US1658879A - Manufacture of alloys - Google Patents

Description

Patented Feb. 14, 1928.

UNITED STATES VERE B. BROWNE, OF BBACKENBIDGE, PENNSYLVANIA.

MANUFACTURE OI ALLOYS.

No Drawing.

This invention relates to the manufacture of alloys, and especially to the manufacture in the electric arc furnace of ferrous-alloys, commonly called alloy steels.

In the production of suchal1oys, and especially those of the type of rustless 1ro n, stainless steel and the like which contain chromium in proportions above about 10 or 12 per cent, it is of great importance that the metal should be. free from oxids and likewise substantially free from dissolved gases. The necessary freedom from oxids is usually attained by removing the oxidizing slag under which the metal has been melted down and refined, and replacing it bya basic slag which, as a final operation, is reduced or brought to a reducing condition either by carbon or by silicon or ferrosilicon. \Vhen lime slags are employed, as is usual ractice, these reducing agents form therewith calcium carbide and calcium silicate, respectively.

Calcium carbide slags are generally employed when the carbon contentof the metal is not of primary importance, or when the carbon specification of the metal is such as to permit a fairly large absorption of carbon by the metal from the slag. When, on the other hand, carbon specifications must be maintained at a low limit, ferrosilicon is commonly employed as the reducing agent. In this way metal 'of the desired low carbon content may readily be made, but other and often serious difficulties are apt to arise.

For example in the productlon of rustless iron, where a chromium content of about 12%'-with a carbon content of 0.12% or less is required, the useof carbide slags is usually.

excluded by the carbon specification. Ferrosilicon has therefore been extensively used; but the product has been seriously affected by certain factors which are directly \a-ttributable to its action. In the first place, as the oxids of iron, manganese, chromium, etc. are reduced back into the metal there is a corresponding formation of silica, SiO,, which combines directly with most oxids yielding silicates, many of which are irreducible at the temperature of the steel-making furnace and consequently remain in the slag.

In the second place a portion of the silicon, on account of its high specific gravity, enters into and alloys with the metal during the reduction of the slag. The metal at this stage of the operation is highly saturated with gases; and silicon, even in proportions Application filed October 21, 1925. Serial No. 64,015.

as low as 0.1% or less, greatly increases the solubility of these gases in the steel at high temperatures. Consequently the gases cannot be effectively removed from the molten metal, but are given up durin solidification in the molds. In exaggerate cases this may give rise to rismg or bleeding ingots; or apparently sound ingots may be formed which nevertheless contain innumerable small blow holes, which cause the ingot to crack during rolling, resulting in heavy losses and high operating costs.

According to the present invention these difliculties are wholly overcome, and I am enabled to produce sound ingots which can be rolled with only such commercial losses as would be encountered in the case of ordinary sound steels.

In practicing my process it is essential first, that the metal to be subjected to the reducing slag should be very low in or practically'free from silicon; and second, that this secondary or reducing slag should be of such a character, and used in conjunction with such-reducing agents, that no substantial quantity of silicon can enter the bath until the deoxidation step, as hereinafter described, has been completed. I use aluminum as my preferred reducing agent, but I use it in conjunction with a substantially silicate-free slag, whereby injurious contamination of the metal bath with silicon is avoided.

My preferred procedure is as follows:

I first subject the metal bath to any of the usual oxidizing slags, continuing this treatment until preliminary tests show that the metal is substantially free from silicon, which should not in any case exceed a few hundredths of one per cent, and is preferably wholly absent. I then remove the oxidizing slag, and replace it by a protective coating of burnt lime. which should contain little or no silica. The use of fiuor spar additions, such as are commonly employed to increase the fusibility of lime slags, should be avoided, especially in case the walls or roof of the furnace contain silicates, for the reason that the attack upon such walls b the fluorides is of a character which wil lead eventually to the contamination of the bath with silicon. A small addition of manganese ore however may be and preferably is made, for a purpose which will appear hereinafter. By reason of the high melting alumina or a mixture of these, the mixture containing the aluminum powder being as intimate as possible in order to protect the aluminum from premature oxidation. The aluminum reduces the oxides present, including those derived from the metal bath; and the small amoiint of manganese oxid. if used, serves to initiate and accelerate this reaction. As the aluminum is added the protective coating becomes progressivel more and more fusible, until finally there 1s formed a highly fluid slag which exhibits a remarkable afiinity for the oxid contents andh also the sulfur contents of the metal While the nature of the chemical and physical action cannot as yet be clearly defined, it is noticeable that following each addition of the aluminum-lime mixture, up to a certain point the slag darkens rapidly, owing to absorption of oxids from the bath, such darkening becoming less and'less and finall disappearing when deoxidation is comp etc. It is apparent that this slag has a much greater affinity for the oxids than has the metal bath, with the result that the oxids, as well as the sulfur compounds, migrate rapidly into the slag. Further additions of aluminum are made until the removal of oxids from the metal is complete, a condition which manifests itself in a very characteristic manner, which permits the process to be controlled with a degree of exactitude impossible with any'other slag with which'l am familiar.

The first manifestation occurs when a metal rod which has been dipped into the slag is plunged into water. A strong evolution 0]. gas occurs, amounting at times almost to an explosion, and causes the slag to fluff up into a feather mass resembling snowflakes. This mani estation appears only toward the completion of the deoxidation; and I have found by experience that when this condition has persisted for a period of about ten minutes, it is quite safe to tap the metal, which will be found free from oxids, low'in sulfur, and of a composition to roll satisfactorily. For convenience and brevity, I will hereinafter refer to this phenomenon simply as the test condition.

I have observed that when the slag has reached the test condition above mentioned, the mass of slag in the furnace appears to be frothing, giving out large volumes of gas which burn with a bluish flame. This gas must of necessity be derived from the metal.

When the slag is dropped into water the evolved gases smell strongly of hydrogen sulfid, indicating the desulfurizing nature of the operation. This is confirmed by the analytical results. In one specific case the sulfur was reduced from .044 to 028% in the space of twenty minutes: and high grade alloys prepared by this process will run consistently with a sulfur content of .Ol to 018%. a result not attainable from the same materials by the silicon reduction method as ordinarily practiced.

As already pointed out, an essential feature of the process is to keep the content of silica in the slag as low as practicable, because the aluminum will reduce the silica to silicon, which passes to the metal bath and produces the same phenomena which occur when silicon is used as a reducing agent in the ordinary manner: in other words, aluminum, in presence of a siliceous slag, acts essentially like silicon. This fact is believed to account in large measure for the many failures which have attended the use of aluminum in the past, because it has been the custom in all such cases to add either sand or fiuor s ar to the secondary lime sla before the reduction, to improve its fusibility. In case sand is added, silicon will be reduced directly into the metal. In case fiuor spar is used, as is the normal practice in most steel melting shops, the fluorin which is evolved attacks the brick lining of the furnace with formation of silicon tetrafluorid, which is taken up by the slag, and breaks up in turn with liberation of more fiuorin, thus repeating the cycle indefinitely; so that even relativel small additions of fluor spar may eventual y result in slags which are quite high in silica. It should be understood however that all precautions to keep the second ary slag low in silica ma be futile, unless the metal to be deoxidize is very low in or substantially free from silicon; and my rocess contemplates the application of a uminum reduction, in a substantially silicatefree slag, overlying a metal bath low in silicon.

My process is applicable generally to the preparation of ferrous alloys such as carbon steels or alloy steels, including such as contain silicon in any proportion: but in this case the process of deoxidation is carried out in absence of silicon, as already'described, the silicon being introduced in any desired proportion after the completion of this ste In the process as described above it will be observed that the aluminum powder performs two quite distinct functions: it reduces the oxid components of the lime slag and thereby provides a reducing slag which deoxidizes the metal bath; and it forms with the lime a fusible aluminate slag which ra idly absorbs both oxids and sulfur compoun s from the metal. Obviously the first of these functions might be performed by other silicon-free reducing agents, including carbon required to construct a suitable aluminate slag should be supplied from the relatively expensive aluminum powder; but I may make a synthetic lime-alumina slag and add thereto only such amount of metallic aluminum as is required for deoxidizing purposes. This is in efi'ect what is accomplished when the aluminum powder is added in admixture with alumina, instead of with lime, as already mentioned.

I I claim:

1. In a process of making ferrous alloys, subjecting a metal bath low in silicon but containing oxids to the action of a basic slag substantially free from silica, and imparting a reducing character to said slag and deoxidizing the metal bath by means of aluminum.

in silicon but containing oxids to the action of a basic slag substantially free from silica, and imparting a reducing character to said slag and deoxidizing the metal bath by means of aluminum.

3. In a process of making ferrous alloys, the steps comprising deoxidizing a. metal bath low in silicon in presence of aluminum and. a basic slag substantially free from silica, and tapping the metal.

4. In a process of making ferrous alloys, the steps comprising deoxidizing a metal bath low in silicon in presence of an aluminate slag substantially free from silica, and tapping the metal. 7

5. Process of makin iron-chromium alloys, com rising deoziidizing a low-silicon al- 10y melt y means of alummum in presence" of a basic slag substantially free from silicaifz- Y 6. Process of makin iron-chromium 8.1- loys, comprising deoxidlzing a low-siliconalloy melt y means of aluminum in presence o ilf an aluminate slag substantially free from s1 me.

In testimony whereof, I aflix my signature.

-. VERE B. BROWNE-