. The eutectic austenite is dendritic
and cannot be distinguished from the primary aus
tenite. Spheroidal graphite precipitates directly
from the melt, becomes enveloped in an austenite
shell that is very soon incorporated into the den
drites, and then grows together with the dendrites
(Ref 110).
Lakeland and Hogan (Ref 111) and then Argo
and Gruzleski (Ref 112) presented composition
G/V diagrams for stable eutectic cast irons. Later,
the complete structural transition, from meta
stable to stable and for different graphite
morphologies, was documented for cast irons of
hypoeutectic composition as a function of
growth velocity and temperature gradients at
the solid/liquid interface as well as cerium
Fig. 35 Microstructures of spheroidal graphite iron found in the same microshrinkage cavity from a cast plate. (a) Primary austenite dendrite. (b) Eutectic austenite dendrite with
encapsulated graphite spheroids. (c) Overall view of microshrinkage. Reprinted with permission from The American Foundry Society. Source: Ref 105
(b)
(a)
Microshrinkage
Fig. 36 Schematic illustration of solidification
mechanisms of continuously cooled (a)
lamellar and (b) spheroidal graphite cast iron
Fig. 37 Schematic representation of the influence of composition (%Ce), temperature gradient (G), and growth velocity
(V) over the eutectic morphology of Fe-C-Si alloys. SG, spheroidal graphite; CG, compacted graphite; LG,
lamellar graphite. Reprinted with permission from Cambridge University Press. Source: Ref 55, 102
74 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
concentration (Fig. 39). Because cerium was
used as a graphite shape modifier, the graphite
was not fully spheroidal. It was found that while
the metastable to stable (white to gray) transi
tion depends mostly on the G/V ratio, the transi
tion between different graphite shapes (lamellar
to compacted to spheroidal) depends mostly on
the cerium concentration.
For solidification of regular eutectics, a number
of relationships have been established between
process and material parameters based on the
extremum criterion (Ref 5). The interlamellar
spacing of LG that has solidified with a planar
interface as a function of cooling rate and compo
sition is summarized in Fig. 40.As the cooling rate
(growth velocity) decreases, the lamellar spacing
increases. Sulfur additions increase the spacing,
even in amounts as low as 0.001%. For the sul
fur containing irons, a sudden decrease in spacing
is observed at growth velocities of approximately
10�5m/s. It can be attributed to the transition from
type A to type D graphite.
Many other DS data were generated for LG
iron. All of these data are positioned above
the theoretical Jackson Hunt model. As dis
cussed in detail in Ref 5, LG iron is an irregular
eutectic, and both an extremum and branching
spacing can be defined. The average spacing is
larger than the extremum one predicted from
the Jackson Hunt theory.
Based on DS experimental information, the
sequence of changes in the eutectic morphology
of DS gray cast iron is summarized in Fig. 41.
As the cooling rate (G � V) increases, the inter
face of LG iron changes from planar to cellular
and then to equiaxed. For relative higher cool
ing rates, the cellular interface may break down
into a dendritic equiaxed mushy zone. Coopera
tive growth of austenite and graphite occurs.
Before the breakdown of the planar interface,
the graphite becomes much finer and twisted.
An increase in the cooling rate may result in
higher nodularity for CG iron and finer struc
ture for SG iron. Increasing the amount of reac
tive impurities,or decreasing the content of
surface active impurities, brings about forma
tion of an irregular interface, with austenite
dendrites protruding in the liquid, and changes
graphite shape from LG to CG and then to
SG. Eutectic growth changes from cooperative
to divorced.
Metastable Solidification of
Austenite-Iron Carbide Eutectic
Metastable solidification of cast iron pro
duces what is commercially called white iron,
whose microstructure consists of austenite den
drites and austenite iron carbide (Fe3C) eutectic
(ledeburite). The white eutectic consists of iron
carbide plates or rods in an austenitic matrix
that becomes pearlite at room temperature
(Fig. 42a). Growth of ledeburite begins with
the development of a cementite plate on which
an austenite dendrite nucleates and grows
(Fig. 43a). This destabilizes the Fe3C, which
then grows through the austenite. As a result,
two types of eutectic structure develop: a lamel
lar eutectic with Fe3C as the leading phase in
the edgewise a direction, and a rodlike eutectic
in the sidewise c direction (Fig. 43b). Under
Fig. 38 Influence of composition and solidification velocity on the morphology of the solid/liquid interface. (a) Spheroidal graphite iron, V = 5 mm/s, magnesium added.
(b) Compacted graphite iron, V = 5 mm/s, magnesium added. (c) Lamellar graphite iron, V = 1.2 mm/s, no magnesium. Reprinted with permission from The American
Foundry Society. Source: Ref 109
Fig. 39 Influence of temperature gradient/growth
velocity (G/V) ratios and percent cerium on
structural transitions in cast iron. CG, compacted
graphite; SG, spheroidal graphite; FG, flake graphite, i.e.,
lamellar graphite. Source: Ref 55
10–3
10–6
10–5
10–8 10–7 10–6 10–5 10–4
 10–4
Growth velocity, m/s
La
m
el
la
r 
sp
ac
in
g,
 m
MK
MK 0.1Si
MK 0.001S
Oh 0.007S
Oh 0.047s
Fig. 40 Effect of growth velocity and composition on the lamellar spacing of lamellar graphite. Data from Magnin
and Kurz (MK) (Ref 113) and Ohira et al. (Oh) (Ref 114). MK: Fe-C eutectic; MK 0.1Si: Fe-C-0.1%Si; MK
0.001S: Fe-C-0.001%S; Oh 0.007S: Fe-C-0.01%Si-0.007%S; Oh 0.047S: Fe-C-0.01%Si-0.047%S. Source: Ref 5
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 75
 
 
 
specific DS conditions, ledeburite behaves like
a regular eutectic, as shown in Fig. 42(b).
The cooling rate has a significant influence
on the morphology of the g/Fe3C eutectic. At
moderate undercoolings, ledeburite structure is
expected. High cooling rates, as obtained in
quenching experiments, produce a degenerated
eutectic structure dominated by Fe3C plates. A
coarse mixture of Fe3C and g fills the spaces
between the Fe3C plates. This structure obvi
ously does not result from cooperative growth.
A platelike carbide structure associated with
equiaxed eutectic grains can be obtained by
increasing the undercooling through superheat
ing, by decreasing the silicon content, or by
adding chromium or magnesium.
A suggested mechanism of the effect of cool
ing rate on the morphology of the metastable
eutectic is summarized in Fig. 44, after
Ref 117. As the cooling rate increases, the lamel
lar part of the original parallelepipedic eutectic
grain becomes larger at the expense of the rod
eutectic, the grain starts bending inward, and
eventually a spherulitic eutectic grain results.
Unalloyed white irons do not have significant
practical applications, but medium and high
alloyed irons are used extensively for their
abrasion resistance. The alloying elements
change the composition and morphology of
the carbides as well as the microstructure of
the matrix. For example, irons having relatively
high carbon, 3 to 5% Ni, and 1.4 to 4% Cr
solidify with a martensitic matrix (Fig. 45a).
Chromium promotes white solidification with
a continuous network of alloyed iron carbides,
(FeCr)3C. When the chromium content is
increased to 7 to 11%, the composition and
morphology of the carbides changes to discon
tinuous Cr7C3 eutectic carbides (Fig. 45b).
Gray-to-White Structural Transition
For the cast iron manufacturer, the stable to
metastable microstructure transition (also known
as the gray to white transition, or GWT) is
of particular interest because it results in the
Fig. 41 Effect of processing parameters (G, V, Co) on the solid/liquid interface morphology and graphite shape in directionally solidified austenite-graphite eutectics. Reprinted
with permission from Elsevier. Source: Ref 28
Fig. 42 Microstructure of iron-iron carbide eutectic (ledeburite). (a) Continuous-cooling solidification. Source:
Ref 115. (b) Longitudinal section of directionally solidified white cast iron. Source: Ref 113
76 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
occurrence of unwanted iron carbides in the gray
iron. In a binary iron carbon alloy, the difference
between the stable (Tst) and metastable (Tmet)
eutectic temperatures is only 3 to 5 K. Thus, dur
ing cooling of an iron carbon casting, the tem
perature of the melt may become smaller than
Tmet before any stable structure nucleates and
grows. This may happen even at very low cool
ing rates. In the Fe C Si system, the Tst Tmet
interval is much larger, and stable solidification
may occur before the temperature reaches Tmet.
The gray or white solidification mode of cast
iron depends on the relative nucleation proba
bility and the growth rates of the graphite and
Fe3C phases. In turn, this will be a function of
the cooling rate and chemistry of the alloy. As
shown in Fig. 46, only the graphite eutectic
can nucleate and grow between the eutectic
temperature for the gray eutectic (1153 �C, or
2107 �F) and that for the white eutectic
(1148 �C, or 2098 �F). Below 1148 �C, both
eutectics can occur. The growth rate of the
g/Fe3C eutectic rapidly exceeds that of the
g/graphite eutectic, and at a temperature of
approximately 1140 �C (2084 �F), GWT occurs
(Ref 119). The intersection of the two curves in
Fig. 46 can be defined as the critical growth rate
for GWT. Alternatively, a critical cooling rate
can be established for the case of continuous
cooling solidification.
Magnin and Kurz (Ref 120) further devel
oped this concept by introducing the role of
nucleation. It is well accepted that nucleation
of white iron is more difficult than that of gray
iron. Consequently, additional undercooling, in
excess of that predicted from growth velocity
considerations, is required for a complete
GWT. This is shown in the figure as DTn
met;
that is, the nucleation temperature of the meta
stable cementite displaces the critical velocity
to the right, from Vcr to Vg�w. Thus, a complete
white iron is obtained only at growth velocities
larger than Vg�w. A similar argument holds for
the white to gray transition, when a complete
transition cannot occur unless the undercooling
is smaller than DTn
st, which is the nucleation
temperature of the stable gray eutectic. Thus,
a region of mixed structure, gray and white,
will exist at growth velocities between Vw�g
and Vg�w. This is the mottled region.
An analytical model proposed by Fras and
Lopez (Ref 121) introduces the concept of a
chilling equivalent. For eutectic iron, the chill
ing equivalent is:
Edgewise growth;
cementite is leading
Crystallographic
-direction of
cementite
Sidewise growth
Edgewise (a)
E
dgew
ise (
) 
S
idew
ise (
)
a b c d
a
b
c
(a) (b)
Fig. 43 Schematic representation of growth of ledeburite eutectic. (a) Lamellar eutectic with cementite as the
leading phase in the edgewise a-direction. Reprinted with permission from Jernkontoret—The Swedish
Steel Producers’ Association. Source: Ref 116. (b) Rodlike eutectic in the sidewise c-direction
Fig. 44 Schematic illustration of change in morphology of austenite-Fe3C eutectic grain at increasing cooling rate.
Source: Ref 117
Fig. 45 Microstructures of nickel-chromium abrasion-
resistant white irons. (a) 3–3.6% C, 3.3–5%
Ni, 1.4–4% Cr. (b) 2.5–3.6% C, 5–7% Ni, 7–11% Cr.
Original magnification: 340�. Source: Ref 118
Microstructure Evolution during the Liquid/SolidTransformation in Cast Iron / 77
 
 
 
E
T1:08
st
1:9�T0:5
pour
1
m1 m23 c5 �Teut þ�Tn
metð Þ10
 !1=6
where DTpour is the superheating above the
eutectic temperature; m1 and m2 are nucleation
and growth coefficients, respectively; and c is
the specific heat. It is seen that the chilling ten
dency (chilling equivalent) decreases as the
number of eutectic grains and their growth rate
(m1 and m2) increase and as the eutectic interval,
DTeut, and the pouring temperature increase.
REFERENCES
1. R.E. Boeri and J.A. Sikora, Int. J. Cast
Met. Res., Vol 13 (No. 5), 2001, p 307 313
2. L.F. Mondolfo, in Grain Refinement in
Casting and Welds, G.J. Abbaschian and
S.A. David, Ed., The Metallurgical Soci
ety of AIME, Warrendale, PA, 1983, p 3
3. B. Bramfitt, Metall. Trans., Vol 1, 1970,
p 1987 1995
4. H. Fredriksson and U. Åkerlind, Solidifi
cation and Crystallization Processing in
Metals and Alloys, Wiley, 2012
5. D.M. Stefanescu, Science and Engineer
ing of Casting Solidification, 3rd ed.,
Springer, 2015
6. P. Basutkar, S. Yew, and C. Loper, Trans.
AFS, Vol 77, 1969, p 321 328
7. G. Ruff and J.F. Wallace, Trans. AFS, Vol
84, 1976, p 705 728
8. E. Moumeni, D.M. Stefanescu, N.S.
Tiedje, P. Larranaga, and J.H. Hattel,
Investigation on the Effect of Sulfur and
Titanium on the Microstructure of Lamel
lar Graphite Iron, Metall. Mater. Trans. A,
Vol 44, 2013, p 5134
9. R. Naro and J.F. Wallace, Trans. AFS,
Vol 77, 1969, p 311 320
10. J.F. Wallace, Trans. AFS, Vol 83, 1975,
p 363 378
11. D. Zeng, Y. Zhang, J. Liu, H. He, and X.
Hong, in Science and Processing of Cast
Iron VIII, Y.X. Li, H.F. Shen, Q.G. Xu,
and Z.Q. Han, Ed., Tsinghua University
Press, 2006, p 86 91
12. A. Okada and H. Miyake, The Unknown
World of Cast Iron, Kansai University
Press, 1996, p 141 (in Japanese)
13. K.B. Wilford and F.G. Wilson, Br. Foun
dryman, Vol 78, 1985, p 301, 364
14. H. Tian and D.M. Stefanescu, Metall.
Trans. A, Vol 23, 1992, p 681
15. W. Patterson and D. Ammann, Solidifica
tion of Lamellar Iron Graphite Eutectic in
Gray Iron, Giesserei, Jan 1959, p 1247
16. H.D. Merchant, Solidification of Cast
Iron A Review of Literature, Recent
Research on Cast Iron, H.D. Merchant,
Ed., Gordon and Breach Science Publish
ers, New York, 1968, p 1 100
17. T. Skaland, T. Grong, and T. Grong,
A Model for the Graphite Formation in
Ductile Cast Iron, Metall. Trans. A,
Vol 24 (No. 10), 1993, p 2321 2345
18. S. Steeb and U. Maier, in The Metallurgy
of Cast Iron, B. Lux, I. Minkoff, and F.
Mollard, Ed., Georgi Publishing, St.
Saphorin, Switzerland, 1974, p 1 11
19. W. Krieger and H. Trenkler, Arch.
Eisenh€uttenwes., Vol 42, 1971, p 175
20. A.A. Vertman and A.M. Samarin, Dokl.
Akad. Nauk SSSR, Vol 134, 1960, p 629
21. A.M. Samarin and V.A. Izmailov, Sov.
Phys. Dokl., Vol 14, 1969, p 392
22. A.A. Vertman and A.M. Samarin, Ed.,
Svoista Rasplavov Jeleza, Nauka, Mos
cow, 1969
23. A. Boyles, The Structure of Cast Iron,
American Society for Metals, Metals
Park, OH, 1947
24. J.T. Eash, AFS Trans., Vol 49, 1941,
p 887 906
25. G.A. Feest, G. McHugh, D.O. Morton, I.S.
Welch, and I.A. Cook, in Proc. Solidifica
tion Technology in the Foundry and Cast
House, The Metals Society, London, 1983
26. C.H. Wang and H. Fredriksson, 48th
International Foundry Congress (Varna,
Bulgaria), 1981
27. H. Fredriksson, Mater. Sci. Eng., Vol 65,
1984, p 137 144
28. D.M. Stefanescu, G. Alonso, P. Larra
ñaga, and R. Suarez, On the Stable Eutec
tic Solidification of Iron Carbon Silicon
Alloys, Acta Mater., Vol 103, 2016,
p 103 114
29. B.Lux,Mod.Cast., Vol 45, 1964, p 222 232
30. B. Lux and H. Tannenberger, Mod. Cast.,
March 1962, p 57
31. D.M. Stefanescu, Comparison between
the Inoculation Efficiency of Sodium and
Barium, Giesserei Prax., Vol 24, 1972,
p 429 433
32. G.X. Sun and C.R. Loper, Trans. AFS,
Vol 91, 1983, p 639 646
33. W. Weis, in The Metallurgy of Cast Iron,
B. Lux, I. Minkoff, and F. Mollard, Ed.,
Georgi Publishing, St. Saphorin, Switzerland,
1974, p 69 79
34. M.A. Gadd and G.H.J. Bennett, The Phys
ical Chemistry of Inoculation in Cast Iron,
The Physical Metallurgy of Cast Iron, H.
Fredriksson and M. Hillert, Ed., Materials
Research Society, North Holland, 1984,
p 99 108
35. L.R. De and Y.J. Xiang, Trans. AFS, Vol
99, 1991, p 707 712
36. M. Chisamera, I. Riposan, and M. Bar
stow, Paper 3, AFS International Inocula
tion Conference (Rosemont, IL), 1998
37. I. Riposan, M. Chisamera, S. Stan, and T.
Skaland, Int. J. Cast Met. Res., Vol 16
(No. 1 3), 2003, p 105 111
38. I.Minkoff,ThePhysicalMetallurgy of Cast
Iron, John Wiley & Sons, New York, 1983
39. T. Skaland, in AFS Cast Iron Inoculation
Conference, Sept 2005 (Schaumburg,
IL), p 13 30
40. S.I. Karsay, Ductile Iron I: Production,
Quebec Iron & Titanium Corp., 1976
41. R.J. Warrick, AFS Trans., Vol 74, 1966,
p 722 733
42. R.L. Naro and J.F. Wallace, AFS Trans.,
Vol 78, 1970, p 229 238
43. K.M. Muzumdar and J.F. Wallace, AFS
Trans., Vol 81, 1973, p 412 423
44. M.J. Lalich and J.R. Hitchings, AFS
Trans., Vol 84, 1976, p 653 664
45. T. Skaland, AFS Trans., Vol 105, 2001,
p 77 88
46. M.M. Jacobs, T.J. Law, D.A. Melford,
and M.J. Stowell, Basic Processes
Controlling the Nucleation of Graphite
Nodules in Chill Cast Iron, Met. Technol.,
Vol 1, 1974, p 490 500
47. Y. Igarashi and S. Okada, Int. J. Cast Met.
Res., Vol 11, 1998, p 83 88
48. S.E. Franklin and R.A. Stark, Further Use
of Secondary Mass Spectrometry in the
Study of Graphite Morphology Control
in Cast Iron, The Physical Metallurgy of
Cast Iron, H. Fredriksson and M. Hillert,
Ed., Materials Research Society, North
Holland, 1984, p 25 35
49. G. Alonso, P. Larrañaga, D.M. Stefanescu,
E. De la Fuente, A. Natxiondo, and R.
Suarez, Kinetics of Nucleation and Growth
of Graphite at Different Stages of Solidifi
cation for Spheroidal Graphite Iron, Proc.
72nd World Foundry Congress, May 21
25, 2016 (Nagoya, Japan)
50. J. Rickard and I.C.H. Hughes, Eutectic
Structure in White Cast Iron, BCIRA J.,
Vol 9 (No. 1), 1961, p 11
Spontaneous change
from white to gray mode
of solidification
Spontaneous change
from gray to white mode
of solidification
0.01 10–3 10–4 10–5 10–6 
2066
2084
2102
2120
1130
1140
1150
1160
1170
Tst
Tmet
1148 °C
1153 °C
Growth velocity, cm/s
(Growth velocity)1/2
Te
m
pe
ra
tu
re
, °
C
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
Vmet
Teut
Tst
Tn
st
Tn
met
Tmet
V
st
Vw–g Vg–WVcr
(b)
(a)
Fig. 46 Gray-to-white eutectic transition as a function
of undercooling and growth velocity. (a) After
Hillert and Rao. Source: Ref 119. (b) After Magnin and
Kurz. Source: Ref 5
78 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
51. R. Elliot, Cast Iron Technology, Butter
worths, London, 1988
52. C. van de Velde, “Development of The
ories on Graphite Formation in Ductile
Cast Iron,” Jan 15, 2004, ceesvande
velde.eu/partone.htm
53. D.M. Stefanescu, Cast Iron, Casting, Vol
15, ASM Handbook, D.M. Stefanescu,
Ed., ASM International, Metals Park,
OH, 1988, p 168 181
54. K.M. Fang, G.C. Wang, X. Wang, L.
Huang, and G.D. Deng, The Microstruc
ture and Metamorphic Regularity of
Graphite in Cast Iron, Science and Pro
cessing of Cast Iron VIII, Y.X. Li, H.F.
Shen, Q.G. Xu, and Z.Q. Han, Ed. (Beij
ing, China), Tsinghua University Press,
2006, p 181 187
55. D.K. Bandyopadhyay, D.M. Stefanescu, I.
Minkoff, and S.K. Biswal, Structural
Transitions in Directionally Solidified
Spheroidal Graphite Cast Iron, Physical
Metallurgy of Cast Iron IV, G. Ohira, T.
Kusakawa, and E. Niyama, Ed., Materials
Research Society Proc., Pittsburgh, PA,
1989, p 27 34
56. M. Hamasumi, A Newly Observed Pattern
of Imperfect Graphite Spherulite in Nodu
lar Iron, Trans. JIM, Vol 6, 1965,
p 234 239
57. P.C. Liu, C.L. Li, D.H. Wu, and C.R.
Loper, SEM Study of Chunky Graphite
in Heavy Section Ductile Iron, Trans.
AFS, Vol 91, 1983, p 119 126
58. G.R. Purdy and M. Audier, in The Physi
cal Metallurgy of Cast Iron, Vol 34, H.
Fredriksson and M. Hillert, Ed., Materials
Research Society Proc., North Holland,New York, 1984, p 13 23
59. K. Theuwissen, M.C. Lafont, L. Laffont,
B. Viguier, and J. Lacaze, Trans. Indian
Inst. Met., 2012, published online TP2602
60. T. Hara, T. Kitagawa, K. Kuroki, S. Sai
kawa, K. Terayama, S. Ikeno, and K. Mat
suda, Mater. Trans., Vol 55, 2014,
p 1500 1505
61. K.M. Fang, Atlas of the Morphology and
Microstructure of the Graphite in Cast
Iron, Science Publishing Co. of China,
2000
62. E. Aguado, D.M. Stefanescu, J. Sertucha,
P. Larrañaga, and R. Suárez, Effect of
Carbon Equivalent and Alloying Elements
on the Tensile Properties of Superfine
Interdendritic Graphite Irons, Trans.
AFS, Vol 122, 2014, p 249 258
63. B. Lux, Giessereiforschung, Vol 19, 1967,
p 141
64. M.Hatate, K.Nakamura, andH. Sumimoto,
Production of Iron Base Alloys with Fine
Graphite and Its Physical Properties, Physi
cal Metallurgy of Cast Iron IV, G. Ohira, T.
Kusakawa, andE.Niyama, Ed., Proc.Mate
rials Research Society, Pittsburgh, PA,
1989, p 149 156
65. D.M. Stefanescu, L. Dinescu, S. Cra
ciun, and M. Popescu, Production of
Vermicular Graphite Cast Iron by Oper
ative Control and Correction of Graph
ite Shape, Paper 37 1, Proc. 46th Int.
Foundry Congress (Madrid, Spain),
CIATF, 1979
66. A. Velichko, C. Holzapfel, and F. Mück
lich, 3D Characterization of Graphite
Morphologies in Cast Iron, Adv. Eng.
Mater., Vol 9 (No. 1 2), 2007, p 39 45
67. H. Itofuji and H. Uchikawa, Formation
Mechanism of Chunky Graphite in
Heavy Section Ductile Cast Irons, Trans.
AFS, Vol 98, 1990, p 429 448
68. D.M. Stefanescu, G. Alonso, P. Larra
ñaga, E. De la Fuente, and R. Suarez,
On the Crystallization of Graphite from
Liquid Iron Carbon Silicon Melts, Acta
Mater., Vol 107, 2016, p 102 126
69. W. Oldfield, G.T. Geering, and W.A. Til
ler, Solidification of Spheroidal and Flake
Graphite Cast Iron, The Solidification of
Metals, Publication No. 110, Iron and
Steel Institute, London, 1967, p 256
70. J.P. Sodocha and J.E. Gruzleski, The
Mechanism of Graphite Spheroid Forma
tion in Pure Fe C Si Alloys, The Metal
lurgy of Cast Iron, B. Lux, I. Minkoff,
and F. Mollard, Ed., Georgi Publishing
Co., St. Saphorin, Switzerland, 1974,
p 443 459
71. B. Dhindaw and J.D. Verhoeven, Nodular
Graphite Formation in Vacuum Melted
High Purity Fe C Si Alloys, Metall.
Trans. A, Vol 11, 1980, p 1049 1057
72. R.W.G. Wyckoff, Crystal Structures, Vol
1, Interscience, New York, 1963
73. W. Bollman and B. Lux, Grain Bound
aries in Graphite, The Metallurgy of Cast
Iron, B. Lux, I. Minkoff, and F. Mollard,
Ed., Georgi Publishing Co., St. Saphorin,
Switzerland, 1975, p 461 470
74. K.I. Vashchenko and A.P. Rudoi, Surface
Tension of Cast Iron, Trans. AFS, Vol 70,
1962, p 855 864
75. R.H. McSwain and C.E. Bates, The
Metallurgy of Cast Iron, B. Lux, I.
Minkoff, and F. Mollard, Ed., Georgi
Publishing, St. Saphorin, Switzerland,
1975, p 423
76. W.C. Johnson and H.B. Smartt, in Solidi
fication and Casting of Metals, Book
192, The Metals Society, 1979, p 129
77. J.S. Park and J.D. Verhoeven, Metall.
Mater. Trans. A, Vol 27, 1996, p 2740
78. H.M. Muhmond and H. Fredriksson, Rela
tionship between the Trace Elements and
Graphite Growth Morphologies in Cast
Iron, Metall. Mater. Trans. A, Vol 45,
2014, p 6187 6199
79. D.M. Stefanescu, F. Martinez, and I.G.
Chen, Solidification Behavior of Hypoeu
tectic and Eutectic Compacted Graphite
Cast Irons, Chilling Tendency and Eutectic
Cells,AFS Trans., Vol 91, 1983, p 205 216
80. X.J. Deng, P.Y. Zhu, and Q.F. Liu, Struc
ture and Formation of Vermicular Graph
ite, AFS Trans., Vol 94, 1986, p 927 934
81. A. De Sy, Met. Prog., Vol 55, 1949, p 138
82. F.H. Buttner, H.F. Taylor, and J.Wulff, Am.
Foundryman, Vol 20 (No. 4), 1951, p 49
83. B. Marincek et al., Giesserei Techn. Wiss.
Beih., Vol 12, 1953, p 587
84. H. Geilenberg, Giesserei Techn. Wiss.
Beih., Vol 16, 1964, p 35
85. K. Herfurth, Investigations into the
Influence of Various Additions on the
Surface Tension of Liquid Cast Iron
with the Aim of Finding Relationships
between the Surface Tension and the
Occurrence of Various Forms of Graph
ite, Freib. Forschungsh., Vol 105,
1965, p 267
86. I. Minkoff and B. Lux, Graphite Growth
from Metallic Solution, The Metallurgy
of Cast Iron, B. Lux, I. Minkoff, and F.
Mollard, Ed., Georgi Publishing, 1975,
p 473
87. S. Amini and R. Abbaschian, Nucleation
and Growth Kinetics of Graphene Layers
from a Molten Phase, Carbon, Vol 51,
2013, p 110 123
88. K. Theuwissen, J. Lacaze, and L. Laffont,
Structure of Graphite Precipitates in Cast
Iron, Carbon, 2015, DOI: 10.1016/j.
carbon.2015.10.066
89. D.D. Double and A. Hellawell, Growth
Structure of Various Forms of Graphite,
The Metallurgy of Cast Iron, B. Lux, I.
Minkoff, and F. Mollard, Ed., Georgi
Publishing Co., St. Saphorin, Switzerland,
1975, p 509 528
90. D.D. Saratovkin, Dendritic Crystalliza
tion, Consultants Bureau, New York,
NY, 1959
91. D.M. Stefanescu, R. Huff, G. Alonso, P.
Larrañaga, E. De la Fuente, and R.
Suarez, On the Crystallization of Com
pacted and Chunky Graphite from Liquid
Multicomponent Iron Carbon Silicon
Based Melts, Metall. Mater. Trans., Vol
47, 2016, p 4012 4023, DOI: 10.1007/
s11661 016 3541 4, published online
May 5, 2016
92. I. Minkoff and I. Einbinder, Official
Exchange Paper Israel, International
Foundry Congress, 1963, p 139 143
93. A.N. Roviglione and J.D. Hermida, From
Flake to Nodular: A New Theory of Mor
phological Modification in Gray Cast
Iron, Metall. Mater. Trans. B, Vol 35,
2004, p 313 330
94. G. Alonso, D.M. Stefanescu, P. Larra
ñaga, and R. Suarez, Understanding Com
pacted Graphite Iron Solidification
through Interrupted Solidification Experi
ments, Int. J. Cast Met. Res., Vol 29
(No. 1 2), 2016, p 2 11, DOI: 10.1179/
1743133615Y. 0000000020, published
online March 8, 2016
95. G.R. Purdy and M. Audier, Electron
Microscopical Observations of Graph
ite in Cast Irons, The Physical Metal
lurgy of Cast Iron, H. Fredriksson
and M. Hillert, Ed. (Stockholm),
Microstructure Evolution during the Liquid/Solid Transformation in Cast Iron / 79
 
 
 
Materials Research Society Symposia
Proc., North Holland, New York,
1985, p 13 23
96. D. Ugarte, Nature, Vol 359, 1992, p 707
97. B. Lux, F. Mollard, and I. Minkoff, On
the Formation of Envelopes around
Graphite in Cast Iron, The Metallurgy of
Cast Iron, B. Lux, I. Minkoff, and F. Mol
lard, Ed., Georgi Publishing, 1975, p 371
98. H. Jones and W. Kurz, Growth Tempera
tures and the Limits of Coupled Growth
in Unidirectional Solidification of Fe C
Eutectic Alloys, Metall. Trans. A, Vol
11, 1980, p 1265
99. W.A.Oldfield,BCIRA J., Vol 9, 1960, p 177
100. E. Scheil and L. Hutter, Research on the
Crystallization of Nodular Graphite Iron,
Arch. Eisenh€uttenwes., May/June 1953,
p 237 246
101. S. Engler and R. Ellerbrok, On the
Growth of an Austenite Shell on the
Graphite Nodule during the Solidification
of Iron Carbon Silicon Melts, Giesserei
forschung, Vol 4, 1977, p 141 145
102. D.M. Stefanescu and B.K. Bandyopad
hyay, Physical Metallurgy of Cast Iron
IV, G. Ohira, T. Kusakawa, and E.
Niyama, Ed. (Tokyo), Materials Research
Society Proc., Pittsburgh, PA, 1989, p 15
26
103. S.E. Wetterfall, H. Fredriksson, and M.
Hillert, Solidification Process of Nodular
Cast Iron, J. Iron Steel Inst., May 1972,
p 323
104. A. Escobar, D. Celentano, M. Cruchaga,
J. Lacaze, B. Schulz, P. Dardati, and A.
Parada, Int. J. Cast Met. Res., Vol 27
(No. 3), 2014, p 176
105. R. Ruxanda, L. Beltran Sanchez, J. Mas
sone, and D.M. Stefanescu, On the Eutec
tic Solidification of Spheroidal Graphite
Iron: An Experimental and Mathematical
Modeling Approach, Trans. AFS,
Vol 109, 2001, p 1037 1048
106. H. Morrogh, Industri Information, No. 11,
Infinitas, Stockholm, 1961, p 35 72
107. F. Henke, Eutectic Solidification of Iron
Carbon Alloys, Giesserei Prax., Nov
1967, p 391 407
108. G.L. Rivera, R.E. Boeri, and J.A
Sikora, Adv. Mater. Res., Vol 4 5,
1997, p 169
109. Y.X. Li, B.C. Liu, and C.R. Loper, Study
of the Solid Liquid Interface during Uni
directional Solidification of Cast Iron,
AFS Trans., Vol. . . . . . . . . . . . . . . . . . . . . 146
Graphitization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Structural Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Cooling Curve (Thermal) Analysis
Ramón Suárez and P. Larrañaga, Maristas Azterlan Engineering
Adrián Udroiu, Metallurgical Quality Assistant . . . . . . . . . . . 149
Evaluation of Carbon Silicon Contents . . . . . . . . . . . . . . . 149
Evaluation of Graphite Shape . . . . . . . . . . . . . . . . . . . . . . 151
Evaluation of Graphite Nucleation . . . . . . . . . . . . . . . . . . 152
Chill Depth Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Amount of Primary Phase . . . . . . . . . . . . . . . . . . . . . . . . 153
Evaluation of Contraction Expansion Balance. . . . . . . . . . . 153
Graphite Distribution Type in Gray Iron . . . . . . . . . . . . . . 154
Tensile Strength and Hardness in Gray Iron . . . . . . . . . . . . 155
Prediction of Nodule Count . . . . . . . . . . . . . . . . . . . . . . . 155
Magnesium Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Control of Compacted Graphite Iron (CGI) . . . . . . . . . . . . 157
Ferrite/Pearlite in Ductile Iron . . . . . . . . . . . . . . . . . . . . . 158
Modification and Inoculation of Cast Iron
Iulian Riposan, Politehnica, University of Bucharest
Torbjorn Skaland, ELKEM Foundry Products . . . . . . . . . . . . . 160
Inoculation of Gray Cast Iron. . . . . . . . . . . . . . . . . . . . . . . . 162
Modification and Inoculation of Ductile Cast Iron . . . . . . . . . 166
Modification and Inoculation of Compacted Graphite
Cast Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Trace (Minor) Elements in Cast Irons
Robert Voigt, Pennsylvania State University . . . . . . . . . . . . . . 177
Effects of Minor Elements on Microstructure and Properties . . 177
Trace Element Testing and Control. . . . . . . . . . . . . . . . . . . . 178
Allowable Levels of Trace and Tramp Elements . . . . . . . . . . 178
Casting Processes
Filling and Feeding Systems for Cast Irons
John Campbell, University of Birmingham. . . . . . . . . . . . . . . . 182
Filling of Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Feeding of Ductile Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Molding and Casting Processes
John Campbell, University of Birmingham
József Tamás Svidró and Judit Svidró, J€onk€oping
University. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Aggregate Molding Materials . . . . . . . . . . . . . . . . . . . . . . . . 189
Binder Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Sand Reclamation Techniques . . . . . . . . . . . . . . . . . . . . . . . 197
Molding and Casting Processes . . . . . . . . . . . . . . . . . . . . . . 199
Surface Quality and Mold Metal Interface Interaction
Doru M. Stefanescu, The Ohio State University and The
University of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Physics and Chemistry of Mold Metal Interaction in
Iron Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
The Casting Skin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Metal Penetration in Sand Molds . . . . . . . . . . . . . . . . . . . . . 212
Computational Modeling of Gas Evolution in Sand Molds
Laurentiu Nastac, The University of Alabama . . . . . . . . . . . . . 218
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Numerical Model Description . . . . . . . . . . . . . . . . . . . . . . . 219
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Heat Treatment
Introduction to Cast Iron Heat Treatment
J.L. Dossett, Consultant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Stress Relief. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Annealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Normalizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Through Hardening and Tempering . . . . . . . . . . . . . . . . . . . 234
Surface Hardening of Cast Irons. . . . . . . . . . . . . . . . . . . . . . 236
Heat Treating of Gray Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Classes of Gray Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Stress Relief. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Examples of Stress Relief . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Annealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Normalizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Transformation Hardening . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Hardenability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Austenitizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Quenched and Tempered Properties . . . . . . . . . . . . . . . . . . . 249
Austempering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Martempering. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Flame Hardening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Induction Hardening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Other Surface Hardening Methods . . . . . . . . . . . . . . . . . . . . 255
Heat Treatment of Ductile Iron
Revised by K. Hayrynen, Applied Process, Inc. . . . . . . . . . . . . 256
Standards for Heat Treatment of Ductile Iron . . . . . . . . . . . . 256
General Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Austenitizing Ductile Cast Iron . . . . . . . . . . . . . . . . . . . . . . 258
Atmospheres for Heat Treatment of Ductile Iron . . . . . . . . . . 259
Annealing Ductile Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Hardenability of Ductile Cast Iron . . . . . . . . . . . . . . . . . . . . 261
Normalizing Ductile Iron. . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Quenching and Tempering Ductile Iron. . . . . . . . . . . . . . . . . 263
Marquenching (Martempering) Ductile Iron . . . . . . . . . . . . . . 263
Austempering Ductile Iron. . . . . . . . . . . . . . . . . . . . . . . . . . 264
Surface Hardening of Ductile Iron . . . . . . . . . . . . . . . . . . . . 265
Stress Relieving of Ductile Iron . . . . . . . . . . . . . . . . . . . . . . 268
Effect of Heat Treatment on Fatigue Strength . . . . . . . . . . . . 268
Heat Treatment of Malleable Irons
Edited by J.R. Keough and K.L. Hayrynen, Applied Process Inc. . . 270
Malleabilizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Hardening and Tempering . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Surface Hardening of Pearlitic Malleable Iron . . . . . . . . . . . . 273
Heat Treatment of High Alloy White Cast Irons
Revised by J.R. Keough and K.L. Hayrynen, Applied
Process Inc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
xii
Alloy Types and Properties . . . . . . . . . . . . . . . . . . . . . . . . . 275
Nickel Chromium White Irons . . . . . . . . . . . . . . . . . . . . . . . 275
High Chromium White Irons . . . . . . . . . . . . . . . . . . . . . . . . 277
Secondary Processing of Cast Iron. . . . . . . . . . . . . . . . . . . . . . 285
Welding of Cast Irons
Reviewed by Charles White, Kettering University . . . . . . . . . . . 287
Fusion Welds98, 1990, p 483 488
110. A. Rickert and S. Engler, Solidification
Morphology of Cast Irons, The Physical
Metallurgy of Cast Iron, Vol 34, H. Fre
driksson and M. Hillert, Ed., Proc. Materi
als Research Society, North Holland,
1985, p 165
111. K.D. Lakeland and L.M. Hogan, in Recent
Research on Cast Iron, H.D. Merchant,
Ed., Gordon and Breach, New York,
1968, p 417 448
112. D. Argo and J.E. Gruzleski, Mater. Sci.
Technol., Vol 10 (No. 2), 1986, p 1019
113. P. Magnin and W. Kurz, Metall. Trans. A,
Vol 19, 1988, p 1955 1963
114. G. Ohira, T. Sato, and Y. Sayama, in The
Metallurgy of Cast Iron, B. Lux, I. Mink
off, and F. Mollard, Ed., Georgi Publish
ing, St. Saphorin, Switzerland, 1974, p 295
115. D.M. Stefanescu and R. Ruxanda, in Met
allography and Microstructures, Vol 9,
ASM Handbook, ASM International,
2004, p 71 131
116. M. Hillert and H. Steinhauser, Jernkon
toret Ann., Vol 144, 1960, p 520
117. K.P. Bunin, I.N. Malinotchka, and I.N.
Taran, Osnovi Metallographyia Tchu
guna, Metallurghyia, Moscow, 1969
118. R.B. Gundlach, in Casting, Vol 15, ASM
Handbook, D.M. Stefanescu, Ed., ASM
International, Metals Park, OH, 1988,
p 678 685
119. M. Hillert and V.V. Subba Rao, “The
Solidification of Metals,” Publication
110, The Iron and Steel Institute, 1968
120. P. Magnin and W. Kurz, in The Physical
Metallurgy of Cast Iron, H. Fredriksson
and M. Hillert, Ed., North Holland, New
York, 1985, p 263
121. E. Fras and H.F. Lopez, Trans. AFS, Vol
101, 1993, p 355
80 / Fundamentals of the Metallurgy of Cast Iron
 
 
 
Principles of Thermal Analysis
Hasse Fredriksson, KTH Stockholm
Doru M. Stefanescu, The Ohio State University and The University of Alabama
THERMAL ANALYSIS (TA) is by defini
tion a method used to study phase transforma
tions in material. It often is used to analyze
solidification processes by recording the tem
perature as a function of time during cooling
or heating of a metal or alloy to or from a tem
perature above its melting point. The first scien
tist who documented that the melting point
could be found from the recording of the tem
perature during heating of metals was Le
Chatelier (Ref 1). Over the following 20 years,
the method was improved by several research
ers (Ref 2 4). They developed a method that
consisted of recording the temperature of a
dummy (reference) sample together with the
alloy investigated. The difference in tempera
ture between the dummy and the investigated
alloy was recorded. This method is called dif
ferential thermal analysis (DTA) and is a stan
dard method for determining several physical
parameters, such as solidification or melting
temperature, heat capacities, and heat of fusion.
It also is used to determine phase diagrams. For
metalcasting applications, DTA was simplified
by using a single sample and performing a com
puter analysis of the cooling curve by simulat
ing the reference sample. In this case, the
reference sample is assumed to have the same
heat capacity as the investigated alloy but with
out any phase transformation (Ref 5 7).
Today (2017), thermal analysis is not only
used for research but also extensively in indus
trial production. In foundries, it is used to con
trol the structure and composition of cast iron
melts. This article describes the use of cooling
curves for analyzing a solidification process,
such as the solidification temperature, structure
analysis, fraction of phases and heat of fusion
with focus on solidification of cast iron, and
the use of cooling curves to control and adjust
the casting conditions.
Basics of Cooling Curves
Cooling curves describe a balance between
the evolution of heat in the sample and the heat
transport away from the sample. A cooling
curve of a cast iron sample with approximately
3.8% C is shown in Fig. 1 (Ref 8) together with
the phase diagram of iron carbon alloys
(Fig. 2). With the help of the phase diagram,
the cooling curve can be divided into four
regions. Region I extends from the initial tem
perature of the superheated region of the melt
to the temperature at which solidification starts
by precipitation of austenite, which is the start
of region II. The slope of the curve in region I
is decided by the heat extraction and the heat
capacity. The change of the slope in region II
can be calculated from the product between
the fraction of solid formed and the latent heat.
Region II extends to the temperature for the
start of the eutectic reaction, region III.
This region continues until the sample has soli
dified, which gives the solidification time.
When the solidification process is finished, the
temperature starts to decrease again; this is
region IV.
The standard terminology used in cooling
curve analysis is introduced in Fig. 3 as follows:
TL, equilibrium liquidus temperature; TE, equi
librium eutectic temperature; TLA, temperature
of liquidus arrest; TEmin, temperature of eutectic
undercooling; TEmax, temperature of eutectic
recalescence; DT, recalescence; DTmax, maxi
mum undercooling; and DTmin, minimum under
cooling (Ref 9).
The heat extraction, dQ/dt, describes the heat
transport away from the sample. For large sam
ples, this often is described by numerical pro
grams. However, for small samples, as is often
used during TA, assuming that the sample is
isothermal, rather simple expressions can be
found for the four different regions presented
in Fig. 1. For region I, the following relation
is found:
dQ
dt
vo rmetal
dT
dt
cmetal
p (Eq 1)
where dQ/dt is the amount of heat emitted from
the sample per unit time, vo is the volume of the
sample, rmetal is the density of the sample, T is
temperature, t is time, dT/dt is the cooling rate,
and cmetal
p is the heat capacity of the metal. If
the heat capacity and the sample weight are
known, and if the cooling rate is recorded, the
heat extraction from the sample can be
calculated and used to analyze the solidification
process in regions I and IV.
In region II, the following relation is
obtained:
dQ
dt
vo rmetal
dT
dt
cmetal
p þ ð �HÞ dfs
dT
� �
(Eq 2)
where DH is the heat of fusion of the sample
material (J/kg), fs is the fraction solid, and
dfs/dT is the fraction of solid formed upon tem
perature decrease. The fraction of solid then is
found by numerical integration of the experi
mental curve.
The volume fraction solidified during the
eutectic reaction, region III, at time t can be
calculated with the following relation:
dQ
dt
vo rmetalð �HÞ dfs
dt
(Eq 3)
During the eutectic reaction, most often one can
disregard the temperature change of the sample,
because the heat of fusion is several hundred
times larger than the heat capacity.
For region IV, a similar relation as the one pre
sented in Eq 1 can be found. However, one should
notice that for cast iron, graphite also is precipi
tated during the cooling process, due to a decrease
of the solubility of carbon in austenite. In a care
ful analysis, the heat evolved by this process must
be added to the heat capacity term.
ASM Handbook, Volume 1A, Cast Iron Science and Technology 
D.M. Stefanescu, editor
DOI: 10.31399/asm.hb.v01a.a0006299
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
1000
1030
1060
1090
1120
1180
1240
1300
1270
1210
1150
0 8040 120 160 200
TL
γ
TE
Te
m
pe
ra
tu
re
, °
F
Te
m
pe
ra
tu
re
, °
C
Time, s
2370
2320
2260
2210
2160
2100
2050
1990
1940
1890
1830
Fig. 1 Cooling curve for an Fe-3.8%C alloy. Source:
Ref 8
 
 
 
The preceding mathematical treatment of the
cooling curve is called Newtonian analysis. In
this approach, it is assumed that the thermal
gradient across the sample is zero and that heat
transfer between the casting and the mold
occurs by convection. Another more accurate
treatment of the heat transfer problem is Four
ier analysis. It requires two thermocouples,
and the mathematics is more cumbersome
(Ref 9 12).
A significant amount of information can be
found from a cooling curve, such as the melt
ing point, the temperature for start of solidifi
cation,. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Weldability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Welding Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
Base Metal Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Repair Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Shielded Metal Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . 296
Gas Metal Arc Welding. . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Flux Cored Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Gas Tungsten Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . 302
Submerged Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
Oxyfuel (Gas) Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Flame Spraying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Braze Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Other Fusion Welding Methods . . . . . . . . . . . . . . . . . . . . . . 308
Solid State Welding Methods . . . . . . . . . . . . . . . . . . . . . . . . 308
Surfacing and Overlaying . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Brazing and Soldering of Cast Irons
Reviewed by Charles White, Kettering University . . . . . . . . . . . 310
Soldering of Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Brazing of Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Brazing Filler Metal Selection . . . . . . . . . . . . . . . . . . . . . . . 311
Cleaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Fixturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Brazing Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Heating Methods for Brazing . . . . . . . . . . . . . . . . . . . . . . . . 316
Application Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Machining
Simon N. Lekakh, Missouri University of Science and Technology
Dika Handayani, Pennsylvania State University
Michael E. Finn, Finn Metalworking and Cutting Solutions . . . 319
Cutting Cast Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Machinability Test Methods. . . . . . . . . . . . . . . . . . . . . . . . . 320
Effect of As Cast Surface Integrity . . . . . . . . . . . . . . . . . . . . 322
Effect of Microstructure on Machinability . . . . . . . . . . . . . . . 323
Spheroidal Graphite Iron Machinability . . . . . . . . . . . . . . . . . 325
Machining Austempered Ductile Irons . . . . . . . . . . . . . . . . . 326
Cutting Tool for Machining Gray Irons . . . . . . . . . . . . . . . . . 327
Machining Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Cutting Lubrication (Coolants) . . . . . . . . . . . . . . . . . . . . . . . 333
Dry Machining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Cleaning and Coating of Cast Irons . . . . . . . . . . . . . . . . . . . . . . 335
General Cleaning of Castings . . . . . . . . . . . . . . . . . . . . . . . . 335
Mechanical Cleaning and Finishing . . . . . . . . . . . . . . . . . . 335
Nonmechanical Cleaning . . . . . . . . . . . . . . . . . . . . . . . . . 337
Cast Iron Organic Coatings
Jayson L. Helsel and Kenneth B. Tator, KTA Tator, Inc. . . . . 338
General Surface Preparation . . . . . . . . . . . . . . . . . . . . . . . 339
Coatings for Atmospheric Exposure . . . . . . . . . . . . . . . . . 341
Architectural Cast Iron Protection . . . . . . . . . . . . . . . . . . . 342
Exterior Coatings for Underground Service . . . . . . . . . . . . 343
Interior Coatings for Underground Service . . . . . . . . . . . . . 344
Repairs/Replacement of Deteriorated Pipe . . . . . . . . . . . . . 345
Cast Iron Inorganic Coatings . . . . . . . . . . . . . . . . . . . . . . . . 347
Electroplating and Electroless Plating . . . . . . . . . . . . . . . . 347
Hot Dip Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Hardfacing and Weld Cladding . . . . . . . . . . . . . . . . . . . . . 351
Thermal Spraying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Conversion Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Porcelain Enameling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Inspection and Quality Control
Casting Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
Common Defects in Ductile Cast Irons . . . . . . . . . . . . . . . . . 361
Defects in Gray Iron Castings . . . . . . . . . . . . . . . . . . . . . . . 363
Surface Defects in Compacted Graphite Iron:
Casting Skin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Examples of Defects in Cast Irons . . . . . . . . . . . . . . . . . . . . 372
Nondestructive Inspection of Cast Irons
John A. Griffin, The University of Alabama . . . . . . . . . . . . . . . 373
The Role of Nondestructive Inspection . . . . . . . . . . . . . . . . . 373
Surface/Near Surface Inspection Methods . . . . . . . . . . . . . . . 373
Volumetric Inspection Methods . . . . . . . . . . . . . . . . . . . . . . 376
Metallography and Microstructures of Cast Iron
Janina M. Radzikowska, The Foundry Research Institute (retired),
Kraków, Poland
George Vander Voort, Struers Inc. (Consultant) . . . . . . . . . . . . 379
Sampling and Specimen Preparation . . . . . . . . . . . . . . . . . . . 379
Grinding and Polishing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Etching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Illumination Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
Fracture Analysis
Diego O. Fernandino and Roberto E. Boeri, National University
of Mar del Plata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Fracture Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Special Cases of Environmentally Assisted Fracture . . . . . . . . 407
Identifying Crack Propagation Direction Using
Fractographic Features . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
Properties of Cast Irons and Effects of Processing . . . . . . . . . . 411
Physical Properties of Cast Irons
Doru M. Stefanescu, The Ohio State University and The University
of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Properties of Liquid Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Thermal Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Conductive Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
Magnetic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Acoustic Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Mechanical Properties of Gray Irons . . . . . . . . . . . . . . . . . . . . . . 430
Classification of Gray Irons . . . . . . . . . . . . . . . . . . . . . . . . . 431
Test Bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Hardness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Elasticity and Deformation . . . . . . . . . . . . . . . . . . . . . . . . . 434
Strength and Ductility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Fatigue Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Stress Life (High Cycle) Fatigue . . . . . . . . . . . . . . . . . . . . . 439
Strain Life (Low Cycle) Fatigue. . . . . . . . . . . . . . . . . . . . . . 446
Fatigue Crack Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Toughness . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . 449
High Temperature Strength . . . . . . . . . . . . . . . . . . . . . . . . . 451
Mechanical Properties of Ductile Irons . . . . . . . . . . . . . . . . . . . . 456
Classes and Grades of Ductile Iron . . . . . . . . . . . . . . . . . . . . 456
Factors Affecting Mechanical Properties . . . . . . . . . . . . . . . . 457
Hardness Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Tensile Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Shear and Torsional Properties . . . . . . . . . . . . . . . . . . . . . . . 462
Damping Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Compressive Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Fatigue Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
Fracture Toughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Impact Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Austempered Ductile Iron . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Mechanical Properties of Compacted Graphite Iron . . . . . . . . . . . 472
Tensile Properties and Hardness . . . . . . . . . . . . . . . . . . . . . . 472
xiii
Compressive and Shear Properties . . . . . . . . . . . . . . . . . . . . 475
Modulus of Elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
Impact Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
Fatigue Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
Elevated Temperature Properties . . . . . . . . . . . . . . . . . . . . . 478
Mechanical Properties of Malleable Irons . . . . . . . . . . . . . . . . . . 481
Classification of Malleable Irons . . . . . . . . . . . . . . . . . . . . . 481
Summary of Grade Mechanical Properties . . . . . . . . . . . . . . . 482
Ferritic Malleable Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Pearlitic and Martensitic Malleable Iron . . . . . . . . . . . . . . . . 483
Damping Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
Wear of Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
General Wear Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 489
Abrasion Resistant Cast Irons. . . . . . . . . . . . . . . . . . . . . . . . 490
Brake Drum and Disk Wear. . . . . . . . . . . . . . . . . . . . . . . . . 497
Piston Rings and Cylinder Liners . . . . . . . . . . . . . . . . . . . . . 499
Grinding Balls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Corrosion of Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
Influence of Alloying and Microstructure . . . . . . . . . . . . . . . 502
Commercially Available Cast Irons. . . . . . . . . . . . . . . . . . . . 503
Corrosion Resistant Cast Irons . . . . . . . . . . . . . . . . . . . . . . . 503
Forms of Corrosion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
Resistance to Corrosive Environments. . . . . . . . . . . . . . . . . . 507
Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
Internal Casting Stresses and Dimensional Stability
Tito Andriollo, Nikolaj Vedel Smith, Jesper Thorborg, and Jesper
Hattel, Technical University of Denmark . . . . . . . . . . . . . . . 511
Macroscopic Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
Dimensional Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
Microscopic Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
Computer Aided Prediction of Mechanical Properties
Ingvar L. Svensson and Jakob Olofsson, J€onk€oping University. . 516
Characterization and Modeling of Microstructure Based
Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Modeling of Hardness in Cast Iron . . . . . . . . . . . . . . . . . . . . 516
Elastic Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Understanding Deformation Behavior Using the Tensile
Test Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Evaluation of Material Parameters from the Tensile
Stress Strain Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
Using Models to Describe the Tensile Test Curve . . . . . . . . . 519
Methods for Evaluating Plastic Deformation . . . . . . . . . . . . . 519
Fitting Model Parameters to Experimental Tensile
Stress Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Shape of Tensile Test Curve and Influences on n and k . . . . . 520
Nature of the Deformation Behavior . . . . . . . . . . . . . . . . . . . 521
Component Behavior Dependent on Microstructure and
Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Computer Aided Prediction of Mechanical Behavior on the
Microstructural Level . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
Computer Aided Prediction of Mechanical Behavior of Cast
Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
Gray Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
Specifications and Selection of Gray Irons
Doru M. Stefanescu, The Ohio State University and The University
of Alabama
Tom Prucha, American Foundry Society . . . . . . . . . . . . . . . . . 527
Classes of Gray Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Effect of Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Effect of Chemical Composition. . . . . . . . . . . . . . . . . . . . . . 530
Effect of Cooling Rate (Section Sensitivity) . . . . . . . . . . . . . 532
Tensile Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
Brinell Hardness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
Transverse Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
Compressive Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
Fatigue Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
Fracture Toughness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
Elevated Temperature Strength. . . . . . . . . . . . . . . . . . . . . . . 542
Dimensional Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Thermal Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
Low Temperature Properties . . . . . . . . . . . . . . . . . . . . . . . . 546
Damping Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
Pressure Tightness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Machinability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Wear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Corrosion Resistant Gray Irons. . . . . . . . . . . . . . . . . . . . . . . 553
Heat Resistant Gray Irons . . . . . . . . . . . . . . . . . . . . . . . . . . 555
Production of Gray Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . 561
Castability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
Cupola Melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564
Induction Melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
Arc Furnace Melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Composition Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Inoculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
Gray Iron Alloying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . 573
Thin Wall Gray Iron Castings
Marcin Górny, AGH University of Science and
Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Cooling Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Solidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Macrostructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575
Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
Chilling Tendency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
Metallic Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Production and Application of Thin Wall Gray
Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
Microstructures and Characterization of Gray Irons
Attila Diószegi and Lucian Vasile Diaconu, J€onk€oping University,
Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
Macrostructure of Primary Austenite. . . . . . . . . . . . . . . . . . . 583
Dendrite Morphology of Primary Austenite . . . . . . . . . . . . . . 584
Eutectic Cells (Colonies) . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
As Cast Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
Ductile Iron Castings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Specification and Selection of Ductile Irons
Niels Skat Tiedje, Technical University of Denmark . . . . . . . . . 593
Designation of Ductile Cast Irons . . . . . . . . . . . . . . . . . . . . . 594
Standard Grade Ductile Cast Irons . . . . . . . . . . . . . . . . . . . . 594
Low Alloy Ductile Cast Irons for Use at Elevated
Temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Austenitic Ductile Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . 596
Austempered Ductile Cast Irons . . . . . . . . . . . . . . . . . . . . . . 596
Castability and Product Design of Ductile Iron
Niels Skat Tiedje, Technical University of Denmark . . . . . . . . . 598
Dimensions, Tolerances, and Precision . . . . . . . . . . . . . . . . . 598
Design for Casting Castability . . . . . . . . . . . . . . . . . . . . . . 598
Solidification Shrinkage. . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
Strength and Stiffness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
Thermal Deformation and Residual Stress . . . . . . . . . . . . . . . 600
Production of Ductile Iron Castings
Revised by Douglas White, Elkem Materials, Inc. . . . . . . . . . . . 603
Raw Materials for Ductile Iron Production . . . . . . . . . . . . . . 603
Control of the Composition of Ductile Iron . . . . . . . . . . . . . . 604
Molten Metal Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
Casting and Solidification . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Metallurgical Controls of Ductile Iron Production . . . . . . . . . 609
xiv
Austempered Ductile Iron Castings. . . . . . . . . . . . . . . . . . . . . . . 612
Production of Austempered Ductile Iron . . . . . . . . . . . . . . . . 612
Composition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Applications for Austempered Ductile Iron . . . . . . . . . . . . . . 615
Thin Wall Ductile Iron Castings
Marcin Górny, AGH University of Science and
Technology
Doru M. Stefanescu, The Ohio State University and
The University of Alabama . . . . . . . . . . . . . . . . . . . . . . . . . 617
Cooling Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Solidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Production and Applications of Thin Wall Ductile
Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
Solidification Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
Heavy Section Ductile Iron Castings
Chantal Labrecque, Serge Grenier, Pierre Marie Cabanne, and
Martin Gagné, Rio Tinto Iron and Titanium . . . . . . . . . . . . . 629
Heavy Section Ductile Iron Production Process . . . . . . . . . . . 629
Market and Applications for Ductile Iron . . . . . . . . . . . . . . . 632
Testing and Standard Requirements . . . . . . . . . . . . . . . . . . . 633
Properties and Microstructures . . . . . . . . . . . . . . . . . . . . . . . 634
Industrial Examples Case Analyses . . . . . . . . . . . . . . . . . . . . 635
Quality Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638
Defects in Heavy Section Ductile Iron Castings . . . . . . . . . . . 639
Metallography and Microstructures of Ductile Irons
George F. Vander Voort, Struers Inc.
Juan Asensio Lozano, University of Oviedo
(Asturias, Spain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
Metallographic Preparation . . . . . . . . . . . . . . . . . . . . . . . . . 645
Grinding and Polishing . . . . . . . . . . . . . . . . . . . . . . . . . . 645
Etchants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
Microstructures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
Graphite Morphology. . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
Nodularity of Graphite in Ductile Iron. . . . . . . . . . . . . . . . 649
Image Analysis of Nodularity Ratings . . . . . . . . . . . . . . . . 653
Matrix Microstructures of Ductile Irons . . . . . . . . . . . . . . . 653
Compacted Graphite (CG) Iron Castings . . . . . . . . . . . . . . . . . 657
Specification, Selection, and Applications of Compacted Graphite Irons
Steve Dawson, SinterCast Limited
Wilson Guesser, Tupy S.A. and State University of
Santa Catarina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Graphite Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Mechanical and Physical Properties . . . . . . . . . . . . . . . . . . . 660
International Specifications . . . . . . . . . . . . . . . . . . . . . . . . . 660
Applications of Compacted Graphite Iron Castings. . . . . . . . . 661
Castability, Product Design, and Production of Compacted Graphite Irons
Steve Dawson, SinterCast Limited
Wilson Guesser, Tupy S.A. and State University of
Santa Catarina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
Castability and Product Design of Compacted
Graphite Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
Production of Compacted Graphite Iron Castings . . . . . . . . . . 669
Microstructure and Characterization of Compacted Graphite Iron
Steve Dawson, SinterCast Limited
Wilson Guesser, Tupy S.A. and State University of
Santa Catarina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
Tensile Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
Hardness and Compressive Properties . . . . . . . . . . . . . . . . . . 678
Impact Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
Fatigue and Fracture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
Thermal Conductivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683
High-Alloy Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
Specification, Selection, and Applications of High Alloy
Iron Castings
Richard B. Gundlach, Element Materials Technology
Harry Tian, GIW Industries, Inc.
Brian Bendig, Penticton Foundry Ltd. . . . . . . . . . . . . . . . . . . . 689
Specification and Selectionof High Alloy Irons . . . . . . . . . . . 689
High Alloy Graphitic Irons . . . . . . . . . . . . . . . . . . . . . . . . . 691
High Alloy White Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
Applications of High Alloy Graphitic Irons . . . . . . . . . . . . . . 702
Applications of High Alloy White Irons . . . . . . . . . . . . . . . . 704
Castability, Product Design, and Production of High Alloy
Iron Castings
Harry Tian, GIW Industries, Inc.
Richard B. Gundlach, Element Materials Technology . . . . . . . . 708
Castability of High Alloy Iron Castings. . . . . . . . . . . . . . . . . 708
Product Design and Processing Factors . . . . . . . . . . . . . . . . . 710
Production of High Alloy Iron Castings . . . . . . . . . . . . . . . . 711
High Alloy Graphitic Irons . . . . . . . . . . . . . . . . . . . . . . . . . 711
High Alloy White Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
Heat Treating High Alloy Iron Castings . . . . . . . . . . . . . . . . 715
Machining and Finishing High Alloy Iron Castings . . . . . . . . 717
Microstructure and Characterization of High Alloy Cast Irons
George F. Vander Voort, Struers Inc.
Juan Asensio Lozano, University of Oviedo (Asturias, Spain) . . 719
Specimen Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719
Grinding and Polishing Procedures . . . . . . . . . . . . . . . . . . 720
Etchants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Microstructures of Austenitic High Alloy Gray Iron . . . . . . . . 721
Microstructures of Austenitic High Alloy Ductile Iron . . . . . . 722
Microstructures of Corrosion Resistant High Silicon
Cast Irons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
Microstructures of Abrasion Resistant Cast Irons . . . . . . . . . . 726
Malleable Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
Malleable Iron Castings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
Melting Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
Current Production Technologies . . . . . . . . . . . . . . . . . . . . . 738
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Ferritic Malleable Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Pearlitic and Martensitic Malleable Iron . . . . . . . . . . . . . . . . 741
Reference Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Abbreviations and Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
xv
A History of Cast Iron
Doru M. Stefanescu, The Ohio State University and The University of Alabama
THE STORY OF CAST IRON is an intricate
part of the saga of metal casting, a glamorous,
fascinating story whose beginnings are traced
to the dawn of human civilization. It is interwo
ven with legends of fantastic weapons and
exquisite artworks, and, as such, it was and is
involved in the two main activities of humans
since they began walking the planet Earth: pro
ducing and stealing/defending wealth. Casting
of iron has emerged from the darkness of antiq
uity, first as magic, later to evolve into an art,
then into a technology, and finally as the com
plex, interdisciplinary science that it is today
(2016). Civilization as we know it would not
have been possible without metal casting in gen
eral and without iron casting in particular. This
article takes the reader through a short time
travel of the evolution of cast iron from witch
craft to virtual cast iron, a road paralleled by
the gigantic stride from a low quality, “corrupt
metal” to the high tech material that it is today.
The Beginnings of Metal Casting
and of the Iron Age
A history incursion in any subject matter
starts with the “who was the first to. . .?” query.
Thus, we should wonder who poured the first
casting and how did this casting look? It
appears that the birthplace of metals can be
traced with some accuracy to the area north of
the Black Sea in the Carpathian Mountains in
today’s Romania, as shown by the arrow in
Fig. 1 (Ref 1). Other sources place the begin
ning in southeastern and central Anatolia,
where shaped copper objects dating from circa
8200 B.C. were found (Ref 2). Our ancestors
started the long road in the mastery of metal
fabrication and use with wrought native copper.
It was not until approximately 5000 B.C. that
metal casting was invented, as humans learned
to melt and cast copper.
A partial chronological list of the progress
achieved by human civilization in the use of
metals and, in particular, of cast iron before the
Modern Era is provided in Table 1 (Ref 2 4).
For a more complete list of metal casting devel
opments, the reader is referred to Ref 5.
The beginning of the iron civilization (Iron
Age) is still subject to controversy. The use of
iron was delayed compared to copper, because
of its lack of availability as a native metal.
Some archeological findings place it at approx
imately 6000 B.C. in Mesopotamia, while
surveys in Anatolia dated it with confidence to
3000 B.C. (Ref 6). Primitive people appear to
have worked with meteoric iron long before
learning to extract iron from iron ore. The
ASM Handbook, Volume 1A, Cast Iron Science and Technology 
D.M. Stefanescu, editor
DOI: 10.31399/asm.hb.v01a.a0006320
Copyright # 2017 ASM InternationalW
All rights reserved
www.asminternational.org
Fig. 1 Birthplace of metals. Source: Ref 1
Table 1 Chronological list of developments and use of cast iron during prehistory,
antiquity, and the medieval ages
Date
Development
LocationPrehistory and antiquity (B.C.)
9000 B.C. Earliest metal objects of wrought native copper Near East,
Anatolia
5000–3000 B.C. Chalcolithic period: melting of copper Near East
3000–1500 B.C. Bronze Age: arsenical copper and tin bronze alloys cast in stone molds Near East
3200 B.C. The oldest casting in existence, a copper frog Mesopotamia
3000 B.C. Iron Age: wrought iron Near East
2000 B.C. Two-part axe head bronze mold in Macon France
600 B.C. First iron casting, a 70 kg (154 lb) tripod China
280 B.C. Colossus of Rhodes built with an iron framework plated with brass to create the skin and
outer structure of Helios. At 30 m (98.4 ft) high, it was one of the tallest statues of the
ancient world.
Greece
233 B.C. Iron plowshares are cast. China
Medieval Ages (5 to 15th century A.D.)
~1122 Theophilus’s On Divers Arts, the first monograph on metalworking Germany
1313 Castings produced from furnace pig iron Germany
~1500 The Basilicas, first famous cast iron gun England
1540 Vannoccio Biringuccio’s De la pirotechnia, published posthumously; the first printed
account of proper foundry practice
Italy
 
 
 
Sumerian word “AN.BAR,” the oldest word
designating iron, is made up of the pictograms
“sky” and “fire.” Similar terminology is found
in Egypt, “metal from heaven,” and with the
Hittites, “black iron from sky” (Ref 7). Genesis
4:22 records that Tubal cain (ancestry line:
Adam, Cain, Enoch, Irad, Mehujael, Methusael,
Lamech, Tubal cain) was the “forger of all
instruments of bronze and iron” or an “instruc
tor of every artificer in brass and iron” and thus
a metalsmith. It is further suggested (Ref 8) that
he “discovered the possibilities of cold forging
native copper and meteoric iron.” In most
ancient cultures, this sky connection led to the
belief that the metallurgist had a direct link to
the divine, if not of divine origin himself. It ele
vated the social status of the early metallurgist
in the tribal hierarchy to that of a chief or sha
man. Metalworkers sometimes rose to the level
of royalty. Genghis Khan was a simple smith
before acceding to power. Even in later history,
the metal worker ranked highly in the social
hierarchy. In Ireland, foundrymen ranked with
the nobility from early