US2488568A - Production of aluminum-silicon alloys - Google Patents

Description

Nov. 22, 1949 Filed May 20, 1948 ELECTRODE CONSUMPTION POWER CONSUMPTTON KW. -HR./LB- ALLOY LB-/ TON ALLOY 2 Shets-Shee-t 1 CARBON CONTENT OF CHARGE, PER CENT OF THEORETICAL REQUIREMENT 4 E 35-145 VOLTS I lO, O0O--II,I5O AMPERES 40o CARBON IN CHARGE,70%THEG1 35o aoo 0.12 0.13 0J4 0.:5 me

ELEOTRODE PERIPHERY RESISTANCE, om" INCH Nov. 22, 1949 Filed May 20, 1948 ELECTRODE CONSUMPTION, POWER CONSUMPTION, LBJTON ALLOY KW. HRJ LB. ALLOY M. M. STRIPLIN, JR, ETAL PRODUCTION OF ALUMINUM-SILICON ALLOYS 2 Sheets-Sheet 2 5 E 45-62 VOLTS I 8,lOO-l0,900 AMPERES 50o CARBON INVCHARGE, 10% THEOR.

ELECTRODE PERIPHERY RESISTANCE, OHM--INGH FIG.3

INVENTOR 7i g WWW Patented Nov. 22, 1949 UNITE D STATES PRODUCT-IGN F ALUMINUM-SILIQON.

ALLO

MarcusM; Striplin, J12, Florence, Ala., andWilliam- M. Kelly, Niagara Falls, N. Y;, assignors to Tennessee Valley Authority, a corporation:

of the United States Application May 20, 1948, Serial No. 28,266

(Granted. under the act ofv March 3, 1883,; asamended April 30, 1928; 370 O. G. 757) 12 Claims.

The invention hereindescribed-maybe manufactured and. used by on for-.theGoyernment for governmental purposes withoutv the. payment to usof. any royalty therefor.

This inventionrelates; to ,improvedmethods for the manufacture of aluminum-silicon alloys by reducing oxides or ores of aluminum and silicon with carbon inelectric furnaces. It relates. particularly to. means for reducingpower and electrodeconsumption.intheprocess..

It. has -long. been known. that aluminum-silicon alloys m y beproducedby heating. clayormixtures of clay, with such materials as bauxite or silica pebble orsand,With-.sufiicient-carbon as reducing agent, preferably in.theformof coke or charcoal, to reduce the.charg e to an alloy of aluminum, silicon. and the other metallic constituents which may be present, usually,-. iron. or titanium. Oxides of aluminum. and silicon from whatever source obtained. might be usedin. this process,

but the: cheapness. and widelayailability of clay makes clay themorev common startingmaterial to beused.

The processis usually carried .outin anelectric furnace because oithe. high temperaturerequired for. reduction, which is. of the. order of about 2000? C. Heat-is produced by the resistance of'thechargeto the fiowof currentand by-arcing. When the required. temperature: is; reached, reduction begins .and-the reduced metals collect. in 2. 13001 in .thebottomof. the.f-urnace from whence the resulting alloyris tapped out. periodically.

Aluminum-silicon. alloys are. useful as. reducing agents in thedeoxidation of steeL. They have alsobeen used asasourceofaluminum and siliconin the manufactureof Silumin, an. alumimum-silicon casting-alloy.containing.about 12 per cent silicon. They have been; found to :have. advantages over ferrosiliconeas a reducing agent in. theproduction of magnesium fromdolomite. In addition to thesea-pplications.thereisa possibility. of obtaining. relatively. pure aluminum. and silicon from. thealloys. A1, method of the. latter type is shown in the U. S. Patent 2,469,418.

Production of, aluminum-silicon alloys is in many respects similar. to production. of. ferro: alloysrsuch as ferrosilicon. The. same types of furnaces are used and. most of the. operating conditionsare similar. Onerof. the diiierencesis in the amount of. electrodecarbon. consumedin the. reduction. For.- example, in. ferrosilicon manufacture, carbon electrode. consumption ranges from 110. to .130. poundszper ton. of prod,- uct when a charge contains. approximately] the amou r arb s chiomet i a equ red amount of carbon monoxide. Iri--.contrast, op.- eration of an aluminum silicon a1loy,.proc ess. in an electric furnace... with. a. charge containing coke in corresponding proportions will result in acarbon. electrode consumption o f more than 500 pounds per ton of alloy. And, inaddition, conditions. will. developin the furnace. which cause ashut-down in .rel atively short;time. This high electrode. consumption and. diiiiculty in getting continuous. furnace. operation I have been the main disadvantagesimthe prior artv processes for the manufacture. of aluminum silicon alloys.

It is common practice in the. production of ferroalloysto useclo'se to or, in'some casejs, s light.- ly more than theamountof. carbon theoretically necessary to reduce all the metaloxidespresent, as is shown in.Electrochemistry,. by.. A Koehler, John Wiley and Sons, New York (1235;). page 414. This practice has also been followed in the manufacture of aluminum:siliconalloys, as is shown by British Patent No. 302,692, December 20, 1927. Attempts have been made-A to-reduceelectrode consumption byincreasing the;proportion ot carbon in thecharge since itis logical to assume;;that oxidation of the carbon on. the electrodes-results from'attack byltheoxidesin the charge. and that this can be reducedbythe provision-oi additional carbon in the charge. In: the-somewhat; similar process of reducing; phosphate-rock to form elemental phosphorus in electric furnaces; such procedure has been; very, efiective as is shownby R. H. Newton in Chemical-and;MetallurgicalEngineering, 45; page 536+540, (l9.38;)-.- Increase in the carbon content of the charge from 100-per cent. to 103 per; cent and; 106 pen cent has. decreased the electrodeconsumptiom from 34. to. 29

and 11. poundsper ton.of.1?2O5;charged respeceduc he ox des: n. the char e. toj lie r' lements "with the formation of an equivalent tively. However, whenincrease-of. carboncontent in the chargein the preparationofr an aluminumsilicon alloy is attempted. no decrease inelectrode consumption followe It is an object of .this inyention to provide; an improved process-for. theproduction of analumihum-silicon alloy in an electric; furn ancl,-

Another object is I to provide a. novel: and; beneficial-ratio of carbon; o} Qxidfis n; thecharge used in such a. process,

Another object isto provideladvantageouselec trical operating conditions-adapted togbeused in connection with such. novel and-advantageous ratio of carbon to, oxide. to secure. lowered. electrode consumption-andlower power consumption per ton of alloy produced;

Still another object isle provide animproyed process forthe roductionof an aluminum -silie con alloy whichiis 'charaeter zedfby.'lgwerecl' electrode consumption;

Other objects and advantages will become apparent as this disclosure proceeds.

We have now found that electrode consumption per ton of alloy produced can be materially decreased and furnace operation improved in the aluminum-silicon alloy process by restricting the amount of carbon used as reducing agent to an amount not greater than 85 per cent of the stoichiometrical amount required to reduce the total oxides in the charge-to their elements and form an equivalent amount of carbon monoxide. It is advantageous in our process to use from 67 to 85 per cent of the amount of carbon theoretically required for reducing the oxides present and for most types of charge there is a particularly advantageous range which is from 70 to 75 per cent of such theoretrical amount. The electrode consumption per ton of product drops ofi sharply with decrease in the proportion of reducing agent from about 85 per cent until a minimum is reached, in the neighborhood of 70 to '75 per cent and steady state furnace operation becomes possible. If the ratio is decreased further, the trend is reversed and there is some increase in electrode consumption.

We have also found that there are certain electrical conditions which may be used in combination with my novel ratio of carbon to oxides in the charge to give maximum decrease in electrode consumption and to give a substantial decrease in power required per ton of alloy produced. These electrical conditions are not effective unless used in combination with our novel ratio of carbon to oxides in the charge.

These electrical conditions may be conveniently expressed in terms of .a factor R, which is defined as follows:

where E equals the potential drop from electrode to hearth measured in volts, I equals current :per electrode in amperes and D equals electrode diameter in inches. This factor is seen to have the dimensions of ohm-inches and is, in effect, the resistance per inch of electrode circumference if the path of the circuit from the electrode to the hearth is considered as consisting of parallel circuits, the number depending on the electrode ciroumference. This factor R is hereinafter in this disclosure and claims referred to as electrode periphery resistance R and is measured in ohminches.

We have found that there is a certain critical range for R which, when combined with use of the proper proportions of carbon, gives a relatively low consumption of electrode. Increase or decrease of R from the critical range even with the optimum amount of reducing agent gives some increase in electrode consumption. On the other hand, use of the proper value of R is of no advantage if the proportion of reducing agent, i. e. carbon, in the charge is too high. Steady state operation of the furnace becomes impossible and electrode consumption is high at any value or R. In the attached drawing, Figure 1 is a graph in which the lower curve A shows the decrease in electrode consumption in relationship to decrease in carbon content of charge. The upper curve B in the same figure shows decrease in power consumption per pound of alloy with decrease of carbon content of charge. Figure 2 is a graph in which curve A shows decrease in electrode consumption with increase in electrode periphery resistance R in ohm-inches when the carbon in the charge is maintained constant at about per cent of the theoretical. Curve B in this graph shows decrease in power consumption under the same conditions. These curves were obtained by the use of a charge composed essentially of clay in admixture with coke and were obtained. by the use of a graphite electrode. Curves of similar shape but differing in the point of minimum were obtained with charges of different composition and with the electrodes of other forms of carbon. It will be seen that greatest eiiiciency is shown when R is between the values of 0.14 and 0.16 when graphite electrodes are used. When electrodes formed from amorphous carbon are used, this figure for R is changed and the greatest efficiency is obtained when the values of R lie between 0.30 and 0.45, with 0.33 being the usual minimum. This is shown by Figure 3, a graph similar to Figure 2, but obtained by the use of electrodes formed from amorphous carbon. Partially graphitized electrodes give curves having optimum values between those shown in Figure 2 and Figure 3. Thus it will be seen that the optimum value for R varies with the type of carbon used in the electrodes, but will lie in the range from 0.141to 0.45.

Since, because of its cheapness, clay is the favored material for producing such alloys, these graphs are confined to the use of clay. It is possible, however, to increase the proportion of aluminum by the addition of a high-alumina raw material such as bauxite to the furnace charge or to increase the proportion of silica by the addition of silica rock. Such variations in the charge require certain variations in the carbon content in order to give the greatest efficiency and the greatest decrease in electrode consumption. Such charges, however, give curves very similar in shape to those shown in the attached graph, although the minimum shown in relationship to the carbon content of the charge may vary considerably from that shown. For instance, when a charge of clay and silica rock in such proportions as to produce an alloy containing approximately per cent silicon is used, we have found that the minimum electrode consumption was obtained when the charge contained 81 per cent of the coke theoretically required for reduction of all oxides present. Since there are a great number of such variations in the charge possible, each having a slightly different point of minimum electrode consumption, it is impractical to show all such possible variations. However, these may be easily determined by methods well-known to the art by selecting ratios between 67 and 85 per cent of the carbon theoretically required and plotting a graph similar to that illustrated.

The ratio of carbon to total oxides must be not greater than the minimum ratio at which the resulting alloy is substantially free from infusible material and must exceed the maximum quantity at which a substantial proportion of slag is formed.

The optimum proportion of carbon present in the charge as reducing agent is not affected by the type of carbon electrodes used. Thus when optimum results are given by a clay charge containing from 70 to 7 5 per cent of the carbon theoretically required for reducing the total metal oxides present when a graphite electrode is used, no change in optimum proportions occurs when electrodes formed from amorphous carbon are substituted for graphite ones.

The electrode periphery resistance R may be varied by raising or lowering the :position of electrodes in the furnace, by changing the sizeor diameter of electrodes or by varying the voltage applied to the electrodes. All these methods are well-known in the art and will not be described fully. No particular type of electric furnace is necessary for carrying out our novel process. Al-

most any arc-resistance electric furnace wellknown in the art may be used. .No particular ma- :ohinery or equipment for preparing the charge is a part of our invention, which is confined to a process for preparing aluminumesilicon alloys. .Detailsand advantages :of our process willlbe- "come apparent from consideration of'the following-.-exam'ples. .Example I illustrates the disadvantages which follow from the use of the methods of the prior art and Examples II and III set forth actual operating data, accumulated in the use of our novel process.

EXAMPLE I Mixtures of clay and coke in proportions sufficient to sup-ply carbon somewhat in excess of the quantity theoretically required for reducing the oxides present were prepared. These mixtures were charged into a small, laboratory-type electric furnace and were heated to approximately 2000 C. Contrary to expectations, electrode consumption was very high under these conditions and other unfavorable results were obtained. Tapping was difficult and the alloy production rate was low, with a consequent increase in the amount of electrical energy required per pound of alloy produced. Other mixtures were then prepared in which the coke content of the charge was reduced to about 90 per cent of the theoretical carbon requirement. These were charged into the same small furnace and smelted by heating to approximately 2000 C. Steady state operation could not be obtained.

EXAMPLE II through a zircon block embedded inthe furnace lining at'approximately the level of'the hearth. Automatic means were used to control thedepth 'of the-electrode in the furnace in order to maintain uniform current. In a part of the work,

agraphite electrode 12 inches in diameter and '60 inches long was used. In further work,-an

electrode, .formed from amorphous carbon, 17

inches in diameter and 60 inches long was used. water cooled electrode holders were used.

The clay used in the tests was calcined in a rotary kiln before charging to the electric furnace. The calcined' clay ranged in size upto'Z- inch lumps and had'the following average chemical composition:

A1 03 sio, F11 TiOQ 333? Per Per Per Per Per cent cent cent cent cent electrodes. are shown as follows:

Both byproductcoke and beehive coke in various sizes were used as the reducing agent. The vquantity of carbon theoretically required to re- .duce thefu'rnace charge was calculated as the iamountneces'sary to reduce the oxides ofalumivnum,silicon,iron and titanium in the .clay and .in the cokeash in accordance withthe equations:

Coke was added to the furnace charge in quantities sufficient to provide some'73.5 per cent of thetotalcarbon theoretically required. The

calculatedamountof coke was mixed with the calcined clay-and the mixture added periodically to the furnace around the electrode. The metal "was tapped from the furnace atregular inter- -vals, usuallyabout two to two and one-half hours "while the furnace-was maintained at a temperature of-the-orderof-2000 C. The electrode peripheryresistanceR was varied and the results shownin the graph inFigure 2 were obtained. The'minimum. consumption of graphite electrodes wasobtained when R was 0.15 ohm-inch. These tests were carried out with 12-inch graphite electrodes.

Other tests were made with 17-inch amorphous carbon electrodes since this type of electrode is used in most commercial ferroalloy operations because of'lower cost. Because of the lower ourrent-carrying capacity of carbon electrodes, however, the optimum value of the electrode periphery resistance R was considerably different from that which had been determined when operating with graphite electrodes. The results of a series of four tests in which the resistance was varied while the carbon content of the charge was held constant showed that the best results are obtainable when the electrode periphery resistance R is maintained in the range from 0.30

to 0.36 ohm-inch, with an optimum at 0.33 ohminch.

EXAMPLE III The same furnace described in Example II was operated for a period of nine days using graphite The operating conditions and results Operating conditions Electrode periphery'resistance,

{ 16B ohm-in 0.15

Operating results *Charge consumption:

Clay, lb. per hr 207 Coke, lb. per hr 60 Clay, lb. per ton of alloy 4954 Coke, lb. per ton of alloy 1444 Coke carbon, lb. per ton of alloy 1364 Alloy production rate, lb. per hr 83 Power consumption, kw.-hr. per lb. of

alloy 5.7 Electrode consumption, lb. per hr 11.4 Electrode consumption, lb. per ton of alloy 274 Total carbon consumed, per cent of theoretical 88 Composition of alloy:

Al, per cent 51.8

Si, per cent 38.7

Fe, per cent s s 4.9

Ti, per cent 4.6 Alloy recovery, lb. per ton of clay 806 Alloy recovery, per cent "a 81 Aluminum recovery, per cent 89 Silicon recovery, per cent 65 (Values involving the weight of the alloy were calculated from the total weight of metal tapped minus the weight of tapping bars consumed. alues involving power consumption include only the power consumed at the furnace electrode. Power consumed in tapping amounted to from 2 to 5 per cent of the total.)

EXAIVIPLE IV The same furnace used in Example III was equipped with electrodes formed from amorphous carbon and was operated for a 5-day period. The operating conditions and results were as follows:

Operating conditions Length of period, hr 120 Total time of operation, hr 118 Operating time, per cent 98.3 Charge proportions:

Clay, lb 1,000 Coke, lb 325 Coke carbon, lb 271 Carbon in charge, per cent of theoretical" 71.3 Secondary voltage 65 Electrode voltage 61 Current, amp s 9,980 Electrode current density, amp. per sq. in- 44 Average k. v. a 649 Average kw 612 Power factor 0.94 Electrode voltage-current ratio (resistance) 0.0061 Electrode periphery resistance,

E 1rD ohm-inch 0.33

Operating results Charge consumption:

Clay, lb. per hr 258 Coke, lb. per hr 84 Clay, lb. per ton of alloy 5,333 Coke, lb. per ton of alloy 1,733 Coke carbon, lb. per ton of alloy 1,447 Alloy production rate, lb. per hr 97 Power consumption, kw.-hr. per 1b. of

alloy 6.3 Electrode consumption, lb. per hr s 13.8 Electrode consumption, lb. per ton of alloy 285 Total carbon consumed, per cent of theoretical 85.7

Composition of alloy Al, per cent s- 46.0 Si, per cent 43.8 Fe, per cent 4.3 Ti, per cent 5.9 Alloy recovery, lb. per ton of clay. 750 Alloy recovery, per cent 75 Aluminum recovery, per cent 71 Silicon recovery, per cent 67 (Values involving the weight of alloy were calculated from the total weight of the metal tapped minus the weight of the tapping bars consumed. Values involving power consumption include total power consumed at the furnace electrode and at the tapping electrode. Power consumed at the tapping electrode amounted to about 2 per cent of the total.)

By use of our novel ratio of carbon to oxides present in the charge, we have obtained lower electrode consumption than has been known in the art of aluminum-si1icon alloy production. Use of our novel and advantageous range of values for the electrode periphery resistance R in combination with our novel ratio of carbon to oxides gives a minimum electrode consumption, lowers the power required per ton of alloy pro duced and gives steady state operation of the furnace. While we do not know the exact causes of these advantageous results, it seems probable that when amounts of carbon approaching the stoichiometric proportion are present in the furnace charge there is a tendency toward the formation of difiicultly fusible carbides of silicon and aluminum and that the accumulation of unfused material in the bottom of the furnace causes the electrodes to rise higher and higher as operation continues, with the result that the hot portion of the electrode is forced out of the charge and exposed to air oxidation, and eventually a furnace shut-down to clean out the unfused material becomes necessary. The use of lo ver resistance to bring the electrode back down is not effective since it does not prevent the accumulation of the unfused material. We have found that the formation of this material does not occur when there is a deficiency of coke prescut. The limit to which the carbon content can be reduced is the point at which the formation of a slag of unreduced alumina occurs.

Having described our invention and explained its operation, we claim:

1. A process for the production of an alumihum-silicon alloy which comprises preparin a charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total metal oxides present in the charge is in the range from 67 to per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one carbon electrode; main taining the carbon electrode periphery resistance R in the range from 0.14 to 0.45 ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

2. A process for the production of an aluminum-silicon alloy which comprises preparing a, charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from '70 to 75 per cent of the carbon theoretically required for reducing said total oxides; smelting the preparedchargeinan' electricfurnace provided with at least one carbon. electrode; main tainingi'the carbon electrode periphery resistance: R in the range from 0.14-to- '0.45-ohm inches during such smelting; and withdrawing aluminum-silicon alloy from th'e'furnace.

3. A. process for. theiproduction -of amaluminum-silicon alloy which comprises preparing a charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 67 to 85 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided With at least one graphite electrode; main taining the carbon electrode periphery resistance R in the range from 0.14 to 0.16 ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

4. A process for the production of an aluminum-silicon alloy which comprises preparing a charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 70 to '75 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one graphite electrode; maintaining the carbon electrode periphery resistance R in the range from 0.14 to 0.16 ohm-inches durin such smelting; and withdrawing aluminum-silicon alloy from the furnace.

5. A process for the production of an aluminum-silicon alloy which comprises preparing a charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 67 to 85 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one amorphous carbon electrode; maintaining the carbon electrode periphery resistance R in the range from 0.30 to 0.45 ohm-inches during such smelting; and withdrawing aluminumsilicon alloy from the furnace.

6. A process for the production of an a1umi num-silicon alloy which comprises preparing a charge containing oxides of aluminum and silicon in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 70 to '75 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one amorphous carbon electrode; maintaining the carbon electrode periphery resistance R in the range from 0.30 to 0.45 ohm-inches during such smelting; and withdrawing aluminumsilicon alloy from the furnace.

7. A process for the production of an aluminum-silicon alloy which comprises preparing a charge containing clay in admixture with carbon as reducing agent; controlling the quantity of riphery resistance R in the rang'e froin 0il4 to 0.16

ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

8. A process for the production of an aluminum-silicon alloy which comprises preparing a charge containing clay in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 70 to 75 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one graphite electrode; maintaining the carbon electrode periphery resistance R in the range from 0.14 to 0.16 ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

9. A process for the production of an aluminum-si1icon alloy which comprises preparing a charge containing clay in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 67 to per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one amorphous carbon electrode; maintaining the carbon electrode periphery resistance R in the range from 0.30 to 0.45 ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

10. A process for the production of an aluminum-silicon alloy which comprises preparing a charge containing clay in admixture with carbon as reducing agent; controlling the quantity of such carbon admixed with other materials in preparing said charge so that the ratio of such carbon to total oxides present in the charge is in the range from 70 to 75 per cent of the carbon theoretically required for reducing said total oxides; smelting the prepared charge in an electric furnace provided with at least one amorphous carbon electrode; maintaining the carbon electrode periphery resistance R in the range from 0.30 to 0.45 ohm-inches during such smelting; and withdrawing aluminum-silicon alloy from the furnace.

11. In the production of an aluminum-silicon alloy where a charge comprising oxides of aluminum and silicon in admixture with carbon as reducing agent is introduced into an electric furnace provided with at least one carbon electrode and is smelted therein, that improvement which comprises reducing consumption of car-bon electrode by controlling the quantity of carbon admixed with said oxides so that the ratio of carbon to total oxides contained in the resulting charge is in the range from 67 to 85 per cent of gthe ratio theoretically required for reducing said 1 1 12 trolled in the range from 70 to 75 per cent of the UNITED STATES PATENTS amount theoretically required for reducing said. total Oxides- 1 2 3 62 mr Aug lszs MABCUS M. STRIBLIN, JR. Y WILLIAM M; KELLY. 5 FOREIGN PATENTS Number Country Date REFERENCES CITED 302,692 Great Britain Dec. 20, 192':

The following references are of record in the file of this patent:

Images