Spheroidization of the Structure of Cast High-Speed Steel

A. N. Dobrovidov and N. D. Tyuteva

Purpose of the Experiments

Ledeburitic tool steels, to which high-speed steel belongs, are characterized by the presence of a complex carbide that forms part of the eutectic.

The carbides precipitating from the melt of high-speed steel are structurally associated with austenite crystals and crystallize in a peculiar form resembling dendrite skeletons.

If the structure of the secondary carbide can be changed by heat treatment, the structure of the ledeburitic carbide has to be changed during forging or rolling of the ingot. During rolling, individual branched crystals of brittle carbide are crushed and transformed into more or less small fragments with sharp corners.

The purpose of subsequent annealing is to spheroidize such fragments, giving them the form of regular rounded grains.

This same annealing should cause phase recrystallization of the secondary carbide and reduce the hardness so that the tool made from this steel can subsequently be machined on machine tools.

The steel entering the tool shop has a structure characterized by a granular eutectoid in which grains of secondary and primary carbide are embedded.

Sometimes high-speed steel has to be forged in the tool shop. Proper forging should not change the structure of the primary carbide, but after forging a softening anneal is mandatory.

The generally accepted regimes for softening annealing of ordinary high-speed steels are given in Table 1.

Table 1

Depending on the duration of holding at high and low temperatures, the hardness obtained lies within the limits indicated in Table 1. Annealed steel with such hardness is machined sufficiently well by chip removal on machine tools.

The hardness of a cast cutting tool made of high-speed steel after annealing according to the regimes given is considerably higher than in the case of rolled or forged blanks, and machining it by cutting becomes difficult.

The greater hardness of an annealed cast tool is explained by the distinctive features of its structure.

Isothermal annealing affects only the basic metallic matrix and the structure of the secondary carbide, and has practically no effect at all on the structure of the primary ledeburitic carbide. The skeleton-like, branched crystals of this carbide permeate the entire structure of the steel and increase the hardness of the steel, irrespective of the hardness of the rest of the structure. However much we soften the basic structure of the steel, its hardness will to a considerable degree be determined by the character of the structure of the primary carbide. The influence of the primary carbide on hardness increases as the carbon content in the steel is raised.

When a high-speed-steel tool is cast in a metal mold, the dimensions of the tool are often obtained with such accuracy that no further machining of the casting by chip removal is required. It is only necessary to sharpen such a tool on an emery wheel. In this way we manufacture fluted cutters for inserted face and cylindrical milling cutters. Such cutters are not annealed and not hardened. They are hardened during cooling in the metal mold and are given only the usual threefold tempering. The pouring of metal molds is carried out on a centrifugal machine. As our experience has shown, some grades of cast steel do not even require tempering.

But when casting in a sand mold, and often when casting a complex tool in a chill mold, the casting has to be machined by cutting. In this case, of course, a softening anneal is necessary. But, as has already been mentioned, ordinary softening isothermal annealing does not give sufficiently good results. Changing the annealing regime—both by changing the temperatures and by considerably increasing the holding time (tens of hours at the high and isothermal holds)—also does not help.

In studying various compositions of steels for cast tools, we became convinced that an increased carbon content in steels always led to an increase in tool life. This was especially pronounced in the case of through roughing lathe cutters.

Table 2 gives the compositions of several high-speed steels from which fluted cutters for face milling cutters were cast on a centrifugal casting machine in metal molds. These steels served for spheroidization experiments.

Table 2

After casting, the hardness of the cutters was within \(R_c = 58\text{–}62\). The cobalt steels had the highest hardness. The cutter made of steel 9 had a hardness of \(R_c = 50\text{–}55\).

For productivity tests in milling, the cutters, as already stated, were not hardened, but only tempered three times at \(570\text{–}600^\circ\) with a holding time of one hour each. The cobalt steels were tempered at elevated temperatures. After the third tempering the hardness was \(R_c = 64\text{–}66\). Steel containing nickel acquired sufficient hardness only after six temperings.

After tempering, the cutters were inserted into the body of the face mill, and the assembled mill was sharpened on a grinding machine.

Isothermal annealing of cast cutters made from the steels listed, at different temperatures and for different durations, did not give good results. The hardness of the cast cutters remained within 320–350 Brinell units. Owing to the great stability of the metastable structures of the nickel steel, its hardness after annealing remained very high \((R_c = 55)\).

Spheroidizing Annealing

On the basis of the foregoing, it was decided to subject the cutters to spheroidizing annealing and to obtain a structure close to that of forged steel. Spheroidizing annealing of large ingots of high-speed steel does not give good results because of the coarse structure of the ledeburitic dendrites. When casting into small metallic molds, the structure of the primary carbides becomes finer as the weight of the casting decreases. But the character of their structure changes little in this process; they nevertheless retain their distinctive skeletal form. The ribs of the skeletons become ever thinner, and the ratio of the total surface of the crystal to its volume continuously increases. An increase in surface leads to an increase in the surface energy of the crystal in comparison with its total energy, and prerequisites are created for the formation, when the temperature is raised, of compact grains.

Gudtsov and Kazeev [2] showed that, by prolonged holding of cast high-speed steel at temperatures close to the hardening temperature, its structure can be very strongly changed, bringing it close to the structure of forged steel. Unfortunately, in their work they did not dwell on the cutting properties of the steel, on its serviceability after ten hours’ holding at high temperatures. However, there are indications that with very prolonged annealing the carbides change their composition, which leads to a change in the properties of high-speed steel [1].

This prompted us to conduct a series of experiments on spheroidizing the structure of cutters cast from the indicated grades of steel, with subsequent testing of their productivity on a milling machine.

The cutters were placed in a steel box, covered with shavings of high-speed steel, and hermetically sealed. Heating was carried out in a furnace with automatic temperature control. As control experiments showed, decarburization was not observed at all.

In selecting the annealing regime, we sought to provide for all steels (primarily cobalt steels) a more or less universal annealing regime, after which the steel could be easily machined by cutting on machine tools.

It should be noted that this general regime was successfully selected, but in some cases it may also be different, as is evident from the following example.

Steel from heats 5 and 8 was annealed as follows: holding at \(1200^\circ\) for about 1.5 hours; holding at \(950^\circ\) for 1.5 hours and at \(740^\circ\) also for 1.5 hours. The hardness after annealing of cutters made of steel 5 was \(R_c = 24\), and of steel 8, \(R_c = 41\).

As can be seen, the steel with a high carbide content gave an unsatisfactory, excessively high hardness. At the same time, cast steel with the usual carbon content gave a very low hardness.

Steels of heats 9 and 7 were subjected to ordinary annealing according to the following schedule: five-hour holding at 900° and twelve-hour holding at 750°. Their hardnesses were, respectively, 55 and 39. The nickel steel did not reduce at all the hardness that it had after casting.

After a number of experiments, a sufficiently universal annealing schedule was found for all steels except the nickel steel. It consisted of preliminary spheroidizing annealing and ordinary isothermal annealing. The spheroidizing annealing was carried out at 1150–1200° for 4 hours, followed by holding at 870° for 6 hours and holding at 740°, also for 6 hours.

The hardness of the cast cutters after the indicated annealing is given in Table 3.

Table 3

No clearly expressed dependence between composition and hardness after annealing can be noted from consideration of the table.

Heats of almost identical composition in all elements except cobalt (heats 4 and 5) gave different hardnesses. The steel with the lower cobalt content (heat 5) even showed a somewhat increased hardness.

On the other hand, heats 3 and 4 have the same hardness, irrespective of cobalt content.

Heats 6 and 7 showed the same, greatest hardness of all the heats, although they differ markedly from one another in the content of carbide-forming elements.

Therefore it may be concluded that the hardness of high-speed steel after annealing depends chiefly on the character of the structure of the primary carbide and only partly on the hardness of the eutectoid formed during annealing and on the structure of the secondary carbide.

Microstructure of the Steel after Annealing

All polished sections of the steels studied were etched with a hot solution of sodium picrate and photographed at a magnification of five hundred times. Fig. 1 gives the microstructure of heat 8 after annealing for one hour at 1200°, one and a half hours at 940°, and six hours at 740°. The hardness of the steel after annealing was $R_c = 36$. In the photograph, the broken network of ledeburitic carbide is clearly visible, which indicates only incipient spheroidization. The hardness is high precisely because of the retained network of ledeburitic carbide.

Figs. 2 and 3 show, respectively, the structures of heats 5 and 1 after annealing for two hours at 930° and then for eight hours at 740°. In the photographs, a network of ledeburitic carbide, slightly broken up, and a finely dispersed mass of secondary carbide are visible. The hardness of these steels was $R_c = 31$ and 37.

The same heats (5 and 1) were annealed according to a schedule that included a one-hour holding at 1040°. This was followed by annealing of the ordinary type: holding at 940° for two hours and at 750° for six hours. The hardness, respectively,

$R_c = 32$ and $38$. In the microphotographs, partial spheroidization is noticeable (Figs. 4 and 5).

And, finally, in Figs. 6, 7, 8, 9, and 10 the microstructures of the steels after the established optimum spheroidizing anneal are shown.

Fig. 1

Fig. 2

In the microphotographs in Figs. 6, 7, 8, and 9 the structures of heats 1, 2, 3, and 4 are shown, while in Fig. 10—the structure of heat 8.

Fig. 3

Fig. 4

As can be seen, upon annealing according to this regime, we have a complete picture of spheroidization of the primary carbide.

The microstructure of cast high-speed steel annealed by the developed method is almost analogous to the microstructure of forged steel after ordinary factory annealing. Only a certain difference is observed in the size of the grains of primary carbides; the coarser grain is formed during annealing of the cast steel.

Fig. 5

Fig. 6

Fig. 7

Fig. 8

In Fig. 11, at the same magnification and etching, the structure of steel RF-1 after factory annealing (“as delivered”) is shown. This structure is quite comparable with the structure in Fig. 9. In Fig. 7 is shown the significant—

...a considerably coarser structure of the primary carbide. This is the structure of heat 2, in which the carbon content is 1.25%. Such a considerable

Fig. 9                Fig. 10

carbon content had a noticeable effect on the structure. After annealing, this heat gave the lowest hardness \((R_c = 21)\).

Fig. 11

Testing of face mills with cast inserts

After spheroidizing annealing, the inserts made from the steels cited above were hardened, tempered, fitted into the body of an assembled face mill, and sharpened. The inserts were heated for hardening in a salt bath at \(1260^\circ\) for three minutes, quenched in a nitrate bath, and tempered two...

times at 560 and 570° for one hour each. The hardness of the cutters after tempering was \(R_c = 64\text{–}65\).

For comparative tests, the same cutters made of forged steel RF-1 were taken. After hardening and tempering their hardness was \(R_c = 64\text{–}65\). The cutters were tested by the method of Professor A. M. Rozenberg. Annealed steel of grade ShKh-15 with a Brinell hardness of about 187 was used for cutting. The test was carried out on a horizontal milling machine at a cutting depth of 3 mm and a feed of 0.298 mm per spindle revolution. On the basis of a series of experiments, the cutting speed selected was 26.7 m per minute.

The durability of the milling cutter was taken to be the number of passes before blunting. Blunting of the cutter was determined by wear of the flank face. When the wear reached 1.5 mm, the cutter was considered finally blunted.

The durability of the milling cutter with forged cutters was taken as 100%. The test results are given in Table 4.

Table 4

As can be seen, all the cutters tested showed increased durability compared with cutters made of RF-1 steel. The cutters made of steels 2 and 4 are particularly noteworthy. The first steel has an increased carbon content and an addition of molybdenum. Its structure after annealing is very coarse. The second steel is ordinary RF-1 with an addition of 3% cobalt. Increasing the cobalt content to 15% did not give especially good results (heat 1).

The results obtained compel a somewhat different assessment of the role of coarse-grained primary carbide in high-speed steel.

Conclusions

1. Positive results, in terms of lowering the hardness of cast high-speed steels and spheroidizing the primary carbide, were obtained by annealing according to the following schedule: holding at 1150–1200° for four hours; holding at 870° for six hours and at 740° also for six hours.

There is every reason to assert that softening spheroidizing annealing of cast high-speed steel is entirely possible, regardless of the carbon content in the steel (up to 1.25%). This makes it possible to manufacture cast tools that also require machining after casting.

1. It is possible that the indicated schedules can be changed and shortened.

2. The cutters of face mills, heat-treated according to the procedure described, showed good cutting properties.

3. It is significant that steels of very diverse composition gave very similar results in cutting tests.

Literature

1. A. P. Gulyaev—Properties and Heat Treatment of High-Speed Steel, p. 31, Mashgiz, 1939.

2. N. T. Gudtsov, S. A. Kazeev—On the Question of the Nature of High-Speed Steel, Metallurg Publishing House, 1930.