US2583227A - Induction heat treating - Google Patents

Description

Jan. 22, 1952 s, NElDlGH 2,583,227

INDUCTION HEAT TREATING Filed Jan. 24, 1947 2 smams smam 1 IQ IO N If) ID ID RAY 5. NE/D/GH INVENTOR.

Jan. 22, 1952 s, NElDlGH 2,583,227

INDUCTION HEAT TREATING Filed Jan. 24, 1947 2 SHEETS-SHEET 2 FIE-n2- PUSH-PULL HIGH FREQUENCY OSCILLATOR s T\ INVENTOR.

Patented Jan. 22, 1952 INDUCTION HEAT TREATING Ray s. Neidigh, Elgin, Ill., assignor to Elgin National Watch Company, Elgin, 111., a corporation of Illinois Application January 24, 1947, Serial No. 723,921

4 Claims. (01. 219 47) 1 It is well known that various alloys may have their behavior modified by raising them to temperatures at which internal changes occur, and then cooling again under conditions for obtaining the desired characteristics. Among such heat treatments are so-called "annealing operations, in which an alloy is raised to a temperature at which all internal strains are relieved and internal changes in ferrous alloys occur by which the hardness disappears, and then a slow cooling is used so that these desirable characteristics may be maintained in the article, to permit easy machining, for example. Another heat treating procedure is the so-called hardening operation by which a ferrous alloy is heated to obtain a conversion of carbides from one to another, followed by a quick cooling past a critical temperature in order to fix the carbide content in the desired form for producing the hardness. Another-type of heat treatment involves the so-called tempering operations, in which an excessively brittle ferrous alloy has its temper drawn to increase the toughness or resistance to fracture, by heating to a closely controlled temperature, which is lower than the temperature utilized for annealing or hardening, and followed by a slow or quick cooling: the course of this temper drawing" usually being followed by observing the oxidation colors appearing at a cleaned surface of the article. Another type of heating is that incident to working, as where a wire is to be hot-drawn, a skelp or tape to be hot-rolled to thickness or to have hot-forming or hot-punching operations performed thereon, etc.

A factor which must be considered and observed during heat treatment, outside of the production or control of the desired type and distribution of carbides in ferrous alloys, is the size of grains or crystals which are present. Thus, it has been found that prolonged heating of a terrous alloy to temperatures in excess of 1450" F. is conducive of the development of large 'grain size, whereas shorter periods of heating or of maintenance at such high temperatures. are conducive of maintaining existent small Brain sizes. Since small grain sizes are preferable for many purposes, it is often desired to keep the grain size as small as possible, and to avoid having the material exposed to high temperature conditions over long periods of time. However, it has been found difficult in practice to attain this result because the heating is a time function of the rate and quantity of heat which is introduced to the alloy. Thus, when the heat is delivered into the body of the alloy by conduction, convection, and

2 radiation methods from externa1 hot bodies, the time required is often great enough to cause serious increase in grain size. When it is sought to heat a ferrous alloy by induction effects of a high frequency electric field, the heating occurs by the resistance of the material to the flow of eddy currents, and by the conversion to heat of the energylosses due to hysteresis in the; alloy. Thus, the more magnetic the material, the greater the hysteresis losses, and correspondingly so much the greater is the heating effect by hysteresis. However, when ferrous alloys are heated to a temperature known as the magnetic critical temperature, around 1450" F. for sprin steels, there is a change in the material so that its magnetic effects are lost. It is sometimes assumed that this is due to the conversion of magnetizable carbide into forms which do not have magnetic dipoles. The effect of this cessation of magnetic properties is that the hysteresis heatin losses cease, and the material is heated under a continuing high frequency field by eddy currents only. If the cross sectional area of the material is large enough, the eddy current heatin effects will be large enough to offset the heat losses of the material by conduction and radiation; and hence for articles of large cross sectional area, any desired temperature above the magnetic critical temperature can be obtained by the induction heating principles. With thin articles, such as spring tapes, it has been suggested that the inability to heat to the desired high temperature (e. g, above magnetic critical for steel) is due to cancellin out of magnetic orientations of the material by overlappin of the eddy current fields within the cross-section.

. When the cross sectional area of the article to be'heated is small, the number of lines "of magnetic force through the material is less, and hence theeddy current effects are less. On the other hand, the ratio of superficial area to volume increases with decrease of area of the similarly shaped bodies, so that the heat losses at the surface are relatively increased.

When it is sought to heat tapes of various metal alloys, capable of heat treatment, suchas in preparing the springs employed for watches .gas under low pressure.

ture of the ferrous alloy or an insuflicient temperature with other alloys. In order to have assurance of the desired heat treatment with ferrous alloys, the temperature maximum should not be limited to the magnetic critical; and with these and other alloys, temperatures up to that of fusion should be capable of production at closely determined levels, in order to follow a determined schedule of treatment. On the other hand, the input of energy, as heating effect, should be at a maximum rate, well in excess of heat losses, in order that the length of time at the higher temperatures shall be a minimum, whereby to avoid grain growth.

It has now been found that by employing and controlling certain inherent characteristics of induction fields which have conductive bodies therein, it is possible to carry the temperature to any desired level, even to the production of actual .fusingrof the tape; and .to accomplish this by (simple and easily controlled apparatus.

An illustrative form of apparatus, by which the procedure may bepracticed, isillustrated .on the accompanying drawings .in which Figure 1 is a conventionalized illustration of the assembly of mechanical apparatus.

Figure 2 is a conventionalized showing of the electrical circuitconnections for Figure 1.

Figure 3 is a detail view, on a largerlscale, of the furnace structures of Figures ,1 and 2.

.In these drawings, the apparatus andmethod are illustrated with respect to .the heat treatment and quenching of a mainspring tape for a watch. In particular,.a so-called 8/0 size tape .of watch mainspring steel will be employed as :a specific example: this particular tape was 0.0044 inchthick and 0. '065inch wide.

Azsupply reel 10 of such tape delivers its contents through an input measuring and feeding assembly conventionally shown as rollers LI, and overguide rollers l2, with the formation of a depending loop or bight 13 in which is supported a weighted roller 1 4 and serving to regularize the feeding and tensioning of the tape T asit moves through Ithefurnace. From the last guide roller I2, the tape passes over a guide roller l5 which preferablyis .of copper for electrically conductive relation with the tape. A pressure roll 16 is pivotallyimounted and energized from acounterweight IT .for maintaining the tape in proper guided and conductive relationship with the roller 15.

From the roller 15, the tape passes vertically downward in the illustrated form and enters the upper end of a furnace chamber .28 whichin the illustrated .form is provided by .a-glass tubing of temperature-resistant glass, such as Pyrex.

.Near the upper end of the furnace tubeifi, it is provided with a side branch 2| through which maybe introduced carbon dioxide or other inert The tube 20 is sur-- rounded .by an induction unit U which is described in detail below,and which is supplied with .high frequency {electric current from the terminals 23, so that .ahigh frequency induction ,ileld .acts upon the tape T as it moves downwardly in the furnace tube 20.

The lower end of the tubeis illustrated as im- .mersed ina quenching bath 25, for example, for

onehalfinch. Immediately below the ,lower'end of the tube are the guide blocks 26 which press against the two faces of the tape and maintain electrical conductive relationship therewith. The tape T passes around the lower guide roller 21 within the quenching bath, and thence travel wardly through a wiper 3% which strips adherent liquid from the tape, and thence over the outlet metering unit conventionalized as the rolls 3|, and thence over other guide rolls and between the cleaning brushes 32, and finally the tape is rewound on a take-up reel 33.

Atthermometer is provided for observing the temperature in the quenching bath 25, and ap-- propriate means (not shown) may be provided for maintaining the bath at a desired temperature. The bath may be of rape seed or other quenching oil, water, or like material, depending upon the selected hardening conditions. The bathmay be supported by a counterweight system; and lowered while threading the tape, or when a quenching is not desired.

"In 'Figure '2, the electrical supply system to thetterminals 23 of Figure 1 is illustrated as comprising a high frequency oscillator 48 which is .conventionalizedas two electron discharge tubes. The anodesof these tubes are connected to the tank inductance ll in parallel with tuning condenser 42 which is illustrated as a three section variable condenser with its central or rotatable plates grounded. It will be underrstood'that when the oscillator is in service, the output circuit elements iii, 32 are to the frequency of the oscillator 4! whereby po tential differences are established in the induc tor 4.1 .inthe usual way. Thus, the resonant com- Ibination of inductance H and capacitance 42 may be regarded as a primary output circuit. A sec ondaryoutput circuit is connected to the induc tor 4| .by appropriate adjustable taps and includesthe blocking condenser 43 to prevent direct currentflow, and the variable three section tun- :ing condenser 44. The condensers as, 45 and the portion of inductor 4! between the adjustable taps, are likewise tuned during operation essen-' ,tially to resonance with the current flowing in the oscillator 40 The potential differences ap peering across the outer terminals of the condenser 44, whose central movable plates are grounded, appears atthe terminals 23 and acts in the furnace induction unit U as aforesaid.

It has beenfound that the tape T demonstrates potential differences along its length, when current flows in the induction unit U. A conductor-5U leads to an adjustable inductor 5:, whose other :terminalisconnected by conductor 52 with a variable condenser 53. This condenser 53 in turn isconnected bya conductor 54 with thee-cmtact blocks 26. It is preferred to include a meter Mysuch as-a hot-wire a'mmeter, in this external or third-circuitxwhich comprises a series connection of theportionof the tape T including that which is'within'the furnace tube 2 the upper conductive contact roll l5, conductor 5!} with the meter v mnductor .5l, conductor 52, condenser 53, conductor 54, and the submerged lower contact blocks 26 iback'to the tape T. The effect of po 'tential differences along the length of the tape T, whichrare in synchronism with the high 1" re "quency :current from the oscillator but not necessarily in phase coincidence, .are thus parent across the tunable elements 5!, 53; and when these elements are adjusted into essential resonance to the oscillator frequency, the potential differenceslead to the exhibition of high current densities in this third circuit.

.InFigure .3, a preferred form of furnace unit :is illustrated. The glassfurnace tube 2D is supported.and;guided by the upper and lower heads 611,161. which have adjusting screws 62 therein. These heads in turn are supported on a bracket 03 and are held spaced by an outer larger fibre tube-64. This figure also shows a chamber 56 connected in the gas supply conduit 2|, with a wire heating coil 61 therein supplied by 60 cycle alternating current and serving to preheat the gas before it contacts the tape. Also, the optical pyrometer 68 is mounted so that the tape T is observed at a point about one inch above the oil level. An opaque tube 69 of non-conductive material is mounted around the lower end of tube 20, and has a window through which the hot tape may be viewed under black body conditions.

The tube of heat-resistant glass had internal and external diameters of 0.68 and 0.78 inch, the wall thickness being 0.05 inch, the tube was inches long. In operation, the distance from the upper end of the tube to the line H (Fig. 3) was about 6 /2 inches, and from line H to the upper end of coil I0 was about 8% inches. The tube was immersed for about a half inch in the oil bath at its lower end. Surrounding the tube 20 beneath its support, is the inductor unit U, which is illustrated in a preferred form which has a total length of 8 inches and comprises a winding 10 of 23 turns of No. 6 copper wire to form a coil wound over the tube 20 for a total length of 8 inches. Immediately at the upper and lower ends of this coil are provided three turns each of 0.25 inch outside diameter copper tubing, wound in a coil of 3 inches outside diameter and length of 1.5 inches. The upper end of the upper large wind ing II is connected to one supply terminal 23, while the lower end thereof is connected by conductor 13 with the upper end of the winding 10. The lower end of the winding 10 is connected by conductor 14 with the upper end of the lower copper winding 1'2, and the lower end of the winding 12 is connected by the conductor 15 with the other terminal 23. This arrangement provides a total inductance of the series-connected windings 10, H, 12 so that the condenser 44 can establish a resonant condition to the primary circuit. The arrangement has been found to provide an accurate impedance matching. These conditions exist when the electrical system is energized in the absence of steel tape T; e. g. with a copper wire or tape. Pre-adjustments may be required when the tape is introduced, and also after the system is in steady moving condition, to compensate the effect of the tape in changing the apparent inductance eflect reflected byit into the driving system.

Figure 3 also shows the supporting member 80 for mounting the lower contact blocks 36. One of these blocks,ad iacent the member 80, may be fixed, and the other block biased for conductive contact with the tape by the action of the spring 8| carried by the yoke 82.

In assembly, attention should be given to the grounding of metal parts of the structure, and adjacent metal. Thus, the feed'reel l0, and the feed drive including rolls ll, l2, are electrically connected to one another and to rolls' 3l and the take-up reel 33 (the conductors not'being shown) and this reel 33 is connected to the'ground. The primary and secondary circuits are also grounded through .the tuning condensers '42, M. It was found that the current demand at'the meter N was reduced to 100 milliamperes with the particular oscillator when this grounding was completed: and that standing waves appeared to be eliminated.

In operation, the liquid level in Figure-3 is preferably at substantially the line b--'b, and it has been found that a desirable heating for hardening and quenching can beeffected when" the tape first begins essentially to increase in temperature at about thef line aa of Figure 3.

In one practice, the vertical distance between lines aa. and b'b was 19 inches. The power input to the high frequency generator was 1 kilowatt and the tape T was of 8/0 size and was moved through the furnace tube 20 at a rateof 60 feet per minute. Upon observation by pyrometer, the'tape had 'a temperature of 1400 F. at a point 11 /2 inches above the oil surface; 1425 F. at a point 6 inches above the surfaceyand 1450 F. at 1 inch above the surface. The measured current in the third 'circ'uit, atthe meter M, was 9 amperes. The frequency generated in the oscillator 40 was about 7 /2 megacycles. When the third circuit was opened, a temperature in excess of the magnetic critical could not be exceeded, and the load conditions on the generator changed, indicated by a drop in the plate current.

In particular, it was found" that when this size of tape was employed, and thespeed reduced to 40 feet per minute, the flow of a current of over 10 amperes as indicated at the meter M would cause the tape to melt. This current was obtained by adjusting the condenser 5-3 more closely to circuit resonance.

Carbon dioxide gas under low pressure was admitted through the connection 2 l The upper end of the tube 20 was circular, with'an opening of 0.10 inch; a rectangular opening 0.01 x 0.09 inch has also been employed.' It was found that the gas was carried along with thistape, and caused to bubble from the lower end of the tube 20 following heating'by radiation and conduction from the tape T.

It is found that within a few seconds after starting from room temperature of all parts, the tape increases in temperature as it travels, until the desired illustrative top temperature of 1450" F. is reached, and thereafter the currents flowing in the several circuits remain essentially constant as long as the speed of the tape remains constant. The temperature of the glass tube gradually rises, and reaches a fixed temperature in about 20 minutes. In a test of the temperature of the glass furnace tube 20 at the center of the tube, it was found to be approximately 550 F. by a thermo-junction-pyrometer, but the leads from this pyrometer passed through the conductive field, so that this reading was probably too high.

This same apparatus has been employed successfully for the heat treatment of main spring tapes ranging in size from 0.0035 inch thick by 0.05 inch wide to 0.095 inch "thick by 0.22 inch wide; with speeds of travel of, say, 5 to feet. per minute dependent upon the size and shape of the tape and upon the power available. The generator has been operated, for such work, with inputs of to 1500 watts and at frequencies of 4 to 10 megacycles. Obviously, apparatus of different size or power, and operating at other high frequencies can be employed. In particular, it has been found that the operation can be so conducted that the tape is quickly raised at an e5 sentially regular rate during its travel between the lines a-a and b-b, so that when a quenching is employed by use of a bath 25-, this quenching is accomplished while thejtemperature is still rising, so to speak-,and thus the; time during 78 which the tape is subjected to temperatures above. room tem'ycraturev is restricted only by he delivery oi; a'suillcient quantity of ener y into the tape for conversion therein to, the form of heat. and iurther it is possible. to effect the. quenchin while the tape is. still rising in temperature, to speak.

In making tapes for main springs-oi watch movements, using the normal steels customary for the purpose, it has been found feasible to pro duce desired conditions along great lengths of the material and with essential uniformity. Because of the close controls which are possible, it has been found preferable for such main spring tapes to use. a heat treatment which is termihated while undissolved carbide grains are still present, as gives a. better wear resistance than when all of the carbon is in essentially uniform solution. Thus, the heating is to a term perature and for a time short of that at which the maximum hardness would be produced by effecting dissolution of all carbide, followed by a quenching. This preferred choice of condition is feasible with the disclosed procedure, for the reason that the. ferrous alloy is carried quickly from room temperature. to the desired quenchin'g temperature. without lag or delay at the temperature corresponding to loss of magnetic prop erties. The time of exposure to heating is favor-- able in that the carbide tends to dissolve more quickly than grain growth occurs. The existence oi the. small grains of undissolved carbide, revealed in metal-lographic studies, appears to limit and greatly restraingrain growth.

It is preferredto employ an upper contact roller 15' of copper or like material, so that this structure can rotate with essential freedom. On the other hand, it is preferred to form the lower contact elements 26 as pressure blocks or contact shoes located close below the surface of the quenching bath. when quenching is to be accomplished. These shoes act as guides for smoothing out and controlling tape warpage. They may be of steel or other metal affording adequate resistance to wear under the conditions of operation and serving to imparta chill block quenching to the tape, in addition to the oil quenching which is accomplished thereupon. Thus, the surface. of the tape is quickly fixed in the desired condition for optimum wear resistance. In operation, the hot tape vaporizes the oil or other comes into contact with the shoes, so that a very good electrical contact is thus established and maintained.

The frequency of the. primary oscillator is adjusted to the efiective distance between the u per contact roller l and the lower contact blocks 25, so that these points are essentially an electrical half-wave apart. The action or the tuning condenser 53 and inductance 5| thus controls the relative phase conditions, wherewith the current nodes are close. to the said contacts, considering the tape itself as a Lecher wire, so that essentially no tendency to current arcing is present. In practice, currents up to 25 amperes have been passed along a carbon steel ribbon without burning the contacts or ribbon surface. Correspo ngly, the current loop, or maximum current effort at the tape, is at a point of the tape within the inductive heating field.

By way of demonstration of the efiect of the flow or current through the. externa rc it neludingthe tuning elementsv 5|, 53;, it maybe pointed out that when this circuit is open, or is tuned greatly out of resonance with the current from the Oscillator Ml, the tape does not become incandescent until it has traveled downward for about one-half of the length of the coil 10 of the furnace unit U, as compared with its attaining incandescence at the line a-a.

While magnetic or inductive coupling may be employed between the primary or secondary circuits and the inductor 5 or other component of the third or external circuit, it has been found that this is unnecessary, and that a more satisfactory control can be attained by the derivation of current flow along the tape as an incident of the magnetic, current and heat effects therein produced from the furnace induction windings Ill, H, l2, with employment of the tuning elements 5!, 53 tov adjust this external circuit to essential resonance and thereby deter mine a current flow as indicated by the meter M. which brings the tape quickly and with constantly rising temperature to the desired condition of upper critical temperature at the moment of its contact with the quenching bath. Thus, if the tape becomes undesirably overheated, a slight detuning permits quick correction for this: while an insufiicient heating is corrected by the converse operation of bringing the external circuit more closely into resonance.

In the illustrative employment above, the admission of carbon dioxide was set out. This gas can be employed under such conditions, provided that. it is pure; that is, the commercial gas should be so pure that it contains less than 1 percent of gaseous impurities, largely hydrogen and nitrogen, insoluble in KOI-I, and at most only traces of water vapor. If the commercial carbon dioxide as is suspect, it should be passed through an externally heated tube containing steel and copper wools, and overcaustic alkali, to decompose and fix the sulphur gases, water vapor, etc. The rate of feed with the stated apparatus was varied from 30 to 50 cubic feet per minute without essential change in behavior. Obviously, rare gases, such as helium, argon, etc, are not troublesome, can be voluntarily added: their cost, however, is not normally justified in treating carbon steels as stated.

The use of carbon dioxide atmosphere, as compared with air, has an advantage even with ribbon materials which are not attacked by air, in that less injurious arcing occurs at the lower contacts upon detuning.

It is obvious that the inventionis not limited to the Specific embodiment illustrated and de scribed, but that it may be employed in many ways within the scope, of the appended claims.

What is claimed is:

1. An apparatus for heatetreatment of long through which the article may be moved in the axial direction, means for supplying high frequency current to said coil, and separate circuit means including at least a part of the article as a series conductor portion of the circuit and including also a device in said circuit means to tune said separate circuit to the condition of essential resonance at the frequency of the current flowing in said coil.

2. An apparatus for heat treatment of a con- ,ductiv tape. c mpr sin an in u ti coil, means for feeding the tape along the axis of the induction coil, contacts adjacent eachend 0f the coil for conductive engagement with the tape, means of supplying high frequency current to said coil at a frequency whose half-wave length is essentially the same as the distance between said contacts, and conductive means connecting said contacts and including a tuning device for establishing said contacts essentially at the current nodes.

3. A high frequency induction heating unit for long metal articles comprising a main hollow coil of conductive material along the axis of which the article can be passed, and additional windings of large diameter located around the ends of the main windings, said coil and windings being connected in series for establishing individual magnetic fields all extending in the same direction along the axis, means for supplying high frequency current to said coil and windings, means efl'ective for bringing the seriesconnected coil and windings into resonance with the said supplied current in the presence of the article, contacts located outside said coil and windings for electrically engaging parts of said article while other parts of said article are with- Y in the coil and windings, conductive means separate from said coil and windings and supplying means, said separate conductive means being connected to said contacts whereby a separate circuit is established including parts of the article located within the coil and windings, and a device in said separate conductive means efiective to tune said separate circuit essentially to resonance with said supplied current.

4. An apparatus for heat-treatment of long conductive articles, comprising an induction coil through which the article may be moved in the axial direction, means for supplying high frequency current to said coil whereby to induce a high frequency magnetic field in at least a part of the length of the article, electrical contact members to engage with the article and spaced 10 essentially an electrical halt-wave apart at said current frequency, and high frequency conductor means connecting said conductor members to form an electrical circuit including said part of the article and including the said conductor means connected in series with said part, the electrical constants of said conductor means having values at which said circuit is essentially in the condition of resonance at the frequency of the current flowing in said coil and with said contact members at current nodes of said circult.

RAY S. NEID IGH.

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

UNITED STATES PATENTS Number Name Date 2,123,776 Hansell July 12, 1938 2,224,998 Wood et a1. Dec. 17, 1940 2,240,019 Quarustrom et al. Apr. 29, 1941 2,299,934 Sherman et a1 Oct. 27, 1942 2,304,225 Wood et a1. Dec. 8, 1942 2,328,225 Morey Aug. 31, 1943 2,383,992 Sherman Sept. 4, 1945 2,394,944 Stanton Feb. 12, 1946 2,417,029 Wilson Mar. 4, 1947 2,437,776 Wilson Mar. 16, 1948 2,459,507 Denham Jan. 18, 1949 2,465,093 Hanson et al Mar. 22, 1949 2,479,346 Goodnow Aug 16, 1949 OTHER REFERENCES Principles of Heat Treatment, by Grossman, page 104, 1935,

Principles of Heat Treatment, M. A. Grossman, 1940, Am. Soc. for Metals, Cleveland, Ohio, page 104.

Images