A Rocking Electric Brass Furnace

tric furnaces give crucible quality of metal without using crucibles. However, not every type of electric furnace can be used for brass melting. If br...
1 downloads 0 Views 2MB Size
June, 1918

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

or I part by volume in 500,ooo. Acetylene a t this concentration would probably not poison platinum. The calculation made above is based on Henry's law. Since t h e acetylene is only I per cent of the ammonia, its partial pressure is 0.01atmosphere and i t follows therefore t h a t I liter of liquor will dissolve a t equilibrium one liter of acetylene a t 0.01 atmosphere, or I O cc. a t normal pressure. I n dissolving ammonia-acetylene in water, the former is absorbed quite rapidly, thereby increasing t h e partial pressure of the acetylene so t h a t the concentration of acetylene in solution is a t first quite high. The gas must be passed in some time after all absorption of ammonia has ceased in order t h a t equilibrium may be established and the acetylene concentration in the liquor reach its minimum value. Laboratory experiments show conclusively t h a t the deductions arrived a t b y application of Henry's law are correct. Ammonia gas containing I t o 2 per cent acetylene was passed into water until the ammonia reached a concentration of 28 per cent. The acetylene in solution was then determined by precipitation as AgzCz with standard silver nitrate, and found t o be 130 cc. C2Hz per liter. T h e ammonia-acetylene mixture was then continued through the solution until it passed freely and no more ammonia was being absorbed. The liquor now contained I O cc. C2H? per liter. The application of this scheme industrially should offer no difficulties. Ammonia absorption apparatus is simple. Two or more absorbers would have t o be employed since equilibrium conditions must be established by blowing t h e gas freely through the first absorber after t h e liquor is saturated. The temperature of the absorber could be adjusted so t h a t t h e strength of the liquor would not be too high after cooling, t o avoid loss of "3. This temperature in winter would probably be in the neighborhood of 35' C. and warmer in summer. The heat of solution makes such a n adjustment easy. Should i t be desirable t o make a liquor absolutely

I

free from acetylene, pure ammonia gas could be blown through the absorber a t t h e end of the operation. About 5 per cent of this very pure liquor could be reserved a s a source of pure gas t o treat the next batch. An ammonia liquor free from all non-reacting foreign gases may be prepared in this way. SUMMARY

I-As little as 0 . 0 2 per cent acetylene in the ammoniaair mixture has a distinctly deleterious effect. The yield drops from about 95 per cent t o 89 per cent or less. 11-The effect of 0.1 per cent acetylene, or its accompanying impurities, is disastrous. The yield may drop as low as 6 5 per cent. 111-A small quantity of acetylene will render t h e platinum so inactive t h a t t h e yield on pure ammonia will be reduced 2 t o 4 per cent for several hours. This means t h a t the ammonia used for manufacture of nitric acid should be free from acetylene a t all times. IV-Operation of oxidizers working on the principle of a self-sustaining reaction without electric heating or preheating, and utilizing sources of ammonia t h a t contain acetylene, is probably impracticable. V-Ammonia gas may be freed from acetylene and other non-reacting gases by dissolving i t in water t o make a strong ammonia liquor. Such procedure involves no difficulty industrially, nor any considerable expense in operating a commercial oxidizing plant. ACKNOWLEDGMENT

The experiments described herein are a part of a n extensive investigation on commercial ammonia oxidation and t h e production of nitric acid thereby, conducted by the Bureau of Mines and the Semet-Solvay Company in co6peration with the General Chemical Company and the Ordnance Department, under the . direction of the Chief Chemist, Dr. Charles L. Parsons. B U R E A U O F MINES WASHINQTON, D.C.

LABORATORY AND PLANT A ROCKING ELECTRIC BRASS FURNACE' By H. W. GILLETT AND A. E. RHOADS Received M a y 15, 1918

It seems inevitable t h a t t h e next few years will see electric furnaces largely replacing crucible furnaces in t h e brass industry, a development comparable t o t h a t which the last few years have seen in t h e steel industry. With Klingenberg clay not available and Ceylon graphite requiring shipping needed for other purposes, crucibles, despite t h e good work done on the problem by crucible manufacturers, the Bureau of Standards, and others, are still, speaking generally, of much poorer quality and many times more costly t h a n they were under pre-war conditions. T h e time is ripe for t h e practical elimination of t h e crucible from the brass industry. 1

Published by permission of the Director of the Bureau of Mines.

459

I

With t h e huge tonnage of brass required for war 200 purposes, the use of t h e small units-averaging lbs. per charge-in which crucible melting is done b y the brass rolling mills, seems, and is, a n anachronism. Besides the avoidance of crucibles and the ability t o melt larger charges, electric melting (in a suitable type of furnace) decreases the loss of metal b y oxidation and by volatilization, prevents the taking up of sulfur from the fuel, gives better and more healthful working conditions, and has many minor advantages such as freedom from handling and storing fuel and ash. Electric furnaces give crucible quality of metal without using crucibles. However, not every t y p e of electric furnace can be used for brass melting. If brass did not differ materially from steel in its behavior during melting, electric furnaces would long ago have superseded crucible

460

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y

furnaces. But brass is made up of copper and zinc, and zinc is volatile a t brass melting temperatures. For this reason, fuel-fired furnaces of the reverberatory type can be applied t o brass only a t the expense of a zinc loss so high as t o prohibit the procedure. Similarly, the direct-arc type of electric furnace used for steel melting, such as t h e HBroult, can be used only on bronzes practically free from zinc, because of the high local temperature of the melt under the arc. Indirect-arc furnaces, such as the Rennerfelt, can be used on brasses carrying up to about 2 0 per cent zinc, but are not suitable for ordinary yellow brass, on account of the formation of a superheated layer on the surface of the melt directly under the arc, and the resulting volatilization of zinc. Induction furnaces of the ordinary horizontal ring type, like the Rochling-Rodenhauser, cannot be used on brass or bronze because the high electrical conductivity of these alloys requires a secondary current so high t h a t the “pinch effect” causes rupture of the secondary ring. Hence it has been necessary t o develop types of furnaces radically different from those in use for steel in order t o meet the requirements of brass. E L E C T R I C BRASS F U R N A C E S I N COMMERCIAL U S E

There are, however, two types of steel furnace which have been applied t o brass (using the term brass loosely t o include bronze, red brass, etc.) : the Snyder, a singlephase, direct-arc furnace; and the Rennerfelt, a twophase, indirect-arc furnace. At the Chicago Bearing Metal Company; Chicago, Ill., two one-ton Snyders and two one-ton Rennerfelts are melting bronze for railroad bearings, high in lead, but practically free from zinc. The metal losses are not much reduced from previous practice in crucibles and open flame oil furnaces, but the furnaces are making savings in melting cost as compared with either the crucible o r t h e open-flame furnaces under present conditions. The Philadelphia Mint is melting nickel and coinage bronze in a rooo-lb. Rennerfelt furnace. The Gerline Brass Foundry Company, Kalamazoo, Michigan, melts Monel metal, red brass, and brass containing up t o about 20 per cent zinc in an 800-lb. Rennerfelt. The furnace at the Gerline plant is run on a 9-hour basis, while the other furnaces mentioned operate 18 t o 24 hours a day. Two other types of furnace designed especially for brass melting have also found commercial use, the Baily and the Ajax-Wyatt. The Baily furnace uses a single-phase granular resistor, the heat from which is reflected down onto the hearth from the roof. It takes charges of about 1000 lbs. Baily furnaces are installed a t the Lumen Bearing Company, Buffalo, N. Y., Hays Mfg. Company, Erie, Pa., Bridgeport Brass Company, Bridgeport, Conn., and the Baltimore Copper Smelting and Rolling Company, Baltimore, Md. The Baily furnace is applicable t o alloys of any zinc content, reduces metal losses, avoids crucibles, and gives good working conditions. The main drawback of this type of furnace is t h a t the source of heat is not close t o the melt and the heat must

Vol.

IO,

No. 6

be reflected down from t h e roof. I n order not t o overheat the roof and cause its prompt failure, as well as t o hold t h e resistor temperature within the limits t h a t allow reasonable life of the resistor trough, the rate of power input is low compared t o the size of the furnace and weight of charge. Hence t h e radiation losses from walls and roof form a large proportion of t h e total power. The furnace is at its best an a4-hour operation. When Io-hour operation is necessary, it is found t h a t the furnace must be heated empty during all or part of the night in order t o give satisfactory output in the daytime. Because of the high heat storage in the walls, a furnace of this type does not respond promptly t o changes in power input, and accurate control of t h e temperature of t h e melt is difficult. The Ajax-Wyatt furnace is a single-phase induction furnace in which t h e secondary ring is in the form of a loop below the level of the hearth proper, so that the hydraulic head of the metal in the hearth opposes t h e rupturing effect of the “pinch” force, thus avoiding the troubles which make horizontal-ring induction furnaces inapplicable t o brass. The metal heated in t h e secondary loop is constantly ejected a t one part of each opening from loop t o hearth, and colder molten metal drawn in a t another part of the opening. These fountains of hot metal issuing from t h e resistor melt the charge in t h e hearth. The constant circulation of metal is a most desirable feature and gives a product of remarkably uniform chemical composition. Because of t h e compactness of the furnace, the generation of heat within the metal itself, and t h e stirring action, vertical-ring induction furnaces are extremely efficient as regards power consumption. The power factor in the sizes so far built is satisfactory. The furnace must be started with a charge of previously melted metal, and sufficient metal t o fill t h e loop must be retained when pouring. The metal in t h e loop must never be allowed t o solidify, or the lining will be ruined. These facts make it difficult to change from one alloy t o another, and require t h a t the furnace be run 24 hours a day, or else receive enough power a t night t o keep the metal in the loop fluid. Ramming up and drying the refractory lining of t h e loop is a job requiring care and experience, as the lining must be perfect or its life will be short. No lining has yet been found which will withstand alloys containing over 3 per cent of lead, and the furnace has been developed mainly for yellow brass. The furnace is fitted for rolling-mill use, where 24-hour operation on yellow brass is t h e rule, but is distinctly less suitable for 10-hour runs or for foundries making a variety of alloys. Several of these furnaces are in use a t t h e Ajax Metal Company, Philadelphia, two a t the American Brass Company, Waterbury, Conn., and twenty-eight a t t h e Bridgeport Brass Company, Bridgeport, Conn. The furnace saves zinc, avoids crucibles, and shows so low a power consumption on 24-hour operation t h a t it can doubtless be used t o advantage in rolling-mill practice even under normal prices of fuel and crucibles.

June, 1918

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

FURNACES UNDER EXPERIMENTAL TEST

Besides the four types mentioned above, each of which has found commercial use where conditions were suitable, there are four other furnaces t h a t have reached a semi-commercial stage, but are still under experimental development. The Bennett furnace a t the Scovill Mfg. Co., Waterbury, Conn., is a three-phase furnace, probably of about 7 5 0 lbs. capacity, and resembles a direct-arc furnace. However, the voltage between electrodes (which are automatically regulated) and bath is kept so low t h a t there is no true arc and the heat is generated b y a sort of contact resistance. This is said t o give low metal losses and t o show a reasonably low consumption of power. The furnace has run mainly on yellow brass and is, therefore, probably applicable t o all brasses and bronzes. The results of t h e work have so far been kept secret and n o detailed data are available. The Foley furnace is a single-phase, vertical-ring induction furnace, similar in general design t o the Ajax-Wyatt, although differing from i t in many points. One such furnace, of about 1000 lbs. capacity, has been in experimental operation a t the Bristol Brass Company, Bristol, Conn., and three 3000-lb. furnaces are under construction. From the small amount of data so far available on this furnace, its metal losses and power consumption will be about the same as in the Ajax-Wyatt; due partly t o its larger size, its power factor is somewhat lower. I t has thesame disadvantages as regards starting, changing from one alloy t o another, and the necessity for z4-hour operation. a s t h a t furnace. The General Electric furnace is a smothered-arc, one- or two- (normally two) phase furnace, of about I 500 lbs. capacity, having four depending electrodes, two on each side of a hearth. Between t h e tips of each pair of electrodes is a carbon block t o which arcs are drawn, t h e arcs being smothered by granular coke. The heat thus generated is reflected down onto the hearth by the roof. The electrodes are automatically regulated. After being tested a t the General Electric Company, Schenectady, N. Y . , this furnace has been installed for further test a t the Chicago Plant of the Crane Company, b u t is not yet considered ready for general commercial use. The heat transfer in this type is similar t o t h a t in the Baily, and the furnace seems theoretically capable of a performance of about the same order as t h e Baily with similar advantages and similar drawbacks. As the General Electric furnace takes a higher power input t h a n the Baily, it may be slightly more efficient in power consumption, but the roof is subject t o even more severe conditions and will require the use of highgrade refractories t o give a good life. The Northrup furnace, being developed by Prof. E. F. Northrup and the Ajax Metal Company, is an induction furnace, heating the charge b y means of eddy currents instead of making the charge, or part of it, the secondary of a transformer. Oscillating current of very high frequency is used instead of alter-

46 I

nating current, and is obtained by the use of condensers or of a special generator. A 60-kw. tapping-type furnace is being tried out. The Northrup furnace has a high power factor, and can take multi-phase current. I t is being developed in order t o produce a furnace suitable for Io-hour operation and for facility in changing from one alloy t o another. Since the heat is generated within the charge itself, the eddy-current furnace should be efficient in power consumption. This type is theoretically very promising, but its development has not yet gone far enough t o show what, if any, mechanical limitations the type will have. Many other types of furnace have been suggested for brass melting, and a number have been tried out more or less thoroughly, but those mentioned above are the most prominent of the types in commercial use or under commercial development. Most of these are either limited in their application, or have some drawbacks, either inherent in the type of furnace, or not yet eliminated by long experience in their design anduse, so t h a t no one type or make of furnace is as yet definitely proven the best for any particular set of conditions, and still less will any one furnace meet all the different conditions found in the whole range of the brass and bronze industry. I n particular, none of these types seems quite fitted t o t h a t common set of conditions where a furnace may be called upon t o melt successive heats of alloys differing widely in composition, t o handle both alloys free from zinc and those high in zinc, and t o operate cheaply on a 9- or Io-hour day. ROCKIA'G

ELECTRIC BRASS FURNACE

I n its study of electric brass melting during the past five years, the Bureau of Mines has tried out a rocking type of furnace, which may perhaps help t o fill this gap. I n the ordinary indirect-arc type of furnace, the heat is applied above the melt and as hot metal is lighter t h a n colder metal, there is little circulation in the bath. If the rate of heat input is a t all rapid, as is necessary for thermal efficiency, heat conduction from the top of the melt downward does not keep pace with the heat supply. Before the melt as a whole reaches the proper pouring temperature, the surface is much superheated. On a n alloy high in zinc the surface will reach the boiling point of the zinc in t h a t particular alloy while the bottom is scarcely melted; such heating creates a high pressure of zinc vapor within the furnace, so t h a t if the furnace is not tightly closed zinc is lost continually. If the furnace is sealed tight, the pressure may even blow out the roof or door. I n case the furnace holds tight and the pressure is not relieved till the spout is opened for pouring a long hissing stream of zinc vapor then shoots out, burning in t h e air. This local overheating is t h e cause of the failure of the indirect-arc furnace t o handle alloys high in zinc without large metal losses. The obvious way t o overcome this trouble is t o stir the melt so vigorously t h a t the temperature of the melt is practically uniform and the superheating of the sur-

46 2

8

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

face prevented. The most practical way t o stir the melt is by the principie of the cement-mixer, b y turning the furnace bodily so as t o stir the contents thoroughly while being heated. Constant rotation. of a cylindrical furnace placed more or less horizontally, but preferably a t a slight angle with the horizontal t o produce endwise motion of the melt during rotation, with electrodes entering a t the ends of the drum and a n arc struck between the electrodes, should not only stir the charge thoroughly, avoid surface overheating and thus prevent zinc losses, but should also give a well-mixed alloy. By washing the walls with metal, the heat stored in the walls and roof should be largely taken up in the metal instead of passing out. The power consumption should, therefore, be low. As the walls are washed with metal their temperature can rise little above the temperature of the metal, which should give a good life of lining. Instead of rotating the furnace through a complete revolution-which would involve difficulty in making brush contacts t o the electrodes and in keeping the metal out of the joints between t h e door and the door opening, as this opening should be on the circumference of the drum rather than on t h e end-it appears simpler t o rock the furnace back and forth so t h a t the molten charge just fails t o reach the door a t either end of its rocking angle. A small furnace of this type was built and tried out. This was rocked back and forth by hand on tracks. It was cheaply constructed from materials a t hand in the laboratory and was not expected t o give very good results on power consumption, as the drum was too small t o allow the refractory lining t o be of desirable thickness. The laboratory furnace held about I O O lbs. of charge, and operated on 5 0 t o 75 volts, 500 t o 7 0 0 amperes, a t a power factor of 85 t o go. The usual power input was about 30 kw. Graphite electrodes z in. in diameter were used. A number of different alloys were melted in the rocking furnace. I n melting 1092.1 lbs. of yellow brass, made up of 45 per cent ingot, 5 5 per cent copper and zinc, the calculated analysis being 65.6 per cent Cu, 34.4 per cent Zn, 1080.4lbs. of ingot were obtained, analyzing 65.9 per cent copper, 34.1 per cent zinc. The metal loss by weight was 1.06 per cent which includes both volatilization and mechanical loss by spatter in pouring. The average pouring temperature was 1080' C. On manganese bronze chips (40 per cent zinc), the furnace gave a net metal loss of 3.0 per cent, while the same lot of chips melted in oil-fired crucible furnaces in commercial practice gave 7.2 per cent loss. Yellow brass chips (25 per cent zinc) gave 1.6 per cent net loss, red brass chips (IO per cent zinc), 1.0 per cent. A fine concentrate ( 2 0 mesh) from brass furnace ashes obtained in the manufacture of brass of So per cent copper, 2 0 per cent zinc, analyzed 71.0 per cent copper and 14.3 per cent zinc, t h e balance being ash, etc., gave on melting in the furnace a recovery of gg per cent of the copper and 50 per cent of

Vol.

IO,

No. 6

the zinc in the concentrate. This material is usually sent t o the smelter and refined in a reverberatory furnace, not all of the copper and none of t h e zinc being recovered. Yellow brass ingot (25 per cent zinc) was remelted with 0 . 5 per cent loss. Red brass (IO per cent zinc) made up from red gates, scrap copper, yellow chips, lead, and tin was melted with 0.5 per cent loss. Heavy German silver scrap (18 per cent nickel, 56 per cent copper, 26 per cent zinc), which gave 1.8 per cent loss on commercial melting in coke fires, was melted with 1.2 per cent loss. Sound copper castings were made from metal melted in the furnace. Red brass of 81.5 fier cent copper, 8.5 per cent zinc, 6 per cent lead, 4 per cent tin, made up from red and yellow ingot and scrap copper, was melted in one series of tests with t h e following results, the furnace being cold a t the start. TABLEI Pouring ~

HEAT

No.

Weight of Charge

Lbs.

Total 639.55

Time Arc was on Min.

Av. 46

Temp. O

c.

Av. 1200

Kw.Hrs. Used

Total 1381/2

Kw. Hrs. per 100 Lbs.

Av. 213/4

The total elapsed time for the five heats, including charging and pouring, was 5 hours. 630.9 lbs. ingot were poured and 7.45 lbs. .metal from spillings, etc., were recovered, giving a gross metal loss of 1.33per cent and a net loss of 0 . 2 per cent.

FIG.I

The power consumption, a t t h e rate of 430 kw. hrs.. per ton on a 5-hour run, starting from the cold, an& a t the rate of 2 9 5 kw. hrs. per ton when t h e furnace is hot, with the metal heated t o 1200' C., is surprisingly low for so small a furnace. The results above show t h a t the rocking furnace is a type capable of giving low metal loss and low power consumption. When the furnace was not rocked while melting alloys high in zinc, pressure built up within the furnace and zinc losses were high.

June, 19I 8

T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y

463

FIG.I1

The laboratory tests having demonstrated t h e probable usefulness of t h e type, a furnace of commercial size was designed. The Detroit Edison Company had long been interested in electric brass furnaces as a possible outlet €or electric power, and offered t o co6perate by constructing a rocking furnace for commercial test without expense t o t h e Bureau of Mines except t h e salaries a n d expenses of its representatives while supervising t h e test. Sketches of t h e furnace design were given t h e Detroit Edison Company, which refined t h e design, made t h e working drawings, constructed and erected t h e furnace. The furnace is shown in Figs. I and 11. The drum is 5 f t . in diameter b y 5 f t . long. The lining is 1 2 in. thick, and consists of silocel brick on t h e outside, special heat-insulating brick in the middle layer, a n d corundite brick (a very refractory firebrick high in A1203) in t h e actual hearth lining. The hearth is 3 f t . long by 3 f t . in diameter, taking charges of 1300 lbs. and upwards. The electrodes are 4 in. diameter graphite, threaded for continuous feed, and are adjusted by screw-operated supports of t h e lathe-slide type. Single-phase, 60 cycle current, stepped down t o 1 2 0 or 1 3 0 volts is used, 300 kv. amp. being available. Electrode adjustment is b y hand, a n d t o stabilize t h e arc a n external reactance is used which brings the power factor of furnace plus reactance,

measured a t the furnace switchboard, t o about 85. T h e open circuit voltage falls t o about 106 t o 116 volts under load. The current varies between 1000 and 2 0 0 0 amperes, 1650 amperes being about the average. The power input can be varied b y altering the length of the arc, and runs from I O O t o 2 0 0 kw., averaging about I 6 5 kw. The flexible leads and the water hose €or electrode cooling are given slack t o allow rocking the furnace, as is clearly shown in Fig. 11. The rocking of the furnace during melting is automatically done by means of t h e control device shown, with cover removed, in the lower left-hand corner of Fig. 11. This can be set t o give a “safe rock” of So0, t h e limit of motion being such t h a t t h e metal just does not run into t h e spout. After the charge has begun t o melt, the “safe rock” is started. It is called the “safe rock” because the angle is such t h a t solid charge will not fall on the electrodes and break them. A complete oscillation on “safe rock” takes 13l/2 seconds. During t h e “safe rock” t h e solid metal is swashed about in the molten part of the charge and is tumbled over, so t h a t fresh surfaces receive direct radiation from the arc. As melting goes on, t h e rocking angle is increased by turning the handle of the control device from time t o time, until, when the metal is all melted, the furnace is on the “full rock” of about 2 0 0 ~ . On “full rock” t h e metal washes t h e whole cir-

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

464

cumference of the hearth save the height of the charging door and a few inches above and below i t , so t h a t metal does not splash into the door joint. A complete oscillation takes 331/* seconds. The reversal of the 5 h. p. motor a t either end of the rocking angle is done by contactors, operated b y solenoids, actuated by the contacts on the control device. When it is desired t o depress the spout past t h e limiting point of the automatic rock, for pouring, the control device is switched out and the solenoids are operated by a reversing switch. The furnace is installed a t the plant of the Michigan Smelting and Refining Company, Detroit, Michigan, which makes brass ingot t o customers' specifications from chips, scrap, and junk of various kinds, b y means of strict chemical control. As the firm makes no sand castings, but ingot only, no observations on the comparative quality of metal melted in the electric furnace and in t h e coke fires were possible. All the metal melted was poured into ingot which went into the regular output of the plant. As,far as could be told b y analysis and appearance, the electrically-melted metal was of a t least as good a quality as from t h e coke fires. On alloys high in lead there was somewhat less segregation t h a n in the metal melted in crucibles, and on charges high in zinc, the zinc content of the metal from the electric furnace was higher t h a n t h a t from the same charges melted in the coke fires. As there is generally much oil on the borings and some non-metallic material in the other scrap, the true metallic content of t h e charge is seldom accurately known. Hence t h e net metal losses cannot be exactly determined. The metal losses were, therefore, compared with those of the coke-fired crucible furnaces operating on the same charge. From 1 0 2 tons of metal melted in strict comparison with the crucible furnaces, the rocking electric furnace produced 3626 lbs. more metal from the same charge than the coke fires, or 1.8 per cent. The alloys melted ran from g o t o 66 per cent copper, I to 9 per cent tin, 1.5 t o 26.5 per cent lead and o t o 30 per cent zinc. The comparative metal losses on a few alloys in the electric and the coke fires are given in Table 11. T A B L EI1 Cu

Composition Per cent Sn Pb

Zn

Weight Per cent Loss Per cent LOSS Charged (Metal, Oil, Dirt) (Metal, Oil, Dirt) Lbs. Coke Fires Electric 3.2 6576 4.6 3.7 7.0 11600 1.8 2.4 14300 2.1 3.6 11790 3.1 7.1 15840 2.4 4.0 11805 2.9 3.7 14392 2.4 3.0 5224 5.1 8.0 7200

The rocking furnace gave alloys and analyzed very close t o the calculated analyses, especially if the difficulty of calculating the analysis of a scrap charge is considered. Characteristic analyses are given in Table 111. There was no difficulty in draining the metal completely from the hearth, and alloys of different composition can be made one after the other without con-

Vol.

No. 6

IO,

tamination by metal left in from the previous heat. TABLSI11 Copper Sought.. . . . . . . . . . . . . . 76 Electric.. . . . . . . . . . . . . 7 5 . 9 Sought.. . . . . . . . . . . . . . 76 Electric., ............ 7 6 . 2 Sought.. . . . . . . . . . . . . . 85 Electric.. . . . . . . . . . . . . 8 5 . 2 Sought.. . . . . . . . . . . . . . 83 Electric.. . . . . . . . . . . . . 8 2 . 9 Sought.. . . . . . . . . . . . . . 67 Electric.. . . . . . . . . . . . . . 6 6 . 6 Coke . . . . . . . . . . . . . . . . 6 8 . 4 S o u e h t . . . . . . . . . . . . . . . 68 Electric. 67.9 Coke . . . . . . . . . . . . . . . . 6 9 . 9 Sought . . . . . . . . . . . . . . 60 Electric. . . . . . . . . . . . . . 5 9 . 7

.............

Tin 8 8.3 8 8 5 4.9 4 4.4

Lead 13 13.1 13 12.4 5 4.8

Zinc 3 2.7 3 3.2 5 5

6

1 1

5.7

6.9

2 1.7 7

30 30.4 29.3 24

2

0.5 1

....

....

t , .

... ... ...

.

I

.

ii'

.

3

'

....

....

The power consumption on Io-hour operation, with no night heating, is shown in Table IV, which gives a r6sum6 of j days' operation. The power consumption on z4-hour operation is shown in Table V for a 4-day run. On the basis of power, read on the high tension side of the transformer, per ton of metal poured, the power consumption on Io-hour operation was 336 kw. hrs. per ton, on red brass poured a t 1180' C. average. For 24-hour operation, the figure is about 260 kw. hrs. per ton for red brass. The electrode consumption was 16.3 lbs. while melting 21,660 lbs. of metal, or I I / ~ lbs. per ton, equivalent t o about 40 cents a t present electrode prices. To this must be added t h e loss due t o accidental breakage. There were nine breakages in melting 7 2 tons, four of which were due t o the charge being so bulky t h a t i t fell against the electrodes when rocking started, and five t o the electrodes being hit while bulky material was being charged. The design of the furnace has now been altered so as t o allow the electrode tips t o be withdrawn into the walls during the charging of bulky material. When a n electrode does break, if nipple joints are used, t h e breakage is usually of the nipple only. I n the n4-hour tests tabulated in Table V, and in a Io-hour run just preceding t h e a4-hour runs, in which the 7 5 . 2 5 Cu, 7.5 ,Sn, 14.25 Pb, 3 Z n alloy was melted, there was charged, for the 75.25 Cu alloy,

......

... ... ... Yellow borings.. ,

252001bs. 11200 lbs. 2 per cent oil = 224 lbs. nonmetallic 15401bs. 10987 lbs. 3906 lbs. 1400 lbs. -

3 per cent oil

=

54805

For the 86 Cu, 6 Sn,

I n g o t , . . . . . . . . . . 16000 lbs. cu... . . 4704lbs. 961bs. Pb . . . . . . . . . . . . .

42 lbs. nonmetallic 266

IO

P b alloy there was charged

20800

TOTAL CHARGE. . 75605 lbs.

266 lbs. nonmetallic 75339 lbs. metallic

There was obtained 53841 lbs. good ingot 75.25 Cu

20149 Ibs. good ingot 86 Cu -

73990 total good ingot, 1349 lbs. gross loss, or 1.8 per cent 63 lbs. scrap 75.25 Cu 43 lbs. scrap 86 Cu 300 lbs. metallics'in 569 lbs. skimmings 53 per cent metallic .in all from 75.25 Cu 130 lbs. metallics in 246 lbs. skimmings skimmings b y assay from 86 Cu 365 lbs. metallics in 429,lbs. ladle skulls from 86 Cu,, 85 per cent metallic

}

__

74891 total metallic recovery, 448 lbs. net loss, or 0.6 per cent

June, 1918

T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY

I n heats 256-313, on over 3 7 l / 2 tons melted, the electrode consumption, including all broken stubs and scrap, was slightly less than 2 lbs. per ton. Since the operation was experimental, i t is not yet possible t o give exact figures on the life of a lining, but as nearly as can be estimated t h e relining cost €or labor and be under so cents per ton with a corundite lining, when melting red brass poured at “so to I 2 O o 0 c. If Only yellow brass poured at 1100’ C., is melted, the lining cost will be still lower. If hot bronze t o be Produced, say at I3OOo C.1 the roof and upper portions of the ends should be lined with zirkite brick.

46 5

possible by electric melting under present conditions, and even a t pre-war prices for crucibles, coke, and metal the rocking furnace will show a distinct though smaller saving. On 24-hour operation the balance in favor of electric melting is still more marked. F~~~ data at hand on the power consumption of other types of electric furnaces, it appears that, when operated on the Same alloy, heating it to the Same ternperature, and running the same number of hours per efficient day, the rocking furnace is somewhat than the direct-arc, and unrocked indirect-arc types, very much efficient than electric furnaces of types in which heat is reflected onto the charge from Accurate temperature control is very easy in the the roof, and very little less so than the induction furrocklng furnace, since a t t h e end of a heat, after t h e naces. These conclusions follow not only from the “full rock,” the walls are no hotter than the metal, data a t hand, but from the method of application and there is no heating up of the charge from hotter of heat in the various types, those mrith the source roof and walls when the power is shut off, as is the case of heat at a distance from the charge being less efiwith those types of furnace where t h e heat is re- cient than those where the heat is developed close t o fleeted downward from the roof. After cutting off the charge. The induction furnaces in which the heat the arc, the temperature falls very Slowly, about 2 is developed in the charge itself should be the most t o 3’ c. Per minute. BY running the arc a minute efficient. On account of t h e washing of t h e walls Or SO every 10 Or 1 5 minutes, a charge Can be held a t with the metal, the rocking furnace should theoretically pouring temperature for a n indefinite period. come next t o the induction type in thermal efficiency. One man can operate the furnace, with the aid Of a I n magnitude of metal losses, the rocking furnace charging’ Were gives a t least as good results as any other type of eleccontrol used, which could easily be done, one man could tric furnace. only possible loss is f r o m the stream probably attend to two furnaces. of metal while pouring, as the furnace is sealed tight The Output per man hour was greater from the while running. Volatilization from the stream while rocking furnace than from the coke fires. The work- pouring is of about the Same in all typesof ing conditions are much less severe and more health- furnaces. ful with the electric furnace t h a n with the coke fires, I n closeness of control of the temperature of the and a man of less rugged physique than is required melt the rocking furnace is superior t o any save t h e for coke fires can readily operate the rocking furnace. induction type. I n thorough mixing of the charge, Various modifications and improvements in design the type is about On the Same plane as the were made during t h e tests, and others t h a t couldnot induction type, and markedly superior t o the other well be made on the first furnace are being incorporated where, in large sizes, segregation in the bath in other furnaces of this type now being built for Detroit firms. The electrodes were a t first introduced may be a serious problem* For example, the following shows the analysis for into the furnace directly through the refractory walls. When making yellow brass from new materials so copper of the first ingot from the first ladle and of t h e t h a t addition of much spelter is required, the zinc, last ingot from the last ladle, when melting 1200-lb. vaporized during the addition of the spelter t o t h e charges of 6 0 Per cent C U , 37 Per cent Zn, 3 Per cent Pb. molten charge, tended t o condense in the clearance HEAT First Ingot, First Ladle Last Ingot Last Ladle Per cent Cu Per c;nt cu No. between the electrode and the hole through which it 59.54 322 59.76 323 59.78 59.66 entered. This would then freeze, solder the electrode in place, and cause breakage. Such trouble was later obviated by the use of graphite sleeves about I n ability t o change from One alloy $0 another, i t the electrodes and by the proper arrangement and is superior t o the vertical-ring induction type, and i n to Operate when used but Io hours operation of the electrode coolers. I t was also found feasible t o charge the zinc with the rest of the charge a day, without night heating, is ahead Of the ring induction type and of the reflected-heat type. instead of speltering a t the end of the heat. Comparing the cost of melting on a Io-hour schedule The rocking furnace can handle alloys of any zinc in the rocking electric furnace and in the coke fires or lead content, being superior on this score t o directof the plant a t which the test was made, the sum of the arc, unrocked indirect-arc, and induction types. The cost per ton of charge for electric power, interest and electrode cost compares favorably with other arc depreciation, electrodes, linings, and for heating ladles, furnaces. With equal conditions of operation, and is just about one-half of t h e cost per ton of charge of suitable refractories in each type, the cost of lining will the single item of crucibles a t present prices and at probably be about the same as with most other types. present crucible life, The value of the metal saved Labor cost should be about t h e same in all handby the electric furnace is about twice t h e cost of t h e regulated arc furnaces. With automatic regulation, coke used by t h e coke fires. Hence a huge saving is which can be applied if desired, the racking type should

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

466

Vol.

IO,

No. 6

TABLEIV-TEN-HOUR OPERATION

Secondary Primary Nature Wt.of Elapsed Sec. Total Weight Pri- Kw. h./ Kw. h./ Heat Per cent Alloy of Charge Time Kw. h. Kw.h. Sec. Pouring Poured mary Cwt. Cwt. DATE No. Cu Sn P b Zn Charge Lbs. Hrs. Min. Arc Motor Power Temp. Lbs. Kw.h.Charged Poured RSJIARKS 1 Average 1314 3 : 40 257 4 261 2OOOOF. 20 . . . . Furnace cooler than usual, not r u n Nov. 5 192 85 5 9 1095' C. previous two days. No. 192 includes 1 hr. 20 min., 100 Kw.hrs. preheat 1 Average 1314 1 : 50 219 3 222 2050'F. . . . . . . . . 17 P-93 85 5 9 11200 c. 2 198 2125OF. ........ 15 .... 2 Little 1304 1 : 35 196 194 79 9 10 1115.5~C. bulkier 1304 2 192 2200'F. ........ 14.5 . . . . 190 2 than 1 : 30 195 79 9 10 1205' C. average L 190 193 2200'F. . . . . . . . . 15 . . . . Time includes 20 min. charging 2 . . . . . 1304 1 : 4 0 3 196 79 9 10 1205' C. Heat No. 197 1066 2115' F. 6360 1162 1052 14 16.3 18.3 6540 10 : 15 DAY TOTAL5 heats 11550 c. ......... 18.5 . . . . 238 2125'F. 2 Little 1304 1 : 45 235 3 9 10 Nov. 6 197 79 1165' C . bulkier 1304 202 2175" F. . . . . . . . . 15.5 . . . . 2 than 1 : 30 199 3 198 79 9 10 1190' C. average 2 188 2250OF. . . . . . . . . 14.5 . . . . 186 2 Little 1 : 40 199 79 9 10 1230' C. 1304 bulkier 2 176 178 2240OF. . . . . . . . . 13.5 2 than (1304 1 : 40 200 79 9 10 1225' C. average 162 2 164 210O'F. ........ 12.5 . . . . 201 79 9 10 11500 c. bulkier 2 162 2125'F. ........ 12.5 . . . . Time includes 20 min. charging 160 202 79 9 10 1165' C. Heat No. 203 average

........

....

/

{

I

D A YTOTAL6 heats Nov. 7

2

203

79

9

10

204

79

9

10

205

79

9

10

206

79

9

10

L

207

84

6

10

0

208

87.5 5

7571

1236

16.3

17

195

2

15

.... ....

14

....

13.5

1304

1 : 40

165

2

167

1304

2 : 30

178

3

181

7824

10 : 30

1106

14

Very bulky

1300

3 : 00

246

2

Very bulky 0 Very bulky 0 Very bulky 0 Very bulky

1300

1 : 40

198

3

1300

2 : 05

188

2

1300

1 : 30

169

3

1300

1 : 50

170

1

6500

10 : 05

971

11

1300

1 : 45

223

4

227

1300

2 : 05

198

3

201

1300

1 : 00

195

2

197

1300

1 : 45

189

3

192

10

0

210

84

6

10

0

211

84

6

10

212

84

6

10

213

84

6

10

0

215

84

6

10

0

216

84

6

10

0

217

84

6

10

0

218

84

6

10

0

DAYTOTAL 5 heats

1132 2185'F. 1195'C.

2

0.5 7

10

14.7

3

2

6

6

14

215

180

84

84

1118

1 : 35

173

209

214

9 : 50

1 : 35

Very bulkv Very bulky

DAYTOTAL 5 heats Nov. 9

7824 (1304 bulkier average

DAYTOTAL 6 heats Nov. 8

....

Very bulky Very bulkv Very bulky Very bulky Very bulky

11304

2150'F. 1175'C. 1950'F. 1065' C.

1120 2120OF. 1160' C. 2 4 8 , 2050" F. 11200 c. 201

2175'F. 1190' C. 190 2175'F. 11900 c. 172 2175OF. 1190' C. 171 2175°F. 11900 c.

982

1300

1 : 25

165

2

167

6500

8 : 50

970

14

984

show a labor cost about the same as t h a t of any other type. From the electrical point of view of desirability of a steady load, the rocking furnace does not have so steady a load and hence, on this score, is not so desirable as the induction furnaces or granular resistor furnaces. I t does not require special transformers, as the granular resistor type does. It lacks the electrical advantages of multi-phase furnaces. I n very large sizes, two arcs could be used in the rocking type, but in sizes up t o one ton, single-phase operation is required, and in a plant so located t h a t the power supply must be of limited capacity, a singlephase arc furnace, with its fluctuating loads, may not

215O'F. 1175' C. 2175°F. 1190' C. 2200° F. 1205' C. 2160'F. 1180" C. 2250'F. 1235' C. 2150'F. 1175' C. 2190'F. 12000 c.

........

13

........

14

.... .... . . . . Includes

7583

14.4

16.0

........

19

..,.

........

15.5

........

14.5

.... . ..

........ ........

13.5

....

13

. . . . Includes 20 min. charging No. 214

6341

1069

15.1

16.9

........ ........

17.5

....

15.5

1210

,

........

15

.... ....

........

14.5

....

........

13

....

6407

15.2

16.7

1073

50 min. delay by broken electrode broken in charging bulky. charge, also 20 min. charging Heat No. 209

Time includes 1 hr. 10 min. delay due to broken electrode caused by bulky charge. Long delay due t o nipple being over-size and requiring t o be filed down Includes 25 min. adjusting electrode holder

be satisfactory from the electrical point of view. Such fluctuation is no drawback in Detroit nor would i t be in most cities or large manufacturing towns. From the results on furnaces of 1 2 5 and 1300 lbs. capacity, i t appears t h a t the rocking type can be built in a wide range of sizes without showing a great loss of efficiency in t h e smaller sizes. This type can doubtless be built in as large sizes a s the brass industry could normally use. I n first cost, the rocking type should be no more expensive t h a n other electric furnaces. While further tests in different plants and under different conditions, which will be made a t least in part, in the near future, are needed t o give accurate

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

June, 1918

DATE Apr. 30

Conseq. Day Weight of Heat Heat Charge Per No. No. Lbs. Cu 75.25 261 1 1305

'

Midnight

267 268 269 270 27 1 272 273

2 3 4 5 6

1305 1305 1305 1305 1305

75.25 75.25 75.25 75.25 75.25

7 8 9 10 11 12 13

1305 1305 1305 1305 1305 1305 1305

75.25 75.25 75.25 75.25 75.25 75.25 75.25

16965

DAYTOTAL May 1

Midnight

Midnight

....

...

.... .... .... .... ....

... ... ... ... ...

.... .... .... .... .... .... ....

... ... ... ...

... ... ...

274

1305

75.25

7.5

14.25

3

2 : 11

1 : 18

143

....

...

275 276 277 278 279

2 3 4 5 6

1305 1305 1305 1305 1305

75.25 75.25 75.25 75.25 75.25

7.5 7.5 7.5 7.5 7.5

14.25 3 14.25 3 14.25 3 14.25 3 14.25 3

1 : 49 1 : 40 1 : 35 1 : 45 3 : 00

1 1 0 1 1

: 12 : 07 : 55 : 12 : 20

147 152 147 142 160

....

...

280 281 282 283 284 285 286 287

7 8 9 10 11 12 13 14

75.25 75.25 75.25 75.25 75.25 75.25 75.25 75.25

7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5

14.25 14.25 14.25 14.25 14.25 14.25 14.25 14.25

3 3 3 3 3 3

1

1305 1305 1305 1305 1305 1305 1305 1305 18270 1305

75.25

7.5

14.25

3

1 : 32 1 : 21 1 : 19 1 : 23 1 : 31 1 : 34 1 : 20 1 : 35 23 : 35 1 : 35

1 0 0 0 0 1 0 0 14 1

: 02 : 57 : 53 : 55 : 56 : 01 : 48 : 53 : 29 : 05

151 146 145 151 152 152 141 147 2076 150

2 3 4 5 6

1305 1305 1305 1305 1305

75.25 75.25 75.25 75.25 75.25

7.5 7.5 7.5 7.5 7.5

14.25 14.25 14.25 14.25 14.25

3 3 3 3

1 : 25 1 : 25 1 : 20 1 : 35 1 : 39

0 0 0 1

: 57 : 55 : 02 0 : 56

: 52

145 143 144 144 141

294 295

7 8

1300 1300

84 84

6 6

10 10

0 0

1 : 31

1 : 18

1 : 04 0 : 50

140 150

296 297 298

9 10 11

1300 1300 1300

84 84 84

6 6 6

10 10 10

0 0 0

1 : 48 1 : 22 2 : 13

0 : 59 0 : 52 1 : 05

158 151 156

299

12 13 14

86 84 84

6 6 6

10 10 10

0

1

1300 1300 1300 18230 1300

84

6

10

0

303

2

1300

84

6

10

0

304

3

1300

84

6

10

300

301 DAYTOTAL May 3 302

Midnight

TABLE V-TWENTY-FOUR-HOUROPERATION Kw. h. Arc Plus Eouiva._ Rocking lent Motor Kw. h. Pri. E1apsed Melting Read on on PriKw.h. Time Secondary mary per Ton Time cent Alloy Hrs. Min. Side Side Charged REMARKS Hrs. Min. Sn P b Zn 7.5 14.25 3 2 : 09 1 : 47 229 Started a t 6 : 30 A.M. Furnace idle since 4 : 30 P.M., Apr. 29 7 . 5 14.25 3 1 : 53 1 : 28 192 1 : 22 7.5 14.25 3 1 : 55 168 7.5 14.25 3 1 : 44 1 : 15 152 1 : 15 144 1 : 44 7.5 14.25 3 3Omin. (included in elapsed time) adding 141 1 : 2 7.5 14.25 3 2 : 12 electrode sections and taking fresh grip 7.5 14.25 3 153 1 : 30 0 : 55 0 : 58 156 7 . 5 14.25 3 1 : 35 151 1 : 00 7.5 14.25 3 1 : 41 7 . 5 14.25 3 1 : 33 1 : 02 146 7 . 5 14.25 3 0 : 59 143 1 : 33 1 : 06 7.5 14.25 3 1 : 47 149 18 min. (included) replacing broken 7.5 14.25 3 2 : 01 1 : 02 150 electrode nipple. End of heat at 5 : 40 A.M. 23 : 17 15 : 11 2270 268 2074

1

DAYTOTAL May2 288

3 3

3

.... ....

: 12 : 59 : 14

162 179 163 2126 167

1 : 43

1 : 22

158

....

0

1 : 55

1 : 09

158

0

1 : 44

1 : 17

159

1 1 1 2

1 1 1 1

: 16 : 10 : 10 : 05

166 162 162 165

0 : 55 1 : 00 0 : 50 1 : 13 13 : 41

152 140 142 144 1875

0 0

305

4

1300

84

6

10

5 6 7

1300 1300 1300 1300

84 84 84 84

6 6 6 6

1 0 ' 0 10 0 10 0 10 0

310 311 312 313

9 10 11 12

1305 1305 1305 1305 15620 69085

75.25 75.25 75.25 75.25

7.5 7.5 7.5 7.5

14.25 14.25 14.25 14.25

DAYTOTAL 4-DAY TOTAL

.... .... .... .... .... .... .... .... .... .... .... .... 2272 .... .... .... .... .... .... .... .... .... .... .... .... 2318 ....

306 307 308 309

8

467

3 3 3 3

1 1 1 22 1

1 1 1 1 19

: 28

: 37 : 49

: 05 : 43

: 50

: 35 : 23 : 15

: 19 : 14 : 21

: 36 : 38

d a t a on the complete performance of t h e rocking type of furnace, i t would seem from t h e results so far t h a t i t may be of distinct value in the brass industry, especially under present conditions as t o crucible prices and quality, fuel supply and prices, and metal prices. At the conclusion of the tests conducted by the Bureau of Mines, which covered over 300 heats, the experimental furnace was p u t on regular production b y the Michigan Smelting a n d Refining Company. This company is having four one-ton rocking furnaces built, and two are under construction for the Electro Bronze Company, of Detroit. The patents taken out by the Bureau of Mines on

0 1 1 13 1

: 59 : 10

I

.

.

.

.... .... .... .... .... I .... ....

.... ....

2005 8865

.*. ... ... ... ... ... ... ... ... ... ... ...

249

... ...

36 min. (included) wait for helpers to pour metal

1 hr. 10 min. (included) replacing broken electrode and altering cooling coil

E n d of heat 5 : 20

A.M.

Heat started a t 6 : 35 A.M. Furnace idle 1 hr. 25 min. between shifts

... ...

... ...

... ... ... ... ... ... ... ... 254 ... ... ... ...

... ... ...

...

...

... ... ...

262 25 7

Much delay in pouring this heat, no helpers Between 294 and 295, furnace idle 45 min. at change of shifts 39 min. (included in elapsed time) replacing broken electrode Heat ended a t 5 : 20

A.M.

Heat started 6:45. Furnace idle 1 hr. 35 min. between shifts Furnace idle 1 hr. between 302 and 303, operator in conference Furnace idle 11/z hrs. between 303 a n d 304. Broke electrode charging 304, none on hand, wait for one from machine shoD 25 min. patching electrode hole between 304 and 305

42 min. delay (included) broke 2 electrode nipples charging

Last ladle poured 6 : 35

A.M.

t h e rocking furnace have been assigned t o the Secret a r y of t h e Interior as trustee, and free licenses t o operate under them can be obtained by making applica. tion through the Director of the Bureau of Mines. Grateful acknowledgment is made t o Cornel1 University for use of the well-equipped Cornel1 electric furnace laboratory in t h e work on the laboratory furnace, t o Dr. J. M. Lohr, formerly of t h e Bureau of Mines, for aid in t h e work on the laboratory furnace, t o the Michigan Smelting and Refining Company for facilities for the test, and t o the Detroit Edison Company, and particularly t o Mr. E. L. Crosby of the latter firm, for never-failing cooperation. A more detailed account of the tests of the rocking

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

46 8

furnace will soon be published the Bureau of Mines.

as

Bulletin 171 of

BIBLIOGRAPHY ADVANTAGES O F E L E F T R I C BRASS MELTING

E. F. Roeber, Manufacture of Brass in the Electric Furnace,” Elcctrochem. and Met. Ind 3 (1905), 4. G . H. Clamer and C. Hering, “The Electric Furnace for Brass Melting,” Trans.,zAm. Inst. Metals 6 (1912) 95. D.D. Miller, T h e Electric Fui&ace for Heating Non-Ferrous Metals,” J . A m . Inst. Metals, 11 (1917), 257. C. A. Hansen, “Electric Melting of Copper and Brass,” T r a n s . A m . Inst. Metals, 6 (1912), 110.

BAILY F U R N A C E

T. F. Baily, “Annealing and Heat-Treating of Steel and Melting of Xon-Ferrous Metals in the Electric Furnace,” Met. and Chem. Eng., 17 (1917). 91.

SNYDER FURNACE

F. T. Snyder, U. S. Patents 1,100,994 and 1,167,026.

G E N E R A L ELECTRIC F U R i i A C E

I. R . Valentine, U. S. Patent 1,242,275. R a N N E R F E L T FURNACE

I. Rennerfelt, U. S. Patent 1,076,518. AJAX-WYATT

FURNACE

G. H Clamer “Melting Brass in the Induction Furnace,” J . A m . I n i t . Metals: 11 (1917), 381. J.R. Wyatt, U. S . Patents 1,201,671, 1,235,628, 1,235,629and 1,235,630. NORTRRUP-AJAX FURN5,CE

E. F. NErthrup,

Production of High Temperature and Its Measurement, Met. & Chem. Enn., 17 (1917), 685.

PINCH EFFECT

”A Practical Limitation of Resistance Furnaces, the Pinch’ Phenomenon,” Trans. Am. Electrochem. Soc., 11 (1907), 329; 15 (1909), 255.

C., Hering,

VOLATILITY O F ZINC,,IN BRASS

Brass Furnace Practice in the United States,” Bureau of Mines, Bull. 73 (19141, 129. J. Johnston, “The Volatility of the Constituents of Brass,” J . A m . Inst. Metals. 12 (1918). 15. H. W Gillett

ROCKING

FURNACG~



’ I

H. W. Gillett, and J. M . Lohr, U. S . Patent 1,201,224, H. W. Gillett, U. S. Patent 1,201,225. MORSEHALL

ITHACA, N. Y.

A SUMMARY OF T H E PROPOSALS F O R T H E UTILIZATION OF NITER CAKE B y JOHN JOHNSTON Received April 1, 1918

Partly owing t o t h e great shortage of sulfuric acid i n Britain, partly in response to an appeal for suggestions made by the Ministry of Munitions, there has been considerable interest in the question of the disposal of niter cake (acid sodium sulfate). Similar interest in this matter will arise here, for there is already a shortage of sulfuric acid; it will consequently be necessary t o economize in acid, and t o substitute niter cake wherever such substitution is feasible. Some time ago I made a search through all recent literature available1 and compiled a summary of t h e various proposals which have been made for the utilization and disposal of niter cake; and i t has been thought desirable t o publish this summary as a means of showing the possibilities and arousing more general interest in this direction. Some of the proposals are obviously not very practical-even in war time; b u t it seemed better not t o exclude a suggestion even although i t does not appear feasible t o us now. Niter cake, a by-product of the production of nitric acid, is an acid sodium sulfate, usually containing only slight impurities. I t s available sulfuric acid content ranges from 3 5 per cent downwards, but is usually from 2 5 t o 30 per cent; this free acid may cause difficulties in handling and transportation, particularly if water, or even moisture, gets access t o it. The annual production of niter cake in the United States was, according t o the 1909 census, about 43,300 tons, 1 I n the literature citations given subsequently, reference t o C . A . , in addition t o the journal reference, signifies that the original article was not available and that the statements made are on the basis of the abstract in Chemical Abstracts.

VoL no, No. 6

of which 27,600 tons were reported t o have a v-Jnte of about $ 2 per ton, t h e remaining 15,700 tam k i a g reported as of no value; t h e amount now available is, however, very much greater and is of the mdeT of 600,000 tons at least. The utilization of this material in place of t h e equivalent quantity of strlfaric acid, in so far as such substitution is possible, womld therefore result in a very appreciable economy 01f sulfuric acid, t h e demand for which is likely ta be in excess of t h e supply available. The substitmtlon C D ~ niter cake for acid would moreover, a t the pmsemt time, result in a considerable money saving, for it can be bought at a price of about $3 per ton at t h e point of shipment, equivalent t o an acid price of abmrt $IO per ton. As an example of t h e expansion of the use of niter cake in Britain since t h e war we may cite a paper by Kilburn Scott.% Before the war it was used to some extent for making hydrochloric acid and sodium SUI fate, and a small amount was sold t o fertilizer plants and t o glass makers. It is now currently used in the following processes: the extraction of grease fram wool suds and from piece scouring suds; refining Q€ grease; stripping color from rags, dyeing of rags, and removing cotton from mixed fabrics in t h e manufacture of shoddy; calico bleaching; paper making; in the mineral water industry; and in making srtlfate of ammonia. He also discusses t h e methods of handling and dissolving niter cake.2 Attention is therefore directed t o the various proposals outlined below, in the hope t h a t niter cake will be used, wherever feasible, as a means of reducing t h e shortage of acid. The proposals have, for convenience of reference, been grouped under a number of headings, but it is obvious t h a t these several categories are not mutually exclusive. A S A PICKLING AGENT-LeChatelier and Bogitch‘ discuss t h e advantages of using niter cake for removing scale from the surface of iron, and recommend a procedure for its use, namely, t o work with a solution at 80” containing 2 5 per cent niter cake, the acidity of which is maintained by further additions of niter cake. Directions are also given in a recent paper.4 The use of niter cake for pickling iron or steel is the subject of a patent granted t o A. K. Eaton,s who claims the process of “removing hammer scale from iron and steel, which consists in subjecting the scale-coated metal t o the action of a bath containing sodium bisulfate.” It is reported t h a t a large tonnage is already used for this purpose, thus releasing an equivalent amount of acid for other purposes. H. W. Brownsdon6 discusses its application in t h e pickling of annealed brass and states t h a t i t works 1

“Economy of Acids in Metal Trades,” J . Sac. Chem. Ind , 36 (1917),

810. J . Sac. Chem. I n d . , 36 (1917), 1216A. C . A , , 10, 2460. 4 “Pickling with Niter Cake,” Iron Trade Review, 1918, 153. I have been informed t h a t solutions of one-half the concentration recommended by LeChatelier and Bogitch are perfectly satisfactory for pickling metals. 5 U. S . Patent 702,050, June 10, 1902, “Method of Removing Scale Oxide from the Surface of Iron and Steel.” 8 J. Sac. Chem. I n d . , 36 (1917), 575. 2

a Rew. MBtall., 12 (1915), 949,