Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
_______________________________
Lecture Notes in Applied Metallurgy
for the course held at Politecnico di Milano for the Master
of Science (MS) Degree in Mechanical Engineering
This booklet is based on bibliography of the course. Several parts of this booklet, as well as pictures and tables, have been freely taken from the original text books in the selected bibliography
of the course. This booklet is provided by the teacher for internal use only to the students attending the course of Applied Metallurgy, course of the M.Sc. in Mech. Eng. at Politecnico di
Milano.
No other external uses and/or scopes that are out of preparation to final exam are therefore allowed.
Any other different uses by the students shall be under their own responsibility.
Table of Content
Chapter 1 - Resuming basic principles of metallurgy ........................................ 8
Introduction ........................................................................................................... 8
Atomic Structure and Interatomic Bonding ....................................................... 8
Atomic arrangements of metals ....................................................................... 12
Vacancies, interstitial spaces and grain boundaries as drive-force for
diffusion in metals ............................................................................................... 20
Elastic and plastic deformation in metal by micromechanical models ........... 31
Slip in polycrystalline crystal........................................................................... 41
Mechanisms of Strengthening in Metals ........................................................... 45
Strengthening by grain size reduction ............................................................. 45
Strain hardening .............................................................................................. 50
Phase boundaries as strengthening sources .................................................... 53
Grain Growth, Recovery and Recrystallization................................................ 57
Recovery .......................................................................................................... 58
Grain growth ................................................................................................... 59
Recrystallization .............................................................................................. 60
Mechanical behavior of metals by macroscopic approach .............................. 64
Tensile test ....................................................................................................... 64
Ductility ........................................................................................................... 67
Resilience ......................................................................................................... 67
Toughness ........................................................................................................ 68
Hardness .......................................................................................................... 70
Resuming the main contents, defining the main paradigm for mechanical
response of metals ................................................................................................ 72
Microstructural features of fracture in metallic materials .............................. 77
Ductile Fracture .............................................................................................. 78
Brittle Fracture ................................................................................................ 81
Chapter 2 - Damage mechanisms and root cause failure analysis basics ....... 85
The contributing factor: the stress concentration ........................................... 92
The energy criterion for fast fracture .............................................................. 94
Review of phase transformations and mechanical properties for ironcarbon alloys................................................................................................. 200
Industrial quenching problems: distortions and quenching cracks ........ 202
Quenching and tempering steels ................................................................. 205
Basics of criteria selection ............................................................................. 215
The manufacturing cycle of quenching and tempering steels .................. 215
Reduction of microstructure and chemical composition inhomogeneity ....... 216
Control of quenching distortion ..................................................................... 220
Criteria for selection ...................................................................................... 226
Spring Steels ................................................................................................. 227
Heat treatments for steels and manufacturing cycle of springs ..................... 229
Patenting ........................................................................................................ 231
Pickling .......................................................................................................... 232
Shot peening................................................................................................... 232
Criteria selection ........................................................................................... 235
Surface hardening of mechanical components .......................................... 235
Surface Hardening by Localized Heat Treatment .......................................... 236
Manufacturing cycle for surface hardened product....................................... 238
Criteria for steel selection for surface hardening .......................................... 239
Carburizing steels ........................................................................................ 240
Carburizing cycle description ........................................................................ 243
Design aspects of carburizing........................................................................ 244
Nitriding steels ............................................................................................. 246
Design features of nitriding parts .................................................................. 249
Manufacturing cycle for nitriding parts......................................................... 250
Chapter 4 Tool Steels ..................................................................................... 253
Classification of tool steels ............................................................................ 253
The AISI classification ................................................................................... 253
Important properties required for various applications ................................ 254
Secondary hardening of tool steels ................................................................ 255
Manufacturing cycle of tool steels ................................................................. 257
Criteria for selection of tool steels for specific applications ......................... 257
Steels for Special Applications Specialty Alloyed Steels ............................. 260
High-Fracture-Toughness Steels ................................................................ 260
Mar-Aging Steels ......................................................................................... 262
High-Strength Low-Alloy Steels ................................................................. 264
Dual-Phase Steels ......................................................................................... 267
TRIP Steels ................................................................................................... 269
Chapter 5 Stainless Steels .............................................................................. 270
Introduction.................................................................................................. 270
(a)
(b)
Figure 1.1 (a) A schematic representation of the Bohr atom; (b) covalent bonding requires that electrons be shared between atoms in such a way that each atom
has its outer sp orbital filled. In silicon, with a valence of four, four covalent
bonds must be formed.
states are occupied. Valence electrons are those ones that occupy the
outermost shell: these electrons are extremely important for establishing
the bonding between atoms. And bonding between atoms are necessary to
form atomic and molecular aggregates.
This implies many of the physical and chemical properties of solids are
based on these valence electrons.
The basics of atomic bonding are best illustrated by considering how two
isolated atoms interact as they are brought close together from an infinite
separation (Fig. 1.2). At large distances, attraction forces exerted between
positive nucleus of one atom and the negative electrons of the other atom
are negligible (mutual attraction): this depend on the fact the two atoms are
too far apart to have an influence on each other. At small separation distances, each atom can actually exerts forces on the other, but when distance decrease too much, namely nuclei get close, repulsive forces between positive nuclei surpass attraction force. These counteracting forces,
attractive FA and repulsive FR, and the magnitude of each depends on the
separation or interatomic distance r (refer to Figure 1.2). The entity of attractive force FA obviously shall depend on the particular type of bonding
that exists between the two atoms, as discussed in brief above. In Fig. 1.2b
the potential energy is shown (as integral of bonding force Fdr ).
Despite the above scheme deals with an ideal situation involving only two
atoms, a similar yet more complex condition exists for solid materials. The
magnitude of this bonding energy and the shape of the energyversus interatomic separation curve (Fig.1.2b) that can vary from material to material influence physical properties of materials. For example, large bonding
energies typically also have high melting temperatures; at room temperature, solid substances are formed for large bonding energies, whereas for
small energies the gaseous state is favored; liquids prevail when the energies are of intermediate magnitude. How much a material expands upon
heating or contracts upon cooling (that is, its linear coefficient of thermal
expansion) is related to the shape of its E vs r curve (a deep and narrow
trough, which typically occurs for materials having large bonding energies, normally correlates with a low coefficient of thermal expansion and
relatively small dimensional alterations for changes in temperature).
10
Figure 1.2 - (a) The dependence of repulsive, attractive, and net forces on interatomic separation for two isolated atoms. (b) The dependence of repulsive, attractive, and net potential energies on interatomic separation for two isolated atoms.
11
Figure 1.3 - The force-distance curve for two materials, showing the relationship
between atomic bonding and the modulus of elasticity, a steep dF/da slope gives a
high modulus.
12
(a)
(b)
13
The material, if it is solid, remains in this state till the equilibrium distance
that is increased by increasing temperature (consider that heating up provides energy for increase atom vibration) is lower than low-bonding energy regime (see again Fig.1.2b: as interatomic atom distance increases, attractive force, i.e. bonding energy decreases). If atoms reaches high interainteratomic distance, their mutual attractive force decreases and atoms
separate each other. While this phenomenon happens in metal, it is observe
a change in status of aggregation: it passes from solid to liquid state.
Now, proceed again in back track: cool down temperature of metal that is
provided in molten state. The vibration energy of atoms progressively decreases, till some atoms can attract each others. These early movements
make some atoms to link each others, and some nuclei of solidified materials appears in the melting phase (see Fig.1.5a)
What distinguishes metals, some polymers and many ceramics is that when
solid state is reached, atoms occupies regular ordering of atoms that extends through the material (Fig.1.5b and c). Particularly here we refer to
metals.
14
For a metal, crystalline solid is built up during cooling down and solidification by way of a periodic and repeated arrangement of atoms that extends throughout the entirety of the specimen. The basic cell that repeat in
the space at is called the unit cell and its geometry varies for metals. In
Fig. 1.6 are show various possibility of arrangement in metals and the
highlighted ones are those we deal with in this course.
Figure 1.6 - The fourteen types of Bravais lattices grouped in seven crystal systems. The highlighted structures are those of our interest.
15
The representation in Fig.1.6 is just schematic; in reality, the lattice structure is made up by atoms that get close each other, compacting and packing in the unit cell above illustrated. In Fig.1.7 is shown the more real
structure of a face centered cubic, or FCC, unit cell (Fig.1.7a) compared
with the schematic representation (Fig.1.7b) of unit cell, and how the unit
cells combine in the space to form solid material (namely, each single
grain of Fig.1.5).
Fig.1.7 - For the face centered cubic crystal structure, (a) a hard sphere unit cell
representation, (b) a reduced-sphere unit cell, and (c) an aggregate of many atoms.
Now, imagine to start from the scheme in Fig.1.7b made of ping pong balls
at the corner and in center of 6 faces of the cube, and pack as you can, trying to realize by outer uniform compression a compact cube. What happens is that some spaces will remain in your compacted cube. These spaces
we call interstitial spaces, or sites. In Fig.1.8, for example, are shown the
16
interstitial spaces present in three type of unit cells. In Fig.1.9 the octahedral site inside the face centered cubic cell.
Figure 1.8 - The location of the interstitial sites in cubic unit cells. Only representative sites are shown. The name of the site (e.g. octahedral, tetrahedral, etc.)
depends on its location in the
Figure 1.9 - The location of the octahedral site in a face centered cubic unit cell.
17
ment) nuclei that are forming inside liquid phase as illustrated in Fig.5, actually we have to consider they are building up meanwhile atoms by atoms
occupy fixed position inside the specific unit cell. Each metal, we know,
has its own specific atomic arrangement it is just like the DNA code for
organisms - that mainly depends on the base element, e.g. body cubic centered, BCC, for Fe (at ambient temperature), face cubic centered, FCC, for
aluminum, highly compact hexagonal for Mg, etc. Thus, when metal cools
down and rapidly solidifies, it is impossible all atoms perfectly arrange to
build up a perfect crystal lattice into perfect grain: in reality many atoms
will occupy wrong positions.
Point defects
We call these wrong positions imperfections of crystal lattice, or more
simply defects. Basic defects of a real crystal lattice consists in a punctual
(single) wrong position occupied by an atom that is missing, thus we call it
a vacancy, or it is inserted in an insufficient space, thus we can define a
self-interstitial atom (see Fig. 1.10). We call this type of imperfections
point defects, because of their effect onto lattice irregularity: they can be
several, but they are punctual imperfections distributed in crystal planes.
If you consider the most generalized case of non-pure metals, namely the
solid solutions of two or more elements, we call metal alloys, like Fe-C alloys, aluminum silicon added alloys, etc. exhibit same type of defects as
vacancies, but additionally to the self-interstitial atoms, crystalline defects
can refer also to irregular occupancy of an alloying element atom that
would have a smaller or wider radius than base element. In Fig.1.11 it is
18
Figure 1.11 - Point defects: (a) vacancy, (b) interstitial atom, (c) small substitutional atom, (d) large substitutional atom,
19
Figure 1.14 Schematic diagram showing small and high-angle grain boundaries
and the adjacent atom positions.
, results from an
20
Figure 1.16 Schematic diagram showing a twin plane or boundary and the
adjacent atom positions (colored circles).
21
Diffusion also occurs for pure metals, but all atoms exchanging positions
are of the same type; this is termed self-diffusion. Of course, self-diffusion
is not normally subject to observation by noting compositional changes.
This was the experiment, to allow us to observe results of diffusion: but
which are mechanisms that causes diffusion into metals?
(a)
(d)
(b)
(e)
(c)
(f)
Figure 1.16 Start experiment time: (a) a coppernickel diffusion couple before a
high-temperature heat treatment; (b) schematic representations of Cu (red circles)
and Ni (blue circles) atom locations within the diffusion couple; (c) concentrations
of copper and nickel as a function of position across the couple; (d) the copper
nickel diffusion couple after a high-temperature heat treatment, showing the alloyed diffusion zone; (e) schematic representations of Cu (red circles) and Ni
(blue circles) atom locations within the couple; (f) Concentrations of copper and
nickel as a function of position across the couple.
22
neighbor atoms and then cause some lattice distortion during the displacement. As stated, this energy is vibrational in nature until cooling down
metals to -273 K. At a specific temperature some small fraction of the total
number of atoms is capable of diffusive motion, by virtue of the magnitudes of their vibrational energies. This fraction increases with rising temperature.
Several different models for this atomic motion have been proposed; of
these possibilities, two dominate for metallic diffusion.
23
Figure 1.17 Schematic representations of (a) vacancy diffusion and (b) interstitial diffusion.
where A denotes the area across which diffusion is occurring and t is the
elapsed diffusion time. In differential form, this expression becomes:
(. 1.2)
The units for J are kilograms or atoms per meter squared per second
(kg/m2. s or atoms/m2. s). If the diffusion flux does not change with time, a
steady-state condition exists.
24
Figure 1.17 (a) Steady-state diffusion across a thin plate. (b) A linear concentration profile for the diffusion situation in (a).
In the present experiment, the concentration profile is assumed to be linear, as depicted in Figure 1.17b, and:
(. 1.4)
The mathematics of steady-state diffusion in a single (x) direction is relatively simple, in that the flux is proportional to the concentration gradient
through the expression:
25
(. 1.5)
(. 1.6)
= (
The above equation is known as Ficks second law. If the diffusion coefficient is independent of composition (which should be verified for each
particular diffusion situation), Equation 1.6 simplifies to:
(. 1.7)
2
=
26
0
2
27
o
o
o
o
28
= 0 exp
Where:
D0 is a temperature-independent pre-exponential (m2/s)
Qd is the activation energy for diffusion (J/mol)
R is the gas constant, 8.31 J/molK
T is the absolute temperature (K)
The activation energy may be thought of as that energy required to produce the diffusive motion of one mole of atoms. A large activation energy
results in a relatively small diffusion coefficient.
29
(. 1.10)
= 0
Q 1
2.3
Because D0, Qd, and R are all constants, Equation 1.10 takes on the form of
an equation of a straight line: y = b+ mx. where y and x are analogous, respectively, to the variables log D and 1/T. Thus, if log D is plotted versus the reciprocal of the absolute temperature, a straight line should result, having
slope and intercept of -Qd/2.3R and log D0, respectively. This is the manner
in which the values of Qd and D0 are determined experimentally. From such
a plot for several alloy systems (Figure 1.20), it may be noted that linear relationships exist for all cases shown.
Figure 1.20 - Plot of the logarithm of the diffusion coefficient versus the reciprocal of absolute temperature for several metals.
30
31
Figure 1.22 - The atom positions around an edge dislocation; extra half-plane of
atoms shown in perspective.
As the point defects above discussed disrupt the perfect arrangement of the
surrounding atoms, irregularities induced by dislocations works in similar
way: within the region around the dislocation line there is some localized
lattice distortion. The atoms above the dislocation line in Figure 1.13 are
squeezed together, and those below are pulled apart; this is reflected in the
slight curvature for the vertical planes of atoms as they bend around this
extra half-plane.
32
(a)
(b)
(c)
Figure 1.23 (a) Tensile external load applied to perfect crystal lattice; (b) tensile
stress leads to elastic deformation, and eventually to (c) a brittle rupture consisting
in simultaneous breaking of atom bonds.
You can read this statement as follows: when atoms are separated by a distance that
determine attractive force drastically reduces, see Fig. 1.2.
33
internal
Slip plane
(a)
(b)
Figure 1.24 (a) Shear stress applied to perfect crystal lattice; (b) on certain plane
shear stresses can develop and induce atoms in two facing rows to slip. The final
shape of the perimeter of the grid irreversibly changes, thus depicting plastic deformation has occurred.
By such a stress field, shear stresses develop onto the intermediate plane,
as shown in the scheme of Fig.1.14b. Let us for the next that shear stress is
higher enough (well better comment how much high it should be), it is
capable to make one crystal lattice plane to slip over the opposite one; the
result of this movement is clearly shown in Fig.1.24b. The atoms of the
upper part of the grid have moved of one step ahead, of the order of an interatomic space. If you imagine to remove the external load, the perimeter
34
of the grid cannot be restored at its original shape. The irreversible deformation has occurred, without implying brittle rupture of atomic bonds.
Thus, the plastic deformation in metal can be effectively explained by this
micromechanical scheme.
But, there is still a but. If we refer to perfect crystal lattice shown in
Fig.1.24, during the slip movement of one step order, 6 atomic bonds have
been broken and 5 have been restored. It appears the net balance of energy
concerns on 1 atomic bonds, but we need to consider the start of the
movement has required in any case external load is sufficient to develop
shear stresses capable to break out 6 atom bonds. You should take into account that were enormously reducing the scale of a grain to a 36 atoms!
But if we consider a grain in reality, the number of atoms from one border
to the opposite border is of the order of millions! This means that to realize
a very insignificant variation of perimeter of grain, shear stress along the
grain should be capable to break millions of atomic bonds, simultaneously,
thus recovering this huge quantity of energy as the atoms onto slip plane
advance one step (see again the scheme in Fg.1. 24b). This cannot be explainable in reality, since too much high would be the minimum shear
stress to activate the slip that no machines on Earth would be adequate to
plastically deform metals. From our knowledge of the metallic bond, it is
possible to derive a theoretical value for the stress required to produce slip
by the simultaneous movement of atoms along a plane in a metallic crystal.
However, the strength actually obtained experimentally on single crystals
is only about one-thousandth (1/1000) of the theoretical value, assuming
simultaneous slip by all atoms on the plane. Obviously, slip does not occur
by the simple simultaneous block movement of one layer of atoms sliding
over another. The modern concept is that slip occurs by the step-by-step
movement of dislocations through the crystal. But, how does it contribute
in plastic deformation of metals?
Dislocation and plastic deformation in metal
Fortunately in metals there are millions of dislocations distributed into
crystal lattice, as that one shown in the scheme of Fig.1.22. To comprehend how dislocation positively works for plastic deformation, let us consider the scheme in Fig.1.25.
When force is applied such that it shears the upper portion of the crystal to
the right the plane of atoms above the dislocation can easily establish
bonds with the lower plane of atoms to its right, with the result that the dislocation moves one lattice spacing at a time. Note that only single bonds
are being broken at any one time, rather than the whole row. The atomic
distribution is again similar to the initial configuration, and so, the slipping
35
of atom planes can be repeated. The movement is much like that of advancing a carpet along a floor by using a wrinkle that is easily propagated
down its length.
This stress required to cause plastic deformation is orders of magnitude
less when dislocations are present than in dislocation-free, perfect crystalline structures.
36
45
Figure 1.26 Permanent deformation occurring at grain boundaries in polycrystalline metals that reveal macroscopic deformation onto surface of specimen under
tensile loading.
37
Figure 1.27 (a) A {111} slip system shown within an FCC unit cell. (b) The
(111) plane from (a) and three slip directions (as indicated by arrows) within that
plane constitute possible slip systems.
Slip occurs along -type directions within the {111} planes, as indicated by
arrows in Figure 1.27. Hence, represents the slip plane and direction combination, or the slip system for FCC. Figure 1.27b demonstrates that a given slip plane may contain more than a single slip direction. Thus, several
slip systems may exist for a particular crystal structure; the number of independent slip systems represents the different possible combinations of
slip planes and directions. For example, for face-centered cubic, there are
12 slip systems: four unique {111} planes and, within each plane, three independent directions.
The possible slip systems for BCC and HCP crystal structures are listed in
Table 1.1. For each of these structures, slip is possible on more than one
family of planes (e.g., {110}, {211}, and {321} for BCC). For metals having these two crystal structures, some slip systems are often operable only
at elevated temperatures.
Table 1.1. - Slip Systems for Face-Centered Cubic, Body-Centered Cubic, and
Hexagonal Close-Packed Metals.
38
Face-centered cubic (fcc) metals have a large number of slip systems (12)
and are therefore capable of moderate-to-extensive plastic deformation.
Although body-centered cubic (bcc) systems often have up to 12 slip systems, some of them, like steel, exhibit a ductile-to-brittle transition as the
temperature is lowered due to the strong temperature sensitivity of their
yield strength, which causes them to fracture prior to reaching their full potential of plastic deformation. In general, the number of slip systems available for hexagonal close-packed (hcp) metals is less than that for either the
fcc or bcc metals, and their plastic deformation is much more restricted.
The hcp structure normally has only three to six slip systems, only onefourth to one-half the available slip systems in fcc.
Since plastic deformation takes place by slip, or sliding, on the closepacked planes, the greater the number of slip systems available, the greater
the capacity for plastic deformation.
Slip in single crystal: the Schmids law
A further explanation of slip is simplified by treating the process in single
crystals, then making the appropriate extension to polycrystalline materials. By mechanics in solid we know that, though an applied stress may be
pure tensile (or compressive), shear components exist. Existence of shear
stress is necessary to activate slip planes, so to move dislocation onto this
plane along its slip direction. However the shear stress acting along this direction shall be sufficiently higher to promote dislocation movement. We
need therefore to calculate the projected shear stress along the slip direction of the external tensile force F applied (see again Fig.1.28a), thus compare the value of this shear stress to the minimum shear stress necessary
for dislocation movement, or critical shear stress, c. The former shear
stress is called the resolved shear stress, r and its magnitude depends not
only on the applied stress, but also on the orientation of both the slip plane
and direction within that plane. To best understand last statement, look at
the scheme in Fig.1.28a. Assume the normal, n, of the slip plane lies at an
angle, , to the tensile axis. The area A of the slip plane considered for all
the specimen will be A0/cos . Similarly, if the slip plane lies at an angle,
, to the tensile axis, the component of the axial force, F, acting on the slip
direction will be Fcos . The resolved shear stress, r, is then given by:
(. 1.20)
cos
cos
= cos cos
39
stress we applied on specimen (see Fig.1.29), the shear stress r for the
Schmid law is 0: this means that also for high applied tensile stress, resolved shear stress does not develop and any dislocations cannot move. Ultimately, slip of planes cannot occur and plastic deformation cannot occur.
(a)
(b)
Figure 1.28 (a) Tensile test of single crystal and the Schmids Law components.
(b) If dislocation is present perpendicular onto slip plane, it can be activated, and
planes slip.
Figure 1.29 - If a slip plane is perpendicular to applied stress , the shear stress r
= 0. Also for high applied stress, no resolved shear stress develop and no dislocation can move. Slip cannot occur, deformation cannot occur.
In general, + 0 because it need not be the case that the tensile axis,
the slip plane normal, and the slip direction all lie in the same plane. A
metal single crystal has a number of different slip systems that are capable
of operating. The resolved shear stress normally differs for each one be-
40
cause the orientation of each relative to the stress axis ( and angles) also
differs. However, one slip system is generally oriented most favorably
that is, has the largest resolved shear stress, r (max):
(. 1.21)
= (cos cos )
(cos cos )
= 2
41
(a)
(b)
3 For randomly oriented grains, the average value of the Schmid factor is ~1/3, which is
referred to as the Taylor factor. It then follows that the yield strength should have a value of
approximately 3c.
42
+45 45
max
max
Fig.1.31 Scheme of polycrystalline plastic deformation initiation: in crystal lattice there are several grains with several slip systems. Because grains have random
orientation, their slip planes have, too. Some grains therefore could result most favorable oriented for activating dislocation movements if slip directions (represented by arrows in the grains) are favorably 45 angles oriented with the load axis.
The 45 orientations represents orientation of such planes where the shear stress
produced by external tensile load reaches its maximum value.
43
Figure 1.32 - Slip lines on the surface of a polycrystalline specimen of copper that
was polished and subsequently deformed.
44
Figure 1.34 - Schematic diagram showing how twinning results from an applied shear
stress . In (b), open circles represent atoms that did not change position; dashed and solid circles represent original and final atom positions, respectively.
Slip and twinning deformations are compared in Figure 8.13 for a single crystal
that is subjected to a shear stress .
Figure 1.35 - For a single crystal subjected to a shear stress , (a) deformation by slip, (b) deformation by twinning.
Mechanical twinning occurs in metals that have BCC and HCP crystal structures, at low temperatures, and at high rates of loading (shock loading), conditions under which the slip process is restricted; that is, there are few operable
slip systems.
45
46
= 0 + 2
Figure 1.36 - The motion of a dislocation as it encounters a grain boundary, illustrating how the boundary acts as a barrier to continued slip. Slip planes are discontinuous and change directions across the boundary.
Grain size may be regulated by the rate of solidification from the liquid phase, and also
by hot plastic deformation, as discussed in Chapter 9.
47
Figure 1.37 - The influence of grain size on the yield strength of a 70 Cu30 Zn
brass alloy. Note that the grain diameter increases from right to left and is not linear.
It should also be mentioned that grain size reduction improves not only the
strength, but also the toughness of many alloys; this fact can be explained
as follows. Considering the scheme in Fig.1.31, if you imagine to reduce
the grain size, many grains will appear in the same round window. This
fact implies several further slip systems will add, because they are pertinent to each single new grain you may consider. The result is the following. On one hand increasing of number of grain boundaries, that leads to
increasing strength as above discussed; on the other hand, the increasing
number of slip systems can contemporarily lead material to be prone dislocation movement inside the grain; this phenomenon results in higher energy absorption during dislocation movement inside the grain, toward the
grain boundaries where dislocation will pile up.
Small-angle grain boundaries are not effective in interfering with the slip
process because of the slight crystallographic misalignment across the
boundary. On the other hand, twin boundaries will effectively block slip
and increase the strength of the material. Boundaries between two different
phases are also impediments to movements of dislocations; this is important in the strengthening of more complex alloys.
48
Figure 1.38 - Variation with nickel content of (a) tensile strength, (b) yield
strength, and (c) ductility (%EL) for coppernickel alloys, showing strengthening.
Alloys are stronger than pure metals because impurity atoms that go into
solid solution typically impose lattice strains on the surrounding host atoms. Lattice strain field interactions between dislocations and these impurity atoms result, and, consequently, dislocation movement is restricted.
For example, an impurity atom that is smaller than a host atom for which it
substitutes exerts tensile strains on the surrounding crystal lattice, as illustrated in Figure 1.39a.
49
Figure 1.39 - (a) Representation of tensile lattice strains imposed on host atoms by
a smaller substitutional impurity atom. (b) Possible locations of smaller impurity
atoms relative to an edge dislocation such that there is partial cancellation of impuritydislocation lattice strains.
50
atoms are present because the overall lattice strain must increase if a dislocation is torn away from them. Furthermore, the same lattice strain interactions (Figures 1.39b and 1.40b) will exist between impurity atoms and dislocations that are in motion during plastic deformation. Thus, a greater
applied stress is necessary to first initiate and then continue plastic deformation for solid-solution alloys, as opposed to pure metals; this is evidenced by the enhancement of strength and hardness.
Strain hardening
Strain hardening is the phenomenon by which a ductile metal becomes
harder and stronger as it is plastically deformed. Sometimes it is also
called work hardening or, because the temperature at which deformation
takes place is cold relative to the absolute melting temperature of the
metal, cold working. Most metals strain harden at room temperature. Fig.
1.41a and 1.41b demonstrate how steel, brass, and copper increase in yield
and tensile strength with increasing cold work. The price for this enhancement of hardness and strength is in a decrease in the ductility of the
metal. This is shown in Fig.1.41c, in which the ductility, in percent elongation, experiences a reduction with increasing percent cold work for the
same three alloys.
Figure 1.41 - For 1040 steel, brass, and copper, (a) the increase in yield strength,
(b) the increase in tensile strength, and (c) the decrease in ductility (%EL) with
percent cold work
51
Figure 1.42 - The influence of cold work on the stressstrain behavior of a lowcarbon steel; curves are shown for 0% cold worked, 4% cold worked, and 24%
cold worked.
52
Dislocations of the same sign will repel each other, while dislocations of
opposite signs will attract each other and, if they meet, annihilate each other. If the two dislocations of opposite signs are not on the same slip plane,
53
they will merge to form a row of vacancies. These types of interactions occur because they reduce the internal energy of the system.
The result, in any case, is increasing resistance to plastic deformation with
increasing dislocation density. Work hardening thus results in a simultaneous increase in strength and a decrease in ductility. Since the work hardened condition increases the stored energy in the metal and is thermodynamically unstable, the deformed metal will try to return to a state of lower
energy. This generally cannot be accomplished at room temperature. Elevated temperatures, in the range of 1/2 to 3/4 of the absolute melting point,
are necessary to allow mechanisms, such as diffusion, to restore the lowerenergy state. The process of heating a work-hardened metal to restore its
original strength and ductility is called annealing. Metals undergoing forming operations often require intermediate anneals to restore enough ductility to continue the forming operation. Approximately 5% of the energy of
deformation is retained internally as dislocations when a metal is plastically deformed, while the rest is dissipated as heat.
The imposed stress necessary to deform a metal increases with increasing
cold work. In the mathematical expression relating true stress and strain
shown by equation 1.25:
(. 1.25)
the parameter n is called the strain-hardening exponent, which is a measure of the ability of a metal to strain harden; the larger its magnitude, the
greater is the strain hardening for a given amount of plastic strain.
54
There are three different types of crystalline interfaces that can develop between two phases (Fig. 1.46): coherent, semicoherent, and incoherent. A
fully coherent phase boundary (Fig. 1.46a, b) occurs when there is perfect
atomic matching and the two lattices are continuous across the interface.
The interfacial plane will have the same atomic configuration in both
planes. Since there is perfect matching at the interface, the interfacial energy is low. When the distances between atoms at the interface are not identical (Fig. 1.46c), coherency strains start to develop. However, since there
is still perfect atomic matching, it is still a coherent phase boundary; only
the interfacial energy will be higher than one with no distortion. When the
mismatch becomes sufficiently large, dislocations form to accommodate
the growing disregistry. The result is called a semicoherent interface (Fig.
1.46d) that has an medium-high interfacial energy. Finally, an incoherent
interface (Fig. 1.46e, f) is an interphase boundary that results when the matrix and precipitate have very different crystal structures, and little or no
atomic matching can occur across the interface. The interfacial energy is
even greater. An incoherent boundary is essentially equivalent to a highangle grain boundary.
55
56
For effective particle strengthening (Fig. 1.48), the matrix should be soft
and ductile, while the particles should be hard and discontinuous. A ductile
matrix is better in resisting catastrophic crack propagation. Smaller and
more numerous particles are more effective at interfering with dislocation
motion than larger and more widely spaced particles. Preferably, the particles should be spherical rather than needlelike to prevent stressconcentration effects. Finally, larger amounts of particles increase
strengthening.
57
58
Recovery
Recovery is the initial stage of the annealing cycle before recrystallization
occurs. During recovery, some of the stored internal strain energy is relieved by virtue of dislocation motion (in the absence of an externally applied stress), as a result of enhanced atomic diffusion at the elevated temperature. There is some reduction in the number of dislocations, and
dislocation configurations are produced having low strain energies.
During recovery, basic types of processes that occur are:
(1) the annihilation of excess point defects, particularly vacancies; the
vacancies that were generated during cold working are annealed
out by migrating to dislocations, grain boundaries, or surfaces;
(2) the rearrangement of dislocations into lower energy configurations, which also annihilates many of them; At slightly higher
temperatures, the rearrangement of dislocations occurs, and, in the
process, the annihilation of dislocations of opposite signs takes
place. The rearrangement of dislocations is assisted by thermal energy, which aids in both climb and slip mechanisms.
(3) the formation of subgrains that grow and interlock into subboundaries (Fig.1.49).
59
Grain growth
It should also be noted that the strengthening effects due to grain size reduction and strain hardening can be eliminated or at least reduced by an elevated-temperature heat treatment. Regarding with the enlarging of grain
size, heating up metals produces diffusivity to increase. The higher is the
diffusivity (refer to eq. 1.9), the higher is the diffusion of species in matrix.
This is true also for the self-diffusion of element. When atoms move from
one position to another obeying to diffusion mechanisms (refer to par.
Vacancies, interstitial spaces and grain boundaries as drive-force for diffusion in metals), they also can move from one grain boundary to another: what could happen is schematically shown in Fig.1.50a. Basing on the
grain boundary migration, in the Fig. 1.50b is shown the mechanisms
that provokes grain growth by small grains disappearing. Ultimately, grain
growth occurs because of metal with its original microstructure is heated
up to certain temperature (Fig.1.51).
(a)
(b)
Figure 1.50 (a) Schematic representation of grain growth via atomic diffusion;
(b) disappearing of small grains.
60
Empirically, it has been shown that grain growth occurs according to:
(. 1.26) =
Where:
D is the average grain diameter,
t is time,
n is a constant,
= 0 2
The constant n increases with temperature and approaches a theoretical
value of 0.5. It should also be noted that the activation energy, Q, also varies with temperature.
Recrystallization
Even after recovery is complete, the grains are still in a relatively high
strain energy state. Recrystallization is the formation of a new set of strainfree and equiaxed grains (i.e., having approximately equal dimensions in
all directions) that have low dislocation densities and are characteristic of
the precold-worked condition. The driving force to produce this new
grain structure is the difference in internal energy between the strained and
unstrained material. The new grains form as very small nuclei and grow
until they completely consume the parent material, processes that involve
short range diffusion. Two stages in the recrystallization process are represented in Fig.1.52a, 1.52b and 1.52c. In these photomicrographs, the small,
speckled grains are those that have recrystallized.
61
62
63
64
65
above as one of strengthening mechanism active in metals. To a good engineering approximation, the volume remains constant during plastic deformation (Al=A0l0), and as the specimen elongates, it decreases uniformly
in cross-sectional area along its gage length. Initially,
strain hardening more than compensates for this decrease in area, and the
engineering stress continues to rise with increasing strain. However, eventually, a point is reached where the decrease in area is greater than the increase in deformation load from strain hardening. This condition will be
reached first at some point in the specimen that is slightly weaker than the
rest. All further plastic deformation is then concentrated in this region, and
the specimen begins to neck or thin locally. Because the cross-sectional area is now decreasing far more rapidly than the deformation load is being
increased by strain hardening, the engineering stress continues to decrease
until fracture occurs.
66
With most metals, there is a gradual transition from elastic to plastic behavior, and the point at which plastic deformation actually begins is difficult to define with precision. The transition from elastic to plastic deformation is illustrated in Fig. 1.57.
(a)
(b)
Figure 1.58 (a) Transition from elastic to plastic behavior; (b) Discontinuous
yielding in plain carbon steels.
67
Ductility
Measures of ductility that are obtained from the tension test are the engineering strain at fracture (f) and the reduction of area at fracture. Both are
usually expressed as percentages, with the engineering strain at failure often reported as the percent elongation. Both of these properties are obtained after fracture by putting the specimen back together and taking
measurements of the final length, lf, and final specimen cross section at
fracture, Af. Percent elongation can be determined by:
where l0 is the original gage length, and lf is the final length of the gage
section. Likewise, reduction in area can be determined by:
where A0 is the original area of the gage section, and Af is the final area of
the gage section at fracture.
Resilience
Resilience is the capacity of a material to absorb energy when it is deformed elastically and then, upon unloading, to have this energy recovered.
The associated property is the modulus of resilience, Ur , which is the
strain energy per unit volume required to stress a material from an unloaded state up to the point of yielding. The modulus of resilience for a specimen subjected to a uniaxial tension test is just the area under the engineering stressstrain curve taken to yielding:
(. 1.27)
=
0
68
(. 1.28)
1
= = [ ]
2
1 2
=
2
Figure 1.59 Schematic representation showing how modulus of resilience (corresponding to the shaded area) is determined from the tensile stressstrain behavior
of a material.
Toughness
69
Figure 1.61 Impact test machine (Charpy test) conducted onto notched specimen.
70
Geometrical features, such as holes, fillets, and radii, produce higher local
stresses than encountered in the body of the material. For example, the tensile
stresses at the top and bottom of the hole shown in Fig. 1.62 are three times
greater than they are in the body of the material. These higher stresses are a result of the inability of the stresses to pass through the hole.
Hardness
Hardness is the resistance to penetration, and the majority of hardness testers force a small sphere, pyramid, or cone into a specimen by means of an
applied load. A number is obtained, and the hardness can often be correlated to the tensile strength of the metal. This proportional relationship depends on the fact both tensile strength and hardness are indicators of a
metals resistance to plastic deformation, that occurs on same micromechanical mechanisms of dislocation movement. Consequently, they are
roughly proportional, as shown in Figure 1.63 for tensile strength as a
71
function of the HB for cast iron, steel, and brass. As a rule of thumb, for
most steels, the HB and the tensile strength are related according to:
UTS (MPa) = 3.34 x HB (or HV for hardness higher than 500 HB)
In any case, the same proportionality relationship does not hold for all
metals, as Figure 1.63 indicates. This depend, in fact, by micromechanisms of dislocation movement in different matrix of different metals.
Figure 1.63 Relationships among hardness and tensile strength for steel, brass,
and cast iron.
72
Resuming the main contents, defining the main paradigm for mechanical response of metals
On an atomic scale level, plastic deformation corresponds to the motion of
dislocations in response to an externally applied shear stress. For example,
an edge dislocation moves by the successive and repeated breaking of
atomic bonds and shifting by interatomic distances of half planes of atoms.
For dislocation edge, in particular, dislocation line motion and direction of
the applied shear stress are thus parallel. We call slip the motion of dislocations in response to an externally applied shear stress. It occurs on specific crystallographic planes and within these planes only in certain directions, depending on crystal structure of the metal. The combination of slip
planes and slip direction, expressed by the Schmids Law, is defined slip
system. The slip plane is that plane that has the densest atomic packing,
and the slip direction is the direction within this plane that is most closely
packed with atoms. Operable slip systems, namely the planes where dislocation are active, depend on the crystal structure of the material.
73
A resolved shear stress is the shear stress resulting from an applied tensile
stress that is resolved onto a plane that is neither parallel nor perpendicular
to the stress
direction. For polycrystalline materials, slip occurs within each grain along
those slip systems that are most favorably oriented with the applied stress.
As favorable oriented we mean that in such grains the resolved shear stress
is highest and it can achieve the critical shear stress, namely the shear
stress necessary to activate dislocation slip along slip plane and in the slip
direction. During deformation, grains change shape and extend in those directions wherein there is gross plastic deformation.
The basic concept is that the ease with which a material is capable of plastic deformation is a function of dislocation mobilitythat is, restricting
dislocation motion leads to increases hardness and strength.
Various modes are practicable to inhibit dislocation mobility. We call
these modes strengthening mechanisms:
- Grain Size Reduction: as grain boundaries are barriers to dislocation motion for two reasons: a) when crossing a grain boundary, a
dislocations direction of motion must change; b) there is a discontinuity of slip planes within the vicinity of a grain boundary. A
metal that has small grains will be stronger than one with large
grains because the former has more grain boundary area, and, thus,
more barriers to dislocation motion. For most metals, yield
strength depends on average grain diameter according to the Hall
Petch equation.
- Solid-Solution Strengthening: the strength and hardness of a
metal increase with increase of concentration of impurity atoms
that go into solid solution (both substitutional and interstitial).Solid-solution strengthening results from lattice strain interactions between impurity atoms and dislocations; these interactions
produce a diminishment in dislocation mobility.
- Strain Hardening: strain hardening is just the enhancement in
strength (and decrease of ductility)of a metal as it is plastically deformed. Degree of plastic deformation may be expressed as percent cold work, which depends on original and deformed crosssectional areas. Yield strength, tensile strength, and hardness of a
metal increase with increasing percent cold work; ductility diminishes. During plastic deformation dislocation density increases, the
average distance between adjacent dislocations decreases, and
because dislocationdislocation strain field interactions, are, on
average, repulsivedislocation mobility becomes more restricted;
thus, the metal becomes harder and stronger.
74
75
Fig.1.65 Basic mechanical properties from tensile test, hardness test and impact
test grouped into: a) resistance capability (left side); b) ductility capability ( right
side).
As we already learned, the left side and right side properties for metal are
strictly correlated, because of the following causal relationship:
i) Dislocations move and reach grain boundary
grain boundary
perimeter modifies macroscopic plastic deformation occurs(1)
Note 1: For example macroscopic deformation is observed in surpassing
the onset yield strength, necking occurs in tensile tests; indenting of test
material by penetrator in the hardness test; highly plastically deformed
regions in the notched section area of Charpy test specimen broken.
76
High stiffness
Low stiffness
UTS
High strength
UTS
UTS
Low strength
Low toughness
r
Low ductility
High toughness
high ductility
Fig. 1.65 - Understanding basic material properties and their mutual correlations.
For example, a high stiffness material (scheme on the left side) is a material with
high slope measured in the origin - of the (stress) and (deformation) curve
obtained by a tensile test on the material. This slope is called the elastic modulus
of the material. Such a high stiffness material usually exhibits further high mechanical strength (i.e. Ultimate Tensile Strength, UTS), low ductility (i.e. deformation at break, r) and low toughness (i.e. energy stored to break, proportional to
area below the tensile test diagram).
77
On the opposite, a brittle fracture occurs with little or no gross plastic deformation and usually occur suddenly, without warning (thus in unsafe
mode - remind the ceramic vehicle axle!).
The tendency for brittle fracture increases as well as any external or internal conditions work for: a) locking dislocations and/or b) reduce slip
planes. Thus, the tendency for a material that has sufficient slip planes
(this is not the case of the Hexagonal Close Packing crystal structure) to
brittle behavior increases as well as its own dislocation population is inhibited in motion. This can happen, for example:
78
79
Fig. 1.67 - When a ductile material is pulled in a tensile test, necking begins and
voids formstarting near the center of the barby nucleation at grain boundaries
or inclusions. As deformation continues, a 45 shear lip may form, producing a final cup and cone fracture.
80
Fig. 1.68 Dimples form during ductile fracture. Equiaxed dimples form in the
center, where microvoids grow. Elongated dimples, pointing toward the origin of
failure, form on the shear slip.
Simple macroscopic observation of this fracture may be sufficient to identify the ductile fracture mode. Examination of the fracture surface at a high
magnificationperhaps using a scanning electron microscopereveals a
dimpled surface (Figure 1.69). The dimples are traces of the microvoids
produced during fracture. Normally, these microvoids are round, or equiaxed, when a normal tensile stress produces the failure (see Figure 1.69a);
however, on the shear lip, the dimples are oval-shaped, or elongated, with
the ovals pointing toward the origin of the fracture (see Figure 1.69b).
81
Brittle Fracture
Brittle fracture occurs in high-strength metals and alloys or metals and alloys with poor ductility and toughness. Furthermore, even metals that are
normally ductile may fail in a brittle manner at low temperatures, in thick
sections, at high strain rates (such as impact), or when flaws play an important role. Brittle fractures are frequently observed when impact, rather
than overload, causes failure. In brittle fracture, little or no plastic deformation is required. Initiation of the crack normally occurs at small flaws,
which cause a concentration of stress. The crack may move at a rate approaching the velocity of sound in the metal. Normally, the crack propagates most easily along specific crystallographic planes by cleavage.
Cleavage fractures are characterized by a planar crack that changes planes
by the formation of discrete steps. River patterns are formed at grain
boundaries (Fig. 1.70) where the cleavage plane in one grain is not parallel
to the plane in the adjacent grain, the difference being accommodated by a
series of steps. The river patterns eventually diminish as the crack propagates and adopts the cleavage plane of the new grain before being reformed at the next grain boundary.
Brittle fracture can be identified by observing the features on the failed
surface.
82
83
Another common fracture feature is the Chevron pattern (Figure 1.72), that
appears at naked eye. They are produced by separate crack fronts propagating at different levels in the material. A radiating pattern of surface
markings, or ridges, fans away from the origin of the crack (Figure 1.73).
Since the Chevron pattern is visible with the naked eye or a magnifying
glass and helps us identify both the brittle nature of the failure process as
well as the origin of the failure. Actually we can say that the Chevron pattern are the naked eye result of the microscopic cleavage fractures, thus
observed at grain scale.
Fig. 1.72 The Chevron pattern in quenched 4340 steel. The steel failed in a brittle manner by an impact blow.
84
Fig. 1.73 The Chevron pattern forms as the crack propagates from the origin at
different levels. The pattern points back to the origin.
In some cases, however, the crack may take an intergranular (along the
grain boundaries) path, when grain boundaries are abnormally weak
(Fig.1.74).
Transgranular fracture
(NORMAL )
Intergranular fracture
(ABNORMAL )
Fig. 1.74 The patter of: transgranular fracture (left side) and intergranular fracture (right side).
85
86
the next: an applied tensile stress is amplified at the tip of a small incision
or notch. Instead this little home-made experience, take a look in the photograph in Fig. 2.1a that shows an oil tanker that fractured in a brittle manner as a result of the propagation of a crack completely around its girth.
(a)
(b)
Fig.2.1 Brittle fracture of: a) an US oil tanker ship during WWII; b) the shipwreck of M V Kurdistan oil tanker ship in 1979.
87
This crack started as some type of small notch or sharp flaw. The catastrophic nature of brittle fracture was dramatically exemplified with frequent brittle failures of Liberty ships during World War II that broke in
half while just sitting at dock. To produce ships quickly for the war effort,
all-welded construction was used for the hulls instead of the traditional
riveted design. Of the 2700 Liberty ships built, approximately 400 sustained fractures, 90 of which were considered serious. In 20 of these, the
hull fractured in half. When the hatch covers were retrofitted with rounded
reinforcements and riveted structures replaced some of the welded structures, the incidence of fracture was greatly reduced. Even if a crack initiated at a defective weld, it would be arrested at a rivet hole before it reached
catastrophic dimensions. After the war, G.R. Irwin and his staff at the Naval Research Laboratory laid the foundation for what is known today as
fracture mechanics.
Liberty ships are just one example of catastrophic brittle fracture. It has also been a recurring problem in aircraft, bridges, train wheels, and other
heavy equipment. Although brittle fracture does not occur today with the
frequency it once did, it can still be a problem if proper design and manufacturing practices are not used. Two important failings of the requirements for ships as First Year Ice Class vessels brought to shipwreck of the
M V Kurdistan oil tanker ship in 1979 South of the Cabot Strait, off Nova
Scotia (Fig.2.1b).
Presence of defect in bilge keel welds combined with high thermal stresses
was the cause of brittle fracture. Studying relationships between material
properties, stress level, the presence of crack-producing flaws, and crack
propagation mechanisms had led in past Century to development of the
discipline named as fracture mechanics. By advancements in fracture mechanics, design engineers are now better equipped to anticipate, and thus
prevent, structural failures. Here in the following some of the fundamental
principles of the mechanics of fracture are addressed.
88
Box Griffith experience and the theoretical critical stress for brittle
materials
In a first attempt to calculate the strength of such a crystalline material,
one atomic plane perpendicular to the tensile load is considered. The bonding force f between
two atoms in two neighboring planes depends on their distance r, as it is
shown in Fig.2.2 (refer to the upper curve that decreases progressively and
asymptotically to zero value for large r distance).
() = sin
; = 0
(eq.2.2) = 0 +
2
89
(2.3)
; = 0
a brittle material exhibits until fracture a linear elastic relationship; thus, during crack formation it is valid also the Hook Law,
= E , or = E(x/r0), if referred to diagram in Fig.2.2.;
b) in the diagram f(x) in Fig.2.2, the variation of distance x to
achieve to rupture (i.e. maximum max value) are so small that the
2
relationship () = sin can be approximated by the
() =
gles: sin ;
90
(2.6) = 0
sin
[1 ]
Actually to form a crack inside material, two surfaces are created: a factor
of 2 is therefore necessary to exploit the result of eq.2.6 to account the total energy per surface unit required for creating two opposite faces formation by atoms debonding:
(2.7) U = 2 =
[1 ] , or =
[1 ]
(2.8)
= =
That represents the Griffith theoretical critical stress value for the fracture
of brittle material (i.e. atoms debonding).
What it was observed experimentally, however, was a great discrepancy
between the theoretical value above calculated and the experimental measured values. For example, The surface energy does not differ much for
various brittle solid materials and approximately equals 1 Jm2 (diamond
is an exception with = 5 Jm2.) The equilibrium distance a0 between atoms, is also almost the same for solids (about 1010 m). The table below
lists values of theoretical strength th and experimental fracture stress b
for some materials. It is clear that there is a large discrepancy between the
two values : the theoretical strength is much too high.
91
The reason for this deviation has been discovered by Griffith in 1921. In
1921 Griffith determined experimentally the fracture stress b of glass fibers as a function of their diameter. For d > 20 m the bulk strength of 170
MPa was found. However, b approached the theoretical strength of 14,000
MPa in the limit of zero thickness (namely zero defect, considering such a
low thickness material do ot contain defects). Griffith was aware in 1913
of work of Inglis 5 who calculated stress concentrations at circular holes in
plates, being much higher than the nominal stress. He concluded that in his
glass fibers such stress concentrations probably occurred around defects
and caused the discrepancy between theoretical and experimental fracture
stress. He reasoned that for glass fibers with smaller diameters, there was
less volume and less chance for a defect to exist in the specimen. In the
limit of zero volume there would be no defect and the theoretical strength
would be found experimentally. However, the Inglis solution poses a
mathematical difficulty: in the limit of a perfectly sharp crack, the stresses
approach infinity at the crack tip. This is obviously nonphysical (actually
the material generally undergoes some local yielding to blunt the cracktip),
and using such a result would predict that materials would have near zero
strength: even for very small applied loads, the stresses near crack tips
would become infinite, and the bonds there would rupture. Rather than focusing on the crack-tip stresses directly, Griffith employed an energybalance approach that has become one of the most famous developments
in materials science. Griffith published his work in 1921 and his paper can
be seen as the birth of Fracture Mechanics. The ingenious insight that
strength was highly influenced by defects has lead to the shift of attention
to the behavior of cracks and the formulation of crack growth criteria.
Ref. Inglis, C.E. Stresses in a plate due to the presence of cracks and sharp
corners. Transactions of the Institute of Naval Architects, Vol 55, 1913, pp 219241
5
92
Fig.2.2 - (a) The geometry of surface and internal cracks. (b) Schematic stress profile along the line XX in (a), demonstrating stress amplification at crack tip positions.
93
(eq.2.9) = 0 1 + 2
(eq.2.10) 2 0
This will yield a value of m that is many times the value of 0. If we want
to represent how much higher is the actual m than nominal stress 0, we
can calculate the ratio between the former and latter values accordingly
with:
(eq.2.11) =
1/2
= 2
94
95
Since the ends of plate cannot move, the forces acting on them can do no
work, and W = 0.
Accordingly, our energy formula in eq.2.12 gives, for the onset of fast
fracture (i.e. sudden crack grow in plate):
(eq.2.13)
Now, as the crack grows into the plate, it allows the material of the plate
near the crack to relax and loses elastic energy. We can estimate in
the way shown in Fig. 2.4.
Fig.2.4 - The release of stored strain energy as a crack grows. The shaded area is
the relaxed area once crack advances from a to a+a.
96
2 2
= 2
= 2 2
97
(eq.2.9bis) = 0 1 + 2
The eq.2.9bis states that the stress at crack tip increase theoretically at infinite value as the crack tip radius t (see scheme in Fig.2.2) approximates
zero value, as actually happens in case of sharp crack.
Furthermore if we are interested to the mathematical relation of the local
stress versus the distance r ahead of a sharp crack in an elastic material
(see the scheme of Fig.2.5), we find that the closer one approaches to the
tip of the crack, the higher the local stress becomes accordingly with the
eq.2.19:
6
Strictly speaking, this result is valid only for a crack through the center of a wide plate of
material. In practice, the problems we encounter seldom satisfy this geometry, and a numerical correction to required to get the strain energy calculation right. In general we
write: = where Y is the numerical correction factor (Values of Y can be found
from tables in standard reference books).
98
(eq.2.19) = 1 +
Notice that the mathematical expression here above representing the local
stress versus distance from crack tip is similar to the mathematical expression of the stress at crack tip as function of the crack tip radius and crack
length a (i.e. eq.2.9bis); on the other and they shall not be confused, because the former may have non-infinite solutions as the crack tip radius
approaches to a non zero value 7, while the latter we read in eq.2.19 has
always infinite values as we want to calculate the stress at crack tip. This
fact means that the expression cannot represent what in reality happens in
regions close to crack tip, namely r equal to zero. This discrepancy with
reality is well solved whether we superimpose to stress curve in eq.2.19
the plastic behavior of ductile material. This implies that, as we consider a
ductile material, until at some distance from the crack tip we call ry the
theoretical stress calculated by the eq.2.19 cannot exceed the yield stress
y of the material: in fact in this area plastic flow occurs and stress here are
fully relaxed, as it is shown in Fig.2.5a.
(a)
(b)
Fig.2.5 - Crack propagation by ductile tearing: a) forming of plastic zone; b) advancement of crack for ductile tearing occurring in plastic zone.
Notice we already verified that for crack tip radius equal to a in the scheme of
Fig.2.2, the value is 3, namely the value typical for a circular hole in a large plate.
7
99
= 1 +
2
2 2
= 2
The crack propagates by fast fracture mode when K is equal to Kc; the
width of the plastic zone, ry, is then given by eq.2.22 with K replaced by
Kc.
And now, note that the zone of plasticity shrinks rapidly as y increases:
cracks in soft metals have a large plastic zone; cracks in hard ceramics
have a small zone, or none at all. Coming back to ductile metals, most
metals contain tiny inclusions (or particles). Within the plastic zone, plastic flow takes place around these inclusions, leading to elongated cavities
(fracture by microdimple, see the scheme in Fig. 2.5b). As plastic flow
progresses, these cavities link up, and the crack advances by means of this
ductile tearing. The plastic flow at the crack tip naturally modifies original
sharp crack (see Fig.2.5a) into a blunt crack (Fig.2.5b): the local stress at
crack tip therefore decrease to values that are just sufficient to keep on
plastically deforming the work-hardened material there, as the scheme in
Fig.2.5b shows.
Moreover, the important thing about crack growth by ductile tearing is that
it consumes a lot of energy by plastic flow: the bigger the plastic zone, the
more energy is absorbed. High energy absorption means that G, is high,
and so is K,. This is why ductile metals are so tough.
100
2
2 2
The local stress at the crack tip is still in excess of the ideal strength and is
thus large enough to literally break apart the interatomic bonds there; the
crack then spreads between a pair of atomic planes giving rise to an atomically flat surface by cleavage, as shown by the scheme in Fig.2.6. The energy required simply to break the interatomic bonds at crack sharp tip is
much less than that absorbed by ductile tearing in a tough material, and
this is why materials like ceramics and glasses are so brittle. It is also why
some steels become brittle and fail like glass, at low temperatures. At low
temperatures in fact metals having b.c.c. and h.c.p. structures become brittle and fail by cleavage, even though they may be tough at or above room
temperature. Only those metals with an f.c.c. structure (like copper, lead,
aluminium, austenitic stainless steel) remain unaffected by temperature in
this way. In metals not having an f.c.c. structure, the motion of dislocations
is assisted by the thermal agitation of the atoms. At lower temperatures
this thermal agitation is less, and the dislocations cannot move as easily as
they can at room temperature in response to a stress. The result is that the
yield strength rises, and the plastic zone at the crack tip shrinks until it becomes so small that the fracture mechanism changes from ductile tearing
to cleavage. This effect is called the ductile-to-brittle transition DBTT; for
steels it can be as high as 0C, depending mainly on the carbon content of
the steel); steel structures like ships, bridges and oil rigs are much more
likely to fail in winter than in summer.
101
(a)
(b)
Fig.2.6 a) Temperature dependence of the Charpy V-notch impact energy (curve
A) and percent shear fracture (curve B) for an A283 steel; b) Photograph of fracture surfaces of Charpy V-notch specimens tested at indicated temperatures (in
C).
Not all materials have a distinct transition temperature (Figure 2.7). BCC
metals have transition temperatures that increases for steesl as the carbon
content increases (see Fig.2.8) , but most FCC metals do not. FCC metals
have high absorbed energies, with the energy decreasing gradually and,
102
Fig.2.7 Schematic curves for the three general types of impact energy
versustemperature behavior.
103
(a)
(b)
Fig. 2.9 - a) The Titanic in the shipyard during her construction. Note the hull
plates, fastened on all sides to the ship's main structure by thousands of rivets; b)
a layout of the watertight compartments and the damage from the collision. The
thick black lines below the waterline indicate the approximate locations of the
damage to the hull.
104
Fatigue failure
Fatigue is a form of failure that occurs in structures subjected to dynamic
and fluctuating stresses (e.g., bridges, aircraft, and machine components).
Under these circumstances it is possible for failure to occur at a stress level
considerably lower than the tensile or yield strength for a static load. The
term fatigue is used because this type of failure normally occurs after a
lengthy period of repeated stress or strain cycling. Fatigue is important inasmuch as it is the single largest cause of failure in metals, estimated to
comprise approximately 90% of all metallic failures. Furthermore, fatigue
is catastrophic and insidious, occurring very suddenly and without warning. Fatigue failure is brittlelike at naked eye observation even in normally
ductile metals, in that there is very little, if any, gross plastic deformation
associated with failure.
The first known catastrophic fatigue failure, involving major loss of life,
was the Versailles (France) railway accident in 1842 (Fig.2.10).
The train was 17 carriages hauled by two steam engines. The front axle of
the leading, four wheeled engine failed due to metal fatigue and the body
of the leading engine fell to the ground.
Fatigue testing of specially designed laboratory specimens started in the
1850s. It is generally accepted that the first fatigue tests on laboratory
specimens were carried out by Wohler. In 1870 Wohler published a final
report. This is cited in Schultz (1996) and contains the conclusions known
as Whlers 8 laws:
Materials can be induced to fail by many repetitions of stress, all
of which are lower than the static strength;
The stress amplitudes are decisive for the destruction of the cohesion of the material;
The maximum stress is of influence only in so far as the higher it
is, the lower are the stress amplitudes which lead to failure.
August Whler, chief superintendent of rolling stock on the Lower SilesiaBrandenberg Railroad, is known for his investigation on the causes of fracture in
railroad axles that led to definition of bending rotating fatigue tests, the work that
is basis of experimental tests today conducted onto materials for the definition of
the S-N curve, also known as Wohler Diagram, in the next described.
8
105
(a)
(b)
Fig.2.10 a) Paint showing Versailles (France) railway accident in 1842; b) description of rupture occurred on train axle.
Investigation on phenomena (early 1900s) led to understand fatigue failures may because of the application of fluctuating stresses that are much
lower than the stress required to cause failure during a single application of
stress. A test apparatus was therefore designed to to duplicate as nearly as
possible the service stress conditions, like stress level, time frequency,
106
(a)
(b)
Fig.2.11 a) Scheme of possible cyclic load for test specimens; from left to right
and up to down: axial test, rotating bending with varying bending stress; rotating
bending with constant bending stress; b) schematic diagram of fatigue-testing apparatus for making rotating-bending tests.
A series of tests are commenced by subjecting a specimen to the stress cycling at a relatively large maximum stress amplitude (max) below the
107
108
Fig.2.12 Variation of stress with time that accounts for fatigue failures. (a) Reversed stress cycle, in which the stress alternates from a maximum tensile stress
() to a maximum compressive stress (min) of equal magnitude. (b) Repeated stress cycle, in which maximum and minimum stresses are asymmetrical relative to the zerostress level; mean stress m, range of stress r, and stress amplitude a are indicated. (c) Random stress cycle.
109
Fig.2.13 Stress amplitude (S) versus logarithm of the number of cycles to fatigue failure (N) for (a) a material that displays a fatigue limit and (b) a material
that does not display a fatigue limit.
For many steels, fatigue limits range between 35% and 60% of the tensile
strength. Most nonferrous alloys (e.g., aluminum, copper, magnesium) do
not have a fatigue limit, in that the SN curve continues its downward
trend at increasingly greater N values (Fig.2.13b).Thus, fatigue will ultimately occur regardless of the magnitude of the stress. For these materials,
110
111
112
Fig.2.16 - Fatigue crack nucleation at notched region in the sharp corners of component.
(a)
(b)
Fig.2.17 a) Schematic view of early stage of crack initiation; b) development of
initiation site onto defect-free surface by extrusions and intrusions phenomena.
113
Fig.2.18 - Fatigue fracture surface: (a) at low magnifications, the beach mark pattern indicates fatigue as the fracture mechanism. The arrows show the direction of
growth of the crack front with the origin at the bottom of the photograph; b) at
high magnification by Scanning Electron Microscope.
114
(a)
(b)
Fig.2.20 a) Schematic view of crack propagation stage; b) development of striations due to mechanisms of ductile tearing driven by plastic radius formation at
crack tip.
115
Cracks grow from an initial size, a0, to a critical size, ac, corresponding to
failure as a function of the number of load cycles (Fig. 2.21). The crack
growth rate, da/dn, can be determined from the slope of the curve. Initially, the crack growth rate is slow but increases with increasing crack length.
Of course, the crack growth rate is also higher for higher applied stresses.
116
117
More specifically, we observe that during a cycle the K value will cyclically vary, due to variation ranging from max to min (refer to Fig.2.22).
Thus, over one cycle, the expression for intensity factor can be represented
by:
(eq.2.22)
= =
The cyclic stress intensity K increases with time (at constant load) because the crack grows in tension. It is found that the crack growth per cycle, da/dN, increases with K in the way shown in Fig. 2.24.
Particularly, in the regime for the crack growth, the crack grows at a constant rate, da/dn, accordingly with the mathematical expression:
(eq.2.23)
118
(eq.2.25)
That is valid for alternate stress (for average stress m different from zero
value, the Paris Law and the eq.2.25 should be corrected by further factors).
STAGE III - Fast Fracture
A subcritical crack of a length a higher than non-propagating threshold (refer to Fig.2.24) will eventually grow to a length at which the metal immediately fails, that is (refer to previous section on fast fracture):
(eq.2.18bis) =
Or:
(eq.2.22bis)
= 2 2
119
Contact fatigue9
Damaging phenomena developing in several mechanical component, such
as cams, gears, and bearings, are related to surface contact. Substantial differences exist in various failure modes, the appearance of worn surface is
different, and the type of contact interaction changes (as for pure rolling,
rolling with sliding, linear sliding). In general, wear modes are basically
classied into four categories, such as: fatigue wear, adhesive wear, abrasion wear, and corrosion wear. Most commonly, fatigue wear is such failure mode that occurs, for example, in a lubricated contact realized between
two parts engaged in gearing. Since in many lubricated contacts surface
adhesion is prevented and adhesive wear ideally avoided, a certain amount
of wear is caused by surface asperities interaction, as discussed in the following. In literature, contact fatigue failure (also called surface fatigue) is
considered a subcategory of fatigue wear, and it includes all those fatigue
surface damaging caused by a repeated rolling contact. In contact fatigue,
cracks are promoted by reversal stress eld generated below the contact
area formed when two elastic bodies are pressed together and they roll.
Depending on various operative conditions, either pure rolling or sliding rolling can be realized. In any case, contact fatigue damage consists
in craters forming (namely, pits) on surface. In literature, it is common to
distinguish in shallow (less than 10 m) or deeper craters (from 10
,m to 200 m); in the rst case, this type of surface damage is called
pitting, in the second case spalling. According to these denitions, it must
be highlighted that pitting and spalling should be distinguished only in
terms of the average dimensions of the craters, that is, they should be distinguished on the basis of the surface damage appearance. Since it is not
easy to clearly distinguish whether surface ruptures result from either increasing damage level (of a single rupture mode) or pitting changing into
spalling, these two terms are frequently used indiscriminately. With reMain reference: G.Stachowiak and A.W. Batchelor, Engineering Tribology,
Third Edition, Elsevier, 2005
9
120
Figure 2.25 Voids created on top surface on gear tooth by cyclic loading (spalling rupture mechanisms).
121
Figure 2.27 Stress status in a static contact; 1, 3 are the principal stresses, p is
the hydrostatic pressure, k is the shear yield stress of the material.
122
If the contact load is sufficiently high then the maximum shear stress will
exceed the resolved shear stress in polycrystalline material, i.e. max > 1/3
y 10, and plastic deformation takes place. Material will then deform (slip)
along the line of action of maximum shear stress. The maximum shear
stress max, also referred to in the literature as 45 since it acts on
planes inclined at 45 to the interface (in static contacts), is given by:
(eq.2.27) = 45 =
1 3
The stresses x, z and vary with depth below the interface. An example of the stress field beneath the surface of two parallel cylinders in
static contact is shown in Fig. 2.28.
Figure 2.28 Subsurface stress field for two cylinders in static contact; pmax is the
maximum contact pressure, b is the half width of the contact rectangle.
10
123
Figure 2.29 - Stresses in a contact with sliding; 1, 3 are the principal stresses, p
is the hydrostatic pressure, k is the shear yield stress of the material, is the coefficient of friction, q is the stress normal to the interface or compressive stress due
to load, is the angle by which the planes of principal stress are rotated from the
corresponding zero friction positions to balance the frictional stress.
124
Figure 2.30- Subsurface stress field for a cylinder sliding on a plane; pmax is the
maximum contact pressure, b is the half width of the contact rectangle.
Box Contact Between Two Parallel Cylinders
The configuration of two elastic bodies with convex surfaces in contact was
originally considered by Hertz in 1881 and is shown in Figure 2.31 and it is
based on the following assumptions:
i.
The surfaces are continuous, smooth, nonconforming and frictionless;
ii.
The size of the contact area is small compared to the size of the bodies, i.e., the strains associated with the deformations are small;
iii.
Each solid can be considered to behave as an elastic half-space in the
vicinity of the contact zone;
iv.
The gap h between the undeformed surfaces can be approximated by
an expression of the form: h = Ax2+By2;
Figure 2.31 - Geometry of two Hertzian elastic bodies with convex surfaces in
125
contact.
The reduced radius of curvature for this case is defined by Hertzian theory as:
Where:
b is the half width of the contact rectangle [m];
l is the half length of the contact rectangle [m];
E' is the reduced Young's modulus is defined as:
R' is the reduced radius of curvature for the two parallel cylinders in
126
contact [m]. For the cylinders: Rax = RA, Ray = , Rbx = RB, Rby =
where RA and RB are the radii of the cylinders A and B respectively:
where:
Since 1/Rx > 1/Ry condition is satisfied, the reduced radius of curvature is:
127
Maximum Deflection:
Figure 2.33 - Hydrodynamic pressure distribution in an elasto-hydrodynamic contact; hc is the central film thickness, h0 is the minimum film thickness.
128
Figure 2.34 - Simplified film geometry and generation of traction in an EHL contact.
129
The traction force has the same effect of increasing the shear stress on top
surface, as shown in the previous scheme of Fig.2.30 (refer to previous
section Hertzian stress status in sliding contact).
Particularly, it is important to note that this condition can occurs not only
at macroscopic scale, namely considering opposite lubricated surfaces as
homogeneous, but the EHL scheme is applicable to microscopic scale.
Consider the effective roughness of a lubricated surface that is put, for example, into rolling contact (as it happens actually for lubricated gears). The
effect of surface roughness on partial EHL is illustrated in Fig.2.35. It has
been found that the sharp peaks, i.e. asperities with high slope or low radius sustained a higher proportion of the contact load than flat peaks, i.e.
asperities with a low slope or large radius. Improved surface finish enables
a diminished fraction of contact load supported by the asperities and the
likelihood of a perfect elasto-hydrodynamic film is enhanced 11.
Figure 2.35 - Effect of roughness and asperity shape on survival of EHL films.
At high levels of surface roughness (see Fg.2.35, left side), the EHL film sustains
a reduction in minimum film thickness and rapid wear of the contact occurs that
produces a corrective wearing of surface in early lifespan of contact system bodies. When asperities are sufficiently smooth, or smoothed after corrective surface
wearing, they are prevented from contacting each other, due to the mechanism of
micro-elastohydrodynamic lubrication or micro-EHL, as schematically
shown in Fig.2.36.
11
When surfaces are polished to an extreme smoothness, however, a contrary trend to
lowered load capacity is probable. It has often been observed in engineering practice that if
the surface is too smooth, e.g. with a surface roughness of 0.001 [m] Ra, then there is a
risk of sudden seizure. In this instance it is commonly believed that small asperities play a
useful role as a reservoir for the lubricant by entrapment between asperities. Under extremes of contact pressures the trapped lubricant can be expelled by asperity deformation to
provide a final reserve of lubricating oil.
130
131
Figure 2.34 - Schematic diagrams of the pivoted bearings: a) a offset line pivot, b) a button point pivot.
132
Pmax
Pmax
2a
tau max
Time
(a)
f (A)
f (B)
micropitting
(pitting/spalling)
(b)
Figure 2.35 (a) Subsurface shear stress produced during cycling of external load
contact; (b) condition that leads to crack initiation (lifetime depends on loading
condition, surface roughness, lubrication conditions, etc.) within the surface, producing subsurface propagation toward surface and void creation (pitting or spalling mechanisms).
133
(a)
(b)
Figure 2.36 (a) Material imperfection as the cause for contact fatigue; (b) illustration of a process of subsurface crack formation by growth and link up of voids.
134
with friction coefcient approximates unity, the material within the 0.1
mm of the surface is stretched along the sliding direction. In these regions,
microstructure becomes hardly oriented, and this severe plastic deformation makes the dislocation density increase. Subsequent work hardening
occurs in these high stretched layers, as they are schematically illustrated
in Fig.2.37.
The high strain induced by traction in EHL contact (or sliding in dry contact) eventually breaks down the original grain structure at the surface to
form dislocation cells, as it occurs in severe work hardening processes.
The example of such a work-hardened like deformed material is shown in
the micrograph of Fig.2.38.
135
The mechanism of surface crack initiated fatigue wear is illustrated schematically in Fig. 2.39. A primary crack originates at the surface at some
weak point and propagates downward along weak planes such as slip
planes or dislocation cell boundaries. A secondary crack can develop from
the primary crack or alternatively the primary crack can connect with an
existing subsurface crack. When the developing crack reaches the surface
again a wear particle is released.
Figure 2.39 Schematic illustration of the process of surface crack initiation and
propagation.
136
To complete the preliminary analysis of cyclically loaded bodies in lubricated contact, typical of gears, the model in the Fig.2.35 is capable to explain also the singular cases of case-crushing failure shown in Fig.2.26. As
shown in Fig.2.41, the case-crushing condition occurs when the shear
stress Herztian profile exceed somewhere at interface case-core interface
the fatigue limit shear stress f of the material.
137
f (C)
f (A)
Case-crushing
Figure 2.41 Condition that leads to crack initiation in depth, at the interface between the hardened layer, also named case, and the base material, also named
core.
138
(static s )
(dynamic k )
How is it that the friction between two surfaces can depend only on the
force P pressing them together and not on their area? In order to understand this behavior, we must first look at the geometry of a typical surface.
139
contact area is therefore distributed between a number of microcontact areas. If the load is raised, the number of contact areas rather than the average individual size of contact area is increased, i.e. an increase in load is
balanced by newly formed small contact areas. A representation of contact
between solids is shown schematically in Figure 10.12.
ai
Fig. 2.43 - Real contact area of rough surfaces in contact; ar = ai is the true area
of contact.
Even if the surface is polished for a long time using the finest type of
metal polish, micro-asperities still survive.
The load pressing the surfaces together is supported solely by the contacting asperities. The real area of contact, a, is very small and because of this
the stress P/ a (load/ area) on each asperity is very large. Initially, at very
low loads, the asperities deform elastically where they touch. However, for
realistic loads, the high stress causes extensive plastic deformation at the
tips of asperities. If each asperity yields, forming a junction with its partner, the total load transmitted across the surface (Fig. 25.3) is:
(eq.2.30)
P ay
where y is the compressive yield stress. The real area of contact is therefore given by:
(eq.2.31) a P/y
140
Fig. 2.44 - The real contact area between surfaces is less than it appears to be, because the surfaces touch only where asperities meet.
Let us now look at how this contact geometry influences friction. If you attempt to slide one of the surfaces over the other, a shear stress Fs / a appears at the asperities. The intense plastic deformation in the regions of
contact presses the asperity tips together so well that there is atom-to-atom
contact across the junction. The junction, therefore, can withstand a shear
stress as large as k approximately, where k is the shear-yield strength of
the material. The asperities will give way, allowing sliding, when Fs/a k.
The distinction between a static and a sliding contact is thought by modern
theory as due to differences in structure and physical processes that occurs.
It was found that while a static contact can be described in terms of a random distribution of point contacts as above described, this model is not
applicable to a sliding contact. A basic feature of sliding contact is that it is
distributed over a lesser number of larger contact areas rather than a large
number of contact points. These areas do not have a fixed location inside
the contact but instead move slowly across the surface as sliding progresses. The frictional interaction appears to be controlled by mechanical interlocking between lumps on opposing surfaces as schematically illustrated
in Figure 2.45.
141
142
Fig. 2.47 - Oxide-coated junctions con often slide more easily than ones
which are clean.
Referring to data in Fig. 2.47, it can be generally observed that:
1. When soft metals slide over each other (e.g. lead on lead), the
junctions are weak but their area is large so is large;
2. When ceramics slide on ceramics friction is lower: ceramics are
very hard, stable in air and water so they have less tendency to
bond and shear more easily.
3. Metals slide on bulk polymers, friction is still caused by adhesive
junctions, transferring a film of polymer to the metal: any plastic
flow tends to orient the polymer chains parallel to the sliding surface, and in this orientation they shear easily, so is low - 0.05 to
0.5.
143
1
2
Fig. 2.47 - Bar chart showing the coefficient of static friction for various
material considerations.
Lubrication
Friction absorbs a lot of work in machinery and as well as wasting power,
this work is mainly converted to heat at the sliding surfaces, which can
damage and even melt the bearing. In order to minimize frictional forces
we need to make it as easy as possible for surfaces to slide over one another. The obvious way to try to do this is to contaminate the asperity tips
with something that: (a) can stand the pressure at the bearing surface and
so prevent atom-to-atom contact between asperities; (b) can itself shear
easily.
Polymers and soft metal, as we have said, can do this; but we would like a
much larger reduction in than these can give, and then we must use lubricants. The standard lubricants are oils, greases and fatty materials such as
soap and animal fats. These 'contaminate' the surfaces, preventing adhesion, and the thin layer of oil or grease shears easily, obviously lowering
the coefficient of friction. What is not so obvious is why the very fluid oil
is not squeezed out from between the asperities by the enormous pressures
generated there. One end of each molecule reacts with the metal oxide surface and sticks to it, while the other attracts one another to form an oriented 'forest' of molecules (Fig. 2.48). These forests can resist very large forc-
144
es normal to the surface (and hence separate the asperity tips very effectively) while the two layers of molecules can shear aver each other quite
easily. This type of lubrication is termed partial or boundary lubrication,
and is capable of reducing by a factor of 10 (Fig. 2.47). As described in
the previous section Contac fatigue, hydrodynamic lubrication is even
more effective, since it aims to separate surfaces.
Wear of materials
Even when solid surfaces are protected by oxide films and boundary lubricants, some solid-to-solid contact occurs at regions where the oxide
film breaks down under mechanical loading, and adsorption of active
boundary lubricants is poor. This intimate contact will generally lead to
wear. Wear is normally divided into two main types: adhesive wear and
abrasive wear.
Adhesive wear
Figure 2.49 shows that, if the adhesion between A atoms and B atoms is
good enough, wear fragments will be removed from the softer material A.
If materials A and B are the same, wear takes place from both surfaces the wear bits fall off and are lost or get trapped between the surfaces and
cause further trouble (see below). The size of the bits depends on how far
away from the junction the shearing takes place: if work hardening extends
well into the asperity, the tendency will be to produce large pieces.
145
Abrasive wear
Wear fragments produced by adhesive wear often become detached from
their asperities during further sliding of the surfaces. Because oxygen is
desirable in lubricants (to help maintain the oxide-film barrier between the
sliding metals) these detached wear fragments can become oxidized to
give hard oxide particles which abrade the surfaces in the way that sandpaper might. Figure 2.50 shows how a hard material can 'plough' wear
fragments from a softer material, producing severe abrasive wear. Abrasive wear is not, of course, confined to indigenous wear fragments, but can
be caused by dirt particles (e.g. sand) making their way into the system, or
- in an engine - by combustion products: that is why it is important to filter
the oil.
The wear rate W related to the wear path s is proportional to the normal
load FN and the hardness of the worn surface HW and a factror k, accordingly with experimental Archard equation:
(2.31)
146
The rate of abrasive wear can be reduced by reducing the load - just as in a
hardness test. The particle will dig less deeply into the metal, and plough a
smaller. Increasing the hardness of the metal will have the same effect.
147
Corrosion 12
Corrosion is a form of degradation caused by chemical and electrochemical reactions that take place at the interface between the surface of a material and the environment with which it is in contact: the result is a gradual
decay of integrity of the metal surface occurring when the main constituents of the alloy (Fe, C, Cr, Ni, etc.) combine with the aggressive agents
present in the environment, thus forming corrosion products.
The corrosion of metals can cause several problems. In the case of piping
and tanks, the degradation can give rise to punctures that lead to leakage of
the fluids contained within; in structural elements corrosion can cause the
reduction of the resistant part with the consequent loss of the components
load capacity. Other problems are also linked to the formation of corrosion
products, which can lead to the alteration of the aesthetic features of the
surfaces or the contamination of the processed substances (for example, in
the case of foods, possible alteration to their organoleptic characteristics).
The corrosion of metals can be divided into two basic forms:
electrochemical corrosion also called wet corrosion in which
the metal alloy undergoes an oxidation reaction in the presence of
an electrolyte (usually water); the oxidation reaction is coupled with
a reduction reaction of the substances present in the environment
(usually oxygen): the combination of the two reactions, anodic (oxidation) and cathodic (reduction), involve both chemical species
(ions and molecules) and electrons;
chemical corrosion (also called high temperature oxidation or dry
corrosion) in which the metal alloy undergoes an oxidation reaction
upon contact with a gaseous atmosphere (usually air); the phenomenon usually occurs at temperatures well above ambient temperature
(>300C) and the corrosion products are oxides and salts with low
melting points.
Wet corrosion and electrochemical corrosion
Main references:
M. Boniardi, A. Casaroli, Stainless Steels, Lucefin Group, 2014 (free English version avaliable at:
http://www.lucefin.com/wp-content/files_mf/stainlesssteels_low.pdf );
M.F.Ashby and M.F.Jones, Engineering Materials 1, Engineering Materials 1, Second Edition. Butterworth Heineman, Oxford, 1996.
12
148
As the copper is a nobler metal than iron (copper has a greater electrochemical potential than iron), a potential difference is established between copper and iron, i.e. an electromotive force that allows the circulation of current (see Box for more details). The analogy using the galvanic
cell is clear: the iron behaves as an anode and the copper as a cathode and
149
the potential difference between the two metals allows the circulation of
current. The oxidation reaction occurs to the anode (iron) and the reduction
reaction of oxygen occurs to the cathode (copper) at the same time.
Box Voltage differences (electrochemical potential difference) as a driving force for wet oxidation
Wet oxidation in metals involves electron flow, as metals are electron conductors. Thus, it is easier to measure the tendency of a metal to oxidise in a solution by using a voltage scale, namely a measure of such a flow rate of electrons. Figure 2.52 shows the voltage differences that would just stop various
metals oxidising in aerated water.
150
anodic regions of a piece of iron, as shown in Fig. 2.53. Then at the cathode,
oxygen is reduced to OH-, absorbing electrons, and the metal therefore becomes positively charged. The reaction continues until the potential rises to
+0.401 V. Then the coulombic attraction between the +ve charged metal and
the -ve charged OH- ion becomes so large that the OH- is pulled back to the
surface, and reconverted to H2O and O2; in other words, the reaction stops. At
the anode, Fe++ forms, leaving electrons behind in the metal which acquires a
negative charge. When its potential falls to -0.440V, that reaction, too, stops
(for the same reason as before). If the anode and cathode are now connected,
electrons flow from the one to the other, the potentials fall, and both reactions
start up again. The difference in voltage of 0.841V is the driving potential for
the oxidation reaction. The bigger it is, the bigger the tendency to oxidise.
151
O2 + H2O + Fe Fe (OH)2
In summary: the iron gradually transforms into solution under the form of
Fe+2 ions, corroding due to the effect of the presence of the copper cathode
where the reduction of oxygen occurs. In the example shown previously,
the anodic and cathodic zones are clearly distinct: the corrosive phenomenon occurs due to a galvanic coupling that creates a potential difference
E because of the different nature of the two metals involved.
However, it should not be believed that corrosion only occurs in these
conditions: the case of degradation phenomena on metal alloys not in contact with dissimilar metals is much more common.
To understand the problem better the experience of Evans should be considered, as shown in Fig. 2.54: his experiment is especially significant as it
explains the phenomenon of corrosion of homogenous metal materials.
152
Initially, the chemical composition of the drop is constant and the oxygen
is totally absent: over time however, due to contact with air, diffusive phenomena will occur near to the external surfaces of the drop (which is the
surface of exchange with the surrounding air) and a variation of the oxygen concentration will take place in the drop. The different concentration
of O2 in the drop will create an anodic zone, i.e. the zone lacking in oxygen in the center of the drop, and a cathodic zone, i.e. the zone rich in oxygen on the outside of the drop. Therefore, a local anode-cathode micro
bond is created which is able to trigger the corrosion process. The E potential existing between the anode and cathode is a kind of electromotive
force that allows the corrosive process to take place: it is called free corrosion potential.
The anodic reaction and the cathodic reaction are all the same as those observed previously for galvanic contact:
(eq.2.35)
Fe Fe+2 + 2e-
(eq.2.36)
overall obtaining:
(eq.2.37)
O2 + H2O + Fe Fe(OH)2
Fe Fe+2 + 2e-
(eq.2.39)
2H+ + 2e- H2
overall obtaining:
(eq.2.40)
2H+ + Fe Fe+2 + H2
For corrosion to occur, the two reactions, anodic and cathodic, must happen at the same time: during the oxidation reaction of the iron, a certain
number of electrons are generated on the anode which, following the reduction reaction, are consumed by the cathode.
153
The above reactions are very common in all metal materials subject to corrosion: the presence, for any reason, of an anodic-cathodic micro bond
provides the electromotive force required to trigger and supply the degradation process.
However, the phenomenon of differential aeration corrosion is not the only type to give rise anodic-cathodic micro bonds able to trigger degradation: often in the corrosive material environment combination there are
already particular local conditions suitable for generating zones with different electrical potential.
The problem arises for various reasons: heterogeneity of chemical composition of the metal mass, in homogenous phases of the microstructure of
the matrix (inclusions, carbides, etc.), Tensile residual stress of high entity
traction, local defects of the piece (micro cavities, blowholes, accentuated
roughness etc.) All these areas act as the anode of the surrounding metal
mass acting as a cathode, causing local micro bonds with preferential corrosion triggers.
Similar situations also arise due to the variability of the electrochemical
characteristics of the corrosive environment, such as that which occurs, for
example, in solutions with various concentrations of noxious species, in
electrolytes with greater or smaller stagnation zones or with non-uniform
temperatures.
Box Potentiodynamic curves
During a corrosive process there is always a certain number of ions in solution
in the electrolyte and a concomitant movement of electrons in the metal, i.e. a
circulation of electrical current, as what occurs in a cell.
The mechanism described is concatenated: the higher the quantity of iron ions,
which will dissolve in the solution, the greater the number of electrons circulating per surface unit exposed to the corrosive environment. This causes the increase of the surrounding current density ic and the corrosion rate Vcorr of
this material will increase in that given environment, i.e.:
corrosion rate: (Vcorr)
ic = ianodic = icathodic
154
process (i.e. how quickly the metal dissolves) and the electrochemical characteristics of the cathodic process (i.e. how quickly the oxygen or the hydrogen
evolves). If the two contributions are separated, i.e. if the correlation E i is
studied separately for the anodic reaction and the cathodic reaction, it is possible
to estimate what the trend of the surrounding current density i will be, in order
to change the applied electromotive force E. This is exactly what Evans did
experimentally and which is still possibly to recreate in the laboratory: to trace
the curves that describe the anodic reaction and cathodic reaction with the variation of the tension conditions E applied, or the reaction existing between the
set tension E and the surrounding current density i both in the anodic area
and in the cathodic area.
The curves obtained experimentally are shown in fig.2.55: they are called potentiodynamic curves, or Evans diagrams, due to the anodic process (anodic characteristics, metal/alloy that is corroded) and the cathodic process (cathodic characteristics, oxygen reduction and hydrogen evolution).
In the very simple case of common carbon steel in an aerated aqueous solution
(the cathodic process is oxygen reduction), the two curves are as shown in
fig.2.55a, whereas for a stainless steel, again in the same solution, the situation
shown in fig.2.55b occurs.
Figure 2.55 - Indicative potentiodynamic curves (anodic characteristic A: material cathodic characteristic B: environment) in aqueous solution 0.05 M of
H2SO4: (a) for a material with active conduct and b) for a material with activepassive conduct.
The intersection point of curves A and B shown in fig.2.55 represents the condition of equivalence between the surrounding current densities (the rate of the
two reactions, anodic and cathodic, is equal): it establishes the functioning point
of the material-environment system and allows the determination of the free cor-
155
rosion potential Ec and the surrounding current density ic, as well as estimating the corrosion rate of the material in that particular environment (Vcorr
ic).In the case of carbon steels, for example, the anodic curve (curve A
fig.2.55a) increases monotonically: the current increases with the increasing tension applied and with it, the corrosion rate increases. With regards to In the case
of carbon steels, the anodic curve (curve A fig.2.55a) increases monotonically:
the current increases with the increasing tension applied and with it, the corrosion rate increases. Concerning stainless steels the curve is that represented in
fig.2.55b, and its different behaviors will be explained in Chapter 5 that is devoted to stainless steels. For the moment, note the effect of the difference of the
two potentio-dynamic behaviors represented in figure 2.55a and fig.2.55b that
represent respectively a carbon steel and a stainless steel: because the surrounding current ic of case a is much greater than the surrounding current ic of
case b, the corrosion rate of a carbon steel, in that determined environment,
will be a lot greater than that of a generic stainless steel.
Vcorr = m/At
156
In industrial applications it is often more interesting to talk about penetration or thinning rate Vthin of the corrosive phenomenon. To convert corrosion rate to the thinning rate simply add the material density (), according to the formula:
(eq.2.42) Vass = Vcorr/ = m/ A t
The formula is commonly used to predict the so-called additional thickness for corrosion, i.e. a thickness in excess on the component which,
during operation, will be gradually removed due to the corrosive action of
the environment.
Galvanic corrosion
An early form of localized corrosion is galvanic corrosion or corrosion due
to galvanic coupling. It occurs when a metal and a metal alloy is connected
(i.e. is in electrical contact) with another metal/metal alloy with greater or
lower thermodynamic nobility: if there is a significant potential difference,
a redox reaction may develop with consequent corrosion phenomena, even
in the presence of just a mildly aggressive environment. The mechanism is
very similar to that shown in Fig.2.51. Every metal or metal alloy has its
own potential that depends on its nature, its chemical composition and on
the environment in which it is placed (temperature, pH, agitation, presence
of oxidants or other harmful species, etc.).
If the difference between the various potentials exceeds a certain threshold,
a significant passage of electrons is created between the donor (the anode, the less noble metal) and the receiver (the cathode, the nobler metal). The intensity of this movement of electrons (actually an electric current) will be greater as the potential difference increases: as a consequence,
the corrosion rate will be higher the further distance between the two materials on the nobility scale. Another relevant aspect that governs the phenomenon of galvanic corrosion is the relationship between the areas of the
two materials in contact: the rate of degradation increases with the increase
of the ratio between the area of the cathodic zone (nobler) to that of the
anodic zone (less noble) exposed to the environment.
In order to assess whether the conditions of galvanic contact between the
two metals/metal alloys can cause problems practically, it is better NOT to
refer to the standard electrical potential scales: this is because there is a
great variability in the behavior of the same material between one corrosive environment and another (and at times also within the same environ-
157
ment) and also because there are never conditions of equilibrium in the actual environments and electrochemical system. Therefore the saltwater
scale of nobility is used (figure 2.56), which is a scale of the potentials
measured under conditions close to actual operation. A typical example of
galvanic corrosion is when it occurs on carbon steel sheet metal (also galvanized) or on aluminum alloy sheet metal in contact with stainless steel
fasteners, placed in marine environments, in aerate aqueous solutions or in
mildly aggressive environments (see fig.2.57).
The opposite situation would be much more serious, i.e. that of stainless
steel sheets fastened with carbon steel or galvanized steel rivets: in this
second case, in addition to the nobility difference between the two materials, the relationship between the areas would be very negative, in favor of
stainless steel.
158
Figure 2.57 Galvanic corrosion of a carbon steel plate with stainless steel fasteners.
159
The degradation occurs when the surface of the stainless steel is sullied
by foreign particles, such as dust from ferrous material that is often generated during grinding operations and from common steel.
The particles deposited on the surface can create conditions for localized
corrosive attacks, even in not very aggressive environments. Marks appear
on the surface (rust-colored in case of ferrous contamination, whitish in the
case of contamination from aluminum or zinc) as a consequence of the
rapid oxidation of the contaminant (steel, aluminum or zinc) and not, as
wrongly believed, of the stainless steel. In the most extreme situation, the
contaminant substance can also cause damage to stainless steel, as it can
hinder the passivation phenomenon, as well as constitute a preferential
trigger zone for other forms of corrosion.
If surface contamination is suspected, the foreign particles must be removed through a chemical passivation operation, using nitric acid based
dilute solutions.
Pitting corrosion
Pitting is the phenomenon of localized corrosion; it produces serious penetrating damage and dangerous holes in the components during operation.
Characteristic elements of pitting are the presence on the surface of the
work piece of multiple small pits (hence the name of this corrosion); pitting usually leads to even deeper pits, i.e. grooves, ulcers, craters etc. The
surface size of the holes is small, between 0.1 and 2 mm, as in the pitting
phenomena the extent of the corrosion products is very limited.
The greatest problem related to this form of corrosion is not the loss of
mass affected by the degradation phenomenon, but rather the damage
caused by penetrating through the resistant section of the component. For
example, consider the case of a stainless steel tank affected by the pitting
phenomenon: the pitting could lead to the creation of a hole in the recipient, the subsequent spilling of the process fluid and the overall disruption
of the system.
Degradation due to pitting has various aspects; fig. 2.59 shows some morphologies typical in sections, as classified by the American standard
ASTM G46.
160
Pitting occurs when discrete areas of a material undergo rapid attack while
most of the adjacent surface remains virtually unaffected. Although the total metal loss may be small, the part may be rendered useless due to perforation. In addition to the localized loss of thickness, corrosion pits can also
act as stress raisers, leading to fatigue or stress-corrosion cracking typical
of austenitic stainless steels (refer to Chapter 5).
Pitting occurs when the anodic or corroding area is small in relation to the
cathodic or protected area. Pitting can occur in protected metals when
there are small breaks in the continuity of the metal coating. Pitting can also occur on bare, clean metal surfaces as a result of irregularities in their
physical or chemical structure. The rate of penetration into the metal by
pitting can be 10 to 100 times that caused by general (uniform) corrosion.
Pitting can cause structural failure from localized weakening while considerable sound metal still remains. Pitting usually requires a rather long initiation period before attack becomes visible. However, once a pit has started, the attack continues at an accelerating rate. Pits tend to grow in a
manner that undermines or undercuts the surface. Typically, a very small
hole is seen on the surface. Poking at the hole with a sharp instrument may
reveal a much larger hole under what had looked like solid metal. Pitting
can cause visible pits, or they may be covered with a semipermeable membrane of corrosion products. Pitting corrosion may assume different
shapes.
Pitting normally occurs in a stagnant environment. Concentration cells can
accelerate pitting. Concentration cells are areas on the metal surface where
the oxygen or conductive salt concentrations in water differ. As a pit becomes deeper, an oxygen concentration cell is started by depletion of oxygen in the pit. The rate of penetration of such pits is accelerated propor-
161
tionately as the bottom of the pit becomes more anodic. Pitting attack increases with temperature. Variations in soil conditions can also trigger pitting.
Crevice corrosion
Crevice corrosion is another form of localized aggression: it occurs in the
presence of cracks, crevice, incrustations, deposits and geometrical discontinuities in which the electrolyte (generally water) is in stagnation with regards to the surrounding environment. A typical case of degradation due to
crevice corrosion regards the seals of flanged piping or in the contact
zones between bolted or riveted plates (fig. 2.60). Similar situations can
occur in welded plates for points on cars, in arc welded seals where penetration is incomplete, between the strands of metal cables, underneath lubricating films such as graphite or molybdenum graphite and on surfaces
coated with Teflon or polyethylene.
Figure 2.60 Schematic of a constructive solution which could cause crevice corrosion.
162
163
Creep 13
At high temperatures, permanent deformation can occur over a period of
time at stresses well below the yield strength. This time-dependent deformation is known as creep and occurs at temperatures greater than approximately 0.3 to 0.5 of the absolute melting point Tm. In creep, thermal activation enables plastic deformation at stresses below those needed to
deform the lattice without thermal activation. When a metal is placed under a constant load, it stretches elastically but also gradually extends plastically. Thus, a metal subjected to a constant tensile load at elevated temperature will creep and undergo a time-dependent increase in length. Since
the mobility of atoms increases with increasing temperature, diffusioncontrolled mechanisms become active. Dislocation mobility increases, slip
becomes easier, new slip systems become available, and dislocation climb
is aided by both increases in temperature and the presence of a greater
number of vacancies. Creep occurs in any metal or alloy at a temperature
where atoms become sufficiently mobile to allow the time-dependent rearrangement of structure. Since the elevated-temperature strength of metals
is closely related to their melting points, it is normal practice to specify the
homologous temperature, which is the ratio of the exposure temperature to
the melting point (T/Tm), on an absolute scale. Creep behavior of a polycrystalline metal or alloy often is considered to begin at approximately 1/3
to 1/2 of its melting point (~0.3 to 0.5 Tm).
The Creep Curve
In a creep test, the time-dependent strain is measured under long-term elevated temperature. Engineering creep is measured by applying a constant
load on a tensile specimen at a constant temperature and measuring the
strain, or extension, of the specimen as a function of time. A schematic of
a test setup for a creep test is shown in Fig. 2.63. Creep tests may be run
for as short as several months to as long as 10 years. Creep tests are usually conducted at constant load rather than constant stress, and, as the specimen elongates, the cross-sectional area decreases, so the applied stress increases with time.
164
165
Since creep is a thermally activated process, it is not surprising that steadystate creep obeys an Arrhenius-type rate equation. The steady-state creep
rate, , in this intermediate temperature range (0.4 Tm<T<0.6 Tm) can be
expressed as a power law function:
(eq.2.43)
where Q is the activation energy for creep; A is the pre-exponential constant; n is a constant, usually between 3 and 10; R is the universal gas constant; and T is the absolute temperature. This equation describes the dependence of creep rate on the key variables, temperature and This equation
describes the dependence of creep rate on the key variables, temperature
and stress. A correlation of creep and diffusion data for pure metals shows
that the activation energy for creep, Q, is equal to the activation energy for
166
The minimum creep rate during steady-state creep is usually the most important design parameter obtained during a creep test. Two criteria are
common: (1) the stress to produce 1% creep in 10,000 h, or (2) the stress to
produce 1% creep in 100,000 h. The first criterion is often used for jet engine components and the second for steam turbines.
Tertiary creep. It is a region of drastically increasing strain rate with rapid
extension to fracture. Tertiary creep is dominated by a number of weakening metallurgical instabilities, such as localized necking, corrosion, intercrystalline fracture, microvoid formation, precipitation of brittle secondphase particles, and dissolution of second phases that originally contributed to strengthening of the alloy. When designing components for service at
elevated temperatures, the steady-state creep rate is usually the significant
design parameter. However, the duration of tertiary creep is also important,
because it constitutes a safety factor that may allow detection of a failing
component before catastrophic fracture.
167
168
Figure 2.67 - Dislocations can climb (a) when atoms leave the dislocation line to
create interstitials or to fill vacancies or (b) when atoms are attached to the dislocation line by creating vacancies or eliminating interstitials.
169
Grain size plays a role in creep of metals. The finer the grain size, the more
rapid the mass transport, causing permanent deformation. Thus, under
conditions where creep is solely due to diffusion, the creep resistance is
improved by increasing the grain size. Note that this is different from lowtemperature behavior, where a fine grain size is almost always beneficial
to strength and ductility.
Creep Life Prediction
A frequent problem encountered in materials selection for component design is that it may be desirable to use a new alloy in a high temperature
structure; however, the alloy is so new that there are not sufficient data to
substantiate its long-term creep resistance. Therefore, it is necessary to extrapolate higher temperature. In order to extrapolate the data to the required times, a considerable amount of effort has been expended in developing empirical prediction methods. One of the most widely used is the
Larson-Miller parameter, P:
(eq.2.44)
170
171
temperatures, because the slightly more open bcc structure results in greater diffusivities.
Design against creep may also involve first determining the dominant
creep mechanism. If diffusion creep dominates, then increasing the grain
size is a beneficial microstructural alteration. Further, for diffusion creep,
improved creep resistance can sometimes be obtained by placing inert particles, such as carbides, on grain boundaries. This helps to pin the grain
boundaries, enhancing creep resistance.
As it will better discussed in Chapter 6, creep-resistant alloys include carbon steels, chromium-molybdenum steels, chromium molybdenumvanadium steels, stainless steels, nickel and cobalt alloys, and superalloys
(a comparison of the stress-rupture properties for a number of different
metallic alloy families is shown in Fig. 2.70).
172
173
letter in Table 3.1), b) then a three digit number nnn signifying the mechanical property (often yield strength) dictated in the standard for that application designation (see again Table 3.1).
Example: the S355 steel refers to structural application with minimum
guaranteed yield strength of 355 MPa.
Application symbol
Meaning
Mechanical Property
Structural steel
Engineering steels
Tinmill Products
Electrical Steel
174
The heat-treated steels, namely all such steels of any classes other than
general purpose steels type, can be designated by three different codes,
depending on the chemical composition, as below described:
ii) Carbon steel, namely steels where no alloying elements are added in
chemical composition:
C nn
where nn is the approximate carbon content multiplied by 100.
For example, C40, means 0.4% of carbon content.
iii) Alloying steel where more alloying elements have been added to chemical composition, but with all the alloying elements less than 5% weight:
nn A B C a b c
where:
- nn is the approximate carbon content multiplied by 100;
- A, B, C, etc. represent the alloying elements; elements are ordered
from left to right as they are sorted by higher to lower content;
- a, b, c (but often only at least two digit number is present) whole
numbers that represent by a multiplying factor the percentage of
the A, B and C elements respectively. Factor is:
o 4 referring to Cr, Co, Mn, Si, W, Ni;
o 10 referring to Al, Be, Cu, Mo, Ti, Pb, V, etc.;
o 100 referring to N, P, S, Ce;
o 1000 referring to B.
For example, a UNI EN 40CrMnNiMo8-6-4 is a low-alloyed steel with
0.40 %C, 2%Cr (=8/4), 1.5% Mn (=6/4), 1% Ni (=4/4) and some percentage of Mo lower than 1%.
iv) High alloying steel with at least one of alloying element more than 5%:
X nn A B C a.b.c
where:
nn is carbon content multiplied by 100;
A, B, C, etc. represent the alloying elements; elements are ordered
from left to right as they are sorted by higher to lower content;
175
Steel for special applications, that are also heat-treated steels that are
distinguished for specific applications. Several families are grouped under such larger class (e.g. high fracture toughness steels, tool steels, low
temperature steels, high temperature steels, etc.);
The Stainless Steels grade, classified into 4 types:
176
177
Table 3.2 Summary of AISI/SAE designations for carbon and low-alloy steels.
178
Classification of steels
The AISI standard groups steels into following main classes:
Plain carbon steels (10xxx series) A plain carbon steel is essentially
an alloy of Fe-C that also contains manganese and a variety of residual elements, in particular S, P (impurities) and Si (Si is added
during the production process for deoxydation stage, refer to
steelmaking section). AISI defines plain carbon steels as Fe-C with:
o 1.65 wt% maximum Mn,
o < 0.6 wt% Si,
o < 0.6 wt% Cu
o no other deliberately added alloying element.
The carbon steels are classified as follows:
o Nonresulfurized carbon steels (10xx or 10xxx series
steels); low, medium and high carbon content steels;
o Resulfurized steels (11xx series);
o Rephosphorized and resulfurized steels (12xx series);
o High manganese carbon steels (15xx series);
Alloy steels, namely Fe-C alloys with the addition of one or more of
the following elements15: Mn, Cr, Ni, Mo, Ti, Co, W, V, Si, Al, B.
Despite plain carbon steels are widely used, they are not adequate
for all engineering applications because of the following limitations:
o
o
o
o
179
180
In bainite, which is obtained by transformation of austenite at a large undercooling, the cementite is more rounded than in pearlite. Tempered
martensite, a mixture of very fine and nearly round cementite in ferrite,
forms when martensite is reheated following its formation.
Fig.3.2 - Micrographs of (a) pearlite, (b) bainite, and (c) tempered martensite, illustrating the differences in cementite size and shape among the three microcostituents.
The microstructures mentioned above in steels are produced by the transformation of the austenite microstructure (obtained by heating up structure
over A1 and A3 temperatures) throughout: a) isothermal transformation
that occurs below A1 temperature; b) continuous cooling from austeinitizing temperature to ambient temperature.
181
182
Fig. 3.3 - Schematic summary of the simple heat treatments for (a) hypoeutectoid
steels and (b) hypereutectoid steels.
16
For annealing, austenitizing of hypoeutectoid steels is conducted about 30C above
the A3, producing 100% ; however, austenitizing of a hypereutectoid steel is done at
about - 30C above the A1, producing austenite and Fe3C. This process prevents the
formation of a brittle, continuous film of Fe3C at the grain boundaries that occurs on slow
cooling from the 100% region. In both cases, the slow furnace cool and coarse pearlite
provide relatively low strength and good ductility. For normalizing, austenitizing is done at
about 55C above the A3 or Acm; the steel is then removed from the furnace and cooled in
air. The faster cooling gives fine pearlite and provides higher strength.
183
Fig. 3.4- The effect of carbon and heat treatment on the properties of plain carbon
steels.
184
the fraction of Fe3C in a steel alloy while holding other microstructural elements constant will result in a harder and stronger material. This is
demonstrated in Figure 3.13a, in which the tensile and yield strengths and
the Brinell hardness number are plotted as a function of the weight percent
carbon (or equivalently as the percentage of Fe3C) for steels that are composed of fine pearlite. All three parameters increase with increasing carbon
concentration. Inasmuch as cementite is more brittle, increasing its content
will result in a decrease in both ductility and toughness (or impact energy).
These effects are shown in Figure 3.13b for the same fine pearlitic steels.
The layer thickness of each of the ferrite and cementite phases in the microstructure also influences the mechanical behavior of the material.
Figure 3.13 - (a) Yield strength, tensile strength, and Brinell hardness versus carbon concentration for plain carbon steels having microstructures consisting of fine
pearlite. (b) Ductility (%EL and %RA) and Izod impact energy versus carbon
concentration for plain carbon steels having microstructures consisting of fine
pearlite.
185
that occur at the Fe3C phase boundaries. First, there is a large degree of
adherence between the two phases across a boundary. Therefore, the
strong and rigid cementite phase severely restricts deformation of the softer ferrite phase in the regions adjacent to the boundary; thus the cementite
may be said to reinforce the ferrite. The degree of this reinforcement is
substantially higher in fine pearlite because of the greater phase boundary
area per unit volume of material. In addition, phase boundaries serve as
barriers to dislocation motion in much the same way as grain boundaries
(see Chapter 1). For fine pearlite there are more boundaries through which
a dislocation must pass during plastic deformation. Thus, the greater reinforcement and restriction of dislocation motion in fine pearlite account for
its greater hardness and strength. Coarse pearlite is more ductile than fine
pearlite, as illustrated in Figure 3.14b. This behavior results from the
greater restriction to plastic deformation of the fine pearlite.
Figure 3.14 (a) Brinell and Rockwell hardness as a function of carbon concentration for plain carbon steels having fine and coarse pearlite as well as spheroidite
microstructures. (b) Ductility (%RA) as a function of carbon concentration for
plain carbon steels having fine and coarse pearlite as well as spheroidite microstructures.
186
Figure 3.15 The CCT diagram (solid lines) for a 1080 steel (the dashed lines
represent the compared line transformation for the TTT diagram).
Referring to heat treatment and CCT diagrams, main effect of alloying elements
that are added to steels: (a) provide solid-solution strengthening of ferrite, (b)
cause the precipitation of alloy carbides rather than that of Fe3C, and (c) improve hardenability. The term hardenability describes the ease with which
steels can form martensite. Plain carbon steels have low hardenability; only very
high cooling rates produce all martensite, despite medium carbon and alloy
steels, which have high hardenability (even cooling in air may produce martensite). This is visible if the CCT curves of low carbon and alloyed steels are
compared (Fig.3.17). The effect of alloying elements, including carbon, is to
shift CCT curves to the right and down in the temperature versus log time diagram.
187
Figure 3.17 CCT diagrams for: (a) red lines, a low-carbon steel; (b) a medium
carbon steel.
It is worth to notice hardenability does not refer to the hardness of the steel. A
low-carbon, high-alloy steel may easily form martensite but, because of the
low-carbon content, the martensite is not hard (Fig.3.18).
188
Figure 3.18 Relationship between hardness of quenched martensite and carbon content; the effect of presence of retained austenite and self-tempering is also considered.
Furthermore, in practical cases, the final hardness developed onto piece is related to how easily we can form martensite in a thick section of steel that is
quenched. While with a more hardenable steel we can get away with a relatively slow cooling rate and still form martensite, in case of low hardenability
steel it is possible that thicker sections cannot be transformed to martensite.
This is caused by the shifting of the cooling curves on thicker sections, as
shown by two different round sections illustrated in Fig.3.19.
Figure 3.19 Shifting of cooling curves for small and large diameters round bar
quenched from austenitizing temperature.
The problem of quenching thicker sections of low hardenability steel is therefore illustrated in Fig.3.20. As it is shown, the smaller radius section piece is
transformed into martensite and bainite structure, across the cross section. However, increasing the section radius causes shift of cooling curves - accordingly
with the Fig.3.19 - thus provoking the formation across the round section of
189
mixture of soft structure, as perlite and ferrite, together with hard martensite and
bainite.
Figure 3.20 Shifting of cooling curves for small and large diameters round bar
quenched causes development of soft structure (perlite and ferrite) mixed with
hard structure (martensite and bainite).
Residual Stresses and Cracking
Further complication we have in the practical case of Fig.3.20. When the austenite in the center later transforms, the hard surface is placed in tension, while
the center is compressed. If the residual stresses exceed the yield strength,
quench cracks form at the surface (Figure 3.21). To avoid this, we can first cool
to just above the Ms and hold until the temperature equalizes in the steel; subsequent quenching permits all of the steel to transform to martensite at about the
same time. This heat treatment is called marquenching or martempering (Figure
3.22). Note that as discussed presently, strictly speaking, the CCT diagrams
should be used to examine non-isothermal heat treatments.
190
Fig.3.23 - Schematic diagram of Jominy end-quench specimen (a) mounted during quenching and (b) after hardness testing from the quenched end along a
ground flat; (c) the hardenability curve.
191
192
(a)
(b)
Fig. 3.25 (a) simple model for non-diffusional transformation from the FCC phase to the Body-Centered Tetragonal (BCT) structure of martensite (diagonal
centered rotated structure) the white atoms are Fe, distinguished from black
spheres representing carbon interstitial position; (b) Photomicrograph showing the
martensitic microstructure.
193
Tempering martensite
In the as-quenched state, martensite, in addition to being very hard, is so
brittle that it cannot be used for most applications; also, any internal stresses that may have been introduced during quenching have a weakening effect. The ductility and toughness of martensite may be enhanced and these
internal stresses relieved by a heat treatment known as tempering. Tempering is accomplished by heating a martensitic steel to a temperature below
the eutectoid for a specified time period. Normally, tempering is carried
out at temperatures between 250C and 650C; internal stresses, however,
may be relieved at temperatures as low as 200C. This tempering heat
treatment allows, by diffusional processes, the formation of tempered martensite, according to the reaction:
Martensite (BCT, single phase)
bides)
where the single-phase BCT martensite, which is supersaturated with carbon, transforms into the tempered martensite, composed of the stable ferrite, cementite phases and carbides (in presence of alloying elements with
great affinity with carbon). The microstructure of tempered martensite
consists of extremely small and uniformly dispersed particles embedded
within a continuous ferrite matrix. The fine and hard particles reinforce the
ferrite matrix being barriers to dislocation motion during plastic deformation. On the other hand, a continuous ferrite phase guarantee adequate
toughness; the double phenomena, hard particles dispersing and Fe matrix formation accounts for the improvement of both two counteracting
properties, hardness and toughness, for an optimal balance. The size of the
precipitates influences the mechanical behavior of tempered martensite;
increasing the particle size decreases the ferritecementite phase boundary
area and, consequently, results in a softer and weaker material, yet one that
is tougher and more ductile. Furthermore, the tempering heat treatment determines the size of the cementite particles. Because carbon diffusion is
involved in the martensitetempered martensite transformation, increasing
the temperature will accelerate diffusion, the rate of cementite particle
growth, and, subsequently, the rate of softening. The dependence of tensile
and yield strength and ductility on tempering temperature for an alloy steel
is shown in Figure 3.27.
194
Fig. 3.27 - Tensile and yield strengths and ductility (%RA) (at room temperature)
versus tempering temperature for an oil-quenched alloy steel (type 4340).
Box Temper Embrittlement
The tempering of some steels may result in a reduction of toughness as measured by impact tests; this is termed temper embrittlement. The phenomenon occurs when the steel is tempered at a temperature above about 575C followed by
slow cooling to room temperature, or when tempering is carried out at between
approximately 375C and 575C. Steel alloyed with Cr are more susceptible to
temper embrittlement have been found to contain appreciable concentrations of
the alloying elements manganese, nickel, or chromium and, in addition, one or
more of antimony, phosphorus, arsenic, and tin as impurities in relatively low
concentrations. The presence of these alloying elements and impurities shifts
the ductile-to-brittle transition to significantly higher temperatures; the ambient
temperature thus lies below this transition in the brittle regime. It has been observed that crack propagation of these embrittled materials is intergranular; that
is, the fracture path is along the grain boundaries of the precursor austenite
phase. Furthermore, alloy and impurity elements have been found to segregate
preferentially in these regions. Temper embrittlement may be avoided by (1)
Molybdenum addition and/or (2) tempering above 575C or below 375C, followed by quenching to room temperature. Furthermore, the toughness of steels
195
Prediction of hardenability
There are large quantities of Jominy hardenability curves available in the
literature, such as the ones shown in Fig. 3.28.
The hardenability of a steel can also be predicted from its composition and
grain size. The ideal diameter (DI) is defined as the diameter of a steel bar
that will harden to 50 vol% martensite when quenched in an ideal quench
medium. The significance of this value is that a bar with a diameter larger
than DI cannot be hardened all the way through its cross section, even in
an infinitely rapid quench. Grossman originally developed a method for
calculating hardenability. The DI calculation starts from the calculation of
the DB, the base diameter. DB is influenced only by the carbon content
and the grain size 17, and it can be read from plot in Fig. 3.29. The DB takes
into accounts only two contributing factors for hardenability of steel, carbon content and grain size, while it considers steel quenched in an ideal
quenching mean.
Fig. 3.28 - Typical Jominy hardenability curves for medium-carbon steels austenitized at 845 C from initial normalized condition.
17 The hardenability increases with larger austenite grain sizes, because larger grain sizes means there are fewer nucleation sites for diffusion-controlled phase transformations
(that is, pearlite, ferrite, or bainite). When the danger of quench cracking is remote (i.e., no
abrupt changes in section thickness) and engineering considerations permit, it may sometimes appear to be more practical to use a coarse-grained steel than a fine-grained steel to
obtain hardenability. However, this is not recommended, because the use of coarser-grained
steels usually involves a serious sacrifice in notch toughness.
196
Fig. 3.29 - Plot of base diameter, DB, used for calculating ideal diameter.
However, each of the alloying elements in the steel contributes by a multiplying factor that increases the depth of hardening. Such multiplying factors are given in Table 3.3.
197
Now, let us consider a steel containing 0.45 wt% C with an ASTM grain
size of 6 and an alloy content of 0.5 wt% Mn, 0.2 wt% Si, and 0.35 wt%
Cr. By the use of the plot in Fig. 3.29, firstly we can calculate the base diameter (DB) of a steel with only 0.45 wt% C and a grain size of 6 ASTM,
obtaining DB= 6.5 mm.
Secondly, we can estimate the multiplying factors mf(element) from the Table
3.3, as it follows:
0.5 wt% Mn= 2.667 = mf(Mn)
0.2 wt% Si= 1.140 = mf(Si)
0.35 wt% Cr= 1.756 = mf(Cr)
By the following relationship, since we already know the DB and the multiplying factors, we can calculate the ideal diameter (DI) for the alloyed
steel:
DI=DB mf(Mn) mf(Si) mf(Cr) =
= (6.5mm) (2.667) (1.140) (1.756) = 34 mm
However, we know that the critical diameter DC depends on the quench
severity. The relationship between ideal diameter (DI) and critical diameter (DC) for various quench severities is given by empirical diagram as it is
shown in Fig.3.30. To use the plot, we need to know the H factor. This is
possible by the use of the Table 3.4: for example, if the steel were
quenched in moderately agitated oil, an H=0.37 can be determined from
Table 3.4.
198
Dc = critical diameter
DI = ideal diameter
For the sake of simplicity, in Fig. 3.31 is reproduced the same diagram of
Fig. 3.30, with all the curves removed except for the one we need for our
final calculation.
199
Fig. 3.31 Simplified plot of Fig.3.30 for determination of critical diameter (DC).
As shown, with the calculated DI, the plot in Fig.3.31 allows us to calculate the critical diameter DC for the specific quench severity factor
H=0.37: the DC is determined to be 7.5 mm.
As last analysis, the section size of bars with square cross sections and
plates can be compared with round bars, by the plot diagram in Fig. 3.32.
There is also an empirical relationship between the diameter of steel and
the Jominy distance at which there will be 50% martensite at the center, as
shown in Fig. 3.32.
(a)
(b)
Fig. 3.32 (a) Comparison of different geometries with the same cooling rate at
the center; (b) Relationship between ideal diameter and Jominy position.
200
Fig. 3.37 - Possible transformations involving the decomposition of austenite. Solid arrows, transformations involving diffusion; dashed arrow, diffusionless transformation.
201
(a)
(b)
Fig. 3.38 - Jominy test on a steel of: (a) low hardenability; (b) high hardenability.
Secondarily, it is important to notice that tempering treatment is often required to recover toughness. This is due to change of the high-strength and
high-brittle medium and high carbon BCT lattice into a tougher BCC ma-
202
trix with fine carbide particles here dispersed. Because of presence of such
particles, the -Fe BCC matrix is strengthened meanwhile it recovers
tough behavior (i.e. dislocations can move again, but small particles oppose to such movement, instead the previous situation of BCT lattice
where dislocations are unmovable).
Industrial quenching problems: distortions and quenching cracks
Quenching is generally considered as the most critical step of heat treatment processes conducted onto steel parts. It may lead to failures that are
generally distortions and, in most severe cases, cracks. Both distortion and
cracks are due to high stresses which develop in the part during quenching.
This occurs because massive parts are subjected to rapid cooling, and this
determine mismatching among the temperatures on top surface and subsurface layers, cross the part sections.
We have already discussed about the variation of cooling curves diagramwe call the cooling islands when two different diameters round bars
made of same steel are quenched in the same quenching mean. For the
reader convenience, here below is shown again the figure 3.20 with the
original caption.
Figure 3.20bis Shifting of cooling curves for small and large diameters round
bar quenched causes development of soft structure (perlite and ferrite) mixed with
hard structure (martensite and bainite).
203
The above figure illustrates one main problem due to limited hardenability
of steels. Hardenability, we remember, is the ability of steels to penetrate
martensite in depth, which depends on the fact cooling curves occurring
inside the part section can intercept the Ms transformation, instead of
crossing other curves representing the transformations of austenite in
something other structures softer than martensite. Actually, for the sake of
precision, also the case illustrated in the Fig.3.20bis is not the ideal case of
fully quenched section; quenching the two steel bars reported in the figure
actually produce martensite mixed to bainite (for 28mm diameter bar) and
small fraction of martensite mixed with very varying structures that range
from ferrite to perlite and bainite (for the 95mm diameter bar).
As discussed, to increase hardenability, we need to increase the alloying
elements in the chemical composition of steel. Alloying elements produces
the lowering of all the critical temperatures and a delay in austenite transformation is produced as a benefit for quenching. Shifting the CCT curves
toward the right, martensite can be obtained onto larger sections by
quenching. This is the basics of quenching operations.
However, one further beneficial effect obtained by alloying steel to increase the hardenability must be highlighted. Let us consider the simplified
scheme in figure 3.39.
A3
A1
Water quenched
Oil quenched
Ms
Mf
Log t
Figure 3.39 - Comparison of two possible quenching operations for a high hardenability steels. Both the water and oil quenching is possible, in terms of martensite
production in part section. The faster is the quenching operation, the larger is the
instantaneous differences between surface and inner section temperatures.
204
The CCT curves in the fig.3.39 show the steel may be quenched either in
water or in oil. Even with water quenching is less expensive, it is not industrially suitable for such a steel; this fact is due observing the instantaneous temperatures achieved in the part sections. In case of water quenching, the temperatures achieved at same time at surface and in the center of
a round section, for example, are very high if compared to the same bar
section when it is quenched in oil. The higher is instantaneous temperature
difference in part section, the higher is the differential volumetric expansion of steel. The volumetric expansion actually accounts for two concurrent contributes (depending on temperatures investigated): a) thermal expansion that is only function of the temperature and the coefficient of
thermal expansion; b) the volumetric expansion due to transformation of
FCC -phase into body-centered tetragonal (BCT) martensite.
In any case, the results is the same: as the sections of a part are interested
by too much temperature variations, for reason a) or b) or both, they are
subjected to differential expansion. If the differential expansion remains in
the elastic regime, the only result is temporary thermal induced stresses.
However this case is very rare, while more frequent is the case that plastic
deformations occur inside the part due to mismatching volumetric expansion. If the plastic deformation caused a beneficial drop of thermallyinduced stresses (the stresses induced by differential thermal expansion finally relax), such local plastic deformations are irreversible local volume
changes, that at least cause macroscopic part distortions (e.g. torsion,
bending, profile and hole ovalization, hole misalignments.
It is worth of noticing that such problems can occur also in heating phase,
especially for large parts; for high speed heating-up process, especially
large parts can suffer of temperature mismatching that can cause distortions. For this reason the temperature ramp-up rate shall be controlled and
equalization stages (Fig.3.40).
205
(a)
Austenitizing or quenching temperature
Temperature
Surface
Core
Time
Soaking
Equalization
(b)
Figure 3.40 - Schematic representation of heating during an austenitizing treatment: a) with final temperatures equalization stage; b) with final and intermediate
equalization.
206
tempered steel (V) will have lower toughness because of its lower strength
level. Toughness is measured in separate tests as impact toughness (e.g.
KV, J/cm2) or as fracture toughness ( KIC, MPam ). The lower the ductility of a material, the more brittle it is. Total brittleness accordingly denotes
zero ductility of the material.
207
quenched BCT structure the high-strength but brittle behavior (remind that
dislocations cannot move in a BCT structure); b) in presence of alloying
elements carbon can form fine and more dispersed carbides, namely fine
particles that are visible to dislocations and that can inhibit them in their
own movement.
Due to such temperature-time driven metallurgical mechanisms, tempering
of martensite would be preferentially conducted at high temperatures, to
speed up the treatment; o the other hand, tempering temperature for
quenching and tempering steels shall be conducted at temperature lower
than 680-700C. In fact, if tempering temperatures are too much close to
A1 temperature of the steel, they would vanish any quenching effects: at
temperature higher than optimized, carbides formed (especially for carbon
steels) would grow and finally coarse instead of fine carbides will be distributed in -matrix. The bigger are carbides, the lower is their capability
to counteract dislocation movement. At least, we could also fail the tempering object by accidentally overpassing the A1 temperature, so that the
microstructure will be turned into martensite mixed to soft structures 18.
This leads to tempering temperature for quenching and tempering steels
range from 600 to 680C to achieve the good balance of strength and
toughness.
The aim of the quenching and tempering process can also be explained by
means of the stressstrain diagram schematically shown in Fig. 3.41. As
hardened, a steel has high yield strength but low ductility, and a small area
below the stressstrain curve (curve 2) indicates low toughness. Ashardened and tempered (curve 3) steel has higher yield strength than in its
normalized condition but also much higher ductility than in its hardened
condition. The greatest area below the stressstrain curve indicates a substantial increase in toughness compared to either normalized or hardened
conditions.
For a certain steel grade, the relation between mechanical properties and
the tempering temperature can be read off from a diagram as shown in
Figure 3.42 for the steel DIN EN 20CrNiMo2 (0.15% C, 0.20% Si, 0.88%
Mn, 0.53% Cr, 0.50% Mo, 0.86% Ni).
208
209
It can be clearly seen from the lower part of the diagram in Fig.3.42 how
the impact toughness increases when the steel is tempered to a temperature
above 550C. Such diagrams enable precise optimization of the strength
level and toughness by selection of the proper tempering temperature.
Maximum toughness values are obtained when tempering a structure that,
after quenching, consists of fine-grained martensite. How different microstructures after different heat treatment processes influence the impact
toughness of 3.5% Ni steel at low temperatures is shown in Figure 3.43.
When testing the impact toughness at low temperatures, the so-called transition temperature (the temperature at which a substantial drop in impact
toughness begins) is of special interest. The lower is the transition temperature, the higher is the toughness.
Certainly, when quenching workpieces of big cross section, not only martensite is obtained, but also other constituents such as bainite, pearlite, and
even pre-eutectoid ferrite, depending on the decrease in cooling rate at
quenching, below the surface toward the core of the workpiece. So after
tempering, besides tempered martensite, other structural constituents having lower toughness are present. From a series of tests with hardened and
tempered steels with about 0.4% C, Figure 3.44 shows a general relation
between the structural constituents and the properties characterizing ductil-
210
Figure 3.44 - Elongation; (b) reduction of area; and (c) impact toughness of hardened and tempered steels having about 0.4% C, as a function of structure constituents and yield strength. F, Ferrite; P, pearlite; B, bainite; M, martensite. Grain
size: ASTM 67.
211
Figure 3.44 - Impact toughness as a function of tensile strength and carbon content
for the structure of tempered martensite. Grain size: ASTM 67.
The great influence of the microstructure achieved after hardening and before tempering on the impact toughness of a steel is evident from Fig. 3.45.
Appearance of pre-eutectoid ferrite or ferrite and pearlite in the structure
results in a substantial decrease in the impact toughness.
When selecting a structural steel for hardening and tempering, the expected microstructure must be considered.
Unalloyed steels for hardening and tempering, because of their low hardenability, exhibit a high degree of section sensitivity with respect to hardness distribution after hardening as shown in Figure 3.46.
When testing the impact toughness of a steel, one should be aware that
toughness is usually higher in the longitudinal direction (rolling direction) than in
the transverse direction. That is because some phases or nonmetallic inclusions
that are present in every steel (carbides, oxides, and sulfides) are stretched during
rolling in the longitudinal direction. In this way a textured structure originates that
has lower impact toughness in the transverse direction than in the longitudinal direction. As a measure of this effect, the factor of isotropy (the ratio of transverse
impact toughness to longitudinal impact toughness) is sometimes used.
19
212
213
Figure 3.46 - Hardness distribution (measured) on the cross section of bars of different diameters made of unalloyed steel (0.52% C, 0.24% Si, 0.90% Mn, 0.06%
Cr) quenched in conventional hardening oil from 860C.
214
Figure 3.48 Applicability of steel grades for hardening and tempering according
to required strength level and bar diameter. Steel designations according to EN
10083.
215
216
Pre-heat treatments
Machining operations
Machining
Machining operations
operations
(e.g hot forging)
Quench
Heat treatments
Temper (600C)
Finishing operations
Finishing operations
(e.g. surface finisihing,
hole drilling)
Figure 3.49 Manufacturing cycle for quenching and tempering steel part.
217
from austenitic grains occur in conditions where atoms can diffuse (Fe and
C predominantly). However, industrial solidification produces macroscopic and microscopic partitioning of chemical elements between parent liquid and the growing solid crystals. A generic scheme of such a problem is
shown in the Fig.3.54. As the grains of such segregated structures recrystallize during secondary hot forming processes (e.g. forging, rolling, etc.;
see Fig.3.50), micro-segregations align with material flowing imposed by
hot working; for example, they align with rolling direction or forging material flowing into dies, leading to appearance of fibers as shown in
Fig.3.55. Particularly, such fibers are macroscopically observable and they
are the result of aligned bands. They are the product of oriented microsegregations that are developed parallel to the main hot working material
flowing. Such bands are observed at microscopic scale by metallographic
analysis (conducted onto polished and acid pickled metallographic samples), as shown in Fig. 3.57.
Full annealing treatment - which means high temperature and prolonged
heating - usually is preferred when micro-segregations, responsible for
banding structures, are willing to be eliminated. Full annealing in fact, instead of normalizing, may produce C and Mn (see box below) re-diffusion
in order to reduce micro-segregation phenomena.
Figure 3.50 Scheme of steel making process, upstream and downstream process steps. The right side of figure shows the semifinished steel made products of larger size (billets, blooms, slabs) and lower size (bars, tubes, sheets, etc.).
Figure 3.56 Forging fibers produced in hot forged crankshaft, compared to longitudinal fibers cut by fully-machined part from round bar.
Fig. 3.57 - Banding in a UNI EN 18CrNiMo7-6 (normalized condition): a) acceptable banding for post heat-treatments; b) non acceptable banding phenomena
after normalizing, due to too much high segregation occurred in solidification process.
220
Fig.3.58 Control of quench distortions by additional material left before the quenching.
221
Figure 3.53The change in structure of a Cu-40% Ni alloy during equilibrium solidification. The nickel and copper atoms must diffuse during cooling in order to satisfy the
phase diagram and produce a uniform equilibrium structure.
On the opposite, if we consider the same simple system but we cool it too rapidly, atoms cannot fully diffuse and produce equilibrium conditions: nonequilibrium structures are therefore produced in the casting. Lets see what happens to our Cu-40% Ni alloy on rapid cooling and refer to Fig.3.54.
222
Figure 3.54 The change in structure of a Cu-40% Ni alloy during non-equilibrium solidification. Insufficient time for diffusion in the solid produces a segregated structure.
Notice the non-equilibrium solidus curve.
The first solid, containing 52% Ni, forms on reaching the liquidus temperature.
On cooling to 1260C, the tie line tells us that the liquid contains 34% Ni and
the solid that forms at that temperature contains 46% Ni. Since diffusion occurs
rapidly in liquids, we expect the tie line to predict the liquid composition accurately; however, diffusion in solids is comparatively slow. The first solid that
forms still has about 52% Ni, but the new solid contains only 46% Ni. We
might find that the average composition of the solid is 51% Ni. This gives a different nonequilibrium solidus than that given by the phase diagram. As solidification continues, the nonequilibrium solidus line continues to separate from the
equilibrium solidus. When the temperature reaches 1240C (the equilibrium solidus), a significant amount of liquid remains. The liquid will not completely solidify until we cool to 1180C, where the nonequilibrium solidus intersects the
original composition of 40% Ni. At that temperature, liquid containing 17% Ni
solidifies, giving solid containing 25% Ni. The last liquid to freeze therefore
contains 17% Ni, and the last solid to form contains 25% Ni. The average composition of the solid is 40% Ni, but the composition is not uniform. The actual
location of the nonequilibrium solidus line and the final nonequilibrium solidus
temperature depend on the cooling rate. Faster cooling rates cause greater departures from equilibrium.
223
224
UTS
Table 3.6 - Jominy test results and main mechanical properties for UNI EN
36NiCrMO16 round bars, various diameters.
As shown by Jominy curve, the hardness is kept at high value also at farend of Jominy test specimen, where normally quenching and tempering
steels exhibit a drop in hardness (at high distances from the quenched end,
the cooling curves cross the austenite-ferrite and austenite- perlite transformations areas on the CCT).
In such a particular case, the high percentage of alloying elements, particularly Ni, Cr and Mo (Ni ranges 3.63-4.27%; Cr ranges 1.55-2.05%; Mo
0.26-0.64) shifts the CCT curves to the right so that air-cooling curve cross
only the Ms and Mf.
225
It is worth noticing that the chemical composition of such a steel is identified in a specific range. Particularly, the sum of percentage of C, Ni and Cr
shall be limited in the range 5 (%C+%Cr+%Ni) 7. The lower limit is
necessary to allow chemical elements to reach minimum content to shift
CCT to the right in order to allow air-quenching. The upper limit is due to
not exceed the alloying elements to such values that would cause martensite transformation also by furnace cooling. In this latter case, in fact, it
would be not possible for such a steel to be heat treated in order to produce
ferrite and perlite, namely the machinable structures.
Comparing the manufacturing cycle for an air-quenched part with the cycle typically employed for a quenching and tempering steel part, only two
are the differences, but so relevant:
- Full annealing is the only one allowed preliminary heat treatment
to realize ferrite and perlite, the machinable structure; normalizing
is not possible since it is an air cooling treatment, that represents
quenching for such a steel;
- The tempering temperature is low, around 200C; this is possible
because as-quenched martensite structure is not brittle as for medium carbon quenching and tempering steel. This fact depends on:
a) a high percentage of Ni added in solid solution to this steel
(high percentage of Ni decrease the ductile-to-brittle temperature);
b) air-quenching determine slow cooling across the large sections,
thus it results in good homogenization of the instantaneous temperatures across the whole part section (i.e. low thermal induced
stresses). As direct consequence of b), the austenite to martensite
transformation starts almost at same time, thus reducing the stresses induced by differential expansion in the section.
226
Pre-heat treatments
Machining operations
Full annealing
Machining operations
Machining
operations
(e.g hot forging)
Quench
Heat treatments
Temper (200C)
Finishing operations
Finishing operations
(e.g. surface finisihing,
hole drilling)
227
Spring Steels
Spring steel is a low-alloy, medium-carbon steel or high-carbon steel with
a very high yield strength. This allows objects made of spring steel to return to their original shape despite significant bending or twisting. To increase yield strength, these steels are characterized by high carbon concentration that ranges 0.60-1.00% (e.g. UNI EN C60, C70 and C100). Due to
high carbon content they are non-weldable steels.
To increase toughness, the carbon content is lowered and the yield strength
increase is therefore pursued by including Si. Silicon has capability to
strengthen the -Fe lattice by solid solution strengthening mechanism,
thus it can raise the elastic limit of the steel and improve the resistance to
permanent set of springs.
The NiCrMo steel, CrMo steel with excellent hardenability were developed as steel for large-sized springs.
Table 3.7 shows the chemical compositions of spring steels designated or
recommended by several countries. Oil quenching and tempering can be
normal heat-treatments, while water quenching can be suitable for low
carbon steels whose the carbon content is limited to 0.5% (e.g. UNI EN
C55), because of risks of quenching cracks.
CrV steels have good hardenability with high toughness, and is used for
higher hardness application.
The SiCr steel is mostly used for oil tempered wire material of coldformed springs. CrMo steel with higher hardenability than the boron added.
228
229
The general manufacturing cycle of steel made heat treated parts (i.e. preheat treatments, machining operations, heat treatments, finishing operations, refer to previous section Quenching and tempering steels) is adjusted for cold and hot worked as illustrated in the fig.3.63.
230
Pre-heat treatments
Cut to lenght
Cut to lenght
Machining operations
Pickling
(Cold) Coil
(Hot) Coil
Patenting
End-forming
Stress relieve
Quench
Heat treatments
Temper (450C)
Finishing operations
Shot peen
Shot peen
Cold Wound
Spring
Hot Wound
Spring
231
Specifically regarding with heat treatments performed, most typical preliminary heat treatment is spheroidizing annealing (sometimes, full annealing and normalizing can be carried out). The spheroidizing annealing is
keeping the materials around the A1 transformation temperature for a certain time and cooling to the A1 transformation temperature by the cooling
speed of 1020C/hr. The purpose is to have the cementite precipitates
changed from the lamellar or net shape to spherical shape. Since this heattreatment can soften the material by a great extent, this can be applied
when the formability of materials is required.
As shown in the scheme of Fig.3.63, quenching and tempering treatment is
conducted to improve the spring characteristics. During quenching, material is heated to the temperature about 3050C higher than the A3 transformation temperature, and immersed into water or oil (sometimes air
cooling in high alloy steel) and cooled rapidly, to have martensite structure. Since the solution content of carbon is high, there are many lattice defects (dislocation) and the crystal structure is fine, the hardness of the martensite of medium-high carbon steel is as high as 6065 HRC. The spring
steel has a high yield point; besides suitable high ductility is required,
therefore, it is tempered at about 450C.
As shown in the fig.3.63 the pre-heating and the quenching and tempering
treatments for the spring steels are only two steps of a more articulated
manufacturing process that requires some further steps, mostly patenting,
pickling and drawing as they are shortly illustrated in the next.
Patenting
Patenting is heat treatment process eventually conducted to restore ductility resources of materials that could be highly work hardened during cold
coiling steps (work hardening causes ductility decrease that could lead to
cracks during cold coiling steps). Most frequently it is conducted on highcarbon steel wires. By patenting, they are continuously transformed to microstructure of fine pearlite, either by isothermal cooling or continuous
cooling. In practice, wire rod, traveling continuously, is first held above
the A3 point, and subsequently cooled down below the A1 point to be
transformed to pearlite. Figure 3.64 shows a schematic of TTT diagrams of
this heat treatment. Several kinds of cooling media, used in the patenting
for the transformation of wire.
232
Pickling
Pickling, either by hydrochloric acid or sulfuric acid solution, is generally
provided on patented wire and hot-rolled rod to remove their surface scale
(ferrous oxide film).
Shot peening
It is well known (refer to Chapter 2) that most fatigue failures initiate at
the surface as a crack and propagate under cycling until failure occurs. For
these cracks to initiate, the surface must be in tensile stress and they will
not propagate through a compressively stressed surface. If a residual compressive stress is produced in the surface, the tensile stress created by the
applied load must first overcome the residual compressive stress before the
resultant surface stress becomes tensile.
Shot peening is the most effective and inexpensive method to produce the
desired residual compressive stress at the surface of components. This involves bombarding the surface with rounded particles propelled at high velocity (fig. 3.51a), 60 m/sec. Each shot particle, with controlled shape and
of uniform size, acts as a peening hammer to cold work the surface and
produce compressive stresses (fig.3.51b and fig.3.52).
233
(a)
(b)
Fig. 3.51 Shot peening: a) process; b) scheme of effect on peened surface.
The commercial origins of the technique go back to the automobile industry of about 1930, and today almost all automobiles use shotpeened valve
springs, and, in many cases, suspension springs. The peening process is
relatively inexpensive and has proven capable of increasing operating life
by five to 10 times or more when compared to unpeened springs
(Fig.3.53).
234
Fig. 3.53 Shot peening applied to surfaces exposed to high bending loads, such as
carburized and hardened gear teeth, increased the life considerably. For example,
the life of teeth subjected to loading of 80,000 psi (about 500 MPa) increased from
270,000 cycles before shot peening to 3 million cycles after shot peening.
235
Criteria selection
As the increasing in alloying element content is key cost-driver, carbon
steel are used when reduced section have to be realized and toughness is
not crucial. Carbon steel with medium and high carbon content are particularly indicated for cold-worked springs.
For hot worked large section size springs, Si and Si-Cr alloyed steel are
used (increased YS and hardenability of steel).
In case high toughness value are targeted, e.g. safe component for transportation, Ni- alloyed spring steels are used.
Surface hardening of mechanical components
There are some applications where it is necessary to have a hard surface
for certain depth below surface but a tough, shock-resistant inner core.
This is the case, for example, of such components that can be subjected to
contact fatigue (refer to Chapter 2) as for example, cams, gears, and shafts.
Such components require hard layers on surfaces to resist but tough inner
cores to resist shock.
As known it is not possible to match two opposite requirements for metal,
namely having high hardness and contemporarily high toughness. In fact,
while a low-carbon steel containing 0.1 wt% C will have a very tough
core, its surface hardness will be low also after quenching. On the other
hand, a euctectoidic steel containing 0.8 wt% C will have a very high surface hardness after quenching, but the core will not be tough and shock resistant.
Thus, to satisfy design requirements there are two approaches to this problem. One is to use a medium-carbon steel and only quench harden the surface through local heat treatment. The other approach is to diffuse some
elements, namely locally adding chemical elements onto surface of a base
steel, so to strengthen the top layer surfaces acting on various metallurgy
strengthening mechanisms we will discuss in depth in the next.
Despite two approaches pursue same objective, namely increase hardness
on surface but keeping tough subsurface, they distinguish since the first
method, called surface hardening of steels, does not change the chemical
composition of steels, while the second approach needs to change local
chemical composition, in order to strengthen lattice. Changes in chemical
analysis on top layers are produced by feeding elements like carbon C and
nitrogen N to steel by a temperature-time process i.e. thermochemical
treatment - that allows to increase diffusion on steel surface. If carbon is
236
237
Induction Hardening. In induction hardening, heat is supplied by surrounding the part with an inductor coil carrying a high-frequency current in the
range of 2 to 500 kHz. Higher frequencies result in a shallower depth of
heating and are therefore used for smaller-diameter workpieces. The coil
acts like the primary winding of a transformer. The oscillating field produced by the induction coil induces magnetic field electrical eddy currents
in the steel within a certain depth of the outer surface, called the skin depth
(Fig.3.53), which decreases as the frequency is increased. The eddy currents produce Joule resistance heating in the skin depth that rapidly raises
the surface temperature. The amount of heat released is proportional to the
square of the current such that QI2Rt also called the Joule heating
(obeying to Joules first law of conductive mean heating).
238
Fig. 3.54 Induction hardening onto a shaft. Coil and water jet system are synchronized in their relative movement with the heated shaft.
Laser hardening. Laser heat treating is a highly selective hardening technique in which a spatially well-defined beam of laser light is absorbed near
the surface, causing rapid heating (heating is limited to the illuminated area) and penetration into the bulk material is limited. Laser hardening allows for a highly defined zone of influence without affecting neighboring
surfaces. High cooling rates make fine structures and high levels of hardness possible. Intricate contours are easily hardened using lasers due to the
flexible beam guidance possibilities. The maximum thickness obtainable
with the laser hardening is 3 mm for steels. In the case of objects of high
precision, limiting the layer thickness to 0.2 mm positively result in controlling local distortions since they are kept so low as they do not require
any post- processing treatment.
Manufacturing cycle for surface hardened product
In the following figure, Fig.3.55, it is shown the manufacturing cycle for a
surface hardened product.
239
Pre-heat treatments
Machining operations
Machining operations
Machining
operations
(e.g hot forging)
Quench
Heat treatments
Temper (600C)
Finishing operations
Finishing operations
(e.g. surface finisihing,
hole drilling)
Surface quenching
Stress relief (optional)
240
which are mostly used are: UNI EN 45Cr2, UNI EN 38 Cr4, UNI EN 37
CrMn4, UNI EN 41CrMo4 and UNI EN 40NiCrMo3, listed by increasing
their own hardenability and their cost. It is important to notice that, when
rapid cooling is guaranteed after heating up the parts, so rapid as critical
interval of temperatures for the tempering brittleness (450-550C) are rapidly passed, Mo can be eliminated from steels for surface quenching, in
order to reduce costs.
Carburizing steels
Carburizing is conducted by heating a low-carbon steel into the singlephase austenitic field, generally around 900C, where the steel has a high
solubility for carbon. Carbon is then absorbed onto surface layers, which
increase progressively carbon content and thus hardened during the following quenching phase, as shown in the scheme of generic carburizing cycle
in Fig. 3.55.
Carbon adsorption
860C
Ac3 (surface )
Carburizing time
Quenching
time
Tempering
time
After quenching, the part is extracted from oil chamber and moved to low
temperature furnace where it is tempered at low temperature, around
180C, in order to relieve stressed but avoiding carbon to diffuse out from
martensite tetragonal lattice, thus vanishing the carburizing and quenching
results in terms of high hardness.
The final result of carburizing process is therefore hard high-carbon case
on top of a tough low-carbon steel core.
241
(eq.3.2)
Where:
C(x,t) is the carbon concentration at certain depth x and fixed time t;
C0 is the initial condition, namely the carbon content before carburizing starts (i.e. the carbon content of base core material);
Cp is the carbon potential of atmosphere, namely the measure of the
concentration of carbon atoms in carburizing outer atmosphere 21,
While carburizing can be done in a solid, liquid, or gaseous medium, gas carburizing processes are mostly industrially used, since they can be highly automated and controlled and for series production (here in the following we refer to this
type of carburizing process). Gas carburizing uses either natural gas, propane, or
butane as carrier gases to realize the carbon potential in atmosphere, accordingly
with following relations:
2CH4 + O2 + 4N2 2CO + 4H2+ 4N2
CO + H2 C + H2O
Fe + CH4 Fe (C) + 2H2
Fe + CO + H2 Fe (C) + H2O
21
242
As shown by the scheme in Fig. 3.56, during the carburizing phase, carbon
diffuses slowly into the bulk of the part, and a carbon concentration gradient below the surface is established. Possible carbon profiles are shown as
examples in Fig.3.57. As the 3 carbon profiles illustrates, within the steel
part, the high-carbon surface has a higher carbon potential than the lowcarbon interior, since carbon tends to diffuse from the surface toward the
center 22.
C atoms
Fe- lattice
243
(b)
Figure 3.57 a) Scheme of carburizing process, where carbon atoms diffuses from
high-carbon concentration atmosphere (i.e. high carbon potential mean) into austenitic lattice; b) 3 different carbon profiles obtained by varying: carbon potential
of external atmosphere (i.e. 0.50%, 0,75% and 1.10%), exposure times (i.e. 2h, 4h,
8h, 16h).
244
3. Once a high carbon content - approximately the euctectoidic content - is achieved on top layers, the part is oil-quenched. The
quenching phase has the scope to produce high-carbon content and
hard martensite on top layers (i.e. case hardened), while maintaining low-carbon content and tough martensite within the part (i.e.
core) - at least, mixed martensite-bainite is produced at core, depending on hardenability of steel and the part size.
4. The final stage consists in reheating quenched parts for tempering
treatment, which shall be conducted at low temperatures (usually
around 150-180C). Tempering is required to reduce residual
stresses develop onto surface layers by quenching.
Design aspects of carburizing
Typical case depths range from few millimeters (e.g. around 1-2 mm) to
3-4mm. in the latter case prolonged carburizing times are required. Since
carbon content on top layers is targeted around 0.8-1.1 wt.%, and because
of the low tempering temperature, surface hardness usually obtained is
about 850-900 HV (about 64-63 HRC). Effective case depth (see Fig.3.57)
is function of both carbon profile and hardenability of base steel.
HV
550 HV
Core hardness
Effective depth
(mm)
Figure 3.57 Typical hardness profile for carburized part and definition of effective depth.
245
Pre-heat treatments
Machining operations
Machining operations
(e.g hot forging)
Carburizing (900C)
Heat treatments
Quenching (860C)
Tempering ( 180C)
Finishing operations
Finishing operations
(e.g. surface finishing,
hole drilling)
246
simple as possible. As size and shape complexity increase, residual stresses produced onto part by rapid cooling might increase the risks of nonrestorable distortions and, at least, cracks (such risks are also increased because of the low mechanical strength of low carbon austenite).
A first conclusion is that, as the part size increases, oil quenching is required, and alloyed steels are therefore used. Considering toughness as major feature together with cost reduction, the UNI EN 16MnCr5 is one possible solution; because of low carbon content, such a steel is easy to hotwork and has good machinability (presence of higher content of Mn).
Addition of Cr increases hardenability, is compared to carburizing carbon
steel, but martensite is limited to low-medium thickness (not more than 20
mm). Since such a steel can suffer of temper embrittlement, this problem is
solved by using the UNI EN 18CrMo 4, which has similar hardenability of
UNI EN 16MnCr5.
Comparable hardenability, but higher toughness, is achieved by the UNI
EN 16NiCr4, due to presence of Ni. Costs in this case increases.
For automotive applications most frequently are used steels with higher
percentage of carbon, alloyed with Cr for increase hardenability (up to 6070mm), and Ni for increasing toughness, like the UNI EN 20NiCrMo2.
This steel has high carbon content (it targets 0.2% wt. limit for carburizing
steels) that increases core and surface strength, while its moderate-high
hardenability allow it is used for complicate shape.
Finally, the UNI EN 18 NiCrMo 5 and UNI EN 18 NiCrMo 7 series are
most used carburizing steel. Despite higher raw material costs, they are
considered safe materials to oil-quench due to high toughness (i.e. high Ni
content) and high hardenability with good machinability.
It is important to highlight that the low-temperature tempering stage limits
the service temperature below the tempering temperature. If higher, service temperature would cause decay in surface hardness due to carbon diffusion from BCT lattice, and its partial transformation into -Fe, the ductile BCC lattice
Nitriding steels
Nitriding process has same scope of carburizing process: develop onto
tough base materials hard surface layers, to certain depth. Similarly, nitriding pursues such an objective by locally changing chemical composition of
the base steel, as it is usual for thermochemical processes. In the specific
case, the chemical specie which is absorbed onto surface is nitrogen N, a
small interstitial atom, too, capable to diffuse in iron lattice. As the N is
absorbed and it diffuses within the part, it combines with Fe to form small
247
particles we call nitrides. Such particles are randomly dispersed into matrix
and they are capable to intercept dislocation, providing a precipitation
hardening strengthening effect. This is the strengthening mechanism by
which nitriding process increase hardness, and mechanical strength, on
surfaces.
By this fact, we can immediately recognize great differences in terms of
metallurgical strengthening phenomena existing between carburizing and
nitriding. While carburizing process strengthens the carbon-enriched surface layers by exploiting lattice transformation, from austenite to martensite BCT lattice, by quenching, nitriding process strengthens the preexisting -Fe (BCC lattice) by precipitates. To make reader clear such
great difference, it is worth mentioning that nitriding is conducted at
around 500C, for one reason that will be explained soon in the next.
By this fact, instead carburizing which requires that the steel be quenched
and then tempered, nitriding is done at temperatures largely below the austenitization temperature. Some direct benefits of this fact are:
a) the distortion inherent in the martensitic transformation during hardening is completely removed; thus, nitriding allows excellent dimensional
control;
b) since the nitriding temperature is below the tempering temperature of
quenched and tempered steels, nitriding can be conducted as final stage to
increase hardness onto very optimized structures that are typical of the
quenched and tempered steels.
Similar to carburizing, nitriding can be done in several different media;
here in the following we refer to industrial gas carburizing. In gas carburizing, nitriding is conducted in atmospheres that decompose ammonia NH3
to provide nitrogen to the surface by catalytic reaction:
(eq.3.2)
NH3
N+3H
The nitrogen is therefore absorbed and it diffuses into the steel. At top surface atomic nitrogen combines with the iron to form iron nitrides Fe2N and
Fe4N types. These phases are clearly visible in the iron-nitrogen phase diagram shown in Fig.3.59. Firstly, by this diagram it is possible to explain
why nitriding temperature must not exceed around 500C. This limitation
is required to nitride steel with safe margin to keep steel below the euctectoidic temperature that is set at 590C. If the euctectoidic temperature is
overpassed, the structure turns into Fe- and the following cooling would
proceed by the euctectoidic transformation from Fe- into lammellar
248
euctectoidic structure, called braunite, that is too much brittle to be accepted onto mechanical parts.
Phases:
N(Fe-): Nitrogen in solid solution of Iron
(Fe4N): lattice structure CFCnitrogen concentrdion max. 6%
(Fe2N): lattice structure EC, high nitrogen concentration over 10%
Figure 3.59 Iron-Nitrogen phase diagram. They are highlighted by circles the
two high nitrogen content phases that form onto surface, the ' (i.e. Fe4N) and the
(Fe2N) phases.
POUROS ZONE
COMPACT ZONE
DIFFUSION LAYER
249
The metallurgical phenomena involved during surface compounds formation are below summarized:
Nitrogen absorbed onto surface form compound layers, nitride
(Fe2N) and nitride (Fe4N) with minor presence of complex nitrides if alloying elements when alloying element with great affinity
with nitrogen are present (mainly Ti, Al, V, Cr, Mo).
The continuous decomposition of (Fe4N) at interface with diffusion zone feed N for diffusion into solid solution Fe- and formation of complex nitrides, when alloying element with great affinity with nitrogen are present (mainly Ti, Al, V, Cr, Mo).
It is clear therefore that the presence of alloying elements with great affinity with nitrogen allow formation of complex and stable complex nitrides;
these nitrides are major responsible for increasing strength and hardness on
top surface layers interested by nitrogen diffusion.
Design features of nitriding parts
Due to low treatment temperature, the diffusion of nitrogen is low. This
fact limit in practice nitriding case depths, that are thin if compared to carburizing. Usually case depth is less than 0.80mm, even though nitriding
times can exceed 100 h.
On the contrary, the surface hardness can achieve very high values, ranging from 900HV to 1200 HV, the latter when Al is added (aluminum forms
very fine AlN precipitates). Molybdenum, in addition to its contribution as
a nitride former.
During single stage nitriding treatments, in which a single nitriding atmospheric composition is maintained, a white layer of iron nitride is formed
(i.e. the porous layer shown in Fig.3.60).
This iron nitride layer is hard but can crack and spall. When this is unacceptable, the layer is removed by surface grinding. Alternatively, a twostage nitriding process can be used in which, after the first stage, the atmospheric conditions are changed so that iron nitride no longer forms at
the surface, and the existing layer is removed as the nitrogen dissolves into
the steel (Fig.3.61).
250
Figure 3.61 Effect of one- and two-stage nitriding on white layer. (a) Single
stage, gas nitrided for 24 h at 525C. (b) Double stage, gas nitrided for 5 h at
525C followed by second stage at 565C for 24 h.
251
600C, the nitriding process can be conducted as final stage, onto finished
parts.
Pre-heat treatments
Machining operations
Machining operations
Machining
operations
(e.g hot forging)
Quench
Heat treatments
treatments
Finishing operations
Temper (600C)
Finishing operations
(e.g. surface finisihing,
hole drilling)
Nitriding (500C)
252
253
254
Table 4.1. Main Groups of Tool Steels and AISI Letter Symbols.
255
256
257
258
Tool steels that are included into the nine groups by Table 4.1, are in the
following classified into 4 main categories of specific applications. Each
application requires specific properties to be enhanced.
1. Steels for plastic molds. During the formation of plastics, dies are subjected to heat and pressure. The temperature of the dies is as high as 250C
and the strength of about 100 MPa. In this case, hardness retention and
strength requirements are of minor importance, instead high machinability
properties and a low degree of distortion in hardening of plastic molds are
very important. The P20 steel (0.30% C, 1.70% Cr, 0.40% Mo, Table 4.1)
is a good choice for molds due to its low degree of distortion and good
machinability. In the case of abrasive plastics, the molds are made of steels
oil-hardening cold working steel (see Table 4.1) as O1 (0.90% C, 1.0
1.4% Mn, 0.50% Cr, 0.50% Ni, 0.50% W) that can be hardened to about
57-61 HRC thanks to a low tempering (between 100C and 350C). The
low tempering temperature defines the low temperature for such steels that
are suitable for tools used to cut or form materials that are at low temperatures.
2. Steels for high-pressure die casting molds. In die casting, tools are heated to about 500C and are subjected to high mechanical forces and erosion.
Most important property in this case is hardness retention at high temperature. Steels that exhibit secondary hardening are required, low carbon and
moderate to high alloy that provide good hot hardness and toughness and
fair wear resistance due to a substantial amount of carbide (class 4 of
fig.4.1). Dies used for light metal casting are therefore commonly made of
H-type steels, as H11 (0.33% C; 4.75%Cr; 0.3% Ni; 1.10% Mo; 0.3% V).
259
260
261
262
Mar-Aging Steels
Maraging steels are a class of high-strength steels with very low carbon
contents (0.030 wt% maximum) and additions of substitutional alloying
elements that produce age hardening of iron-nickel martensites. The term
maraging was derived from the combination of the words martensite and
age hardening.
Maraging steels have high hardenability and high strength combined with
high toughness. The maraging steels have a nominal composition by
weight of 18% Ni, 7 to 9% Co, 3 to 5% Mo, less than 1% Ti, and very low
carbon contents.
The commercial maraging steels, 18Ni(200), 18Ni(250), 18Ni(300), and
18Ni(350), have nominal yield strengths after heat treatment of 1380,
1725, 2070 and 2415 MPa (200, 250, 300, and 350 ksi), respectively.
The name derives from the combination of nominal content of Ni and the
average UTS in ksi. Typical properties of maraging steels are shown in
Table 4.3
263
In these Fe-alloys, carbon is considered an impurity and kept to as low a level as possible to minimize the formation of titanium carbide (TiC), which can adversely affect ductility and toughness.
25
264
Fig.4.5 - Fine grain size in; a) common plain carbon steel; b) high-strength lowalloy (HSLA) steel.
265
These fine precipitates pin the austenite grain boundaries, hindering grain
growth. At still lower rolling temperatures, they inhibit recrystallization of
the severely deformed austenite grains. The elongated and pancaked grains
can then rapidly transform to fine ferrite. These fine precipitates also provide additional locations for ferrite nuclei to form during cooling, resulting
in an even finer ferrite grain size. In addition, as shown in Fig. 4.6, nuclei
locations are also formed within the austenite grains at locations of deformed shear bands.
High-strength low-alloy steels are primarily hot rolled into the usual
wrought product forms (sheet, strip, bar, plate, and structural sections) and
are commonly furnished in the hot rolled condition. In addition to hot
rolled products, HSLA steels are also furnished as cold rolled sheet and
forgings.
Since HSLA steels do not contain a large percentage of alloying elements,
they can be competitively priced against plain carbon steels.
The HSLA steels are extensively used as structural beams for bridge construction, off-shore oil and natural gas platforms, ship hull and deck plate,
and electrical transmission towers and poles. In automobiles, HSLA steels
are used for safety applications such as ultrahigh-strength impact door
beams and energy-absorbing bumper assemblies and for increasing fuel
economy through thinner and lighter-weight chassis sections (Fig.4.7).
266
42CrMo4
Forging
LMUmachining
Quenching
and
tempering
LMUfinishing
Nitriding
25MnCrSiVB6
Forging
LMUmachining
267
Dual-Phase Steels
A dual-phase steel is one that consists of islands of hard martensite embedded in a tougher continuous ferrite matrix. This mixture of fine ferrite
and austenite grains is produced by heating into the two-phase a+c field,
followed by quenching to convert the austenite to martensite (Fig.4.9). The
microstructural constituents that can be present in dual-phase steels after
processing are therefore a fine grained ferrite matrix, some proeutectoid
ferrite that formed from the austenite during cooling, and martensite or
lower bainite, depending on the alloy content and the Ms temperature.
268
Fig.4.9 Stress-strain curves for plain carbon, high strength low-alloy (HSLA),
and dual-phase steels.
To be successful, the hat treatment of these steels requires suitable continuous cooling transformation characteristics and controlled cooling rates, as
shown in the scheme of Fig.4.10.
269
TRIP Steels
The term TRIP is derived from the mechanism of transformation-induced
plasticity. The nominal composition by weight of these steels is 0.25% C,
2% Mn, 2% Si, 10% Cr, 9% Ni, and 5% Mo. They contain a high percentage of retained austenite (10 to 15%). The austenite transforms to martensite during forming of the part, thus providing enhanced formability, or it
transforms on impact, as in an automotive crash.
Mechanical properties, heat treatments and applications
Similar to dual-phase steels, TRIP steels are annealed in the intercritical
region, but instead of direct cooling to room temperature to form martensite, they are isothermally treated. The isothermal hold during heat treatment is specifically designed to produce large, dispersed volume fractions
of retained austenite in the ferrite matrix after intercritical annealing. Thus
products of transformation therefore include bainite and retained austenite.
The scope of dispersed retained austenite is comprehensible when these
steels are cold worked or plastically deformed. During cold plastic deformation, the unstable retained austenite transforms to martensite, hence the
term transformation-induced plasticity.
The strain induced martensitic transformation increases strain hardening;
furthermore, it helps to delay necking instability during forming operations.
Due to dynamic increase of strength during impact, similarly to DualPhase steels they are suitable for safety components to absorb impact energy in automotive crash.
270
271
Figure 5.1 - Schematic of the evolution of stainless steels starting from plain carbon steels: the stainless steel tree structure.
272
273
In table 5.1 some typical ferritic steels are The relative designation in accordance with the standard EN 10088 refers to alloyed steel type, with X
followed by carbon content in percentage and other alloying elements, e.g.
Cr, with their own actual weight percentage. The last column provides the
most appropriate AISI designation for such steels.
27 The phase diagram shows that for some compositions with Cr below 12%wt., solid
solutions can be transformed into austenite by heating them into the gamma loop, and,
when cooled, austenite transforms into ferrite. As shown in the following, this transformation forms the basis for the heat treatable martensitic stainless steels.
274
Table 5.1 - Chemical composition of some of the main types of ferritic stainless
steels from EN 10088 and AISI designations.
Figure 5.2 illustrates the metallurgic logic that led to the creation of the
main types of ferritic stainless steels present on the market, starting from
type X6Cr17. The first family, to which the X6Cr17 belongs, has a chromium content ranging from 15.5% to 18%: this family is the most numerous in terms of the quantity of alloy types present on the market. There are
a further two families, one characterized by chromium content ranging
from 11.5% to 14.5% (ferritic stainless steels with low chromium content)
and the other with a chromium content of more than 18% (known also as
superferritic stainless steels).
In all three of the above families - with low, medium and high carbon
chromium content - other alloy elements may also be present: these include aluminum and silica, aimed at stabilizing the ferritic structure and
improving resistance to oxidation under heat, molybdenum, necessary to
improve corrosion resistance, titanium and niobium, to prevent precipitation of the harmful chromium carbides, as below described.
275
Figure 5.2 - Schematic of the main types of ferritic stainless steels, starting from
X6Cr17 (similar to AISI 430), progenitor of this family.
Box The intergranular corrosion induced by carbide precipitations
A metallurgical problem typical in stainless steels, including ferritic, is related
to the precipitation of chromium carbides, type Cr23C6. This occurs due to the
high level of affinity between the chromium and carbon. Chromium carbides
usually deposit on the grain boundary and leading to a localized, i.e. within the
boundary of the grain itself, depletion of chromium: a consequence of this process is that the chemical composition of the metal mass falls below the passivation threshold (10.5% chromium) giving rise to intergranular corrosion, also
in only slightly aggressive environments. The scheme of such a phenomenon is
shown in Fig. 5.3. The entity of corrosion depends both on the quantity of carbide precipitation and the lesser or greater ability of the chromium to spread in
the steel lattice: in the crystalline structure of the steel, this causes short or long
range depletion of the chromium with differing effects in terms of material deterioration. As the solubility of carbon in the body-centered cubic lattice is significantly limited, the precipitation of chromium carbides in ferritic stainless steels
is practically impossible to eliminate. On the other hand, due to the high diffusivity of chromium in the lattice of phase , the chromium gradients in the vicinity of the ferritic grain boundary are much less pronounced with respect to
what occurs in austenitic stainless steels (refer to Fig.5.3a): as a consequence
the problems of intergranular corrosion in ferritic stainless steels, under the ef-
276
fect of carbide precipitation on the grain boundary, are less marked with respect
to the same phenomena in austenitic stainless steels.
With the exception of the case in which high mechanical properties are required
(such as the case of martensitic stainless steels), this problem is avoided by minimizing the carbon content or attempting to neutralize the effect of the latter, by
combining it with other chemical elements, such as titanium and niobium. Also,
as explained further in the section below, the semi-finished products should also
undergo a complete annealing heat treatment before being used.
Heat treatments
The only heat treatment possible for the family of ferritic stainless steels is
complete annealing or simply annealing: its role is to optimize corrosion
resistance of the steel, aiding a uniform distribution of the chromium in the
crystalline structure of the semi-finished part. Great care is taken in selecting the temperature and hold time, as this family of materials is particularly sensitive to the phenomena of crystalline grain enlargement. Annealing
277
278
(a)
(b)
Figure 5.5 a) Effect of chromium content on transition curves of iron-chromium
alloys (the carbon content is ~ 0.01); b) Effect of thickness of semi-finished part
on the transition temperature of two ferritic stainless steels.
Main applications
Steel grade X6Cr17 (similar to AISI 430) is used for machine parts in the
industry for the production and transformation of nitric acid (tanks, condensers, pipelines, coils, etc.); it is also widely used for cracking and reforming plants in the oil sector, for the production of low cost cutlery,
household applications, decorative elements and furnishing accessories for
interiors, coverings for bar counters and tables, and kitchen extractor
hoods. It is also used in a number of applications in the cheese/milk dairy
sector and for manufacture of train carriages and buses. Ferritic stainless
steel grade X6Cr17 is also used in components operating at temperatures
up to 750-800C due to its optimal resistance to oxidation under heat. In
more aggressive environments, the ferritic stainless steel X6CrMo17-1
(similar to AISI 434) can be used, in which the presence of molybdenum
279
improves resistance to corrosion due to pitting; as well as for exhaust systems in the automotive sector, these steels are also used for internal cladding of buildings, on escalators, lifts and conveyor belts, for drinking water pipelines or as supports for photovoltaic cells.
280
To limit carbide formation and detrimental effect on intergranular corrosion (refer to Box The intergranular corrosion induced by carbide precipitations) the carbon content is fixed to very low values, usually it ranges
0.02-0.06%. The combination of chromium and nickel in the alloy obtains
a stainless steel with an austenitic structure, i.e. with a face-centered cubic
lattice, highly resistant to atmospheric corrosion and water-based solutions 28.
In austenitic stainless steels, added corrosion resistance can be achieved by
increasing the nickel content up to 11-12% or adding molybdenum to values of 2-3%. The latter action leads to another widely used stainless steel,
i.e. the X5CrNiMo17-12-2 (similar to AISI 316), progenitor of the second
family of austenitic stainless steels, i.e. chromium-nickel molybdenum 29.
Finally, a further possibility to obtain an austenitic structure in a stainless
steel consists in replacing the nickel part with manganese 30, the element
which otherwise would only act as a deoxidant in the alloy. This solution
was developed in the 1950s, with the aim of limiting production costs; the
progenitor of the family of austenitic stainless steels with chromiummanganese-nickel is the X12CrMnNiN17-7-5 (similar to AISI 201).
Figure 5.6 illustrates the metallurgical logic, which led to the creation of
two main types of austenitic stainless steels available on the market: one is
chromium-nickel stainless steels such as X5CrNi18-10 and the other is the
chromium-nickel-molybdenum such as X5CrNi17-12-2 and their respective by-products.
29
The austenitic chromium-nickel-molybdenum stainless steel X5CrNiMo17-12-2 (similar to AISI 316) is suitable for applications in contact with sea water (Cl- 20,000ppm =
2%) or with water-based solutions strongly contaminated by chlorides.
30 The alternative to manganese with respect to nickel is based on the austenitizing effect that both elements show if added to Fe-C alloys: as the austenitizing effect of the manganese is equal to half that of nickel, around 2% of Mn needs to be added for each 1% of
replaced Ni.
281
Fig. 5.7 - Schematic of main types of austenitic chromium-nickel and chromiumnickel-molybdenum starting from X5CrNi18-10 and X5CrNiMo17-12-2.
282
On the basis of the C curves austenitic steels with low carbon content
have been developed, such as X2CrNi18-9 (similar to AISI 304L27) or
X2CrNiMo17-12-2 (similar to AISI 316L) which are virtually insensitive
to the phenomena of chromium carbide precipitation.
Fig.5.7 - Time-temperature diagram (C curve) showing precipitation of chromium carbides on variation of the carbon content, for austenitic stainless steels with
18% chromium and 9% nickel.
283
284
pipes, plumbing, elevators and signs. It is also used in building and urban
furnishings, in domestic utensils (cookware, cutlery, appliances), in sinks
and cabinets, as well as in the chemical, petrochemical, nuclear and pharmaceutical industries as well as for the production of milk, beer, canned
food, paper and pulp, colorants and explosives. Due to its high toughness
at low temperatures, it is also widely used in the cryogenics sector for the
storage and transport of liquefied gases.
the types X6CrNiTi18-10 (similar to AISI 321) and X6CrNiNb18-10 (similar to AISI 347) are the stabilized versions with titanium and niobium. As
these materials have a high resistance to corrosion comparable with that of
X5CrNi18-10, they are used in the same applications, especially for the
production of large sized and/or very thick items that are subjected to
welding operations.
X2CrNi18-9 (similar to AISI 304L) is actually the low carbon version of
X5CrNi18-10: this composition is also used to solve issues of the precipitation of chromium carbides during welding and is a good alternative to
the previously indicated stabilized types.
The stainless steel X5CrNiMo17-12-2 has high resistance to corrosion in
seawater, in contact with process waters contaminated with halides (Cl-, I-,
F-), with acid condensates and with waste water; it has excellent behavior
in the presence of organic acids or alkaline solutions, also showing a small
resistance in dilute sulphuric, hydrochloric or phosphoric acid solutions.
mium content (normally between 11.5% and 18%) and carbon contents
among the highest of the stainless steels most commonly used (generally
between 0.1% and 1%).
Chromium, a highly ferritising element, and carbon, an austenitising element, are balanced so that the steel has an austenitic structure at high temperature and a martensitic structure at ambient temperatures after tempering.
It must be pointed out that the presence of a high content of carbon, required to ensure a good level of hardness and mechanical resistance for the
steel, also tends to aid the formation of chromium carbides. Consequently
martensitic stainless steels are, among all types, the least resistant to corrosion; their field of application in fact is limited to only slightly aggressive
environments. Fig. 5.8 illustrates the metallurgical logic that has led to
the creation of the main types of martensitic stainless steels, starting with
285
X30Cr13 (similar to AISI 420B), the progenitor of this family, and still today widely used in the knife sector.
Fig. 5.8 - Schematic of the main types of martensitic stainless steels, starting from X30Cr13 (similar to AISI 420B), progenitor of this family.
When such steels are heated to temperatures above the critical points, their
original microstructure transforms completely into austenite plus carbides,
to then become martensite plus carbides under the effect of cooling the
steel in oil or in air. Given the existence of critical points, it is therefore
possible to trace the isothermal transformation curves (TTT) and anisothermal curves (CCT) of the austenite: these curves are shown in fig. 5.9
for a martensitic stainless steel grade.
Due to similarities of heat treatments conducted onto martensitic stainless
steels and to hardened and tempered steels, their manufacturing cycle are
very similar, too.
286
Figure 5.9 - Isothermal transformation curves (TTT) and anisothermal transformation curves (CCT) for martensitic stainless steel type X39Cr13 (similar to AISI
420C).
Heat treatments
If no substantial differences exist in heat treating martensitic stainless
steels and common quenched and tempered steels, actually some peculiarities should be taken into account. In case of martensitic stainless steels,
heating phase must be carefully conducted since stainless steels has very
low conductivity. To avoid permanent deformation of parts, or crack formation, a worst case, the heating phase must be controlled with pre-heating
287
between 550C and 800C, especially in the case of large size parts and/or
parts with complex geometry.
As regards the subsequent austenitization phase, a higher temperature must
be envisaged with respect to common quenching and tempering steels, thus
it should be as higher as possible to solubilize carbides. However, grain
coarsening phenomena may occur at high temperatures, that leads to consequent reduction in resistance properties of the steel. For these conflicting
objectives, the optimized austenitizing temperature shall be chosen, depending on specific steels, in the range 950 C 1100 C.
Martensitic stainless steels can be tempered either at low temperatures (below 400C) or at high temperatures (over 640C). However, temperature
range of 450- 600C must be avoided. This interval is considered critical
as it reduces resistance
to brittle fractures (a type of temper brittleness) and significantly reduces
corrosion resistance.
Physical and mechanical properties
The physical properties of martensitic stainless steels are very similar to
those of the common quenching and tempering steels, sharing the same
martensitic tempered microstructure. In the hardened state most martensitic stainless steels have a unit resistance under traction Rm of between 700
MPa and 1700 MPa depending on the chemical composition and tempering conditions; the fatigue limit in air remains around the value 0.45Rm.
The ductile to brittle transition behavior is similar to that of special heattreatable steels used frequently in industrial applications.
Main applications
The martensitic stainless steels most commonly used in industrial sectors
are X30Cr13 (similar to AISI420B) and X12Cr13 (similar to AISI 410):
among the two, the first will guarantee hardness with a higher carbon content, while the second features increased strength.
These two steels are widely used where high mechanical resistance and
wear resistance is required. Corrosion resistance is good, especially in rural environments, fresh water not contaminated by chlorides, in contact
with foodgrade substances, or with weak acids (such as organic acids),
with petrol products (crude and intermediate) and with oxidating saline solutions (chromates, permanganates, etc.).These are frequently used for the
production of table knives, industrial cutting knives, surgical instruments,
rasors, scissors, calipers, machine gun barrels, brake disks for motor vehicles; they are also used in the sector of moulds for plastic materials and for
springs.
288
Type X17CrNi16-2 (similar to AISI 431) is must more resistant to corrosion with respect to those described above, due to the increased chromium
content in the alloy: it also offers increased strength due to the presence of
nickel (not present in types X30Cr13 and X12Cr13). The steel type
X17CrNi16-2 is used in particular for shafts on marine engines, propeller
shafts, pump parts, hydraulic machinery shafts, valves, turbine blades and
wear-resistant components in reforming oil plants.
Austenitic-Ferritic (Duplex) Stainless Steels
Austenitic-ferritic stainless steels (also called duplex or biphasic) owe their
name to the presence of a mixed structure of austenite and ferrite at ambient temperature. This is the result of the combination of chromium and
nickel suitably balanced in the chemical composition of the alloy: typically, duplex stainless steels have a chromium content between 22% and
25%, nickel content between 4% and 7%, with added molybdenum (3-4%)
and nitrogen (0.1-0.25%) in many cases.
The commercial name originally developed by Sandvik is most commonly
used: it includes two pairs of numbers of which the first indicates the
chromium content and the second the nickel content (for example 2205 indicates a biphasic stainless steel containing 22% chromium and 5% nickel). Duplex stainless steels allow the combination of the specific properties
of resistance to corrosion of the austenitic and ferritic stainless steels;
however, they also share a number of specific problems, typical of both,
such as the precipitation of chromium carbides and the presence of a harmful phase that forms at around 800C as well as an embrittlement issue at
475C.
Heat treatments
An annealing treatment before being put into operation is required in temperature interval of 1050-1150 C; it is then rapidly cooled in water to optimize the austenite ferrite ratio that should be close to unity.
Physical and mechanical properties
Most of the physical properties are all comparable with those of austenitic
stainless steels (specific heat, conductivity, resistance and elastic modulus); the only exception is thermal expansion between austenitic and ferritic stainless steels. When it is exposed to magnetic fields, the duplex component is ferromagnetic, similar to what happens for ferritic stainless
steels.
The peculiar biphasic microstructure of such steels allows high values to
be obtained of both the yield stress (which is around double that of austen-
289
itic stainless steels) and the tensile strength. Furthermore, they have an excellent behavior in terms of DBTT, due to transition temperature set to
around -80C.
Main applications
X2CrNiN23-4 (type 2304) was the first of the low alloy biphasic stainless
steels: this steel was developed to economically compete with the most
common austenitic stainless steels used in applications where mechanical
resistance and resistance to corrosion in chloride environments are essential. It is used in the production plants of nitric acid and for plants in contact with caustic substances (concentrations <30%), with organic acids or
aqueous solutions rich in chlorides.
The steel X2CrNiMoN22-5-3 (type 2205) is most used duplex stainless
steel, thanks to its high mechanical resistance and excellent anti-corrosion
characteristics. It is largely used in the petrochemical and chemical field,
in environments contaminated by chlorides and in the presence of carbon
dioxide CO2 or hydrogen sulphide H2S; it is widely used in desalinisation
plants and plants where steel parts shall be in contact with dilute and concentrate solutions of sulphuric, phosphoric, acetic and formic acids.
290
291
Semi-austenitic precipitation hardening stainless steels, such as for example X7CrNiAl17-7 (also called 17-7 PH), have a predominantly austenitic
structure after annealing at ambient temperature and cold-working operations, such as cold rolling, can be more easily conducted. After annealing
the austenitic structure is unstable so that either cold plastic deformation
operations or post heat treatment transform microstructure into martensite.
The aging heat treatment will be performed on the martensitic structure obtained in this way, similarly to what happens in normal martensitic PH
stainless steels.
Main applications
Precipitation hardening stainless steels are generally commercialized onto
market in form of long products, round or hexagonal section bars, plates
and sheets. Thanks to high mechanical resistance and resistance to corrosion, they are employed in many sectors to fabricate valves, shafts, bearings and turbine and compressor blades. They are also used in applications
for load cells, sprockets, firearms, utensils, springs and surgical instruments.
292
293
34
When designing with oxidation-prone materials, it is obviously vital to know how fast
the oxidation process is going to be. Intuitively one might expect that, the larger the energy
released in the oxidation process, the faster the rate of oxidation. For example, aluminium
oxidizes much more slowly than iron. If you heat a piece of bright iron in a gas flame, the
oxygen in the air reacts with the iron at the surface of the metal where the oxygen and iron
atoms can contact, creating a thin layer of iron oxide on the surface, and making the iron
turn black. The layer grows in thickness, quickly at first, and then more slowly because iron
atoms now have to diffuse through the film before they make contact and react with oxygen. If you plunge the piece of hot iron into a dish of water the shock of the quenching
breaks off the iron oxide layer, and you can see the pieces of layer in the dish. The iron surface now appears bright again, showing that the shock of the quenching has completely
stripped the metal of the oxide layer which formed during the heating; if it were reheated, it
would oxidize at the old rate. Thus, the important thing about the oxide film is that it acts as
a barrier which keeps the oxygen and iron atoms apart and cuts down the rate at which
these atoms react to form more iron oxide. Aluminum, and most other materials, form oxide barrier layers in just the same sort of way - but the oxide layer on aluminum is a much
more effective barrier than the oxide film on iron is.
294
295
since molybdenum is effective for these steels against coarsening of carbides in such temperature range.
Among chromium-molybdenum structural steels, low alloyed CrMoNiMn
are weldable ferritic steels. The 8CrMoNiNb9-10 (equivalent to ASTM
A335 grade P24) series are the mostly used for seamless alloy-steel pipe
for high-temperature service.
The increase of chromium to above 7% in CrMo steels leads to a group of
steels containing also martensite. This microstructure introduces a new element of structural hardening, despite ferritic steels, due to its a fine lath
structure with stabilized fine carbides dispersed in matrix. Thus, structural
hardening by martensite is responsible for the large increase in strength of
X11CrMo9-1, as compared to 8CrMoNiNb9-10, as shown by Fig.6.2.
Further improvements, especially of the creep strength, have been
achieved by alloying with vanadium, niobium, tungsten and boron, as also
shown in Fig. 6.2b. The introduction of X20CrMoNiV11-1 at the beginning of the sixties allowed major increases in power plant efficiency. The
transformation behavior and microstructure of this alloy are comparable to
those of X11CrMo9-1. The higher creep-rupture strength of
X20CrMoNiV11-1 results mainly from the larger volume of dispersed
M23C6 carbides in the microstructure, a result of the alloy's higher carbon
content. The best results in terms of elevated-temperature strength are obtained by quenching and tempering to produce a microstructure consisting
of upper bainite.
Both the chromium molybdenum and the chromium-molybdenum vanadium steels extend the operating range up to 540C.
296
(a)
(b)
Fig.6.2 - Creep-rupture strength of some Cr-Mo and Cr-Mo-V heatresistant steels.
Stainless steels for high temperatures
Stainless steels are also used for elevated temperature applications. Martensitic stainless steels are used at temperatures up to 540 to 650C, but
they must be tempered at approximately 55C higher than the operating
temperature so to prevent softening in service (due to coarsening of carbides formed during previous tempering treatment).
Due to their higher chromium contents, ferritic stainless steels have the
best scaling resistance, but they are limited to service temperatures of approximately 370C to avoid the precipitation of the embrittling phase
(refer to Chapter 5, Fig. 5.2).
Austenitic stainless steels have the best creep resistance of the stainless
steels and are used at temperatures up to 870C. In austenitic stainless
steels, strengthening mechanism for enhancing creep resistance is related
297
298
ments. Although nickel itself is not a carbide former, the presence of carbon as an alloying element leads to the formation of carbides. The carbides
most frequently found in nickel alloys include MC, M6C, M7C3, and
M23C6, where M is the carbide-forming element or elements, such as
chromium, which forms Cr7C3. In general, carbides and fine particles improve the elevated-temperature strength if they are located on the grain
boundaries as discrete particles. However, if they form continuous grainboundary films, they can cause embrittlement.
Nickel-Chromium-Iron Alloys.
This family of alloys was developed for high-temperature oxidizing environments. They typically contain 50 to 80 wt% Ni, which permits the addition of other alloying elements to improve strength and corrosion resistance while maintaining toughness. The Ni-Cr-Fe alloys contain
chromium ranging from 14 to 30 wt% and they form a protective surface
film of Cr2O3, they have excellent corrosion resistance in many severe environments, including immunity to chloride ion stress corrosion cracking.
They also have good oxidation and sulfidation resistance along with good
strength at elevated temperatures.
Among variety, the alloy 800 commercially known as INCOLOY is an NiCr-Fe alloy austenitic, solid-solution alloy. Titanium nitrides, titanium carbides, and chromium carbides normally appear in the alloys microstructure. The nitrides are stable at all temperatures below the melting point and
are therefore unaffected by heat treatment.
At elevated temperatures it offers resistance to oxidation, carburization,
and sulfidation along with rupture and creep strength. It is widely used material for construction of equipment requiring corrosion resistance, heat resistance, strength, and stability for service up to 816C. For applications
requiring greater resistance to stress rupture and creep, especially at temperatures above 816C, INCOLOY alloys 800H and 800HT 35 are used.
INCOLOY is approved under the Boiler and Pressure Vessel Code of the
American Society of Mechanical Engineers (ASME).
Incoloy 800, 800H and 800HT are identical except for the higher level of
carbon in alloy 800H compared to 800 series and the addition of up to 1.20 percent aluminum and titanium in alloy 800HT, compared to 800 and 800H. Incoloy
800 was the first of these alloys and it was slightly modified into Incoloy 800H.
This modification was to control carbon (0.05-0.10%) and grain size to optimize
stress rupture properties. Incoloy 800HT has further modifications to the combined titanium and aluminum levels (.85-1.20%) to ensure optimum high temperature properties.
35
299
Ni-Superalloys
Ni-superalloys are heat-resistant alloys of nickel, iron-nickel that are frequently used at temperatures exceeding 540C. However, some superalloys are capable of being used in load-bearing applications in excess of
85% of their incipient melting temperatures. They exhibit the best combination of high strength, good fatigue and creep resistance, good corrosion
resistance, and the ability to operate at elevated temperatures for extended
periods of time (i.e., metallurgical stability). Their combination of elevated
temperature strength and resistance to surface degradation is unmatched by
other metallic materials. Superalloys have been used in aircraft, industrial,
and marine gas turbines, nuclear reactors, aircraft skins, spacecraft structures, petrochemical production, orthopedic and dental prostheses, and environmental protection applications. Todays modern, high-performance
aircraft jet engine could not operate without the major advances made in
superalloy development over the past 50 years. Thanks to superalloys, increases in engine operating temperatures over the last 50 years have been
gradual but significant, from approximately 400C to 1290C, that resulted increase in operating temperature translates into improved engine efficiency. Typical applications for superalloys in a jet engine are shown in
Fig. 6.3.
300
301
Hastelloy series, solid-solution-strengthened nickel-base superalloys primarily developed for low-temperature service, often in corrosive environments.
Inconel 718, a superalloy that contains niobium which is strengthened primarily by phase (i.e. Ni3Nb particles); it rapidly loses
strength in the range of 650 to 815C because of the instability of
the precipitates. However a widespread use of Inconel 718 is
due to its excellent combination of mechanical properties, moderate price, and ease of processing, including weldability. Inconel
718 may account for as much as 60 wt% of the total consumption
of superalloys.
An example of application: Ni-superalloy made turbine blade
Improved high-temperature material properties are largely responsible for
the increase in the engine operating temperature. Some improved material
performance has come about from minor alterations in alloy chemistry,
and some has resulted from processing changes. For example, the major
stress axis in a turbine blade is parallel to the blade axis. In a polycrystalline blade, this stress is normal to some grain boundaries, and this causes
voids, precursors to fracture, to initiate on the boundaries (Fig.6.5).
302
For further increase in creep resistance, the turbine blades made of superalloys are coated with thermal barrier coating. To insulate blade of NiSuperlloy, overlay protective coatings are generally deposited on deposited
on top surface of base Ni-superalloy metal; MCrAl or MCrAlY, zirconia
layers (0.3-0.4mm thick) coatings, are derived directly from the deposition
process. They do not require diffusion for their formation. The constituent
denoted M in these designations has, at various times, been iron, cobalt,
nickel, or combinations of nickel and cobalt. Overlay coatings are usually
applied by physical vapor deposition (PVD) in vacuum chambers. They also may be applied by plasma spray techniques.
Temperature C
creep is a diffusion creep mechanism that consists in diffusion of atoms in a material along the grain boundaries, which produces a net flow of material and a sliding of the
grain boundaries.
303
Table 7.1 The major driving forces for competing on today automotive market.
Recent changes to the Corporate Average Fuel Economy (CAFE) are driving automakers to seek more aggressive methods for fuel consumption reductions. Lightweighting of vehicles will be a factor in meeting these re37 Main reference: F.C. Campbell, Elements of Metallurgy and Engineering Alloys,
ASM International, 2008 (Chapters 27).
304
quirements due to the inherent relationship between mass and fuel consumption. In addition, lightweighting may benefit other advanced fuelsaving but load constrained technologies, such as battery-powered vehicles.
The ability to introduce new lightweight materials into vehicles is not a
trivial matter. Many see a new concept, or limited production, vehicle introduced to the market with lightweight space-aged materials and feel
that adoption by mass produced vehicles is a simple matter of remove and
replace. However, this is not the case; factors such as existing infrastructure, material cost, and high volume capacity become of great importance
for mass production vehicles. In addition, many of the low production vehicles incorporate these lightweight materials as a method for gaining experience on their performance. Without significant data to support durability, the risk-averse automotive culture will not adopt new materials.
Therefore it often takes many years to implement lightweight technology
in mainstream production vehicles.
Today, there is a high emphasis on greenhouse gas reductions and improving fuel efficiency in the transportation sector, all car manufacturers, suppliers, assemblers, and component producers are investing significantly in
lightweight materials Research and Development and commercialization.
All are moving towards the objective of increasing the use of lightweight
materials and to obtain more market penetration by manufacturing components and vehicle structures made from lightweight materials. Because the
single main obstacle in application of lightweight materials is their high
cost, priority is given to activities to reduce costs through development of
new materials, forming technologies, and manufacturing processes. Yet the
weight reduction is still the most cost-effective means to reduce fuel consumption and greenhouse gases from the transportation sector. The reasoning behind this is simple; it takes less work to accelerate and move a lighter object. It has been estimated that for every 10% of weight eliminated
from a vehicle's total weight, fuel economy improves by 7%. This also
means that for every kilogram of weight reduced in a vehicle, there is
about 20 kg of carbon dioxide reduction. In terms of lifecycle management, the use cycle of the vehicle typically dominates the overall CO2
emissions generation of the vehicle (see Figure 4). Lighter weight materials have the advantage of providing sustained greenhouse gas emission reductions over the use cycle of the vehicle. Conventional steels have been a
dominant force in the material selection of automotive parts. A combination of high-strength, low cost and capacity to be produced in high volumes has made mild steel a very attractive material in automotive applications. However, as technology changes and fuel economy increases in
305
importance, other materials are gaining acceptance in the automotive industry. Materials such as high-strength steel, aluminum, composites, and to
date to a much lesser extent magnesium have increased their overall utilization over the past 30 years, while mild steels have seen a steady decline
(see Fig.7.1). At the same time, regular steel and iron castings have also
seen a steady decline.
306
307
308
Figure 7.2 - Total Greenhouse Gas Emissions during Various Phases of the Vehicle
Lifecycle.
309
which have key role in different stages of crash. However the materials deformation and progressive failure behavior in terms of stiffness, yield, strain hardening, elongation and strain at break are also
very important in the energy absorption capacity of the vehicle.
[Witteman, 1999].
Recycling and life cycle considerations. One of the major growing
concerns in all the industries including automotive, is an increased
awareness for environment. Issues such as protection of resources, reduction of CO2 emissions, and recycling are increasing the topics of consideration. While the United States has not issued regulations concerning automotive end-of-life requirements,
European Union (E.U.) and Asian countries have released stringent
guidelines. European Union legislation implemented in 2006 dictates that a significant percentage of the vehicle should be re-used
or recycled. The End of Life Vehicles (ELV) Directive from environment agency aims to reduce the amount of waste produced from
vehicles when they are scrapped. Around two million vehicles
reach the end of their life in the UK each year. These vehicles are
classed as hazardous waste until they have been fully treated. The
directive requires ELV treatment sites to meet stricter environmental standards. This would mean that the last owner of a vehicle must
be issued with a Certificate of Destruction for their vehicle and they
must be able to dispose of their vehicle free of charge. Vehicle
manufacturers and importers must cover all or most of the cost of
the free take-back system. It also sets higher reuse, recycling and
recovery targets and limits the use of hazardous substances in both
new vehicles and replacement vehicle parts. [Environment agency,
2010]. As the results of the new legislations, no discussion of new
materials in the automotive industry should conclude without a
consideration of recycling. Considerable R&D efforts are now focused on developing materials with greater potential of recycling
and re-use or developing ways of recycling and re-use of the current materials. This includes both metal and composite materials.
The composition and forming processes of the metal materials are
changing to accommodate this recycle and re-use demand. This also justifies the great attention towards natural fiber based composites and new high temperature resistant thermoplastic resins. Traditionally the design of the part/components has mainly been the
responsibility of the OEMs. Recently, however, manufacturers have
begun to shift design responsibility to part/component suppliers.
After the manufacturing stage, the vehicle spends a long period of
time in the use phase of the life cycle. In the past, the median life of
310
311
One other major drawback that has been considered over the years is the supposed
high ignitability of magnesium. If it is dust or fine particles of magnesium are highly ignitable in air, due to high reactivity of magnesium to oxygen, it is not quite different from the
affinity and reactivity of aluminum fine powders or dust to oxygen. What is really different
and more hazardous for magnesium is that, when ignited, it cannot be extinguished simply
removing heat sources that has brought magnesium to melt point and flame ignition condi-
312
Figure 7.3 The hexagonal close-packed crystalline structure, schematic representation on the left, actual spatial occupancy of Mg atoms o the right.
Magnesium also has a rather low melting point (650C), which increases
its susceptibility to elevated-temperature creep. However, through im-
tion. However, recent developments by Koreans researchers has led to put onto marketplace
new magnesium series added with CaO that make magnesium alloy self-extinguishable in
case of flame ignition. These magnesium modified alloy series are called Eco-Magnesium
and when they are produced by adding over 0.5~0.7wt% CaO they can be considered fireretardant and fire-proof.
313
314
solution, and since the compound is brittle, alloys with a eutectic network
are also brittle. When the aluminum content exceeds 8%, discontinuous
precipitation of the phase occurs at the grain boundaries as shown in
Fig.7.5, leading to a decrease in ductility. The phase diagram also shows
that aluminum has a maximum solubility of 12.7 wt% at 436C, which decreases down to approximately 2 wt% at room temperature. While this
type of solubility would seem to indicate that these alloys would be amenable to precipitation hardening, unfortunately the resulting precipitate is
coarse and results in only moderate hardening. While many of the alloying
additions to magnesium form eutectics with decreasing solid solubility that
lead to precipitation hardening, the strength that results from this hardening mechanism is much less than that observed with the precipitation hardened aluminum alloys. Since the response to precipitation heat treatment is
poor, most of these alloys are used in either the as-cast.
Zinc behaves in a manner similar to aluminum; the ductility reaches a
maximum at a 3 wt% addition, and a good combination of strength and
ductility occurs with 5 wt% Zn. It also improves the corrosion resistance
by combining with the harmful impurities iron and nickel. Zinc is also
used in conjunction with zirconium, rare earths 40, or thorium to produce
precipitation-hardenable alloys.
Zirconium is a powerful grain refiner. However, zirconium cannot be
used in combination with aluminum or manganese because it forms brittle
intermetallic compounds that destroy ductility.
Harmful elements
The elements iron and nickel are harmful impurities that greatly reduce
corrosion resistance. Copper is also often considered, along with iron and
nickel, as an impurity but is actually used as an alloying addition in some
magnesium alloys. Iron is by far the most troublesome of the three, because nickel and copper are more readily controlled by selecting the purity
of the starting materials.
The ASTM designation of Mg alloys and most used Mg alloys
In the following, an example of ASTM designation of magnesium alloys is
shown and thus commented as practical example (all magnesium alloys are
designated by this scheme):
AZ91D
Rare earth (RE) elements are potent solid solution strengtheners in magnesium alloys.
40
315
316
317
318
gas-shielded arc welding processes (MIG/TIG), although processing parameters and equipment used for aluminum welding have to be modified. In automotive sector, the most promising method for welding thin Mg sheets is
laser or electron beam, while friction stir welding is a more recent development suitable for joining materials that are more difficult to weld or even for
joining dissimilar materials like magnesium or aluminum alloys.
Corrosion protection. For ordinary outdoor applications, substantial improvement over the earlier alloys are sufficient. On other side, seacoast locations involving direct contact with salts are definitely corrosive to magnesium alloys. The optimal corrosion protection (in terms of thickness and
resistance) is provided by an anodizing treatment, followed by the application of an organic paint system chemical conversion coatings or anodizing
treatments, are required. A number exist, such as the older Dow 17 and HAE
and the newer and improved Tagnite, Keronite, and Magoxid treatments. In
many cases it is sufficient to simply coat the counterpart and leave the magnesium uncoated (if it is no view part). This to prevent a defect in a coating
on magnesium would result in an enhanced localized corrosion attack of the
magnesium component. No contact corrosion of magnesium is caused by
anodized AlMg3 alloy. Conventionally galvanized steel bolts can be attached to magnesium by using a silicate sealing. The silicate sealing of galvanized steel bolts was successfully applied by Audi and VW at the B80
magnesium gear housing.
319
adding CaO to conventional Mg alloys; ii. a non SF6 (and other GHGs cover
gases) die-casting is possible; iii. just safe and clean nitrogen enriched atmosphere is required for prolonged time, while for short process cycle time
casting can be done in air. These novel Mg-CaO added alloys are available
on the market, known as Eco-Magnesium alloys.
As alternative shaping technology, semisolid processes have been researching to produce parts that can compete with die-cast aluminum parts.
A semisolid process relies on the behavior of metals that, stirred in intermediate liquid and solid phase, exhibit a remarkable decrease of its viscosity
[1]. Scientifically referred as thixotropic behavior [2-6], this phenomenon in
metals is well explained because spheroidal rather than a dendritic microstructure forms during agitation of semisolid slurry [7]: rounded solid particles are suspended in the liquid fraction, thus viscosity drops. Consequently,
a laminar not turbulent flow is possible to fill semisolid material into high
complexity cavities with smooth mold filling preventing air entrapment, as it
is usual in conventional die-casting. The absence of air/gas entrapment results in:
higher integrity of the parts, due to absence of internal gas porosities, with superior mechanical properties;
possibility to conduct post-heat treatments to increase furthermore
mechanical properties, not possible for the cast alloys (gas entrapped causes blistering phenomena during heat).
The state-of-the art recognizes for aluminium alloys two semisolid routes,
Rheocasting and Thixocasting, while for magnesium alloys, only a
commercial semisolid process route is possible, and it is called Thixomolding:
i.
Thixocasting: consists in the manufacture of billets of the desired
globular microstructure - usually by continuous casting and induction stirring. The billet are fully solidified in globular shaped microstructure and used as feedstock material for a subsequent reheating in semisolid state and final shaping.
ii.
Rheocasting: the semisolid slurry is prepared directly by spilling
out from furnace into a heated crucible equipped with a stirrer;
once the semisolid slurry is ready, it is gravity poured into the die
or pressure injected. The advantage against Thixocasting is that
avoids the re-heating of thixotropic state billet, as happens in the
Thixocasting process. The advantage is reduced energy per kg of
finished product.
iii.
Thixomolding: employs a machine similar to injection molding of
polymers. Room temperature magnesium alloy feedstock (i.e.
chips) are fed into the a heated barrel through feeder. A screw inside the barrel agitates and shears chips while they are heated up to
semi-solid temperature range. The semisolid slurry once ready is
320
321
Titanium is considered the most biocompatible of all metals due to its ability
to withstand attack from bodily fluids, stay inert in the human body, be compatible
with bone growth and stay strong and flexible during use. The TiO2 protective
film that forms onto titanium metal when titanium is exposed to air (i.e. oxygen in
air) is insoluble and chemically non transportable so that it makes such titanium
highly resistant to body environments, preventing any immune reaction from occurring.
42
322
C), ingot casting and primary fabrication procedures are complicated and
expensive. Secondary fabrication processes, such as forming and machining, are also usually more costly than those for other competing metals.
Titanium Metallurgy
Pure titanium at room temperature has an alpha () hexagonal closepacked (hcp) crystal structure, which transforms to a beta () bodycentered cubic (bcc) structure at a temperature of approximately 885C.
This transformation temperature, known as the beta-transus temperature,
can be raised or lowered depending on the type and amount of impurities
or alloying additions.
As a result of the hcp crystalline structure, alloys with appreciable amounts
of alpha must be formed at elevated temperatures, while those with predominantly bcc structures exhibit varying degrees of room temperature
formability. At room temperature, commercially pure titanium is composed primarily of the alpha phase. As alloying elements are added to titanium, they tend to change the amount of each phase present and the betatransus temperature in the manners shown in Fig. 8.1.
323
Specifically:
Addition of alpha-stabilizer elements, such as aluminum, oxygen, nitrogen, and carbon, increases the beta-transus temperature by stabilizing the
alpha phase (see Fig.8.1, upper left box);
Addition of beta-stabilizer elements, such as molybdenum, vanadium,
tantalum, and niobium, that are miscible in the beta phase decreases the
decrease the beta-transus temperature and stabilizes the beta phase (see
Fig.8.1, upper right box);
Addition of eutectoid-forming elements, such as manganese, iron, chromium, cobalt, nickel, copper and silicon, that forms an eutectoid reaction
with Ti (see Fig.8.1, bottom left box).
Titanium has a great affinity for interstitial elements, such as oxygen and nitrogen,
and readily absorbs them at elevated temperatures, which increases strength and
reduces ductility.
Hydrogen has detrimental effect, and it always must be minimized in titanium alloys because it causes hydrogen embrittlement by the precipitation of hydrides.
The maximum limit allowed is approximately. 0.015 wt% (~100 ppm). Absorption of several hundred ppm of hydrogen results in embrittlement (Fig. 8.2) and
the possibility of stress cracking. Note that the addition of 20 ppm does not cause
embrittlement, but when the hydrogen content goes up to 250 ppm, the reduction
in area is seriously impaired.
324
Titanium Alloys
Titanium alloys are classified according to the amount of alpha and beta
retained in their structures at room temperature. Classifications include
commercially pure, alpha and near-alpha, alpha-beta, and metastable beta.
While these classifications are useful, many of them are actually very close
to each other in the total amount of beta stabilizer present, as illustrated
in the Fig. 8.3 phase diagram. For example, Ti-6Al-4V is classified as an
alpha-beta alloy, and Ti-6Al-2Sn-4Zr-2Mo is classified as a near-alpha alloy, yet they differ little in the total amount of beta stabilizer concentration.
The properties of a number of commercially important alloys are given in
Table 8.1.
325
326
only alpha, it cannot be strengthened by heat treatment. Except for cryogenic applications, the use of Ti-5Al-2.5Sn has declined as alloys with better forming properties and higher creep resistance have been developed.
Near-alpha alloys contain small amounts beta phase dispersed in an otherwise all-alpha matrix and they generally contain 5 to 8 wt% Al. The
amount of aluminum that can be added as an alloying element is limited 43
since an excess of aluminum content forms the brittle intermetallic compound -2 (Ti3Al) forms, which adversely affects ductility.
The near-alpha alloys (such as Ti-6Al-2Sn-4Zr-2Mo-0.25Si) retain their
strength to high temperatures and have good creep resistance in the range
of 315 to 595C; furthermore they have good cryogenic behavior.
Alpha-Beta Alloys
They contain both the alpha and beta phases. Again, aluminum is the principal alpha stabilizer that strengthens the alpha phase but beta stabilizers,
such as vanadium, are added and they also provide strengthening and allow these alloys to be hardened by solution heat treating and aging. An example is the widely commercial available Ti-6Al-4V, which is the workhorse of the aerospace industry.
Although the metallurgy of titanium heat treatment is complex, the response to heat treatment of alpha-beta alloys is a result of the instability of
the high-temperature beta phase at lower temperatures. You may refer
again to Fig.8.3, and consider the Ti-6Al-4V as major example. Heating
such an alloy, to the solution treating temperature (above martensite start,
Ms, line), produces a higher ratio of beta phase. During rapid cooling 44, i.e.
quenching, the beta is transformed to beta and titanium martensite or
phase, the acicular structure shown in Fig.8.4a.
During subsequent aging (i.e. reheating at low temperature) treatment,
strengthening effects may occur together 45:
a. the decomposition of the unstable martensite into beta-plus-alpha
microstructure (it is defined as the first stage of hardening) visible
in Fig.8.4b;
43 The maximum aluminum content is actually calculated considering other elements
that promote a-2 phase formation; thus, the maximum aluminum content must be account
considering the aluminum equivalent content that is determined by summing the following
weight percentages of alloying elements: Aleq = Al+1/3Sn+1/6Zr+10 (O+C+2N).
44 Since the response to heat treatment is a function of cooling rate from the solution
temperature, the section sizes that can be through hardened are limited.
45 The martensite formed in titanium alloys is not like the extremely hard and strong
martensite formed during the heat treatment of steels.
.
327
(a)
(b)
Figure 8.3 - Microstructure of Ti-6Al-4V: a) after quenching from solution treating temperature; b) after aging treatment.
The weldability of the alpha-beta alloys is not as good as the near-alpha alloys, but their formability is better. The alloys that contain smaller percentages of beta stabilizers, known as lean alloys, are more weldable. As
the amount of beta stabilizers increases, the weldability decreases.
Beta Alloys
Beta alloys are sufficiently rich in beta stabilizers and lean in alpha stabilizers that the beta phase can be completely retained with appropriate
cooling rates. Beta alloys are metastable, thus the precipitation of alpha
phase in the metastable beta is usual and it is also a method used to
strengthen the alloys. Beta alloys contain small amounts of alpha stabilizing elements as strengthening agents, such as Ti-10V-2Fe-3Al. such alloys
are high-strength alloys that can be heat treated to tensile strength levels
approaching 1380 MPa. Heat treatment consists in solution treating and
aging hardening treatment, as for the alpha-beta alloys. Since their better
response to solution and aging treatment than the alpha-beta alloys, heavier
sections can be treated. The aging sequence, including temperature and
time, is important in producing a uniform precipitation without the occurrence of grain-boundary alpha. In fact, an excessive grain-boundary alpha
precipitation is detrimental to alloy ductility, fatigue strength, and stresscorrosion cracking resistance.
328
329
330
terms of chemical alloying elements) into further furnace, the Ladle Furnace or LF (see again Fig.9.1, refer to unit 3);
Primary steelmaking production from steel scraps. Because the raw material used is already steel, the upstream stages conducted into Blast Furnace and Basic Oxygen Furnace are not necessary. The steelmaking process in this case starts from the melting of steel scraps into an Electric
Furnace (EF); melt steel is therefore spilled out to the Ladle Furnace (see
again Fig.9.1, refer to unit 3), where chemical composition refinement
takes place.
Note from above that, despite two different steelmaking processes are required if
you start from either iron ore or recycling scraps, the steps that follow the unrefined steel production in either BOF or EF are the same: the unrefined steels shall
be turned into desired chemical composition in the Ladle Furnace.
When steel is refined, it can be cast into ingots or continuously cast into billets,
blooms or slabs, depending on section shape and size 48 (see again the scheme in
Fig.9.1).
Cast semifinished products are therefore hot worked to improve homogeneity, refine the as-cast microstructure, and fabricate desired product shapes. All these
downstream steps are those we call secondary steelmaking process and, by generalizing, they start from initial hot deformation processes (e.g. forging or rolling)
can proceed in further refinement of section, shape and size of products. Some
products therefore derives directly from hot rolling operations, and they are used
in the hot rolled condition, some others are heat treated to obtain specific properties.
Some other steel products are fabricated by secondary working by hot forging into
semifinished shapes or hot extruding.
For some other products thin sections, close clearances and good surface quality is
required for example to fabricate cold drawn pans, wound cold springs, etc. In this
case the hot deformed semifinished products are further cold worked to achieve
final shape, size and surface quality.
In the following sections, some further details on main units of the upstream primary steelmaking production for both the iron-ore and recycled steel scraps pattern routes are provided.
48 Continuous casting process can produce semifinished products that can be: a) billets,
namely long semifinished products with round or square cross-section and with section area less than 230 cm2; b) blooms, similar to billets except the cross-sectional area that is
greater than 230 cm2; c) slabs, long semifinished product with rectangular cross-section area.
1
2
Secondary production
Blast Furnace
As mentioned, the first step in making steel from iron ore is to make iron
by chemically reducing the ore (iron oxide) with carbon, in the form of
coke, according to the general equation:
(eq. 9.1) Fe2O3+3CO
2Fe+3CO2
333
Blast furnace main units - 1: Iron ore + Calcareous sinter; 2: coke; 3: conveyor belt; 4: feeding opening, with a valve that prevents direct contact with the internal parts of the furnace; 5: Layer of coke; 6:
Layers of sinter, iron oxide pellets, ore; 7: Hot air (around 1200C); 8: Slag; 9: Liquid pig iron; 10:
Mixers; 11: Tap for pig iron; 12: Dust cyclon for removing dust from exhaust gasses before burning
them in 13; 13: air heater; 14: Smoke outlet (can be redirected to carbon capture & storage (CCS)
tank); 15: feed air for Cowper air heaters; 16: Powdered coal 17: cokes oven 18: cokes bin 19: pipes
for blast furnace gas.
The stack inside the blast furnace is kept full with alternating layers of
coke, ore, and limestone admitted at the top during continuous operation
(see Fig.9.2, details 1, 2 and 4). Coke is ignited at the bottom and
burned rapidly with the forced air from the tuyeres 49 (see Fig.9.2, detail
7). The iron oxides in the ore are chemically reduced to molten iron by
carbon and carbon monoxide from the coke. The slag formed consists of
the limestone flux, ash from the coke, and substances formed by the reaction of impurities in the ore with the flux; it floats in a molten state on the
top of the molten iron; see the detail 8 in Fig.9.2, as it represents the slag
that floats it is lighter than metal - onto upper part of molten metal. The
slag is periodically purified from molten metal by the opening outlet in
8, thus eliminated for recycling by 8. On the opposite side, from the
49 From the French, tuyer is an opening water-cooled nozzle through which air is blown
into a blast furnace to facilitate combustion.
334
outlet 9 the molten pig iron is spilled out and loaded onto torpedo cars
(see Fig.9.2, detail 11) for transferring it quickly to the next station, the
Basic Oxygen Furnace, where conversion from pig iron into steel takes
place.
Hot gases rise from the combustion zone, heating fresh material in the
stack and then passing out through ducts near the top of the furnace.
Finally exhaust gases must be captured by pipes (see Fig.9.2, detail 19),
cleaned by dust by use of cyclone for removing dust (see Fig.9.2, detail 12)
so to burn them in an air heater (see Fig.9.2, detail 13) and finally transferred to smoke outlet (in modern plant, this can be redirected to carbon
capture & storage CCS tank to reduce CO2 emissions).
Modern blast furnaces range in size from 20 to 35 m, have hearth diameters of 6 to 14 m, and can produce from 1,000 to almost 10,000 tons of pig
iron daily.
Basic Oxygen Furnace
Basic Oxygen Furnace (BOF) or Basic Oxygen Converter is a steel making
furnace, in which molten pig iron and steel scrap convert into steel due to
oxidizing action of oxygen blown into the melt under a basic slag. Typical
basic oxygen furnace is shown in fIg.9.3a. It has a vertical vessel lined
with refractory lining. Actually only 8-12% of the furnace volume is filled
with the treated molten metal. A careful balance between the relative
amounts of pig iron and scrap charged into the converter is maintained as a
means of controlling the temperature and to ensure that steel of the required specification is produced.
A water-cooled lance is then lowered into the vessel (see Fig.9.3), through
which very pure oxygen is blown at high pressure. The oxygen interacts
with the molten pig iron to oxidize undesirable elements, including excess
carbon, manganese, and silicon from the ore, limestone, and other impurities such as sulfur and phosphorus.
Oxidation of the molten metal and the slag is complicated process proceeding in several stages and occurring simultaneously on the boundaries between different phases (gas-metal, gas-slag, slag-metal), however finally
the reactions may be presented as follows. The lance "blows" 99% pure
oxygen over the hot metal, igniting the carbon dissolved in the steel, to
form carbon monoxide and carbon dioxide, and causing the temperature to
rise to about 1700C. This melts the scrap, lowers the carbon content of the
molten iron and helps remove unwanted chemical elements.
After a sample has been taken to verify the chemical composition of the
steel, the vessel is tilted to allow the molten steel to flow out. The steel is
335
(a)
(b)
Fig. 9.3 A Basic Oxygen Furnace: a) scheme of equipment; b) the BOF, the
converter, while it is loaded by pig iron transferred from torpedo car (see upperright corner) and when steel is spilled out from converter to steel ladle, to reach
the next station, the Ladle Furnace. For a video describing Basic Oxygen Furnace
operation click here.
336
50 Since the electric arc furnace has a relatively low capital equipment cost and uses
steel scrap, this process is used where local supplies of steel scrap are available and has
given rise to what are known as mini mills.
51 The electric arc furnace is also used for producing alloy steels that contain appreciable amounts of easily oxidized alloying elements, such as chromium, tungsten, and molybdenum. It can also be used to make steels requiring very low sulfur and phosphorus contents. Special slags are used to lower the sulfur and phosphorus levels and to protect against
oxidation of alloying elements.
337
Fig. 9.4 Electric arc furnace main operations. For a video describing Electric Arc
Furnace gen Furnace operation click here.
Ladle Furnace
Molten steel have to be refined to desired chemical composition in a device called Ladle Furnace (LF). From BOF the unrefined steel is poured
into a ladle that moved quickly to next stand, the Ladle Furnace, where it
is placed under a cover equipped with three graphite electrodes connected
to a three-phase arc transformer. The scheme of a Ladle Furnace stand is
illustrated in Fig.9.5; notice the ladle visible under the cover with three
electrodes.
(a)
(b)
Fig.9.5 (a) Scheme of a Ladle Furnace stand; (b) section of Ladle Furnace with
main functions.
338
The ladle bottom has a porous refractory plug, which is connected to the
argon supply pipe at the Ladle Furnace stand. The LF stand is also
equipped with an addition hopper mounted on the cover and a lance for injection of desulfurizing agents. Fumes formed during the operation are extracted through the cover.
Molten steel treated in Ladle Furnace is covered by a layer of desulfurizing
slag. The graphite electrodes are submerged into the slag, which protects
the ladle lining from overheating produced by the electric arcs. The arcs
are capable to heat the steel at the rate about 3C/min. During the treatment process argon is blown through the bottom porous plug providing
continuous metal stirring. Stirring results in distribution of heat produced
by the arcs, chemical homogenization and desulfurization of the steel by
the slag. Alloying elements are therefore added through the addition hopper.
Ladle Furnace primary functions are those to permit the secondary refining
of molten steel before to final cast, such as: temperature homogenization
or adjustment; chemical adjustments for carbon, sulfur, phosphorus, oxygen and precise alloying; inclusion control; degassing, and others.
The function of the porous plug is to provide inert gas stirring of the molten metal to promote homogenization. Normal stirring operations are performed by percolating argon gas through a purge plug arrangement in the
bottom of the ladle.
Vacuum degassing is also possible with the steel in a ladle, and argon
lances can be used to stir the steel to make the composition more homogeneous. Vacuum degassing produces ultra low-carbon steels, with carbon
contents as low as 0.002 wt%. Vacuum degassing also removes hydrogen
that can result in hydrogen flaking and porosity.
From Ladle Furnace to Casting Process
Whichever is the raw material used and relative patter route employed, either the iron-ore or recycled steel scrap, the ladle refining operations must
be completed in the Ladle Furnace (such phase, crucial for steel refining, is
also called ladle metallurgy).
From the ladle furnace station, the liquid steel is moved by ladle to be cast
to produce ingots or continuously cast long product (in a continuous casting machine).
Ingot Casting
During ingot casting, the ladle is moved by an overhead crane so that is
can be tapped or teemed into individual upright-standing molds on rail
339
cars. The metal can be poured into the mold either from the top of the
mold or from the bottom through a connecting channel. In the first case,
the steel is poured from the ladle directly into the mold (Figure 9.6a). After
the mold is filled, the ladle opening is closed and the ladle is moved by
crane to the next mold, where the process is repeated. In bottom pouring,
several molds (from two to 60) can be filled with steel simultaneously.
Here, the molds are mounted on a stool having channels lined with refractory brick. The steel from the ladle descends through the fountain into the
channels of the stool and then enters the mold from the bottom (Figure 9-6
b). The pouring method used depends on such factors as the steels grade
and weight and the intended use of the ingots.
(a)
(b)
Fig.9.6 Ingot casting through: (a) top-pouring and (b) bottom-pouring processes;
(1) ladle with metal, (2) mold, (3) stool, and (4) fountain.
Continuous Casting
Although ingot casting has been the traditional method, continuous casting
has rapidly evolved as the method of choice because of cost and quality
advantages.
In the continuous casting process (Fig. 9.7), the ladle of molten steel is
transported to an elevated casting platform above the casting machine. The
molten steel is poured into a rectangular trough, called tundish, which acts
as a reservoir for the steel. From a spout in the bottom of the tundish, the
molten steel is poured into a water-cooled mold with a movable temporary
bottom. As the molten steel enters the mold, the metal at the surface of the
mold solidifies, forming a thin skin. The skin thickens as the metal passes
through the mold, and the temporary bottom is slowly lowered to allow
340
metal to be continuously poured into the mold. The remaining metal in the
center of the ingot is solidified by spraying cold water onto the ingot as it
leaves the mold. The solid metal billet is pulled by rollers so that a long,
continuous steel slab is produced.
At the end of the machine, it is straightened and cut to the required length.
Fully formed slabs, blooms, and billets emerge from the end of this continuous process. The continuous casting process runs for days or weeks as
ladle after ladle of molten steel feeds the casting machine.
The advantages of the continuous casting process include reduced costs,
improved quality, increased yield, lower energy costs, and less pollution. It
is now the process of choice for high-volume, low-cost plain carbon steels.
Quality improvements include less variability in chemical composition,
both through the thickness and along the length of the continuously cast
slab. The surface quality of the slab is also higher than for an ingot, having
fewer surface defects such as seams and scabs.
The yield for continuous casting is also higher, since it is not necessary to
crop the ends of continuously cast slabs. Energy savings are achieved,
since the continuously cast slabs are sent directly to rolling mills and do
not require soaking pits for reheating. In addition, since the thickness of
continuously cast slabs is approximately half the thickness of individual
ingots, much less hot rolling is required.
Fig.9.7 Continuous steel casting, architecture and main equipment for the operations. For animation video describing all process steps from recycling scrap to
continuous casting click here.
341
342
As freezing progresses, the thermal gradient decreases, and this causes the dendrites to become very long. Breakdown of columnar growth may occur as a result
of fracturing of the very long dendrite grains by convection currents in the melt.
These broken arms can then serve as nuclei for new grains, as shown in Fig. 9.10
that refers to typical ingot product that solidifies.
343
especially in alloys with wide freezing ranges and a large mushy zone, it is particularly difficult to eliminate.
The primary casting microstructure is thus made of:
Large grains;
Non-equiaxic structure;
Non homogeneous chemical composition due to macrosegrations and microsegregations;
Gas porosities and shrinkage porosity.
As result, the mechanical properties of primary casting structure are:
Very low mechanical strength (low YS and UTS);
Very low toughness (low KV and KIc);
namely those features that do not allow products to be put in service directly cutting from ingots, blooms, slabs and billets.
344
345
346
347
348
(a)
(b)
(c)
Fig. 9.15 Example of bulk hot metalworking: a) rolling mill plant for plate rolling (click on figure or click here for video); b) forging of primary continuously
cast billet into rolls (click on figure or click here for video); c) ring rolling produced from slab (click on figure or click here for video).
349