Tempering of Engineering Steels
Chen Zhu Trinity College Literature Review
09/2005
Table of Contents 1ˊ Introduction............................................................................................................... 1 2ˊ Hardening of steels.................................................................................................... 1 3ˊ Martensitic transformation...................................................................................... 3 3.1
The crystal structure of martensite ........................................................................ 4
3.2
The crystallography of martensitic transformations............................................ 4
3.3
The morphology of ferrous martensites................................................................. 7
3.4
Retained austenite in martensitic transformation ................................................ 9
4ˊTempering of martensite .......................................................................................... 10 4.1
Quench ageing of carbon steels............................................................................. 10
4.2
Tempering Process of carbon steels.......................................................................11
4.2.1
Room temperature ageing before tempering.....................................................11
4.2.2
Tempering stage 1 ............................................................................................... 12
4.2.3
Tempering stage 2 ............................................................................................... 12
4.2.4
Tempering stage 3 ............................................................................................... 13
4.2.5
Tempering stage 4 - secondary hardening ........................................................ 15
4.3
Role of carbon content........................................................................................... 17
4.4
Mechanical properties of tempered plain carbon steels ..................................... 17
4.5
The effect of alloying elements on the formation of iron carbides .................... 18
4.6
Nucleation and growth of alloy carbides ............................................................. 19
4.7
Effects of alloying elements ................................................................................... 20
4.8
Tempering parameter ............................................................................................ 26
5ˊ Remaining questions............................................................................................... 28 References ........................................................................................................................ 29
1.
Introduction The hardening of steels, by plunging the metal red-hot into water (to produce martensite), and its
toughening, by tempering the quench-hardened metal at a moderate temperature, have been known empirically and used for thousands of years.[1] Tempered martensite is a very hard yet tough material, which finds many engineering applications where wear resistance is vital. Typically its uses include gears, pinions, shafts, crankshafts and piston rods for engines.[2] The process of tempering, a heat treatment which reduces brittleness and increases the toughness of hardened steel, has been studied in great detail during the past century. In particular, the addition of alloying elements to maintain hardness at higher tempering temperatures has been analyzed. As a result, it is known that common alloy additions to steel, such as chromium, manganese and nickel, retard the kinetic of softening during tempering. However, the exact atomic mechanisms of these retardation processes are not fully understood. The emphasis of this report will be the introduction of martensitic transformation, tempering of martensite and the roles of the added alloy elements during the martensite tempering.
2.
Hardening of Steels Steels are usually heat-treated by raising them through the eutectoid transformation to a temperature
within the single-phase austenitic field, holding them there long enough to dissolve the cementite and disperse the carbon uniformly, and then cooling to room temperatures. The rate of this cooling determines the resultant microstructure of the material.[3] Slow cooling in a furnace, referred to as annealing, results in a coarse ferrite-pearlite structure; somewhat faster air cooling know as normalizing gives a fine ferrite-pearlite or bainitic structure; and fast cooling or quenching in a liquid bath (oil, brine or water) gives a martensite microstructure.[3] The effect of cooling rate on microstructure is illustrated in Figure 1.1.[4]
1
Figure 1.1 the variation of microstructure as a function of cooling rate for an eutectoid steel.[4]
Figure 1.2 Isothermal Transformation diagram of a hypo-eutectoid steel: 0.35wt% C, 0.37wt% Mn, A=austenite, F=ferrite, C=cementite, M=martensite. [5] 2
Figure 1.3 Isothermal Transformation diagram of a low alloy steel(4340): 0.42wt% C, 0.78wt% Mn, 1.79wt% Ni, 0.80wt% Cr, 0.33wt% Mo. A=austenite, F=ferrite, C=cementite, M=martensite.[5] The critical cooling rate shown in Figure 1.1 must be achieved to produce a 100% martensite microstructure.[4] This critical cooling rate, however, varies between ‘high purity’ plain carbon steels and alloy steels because of the effect of alloy additions. The Time Temperature Transformation (TTT) curve shown in Figure 1.2[5] is for the isothermal transformation of a 0.35wt%1 C hypo-eutectoid steel. A comparable TTT curve for the isothermal transformation of a low alloy SAE 4340 steel (0.4wt% C) is shown in Figure 1.3.[5] In plain carbon steels the reaction near the pearlite nose of the TTT curve is rapid, so a fast cooling rate is required to achieve an effective quench and martensite microstructure. This has practical disadvantages, as it is not possible to quench the interior of thick sections of materials. Large temperature gradients are also established across the material that give transformation stresses and can lead to quenching cracks.[6] The addition of alloying elements (Mn, Ni, Cr and Mo) to the steel retard the pearlite reaction, so martensite can be achieved at lower quenching rates. This is seen as a shift in the nose of the pearlite reaction to the right in the TTT curve shown in Figure 1.2. Other changes, such as separation of the pearlite and bainite reactions into two distinct ‘c-shaped’ curves, can be seen in the TTT curve. Detailed analysis of these effects can be found in the literature and textbooks but it is of little relevance to this investigation.[1, 4]
3. 1
Martensitic Transformation All compositions in this review are in weight percentages. 3
3.1.
The Crystal structure of martensite Martensite in steels is a supersaturated solid solution of carbon in ferritic iron. For alloys which have a
low martensite-start temperature or a high carbon concentration, the carbon atoms tend to be order in such a way that the crystal structure changes from body-centered cubic to body-centered tetragonal. The tetragonality of the ordered martensite, measured by the ratio between the axes, c/a, increases with carbon content:
c 1 0.045wt %C [7, 8] a The formula is not applied to the alloys which have a high martensite-start temperature (Ms) or a low carbon concentration.[9] Under those circumstances, the structure of the martensite will remain body-centered cubic. There used to be two alternative explanations: one was that the carbon distribution along the three axes tended to become the same due to the thermal disordering, the other, the carbon atoms cluster on defects. It has now been observed directly that the carbon atoms moved and segregated to dislocations during quenching.[10] In steels less that 0.2wt% C, the carbon content left in the martensite is small. As a result, their structure is cubic.[11]
Figure 2.1 Variation of the lattice parameters of martensite and austenite as a function of carbon content.[12]
3.2.
The crystallography of martensitic transformations Martensitic transformations are first order, diffusionless, shear (displacive) solid state structural changes.
The change in crystal structure is achieved by a homogeneous lattice deformation of the parent phase. To minimize the strain energy the martensite forms as thin plates on particular crystallographic planes known as the habit planes. The consequences of this mechanism can be seen macroscopically because the shape of the transformed region changes, the strain being a combination of shear (~0.25) parallel to, and a dilatational strain 4
(~0.03) normal to the habit plane[13] .
Figure 2.1 Bain distortion for a face-centered cubic lattice transforming to a body-centered cubic lattice. The body-centered tetragonal cell is outlined in the face-centered cubic structure in (A), and shown alone in (B). The Bain distortion converts (B) to (C)[14] The basis of crystallographic pheonomenological theory of martensitic transformation is that in a martensitic transformation there should be an undistorted and unrotated interface between the martensite and the parent phase formed as a result of an invariant plain strain. Invariant plain strain is a homogeneous distortion in which the displacement is proportional to the distance from the invariant plane (habit plane). The Bain strain[15] implies the following orientation relationship between the parent and product lattices:
[011] fcc //[001]bcc [110] fcc //[100]bcc [110] fcc //[010]bcc Figure 2.1 shows the Bain distortion in steels. But in fact, the experimentally observed orientation relationships are irrational, e.g., close to the Kurdjumov-Sachs orientation relationship[16],
{111} fcc //{011}bcc 5
101 ! fcc // 111 !bcc A combination of the Bain strain and a slight rotation reduces the strain associated with the transformation. The Bain stain (B) and rotation (R) constitute the homogeneous transformation strain. The observed surface relief is an invariant-plane strain (P) involving a shear on the habit plane, and a small expansion normal to the habit plane.[17, 18] To reconcile the experimental surface relief with theory, the existence of a homogeneous invariant-plane strain (P2) is postulated:
BR
PP2
The displacements due to this additional deformation are not observed macroscopically because they are cancelled out by periodically slipping or twinning the martensite. This is called the lattice invariant deformation because neither slip or twinning change the nature of the lattice.[19] Figure 2.2[20] shows schematically the two type of lattice invariant deformation occurring within a martensite plate.
Figure 2.2 Formation of martensite plate, illustrating two types of lattice deformation: slip and twinning
6
3.3.
The morphology of ferrous martensites The different crystallographic features result from a number of transformation mechanisms that operate
to establish the interface between the austenite and martensite structures under varying conditions. These various mechanisms must be consistent with the description of the transformation provided by the crystallographic theory. The crystals of martensite may be arranged in one of two major morphologies: lath or plate.[21, 22, 23] Lath martensite crystals are aligned parallel to each other in volume elements termed blocks. In contrast, plate martensite crystals form nonparallel arrays and have irrational habit planes of high multiplicity, and thus have much more irregular microstructural arrangements.[24] The morphology of ferrous martensites is characterized by its complex variability, other substructural features have been found in ferrous carbon-bearing martensite. [25, 26].
Low carbon martensite[27] Habit plane close to {111} fcc Kurdjumov-Sachs relationship: {111} fcc //{011}bcc , 101 ! fcc // 111 !bcc Referred to as lath martensite This type of martensite is found in plain carbon and low alloy steels up to about 0.5% carbon. The morphology is lath or plate like, where the laths are very long and about 0.5 m wide. These are grouped together in packets with low angle boundaries between each lath, although a minority of laths is separated by high angle boundaries. In plain carbon steels practically no twin-related laths have been detected, while in iron-nickel alloys adjacent laths are frequently twin-related. Internally, the laths are highly dislocated and it is frequently difficult to resolve individual dislocations which form very tangled arrays. Twins are not observed to occur extensively in this type of martensite.
Medium carbon martensite[27] Habit plane close to {225} fcc Kurdjumov-Sachs relationship: {111} fcc //{011}bcc , 101 ! fcc // 111 !bcc Referred to as acicular martensite It’s characteristic morphology is that of lenticular plates, a fact easily demonstrated by examination of
7
plates intersecting two surfaces at right angles. These plates first start to form in steels with about 0.5wt% C, and can be concurrent with lath martensite in the range 0.5wt%- 1wt% C. Unlike the laths, the lenticular plates form in isolation rather than in packets, on planes approximating to {225} fcc and on several variants within one small region of a grain, with the result that the structure is very complex. The burst phenomenon probably plays an important part in propagating the transformation, and the austenite is thus not as uniformly or as efficiently eliminated as with lath martensite. The austenite is not as uniformly or as efficiently as with lath martensites. The physical difference could be connected with the fact that higher percentages of retained austenite occur as the carbon level is increased, as shown in Figure 2.3[28], and the martensite is predominantly lenticular. The micro-twinning is found predominantly in this type of martensite, which forms at lower Ms temperatures, as the carbon content increases.
Figure 2.3 The effect of carbon concentration on the relative fraction of lath martensite, the Ms temperature and the volume fraction of retained austenite.[29]
High carbon martensite[27] Habit plane close to {259} fcc Nishiyama-Wasserman relationship: {111} fcc //{110}bcc , 112 ! fcc // 110 !bcc When the carbon content is more than 1.4wt%, the orientation relationship changes from K-S to N-W, and the habit plane changes to around {259} fcc . The change is not detectable microscopically as the morphology is
8
still lenticular plates which form individually and are heavily twinned[30, 31]. This type of martensite obeys more closely to the theoretical predictions than the mechanism.[32] The
{225} martensite. The plates are formed by the burst
{259} martensite only forms at very high carbon levels in plain carbon steels or caused to
occur at much lower carbon contents by adding metallic alloying elements.[33] These high carbon martensite steels are not widely used in engineering according to their low Ms temperature and brittleness in tempered form..
3.4.
Retained Austenite in martensitic transformation In alloys with Mf below room temperature, part of the austenite phase remains untransformed after quenching to room temperature, which is referred to as retained austenite. The amount of retained austenite depends on the conditions of quenching.[34] In earlier days, Tamaru and Sekito[35] studied the problem using carbon steels and obtained the results shown in Figure 2.7, and their findings reveal that the retained austenite content increases with increasing carbon content. This effect is obviously due to the lowering of Ms and Mf with increasing carbon content, as shown in Figure 2.3 above (page 8).
Figure 2.7: Change in amount of retained austenite with quenching temperature (in carbon steels) The amount of retained austenite is maximum for a certain austenitizing temperature. It is readily concluded, that the amount of retained austenite is limited for too low austenitizing temperatures because of insufficient dissolving of iron carbide. The decreasing amount of retained austenite above 1000°C is probably associated with austenite gain growth. (Grain boundaries provide the barriers to martensite 9
growth) More austenite is retained with oil quenching that water quenching. It is known[36] to be due to the effect of cooling rate. As shown in Figure 2.8[37], except for cooling rates so slow that extremely small amounts of retained austenite result because of insufficient quenching, the amount of retained austenite decreases with increase in cooling rate.
Figure 2.8: Effect of cooling rate on the amount of retained austenite (carbon steel)
4.
Tempering of martensite Martensite is normally very brittle so it is necessary to modify the mechanical properties by heat treatment
in the range of 150-700°C. This process, which is called tempering, is one of the oldest heat treatments applied to steels, although it is only in recent years that a detailed understanding of the phenomena involved has been obtained. Essentially, martensite is a highly supersaturated solid solution of carbon in iron which, during tempering, precipitates carbon in the form of finely divided carbide phases. The end result of tempering is a fine dispersion of carbides in an Į-iron matrix which often bears little structural similarity to the original as-quenched martensite. Retained austenite does not remain stable during the tempering process and decomposes either to bainite or to other mixtures of ferrite and carbide phases.
4.1.
Quench ageing of carbon steels Most commercial quenched and tempered steels have Ms temperatures that are considerably above room
temperature, also, many products manufactured from these steels have appreciable dimensions cooling through the Ms-Mf region is sufficiently slow that extensive redistribution of carbon atoms may occur after 10
transformation while the steel is still cooling. In low-carbon, low alloy steels, carbon segregation to dislocations occurs during rapid quenching even though no actual carbide precipitation occurs. Also, segregation of carbon from the martensite interlath austenite films may occur during quenching. Carbon-atom redistribution prior to carbide precipitation is likely to occur both through clustering and segregation to lattice defects. Clustering, which occurs on a scale much finer than that of the martensitic dislocation substructure and hence requires diffusion of carbon atoms over shorter distances, is likely the dominant process in initially virgin martensites, especially when the carbon content is high.[38] More recent APFIM (Atomic probe field ion microscopy) studies of high Ms medium-carbon steels by Sarikaya et al[39] have indicated considerable carbon enrichment to the thin interlath retained austenite films during quenching, which is another possible mode of carbon redistribution. Such carbon redistribution has been detected in lath martensites (Ms between 400 and 250 °C) even during very rapid water quenching.
4.2.
Tempering Process of carbon steels On reheating as-quenched martensite, the tempering takes place in four distinct but overlapping
stages:[40] 1.
The formation of a transition carbide and the lowering of the carbon content of the matrix matensite.
2.
The transformation of retained austenite to ferrite and cementite;
3.
replacement of Į-iron carbide by cementite; martensite loses tetragonality
4.
The development of alloy carbides or secondary hardening in alloy steels.
4.2.1.
Room temperature ageing before tempering
Before the first stage of tempering, at room temperatures, the carbon atoms cluster before the precipitation of Į-iron carbide. The process of carbon atoms clustering before the epsilon carbide precipitation is sometimes termed the room temperature ageing of martensite steels. Winchell and Cohen[8] first observed an increase and peak in hardness and electrical resistivity of initially untempered martensite with increasing ageing temperature by studying a wide range of Fe-Ni-C alloys with low Ms (about -35°C) so that autotempering was avoided. Izotov and Utevskiy[25] presented TEM evidence that the diffuse scattering and super lattice spots in the electron diffraction patterns of aged very high-carbon martensite were produced by carbon clustering during ageing. Chow and Kaplow[41] in a Mossbauer study of Fe-1.85wt% C martensite obtained by splat quenching 11
interpreted the results of room-temperature ageing in terms of Fe4C cluster formation. The appearance of carbide is controlled by the solubility limit of the ferrite, which, in return, is influenced directly by the dislocation density.[42] The FIM/AP study of Fe-15Ni-1C steels by K. A. Taylor and L. Chang[43]and the Three dimensional ECOPoSAP research by Wilde and etc.[10] revealed that a periodic tweed structure consisting of carbon modulations develops during ageing of martensite at room temperature. They found the initial decomposition proceeds in a spinodal matter such that the carbon level in the carbon-rich regions increases towards the composition of (Fe,Ni)8C with ageing time, while the matrix is progressively depleted of carbon to a very low level. The decomposition is followed by a coarsening process during which the wavelength increases with ageing time. No diffraction reflections corresponding to an Į-iron carbide were observed in the Fe-C system, which is explained as the incomplete ordering of carbon atoms on the interstitial sublattice.
4.2.2.
Tempering stage 1
Martensite formed in medium and high carbon steels (0.3-1.5% C) is not stable at room temperature because interstitial carbon atoms can diffuse in the tetragonal martensite lattice at this temperature. This instability increases between room temperature and 250°C, when İ-iron carbide precipitates in the martensite.[44] This carbide has a close-packed hexagonal structure, and precipitates as narrow laths or rodlets on cube planes of the matrix with a well-defined orientation relationship: [45, 46, 47, 48]
(101)D ' //(1011)H (011)D ' //(0001)H [111]D ' //[1210]H To complicate this issue further, several studies[49, 50, 51, 52, 53] employing dark-field electron microscopy indicated that what appeared to be rodlike carbides were actually composed of arrays of much smaller particles. The disparity among the above observations suggests that alloy composition exert an important influence on the actual carbide morphology. In Fe-25Ni-C steels, the carbide is observed to exhibit extensive faulting on the basal plane, which funtions as an internal accommodation mechanism to achieve an IPS that minimizes the elastic strain energy.[54] In the higher carbon steels, an increase in hardness has been observed on tempering in the range 50-250°C, which is believed to be attributed to precipitation hardening of the martensite.
4.2.3.
Tempering stage 2 12
During stage 2, austenite retained during quenching is decomposed. Cohen and coworkers detected this stage by X-ray diffraction measurements as well as dilatometric and specific volume measurements. However, the direct observation of retained austenite in the microstructure has always been rather difficult, particularly if it is present in low concentrations. In martensitic plain carbon steels below 0.5wt% carbon, the retained austenite is often below 2%, rising to around 6% at 0.8wt% carbon and over 30% at 1.25wt% C. Thomas[55] and Barnard [56]have shown that retained austenite is present as thin, continuous layers between martensite laths in steels in which the lath martensite morpholopy develops. This retained austenite survives the aging and the tempering stage 1 and then decomposes to large, relatively continuous interlath carbides. The interlath carbides which have formed in the 4340 steel tempered at 350°C are shown in Figure 3.1. These interlath carbide arrays are detrimental to toughness and are associated with the transgranular mode of tempered martensite embrittlement which develops in medium carbon steels tempered between 200 and 400°C.[39, 55, 57, 58]
Figure 3.1 Interlath Cementite formed in martensitic structure of 4340 steel tempered at 350°C. (a) Bright field electron micrograph (b) Dark field electron micrograph taken with (210) cementite diffracted beam 4.2.4.
Tempering stage 3
During the Third stage of tempering, cementite first appears in the microstructure as a Widmanstatten distribution of plates which have a well-defined orientation relationship with the matrix which has now lost its
13
tetragonality and become ferrite. The relationship is that due to Bagaryatskii[59]:
(211)D ' //(001) Fe3C , [011]D ' //[100]Fe3C , [111]D ' //[010]Fe3C This reaction commences as low as 100°C, and is fully developed at 300°C, with particles up to 200 nm long and ~15 nm in diameter. Similar structures are often observed in lower carbon steels as quenched, as a result of the formation of Fe3C during the quench. During tempering, the replacement of transition carbides and low-temperature martensite by cementite and ferrite. During tempering, the most likely sites for the nucleantion of the cementite are the ¦-iron carbide interface with the matrix, and as the Fe3C particles grow, the¦-iron carbide particles gradually disappear. The twins occurring in the higher carbon martensites are also sites for the nucleation and growth of cementite which tends to grow along the twin boundaries forming colonies of similarly oriented lath-shaped particles of {112}D habit. The orientation relationship with the ferritic matrix is the same in both cases. A third site for the nucleation of cementite is the grain boundary regions, both the interlath boundaries of the martensite and the original austenite grain boundaries. The cementite can form as very thin films which are difficult to detect but which gradually spheroidize to give rise to well-defined particles of Fe3C in the grain boundary regions. These grain boundary cementite films can adversely affect ductility. However, they can be modified by addition of alloying elements. During the third stage of tempering the tetragonality of the matrix disappears and it is then essentially ferrite, not supersaturated with respect to carbon. Subsequent changes in the morphology of the cementite particles occur by an Ostwald ripening type of process, where the smaller particles dissolve in the matrix providing carbon for the selective growth of the larger particles. It is useful to define a “three-B” stage of tempering in which the cementite particles undergo a coarsening process and essentially lose their crystallographic morphology, becoming spheroidized. The coarsening commences between 300 and 400°C, while spheroidization takes place increasingly up to 700°C. At the higher end of this range of temperature, the martensite dislocations, lath and twin boundaries are replaced by more equiaxed ferrite grain boundaries by a process which is best described as recovery and recrystallization. The final result is an equiaxed array of ferrite grains with coarse spheroidized particles of Fe3C partly, but not exclusively, in the grain boundaries. The spheroidization of the Fe3C rods is driven by the resulting decrease in surface energy. The particles, which preferentially grow and spheroidize are located mainly at interlath boundaries and prior austenite 14
boundaries,[60, 61] although some particles remain in the matrix. The boundary sites are preferred because of the greater ease of diffusion in these regions, also greater overall lowering of interfacial area. The original martensite lath boundaries remain stable up to about 600°C, but in the range 350-600°C, there is considerable rearrangement of the dislocations within the laths and at those lath boundaries which are essentially low angle boundaries. 4.2.5.
Tempering stage 4 - secondary hardening
Druing the 4th stage of tempering, a number of the familiar alloying elements in steels form carbides, which are thermodynamically more stable than cementite. It is interesting to note that this is also true of a number of nitrides and borides. Nitrogen and boron are increasingly used in steels in small but significant concentrations. The alloying elements Cr, Mo, V, W and Ti all form carbides with substantially higher enthalpies of formation, while the elements nickel, cobalt and copper do not form carbide phases. Manganese is weak carbide former, usually found in solid solution in cementite and not in a separate carbide phase. It would, therefore, be expected that when strong carbide forming elements are present in steel in sufficient concentration, their carbides would be formed in preference to cementite.. However, the metallic elements diffuse substitutionally, in contrast to carbon and nitrogen which move through the iron lattice interstitially, with the result that the diffusivities of carbon and nitrogen are several orders of magnitude greater in iron, than those of the metallic alloying elements. Consequently, higher temperatures are needed for the necessary diffusion of the alloying elements prior to the nucleation and growth of the alloy carbides and, in practice, for most of the carbide forming elements this is in the range 500-600°C. The coarsening of carbides in steels is an important phenomenon which influences markedly the mechanical properties. We can apply the theory for coarsening of a dispersion due to Lifshitz and Wagner, which gives for spherical particles in a matrix:
rt 3 r03
k Vm 2 DG t RT
where
rt = the mean particle radius at time t r0 = the mean particle radius at time 0
D = diffusion coefficient of solute in matrix
G
= interfacial energy of particle/matrix interface per unit area
15
Vm = molar volume of precipitate k = constant. The coarsening rate is dependent on the diffusion coefficient of the solute and, under the same conditions, at a given temperature, cementite would coarsen at a greater rate than any of the alloy carbides formed. The formation of alloy carbides between 500 and 600°C is accompanied by a marked increase in strength, often in excess of that of the as-quenched martensite, as shown in Figure 3.2. This phenomenon, which is known as secondary hardening, is best shown in steels containing molybdenum, vanadium, tungsten, titanium, and also in chromium steels at higher alloy concentrations.
Figure 3.2 The effect of molybdenum on the tempering of quenched 0.1wt% steel[62] This secondary hardening process is a type of age-hardening reaction, in which relatively coarse cementite dispersion is replaced by new and much finer alloy carbide dispersion. On attaining a critical dispersion parameter, the strength of the steel reaches a maximum, and as the carbide dispersion slowly coarsens, the strength drops.
16
4.3.
Role of carbon content Carbon has a profound effect on the behavior of steels during tempering. Firstly, the hardness of the
as-quenched martensite is largely influenced by the carbon content, as is the morphology of the martensite laths which have a {111} habit plane at low carbon contents, changing to {225} at higher carbon contents.
Figure 3.3 Hardness of iron-carbon martensites tempered 1h at 100-700°C.[63] The Ms temperature is reduced as the carbon content increases, and thus the probability of the occurrence of auto-tempering is less. On subsequent tempering of low carbon steels (<0.2wt% C) up to 200°C further segregation of carbon takes place, but no precipitation has been observed. Under normal circumstances it is difficult to detect any tetragonality in the martensite in steels with less than 0.2 % C, a fact which can also be explained by the rapid segregation of carbon during quenching. The hardness changes during tempering are also very dependent on carbon content, as shown in Figure 3.3.
4.4.
Mechanical properties of tempered plain carbon steels The intrinsic mechanical properties of tempered plain carbon martensitic steels are difficult to measure for
several reasons. Firstly, the absence of other alloying elements means that the hardenability of the steels is low, so a fully martensitic structure is only possible in thin sections. However, this may not be a disadvantage where shallow hardened surface layers are all that is required. Secondly, at lower carbon levels, the Ms temperature is rather high, so tempering is likely to take place during cooling. Thirdly, at the higher carbon levels the presence of retained austenite will influence the results. Added to these factors, plain carbon steels can exhibit quench 17
cracking which makes it difficult to obtain reliable test results. This is particularly the case at higher carbon levels, i.e. above 0.5% carbon. Provided care is taken, very good mechanical properties, in particular proof and tensile stresses, can be obtained on tempering in the range 100-300°C. However, the elongation is frequently low and the impact values poor. Plain carbon steels with less than 0.25% C are not normally quenched and tempered, but in the range 0.25-0.55 % C heat treatment is often used to upgrade mechanical properties. The usual tempering temperature is between 300 and 700°C allowing the development of tensile strengths between 1700 and 800 MPa, the toughness increasing as the tensile strength decreases. This group of steels is very versatile as they can be used for crankshafts and general machine parts as well as hand tools, such as screwdrivers and pliers. The high carbon steels (0.5-1.0%) are much more difficult to fabricate and are, therefore, particularly used in applications where high hardness and wear resistance are required, e.g. axes, knives, hammers, cutting tools. Another important application is for springs, where often the required mechanical properties are obtained simply by heavy cold work, i.e. hard drawn spring wire. However, carbon steels in the range 0.5-0.75% C are quenched, and then tempered to the required yield stress.
4.5.
The effect of alloying elements on the formation of iron carbides Carbides will always remain, even above 400°C, but the Fe3C phases tends to coarsen rapidly, so strength
drops. Silicon additions are found to inhibit Fe3C coarsening. Silicon is relatively insoluble in Fe3C, so silicon is rejected by carbon diffusion in the rate controlling step.[64] While the tetragonality of martensite disappears by 300°C in plain carbon steels, in steels containing some alloying elements, e.g. Cr, Mo, W, V, Ti, Si, the tetragonal lattice is still observed after tempering at 450°C and even as high as 500°C. It is clear that these alloying elements increase the stability of the supersaturated iron-carbon solid solution. In contrast manganese and nickel decrease the stability. Alloying elements also greatly influence the proportion of austenite retained on quenching. Typically, a steel with 4% molybdenum, 0.2% C, in the martensitic state contains less than 2% austenite, and about 5% is detected in a steel with 1% vanadium and 0.2% C. On tempering each of the above steels at 300°C, the austenite decomposes to give thin grain boundary films of cementite which, in the case of the higher concentrations of retained austenite, can be fairly continuous along the lath boundaries. It is likely that this interlath cementite is responsible for tempered martensite embrittlement, frequently encountered as a toughness minimum in the range 18
300-350°C, by leading to easy nucleation of cracks, which then propagate across the tempered martensite laths. Alloying elements can also restrain the coarsening of cementite in the range 400-700°C, a basic process during the third stage of tempering. Several alloying elements, notably silicon, chromium, molybdenum and tungsten, cause the cementite to retain its fine Widmanstatten structure to higher temperatures. With the exception of silicon, the mechanisms of these effects are unknown, although it is thought that the elements act as either by entering into the cementite structure or by segregating at the carbide-ferrite interfaces. Whatever the basic cause may be, the effect is to delay significantly the softening process during tempering. This influence on the cementite dispersion has other effects, in so far as the carbide particles, by remaining finer, slow down the reorganization of the dislocations inherited from the martensite, with the result that the dislocation substructures refine more slowly. The cementite particles are also found on ferrite grain boundaries, where they control the rate at which the ferrite grains grow.
4.6.
Nucleation and growth of alloy carbides The dispersions of alloy carbides which occur during tempering can be very complex, but some general
principles can be discerned which apply to a wide variety of steels. The alloy carbides can form in at least three ways: In-site nucleation at pre-existing cementite particles. It has been shown that the nuclei form on the interfaces between cementite particles and the ferrite. As they grow, carbon is provided by the adjacent cementite, which gradually disappears. By separate nucleation within the ferrite matrix, usually on dislocations inherited from the martensitic structure. At grain boundaries and sub boundaries-these include the former austenite boundaries, the original martensitic lath boundaries (now ferrite), and the new ferrite boundaries formed by coalescence of sub boundaries, or by recrystallization. In-site nucleation at pre-existing cementite particles are a common occurrence but because these particles are fairly widely spaced at temperatures above 500°C, the contribution of this type of alloy carbide nucleation to strength is very limited. The nucleation of carbides at the various types of boundary is to be expected because these are energetically favourable sites, which also provide paths for relatively rapid diffusion of solute. Consequently the ageing process is usually more advanced in these regions and the precipitate is more massive. In many alloy steels, the 19
first alloy carbide to form is not the final equilibrium carbide and, in some steels, as many as three alloy carbides can form successively. In these circumstances, the equilibrium alloy carbide frequently nucleates first in the grain boundaries, grows rapidly and eventually completely replaces the Widmanstatten non-equilibrium carbide within the grains.
4.7.
Effects of Alloying Elements The presence of more than one carbide-forming element can complicate the precipitation processes during
tempering. In general terms, the carbide phase which is the most stable thermodynamically will predominate, but this assumes that equilibrium is reached during tempering. This is clearly not so at temperatures below 500-600°C. The use of pseudo-binary diagrams for groups of steels, e.g. Cr-V, Cr-Mo, can be a useful guide to carbide phases likely to form during tempering. Certain strong carbide formers, notably niobium, titanium and vanadium, have effects on tempering out of proportion to their concentration. In concentrations of 0.1 wt % or less, provided the tempering temperature is high enough, i.e. 550-650°C, they combine preferentially with part of the carbon and, in addition to the major carbide phase, e.g. Cr7C3, Mo2C, they form a separate, very much finer dispersion, more resistant to over-ageing. This secondary dispersion can greatly augment the secondary hardening reaction, illustrating the importance of these strong carbide forming elements in achieving high strength levels, not only at room temperature but also at elevated temperatures, where creep resistance is often an essential requirement. The effect of adding alloy elements on the hardness of martensite in medium carbon steels tempered for one hour at 56°C intervals in the range 200 to 600°C has been studied by comparing the hardness of high purity 0.4wt% carbon steel and 4340 alloy steel by Richard Hardwicke. As shown in figure 3.4, in the range 200 to 600°C, the hardness curves diverge with the carbon steel showing significantly less hardness than 4340 alloy.[65] Note the curves begin to diverge of temperature as low as 200°C. The processes occurring in the low temperature range are particularly poorly understood.
20
Hardness Data - 1hr Temper 700
High Purity 0.4%C Steel 4340 Alloy Steel
600
Hardness (Hv)
500 400 300 200 100 0 0
100
200
300
400
500
600
700
Tempering Temperature (C)
Figure 3.4 Change in hardness with tempering temperature for SAE 4340 alloy steel and high purity 0.4wt% C steel 4.7.1 Elements effect on Ms temperatures Both the Ms (martensite start) and Mf (martensite finish) temperatures in steel are functions of the carbon content, as shown in Figure 3.5. [66] Notice in Figure 3.5 that the martensite finish temperature, Mf, occurs at room temperature (20°C) near 0.6wt% C. The amount of retained austenite under these conditions is over 3%. For the majority of steels containing more than 0,50% C, Mf lies below room temperature. This implies that after hardening these steels practically always contain some residual austenite.
21
Figure 3.5 The effect of carbon on Ms and Mf
Subsititional alloying elements in steels also affect the martensitic transformation. All alloying elements with the possible exception of Co, lower Ms, as well as Mf.
Ms may be calculated from the equation given below, by inserting the percentage concentration of each alloying element in the appropriate term. The equation is valid only if all the alloying elements are completely dissolved in the austenite.[67]
Ms (°C) = 561 - 474C - 33Mn - 17Ni - 17Cr - 21Mo
For high-alloy and medium-alloy steels Stuhlmann has suggested the following equation:
Ms (°C) = 550 - 350C - 40Mn - 20Cr - 10Mo - 17Ni - 8W - 35V - 10Cu + 15Co + 30Al
Where all symbols refer to weight percentage of the elements concerned.
It can be noted that carbon has the strongest influence on the Ms temperature. Figures 3.6 and 3.7 show diagrams with an example of experimental results of the effect of alloys on the Ms temperature of various types of Fe-bases binary steels.
22
Figure 3.6
Figure 3.7
Effect of alloys on the Ms - temperature[68]
Effect of Ni on the Ms – temperature[69]
4.7.2. Effect of Different Amount of Manganese Researches [70] have shown that the effect of manganese on the hardness of tempered martensite is increasing from zero at 200°C in a regular manner with tempering temperature to 420°C. In the region from 400°C to 700°C, the hardness increase varies about an average value. The variation is ±10HV Vickers harness number in the region of 0.4pct manganese and the amount of variation is a minimum at 1.6pct manganese. The
23
changes of tempered martensite microstructure as manganese increased suggest that manganese increases the hardness of tempered martensite principally by retarding the coalescence of carbides, and thus provides a resistance to grain growth in the ferrite matrix. The combination of more and smaller carbides and the apparent lower state of recovery of the martensite (finer packets of ferrite) causes the observed substantial increase in the hardness of tempered martensite as the percentage of manganese in steel increases.
4.7.3. Effect of Phosphorus Phosphorus increased the hardness of tempered martensite at all tempering temperatures except 200°C. Phosphorus is considered to have the same effect at all tempering temperatures in the range 260°C to 650°C. phosphorus was assumed to increase the hardness of tempered martensite
4.7.4. Effect of Silicon Silicon increased the hardness of tempered martensite at all tempering temperatures. Silicon was found to have a much greater effect at 300°C than at other tempering temperatures. This is in agreement with the effect of silicon on coarsening of Fe3C, a change that occurs at about 316°C. Hobbs found[61] that by adding the silicon to a medium-carbon steel, the tempering process in the temperature range 400 – 700°C could be retarded markedly, and with the exception of the 2wt% Si steel tempered at 400°C, the structural changes leading to the formation of the ferrite and cementite were complete within 1 hour and that the ferrite was no longer supersaturated with carbon. In the 0.86 Si steel tempered at 650°C, The carbides were smaller and the ferrite tended to be divided into smaller lath-like regions comparing to 0.09 Si steel [70].
4.7.5. Effect of Nickel Nickel has a relatively small effect on the hardness of tempered martensite which is essentially the same at all the tempering temperatures. Accordingly, nickel has no apparent effect to microstructures. The effect of nickel to the hardness is probably due to weak solid-solution hardening.
4.7.6. Effect of Chromium In Chromium steels, two chromium carbides are very often encountered, Cr7C6 (trigonal) and Cr23C6 (complex cubic). The normal carbide sequence during tempering is Matrix ĺ (FeCr)3C ĺ Cr7C3 ĺ Cr23C6. 24
While this sequence occurs in higher Chronium steels, below about 7 wt% Cr, Cr23C6 is absent unless other metals such as Mo are present. Chromium is a weaker carbide former than Vanadium, which is illustrated by the fact that Cr7C3 does not normally occur until the Chromium content of the steel exceeds 1 wt% at a carbon level of about 0.2 wt%.
Figure 3.8
The effect of Chromium on the tempering of a 0.35wt% C steel.[62]
4.7.7. Effect of Molybdenum and tungsten Molybdenum is a strong carbide forming element that can be expected to produce substantially higher hardness than an Fe-C alloy when the alloys are tempered at higher temperatures. Molybdenum is a potent addition to steels quenched and tempered at 1000°F (538°C) or above. It partitions to the carbide phase at elevated temperatures, and thus keeps the carbide particles small and numerous. When molybdenum or tungsten is the predominant alloying element in a steel, a number of different carbide phases are possible, but for composition between 4 and 6 wt% of the element the carbide sequence is likely to be: Fe3C- Mo2C- M6C.[27] The carbides responsible for the secondary hardening in both the case of tungsten and molybdenum are the isomorphous hexagonal carbides Mo2C and W2C, both of which, in contrast to vanadium carbide, have well-defined rod let morphology. According to Figure 3.2, increasing added Mo significantly improve the hardness of steels with the same carbon content. 25
4.7.8. Effect of Vanadium Vanadium is a stronger carbide former than chromium or molybdenum. It can be expected to have a potent effect on the hardness of tempered martensite. Vanadium carbide forms in steel containing relatively small amounts of vanadium. in steel with as little as 0.1% V, the face-centered cubic vanadium carbide VC is formed. It is often not of stoichiometric composition, being frequently nearer V4C3, but with other elements in solid solution within the carbide. Normally, this is the only vanadium carbide formed in steels, so the structural changes during tempering of vanadium steels are relatively simple. The large effect of vanadium is probably due to the formation of an alloy carbide (V4C3 or VC), which replaces cementite type carbide at high tempering temperatures and persists as a fine dispersion up to the 1300°F (704°C).
4.8. Tempering Parameter The effect of tempering temperature and time upon the properties of quenched steel is a subject of great practical importantce. It would be very desirable to be able to predict the properties of quenched steel when given any selected tempering cycle, with a minimum of experimental work upon that heat treatment or even type of steel. Tempering charts in metallurgical literature usually give hardness as a function of tempering temperature for only one tempering time. However, in practice, tempering times vary frequently, and a method of converting tempering curves for one time to curves for another time would be very valuable. It is also advantageous, in determining tempering parameters for a specific application, not to have to temper several temperatures for the time used in the commercial operation. A more satisfactory method would be to temper for short times and only one temperature still be able to predict properties for longer tempering conditions.
Work by Hollomon and Jaffe[71] showed that temperature and time are independent variables in tempering of steels, and that one can obtain the same result such as tempered hardness either by decreasing temperature and increasing time or by raising temperature and decreasing time. Both variables could be combined in a parameter named as Hollomon-Jaffe tempering parameter P:
P
T (C log t )
26
Where T is the absolute temperature and t the tempering time in hours, while C is materials constant in the range of 10 to 20. The relationship between hardness and parameter P could be defined by the equation:
Hardness
f ( P)
In other words, as long as the parameter has a constant value for a specific materals, the same hardness can be produced with a short tempering time and a high temperature as with a long tempering time and low temperature. The relationship can be used reliably to predict hardness of tempered plain carbon and alloy steels containing 0.2-0.85%wt C and less than 5%wt total alloying elements.
Figure 3.9 Effect of tempering temperature to the hardness of carbon steels with different carbon content[62]
27
Figure 3.10 Effects of tempering temperature and tempering time to the hardness of carbon steels with different carbon content[72] 5. Remaining questions: Although the tempered martensite has been studied for decades, the atomic scale redistribution of alloy elements during the tempering process is still not fully understood. It is necessary to know the mechanisms of the 3D alloy redistribution across the ferrite/cementite interface in order to understand the retardation of softening in medium carbon steels. This forms the first part of my thesis project. Another interesting problem is the carbon redistribution at the very first stages of tempering during low temperature tempering process. The process of carbon atom clustering before the epsilon carbide precipitation is going to be studied as the second part of my project, in order to learn the carbon effects of carbon redistribution on the mechanical properties.
28
Reference 1.
A.Cottrell, An Introduction to Metallugy, 2nd Edition. 1995. 365.
2.
Metals Handbook. 10th Edition ed. Vol. Vol. 1. 1990: ASM International. 135.
3.
A.Cottrell, An Introduction to Metallugy, 2nd Edition. 1995. 372.
4.
Reza Abbaschian Robert E. Reed-Hill, Physical Metallurgy Principles, 3rd Edition. 1994, Boston: PWS Publishing Company. p634.
5.
Atlas of Isothermal Transformation and continuous Cooling Diagrams. 1977, Ohio: American Society of Metals.
6.
A.Cottrell, An Introduction to Metallugy, 2nd Edition. 1995. 375.
7.
C. S. Roberts, B. L. Averbach, and M. Cohen, The Mechanism and Kinetics of the First Stage of Tempering. Transactions of the American Society for Metals, 1953. 45: p. 576-585.
8.
P. G. Winchell, and M. Cohen, The Strength of Martensite. Transactions of the American Society for Metals, 1962. 55: p. 347-361.
9.
R. E. Reed-Hill, and R. Abbaschian, Physical Metallurgy Principles, 3rd Edition. 1994, Boston: PWS Publishing Company. 634.
10.
J. Wilde, A. Cerezo, and G. D. W. Smith, Three-dimentional Atomic-scale Mapping of a Cottrell Atmosphere around a Dislocation in Iron. Scripta Materialia, 2000. 43: p. 39-48.
11.
L. Chang, S. J. Barnard, and G. D. W. Smith. The Segregation of Carbon atoms to Dislocations in Low-Carbon Martensite: Studies by Field Ion Microscopy and Atom Probe Microsanalysis. in Speich Symposium. 1992. Montreal, Canada: Iron adn Steel Society.
12.
C.S.Roberts, Effect of Carbon on the Volume Fractions and Lattice Parameters of Retained Austenite and Martensite. Transactions of the American Institute of Mining and Metallurgical Engineers, 1953. 197: p. 203-204.
13.
H.K.D.H.Bhadeshia, Encyclopedia of Materials: Science and Technology. 2001: Elsiver Science. 5203-5206.
14.
M. S. Wechsler, D. S. Lieberman, and T. A. Read, On the Theory of the Formation of Martensite. Transactions of the American Institute of Mining and Metallurgical Engineers, 1953. 197: p. 1503-1515.
15.
E.C. Brain, The nature of martensite. Transactions of the American Institute of Mining and Metallurgical Engineers, 1924. 70: p. 25.
16.
Sachs G. Kurdjumov G, Z. Phys., 1930. 64: p. page 325.
17.
K. A. Johnson, and C. M. Wayman, The Crystallography of the Austenite-Martensite Transformation in an Fe-Cr-C Alloy. Acta Crystallographica, 1963. 16: p. 480-486.
18.
J. S. Bowles, and J. K. Mackenzie, Crystallography of Martensite Transformations I. Acta Metallurgica, 1954. 2: p. 129.
19.
J. M. Rigsbee, and H. I. Aaronson, The Interfacial Structure of the Broad Faces of Ferrite Plates Acta Metallurgica, 1979. Vol. 27: p. 365-376.
20.
J. W. Christian, Chapter in Martensite: Fundamentals and Technology, ed. E. R. Petty. 1970: Longmans.
21.
P. M. Kelly, and J. Nutting, Journal of the Iron and Steel Institute, 1961. 197: p. 199-211.
22.
A. R. Marder, and G. Krauss, The Morphology of Martensite in Iron-Carbon Alloys. Transactions of the American Society for Metals, 1967. 60: p. 651-660.
23.
A. R. Marder, and G. Krauss, The Formation of Low Carbon Martensite in Fe-C Alloys. Transactions of the American Society for Metals, 1969. 62: p. 957-964.
24.
G. Krauss, Deformation and Fracture in Martensitic Carbon Steels Tempered at Low Temperatures.
29
Metallurgical and Materials Transaction A, 2001. 32A: p. 861-877. 25.
V. I. Izotov, and L. M. Utevskiy, Physics of Metals and Metallography - USSR, 1968. 25: p. 86.
26.
R. Oshima, and C. M. Wayman, Transactions of the Japanese Institute for Metals, 1975. 16: p. 73-86.
27.
R. W. K. Honeycombe, and H. K. D. H. Bhadeshia, Steels: Microstructure and Properties. 3rd ed. 1995, London: Edward Arnold.
28.
W. C. Leslie G.R. Speich, Tempering of Steel. Metallurgical Transactions 1972. 3: p. 1043.
29.
G. R. Speich, and W. C. Leslie, Tempering of Steel. Metallurgical Transactions 1972. 3: p. 1043.
30.
G. Krauss, and A. R. Marder, the Morphology of Martensite in Iron Alloys. Metallurgical Transactions, 1971. 2: p. 2343-2357.
31.
Tadashi Marki, Kaneaki Tsuzuki, and Imao Tamura, The Morphology of Microstructure Composed of Lath Martensite. Transactiosn ISIJ, 1980. 20: p. 207-214.
32.
A. R. Entwisle, The Kinetics of Martensite Formation in Steel. Metallurgical Transactions, 1971. 2: p. 2395-2407.
33.
B. P. J. Sandvik, and C. M. Wayman, Charqacterisitics of Lath Martensite: Part I. Crystallographic and Substructural Features. Metallurgical Transactions A, 1983. 14A: p. 809-822.
34.
Zenji Nishiyama, Martensitic Transformation, ed. Morris E. Fine, M. Meshii, andC. M. Wayman. 1978, London: Academic Press.
35.
K. Tamaru, and S. Sekito, Kinzoku no Kenkyu, 1931. 8: p. 595.
36.
J. A. Mathews, American Society of Steel Treating, 1925. 8: p. 565.
37.
H. Esser, and H. Cornelius, Arch. Eisenhuttenwes, 1934. 7: p. 693.
38.
M. K. Miller, P. A. Beaven, and G. D. W. Smith, A Study of the Early Stages of Tempering of Iron-Carbon Martensites by Atom Probe Field ion Microscopy. Metallurgical Transactions A, 1981. 12A: p. 1197-1204.
39.
M. Sarikaya, A. K. Jhingan, and G. Thomas, Retained Austenite and Tempered Martensite Embrittlement in Medium Carbon Steels. Metallurgical Transactions A, 1983. 14A: p. 1121-1133.
40.
Martensite, ed. G. B. Olson, andW. S. Owen. 1992: ASM International.
41.
W. K. Choo, and R. Kaplow, Mossbauer Measurements on the Aging of Iron-Carbon Martensite. Acta Metallurgica, 1973. 21: p. 725-732.
42.
David Kalish, and E. M. Roberts, On the Distribution of Carbon in Martensite. Metallurgical Transactions, 1971. 2: p. 2783-2789.
43.
K. A. Taylor, L. Chang, G. B. Olson, G. D. W. Smith, M. Cohen, and J. B. Vander Sande, Spinodal Decomposition during Aging of Fe-Ni-C Martensites. Metallurgical Transactions A, 1989. 20A: p. 2717-2737.
44.
M. K. Miller, P. A. Beaven, S. S. Brenner, and G. D. W. Smith, An Atom Probe Study of the Aging of Iron-Nickel-Carbon Martensite. Metallurgical Transactions A, 1983. 14A: p. 1021-1024.
45.
M. G. H. Wells, An Electron Transmission Study of the Tempering of Martensite in an Fe-Ni-C Alloy. Acta Metallurgica, 1964. 12: p. 389.
46.
C. J. Barton, The Tempering of a Low-Carbon Internally Twinned Martensite. Acta Metallurgica, 1969.
47.
S. Murphy, and J. A. Whiteman, The Precipitation of Epsilon-Carbide in Twinned Martensite.
17: p. 1085. Metallurgical Transactions, 1970. 1: p. 843. 48.
R. Padmanabhan, and W.E. Wood, Precipitation of İ-Carbide in Martensite. Materials Science and Engineering, 1984. 65: p. 289.
49.
Y. Tanaka, and K. Shimizu, Carbide Formation upon tempering at Low Temperatures in Fe-Mn-C
30
Alloys. Transactions of the Japan Institute of Metals, 1981. 22: p. 779. 50.
Y. Hirotsu, and S. Nagakura, Crystal Structure and Morphology of the Carbide Precipitated from Martensitic High Carbon Steel During the First Stage of Tempering. Acta Metallurgica, 1974. 20: p. 645.
51.
Y. Hirotsu, Y. Itakura, K. Su, and S. Nagakura, Electron Microscopy and Difffraction Study of the Carbide Precipitated from Martensitic Low and High Nickel Steels at the First Stage of Tempering. Transactions of the Japan Institute of Metals, 1976. 17: p. 503.
52.
K. Nakazawa D. L. Williamson , and G. Krauss, A Study of the Early Stages of Tempering in an Fe-1.2
53.
D. L. Williamson, K. Nakazawa, and G. Krauss, A Study of the Early Stages of Tempering in an Fe-1.2
Pct C Alloy. Metallurgical Transactions, 1979. 10A: p. 1351. Pct C Alloy. Metallurgical Transactions, 1979. 10A: p. 1351. 54.
K. A. Taylor, G. B. Olson, M. Cohen, and J. B. Vander Sande, Carbide Precipitation during Stage 1 Tempering of Fe-Ni-C Martensite. Metallurgical Transactions A, 1989. 20A: p. 2749-2765.
55.
G. Thomas, Retained Austenite and Tempered Martensite Embrittlement. Metallurgical Transactions A, 1978. 9A: p. 439-450.
56.
S. J. Barnard, G. D. W. Smith, M. Sarikaya, and G. Thomas, Carbon Atom distribution in a Dual Phase Steel: an Atom Probe Study. Scripta Metallurgica, 1981. 15: p. 387-392.
57.
J. P. Materkowski, and G. Krauss, Tempered Martensite Embrittlement in SAE 4340 Steel. Metallurgical Transactions A, 1970. 10A: p. 1643-1651.
58.
F. Zia-Ebrahima, and G. Krauss, The Evaluation of Tempered Martensite Embrittlement in 4130 Steel by Instrumented Charpy V-notch Testing. Metallurgical Transactions A, 1983. 14A: p. 1109-1119.
59.
D. H. Jack, The Orientation Relationship of Interstitial Phases in Iron. Materials Science and Engineering, 1974. 13: p. 19.
60.
R. W. Caron, and G. Krauss, The Tempering of Fe-C Lath Martensite. Metallurgical Transactions. 3: p. 2381-2389.
61.
R. M. Hobbs, G. W. Lorimer, and N. Ridley, Effect of Silicon on the Microstructure of Quenched and Tempered Medium-Carbon Steels. Journal of the Iron and Steel Institute, 1972. 210: p. 757-764.
62. 63.
E. C. Bain, and H. W. Paxton, The Alloying Elements in Steel. 2nd ed. 1961. G. R. Speich, Tempering of Low-Carbon Martensites. Transactions of the Society of Mining Engineers of AIME, 1969. 245: p. 2553.
64.
S. J. Barnard, G. D. W. Smith, A. J. Garratt-Reed, and J. Vander Sande, Advances in the Physical Metallurgy and Applications of Steels. Influence of Silicon on the Tempering of Steel. 1982, London: The Metals Society.
65.
Richard Hardwicke, Atom Probe Study of the Role of Alloy Elements on the Tempering of Steel, Part II Thesis in Department of Materials. 2001, University of Oxford: Oxford.
66.
A. R. Troiano, and A. B. Greninger, Metal Progress, 1946. 50: p. 303.
67.
W. Stevens, and A.G. Haynes, The temperature of forming martensite and bainite in low-alloy steels. Journal of the Iron and Steel Institute, 1956. 183: p. 349-359.
68.
M. Izumiyama, M. Tsuchiya, and Y. Imai, Journal of the Japan Institute of Metals, 1970. 34: p. 291.
69.
L. Kaufman, and M. Cohen, The Martensitic Transformation in the Iron-Nickel System. Transactions of the American Institute of Mining and Metallurgical Engineers, 1956. 206: p. 1393.
70.
C.R.Hribal R.A.Grance, and L.F.Porter, Hardness of Tempered Martensite in Carbon And Low-Alloy Steels. Metallurgical Transactions A, 1977. 8A: p. 1775-1785.
71.
J. H. Hollomon, and L. D. Jaffe, Time-Temperature Relationships in Tempering Steel. Transactions of
31
the American Institute of Mining and Metallurgical Engineers, 1945. 162: p. 2023. 72.
E. C. Bain, Functions of the Alloying Elements in Steel. 1939, American Society for Metals. p. 233.
32