MICROSTRUCTURE AND PROPERTIES OF IRON–COBALT–TUNGSTEN ALLOYS HARDENED DURING CASTING

A. D. KLEMENT'EV

(Presented by the scientific seminar of the departments of physical metallurgy, equipment and technology of heat treatment of metals; technology of metals; equipment and technology of welding production)

In studies [1, 2] devoted to the development of cast cutting dispersion-hardening iron–cobalt–tungsten alloys, certain regularities were found in the relationship between the structure and the properties of these alloys. Subsequent investigations [3, 4] indicated the presence of a more complex dependence of the properties on the structure.

The present article is devoted to studies of the microstructure, of the factors affecting it, and of its relationship to properties important for evaluating the cutting ability of alloys hardened during casting.

The alloys were melted in a vacuum induction furnace; pouring was carried out inside the furnace into a metallic mold. The alloys hardened during casting were subjected only to tempering. An indispensable condition of such a technological process is self-hardening of the alloys, which imposes special requirements on the chemical composition and on the cooling conditions of the alloys during casting.

The general picture of the influence of chemical composition and transformations in the alloys on the microstructure can be represented by a schematized spatial phase diagram of the principal components of the alloy, constructed from the data of [5], shown in Fig. 1 for the temperature interval from 20 to 1300°C.

The extensive regions of the α- and γ-states of the alloys are separated by a narrow intermediate region, which has a broad bend at a temperature level of about 950°C. In this case the γ-region proves to be greatly expanded at high temperatures, and the α-region, correspondingly, expanded at low temperatures. As a result, a large region of alloys is obtained that undergo a γ–α transformation, and a very narrow region, adjoining the Fe–W side of the diagram, of purely ferritic alloys.

The surface of limiting solubility of tungsten in the complex solid solution forms a narrow region located along the entire Fe–Co side of the diagram. The narrowing of this region with decreasing temperature indicates the presence of a large region of alloys prone to dispersion hardening due to the precipitation of iron and cobalt tungstides (Θ phase). The varieties of phase constitution of dispersion-hardening alloys in the quenched state may be the following (Fig. 1): alloys without a γ–α transformation, of ferritic type 1—1 and austenitic type 3—3; alloys with a γ–α transformation without an excess phase, type 2—2, and with an excess phase, type 4—4. In addition, there may be allo-

alloys containing simultaneously the $\alpha$- and $\gamma$-phases, as well as with an excess phase or without it. To this must be added the broad possibilities for varying the structural constituents and, consequently, the diversity of microstructures and of the alloy properties determined by them.

Let us consider the effect of additions on the alloys under study. As was shown in [1], alloys whose structure can be improved by recrystallization, i.e., alloys undergoing the $\gamma$–$\alpha$ transformation, should be used as cutting alloys.

Fig. 1. Schematic ternary phase diagram
(according to [5])

As the starting alloy for introducing additions, an alloy was chosen whose composition corresponds to point $c'$ of the phase diagram (Fig. 1). This alloy undergoes the $\gamma$–$\alpha$ transformation at a relatively low concentration of cobalt and tungsten (20% each), but possesses a sufficiently high degree of dispersion hardening. Its position at the junction of the principal regions of the alloy phase diagram makes it possible to evaluate readily the influence of additions on the microstructure. Upon alloying, the alloy will, as it were, move into adjacent regions of the diagram; at the same time the temperature of the $\gamma$–$\alpha$ transformation will also change.

An important restriction must be imposed on the choice of alloying elements: they must not worsen the conditions of dispersion hardening of the alloys. For this purpose a hypothesis was developed [6] that makes it possible to evaluate elements according to their influence on the amount of excess phase participating in dispersion hardening of the alloy, and that provides the possibility of expediently changing the ratio of elements in the complex solid solution. The hypothesis has found good qualitative confirmation in a wide range of systems of dispersion-hardening alloys and in our experimental studies.

It has been shown that nickel can be used without limitation, as can significant concentrations of manganese, chromium, silicon, and vanadium; small additions of titanium, niobium, tantalum, and beryllium; and boron and zirconium only as modifiers. Provided that the tungsten concentration in the alloy is reduced, the amounts of chromium, manganese, silicon, and vanadium additions may be increased and, what is important from the standpoint of replacing tungsten, so too may titanium, niobium, tantalum, and beryllium. Molybdenum is analogous to tungsten and may be used over broad ranges to replace it.

Another important criterion for evaluating additions is their influence on the stability of the regions of existence of the $\alpha$- and $\gamma$-phases. For the alloys under study, the proposition [7] is fairly well justified that elements with a smaller number of electrons in an unfilled $d$ shell stabilize the low-temperature phase, whereas those with a larger number stabilize the high-temperature phase, provided that the crystal-lattice types of the solvent modification and the alloying element correspond. Thus the $\alpha$-phase is stabilized by W, Mo, Si, Cr, V, Ti, Be, and others; the $\gamma$-phase is stabilized by Co, Ni, Mn (Mn is an exception to the rule cited).

Elements that expand the stability range of the $\gamma$-phase and have atomic sizes close to that of iron (Ni, Mn) promote some reduction in the amount of the second phase, which corresponds (Fig. 1) to a shift of the alloy from point $c'$ into the $\gamma$-region (alloy of type 2—2). In this case the temperature of the $\gamma$-$\alpha$ transformation should decrease. With substantial alloying by these elements, complete suppression of the $\alpha$-phase is possible (alloy of type 3—3). With the introduction of elements that reduce their overall mutual solubility in the solid solution, an excess phase should appear in the quenched state (alloy of type 4—4).

Elements that expand the stability range of the $\alpha$-phase and only slightly exceed iron in atomic size (Cr, V, Si, etc.) cause a tendency toward an increase in the amount of the second phase, which corresponds to displacement of the alloy into the $\alpha+\Theta$ region, while at the same time shifting the $\gamma$-$\alpha$ transformation into the range of higher temperatures. The same elements with substantially larger atomic sizes (Ti, Nb, Ta) sharply increase the amount of excess phase. When the $\gamma$-phase is completely suppressed, purely ferritic alloys are obtained. By combining elements having opposite effects on the stability of the $\alpha$- and $\gamma$-phases, it is possible to vary the phase composition of the alloys.

The gamma-alpha transformation affects the hardenability of the alloys. During this transformation, structural stresses arise, promoting decomposition of the solid solution. The difference in the solubility of the elements in the $\alpha$- and $\gamma$-phases must also be taken into account. In addition, shifting the $\gamma$-$\alpha$ transformation into the region of higher temperatures, where atomic diffusion mobility is greater, stimulates premature decomposition of the solid solution, whereas lowering this temperature helps preserve the supersaturated state of the solid solution. The hardenability of the alloys also depends on the casting technology [4].

As studies have shown, the more complete the quenching, the lower the initial hardness and the higher the degree of dispersion hardening, the greater the bending strength and the better the cutting properties.

The microstructure of the alloys in the initial state reflects primary dendritic crystallization (Fig. 2). The grain boundaries are tortuous; in some alloys, regions of finely branched dendrites appear (Fig. 2, a). The excess phase is located in the form of continuous grain-boundary precipitates or point-like accumulations along grain boundaries, especially in etched areas of the structure (Fig. 2, b). In a number of cases the pattern is masked by decomposition of the solid solution.

The relationship of microstructure to properties is as follows: modification, which leads to the formation of a finer-grained structure with a compact form of primary crystals [3], most effectively improves the mechanical and cutting properties. The presence in the alloy of an excess continuous boundary phase lowers the bending strength and sharply impairs the cutting properties. Large chips of the tool cutting edge arise. Branching of the crystals worsens the properties and causes their nonuniformity [4].

When fine-branching regions appear at the junctions of dendrites (Fig. 2), a considerable nonuniformity of properties arises (especially readily detected in microhardness studies), both in individual regions and over the cross section of the casting. The mechanical and cutting properties deteriorate in the direction toward the central part of the casting.

The regions of dendrite branching are sites of concentration of particles of the excess phase and of all possible defects, in particular shrinkage and gas porosity. Another factor contributing to deterioration of properties in these regions is the tendency toward premature decomposition of the solid solution in the thin dendrite branches. Even when good hardening of the continuous dendrite regions is achieved at the surface of the casting, partial premature decomposition of the solid solution occurs in the branched regions. Under these conditions, even a very slight change in the hardenability of the alloy over the cross section of the casting leads to the microinhomogeneity of properties being expressed more sharply in its inner part, and with it the difference in properties in different regions of the casting cross section increases. This nonuniformity of properties is intensified after tempering of the alloys.

In alloys with a branched dendrite structure (Fig. 2), releasing the vacuum during pouring of the alloys reduces the nonuniformity of properties over the cross section of the casting.

Fig. 2. Microstructure of alloys in the initial as-cast state, ×86:
a — 20% Co, 10% W, 10% Mo, 0.1% Ti;
b — 20% Co, 10% W, 10% Mo, 0.5 Nb.

This may be explained both by a decrease in porosity as a result of increased pressure during crystallization and, possibly, by improved through-hardening of the alloy owing to a somewhat higher cooling rate of the casting in air compared with cooling in vacuum.

It follows from this that it is expedient to use casting methods involving the application of pressure to the crystallizing metal, as well as build-up technology, which has obvious economic advantages, since improvement in the properties of the alloys studied is more easily achieved in the thin layers of the casting surface.

In cases where the microstructure indicates premature decomposition of the solid solution during hardening, the mechanical and cutting properties deteriorate (the hardness after hardening increases, but the degree of dispersion strengthening decreases). The reduced hardenability has the most unfavorable effect on alloys with a nonuniform structure, especially when an excess phase is present in the structure.

Tempering in alloys with a favorable structure reduces the bending strength by 25–30%, and by considerably more in alloys with a nonuniform structure.

REFERENCES

1. A. D. Klement'ev. Investigation of cast dispersion-hardening cutting alloys. Izvestiya vuzov. MVO SSSR, Ferrous Metallurgy, No. 3, Publishing House of the Siberian Metallurgical Institute, 1959.

2. A. D. Klement'ev. Cast dispersion-hardening cutting alloys. Modern designs of cutting tools. Proceedings of the Conference of Toolmakers of Western Siberia, collection 1, Moscow, TsINTIAM, 1962.

3. A. D. Klement'ev. Modification of iron-cobalt-tungsten alloys. Izvestiya TPI, vol. 157, Tomsk, 1970.

4. A. D. Klement'ev. Investigation of the nonuniformity of properties in the cross section of castings made of iron-cobalt-tungsten-molybdenum alloys. Izvestiya TPI, vol. 173, Tomsk, 1970.

5. W. Köster und W. Tonn. Das System Eisen-Kobalt-Wolfram. Archiv für das Eisenhüttenwesen, No. 8, 1932.

6. A. D. Klement'ev. On the solubility of elements in dispersion-hardening alloys. Izvestiya TPI, vol. 197, Tomsk, 1963.

7. A. D. Shvets. The role of structure in matters of alloy design. Structure of Metals and Alloys (from English), collection, Moscow, Metallurgizdat, 1957.