Material and Powder Properties

1
Material and Powder Properties

Hgans Handbook for Sintered Components
Material and powder properties

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 and Powder Properties

1. 2. 3. Material Science Production of Iron and Steel Powders Characteristics of Iron and Steel Powders
Production of Sintered Components

4. 5. 6. 7. Compacting of Metal Powder Compacting Tools Sintering Re-pressing, Coining and Sizing
Design and Mechanical Properties

8. Designing for P/M Processing 9. Sintered Iron-based Materials 10. Supplementary Operations
Material and powder properties

Material Science . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .7
1.1 The Natural States of Matter . . . . . . . . . . . . . . . 8 1.2 The Crystalline Structure of Metals . . . . . . . . . . . . 10 1.3 Diffusion in Metals . . . . . . . . . . . . . . . . . . . . 26 1.4 Binary Phase Diagrams of Metals . . . . . . . . . . . . 33 1.5 The Iron-Carbon System . . . . . . . . . . . . . . . . . 46 1.6 Transformation Diagrams of Steels . . . . . . . . . . . . 58 1.7 Influence of the Microstructure on the Properties of Steel 67
Production of Iron and Steel Powders . . . . . . . . . .75

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 76 2.2 The Hgans Sponge Iron Process . . . . . . . . . . . 78 2.3 The Hgans Water-Atomizing Process . . . . . . . . . 82 2.4 Hydrogen-reduced Iron Powder . . . . . . . . . . . . . 84 2.5 Alloying Methods . . . . . . . . . . . . . . . . . . . . . 85 2.6 Distaloy and Starmix . . . . . . . . . . . . . . . . . . 87
Characteristics of Iron and Steel Powders . .. .. .. .. .. .. ..95

3.1 General Aspects . . . . . . . . . . . . . . . . . . . . . 96 3.2 Properties of Hgans Iron Powders . . . . . . . . . . 100
This chapter presents a selection of metallurgical concepts and methods essential to the understanding of basic processes in powder metallurgy in particular iron powder metallurgy. The somewhat shorthanded presentation of some items may stimulate the reader to collect further information from pertaining special literature.
Material Science
1.1 1.2 1.3 1.4 1.5 1.6 1.7
The Natural States of Matter . . . . . . . . . . . . . . 8 The Crystalline Structure of Metals . . . . . . . . 10 Diffusion in Metals . . . . . . . . . . . . . . . . . . . . . . 26 Binary Phase Diagrams of Metals . . . . . . . . . 33 The Iron-Carbon System . . . . . . . . . . . . . . . . . 46 Transformation Diagrams of Steels . . . . . . . . 58 Influence of the Microstructure on the Properties of Steel . . . . . . . . . . . . . . . . 67
MATERIAL SCIENCE
1.1 The Natural States of Matter

All matter surrounding us consists of atoms, ions and molecules, which, depending on their mutual distances, are under the influence of stronger or weaker reciprocal forces. Depending on the strength of their reciprocal forces, these particles form gases, liquids or solids. In gases. The particles (i.e. atoms, ions or molecules) move about highly irregularly at very high speeds (some 100 m s-1 at Room Temperature). Their free length of way, , (i.e. the distance a particle can move before colliding with another one) is large compared to the particle diameter. Their density is in the order of 1019/cm3. This situation is called random order. See Fig. 1.1 a. In liquids. The particle density is in the order of 1022 / cm3. The particles move about rather irregularly but are more crowded than in the gaseous state, each particle having approximately ten very close neighbours. Significant for liquids is the frequent local formation of crystal-like agglomerates extending over several particle diameters. Because of the high thermal energy of the particles, these agglomerates are very short-lived and there is no correlation between more distant particles. This situation is called instable short-range order. See Fig. 1.1b. In contrast to gases and liquids, amorphous and crystalline materials exhibit, within a certain range of temperatures, a considerable rigidity they are solids. In amorphous solids. The particle density is of the same order of magnitude as in liquids, i.e. 1022/cm3. Similar to liquids, a short-range order exists between particles. But in contrast to liquids, due to the heavily restricted mobility of the particles in amorphous solids, this order is stable and, logically, it is called stable short-range order. See Fig. 1.1 c. Amorphous materials have no sharply defined melting point. The transition from their solid to their liquid state occurs gradually. They can, in fact, be looked at as liquids with an extremely high viscosity at room-temperature. Classic examples are glasses. But today, by means of special processes, also metals can be transformed into an amorphous state.
THE NAtURAL StAtEs OF MAttER
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
1.2 The Crystalline Structure of Metals

Our further considerations are focused mainly on the crystalline in particular the metallic state of matter. The formation of crystalline structures is controlled by certain binding-forces between the atoms. Geometrical descriptions of crystalline structures refer to so-called ideal crystals only and do not take into account disturbances occurring in real crystals, such as surfaces, grain-boundaries, dislocations, vacancies, foreign atoms and other irregularities. In the following paragraph, we will discuss; 1) various types of bonding between atoms, 2) structures of ideal crystals, 3) stacking faults and disturbances in real metal crystals, and 4) plastic deformation in metal crystals.
1.2.1 Bonding Between Atoms
Crystal structures owe their existence to various types of interatomic forces between neighbouring atoms. Four characteristic types are recognized: Van der Waals, covalent, ionic and metallic. Van der Waals force. This is the only appreciable force exerted between well separated atoms and molecules. It is a weak attractive force that acts between all atoms and is responsible e.g. for the condensation of noble gases and chemically saturated molecules to liquids and solids at low temperatures. Its explanation requires quantum mechanics because it involves fluctuations of the electronic charge in the atom. When two atoms approach fairly closely these fluctuations can occur in unison so that one atom has its electrons slightly nearer the other nucleus whenever this nucleus happens, through the movements of its own electrons, to be more exposed in this direction than usual.
THE CRYstALLInE StRUctURE OF MEtALs
11
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.
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
BCC
Metal
-Fe -Fe Cr Mo W -Ti
Temperature C
Lattice Constants
a) 10-10 m; c) 10-10m 2,86
Ratio of Lattice Constants c/a
1390
2,932 2,884 3,147 3,165
900
3,306 3,524
FCC
Ni -Co Cu -Fe Ir Pt Al Au Ag Pb 1000 467
3,560 3,619 3,651 3,839 3,923 4,049 4,078 4,086 4,950 2,506; 4,068 2,665; 4,947 2,950; 4,679 2,979; 5,617 3,209; 5,210 1,62 1,86 1,59 1,89 1,62 1,633
HCP
-Co Zn -Ti Cd Mg (ideal)
* W. Schatt: Einfhrung in die Werkstoffwissenschaft. VEB Deutscher Verlag fr Grundstoffindustrie, Leipzig 1981, S.55.
13
c
Figure 1.2. Unit cell in a crystal cell; lattice constants: a, b, c.
1.2.2 Ideal Crystals
A crystal structure has three general properties: periodicity, directionality and completeness. Periodicity is the regular repetition in space of the atomic unit of the crystal (wallpaper pattern). It is the basis of the remarkable plastic properties of crystals. Directionality of the crystal structure appears in the fact that properties like conductivity and elasticity vary with the direction of their measurement. Completeness is simply the filling of all crystal sites defined by the periodic structure with the required atoms. In crystallography, it is customary to present a crystal structure graphically as a space lattice within which the sites of atoms are marked by points (or small circles). These points are referred to as lattice points. The basic unit of the crystal structure which repeats itself in all three directions of space is called unit cell. The defining edges of this cell are referred to as lattice constants, (usually a, b, c). See Fig. 1.2. The lattice points lie at the intersections of three families of parallel planes, called lattice planes. The distance between neighbouring lattice planes is usually designated by the triple-indexed symbol d h k l. See Fig. 1.3. The spatial orientation of any lattice plane is clearly defined by its so-called Miller indices (h k l), which are determined in the following way:
14
MATERIAL SCIENCE
LP
LP LP LP-distance LP-family
Figure 1.3. Lattice planes (LP) in a crystal lattice.
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).
15
Figure 1.4. Derivation of Miller indices.
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
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.
Figure 1.6. Lattice points and lattice directions in the cubic crystal system.
For the characterization of lattice planes, directions and points in hexagonal crystal systems, four instead of three Miller indices are required. Their detailed description can be found in pertaining special literature and has no bearing on our further discussions. As already mentioned, all metals crystallize either in a body-centered-cubic (BCC), in a face-centered-cubic (FCC) or in a hexagonal close-packed (HCP) structure. The respective unit cells of these structures are shown in Fig. 1.7. The dice-shaped unit cells of the BCC- and of the FCC-structure are each presented in two different ways: to the left as abstract point lattice, and to the right as conglomerate of spheres. The sphere-model conveys a fairly realistic picture of the packing structure of the atoms inside the crystal.
17
BCC
r
NA = 2 NN = 8 PD = 68%
ar = 4 r / 3
FCC
NA = 4 NN = 12 PD = 74%
r
ar = 4 r / 2
HCP
NA = 6 NN = 12 PD = 74%
d A
B A
d= 2r h = d 3
FCC
B A
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:
18
MATERIAL SCIENCE
In the case of the BCC-cell, putting together an eighth of an atom from each corner of the cell, and adding the one whole atom from the center yields 2 whole atoms per unit cell: NA = 2. In the case of the FCC-cell, putting together an eighth of an atom from each corner of the cell, and putting together to 3 whole atoms the 6 halves of an atom from the faces of the cell, yields 4 whole atoms per cell: NA=4. In the case of the HCP-cell, putting together the 12 sixth of an atom from the corners of the cell, putting together to 1 whole atom the two halves of one atom from the hexagonal faces, and adding the three atoms from inside the cell yields 6 atoms per cell: NA = 6. Obviously, for the BCC-cell, the number of nearest neighbours is NN = 8. The number of nearest neighbours for the FCC- and for the HCP-cell becomes evident when looking at the plane of close-packed atoms (shaded), within which each atom has 6 nearest neighbours. In addition, each atom has 3 nearest neighbours each in the two planes above and beneath the shaded plane. All together, they add up to 12 atoms: NN = 12. Both, in the FCC- and in the HCP-structure, atoms are packed at maximal possible density, viz. PD 74%. The packing density of the atoms in the BCCstructure is only PD 68%. The reader can easily verify these two figures by dividing the volume of atoms belonging to one unit cell (NA/4 r3/3) with the volume of this cell (a3 ). In Fig. 1.7, for comparison, fully close-packed lattice planes in the HCP- and in the FCC-structure are marked by shading. The diagram in Fig. 1.8 (a) shows a fully close-packed plane. The spheres or atoms lie in three sets of close-packed lines, physically equivalent and symmetrically orientated to one another. Both, the HCP- and the FCC-crystal structure are formed by stacking a number of close-packed planes on one another in a certain stacking sequence.
19
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.
20
MATERIAL SCIENCE
The stacking sequence ABABAB forms the HCP-structure, and ACBACBACB forms the FCC-structure. In FCC-structures, the stacking sequence occasionally forms mirror images of itself like ABC(A)CBA(C)ABC. This kind of stacking faults is called twinning. Crystal twins occur frequently in copper, nickel and austenitic steels and appear on micrographs as parallel stripes inside the crystal grains.
1.2.3 Real Metal Crystals
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 crys tallites).
21
2 1
Figure 1.10. Zero-dimensional lattice disturbances:
(1-1) vacancy,
(2-2) interstitial atom, (3-3) bigger and, (4-4) smaller substi tutional atoms.
Zero-dimensional lattice disturbances are 1) missing atoms in the regular lattice structure, (so-called vacancies), 2) foreign atoms occupying interstices between host atoms (interstitial atoms, and 3) larger or 4) smaller foreign atoms substituting atoms of the host lattice (substitutional atoms). See Fig. 1.10.
One-dimensional lattice disturbances are so-called dislocations which occur in two different varieties, edge dislocations and screw dislocations, forming zones of lower packing-density which like filaments stretch through the crystal structure. The schematic drawings in Fig. 1.11 shows one edge dislocation (a) and one screw dislocation (b).
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.
22
MATERIAL SCIENCE
Two-dimensional lattice disturbances are the above mentioned grain boundaries, which can also be imagined as arrays of edge dislocations stacked on top of one another. See Fig. 1.12.
Figure 1.12. Two-dimensional lattice distur bance: a grain boundary (composed of an array of dislocation lines stacked on top of each other) the grain boundary face is perpendicular to the drawing plane.
Three-dimensional lattice disturbances are non-metallic inclusions, dislocation tangles, and small pores as occur in certain diffusion processes (Kirkendalleffect) and in sintering processes.
1.2.4 Plastic Deformation of Metal Crystals
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.
23
Figure 1.13. Atomic bonds trans iently being ruptured as dislocation travels through crystal under the influence of shearing stresses.
Along a dislocation line perpendicular to the direction of slip, bonds are temporarily suspended. As shown schematically in Fig. 1.14, the dislocation line, driven by shearing-stresses, travels through the crystal achieving a small displacement of the upper half of the crystal against its lower half.
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.
In order to achieve a noticeable plastic deformation, a great many dislocation lines must travel through the metal. The more crowded the dislocations get, the more do they interfere with one another: They get entangled, pile up against grain boundaries (ends of slip planes) and catch on with inclusions and other lattice disturbances. See Fig. 1.15. The jamming and entangling dislocations build up a growing resistance against further deformation which eventually ceases, unless shearing-stresses are increased. This is the cause of the phenomenon of deformation hardening and a contributing cause to dispersion hardening and other hardening processes.
24
MATERIAL SCIENCE
When the temperature is raised, dislocations gradually dissemble as atoms escape from them via vacancies. This dissembling mechanism is called dislocation climbing.
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
See Fig. 1.16. Dissembling of jammed and entangled dislocations is the mechanism behind the effect which soft-annealing has on cold-deformed metals. The fact that metals deform easier at elevated temperatures is due the following two processes: 1) the increased number and greater mobility of vacancies in the metal facilitate the dissembling of dislocations, and 2) the thermal vibration of the atoms in their lattice sites is intensified. Both processes have a lowering effect on the metals resistance to deformation, i.e.on its yield point.
Figure 1.16. Dislocation climbing: A dislocation dissembles gradually as atoms escape from it via vacancies.
26
MATERIAL SCIENCE
1.3 Diffusion in Metals

All processes and reactions taking place in gases, liquids or solids are depending on the thermally activated exchanges of atomic particles controlled by laws of thermodynamics and statistics. In homogeneous materials, the thermally activated particles move around in all directions without any preference. When gradients of temperature, pressure, concentration or electrical potential occur in the material, the irregular movement of particles is superimposed by a drift in the direction of the gradient (similar to a mosquito swarm drifting in the wind). This drift of atomic particles is called (directional) diffusion. Diffusion usually accomplishes a noticeable transport of material as, for instance, in the formation or transformation of alloys and in the sintering of metal powders. The movement and exchange of atoms in solid metals is feasible in various ways. Inside the crystal lattice of the metal, small foreign atoms can move relatively easily from interstice to interstice between the host atoms. This kind of atom movement is called interstitial diffusion. See Fig. 1.17 (a). The host atoms themselves and substitutional atoms can move only from one vacancy to another. This kind of atom movement is called vacancy diffusion or, pertaining to the host atoms, self-diffusion. See Fig. 1.17 (b). More easily than in the close-packed lattice structure, atoms can diffuse along the less closely packed grain boundaries and most easily along surfaces. Accordingly one speaks of volume-diffusion, grain-boundary-diffusion and surface-diffusion.
Figure 1.17. Interstitial diffusion (a) and vacancy diffusion (b); (schematically).
DIFFUsIOn In MEtALs
27
1.3.1 Laws of Diffusion
In heterogeneous materials, concentration differences tend to level, because an even distribution of concentration constitutes a state of minimal free enthalpy. According to Ficks first law of diffusion, the number dn of atoms passing through an area A perpendicular to the direction of their movement, during a short time interval dt, is proportional to the concentration gradient -dc/dx residing at the location of area A:
(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.
28
MATERIAL SCIENCE
1.3.2 A Particular Solution of the Diffusion Equation
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
29
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 accor ding to equ. (1.4) and represent the distribution of A-atoms across the two bars after different periods of time.
Remarkable with equation (1.4) is the circumstance that the concentration c (x,t) does not depend explicitly of the individual parameters x and t but on their combination as in the expression . This means that any given concentration of A-atoms in B moves in x-direction (i.e. from A to B) according to the time law: . This time law is significant of all diffusion processes and controls e.g. the speed of growth of oxide layers on metal surfaces and the travelling speed of a carburized zone from the surface to the center of a work piece of steel.
30
MATERIAL SCIENCE
1.3.3 The Diffusion Coefcient
The diffusion coefficient D is a measure of the diffusion rate. D depends on the mechanism of atom movement inside the metal and particularly strongly of temperature according to the following relationship:
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
31
Temp. 10 -10
14
11
9 x 100 C
Surface 10 -12 D, m2/s (log-scale) Grain Boundary 10 -14
10 -16 -Fe 10 -18
Volume
7 1/T (1/K)
9 x 10 -4
Figure 1.19. Temperature dependence of the coefficients of self-diffusion of iron, for diffu sion along surfaces and grain boundaries and inside the crystal volume (In D over1/T).
32
MATERIAL SCIENCE
Diffusion coefficients D, material constants D0 and activation energies Q for some important diffusion systems are presented in Table 1.2.
Table 1.2. Physical Data on some Important Diffusion Systems
Diff. Atom
Fe
HostCrystal
-Fe (BCC) -Fe (FCC) -Fe
D0 (m2/s)
2,0 x 10 -4
Activation Energy Q KJ/mol kcal/mol
T (C)
D (m2/s)
241
57,5
500 900 900 1100 500 900 900 1100 500
1,1 x 10-20 3,9 x 10-15 1,1 x 10-17 7,8 x 10-16 2,3 x 10-12 1,6 x 10-10 9,2 x 10-12 7,0 x 10-11 4,4 x 10-19
Fe
5,0 x 10 -5
284
67,9
6,2 x 10 -7
80
19,2
-Fe
1,0 x 10 -5
136
32,4
Cu
Cu
7,8 x 10 -5
211
50,4
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
BInARY PHAsE DIAGRAms OF MEtALs
33
1.4 Binary Phase Diagrams of Metals
1.4.1 Thermodynamics
First of all, some definitions are required: A thermodynamic state is given through the following four parameters of state: P, V, T and c (pressure, volume, temperature and concentration). A heterogeneous system consist of several phases separated by interfaces and having different properties (e.g. ice-crystals or droplets of oil in Water). A phase is in itself homogeneous and can consist of several components present in varying concentrations (e.g. salt water consists of H-atoms, O-atoms, Na-atoms and Cl-atoms). Above-mentioned parameters of state given, number, type and relative amount of different phases present in a heterogeneous system are stable, when the system is in thermodynamic equilibrium. What is meant by this statement, will emerge from the following consideration. A systems (e.g. a metals) internal energy E is the sum of all individual kinetic and potential energies of its particles (metal atoms). The external energy PV imposed on the system is the product of the systems volume V and the external pressure P. The sum of these two energies E + PV can also be split up according to a different aspect, namely into one part which constitutes irreversible thermal energy TS (S = entropy) and another part G which is available to accomplish work (e.g. transformation work). G is called Gibbs free energy. This interrelationship is expressed by the following equation:
E + PV = TS + G (1.6)
Thus, Gibbs free energy is given by the expression:
G = E TS + PV (1.7)
34
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:
d G = d(E - ST + PV) = d F + d(PV) (1.8)

From the First and Second Law of Thermodynamics, in case T = constant and P = constant:
dG = d(E ST + PV) 0 (1.9)

and, in case T = constant and V = constant, it follows: dF = d(E TS) 0 (1.9a) Since equilibrium reactions between metal phases usually take place at constant pressure and constant temperature but sometimes are accompanied by volume changes, equation (1.9) applies. This means that between phases striving for equilibrium, only such reactions can take place that do not increase the free energy, G. All reactions cease when the phases are in equilibrium and the free energy, G has reached a minimum. The schematic diagram in Fig. 1.20 illustrates this situation.
Free Energy G
Solid Liquid Solid m.p Liquid Temperature
Figure 1.20. Change of free energy with temperature for a pure metal.
35
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.
1.4.2 Derivation of Binary Phase Diagrams
Applying the thermodynamic rules discussed in the preceding paragraph to a heterogeneous system involving two components A and B and two phases 1 and 2, equilibrium, at given temperature and pressure, can be achieved only by changing the concentrations of A and B in the two phases. The component A will diffuse from phase 1 to phase 2, or vice versa, until the changes of free energy with changing concentration are equal with respect to both phases, viz.: (1.10) The implication of this equation is that the two functions G1(c) and G2(c), plotted in a G-c-diagram, have a common tangent when they are in equilibrium. On the basis of these considerations, any binary diagram can be derived theoretically as long as the respective free energies are known as functions of the concentration. These functions can be deduced from the entropy of mixing of the involved components. A discussion of entropy of mixing would exceed the frame of this chapter, and is not required for the understanding of succeeding paragraphs.
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:
Systems with Complete Miscibility in the Solid and in the Liquid State.
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).
37
GS a) Free Energy of Phases G(c) GL
T>TMB
T=TMB GS GL b) GS CL GL GS d) CS TMA>T TMB>T>TMA
GL c)
Figure 1.21. Deriving a binary phase diagram from the free energy curves for the liquid and solid phases at different temperatures. A and B are the components of the system which are completely miscible both in the liquid and solid state. GL and GS are the free energies of the liquid and of the solid state resp.
Temperature T
Liquidus
TMB
TMA
Solidus
TMA and TMB are the melting points of A and B resp. CL and CS are the concentrations of B in the liquid and in the solid state resp.*
A e)
Concentration c
* G.E.R. Schulze: Metallphysik, Akademie-Verlag, Berlin 1974.
38
MATERIAL SCIENCE
Experimentally, binary diagrams can be established by means of thermal analysis. To this effect, from the two components of the system to be investigated, a number of alloys of different composition are smelted and subsequently cooled. By means of a thermocouple, the temperature changes during cooling are registered. On the so-obtained curves, phase transformations show up in the form of kinks. These kinks are caused by thermal energy being released during phase transformation (lowering of free energy) and slowing down the rate of cooling. The principle of this procedure is illustrated in Fig. 1.22. Typical representatives of this simple binary system are the following pairs of metals: Cu-Ni, Ag-Au and Ag-Pd.
B 1 2 3 A
Li Temperature T
Temperature T
So
A a) Time t b)
c3
c2
c1
Concentration c
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.
More complicated phase diagrams can be derived theoretically and established experimentally by methods analogous to the ones described above. As an example, we will study the binary T-c-diagram of a so-called eutectic system as presented in Fig. 1.23.
* W. Schatt: Einfhrung in die Werkstoffkunde, VEB Deutscher Verlag fr Grundstoffindustrie, Leipzig 1981.
39
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.
L
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.
A remarkable feature of eutectic systems is the circumstance that at a certain temperature TE (eutectic temperature) not two but three phases are in equilibrium with one another, viz. a solid A-rich -phase, a solid B-rich -phase, and a liquid phase L (the melt). The equilibrium relation between the three phases is represented in the diagram by a horizontal tie-line (also called eutectic line) at the level of TE. Above this line, there are two so-called 2-phase areas (+L) and (+L) as well as one 1-phase area (L). All three areas join in one point E (the eutectic point) on the eutectic line. Below the eutectic line, there is one 2-phase area (+). To their left, the 2-phase areas border a single-phase area () and to their right, they border a single-phase area (). In the (+L)-area, the melt is in equilibrium with the solid -phase and in the (+L)-area with the solid -phase. In the (+)-area, the two solid phases and are in equilibrium with one another. No additional
+ L
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.
41
Pb Sn
327C
300 Temperature C
Liquid (L)
200
+L
187C
232C
+L
97,5
19,2 100
0 0 20 40 60 80 100% Sn
Pre-eutect (19,2% Sn)
Eutecticum (61,9% Sn)
Dark (19,2% 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).
42
MATERIAL SCIENCE
Peritectic Systems.
An equilibrium between three phases occurs also in a family of binary diagrams called peritectic. In these systems, the melt reacts with an already precipitated solid phase, say , in such a way that a different solid phase, say , is formed according to the peritectic reaction L+. Fig. 1.25 shows the binary diagram of a peritectic system and the cooling curves by means of which it is established.
Temperature
Temperature
Concentration
Time
Figure 1.25. Experimental derivation of the phase diagram of a peritectic system A-B
Eutectoid Systems.
In some binary systems, as for instance in the iron-carbon-system, in addition to the eutectic reaction between the melt and two solid phases L +, an analogous reaction occurs between one solid phase and two other solid phases: + which is called eutectoid reaction. The iron-carbon-system is of central interest not only to the metallurgy of conventional steel but also to the powder metallurgy of iron. But before concentrating on this important system, we want to give a more detailed explanation of the above-mentioned lever-rule of phases and, in this context, describe the phenomenon of crystal segregation which frequently occurs in the solidification process of alloys.
43
1.4.3 The Lever-Rule of Phases
Guided by the simple binary diagram for two components A and B as shown in Fig. 1.26, this rule is readily explained.
Liquid
m0 mL
TB mS
P0
y
Figure 1.26. Lever-rule of phases; mL = amount of residual melt with concentration cL. mS = amount of solid phase with concentration cS precipitated from
TA Solid Solution
the melt. x = (c0 - cL) and y = (cS - c0), c0 = initial concentration of the melt = average concentration of the solid alloy. r-rule: mLy = mSx.
CL
C0
CS
Concentration of B in A
We assume that a given amount m0 of a molten alloy of concentration c0 (concentration of B-atoms) is rapidly cooled down to a temperature T0 inside the 2-phase area (liquid + solid). Through P0 we draw a horizontal line (the so-called tie-line) which cuts the two phase-borderlines at the points mL and mS respectively. In this undercooled state, the melt of concentration c0 has become thermodynamically instable. According to the binary diagram, a melt of concentration cLshould now be in equilibrium with a solid phase of concentration cS. In order to establish this new equilibrium, crystallites of concentration cS > c0 must precipitate depriving the melt of B-atoms. In this reaction, both, the total amount m0 = (mL + mS) of alloy and the total amount c0m0 of B-atoms must remain unchanged.
44
MATERIAL SCIENCE
Thus, the following two laws apply: (a) Conservation of total mass: (b) Conservation of mass of B-atoms: Inserting (a) into (b): Evident from the diagram:
m0 = mL + mS c0m0 = cLmL + cSmS mL (c0 cL) = mS (cS - c0)
x = (c0 cL) and y = (cS c0) mLx = mS y
Lever-rule of phases: m0 mL mS c0m0 cLmL cSmS
= mass of melt before solidification = total mass of alloy = mass of residual melt after established equilibrium = mass of solidified phase after established equilibrium = mass of B-atoms before solidification = mass of B-atoms in residual melt = mass of B-atoms in solidified phase
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).
Frequently, during solidification of molten alloys, a characteristic phenomenon called crystal segregation occurs, i.e. a heterogeneous distribution of concentration inside individual crystal grains of the solidified alloy. Crystal segregation comes about as follows: As can be seen from the binary diagram shown in Fig. 1.27, solidification starts with the precipitation of small B-rich dendrites of concentration c1 > c0. As solidification continues, thin layers of decreasing concentration of B-atoms are successively being deposited on the dendrites, wrapping them in onion skins so to speak. In the outermost layer, the concentration of B-atoms is cn < c0. The varying amount of B-atoms in the individual layers is a direct consequence of the above-mentioned lever-rule of phases.
1.4.4 Crystal Segregation
45
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 Tn TA
Solid Solution
C0
TB
Concentration of B in A
Figure 1.27. Crystal
Cn
C0
C1
segregation; c0= initial concentration of the melt; c1 = concentration of den dritic crystal nuclei first precipitated from the melt; cn = concentration of resi dual melt solidifying last.
When a molten alloy is cooled sufficiently slowly, due to diffusion processes, concentration differences caused by crystal segregation level to a certain degree already during continued cooling. In some cases, in order to improve the properties of the solidified alloy, remaining concentration differences have to be eliminated by means of a subsequent heat treatment called diffusion annealing or homogenizing.
46
MATERIAL SCIENCE
1.5 The Iron-Carbon System
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
1.5.1 The Equilibrium Diagram: Iron-Carbon
THE IROn-CARbOn SYstEm
47
a lamellar structure called pearlite. An austenite containing exactly 0.8% C transforms directly and entirely to pearlite when it passes through the eutectoid point S. An austenite containing more than 0.8% C transforms partly to cementite (secondary cementite) when it passes the phase-borderline SE (also called Acm) and below the eutectoid line SK, its residue transforms directly to pearlite.
Figure 1.28. Equilibrium phase diagram of the system Iron-carbon: fully drawn lines = metastable system (Fe-Fe3C), broken lines = stable system (Fe-graphite).
48
MATERIAL SCIENCE
With reference to these three model cases, one distinguishes between hypoeutectoid, eutectoid (pearlitic) and hypereutectoid iron-carbon-alloys or steels. The respective microstructures are presented schematically below the diagram. Real microstructures of austenite, ferrite, pearlite and cementite are shown in Fig. 1.29.
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.*
* Jrnets och stlets metallografi, Sandvikens Jernverks Aktiebolag 1964.
49
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.
1.5.2 The Non-Equilibrium Diagram: Iron-Carbon
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 0 (equilibrium). Undercooling step I.
Equilibrium transformation at very low cooling rate. Already at a cooling rate of 1 Ks-1, a reduced formation of pre-eutectoid ferrite in under-eutectoid steels and of pre-eutectoid cementite in over-eutectoid steels becomes noticeable. At further increased cooling rate, the formation of preeutectoid ferrite is gradually hampered to such a degree that it occurs only in steels of very low carbon contents. The austenite transforms to pearlite of increasingly finer lamellar structure which, when just about resolvable by light-microscopic methods, is called sorbite. In hypoeutectoid steels, the formation of pre-eutectoid cementite is completely suppressed.
Undercooling step II.
An extremely fine-lamellar pearlite, called trostite, is formed. Its lamellar structure is resolvable by means of electron-microscopic methods only. The united undercooling steps I and II are also called pearlite step.
* W. Schatt: Einfhrung in die Werkstoffkunde, VEB Deutscher Verlag fr Grundstoffindustrie, Leipzig 1981.
51
Undercooling step III.
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
G
E 1 Ks-1 15 Ks-1 40 Ks-1

Equilibrium
I Pearlite formation whith recalescence
900 800 700 600 Temperature

P
0 50 100 150 200 250 300 400 500 Ks-1
II Pearlite formation whithout recalescence
500 400 300 200

IV Martensite formation III Bainite 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
52
MATERIAL SCIENCE
Undercooling step IV.
In this so-called martensite step, the undercooling of the austenite is so extreme that the diffusion of carbon atoms has ceased and, hence, the formation of cementite has become impossible. The FCC lattice of the extremely undercooled austenite shears diffusionless into a tetragonally distorted (BCT) ferrite lattice. Since the so formed ferrite is extremely supersaturated with carbon, high internal stresses occur in its lattice. The tetragonal distortion of the lattice and the internal stresses increase with increasing carbon content. See Fig. 1.32. This is the cause of the high hardness of martensite.
Figure 1.32. Influence of carbon content on the axis lengths of the tetragonal martensite lattice.*
* K. Honda and Z. Nihiyama.
53
1.5.3 Mechanisms of Austenite Transformation
To further the understanding of austenite transformations, it will be helpful to discuss in more detail the various mechanisms of diffusion and structural changes involved in the formation of ferrite, pearlite, bainite and martensite. In between the close-packed iron atoms in the FCC lattice of austenite, there are two types of interstices: the smaller tetrahedral and the larger octahedral interstices as shown in the schematic diagrams in Fig. 1.33. Carbon atoms dissolved in austenite occupy octahedral interstices only, where they cause the least possible lattice distortions and internal stresses. To each unit cell of the austenite lattice belong four octahedral interstices (one at its center and a quarter of one on each of its twelve edges). Thus, in each unit cell, there are four positions available to be filled by carbon atoms. Since also four iron atoms belong to each austenite cell, the theoretically possible concentration of carbon atoms in austenite would be 50 atomic-% compared to their maximal concentration of 9.51 atomic-% in reality. This means that, even at maximal carbon concentration, most octahedral interstices in the austenite lattice remain unfilled.
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,
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).
Figure 1.34. Geometrical interrela tion between the FCC lattice of aus tenite and the BCC lattice of ferrite. = possible sites of carbon atoms.
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.
55
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.
Below the eutectoid temperature (723C), austenite separates into cementite and ferrite, thereby lowering the free energy of the alloy. Nucleated at the austenite grain-boundaries, lamellas of cementite and ferrite grow side by side gradually towards the center of the austenite grains as shown in the schematic diagrams in Fig. 1.35. The lamellar structure occurs because the precipitating cementite completely deprives its austenitic neighbourhood of carbon triggering its transformation to ferrite. With increasing degrees of undercooling, i.e. decreasing temperature, the diffusion rate of carbon decreases and consequently, the lamellar structure of the pearlite becomes finer (and harder).
Figure 1.35. Nucleation of pearlite colonies at austenite grain boundaries (left) and growth of pearlite lamel las in austenite grains (right).*
Martensite Formation.
Martensite formation is based on the same mechanism as ferrite formation. The difference lies in the cooling rate. At low cooling rates, carbon atoms have sufficient time to diffuse and form cementite before the lattice of
*A.H.Cottrell: An introduction to Metallurgy, Edward Arnold (Publishers), London 1968.
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.
a
Figure 1.36. Acicular martensite (a) and tempe red martensite with finely dispersed cementite (b).*
Bainite formation.
Bainite is of great technical interest because it offers a very favourable combination of hardness and toughness not easily attainable with martensite. According to differences in cooling conditions, one distinguishes between two kinds of bainite: upper bainite and lower bainite. Upper bainite occurs at temperatures around 500 C where diffusion rates of carbon are relatively higher. Lower bainite occurs at temperatures around 300C (but above MS) where diffusion rates of carbon are relatively lower.
* K.J. Irvine, Symposium: Steel-Strengthening Mechanisms, Zurich, May 5th and 6th 1969.
57
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.
a
0.21% Cr; 0.39% Ni).*
Figure 1.37. Lower (a) and upper (b) bainite in a bainitic steel (0.87% C; 0.44% Mn; 0.17% Si;
* I.J.Habraken and M.Economopoulos, Symposium: Transformation and Hardenability in Steels, Ann Arbor, Michigan, USA 1958.
58
MATERIAL SCIENCE
1.6 Transformation Diagrams of Steels

As briefly mentioned in paragraph 1.5.2, the goal of commercial heat treatment of steel is attaining optimal combinations of properties adaptable to a large variety of applications. This can only be achieved by steering the formation of adequate microstructures via strictly controlled cooling rates and holding times. Adequate heat-treating procedures cannot be designed on the basis of equilibrium diagrams. To this end, a particular kind of empirically established transformation diagrams are required, usually called Time-Temperature-Transformation diagrams, or TTT-diagrams for short. According to the procedure by which they are established, one distinguishes between Isothermal Transformation diagrams (ITT-diagrams) and Continuous Cooling Transformation diagrams (CCT-diagrams). Before going into details, a general feature of phase transformations must be briefly explained. Phase transformations in metals (like any kind of phase transformations) pass through two characteristic periods: a period of nucleation and a period of growth. Nucleation is a process in which small nuclei of a new phase precipitate at random from the mother-phase. Growth means that nuclei, which incidentally exceed a critical minimum size, begin to grow on account of the mother-phase, eventually replacing it entirely. The growth rate of the new phase is proportional to the residue of untransformed mother-phase, i.e. at the beginning, the new phase grows fast, but as the amount of mother-phase shrinks, the growth rate of the new phase slows down accordingly. This mechanism yields the characteristic s-shaped transformation curve as shown schematically in Fig. 1.38.
TRAnsFORmAtIOn DIAGRAms OF StEELs
59
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.
In order to experimentally derive an ITT-diagram, a series of small specimens of the steel to be investigated are austenitized at a temperature above A3. From austenitizing temperature, the specimens are individually quenched to certain preselected temperatures, where they are held for various lengths of time. Subsequently they are quenched to Room Temperature. On the so-treated specimens, the amount of residual austenite is determined by means of quantitative metallographic methods. Plotting the so-obtained amounts of residual austenite against respective holding times, characteristic s-curves (of type as discussed above) emerge one for each holding temperature. On each of these curves, three points of time are marked: 1) the point of beginning transformation (100% austenite), 2) the point where the amount of residual austenite has decreased to 50% (50% transformation) and 3) the point where the amount of residual austenite has dropped to 10% or less (90 - 100% transformation). Transferring these three points from each individual S-curve to a temperature-time-diagram, as illustrated schematically in Fig. 1.39, one obtains an ITT-diagram with three characteristic nose-shaped curves, representing beginning, half-completion and end of austenite transformation respectively.
1.6.1 ITT-Diagrams
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.*
The following details are noted: At low degrees of undercooling, i.e. at temperatures close below A1, rather long periods of time elapse before transformation begins (hours or days). Transformation product: pearlite (P) and residual austenite (A). With increasing degrees of undercooling, transformation begins correspondingly sooner. Transformation product: P + A. At medium degrees of undercooling, i.e. at temperatures between approx. 550 and 500 C, transformation begins almost instantly (after less than 1 second). Transformation product: P + A. With further increasing degrees of undercooling, the beginning of transformation is delayed again ( up to approx. 20 seconds). Transformation product: A + B (bainite).
* Sandvikens Jernverks Aktiebolag: Stl - struktur och vrmebehandling, 1958.
61
At quenching to temperatures below approx. 280 C, large parts of austenite transform instantly (without preceding period of nucleation) into martensite (M). The lower the quenching temperature, the smaller is the amount of residual austenite. Transformation product: x% A + (100 - x)% M. The nose of the transformation curves between 500 and 600 C is called pearlite-nose because above its tip, austenite transforms to pearlite. The noseshape of the transformation curves has a simple explanation. Close below A1, the diffusion rate of carbon is high, but, due to the low degree of undercooling, the probability of cementite nucleation is very low. With increasing degrees of undercooling, the probability of cementite nucleation increases accordingly, but the diffusion rate of carbon decreases. As a result of these two counteracting processes, transformation times adopt a minimum between low and high degrees of undercooling. Carbon steels with less than 0.25% C are not well suited for heat-treating procedures aiming at the formation of martensite or bainite. Their pearlitenose lies too far left in the ITT-diagram and cannot be bypassed at technically practical cooling rates. A cooling curve just touching the pearlite-nose is called critical cooling curve. When the carbon content is increased or when alloying elements like Cr, Ni, Mn, Si, or W are added, the pearlite-nose is shifted toward the right in the diagram, the critical cooling rate is correspondingly decreased, and the formation of martensite and bainite facilitated. Certain combinations of Cr-, Ni-, Mn-, and Mo-additions entail transformation curves with two noses. Underneath the pearlite-nose a so-called bainite-nose occurs as in the ITT-diagram shown in Fig. 1.40.
62
MATERIAL SCIENCE
C 800 700
F A 1400 1200
A0 3 A0 1
A A+F A+P P
600 Temperature 500 400 300 200 100 0 1000 800 600 400 ITT DIAGRAM 200
1 min. 1 hour 1 day MS M 90 M 90 50%
32 0.5 1 2 5 10
10 2
10 3
10 4
10 5
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.*
When cooling a big work piece of steel from austenitizing temperature, center and surface are cooling at different rates. Then, it may happen that the cooling curve pertaining to the surface bypasses the pearlite-nose, while the curve pertaining to the center passes through it. In such a case, a martensitic microstructure occurs at the surface and a pearlitic-bainitic microstructure at the center of the piece. See schematic diagram in Fig. 1.41. An analogous problem arises, when a heat of smaller pieces is cooled, because pieces located in the outer parts of the heat are cooling faster than those buried deeper inside. This kind of problem is avoided if the heat is quenched to a temperature somewhat above martensite formation (MS) and held there until transformation is complete.
* Atlas of Isothermal Transformation Diagrams, US Steel, Pittsburgh, 1951.
Hardness - RC 13 24 31 26 30 37 44 50 62
63
Figure 1.41. Cooling curves for surface and center of a small (fully drawn lines) and of a big (dotted lines) work piece (schematically).*
1.6.2 Heat Treatment of Steel based on ITT-Diagrams
The properties of steel can be improved considerably through special heat treatments, i.e. certain sequences of heating, quenching and tempering. Properties like strength, hardness, toughness and ductility, depend essentially, if not entirely, on the microstructure of the steel. Heat-treating procedures must, therefore, be especially designed to attain the required microstructures.
* Sandvikens Handbok, Del 7, Vol II, Sandvikens Jernverks Aktiebolag, 1964.
64
MATERIAL SCIENCE
For the design of reliable heat treatments, ITT-diagrams are indispensable. Depending on the kind of application, one or the other of the following four standard heat treatments are frequently applied: a) Quenching and tempering. b) Martempering. c) Austempering (Bainitizing). d) Isothermal soft-annealing. The schematic ITT-diagrams shown in Fig. 1.42 illustrate these procedures which, in detail, can be described as follows: a) Quenching and tempering: On quenching to temperatures well below the martensite temperature (MS), the aus tenite transforms to a structure consisting of martensite and smaller amounts of retained austenite. In this condition, the steel is extremely hard but also extremely brittle. In order to reduce its brittleness and increase its toughness, the steel is subsequently tempered at temperatures above MS. Due to temperature differences between surface and core or between thinner and thicker sections of the steel piece, cracks may arise on quenching. Hence, this type of heat treatment is less suited for big work pieces or pieces of complicate shape. b) Martempering: In this heat treating process, the risk of crack formation in the steel is avoided. The steel is quenched in a bath at a temperature slightly above MS and held in the bath until the core of the piece reaches bath temperature. During this period, the bath temperature falls gradually below MS, and the austenite transforms uniformly into martensite, whereupon the steel is removed from the bath and allowed to cool in air. Subsequently the steel is tempered to desired hardness. c) Austempering: In this heat treatment, the steel is quenched to a temperature below the bainitenose but above MS, at which temperature the austenite is transformed isothermally to bainite. Subsequent tempering is not required, since, even without tempering, bain ite offers a favourable combination of hardness and toughness. d) Isothermal annealing: In many alloyed steels, there is a pronounced minimum in the ending line of the ITT-diagram at relatively high temperatures. Assuming that a soft ferritic-
65
pearlitic structure is to be obtained, advantage may be taken of the timetemperature coordi nates 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 tempe rature 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.
Temperature
Temperature
Figure 1.42. Four standard procedures for the heat treatment of steels: a. Quenching and Tempering; b. Martempering; c. Austempering; d. Isothermal Annealing.*
* Atlas of Isothermal Transformation Diagrams, US Steel, Pittsburgh, 1951.
Temperature
Temperature
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.
InFLUEncE OF tHE MIcROstRUctURE On tHE PROpERtIEs OF StEEL
67
1.7 Inuence of the Microstructure on the Properties of Steel
1.7.1 Pearlite and Spheroidite
In pearlitic structures, lamellas of soft ferrite alternate with lamellas of hard cementite. Under the influence of shearing stresses, plastic deformation occurs essentially only in the soft ferrite lamellas where dislocations can move relatively easily. The thinner the ferrite lamellas are, the more restricted, by the rigid cementite lamellas, is the mobility of the dislocations. This explains why the ductility of pearlite increases and its hardness correspondingly decreases with the thickness of its ferrite lamellas as shown in the diagram in Fig. 1.43. By means of elongated soft-annealing, the cementite lamellas can be spheroidized. In the so transformed pearlite, also called spheroidite, small spherical cementite particles are embedded in a coherent ferrite matrix. See microstructure in Fig. 1.44. Spheroidite is the most ductile variety of pearlite because, in its coherent ferrite matrix, dislocations can move more freely, restricted only to a lesser degree by the spherical cementite inclusions.
Figure 1.43. Hardness of pearlite as a function of the distance between its lamellas.
Figure 1.44. Spheroidized cementite (spheroi dite) in a carbon steel with 1.30% C.
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 hard ness (a) and ductility (b) for a plain carbon steel having fine and coarse pearlitic as well as spheroidi tic microstructures.* Figure 1.46. Influence of carbon content on the mechanical properties of a plain carbon steel having a fine pearlitic microstructure.**
* ** Metals Handbook, Vol.4, 9.Edition, ASM, 1981. **
69
1.7.2 Martensite and Bainite
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.
* E.C.Bain: Functions of the Alloying Elements in Steel, AMS, 1939.
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
Figure 1.48. Mechanical properties as functions of tempering temperature for a steel, quenched from 850 C in oil, and containing 0.30% C; 0.25% Si; 0.60% Mn; 0.30% Cr; 3.30% Ni and 0.25% Mo.*
* Sandvikens Handbok, Del 7, Vol II, Sandvikens Jernverks Aktiebolag, 1964.
72
MATERIAL SCIENCE
Figure 1.49. Hardness as a function of tempering time and temperature for a waterquenched eutec toid plain carbon steel.*
Since the microstructure of bainite - similar to that of tempered martensite - is composed of finely dispersed cementite particles in a ferritic matrix, bainitic steels offer favourable combinations of hardness, strength and toughness. The diagram in Fig. 1.50 shows the influence of transformation temperature on microstructure and tensile strength of a low-carbon bainitic steel.
Figure 1.50. Relationship between tempera ture of maximum transformation rate and strength of low carbon bainitic steels.**
* E.C.Bain: Functions of the Alloying Elements in Steel, AMS, 1939. ** K.J. Irvine, Symposium: Steel-Strengthening Mechanisms, Zurich, May 5th and 6th 1969.
73
1.7.3 Temper Brittleness
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.
This chapter presents a short history of iron powder and a brief account of how world production and consumption of iron powder have developed over the last four decades. It describes in detail two production methods which together account for more than 90% of todays world production of iron and steel powders.
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
PRODUCTION OF IRON AND STEEL POWDERS
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 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Tons of powder 450 475 500 565 590 600 650 670 725 775 715 785 820 890 855 895 930 850 680 945
INTRODUCTiON
77
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.
78
2.2 The Hgans Sponge Iron Process

The Hgans sponge iron process, is essentially a chemical process in which finely divided iron ore is being reduced with coke breeze yielding a spongy mass of solid metallic iron, which can readily be comminuted to iron powder. The iron ore used at Hgans is a magnetite (powder Fe3O4) obtained from selected sources. This magnetite, which by nature contains only very small amounts of gang and has extremely low contents of sulfur and phosphorus, is being dressed and concentrated while still at the mining location and then delivered to Hgans in a highly pure state. As shown in the flow-sheet in Fig. 2.1, the process of transforming the magnetite to iron powder proceeds as described on the next page.
THE HGANS SPONGE IRON PROCESS
79
1. Reduction Mix of Coke Breeze and Limestone 2. Iron Ore 3. Drying 4. Crushing 5. Screening 6. Magnetic Separation 7. Charging in Ceramic Tubes 8. Reduction in Tunnel Kilns, Approximately 1200C 9. Discharging
10. Coarse Crushing 11. Storage in Silos 12. Chrusing 13. Magnetic Separation 14. Grinding and Screening 15. Annealing in Belt Furnace, Approximately 800-900 C 16. Equalising 17. Automatic Packing 18. Iron Ore 19. Reduction Mix
Figure 2.1. Flow-sheet for the Hgans Sponge Iron Process.
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).
THE HGANS SPONGE IRON PROCESS
81
Subsequently, the powder is passing a belt furnace (15) where it is soft-annealed at 800 - 1000C in hydrogen, and its remaining contents of carbon and oxygen are reduced to a very low level. During annealing, the powder agglomerates to a very crumbly cake which is gently comminuted again in a special mill (not shown on the flow-sheet). The so treated powder has good compressibility and high green strength. Many belt furnaces are, of course, available to take care of the huge volume of iron powder to be treated. The powder from the belt furnaces is collected and homogenized (16) in lots of 60 or 120 tons. Each lot is carefully checked with respect to specified properties and packed (Fig. 2.2) and stored, ready for shipment (17).
Figure 2.2. Hgans standard packing, flexbag 1000 kg.
82
2.3 The Hgans Water-Atomizing Process

A modern atomizing plant of Hgans AB is located at Halmstad, a small coastal town 80 km north of Hgans. The water-atomizing process, as practiced there, can be followed on the flow-sheet shown in Fig. 2.3, and is described in detail on the next page.
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
Figure 2.3. Flow-sheet for the Hgans Water-Atomizing Process.
THE HGANS WATER-ATOMiZiNG PROCESS
83
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.
84
2.4 Hydrogen-reduced Iron Powder

Hydrogen-reduced iron powder is manufactured from selected iron oxides (Fe203) and is widely used for low density structural parts and bearings as well as sintered friction materials. The powder particles are markedly irregular in shape with a sponge-like microstructure. The irregular particle shape creates very good mechanical interlocking which results in very high green strength. The sponge-like particle structure provides a high surface area that promotes sintering, resulting in increased shrinkage. Therefore, the hydrogen-reduced iron powders provide high green strength, introduce porosity with fine pores while maintain good strength as well as provide an alternative method for controlling the dimensions of parts.
Figure 2.4 Flow-sheet for the Hydrogen-reduced Iron Powder.
Figure 2.5 Irregular particle shape of hydrogen-reduced iron powder.
ALLOYiNG METHODS
85
2.5 Alloying Methods

In order to achieve hardenable sintered ferrous materials, carbon and other suitable alloying elements, like e.g. copper, nickel, and molybdenum, have to be introduced. While carbon is normally admixed to the iron powder in the form of graphite, metallic alloying elements are commonly introduced by either of the following two methods: Method 1: Water atomization of the liquid iron alloy, resulting in a homogeneously alloyed powder. Method 2: Mechanically blending plain iron powder with the respective alloying elements in powder form, and letting the actual alloying process take place during sintering of the parts compacted from the powder mix. Both methods have their advantages and disadvantages:
Homogeneously alloyed powders.
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.
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.
Figure 2.6. Influence of some alloying elements on the hardness of iron.
DISTALOY AND STARMIX
87
2.6 Distaloy and Starmix

In order to eliminate the segregation problem with powder mixes, Hganshas developed two special processes for the production of segregation-resistant iron powder mixes. A large variety of standard and tailor-made powder mixes produced according to these processes are offered under the trade-names Distaloy and Starmix. The Distaloy process can be described as follows: Alloying elements used in the Distaloy process are mainly copper, nickel and molybdenum (but not graphite!) in the form of very finely grained powders. The process starts with weighing-in a production lot of 30 tons of iron powder and alloying powders in exactly controlled proportions. This lot is mixed in a double-cone mixer. Special precautions are taken to prevent segregation of the mix when discharged from the mixer. The so produced powder mix is heat-treated in a continuous furnace under a reducing atmosphere at a temperature somewhat below the melting point of the lowest-melting alloying element. During this heat-treatment, the fine particles of the added alloying elements are safely bonded to the surfaces of the coarser iron powder particles.
88
See SEM-photographs in Fig. 2.7. Due to inter-diffusion between the diffusion bonded alloying particles and the iron particles, the latter become, to a certain extend, locally pre-alloyed.
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.
89
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.
90
The Starmix processes uses special types of binders to glue graphite and lubricants to the iron powder particles during the mixing procedure. See SEM-photograph in Fig. 2.9.
Figure 2.9. SEM-photograph showing fine graphite particles glued, in the Starmix process, to the surface of an iron powder particle (NC100.24).
91
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. Loss of Graphite, %
Conventional mix Starmix 500
Figure 2.10. Graphite loss in air de-dusting test for a Starmix treated powder mix and for a conventional powder mix.
Figure 2.11. Relative fre quency of dimensional changes during sintering for a Starmix treated powder mix and for a conventional powder mix.
Frequency, %
Conventional mix
Starmix 500
92
Combined carbon, %
1.20 1.10 1.00 0.90 5000
Starmix 500
Figure 2.12. Variation of carbon content in sintered parts during mass production, comparing Starmix treated mixes and conventional mixes of NC100.24 + 1.2%
Conventional mix 10000 15000
graphite.
Number of parts
Apparent Density, g/cm3

2.80 2.75 2.70
32 30 28 26 24
Flow, sec/50 g
Green Density, g/cm3

6.70 6.60 13 12
Green Strength, N/mm2
Tensile Strength, N/mm2

400 390 380 4.0 3.0 2.0
Elongation, %
Figure 2.13. Comparison of powder properties and sintered properties for Starmix treated mixes and conventional powder mixes of NC100.24 + 0.8% graphite + 0.8% Zn-stearate.
Hardness, HV10
140 130 +0.20 +0.10
Dimensional Charge, %
Conventional mix
Starmix 500
93
Our Starmix Grades

Starmix 500
This is the standard Starmix for improved tolerances. Better control of additives such as carbon crucial for heat-treated parts For improved tolerances in less demanding applications
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
This Starmix provides tighter tolerances in components with high graphite contents. Delivers properties particularly suitable for valve guide applications
Starmix BOOST
Starmix BOOST is designed for parts with highly complex shapes. Can be tailored for very demanding applications to meet tough requirements on productivity, consistency and part precision Excellent lubrication and high AD New lubrication system without Zinc ensures a stain-free finished component
The most important prerequisite of a successful mass production of high-quality sintered parts is a high and consistent quality of the powder from which the parts are to be made. The characteristics of the powder decide its compacting and sintering behavior and, eventually, about the properties of the nished product.
Characteristics of Iron and Steel Powders
3.1 3.2
General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . 96 Properties of Hgans Iron Powders . . . . . . . 100
96
CHARACTERISTICS OF IRON AND STEEL POWDERS
3.1 General Aspects

Iron and steel powders - as well as other metal powders - used in the production of sintered parts can be characterized by three categories of properties:
1. Metallurgical properties
chemical composition and impurities microstructure microhardness particle size distribution external particle shape internal particle structure (particle porosity) flow rate bulk density compressibility, green strength and spring back
2. Geometrical properties
3. Mechanical and physical properties
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
97
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).
Particle Size Distribution
is determined by sieve analysis or laser diffraction if particle sizes are above 45 m (minimum screen aperture). Finer powders are analyzed by means of laser diffraction methods.
External Particle Shape
is analyzed by means of scanning electron microscopy (SEM). (See Fig. 3.1 left).
Internal Particle Structure
is analyzed by means of metallographic techniques. (See Fig. 3.1 right).
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).
98
CHARACTERISTICS OF IRON AND STEEL POWDERS
Flow Rate
or rather its reciprocal value, is the time in seconds which an amount of 50 g dry powder needs to pass the aperture of a standardized funnel. Flow rate is influenced by type and amount of lubricant admixed to the powder. The flow rate indicates how fast a compacting tool can be filled with powder, and thus is a limiting factor in the compacting cycle of the powder press.
Bulk Density (Apparent Density)
is determined by filling the powder through a standardized funnel into a small cup, leveling-off the surplus powder on top of the cup, and dividing the weight of powder contained in the cup by the cup volume (25 cm3). Apparent density is influenced by type and amount of lubricant admixed to the powder. The apparent density of the powder decides about the required filling depth of the compacting tool.
Compressibility
is the name of a curve obtained by plotting the compact densities of a series of small cylindrical powder compacts ( 25 mm), over the applied pressures. Compact density is the weight of a powder compact divided by its bulk volume. Compact density is influenced by type and amount of lubricant admixed to the powder. Green density is the compact density of a small cylindrical powder compact ( 25 mm) pressed with a standardized pressure (typically 4.2 tons/ cm2 or 600 N/mm2). The compressibility of the powder decides about how high a compacting pressure is needed to achieve a desired compact density.
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
99
Spring Back
is the elastic expansion of a cylindrical powder compact ( 25 mm) after ejection from the compacting die. Its value is expressed as the difference between the Outer Diameter of the compact and the Inner Diameter of the (empty) die divided by the ID of the die. Spring back increases with increasing compacting pressure and is influenced by type and amount of lubricant admixed to the powder and by the elasticity coefficient of the die material in which the powder is compacted. The spring back value is important for calculating the exact dimensions of the compacting tool in relation to the required dimensions of the compact.
100 CHARACTERISTICS OF IRON AND STEEL POWDERS
3.2 Properties of Hgans Iron Powders

As has been mentioned in the preceding chapter, Hgans produces two different kinds of ferrous powders: sponge-iron powders, and water-atomized (unalloyed and low-alloyed) iron powders. Typical structural differences between these two kinds of powders appear on SEM-photographs and cross-sections of representative powder particles shown in Fig. 3.1. The external shapes of both particles are irregular and fairly similar to one another. But the sponge iron particle has, as its name suggests, a spongy internal structure, while the water-atomized particle is internally compact. Both kinds of powder are specially treated to yield various standard grades having different properties to fit different applications. These standard grades are also used in various press-ready powder mixes, i.e. blended with lubricants and alloying additives like graphite, copper, nickel, molybdenum, phosphorous and others. Properties of three sponge-iron grades NC100.24, SC100.26, MH80.23, and of two water-atomized iron powder grades ASC100.29, ABC100.30, are presented in the diagrams in Figs. 3.2 and 3.3 and in Table 3.1.
PROPERTIES OF HgANS IRON POWDERS 101
Table 3.1. Properties of some Hgans Iron Powders

Powder grade Approx. particle size range m
20 180 20 180 40 200 20 180 30 200
Apparent density g/cm3
Flow s/50g
H2-loss %
C %
Green density
1)
Green strength
1)
g/cm3
2.45 2.65 2.30 2.95 3.00 31 29 34 24 24 0.20 0.12 0.32 0.08 0.06 <0.01 <0.01 0.08 0.002 0.001 7.02 7.12 6.292) 7.20 7.28
N/mm2
47 40 24 3) 41 44
NC100.24 SC100.26 MH80.23 ASC100.29 ABC100.30

1) 2)
compacted at 600 N/mm2 in a lubricated die. compacted at 4.2 t/cm2 in a lubricated die. 3) measured at a green density of 6.0 g/cm2
Figure 3.2. Particle size distribution, flow and apparent density of five different iron powder grades.
Compressibility at 600 MPa

g/cm3 MPa
Green strength at 600 MPa (1) (2) (3) (4) (5) ABC 100.30 ASC 100.29 SC 100.26 NC 100.24 MH 80.23
7,5
35
7,0
30
6,5
25
6,0
20
5,5
15
5,0
(1)
(2) (3) (4) Powder type
(5)
10
(1)
(2) (3) (4) Powder type
(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
is especially designed to match the requirements for self-lubricating bearings. Its particle size range is chosen to give a pore structure optimal for this application.
MH80.23 can also be added to powder mixes in small quantities to dramatically improve green strength.
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
is a water-atomized iron powder of outstanding compressibility and chemical purity. It is especially well suited for the production of high-density structural components. Densities of up to 7.4 g/cm3 are achievable in one single pressing operation.
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 lubri cants.
Apart from unalloyed iron powder grades, Hgans produces water-atomized low-alloyed steel powders trade-name AstaloyTM and a variety of diffusion alloyed steel powder trade-name Distaloy. (The Distaloy process has been described in some detail in the preceding chapter).
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
20 180 20 180 20 180 20 150 20 180 20 180 20 180
Apparent density g/cm3
Flow s/50g
H2-loss %
C %
Green density
1)
Green strength
1)
g/cm3
3.00 3.00 2.85 2.80 3.05 3.05 3.08 26 25 27 27 25 27 25 0.10 0.10 0.13 0.12 0.10 0.10 0.10 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 7.07 2) 7.10 7.04 7.10 7.10 7.17 7.07
N/mm2
20 2) 22 26 40 31 36 29
Astaloy LH Astaloy Mo Astaloy CrA Distaloy SA Distaloy DC Distaloy AQ Distaloy HP

1) 2)
compacted at 600 N/mm2 in a lubricated die. with 0.6% Kenolube admixed
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
is a water-atomized steel powder offering very good performance for a wide range of applications. Pre-alloyed with 1.8% Cr it offers good hardenability, which can be further increased by adding Cu in the mix.
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 DC (DC = Dimensional Control)
is based on Astaloy Mo to which 2% finely dispersed nickel powder has been diffusion-bonded. It thus contains 2% Ni + 1.47% Mo. Parts compacted from Distaloy DC admixed with graphite exhibit high strength and very small scattering of dimensions after sintering, independent of compact density; and the sintered parts respond very well to subsequent heat-treatment.
Distaloy AQ
is based on the water-atomized grade ASC100.29, to which 0.5% Mo and 0.5% Ni have been diffusion bonded. This alloy has been specifically developed for subsequent heat-treatment, such as case hardening and through hardening.
Distaloy HP (HP = High Performance)
is based on Astaloy Mo to which 4% finely dispersed nickel powder and 2% finely dispersed copper powder have been diffusion-bonded. It thus contains 4% Ni + 2% Cu + 1.41% Mo. Parts compacted from Distaloy HP admixed with graphite exhibit dimensional changes close to zero during sintering and adopt very high strength values. More detailed information about all powder grades described above, as well as about many other interesting powder grades and powder mixes can be obtained from Hgans in form of special brochures and electronic data.
107
108 HGANS HANDBOK FOR SINTERED COMPONENTS
Index
A
activation energy. . . . . . . . . . . . . . 30, 32 alloying methods . . . . . . . . . . . . . . . . . 85 amorphous solids . . . . . . . . . . . . . . . . . . 8 stable short-range order . . . . . . . . . . . . 8 apparent density . . . . . . . . . . . . . . . . . 98 atoms per unit cell . . . . . . . . . . . . . . . . 17 austempering. . . . . . . . . . . . . . . . . 64, 65
cast-iron. . . . . . . . . . . . . . . . . . . . . . . . 46 CCT-diagrams. . . . . . . . . . . . . . . . 58, 66 cementite formation . . . . . . . . . . . . . . . 54 ceramic . . . . . . . . . . . . . . . . . . . . . . . . . 80 chemical composition and impurities 96 coke breeze . . . . . . . . . . . . . . . . . . . . . . 78 compact density . . . . . . . . . . . . . . . . . . 98 compressibility . . . . . . . . . . . . . . . . . . . 98
compressibility green strength. . . . . . 96 continuous cooling transformation diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . 58 cooling curves . . . . . . . . . . . . . . . . . 38, 61 cooling rate . . . . . . . . . . . . . . . . . . . 50, 51 crack propagation. . . . . . . . . . . . . . . . 73 crude powder . . . . . . . . . . . . . . . . . . . . 80
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 . . . . . . . . . . . . . . . . . . . . .
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
carbon-steel . . . . . . . . . . . . . . . . . . . . . 46
electric arc furnace . . . . . . . . . . . . . . . 83
INDEX 109
electron gas . . . . . . . . . . . . . . . . . . . . . . 11 entropy of mixing . . . . . . . . . . . . . . . . . 35 eutectic system . . . . . . . . . . . . . . . . 38, 39 equilibrium diagram. . . . . . . . . . . . . . 46
external particle shape . . . . . . . . . 96, 97
isothermal . . . . . . . . . . . . . . . . . . . . . . . 64 annealing . . . . . . . . . . . . . . . . . . . 64, 65 transformation . . . . . . . . . . . . . . . . . . 59 transformation diagrams. . . . . . . . . . 58 isothermal soft-annealing. . . . . . . . . . 64 ITT-diagrams . . . . . . . . . . . . . . . . . 58, 59
interstitial atoms . . . . . . . . . . . . . . . . . 21
FCC . . . . . . . . . . . . . . . . . . . . . . . . 17, 19
Fe- Fe3C - diagram . . . . . . . . . . . . . . . .51
ferrite. . . . . . . . . . . . . . . . . . . 46, 49, 50
lever-rule of phases. . . . . . . . 40, 42, 44
ferrite formation . . . . . . . . . . . . . . . . . 53
first and second law of thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 free energy. . . 33, 34, 36, 37, 38, 54, 55
magnetic separators . . . . . . . . . . . . . . . 80 martempering . . . . . . . . . . . . . . . . 64, 65
flow rate . . . . . . . . . . . . . . . . . . . . . 96, 98
magnetite . . . . . . . . . . . . . . . . . . . . . . . 78 martensite . . . . . . . . . . . . . . . . . . . . 49, 53
mechanical and physical properties .. 96 metallurgical properties . . . . . . . . 96, 97 microhardness. . . . . . . . . . . . . . . . . . . 96 microstructure . . . . . . . . . . . . . . . . . . . 96 Miller indices. . . . . . . . . . . . . . . . . . . . 13
gases. . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 grain boundaries. . . . . . . . . . . . . . 20, 31
geometrical properties . . . . . . . . . . 96, 97 green density . . . . . . . . . . . . . . . . . . . . 98 green strength . . . . . . . . . . . . . . . . . . . 98 growth . . . . . . . . . . . . . . . . . . . . . . . . . . 59
non-equilibrium diagram. . . . . . . . . . 49
non-metallic inclusions . . . . . . . . . . . . 22
hardness of pearlite . . . . . . . . . . . . . . . 67 HCP . . . . . . . . . . . . . . . . . . . . . 12, 17, 19 heat treatment of steel . . . . . . . . . . . . . 63 heterogeneous system . . . . . . . . . . . . . 33 homogeneously alloyed powders . . . . . 85 hydrogen-reduced iron powder . . . . . 84
O P
octahedral . . . . . . . . . . . . . . . . . . . . . . . 53
Helmholtz free energy . . . . . . . . . . . . . 34
packing density . . . . . . . . . . . . . . . . . . 17
parameters of state . . . . . . . . . . . . . . . 33 pearlite. . . . . . . . . 48, 61, 62, 67, 68, formation . . . . . . . . . . . . . . . . . . . . . . lamellas. . . . . . . . . . . . . . . . . . . . . . . nose. . . . . . . . . . . . . . . . . . . . . . . . . . phase diagrams . . . . . . . . . . . . . . . . . .
hypoeutectoid steels . . . . . . . . . . . . . . . 53
particle size distribution. . . . . . . . 96, 97
internal energy . . . . . . . . . . . . . . . . . . . 33 interstices . . . . . . . . . . . . . . . . . . . . . . . 53
inter-metallic phase Fe3C . . . . . . . . . . 46
internal particle structure . . . . . . . 96, 97
69 55 55 61 33 plastic deformation metal crystals. . . 22 polycrystalline structure . . . . . . . . . . . 20
110 HGANS HANDBOK FOR SINTERED COMPONENTS
powder mixes . . . . . . . . . . . . . . . . . . . . 86
tie-line . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Q R
transformation diagrams . . . . . . . . . . 58 transition points A3 and A1 . . . . . . . . . 50 trostite . . . . . . . . . . . . . . . . . . . . . . . . . . 50 TTT-diagrams . . . . . . . . . . . . . . . . 49, 58 twinning. . . . . . . . . . . . . . . . . . . . . . . . 20
quenching and tempering. . . . . . . . . . 64
reduction mix . . . . . . . . . . . . . . . . . . . . 80 rotary dryers . . . . . . . . . . . . . . . . . . . . 80
ring-shaped nozzle . . . . . . . . . . . . . . . . 83
U V
undercooling . . . . . . . . . . . . . . . . . . 50, 51 unit cell . . . . . . . . . . . . 13, 15, 17, 18, 53
self-tempered martensite . . . . . . . . . . . 57 solidus . . . . . . . . . . . . . . . . . . . . . . . . . . 36 sorbite . . . . . . . . . . . . . . . . . . . . . . . . . . 50 space lattice . . . . . . . . . . . . . . . . . . . . . 13 spheroidite. . . . . . . . . . . . . . . . . . . . . . 67 spheroidized cementite . . . . . . . . . . . . 67 sponge-iron powders. . . . . . . . . . . . . 100 spring back . . . . . . . . . . . . . . . . . . . . . . 99 sponge-iron process . . . . . . . . . . . . . . . 78 stable system . . . . . . . . . . . . . . . . . 46, 47
slip plane . . . . . . . . . . . . . . . . . . . . . . . 22
vacancies . . . . . . . . . . . . . . . . . . . . . . . . 21
W Y
water-atomized . . . . . . . . . . . . . . . . . 100
water-atomized process . . . . . . . . . . . . 82
yield point . . . . . . . . . . . . . . . . . . . . . . . 25
stacking sequence . . . . . . . . . . . . . . . . .18
starmix . . . . . . . . . . . . . . . . . . . . . . . . 93 stationary case. . . . . . . . . . . . . . . . . . . 27 substitutional atoms . . . . . . . . . . . 21, 26 supersaturation . . . . . . . . . . . . . . . . . . 54 surfaces . . . . . . . . . . . . . . . . . . . . . . . . . 31
temper brittleness . . . . . . . . . . . . . . . . 73 tetragonal distortion. . . . . . . . . . . . . . 52 tetragonal martensite lattice . . . . . . . . 52
tempered martensite. 56, 57, 70, 72, 73 tetragonally distorted ferrite . . . . . . . 56 tetrahedral. . . . . . . . . . . . . . . . . . . . . . 53 thermal analysis . . . . . . . . . . . . . . . . . . 38 thermodynamic . . . . . . . . . . . . . . . . . . 33 equilibrium . . . . . . . . . . . . . . . . . . . . 33 state. . . . . . . . . . . . . . . . . . . . . . . . . . 33
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. With its leading position in metal powder technology, Hgans is perfectly placed to help you explore those possibilities as your application project partner. Power of Powder is being applied far beyond its traditional role in the production of components for vehicles. Iron powder is used in food fortification to combat anaemia. Nickel powders are vital ingredients in valve coatings to enhance wear resistance. Specially formulated iron-based powders offer new solutions for high-temperature brazing. Soft Magnetic Composites with 3D magnetic properties are opening the way for innovative electric motors. In fact, metal powder technology generates virtually endless possibilities. To find out how you can apply the Power of Powder, please contact your nearest Hgans office.
www.hoganas.com/pmc

Material and Powder Properties - Dec - 2013

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Material and Powder Properties - Dec - 2013

Загружено:

Авторское право:

Доступные форматы

1

Material and powder properties

Hgans Handbook for Sintered Components

Material and Powder Properties

Production of Sintered Components

Design and Mechanical Properties

Material and powder properties

Production of Iron and Steel Powders . . . . . . . . . .75

Characteristics of Iron and Steel Powders . .. .. .. .. .. .. ..95

1.1 1.2 1.3 1.4 1.5 1.6 1.7

1.1 The Natural States of Matter

THE NAtURAL StAtEs OF MAttER

1.2 The Crystalline Structure of Metals

1.2.1 Bonding Between Atoms

THE CRYstALLInE StRUctURE OF MEtALs

Ratio of Lattice Constants c/a

2,932 2,884 3,147 3,165

Ni -Co Cu -Fe Ir Pt Al Au Ag Pb 1000 467

-Co Zn -Ti Cd Mg (ideal)

THE CRYstALLInE StRUctURE OF MEtALs

1.2.2 Ideal Crystals

Figure 1.3. Lattice planes (LP) in a crystal lattice.

THE CRYstALLInE StRUctURE OF MEtALs

Figure 1.4. Derivation of Miller indices.

THE CRYstALLInE StRUctURE OF MEtALs

THE CRYstALLInE StRUctURE OF MEtALs

1.2.3 Real Metal Crystals

THE CRYstALLInE StRUctURE OF MEtALs

1.2.4 Plastic Deformation of Metal Crystals

THE CRYstALLInE StRUctURE OF MEtALs

THE CRYstALLInE StRUctURE OF MEtALs

1.3 Diffusion in Metals

1.3.1 Laws of Diffusion

1.3.2 A Particular Solution of the Diffusion Equation

1.3.3 The Diffusion Coefcient

Surface 10 -12 D, m2/s (log-scale) Grain Boundary 10 -14

10 -16 -Fe 10 -18

Activation Energy Q KJ/mol kcal/mol

500 900 900 1100 500 900 900 1100 500

BInARY PHAsE DIAGRAms OF MEtALs

1.4 Binary Phase Diagrams of Metals

d G = d(E - ST + PV) = d F + d(PV) (1.8)

dG = d(E ST + PV) 0 (1.9)

Solid Liquid Solid m.p Liquid Temperature

BInARY PHAsE DIAGRAms OF MEtALs

1.4.2 Derivation of Binary Phase Diagrams

BInARY PHAsE DIAGRAms OF MEtALs

GS a) Free Energy of Phases G(c) GL

T=TMB GS GL b) GS CL GL GS d) CS TMA>T TMB>T>TMA

* G.E.R. Schulze: Metallphysik, Akademie-Verlag, Berlin 1974.

BInARY PHAsE DIAGRAms OF MEtALs

BInARY PHAsE DIAGRAms OF MEtALs

Pre-eutect (19,2% Sn)

Eutecticum (61,9% Sn)

Dark (19,2% Sn)

Light (97,5% Sn)

BInARY PHAsE DIAGRAms OF MEtALs

1.4.3 The Lever-Rule of Phases

m0 = mL + mS c0m0 = cLmL + cSmS mL (c0 cL) = mS (cS - c0)

x = (c0 cL) and y = (cS c0) mLx = mS y

Lever-rule of phases: m0 mL mS c0m0 cLmL cSmS

1.4.4 Crystal Segregation

* Metals Handbook, Vol.4, 9.Edition, ASM, 1981.