Effect of Vacuum Melting on Nonmetallic Inclusions in Ball-Bearing Steel

M. I. Slyusman

(Presented by Professor, Doctor A. N. Dobrovidov)

Bearing production imposes very stringent requirements on steel with respect to its cleanliness in nonmetallic inclusions. In ball-bearing steel, the presence of nonmetallic inclusions and carbide heterogeneity are sources of spalling and the formation of the finest cracks on the working surfaces of bearing parts. The emergence of inclusions onto the surface of the raceway of a ring or onto the surface of a ball leads to the so-called flaking of the metal [1]. In addition, the high requirements for bearing steels with respect to metal cleanliness and, in particular, for ensuring a dispersed distribution of carbides are determined by the specific operating conditions of bearing elements: high specific pressures, alternating loads, and high speeds, which cause wear of the working surfaces. Therefore, the requirements imposed on ball-bearing steel with respect to freedom from inclusions and carbide segregation are among the highest of those imposed on special steels. These requirements have made it necessary to carry out a large amount of work on the study and improvement of methods for melting ball-bearing steel.

In recent years, enormous work has been done to improve methods for producing ball-bearing steel; the technology of melting, teeming, rolling, and other elements of the process has been significantly improved.

Recently, new methods of melting steels and alloys have become increasingly widespread; among these, first of all, one should name methods based on the use of vacuum.

In vacuum melting of ball-bearing steel, it is especially important to remove nonmetallic inclusions from the metal as completely as possible and to free it from dissolved gases. The solubility of such gases as nitrogen and hydrogen in iron is determined by Sieverts’ law:

\[ S = k_1 \sqrt{P} \]

where

\(S\) — solubility of gases in the metal,
\(P\) — partial pressure of the gas in the working space of the furnace.

It is quite obvious that in a vacuum furnace the partial pressure of gases decreases, which causes the dissolved gases to be released from the metal. The gases released are pumped out of the furnace, and thus a very low content of dissolved gases can be obtained in the metal. As for oxygen, its removal from the metal occurs as a result of its interaction with carbon and the pumping out of the carbon monoxide that is formed. It has been established that the deoxidizing capacity of carbon is inversely proportional to the pressure of the gas phase. Consequently, there is no doubt that the most complete degassing of the metal is possible when vacuum acts on the liquid metal.

As for the influence of vacuum on the content of nonmetallic inclusions in steel, there exist various, and often contradictory, opinions on this question. This can be explained by the fact that, in addition to the melting method, the content of inclusions in the metal is affected by other very significant but still insufficiently studied conditions of steel melting or casting. It is quite obvious that, in addition to the residual pressure in the furnace, the oxygen content and the nonmetallic inclusions in the metal are influenced by such factors as air leakage into the vacuum system due to leaks in the vacuum joints and gas evolution from the refractory materials of the furnace. At present, crucibles in vacuum induction furnaces are lined with magnesium oxides, as well as aluminum oxides and zirconium oxides. As a result of interaction of the molten metal with the crucible lining during melting, an increase in the oxygen content in the metal is possible, as is contamination of the metal with elements entering into the composition of the lining material. From this point of view, magnesium oxide is the most expedient material for lining the crucible in a vacuum induction furnace. It is true that, when the metal interacts with a lining made of MgO, the oxygen content in the metal may increase. However, in this case contamination of the metal by the crucible materials does not occur, since the reduced magnesium, owing to the high pressure of its vapors, volatilizes from the metal. When, however, the metal interacts with a lining made of aluminum oxide or zirconium oxide, the transfer of aluminum and zirconium into the metal is observed [2].

The aim of the present work was to investigate the question of the nature of the inclusions and the degree of contamination by inclusions of ball-bearing steel melted in a vacuum induction furnace of the IV-52 type.

The starting material for melting was rod steel ShKh-15 of the following composition:

\[ \mathrm{C} = 1.03\%,\quad \mathrm{Cr} = 1.45\%,\quad \mathrm{Mn} = 0.36\%,\quad \mathrm{Si} = 0.34\%. \]

The crucibles were made of magnesium oxide. When the charge was loaded into the crucible, the required amount of electrode scrap was placed on its bottom, which to some extent ensured preliminary deoxidation of the bath with carbon. The pressure in the furnace was controlled by a VT-2 thermocouple vacuum gauge. Melts were carried out at various pressures: \(5 \cdot 10^{-2}\), \(5 \cdot 10^{-1}\), and \(10\) mm Hg. For comparison, a melt was carried out in an open furnace. The rarefaction in the furnace was produced by means of a VN-4 fore-vacuum pump. In the case of melting at the minimum pressure (\(5 \cdot 10^{-2}\) mm Hg), the entire melt, up to the end of solidification of the steel, took place with the pump operating. In all melts, the pressure in the working space of the furnace was kept constant. The temperature of the metal before tapping was determined from the readings of an OPPIR-09 optical pyrometer and varied within the limits of 1520–1540°. The liquid steel was poured into a metal mold of circular cross section with an average diameter of 60 mm and a casting taper of 2°. The height of the mold was 250 mm, including the height of the working part of the casting—170 mm;

the remaining part of the mold height was set aside for the feeder head, molded from sand on the refractory bottom (Fig. 1). The melt weight was 10 kg. The metal was poured by tilting the furnace body together with the crucible into a mold installed in a branch pipe welded to the furnace body.

Fig. 1.

The chemical composition of the experimental melts is given in Table 1.
For the above-mentioned melts, gas analysis was carried out, the results of which are summarized in Table 2.

The data in Table 2 confirm that the nitrogen and hydrogen content in the steel decreases as the partial pressure of these gases decreases. As for oxygen, this proposition is valid only up to a certain value of the residual pressure in the furnace. This can be explained by the following circumstance. As was already indicated above, the removal

Table 1

Table 2

of oxygen from the metal occurs mainly as a result of the deoxidizing capacity of carbon, which increases as the pressure in the furnace decreases. It has been established that the minimum pressure in the furnace at which the metal is practically completely deoxidized by carbon is 1–2 mm Hg [3]. Thus, lowering the pressure in the furnace below that indicated, from the point of view of completeness of deoxidation and reduction of the oxygen content in the metal, is irrational.

The nonmetallic inclusions of the heats under investigation were evaluated according to GOST 801-60 after preliminary forging to a diameter of 12–14 mm. From the billet of each heat, five specimens were taken, and the character and rating of the nonmetallic inclusions were determined by the metallographic method. The average ratings of the nonmetallic inclusions of the tested heats are given in Table 3.

Table 3

Fig. 2. Nonmetallic inclusions in ShKh-15 steel, melted in an open furnace.

Fig. 3. Nonmetallic inclusions in ShKh-15 steel, melted at a pressure of 10 mm Hg.

Fig. 4. Nonmetallic inclusions in ShKh-15 steel, melted at a pressure of \(5 \cdot 10^{-1}\) mm Hg.

Fig. 5. Nonmetallic inclusions in ShKh-15 steel, melted at a pressure of \(5 \cdot 10^{-2}\) mm Hg.

It should be noted that, as a result of melting in vacuum, not only are the quantity and size of nonmetallic inclusions in the metal reduced, but they are also distributed more uniformly. As a result, the inclusions appear as separate crystallites that do not give a high rating (Figs. 2, 3, 4, 5).

A study of the influence of various technological factors on the contamination of steel by inclusions showed that the number of oxide inclusions increases very substantially during teeming and especially toward the end of casting a heat [4]. This is explained by oxidation of the stream of metal during casting. When melting in vacuum, the metal throughout the entire melting process, up to solidification, remains isolated from the action of the atmosphere.

Conclusions

1. When vacuum acts on the liquid metal, the most complete degassing of it is possible; this, undoubtedly, contributes to improvement of the mechanical, physical, and other properties of the steel.

2. Melting in vacuum makes it possible to bring the chemical reactions between oxygen and carbon to the fullest possible completion, which permits good deoxidation of the metal.

3. The use of vacuum furnaces makes it possible to obtain ShKh-15 steel of a very high degree of purity with respect to nonmetallic inclusions, which cannot be achieved when steel is melted in open furnaces.

4. Melting in vacuum is the most reliable and radical method for obtaining high-quality ball-bearing steel.

References

1. S. S. Shteinberg. Steel for Ball and Roller Bearings. OGIZ, 1932.

2. S. G. Voinov, A. G. Shalimov. Ball-Bearing Steel. Metallurgizdat, 1962.

3. A. M. Samarin, R. A. Karasev. Reports of the Academy of Sciences of the USSR, vol. 119, No. 5, 1958.

4. M. I. Vinograd. Nonmetallic Inclusions in Ball-Bearing Steel. Metallurgizdat, 1954.