US2768075A

US2768075A - Melting, remelting and recovering of aluminium and its alloys

Info

Publication number: US2768075A
Application number: US257047A
Authority: US
Inventors: Sterental Volf
Original assignee: Individual
Current assignee: Individual
Priority date: 1942-12-12
Filing date: 1951-11-19
Publication date: 1956-10-23
Anticipated expiration: 1973-10-23
Also published as: GB640195A; FR918918A; CH263331A

Description

Oct. 23, 1956, v. STERENTAL 2,768,075

MELTING, REMELTING AND RECOVERING OF ALUMINIUM AND ITS ALLOYS 5 Sheets-Sheet 1 Filed Nov. 19, 1951 Fig.1

Oct. 23, 1956 v. STERENTAL 2,763,075

, MEL-TING, REMELTING AND RECOVERING OF ALUMINIUM AND ITS ALLOYS Filed Nov. 19, 1951 5 Sheets-Sheet 2 Fig.2

v. STERENTAL 2,768,075 MELTING, REMELTING AND RECOVERING OF ALUMINIUM AND ITS ALLOYS Oct. 23, 1956 5 Sheecs-Sheec 3 Filed Nov. 19, 1951 Oct. 23, 1956 v. STERENTAL 2,768,075

MELTING, REMELTING AND RECOVERING OF ALUMINIUM AND ITS ALLOYS Filed NOV. 19, 1951 5 Sheets-Sheet 4 Fig.10

' Oct. 23, 1956 v. STE NTAL 2,763,075

MELTING, REMELTI AND RECOVERING OF ALUMINIUM AND ITS ALLOYS 5 Sheets-Sheet 5 Filed Nov. 19,- 1951 Patented Oct. 23, 1956 MELTING, REMELTING AND RECOVERING OF ALUMINIUM AND ITS ALLOYS Volf Sterental, Turin, Italy Application November 19, 1951, Serial No. 257,047 Claims priority, application Italy December 12,

'13 Claims. (CI. 7568) This invention relates to a method of smelting, resmelting, and recovering aluminum, its alloys and their various residues from machining and foundry, with minimum melting losses lower than those inherent to other smelting process and a satisfactory purification of the metal bath from the oxide.

In smelting aluminum and its alloys and in re-smelting and recovery of their residues from machining and foundry, it is necessary to take various precautions in order to avoid oxidising of the metal during its liquefaction and prevent absorbtion of soluble gases and oxidation of the molten metal bath, enhanced by an accidental superheating of the latter. Moreover, it is convenient to improve purification of the metal bath from oxide, either from the superficial layer covering the solid metal, of which the quantity in proportioned to the width of its surface with respect to volume, or formed during smelting and incorporated with the liquid metal on account of the greater specific weight of the oxide with respect to the molten bath.

The metal is usually protected against oxidation by smelting it under a bath of salt or salt mixture having a higher melting point and a lower specific weight than the aluminum or alloy be melted so that, when smelting has been carried out, the salt bath floats over and covers the molten metal bath.

The types of furnaces employed for smelting, resmelting and recovery of aluminum, affording results which are at present considered to be the best, are rotary furnaces with a'constant speed of rotation and reverberating (pit) furnaces.

These results are not, however, fully satisfactory, either in respect of the metal smelting loss, nor in respect of its degree of purification from oxide and degassing.

I have ascertained through various experiments and laboratory as well as commercial tests that during smelting of aluminum and its alloys, a number of factors either alone and more particularly jointly, as well as the power of the heat source employed and the presence of a protecting salt mixture, are essential, in order to obtain the highest smelting output and best properties of the resulting metal, said essential factors being as follows:

(a) the power of the heat source being equal, the manner of transmitting heat from the source to the salt mixture and charge, with special reference to variation with time of the heat quantities transmitted and of the temperatures at which said transmissions take place;

(b) the relative state of the metal charge and salt mixture during the smelting process and, more particularly, during the stage beteween incipient smelting and total liquefaction;

(c) the way in which the salt mixture affects the aluminum oxide contained in the metal charge.

In smelting of aluminum under the protection of a salt mixture in a rotary furnace, all these factors act interdependently of one another, largest influence being due to the first mentioned one.

In the course of my research work I could ascertain that the metal loss by smelting is indirectly proportioned to the speed at which the charge to be smelted covers the interval in temperature between room temperature and a temperature near incipient melting of metal, and to the oscillations in temperature of the charge in a unit of time between incipient melting temperature and total liquefaction and, sinks as the always rapid increase in temperature of the bath from liquefaction to pouring temperature is more gradual, continuous and free from sharp changes.

These improved results may be obtained because, by proceeding in the above described manner, during the first stage the substances (moisture, fats, hydrocarbons, nitrides, nitrates, carbides, carbonates, hydrates, etc.) incorporated in the superficial oxide film covering the metal are rapidly decomposed and volatilised by subtracting the latter quickly from the objectionable action of these substances. During the first stage, moreover, the growth and thickening of the superificial oxide film is minimised by an increased extent as the temperature of the solid metal is raised till it reaches the incipient melting temperature.

During the second stage, in which rapid liquefaction takes place, a further purification of the metal from the superficial oxide is effected by minimising the formation of new oxide films after elimination of the pre-existing ones.

During the third stage the absorption of gas by the molten metal bath is minimised.

j These conditions cannot be fulfilled in smelting processes presently in use for the following reasons.

(a) In order to smelt quickly the furnace should be provided with a source of heat of high power which transmits its heat at a considerable difference in temperature with respect to the charge.

The salt mixture being, however, a poor conductor of heat, prevents rapid transmission of heat to the metal, more particularly during the first stage, when the still solid mixture is subdivided into lumps by thin air films which further increase the resistance against heat transmission. Moreover, this mixture due to considerable superficial over heating strongly evaporates and becomes burnt.

(b) In order to obtain the highest heat efiiciency from the source for rapid melting, the furnace should be provided with a satisfactory heat insulation in the form of a thick fireproof lining, this preventing a gradual and controlled rise in temperature during the final stage of the process.

According to this invention the optimum conditions described above may be obtained by conveniently varying the thermic inertia of the furnace at each stage of the process.

I shall now explain what I understand by thermic inertia and the manner of varying it.

I understand by thermic inertia of a system the characteristic aptitude of the system to tend to maintain its status quo of heat balance under external heat stresses,

to oppose the action of further heat stresses which tend to give rise to a variation in conditions of pre-existing heat balance and to hold up variation of its thermic state, temperature, quantity of stored heat, quantity of absorbed heat yielded or transmitted in a time unit under the action of further external positive heat stresses (heating) or negative stresses (cooling) other than those with which the system was previously balanced.

The greatness of thermic inertia of a system may be evaluated from the greater or smaller delay incurred 'in 'varying' its thermic state on variation of outer stresses.

Thelgreaterthe delay in reaching the new state of M, mass of the system K, total coefficient of heat conductivity C, specific heat of the system T, absolute temperature As a result:

Thethermic inertia of a system increases as its mass M grows, .its total coefficient of heat conductivity K sinks its specific heat C and the absolute temperature T rise.

The following example will further explain the meaning of thermic inertia.

On introducing simultaneously into a furnace for heat .treatment, say to 500 C., one kilogram of sheet aluminum millimeter thick, and a cube of aluminum weighing likewise one kilogram, the sheet aluminum will reach a temperature of 500 C. throughout its width and thickness after five minutes, while the aluminum cube reaches this uniform temperature throughout only after about one hour, although the quantity of heat necessary for heating to 500 C. one kilogram of aluminum sheet is the same as required by the aluminum cube of the same weight, and their respective heat capacities are the same.

The considered aluminum cube is of a thermic inertia twelve times greater than the same weight of millimeter sheet aluminum.

Various effects derive from thermic inertia and may be differently utilised.

For instance, in transmission of heat from one source of heat S at a temperature ii to a substance Q at a temperature t3 lower than tr through an intermediate wall P, if the thermic inertia of P is lowest, the quantities of heat supplied by S shall be transmitted with a minimum delay to Q and at a temperature t2 of P very near ts and, consequently, also in small regularly supplied quantities proportioned to the difference in temperature tz-ta, which is negligible.

If, on the contrary, the thermic inertia of P is great, the quantities of heat supplied by the source S shall be transmitted to Q with a greater delay and, owing to the consequent rise in temperature t2 of P, at a temperature t2 of P which rises nearer to -t1 as the thermic inertia of P increases; consequently, even if delayed the quantity of heat transmitted is greater in time unit proportionately to the greater difference in temperature iz-ta.

The systems distinguished by a very low thermic inertia admit of an easy and quick adjustment of temperature and of the quantity of heat transmitted in a time unit, while the systems of high thermic inertia may serve as intermediate heat accumulators capable of supplying at .given moments and in a time unit quantities of heat considerably greater than those that may be supplied by the source of heat, which is thereby advantageously supplemented in case of need.

It should be pointed out that in the first mentioned case of lowest thermic inertia of the wall P, owing to the short delay between reception of heat from the source S at a temperature t1 and yielding of the corresponding quantity of heat to the substance Q at a temperature tz lower than t1, the intermediate wall P cannot accumulate large quantities of heat; consequently the temperature t2 of the wall P cannot rise considerably and the difference l2t3 between the temperatures of P and Q is negligible.

Vice versa, in the case of a high thermic inertia of the wall P, before reaching the heat balance with the substance Q and yielding to the latter the quantities of heat absorbed from the source S, P accumulates considerable quantities of heat, its temperature t2 rises and comes nearer the temperature t1 of S, and the difference t2t3 becomes considerably greater.

It may therefore be stated that in the case of heat transmission between a source of heat and a substance to be heated through :an intermediate body, the thermic inertia of the intermediate body may be evaluated not only on th basis of the delay incurred in transmitting heat, but also of the difference in temperature which is established between the intermediate body and the substance to be heated, this difference in temperature increasing as the thermic inertia of the intermediate body uses.

Since th thermic inertia depends upon K, M, C, T, it will be obvious that the thermic inertia of a system may be made variable with time, if one or more factors K, M, C, T are conveniently varied.

The above will be clearly explained by describing my improved process for smelting aluminum under a bath salt in a rotary furnace of variable thermic inertia, in which variation in thermic inertia of the furnace is obtained by varying, according to the various stages of the process and in the most suitable manner, the mass M of the furnace walls which are involved in transmitting heat from the source to the charge.

The invention shall be described with reference to the accompanying drawings, wherein:

Figure l is an axial vertical section of a rotary furnace according to my invention,

Figure 2 is an axial vertical section of a modification,

Figure 3 is a section on line III-III of Figure 2,

Figures 4, 5 and 6 are diagrams showing the variations of thermic inertia in the furnace according to Figures 2 and 3,

Figures 7, 8 and 9 are diagrams showing the variations of thermic inertia in a modified furnace construction.

Figure 10 is an axial vertical section of a third furnace construction, and

Figure ll is a horizontal section on line XIXI of Figure 10.

In the example shown in Figure 1 a sheet metal drum 1 carries externally two race rings 2, that may be centered on the drum by means of the bolts 3, said race rings rolling with the drum 1 on rollers 4 by which the rotational movement may be transmitted.

The inside of the drum is lined with a heat insulating layer of sutficient thickness for maintaining the temperature of the outer jacket constant and near the room temperature. A refractory lining 6 is arranged within the heat-insulating layer and is of sufiicient thickness for effectiveness of the heat-insulating layer whatever the temperature inside of the furnace is.

Internally of the refractory lining 6 a drum of cast iron 7 of great thickness (10 centimeters) is arranged.

Concentrically with the drum 7 I arrange a further cast iron drum of mean thickness (about 5 centimeters) and concentrically with the latter I arrange a third cast iron drum of small thickness (2 centimeters). The diameters of the three cast iron drums are such as to leave a small clearance 10, 11 between each of them of about 5 millimeters, the said clearances accommodating spacers 12 of heat-insulating material arranged at given intervals for the purpose of centering the three drums relatively to one another.

An axial hole is bored in the front wall of all the drums, including the fire-proof lining, heat-insulating and sheet metal jacket of the furnace and serves as furnace mouth.

Along the contour of the furnace mouth three cast iron cylinders are kept spaced by two heat-insulating tight-seating rings 14, 15, so that the two clearances 10, 11 are tightly sealed.

Each of the two clearances, which permit rotation of the furnace, communicates by means of

tubes

16, 17 externally of the furnace with either a vacuum means of

cocks

18, 19 or with the cocks 20, 21. The mouth of the furnace may be closed by means of a door 22 lined with fire-proof material and mounted on a carriage 27 movable on rails 28 in the axial direction of the furnace. A burner 23 is mounted on the door 22 and, in order to increase the effect of the variation in thermic inertia, is arranged in such manner as to direct the flame, during the smelting process, obliquely upwards and towards the portion of the inner cast iron drum which during rotation of the furnace is about to plunge under the molten charge. A hole 24 is bored in the lower portion of the door (22) and communicates through a movable channel 25 and a stationary channel 26 with the flue.

In order to obtain the best results from variation in thermic inertia of the furnace, it is convenient for the bore in the inner drum 9 of the furnace to be about 4 to 6 times greater than the greatest depth of the molten bath, and for the diameter of the drum to be 0.4 to 1.4 times its length.

While in this manner the capacity of the furnace is proportionate to the volume of the charge to be melted, it is convenient for the power of the heat source to be greatest consistently with the size of the furnace and kind of the heat source.

The smelting process is as follows.

The mixture of salt should in a molten state be of a fluidity such as to avoid a viscous and expansible consistency, to prevent during rotation of the furnace the bath of molten salt from adhering to the furnace walls and form a continuously pliant lining adhering to the furnace walls.

The salt mixture is introduced into the rotary furnace described above in a quantity equalling approximately the weight of the metal to be melted, rotation of the furnace at constant speed is started and heating is commenced by adjusting the source of heat to its maximum value and directing the flame obliquely upwardly and to the left if the furnace rotates anti-clockwise.

At this stage the clearances 10, 11 are filled with air, while the

cocks

18, 19, 20, 21 are kept closed. In this manner the thermic inertia of the furnace is greatest. The heat supplied by the source of heat at a constant rate is not directlyabsorbed by the salt charge (temperature t3), both because the latter is separated from the hottest flames, by the layer of coolest combustion products re flected by the bottom wall furnace and directed over the charge surface towards the opening 24, and because the greater quantity of heat supplied by the source is ab sorbed by the wall of the drum 9 and is transmitted through the thin layer of air in the clearance 11 to the drum 8 and from the latter through the layer of air in the clearance 10 to the drum 7. The rise in temperature t3 of the salt charge is therefore delayed till the three cast iron drums have accumulated a quantity of heat such as to rise in temperature from t2 beyond is by a steady though slow increase.

When the salt charge reaches its point of incipient melting, the cock 19 is opened and the clearance 11 is evacuated. This practically cuts off heat transmission from the inner drum, and the other two drums which are also insulated externally by the refractory and heat-insulating linings substantially maintain all the heat stored and the temperature attained over the period of the successive stages.

In this manner, the thermic inertia of the bottom is lowest, because the wall of the inner drum which is alone involved in the further stages by the heat transmission process is thin (reduction of the mass M to a minimum valve).

Heating of the molten salt is pursued till it is superheated to a maximumof 100200 C. beyond its melting point. Melting and superheating of the salt bath proceed rapidly because the thin wall of the inner drum pufi, surrounding air through 6 absorbs and quickly yields heat during uniform rotation of the furnace from the source to the charge (minimum thermic inertia). Similarly, adjustment and control of temperature are very easy and burning as well as objectionable evaporation of considerable salt quantities are avoided.

Heating is held up, the metal charge, for instance aluminum alloy turnings, to be melted is introduced into the furnace, rotating the furnace steadily on, till the metal charge is amalgamated with the liquid salt and cools it, forming therewith a solid conglomerate, of which the temperature is intermediate between the temperature of the cold metal and bath of molten salt, and in any case lower by about 200 C. than the temperature of incipient melting of the salt.

Since the furnace is of a lowest thermic inertia during this stage, the heat balance is quickly established, so that even the wall of the inner drum cools down rapidly, acquiring the temperature of the conglomerate.

Heating is applied again and adjusted to highest power, and the flame is obliquely directed as mentioned above, the cock 19 is closed, the cock 21 is opened till the clearance 11 is filled with air, and the cock 21 is re-closed. In this manner, the heat transmission between the drum 9 and the other two drums being re-established, the thermic inertia of the furnace becomes greatest (increase of mass M), the inner drum receives large quantities of heat from the source of heat, as well as the heat previously stored at a temperature considerably higher in the masses of the two outer drums, and appreciably rises in temperature, transmitting to the conglomerate during rotation of the furnace large quantities of heat at a considerable difference in temperature, which is characteris tic of a high thermic inertia. Conditions of heat transmission are therefore considerably improved notwithstanding the low transmission coefiicient of the salt surrounding the turnings and air cushions separating the lumps of conglomerate. As heating proceeds, the conglomerate quickly overcomes the temperature gap separating it from incipient melting temperature.

When the conglomerate reaches its incipient melting temperature, the cock 18 is opened and the clearance 10 is evacuated, whereby the outer drum 7, which has in the meantime yielded most of the heat previously stored therein, is thermically insulated from the two internal drums and the furnace is now of a mean thermic inertia (mass M reduced to less than one-half).

On pursuing heating and rotary movement the charge melts entirely and separates into two layers, viz. melted salt above and metal bath beneath.

At this stage, the cock 19 is opened and the clearance 11 also is evacuated. The furnace now acquires a minimal thermic inertia (mass M reduced to less than A3). Heating is pursued till the metal bath reaches the desired pouring temperature (50 to 200 C. higher than melting temperature).

At this stage of the process which is adjusted to minimum thermic inertia, the difference in temperature between the inner furnace walls and metal bath is smallest, however, transmission of heat is quick on account of the transmission coeflicient between walls and melted-metal which is considerably higher than that existing during the stage of highest thermic inertia between the walls and conglomerate.

When the bath has reached its predetermined pouring temperature, rotational movement is stopped, heating is held up and metal is poured through the pouring spout 29.

The thermic inertia of a system may be varied in manners other than mentioned above.

In order to further clear what is going to be described I will supply the following preliminary explanations.

I understand by static heat balance, the balance established in a system under the action of outer heat stresses constant in time, and by dynamic heat balance the balance corresponding to outer stresses variable in time and, the latter case including only conditions of transmission and heat balance existing at a given moment;

I understand by specific thermic inertia of a system its characteristic corresponding to attainment of a static heat balance which will be measured for instance from the delay incurred by the system in reaching a new state of heat balance under the action of new heat stresses that are constant and different from those with which the system was already in a state of balance.

I understand by instantaneous thermic inertia of a system the inertia set up during conditions of dynamic balance.

When the-system is submitted to the action of variable external heat stresses periodically varying between a maximum and minimum, the system periodically varies its thermic state, but with a phase displacement with respect to outer heat oscillations which increases as thermic inertia rises.

Considering now the case in which the half-period of oscillations of external heat stresses, that is, the time these oscillations take during their periodic variation to reach their highest value starting from the immediately preceding minimum value, is shorter than the time required for the system under consideration, starting from a state of balance with the minimum value of external oscillations, to reach, as a function of its heat inertia, the thermic balance with the maximum value of the said external stresses, further considering that the system is of a high thermic inertia due to a large mass M and, above all, to a low coefficient of conductivity K.

In this case, the shorter the period of variations of outer heat stresses, the less capable is the system to follow them even with delay and to reach a heat balance throughout its mass M with the extreme maximum and minimum values of outer heat stresses.

Under these conditions, part only of the total mass M, which is referred to as part mass M and of which the thickness, starting from the surface of the system exposed to the action of outer heat stresses, decreases as their period of oscillation becomes shorter, is affected by these outer stresses and similarly undergoes periodic heat oscillations. Moreover, as the period of the outer heat stresses is shortened, the lower the maximum temperature attained by the exposed surface of the system sinks the width of the oscillation of temperature of the exposed surface about a mean value is reduced, the delay of said oscillations with respect to that of outer oscillations is shortened that is, their phase displacement, hence their instantaneous thermic inertia are reduced, the said part mass M ultimately behaving like the mass of a new system of thermic inertia lower than the specific thermic inertia of the total system, which succeeds in following with a shorter delay the variation in outer heat stresses. Moreover, said outer heat stresses may be considered from a practical standpoint and in respect of the new system of which the mass is M to be made up of fictitious heat oscillations of the same period of oscillation as the actual ones, but of reduced width, assuming as maximum and minimum temperatures of the oscillations of the said imaginary stresses, the maximum and minimum temperatures attained by the system of which the mass is M during its consequent periodic oscillation.

Finally, it may occur that, in the specific case in which the outer heat oscillations undergo periodic oscillations of variable frequency and provided the half-period of said oscillations is shorter than the time the whole system would take on account of its specific thermic inertia to attain the state of static thermic balance with the maximum absolute value of the outer stresses starting from a previous static balance with anabsolute minimum value of. said stresses, the variation in instantaneous thermic inertia of the system becomes automatic as a result of the corresponding automatic variation in part massM' of said system.

More particularly, owing to, the variation in. the. part mass Mp indirectly proportionate to the frequency of periodical oscillations of the outer heat stresses, the instantaneous value of the thermic inertia is likewise indirectly proportioned to the frequency of said oscillations. At the same time, the system may be considered practically, more particularly in respect of its. part mass Mp as subjected to the action of fictitious outer heat stresses of a width of oscillation and of maximum values of the temperatures attained all the more reduced with respect to mean temperature and as compared with actual values of the said stresses, the greater their increase in frequency is.

The. above indicates further manners of. varying the thermic inertia of a rotary furnace for carrying out the novel process.

Figures 2 and 3 show a rotary furnace, consisting of a sheet metal drum 30, lined inside with a layer of insulating bricks 31 and a refractory layer 32 with the interposition between the two layers of a packed heatinsulation 13. The drum is rotated by means. of the race rings 33 on rollers 34. Axial holes 35, 36 are bored in the side walls of the drum and serve when open as months for the furnace.

During operation the. two

months

35, 36 are closed by the

doors

37, 38, respectively, mounted on

carriages

39, 40 movable on

rails

41, 42 in, the axial direction of the furnace, permitting by this movement opening and closure of the furnace mouths.

The furnace is provided with a source of heat of considerably higher power than ordinarily employed in rotary furnaces of the conventional type for melting aluminum under a salt bath. Moreover, the source may transmit its considerable heat power at a very high temperature. In the present case, the source of heat is represented by the arc struck between a pair of

graphite electrodes

43, 44. reaching within the furnace through openings arranged eccentrically in the closing doors.

In order to enhance the effect of variation in thermic inertia of the furnace, both

electrodes

43, 44 should be arranged obliquely and, preferably, eccentrically with respect to the furnace axis, so that the arc strikes in proximity of the wall which during rotation of the furnace is about to plunge beneath the melting charge.

The heat power of the arc may be two to five times greater than those generally installed in furnaces of the same capacity.

Two

holes

45, 46 are bored in the lower portion of the

doors

36, 37, 38 and communicate with two

channels

47, 48 through which a gas stream controlled by two valves 49, 50 may enter the furnace and issue therefrom.

The melting process is as follows.

The salt mixture is introduced into the rotary furnace as described above in a quantity equalling approximately the weight of the metal to be melted, rotation of the furnace is started at constant speed as usual, heating is commenced by setting the arc to maximum power and a gas stream is simultaneously caused to flow through the furnace chamber. Combustion products of other furnaces may be utilised inexpensively. The gas stream at a temperature lower by ZOO-to 800 C. thanin direct proximity of the arc (temperature 25003000 C.) enters for instance through 15 andflows on account of its lower temperature over the surface of the salt charge and protects it against direct radiation of the heat from the are before issuing through the opening 16. During this stage the furnace walls slowly absorb (great thermic inertia) large quantities of heat while the temperature of the surface of the refractory lining considerably rises.

After a given interval of time, when it would be objectionable (excessive power of the source). for the surface of therefractory lining to further rise in temperature, for. its-high. thermic: inertiawould delay PC1169 trationvof the-heat through: then-thickness of 'thei masonry;

heating is held up during a period of time equaling the former interruption. The inner surface of the refractory lining is thereby submitted to the cooling action of the inside of the furnace, of the gases and cooler charge. After a similar period of time, the arc is struck and cut out after a similar period, heating being applied and held up periodically till the salt reaches its incipient melting temperature. When the salt charge reaches its incipient melting point, the thermic inertia of the furnace is reduced to a minimum, increasing the frequency of the successive periods of operation and interruption of the are. More particularly, the period during which the arc is connected, is reduced to a minimum interval, such that, during the period of operation of the are, the rise in temperature of the inner surface of the masonry cooled in part during the period immediately preceding interruption of heating, becomes slightly higher than the corresponding temperature of the charge and is kept within a range such that at the end of this stage of the process (melting and over heating of the bath salt) the difference in temperature between the inner surface of the masonry and the molten super-heated salt bath, is lower than the admissible allowance in temperature (about 50-100 C.) with respect of the maximum heating temperature of the salt bath (about 850 C.). Heating of the salt bath is pursued in this manner till the bath is over heated by about 100 C. above its melting point. Heating is interrupted and the gas stream is cut off, the furnace mouths are opened and the metal charge to be melted is introduced into the furnace, further rotating the latter. Upon completion of charging, the furnace mouth is closed and the furnace is rotated further, till the metal charge amalgamates with the liquid salt, cools the latter and forms therewith a solid conglomerate at a temperature intermediate between the temperatures of the cold metal and of the molten salt bath, in any case lower by about 200-300 C. than the incipient salt melting temperature. The gas stream is again admitted and heating is repeated at periodic intervals which are again adjusted in such manner that the furnace is during this stage at its highest thermic inertia. For greater clearness, the maximum degree of thermic inertia which should conveniently be reached at this stage of the process, may be defined by the equivalent rule of the highest admissible temperature on the inner surface of the refractory linings (1300l500 C. according to the quality of the refractory) in the following manner.

The most convenient maximum thermic inertia of the furnace is that at which delay in reaching the heat balance between the furnace walls and charge brought to its incipient melting temperature, equals the time required for heating the charge from its starting temperature on adjustment of the furnace to maximum inertia up to incipient melting temperature.

In other words, the most convenient maximum thermic inertia is given when the furnace walls start yielding back to the charge, by virtue of the greater mass of the furnace walls which has slowly been affected by the stress of the heat source and the greater difference in temperature between the walls and charge, large quantities of the heat previously absorbed from the source of heat when the charge is about to reach its incipient melting temperature.

When the conglomerate reaches its incipient melting temperature, the thermic inertia of the furnace is reduced to an intermediate value between the maximum and minimum values by increasing the frequency of the successive periods of operation and interruption of the are, respectively.

Heating is periodically pursued and the furnace is rotated further till the charge is fully melted and separates into two layers, viz. molten salt above and metal bath beneath.

The thermic inertia of the furnace is then reduced to a minimum by increasing the frequency of the successive periods of operation and interruption of the are, so that the difference in temperature between the inner surface of the masonry and that of the molten bath, when the latter reaches its pouring temperature, is lower than the allowance in temperature (50 to C.) with respect to the predetermined pouring temperature of the metal.

Heating of the furnace set to minimum thermic inertia is pursued till the metal reaches its predetermined pouring temperature (from 50 to 200 C. higher than the melting temperature), whereupon rotation of the furnace is stopped, heating is held up and the metal is poured.

For the sake of clearness the degree of lowest convenient thermic inertia to be attained during this last stage of the process may be defined by the equivalent rule of the diiference in temperature between the furnace walls and metal bath within the temperature allowance admitted on the pouring temperature of the metal, as well as in the following manner.

The most suitable thermic inertia of the furnace is the value at which, when the charge has reached a given temperature and heating from the source of heat is interrupted, the successive increase in temperature of the charge due to heat yielded with delay by the walls of the furnace on account of its instantaneous thermic inertia, is lower than the allowance admitted on the predetermined temperature reached by the charge on interruption of heat supply from the source of heat. The reasons for variation of the thermic inertia of the furnace will be obvious.

In Figures 4, 5 and 6, X denotes the wall of a furnace, Y the source of heat, Z the metal charge, W the protecting salt. In Figure 4 it is assumed that the source of heat of considerably higher power than reasonably admissible in view of the proportions of the furnace and transmitting heat at very high temperature, is continuously operated throughout the process. The curve A is a circular diagram of the temperatures at the various points on the inner surface of the chamber referred to instantaneous temperature of the charge to be melted. For instance the line (To) denotes the difference in temperature between the point 0 and charge W. The dash line S which is assumed to coincide with the outer wall of the chamber, is the line connecting the points of the wall which are at that moment at 100 C.

Under these conditions, the furnace is at its specific very high thermic inertia on account of the thickness of the fire-proof lining necessary for a satisfactory heat insulation, so that the outer surface of the masonry is kept at a temperature of 100 C. The inner surface of the masonry is considerably over heated, exceeding the softening temperature of the refractory lining, the salt bath becomes burnt on account of excessive overheating and a large quantity thereof evaporates, appreciably lowering its protecting and purifying action, and the metal bath undergoes a highly injurious final overheating.

In Figure 5 it is assumed that the source of heat is intermittently operated as described in connection with adjustment of the furnace to its instantaneous thermic inertia which is considered to be the highest convenient one for that stage of the process.

The furnace walls are therefore subjected to the action of periodically oscillating outer heat stresses and the part mass Mp1 of the walls which follows and undergoes their effect is confined between the internal surface and dot line S1 connecting the points on the wall which under the conditions described are at that moment at 100 C. The furnace corresponds from the standpoint of thermic inertia to a furnace of a smaller mass limited by the dotted line S.

The inner surface of the furnace does not undergo any injurious overheating and the furnace walls yield to the large heat quantities. Consequently, it may be imagined that the source of heat, from the standpoint of the part mass Mp1, is replaced by another one operating at a lower temperature as compared with its maximum actual one and nearer the maximum temperature attained by the inner furnace walls as frequency rises during intermittent furnace operation. This consideration and consequence of variableness of thermic inertia in the manner described permit of using heat powers considerably higher than usual and operating at very high temperatures which would otherwise be injurious.

In Figure 6 it is assumed that the source of heat is intermittently operated at a higher frequency as described in connection with the adjustment of the furnace to its instantaneous thermic inertia, which is considered to be the lowest convenient for that stage of the process. The part mass M 2 becomes considerably lower and is limited by the dotted line S2. As shown by the diagram, during this stage the walls rapidly yield the heat absorbed from the source and temperatures of the inner surface of the furnace and the charge become approximately equal. The furnace corresponds from the standpoint of thermic inertia to a furnace of wall thickness a1.

The advantages obtainable by the above described smelting process (variation of the thermic inertia of the furnace by variation of the part mass involved in heat transmission from the source of heat to the charge) may be enhanced by the concurrent effect of a salt bath, distinguishing by a melting point substantially equal to that of the batch to be melted, and invariable over the whole process.

The methods presently in use for melting and recovering residues of aluminum alloys proceed in the following manner. After the furnace has been preheated, the salt mixture is charged and melted as the furnace is rotated steadily at constant speed. In practice I adopt a speed of rotation of 1.5 to 2 revolutions per minute. The melting point of these mixtures is always higher than that of the aluminum alloy to be melted, both when the mixture is fresh and still more so when it has been melted several times owing to the alumina dissolved therein. When the salt mixture is liquefied, heating is interrupted and the metal charge, for example chips to be melted are introduced as the furnace rotates further. On account of the low heat conduction of the salt mixture, the chips produce its partial solidification and, on completion of charging, ail the turnings are amalgamated with the salt set around them. Heating is then resumed until the aluminum alloy has reached the pouring temperature. During the latter period, chip and salt blocks agglomerated together roll within the furnace; by reason of the heat supplied, the chip and salt conglomerate melts, but aluminum will, of course, melt first and is pressed out of the conglomerate in small drops. These drops are immediately covered by an oxide fihn and on account of the continuous rotation of the furnace and consequent grinding action of the not yet melted conglomerate lumps uninterruptedly break and assemble again constantly increasing the oxidation. The optimistic reported data of low melting losses resulting from the known processes are mere illusions because the presence in the cast metal of a remarkable quantity of a finely subdivided oxide makes the practical results indicated deceptive.

Somewhat improved conditions might be obtained by employing salts with a melting point lower than that of the alloy. However, even in this case a remarkable quantity of oxide will be produced, since the chips of which the apparent volume is much larger than that of the salt, will at every rotation emerge from the bath into the furnace chamber periodically oxidizing themselves.

The salts employed in the present process should have definitely specified characteristics. Considering that the bath is a salt mixture, its melting point will not be a fixed one, but will be included within a certain range. In accordance with the above, it is essential that the higher 12 melting temperature limit be the same as that of the alloy to be melted. Moreover, considering that this higher temperature should be constant, it is essential that the alumina may not dissolve in the mixture, of which the melting point would otherwise vary, but that it should remain in suspension therein.

Therefore, the salts employed in previously known mixtures are not suitable for use in applicants process. Consequently, mixtures containing the following salts should be avoided: sodium carbonate, sodium hydrate, sodium silico-iiuoride, cryolite, sodium fluoride, calcium fluoride, aluminum fluoride, beryum fluoride, etc., because alumina dissolves in them. Furthermore, ammonium chloride, aluminum chloride, zinc chloride, manganese chloride, etc. should be avoided because of their unstableness.

As distinct therefrom, a mixture of sodium chloride, potassium chloride, magnesium chloride, calcium chloride, beryum chloride, strontium chloride, in different proportions, will have melting intervals of which the higher temperatures of ultimate melting may be lowered as desired to make them equal to those of the alloys to be melted, and will not contain unstable ingredients or solvents of A1203; they will incorporate the aluminum oxide in the form of a mechanical suspension, and keep their melting range constant and unvariable over the full period of the process.

With regard to constant speed of rotation, as mentioned above, this should conveniently be lower than normally adopted in rotary furnaces presently in use (1.5 to 2 rev./min.) and within a range of 0.25 to 0.75 rev./min.

In order to avoid not only a possible grinding effect of the conglomerate being melted on the aluminum drops, this action being annulled by the use of a salt mixture substantially equal in melting point to the alloy to be melted, but also that lumps of agglomerate in the incipient melting state may stick to the furnace walls and, before reaching a melted condition, may be carried along upwardly, out of the range of the protecting gas stream and exposed to the direct action of the source of heat, which is extremely injurious within a short range.

The consideration of the rotational speed of the furnace suggests a further manner of carrying out the present process, by varying the thermic inertia of the furnace.

In fact, during rotation of the furnace, the inner portion of the masonry is subjected to periodically variable heat stresses, more particularly heating stresses due to the source of heat (assumed to operate steadily at its maximum power) and cooling stresses by effect of the cold charge to be melted, with which the successive portions of the masonry alternately come into contact.

By varying the speed of rotation of a rotary furnace from any desired minimum speed to a top value of which the top limit is given by the speed at which the charge becomes centrifuged, because under these circumstances the charge would by effect of the centrifugal force steadily adhere to the furnace walls and operation of the latter would be equivalent to that of a stationary furnace without a relative motion between its walls and the charge, the instantaneous thermic inertia of the furnace may be practically varied from a maximum value corresponding to the lowest speed of rotation, down to a minimum value depending upon the highest speed attainable before the charge becomes centrifuged. The said variation in thermic inertia is due to the fact that during rotation of the furnace at variable speed, the inner surface of the masonry of the furnace is subjected to the action of periodically variable heating stresses due to the action of the source of heat and of cooling stresses due to the charge.

The period of the oscillations of the said heat stresses is dependent upon the speed of rotation of the furnace and their frequency equals the number of revolutions of the furnace per minute. The width of the oscillatioiis of the thermic stresses acting on the inner surface of the masonry is also variable in time, its maximum being determined by the characteristics, which are supposed to be kept constant, of the source of heat, while the minimum values vary dependently upon the steadily increasing temperature of the charge from a minimum value as the charge is introduced into the furnace up to a maximum value as the charge attains its pouring temperature.

These stresses which are periodically variable also in frequency and width, result in variations of the part mass Mp of the masonry of the furnace, which is determined by a thickness starting from the inner surface of the masonry, variable dependently upon the speed of rotation of the furnace and indirectly proportionate thereto, said part mass Mp undergoing periodic variations of its thermic state, thereby determining the extent of its instantaneous thermic inertia.

The period of the said oscillations of the part mass Mp is likewise variable dependently upon the speed of rotation of the furnace, the frequency of said oscillations equalling the number of revolutions per minute of the furnace. Their width varies steadily as a result of the variation in maximum temperatures attained on periodic oscillations of the temperature of the inner surface of the masonry, the said maximum temperatures varying in an indirect proportion to the speed of rotation of the furnace,

and as a result of the steady increase in temperature of the charge, in such manner that as melting of the charge proceeds, the speed of rotation of the furnace increasing adequately, the width of the thermic oscillations of the masonry about the temperature considered as mean temperature of the charge, as the latter reaches its predetermined pouring temperature steadily sinks down to a temperature within admissible range of allowance in respect of the maximum speed of rotation attainable.

Finally, the heat stresses due to the source of heat may be considered from the standpoint of the action on the furnace walls and charge, to be replaced by fictitious stresses equalling the actual ones in power, and as the temperature sinks nearer the temperature of the charge, when the latter is about to attain its pouring temperature, the rotational speed of the furnace increases.

This consideration leads to a further advantage by making possible the use of sources of heat of powers exceeding those normally adopted in rotary furnaces and operation thereof at high ranges of temperature without damaging either the masonry or the charge.

In Figures 7, 8 and 9, X is the well of a furnace, Y is a source of heat, Z is the charge, 4 is the protecting bath. In Figure 7 the furnace is assumed to be stationary. The curve A is a circular diagram of the temperature of various points of the internal surface of a chamber referred to the instantaneous temperature of the charge to be melted. So, for instance, the line To shows the difference in temperature between the point 0 and charge Z. The dash line S, which is supposed to coincide with the external chamber wall, is the line connecting the points of a wall which are at that moment at a temperature of 100 C.

In this case the thermic inertia is highest, the earth subtracts heat from the bath, the vault is overheated and a strong final heating of the bath is aiforded.

In Figure 8, the furnace is supposed to be imparted a slow rotation. The earth yields to the charge large quantities of heat, melting is expedited, the thermic inertia is of the mean value. Under the aspect of thermic inertia, the furnace equals a stationary furnace having a thickness s.

In Figure 9 the furnace is supposed to be rapidly rotated. As will be seen from diagram A, during rotation the vault receives but little heat, the earth yields little heat and the earth and charge temperatures are approximately equal.

The thermic inertia is minimized and temperature may be satisfactorily adjusted.

Under the aspect of thermic inertia, the furnace corresponds to a furnace of the thickness s1.

In order to increase the effect of variation in thermic inertia during the melting process and subtract as far as possible the charge from direct effects of the source of heat, contrarily to the process in rotary furnaces of the known type, it is convenient for the maximum diameter of the inner furnace chamber to be four to six times greater than the depth of the molten bath and for the ratio between said bore and the actual length L (that is, exclusive of the conical connections between the inner chamber of the above said maximum diameter and the side walls of the furnace or the bores therein) to be greater than that currently used in rotary furnaces.

Generally the ratio L/= 2 to 5 is adopted.

In order to obtain the best results by the present process it is convenient to adopt a ratio L/=0.4 to 1.4.

The purpose in view is to cause the heat transmission to the charge by direct conductivity from the furnace walls which plunge beneath the charge during rotation to prevail over transmission by radiation from the furnace walls overlying the charge.

Moreover, since the salt, which is a poor conductor of heat, by eifect of the radiating heat of the source is overheated to a greater extent at the surface, largely evaporates and becomes burnt, a higher sensitivity of the salt charge results in respect of variation in thermic inertia of the furnace walls, the eifectiveness of the salt charge being prolonged by interposing between the free surface of the salt and the source of heat a source of gas at a temperature lower than that of the source of heat, so that the gas stream determines a protection of the salt surface against radiating heat of the source of heat.

In case an electric arc or an electric resistance are used as a source of heat for the furnace, a gas stream may be advantageously employed, and the still hot cornbustion products of other flame furnaces may be used as such gas.

When a flame is used as a source of heat, the combustion products may be interposed, as protecting gaseous stream, by utilising the fumes, after they have cooled to a temperature lower than the flame temperature.

In order to enhance the effect of the thermic inertia, it is convenient for the source of heat to evolve its power in an eccentric position with respect to the axis of rotation of the furnace and in proximity of the part of the furnace wall which on rotation of the latter is about to plunge beneath the charge.

When a flame is used as a source of heat, it is convenient to bias the flame with respect to the axis of rotation of the furnace in a direction other than the one establishing the shortest Way between the inlet opening of the flame into the furnace and the outlet opening for the combustion products. It is preferable to direct the flame obliquely upwards during melting and towardspthe part of the masonry which on rotation of the furnace is about to plunge beneath the charge, and in such manner that, if the outlet opening is not promptly reached by the combustion products, at least part of them, after having partly cooled against the furnace walls, on rebounding against the side walls of the inner furnace chamber, is reflected downwardly and sucked through a suitable opening situated on the same side and underneath the flame inlet opening, whereby the combustion products, being cooler than the flame which should be set to a shorter size than the furnace length, form a protective gas stream for the salt charge against direct radiation by the flame.

The present process as integrally applied utilising automatic variation of the part mass Mp which varies dependently upon the speed of rotation of the furnace, for varying the thermic inertia, is described in the following examples.

Figures 10 and 11 show a rotary furnace comprising a sheet metal drum 51 carrying externally two race rings '15 52 that may be centered on the drum by means of bolts 53 and roll together with the drum 51 on rollers 54, which transmit to the furnace a movement of rotation imparted to the rollers by a motor provided with a device for varying the rotational speed of the furnace.

The drum is provided with an inner lining comprising three concentric layers, namely, an external layer 55 of insulating bricks, an intermediate layer 56 of packed refractory earth and in inner layer 57 of fire-proof bricks.

An axial cylindrical opening 58 is bored in one of the side walls of the furnace and serves as a charging opening and fume escape opening. In front of the opening 58 a detachable tubular connection 59 with a fire-proof lining is provided which, when facing the opening 58, serves to convey the fumes from the furnace to the flue 60; for charging the furnace, the connection 59, is removed, leaving the opening 58 free.

A gate 61 provided with a refractory lining is arranged between the opening 58 and connection 59 and serves for reducing or fully closing the opening 58.

A further conical axial opening 62 larger in diameter at its base than the opening 58 is bored in the opposite side wall, which serves at the same time for introducing the flames of a burner and evacuating at least in part the combustion products.

A movable device 63 is arranged in front of the opening 62 and consists of a carrying structure provided with a refractory lining and formed internally with two superposed conduits, of which the upper conduit 64 is tapered and faces by its larger base the top portion of the opening 62, while it carries at its smaller base a burner 65, said conduit serving as diffuser cone and guide for the flame. The axis of the truncated cone 64 is orientated in such manner that the flames may be directed during operation of the burner into the furnace biassed with respect to the axis of rotation obliquely upwards on the side on which the masonry of the furnace plunges during rotation beneath the charge to be melted.

The device 63 is provided beneath the tapered opening 64 and with the interposition of a layer of refractory material with a tubular connection or conduit 66 controlled by a valve 67. The connection 66, which faces at one end the lower portion of the opening 62 and communicates at its other end with the flue 60, serves for sucking through the opening 62 at least a part controlled by means of the valve 67 of the combustion products.

The present process is carried out in the following manner.

Assuming the charge to be melted is composed by thin sheet metal cuttings of aluminum-copper alloy, this alloy starts melting at 525 C. A mixture of salts of the following composition is suitable for the above-mentioned alloy:

Percent Sodium chloride 20.5 Potassium chloride 23.7 Calcium chloride 29.1 Barium chloride 21.5 Strontium chloride 5.2

in 20 minutes, in orderto be set to a maximum thermic" inertia, and a maximum speed of two turns per minute for setting to minimum thermic inertia. These minimum and maximum speeds have been determined by the above described method of determining the most convenient maximum and minimum thermic inertias for carrying out the process.

Generally, according to the methods described, in order to obtain the most convenient thermic inertia the furnace shall perform at least one to two turns during the period from the moment heating of the charge starts to the moment the latter reaches its incipient melting point, at a speed of rotation between one turn in 25 minutes and 0.7 turns/min. and complying with the requirement of preventing any objectionable grinding effect during rotation.

In order to obtain the most convenient minimum thermic inertia the speed of rotation shall be one to five turns per minute. The speed shall be the maximum practicable speed at which the charge cannot be centrifuged. Practically, a satisfactory adjustment of the furnace is obtained to a minimum thermic inertia meeting the above mentioned requirements when the maximum speed of rotation of the furnace reaches one tenth of the speed at which centrifuging of the charge in the furnace would occur, and is comprised between one and five turns per minute according to the diameter of the furnace and power of the heat source.

The heat power may be assumed to be 1.5 to 2.5 times greater than currently employed for furnaces of the same capacity.

The melting process is carried out in the following manner.

The salt mixture of the above described composition is introduced into the rotary furnace as described above in a quantity equalling about 2/ 3 up to 1/ 1 of the weight of the metal charge to be melted, in this case, 700 kgs. mixture.

Rotation of the furnace is started by adjusting the variable transmission to rotate the furnace at its minimum speed (one turn in 15 minutes) so that during this stage the thermic inertia of the furnace is at its maximum value. Heating is commenced by adjusting the burner to its maximum power (2,000,000 caL/h.) and directing the cone 64 by means of the movable device 63 brought near the opening 62 in such manner that the flame propagates within the furnace obliquely upwards and to the left, if the furnace rotates anti-clockwise and by setting the live flame in such manner that its maximum length equals the maximum length of the furnace.

The bore of the opening 58 is adjusted by means of the door 61 and the suction of the conduit (connection 66) is controlled by means of the valve 67 in such manner as to cause a gas stream to flow through the furnace chamber, the gas flowing over the surface of the salt charge and protecting it against direct radiation by the flame, causing at least part of the combustion products that have cooled in part against the furnace walls to reverse their direction and issue from the furnace at the lower portion of the opening 62, instead of issuing through the opening 58.

When the salt mixture reaches its incipient melting temperature, (470 C.) the thermic inertia of the furnace is gradually reduced to a minimum value, by increasing the speed of rotation, whereupon it is raised to its maximum value (two turns per minute) when the mixture has fully liquefied (525 C.).

Heating with the furnace set to minimum thermic inertia is pursued until the salt is overheated at the temperature of 725 to 750 C.

Heating and flow of the gas stream are then interrupted, the furnace opening is opened, the metal charge to be melted is introduced (1,000 kgs. sheet metal cuttings of aluminum-copper alloy),-settin g-the furnace during charging to its minimum thermic inertia, as maxi- 17 mum speed of rotation (two turns per minute) is maintained.

The furnace door is there'tpon closed and the furnace is rotated further at its maximum speed till the metal charge, by amalgamating with the previously overheated salt bath, cools it and forms therewith a solid conglomerate. The highest speed will assist uniform formation of the conglomerate and quick transfer of the accumulated superheating heat, the highest speed corresponding to lowest thermic inertia of the furnace producing during this stage the decomposition and volatilisation of the humidity, hydrocarbons, hydrates, carbonates, intrudes etc. incorporated in the superficial oxide film which covers the solid aluminum.

Heating is again applied by setting the burner to its maximum power, directing the flame again obliquely upwards to the left and maintaining such conditions till the end of the process. The gas stream is caused to flow again as during melting of the salt mixture till the end of the process and the furnace is set to its maximum thermic inertia, reducing the speed of rotation to a minimum value (one turn in fifteen minutes), increasing the melting rate to its practically highest value. This Will give the best results from the point of view of melting output, refining structure, compactness and grain fineness. In fact, by operating as above, the not yet liquefied charge is rapidly withdrawn from the injurious action of the products of decomposition of the substances incorporated in the superficial oxide film and the growth and thickening of this film is minimized, concurrently therewith, the lowest furnace rotational speed, which corresponds to the highest thermic inertia, minimizes the grinding action which is so detrimental in previous processes.

When the lower temperature of the melting range of the salt mixture is reached, the particles of the conglomerate begin to stick and unite together, instead of undergoing the self-grinding action peculiar to temperatures lower than the one mentioned above.

When the conglomerate has reached the incipient melting temperature, the thermic inertia of the furnace is reduced to an intermediate value by increasing the rotational speed of the furnace in order to bring it to the maximum speed, when the charge has fully melted.

Just before the melting point is reached, the whole bulk is like one mollified and plastic salt mass in which the sheet metal chips are suspended and uniformly distributed. On account of the above mentioned conditions and of the decreasing thermic inertia which characterizes this stage of the process, both the salt and incorporated sheet metal chips simultaneously melt over the whole mass. Metallic drops are formed in the melting salt mass simultaneously and out of contact with the gas within the furnace. Moreover, these drops slip as they are formed out of the oxide film which covers every aluminum sheet metal chip in a solid state, leaving these films in suspension in the salt mass.

When melting of the charge is completed, thermic inertia of the furnace is reduced to its minimum by imparting to the furnace its maximum rotational speed.

Heating of the furnace set to its minimum thermic inertia is pursued until the metal reaches its predetermined pouring temperature (which is 50 to 200 C. higher than the pouring temperature).

Rotation of the furnace is stopped, heating and gas flow are held up and the metal is poured.

While in previous processes melting of 1000 kgs. sheet metal chips of aluminum-copper alloy 8 to 10 millimeters thick give at the utmost 960 kgs. metal, the present process 994 kgs. perfectly purified metal are obtained.

According to a modification of the present process, variation of thermic inertia of the furnace may be obtained by replacing the continuous rotational movement by a discontinuous rotational movement, by alternating movements of rotation at variable angular speed and width comprised in variable fractions of a full turn with intermittent hold up intervals, during which the furnace is maintained stationary.

In order to set the furnace so maximum thermic inertia, the intermittent movements of rotation shall be short and, at will, of low angular speed so that, during each period of movement, the furnace performs small fractions of a full turn, while the duration of intervals therebetween during which the furnace is maintained stationary shall be of a period such as to permit any parts of the charge which have stuck to the furnace walls during the directly preceding fraction of a turn and slightly lifted over the level of the charge, to become loosened and fall towards the mass of the charge; moveover, the total time during which the furnace performs by its abovementioned discontinuous movement a full turn, should be equivalent to the time in which the same furnace should perform a full turn by a continuous movement of rotation in order to meet the requirements mentioned above for reestablishing the most convenient thermic inertia.

In order to reduce the thermic inertia of the furnace the width of discontinuous rotational movements is increased and, if desired, the angular speed of said movements, and the period duration of the intervals during which the furnace is kept stationary, always with the object of obtaining an equivalence between the number of turns of the discontinuous movement and the number of turns of the continuous movement.

Moreover, in' order to obtain the smallest values of thermic inertia and optimum minimum thermic inertia according to the rules mentioned above, it is possible to adopt a continuous rotational movement, as already described.

In case the salt or salt mixture employed as protecting and purifying bath for the aluminum is of a melting point or gap other than the alloy to be melted, it is possible to obtain superior results more satisfactory than those obtainable by the present processes, by suitably varying the thermic inertia of the furnace during the melting process. In this case, a modification of the proc ess permits to obtain best results provided. the molten salt mixture is of satisfactory fluidity and during all the stages of the process the centrifuging speed of. the charge is not exceeded, and during the stage ofthe process comprised between incipient metal melting and total liquefaction of the salt the speed at which the charge may undergo an objectionable grinding effect is not reached.

The modified process is as follows.

I introduce into a rotary furnace of variable thermic inertia in one of the manners described a salt mixture of a satisfactory fluidity in a molten state and a specific weight lower than the alloy to be melted (for instance, an aluminum alloy containing 13% of silicium, meltmg temperature 565 C.) and :a melting gap other than the alloy to be melted (for instance, the mixture given in the previous example of a melting gap between 470 and 525 C.) in a quantity equalling the weight of the charge of metal to be melted (for instance, 1,000 kgs.).

Rotation of the furnace is started maintaining the rotational speed during the Whole process at a value lower than corresponding to centrifugation of the charge.

Heating is commenced adjusting the source of heat to its maximum power and directing it into the inside of the furnace to cause it to develop its heat power in a position eccentric to the axis of rotation of the furnace and in proximity of the portion of the wall furnace which during rotation of the latter is about to plunge beneath the charge, and, at the same time, a gas stream is caused to flow through the furnace chamber at a temperature lower than the source of heat and the thermic inertia of the furnace is adjusted so that during this initial stage it is of mean value (for instance by rotating the furnace at a speed ranging between 0.5 and 2.8 rev./min.).

When the salt mixture reaches its incipient melting temperature the thermic inertia of the furnace is reduced,

adjusting it to the minimum value (for instance by ro- 19 tating the furnace at a speed ranging from 2 and rev./min.), when the mixture has fully liquefied.

Heating is pursued in this manner, the furnace being set to its minimum thermic inertia till the salt bath is overheated (maximum temperature 850 C.).

Heating and the gas flow are then reduced to a minimum and may even be held up when the metal charge to be melted (for instance, 1000 kgs. castings of A1+l3% Si alloy of an average 3 millimeter thickness) is introduced, adjusting the furnace during charging to its minimum thermic inertia and maintaining said adjustment till the metal charge has amalgamated with the salt bath and has cooled it, forming a conglomerate therewith.

Heating is started again, adjusting and directing the source of heat, as already mentioned, for melting the salt mixture, the gas stream is again caused to flow and the furnace is set to its mean thermic inertia.

Heating is pursued till the metal reaches its incipient melting temperature. Thermic inertia is then adjusted to its maximum value (0.04'0.7 R. P. M.) and is maintained such till the temperature of full liquefaction of the salt mixture is reached. Thermic inertia of the furnace is then reduced to its minimum value and the process is carried out further till the metal reaches its predetermined pouring temperature, whereupon the metal is poured.

While previous processes will yield from 1,000 kgs. castings of an average 3 millimeter thickness a maximum of 950 kgs. re-melted ingots, the above described modified process gives 996 kgs. ingots, while if a salt mixture of a melting point substantially equalling that of the alloy (for instance, NaCl 30% +KCl 5% +CaClz 30% +BaCl2 31% +StrCl2 4%rnelting temperature 565 C.) the nonmodified process gives 998.5 re-melted ingots.

What I claim is:

1. In a method of melting aluminum and aluminum alloys in a rotary furnace the walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a suitable salt mixture; heating the mixture to its incipient melting temperature while maintaining the thermic inertia of the furnace at its maximum value; reducing the thermic inertia and setting it at its minimum value when the salt mixture is fully liquefied; over-heating the salt at minimum thermic inertia of the furnace, reducing the heat supply and introducing into the furnace the metal charge to be melted; agglomerating the metal charge with the molten salt bath while maintaining the thermic inertia at its minimum value until metal salt conglomerate is formed; raising the thermic inertia of the furnace and starting heating again to bring the conglomerate to its incipient melting state; reducing the thermic inertia and setting it to the minimum value when all the charge is fully liquefied; pursuing heating of the molten charge until the metal reaches a predetermined pouring temperature; cutting off the source of heat and pouring the molten metal.

2. A method according to claim 1, and in which the rises and falls, respectively, of the thermic inertia of the furnace are obtained by lowering and raising, respectively, the frequency of periodic oscillations of heat stresses transmitted from said source of heat to the walls of the furnace.

3. A method according to claim 1, and in which the rises and falls, respectively, of the thermic inertia of the furnace are obtained by lowering and raising, respectively, the rotational speed of the furnace.

4. In a method of melting aluminum and aluminum alloys in a rotary furnace the walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a suitable salt mixture; having a solidification interval equal to that of the metal to be melted;

heating the mixture to its incipient melting temperature while rotating the furnace at a speed of 0.04-0.7 rounds per minute; increasing the rotary speed of the furnace and setting it at a value of l to 5 R. P. M. when the salt mixture is fully liquefied; overheating the salt at said last specified speed of the furnace; reducing the heat supply and introducing into the furnace the metal charge to be melted; agglomerating the metal charge with salt bath while maintaining the said last specified speed of the furnace until metal-salt conglomerate is formed; reducing the speed of the furnace down to the first specified value while starting heating again to bring the conglomerate to its incipient melting state; raising the rotary speed of the furnace to bring its value up to the second specified value when all the charge is fully liquefied; pursuing heating of the molten charge until the metal reaches a predetermined pouring temperature cutting off the source of heat; stopping the furnace and pouring the molten metal.

5. In a method of melting aluminum and aluminum alloys in a rotary furnace the walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a suitable salt mixture having a solidification interval other than that of the metal to be melted; heating the mixture to its incipient melting temperature while rotating the furnace at a speed of 0.5-2, 8 rounds per minute; increasing the speed of the furnace up to a value not higher than 5 R. P. M. when the salt mixture is fully liquefied; overheating the salt at said last specified speed of the furnace; reducing the heat supply and introducing into the furnace the metal charge to be melted; agglomerating the metal charge with the salt bath while maintaining said last specified speed of the furnace until metal-salt conglomerate is formed; reducing the speed of the furnace down to the first specified value while starting heating again to bring the conglomerate to its incipient melting state; reducing the rotary speed of the furnace to a value comprised between 0.04 and 0.7 R. P. M. until all the salt mixture is liquefied; raising the said speed up to a value of 15 R. P. M. until all the charge is liquefied; pursuing heating of the molten charge until the metal reaches a predetermined pouring temperature; cutting off the source of heat; stopping the furnace and pouring the molten metal.

6. Method of melting aluminum and aluminum alloys as claimed in claim 1 and further comprising moreover the steps of interposing between the source and melting charge a gas stream at a temperature lower than the source of heat but higher than the charge during heating of the salt charge until its melting and during heating of the salt-metal agglomerate until its melting.

7. Method as claimed in claim 1, and further comprising moreover the steps of adjusting the source of heat to its maximum power and directing it into the furnace chamber in such manner that the quantity of heat transmitted directly by the source to the charge is considerably smaller than the quantity of heat transmitted indirectly to the charge by the walls of the furnace chamber during heating of the salt charge until its melting and during heating of the salt-metal agglomerate up to its melting.

8. In a method of melting aluminum and aluminum alloys in a rotary furnace the walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a salt mixture having a solidification interval other than that of the metal to be melted; adjusting the thermic inertia of the furnace to its maximum value by raising the part mass of the furnace walls and heating the mixture to its incipient melting temperature; lowering the thermic inertia of the furnace by reducing the part mass and setting to minimum thermic inertia when the salt mixture is fully liquefied; overheating the mixture at minimum thermic inertia of the furnace; cutting-oil the source of heat and introducing into the furnace the metal charge to be melted in a quantity such that the heating time necessary for successively heating all the charged salt and metal conglomerate to its incipient melting temperature equals the time previously taken for heating the salt mixture to its incipient melting temperature; agglomerating the metal charge with the salt bath while maintaining the thermic inertia at its minimum value until metal salt conglomerate is formed; setting the thermic inertia at a medium value and starting heating again to bring the conglomerate to its incipient melting state; reducing the thermic inertia when the salt mixture is fully liquefied and setting it to the minimum value when all the charge is fully liquefied; allowing the charge to separate into salt and metal layers while pursuing heating until the metal reaches the predetermined pouring temperature; cutting off the source of heat and pouring the molten metal.

9. In a method of melting aluminum and aluminum" alloys in a rotary furnace the Walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a salt mixture having a solidification interval other than that of the metal to be melted; starting rotation of the furnace and maintaining its speed of rotation during the whole process at a value lower than the centrifuging speed of the charge; heating the mixture by subjecting it to heat stresses at a minimum frequency until an incipient melting temperature of the mixture is reached; increasing the frequency of the heat stresses and bringing it to its maximum value when the salt mixture is fully liquefied; overheating the mixture; cutting off the source of heat and introducing into the furnace the metal charge to be melted in a quantity such that the heating time necessary for successively heating all the charged salt and metal conglomerate to its incipient melting temperature equals the time previously taken for heating the salt mixture to its incipient melting temperature; agglomerating the metal charge with the salt bath until metal salt conglomerate is formed; starting heating again while lowering the heat stress frequency to bring the conglomerate to its incipient melting state; increasing said frequency when the salt mixture is fully liquefied and setting it at its maximum value when all the charge is fully liquefied; pursuing heating until the metal bath reaches its predetermined pouring temperature; cutting off the source of heat, stopping rotation of the furnace and pouring the molten metal.

10. In a method as claimed in claim 8, further steps of interposing between the source of heat and the charge a gas stream at a temperature intermediate between the source and charge during heating of the salt mixture until its melting and during heating of the conglomerate until its melting.

11. In a method as claimed in claim 8, the steps of heating the salt mixture until its melting and the conglomerate until its melting while adjusting the source of heat to its maximum power and controlling the heat evolved in such a manner that the quantity of heat transmitted directly by the source to the charge is considerably smaller than the quantity of heat transmitted indirectly to the charge through the furnace walls.

12. In a method as claimed in claim 1, wherein the depth of the charge is comprised between A1 and ,4; of the maximum inner diameter of the furnace.

13. In a method of melting alumnium and aluminum alloys in a rotary furnace the walls of which are involved in heat transmission from the source of heat to the charge to be melted under a salt bath, the steps of: introducing into the furnace a salt mixture having a solidification interval other than that of the metal to be melted; starting rotation of the furnace and maintaining it between 0.5 and 2.8 revolutions per minute; adjusting the source of heat to its maximum power and controlling the heat evolved in a manner that the quantity of heat transmitted directly to the charge is smaller than the quantity of heat transmitted indirectly to the mixture through furnace walls and interposing between the source and the mixture a gas stream at a temperature intermediate between the source and mixture; increasing the speed of the furnace to a value of 1 to 5 revolutions per minute when the salt mixture is fully liquefied; overheating the mixture while maintaining the rotational speed between 2 and 5 revolutions per minute; substantially reducing heating and flow of the gas stream and introducing into the furnace the charge of metal in a quantity such that the heating time necessary for successively heating all the charged salt and metal conglomerate to its incipient melting temperature equals the time previously taken for heating the salt mixture to its incipient melting temperature, and adjusting the total quantity of the charge in a manner that in melted condition the depth of the bath is four to six times lower than the maximum inner diameter of the furnace; agglomerating the charge while maintaining the revolutional speed between one and five revolutions per minute; starting heating again while passing a gas stream between the source and charge and while rotating the furnace at a speed between 0.5 and 2.8 revolutions per minute until the charge reaches its incipient melting temperature; reducing the speed to an interval between 0.04 and 0.7 revolutions per minute until the salt mixture is fully liquefied; increasing the speed so as to bring it to an interval ranging between 1 and 5 revolutions per minute when all the charge is fully liquefied; pursuing heating until the metal bath reaches the predetermined pouring temperature; cutting off the source of heat, stopping rotation of the furnace and pouring the molten metal.

References Cited in the file of this patent UNITED STATES PATENTS 451,405 Langley Apr. 28, 1891 1,380,767 Booth June 7, 1921 1,904,781 Crawford Apr. 18, 1933 2,128,444 Vroonen Aug. 20, 1938 2,312,811 Gentil Mar. 2, 1943 2,337,072 Tarbox Dec. 21, 1943 2,382,723 Kirsebom Aug. 14, 1945 2,387,014 Gibson Oct. 16, 1945 2,599,158 Brassert June 3, 1952 FOREIGN PATENTS 404,467 Italy June 15, 1943

Claims

1. IN A METHOD OF MELTING ALUMINUM AND ALUMINUM ALLOYS IN A ROTARY FURNACE THE WALLS OF WHICH ARE INVOLVED IN HEAT TRANSMISSION FROM THE SOURCE OF HEAT TO THE CHARGE TO BE MELTED UNDER A SALT BATH, THE STEPS OF: INTRODUCING INTO THE FURNACE A SUITABLE SALT MIXTURE; HEATING THE MIXTURE TO ITS INCIPIENT MELTING TEMPERATURE WHILE MAINTAINING THE THERMIC INERTIA OF THE FURNACE AT ITS MAXIMUM VALUE; REDUCING THE THERMIC INERTIA AND SETTING IT AT ITS MINIMUM VALUE WHEN THE SALT MIXTURE IS FULLY LIQUEFIED; OVER-HEATING THE SALT AT MINIMUM THERMIC INERTIA OF THE FURNACE, REDUCING THE HEAT SUPPLY AND INTRODUCING INTO THE FURNACE THE METAL CHARGE TO BE MELTED; AGGLOMERATING THE METAL CHARGE WITH THE MOLTEN SALT BATH WHILE MAINTAINING THE THERMIC INERTIA AT ITS MINIMUM VALUE UNTIL METAL SALT CONGLOMERATE IS FORMED; RAISING THE THERMIC INERTIA OF THE FURNACE AND STARTING HEATING AGAIN TO BRING THE CONGLOMERATE TO ITS INCIPIENT MELTING STATE; REDUCING THE THERMIC INERTIA AND SETTING IT TO THE MINIMUM VALUE WHEN ALL THE CHARGE IS FULLY LIQUEFIED; PERSUING HEATING OF THE MOLTEN CHARGE UNTIL THE METAL REACHES A PREDETERMINED POURING TEMPERATURE; CUTTING OFF THE SOURCE OF HEAT AND POURING THE MOLTEN METAL.