Metallurgy for the Non-Metallurgist®

Solidification of Metals

Revised by
McIntyre (Mac) R. Louthan, Jr., Ph.D., FASM

3/2013
 Copyright © 2013
by
ASM International®
All rights reserved

No part of this lesson may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photo-
copying, recording, or otherwise, without the written permission of the copyright owner.

Great care is taken in the compilation and production of this book, but it should be made clear that NO WARRANTIES, EXPRESS OR IMPLIED,
INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE
GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot 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 publication, 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. ASM does not endorse or guarantee the content contained in this book. The
conclusions and opinions contained in this book and expressed by the instructor are those of the author and not those of ASM. ASM is not respon-
sible for the use or misuse of the information contained herein. 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 lesson 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 book 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.

ASM staff who worked on this project include Pete Rottier, Diane Whitelaw, and Bonnie Sanders. Copy edit by Jo Hannah Leyda.

ASM International®
Materials Park, OH 44073-0002
www.asminternational.org

Printed in the United States of America

 Metallurgy for the Non-Metallurgist®

Solidification of Metals
Lesson Overview

Metals are useful for engineering purposes because they can be fabricat-
ed into useful shapes. One method of fabrication involves a casting. This
is where molten metal is introduced into a mold, die, or some other type of
form that contains the molten metal until solidification (freezing) occurs, af-
ter which the metal is removed from the mold and either has the desired shape
or is machined into the desired shape. In the second method of fabrication
parts are made from wrought material. This begins with molten metal that is
allowed to solidify as an ingot or a continuously cast bloom. This intermedi-
ate product is then mechanically formed into a semi-finished product, such as
bar or plate. The bar or plate is then forged, or machined into the component’s
final geometry. In either event, the process of manufacturing the part begins
with molten (liquid) metal that is cooled and undergoes solidification dur-
ing cooling. During the solidification, the atoms in the metal pack together
in specific ways, much as pool balls are packed in a specific way when they
are placed in the rack, except that the packing of atoms in a metallic solid
results in a regular, three-dimensional pattern that is called a crystal structure.
The crystal structure of a metal simply describes the long-range arrangement
of atoms. There are several different crystal structures (atom arrangements)
that metals generally develop. Copper, aluminum, nickel, gold, silver, and
many other metals and alloys have a crystal structure that differs from that
of chromium and molybdenum. Zinc has another crystal structure, while iron
may have the same structure as copper or the same structure as chromium

Lesson Learning Objectives

After completing this lesson, students will be able to:

► Describe the basic metal crystal structures

► Explain the process of solidification in metals
► Read a basic binary phase diagram
2 ► Solidification of Metals

depending on the temperature. However, in virtually all cases, solid metals

have a specific crystal structure. Copper will always solidify with a specific
atom arrangement and the atom arrangement in solid chromium will always
differ from the atom arrangement of copper. Metals are crystalline solids
even though metallic solids typically do not have the crystalline look that
we see in rubies, sapphires, and diamonds. Crystalline solids simply have
their atoms arranged in a definite order or pattern that is uniformly repeated
in three dimensions. Another type of solid is termed amorphous, and many
polymers and glasses have amorphous structures. A complex pattern of at-
oms can be found in amorphous materials, but the pattern is not repeated
uniformly throughout the solid. It is this lack of homogeneity of pattern that
differentiates amorphous from crystalline materials. This lesson introduces
basic concepts of the characteristics of atom arrangements that provide the
crystal structure of metals, describes the solidification process, and introduc-
es phase diagrams. Phase diagrams provide a graphical description of what
happens to a metal when two metals are melted together and solidified as an
alloy. In the diagram, temperature is important because the alloy may change
from a crystalline (solid) phase to a liquid phase as the temperature increases.
The relative amount of the two metals is also important because the crystal
structure may change as the composition changes.

Basic Crystal Lattices

Virtually all solid metals are crystalline, but all metals do not have the
same long-term pattern in their atomic arrangements. The most common pat-
tern of atoms in solids involves a cubic arrangement.

A simple cube is the shape of a box and the atoms are located at the corners
of the box (Fig. 1). This type of atomic arrangement is called simple cubic.

Fig. 1 Simple cubic lattice. (No metal has this structure.)

 Solidification of Metals ◄ 3

Most metals do not form simple cubic crystal structures, but many metals
form atom arrangements that are slight modifications of the simple cubic.
One modification is called body-centered cubic, abbreviated bcc. Solid chro-
mium has a bcc crystal structure. Figure 2 shows the arrangement of atoms
in a bcc lattice. There is an atom located at each corner of the cube and an
additional atom in the center of the cube. In addition to chromium, niobium,
lithium, molybdenum, tungsten, and vanadium are examples of metals whose
atoms are arranged in a bcc pattern. Iron also has a bcc crystal structure at
temperatures below 912 °C (1674 °F) and at temperatures between 1394 to
1538 °C (2541 to 2800 °F). Iron melts when heated above 1538 °C. Molten
iron is said to be in the liquid phase. When iron solidifies, a solid phase forms
and that phase has a bcc lattice. The bcc phase is stable between 1394 and
1538 °C, and when the temperature falls below 1394 °C the atom arrange-
ment changes. The new arrangement that forms is called face-centered cubic.

The face-centered cubic lattice, abbreviated fcc (Fig. 3), is another modifi-
cation of the simple cubic lattice or crystal structure. In the fcc lattice an atom

Fig. 2 Body-centered cubic (bcc) lattice

Fig. 3 Face-centered cubic (fcc) lattice

4 ► Solidification of Metals

is located at each corner of the cube and additional atoms area at the center of
each face of the cube, hence the name face-centered cubic. Aluminum, cop-
per, gold, lead, nickel, platinum, and silver all have their atoms arranged in
the fcc pattern, as does iron at temperatures between 912 and 1394 °C (1674
and 2541 °F).

Another common arrangement of atoms in solid metals is the hexagonal

close-packed lattice, abbreviated hcp (Fig. 4). At the top and bottom, there
are six atoms so arranged that when straight lines are drawn between them,
they form a hexagon. There is one additional atom in the center of each hexa-
gon, and there are three atoms in the form of a triangle between the top and
bottom hexagons. Metals that have an hcp structure are magnesium, cad-
mium, and zinc.

Many pure metals, iron for example, have two or more different crystal
structures depending on temperature. Iron can be either bcc or fcc, while
zirconium and titanium can be either hcp or bcc. Uranium may have any of
three different crystal structures, depending on temperature, and plutonium
and manganese have more than three different crystal structures. The exis-
tence of more than one crystal structure in a metal is termed allotropy. The
existence of diamond, graphite, and soot make the existence of allotropy in
carbon familiar to most people. In addition to temperature, pressure may also
introduce allotropic transformations. However, metals are commonly used at
pressures that do not introduce allotropic transformations.

The atoms shown in Fig. 1 through 4 have been represented by dots, which
are connected with straight lines to give a visual idea of the general geometry.

Fig. 4 Hexagonal close-packed (hcp) lattice. Solid atoms are at unit cell
boundaries. Open atoms are at interstitial positions.
 Solidification of Metals ◄ 5

In reality the atoms should be represented as spheres that are packed against
each other. The illustrations of the bcc, fcc, and hcp lattices show the building
blocks that can be stacked together to represent a metallic solid. Each block
is termed a unit cell and atoms may be contained in more than one unit cell.
For example, the cube corners in the bcc or fcc lattices are points where eight
unit cells touch each other; thus, the atoms at the cube corners are shared by
eight unit cells.

Atoms do not remain firmly fixed in a specific position but are constantly
vibrating. The vibrations are caused by the thermal energy contained within
the metal. When heat is added to a solid metal, the amplitude of atom vi-
bration increases and the metal expands. This expansion is termed thermal
expansion and will cause a metal rod to lengthen as its temperature increases
and contract when the temperature decreases. When sufficient heat is added,
the atoms obtain enough energy to overcome the attractive forces that have
held them together and the metal melts or becomes a liquid. A pure metal
changes from a solid to a liquid at a specific temperature, the melting point.
However, when the melting point is reached, heat must be continually added
to cause the metal to melt. The amount of heat necessary to convert solid
metal at its melting temperature to liquid metal at the same temperature is
known as the heat of fusion. We take advantage of the heat of fusion when we
use ice to cool a beverage. The heat that is needed to convert the solid ice to
water is taken from the beverage and the temperature of the drink is lowered.

Solidification of Pure Metals

The term pure metal generally refers to commercially pure metals, which
contain measurable amounts of several types of impurity atoms. To obtain a
metal from its ore and refine it so that it is absolutely free from impurities is
practically impossible. The impurities in the metal must be incorporated into
the lattice when the metal solidifies, thus relatively pure metals may solidify
into relatively impure metals.

Freezing or solidification does not occur simultaneously throughout the

liquid. Tiny solid particles nucleate in the liquid metal after the freezing point
has been reached. As heat is continually extracted, the atoms lose their vi-
brational energy and become attached to the nuclei and cause it to increase
in size. Initially the nuclei are far too small to be seen (even if metals were
transparent), but they grow as the metal freezes. These multiple growing nu-
clei eventually become large enough to be observed, and when the solidifica-
tion is complete, each nuclei will become a grain. However, even when the
metal is solid throughout, the observation of these grains and their structure
may require the use of a microscope.

The process of solidification requires that the heat of fusion be removed

from the liquid as well as the formation of solid nuclei. Thus, when a pure
6 ► Solidification of Metals

metal is heated well above its melting point and cooled, if the temperature is
read at regular intervals and plotted as a function of time, a curve similar to
those shown in Fig. 5 will be obtained.

In these curves, point A represents the temperature of the metal at the first
measurement. As time goes by, the temperature of the liquid falls to point B.
Point B actually represents the freezing point of the metal, but as shown in the
figure, the metal may not solidify at this temperature because the nuclei nec-
essary for solidification may not have developed. Therefore, the metal will
continue to cool to the temperature represented by point C. This phenomenon
is called supercooling. By the time the metal reaches this temperature, the
embryos have developed and stable solidification nuclei are present. Once
the nuclei are formed, the metal particles grow at a rate that depends on the
rate of heat extraction.

When supercooling of a pure metal occurs, the temperature may rise (as
from C to D) because of the necessity to reclaim the heat that is released dur-
ing solidification below the melting point. The total heat released from the
beginning of solidification, point B, to the end of solidification, point D, is
called the latent heat of fusion. Point D is the same temperature as point B,
which is the freezing point of the metal.

From point D to point E, the curve is perfectly flat and there is no change
in temperature. The flat part of the curve represents a constant solidification
temperature. Thus the solidification of a pure metal may be considered to be
isothermal freezing. It is analogous to the constant temperature freezing of
water. At point E, the metal is completely solid and freezing is completed.
The portion of the curve from E to F represents the temperature of the solid
metal as it cools down to room temperature.

To the unaided eye, a metal in the solid form has a characteristic metallic
luster; however, little can be observed about the various nuclei that grew to
form the crystalline solid. The configuration of the crystals that developed

Fig. 5 Cooling curves for a pure metal, showing (a) an ideal cooling curve
and (b) an actual cooling curve
 Solidification of Metals ◄ 7

from the nuclei is termed the microstructure of the metal. To obtain micro-
structural information, a specimen is cut from the solid metal and carefully
polished until the surface is flat, smooth, highly reflective, and contains few
(preferably no) scratches. The polished surface is then etched with a chemical
solution called an etchant. The etchant attacks each surface grain and its con-
stituents differently. The purpose of the polish/etching process is to provide
a surface that, when examined under an optical microscope, using reflected
light, will provide information about the microstructure of the metal. The
science of developing and interpreting metal structures via the microscope is
known as metallography.

After a pure, slowly cooled, as-solidified metal has been etched, a struc-
ture similar to the one shown in Fig. 6 may be seen even with the unaided
eye. In this structure, long columnar crystals or grains extend from the out-
side to the center of the metal casting. (Although “crystal” and “grain” are
used interchangeably by many metallurgists, the term grain is preferred.)

When molten pure metal is poured into a mold, the metal in contact with
the mold wall cools quite rapidly, while the metal near the center cools at a
slower rate. Because the metal near the mold wall is coolest, nucleation (the
formation of nuclei) takes place first at the mold wall. In part A of Fig. 7,
several nuclei have formed in the liquid metal at the mold wall.

With time, as the metal cools, the nuclei grow in size (Fig. 7, part B). Note
that the crystals that have grown from the nuclei have become so large that
they touch each other. Soon all the crystals along the edge of the mold are in

Fig. 6 Columnar structure in solidified metal. The grains growing from two
walls intersect in planes; a clearly defined line of intersection is visible on the
polished and etched surface.
8 ► Solidification of Metals

contact with each other. If the liquid metal is poured off at this point, a shell
of solid metal will be found next to the mold walls.

Because the crystals are in contact with each other over the entire mold
surface, the only direction that the crystals can grow is toward the center. Part
E in Fig. 7 shows the appearance of the metal after it has completely solidi-
fied, which resembles the micrograph of an as-cast ingot on the right side of
Fig. 7. Figures 6 and 7 show the microstructure of similar castings but differ
somewhat because Fig. 6 is a horizontal section and Fig. 7 is a vertical sec-
tion.

An examination of the solid metal shell adhering to the mold wall after the
liquid metal was poured from a partially solidified casting would show little
crystalline needles protruding from the surface of the solid. These needles
are known as dendrites. Dendrites are crystals that form in a treelike pattern
during solidification (Fig. 8). However, after solidification, liquid freezes be-
tween the dendrite arms and the dendritic pattern of pure metals is difficult
to recognize. In Fig. 8(a), a crystal is just starting to grow from the mold
wall into the liquid. This type of growth is called dendritic growth. Within
the same mold, many dendrites can be growing at the same time. They grow

Fig. 7 Growth of columnar grains. As metal cools: (A) nucleation begins

at the mold walls; (B) to (D) columnar crystals grow toward the center of
the melt as heat continues to be extracted through the walls; (E) freezing is
complete.
 Solidification of Metals ◄ 9

quite freely until they come in contact with each other and growth of the den-
drites is restricted. The contact between two dendrites results in the formation
of a boundary.

The surface formed by the contact between two adjacent dendrites is called
a grain boundary or an interdendritic boundary. In an as-cast metal these
boundaries are the areas where the liquid froze last. In Fig. 9, the boundaries
are represented on an atomic scale and may be seen to be a zone of disar-

Fig. 8 The progressive growth of dendrites into the liquid during solidifica-
tion. Time gradually increases from (a) to (d). Note the formation of bound-
aries as two dendrites come in contact with one another.

Fig. 9 Appearance of grain boundaries

10 ► Solidification of Metals

ranged atoms. Atoms within each dendrite or grain are perfectly aligned, but
the alignment at the boundary is not perfect. As a result, atoms along the
boundary are packed in a less orderly fashion than atoms within either sur-
rounding grain and try to assume a compromise position with the two adjacent
lattices. The atoms at the grain boundary are shaded and the way they align to
form grain boundaries is clearly evident. Grain boundaries are approximately
three atoms wide. However, because the atoms are so small, etched grain
boundaries appear as thin lines on the micrograph and non-etched boundar-
ies are generally indistinguishable. Because the atoms in the grain boundary
are not properly aligned, the boundary atoms form a high-energy area that is
more easily etched by acid than the adjacent grains. It is that chemical attack
that makes the region visible when examined using a microscope.

Alloying and Impurities

Virtually all metals and all structural metals contain more than one ele-
ment. Regardless of the effort made to refine a metal, gold for example, some
impurities remain. The impurities may be atoms of copper, silver, carbon,
sulfur, or any other element. Gold that is greater than 99.9% pure by weight
is considered pure gold and is termed 24 karat gold. This grade of gold is not
very strong and jewelry made from 24 karat gold deforms relatively easily.
The strength of gold can be increased if copper atoms are mixed with the
gold atoms. A gold-copper mixture that contains 75% Au (gold) and 25% Cu
(copper) by weight is termed 18 karat gold. Copper atoms are not considered
impurities in 18 karat gold. The copper atoms were intentionally added to
improve the strength and decrease the cost of the material and are considered
alloy additions. Alloys are mixtures of two or more elements, and when the
additions are made to a metal with the object of improving some property,
the process is called alloying. The primary difference between alloy additions
and impurities is that the alloy additions are generally desirable while impuri-
ties are typically unwanted.

Impurities may be divided into two classes: elemental impurities that may
be dissolved in metal and impurity compounds such as oxides, sulfides, and
silicates, which do not dissolve in the metal. The compounds are termed in-
clusions and always exist in the metal (or alloy) as particles. Sources of in-
clusions include the refractories in which the metal is melted, dirt that may
be from the original metal stock, residual particles that could not be removed
from the ore during processing, or particles from the fuels and slags added
during refining.

When an ore is processed, it is usually not economically feasible to even

attempt to completely remove all of the impurities. More often, it is important
only to limit the amount of impurity. There are a few cases where an impurity
may actually be beneficial. When tiny particles are present in liquid steel,
they are randomly scattered throughout the melt and form excellent sites for
the nucleation of a solid phase during solidification. Each nucleus eventually
 Solidification of Metals ◄ 11

becomes an individual crystal or grain. If there are many nuclei, there will be
many grains. The more grains present in a given volume of metal, the smaller
the grain size. Small grain size is desirable for most engineering applications
of metals and alloys. Some of the advantages are discussed in other lessons
in this course.

The effect of impurities on nucleation is illustrated in Fig. 10. In compar-

ing these two cooling curves, it is apparent that the amount of supercooling
necessary before solidification is considerably less in the case of an impure
metal and may even be eliminated. The impurities present offer nucleation
sites, providing a base on which the atoms of the liquid metal can deposit to
form the growing solid crystal (grain). The greatest objection to impurities
is that they generally segregate at the grain boundaries. This segregation can
lead to embrittlement or weakness and is therefore undesirable.

Metal solutions have many of the same properties as chemical solutions.

A solution is made up of two parts—the solvent and the solute. The solute is
material that is dissolved, while the solvent is the material that does the dis-
solving. In steel, iron is the solvent and carbon is the solute, and in 18 karat
gold, gold is the solvent and copper is the solute.

Alloying is the intentional mixing of two metals or a metal with a non-

metal. Some metals, when mixed together, are completely soluble in each
other. They can be mixed in any proportion and will be fully soluble in the
solid and liquid states. Copper and nickel are two metals that act in this way.
Any amount of copper can be alloyed with nickel to give a simple solution.
One reason for this solubility is that the atoms of copper are almost the same
type and size as the nickel atoms and both copper and nickel are fcc metals.

Fig. 10 Characteristic cooling curves for pure metals. Less supercooling de-
velops when foreign matter is present, (b), because impurities have a strong
nucleating effect. (The true melting point, however, is not affected.)
12 ► Solidification of Metals

Most metals are completely soluble in each other in the liquid state. The
exceptions include metals whose atoms have large differences in size, chemi-
cal properties, electronic structure, and/or some other fundamental property.
Pairs of liquid metals that show immiscibility (insolubility) at particular tem-
peratures and compositions include aluminum-lead and iron-silver. Liquid
immiscibility in a two-metal system is illustrated in Fig. 11.

Consider the two metals to be metal A and metal B. Pure A is positioned in

the lower left-hand corner of the figure, and as B is added to A, the percentage
of B increases until 100% B is reached at the lower right-hand portion of the
figure. The figure shows that the solubility of the two liquids in each other
increases with increasing temperature and that there is a temperature above
which A and B are completely soluble in each other. This temperature is T
and is represented by the line TY. Above this temperature, any ratio of A to B
could be mixed and only one liquid would be present.

Solutions can also form when the alloys are in the solid state. In these
solutions, the solute atoms replace, or substitute for, the atoms on the solvent
metal crystal structure. These solutions are called “substitutional solid solu-
tions.” In the example mentioned previously, copper and nickel can replace
each other in all proportions. It is more common in metal systems, however,
that each component shows limited solubility for the other, but not complete
solubility.

In some cases, the size of the solute atom is small enough that it can fit in
the small “holes” that exist between the atoms of the solvent crystal structure.
Such solutions are called “interstitial solid solutions.” Atoms of hydrogen,
carbon, boron, and nitrogen are small enough to form interstitial solid solu-
tions in iron. Interstitial solid solutions always show limited solid solubility.

Fig. 11 Diagram showing that two liquids, 1 and 2, are totally immiscible at
low temperatures, and that mutual solubility increases at higher temperature,
until at T′ complete solubility results
 Solidification of Metals ◄ 13

Diagrams are often used to graphically represent the effects of temperature

on solid solubility, melting, and to describe what phases should be present for
various combinations of metal mixtures and temperatures. These diagrams
are termed phase diagrams and are very useful in the metallurgical industry.
Phase diagrams are also referred to as equilibrium diagrams (not always cor-
rectly) or constitutional diagrams.

Phase Diagrams
Phase diagrams are graphs that show the interrelationship between phas-
es, temperature, and composition of an alloy system. Such diagrams are very
important tools in metallurgical study because phase diagrams can be used
to determine what phases should be present, the equilibrium composition of
each phase, and the relative amount of each phase in a given alloy at a spe-
cific temperature. The effects of changes in temperature and/or alloy compo-
sition on the phases present can also be understood. The properties of alloys
are controlled by the nature, amount, and distribution of the phases; thus, an
understanding of phase diagrams is important to the understanding of many
metallurgical processes.

The concept of a phase diagram can generally be appreciated by consider-

ing the addition of sugar to tea. The amount of sugar that can be added to and
dissolved in the tea is temperature dependent and increases as the tempera-
ture increases. An “empty” glass that held sweetened tea will frequently have
sugar crystals remaining on the glass bottom. The sugar crystals represent
one component of the tea-sugar “alloy.” The liquid tea represents another
component. This tea-sugar system can be considered a binary alloy system. A
binary alloy is a system that contains two components. A hypothetical phase
diagram for the tea-sugar binary alloy is illustrated in Fig. 12.

Fig. 12 Hypothetical phase diagram for tea-sugar system

14 ► Solidification of Metals

The line AB separates the liquid region from the liquid-plus-sugar region
and is therefore a boundary between a single-phase and a two-phase region.
The line AB also describes the temperature dependence of the solubility of
sugar in tea. Note that at room temperature the maximum amount of sugar
that can be dissolved in the tea is given by S1. Increasing the temperature in-
creases the solubility of sugar in tea so that the quantity of sugar represented
by S2 could be dissolved. At temperature T3, any amount of sugar less than S3
could be added to the tea and dissolved in the liquid.

If a tea-sugar system containing S2 amount of sugar were held at T2, a sin-

gle-phase solution of sugar in tea would result. If the amount of sugar were
increased to S3, all the sugar would not dissolve unless the temperature were
increased. Because the sugar cannot dissolve, sugar crystals will exist at the
bottom of the glass and a two-phase system (tea with sugar dissolved in the
tea and sugar crystals) will be present. To dissolve all the sugar, the system
would have to be heated. It should also be noted that if the system were at
temperature T3, the amount of sugar dissolved in the tea would not increase
above S3. The addition of more sugar (beyond S3) would simply result in the
presence of more sugar at the bottom of the glass. Many of the characteristics
of this hypothetical tea-sugar phase diagram are typical of some binary phase
diagrams for metallic alloys.

Most engineering alloys do not undergo composition changes during use;

thus, changes in temperature do not result in changes in the concentration of
the alloy elements. The significance of this fact can be seen by consideration
of a portion of the copper-silver phase diagram. The low-temperature portion
of the copper-silver phase diagram is shown in Fig. 13, which is very similar
to the hypothetical phase diagram in Fig. 12.

The left-hand side of the diagram shows that the solubility of silver in cop-
per is limited. It is important to realize that the phase diagram shown in Fig.
13 is describing the solubility of copper in solid silver. The dissolved silver
atoms have replaced copper atoms in the crystal structure, thus silver is a sub-
stitutional solute in the solid copper lattice. The silver solubility refers to the

Fig. 13 Low-temperature portion of the copper-silver phase diagram

 Solidification of Metals ◄ 15

maximum amount of silver that can be dissolved in the copper lattice without
a new phase forming. The line AB represents the temperature dependence of
silver solubility in copper. The right-hand side of the diagram shows that the
solubility of copper in silver is also limited, and the line CD represents the
temperature dependence of copper solubility in silver.

The similarity between lines AB and CD in Fig. 13 and the line AB in Fig.
12 should be emphasized. These lines on a phase diagram represent the ter-
minal solid solubility of silver in copper (AB) and the terminal solid solubil-
ity of copper in silver (CD). The copper phase that has dissolved silver and
the silver phase that has dissolved copper are solid solutions. Note that the
terminal solid solutions originate at the sides of the phase diagram (100%
Cu side and the 100% Ag side). The phase on the left-hand side of the phase
diagram is usually designated by the Greek letter alpha (α); the next phase
encountered as one moves from left to right across the phase diagram is gen-
erally termed the beta phase (β).

Thus, for the phase diagram shown in Fig. 13, the copper-rich phase would
be the α phase and the silver-rich phase the β phase. Another convention for
designating or naming the phases on the phase diagram would be to term
the copper-rich phase (Cu) and the silver-rich phase (Ag). The portion of the
phase diagram shown in Fig. 13 is divided into three regions. The copper-
rich region is to the left of the solvus line (AB). This region is a single-phase
region with only the α phase present for alloys having compositions that lie
within that single-phase region. The α-phase region is shaded by dots in Fig.
14. The silver-rich region to the right of the solvus line (CD) is also a single-
phase region with only the β phase present for alloys having compositions
that lie within the boundaries developed by the vertical line representing
100% Ag and the solvus line CD. The β-phase region is shaded by cross-
hatching in Fig. 14.

The zone between the two single-phase regions contains both the α and
the β phases. This is a two-phase region. Alloys of copper and silver that
have compositions and temperatures that fall within the two-phase zone are

Fig. 14 Labeling of regions in phase diagrams

16 ► Solidification of Metals

mixtures of both phases and are therefore two-phase alloys. Sterling silver
is an example of a two-phase alloy. The improved mechanical properties of
sterling silver over those of pure, single-phase silver are due to the effects of
the second phase.

To consider the silver side of the copper-silver phase diagram, it is conve-

nient to redraw the diagram so that the silver-rich side is on the left-hand side
(Fig. 15). The composition of sterling silver is 92.8% Ag with approximately
7.2% Cu. The composition is shown as the heavy vertical line in Fig. 15.
This line can be used to illustrate an important fact for many alloy systems.
Note that at high temperatures (above approximately 740 °C, or 1364 °F), the
silver alloy is a single-phase alloy, β, while below that temperature the alloy
should be a two-phase β + α. Because of this, when sterling silver is heated
to temperatures above 740 °C, the alloy is all β, but when alloy is cooled to
room temperature the α phase should precipitate and the alloy would convert
from a single-phase alloy into a two-phase alloy through the precipitation of
the α phase. Precipitation occurs because as the alloy is cooled the solubility
of copper in silver decreases. Because of the presence of “excess” copper
(copper that should not be dissolved in the silver) at temperatures below 740
°C, the single-phase alloy becomes a supersaturated solid solution and the
copper-rich phase, α, will precipitate within the silver-rich lattice or matrix.
As the copper-rich phase precipitates, the amount of copper dissolved in the
silver-rich lattice decreases.

At 700 °C (1292 °F) the silver-rich phase contains approximately 6% Cu

and at 600 °C (1112 °F) it contains only 4% Cu. The remainder of the copper
is present in the copper-rich second-phase precipitate. This is shown sche-
matically in Fig. 16(a). As the temperature decreases, the amount of copper

Fig. 15 Silver-rich side of copper-silver phase diagram

 Solidification of Metals ◄ 17

Fig. 16 (a) Schematic representation of microstructure of sterling silver and

actual micrograph of Ag-10Cu alloy. The grains are the β or silver-rich ma-
trix, and the dots are used to represent the α or copper-rich phase. (b) A
longitudinal section of 90Ag-10Cu strip that has been cold drawn then an-
nealed at 732 °C (1350 °F). The section is prepared for examination under a
microscope and the microstructure seen consists of spheroidal particles of a
copper-rich phase present in bands in a silver-rich matrix. Etchant, HNO3 +
K2Cr2O7 and CrO3 + Na2SO4; original magnification, 800×

in the silver matrix decreases and the amount of copper-rich phase increases.
However, this decrease in the amount of copper in the silver matrix cannot
take place in an arbitrary manner. As the alloy cools, copper atoms must
move out of the silver-rich phase and form a specific amount of the copper-
rich phase (Fig. 16b). Movement of the atoms is termed diffusion and re-
quires time. Therefore the precipitation of the copper-rich phase does not
occur instantly, but the precipitation process occurs over a period of time.

Consider the hypothetical situation where a sterling silver alloy sample is

cooled instantly from 800 to 600 °C (1472 to 1112 °F). Because the forma-
tion of the α phase is time dependent, when time equals zero, no α phase will
form. As time increases, the amount of α phase present will increase until the
specific amount required by the phase diagram has formed. The time neces-
sary for the formation of this required amount of α phase is the time neces-
sary for equilibrium to be obtained. The time to reach equilibrium may be ex-
cessive and many studies are stopped when the system is at near equilibrium
rather than at actual equilibrium. Phase diagrams are often near-equilibrium
diagrams and describe what phases would be present if sufficient time were
provided to allow the required diffusion to take place. For many reasons,
including time, commercial operations are often stopped before precipitation
processes are completed.

There are many features common to phase diagrams for metallic systems.
One of the general features of phase diagrams is the lines that show the limit
of solid solubility of one element in another. This feature is termed the solvus
18 ► Solidification of Metals

line. A solvus line for silver dissolving copper is shown in Fig. 15. There are
also a few systems that do not have solvus lines and one feature of such a
system is that the two metals are mutually soluble in each other.

A binary system where the two metals are mutually soluble in the gas-
eous, liquid, and solid states is termed an isomorphous system. However,
from a practical standpoint only the liquid and solid states are important to
most metallurgical processes and most metallurgical phase diagrams show
only the liquid and solid phases. The copper-nickel system provides an ex-
ample of a binary isomorphous system. The copper-nickel phase diagram is
shown in Fig. 17. Note that nickel melts at the specific temperature of 1455
°C (2651 °F) and that copper melts at ~1085 °C (~1985 °F), but that all alloys
of copper and nickel melt over a temperature range rather than at a specific
temperature. The temperature at which melting starts is termed the solidus
temperature. The solidus line or curve in Fig. 17 therefore represents the
temperatures below which the alloy would be a solid under near-equilibrium
conditions.

The liquidus curve represents the temperatures above which, under near-
equilibrium conditions, the alloy would be liquid. The fact that freezing takes
place over a temperature range establishes several interesting problems dur-
ing the solidification of an alloy. An enlarged view of the nickel-rich side of
the copper-nickel phase diagram (Fig. 18) shows that during the cooling of
a 90% Ni, 10% Cu alloy (this type of alloy can be termed a 90Ni-10Cu, Ni-
10Cu, or 90-10 Ni-Cu alloy), the alloy would be liquid at 1465 °C (2669 °F)
and would remain liquid as the temperature was reduced to 1449 °C (2640
°F). At 1449 °C (2640 °F), the alloy would begin to solidify. The first solid
to form would have approximately 3% Cu, as labeled on the phase diagram,
and contain less copper than the alloy contained. The composition of the first
solid to form during the solidification of virtually any alloy will differ from
the composition of the alloy if the solidification process allows time for atom
movement.

Fig. 17 The copper-nickel phase diagram

 Solidification of Metals ◄ 19

Fig. 18 Nickel-rich end of copper-nickel phase diagram

The significance of this observation can be seen if the alloy is cooled to

1420 °C (2588 °F). At that temperature the composition of the solid would be
approximately 7.6% Cu, while the composition of the liquid present would
be approximately 19.6% Cu. This provides the interesting situation that dur-
ing solidification of the 10% Cu alloy, neither the solid nor the liquid contain
10% Cu. The solid is depleted in copper, while the liquid is copper-rich com-
pared to the concentration of copper in the alloy. This effect leads to two po-
tential metallurgical problems with cast structures, coring and hot shortness.
Coring is the development of solidification-induced composition gradients
within the casting, and hot shortness is the development of high-temperature
weaknesses due to the composition gradients.

Coring results during solidification when the rate of cooling is too fast for
diffusion to allow the composition of the solid and liquid phases to adjust to
near-equilibrium conditions. The near-equilibrium freezing of the 10% Ni-
Cu just described would require that the copper concentration of the first
solid to form increase as subsequent solidification took place. For example,
the solid that formed at 1430 °C (2606 °F) would, under near-equilibrium
conditions, contain approximately 4.1% Cu, but when the temperature
reached 1410 °C (2570 °F), the near-equilibrium copper concentration in the
solid phase would be approximately 9.5%. Thus, during cooling from 1430 to
1410 °C (2606 to 2570 °F), copper diffusion into the solid would be required
to increase the copper concentration from 4.1 to 9.5%. Such copper diffu-
sion would require more time than is normally available in the typical so-
lidification process during casting. When the solidification process is too fast
for significant diffusion, the first solid to form will contain less copper than
anticipated from the near-equilibrium phase diagram, and the last liquid to
solidify would contain more copper than anticipated. This phenomenon leads
20 ► Solidification of Metals

to a composition change in the microstructure. This composition change is

termed coring and is illustrated in Fig. 19 for a 70Ni-30Cu alloy. Coring is
common in cast structures.

The composition gradients induced by coring also change the microstruc-

ture of alloys as illustrated in Fig. 19. Additionally, the properties of alloys
are controlled by the composition and microstructure, thus coring changes
the properties of the alloy. Coring can be minimized by very slow cooling
during solidification or by heating the cast alloy to an elevated temperature
and holding for sufficient time to allow diffusion to eliminate the concentra-
tion differences. This process of heating and holding to eliminate composi-
tion differences is termed homogenization or a homogenization anneal. An
example of a homogenized 70Ni-30Cu alloy is shown in Fig. 20. Note the
relatively uniform distribution of copper in the alloy causes the microstruc-
ture to etch uniformly. The higher the temperature of a homogenization an-

Fig. 19 Typical microstructure of an as-cast alloy showing coring. The first-

to-freeze areas appear light; the last-to-freeze areas are gray.

Fig. 20 Homogenized 70Ni-30Cu alloy. Etchant, NaCN(NH4)2S2O8; original

magnification, 100×
 Solidification of Metals ◄ 21

neal, the shorter the time required for homogenization to be complete. Me-
chanical deformation of the alloy at elevated temperatures can also assist the
homogenization process.

The phase diagram in Fig. 18 shows the near-equilibrium solidus and liq-
uidus curves for the nickel-copper alloy system. This diagram shows that for
near-equilibrium solidification the 90Ni-10Cu alloy would totally freeze at
1410 °C (2570 °F). However, actual casting practices cause cooling to be too
fast for near-equilibrium conditions to exist. The actual solid phase contains
less copper than would be predicted under near-equilibrium cooling condi-
tions. Therefore, for practical purposes the solidus curve for the real system
does not coincide with the near-equilibrium solidus curve. Because the solid
contains less than the near-equilibrium amount of copper at any given tem-
perature, the actual solidus curve is displaced (Fig. 21). The liquidus curve is
also displaced (as illustrated) after solidification begins; however, it is impor-
tant to recognize that the displacement does not occur until after the initiation
of the solidification process. The displacement of the liquidus curve has little
effect on any metallurgical process but the significance of the displacement
of the solidus curve is very important. This importance can be seen if the
solidification of a 90Ni-10Cu alloy is considered.

Under near-equilibrium conditions the alloy would solidify at 1410 °C

(2570 °F); however, because rapid nonequilibrium cooling causes a displace-
ment of the solidus curve, the solidification temperature is lowered. For the
conditions illustrated in Fig. 21, the alloy would not be totally solid until

Fig. 21 Displacement of solidus and liquidus curves due to coring in solidification

structures
22 ► Solidification of Metals

the temperature was lowered to 1390 °C (2534 °F). Thus, the actual solidi-
fication temperature is 20 °C (36 °F) lower than anticipated from the near-
equilibrium phase diagram. The amount of displacement of the solidus curve
will depend on the cooling rate and the nature of the alloy. The solidus curve
depends on the average composition of solid present during freezing. Shifts
of the solidus curve as much as several hundred degrees have been observed
for nonequilibrium cooling conditions.

If an alloy that is substantially cored is heated to temperatures close to the

near-equilibrium melting temperature, the grain boundaries may melt (this
is termed grain-boundary liquation or incipient melting). Such melting will
immediately lower the strength of the alloy and the alloy could easily fall
apart if mechanical forces acted on the alloy while the grain boundaries were
liquid. This lack of strength and tendency to fracture during mechanical pro-
cessing is termed hot shortness.

Eutectic Systems

The isomorphous phase diagram shown in Fig. 17 is typical of several

binary systems. Examples of isomorphic binary metallic systems are pro-
vided in Table 1. For alloys to show this behavior, both metals in the binary
alloy must have similar crystal structure, similar chemical properties, and
the size of the atoms must also be similar. The number of possible binary
alloy systems satisfying all of these conditions is very limited. Most metal
combinations do not show complete mutual solid solubility. Therefore, phase
diagrams that show limited solid solubility such as found in the copper-silver
alloy system are quite common. One of the common types of phase diagrams
for binary metal systems showing limited solid solubility is the eutectic phase
diagram.

In eutectic systems, the solute additions lower the liquidus curve of the
solvent. On the right-hand side of the phase diagram one element (element
A) is the solvent and another element is the solute (element B), while on the
left-hand side of the phase diagram element B is the solvent and element A
is the solute. Therefore the liquidus curve is lowered from both sides of the
phase diagram. The two liquidus curves will meet at a minimum temperature
known as the eutectic temperature. A typical eutectic is the copper-silver al-

Table 1 Binary alloy systems having isomorphic phase diagrams

Silver-gold (Ag-Au) Hafnium-zirconium (Hf-Zr)
Silver-palladium (Ag-Pd) Molybdenum-tantalum (Mo-Ta)

Bismuth-antimony (Bi-Sb) Molybdenum-tungsten (Mo-W)

Calcium-strontium (Ca-Sr) Nickel-palladium (Ni-Pd)

Niobium-tungsten (Nb-W)
 Solidification of Metals ◄ 23

loy system represented by the phase diagram shown in Fig. 22. Copper ad-
ditions to silver lower the liquidus temperature of silver, and silver additions
to copper lower the liquidus temperature of copper. Such lowering of the
liquidus temperature by alloy additions is common to many materials.

The chemical term used for lowering the temperature is molar freezing
point depression. Most motorists are familiar with the freezing point depres-
sion effect of “antifreeze” on the freezing point of water. Also, salt added
to water lowers the freezing point of water. This depression of the liquidus
temperature by alloy additions to metals is identical to the depression of the
freezing point of water by additions of antifreeze or salt. Because the eutectic
point is the minimum temperature in the liquidus curve, the eutectic alloy has
the lowest melting point of any alloy in a eutectic system. The low melting
point of eutectic alloys is useful in metallurgical practices. Eutectic mixtures
of salts can be melted at low temperatures relative to any of the component
salts, thus molten eutectic salt baths are used as heating media for many met-
al and alloy parts in heat treating. Eutectic alloys have lower melting points
than any of their components. Such alloys are useful for joining operations
such as soldering and brazing. Thus, lead-tin alloys are frequently used as
solders and copper-silver alloys are used as brazes. (The primary difference
between a solder and a braze is the melting temperature, with brazes melting
at temperatures above approximately 450 °C, or 842 °F, and solders melting
at temperatures below 450 °C.)

The solidification (or freezing) of a eutectic alloy provides several inter-

esting features. Consider both an 80Ag-20Cu and an 80Cu-20Ag alloy. The
compositions of these alloys are indicated by the heavy vertical lines on Fig.
22. The 80Ag-20Cu alloy is all liquid at approximately 820 °C (1508 °F),
while the 80Cu-20Ag alloy must be heated to approximately 960 °C (1760
°F) before it is all liquid. During cooling the first solid to form in the 80Cu-

Fig. 22 The copper-silver eutectic system

24 ► Solidification of Metals

20Ag alloy is the α phase, which is copper rich and contains 93% Cu. If the
alloys were very slowly or “near-equilibrium” cooled, the composition of the
solid would follow the solidus line. The first solid phase to form in the 80Ag-
20Cu alloy is silver-rich solid, β, that will contain 91.2% Ag and 8.8% Cu
when the eutectic temperature 780 °C (1436 °F) is reached. The copper-rich
solid, α, formed in the 80Cu-20Ag alloy will contain 92.1% Cu and 7.9% Ag
when the eutectic temperature is reached. The composition of the remaining
liquid in both alloys follows the liquidus lines during solidification. Thus, the
last liquid to freeze in both alloys will be identical and contain approximately
28.1% Cu and 71.9% Ag. This last liquid to freeze has the eutectic composi-
tion. Freezing of this eutectic liquid requires the formation of two solids with
very different chemical compositions. One solid must contain 92.1% Cu and
therefore be copper rich while the other solid must contain 91.2% Ag and
therefore be silver rich. This nearly simultaneous formation of two solids of
different compositions from a single liquid can be described by the reaction:

L → Solid I + Solid II

or for this particular case:

L71.9% Ag → α7.9% Ag + β91.2% Ag

where α is the copper-rich phase, which is copper with silver in substitutional

solid solution, and β is the silver-rich phase, which is silver with copper in
substitutional solid solutions. Diffusion is required for a liquid containing
71.9% Ag to solidify into two solids, one containing only 7.9% Ag and an-
other containing 91.2% Ag.

The solidification of a eutectic alloy therefore requires significant changes

in the local chemical compositions. The process required for such composi-
tion changes is schematically illustrated in Fig. 23. Note that according to the
illustration, copper must be rejected during the formation of the silver-rich
solid. This rejection causes the copper concentration in the liquid adjacent to
the solid to increase and thus promotes the formation of the copper-rich solid.
Continuation of this solidification process would cause silver to be rejected
when the copper-rich region formed. This rejection of silver would promote
the formation of the silver-rich phase, and the solidification process would
continue. This type of solidification requires that the two eutectic phases be
intermixed so that they “surround” each other.

There are several types of eutectic microstructures. However, the common

feature of each type is that the two phases are intermixed. Micrographs of
several types of eutectic microstructures are given in Fig. 24.
 Solidification of Metals ◄ 25

Fig. 23 Illustration of composition changes associated with solidification of

a eutectic liquid

These microstructures all show that one phase is surrounded by another.

The phase surrounding the rods, platelets, or angular grouping of the other
phase is termed the matrix phase. (The matrix phase is often referred to as the
continuous phase, and the other phase, correspondingly, the discontinuous
phase.) Generally, the matrix phase is also the most prevalent phase in the mi-
crostructure. Fast cooling (or more accurately, fast solidification) promotes a
fine distribution of the eutectic phases, while slow cooling promotes a more
coarse distribution. It is now worthwhile to reconsider the solidification of
the 80Cu-20Ag and the 80Ag-20Cu alloys.

As the 80Cu-20Ag alloy is cooled from 1000 °C (1832 °F), solidification

begins at 960 °C (1760 °F). This first solid to form is the copper-rich phase.
This phase is termed the proeutectic (meaning before the eutectic) solid, and
the proeutectic solid continues to form as the temperature is lowered and the
solidification process continues. When the eutectic temperature is reached,
the remaining liquid undergoes the eutectic reaction and develops a micro-
structure similar to that shown in Fig. 24(c). However, the proeutectic phase
that formed before the eutectic temperature was reached remains in the al-
loy and can be seen in the microstructure. Micrographs illustrating alloys
containing proeutectic phases as well as the eutectic structure are shown in
Fig. 25. Solidification of the 80Ag-20Cu alloy would be similar except that
in this case the proeutectic phase, β, begins to form at 810 °C (1490 °F) and
is silver rich.
26 ► Solidification of Metals

Fig. 24 Examples of various eutectic microstructures. (a) Aluminum-silicon

eutectic. Note that the silicon-rich particles (dark phase) are surrounded by
the aluminum-rich matrix. (b) Cadmium-tin eutectic. Note that the cadmium-
rich phase (dark lines) is surrounded by the tin-rich matrix. (c) Aluminum-
magnesium eutectic. Note that the Mg2Al3 phase (dark platelets) is surround-
ed by the aluminum-rich matrix. (d) Manganese-tin eutectic. Note that the
Mg2Sn phase is surrounded by the tin-rich matrix.

The amount and nature of the proeutectic phase influences the properties
of the alloy as do the relative amounts and distribution of the two phases in
the eutectic. Thus, knowledge of the relative amounts of the various phases
present is useful for predictions of properties. The phase diagram can be used
to estimate the relative amounts of the phases present.

The Lever Rule

The lever rule or lever principle can be applied to a phase diagram to de-
termine the composition and relative amounts of phases present (under equi-
librium conditions) for any combination of alloy content and temperature.
Consider, for example, the portion of the phase diagram shown in Fig. 26 and
 Solidification of Metals ◄ 27

Fig. 25 Microstructures of alloys showing proeutectic phases and eutectic structures.

(a) Aluminum-silicon eutectic with proeutectic aluminum-rich phase (light phase). (b)
Lead-tin eutectic with proeutectic lead (dark phase). (c) Aluminum-zinc eutectic with
proeutectic zinc phase (light phase)

an alloy of A and B that contains 30% B. For general purposes, the composi-
tion is considered CA. If this alloy is at a temperature above temperature T1,
the alloy is all liquid. If the temperature is below T2, the alloy is all solid.
Under near-equilibrium conditions, the solid at temperatures below T2 would
have a homogeneous composition of 70A-30B. At temperatures between T2
and T1, the alloy would be partially liquid and partially solid.

If the temperature of the alloy is T3, information about the composition

of the phases and the relative amounts of the phases can be obtained as fol-
lows. A constant temperature or isothermal line is drawn horizontally within
the two-phase field or region. This line, labeled ab in Fig. 26, is termed a tie
line. It intersects the vertical line representing the composition of the alloy at
point c, and the liquidus and solidus curve, at points b and a, respectively.
The composition of the solid is given by dropping a vertical line from a to the
28 ► Solidification of Metals

Fig. 26 Illustration of lever principle as applied to a phase diagram

composition axis. In this case, the composition of the solid is approximately

18% B; however, for generalization the composition will be considered CA.
The composition of the liquid is determined by dropping a vertical line from
b to the composition axis. In this case, the composition of the liquid CL is
determined to be 52% B.

The use of a tie line enables the determination of the composition of the
phases present in any given binary alloy at any temperature in a two-phase
field. The relative amounts of the phases present can also be established.

The relative amounts of solid and liquid present are determined by divid-
ing the tie line (ab) into two parts, ac and cb. The length of the line cb com-
pared to the total length of the tie line establishes the fraction of solid in the
alloy, while the length of the line ac compared to the total length of the tie line
establishes the fraction of liquid. This can be expressed as:

%Liquid = [(CA – CC)/(CL – CC)] 100

and

%Solid = [(CL – CA)/(CL – CC)] 100

where CA is the composition of the alloy expressed in %B, CL is the compo-

sition of the liquid expressed in %B, and CC is the composition of the solid
 Solidification of Metals ◄ 29

expressed in %B. For the example given in Fig. 26, the percentage solid
would be:

% = [(52 – 30)/(52 – 18)] 100 = (22/34) 100 = 64.7%

and the percentage liquid would be:

%L = [(30 – 18)/(52 – 18)] 100 = (12/34) 100 = 35.3%

The %L + % solid α must equal 100 because α and liquid are the only
phases present. Thus, after the calculation is made for one phase, the percent-
age of the other phase can be determined by subtraction.

If equilibrium conditions do not exist, but the compositions of the alloy

and the phases present are known, the lever rule can be used to calculate the
amount of phases present. The lever rule, however, cannot be used to predict
the composition of phases under nonequilibrium conditions.

The general rules for phase determinations illustrated in the example just
given can be applied to any binary phase diagram, regardless of how com-
plicated the phase diagram might be. To illustrate this, the copper-zinc phase
diagram is reproduced as Fig. 27.

This diagram has several features that will not be discussed in this lesson;
however, the lever principle can be applied to any region of the diagram.
Before the lever principle is applied, several other points should be made.

The phase diagram shown in Fig. 27 is divided into several regions; the
labeled regions, L, (Cu), β, β′, γ, δ, e, and (Zn) are single-phase regions in
the diagram. Any time the combination of alloy content and temperature lies
within a single-phase region, only one phase should be present and the com-
position of that phase will equal the composition of the alloy. For example,
the 65Cu-35Zn alloy (indicated on the diagram as Alloy 1 on the composition
axis) would be a single-phase liquid at temperatures above 920 °C (1688 °F)
and would be a single-phase solid (Cu) at temperatures between 760 and 210
°C (1400 and 410 °F). At all other temperatures, the alloy should exist as a
two-phase alloy.

The two-phase regions on the phase diagram separate the single-phase

regions. The zone between the (Cu) and the L regions is a (Cu) + L zone. The
zone between the (Cu) and β region is a (Cu) + β zone. The lever rule can be
applied in any two-phase zone. Consider, for example, a 75Zn-25Cu alloy
(Alloy 2 on the diagram) that is held at 400 °C (752 °F) until equilibrium is
obtained. The tie line is again shown as line ab with point c representing the
30 ► Solidification of Metals

Fig. 27 The copper-zinc phase diagram. (Source: T.B. Massalski, P.R. Subra-
manian, H. Okamoto, and L. Kacprzak, Ed., Binary Alloy Phase Diagrams,
2nd ed., Vol 2, ASM International, 1990, p 1509)

intersection of the vertical line at the alloy composition with the tie line. In
this case, Cc = 75% Zn, Ca = 68% Zn, and Cb = 79% Zn. Both the γ and the
ε phases are solids in this case. The fraction of γ phase will equal the length
of line cb divided by the length of line ab; thus, the percentage can be given
by the lever rule as:

% γ = [(Cb – Cc)/(Cb – Ca)] 100 = [(79 – 75)/(79 – 68)] 100 = (4/11) 100
= 36.4%
Calculations of this type can be made at any temperature for a two-phase
region of the phase diagram, provided the alloy composition is known. The
primary reason to make such calculations is the dependence of properties on
the amount of the various phases present in the microstructure.

Eutectoid Reactions

Relatively careful examination of the phase diagram shown in Fig. 27

shows that although there are no eutectic reactions, there are alloy changes
that have a similarity to the eutectic reaction. The eutectic reaction is:
 Solidification of Metals ◄ 31

L → (Solid)I + (Solid)II
If an alloy with 74% Zn is held just above 560 °C (1040 °F), the alloy
would be all δ. If the temperature were reduced to 555 °C (1031 °F), the alloy
would transform from all δ to two other phases, γ and ε, when 558 °C (1036
°F) was reached. The reaction to describe this change is:

(Solid)I → (Solid)II + (Solid)III

This type of reaction is termed a eutectoid reaction and is similar to the
eutectic reaction except that the high-temperature phase is a solid rather than
a liquid.

Peritectic Reactions

The peritectic reaction occurs on cooling when the liquid phase (L) reacts
with one solid phase (S1) to form another solid phase (S2). The peritectic reac-
tion can be summarized by:

L + S1 → S2

An equilibrium diagram of metals A and B that has a peritectic reaction is

shown in Fig. 28. An alloy of composition X (the peritectic composition) will
undergo the peritectic reaction during near-equilibrium cooling.

Fig. 28 Constitution diagram of the binary alloy X of the metals A and B with
a peritectic reaction
32 ► Solidification of Metals

Above the liquidus, the alloy will exist as a homogeneous liquid solution
of composition X. After cooling to temperatures slightly below the liquidus
line (at temperature T1, for example) a solid solution (of composition α1 will
start to crystallize. From temperature T1 at the liquidus to slightly above the
peritectic reaction temperature T2, the solid solution will continue to crystal-
lize, with the liquid composition changing from L1 to L2 and the solid com-
position changing from α1 to α2.

Just slightly above temperature T2, the following amount of each phase
will be present:

% α2 Solid solution = [(L2 – β2)/(L2 – α2)] 100

= [(70 – 55)/(70 – 25)] 100
= (15/45) 100
= 33%
% Liquid= [(β2 – α2)/(L2 – α2)] 100
= [(55 – 25)/(70 – 25)] 100
= (30/45) 100
= 67%
At temperature T2, when equilibrium conditions prevail, liquid of com-
position L2 and solid solution of composition α2 react to form a single new
solid-solution phase of composition β2. From the temperature T2 to room
temperature, T3, the solid cools without further change in the composition of
the solid. The solidification of alloys with other compositions is more com-
plicated and is not discussed in this lesson.

Nonequilibrium Solidification of Metals

The solidification of alloys under nonequilibrium conditions, which is

quite common in commercial casting, produces several microstructural fea-
tures in addition to coring. When an alloy solidifies under equilibrium condi-
tions, the grains formed are generally equiaxed (equal in size in all directions)
and the composition throughout the individual grains is uniform. However,
when temperature gradients exist during solidification, even in a pure metal,
the grains that form will tend to be elongated in the direction of the largest
gradients (rapid heat flow). This type of grain is formed when a pure metal
is poured (cast) into a mold that has a temperature significantly lower that of
the molten metal. These grains have their long axis perpendicular to the mold
wall and are called columnar grains because of their shape. In the case of the
solidification of a pure metal (or a eutectic alloy) the freezing front is rela-
tively smooth, as illustrated in Fig. 29. The cooling rate and crystal-forming
 Solidification of Metals ◄ 33

Fig. 29 Columnar grains formed during solidification under nonequilibrium

conditions in a pure metal or eutectic alloy

tendency of the metal will determine the degree of smoothness and the speed
at which the solidification front advances. In general, the faster the rate of
cooling, the smoother the solidification front.

When an alloy solidifies under nonequilibrium conditions the freezing

front is rough. The reasons for this roughness are the differences in tempera-
ture and in composition that occur as solidification proceeds. The composi-
tion of the first solid to form differs from that of the liquid and this difference
imposes compositional gradients in the liquid. The compositional gradients
superimposed on the thermal gradients cause solidification to proceed at dif-
ferent rates in different locations and directions. These differences in solidi-
fication rates cause the solidification front to become rough and irregular, as
illustrated in Fig. 30.

A condition frequently encountered in a commercial casting in which the

mold wall acts as a large heat sink is illustrated in Fig. 31. The mold wall
may be water cooled, have a high thermal conductivity, and/or a large mass.
In any case, the liquid next to the mold wall is suddenly plunged well below
its freezing point, and a large number of solid grains are nucleated in this
undercooled region. This produces a “chill zone” of fine equiaxed grains near
the mold wall. As these grains form, they release the latent heat of solidifica-
tion (heat of fusion) that raises the temperature of the adjacent liquid, slowing
the solidification process and causing freezing to continue by the growth of
columnar grains as described by Fig. 30.
34 ► Solidification of Metals

Fig. 30 Solidification of a solid-solution alloy under nonequilibrium condi-

tions

Fig. 31 Solidification of an alloy cast against a chilled mold wall. A chill zone
of fine equiaxed grains forms next to the mold wall.

During solidification of an alloy a mushy mixture of liquid and solid de-

velops. The mushy mixture of phases exists over the temperature range where
solidification is occurring. For near-equilibrium conditions this mushy zone
can be determined from the equilibrium diagram of the alloy. The solidifica-
tion range is defined as the temperature difference between the start and fin-
ish of solidification of the alloy. In other words, it is the difference between
the liquidus and solidus temperatures of a given alloy. Under the conditions
of actual solidification processes the mushy range may extend beyond that
 Solidification of Metals ◄ 35

shown in the equilibrium diagram. Hot shortness may occur any time a cast-
ing is mechanically processed in the mushy range.

Grain Refining

The properties of metals and alloys are influenced by the size of the grains
in the product. Fine grains are generally desirable, thus effort is frequently
made to minimize the grain size in an individual casting. The grain size of
cast aluminum may be refined by adding an element or compound to the
molten aluminum before the casting process is initiated. In order to act ef-
fectively as a grain refiner, the addition must promote nucleation of the solid
phase and remain relatively uniformly dispersed throughout the molten met-
al. Titanium alone, or in combination with boron-bearing compounds, is the
most common grain refiner for aluminum castings. The grain refiner is usu-
ally added immediately before pouring the molten metal. Titanium and boron
can be added as salts (K2TiF6 or TiBF4) or as binary alloys. These additions
are generally termed hardeners and a 5% Ti-Al hardener is commonly used.
The term hardener is used because the resulting fine-grain casting is harder
than a casting with larger grains. Grain refinement by the addition of harden-
ers is used in many alloy systems because only small amounts of the hardener
are required to provide significant improvements in mechanical properties.

Segregation and Shrinkage

The properties of a casting are influenced by composition as well as grain

size. However, as we have seen, the composition of a casting may vary from
place to place if the solidification process is rapid and if no homogenization
anneal has been given to the as-cast product. Segregation is the term used to
describe the condition where differences in chemical composition exist in
different areas of an ingot or casting. If these differences are present within
or around the grains themselves, then the term microsegregation is applied.
Macrosegregation is the term applied to segregation across a whole ingot or
casting. Various types of segregation are summarized in Fig. 32.

Fig. 32 Classification of segregation

36 ► Solidification of Metals

Microsegregation can be the result of coring or caused by the segregation

of impurities and insoluble constituents at the grain boundaries. During so-
lidification of an alloy, the insoluble constituents and impurities are pushed
ahead of the solidification front and accumulate at the edges of the freezing
grains. When solidification is completed the grain boundaries are typically
areas rich in impurities.

Macrosegregation can occur before or during solidification. Before solidi-

fication begins, gravity may cause segregation of constituents to the bottom
or top of the casting if not stirred. During solidification, direct segregation,
inverse segregation, and liquation can take place.

Direct (or normal) segregation requires a temperature gradient between

the mold and the freezing alloy. When the liquid alloy is poured into a mold,
the first metal to freeze will form a shell on the cold mold wall. The metal will
then proceed to freeze toward the center of the ingot. Because the last metal
to freeze will be richer in the alloy additions, the center of the ingot will have
a higher concentration of these constituents than will the outside.

Inverse segregation is the opposite of direct segregation and is very com-

mon in aluminum alloys. In inverse segregation the high concentration of alloy
elements is found at the outer areas of the casting instead of at the center. The
occurrence of coring and the presence of a temperature gradient are necessary
for inverse as well as direct segregation. The chief factor causing inverse seg-
regation is the shrinkage of the metal during solidification. The cooling that
occurs during solidification causes the solid dendrite arms that extend into the
casting to shrink. This shrinkage creates gaps between the arms. The alloy-rich
liquid metal at the center of the ingot is forced into the gaps. Aluminum alloy
ingots and castings have a tendency toward inverse segregation because they
have such large solidification shrinkage, approximately 6%.

The term liquation is applied to the process whereby the alloy-rich con-
stituents ooze to the outside surface of the cast ingot. Liquation occurs by the
same process as inverse segregation in combination with partial remelting
of the metal in contact with the cold mold surface. The remelting may occur
because of the release of the heat of fusion during freezing.

Even if significant segregation does not occur, the casting or ingot will
be left with shrinkage porosity if the liquid metal is unable to fill the cavi-
ties formed by the solidification shrinkage. The amount and distribution of
shrinkage porosity depends on the solidification range of the alloy being cast
and the manner in which the liquid metal feeds the gaps and holes left by so-
lidification shrinkage. In general, it is more difficult to feed the shrinkage in
alloys with wide solidification ranges because it is harder for liquid metal to
flow through the network of dendrite arms, in the mushy zone of liquid plus
solid, in order to feed the shrinkage.
 Solidification of Metals ◄ 37

Ingot Solidification

Freezing of ingots proceeds from the mold walls and bottom toward the
center of the ingot; thus the direction of freezing is perpendicular to the mold
walls and bottoms. As would be expected, freezing of the liquid steel pro-
ceeds selectively with an increasing amount of the alloying elements col-
lecting in the last liquid to freeze. The crystals grow (from the melt) perpen-
dicular to the mold walls, and because there are a myriad of crystals growing
simultaneously in an ingot, small microscopic-sized pools of alloy-rich liquid
are trapped between the dendrites. The formation of dendrites and the ten-
dency for segregation have a significant effect on ingot solidification.

The initial crystals to form are numerous small treelike figures (dendrites
or dendritic structure) pushing their branches farther and farther into the liq-
uid metal and trapping small pools of liquid metal made up of atoms with de-
creasing mobility (or energy). This growth process causes minor gradations
in the concentration of any element, even within a given grain. The trunks
and branches of the dendrite will be rich in the base metal, while the space
between them will be richer in the alloy elements.

Because the diminishing pool of liquid steel at the center of the ingot has
been enriched in alloy additions, it will have a lower melting point than any
other zone of the ingot. If the ingot were sectioned longitudinally and etched,
three distinct types of crystals would be revealed, proceeding from the ingot
surface toward the ingot center: (1) at the ingot surface a very thin zone that
may contain small equiaxed crystals (this region is termed the chill zone and
may not occur in some ingots); (2) a zone of “columnar” crystals, perpen-
dicular to the ingot surface; and (3) a zone of equiaxed crystals in the center
portion of the ingot.

The columnar crystals are formed because the rate of cooling is rapid due
to the initial temperature difference between the liquid metal and the mold
walls. Columnar grains in steel ingots may be apparent without magnification
or at low magnification. The columnar zone ceases to grow when the mold
wall is sufficiently heated and the heat extracting ability of the mold suffi-
ciently diminished, so solidification of the center of the ingot proceeds at a
comparatively leisurely pace and the central zone of the casting is developed.

When fully deoxidized steel solidifies from the liquid state, it contracts to
a certain degree. Deoxidation of a steel is the removal of oxygen from the
liquid metal. This is often accomplished by adding aluminum to the steel to
form aluminum oxide particles. These particles are termed inclusions and
a steel that has experienced oxygen removal by aluminum additions is said
to have been aluminum killed. The final stages of contraction occur in the
part of the ingot that solidifies last (i.e., the center of the ingot) and this final
shrinkage gives rise to a shrinkage cavity in ingots of many killed steels. This
38 ► Solidification of Metals

shrinkage cavity is known as a “pipe,” and steel exhibiting this condition is

known as “piped” steel.

The segregation and piping that may occur in ingots may cause “planes of
weakness” in the cast ingot. Molds design can influence, and even minimize,
the planes of weakness. For example, a square ingot would have planes of
weakness at the ingot corners due to the effect of this design on the growth
of columnar crystals. Crystals growing from the side walls meet on a plane
projected toward the opposite corner. This type of ingot structure tends to
rupture during rolling or forging. For this reason, mold corners are rounded.
However, these planes of weakness persist to some degree in all ingots and
must be considered in initial rolling or forging operations to avoid rupture.
Very light reductions must be made initially when rolling or forging the ingot
until the crystals are refined and the weak areas are removed. A relatively low
pouring temperature and a consequent rapid rate of solidification of the steel
in the ingot mold minimize the growth of columnar crystals by maximizing
the size of the chill zone in the ingot. Because the chill zone has no “planes of
weakness” the processing of ingots with relatively large chill zones is easier
than the processing of ingots with primarily columnar grains. The rate at
which the metal freezes is dependent on the pouring temperature and the rate
of heat transfer from the liquid metal through the mold wall. Constant effort
is made to improve mold design and thus improve the properties of the cast
product from the mold.

Summary

This lesson describes a metallic or crystalline solid and develops the con-
cepts required for the metallurgical use of phase diagrams. The techniques
necessary to use a phase diagram to determine the phases present, the com-
position of each phase, and the relative amounts of each phase for a specific
alloy at a specific temperature are presented. Microstructural development
during solidification and during phase transformations is introduced. Eutectic
and peritectic systems are described and typical examples of microstructures
are given. The role of diffusion in microstructural development is empha-
sized and because time and temperature are required for equilibrium condi-
tions to exist, the concept of nonequilibrium structures is introduced. This
lesson provides the background necessary for the introduction of the effects
of heat treatments and thermomechanical processing on the properties of
metals and alloys.

ACKNOWLEDGMENTS

This lesson is revised and updated with the help of:

Kenneth W. Burris, Sr.

Engineering Specialist, Gear Engineering
 Solidification of Metals ◄ 39

Technical Reviewers
Dr. Richard Zordan
Dr. Robert L. Freed
40 ► Solidification of Metals

Metric Conversion Factors

To convert from To Multiply by
in. mm 25.4
in. m 25.4 × 10−3
mil µm 25.4
µin. µm 25.4
in.2 m2 6.45 × 10−4
in.3 m3 1.64 × 10−5
ft m 3.048 × 10−1
ft2 m2 9.29 × 10−2
ft3 m3 2.831 × 10−2
oz g 2.834 × 10
lb kg 4.536 × 10−1
Btu J 1.054 × 103
Btu/lb · °F J/kg · K 4.18 × 103
Btu/ft · h · °F W/m · K 1.73
in./in. · °F m/m · K 1.8
psi Pa 6.895 × 103
psi kPa 6.895
ksi kPa 6.895 × 103
ksi MPa 6.895
ksi · in.1/2 MPa · m1/2 1.099
ksi√in. MPa√m 1.099
ozf gf 28.4
lbf kgf 4.536 × 10−1
lbf N 4.448
ft • lbf N • m (or J) 1.356
lbf/in.2 kgf/cm2 14.223
lbf/in.3 kgf/m3 2.768 × 104
lb/ft3 kg/m3 16.019
lb/in.3 g/cm3 2.768 × 10
lb/in.3 kg/m3 2.768 × 104
gal (U.S. liquid) L 3.785
gal (U.S. liquid) m3 3.785 × 10−3
lb/gal g/L 119.826
ft/gal ml/L 748
°F °C (°F − 32)/1.8
°F K (°F + 459.67)/1.8
°C °F (°C • 1.8) + 32
°C K °C + 273.15
K °C Κ − 273.15

Abbreviations
J joule m meter Pa pascal
kgf kilogram force mm millimeter K kelvin
L liter N Newton W watt

Multiple and submultiple units

1012 tera 10−1 deci
109 giga 10−2 centi
106 mega 10−3 milli
103 kilo 10−6 micro
102 hecto 10−9 nano
10 deka 10−12 pico