CHINA FOUNDRY Vol.6 No.

Colour Metallography of Cast Iron

By Zhou Jiyang, Professor, Dalian University of Technology, China
Translated by Ph.D Liu Jincheng, Fellow of Institute of Cast Metal Engineers,UK

*Note: This book consists of five sections: Chapter 1 Introduction, Chapter 2 Grey Iron, Chapter 3 Ductile Iron, Chapter
4 Vermicular Cast Iron, and Chapter 5 White Cast Iron. CHINA FOUNDRY publishs this book in several parts serially,
starting from the first issue of 2009.

Chapter 1
Introduction (Ⅱ)
1.5 Colour metallography technique
of cast iron
Colour metallography possesses better differentiation ability and
is more sensitive to segregation, grain orientation and stress state
than black-white metallography. It adds functions to traditional
metallography and displays wide application perspectives.
Colour metallography technique belongs interference film
metallography. By using chemical or physical methods, a
thin interference film can be formed on the surface of alloy
metallographic specimen; with the aid of extinction effect in thin
film interference, the microstructures are displayed with different Fig. 1-24: Interference effect of air-thin film-metal
colours. Colour metallography methods producing interference
film-metal interfaces, reflection and refraction occur. The path
film include polarized light, chemical deposition, constant
and phase of reflected ray and refracted ray each experienced are
potential, vacuum coating, ion sputtering and heat tinting [21].
different. When the two rays meet, the reflected ray produces thin
Among them, chemical deposition (also called etching method) is
film interference effect, the intensity of reflected ray is weakened,
simple, convenient, therefore obtains wide application [22].
even disappeared, that is, destructive effect. If no absorption and
The colour metallography technique used in this book is a hot
destructive effects for certain wavelength light, men will not see
alkaline etching display method, belonging category of chemical
colours.
deposition interference film. In this method, metallographic
Optical phase has important effect on destructive interference
sample is put into hot sodium hydrate solution and heated. When
effect; the phase condition of destructive effect is influenced
solution reaches a certain temperature, hold the sample for certain
by inherent optical parameter of metal structure, and more
time at the temperature, take out, clean and dry it; microstructures
importantly is determined by thickness of a film. Different film
with different colours can be observed under microscope. In the
thickness produces different destructive effect. For example, under
early time, this method was used in alloy steel, white iron to colour
white lighting, if a certain film thickness of metal structure is just
carbide and phosphide, and in the study of segregation within grain
destructive with blue light wavelength (440 µm), thus, reflected
in grey iron [23-27]. From the middle of 1980s, the author [28-30] and
other researchers [31-34] have continuously improved and perfected light consists of mixed lights without blue light. At this time the
this hot alkaline etching method and gradually developed this reflected light is not a white light, but displays another wavelength
method into a method for studying the solidification structure and colour. Under the same principle, different film thicknesses are
phase transformation of cast iron. produced on the same structure with un-uniformed composition
caused by segregation; thus different locations of the same phase
1.5.1 Principle of hot alkaline etching method can display different colours.
Hot alkaline etching method can cause cast iron to produce
Using chemical method to deposit a thin film onto a metal surface,
different thickness films on the sample surface with different local
an optical interference effect occurs through the film. This is the
composition and structure. When the metallographic sample is
fundamentals of optics for hot alkaline method displaying colour.
merged into alkaline solution, SiO2 formed by Si in cast iron will
The principle of optical interference effect is described in
react with NaOH following:
Fig.1-24. Suppose a ray of light consisting of continuous spectrum
2NaOH + SiO2 = Na2SiO3·H2O (colloid)
is incident on an air-film-metal system. The white ray incident to
the surface of the film passes through two interfaces, air-film and The picronitric acid in alkaline solution, C 6 H 2 (NO 2 )OH,

152
 Serial Report
May 2009

simultaneously reacts with NaOH:

C6H2(NO2)3OH + NaOH = C6H2(NO2)3ONa + H20
Sodium metasilicate Na2SiO3·H2O, as colloid, is absorbed and
deposited onto the surface of sample and formed a thin film. For
silicon rich region containing more SiO2, the film forms quicker
and thicker; for silicon poor region, the formed film is thinner.
In the end, films with different thickness are formed. Except for
Si, segregation of other elements also accelerates the forming of
different thick films.

1.5.2 Characteristics of colour metallography with

hot alkaline etching
Based on author's many years of practical experiences, colour
metallography using hot alkaline etching technique in studying
solidification of cast iron has following advantages:
(1) Can display solidification development process
The solidification development process displayed by colour
metallography with hot alkaline etching method is achieved with
etching austenite. The negative segregation of Si in austenite
is displayed by etching austenite. Si is an element enriching in
austenite. According to the phase diagram, the early crystallized
austenite forms at high temperature containing lower carbon
content; high temperature is beneficial for enhancing diffusion
rate of Si; while low carbon can decrease the repulsion force
of Si→C, thus, is favorable for Fe-Si forming strong chemical
bond, causing Si atoms easily to migrate towards the inside of
austenite grain. Therefore, the earlier the austenite forms, the
higher the Si content; the later the austenite forms, the lower
the Si content. Beside, the Si content of austenite has a direct
relation with colours. At constant etching condition, Si directly
influences the film thickness, thus displays different colours. The
relations between Si content and qualitative or semi-quantitative
measurement by colours for some actually examples are shown
in Fig. 4-37 and Table 4-3 (in Chapter 4). By connecting and
relating the time austenite starting to solidify, Si content and the
displayed colours, Table 1-3 is obtained. Since etching process
and colour photo developing conditions are different, the obtained
general colour tones of metallographic photos are different(see
Fig. 1-25), but the colour variation sequence is the same with the
variation of Si content. Some time, certain colour is not obviously
displayed; after a longer time etching or using solution with （a) tone 1 （b) tone 2 （c) tone 3
higher alkalinity, all the colours in Table 1-3 can be observed. If
etching time is short, the colours presented in Table 1-3 will not Fig. 1-25: Relation between Si content of solidified
start from serial number 1. The shorter the etching time, the lower structure and colour under different colour tones.

Table 1-3: Relation of colour, Si content and solidification time

Number Colour Silicon content Time to solidify Etching time
1 Yellow high early long
2 Blue green
3 Purple
4 Orange red
5 Pale yellow
6 Pale blue
7 Blue
8 Brown
9 Cream white or pale yellow low late short

153
 CHINA FOUNDRY Vol.6 No.2

the produced colour sequence moves to. Thus, as long as control interference film, there are no interference films deposited on these
the etching time in alkaline solution, gradually etch and show the cathode (positive) region, thus these structures are not easy to be
colour, operate in steps, the colour variation can be observed at a coloured in normal alkaline solution.
particularly selected position. With increasing etching time, colour In addition to caustic alkali, a certain amount of picronitric
changes following sequential number 9 →1. Then, check the acid is added in hot alkali solution. Picronitric acid is a kind
colour in Table 1-3 and find the colour which is the same or close of weak acid which can increase concentration of H + in the
to it to determine the solidification time at that position and to solution, change the stable potential of anode and electrochemical
compare with solidification progress of other positions. In addition, polarisation, causing cathode, which originally cannot be coloured,
using this technique, it can also study solidification according to to be coloured. Experiments [36] have confirmed that if the reagent
colours, rather the fixed position. The region having the same colour is well prepared and etching time and temperature is properly
will start to solidify at the same time: the closer to the region shown controlled, these cathode phase structures in alkali solution can
by the serial number 1, the earlier the time starting to solidify. also be coloured and displayed. If no picronitric acid added in the
Therefore, the solidification information on nucleation, growth and hot alkali solution, pearlite, ferrite and carbide structure cannot be
relations between solidification structures can be obtained from the displayed
types, distribution and shape characteristics of colours. The above features are beneficial for studying high temperature
(2) Can display solidification structure existing at solidification and solid phase transformation. For example, from Fig.
high temperature and transformation phases at room 1-26 and Fig. 3-47(in Chapter 3), it can be seen clearly that as-cast
temperature simultaneously pearlite occurs at the last region to solidify and the lower ausferrite
Using hot alkaline etching, except for colouring solidification formed during austempering has certain inheriting relationship with
structure, room temperature structure of cast iron can also be its originally solidified structure, as shown in Fig.1-27.
displayed. (3) Can visually and quickly display segregation of cast
According to electrochemical etching principle [35], whether the iron.
structures in iron and steel materials are etched or not, depend on Cast iron is a kind of metal material containing large amounts
their polarity, that is, the height of stable potential in solution. The
phase having low potential becomes anode (positive) and is etched;
the phase having high potential becomes cathode (negative) and
is protected. Under effect of acid reagent, the room temperature
structures in iron and steel, such as ferrite, martensite, austenite
solid solution all have lower stable potential and are etched as
anode. While, stable potential of carbides and nitrides is relatively
high and, as cathode (negative) they are protected. However,
when iron and steel specimens are put into alkaline solution,
these phases’ electrochemical behaviours are changed. Originally,
carbides, nitrides and phosphides are cathode (negative) phase
in acid solution; but in caustic alkaline solution, their potential is
even lower than that of matrix phase. Thus, they become anode
(positive) and are preferentially etched. While solid solutions, such
as ferrite, martensite and austenite etc, become negative phase in Fig. 1-26: High temperature structures and room temperature
alkaline solution. According to anode deposition mechanism of transformed structures can be displayed
simultaneously using hot alkaline method

Solidification structure Austempered structure

Fig.1-27: Relation between low ausferrite and solidification structure

154
 Serial Report
May 2009

and various types of solute elements, with serious positive and be complemented if combining with liquid quenching method. In
negative segregation. The hot alkali method can be used to quickly addition, if colours can be pre-calibrated by electron probe, this
display face distribution information of elements in macro or method is possible to be developed from qualitative to semi or full
micro region. quantitative display.
(4) Presented information is very sensitive
Experiments have shown that the variation in composition, 1.5.3 Operational procedure of colour metallography
which cannot be identified with electron probe, can be displayed using hot alkaline etching method
with different colour using hot alkaline method. The important contents of hot alkaline etching process include
(5) Operation is simple and convenient; no special etching reagent, etching temperature and time; these all result
experimental apparatuses needed in decisive influence on colour effect. Figure 1-28 is an actual
The disadvantage of this method is that it cannot display the example showing the influence of etching time on colour
precipitation process (nucleation and growth) of graphite; it can development.

(a）5 min, (b) 10 min, (c）15 min, (d）25 min, (e）35 min, (f）45 min (with polarized light)
Fig: 1-28 Relation between colour evolution and etching time (etching solution 1)

155
 CHINA FOUNDRY Vol.6 No.2

(1)Preparation of etching reagent strong alkalinity, reaction is violent, sharp colour can be obtained
The composition of hot alkaline reagent consists of sodium in a short time. When sample has low silicon content, number
hydrate (NaOH), potassium hydrate (KOH), picronitric acid and 2 and 4 are more suitable. Reagent number 5 is more violent in
distilled water. Table 1-4 lists several actual formulas of the reaction, its etching time is relatively short, and especially suitable
etching reagents. Reagent number 1 has moderate alkalinity and for ductile iron containing Al. Reagent number 3 is weaker in
reaction is mild; the displayed colours show good gradation of alkalinity than number 1, needs lower temperature, suitable for
colours and good colour transition; the reaction is slow, etching naked-eye observation of macro solidification morphology with
time is long and it is easy to operate. Formula number 2 and 4 have low magnification.

Table 1-4: Formula of hot alkaline etching reagent and etching process

Etching reagent Formula Etching process

Literature
No. NaOH (g) KOH (g) Picronitric acid (g) Distilled water (mL) Temperature (oC) Time (min)
1 28 1 4 200 95–100 10 –30 [28, 29]
2 100 8 200 95 –100 1.5 –4 [31, 32]
3 20 3 4 200 65 10– 65 [28]
4 40 160 40 200 120 1–1.5 [37]
5 28 56 28 200 95– 100 1.5–3 [38]

Preparation procedure of the etching reagents is as follows: progress. The colour changing sequence in the same position is:
first add NaOH into water, stir to accelerate dissolution; then add brown, blue, pale blue, pale yellow, yellow brown, orange, blue
pirotrinic acid, stir to make solution uniform. After fully dissolved, green, red yellow.
heat the solution (best to heat in a constant temperature bath) to The promise for satisfied result is well controlled etching
required temperature, then put sample in the solution. Place the solution temperature. The higher the temperature, the shorter the
polished face up to protect from scratching. After holding the etching time. Often, in order to obtain good colour contrast for
sample at the constant temperature for required time, take the different samples or obtain the same or similar colour for a certain
sample out, wash with distilled water, then wash away water with phase, constant temperature is used by changing etching time. For
absolute ethylalcohol. Dry the sample, then the sample can be used certain structure, the time demanded for displaying special colour
for observation. is related to composition of cast iron, (particularly silicon content).
(2) Etching and display process So, the corresponding relation is varying with variation of cast iron
Important etching parameters are etching temperature and time. composition.
At constant temperature etching time exerts significant effect For samples difficult to etch, first use normal etching reagent
on film forming rate. With extension of time, colour gradually to pre-etch to increase surface activity. This can accelerate film
changes (see Fig.1-28) and displays solidification development formation.

References
[1] Daris J R．Cast Irons．In: ASM Specialty Handbook, ASM International Materials Park, OH, 1996.
[2] Stefanescu D M．Cast Iron．In: Metals Handbook, 9th ed.，ASM, Metal Park, OH, 1988(15):168—181.
[3] Steeb S, Maier U．The Molten Structure of Fe-C Alloys, The Metallurgy of Cast Iron, ed. by B. Lux, I. Minkoff and F. Mollard. In: Proc. 2nd Int.
Symp. on the Physical Metallurgy of Cast Iron. St Saphorin: Georgi Publishing Company, Switzerland,1975.
[4] Benecke T, Ta A T, Kahr G, Schubert W D and Lux B．Aufloseverhalten und Vorimpfeffekt von SiC in Gusseisenschmelzen．Giesserei，1987(10
／11)：301—306.
[5] Xудокормов Д Н, Kалиниченко A C, Жвавый H П. Испольэование Быстрооxлажденного Mодификатора для yстранения Oтбела в Чугуне.
Литейное Проиэводство,1989 (12): 9—10.
[6] Margerie J C. The Notion of Heredity in Cast Iron Metallurgy, The Metallurgy of Cast Iron, ed. by B. Lux, I. Minkoff and F. Mollard. In: Proc. 2nd
Int. Symp. on the Physical Metallurgy of Cast Iron. St Saphorin: Georgi Publishing Company, Switzerland,1975: 545-558.
[7] Hикитин B M. Oсновные эакономерности для структрной наследственности в системе “шихта-расплав-отливка”. Литейное
Проиэводство, 1991(4): 4.
[8] Bian Xiufang, Liu Xiangfa and Ma Jiayi. Heredity of Cast Metals. Jinan: Shandong Science & Technology Press, China, 1999. (In Chinese)
[9] Masaalski T B．Binary Alloy Phase Diagrams．2nd ed., Vol. 1, America: William W: Scott, Jr., 1996.
[10] Lu Wenhua，Li Longsheng and Huang Liangyu. Cast Alloys and Smelting. Beijing: China Machine Press, China, 1997. (In Chinese)
[11] Oldfield W．Chill-reducing Mechanism of Silicon in Cast Iron．BCIRA Journal, 1962(1):17—27.
[12] Janovak J F, Gundlach R B．A Modern Approach to Alloying Gray Iron．AFS Transactions，1982(90)：847—863.
[13] Heine R W．The Fe-C-Si Solidification Diagram for Cast Iron．AFS Transactions, 1986(94): 391—402.
[14] Labrecque C， Gagne M．Interpretation of Cooling Curves of Cast Irons：A Literature Review．AFS Transactions, 1998(106): 83—90.
[15] Strong G R．Thermoanalyse als Mittel zur Quanlitätssicherung bei der Erzeugung von Gusseisen mit Kugelgraphit．Giesserei-Praxis, 1985(13／
14)：210—215.
156
 Serial Report
May 2009

[16] Elliott R. Cast Iron Technology. London: Butterworths & Co., 1988.
[17] Nagayoshi H, Imanishi K. Effect of Mg Content of Spheroidizer on the Chilling Tendency of SG Melt. AFS Transactions, 1996(104): 75—78．
[18] Mouquet O, Delpeuch F, Godinot P. Ermittlung des Metallurgischen Zustandes von Induktionsofeneisen durch Thermische Analyse. Giesserei-
Praxis, 2001(5): 199—213.
[19] Deike R. Einfluβ von Spurenelementen auf die eutektische Erstarrung und die eutektoide Umwandlung von Guβeisen. Giesserei, 1999(6): 175—182.
[20] Wallace J F. Effects of Minor Elements on the Structure of Cast Irons. AFS Transactions, 1975 (83): 363—378.
[21] Beraha E, Shpigler B. Color Metallography. 1st ed., ASM, Metals Park, OH, 1977.
[22] Weck E, Leistner E. Metallographic Instructions for Colour Etching by Immersion. Deutscher Verlag für Schweiβtechnik GmbH-Düsseldorf, 1982.
[23] Petzow G. Metallographic Etching. 1st ed., ASM, Metals Park, Ohio, 1978.
[24] (Ukraine)Коваленко B C. Handbook of Metallurgical Phase Reagents. Translated by Li Chunzhi, et al, Beijing: Metallurgical Industry Press, 1973.
(In Chinese)
[25] Mалиночка Я Н, Oсава H Г. Заводская ЛабораториЯ, 1962(3): 315.
[26] Motz J, Rohrig K. Untersuchung einiger Gefügezusammenhange in phosphorarmen Gieβereiroheisen. Giesserei-Forschung, 1970(4): 142—152
[27] (Russa)БунинК П, et al. The Structure of Cast Iron. Translated by Harbin University of Technology, Beijing: China Machine Press, 1977. (In Chinese)
[28] Zhou Jiyang. Gefügebildung von Guβeisen mit Kugelgraphit bei langsamer Erstarrung. Germany Aachen. Fotodruck. J. MAIN GmbH, 1986
[29] Zhou Jiyang, Schmitz W und Engler S. Untersuchung der Gefügebildung von Guβeisen mit Kugelgraphit bei langsamer Erstarrung. Giesserei-
Forschung, 1987(2): 55—70
[30] Zhou Jiyang, Zhong Fengqi, Schmitz W, Engler S. Application of a New Colour Metallorgraphy Technique to Cast Iron. Practical Metallography,
1993(3): 122—128.
[31] Wang Y, Fan Z, Gan Y, et al. Eutectic Solidification Characteristics of Commercial Gray Cast Iron. In: Proc. Conf. on Physical Metallurgy of Cast
Iron, Cast Iron IV, America Materials Society, 1990: 95—102.
[32] Tian H and Stefanescu D M. Application of a Coloration Etching Method to the Study of Microstructures in Primary and Eutectic Solidification in
Cast Iron. Materials Characterization, 1992(3): 329—333.
[33] Van C A de Velde. Untersuchungen zum Erstarrungsprozeβ vom GGG. In: der Deutsche Meehanite-Tagung, 1993. Gieβerei-Erfahrungsaustausch,
1994(3): 105—108.
[34] Motz J M, Wolters D B. Über Erstarrungsgefüge und Eigenschaften in dickwandigen Guβstücken aus ferritischem Guβeisen mit Kugelgraphit-Teil 1:
Mikroseigerungen. Giesserei-Forschung, 1988(2): 69—79.
[35] Colour Metallography Technical Group. Colour Metallography Techniques: Principle and Method. Beijing: National Defence Industry Press,
China, 1987. (in Chinese)
[36] Zhao Yanhui. Study on Colour Metallography of Cast Iron Solidification[Dissertation]. Dalian: Dalian University of Technology, China 1999.(in
Chinese)
[37] Motz J M. Microsegregations Easily Unnoticed Influencing Variable in the Structural Description of Cast Materials. Practical Metallography,
1988(25): 285—293.
[38] Boutorabi S M A, Campbell J. An Etching Technique for Primary Austenite Dendrites in Ductile Cast Iron. Materials Characterization, 1993(1):
127—132.

Chapter 2
Grey Iron (Ⅰ)
Grey iron is a type of cast iron with grey color fracture and solidification, due to significant difference in composition in the
carbon precipitated as flake graphite. According to its chemical region last to solidify, the formed structures are quite different. For
composition in Fe-C phase diagram, grey iron is categorised convenient discussion, this solidification is separately listed as the
into three types: hypoeutectic, eutectic, hypereutectic irons. In third stage solidification in this book.
order to satisfy strength demand, most engineering grey irons are Thus, the solidification structures of grey irons comprise
hypoeutectic composition. However, for achieving combinations primary austenite, primary graphite, eutectics (eutectic graphite +
of good castability, strength, conductivity and dumping capacity, eutectic austenite) and the structure formed around eutectic cells.
eutectic and even hypereutectic grey irons attracted more and more
attention in research. 2.1 Graphite in cast iron
In a hypoeutectic cast iron with carbon saturation degree (the
Graphite is an extremely important constituent in cast iron,
ratio of carbon content in the iron to the actual carbon of eutectic
accounting for about 10% volume in cast iron. Graphite itself
point) Sc < 1, the first precipitated phase is primary austenite
has almost no strength, but its shape or morphology, amount and
during solidification; for a hypereutectic iron with Sc > l, the
distribution exert significant influence on the properties of cast iron.
first precipitated is primary graphite. For these two type irons, at
Commonly observed graphite shapes are: flake, vermicular or
the second solidification stage, eutectic crystallization will take
compacted, spheriodal and quasi spheriodal-temper graphite. The
place. While, a cast iron with Sc = l only has the second stage
former three types are the graphite directly precipitated from liquid
eutectic crystallization during solidification. In the last stage of

157
 CHINA FOUNDRY Vol.6 No.2

iron, while, the last one is obtained from solid tempering treatment. In China, F type graphite is added based on the classification.
The cast irons with above types of graphite are correspondingly Characteristics of the various types of flake graphite are listed
named grey iron, vermicular iron, spheriodal iron and malleable in Table 2-1, among them, Kish graphite (type C) and star-
iron. like graphite (type F) belong to primary graphite, the rest types
Internationally, according to their shape and distribution, flake precipitate during eutectic stage and belong to eutectic graphite.
graphite is classified into A, B, C, D and E five types, see Fig. 2-1.

(a) type A (b) type B

(c) type C (d) type D (e) type E

Fig.2-1: Classification of flake graphite

Table 2-1: Classification of flake graphite forms in grey iron

Graphite Type Feature
Flake graphite A uniform distribution, random orientation
Rosette graphite B Rosette grouping, random orientation
Kish graphite C Coarse and large primary graphite flakes with small flakes around
Interdendritic dot-like graphite D Randomly orientated interdendritic undercooled short flake graphite
Interdendritic flakes graphite E Interdendritic orientated undercooled short flake graphite
Star-like graphite F star or spiky graphite mixed with short flake graphite

2.1.1 Crystal structure of graphite between atoms, similar to that of metallic bond. The adjacent layers
are connected by a weak attractive force-Van der Waals force-
In nature world, there are two types of carbon: crystalline carbon the interaction force between molecules. The huge difference in
(graphite, diamond) and amorphous carbon (coke, coal etc.). bonding forces makes graphite strong anisotropy. Due to a very
Graphite is a hexagonal crystal structure, comprising six prismatic weak bond between layers, it is easy to split and slip between layers.
faces and two close-packed basal planes; its crystal structure is Consequently, the strength and hardness in basal plane are markedly
illustrated in Fig. 2-2. The plane between layers is called basal plane higher than that in prism plane; while, the electrical resistivity is
(0001), the plane perpendicular to the basal plane is a rectangular opposite, the resistivity in basal planes is much lower than that in
face, called prism face (10 0). The orientation parallel to the
prismatic planes.
basal plane is crystallographic orientation α[10 0], called α-axis;
the orientation perpendicular to basal plane is crystallographic 2.1.2 Defects in graphite crystal
orientation c[0001], called c-axis. The atom distance within the
layer (basal plane) is 0.142 nm; the distance between layers (basal There are large amounts of defects or imperfections in graphite
planes) is 0.335 nm. Within each layer the carbon atoms are packed crystal, such as lattice distortions caused by interstices or lattice
in hexagon; each carbon atom is bonded with three nearest-neighbor vacancies, staggered pack arrangements of one row or several
atoms by 400—500 kJ/mol connecting energy, a kind of energy rows. The crystal defects directly affect growth of graphite.

158
 Serial Report
May 2009

(a) Schematic, (b) Lattice fringe image under high resolution electrical microscope[1], fringe spacing 0.335 nm
Fig. 2-2: Crystal structure of graphite
There are following defects in graphite crystal: more atoms as possible corresponding to each other. The most
(1) Rotating twin: rotating twins are common defects in angles forming rotating twins are 22o, 28o [2] or 13.2o , 21.8o and
graphite crystal; the carbon atomic layers remain parallel to each 27.8o [3].
other. The top and bottom layers rotate an angle around c-axis The big difference in bonding force between basal and prismatic
each other (see Fig. 2-3), this angle is called angle θ . The rotating planes is the important reason causing rotating twin defects to take
angle complies with the interface response theory to ensure as place frequently.

(a) Schematic (b) Scanning microscopy morphology [1]

Fig. 2-3: Rotating twin defect in graphite

(2) Screw dislocation: although the carbon atoms in basal i =2r cos a
plane are connected by strong covalent bond, there still exist 2
vacancies in certain local regions in the hexagonal network. The Where α — apex angle, possible range for spiral pack is α= 0-180o.
resulted plier shaped defects from these vacancies are the bases (b) Typical screw: pure screw dislocation perpendicular to
for screw or spiral dislocations. When there is a shearing stress plane (0001)
perpendicular to basal plane, this stress can cause graphite crystal (c) Cylinder screw: apex angle is 0o for this case.
to split along Q plane, leading to screw dislocations. The CD line Actual often existing screw dislocations are circular cone
is the centre of spiral face, as illustrated in Fig. 2-4. There are screw; the apex angle is between 0-180o. Nodularizing elements
much less screw dislocation defects than rotating twin defects in are embedded in the layers or absorbed in basal plane causing
local barrier effect on (10 0), easily creating screw dislocation.
graphite crystal.
Double and Hellawell divide screw dislocations into three types, Besides, liquid iron undercooling, especially the undercooling
as illustrated in Fig. 2-5. at existence of low S and O-active elements, will significantly
(a) Circular cone screw: formed by the folded graphite flakes increase disturbing sources and accelerate the formation of screw
rotatingθ angle round c-axis. The rotating angleθ is calculated by: dislocations [4].

159
 CHINA FOUNDRY Vol.6 No.2

(a) schematic (b) scanning microscopy dislocation morphology of graphite spheroid [1]

Fig.2-4: Screw dislocation defect in graphite

two conditions:
(1) Conform to certain relation of lattice mismatch or
disregistry: a certain lattice face of the foreign particle (inclusion)
substrate has a certain corresponding relation in lattice parameter
and orientation with graphite crystal lattice.
A characteristic parameter of the lattice matching relationship is
plane mismatch δ.
According to planar lattice corresponding model by Bramfitt[6],
the mathematical expression for plane lattice mismatch is:
(d[UVW] COS α )-d[UVW]
(a) Circular cone type， (b) typical screw (α=180o)，
d[UVW]
(c) cylinder screw (α=0°)
Where,
Fig. 2-5: Type of spiral structure in graphite [3]
d [UVW] - atomic spacing of foreign nucleus along [UVW]
2.1.3 Conditions of graphite nucleation direction, UVW is the low-index direction of foreign nucleus in
(hkl).
Graphite nucleation can be homogeneous or heterogeneous d[UVW] - the inter-atomic spacing of graphite along [UVW]
nucleation. direction, UVW is the low-index direction of graphite in (hkl)
Homogeneous nucleation relies on large amounts of fluctuated α -angle between heterogeneous nucleus and graphite in
micro-carbon atom clusters of (C6)n, which transform to crystal [UVW] direction
structure the same as graphite, forming germs or embryos, then The smaller the mismatch of the two substances δ, the stronger
finishing nucleation process. Only when a liquid iron reaches the nucleation ability between them. The general relation is [7,8]:
enough undercooling, the germs with certain sizes can transform δ<6％ has the strongest nucleation ability
to stable nuclei. δ=6％-12％ has nucleation ability
The nature of these micro clusters is not very clear at present, δ>12％ has a weak nucleation ability
however, they do exist [5]. Often, the (C6)n clusters in iron melt are Mismatch δ (%) between a certain special lattice face of some
in unstable state, forming dynamic equilibrium with carbon-rich substances and the (0001) face of graphite is listed in Table 2-2.
liquid regions. When liquid iron reaches high enough undercooling Except for a small lattice mismatch, the good matching
(200-230oC), the smaller sizes of micro cluster (C6)n are possible relationship also means that the two substances have the same
to exist as stable nuclei. Under normal conditions, such high lattice structure, similar or proportionally similar lattice constants;
undercooling is very difficult to achieve; therefore, the nucleation the two substances have different lattice structure, but have similar
of graphite mainly is heterogeneous nucleation. atom pack arrangements and close or proportionally similar lattice
Heterogeneous nucleation refers to that carbon atoms constants on certain lattice plane. It is obvious that if the substrate
preferentially attach to the surface of existing particles (substrates) and growing crystal are the same substance, the growing crystal
and form nuclei. There exist large amounts of foreign particles has the best nucleation condition.
in liquid cast iron. For example, in 1 cm3 of iron melt, just the (2) Satisfy interfacial energy requirement between
number of oxide particles can reach five million. That does not heterogeneous nucleus and graphite: only when the foreign
mean all the foreign particles can become nuclei of graphite. particles can be wet by graphite, graphite can form nuclei on the
Becoming a nucleation substrate, foreign particles must satisfy foreign substrates.

160
 Serial Report
May 2009

Table 2-2 Mismatch δ(%) between a special lattice face of some substances and the (0001) face of graphite

Substance Mismatch δ (%) Reference Substance Mismatch δ (%) Reference

graphite 0 CaS 8.3
BN 2.0 CeS 2.9
[10]
CaS 8.3 LaS 1.5
CaC2 9.9 BaS 7.5
MnS 10.2 [9] CaO • SiO2 7.5
α-Al2O3 11.4 SrO • SiO2 3.5
CaO 13.3 BaO • SiO2 1.5
[6]
β-SiO2 37.1 CaO • Al2O3 • 2SiO2 3.7
β-Si3N4 43.9 SrO • Al2O3 • 2SiO2 6.2
MgS 12.5 BaO • Al2O3 • 2SiO2 7.1
[10]
MnS 12.1 AlN 9.47 [11]

According to the most recent study [12], the interfacial energy independently along (1010) face; the steps on (1010) face can only
between iron liquid - nucleus σ LN must be higher than the advance to the edge of crystal. Due to the weak connecting force
interfacial energy between iron liquid - graphite σLG. In addition, (bond) between a single atom and (1010) face, graphite growth
σLN needs to be higher than the interface energy between nucleus- is very difficult, even almost impossible to grow with this type of
graphite σNG[13]. Only under these conditions, it can be sure that mechanism.
the two phases have stable coherence and adhere each other. At (2) Rotating step growth: graphite deposits on the rotating twin
this time, the small mismatch can fully play a role for nucleation. steps, the (1010) face grows by step expanding laterally. Since the
Thus, only lattice match is not enough to fully explain the efficient deposited atom has two touching faces, this growth is more stable
nucleation; the energy requirement between melt - nucleus than the two dimensional growth and has large growth probability.
interfacial energy needs to be satisfied. (3) Screw dislocation growth: carbon atoms deposit on the
Nevertheless, no all the inclusions existing in the centre of edge of steps caused by screw dislocation, the steps continuously
graphite satisfy above principles, since inclusions in the centre expand, sweeping the crystal plane. The angle-velocity sweeping
of graphite may comprise several inclusions. However, one kind crystal plane at dislocation centre is faster than that in the region
of inclusions satisfying lattice mismatch and interfacial energy far from the centre, giving rise to spiral, tower-tip-like crystal
requirement with graphite is enough. At most cases, the nuclei of surface. Due to having three touching surfaces, carbon atoms enter
graphite are about 1-2 µm[14]. step without need of forming a new surface.
Two-dimensional growth and rotating step growth mechanisms
2.1.4 Mechanism of graphite growth adapt prism face (1010) growth, the growth velocity v a and
undercooling △T follow exponential rule:
The nucleation of graphite influences the amounts of graphite
nuclei, while the growth process determines the final morphology va = μ ae-b/ΔT
of graphite. The different morphologies of flake, vermicular even Where μ a, b - coefficients
spheriodal graphite are not coming from different types of nuclei, When graphite basal face (0001) grows, carbon atoms deposit
but from different growth processes. on the steps of screw dislocations, the growth velocity vc of basal
The growth dynamical mechanism of graphite is closely related face and undercooling ΔT follow parabolic relation:
2
to melt undercooling and crystal defects of graphite. vc = μ cΔT
There exist three mechanisms of graphite growth (that is, Where μ c - coefficient
graphite atom packing onto nucleus), see Fig. 2-6. Summarizing the relationship of prism face growth velocity va,
(1) Two dimensional growth: graphite forms nuclei and grows and basal face growth velocity vc with undercoolng ΔT , the cross
of the two curves is called critical undercooling ΔT c; when△T ＞
△T c, vc＞va and when△T ＜△T c, vc＜va.
When graphite grows under an ideal state without lattice defects,
carbon atoms are packed in the form of single crystal graphite
flake with the same crystalline orientation. Kozlov observed that
without of influences undercooling and subversive elements, the
equilibrium shape of graphite is flake [15]. In the same way, Nakae
also found that the graphite in a pure Fe-C-Si alloy is a straight,
thin and long shape [16].
However, the actual growth state of graphite is very different
from the ideal state. There exist various defects in the actual crystal
germs or embryos, these defects destroy the neat arrangement of
Fig. 2-6: Three mechanisms of graphite growth graphite lattice, causing braches and many crystalline orientations.

161
 CHINA FOUNDRY Vol.6 No.2

2.1.5 Branches in graphite growth undercooling. Large KF means that graphite branching is further
severe. Since only large amount of branches can increase the ratio
There are two types of graphite branches [l7], as illustrated in Fig. 2-7: of surface area S to volume V.
graphite branches along a-axis, basal plane (0001) split; c-axis
(3) Austenite near to graphite: the austenite near to graphite
branches, prism plane (1010) split. can inhibit graphite branching.
The factors affecting the types and severity of graphite The types and severity of graphite branching is directly related
branching include types of graphite defects, undercooling and to final graphite morphology. Their relationship is summarised in
austenite growth near to the graphite. Table 2-3.
(1) Crystal defects: the defects existing in graphite directly
influences the types of formed branches. The packing of carbon
2.2 Crystallization of primary graphite
atoms on rotating twin steps is equivalent to splitting of graphite
along basal plane forming a-axis branch. If carbon atoms pack on 2.2.1 Crystallization thermodynamics of primary
the step of opening of screw dislocation, because the one end of (type C and F) graphite
step is fixed at dislocation line, steps grow around dislocation line,
When hypereutectic iron melt of composition X is undecooled to
resulting in c-axis branch perpendicular to basal plane (0001).
temperature 1 in Fig. 2-8, the composition of the melt changes to
(2) Undercooling: undercooling is a important factor
Xα and its free energy is G1. At this time, the free energy of the
influencing branching of graphite. The relationship [18] between
melt is ΔG higher than that of the mixture of the melt and graphite.
shape factor K F and undercooling ΔT derived from fundamental
Thus, the mixture of the melt with graphite precipitated is more
theory of crystal growth is:
stable, causing primary graphite to precipitate from the iron melt.
KF = QρΔT /tKM σ
The growth of primary graphite is driven by chemical potential
Where, F-is crystal shape factor (S/V) ratio of surface area S
K
gradient resulted by carbon concentration difference, since in the
of a crystal to its volume V;
region near to primary graphite, carbon is poor, the carbon content
Q-formation heat of graphite crystal;
is Xα. While, in the region far from graphite, the melt remains its
ρ -density of graphite;
original composition X. The difference in carbon concentration
tK-eutectic temperature;
results in carbon chemical potential gradient, causing carbon atoms
M-relative molecular mass of graphite;
to diffuse towards the growing crystal, graphite will continuously
σ -surface energy of touch surface between graphite and melt;
precipitate until the remaining melt reaches composition Xα. At
△T -undercooling.
this time, the remaining melt with composition Xα is in equilibrium
In above equation, except for △T , other parameters are
with precipitated graphite.
all constants. Thereby, K F is increased with increasing of

(a) (b)
(c)

Fig. 2-8: Free energy change of hypereutectic

cast iron when precipitated
primary graphite

2.2.2 Precipitation of primary

graphite
The type C and F graphite are primary flake
graphite, and present in hypereutectic grey
iron. The two dimensional characteristics of
(a) Branch along a-axis, (b) Branch along c-axis, (c) high magnification lattice primary graphite is less branched, straighter
fringe image of branch along c-axis. and flatter, coarser and larger. This type of
Fig. 2-7: Types of graphite branches graphite is formed under small undercooling,

162
 Serial Report
May 2009

Table 2-3 Relationship between branch types and severity and graphite morphology

Influencing factor
Branch type Degree of branching Graphite morphology
lattice defect undercooling
Small Small Flake
a-axis branch Rotating twin ↓ ↓ ↓
Medium More Intermediate
Medium More Intermediate
c-axis branch Screw dislocation ↓ ↓ ↓
Big Severe Spheriodal

and the growth temperature is in the region between liquidus and directly formed on the surface gathers and forms surface blocks on
eutectic temperatures, and diffusion of carbon at this temperature the melt surface.
region is fast. In addition, austenite is very difficult to form (2) Formation of Kish graphite[20]: for Kish graphite, oxygen
around graphite. Graphite growth is not affected by austenite; plays an important role in either the nucleation or growth by
therefore, graphite is easy to grow to C type graphite flakes. When flake-like dendrite manner. Practices have demonstrated when
undercooling is increased (for example thin wall casting), graphite hypereutectic iron melt is exposed in air, Kish graphite is easy to
branches are increased, forming type F, star-like graphite. form; disturbing iron melt, thus continuously creating new surface
The branching of primary graphite can be classified to two types: will accelerate the formation of Kish graphite. The role of oxygen
small angle and large angle branching [19]. The large branching is forming SiO2 as substrate for Kish graphite nucleation. When
angle is due to the development of rotating twin defect; the defects iron melt is below 1,400℃, the Kish graphite forms more than in
cause more growth front and change in growth directions. The higher temperature, since SiO2 is not stable above 1,400℃. In the
formation of small angle branching is depended on solidification melt with low or no silicon, the precipitation phenomenon of Kish
conditions (such as the variation of undercooling in solidification graphite is not observed. This shows that existence of SiO2 has
front, embedded inclusions etc.) non-ignored influence on the formation of Kish graphite.
Without oxygen (for example protected by argon gas), the
2.2.3 Kish graphite cooling rate is a basic condition influencing the formation of Kish
Kish graphite is a type of special structured graphite, which graphite. Kish graphite can only form under slower cooling rate
crystallized in carbon saturated iron melt with flake dendrite (<180℃/h). At higher cooling rate, the formation of Kish graphite
growth. The graphite flakes are gathered on the iron melt surface, is inhibited, since high cooling rate increases crystal growth
either mixed with slag and exist in slag layers, or leave melt and defects and accelerates graphite branching trend, causing planar
float in surrounding. growth change to blocky growth. However, the cooling rate cannot
(1) Structure of Kish graphite: the crystal characteristics of Kish be too slow, if cooling rate is decreased below 50℃/h, massive
graphite is flake-like dendrite structure, as illustrated in Fig. 2-9. single crystal graphite will form which is even bigger, thicker and
When observing normally from c-direction, it is seen that the hexagon more compacted than Kish graphite.
rings are almost overlapped and have branches; and graphite flakes There exist two types of Kish graphite: (a) severely branched
are connected in layer in a-direction. Kish graphite precipitated from and thin flake graphite. This type of graphite is formed under fast
melt first gathers together in melt, and then floats to the surface of cooling and on the iron melt surface directly in contact with air,
the melt, forming bulk or blocky volume. While, the Kish graphite often observed during cleaning slag operation or occur on the
flowing iron melt surface. (b) less branched,
thick and disc-like graphite. This type of
graphite is formed under slow cooling and on
the iron melt surface in less contact with air,
often observed in the slag layers of liquid iron
conveying tank or in the Kish graphite layers.

(Be continued)

(a) Observation from c -direction (b) Observation from a-direction

Fig. 2-9: Crystal structure characteristics of Kish graphite

163