It was originally my intention to cover heat treatment here, but I quickly realized that the level of detail I wished to achieve made combining the prerequisite information with those processes nothing short of cumbersome. So, the ultimate purpose of this exploration of knowledge must once again wait for a future post. Until then, below is a brief summary of those topics.
Continued from
Part I here
Molecular Geometry
Steel, or more specifically Iron,
has the ability to take on various arrangements in its metallic lattice structure
(as a solid), which are called allotropes. When pure iron is alloyed with other
elements (namely carbon, amongst others), the amount of those elements changes
the material properties of the alloy, and as a result, the temperature at which
the allotropes exist. Those temperatures will be specific to each alloy (which
includes the percentage of alloyed elements).
There are seven atomic crystal
systems, although iron exists only in the Cubic system. Each of the seven
systems have multiple variations depending on the non-corner nuclei, which
define the geometry. This principle is important for the phases of steel. The
seven systems are, for only the point of distinction, summarized below.
-Triclinic- Each length of the
sides is exclusively different, and the angles between lengths are also
exclusively unequal.
-Monoclinic- A prism where the ends are two parallelograms,
where the angles of the non-parallelogram faces are 90 (perfect rectangles).
-Orthorhombic- A rectangular prism where the faces have
side-pairs of exclusively unequal lengths, resulting in three orthogonal pairs
of parallel sides. (None of the faces are squares, but are all rectangles). The
Cementite phase of steel exists in this form.
-Tetragonal- The two end faces are square and parallel,
resulting in the four side faces being equal rectangles. The martensitic phase
of steel forms a body centred tetragonal lattice.
-Hexagonal- A hexagonal prism with the ends being regular
hexagons. An additional nucleus rests in the centre of each hexagonal face.
-Trigonal- A system which is composed of a rhombus
superimposed with an extruded regular hexagon, called a rhombohedron.
Body Centred Tetragonal Orthorhombic
(Martensite) (Cementite)
Cubic Crystal System- This is
the important system for the context of steel. The unit cell is an isometric
lattice, meaning each side is of equal length and adjoining angle (perfect
cube). There are three variations of the cubic system, of which the later two
are found in steel.
Primitive- Only the eight corner
nuclei exist in the structure. It is the simplest of the cubic crystals.
Body Centred Cubic- (cB, or bcc)
A ninth nucleus rests in the centre of the eight corners. This is the predominant
stable form of iron at room temperature (more will be discussed later). Due to
the packing of the nuclei in this fashion, the cB unit cell has a relatively
small packing factor of .68 (a measure of how tightly spheres are packed in a
space, with 1 (impossible) being 100% of the space). In steel, different phases
may transition to body centred tetragonal (as martensite), which is similar to
cB excepting that one dimension is stretched to form a rectangle, while the
angles and remaining two dimensions are unchanged.
Body Centred Cubic Face
Centred Cubic
(Ferrite) (Austenite)
Face Centred Cubic- (cF, or fcc) In addition to the eight
corner nuclei, there is an additional one in each of the centre of each face
(but not centre of the entire cube), for a total of fourteen nuclei. Because
the atom volume is constant between cells, the atoms in a cF cell are more
tightly packed. This arrangement yields a packing factor of .74048 (which is
the theoretical maximum for any sphere packing orientation). Because the atoms
are so tightly packed, the cF cell is the most dense, which means it also occupies
a smaller volume than the same number of atoms in a cB lattice cell. Iron in
this form is called Austenite.
Phases of Steel
Along with molecular geometry,
steel exists in various phases depending on its temperature. These allotropes
are significant when it comes to heat treatment and the forging of steel.
Temperatures at which these transitions occur (assuming no change in pressure)
is dependent on the alloy of the particular grade of steel.
Delta Iron- (δ-Fe) When pure (elemental) liquid iron
solidifies, which is approximately 1,540 °C (2,800 °F), this is the first form
it takes. The molecular geometry is a cB lattice, and it remains in this form
until cooing an additional 144 °C.
Gamma Iron- (γ-Fe) After reaching 1,394 °C, the iron becomes
austenitic, meaning it transitions to a face centred cubic lattice, which is
more closely packed than the body centred lattice and thus reduces in volume
slightly (~8%). γ-Fe is capable of dissolving over 2% more carbon than other,
lower temperature phases of iron.
Alpha Iron- (α-Fe) Below 912 °C (1,674 °F), pure iron transitions back
to a cB lattice structure. If pure α-Fe exists in this form, it is also called
Ferrite. α-Fe saturation for carbon is only .021% by weight, which is
significantly lower than the previously mentioned γ-Fe. Depending on the
temperature, Alpha Iron can exist in two forms, depending on its magnetic
properties. Below the Curie Temperature (Tc), the temperature at
which the metal transitions between being paramagnetic and ferromagnetic, the
iron is considered only to be α-Fe. For pure iron, this temperature is 770 °C
(1,418 °F), and below which the Ferrite is ferromagnetic, meaning it will
respond to a magnetic field.
Beta Iron- (β-Fe ) Above Tc
but below the phase transition to Ferrite, the lattice (still body centred
cubic), becomes paramagnetic. Beta Iron is identical to Alpha Iron excepting
this difference. Due to the loss of magnetic response above the Curie
Temperature, it is often used as a reference for simple high carbon steels
during heat treatment. However, the difference (in pure iron, which decreases
in eutectic steels) is still 142 °C (256 °F), which is fairly significant given
the impossibility of hardening below the cB/cF transition.
Alloying Iron and the Crystal Structure
To alloy steel, additional
elements must be introduced to the lattice. Because the distance between
lattice points (centres spheres that represent the entirety of the atom which
occupies that space) is dependent on the atoms that form the lattice,
introduction of other atoms will change the crystal structure around that flaw.
Several types of point defects exist within the crystal lattice, where two general
types of involve the introduction of non-lattice elements. How the lattice is
effected by the defects will determine many of the material properties in
addition to what the particular atom is that causes that defect. Defects by
definition do not change the classification of the parent crystal lattice system.
It is also worth mentioning that a metallic alloy, rather than a specific
repeated molecule like SiO2 in quartz which has a tetragonal base
crystal lattice, there no base lattices that uniformly include all of the
elements found within it. However, in certain situations, such as in steel,
there can be centralized lattices
between some of the alloyed elements
(namely in the iron-carbon system, while the additional alloyed elements remain
substitutional or interstitial defects).
Vacancies- A type of lattice
defect that exists in all lattices, whether or not there is a non-base atom
introduced. This is a hole in the lattice where a lattice point would regularly
have an atom reside. Because these defects always exist, all ductile metallic
lattices can be work hardened, meaning the mechanical deformation of the
lattice which results in the pinning of these defects in positions which
require high amounts of energy to move.
Substitutional- Point defects
where a single lattice point of a uniform lattice is replaced with another
element. In the case of steel, this will not replace iron with carbon, instead
being additional alloyed elements such as vanadium or chromium. Depending on
the relative size of the substituted element, the lattice will undertake compressive
and tensile stress to accommodate the difference. The immediate neighbours of
the substitutional point defect will experience the largest stress, while those
farther out will experience little change.
Interstitial- Point defects
where additional lattice points are introduced between the normal points. In steel formed with ferrite, carbon
is an interstitial defect, and occupies a space approximately twice the
equivalent spherical volume of the interstitial holes present in the carbon,
which introduces a large amount of stress to the structure and why the maximum
dissolved carbon content is .021%.
Pinning of dislocations, whether
vacancies, substitutions, or interstitial defects, will cause the material to
become harder and more brittle, as the dislocations will require more energy to
shift to an unpinned position. As a result, all metals which are capable of
undertaking plastic deformation without failure are capable of work hardening.
This also means that work hardening of a hardenable alloy of steel is
distinctly different from the phase transformation which can harden it via
quenching. Additionally, it should be noted that defects in steel will undergo
diffusion, so alloy gradients will slowly equalize over time. At higher
temperatures, this occurs more rapidly, as the defect atom (interstitial or
substitutional) is given more energy to move out of the lattice position it
occupies. Similarly, pinned defects will migrate to equilibrium within the
material over time, again more rapidly at higher temperature, which may
re-locate them from their pinned, dislocated position.
Eutectoid and Eutectic principles
A eutectic system is one in which
two components at a given state (i.e. solid) can coexist at a certain
temperature with their combined liquid form.
For the purposes of explanation,
[A] and [B] will be two constituent materials which, when both melted, form a
homogeneous material [C]. If the ratio of [A] to [B] is too low, [A] will begin
to precipitate out of the liquid solution before [B], meaning [A] and [C] will
exist at the same time. However, if there is too much of material [B], then it
too will solidify first, leaving [B] and [C] to coexist. Now, if the ratio of
[A] and [B] is at the eutectic composition, then [A], [B], and [C] will all
exist at the same critical, eutectic temperature. At that ratio, the material
is said to be in a eutectic system. It should be noted that not all systems
have a eutectic ratio, most notable a mixture of silver and gold (Electrum).
When the phase of [C] is a solid
instead of a liquid, but at some temperature a solid phase transition still
occurs (of a crystal lattice), the material has a eutectoid temperature (as opposed to eutectic). This sort of system, which is illustrated above, exists
for the forms of an iron/carbon mixture. At the eutectoid temperature 727 °C
(1,341 °F) for the iron-carbon (Fe-C) system, Austenite, Ferrite, and Cementite
all coexist simultaneously. For the Fe-C system, the eutectoid composition
ratio is .76% Carbon by weight. Above the eutectoid temperature, Austenite (γ-Fe)
exists, while below, a mixture of Ferrite (α-Fe) and Cementite forms. Should
there be an excess of Ferrite, and thus iron, the steel will be a hypoeutectoid
(below the eutectoid ratio); and if there is an excess of Cementite, and thus
carbon, it will be a hypereutectoid.
Importantly, if a system has both
a eutectoid and eutectic temperature, which the Fe-C system does, those two
temperatures are not necessarily the same (and are quite different for steel).
A final note on eutectoids as
they pertain to steel is that the presence of additional alloying elements
changes the eutectoid ratio and temperature for that particular alloy. For
example, Nickel and Chromium both lower the eutectoid ratio for carbon in
steel.
Phases and Microstructures of Steel
Before understanding the heat
treatment process, it is important to examine the forms of steel as it exists
within the crystal structure. There is a distinct difference between phases and
microstructures, although the terminology common between them is often confused.
Phases are a product of the intermolecular kinetic energy distribution, whereas
microstructures are, as mentioned in the discussion of native and meteoric
iron, a relative arrangement of macro-lattices (as opposed to the crystal
system micro-lattices). Both the phase and microstructure can have dramatic
effects on the working properties of the steel, whether ductility, hardness,
brittleness, strength, reaction to strain, conductivity, magnetic reaction,
amongst many others. The following list is not meant to be all inclusive, but
rather comprised of those relevant to bladesmithing.
-Phases-
Ferrite- As aforementioned,
ferrite is the cB crystal structure of pure iron. Ferrite does not exist as an
alloy, although it can and often does exist in localized concentrations of iron
based metals. When combined with other phase
concentrations, ferrite can form certain microstructures.
Maximum carbon content of .021% by weight.
Austenite- Also mentioned above,
γ-Fe is the cF, paramagnetic allotrope of iron that exists above 1,394 °C. This
phase transition begins in pure iron at 912 °C where the first cB lattices
change to cF, which, due to the change in atomic density, is able to dissolve
approximately 2% more carbon. This is important when concerning heat treatment
of steel and will be later discussed.
Cementite- Iron Carbide (Fe3C),
which is an orthorhombic base crystal lattice form. When combined with Ferrite,
it commonly forms the Pearlite microstructure. In the orthorhombic system in
which the lengths of the sides are stretched from their cubic form (three side
sets are all inequal length but intersect at right angles), carbon is more
easily accommodated in the interstitial holes. As a result, the interstitial
stress is less significant and the lattice can hold more carbon before becoming
unstable. Cementite is, in its pure form, 6.67% carbon. As Austenite cools,
carbon diffuses into pockets that form cementite to take in what the Ferrite
cannot hold. Cementite is extremely hard and brittle (a ceramic if pure Fe3C).
Martensite- When Austenite is
cooled with sufficient speed (varies with the alloy of steel), instead of being
converted back to ferrite and cementite, the carbon is trapped in the lattice
and forms a body cantered tetragonal structure that is, in effect, a super
saturated form of ferrite. Due to the excess carbon held here instead of in
localized pockets as cementite, there is a comparatively high amount of
internal stress due to these interstitial carbon atoms being pinned in the
tetragonal lattice. This is where Martensite gains its hardness, as the pinned
shear dislocations do not permit the movement of defects across the lattice.
Martensite is approximately 1.75x harder than pearlite on the Brinell scale.
Different alloys of steel have different hardenability depending on the carbon
concentration, as that is the determining factor of how much of the lattice is
held in the martensitic form rather than lesser concentrations of carbon which
return instead to ferrite or cementite. Past the eutectoid point, larger
amounts of cementite will precipitate due to the larger stable ratio of carbon
to iron (Fe3C). Since
Martensite is in a state unattainable in thermodynamic equilibrium, the
introduction of thermal energy allows the Martensite lattice to convert back to
cementite and ferrite due to the increased diffusion rate of dislocations. Of a
lower density per lattice cell than Austenite, Martensitic conversion results
in a volumetric expansion when formed. This mostly accounts for the curvature
achieved through differential hardening of single edged blades and the warps
and bends of thin cross sectional areas.
-Microstructures-
Grains- All metals have grains.
When cooled from a liquid state, imperfections in the metal called nucleation
sites (non-alloying elements, foreign molecules, alloys of a higher melting
temperature, etc.) are where crystalline lattices begin to grow. Since there is
nothing that forces them to align in any particular fashion, the spatial
orientation is random. The more energy that is transferred from the cooling
metal to the growth of solid via metallic bonds, the larger the crystals grow
outwards or, if the conditions are right, the growth at new nucleation sites.
When the randomly oriented crystals meet, they form grain boundaries and stop
growing. Since the crystal faces are not aligned, the bonds there will be
weaker than inside the newly formed grains and of a higher stress. Because of
this, the fewer the total number of grains and thus larger in size, the more
ductile the material (as it will be more accepting of dislocation diffusion).
Due to the stress concentrations of grain boundaries, materials tend to fail
along those boundaries rather than through the grains themselves. Grain size
and growth can be determined during many formative and heat treating processes,
which will later be discussed in detail, whereas the shape of the grain boundaries is determined by other material properties (lamella, lenticular, spheroidic, acicular, etc.).
Pearlite- Perhaps the most
common microstructure of steel, a quasi-homogeneous combination of 88% Ferrite
and 22% Cementite by weight. As shown in the above eutectic phase diagram,
pearlite is formed by the cooling of Austenite below the critical temperature.
In a given alloy, pearlite can exist along with cementite and ferrite,
depending on the carbon concentration relative to the percent utilization by
the pearlite (~1.5906% Carbon by weight). Within the pearlite complex,
alternating layers of the carbon rich cementite and carbon deficient ferrite
form a lamellar type microstructure (discussed previously with native iron).
Bainite- Very similar to
pearlite in that it is composed of ferrite and cementite. Bainite is formed
from a more rapid cooling of austenite than when pearlite is formed, yet not so
quickly that it instead converts to Martensite. However, instead of the lamella
that pearlite grows, Bainite produces a radiating needle like patterns of
ferrite clustered at a central nucleation point, called acicular crystals. Between
the sheaves, cementite fills the voids, or in other phase transformations,
Martensite or retained Austenite. Due to this structure, it is not as ductile
as pearlite, although it is measurably harder. For this reason, and the fact
that the tremendous internal stress of Martensite is not present, there is no
need to temper or otherwise heat treat Bainite after its formation to retain
its hardness.
Widmanstätten Patterns- First
introduced during the discussion of native and meteoric iron, Widmanstätten
patterns are the product of iron-nickel crystal growth from the alternating
layers of Taenite and Kamacite. The lamella meet at 60 degree angles and, due
to the difference in diffusive properties between the iron and nickel, become
easily visible on the surface. Beyond native and meteoric iron, these are
relatively uncommon to metalworking.
Tempered Martensite- In the
formation of Martensite, the steel is subjected to a high internal stress. To
relive that stress yet retain some of the hardness of the martensitic crystals,
the material is tempered, allowing the decomposition of the body cantered
tetragonal structure into a more stable form (cementite, ferrite, or together
to become pearlite). While this process occurs, layers of Martensite form
beside the other two phases of steel or beside pearlite. This forms, along with
the intermediate microstructures, a final layered composition of lower internal
stress while retaining hardness proportional to the temperature at which the
martensitic steel was tempered (how much thermal energy was given to the steel
determines the threshold for the energy required for a dislocation to be
unpinned) and the time at which that temperature was held (unpinning of
dislocations is a measure of probability, meaning longer times at that energy
increases the chance for dislocations being freed at or below that energy also
increases). In this sense, the degradation of Martensite is the result of the
martensitic lattice, reaching thermodynamic equilibrium through the collapse of
its structure into new, lower stress crystals.