Documente Academic
Documente Profesional
Documente Cultură
www.hoganas.com/pmc
Material
and Powder
Properties
Hgans Handbook for Sintered Components
Material
and powder
properties
Hgans Handbook for Sintered Components
Copyright Hgans AB
December 2013
0674HOG All rights reserved
Hgans Handbook for Sintered Components is intended for customer use.
The data presented in the handbook has been obtained from test specimens,
sintered under well-controlled conditions, in the Hgans AB laboratory.
Note that data established for any particular production equipment
or conditions may differ from those presented in this handbook.
All trademarks mentioned in this handbook are owned by Hgans AB,
Sweden and registered in all major industrial countries.
PM-SCHOOL HANDBOOK 1
Material Science
Production of Iron and Steel Powders
Characteristics of Iron and Steel Powders
PM-SCHOOL HANDBOOK 2
PM-SCHOOL HANDBOOK 3
Index
108
Material Science
1.1
1.2
1.3
1.4
1.5
1.6
1.7
MATERIAL SCIENCE
In crystalline solids, such as metals, the particle density is in the order of 1023 /cm3 ,
i.e. ten times higher than in liquids and amorphous materials. The particles
are homogeneously distributed and form a strict geometrical structure. See
Fig. 1.1d. Their mobility is restricted to very small vibrations around fixed
equilibrium sites. These vibrations increase with temperature. This situation is
called crystalline long-range order. The crystalline state possesses the lowest
internal energy of all states of matter and, therefore, is stable below a sharply
defined melting point. The above mentioned states are not specific to any
particular kind of material. In principle, but not always in practice, any type of
material can adopt any of these states.
a)
c)
b)
d)
Figure 1.1 Distribution of atoms in gases, liquids, amorphous and crystalline solids
(schematically);
a. Gas: statistical distribution, Number of atoms per cm3 = 1019, = free length of way.
b. Liquid: instable close-range order, Number of atoms per cm3 = 1022.
c. Amorphous solid: stable close-range order, Number of atoms per cm3 = 1022.
d. Crystal: crystalline order, Number of atoms per cm3 = 1023.
10
MATERIAL SCIENCE
Ionic bond. Neighbouring atoms exchange electrons and the so formed positive
and negative ions are pulled together electrostatically as, for example, in
NaCl-crystals.
Covalent bond. Partly empty electron shells of neighbouring atoms overlap so
that their electrons belong to both atoms. This leads to strong bonds between
the atoms because in the overlapping shells their electrons are in a lower state
of energy than in separate shells. One important feature of these bonds is that
they can exist between atoms of the same type between which there can be no
ionic bonding. Examples are H2 ,O2 and Cl2. One other important feature of these
bonds is that they act in preferential directions as, for example, in the tetrahedral
structures of the CH4-molecule and of diamond.
Metallic bond. This type of bond can be explained only on the basis of the
following quantum-mechanical principle. All atoms in the metal crystal form
one common band of electron shells to which each individual atom contributes
one electron. These electrons, not being bonded to individual atoms, can move
freely inside the entire metal crystal. They form a so-called electron gas.
According to an obsolete electrostatic theory, the negatively charged electron
gas presses the negatively charged metal ions closely together. But the correct
explanation is that the mentioned arrangement constitutes the lowest possible
quantum state of energy for the metal crystal and allows the atoms to be packed
in the closest possible way. To this kind of close packing, metals owe their good
plastic formability. The freely moving electrons are responsible for the high
electrical and thermal conductivity of metals and for the characteristic metallic
gloss.
The overwhelming majority of all chemical elements crystallize in metal
structures, approximately one third of them each in BCC-, in FCC- and in
HCP-structure.
11
12
MATERIAL SCIENCE
See Table 1.1. The geometrical characteristics of these particular structures are
described in the following paragraph.
Table 1.1. Crystal structures and lattice constants of some metals.*
Structure
Metal
BCC
-Fe
-Fe
1390
2,932
Mo
3,147
3,165
900
Ni
3,306
3,524
467
Cu
-Fe
Ratio of Lattice
Constants c/a
2,86
2,884
-Co
HCP
Lattice Constants
a) 10-10 m; c) 10-10m
Cr
-Ti
FCC
Temperature C
3,560
3,619
1000
3,651
Ir
3,839
Pt
3,923
Al
4,049
Au
4,078
Ag
4,086
Pb
4,950
-Co
2,506; 4,068
1,62
Zn
2,665; 4,947
1,86
-Ti
2,950; 4,679
1,59
Cd
2,979; 5,617
1,89
Mg
3,209; 5,210
1,62
(ideal)
1,633
c
Figure 1.2. Unit cell in a crystal
cell; lattice constants: a, b, c.
13
14
MATERIAL SCIENCE
LP
LP
LP-family
LP
LP-distance
LP-distance
LP
LP
LP
Choose a coordinate system with the origin at the corner of one unit
cell and with its axes parallel to the edges a, b, c of this unit cell.
Determine the intersections of the plane with the tree axes, x y z, of this
coordinate system.
Express, in units of a, b, c, the distances of these intersections from the
origin of the coordinate system: ma, nb, pc.
Take the reciprocals 1/m, 1/n, 1/p and find three natural numbers
related to one another in the same way as the reciprocals: 1/m : 1/n : 1/p
= h : k : l. Put the result between round brackets: (h k l).
Example: The plane, shown in Fig. 1.4, intersects the coordinate axes
at 2a, 4b and 1c respectively. The above procedure applied to this plane
yields m = 2, n = 4, p = 1. Thus, it follows: 1/2 : 1/4 : 1 = 2 : 1 : 4, and
the Miller indices of this plane are:
(h k l) = (214).
Lattice planes and their Miller indices play an important roll e.g. in X-ray
structural analysis. Each family of planes is responsible for specific X-ray
reflections from which the respective Miller indices can be calculated and
conclusions can be drawn as to the type of crystal structure.
Some significant lattice planes of the cubic crystal system and their respective
Miller indices are shown schematically in Fig. 1.5.
Figure 1.5. Miller indices for some important lattice planes in cubic crystals.
Not only lattice planes but also lattice directions can be described by Miller
indices. A lattice direction is defined as the direction of a local vector
r = u a + v b + w c, originating from the corner of a unit cell, a, b, c being the
edge vectors of the cell (unit vectors of the coordinate system x, y, z).
The statement of the three coordinates of the vector in square brackets [uvw]
definitely identifies a given lattice direction. In the cubic system (and only
15
16
MATERIAL SCIENCE
there), a lattice direction [uvw] is always perpendicular to the lattice plane (hkl)
of same indices (h = u, k = v, l = w). The coordinates of the local vector r, put in
double square brackets [[uvw]], specify a lattice point. Some lattice directions
and lattice points are indicated in the schematic drawing in Fig. 1.6.
BCC
r
ar = 4 r / 3
ar = 4 r / 2
NA = 2
NN = 8
PD = 68%
FCC
NA = 4
NN = 12
PD = 74%
HCP
NA = 6
NN = 12
PD = 74%
d= 2r
FCC
B
A
h = d 3
Figure 1.7. Unit cells of some important crystal structures: body-centered cubic (BCC); face
centered cubic (FCC); and hexagonal close-packed (HCP). NA = number of atoms/elementary
cell; NN = number of nearest neighbours; PD = packing density.
From the unit cells presented in Fig. 1.7, three significant parameters can
be derived, viz. the number of atoms per unit cell, NA, the number of nearest
neighbours, NN, and the packing density PD. The number of atoms, NA, is easily
obtained in the following way:
17
18
MATERIAL SCIENCE
FCC
FCC
A
C
B
A
C
B
A
HCP
HCP
A
B
A
B
A
B
A
FCC-Twinn
FCC-Twinn
A
B
C
A
C
B
A
Figure 1.8. Arrangement of atoms in a close-packed plane (a), and in successive closepacked planes (b) and (c). Below: stacking sequences of close-packed planes in FCC and
HCP crystals.
The diagram in Fig. 1.8 (b) shows an element of the stacking sequence, a layer
B laid on a similar layer A in the most closely packed way. The layers have the
same orientation and each B atom rests symmetrically in a hollow provided by
three adjoining A atoms. We notice that only one-half of the hollows are used
by B atoms and that an equally close-packed arrangement would result if we had
used the other hollows, i.e. the C positions in diagram (c), instead.
Any stacking sequence chosen from the A, B and C positions is fully closepacked provided positions of the same type are not used by neighbouring planes.
For example, ABACBABAC is close-packed but ABBAAACCBBC is not.
19
20
MATERIAL SCIENCE
All metals used in technical applications (apart from single crystals and whiskers)
have a polycrystalline structure, i.e. they are composed of many randomly
oriented small crystal grains agglomerated to one another. When cooling down
a molten metal below its melting point, at first, very small dendrites (crystal
nuclei) precipitate from the melt. These dendrites gradually develop into small
crystallites which randomly oriented float around in the melt. Independently of
one another, the floating crystallites continue to grow, gradually depriving the
melt of atoms. Eventually, they get so crowded in the diminishing melt that they
begin to agglomerate. See Fig. 1.9.
Because of their original random orientation, a misfit between crystal lattices
occurs at the boundaries of adjoining crystallites (or crystal grains). In these
grain boundaries, atoms are not quite so closely packed as in the undisturbed
lattice inside the crystal grains. On a polished and etched metallographic section,
grain boundaries usually appear as thin dark lines.
Apart from grain boundaries, there are several other kinds of lattice
disturbances in real crystals. One distinguishes between zero-, one-, two-, and
three-dimensional lattice disturbances.
Figure 1.9. Formation of a polycrystalline
structure during solidification of a liquid
metal (schematically);
left: nucleation period (dendrites floating
in the melt);
right: end of solidification
( agglomerated crystallites).
2
1
Figure 1.10.
Zero-dimensional lattice
disturbances:
(1-1) vacancy,
atoms.
Figure 1.11. One-dimensional lattice disturbance: (a) edge dislocation, (b) screw dislocation;
in both cases, the dislocation line is perpendicular to the drawing plane.
21
22
MATERIAL SCIENCE
Dislocations play the most essential part in plastic deformation of metals. Under
the influence of shearing-stresses, dislocations are created, and slip occurs on
certain lattice planes (mainly (111)-planes). On these slip planes, dislocation
lines move like waves along energetically favorable directions (mainly
[111]-directions).
This kind of slipping is facilitated by the circumstance that bonds between
adjacent atoms on both sides of the slip plane are not broken all at the same time
but successively one after the other. See Fig. 1.13.
Figure 1.14. Dislocation line, stretching from edge dislocation (a) to screw dislocation (b),
travels through a crystal, shifting its upper half relative to its lower half by atomic distance.
23
24
MATERIAL SCIENCE
Figure 1.15. Obstacles hampering the movement of dislocations; Top left: dislocations pile
up against grain boundaries (schematically); Top right: dislocation lines traveling from left to
right catch on with inclusions (schematically); Bottom: single dislocations (black lines) and
dislocation tangles (diffuse black spots) ina-iron with 0.91 at.-% Cu, plastically deformed by
4%, 50 000:1 replica.*
* E. Hornbogen, Symposium: Steel-Strengthening Mechanisms, Zurich, May 5th and 6th 1969.
25
26
MATERIAL SCIENCE
Figure 1.17. Interstitial diffusion (a) and vacancy diffusion (b); (schematically).
Diffusion in Metals
(1.1)
The proportionality factor D is called diffusion coefficient and is usually
measured in cm2 s-1. Equation (1.1) applies only if the concentration gradient
-dc/dx does not change during the whole diffusion process (stationary
case). Such is approximately the case, for instance, when carbon from a hot
carbonaceous gas diffuses through the thin steel wall of a furnace muffle. On
its inside, the muffle is being carburized by the carbonaceous gas, on its outside
decarburized by the surrounding air. As a consequence, a constant gradient of
carbon concentration establishes itself in the wall and carbon diffuses through
the muffle wall at constant rate.
If, in the course of diffusion, concentration changes everywhere in the material
(non-stationary case), Ficks second law of diffusion applies:
(1.2)
In cases where the diffusion coefficient D can be considered being independent
of the concentration c, this equation is simplified to:
(1.3)
Given the concentration distribution at the beginning of the diffusion process,
this partial differential equation can be solved by means of Gau error function
erf ( ). A relatively simple particular solution, applicable to many practical cases,
is presented in the following paragraph.
27
28
MATERIAL SCIENCE
In many technical applications, the surfaces of two different metals are in such
intimate contact with one another that they can exchange atoms. Such is the case
e.g. when iron surfaces are protected with thin layers of zinc, copper, nickel or
chromium, or in powder metallurgy where powder particles of different metallic
identity are intimately pressed together in a compacting process.
In order not to confuse the understanding with complicated geometrical
parameters, we choose a solution of equation (1.3) applying to the simple
geometrical arrangement shown in Fig. 1.18: The end-faces of two oblong
metal bars, A and B, are closely pressed together. This simple model illustrates
the gradual leveling of concentration between two different metals which is
significant of all diffusion processes. The concentration distribution at the
beginning of the diffusion process is extreme, viz. 100% A-atoms in the left bar
and 100% B-atoms in the right bar. For this model, the particular solution of
equation (1.3) is:
(1.4)
where c (x,t) is the concentration of A-atoms in B at time t and distance x from
the contact surface between A and B. A large variety of numerical values of
Gau error-function erf ( ) can be found in relevant tables of mathematical
textbooks.
Diffusion in Metals
Figure 1.18. Course of diffusion between two bars of metals A and B (schematically).
Assumed are initial concentrations of 100% A-atoms in the left and 100% B-atoms in the
right bar. The curves are drawn according to equ. (1.4) and represent the distribution of
A-atoms across the two bars after different periods of time.
29
30
MATERIAL SCIENCE
D = D0 exp ( Q / RT ) (1.5)
D0 is a characteristic property of the metal, R is the universal gas constant, T is
the absolute temperature, and Q is the activation energy.
D0 depends on the atomic weight of the diffusing atoms and on the melting
temperature of the metal in which diffusion takes place. In the case of vacancydiffusion, the activation energy is the sum of formation energy and migration
energy of vacancies. In the case of interstitial-diffusion, the activation energy is
identical with the traveling energy of interstitial atoms. The respective travelling
energies of atoms are lower in grain boundaries and lowest on surfaces.
Plotting the logarithm of D from equation (1.5) over 1/T , one obtains a straight
line with the slope Q/R. See Fig. 1.19. Thus, from the slope of this straight line,
the activation energy of the diffusion process can easily be calculated. Deviations
from the straight line indicate that several kinds of atomic movements with
varying activation energies are involved in the diffusion process.
Diffusion in Metals
Temp.
14
11
9 x 100 C
10 -10
Surface
D, m2/s (log-scale)
10 -12
Grain Boundary
10 -14
Volume
10 -16
-Fe
10 -18
9 x 10 -4
1/T (1/K)
Figure 1.19. Temperature dependence of the coefficients of self-diffusion of iron, for diffusion
along surfaces and grain boundaries and inside the crystal volume (In D over1/T).
31
32
MATERIAL SCIENCE
Diff.
Atom
HostCrystal
Fe
-Fe
(BCC)
2,0 x 10 -4
241
57,5
500
900
1,1 x 10-20
3,9 x 10-15
Fe
-Fe
(FCC)
5,0 x 10 -5
284
67,9
900
1100
1,1 x 10-17
7,8 x 10-16
-Fe
6,2 x 10 -7
80
19,2
500
900
2,3 x 10-12
1,6 x 10-10
-Fe
1,0 x 10 -5
136
32,4
900
1100
9,2 x 10-12
7,0 x 10-11
Cu
Cu
7,8 x 10 -5
211
50,4
500
4,4 x 10-19
Zn
Cu
3,4 x 10 -5
191
45,6
500
4,3 x 10-18
Al
Al
1,7 x 10 -4
142
34,0
500
4,1 x 10-14
Cu
Al
6,5 x 10 -5
135
32,3
500
4,8 x 10-14
Mg
Al
1,2 x 10 -4
131
31,2
500
1,8 x 10-13
Cu
Ni
2,7 x 10-5
255
61,0
500
1,5 x 10-22
D0 (m2/s)
T (C)
D (m2/s)
1.4.1 Thermodynamics
E + PV = TS + G (1.6)
Thus, Gibbs free energy is given by the expression:
G = E TS + PV (1.7)
33
MATERIAL SCIENCE
The expression:
F = E - TS (1.7a)
is called (Helmholtz) free energy. The differential of (1.7), constitutes the
change dG of free energy:
Free Energy G
34
Solid
Liquid
Solid
Liquid
m.p
Figure 1.20. Change of free energy with temperature for a pure metal.
Temperature
The free energies of the liquid and of the solid state of a pure metal are plotted as
functions of temperature. The two respective curves cross at the metals melting
point. Below the melting point, the free energy of the solid state is lower than
the one of the liquid state. Above the melting point, the relation between the two
free energies is reversed.
For a metal system as defined in the heading to this paragraph, the free energies
GL(c) and GS(c) of the liquid and of the solid phase are plotted, as functions of
concentration and temperature, in a schematic diagram as shown in Fig. 1.21.
The melting point of the components A and B are denoted TMA and TMB (> TMA )
respectively. In the diagram, we distinguish between the following temperature
ranges:
35
36
MATERIAL SCIENCE
a) T > TMB
Over the entire range of concentration, the free energy curve GL(c) of the melt
lies below the free energy curve GS(c) of the solid. Over the entire range of
concentration, only the liquid phase is thermodynamically stable.
b) T = TMB
For c = 1 (100 %B), the free energy of the melt and the free energy of the solid
are equal: GL(c) = GS(c).
c) TMA < T< TMB
In this temperature range, the two free energy curves cross, and each curve
exhibits a minimum. The equilibrium of the two phases lies on the common
tangent of the two curves in accordance with equation (1.10). The abscissas of
the tangent points cL and cS correspond to the concentration of B in the melt and
in the solid phase respectively that are in thermodynamic equilibrium.
d) T < TMA
In this temperature range, over the whole range of concentrations, the free
energy curve of the solid lies below the free energy curve of the melt, viz. in
this temperature range, only the solid phase (composed of A- and B-atoms) is
thermodynamically stable.
Thus, the ranges of thermodynamic stability of the various phases in the binary
T-c-diagram (phase diagram) are separated by two theoretically deduced phase
borderlines (Liquidus and Solidus).
T>TMB
GS
GL
a)
Free Energy of Phases G(c)
T=TMB
GS
GL
b)
TMB>T>TMA
GS
GL
CL
c)
CS
TMA>T
GL
GS
d)
different temperatures.
Temperature T
TMB
Liquidus
Solidus
TMA
A
e)
Concentration c
37
MATERIAL SCIENCE
Li
Temperature T
Temperature T
38
So
A
a)
Time t
c3
c2
c1
Concentration c
b)
Figure 1.22. Experimental derivation of a phase diagram by means of thermal analysis, Left:
cooling curves for the two pure metals and for three of their alloys. Right: phase diagram
derived from cooling curves.*
Eutectic Systems.
* W. Schatt: Einfhrung in die Werkstoffkunde, VEB Deutscher Verlag fr Grundstoffindustrie, Leipzig 1981.
In this system, two components (metals) A and B are completely miscible in the
liquid state but only to a limited degree in the solid state.
Temperature
Temperature
Figure 1.23. Experimental derivation of the phase diagram for a eutectic system of two
components A and B with limited miscibility in the solid state.
39
40
MATERIAL SCIENCE
solid phases exist. To the right of the binary diagram, the cooling-curves for
three different alloys 1, 2, 3 are shown as obtained by thermal analysis. On the
cooling-curves for alloys 1 and 2, the eutectic reaction shows up in the form of
sharp kinks. Below the binary diagram, microstructures of the solidified alloys
are shown (schematically).
Alloy 1 of concentration cE (eutectic alloy) remains liquid until it passes the
eutectic line. Then, it spontaneously precipitates two different solid phases
and with concentration c and c respectively. The separation of the melt takes
place according to the eutectic reaction L + . The ratio of the relative
amounts of and in the microstructure of the solidified alloy is exactly equal
to the ratio of distance E - M2 and distance E - M1 on the eutectic line.
This relationship follows from a law called lever-rule of phases which will be
explained in paragraph 1.4.3.
Alloy 2 begins to solidify when passing the borderline between the melt (L)
and the (L+)-area. As temperature falls further, -crystallites of gradually
increasing A-concentration precipitate from the melt, depriving it of B-atoms,
according to the reaction L. As the alloy passes the eutectic line, the residual
melt spontaneously solidifies, precipitating -crystallites of concentration c
and -crystallites of concentration c, according to the eutectic reaction L+.
Also in this case, the above-mentioned lever-rule of phases applies.
Alloy 3 begins to solidify in a similar fashion as alloy 2, gradually reducing the
amount of the remaining melt. But on arrival at the borderline between the (L+)area and the ()-area, the residual melt has disappeared and the alloy is completely
solidified as -phase. On further cooling, the alloy passes the borderline between
the ()-area and the ( + )-area and is now tending to establish an equilibrium
between -phase and -phase according to the reaction + . Consequently,
-crystallites precipitate along the grain boundaries of the solid -structure.
Binary eutectic systems are formed e.g. by the following pairs of metals: Pb-Sn
(soldering alloys), and Cu-Ag (brazing alloys). In both cases, the eutectic allows
to make alloys that melt at considerably lower temperatures than any of the pure
metals they are composed of. Fig. 1.24 shows the binary system Pb-Sn and two
typical microstructures.
In the microstructure at left, the pre-eutectic Pb-rich -phase stands out as small
potato-shaped dark areas embedded in the subsequently formed eutectic. The
microstructure at right exhibits a plain eutectic with its typical arrangement of
alternating lamellae of - and -phase.
Pb Sn
327C
Liquid (L)
Temperature C
300
232C
+L
200
+L
187C
19,2
97,5
100
0
0
20
40
Dark
(19,2% Sn)
60
80
100% Sn
Light
(97,5% Sn)
Figure 1.24. Phase diagram of the system Pb - Sn. Microstructures: -phase embedded in
eutectic (left) and pure eutectic (right).
41
MATERIAL SCIENCE
Peritectic Systems.
Concentration
Temperature
Temperature
42
Time
Figure 1.25. Experimental derivation of the phase diagram of a peritectic system A-B
Eutectoid Systems.
Guided by the simple binary diagram for two components A and B as shown in
Fig. 1.26, this rule is readily explained.
Liquid
TB
m0
mL
P0
mS
y
Figure 1.26. Lever-rule of phases;
mL = amount of residual melt with
concentration cL.
TA
Solid Solution
CL
C0
Concentration of B in A
CS
43
44
MATERIAL SCIENCE
m0 = mL + mS
Lever-rule of phases:
m0
mL
m S
c0m0
cLmL
cSmS
mLx = mS y
The lever-rule of phases expressed in words: (amount of residual melt) x (tieline from balance point to liquidus line) = (amount of precipitated solid phase)
x (tie-line from balance point to solidus line).
Although the concentration of B-atoms in the core of each crystal grain is higher,
and in its outermost zone lower, than the original concentration of the melt, the
average concentration of the alloy has not changed during solidification.
Concentration of A in B
Liquid
T1
Temperature
TB
C0
Tn
TA
Solid Solution
Concentration of B in A
Cn
C0
C1
segregation;
c0= initial concentration
of the melt;
c1 = concentration of
dendritic crystal nuclei
first precipitated from
the melt;
cn = concentration of
residual melt solidifying
last.
45
46
MATERIAL SCIENCE
This binary diagram applies to the equilibrium states of carbon-steel and castiron. The term carbon-steel refers to iron-carbon alloys containing less than
approximately 2% C. The term cast-iron refers to alloys containing between 2
and 4% C where the Fe-C-eutectic occurs. Of interest for technical applications
is mainly the iron-rich side of the diagram stretching to the concentration of the
inter-metallic phase Fe3C (6.67% C).
This part of the diagram is shown in Fig. 1.28 and can be looked at as being
composed of three sub-diagrams: one peritectic, one eutectic and one eutectoid
diagram. In connection with the heat treatment of steel, only the eutectoid subdiagram is of importance. The range up to temperatures of approx. 1150 C and
up to concentrations of approx. 1.5% C is of particular interest to the powder
metallurgy of iron.
Pure iron melts at 1536 C. Below this temperature, it solidifies in BCC crystal
structure, called -iron, which can dissolve max. 0.10% carbon interstitially (at
1493 C). On further cooling, at 1392 C, -iron transforms to FCC crystal
structure, called -iron, which can dissolve max. 2.06% carbon interstitially (at
1147 C). Carbon-containing -iron is called austenite. At 911 C, pure -iron
transforms again to a BCC structure, which is called -iron or ferrite, and which
is ferro-magnetic below 769 C (Curie-point). Ferrite can dissolve max. 0.02%
carbon (at 723 C).
Iron together with carbon forms the inter-metallic phase Fe3C, a carbide called
cementite, which has an ortho-rombic structure and a carbon content of 6.67%.
Cementite is a so-called meta-stable phase because, when annealed for long
times at high temperatures, it dissembles into iron and graphite.
Depending on whether carbon occurs bound in cementite or as free graphite,
one speaks of the meta-stable or of the stable system. Transformations in steel
and in white cast-iron proceed according to the meta-stable system (fully drawn
lines). Transformations in graphite-containing gray cast-iron proceed according
to the stable system C9D91.
An austenite containing less than 0.8% C transforms partly to ferrite when it
passes the phase-borderline GS (also called A3), and below the eutectoid line
PS (also called A1), its residue separates into ferrite and cementite, forming
Figure 1.28. Equilibrium phase diagram of the system Iron-carbon: fully drawn lines =
metastable system (Fe-Fe3C), broken lines = stable system (Fe-graphite).
47
48
MATERIAL SCIENCE
Figure 1.29. Microstructure of steel; a. Austenite (stainless steel), b. Ferrite (< 0.02% C),
c. Ferrite + Pearlite (0.30% C), d. Pearlite + Ferrite (0.60% C), e. Pearlite (0.85% C),
f. Pearlite + grain boundary cementite.*
The iron-carbon diagram as shown in Fig. 1.28 like the binary diagrams treated
further above pertain to states of equilibrium and can, in principle at least, be
deduced theoretically from classical thermodynamics. With cooling rates as usually
practiced in the steel-producing and steel-working industry, states of equilibrium are
not always easily achievable and, in many cases not even desirable.
On the contrary, it is just the aimed provocation of certain meta-stable states that
offers outstanding technical advantages. With their corresponding microstructures,
the value scale of properties like tensile strength, elongation, hardness and
toughness can be largely extended. By varying the additional parameters time
and cooling-rate, iron-carbon-based materials can be optimally adapted to a great
variety of applications.
At the present state of development of thermodynamic theories, it is still not
possible to quantitatively describe the states of non-equilibrium as occurring
during forced cooling of steels. Therefore, empirically-experimentally established
transformation diagrams are still the only basis on which heat treatments of steel
can be intelligently planned.
Such transformation diagrams, known as TTT-diagrams (for Time, Temperature,
Transformation), exist today for all common types of steel and can be found in
brochures made available by the steel producers. In paragraph 1.6, these diagrams
will be discussed in detail.
Transformations in the solid state require rearrangement of atoms by diffusion and
in some cases also certain mechanisms of shearing the crystal lattice. Since diffusion
is a time- and temperature-dependent process, the attainment of equilibrium is
the more retarded the more the steel is undercooled. At sufficient degrees of
undercooling, meta-stable and instable crystal structures (and corresponding
microstructures) occur which considerably deviate from equilibrium structures.
With increasing cooling rate, the phase-borderlines in the equilibrium diagram:
iron-carbon (Fig. 1.28), are moving towards lower temperatures, particularly the
borderlines GSE (A3 and Acm) and PSK (A1) of the austenite area. The diagram
in Fig. 1.30 shows the displacement of the A3-point and the of the A1-point with
increasing cooling rate for a steel with 0.45% C. As can be seen from this diagram,
the A3-point moves faster towards lower temperatures than the A1-point.
As a consequence, with increasing cooling rate, the formation of pre-eutectoid
ferrite is more and more reduced and, eventually, entirely suppressed as A3 and
A1 join in a common A-point. At still higher cooling rates, Bainite (AB) and
Martensite (MS) are formed.
49
50
MATERIAL SCIENCE
Figure 1.30. Influence of cooling rate on the position of the transition points A3 and A1 for a
steel with 0.45% C.*
The influence of cooling rate on the displacement of A3 and the A1, as just
described for a steel with 0.45% C, applies analogously to the entire eutectoid
part of the iron-carbon diagram, i.e. to all plain carbon steels.
The diagram in Fig. 1.31 illustrates the situation. Pertaining cooling rates are
stated along the right ordinate of the diagram. With reference to cooling rates,
one distinguishes between the following five steps of undercooling:
Undercooling step I.
In this so-called bainite step (below approx. 450 C), the diffusion of iron atoms
has practically ceased and a formation of pearlite is no longer possible. Only
the interstitial carbon atoms can still diffuse in this temperature range. Already
before the FCC austenite lattice shears into the BCC ferrite lattice, cementite
begins to precipitate. The resulting microstructure consists of finely dispersed
cementite particles in a matrix of acicular ferrite. It is a kind of dispersionhardened ferrite, so to speak. The orientation of the ferrite needles is related to
the initial austenite lattice.
1000C
1 Ks-1
15 Ks-1
40 Ks-1
900
800
700
50
100
150
200
250
300
400
500
Temperature
600
II Pearlite formation whithout recalescence
500
400
Ks-1
300
200
IV Martensite formation
100
0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1,6
1,8 %
Carbon Content
Figure 1.31. Influence of cooling rate on the position of phase boundary lines in the
eutectoid part of the Fe- Fe3C - diagram.
Cooling rate
Equilibrium
51
52
MATERIAL SCIENCE
Figure 1.32. Influence of carbon content on the axis lengths of the tetragonal martensite
lattice.*
* K. Honda and Z. Nihiyama.
Figure 1.33.
Tetrahedral (a)
and octahedral (b)
interstices in the FCC
lattice of austenite.
= possible sites of
carbon atoms.
a
Ferrite Formation.
Pure -iron does not contain any carbon atoms at all. In hypoeutectoid steels
(< 0.8 weight-% = 3.7 at.-%), at best, only one out of seven unit cells of austenite
contains one carbon atom. Between 911C and 723C, because of the sparse
distribution of carbon atoms, parts of the FCC austenite lattice can easily shear
into the BCC ferrite lattice, while carbon atoms, not dissolvable in ferrite, get
enriched in the residual parts of the austenite. This shearing mechanism requires
a minimum of rearrangement of iron atoms in the respective crystal lattices,
53
54
MATERIAL SCIENCE
as becomes evident from the schematic diagram in Fig. 1.34. In the drawing
representing two adjacent FCC cells, certain iron-atoms (filled circles) are
connected by tie-lines in such a way that they form a body-centered-tetragonal
(BCT) cell. It does not take much imagination to comprehend that a slight
stretching of its horizontal and/or slight upsetting of its vertical edges turns
this BCT cell into an ordinary BCC cell. The theoretically possible positions of
carbon atoms are indicated by small black spots on the edges of the cell. Since
the described shearing mechanism occurs almost instantly, one could say that,
on cooling, the FCC lattice of austenite snaps into the BCC lattice of ferrite
(or vice versa on heating).
Cementite Formation.
The maximal solubility of carbon in austenite is 2.06 wt.-% = 9.51 at.-% and
occurs at 1147C (eutectic temperature). This means that scarcely twenty
percent of all theoretically possible positions are filled with carbon atoms (i.e.
less than one carbon atom per austenite cells). A denser filling with carbon
atoms (supersaturation) would increase internal stresses in the austenite lattice
to such extent, that the corresponding increase in elastic energy would exceed
the formation energy of cementite.
Consequently in case of supersaturation with carbon, the austenite tends to
lower its free energy by transforming to cementite. The tendency increases with
increasing degree of undercooling. The concentration of carbon in cementite is
25 at.-%, i.e. at least close to three times higher than in austenite. Thus, in order
to form cementite, considerable amounts of carbon atoms must be adequately
displaced by diffusion.
Not enough with carbon diffusion, iron atoms too must be rearranged by
diffusion in order to enable the austenite to transform partly into cementite. Thus,
the formation of cementite depends not only on the austenites opportunity to
lower its free energy but also on the actual mobility of the carbon atoms (which
decreases with increasing degree of undercooling). At temperatures where the
mobility of carbon atoms in the austenite lattice has ceased, cementite cannot be
formed, and martensite occurs instead. See further down.
Pearlite Formation.
Figure 1.35. Nucleation of pearlite colonies at austenite grain boundaries (left) and growth of
pearlite lamellas in austenite grains (right).*
Martensite Formation.
55
56
MATERIAL SCIENCE
the remaining austenite shears into the ferrite lattice. At very high cooling
rates, as in quenching procedures, carbon atoms cannot move fast enough to
avoid getting trapped inside the spontaneously forming ferrite. Since ferrite,
under equilibrium conditions, cannot dissolve more than 0.02 wt.-% = 0.09
at.-% carbon, it is now extremely supersaturated with carbon, and its lattice
correspondingly (tetragonally) distorted (ref. Fig 1.32).
The high internal stresses arising from this distortion are the cause of the high
hardness of martensite. In microstructures, the tetragonally distorted ferrite, i.e.
martensite, appears in the form of relatively coarse needles oriented at certain
angles inherited from the initial austenite lattice.
The small areas in between the needles are remaining austenite. See micrograph
in Fig. 1.36 a. When annealing martensite at temperatures above MS, finely
dispersed cementite precipitates from the supersaturated ferrite. See micrograph
in Fig. 1.36 b. Thus, the high level of internal stresses in the tetragonal ferrite
lattice is reduced, and the brittleness of the martensite correspondingly lowered.
Figure 1.36. Acicular martensite (a) and tempered martensite with finely dispersed
cementite (b).*
Bainite formation.
* K.J. Irvine, Symposium: Steel-Strengthening Mechanisms, Zurich, May 5th and 6th 1969.
In the case of upper bainite, small parallel ferrite platelets, nucleated by shearing
processes, grow gradually inside the austenite. Moving ahead of the growing
ferrite platelets, carbon atoms accumulate in the remaining austenite which gets
increasingly more restricted. Eventually, the remaining austenite becomes so
restricted and so much enriched in carbon that it transforms to cementite jammed
between the ferrite platelets. See micrograph in Fig. 1.37a.
In the case of lower bainite, ferrite platelets form in much the same way as
in upper bainite but faster. At the same time, the diffusion rate of carbon has
dropped so much that carbon atoms cannot move fast enough to avoid getting
trapped inside the fast growing ferrite platelets. In this respect, the mechanism
of lower bainite formation is quite similar to the mechanism of martensite
formation.
But in the case of lower bainite, temperatures are high enough to initiate
precipitation of finely dispersed cementite particles inside the ferrite platelets
immediately after they have been formed. See microstructure in Fig. 1.37 b.
One could say that lower bainite is a kind of self-tempered martensite. Indeed,
comparing Fig. 1.36 b with Fig. 1.37 b, it can be seen that the microstructures of
lower bainite and tempered martensite have a certain resemblance.
Figure 1.37. Lower (a) and upper (b) bainite in a bainitic steel (0.87% C; 0.44% Mn; 0.17% Si;
0.21% Cr; 0.39% Ni).*
* I.J.Habraken and M.Economopoulos, Symposium: Transformation and Hardenability in Steels, Ann Arbor,
Michigan, USA 1958.
57
58
MATERIAL SCIENCE
Degree of Transformation
Nucleation
Growth
Time
Figure 1.38. Typical s-shaped course of isothermal transformation passing from a period of
nucleation to a period of growth.
1.6.1 ITT-Diagrams
59
60
MATERIAL SCIENCE
Figure 1.39. Empirically derived ITT-diagram for a hypoeutectoid carbon steel (right) and its
relation to the equilibrium phase diagram Fe-Fe3C (left). Austenitizing above A3.
A = austenite, F = ferrite, P = pearlite, B = bainite.*
61
C
800
Hardness - RC
MATERIAL SCIENCE
F
A
1400
A0 3
A0 1
13
700
1200
24
600
Temperature
62
500
400
31
1000
A+F
A+P
26
30
800
300
600
200
400
37
B
MS
M 90
M 90
44
50%
50
ITT DIAGRAM
100
200
1 min.
32
0.5 1 2 5 10
1 hour
10 2
10 3
1 day
10 4
10 5
62
10 6
Time - seconds
Figure 1.40. ITT-diagram with pronounced bainite-nose for a steel with 0.42% C;
0.68% Mn; and 0.93% Cr.*
Figure 1.41. Cooling curves for surface and center of a small (fully drawn lines) and of a big
(dotted lines) work piece (schematically).*
63
64
MATERIAL SCIENCE
Temperature
Temperature
Temperature
Temperature
pearlitic structure is to be obtained, advantage may be taken of the timetemperature coordinates of this minimum to design a short annealing cycle. This
is accomplished by cooling the steel initially in the austenitic state as rapidly
as convenient to the temperature of the minimum and holding it approximately
at this temperature for the time required to transform the austenite completely.
Subsequently, the steel may be cooled as convenient.
Figure 1.42. Four standard procedures for the heat treatment of steels: a. Quenching and
Tempering; b. Martempering; c. Austempering; d. Isothermal Annealing.*
65
66
MATERIAL SCIENCE
These examples have shown that ITT-diagrams are helpful guides in designing
adequate heat treating procedures. In industrial practice, heat treatments rarely
proceed under strictly isothermal conditions. In many cases, continuous cooling
procedures are even more practical. Thus, one might think that CCT-diagrams
would be more adequate. But the derivation of CCT-diagrams is a rather tedious
task, and would, even if feasible, rarely be warranted since a particular CCTdiagram exactly presents but one sample; samples from other heats, or even
from other locations of the same heat, are likely to have slightly different
CCT-diagrams.
Thus, when used with discrimination and with its limitations in mind, the
ITT-diagram is most useful in interpreting and correlating observed phenomena
on a rational basis, even though austenite transforms during continuous cooling
rather than at constant temperature.
67
68
MATERIAL SCIENCE
See diagrams in Fig. 1.45. In mixed structures of pearlite and ferrite, strength
and hardness increase with increasing proportion of cementite, while toughness
and ductility decrease. See diagram in Fig. 1.46.
Figure 1.45.
Figure 1.46.
Figure 1.45 Influence of carbon content on hardness (a) and ductility (b) for a plain carbon steel
having fine and coarse pearlitic as well as spheroiditic microstructures.*
Figure 1.46. Influence of carbon content on the mechanical properties of a plain carbon steel
having a fine pearlitic microstructure.**
Martensite is the hardest and most brittle of all phases occurring in microstructures
of steel. It has virtually no ductility at all. Its hardness increases with increasing
carbon content - first faster (up to approx. 0.4% C) then slower, approaching
a maximum at approx. 1% C. The diagram in Fig. 1.47 shows the influence of
carbon content on the hardness of martensite and pearlite.
Martensite owes its high hardness not to the presence of cementite, as is the
case with pearlite, but to the high internal stresses caused by its supersaturation
with carbon. In its heavily distorted crystal lattice, the number of sliding systems
available for plastic deformation is extremely low.
Figure 1.47. Influence of carbon content on hardness for plain carbon martensitic and fine
pearlitic steels.*
Since austenite has a higher density than martensite, a noticeable volumeincrease occurs on quenching. Due to the accompanying shearing stresses, big
or intricately shaped work pieces may crack when quenched. This constitutes
a serious problem in the heat treatment of steels containing more than approx.
0.5% C.
69
70
MATERIAL SCIENCE
In the quenched state, martensite is too hard, too brittle and too crack-sensitive for
most applications. Its brittleness can be reduced and its toughness considerably
increased through tempering at temperatures between 250 and 650C. Internal
stresses in the martensite lattice dissolve already at approx. 200 C. During
tempering, carbon atoms have sufficient time to diffuse and form cementite
which precipitates in the form of finely dispersed particles inside the martensite
needles (ref. Fig. 1.36 b). Tempered martensite has a certain resemblance to
lower bainite (ref. Fig. 1.37 b). The size of the precipitated cementite particles
in martensite increases with tempering time and temperature.
By means of varying these two parameters, the properties of hardened steel
can be optimally adapted to many different applications. The diagrams in Fig.
1.48 and Fig. 1.49 illustrate these possibilities.
71
19
190
18
180
17
170
16
160
15
150
8
7
6
5
70
re
120
ng
th
60
110
100
Contra
ction
50
400
300
200
100
th
90
80
70
60
40
30
50
40
30
20
10
St
80
ng
ile
500
ss
re
10
oint
90
St
11
130
ns
600
ct
12
ld P
dne
pa
13
Yie
140
Har
Im
14
Te
100
0
100
20
Elongation 5
10
Elongation 10
200
300
400
500
600
0
700
Tempering Temperature, C
Hardness, HB
200
20
72
MATERIAL SCIENCE
Figure 1.49.
Hardness as
a function of
tempering time
and temperature
for a waterquenched eutec
toid plain carbon
steel.*
With some types of steel, tempering may cause a marked decrease of impact
strength. This phenomenon, which is called temper brittleness, occurs either
when the steel is tempered above 575 C and subsequently slow-cooled to R.T.,
or when tempered between 375 and 575 C. Steels containing Mn, Cr or Ni and,
in addition, small amounts of impurities, like Sb, P, As or Sn, are especially
sensitive to temper brittleness.
When the mentioned alloying elements and impurities are present, the border
between tough and brittle tempered martensite is shifted noticeably to higher
temperatures. Crack propagation in temper-brittle steel proceeds along former
austenite grain boundaries where the impurities preferrably have segregated
during solidification of the melt. Temper brittleness can be avoided by reducing
the amount of impurities or by tempering above 575 C or below 375 C with
subsequent quenching to R.T.
73
Production of Iron
and Steel Powders
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 The Hgans Sponge Iron Process . . . . . . . . . .
2.3 The Hgans Water-Atomizing Process . . . . . .
2.4 Hydrogen-reduced Iron Powder . . . . . . . . . . . . .
2.5 Alloying Methods . . . . . . . . . . . . . . . . . . . . . . . . .
2.6 Distaloy and Starmix . . . . . . . . . . . . . . . . . . . .
76
78
82
84
85
87
76
2.1 Introduction
Iron and steel powders for the manufacturing of sintered structural components
(including sintered porous bushings) are produced in many parts of the world. The
worldwide consumption of such powders has been growing increasingly fast over
the last four decades and reached 900,000 metric tons by the end of 2010.
Over the last forty years, the quality of iron and steel powders has been
continuously improved and the spectrum of available grades has been widened.
During the same period, compacting- and sintering-techniques have become
more and more sophisticated. This development has lead to a substantially
widened range of applications for sintered and iron steel parts.
Table 2.1. Global Sales of all Iron Powders
Year
Tons of powder
1991
450
1992
475
1993
500
1994
565
1995
590
1996
600
1997
650
1998
670
1999
725
2000
775
2001
715
2002
785
2003
820
2004
890
2005
855
2006
895
2007
930
2008
850
2009
680
2010
945
Introduction
Table 2.1 shows that the worldwide sales of iron powders have more than
doubled over the last 20 years. It must be mentioned that, at present (2011), the
worldwide production capacity of iron and steel powders is considerably larger
than the consumption. Thus, there is no risk of shortage for many years to come.
Currently, there are two basically different production methods which together
account for more than 90% of the world production of iron and steel powders,
viz. the Hgans sponge-iron process and the water-atomizing process. The
former process is based on reduction of iron ore, yielding a highly porous
sponge-iron which subsequently is comminuted to powder. The latter process
is based on atomization of a stream of liquid iron (or steel) by means of a jet of
pressurized water. Both processes will be described in detail further below.
In the manufacturing of sintered parts, iron powders are always used admixed
with a small amount of lubricant in powder form in order to minimize the friction
in the compacting tool. In many cases, they are also blended with alloying
elements in powder form, like graphite, copper, nickel, molybdenum and others
(in order to achieve increased strength properties).
Since powder blends tend to segregate when transported and handled, Hgans
has developed special blending processes in which the alloying additives are
safely bound to the iron powder particles. Powder mixes produced according
to these processes are known as the trade names Distaloy and Starmix. These
two processes are treated in detail further below.
77
78
79
80
The process starts with two raw materials: a reduction mix consisting of
coke breeze blended with ground limestone (1), and a pre-processed iron ore (2).
The iron ore and the reduction mix are being dried separately in two rotary
dryers (3). The slightly agglomerated dried reduction mix is crushed (4) and
screened (5), and the dried iron ore is passed through a magnetic separator (6).
Then, both materials are charged by means of an automatic charging device
into tube-like ceramic retorts as illustrated (7), (18), (19). These retorts have
an Inner Diameter of 40 cm, are 2 m long, and consist of four tube segments
of silicon carbide being stacked on top of each other.
These retorts are standing, 25 each, on rail-bounded cars which are clad with a
thick layer of refractory bricks. These cars are traveling slowly through a tunnel
kiln of approx. 260 m length (8) within which the retorts are gradually heated
to a maximum temperature of approx. 1200C. As the temperature inside the
retorts increases, the coke breeze begins to burn forming CO which, in turn,
begins to reduce the magnetite to metallic iron while itself oxidizing to CO2.
The so generated CO2 reacts with the remaining coke breeze forming new
CO, which again reduces more magnetite to metallic iron. This reaction cycle
continues until all magnetite has been reduced to metallic iron and the major part
of coke breeze is burned up. Parallel to the reduction cycle, the limestone in the
reduction mix binds the sulfur arising from the burning coke breeze.
After completed reduction, the retorts are slowly cooled down again to approx.
250C before leaving the kiln. Inside each retort, there is now a tube-like sponge
iron cake with a porosity of about 75%, a residue of unburned coke breeze, and
a sulfur-rich ash. At an automatic discharging station (9), the sponge iron tubes
are pulled out off and the remaining coke breeze and ash are exhausted from the
retorts. Thereafter, the retorts are ready to be charged again and go on a new trip
through the tunnel kiln.
The sponge iron tubes (after having been cleaned from adhering coke breeze
and ash) are in several steps crushed and comminuted to a particle size below
3 mm (10). The so obtained crude powder is intermediately ensilaged before
further processing. From the intermediate silo (11), the crude powder is passed
through a specially designed chain of magnetic separators (12), mills (13) and
screens (14), in order to be refined to a particle size below 150 m (<100 Tylermesh) and a well defined bulk density (apparent density).
81
82
1. Selected scrap
2. Arc furnace
3. Liquid Steel
4. Injection
5. Atomizing (see below)
6. Dewatering
7. Drying
8. Magnetic Separation
9. Screening
10. Equalizing
11. Transportation
5. ATOMIZING
A. Tundish
B. Steel Stream
C. High Pressure Water
D. Nozzle
E. Atomized Iron Powder
The raw material for this process is carefully selected iron scrap and sponge iron
from the process described in the preceding paragraph. This raw material (1) is
melted down in an electric arc furnace of 50 tons capacity (2) where, if desired,
alloying elements can be added.
The melt is teemed slag-free through a bottom hole into a ladle (3) where it is
refined with an oxygen lance (4). The ladle is then transferred to a ladle furnace
and on to the atomizing station (5), where the liquid iron (or steel) is teemed
slag-free through a bottom hole in the ladle into a specially designed tundish (A).
From there, the liquid iron (or steel) flows in a well controlled stream (B)
through the center of a ring-shaped nozzle (D) where it is hit by jets of highly
pressurized water (C). The stream of liquid iron (or steel) explodes into fine
droplets (E). Some of these droplets freeze immediately to small spheres, others
unite in small irregularly shaped agglomerates while freezing.
Air, swept along by the water jet, and water vapor arising in the atomizing
process, cause superficial oxidation of the small droplets. The solidified droplets
and the atomizing water are collected in a huge container, where they are settling
as a mud. This powder mud is de-watered (6) and dried (7).
The dry powder is magnetically separated from slag particles (8), screened (9)
and homogenized (10), and eventually transported in special containers (11) to
the works at Hgans for further processing.
In the state leaving the atomizing plant, the atomized powder particles are not
only superficially oxidized but also very hard because, due to the extremely
high cooling rates residing in the atomizing process, they have solidified in the
martensitic state despite their low carbon content.
The powder is, therefore, soft-annealed, and its surface oxides and residual
carbon are reduced in belt furnaces of the same type as described in the preceding
paragraph. Routines for homogenizing, quality checking, packaging and storing
are the same as for sponge iron powders.
83
84
Alloying Methods
Advantages:
Alloying elements do not segregate when the powder is handled .
Yield fully homogeneously alloyed sintered parts.
Disadvantages:
Have low compressibility, because their particles are solutionhardened. (See Figs. 2.6 and 2.7).
In order to change or correct the composition of a fully alloyed powder,
if ever so little, a new melt (usually 50 tons at time) will have to be
atomized.
85
86
Powder mixes.
Advantages:
Have higher compressibility. (See Fig. 2.8).
No additional mixing operation is required as the powder has to be
admixed with a lubricant anyway.
The composition of a powder mix can very easily be changed or
corrected by re-mixing it with additional amounts of either iron powder
or alloying elements.
Disadvantages:
Yield less homogeneously alloyed sintered parts, because the admixed
alloying elements (except carbon) diffuse very slowly in solid iron.
(ref. Chapter 6, Figs. 6.9 and 6.10).
Alloying elements tend to segregate when the powder mix is transported and handled. (However, powder mixes can be made segregationresistant by means of special treatments as described in the following
paragraph).
The diagram in Fig. 2.6 shows the influence of alloying elements on the hardness
of iron.
87
88
Figure 2.7. SEM-photographs showing fine particles of copper, nickel and molybdenum
diffusion bonded in the Distaloy process, to the surface of an iron powder particle.
The so treated powder mix contains the alloying additives as finely and evenly
dispersed as possible . The fact that the iron powder particles are locally prealloyed has practically no negative effect on the compressibility of the mix.
See Fig. 2.8, where the compressibility curve for a Distaloy grade is shown
in comparison with those for an ordinary powder mix and for an atomized
homogeneously alloyed powder of identical compositions.
Figure 2.8. Compressibility curves for three ferrous powders produced by different methods
but having the same chemical composition: 1.75% Ni, 1.5% Cu, 0.5% Mo, remainder Fe.
Graphite and lubricants have to be excluded from the Distaloy process because,
during heat-treatment of the powder mix, graphite would carburize the iron
particles and spoil the compressibility of the powder mix, and lubricants would
burn-off.
89
90
Figure 2.9. SEM-photograph showing fine graphite particles glued, in the Starmix process,
to the surface of an iron powder particle (NC100.24).
Loss of Graphite, %
The Starmix processes can, of course, be applied to ordinary iron powder mixes
as well as to Distaloy powders; in both cases, it yields segregation-resistant
press-ready powder mixes. The gain in product consistency achieved through
the Starmix 500 process is illustrated in the diagrams in Figs. 2.10 to 2.13.
Conventional
mix
Starmix 500
Figure 2.10. Graphite loss in air de-dusting test for a Starmix treated powder mix
Frequency, %
Conventional mix
Starmix 500
91
Combined carbon, %
92
Starmix 500
1.20
of carbon content in
sintered parts during
1.10
mass production,
comparing Starmix
1.00
Conventional mix
0.90
conventional mixes of
NC100.24 + 1.2%
5000
10000
15000
graphite.
Number of parts
32
30
28
26
24
Flow, sec/50 g
6.70
13
6.60
12
Elongation, %
400
4.0
Figure 2.13.
390
3.0
Comparison of powder
380
2.0
Hardness, HV10
Dimensional Charge, %
140
+0.20
130
+0.10
Conventional mix
Starmix 500
Starmix 400
Starmix 400 has tailored properties for tolerance control in more complex parts.
Key properties can be fine-tuned to meet specific requirements
Improved performance in more demanding applications
Starmix 820
Starmix BOOST
93
Characteristics
of Iron and
Steel Powders
3.1
3.2
General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . 96
Properties of Hgans Iron Powders . . . . . . . 100
96
1. Metallurgical properties
2. Geometrical properties
flow rate
bulk density
compressibility, green strength and spring back
All these powder properties are inherited from and specific to the process by
which the powder was produced. Some of them are interrelated with each other.
For instance:
microstructure and microhardness are depending on chemical
composition;
compressibility decreases with increasing microhardness,
increasing particle porosity and decreasing particle size;
coarser powders and powders of regular particle shape flow better
than fine powders and powders of irregular particle shape;
powders of irregular particle shape have better green strength after
compacting than powders of regular particle shape.
In the following, we present a brief definition of the aforementioned powder
properties and their relevance to processing steps in the production of sintered
parts.
General Aspects
Metallurgical Properties
are determined by chemical analysis and metallographic procedures. The chemical composition of a ferrous powder has a great influence upon the final strength
properties of the sintered parts. Non-metallic impurities may have an adverse
effect upon compressibility and upon the life of compacting tools.
Geometrical Properties
viz. particle size distribution, particle shape and particle porosity, determine the
powders specific surface which is the driving force of the sintering process
(Chapter 6).
is analyzed by means of scanning electron microscopy (SEM). (See Fig. 3.1 left).
Figure 3.1. External particle shape (SEM) and internal particle structure (cross-section) of
sponge iron powder (NC100.24) and water-atomized iron powder (ASC100.29).
97
98
Flow Rate
Compressibility
Green Strength
is the bending strength of a green (i.e. compacted but not sintered) rectangular
test bar. Green strength increases with increasing compact density and is
influenced by type and amount of lubricant admixed to the powder. Sufficient
green strength is required to prevent compacts from cracking during ejection
from the compacting tool and prevent them from getting damaged during
handling and transport between press and sintering furnace. The more complex
and delicate the shape of a compact, the higher its green strength should be. If
the green strength of compacts is high enough, they may even be machined prior
to sintering (e.g. undercuts, traverse slots and holes).
General Aspects
Spring Back
99
Approx.
particle
size
range m
Apparent
density
g/cm3
Flow
s/50g
H2-loss
%
NC100.24
20 180
2.45
31
0.20
SC100.26
20 180
2.65
29
MH80.23
40 200
2.30
ASC100.29
20 180
ABC100.30
30 200
C
%
Green
density
Green
strength
1)
1)
g/cm3
N/mm2
<0.01
7.02
47
0.12
<0.01
7.12
40
34
0.32
0.08
6.292)
24 3)
2.95
24
0.08
0.002
7.20
41
3.00
24
0.06
0.001
7.28
44
Figure 3.2. Particle size distribution, flow and apparent density of five different iron powder
grades.
g/cm3
MPa
7,5
35
7,0
30
6,5
25
6,0
20
5,5
15
5,0
(1)
(5)
10
(1)
(2)
(3)
(4)
(5)
(1)
ABC 100.30
ASC 100.29
SC 100.26
NC 100.24
MH 80.23
(5)
Figure 3.3. Compressibility and green strength of five different iron powder grades.
NC100.24
is one of the most widely used grades in the manufacturing of sintered parts.
Due to the irregular surface and the spongy structure of its particles, it has high
green strength; and due to its low contents of oxygen and carbon, it has good
compressibility.
SC100.26
has the best compressibility of all Hgans sponge-iron powder grades. Its
green strength is slightly lower and its apparent density slightly higher than
for NC100.24. It is very useful in cases where parts with high density are to be
achieved in one single pressing operation.
MH80.23
ASC100.29
is a water-atomized iron powder which due to its high purity and its compact
particle structure has very high compressibility. ASC100.29 can be compacted
in one single pressing operation at moderate pressures to densities of up to 7.2
g/cm3.
ABC100.30
1. Zn-st
2.Kenolube
3. Amide wax PM
a) sponge grade
i.e NC100.24
1. Zn-st
2.Kenolube
3. Amide wax PM
b) atomized grade
i.e ASC100.29
Figure 3.4 Influence of mixing time upon apparent density of two different iron powder
grades admixed with three different lubricants.
Properties of some AstaloyTM and Distaloy powders are presented in Table 3.2.
Table 3.2. Properties of some AstaloyTM and Distaloy Powders
Powder
grade
Approx.
particle
size
range m
Apparent
density
g/cm3
Flow
s/50g
H2-loss
%
Astaloy LH
20 180
3.00
26
0.10
Astaloy Mo
20 180
3.00
25
Astaloy CrA
20 180
2.85
Distaloy SA
20 150
Distaloy DC
C
%
Green
density
Green
strength
1)
1)
g/cm3
N/mm2
<0.01
7.07 2)
20 2)
0.10
<0.01
7.10
22
27
0.13
<0.01
7.04
26
2.80
27
0.12
<0.01
7.10
40
20 180
3.05
25
0.10
<0.01
7.10
31
Distaloy AQ
20 180
3.05
27
0.10
<0.01
7.17
36
Distaloy HP
20 180
3.08
25
0.10
<0.01
7.07
29
1)
2)
Astaloy LH
is a water-atomized steel powder alloyed with 0.90% Ni, 0.90% Mo and 0.20%
Mn. It has very good hardenability and can be used for case-hardened as well as
through-hardened parts with non-critical tolerances. Admixing Cu enables very
good sinter-hardening properties.
Astaloy Mo
is a water-atomized steel powder alloyed with 1.50% Mo. Astaloy Mo has high
compressibility, exhibits an homogeneous microstructure after sintering and has
optimal hardenability. These properties make it an excellent choice for surfacehardened components requiring high surface hardness in combination with good
core toughness.
Astaloy CrA
Distaloy SA
is based on the sponge-iron grade SC100.26, to which 1.75% Ni, 1.5% Cu and
0.5% Mo in form of very finely dispersed powders have been diffusion-bonded.
Distaloy SA is recommended for densities up to 6.9 g/cm3 after single pressing.
With additions of graphite, a tensile strength of 600 N/mm2 can be achieved
in one pressing and sintering operation and so produced parts respond well to
heat-treatment.
Distaloy AQ
107
Index
cast-iron . . . . . . . . . . . . . . . . . . . . . . . . 46
CCT-diagrams . . . . . . . . . . . . . . . . 58, 66
cementite formation . . . . . . . . . . . . . . . 54
ceramic . . . . . . . . . . . . . . . . . . . . . . . . . 80
amorphous solids . . . . . . . . . . . . . . . . . . 8
stable short-range order . . . . . . . . . . . . 8
apparent density . . . . . . . . . . . . . . . . . 98
atoms per unit cell . . . . . . . . . . . . . . . . 17
austempering . . . . . . . . . . . . . . . . . 64, 65
austenite . . . . . . . . . . . . . . . . . . . . . . . . 53
grains . . . . . . . . . . . . . . . . . . . . . . 55, 73
retained . . . . . . . . . . . . . . . . . . . . . . . 64
transformations . . . . . . . . . . . . . . . . . 53
56
51
56
56
62
56
57
bainitizing . . . . . . . . . . . . . . . . . . . . . . . 64
BCC . . . . . . . . . . . . . . . . . . . . . . . . 12, 17
BCT . . . . . . . . . . . . . . . . . . . . . . . . . 52, 54
belt furnace . . . . . . . . . . . . . . . . . . . . . 81
binary phase diagram . . . . . . . . . . . . . 37
bonding between atoms . . . . . . . . . . . . 10
covalent bond . . . . . . . . . . . . . . . . . . 11
ionic bond . . . . . . . . . . . . . . . . . . . . . 11
metallic bond . . . . . . . . . . . . . . . . . . . 11
Van der Waals force . . . . . . . . . . . . . . 10
bulk density . . . . . . . . . . . . . . . . . . 96, 98
bainite . . . . . . . . . . . . . . . . . . . . . . . . . .
bainite step . . . . . . . . . . . . . . . . . . . . .
formation . . . . . . . . . . . . . . . . . . . . . .
lower . . . . . . . . . . . . . . . . . . . . . . . . .
nose . . . . . . . . . . . . . . . . . . . . . . . . . .
upper . . . . . . . . . . . . . . . . . . . . . . . . .
bainitic steel . . . . . . . . . . . . . . . . . . . . .
carbon-steel . . . . . . . . . . . . . . . . . . . . . 46
coke breeze . . . . . . . . . . . . . . . . . . . . . . 78
compact density . . . . . . . . . . . . . . . . . . 98
compressibility . . . . . . . . . . . . . . . . . . . 98
crude powder . . . . . . . . . . . . . . . . . . . . 80
crystal . . . . . . . . . . . . . . . . . . . . . . . . . . 12
grains . . . . . . . . . . . . . . . . . . . . . . . . . 20
lattice . . . . . . . . . . . . . . . . . . . . . . . . . 14
segregation . . . . . . . . . . . . . . . . . . . . 44
twins . . . . . . . . . . . . . . . . . . . . . . . . . 20
crystalline solids . . . . . . . . . . . . . . . . . . . 9
crystallites . . . . . . . . . . . . . . . . . . . . . . 20
Curie-point . . . . . . . . . . . . . . . . . . . . . . 46
dendrites . . . . . . . . . . . . . . . . . . . . . . . . 20
diffusion . . . . . . . . . . . . . . . . . . . . . . . . 26
coefficient . . . . . . . . . . . . . . . . . . 27, 30
equation . . . . . . . . . . . . . . . . . . . . . . . 28
Ficks first law . . . . . . . . . . . . . . . . . . 27
Ficks second law . . . . . . . . . . . . . . . 27
rate . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
self . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
surface . . . . . . . . . . . . . . . . . . . . . . . . 26
systems . . . . . . . . . . . . . . . . . . . . . . . 32
vacancy . . . . . . . . . . . . . . . . . . . . . . . 26
volume . . . . . . . . . . . . . . . . . . . . . 26, 31
dislocations . . . . . . . . . . . . . . . . . . . 21, 23
dispersion hardening . . . . . . . . . . . . . . 23
INDEX 109
electron gas . . . . . . . . . . . . . . . . . . . . . . 11
entropy of mixing . . . . . . . . . . . . . . . . . 35
equilibrium diagram . . . . . . . . . . . . . . 46
FCC . . . . . . . . . . . . . . . . . . . . . . . . 17, 19
Fe- Fe3C - diagram . . . . . . . . . . . . . . . .51
ferrite formation . . . . . . . . . . . . . . . . . 53
gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
growth . . . . . . . . . . . . . . . . . . . . . . . . . . 59
hardness of pearlite . . . . . . . . . . . . . . . 67
interstitial atoms . . . . . . . . . . . . . . . . . 21
isothermal . . . . . . . . . . . . . . . . . . . . . . . 64
annealing . . . . . . . . . . . . . . . . . . . 64, 65
transformation . . . . . . . . . . . . . . . . . . 59
transformation diagrams . . . . . . . . . . 58
isothermal soft-annealing . . . . . . . . . . 64
ITT-diagrams . . . . . . . . . . . . . . . . . 58, 59
magnetic separators . . . . . . . . . . . . . . . 80
magnetite . . . . . . . . . . . . . . . . . . . . . . . 78
martempering . . . . . . . . . . . . . . . . 64, 65
martensite . . . . . . . . . . . . . . . . . . . . 49, 53
microhardness . . . . . . . . . . . . . . . . . . . 96
microstructure . . . . . . . . . . . . . . . . . . . 96
Miller indices . . . . . . . . . . . . . . . . . . . . 13
non-equilibrium diagram . . . . . . . . . . 49
non-metallic inclusions . . . . . . . . . . . . 22
octahedral . . . . . . . . . . . . . . . . . . . . . . . 53
parameters of state . . . . . . . . . . . . . . . 33
heterogeneous system . . . . . . . . . . . . . 33
hydrogen-reduced iron powder . . . . . 84
hypoeutectoid steels . . . . . . . . . . . . . . . 53
internal energy . . . . . . . . . . . . . . . . . . . 33
packing density . . . . . . . . . . . . . . . . . . 17
69
55
55
61
33
plastic deformation metal crystals . . . 22
polycrystalline structure . . . . . . . . . . . 20
powder mixes . . . . . . . . . . . . . . . . . . . . 86
reduction mix . . . . . . . . . . . . . . . . . . . . 80
ring-shaped nozzle . . . . . . . . . . . . . . . . 83
tie-line . . . . . . . . . . . . . . . . . . . . . . . . . . 39
transformation diagrams . . . . . . . . . . 58
transition points A3 and A1 . . . . . . . . . 50
trostite . . . . . . . . . . . . . . . . . . . . . . . . . . 50
TTT-diagrams . . . . . . . . . . . . . . . . 49, 58
twinning . . . . . . . . . . . . . . . . . . . . . . . . 20
rotary dryers . . . . . . . . . . . . . . . . . . . . 80
self-tempered martensite . . . . . . . . . . . 57
slip plane . . . . . . . . . . . . . . . . . . . . . . . 22
solidus . . . . . . . . . . . . . . . . . . . . . . . . . . 36
sorbite . . . . . . . . . . . . . . . . . . . . . . . . . . 50
space lattice . . . . . . . . . . . . . . . . . . . . . 13
spheroidite . . . . . . . . . . . . . . . . . . . . . . 67
spheroidized cementite . . . . . . . . . . . . 67
sponge-iron powders . . . . . . . . . . . . . 100
sponge-iron process . . . . . . . . . . . . . . . 78
spring back . . . . . . . . . . . . . . . . . . . . . . 99
stable system . . . . . . . . . . . . . . . . . 46, 47
starmix . . . . . . . . . . . . . . . . . . . . . . . . 93
stationary case . . . . . . . . . . . . . . . . . . . 27
substitutional atoms . . . . . . . . . . . 21, 26
supersaturation . . . . . . . . . . . . . . . . . . 54
surfaces . . . . . . . . . . . . . . . . . . . . . . . . . 31
temper brittleness . . . . . . . . . . . . . . . . 73
tetrahedral . . . . . . . . . . . . . . . . . . . . . . 53
thermal analysis . . . . . . . . . . . . . . . . . . 38
thermodynamic . . . . . . . . . . . . . . . . . . 33
equilibrium . . . . . . . . . . . . . . . . . . . . 33
state . . . . . . . . . . . . . . . . . . . . . . . . . . 33
undercooling . . . . . . . . . . . . . . . . . . 50, 51
vacancies . . . . . . . . . . . . . . . . . . . . . . . . 21
water-atomized . . . . . . . . . . . . . . . . . 100
water-atomized process . . . . . . . . . . . . 82
yield point . . . . . . . . . . . . . . . . . . . . . . . 25
INDEX 111
112 NOTES
Power of Powder
Metal powder technology has the power to open up a world
of possibilities. The inherent properties of metal powders provide
unique possibilities to tailor solutions to match your requirements.
This is what we call Power of Powder, a concept to constantly
widen and grow the range of metal powder applications.
To find out how you can apply the Power of Powder, please
www.hoganas.com/pmc