US2247370A - Powder metallurgy - Google Patents

Description

July 1, 1941. c. e. GOETZEL POWDER METAIILURGY 2 Sheets-Sheet 2 Original Filed Oct. 18, 1938- INVENTOR 606526 Z BY MM Ciao; 64/622562 K g .l

ATTORNEYS Patented July 1, 1941 POWDER IVIETALLURGY Claus Guenter Goetzel, New York, N. Y., assignor to Hardy Metallurgical Company, New York, N. Y., a corporation of Delaware Original application October 18, 1938, Serial No.

1939, Serial No. 276,527

6 Claims.

This invention relates to powder metallurgy andparticularly to the heat treatment of masses of metal powder particles to bond them together into a continuous metal structure. tion contemplates an improved method and apparatus for determining or establishing and maintaining appropriate conditions for such heat treatment. 7

This application isa division of my copending application Serial No, 235,597, filed October 18,

It has been customary heretofore to heat masses of metal powders, particularly coherent masses or briquettes that have been compacted by pressure with or without the aid of 'a binder, in order to establish metallic bonds between the particles. The briquettes of metal powder may be heated to the point of incipient fusion so that grain boundaries are destroyed, but such practice (especially if the surfaces of the'briquettes are relatively unsupported or unconfined) tends to bring about undesirable distortion. Adequate bending of metal powder particles may also be accomplished by maintaining the briquette at an elevated temperature but below the fusion point until suiiicient diffusion of metal occurs between particles. This is known as diffusion welding and is preferred when articles of accurate configuration are to be made because it involves less distortion of'an unsupported briquette, i. e.,

one that is heat treated while unconfined in a mold. Optimum conditions for difiusion welding vary depending upon the metallic constituents and thedegree of consolidation of the briquette prior to heat treatment. Hence, it is frequently necessary to carry out a series of experiments with various time periods and temperatures in order to determine a desirable set of conditions for heat treating a given kind of briquette. Even then optimum conditions may not be determined unless a very extensive series of tests is made.

As a result of my investigations, I have developed a method and apparatus whereby optimum or desirable conditions for heat treatment of any given mass of metal powder may be determined quickly and accurately and without re- ;sort to the cut and try expedients of the prior art. The method is based upon my observation thatamass of metal powder undergoing The inven- Divided and this application May 31,

heat treatment exerts a varying influence on an alternating current flowing in a circuit with which the mass is inductively associated. Thus, if the metal powder mass, preferably a briquette, is heated while disposed as the core of an induction coil energized by alternating current, I have found that the effective inductance of the coil changes gradually until the mass reaches a temperature at which diffusion welding occurs rapidly and thereafter remains substantially constant until the mass is near to being fused. Changes in the impedance or effective inductance of the coil may be observed by noting variations in either the characteristics of the circuit or the character of the alternating current or both which the changes bring about, and my invention contemplates making such observations and employing them as an index and ultimately as a control of proper conditions of heat treatment.

Thus, in accordance with my invention a mass of metal powder particles which are to be bonded together into a more or less continuous metal structure is subjected to heating while disposed in inductive relationship with a conductor in a circuit energized by an alternating current, and

the changes in the characteristics of the circuit Y or the character of the current, or both, are noted as the temperature of the mass is raised. Depending upon the circuit employed, the impedance (or effective inductance) change may be manifested by a change in any one or more of a number of current characteristics, such, for example, as amperage, phase relationship between amperage and voltage waves, wave form, frequency or wave length. One or more of these characteristics is observed while the heating progresses so that an index to the effective inductance of the conductor (coil) is obtained. When the mass attains a proper temperature for good diffusion welding, the effective inductance of the conductor which changes as the temperature rises, becomes substantially constant and the aforementioned variations substantially cease. When this'occurs the temperature of the mass may be observed. Having determined this temperature for a specimen briquette, like briquettes may without further testing be raised to and maintained at this-temperature with the assurance that diffusion welding will occur relatively promptly and thoroughly. The same procedu e m y be used as a basis for automatic heat treatment control by employing means for maintaining the temperature of the mass substantially constant or for interrupting the heating or both when the effective inductance of the coil reaches a constant value.

In a preferred modification of my invention, the temperature of a metal powder briquette is raised while it is disposed in inductive relationship with a conductor (preferably a coil in which the mass is disposed as a core) energized by an alternating current, the frequency (or conversely, the wave length) of which varies with the effective inductance of the conductor. The frequency or wave length of the current is determined as the temperature is raised. When the frequency (or wave length) attains a substantially constant value, the proper treatment temperature has been reached, and the mass should be maintained at or about this temperature until adequate diffusion has occurred.

The mass may be heated by any appropriate means such as a furnace heated by fuel or electrical resistance. However, as explained in greater detail hereinafter, I prefer to raise the temperature of the briquette by means of eddy or short circuit currents induced within it by means of the field set up by the alternating current flowing in the conductor. These eddy currents preferably are of high frequency, because (as disclosed by Charles Hardy in his copending application Serial No. 233,387, filed October 5, 1938) the presence of these eddy currents in the briquette not only brings about a rapid rise in its temperature but also accelerates the rate at which diffusion welding occurs. Automatic means may be employed for controlling the current input into the apparatus when the frequency of the current of the high frequency induction coil becomes constant to the end that the heat treatment of the briquette proceeds at a proper temperature. These and other features of my invention will be more thoroughly understood in the light of the following detailed description taken in conjunction with the accompanying drawings, in which Fig. 1 is a diagrammatic representation of apparatus of my invention adapted to the determination of the proper treatment temperature of a powder metal mass in order to bring about diffusion welding; and

Fig. 2 is a modified form of the apparatus of Fig. 1 adapted for automatic heat treatment control.

Referring now to Fig. 1, the apparatus comprises a spark gap type high frequency generator or converter I energized by a source I l of alternating current of commercial frequency, say 60 cycles per second, and commercial potential, say 220 volts. The source is connected to a stepup transformer l2 having a primary coil l3 and a secondary coil IS. The primary is connected with the power source through a. switch 14 and a variable reactance MA. The secondary coil l5 of the transformer is connected to an oscillatory circuit including a high frequency induction coil 16 of a furnace IT, a condenser l3 and a spark gap [8; The spark gap is in series with the coil [5 in a low frequency circuit ISA and is also in series with the coil H5 in a high frequency circuit ISA. The condenser is connected between the spark gap and the coil IS. The capacitance of the condenser is chosen to provide the required wave length for the furnace and type of briquette to be used.

The high frequency induction coil should be of low ohmic resistance and may be made conveniently as a helix of copper tubing through which a cooling fluid is circulated from a source (not shown). Disposed within the induction coil and preferably concentric therewith is a mufile made of refractory non-conductive material such as quartz and having end portions extending substantially outside the ends of the coil. One end portion is provided with an integrally formed observation port 2| closed by a quartz window 22. Below the port is an inlet nipple 23 adapted to be connected to a source of nonoxidizing, inert, or reducing gas in order that a suitable atmosphere may be maintained in the muflie. The other end of the mufile is closed by a removable but tight fitting plug 24 provided with an outlet nipple 25 through which the gas may be withdrawn either to evacuate the muffle or assure a circulation of gas therethrough. The plug is also provided with a seal 26 through which projects a thermocouple 21 that extends to theheating zone within the muffle, and is connected to a pyrometer type voltmeter 21A.

A metal powder briquette 28 to be heat treat ed is disposed within the muffle adjacent the hot junction of the thermocouple.

Frequency may, of course, be determined by observing wave length. A radio wave meter 29 is provided. It has a pick-up coil 30 placed within the field of the induction coil. The pick-up coil is connected in series with a rectifier 3| and an indicator 32, such as a hot wire ammeter. A variable condenser 33 for tuning purposes is shunted across the pick-up coil. Any suitable wave or frequency meter may be employed in place of the simple one illustrated.

It is important that the means provided for supplying high frequency current to the induction coil be such that a. change in the impedance or effective inductance of the coil brings about a substantial change in the wave length of the high frequency circuit. The oscillatory circuit illustrated in Fig. 1 is only one of a number which will permit change of the impedance to affeet the wave length.(and conversely the frequency) but it is particularly well adapted to the present purpose because it permits a current of suitable wave length (say of the order of 3000 meters, or less) to be employed, and thus shortens markedly the time for heating the briquette up to optimum temperature and also shortens the following interval during which the eddy currents complete the diffusion welding.

In the operation of the apparatus of Fig. 1, one or more relatively cold briquettes (formed by compressing a mass of metal powders in any of the heretofore customary manners into a coherent mass of desired configuration, such for example as a gear) is placed in the muflle, and a suitable inert, reducing or non-oxidizing atmosphere, say one of dry hydrogen, is established within the mufile, although with certain materials an atmosphere of air or a vacuum may be employed. The induction coil is then energized by the high frequency alternating current. As a result, myriads of eddy currents are induced in the briquette. The temperature of the briquette rises rapidly. As the temperature rises the setting of the variable condenser of the radio wave meter is altered by the operator to keep the meter in tune with the current of the nduction coil. .It will be observed that the wave length gradually decreases, which is to say that the frequency gradually increases as the briquette is heated, until at a certain temperature dependent upon the metallic constituents and degree of consolidation of the briquette, a substantially constant wave length is attained. When this condition is reached, the temperature of the briquette is determined by means of the .thermocouple. This temperature is the one at whicn heat treatment should be conducted.

When it is attained, the variable reactance on the input side of the transformer is increased to cut down the power input to the coil and consequently the heat supplied to the-furnace. The power input is reduced until the heat input approxlmately equals the heat losses from the furnac so that no further rise in temperature of the briquette occurs, and the briquette is held for a short time, usually only a few minutes, until diffusion welding is complete. With certain metals having a high diffusion rate, welding is complete practically as soon as the wave length attains a constant value, and the current may be turned off the induction coil immediately by means of the switch. Otherwise the heating should be continued at the reduced current input for a time which will vary depending upon the type of briquette employed, and will in general be short for difiicultly fusible metals and relatively longer for low-melting point metals and When the only quantity sought is the allows. optimum treatment temperature or the constant wave length and adequate welding is unnecessary, the current may be turned oil assoon as the temperature of the wave length has been observed.

My invention is applicable to the heat treatture. The change of efiective inductance of the induction coil is apparently due to a change in the nature of its core, which is the briquette undergoing heat treatment. At first the briquette is composed of numerous particles or small asgregates of particles at least partially insulated from each other by films of occluded or adsorbed gas, or films of oxidation products or binder. When an appropriate temperature is reached, however, welding begins to occur very rapidly,

especially when diffusion is accelerated by high I trated in the following tests, in which the induction coil comprised about 60 turns of inch 0. D. water cooled copper tubing wound in a helix about 3 inches through. The coil was energized by a Lepel high'frequency spark gap converter (type C-6 in Test 1, type -2 in Tests II to V, inclusive) which took current. of 60 cycles, 220 volts and converted it to current of ment of briquettes of any metals or metal mixslightly less than 3000 meters. The wave length of the current in the coil was determined with a standard wave meter which had no direct connection to the energizing current, but received waves on a pick-up coil placed about 18 inches from the induction coil. Amuiile of the type described hereinbefore was placed within the were subjected to physical tests and microscopic coil, and the thermocouple w'as disposed in the muille for determining the temperature of .the

briquettes.

From electrolytically deposited copper powder of minus 150 mesh and with relatively clean surfaces, twelve briquettes inch square and 3 inches long were made by compression in a mold under a force of 50 tons per square inch. The briquettes thus formed were removed from the mold and placed in the muilie separated by thin layers of alundum powder. An atmosphere of dry hydrogen was produced in the furnace and the current was turned on. The following table gives temperatures of the briquettes, wave lengths and frequency of th current employed to heat the briquettes, and the power input in kilowatts.

Durlng this and the following tests, the voltage Tempem' Wave Fre uenc Power Y, Time, minutes g ggs: length, eye as per 1%? t meters second Watts 20 2590 115, 800 2. 1 593 2560 117, 200 2. 1 782 2550 117, 700 1. 8 815 2540 118, 200 1. 8 827 2540 118, 200 1. 8 827 2540 118, 200 1. 8 827 2540 118, 200 Oil 1) 1 Current supplied. i Current reduced. Current of]. Briquettes cooled in muflle.

It will be observed that when the briquettes reached a temperature of approximately 815 C., the wave length of the energizing high frequency current attained a substantially constant value of 2540 meters and that additional heating didnot substantially change the wave length. HOW-u ever, if the briquettes had been heated to the point at which they became plastic and distortion occurred, further reduction in wave length would have been encountered. The bars which were made from the briquettes examination and were found to be mechanically sound and comparable in strength to bars of cast copper.

'rrsr No. II.-Heat treatment of copper-tin 1111- quettes to produce bronze bars Copper powder parts by weight) and tin powder (10 parts by weight) in a dry condition sure a uniform mixture. The mesh size of the powder mixture was all minus (Tyler scale); The mixture was pressed into briquettes inch square by 3 inches long in molds under a force of 50- tons per square inch and twelve of the briquettes were heat treated in the muflie. An atmosphere of dry hydrogen was maintained in the mufile and the briquettes were separated from each other by thin layers of.alunclum powder. The following table gives values for time,-temperature, wave length, frequency and power input during the heat treatment.

1 Current on.

1 Current increased.

1 Current decreased.

4 Current 03.

I Briquettes cooled in muflle.

It will be noted that in Test No. II, the conditions were substantially like those in the previous test except that the optimum temperature of heat treatment was approximately 821'C. instead of 815 C. for copper, and that the presence of tin was apparently effective in reducing the wave length.

The resulting bronze bars were subjected to physical tests and microscopic examination. It was found that the copper and tin had alloyed and formed a homogeneous mass. Segregations of impurities such as oxides, etc., originally present upon the surfaces of the powder particles had occurred. The segregations were substantially spherical in form and did not aflfect appreciably the strength of the bars, which was comparable to that of bars made from cast bronze. In short, the products 01 the test were in every respect satisfactory.

Tzs'r No. ,III.- -Heat treatment of iron powder briquettes The briquettes in this case were made from iron powder obtained by hydrogen reduction oi. iron oxide. The powder had a mesh size of minus 150 and had clean metallic surfaces. The briquettes were formed under a force of 50 tons per square inch. Twelve of the briquettes in relatively unsupported condition, 1. e., with unconfined sur- 'faces, were placed in the muilie separated by layers of alundum powder and subjected to the e1- iect of the eddy currents produced as described hereinbeiore. The following table gives data observed during the progress of the heat treatment. During heat treatment, an atmosphere of. dry hydrogen was maintained within the-muille.

Temperaw Power ture of the W in t Time minutes len th. are per tg meters second T #533 1 20 2830 106, 100 6. 0 3. 593 2700 111,211) 6.0 6 871 2610 115,001 6.0 6 L 1060 2570 116, 800 4. 7... 1093 2560 117,10 4. 5 10 1009 2500 117, 31) 4. 6 13 1093 2560 117, 200 4. 5 17 1099 2660 117,11) Oi! 51 4 2) I ower om 1 gower redub dby regulating reactence. Power oil.

4 Briquettes cooled in mutlle.

peared, and no evidence 01' distortion was apparent.

' Trsr No. IV.-Heat treatment of iron nickel powder mixture to produce binary alloy Nickel powder (4% by weight) produced from nickel carbonyl and iron powder (96% by weight) produced by hydrogen reduction of iron oxide were mixed in a dry condition for eight hours. The mixture had a mesh size of minus 150 and was formed into twelve briquettes inch square by 3 inches long under a force of 50 tons per 'square inch. The twelve briquettes, separated from each other by layers of alundum powder, were disposed in the muille in an atmosphere of dry hydrogen and subjected to heating by eddy currents induced in them. The following table givesdata observed during the progress 01' the heat treatment.

The resulting bars of iron nickel alloy were entirely satisfactory. No appreciable distortion occurred in heat treatment so that the bars wereoi accurate configuration. The micrographic structure and the physical properties, such as tensile and compressive strengths, of the bars were good.

It will be observed that the optimum treatment temperature was in the neighborhood of 1100' 0., as indicated by the thermocouple disposed within the muflle, and that this temperature corresponded to a wave length of 2510-2520 meters. The power input of 4.5 kilowatts which was established after six minutes of treatment was sumcient to maintain the briquettes approximately at optimum treatment temperature. In other words, enough heat was generated by this power to make up for the heat lost by conduction, convection and radiation from the briquettes.

Txsr No. V.Heat treatment of powder metals to produce ternary stainless steel alloys In this test, the powder mixture was made from iron powder produced by the hydrogen reduction of iron oxide and nickel powder produced irom nickel carbonyl, plus finely divided low carbon ierrochrome having a chromium content of approximately 'l0%. The mixture was all minus 100 mesh in size and most of it was minus 200. The mixing period was twelve hours. The

briquettes were made in molds under a compres-' sive force of tons per square inch. The twelve briquettes were nrst heated in the muiile in an atmosphere of dry hydrogen. Approximately 15 Y minutes were consumed in raising the briquette The resulting iron bars were entirely satisfactory. They were reduced to f of their original to optimum heat treatment temperature which was maintained for about an hour. Thereafter, the were allowed to cool for about 40 minutes in the muiiie. At room temperature these bars were repressed under a force of '10 tons per square inch and then annealed in the tained for about an hour in an atmosphere of dry hydrogen. The briquettes were then cooled immediately by quenching them from their elevated temperature in water. The log of the primary heat treatment is as follows;

Temperature (C.)

asob- Tempera- Time 3511.? $1513? Wave Frequency 531% 1 minutes optical cated by ggig i gg kilopyrometer thermowatts through couple muflie port 20 2820 106,400 5.0 150 2800 107,200 5.0 310 2770 108, 300' 5. 440 2760 108, 700 5. 0 590 2740 109,600 5.0 750 2730 109.900 5.0 880 2710 110,800 5.0 990 2690 111,600 5.0 1070 2670 112, 300 5. 0 1100 2660 112,800 5.0 1130 2640 113,700 5.6 1170 2620 119, 600 5. 6 1225' 2600 115,400 5.6 1% 2590 115,900 6.0 1310 2570 116, 800 6.0 1340 2550 117, 300 6. 0 1390 2530 118,600 6.0 1395 2520 119. 100 4. 1400 25%) 119, 100 4. 5 1400 2520 119,100 4.5 1395 2520 119,100 4. 5 1400 2530 118,700 4.5 1410 2520 119.100 4. 5 1400 2530 118, 700 4. 5 1390 2530 118.700 4.5 1390 2530 118, 700 Off 1240 1 Power on.

1 Power increased by reducing reactance.

1 Power increased again by reducing reactance.

4 Power decrewed by increasing reactance.

5 Power off.

' Bars cooled in muillc.

The bars produced in these tests were excellent. Their corrosion resisting properties were high. Their mechanical properties were comparable in every respect to those of.stainless steel produced by processes involving fusion. There was no distortion.

As the foregoing results show, optimum treatment temperatures corresponded .to a wave length of 2520-2530 meters. Optimum treatment temperature was in the neighborhood of 1400 C. The temperatures observed with the optical pyrometer and the thermocouple did not check exactly, but it is believed that the optical pyrometer readings are less exact in the high temperature range at which the work was done.

When my invention is employed for control purposes it is not necessary to observe the temperature at which frequency and wave length become constant, as will be apparent from the following description of the apparatus of Fig. 2, which is a modification of that of Fig. 1, like rent flow of the meter circuit approaches a maximum. which will be when the meter circuit is tuned to'resonance with the current in the high frequency induction coil. The relay is connectable in the wave meter circuit in place of the indicator by a double pole double throw switch 4 I. Three leads 42', 43, 44 are provided on the outlet side of the relay. The outer lead 44 is connected to an adjustable tap 45 on the coil of the variable reactance 14A. The center lead 43 is connected to the lower end of said coil and the inner lead 42 is connected to the power source through a time switch 46 which is also connected to the center lead. The time switch may be of any suitable form designed to open a switch after a predetermined interval. For example, it may consist of a solenoid connected in the lead 43 which pulls out a clock catch when it is energized, thus permitting the clock to start turning a switch to a cut-out point.

In the operation of the apparatus of Fig. 2, the appropriate wave length value for the heat treatment of a given specimen powder mass is determined preliminarily either with the apparatus of Fig. l, o by heating the specimen in the apparatus of Fig. 2 with the indicator 32 connected in the wave meter circuit through switch while adjusting the variable condenser 33 so that this circuit is tuned and a maximum current value is obtained. The relay is adjusted to operate under the influence of a current approaching this value, and the switch 4| is thrown to connect the relay to the meter circuit; the tap on the coil of the variable reactance is adjusted so that when the relay operates the power input to the furnace will be just enough to balance heat losses and hold a constant heat; and the time switch is set so that it will operate after an interval sufficient to permit adequate bonds to be formed between the particles of the briquette. With the apparatus thus set, another briquette or a plurality thereof is introduced into the muffle; current is applied by closing the switch 14; and thereafter heat treatment control is automatic. Thus, when the current in the high frequency coil attains a constant wave 1 length, an increase in current in the meter cirparts of the apparatus being designated'by the same numerals as in Fig. 1. Hence, difliculties in obtaining exact readings of high temperatures may be eliminated.

The apparatus of Fig. 2 differs from that.of

cuit causes the relay to operate. The relay, which. at rest makes a contactbetween inner lead 42 and outer lead 44, operates to break this contact and etsablish one between lead 42 and middle lead 43. When this occurs, the time switch is energized and begins to operate and at the same time the variable reactance is increased so that the power input to the high frequency induction coil is decreased to the proper degree. The temperature of the briquette is thus held constant at the "proper point for a predetermined interval of time until the time switch opens the circuit, deenergizes the high frequency induction coil and allows the briquette to cool. The settings of the variable condenser in the wave meter circuit, the tap on the variable reactance coil and the time switch necessarily are predetermined for any type of powder metal briquette, but when once determined permit completely automatic control of heat treatment.

As indicated hereinbefore, determination of proper heat treatment conditions and automatic control thereof are preferably based upon observations of constant frequency of a current flowing in a circuit having a conductor in inductive relationship with a briquette to be heated. However, other manifestations of achange in effective inductance of the conductor may also termination and control of proper heat treatment conditions by such means.

From the foregoing test data it will be seen that the desirable or optimum treatment-tem perature indicated by the method of my invention (i.e., the temperature attained by a powder metal mass when changes in effective inductance of an inductively associated conductor substantially cease to occur) approaches but is lower than the melting point of the, metal mass or product which results from the bonding together oi' the metal particles of the briquette by diffusion welding. This temperature relation is shown by a comparison of the indicated proper treatment temperatures and the melting points of the products of the foregoing tests:

Desirable treatment temp eatui-e, Pmduct Copper Bronze Nickel steel... Btsinless steel..

It is not necessary substantially to exceed the indicated treatment temperature in order to obtain rapid bonding together of the particles, and distortion of the briquettes tends to increase as this indicated temperautre is exceeded. Consequently, efforts should be made to hold the mass at or about the indicated temperature and in no case should the melting point be approached too closely.

Iclaim: 1. In apparatus for heating a mass of metal powder particles to bringwabout metallic bonding therebetween, the combination which comprises a mass of metal powder particles to be heated, an induction coil disposed in inductive relationship with the mass, a source of high frequency current connected to said induction coil in a circuit which permits the frequency of the current to vary substantially with the effective inductance of the coil, means for determining the frequency of the current during the heating, means for determining the temperature of the mass when said frequency attains a substantially constant value during the heating and means for preventing the temperature of the mass from increasing substantially after the frequency has attained the substantially constant value.

2. In apparatus for heating amass of metal powder particles to bring about metallicbonding coil which permits the frequency of the current to vary substantially with the effective inductance of the coil, means for determining the frequency of the alternating current, and means for reducing the input of the current to the coil in response to the attainment of a frequency of substantially constant value as indicated by the frequency determining means during the heating.

3. In apparatus for heating a mass of metal powder particles to bring about metallic bonding therebetween, the improvement which comprises an induction coil disposable in inductive relationship with the mass, a source of high frequency current, a circuit connecting said source and said coil and such as to permit the frequency of the current to vary substantially with the effectiveinductance' of the coil, means for determining the frequency of the current, and automatic means for holding the temperature of the, mass substantially constant when the frequency of the current attains a substantially constant value.

4. Apparatus according to claim 3 in which the -means for holding the temperature of the mass .constant includes a device for reducing the input of high frequency current to the coil.

5. In apparatus for heating a mass of metal powder particles to bring about metallic bonding therebetween, the combination which comprises a mass of powder particles to be heated, an induction coil disposed in inductive relationship with the mass, a source of alternating current connected in a circuit to said coil, means operatively associated with the circuit for determining a change in the impedance of' said circuit while the coil is disposed in inductive relationship with the mass being heated and means-for changing the input of current to the coil in response to the change in said factor.

6. In apparatus for heating a mass of metal powder particles to bring about-metallic bonding therebetween, the combination which comprises a mass of metal powder particles to be heated,

an induction coil disposed in inductive relationship with the mass during the heating, a source of high frequency alternating current, a circuit connected to the coil and to the source and such as to permit the frequency of the current to vary substantially with the effective inductance of the coil, means operatively associated with the circuit for determining the frequency, means for reducing the input of electrical energy by the current to the coil in response to the attainment of a substantially constant frequency during the heating, and means for interrupting the supply of current to the coil at the end of a predetermined interval of .time after the frequency becomes substantially constant.

CLAUS GUENTER GOETZEL.

Images