Cemented carbide phase diagrams: A review

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

Contents lists available at ScienceDirect

Int. Journal of Refractory Metals and Hard Materials j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / I J R M H M

Review

Cemented carbide phase diagrams: A review C.M. Fernandes, A.M.R. Senos ⁎ CICECO, University of Aveiro, 3810-193 Aveiro, Portugal

a r t i c l e

i n f o

a b s t r a c t One of the main topics of the actual research in the field of cemented carbides concerns the development of new composites, with partial or total substitution of the traditional cobalt binder by other more economic and less toxic materials. Composites with partial substitution of cobalt by nickel and iron are currently entering in industrial production. However, the total cobalt replacement is envisaged and Ni–Fe or Ni–Fe–Cr alloys are being currently investigated for such a purpose. The actual knowledge on phase diagrams for WC and different binders will be extremely useful and opportune regarding the need to choose initial compositions leading to a desired final phase composition and to select adequate sintering cycle conditions. In the present review, the existent phase diagrams of W–C–M with M = (Co, Fe, Ni, Fe–Ni, Fe–Al, Co–Fe–Ni, Cr and Cr–Fe) are presented and discussed. © 2011 Elsevier Ltd. All rights reserved.

Article history: Received 8 September 2010 Accepted 5 February 2011 Keywords: Tungsten carbide Phase diagrams Cemented carbide Metallic binders

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . Phase diagrams . . . . . . . . . . . 2.1. System W–C . . . . . . . . . 2.2. WC cemented carbides diagrams 2.2.1. System W–C–Co . . . 2.2.2. System W–C–Fe . . . 2.2.3. System W–C–Ni . . . 2.2.4. System W–C–Fe–Ni . . 2.2.5. System W–C–Co–Fe–Ni 2.2.6. System W–C–Fe–Al . . 2.2.7. System W–C–Cr . . . 2.2.8. System W–C–Cr–Fe . . Concluding remarks . . . . . . . . . Acknowledgments . . . . . . . . . . References . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

1. Introduction Tungsten is usually found in the form of tungstates in nature. The most important minerals are wolframite ((Fe,Mn)WO4) and scheelite (CaWO4) [1]. China has the world's largest deposits of tungsten, followed by Canada, Russia, USA, Australia, Korea, Turkey, Bolivia and others. The largest European reserves are in Portugal, France, Austria, Sweden and southern England; smaller deposits are found in the Erz Mountains, near Reichstein and also Baden-Württemberg, Germany. In World War II, tungsten played a role of enormous importance in diplomatic relations. Portugal, as the main source of this element in

⁎ Corresponding author. Tel.: +351 234 370257; fax: +351 234 370204. E-mail address: [email protected] (A.M.R. Senos). 0263-4368/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijrmhm.2011.02.004

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

405 406 406 407 407 408 410 410 414 414 416 416 417 417 417

Europe, was placed under great pressure of both sides in the dispute, since this element was essential for the wolfram production as a material in the armament industry. The first sintered tungsten carbide was produced in 1914 (see Fig. 1) for use in drawing dies and rock drills [2]. The year 1923 represents an important milestone in the tungsten chronology, Fig. 1. It marks the invention of the first hard metal tool by Karl Schröter of OSRAM Studiengesellschaft and the corresponding application for a patent (Patent-Treuhand-Gesellschaft für elektrische Glühlampen mbH in Berlin) sold to Krupp in Germany in 1926. The two inventions claimed by the patent were [2,3]: ▪ A hitherto unique hard metal alloy composition, namely the combination of the very hard tungsten carbide, WC, with small amounts (10–20 wt.%) of an iron group metal (Fe, Ni and Co);

Author's personal copy 406

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

1890 W2C

1900

WC

1910 Cast WC-W2C

1920 WC-Co

1930

WC-Mo2C-TiC-Co/Ni WC-TaC-Co, WC-TiC-Co WC-TiC-TaC-Co

1940

WC-Cr3C2-Co Development of standard grades

1950 1960 1970 1980 1990

WC-Ni/Cr WC-TiC-(Ta,Nb)C-Cr3C2-Co WC-TiC(TaC)-HfC-Co Micrograins WC-Fe(Co,Ni) Cast (W,Ti)C (W,Mo)C-Co WC-Fe/Co/Ni

Coatings TiC TiN Ti(C,N) Al2O3 Hf(C,N)

Fig. 1. Historical development of the cemented carbide cutting tools during the last century [1].

▪ The manufacture of the hard metal alloys by the application of the process of Powder Metallurgy (PM), namely the pressing and sintering of mixed powders of tungsten carbide and binder metal. The name «WIDIA-N» (WC–6% Co) was chosen by Krupp for this cemented carbide (hard metal) a trademark that was created from the German words “Wie” and “Diamond”, meaning “like diamond”, which survives until today [1,3]. Other proprietary names were assigned to the General Electric Company in United States with the name Carboloy and to the British Thomson-Houston Company, Ltd in England with the names Wimet and Ardoloy. Cemented carbide was used to designate a metal matrix composite constituted by hard ceramic particles, normally WC, into a metallic matrix. Although the term cemented carbide is widely used in the United States, these materials are better known as hard metals in Europe. The importance of the invention is confirmed by noting that today, ninety years later, the same compositions, made by essentially the same process, are still a very significant product of tool materials industry. The next developments arose only a few years later with the discovery by Schwarzkopf that solid solutions of more than one carbide, particularly TiC, TaC, NbC, Mo2C and WC, have superior mechanical properties to the individual carbides in cemented hard metals (see Fig. 1). Another important advance in the development of hard metals happens in the late 1960s and early 1970s with the application of coatings on hard metal tools like TiC, TiN, TiCN and Al2O3 which are extremely hard and thus very abrasion resistant [2]. The development of metal cutting tools, which are suitable to cut a variety of materials such as gray cast iron, ductile nodular iron, austenitic stainless steel, nickel-base alloys, titanium alloys, aluminum, free-machining steels, plain carbon steels, alloy steels, and martensitic and ferrite stainless steels, has been very fast over the last four decades, having been greatly stimulated by much improved design and manufacturing techniques, such as coating technologies (Fig. 1) [4]. However, the development of new coatings and improved design are only one side of the coin. Continuous improvement of compositions and other manufacturing techniques led to improve the performance of hard metals and opened new areas of applications. Nowadays, the hard metal products have covered various industry sectors; almost 67% of total production of cemented carbides goes into metal cutting tools, about 13% for mining, oil drilling and tunneling

industries and 11% and 9% for wood working and construction industries, respectively [5]. New applications are constantly being identified for carbides, largely because of their excellent combinations of abrasion resistance, mechanical impact strength, compressive strength, high elastic modulus, thermal shock resistance and corrosion resistance. A special application for fine or ultrafine cemented carbides is in drills for the drilling of the very fine holes in printed circuit boards for the computer and electronic industries. Although other metal carbides, such as TiC, are also used in cutting tools, around 95% of all cemented carbide cutting tools are tungsten carbide-based [4]. There has been a continuous expansion in the consumption of cemented carbide from an annual world total of 10 t in 1930; to 100 t around 1935; 1000 t in the early 1940s; through 10 000 t in the early 1960s and up to nearly 30 000 t at present [6]. Annual production of tungsten for use in tungsten carbide worldwide was about 62 000 t in 2007. Nowadays, the worldwide cemented carbide industry production is estimated to be worth more than 10 billion € [7]. Future developments in the field of hard materials are mainly related to: (i) the raw materials supply and its scarcity resulting in high prices, owing to the main ore deposits being in areas less accessible to the “industrial world”, as is the case of Co; (ii) other economic factors; (iii) the general objective to save energy and (iv) the increasing demand to have special materials with defined and optimized properties. Besides those demands concerning economic and technological fields, another very important one, related to health, aims at the replacement of the traditional Co binder. The chronic inhalation of hard metal particles can produce an interstitial lung disease (hard metal disease). Recent studies have demonstrated that this disorder can be explained by the interaction between cobalt metal (Co) and tungsten carbide (WC) particles, which represent the main constituents of hard metal [8]. The inhalation effects of both particles (WC–Co) are more severe than those induced by Co metal particles alone, while WC particles could be considered as innocuous [8]. By the reasons pointed out before, one of the main topics of the actual research in the field of cemented carbides concerns the development of new composites having comparable or superior properties with partial or total substitution of the traditional cobalt binder by other more economic and less toxic materials. Composites with partial substitution of cobalt by nickel and iron are currently entering in the industrial production but the total cobalt replacement is envisaged. The Fe–Ni and Fe–Cr–Ni alloys are being currently investigated for such a purpose, showing highly promising results in terms of the technological properties (comparable or even superior properties can be obtained in these systems if the initial composition and processing conditions are well established) [9–13]. Nickel aluminides (Ni3Al and NiAl) and iron aluminides (Fe3Al and FeAl) are being also considered as binder materials for WC composites in high-temperature applications and extremely corrosive environments [14,15]. The knowledge of the phase diagrams for WC and different binders will be, therefore, extremely useful and opportune regarding the need to choose initial compositions and to select adequate sintering cycle conditions. For example, the modification of the binder composition through the increase of the nickel content reduces the carbon amount necessary to keep the alloy in the region free of η-phase [16,17]. Taking this into account, the objective of this work is to review the existent phase diagrams of W–C–M, where M is a binder metal or alloy for WC, such as M = Co, Fe, Ni, Fe–Ni, Fe–Al, Co–Fe–Ni, Cr–Fe and others. 2. Phase diagrams 2.1. System W–C The phase WC, also called δ-WC [18], is a line compound with insignificant deviation from stoichiometric composition, in contrast to

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

the cubic MC carbides which exist over a wide composition range MCx, where 1 N x N 0.5 [19]. Tungsten also forms two other carbides: WC0.5 (normally named β-W2C) with a hexagonal compact (hcp) metal sublattice with the C atoms partly filling the octahedral interstices; the second form is stable in a rather small temperature range at high temperatures and is a cubic NaCl type phase composition γ-WC1 − x, where x ≈ 0.4, in which the face centered cubic (fcc) interstitial sublattice is partly occupied by C [19,20]. β-W2C crystallizes in three structure types: PbO2, Fe2N and CdI2 types, denoted by β, β′ and β″, respectively. These polymorphs are stable at different temperatures and compositional ranges. Recently [18] a phase diagram in the W–C system was proposed in order to more accurately assess this system (Fig. 2). The W2C phase is originated from a eutectoidal reaction between elemental W and δ-WC at 1250 °C, melts congruently at approximately 2758 °C and forms eutectic melts with the W solid solution at 2215 °C and with γ-WC1 − x at near 2715 ± 5 °C [18]. Phases of W2C stoichiometry are obtained as intermediate products during WC production. The γ-phase results from a eutectoidal reaction between β and δ at 2516 °C and melts at approximately 2785 °C. δ-WC is the only binary phase stable at room temperature and has almost no compositional range up to 2384 °C but may become slightly carbon deficient between this temperature and its incongruent melting point. 2.2. WC cemented carbides diagrams The knowledge of the phase equilibria diagram of cemented carbides is an important tool to predict the phases after the sintering step and the adequate sintering temperature. The phase equilibria of cemented carbides are characterized by the existence of a two phase region between the carbide phase and the binder metal phase (Fig. 3, gray region). The best properties in WC–Co system have been obtained within the two phase region. The carbide-binder metal sections are eutectic systems and the special character of the WC-based materials is due partly to the constitution of this system. The systems of WC with Fe, Co or Ni show eutectic compositions richer in the carbide component (see Fig. 4) than other systems, namely TiC-based materials [21]. The sintering temperature of the hard metals lies above the eutectic temperature and the relatively high carbide content in solution leads

Fig. 2. Phase diagram of the W–C system, adapted from Gusev [18].

407

to good densification by solution–precipitation processes and contributes to obtain the desired microstructure. Since the hard metal compacts are predominantly densified by liquid phase sintering, the sintering temperature must be selected high enough that at least one phase of the systems is in the liquid state. Besides that, satisfactory combinations of hardness and transverse rupture strength of the cemented carbides are achieved when precipitation of other phases besides fcc binder metal and WC (e.g. graphite or the M6C carbide) on cooling from the sintering temperature is avoided.

2.2.1. System W–C–Co Cobalt has been the traditional choice as binder in the hard metal production, since the first marketable hard metal of WC–6 wt.% Co was produced in 1926 under the name of «WIDIA-N». Nowadays more than 90% of all WC-hard metals utilize Co as the preferred binder metal with contents between 3 and 30 wt.%. The apparent superiority of cobalt, relatively to other binders, is related to its best comminution characteristics in milling, superior wettability for WC, higher solubility of WC in cobalt at sintering temperatures and excellent properties [22]. Cobalt is a ferromagnetic metal with a hcp structure as the most stable phase at room temperature. It undergoes a phase transition from hcp into a fcc structure at 450 °C. The reasons for the dominant role of Co are also some unique properties of Co and the W–C–Co ternary system. The ternary diagrams for (Co, Ni and Fe) WC in the binder-rich corner were reported by Holleck [23] (Fig. 5). The width of the two-phase zone is smaller for iron and almost equal for cobalt and nickel. Over the last century, there have been several investigations concerning the W–C–Co system. The investigations have shifted from using only classic metallography and X-ray diffraction (XRD) toward combining experimental information with thermochemical descriptions to define equilibrium phase boundaries. Takeda [24] presented in 1931 the first approach to the W–C–Co diagram. Later, in 1952, Rautala and Norton [25] reported for the W–C–Co system two additional carbide phases, named θ and κ, having compositions Co3W6C2 and Co3W10C4, respectively. These authors [25] also proposed a pseudobinary WC–Co diagram, but the presence of the η-phase after rapid cooling could not be explained. Almost thirty years later Pollok and Stadelmaier [26] calculated an isothermal section through the W–C–Co system at 1400 °C (Fig. 6), which represents a usual sintering temperature. In particular, they identified an eta-carbide of composition Co2W4C and a carbide CoW3C, which have been reported by Rautala and Norton [25] as carbides θ and κ, respectively. From this figure it is obvious that WC is in stable equilibrium with liquid cobalt only within a rather narrow region of compositions. Gruter [27] constructed a pseudobinary WC–Co phase diagram with a eutectic temperature of 1280 °C. This author tried to explain the presence of the η-phase after rapid cooling by proposing that the ηphase remains in equilibrium with WC and liquid, even at stoichiometric compositions, at temperatures ranging from approximately 1280 to 1450 °C. The solubility of tungsten carbide in cobalt at the sintering temperature is high but decreases during cooling with reprecipitation on existing carbide grains. This preliminary work was followed by an exhaustive review of the thermodynamic properties of alloys and phase equilibria in the W–C–Co system by Guillermet [28], who calculated several isothermal sections of the system through thermodynamic methods. However, tradition and experience are strong tools too and most cemented carbide producers can make high quality cemented carbides without a detailed knowledge of the phase diagrams. This is particularly true for the W–C–Co system which is a very straightforward ideal system for cemented carbides, with a eutectic at 1280 °C which is well below cobalt's melting point of 1490 °C.

Author's personal copy 408

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

Fig. 3. Hard metal phase equilibria schematically [21].

Guillermet [28,29] reported a WC–Co phase diagram and a vertical section of the W–C–Co phase diagram calculated at 10 wt.% Co, Fig. 7 (a) and (b), respectively, which allow to easily follow what occurs at each temperature during cooling. The filled symbol on the composition axis represents the carbon content of the system corresponding to contents of W and C in the stoichiometric proportion for WC. Two other compositions are indicated by a and b. They correspond to the minimum and maximum carbon contents of alloys which will consist, just after solidification, of a mixture of only fcc + WC, assuming that there is no segregation. Considering the extremely low dissolution rate of the unstable M6C phase after the solidification [28], a and b are considered as the limits of the region of favorable carbon contents, regarding the absence of graphite or M6C in composites sintered in non equilibrium conditions. Regarding Fig. 7(b), the precipitation of graphite will not occur during solidification of a stoichiometric alloy made from pure cobalt and WC. However, increasing the carbon content to an excess of 0.05 wt.% will cause graphite precipitation during the cooling of the liquid. The effect of Co content in the extension of the critical carbon range was investigated by Guillermet, too [28]. This author reported that the change of Co content from 6 to 10 wt.% induces an increase of about 70% in the range of favorable carbon amounts. Some improvements in the description of the four-phase equilibria with the liquid phase within the system of W–C–Co have been recently reported [30]. The solubility of WC in Co is high but also strongly varies depending upon the temperature. The solubility of tungsten in cobalt increases with decreasing carbon content and has been reported to

Fig. 4. Schematical drawing of the pseudobinary systems WC–(Fe,Co,Ni). HM indicates the composition of commercial hard metals [21].

vary, accordingly, between 2 and 15 wt.% at around 1250 °C [31,32]. The solubility of tungsten at ambient temperature has been reported to be 3.5 wt.% in cobalt binder [33]. The solubility of cobalt in tungsten carbide is very small and can be neglected. With regard to the solid solubility of carbon in the binder, a typical range reported for WC–Co was 0–0.2 wt.% at elevated temperatures, with the higher values at the lower tungsten levels. Solid solubilities of tungsten and carbon in cobalt binders are inversely related [33]. 2.2.2. System W–C–Fe Iron presents a magnetic bcc crystal structure, also known as α-Fe (ferrite). It is, like Ni, a soft metal that can dissolve only a small concentration of carbon (0.021 wt.% at 910 °C). In the presence of higher amount of C, at temperatures above 912 °C and up to 1400 °C, α-Fe undergoes a phase transition from bcc to fcc, also called γ-Fe (austenite). This phase is metallic and non-magnetic and can dissolve considerably more carbon (2.04 wt.% at 1146 °C). Tungsten has already been used as an alloying element (2 up to 18 wt.%) in a steel category usually referred to as high speed steel (HSS). The tungsten increases efficiently the hardness and toughness of HSS and maintains these properties at the high temperatures generated in service [34]. Another characteristic of tungsten in Fe matrix is the ability to induce martensitic transformation during cooling down cycles, especially when other austenitization elements are present, like Ni or Mn. The increase of tungsten, as a hardening element in the transformed matrix, decreases the martensite strain and increases the lattice parameter of martensite while it has no effect on the martensite transformation time [35]. The W–C–Fe system has been subjected to many experimental and theoretical investigations over the years [20,36], because iron forms a ternary eutectic melt at only 1143 °C and is also a possible substitute for Co [33]. However, iron dissolves a much lower quantity of WC at the eutectic temperature (1143 °C) than cobalt at the eutectic temperature of 1320 °C [17]. Besides that, the width of the twophase zone is smaller for iron when compared with cobalt (Fig. 5). C– Fe–W isothermal sections at 1000 °C have been reported for the Fe rich corner (Fig. 8). Three ternary phases are discernible: two cubic η phases and hexagonal FeW3C. The η carbides are (FeW)12C with a large stoichiometric range for Fe and W and (FeW)6C with a smaller one. Additionally, the cementite phase was also reported as a ternary phase, (Fe,W)3C, because of the metastability of cementite relative to graphite could be stabilized with small contents of W (0.7%) up to 1000 °C [37].

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

409

C 1000 0C

14 WC + γ Fe(W,C)+C

12

at%C

10 8

WC γ·Fe(W,C)C

WC + γ Fe(W,C) +η·W 3 Fe 3 C

6 4

η·W3 Fe 3 C + γ·Fe(W,C)

W

12

14

2

γ

8

10

6

2

4

Fe

at%W C

C 1200 0 C

1100 0C

12 at%C

WC·Co(W,C)·C

12 WC·Ni(W,C)·C

10 Co(W,C)·C

10 Ni(W,C)·C

8

WC·W 3 Co 3 C ·Co(W,C)

WC·Ni(W,C)

10

8

6

4

2

Co

at%W

4 2

W 6Ni3 C·Ni(W,C)

W 3Co 3 C·Co(W,C)

12

6

4 2

14

8

WC·η·W 6 Ni3 C ·Ni(W,C)

6 WC·Co(W,C)

W

at%C

W

18

16

14

12

10

8

6

4

2

Ni

at%W

Fig. 5. Isothermal sections of the binder metal rich corner for the systems: a) W–Fe–C, b) W–Co–C and c) W–Ni–C. The dotted lines represent the theoretical WC-binder metal compositions.

The interesting sections for hard materials investigation concern iron amounts in the range between 4 and 20 wt.%. For such a purpose, Guillermet [38] calculated a section for the phase diagram of Fe–W–C with 10 wt.% Fe (Fig. 9). Comparing with Fig. 7(b), corresponding to the

Fig. 6. Isothermal section of the W–C–Co phase diagram at 1400 °C [28].

W–C–Co phase diagram for 10 wt.% Co, a decrease in the solid/liquid equilibrium temperatures is now noted, and a displacement of the favorable region towards values higher in carbon than the stoichiometric composition. On the other hand, the calculations suggest that the favorable effect of Fe additions on decreasing the solid/liquid equilibrium temperatures is coupled with an unfavorable reduction in the extent of the region of favorable carbon contents (between a and b). According to literature, four ternary carbide phases have been reported in the ternary W–C–Fe system, M4C, M6C, M12C and M23C6 [20]. Only M6C in the W rich part has been studied under well-controlled experimental conditions. The M6C carbide, which has many isomorphs among ternary carbides, contains at least two types of metallic atoms. Its formation is favored by the combination of one weak and other strong carbide formers, e.g. Fe and W. The M6C carbide is stoichiometric with respect to C but has a measurable range of existence with respect to Fe and W. Experimentally, the carbide has been found for compositions within the range Fe2W4C and Fe4W2C [26,36,39]. The sintering of an alloy with stoichiometric composition and 10 wt.% of iron binder can be possible within the two-phase field WC + liquid if the temperature is above 1400 °C, see Fig. 9. However, on cooling, the stoichiometric composition is within the three-phase

Author's personal copy 410

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

binder material almost exclusively used in die and wear parts because of its good corrosion resistance [43]. The melting point of nickel at 1455 °C is appreciably lower than that of cobalt at 1495 °C, however to obtain satisfactory densification it is necessary to use increased sinter time and temperature [22,44]. This leads to a tendency to pick-up carbon from the graphite vacuum furnaces leading to graphite precipitation. The higher vapor pressure of nickel (ten times that of cobalt) at sintering temperature also causes considerable loss of the nickel binder and it is, therefore, necessary to control the working pressure [22]. The loss of nickel in practice has been reported to be 10 wt.% or more at sintering temperatures [22]. Comparing the W–C–Ni phase diagram, Fig. 10(b), with the corresponding section of W–C–Co, Fig. 10(a), it is possible to examine the consequences of a full substitution of Co by Ni. The width of the fcc/WC region remains essentially unchanged, but the range of favorable carbon contents moves towards lower values as can be seen by comparison with the stoichiometric composition [38]. Besides that, the change from W–C–Co to W–C–Ni involves an appreciable increase in the equilibrium temperatures of eutectic and peritectic points. The partial substitution of Co by Ni induces an increase in the as referred equilibrium temperatures [38] that depend on the Co:Ni ratio. However, according to Guillermet's [38] conclusions, the predicted effects on the solid/liquid equilibrium temperature remain relatively small, until full substitution of Co by Ni. At sintering temperatures, the stoichiometric composition is within the two-phase field WC + liquid, but on cooling, graphite will precipitate as indicated in the section of the phase diagram shown in Fig. 10(b). This situation is opposite of iron, so attempts have been made in using Fe–Ni binders with more favorable conditions to stoichiometric composition. The tungsten solubility in nickel has been reported to be of the same order as that in cobalt, but with the inclination to be somewhat higher [33,44]. The solubilities of W and C in the liquid phase of the ternary W–C–Ni system were reported by Uhrenius [42] to be about 5.0 and 2.0 wt.%, respectively at 1350 °C. At ambient temperature the solubility of tungsten in nickel has been reported to be 5.4 wt.% [33].

Fig. 7. Vertical section of the W–C–Co phase diagram (a) between stoichiometric WC and Co [29]; (b) calculated at 10 wt.% Co [28]. The solid symbol on the composition axis indicates the stoichiometric composition (5.5 wt.% C). The points denoted by a and b define, respectively, minimum and maximum carbon contents of alloys which are in two-phase state of fcc + WC just after the equilibrium solidification.

field WC + M6C + Fe (fcc) and the embrittling η-phase (M6C) will form. However, additions of free carbon to Fe–WC alloys, exceeding the stoichiometric amount required for the tungsten carbide, can inhibit the formation of large grains of the brittle η-phase [40,41]. The solubility of Fe in the WC carbide is reported [36] to be very limited, about 0.6 wt.% at 1250 °C, while the solubility of W in ferrite phase was found to be about 4.5 wt.% at 700 °C and 5.0 wt.% at 1250 °C [36]. Uhrenius [42] reported the solubility of W and C in the liquid phase, in equilibrium with WC at 1350 °C, and a carbon activity close to ac = 1, to be about 9.0 and 4.7 wt.%, respectively. 2.2.3. System W–C–Ni Nickel has received the most attention as an alternative binder to cobalt. Its structure and properties are similar and the lattice parameter of fcc nickel is only slightly less than fcc cobalt. The principal differences are (i) fcc cobalt is metastable and can transform to hcp and (ii) cobalt is much more strongly magnetic [33]. Nickel is a

2.2.4. System W–C–Fe–Ni The satisfactory properties of Fe–Ni bonded WC carbide tools are associated with an accurate control of the composition, in order to guarantee the absence of M6C and graphite in the sintered alloy [45– 47]. This means that the global composition of the system at the sintering temperature should fall inside the WC + liquid region of the respective phase diagram. During cooling, the composition should fall within the fcc + WC region and should not precipitate graphite or M6C. So, in order to achieve satisfactory results one should be able to accurately define the region of critical carbon contents once the Fe:Ni ratio has been selected. For specific applications, the metastability of the binder is explored using Fe–Ni–C binders in order to obtain martensitic transformation [48,49]. Some constraints, like the binder composition imposed by the need to avoid graphite and M6C, the binder deformation imposed by the surrounding carbide grains, the elastic and thermal mismatch stresses and strains between the carbide and the binder can significantly modify the course and characteristics of the martensite transformation in the binder. A assessment based upon the use of thermodynamic models has been attempted by Guillermet [50,51]. The author [50,51] calculated sections of temperature and composition from the W–C–Fe–Ni system, for Fe + Ni = 20 wt.% and Fe:Ni ratios of 3:1, 1:1 and 1:3. For comparison between the graphs, a general increase in the temperatures of the solid–liquid equilibria when the Ni content increases is observed. Simultaneously, the two-phase field fcc + WC moves towards lower carbon contents with respect to the stoichiometric composition. It is also observed, that the width of the twophase field varies with the content of Ni, diminishing first, but increasing again when it approaches the ratio Fe:Ni of 0:1.

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

Fig. 8. Fe–W–C, isothermal sections at 1000 °C and respective amplifications for the Fe rich corner.

411

Author's personal copy 412

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

b) increasing the Ni content in the initial composition decreases the necessary carbon excess. These observations can be explained by the position of the twophase region and by the displacement of both lines in Figs. 11(b) and 12(b) to lower carbon contents when the Ni content is increased. In particular, a low excess carbon content is required if an Fe:Ni ratio 3:1 is adopted. This may explain the satisfactory results obtained by adopting this ratio in early attempts [54,55] to substitute Co by Fe:Ni. It should be emphasized that due to the scarcity of information in this quaternary system some carbide phases are not enclosed in the curves calculated by Guillermet [50]. However, they had been detected in Fe–W–C and Ni–W–C systems, namely the M12C carbide which may form at low values of the carbon activity and high Ni contents, the M4C and M2C carbides [50].

Fig. 9. Vertical section of the Fe–W–C phase diagram, calculated at 10 wt.% Fe [38]. The solid symbol on the composition axis indicates the stoichiometric composition (5.5% C). The points denoted by “a” and “b” define, respectively, minimum and maximum carbon contents of alloys which are in two-phase state of fcc + WC just after the equilibrium solidification.

In order to exemplify the previous interpretation, the calculation of a vertical section of the W–C–Fe–Ni diagram from the work carried out by Guillermet [50,51] is represented in Fig. 11(a). The calculation was performed for an alloy containing 20 wt.% of binder and Fe:Ni ratio equal to 3:1. It can be seen from that figure that the stoichiometric mixture of WC would precipitate M6C during sintering at 1350 °C. Even though the alloy after cooling would arrive at a stable two-phase field of FeNi-binder (fcc) and WC at temperatures below 1300 °C, the precipitates of M6C formed at sintering temperature, would not dissolve during cooling and will stay in a metastable form at lower temperatures. The calculated section shows that it would be better to increase the carbon content of the alloy to a value between 5.0 and 5.1 wt.% in order to avoid either M6C precipitation or the graphite precipitation which would occur if the carbon content exceeds 5.1 wt.% [52,53]. The positions of these points change with the Fe + Ni content and the Fe:Ni ratio, but the lines defined by their displacement in the temperature-composition space can be directly calculated. The temperature projection of the areas in the Fe–Ni–W–C system which would give graphite or M6C phase in equilibrium with the liquid phase during sintering or subsequent cooling was calculated by Guillermet [50,51]. The zone in the middle, in Fig. 11(b), shows the most favorable compositions, where none of these two phases will precipitate. The temperature projection presented in Fig. 11(b) corresponds to an amount of Fe + Ni equal to 20 wt.%. From the figure it is observed that stoichiometric mixtures of WC and the pure metals will give WC + fcc-metal only if the ratio Ni/(Ni + Fe) falls in the region between 0.45 and 0.65. When the (Fe + Ni) content is decreased, the region of interest becomes considerably narrower. This is illustrated in Fig. 12(a) and (b) were the results obtained by the calculation for Fe + Ni = 10% are presented. The analysis of these figures and Fig. 11(b), may offer a simple explanation of an empirical observation reported by Agte [54] and more recently by Moskowitz [55]. They observed that: a) satisfactory properties may be obtained if the carbon content of the Fe–Ni–W–C alloy exceeds the stoichiometric composition;

Fig. 10. (a) Vertical section of the W–C–Co phase diagram, calculated at 10 wt.% Co [31]. (b) Vertical section of the W–C–Ni phase diagram, calculated at 10 wt.% Ni [38]. The solid symbol (•) on the composition axis indicates the stoichiometric composition (5.5 wt.% C). The points denoted by a and b define, respectively, minimum and maximum carbon contents of alloys which are in two-phase state of fcc + WC just after the equilibrium solidification.

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

413

C increases up to 9.0 and 4.7 wt.%, respectively, as the Fe:Ni ratio increases until 1:0; (c) the solubility of these elements is more limited when the liquid phase is constituted only by Ni. Among the interesting findings of Guillermet [38,50] the following are emphasized: a) the substitution of cobalt by nickel does not alter the width of the two phase region but shifts it to an area of lower carbon content; b) there is an appreciable increase in the four phase equilibrium temperature L ↔ WC + β + C; c) adding iron instead of nickel displaces the favorable region to higher carbon contents, lowers the liquidus temperature and narrows the width of the two phase region WC + β;

Fig. 11. (a) Vertical section of the Fe–Ni–W–C phase diagram calculated at Fe + Ni = 20 wt.% and %Fe:%Ni = 3:1 [50,51]. The stoichiometric composition is represented by the traced line. (b) Temperature projection calculated at Fe + Ni = 20 wt.%. The lines describe the compositions of a mixture of WC + liquid in equilibrium with fcc + M6C (left) or fcc + graphite (right). The asterisk on the composition axis indicates the stoichiometric composition.

The solubility of WC in Fe–Ni–W–C liquid at 1500 °C for various Fe: Ni ratios has been calculated by Guillermet [50] and compared with the experimental data according to Gabriel [56] (Fig. 13). The projections have four almost parallel curves corresponding to the Fe–W–C system and to the Fe–Ni–W–C system with Fe:Ni ratios of 3, 1 and 0.3. A fifth curve, for the Ni–W–C system, intersects the reported curve for Fe:Ni = 1. All these curves extend between two limits: the left-hand limit represents the appearance of graphite and the righthand limit represents the appearance of M6C carbide. Small discrepancies between the calculated and the experimental points were noted (Fig. 13). These results are in agreement with the ones reported by Uhrenius [42] at 1350 °C: (a) the solubilities of W and C in the liquid phase depend on the Fe:Ni ratio; (b) the solubility of W and

Fig. 12. (a) Vertical section of the Fe–Ni–W–C phase diagram calculated at Fe + Ni = 10 wt.% and %Fe:%Ni = 1:1 [50]. (b) Temperature projection calculated at Fe + Ni = 10 wt.%. The lines describe the compositions of a mixture of WC + liquid in equilibrium with fcc + M6C (left) or fcc + graphite (right). The asterisk on the composition axis indicates the stoichiometric composition.

Author's personal copy 414

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

Fig. 15. Plot of the Fe–Al–W–C section, at 1450 °C, showing the solubility of W and C in liquid iron aluminide (Fe–40 at.% Al). The results are from calculations using Thermocalc™ [57].

Fig. 13. Projection of the calculated solubilities of WC in Fe–Ni–W–C liquid, at 1500 °C and various Fe:Ni ratios, in comparison with experimental data (Gabriel [56]); Fe–W–C (□), Ni–W–C (×), Fe:Ni = 3:1 (O), Fe:Ni = 1:1 (Δ), Fe:Ni = 1:3 (+).

d) by the addition of iron and nickel together it is possible to arrange for the favorable zone to remain centered roughly on the stoichiometric composition.

2.2.5. System W–C–Co–Fe–Ni A section calculated by Guillermet [38] for a composition of WC with 10 wt.% of binder, containing 5 wt.% of Fe and equal parts of nickel and cobalt is represented in Fig. 14. The effect of Fe additions depends on the Ni:Co ratio [38]. For example, keeping the same percentage of Fe (5 wt.%) and modifying the Ni:Co ratio between 1:4 and 1:1, a decrease in the width of the favorable zone is observed, being approximately 50% of the width observed in the W–C–Co–Ni

Fig. 14. Vertical section of the W–C–Co–Fe–Ni phase diagram calculated at 5 Fe, 2.5 Ni and 2.5 wt.% Co: the stoichiometric composition is indicated by the solid symbol (•) at 5.5 wt.% C; the alloys with carbon amount between a and b solidified with the formation of only WC + fcc [38].

diagram, for the same Ni:Co ratio and binder percentage. This effect is more accented when the iron addition is made to the alloy with the highest Ni:Co ratio, where reductions of about 70% of the width of the favorable zone in the diagram W–C–Co–Ni are observed [38].

2.2.6. System W–C–Fe–Al Iron aluminide received attention as a binder material, since it can easily form an aluminum oxide layer and can provide excellent corrosion resistance in a wide range of corrosive environments [14,15]. Besides that, aluminum possesses lower density, lower melting point (660 °C), is easily recycled and is abundant in the earth's crust. The information concerning the phase equilibria in the quaternary Fe–Al–W–C system is scarce. Subramanian et al. [57] reported the solubility of the carbide phases in liquid iron aluminide at 1450 °C. The results show that molten Fe–40 at.% Al in equilibrium with graphite, dissolves ≈5 at.% carbon and 1 at.% tungsten (Fig. 15). The low solubility values of the carbide phases in liquid iron aluminide restrict the liquid phase sintering of mixed powders, since it only yields a dense and homogenous microstructure for carbide volume fractions smaller than 0.8. Further research is aimed at increasing the WC content in these materials and improving creep strength and lowtemperature ductility in air.

Fig. 16. The calculated isothermal section of the W–C–Cr system at 1350 °C. The Greek letters α1 and α2 are used to denote the W and Cr-rich side of the BCC phase [62].

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

415

Author's personal copy 416

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

Fig. 17. (a) Isopleth W–C–Fe–0Cr (1250 °C) [64]; (b) Isopleth W–C–Fe–2Cr (1250 °C) [64]; and (c) Isopleth W–C–Fe–10Cr (1250 °C) [64].

2.2.7. System W–C–Cr Chromium carbide (Cr3C2) is normally added to the WC–Co system as a grain-growth inhibitor, but it also improves corrosion and oxidation resistance, because it forms a tenacious, self-repairing film [43]. Besides that, hard metals with chromium in the binder are reported to have higher hardness and work hardening rate and a marked improvement in abrasion resistance [22,58]. However, chromium is also a strong carbide former and the solubility of chromium in the Co binder is thus very much limited by the formation of carbides. Chromium restricts the solution/re-precipitation stage and as a result higher temperatures and longer times were required to sinter these materials [59]. The information in the literature is restricted to the ternary W–C– Cr system. Isothermal sections at temperatures higher than 1300 °C, give the same kind of phase diagram even, although details differ [60,61]. No information on the system at lower temperatures has been found in the literature. Fig. 16 shows a calculated section at 1350 °C in the ternary W–C– Cr system calculated by Gustafson [62]. A maximum solubility of 8.6 at.% W in M23C6 was determined, whereas the other stable Cr carbides, M7C3 and M3C2, also dissolve low amounts of W, about 3.8 and 8.8 at.%, respectively [62]. The C-rich side is dominated by the two-phase equilibrium WC + M3C2. The temperature for the invariant four-phase equilibrium WC + M2C + M3C2 + graphite was placed at 1498 °C [62]. 2.2.8. System W–C–Cr–Fe There is relatively little experimental information available on the quaternary system W–C–Cr–Fe in the literature. Both Cr and W form carbides and many different types of carbide may form in steels

containing these elements. The first schematic diagram of the quaternary W–C–Cr–Fe system was presented by Goldschmidt [39] in the early 1950s. Only one new carbide, not present in any of the binary systems, was included in that diagram. The ternary M6C carbide, also stable in the W–C–Fe system, was shown to possess a limited solid solubility for Cr. The diagram also indicated that all Cr in the Cr23C6 carbide can be replaced by Fe and some W to attain the composition Fe21W2C6. Uhrenius et al. [63] presented experimental data on equilibria between Fe-rich fcc phase (austenite) and carbides in the W–C–Cr–Fe system at 900, 1000 and 1100 °C. Their work was mainly concerned with equilibria involving M23C6, but some results on M6C and cementite were also presented. The solid solubility for Fe in the M23C6 carbide is always less than 50% in the ternary C–Cr–Fe system, but the Fe content in the quaternary M23C6 carbide can be higher than 75% [62]. The region of existence of the M6C carbide has been studied by Bergström [64]. He reported the existence of two additional carbides, M12C and M4C, also found in the ternary system C–W–Fe. The M6C and M12C carbides were found to dissolve up to 8 and 4 at.% Cr, respectively at 1250 °C. The maximum solubilities at 1250 °C of Cr in other phases, like Fe and W, were reported to be 10 at.% and 2 at.%, correspondingly. The isothermal sections at 1250 °C of the W–C–Fe and W–C–Fe–Cr at Cr levels of 2 and 10 at.% constructed by Bergström [64] are represented in Fig. 17. The isopleth W–C–Fe–2Cr (Fig. 17(b)) contains the same regions as the isopleth Fe–W–C–0Cr (Fig. 17(a)) since no phase, except WC has a Cr solubility limit below this section. The isopleth W–C–Fe–10Cr (Fig. 17(c)) is positioned above the maximum Cr solubilities of all phases. The maximum solubility of Cr in Fe lies on this level, which implies that the four-phase tetrahedron,

Author's personal copy C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

M6C + M12C + μ(Fe3W2) + Fe, the three-phase volume, M6C + M12C + Fe, and the two-phases volume M6C + Fe and M12C + Fe, touch the section only in one corner [64]. The existence of an additional phase not covered in any of the above-mentioned papers has been reported by Goldschmidt [65]. He found a new intermetallic phase, possessing the beta-manganese structure, at high temperatures. The phase occurred over an appreciable composition range in the W–C–Cr–Fe system, a representative composition being Fe10Cr9W2C. The phase was stable only above about 1350 °C and it is decomposed into a mixture of Fe-rich bcc and chi phases on cooling. The quaternary systems Co–Cr–W–C and Ni–Cr–W–C have been scarcely explored and little information is available. For the Co–Cr– W–C a new chi-phase named Co25Cr25W8C2 with α-manganese type structure has been reported [66]. 3. Concluding remarks Phase diagrams of cemented carbides have been established for the ternary systems of W–C–M with M = Co, Fe, Ni and Cr. The W–C– Co system is the best documented one in correspondence to the intensive research and to the large scale production of the hard metal components within these compositions. In this case, the equilibrium compositions and phase structures are reasonable known, while for the other systems, especially for W–C–Cr, this knowledge is more limited. Phase diagrams of systems with more than three components have also been developed for W–C–Fe–Ni, W–C–Fe–Al, W–C–Fe–Cr and W– C–Co–Fe–Ni, fixing the amount of one or more components. Therefore, the information is usually restricted to one or two percentages of metallic binder (usually 10 and/or 20 wt.%) and to constant ratios between the metallic components, even for actual cemented carbide composites with increasing industrial production, such as W–C–Co–Ni–Fe. There is also a lack of information on the equilibrium phase compositions and related phase structures. Consequently, the research and the technological development of cemented carbides are still being frequently done by trials, based on phase diagrams of similar systems. New binders for cemented carbide tools are currently being investigated for instance stainless steels, Al–Ni and Al–Fe alloys, making necessary the development of related phase diagrams. This knowledge would be extremely useful to guide the selection of initial compositions and thermal processing conditions and to reach final phase compositions which may present synergies in terms of the composite properties. Acknowledgments The author C.M.F. gratefully acknowledges the financial support of the POCTI program of the Portuguese Foundation for Science and Technology (FCT) and the European Social Fund (FSE). References [1] Pastor H. Centenaire de la découverte du carbure de tungstène par Henri Moissan; historique du développement de ce matériau. La Revue de Métallurgie-CIT/Science et Génie des Matériaux; 1997. p. 1537–52. [2] Cornwall RG, German RM. WC–Co enjoys proud history and bright future. Metall Powder Rep 1998;J-A53(7–8):32–3. [3] Schwarzkopf P, Kieffer R. Cemented carbides. Macmillan Co.; 1960. [4] Yao Z, Stiglich JJ, Sudarshan TS. Nano-grained tungsten carbide–cobalt (WC/Co). Mater Modif 1999:1–27. [5] Houping ZLW. Status of cemented carbide industry both at home and abroad. Cemented Carbide 2009;26(2):122–7. [6] Schubert WD, Lassner E, Böhlke W. Cemented carbides—a successful story, Itia. http://www.itia.org.uk. 2010 accessed June 2010. [7] Cornwall RG, German RM. Think bigger! The future is bright for MIM. Met Powder Rep 2004;59(11):8–11.

417

[8] Lison D, Lauwerys R. The interaction of cobalt metal with different carbides and other mineral particles on mouse peritoneal macrophages. Toxic Vitro 1995;9(3): 341–7. [9] Fernandes CM, Senos AMR, Vieira MT. Sintering of tungsten carbide particles sputterdeposited with stainless steel. Int J Refrac Met Hard Mater 2003;21:147–54. [10] Fernandes CM, Senos AMR, Vieira MT, Fernandes JV. Composites from WC powders sputter-deposited with iron rich binders. Ceram Int 2009;35:1617–23. [11] Fernandes CM, Senos AMR, Vieira MT, Antunes JM. Mechanical characterization of composites prepared from WC powders coated with Ni rich binders. Int J Refrac Met Hard Mater 2008;26:491–8. [12] González R, Echeberría J, Sánchez JM, Castro F. WC–(Fe, Ni, C) hardmetals with improved toughness through isothermal heat treatments. J Mater Sci 1995;30 (13):3435–9. [13] Suzuki H, Yamamoto T, Chujo N. Properties of WC–10% (Ni–Fe). Jpn Soc Powder Powder Metall J 1967;14:26–31. [14] Mosbah AY, Wexler D, Calka A. Abrasive wear of WC–FeAl composites. Wear 2005;258:1337–41. [15] Ahmadian M, Wexler D, Chandra T, Calka A. Abrasive wear of WC–FeAl–B and WC–Ni3Al–B composites. Int J Refract Met Hard Mater 2005;23:155–9. [16] Meredith B, Milner DR. Densification mechanisms in the tungsten carbide–cobalt system. Powder Metall 1976;1:38–45. [17] Upadhyaya GS, Bhaumik SK. Sintering of submicron WC–10 wt.% Co hard metals containing nickel and iron. Mater Sci Eng A 1988;A105-106:249–56. [18] Gusev AI, Kurlov AS. Tungsten carbides and W–C phase diagram. Inorg Mater 2006;42(2):121–7. [19] Cottrell AH. Cohesion in tungsten carbide. Mater Sci Technol 1995;11(3):209–12. [20] Gustafson P. A thermodynamic evaluation of the C–Fe–W system. Metall Trans 1987;18 A:175–88. [21] Holleck H. Constitutional aspects in the development of new hard materials; 1983. p. 849–61. [22] Penrice TW. Alternative binders for hard metals. Carbide Tool J 1988;20(4):12–5. [23] Holleck H. Annual Report, Aufbau, Herstellung, und Eigenschaften hochschmelzender Verbindungen and Systeme (Harstoffe und Hartmetalle), KfK-Ext. 6/78-1, Institute for Materials and Solid State Research, Kernforschungszentrum in Karlsruhe, Germany 1977; 57–65 including English translation of Oberacker R. et al., “Properties of tungsten carbide hard metals with Fe–Co–Ni binder alloys, part I: effect of the composition, including carbon content,” 57–65. [24] Takeda S. Technol Rep Tohoku Imp Univ 1931;10:42–92. [25] Rautala P, Norton JT. Tungsten–cobalt–carbon system. Trans AIME 1952;194: 1045–50. [26] Pollock CB, Stadelmaier HH. The eta carbides in the Fe–W–C and Co–W–C systems. Metall Trans 1970;1:767–70. [27] Gruter M. Thesis: Untersuch ungen in den Systemen Co–C, Co–WC. Germany, Münster 1959. [28] Guillermet AF. Thermodynamic properties of the Co–W–C system. Metall Trans 1989;20A:935–56. [29] Guillermet AF. Composition dependence of the Curie temperature of Co–W–C alloys predicted from thermodynamic data. Z fuer Metallkunde Mater Res Adv 1989;80(8):549–55. [30] Markström A, Sundman B, Frisk K. A revised thermodynamic description of the Co–W–C system. J Phase Equilib Diffus 2005;26(2):152–60. [31] Chaporova IN, Shchetlina YeA. Investigations of the carburizing process in tungsten carbide hard alloys with cobalt and nickel. Hard Met Prod Technol Res USSR 1964:196–211. [32] Jia K, Fischer TE, Gallois B. Microstructure, hardness and toughness of nanostructured and conventional WC–Co composites. Nanostruct Mater 1998;10(5):875–91. [33] Tracey VA. Nickel in hardmetals. Refract Met Hard Mater 1992;11:137–49. [34] Tekin E, Richman MH. An electron-microscopical study of the tempering of tungsten high-speed-steels. Metallography 1970;3(3):327–35. [35] Kenawy MA, Nagy MR, Sakr EM. Martensite–austenite transformation in maraging steel alloys during tensile deformation and thermal cycling. J Mater Sci 1986;21: 3071–7. [36] Bergström M. The eta-carbides in the ternary system Fe–W–C at 1250 °C. Mater Sci Eng 1977;27:257–69. [37] Raynor GV, Rivlin VG. Phase equilibria in iron ternary alloys. Inst Met 1988;4: 200–12. [38] Guillermet AF. The Co–Fe–Ni–W–C phase diagram: a thermodynamic description and calculated sections for (Co–Fe–Ni) bonded cemented WC tools. Z Metallkunde 1989;80(2):83–94. [39] Goldschmidt HJ. The structure of carbides in alloy steels part II—carbide formation in high-speed steels. J Iron Steel Inst 1952;170:189–204. [40] Moskowitz D, Ford MJ, Humenik M. High-strength tungsten carbides. Mod Dev Powder Metall 1970;5:225–34. [41] Moskowitz D, Ford MJ, Humenik M. High-strength tungsten carbides. Int J Powder Metall 1970;6(4):55–64. [42] Uhrenius B, Forsen K, Haglund BO, Andersson I. Phase-equilibria and phasediagrams in carbide systems. J Phase Equilib 1995;16(5):430–40. [43] Raghunahan S, Caron R, Sandell P. Tungsten carbide technologies. Adv Mater Processes 1996;4:21–3. [44] Ekemar S, Lindholm L, Hartzell T. Aspects on nickel as a binder metal in WC-based cemented carbides. Proc. 10th Plansee Seminar - Metallwerk Plansee, 1; 1981. p. 477–824. [45] Prakash L, Holleck H, Thümmler F, Walter P. The influence of the binder composition on the properties of WC–Fe/Co/Ni. Mod Dev Powder Metall 1981;14: 255–68.

Author's personal copy 418

C.M. Fernandes, A.M.R. Senos / Int. Journal of Refractory Metals and Hard Materials 29 (2011) 405–418

[46] Chaporova IN, Kudryavtseva VI, Sapronova ZN, Sychkova LV. An investigation of the WC–Fe–Ni system. Russ Metall 1981;4:211–5. [47] Chaporova IN, Kudryavtseva VI, Sapronova ZN. Structure and properties of a hard alloy based on tungsten carbide with Fe–Ni binder. Met Sci Heat Treat 1978;20(5– 6):402–4. [48] Viswanadham RK, Lindquist PG. Transformation-toughening in cemented carbides: part I. Binder composition control. Metall Trans A 1987;18A:2163–73. [49] Kakeshita T, Wayman CM. Martensitic transformations in cermets with metastable austenitic binder I: WC–(Fe–Ni–C). Mater Sci Eng 1991;A141:209–19. [50] Guillermet AF. An assessment of the Fe–Ni–W–C phase diagram. Z Metallkunde 1987;78(3):165–71. [51] Guillermet AF. Use of phase-diagram calculations in selecting the composition of Fe–Ni bonded WC tools. Int J Refract Hard Met 1987;6(1):24–7. [52] Uhrenius B. Phase diagrams as a tool for production and development of cemented carbides and steels. Powder Metall 1992;35(3):203–10. [53] Uhrenius B. Contribution to the knowledge of phase equilibria in tungsten–carbon based systems. Scand J Metall 1991;20(1):93–8. [54] Agte C. Development of the tungsten carbide technology in the GDR. Neue Hütte 1957;9(537):544. [55] Moskowitz D. Abrasion resistant iron–nickel bonded tungsten carbide. Mod Dev Powder Metall 1977;10:543–51.

[56] Gabriel A, Gustafson P, Ansara I. A thermodynamic evaluation of the C–Fe–Ni system. TRITA-MAC-0285. Stockholm: Royal Institute of Technology; 1985. [57] Subramanian R, Schneibel JH. FeAl–TiC and FeAl–WC composites—melt infiltration processing, microstructure and mechanical properties. Mater Sci Eng 1998;A244: 103–12. [58] Uhrenius B, Pastor H, Pauty E. On the composition of Fe–Ni–Co–WC-based cemented carbides. Int J Refract Met Hard Mater 1997;15(1–3):139–49. [59] Roebuck B, Bennett EG. Phase size distribution in WC/Co hardmetal. Metallography 1986;19:27–47. [60] Rudy E, Chang YA. Plansee Proceedings 1964; 1965. p. 786–822. [61] Stecher P, Benesovky F, Nowotny H. Chromium–tungsten–carbon system. Planseeber Pulvermet 1964;12:89–95. [62] Gustafson P. A thermodynamic evaluation of the C–Cr–Fe–W system. Metall Trans 1988;19A:2547–54. [63] Uhrenius B, Frondell S. Met Sci 1977;11:73–81. [64] Bergström M. The eta-carbides in the quaternary system Fe–W–C–Cr at 1250 °C. Mater Sci Eng 1977;27(1):271–86. [65] Goldschmidt HJ. Occurrence of the β-manganese structure in transition metal alloys and some observations on x-phase equilibriums. Metallurgia 1957;56:17–26. [66] Ettmayer P, Seifahrt H. Über eine neue Chi-Phase im System Co–Cr–W–C. Monastshefte Chem 1969;100:616–23.