SOLIDIFICATION OF METALS

When metals are in liquid state there is no orderly arrangement of the atoms which are free
to move about with respect to each other, thus in the liquid state, metals possess mobility.
As the temperature of a molten metal falls a point is reached where the metal starts to
solidify.
At this point the atoms change from a disordered or amorphous state to an ordered or
crystalline state.
Like all pure crystalline substances, pure metals solidify at a fixed temperature as shown in
Fig. 1 (a) below.

However, under industrial conditions, such purity will not be obtained and crystal nucleus
will form around an impurity particle such as a particle of slag.
If the purity is of high order, some under-cooling may occur before nucleation (the
formation of the nucleus of the crystal) commences. (Refer to Fig. 1 (b) below).

Amorphous (non-crystalline) solids such as glass, pitch and some polymers (plastics)
exhibit no such change point and there is a gradual change from the liquid state to the solid
state. (See Fig. 1 (c) below)

Once the nucleus of the crystal forms, it provides a solid/liquid interface where
crystallization will proceed.
For metals, the nuclei will generally be made up of face-centred cubic, body-centred or
close-packed hexagonal unit cells.
As the crystals grow they tend to develop spikes and change into “tree-like” shapes called
dendrites.
Fig. 2 below shows a typical metal dendrite.
The dendrite crystal grows until the spaces between the branches fill up.
Growth of the dendrite ceases when the branches of one dendrite meet those of an adjacent
dendrite.
This process continues until eventually the entire liquid solidifies.
At this point there is little trace of the dendrite structure left, and it is only possible to see
the grains into which the dendrites have grown.

~Page 1 of 18~
The steps in the growth pattern of a crystal from nucleus to grain are shown in Fig. 3 below.

The reason for dendritic growth is as follows.

When a solid metal is heated, its atoms vibrate about fixed points called lattice points, each
atom being held in place by electrostatic forces of attraction.
As the temperature of the metal increases the energy of each atom increases.
At a certain temperature (depending upon the metal) the atoms become so agitated that
they escape from their fixed positions.
Thus the structure begins to lose rigidity, that is, it begins to melt.
In moving from their fixed positions the atoms and molecules work against their binding
forces and energy is used up.
This energy is replaced by the heat source of the furnace.
The replacement heat energy producing fusion (melting), rather than causing a rise in
temperature, is called latent heat of fusion.

~Page 2 of 18~
When a molten metal at its fusion (melting) point solidifies it gives up its latent heat
energy, that is, it gives up the heat energy originally required to cause melting.
Thus, during solidification, the metal/liquid interface is warmed up by the release of latent
heat energy.
This slows down or stops further solidification occurring in that direction, and
solidification recommences in some other direction where the temperature is sufficiently
low.
The result of this action is for spikes of solid metal to occur in regions where the liquid is
coolest.
As these new spikes warm up, forward growth is again retarded and secondary and even
tertiary spikes are formed to produce the typical dendrite.

GRAIN STRUCTURE
Although metals are crystalline solids, this is not immediately apparent when they are
examined under a microscope.
Fig. 4 (a) below shows the appearance under a microscope of a typical metal specimen
which has been polished and etched.
Although obviously granular, it is difficult to identify the geometric regularity when they
are free to grow without restraint or interference.
In a metal, many crystals commence to grow at the same time and eventually collide with
each other so that growth becomes restricted and their boundaries become distorted.
The term grain is used to describe crystals whose geometric shape has been distorted by
contact with adjacent crystals so that their growth is impeded.

Fig. 4 (b) below shows, diagrammatically, how the atoms within a grain can have the
regular geometric space lattice expected of a crystal, and how that pattern breaks down at
the grain boundary to make way for the geometric space lattice in an adjacent grain.

~Page 3 of 18~
EFFECT OF COOLING RATE ON GRAIN GROWTH
Although it is hard to relate a dendrite to the well-ordered crystal structures previously
considered, remember that the unit cells and space lattices are very, very small even when
compared in size with a dendritic spike.
Thus, during solidification and crystallization, the ordered pattern of the space lattice is
still being built up, but the rate of growth is not uniform in all directions.
Another factor that influences the size and shape of the grains as a metal cools is the rate of
cooling.
Fig. 5 below shows a section through the corner of a large casting.

Notice that the grain structure is not uniform but varies with the shape of the mould and
the rate of cooling.
Fine chill, grains occur at the surface of the casting where the molten metal first comes into
contact with the cold surface of the mould.
At the centre of the casting, where the rate of cooling is relatively slow, large equi-axed
grains are formed.
At an intermediate position, long columnar grains are formed at right angles to the surface
of the casting.
These result in a “plane of weakness” at each corner of the casting which can overcome, to
some extent, if the corners are given a large radius.

Sand and metal oxides (slag) adhering to the surface of the casting, together with fine, hard
chill crystals just below the surface can cause premature wear and even serious damage to
cutting tools.
It is essential, therefore, for the first cut to be sufficiently deep for the tip of the cutting tool
to operate below this surface zone as shown in Fig. 6 below.
To ensure that the machined faces of the casting will “clean up” during machining the
pattern maker must allow sufficient additional metal (machining allowance) to be left on
such faces when the casting pattern is manufactured.

The pattern must also allow for the shrinkage of the metal, as the casting cools, by making
the pattern oversize by a controlled amount.
To do this the pattern maker uses a contraction rule to suit the metal being cast.
This rule is marked with normal dimensions but they are all oversize by the required
amount.

~Page 4 of 18~
~Page 5 of 18~
EXERCISE

1. Explain the mechanism of crystal formation and growth as a metal cools from the liquid
to solid state.
2. Explain with the aid of sketches:
(a) how dendritic growth occurs;
(b) the difference between the crystal structure of a metal and its irregularly shaped
grains.
3. Explain what is meant by the ‘latent heat of fusion’ and why the temperature of the
metal remains constant whilst fusion occurs.
4. Explain the essential difference between crystals and grains when referring to the
structure of a metal.
5.

~Page 6 of 18~
SOLIDIFICATION DEFECTS
Under practical industrial conditions impurities will inevitably be present as pure metals
solidify.
Also the dendrites will be growing in highly competitive conditions within the confines of
the mould and each branch of each dendrite will be ‘fighting’ for space to develop.
This overcrowding causes the dendritic branches or spikes to press against each other
causing deformation.
The more important solidification defects which can occur are:
o Shrinkage;
o Misorientation;
o Segregation and Inclusions;
o Gas porosity

1. SHRINKAGE
This can occur on two scales:
o Very fine shrinkage caused by over-rapid cooling of the cast metal leading to porosity
between the dendrite branches.
o Large-scale shrinkage cavities resulting from poor mould design which prevents the
feeding of molten metal to compensate for the normal volumetric shrinkage that occurs
whenever metal solidifies and cools.

Shrinkage cavities occurring between the dendritic branches and spikes are known as
interdendritic porosity and are caused by over-rapid cooling of the cast metal.
As the molten metal between the dendrite branches and spikes cools and solidifies, it
shrinks.
Under normal cooling conditions additional molten metal has time to flow into the cavities
so formed and no discontinuity occurs.
If, however, cooling is too rapid there is no sufficient time for the shrinkage cavities to fill
and porosity occurs along the dendrite branches.
These fine cavities should not be confused with those due to gas porosity.

When a casting cools and solidifies there is nearly always considerable volumetric
shrinkage.
(A notable exception is cast iron, which expands as it solidifies due to structural changes in
the crystal lattice and then contracts normally as cooling continues in the solid state.)
In a well-designed mould there are sufficient runners and risers to feed additional molten
metal back into the mould as shrinkage takes place.
If the feeding of metal is inadequate during solidification, drawing may occur.
This is where the solidifying and shrinking metal in the smaller sections of the casting
draws molten metal from adjacent, thicker sections of the casting instead of from the
runners and risers.
This leaves large cavities which weaken the casting.
It also leaves unsightly sunken surfaces which may not clean up during machining.

2. MISORIENTATION
Dendritic growth forms a highly competitive environment and that the outermost branches
and spikes of the dendrites are distorted as they interfere with each other at the grain
boundaries.

~Page 7 of 18~
This results in the strict crystalline geometry of the space lattice breaking down at the
grain boundary and producing an amorphous layer some two or three atoms thick, as
shown in Fig. 1 below.
This amorphous layer behaves like a highly viscous liquid and allows slight movement to
occur between the grains.
This accounts for the creep which occurs in metals stressed over long periods of time, and
why creep is greater at elevated temperatures when the viscosity of the amorphous layer is
reduced.
This also accounts for the reason why metals tend to fracture by transverse cracking of the
crystals at low temperatures when the amorphous layers are more viscous and resistant to
intercrystalline movement.
At high temperatures failure is more likely to occur by intercrystalline cracking than by
transverse cracking since the amorphous layer is less viscous, weaker and more likely to
allow intercrystalline movement.

3. SEGREGATION AND INCLUSIONS

Once solidification is complete, there will be no evidence of dendritic growth providing the
metal is absolutely pure and no shrinkage cavities or inclusions are present.
However, such conditions never occur commercially and it is necessary to examine the
effects of such inclusions.
Various types of segregation and inclusions may be present such as:
o Dissolved inclusions with minor segregations
o Dissolved inclusions with major segregations
o Undissolved inclusions

(a) DISSOLVED INCLUSIONS

These tend to remain in the molten metal as long as possible so that the inclusions finally
solidify in the spaces between the branches and spikes of dendrites along with the residual
host metal.
The presence of these dissolved inclusions often causes discoloration of the host metal
outlining the shape of the original dendritic formation when examined under the
microscope.

~Page 8 of 18~
Thus the dissolved inclusions are segregated from the host metal and are referred to as
minor segregations, as shown in Fig. 2 below.
The figure also shows interdendritic porosity.

(I) DISSOLVED INCLUSIONS WITH MINOR SEGREGATIONS

These are the most harmful since they lead to brittleness and cracking in the casting and
may result, in the case of an ingot, to cracking and crumbling during subsequent forging
and rolling.
This is referred to as ‘shortness’ in the metal.
Metal which is hot short crumbles when hot worked, whilst metal referred to as cold short
crumbles and cracks when cold worked.

(II) DISSOLVED INCLUSIONS WITH MAJOR SEGREGATIONS

These are likely to be found in large castings such as ingots where large columnar crystals
are present.
As these move inwards they push the residual molten metal and dissolved inclusions ahead
of them.
Thus all the inclusions tend to become concentrated in the central ‘pipe’ where final
solidification takes place.
The appearance of these segregated inclusions in a sectioned and etched ingot casting is
shown in Fig. 3 below.

~Page 9 of 18~
(b) UNDISSOLVED INCLUSIONS
These are substances such as sand particles washed from the sides of the mould by the
molten metal and also metal oxide particles (scale).
Generally such inclusions are less dense than the molten metal so they tend to float to the
surface of the runners and risers where they can be discarded.
However, they sometimes become trapped in the casting where they form discontinuities
from which fatigue cracks can originate.
Such inclusions also adversely affect the strength and machining properties of the casting.
A cutting tool hitting a hard inclusion will have its cutting edge blunted or chipped.

4. GAS POROSITY
Gases are frequently dissolved in the hot, molten metal but are expelled as the metal cools
and solidifies.
The sources of these gases may be from the furnace atmosphere, or from chemical
reactions which take place during the melt.
Most aluminium alloys and some copper alloys are susceptible to gassing.
These metals tend to absorb hydrogen gas from the furnace atmosphere or from the
moisture in the foundry atmosphere.
The hydrogen is driven off by the addition of suitable chemicals immediately prior to
pouring.
This is known as degassing.
Adequate ventilation is required during this operation as the gas driven off is often in the
form of hydrogen chloride which turns into hydrochloric acid on contact with atmospheric
moisture.

~Page 10 of 18~
When ferrous metals such as steel and cast iron are cast, carbon monoxide gas may be
present as the carbon in these metals tends to combine with the oxygen in the air.
Any gases generated will bubble out as the metal cools and become trapped between the
branches of the dendrites to form small random cavities.

The moisture in green-sand moulds may boil to steam which will be trapped at or just
below the surface of the casting to form larger blow-holes if adequate venting of the mould
is not provided.
When exposed during machining, blow-holes appear as spherical or oval cavities with
shiny surfaces.

~Page 11 of 18~
EXERCISE

1. Name any three (3) solidification defects and write brief notes on the cause and effect of
each one.
2. Explain the following terms associated with the solidification of a molten metal:
(a) interdendritic porosity
(b) drawing
(c) segregations
(d) undissolved inclusions
(e) gas porosity
3. Explain what is meant by the term ‘misorientation,
4. State how misorientation can affect the ‘creep’ resistance of a metal.

~Page 12 of 18~
MACRO- AND MICROSCOPICAL EXAMINATION
1. MACRO-EXAMINATION
Macro-examination implies the use of the unaided eye or the use of a low power
magnifying glass.
The sample component is sectioned and the surface is ground smooth.
Since grinding tends to ‘drag’ the surface slightly, it is usual to hand finish the surface using
a grade 0 or 00 abrasive paper.
To reveal the grain structure it is necessary to etch the specimen.
This is done by dripping a suitable etchant onto the surface of the specimen.
The etchant will eat away the grain boundaries so that the individual grains stand out in
relief.
The specimen is then washed and examined.
Details can frequently be seen more clearly whilst the surface is wet.
Table 1 below lists a number of suitable etchants for macro-examination, whilst Fig. 1
shows a typical example of the appearance of a component prepared for macro-
examination where a slag inclusion in the weld is just visible.

~Page 13 of 18~
2. MICROSCOPICAL EXAMINATION
Microscopical examination (microscopy) requires much more careful preparation of the
specimen.
Since this has to be mounted on a microscope slide, it is usual to cut a small specimen from
the component to be examined.
The specimen is then mounted in a low melting point alloy or a thermosetting plastic
matrix for ease of handling as shown in Fig. 2 below.
Initial grinding must proceed with the utmost care to avoid overheating the specimen and
altering the micro-structure.
Finishing is then carried out using progressively finer grades of abrasive paper.
The paper is placed flat on a piece of plate glass and the specimen is moved back and forth
so that the abrasive marks are a series of straight lines.
At each change to a finer grade of abrasive the specimen is worked so that the new abrasive
marks at right-angles to the previous ones.
Treatment continues until the previous marks are no longer visible.
Finally, the specimen is polished on a rotary metallurgical polishing machine.
A suspension of jeweller’s rogue or a suspension of diamantine is dripped onto the rotating
pad of ‘selvyt’ cloth.
Absolute cleanliness is essential to avoid scratching the polished surface and all traces of
polishing agent must be removed before etching.
Polishing continues until no marks can be seen on the unetched specimen under the
microscope.

~Page 14 of 18~
The etchant used and the method of application depends upon the metal being examined
and the particular characteristics of the grain structure to be exposed.
Table 2 below lists a few of the more widely etchants and their applications for
microscopical examination.
Since metallurgical specimens are opaque, metallurgical microscopes use reflected light.
The light source is adjustable for intensity and colour filters are provided so that the light
can be tinted to give the optimum contrast and clarity.
Usually a turret of objective lenses is provided so that the magnification can be quickly and
easily changed.
Fig. 3 below shows the appearance of a typical etched specimen as seen under a
microscope.
The slag inclusion and the grain structure of the metal, in this case, are more easily visible
than is the case under macro-examination.

~Page 15 of 18~
Table 2 Etchants for Microscopical Examination
S/N Material Composition Application
1 Steel Nital General purpose etching reagent for cast
2-5% irons and wrought steels.
Concentrated Reveals pearlite, martensite, troostite and
Nitric Acid ferrite boundaries.
95-98% Ethyl or
Methyl Alcohol
Picral Reveals details of pearlitic and spheroidised
4% Concentrated structures – does not attack ferrite
Picric Acid boundaries.
96% Ethyl or Excellent for most cast irons except alloy
Methyl Alcohol and ferritic cast irons.
2 Aluminium 0.5% General purpose cold-swabbing etchant.
and Concentrated
Aluminium Hydrofluoric Acid
Alloys 99.5% Water
Keller’s Reagent Immersion etching of wrought, heat-
1.0% treatable alloys such as ‘duralumin’.
Concentrated
Hydrofluoric Acid
1.5%
Concentrated
Hydrochloric Acid
2.5%
Concentrated
Nitric Acid
95% Water
3 Copper and 10 g Ammonium General purpose etchant which reveals the
Copper Persulphate grain boundaries of brasses, tin bronzes
Alloys 20% Ammonium and cupro-nickel alloys.
Hydroxide Must be freshly made before use.
80% Water
10 g Ferric Suitable for ⍺β brasses, bronzes, aluminium
Chloride bronzes and cupro-nickel alloys.
35% Concentrated Darkens the β-phase and improves the
Hydrochloric Acid contrast.
65% Water

~Page 16 of 18~
EXERCISE

1. Explain the difference between the microscopical and macro-examination of metals.

2. Explain how a metal specimen is prepared for:
(a) microscopical examination of metals.
(b) macro-examination of metals.
3.

~Page 17 of 18~
RESOURCES
1. Engineering Materials, Vol. 1, R. L. TIMINGS
2. Materials for Engineers and Technicians, RAYMOND A. HIGGINS
3. Engineering Materials, W. BOLTON
4. Introduction to Engineering Materials, VERNON JOHN
5. Material Science, K. G. ASWANI
6. Materials Science, R. S. KHURMI & R. S. SEDHA
7. Engineering Materials, R. K. RAJPUT
8. The Science and Engineering of Materials, DONALD R. ASKELAND
9. Materials for Engineering, JOHN W. MARTIN
10. Mechanical Engineer’s Pocketbook, ROGER L. TIMINGS
11. Engineering Materials (e-book), SHAYMAA MAHMOOD
12. www.bookboon.com
13. www.google.com
14. www.woodheadpublishing.com
15. www.elsevier.com
16. www.elsevierdirect.com
17. www.the5thwave.com
18. www.bookselsevier.com
19. www.gigapedia.com
20. www.schandgroup.com
21. www.newnespress.com
22. www.pearson.com
23. www.mcgraw-hill.com
24. www.phindia.com
25. www.ebooks.me-mechanicalengineering.com
26. www.ebooksgo.org
27. www.bookaid.org
28. www.wiley.com
29. www.colorado.edu
30. www.integra-india.com
31. www.sc.edu
32. www.substech.com
33. www.scientific.net
34. www.uddeholm.com
35. www.newagepublishers.com
36. www.academicpress.com
37. www.bh.com
38. www.dummies.com
39. www.tutorial.mathlamar.edu
40. www.agilant.com
41. www.youtube.com
42. www.ibiblic.org
43. www.bipm.fr
44. www.NIST.gov
45. www.harcourt-ap.com
46. www.charontec.com
47. www.sabre.org
48. www.item24.de/en/mechanicalengineering-glossary

~Page 18 of 18~