RECONSTRUCTION OF SEMI-INDUSTRIAL VACUUM FURNACES FOR THE STUDY OF CENTRIFUGAL CASTING IN VACUUM

A. I. Slyusman

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

Remelting metals and alloys in vacuum is at present one of the most advanced and reliable means of improving their quality. Of great theoretical and practical interest at the present time is centrifugal casting in vacuum, which is still almost unused. This method of producing castings makes it possible to combine the advantages of vacuum remelting—obtaining a metal of strictly specified composition, clean with respect to gases and nonmetallic inclusions, with high mechanical and physical properties, etc.—with the advantages of centrifugal casting, namely obtaining a dense casting with a good structure and a clean surface [1]. Of particular interest is the study of the advisability of applying this method to the production of cast cutting tools from high-speed steels, cast parts from die steels and ball-bearing steels, and from alloys with special physical properties, etc.

Experimental heats can be carried out in a vacuum induction furnace of small capacity (3–7 kg). The use of large-capacity furnaces for preliminary heats leads to an irrational expenditure of a large quantity of costly scarce metals (Ni, Co, W, Ti, etc.). Melting small quantities of metal in such furnaces is also irrational, since the melting time is then greatly increased, which adversely affects the quality of the metal being melted. It is therefore expedient to modify existing vacuum furnaces for melting and casting small quantities of metal.

To study centrifugal casting in vacuum, we used a semi-industrial induction vacuum furnace of the IV-52 type with a crucible capacity of 60 kg. The furnace was reconstructed: a new inductor was calculated and wound (the calculation was made for melting a charge weighing 6 kg), and a centrifugal mechanism for casting metal into a rotating mold was manufactured and installed in the furnace.

The calculation of the inductor consisted in determining its diameter, height, and number of turns. These quantities were calculated on the assumption that, for normal operation of the installation, the inductance of the new inductor must be close to the inductance of the factory-made inductor.

When calculating inductors with crucibles of capacity up to 1 liter, the diameter and height of the inductor are determined by the formulas:

\[ d_{\mathrm{t}}=\sqrt[3]{\frac{4G}{\pi\gamma}\left(\frac{d_{\mathrm{t}}}{h_{\mathrm{t}}}\right)};\quad d_{\mathrm{i}}=d_{\mathrm{t}}+2\delta;\quad h_{\mathrm{i}}=h_{\mathrm{t}}\cdot(1.1\div1.3), \]

where \(d_{\mathrm{i}}\) is the inductor diameter,
\(d_{\mathrm{t}}\) is the crucible diameter,
\(\delta\) is the crucible wall thickness,
\(G\) is the charge weight,
\(\gamma\) is the metal density,
\(h_{\mathrm{t}}\) is the crucible height,
\(h_{\mathrm{i}}\) is the inductor height.

We take \(\delta\) to be 4–5 cm; we take \(\dfrac{d_{\mathrm{t}}}{h_{\mathrm{t}}}\) to be 0.5.

The number of turns \(\omega\) is determined from the equality of the inductances of the inductor being made and the factory inductor: \(L=L_{\mathrm{fac}}\).

\[ L=\frac{\mu\pi\left(d_{\mathrm{i}}^{2}-d_{\mathrm{t}}^{2}\right)\omega}{4h_{\mathrm{i}}}; \quad \text{hence}\quad \omega=\omega'\sqrt{\frac{\left(D_{\mathrm{i}}^{2}-D_{\mathrm{t}}^{2}\right)h_{\mathrm{i}}} {\left(d_{\mathrm{i}}^{2}-d_{\mathrm{t}}^{2}\right)H_{\mathrm{i}}}}, \]

where \(\omega'\) is the number of turns of the factory inductor,
\(D_{\mathrm{i}}\) is the diameter of the factory inductor,
\(H_{\mathrm{i}}\) is the height of the factory inductor,
\(D_{\mathrm{t}}\) is the diameter of the factory crucible [2].

According to the calculation results, an inductor of the following dimensions was made: \(h_{\mathrm{i}}=180\) mm, \(d_{\mathrm{i}}=340\) mm, \(\omega=18\) turns.

The inductor was made of copper tubing 16 mm in diameter with a wall thickness of 1 mm. To prevent short-circuiting between them, the turns were coated with heat-resistant varnish and mica-glass fabric. To protect the insulation and to prevent the crucible material from spilling between the turns during ramming, the inductor was coated on the inside with refractory clay.

Chromomagnesite, magnesite, and corundum were used as the material for ramming the crucible. In the study of high-speed steel, the most expedient material for making crucibles was found to be electrocorundum, since such crucibles showed maximum resistance. The reaction of the liquid metal with the crucible material during holding for up to 20 minutes and under the vacuum that we used does not develop significantly; therefore, contamination of the metal by reaction products practically does not occur.

The crucible was rammed and sintered around a turned graphite template. The composition of the ramming mixture was:

Boric acid was added to improve the sinterability of the mixture. To facilitate ramming, the mixture was slightly moistened. The crucible collar was formed by hand from the same mixture with the addition of 30% refractory clay. The template was slowly heated with the aid of the inductor. After several hours the temperature of the template became close to the melting temperature of aluminum oxide (about \(2000^\circ\)); it was then carefully removed, and the sintered crucible, in order to reduce the cooling rate, was covered with a sheet of asbestos.

The furnace design permits the inductor with the crucible to be tilted during pouring only by 67°. Therefore, for the liquid metal to be poured completely out of the crucible, its inner walls must form an angle of not less than 23° with the vertical axis. With the factory inductor, designed to operate with a crucible of 60 kg capacity, this can be achieved by making the pattern conical. With the new inductor, such a method would have caused a dangerous reduction in the thickness of the upper part of the crucible walls. It was therefore necessary to arrange the inductor not vertically, but at an angle of approximately 20° to the vertical axis. This somewhat complicated ramming of the crucible, but ensured complete pouring of the liquid metal during casting.

Centrifugal casting was carried out by means of a special mechanism located in the neck of the furnace. The mechanism consists of a 300-watt direct-current motor, a shaft on which a faceplate for mounting the mold is fastened, and a frame. Since the mechanism operates in vertical and inclined positions, the shaft is mounted on thrust and radial bearings. The housing in which the shaft bearings are located, and the motor, are rigidly fastened to the frame, and the entire mechanism is secured in the neck of the furnace by means of expansion screws located in the frame (in Fig. 1 only the axes of these screws are shown). Two holes are drilled in the neck housing, into which contacts are inserted for

Fig. 1.

supply of direct current to the motor. The tightness of the contacts is ensured by means of a special seal. Current is supplied to the motor from a Umformer through a rheostat, by means of which the number of revolutions is regulated, and an ammeter, by which the operation of the motor is monitored, since during pouring direct observation of the operation of the mechanism is impossible. The rotational speed of the mold was regulated by means of a rheostat from 150 to 900 rpm.

The molds are fastened to the faceplate with M8 bolts. Figure 2 shows one of the molds used for studying centrifugal casting in vacuum of high-speed steel. The mold is made of low-carbon steel. A recess is machined in its bottom, into which a ceramic insert was placed. This was done to prevent the casting from burning onto the mold at the point where the metal stream enters during pouring. The inner surface of the mold was coated with a thin layer of a mixture consisting of 80% zirconium oxide and 20% refractory clay. The coating was applied by spraying the heated mold from an atomizer with the mixture suspended in water. In addition, the inner surface was coated with soot. As a result of such treatment of the mold, castings were obtained with a good surface.

Fig. 2.

Vacuum in the furnace was produced by a VN-4 mechanical pump and was of the order of \(10^{-1}\) mm Hg. The residual pressure was measured with a VT-2 thermocouple vacuum gauge. The temperature of the liquid metal was determined with an OPIR-09 optical pyrometer. To make it possible to vary the pressure in the furnace during melting and pouring, it was connected to an argon cylinder.

Thus, the installation makes it possible to study the influence of residual pressure in the furnace, leakage, temperature, the number of revolutions of the mold, and other conditions during melting and pouring in vacuum on the structure and properties of various steels and alloys.

LITERATURE

1. D. G. Lyubetskii. Production of Steel and Castings in Vacuum. Mashgiz, 1959.
2. A. V. Donskoi, G. S. Ramm, Yu. B. Vigdorovich. High-Frequency Electrothermal Installations with Tube Generators. Gosenergoizdat, 1957.