US2599850A - Process of controlling and placing of radio-frequency heat in a dielectric - Google Patents

Description

June 10, 1952 J w MANN ETAL 2,599,850

PROCESS OF CONTROLLING AND PLACING OF RADIO-FREQUENCY HEAT IN A DIELECTRIC Filed May 27, 1947 INVENTORS JULJUS W MANN BY GEORGE 1?.EU55ELL 14 T TOENE Y5 Patented June 10, 1952 PROCESS OF CONTROLLING AND PLACING ENCY HEAT IN A DI- OF RADIO-FREQU ELECTRIC Julius W. Mann and George F. Russell, i Tacoma, Wash.

Application May 2'7, 1947, Serial No. 750,836

5 Claims. ('01. 219-47) This case is a continuation-in-part application of the case filed by us on November 26, 1943, for a Process of Controlling the Placing of the Heat in the Interior of Material by a Radio Frequency Standing Wave, Serial No- 511,882 now abanone 1 .In our patent on asingle Standing Wave Radio Circuit, No. 2,506,158, issued May 2, 1950, We disclosed how it is possible to set up a single standingiwave in a radio circuit and make the load or work an inherent part of the inductance and capacity, thus eliminating the necessity of matching impedance and resonance. as between a master oscillatorand a separate resonant output circuit. We further showed that substantially the entire length of the half-standing wave was in the inductance portion in the fundamental circuit and substantially the entire length of the other onehalf standing wave was in the capacitance portion of the circuit. It was pointed out in the CO1- pendingcase that the inductance portion of the fundamental circuit had two points of greatest energy lossandit was at these two points that the tapsof the grid, circuit were connected. This caused the oscillator tubes in the circuit to be biased properly with a minimum of effort and q p nt. f

There are two; additional power wave crests in the capacitance portion of the circuit and the present invention deals mainly with this feature of the circuit and sets forth how the power wave crests can be altered to heat dielectrics in different manners. For example, if the opposite walls of a tube are to be heated to a greater extent than the remaining portions of the tube wall, the radio circuit and associate apparatus may be adjusted to heat the two portions of thetube wall to a greater degree than the rest of the tube.

Again it might bedesirable to heat the dielectric more uniformly and to accomplish this the two spotsofgreatest' heat energy are developed in the dielectric 'at points spaced a sufiicient distance apart so that the .conduction of heat through the'diele'ctric'and between the spots will raise the temperature of this portion of the dielectric substantially to that of the temperature ofthe hot spots themselves. Also it is desirable sometimes to heat the dielectric to a higher temperature than that created by the'two individual power crests and this is accomplished by bringing the current and voltage more into phase with each other and merging the two power crests into'what appears as one and thus obtaining the result of the increased center peaked temperature.

Other objects and advantages will-appear in the following specification, and the novel features of the device will be particularly pointed out in the appended claims.

Our invention is illustrated in the accompany-- ing drawing forming a part; of this application, inwhich: 1

Figure 1 is a wiring diagram of the single standing waveradio circuit shown in our Patent No. 2,506,158; No. 510,566, filed November 16, 1943.

Figures 2 and 3 show graphs of current and voltage curves;

Figure 4 illustrates a power curve when current and voltage are out of phase;

Figure 5 illustrates a different power curve when the current and voltage curves are partially in phase;

Figure 6 by graph and illustration indicates heating in the capacitance or inductance and shows the merging or the separating of the two hot spots; and v Figures 7, 8 and 9 illustrate three different types of heating on three different types of dielectric.

While we have shown only the preferred forms of our invention, it should be understood that various changes or modifications may be made within the scope of the appended claims without departing from the spirit and scope of the invention.

In carrying out our invention we make use of the single standing wave radio circuit illustrated and claimed in our Patent No. 506,158. In our patent, the fundamental circuit comprises the inductance L, see Figure 1, and the capacitance W. The plate tank or restorative force circuit is inductively coupled to the fundamental circuit by the center of the inductance L of the restorative force circuit being inductively coupled to the current anti-nodal point of the inductance L of the fundamental circuit. The grids of the tubes T and T have their taps l and 2 adjustably connected to the inductance L at points of great energy reaction. The results obtained from the proper adjustment of these points with reference to the energy reaction are indicated at 4a 4a in Fig. 6 and will be referred to more particularly hereinafter. An energy consuming dielectric placed in the capacitance W will bring the voltage and current into phase'and change the position of the crests of the power curves. The grid taps I and 2 will be connected to the inductance L at energy crest positions which means the grids will receive proper excitation and bias in con- ,nection'with the grid bias resistance. Because but a single standing wave resides in the circuit, grid and. plate swings will increase simultaneously and absorb a greater restorative force, which in turn allows a greater power consumption in the dielectric.

In Figures 2 and 3 we graphically illustrate the current and voltage curves of a single standing wave in the capacitance W and inductance L of the circuit shown in Figure 1. The plates P and P of the capacitance W of Figure 1 are shown spaced along the X-axis which indicates time in both Figures 2 and 3 with the capacitance being indicated by W and the inductance by L in these figures so as to agree with the showing in Figure 1. The current curve I is illustrated as leaving the plate P (the plate P is shown twice in Figures 2 to 6, inclusive, in order to graphically illustrate the endless circuit) and simultaneously flowing through both the capacitance W and the inductance L to the plate P and then returning to the plate P, note the arrows in Figure 2.

The voltage curve is 90 out of phase with the current curve when there is theoretically no resistance in the circuit. The voltage curve is indicated at E in Figure 3 and is at maximum when the current curve is zero. The voltage is zero in both the capacitance W and inductance L of Figure 1 when the current is maximum and vice versa. The arrows on the curve E portion lying in the capacitance W in Figure 3 indicate the changing of the voltage from maximum at the plate P to zero midway between the plates P and P and back to maximum at the plate P, and also indicate the return or other half cycle in the same voltage line E. Simultaneously the voltage in the inductance L starts at maximum at the plate P, drops to zero at the current antinodal position and again rises to maximum at the plate P. The second half cycle is also indicated by the return arrow to the plate P in the inductance portion of the circuit L and on the same line E in Figure 3. In both Figures 2 and 3 the arrows above the X-axis time line show direction in one half the cycle while the arrows below the X-axis show directional efiects during the second half of the cycle.

The power distribution curve Y shown in Figure 4, illustrates what happens when the current curve I and the voltage E are 90 out of phase. There are two power crests in the capacitance portion W of the circuit and two power crests in the inductance portion L. This is the condition of wattless power where all of the energy put into the circuit during the first quarter cycle is returned to the generator in the second quarter of the cycle, and likewise all of the energy put into the circuit during the third quarter is returned to the generator in the fourth quarter cycle. Only the first and second quarters are here shown. In actual practice the current and voltage curves are never 90 out of phase. Figure 4 illustrates a theoretical condition when there is no resistance in the circuit. Figure 5 illustrates the change in the standing wave distribution curves Y when current and voltage are brought more into phase by the placing of an energy consuming dielectric in the capacitance W or by placing a conductor in the inductance L. The small loops Y are the negative power loops. It will be seen that much more power is consumed in the circuit during the first quarter than is returned to the generator in the second quarter of the cycle and also more is consumed in the circuit during the third quarter than is returned to the generator in the fourth quarter of the cycle. The whole cycle is depicted here, rather than only a half cycle as in Figure 4 described above.

In Figure 6 the power distribution curves Y lying below the X-axis shown in Figure 5 are swung through an arc of and are indicated by the dotted lines. The resultant heat placement in the capacitance W and in the inductance L are indicated by the heat curves H. The two curves will be similar in the inductance and the capacitance portions and their overlapping will be in a similar proportion. The heated areas, therefore, in the dielectric portion W will be similarly disposed along the X-axis to those in the inductance portion L. The inductance portion L may be variable as may be the capacitance portion W, but in any event the entire composition of L and W being a resonant circuit will have matched impedance in the condition of proper operating efficiency. Under such a circumstance, therefore, the heat placement will be similarly disposed in each section. Below the time X-axis line in Figure 6 in the capacitance portion W is a representation of the dielectric 3 disposed in such section with the area 3a of greatest heating effect centered as if a single burn. Under the X-axis in the inductance portion L is a representation of the inductance 4 showing the two spot heat effect at 4a respectively merged into a single peaked effect which, in an inductance, is equivalent to the apparent single burn in the dielectric disposed in the capacitance portion W.

The spots of greatest heat in the dielectric or conductor may be altered in their positions with respect to each other by: one, controlling the length of the inductance and the size of the electrodes or plates P and P; two, altering the total mass and shape of the dielectric and its dielectric constant; three, altering the phase angle between the current and voltage which in turn is governed by changing conditions one or two, and; four, employing cauls of thicknesses and dielectric constant. The heating effect on the dielectric is controlled by the phase angle shift and the shift is controlled by the varying of the inductance and capacity and the relative resistance of the load.

We show three examples of different types of heating dielectric for various commercial purposes in Figures 7, 8 and 9. Figure 7 illustrates a thermo plastic tube A to be heated on its two sides in order to be bent. Internal pressure is built up within the tube by an air inflated bag or hose, not shown, and the tube is placed between a pair of cauls 5-5. The cauls are used as spacers between the plates P and P and the tube A and prevent physical contact of the plates with the tube. Condenser plates P and P are disposed at the sides of the cauls and are connected to the radio frequency inductance L. The centers of the power distribution curves Y coincide with the heat curves H which in turn have their peaks falling at the two points 66 at opposite sides of the tube. The portions of the tube lying between these two points will not be heated to the same extent and therefore the plastic tube wall will be softened at the two points and the tube when removed can be bent with the greatest heated portions forming the top and bottom portions of the tube during the bending process.

Where a more uniform heating is desired such as the bonding of material forming a resin bonding grinding wheel B in Figure 8, the cauls 11 are placed between the sides of the wheel and the plates P, P so that the crests of the power disgases-t "5 tribuutn curves will fail 'e 'ift t6" arena of the wheel. e ptwerpea scante made to eveuap each otherto a greater Xtentft Figure 7. The overlapping portions of the power curves will create a combined heat to cause the central peak of the heat curve H to be ateen; many as greatas the hotsp'ots ar ea y the power crests. The result is a substantiallyuhiform heating across the entire width of the wheel. Note that the uniform heatihgitdos notext'end from plate to plate the condenser.

heating shown at H in the'same figure. The high peaked effect indicates that the two hot spots are merged sufficiently to heat the interioro f the preform to a greater extent than the individual heat createdby each spot. 7

By actual experiment with the above three ex amples of Figures 7, 8 and 9, we found that a two spot separation of heat was accomplished by in-' creasing the'spacing between the plates, thereby lessening the capacitance and by physically lengthening the inductance: to increase. it, as in Figure 7. Thedielectric mass Ain this figure created less capacitance" than the dielectric masses B" and 'D in Figures Sand 9, respectively. In obtaining the two'spot overlap for the single center burn as in Figure 9; the capacity was made larger and the inductance relativelysmaller than the same components in Figure 7. The more ev'en'distributiori of Figure 8 was accomplished; by a compromise of the capacitance and inductance from the more extreme cases'show'n'in Figures 7 and 9.

The two burned area's shown at 4a in Figur 6 and at 6 in Figure 7 is' a result of'a shift in phase of the current and voltage which is less than.

that which causes a center burn. The burned spots'are close together'at 3am Figure 6 and also in Figure'Q, when the phase shift 'is'greater and the in phase relationship between the electrical components is closer than where the burned spots are separated. When the phase angle between the current and voltage'approaches 90?, there is very little energy absorption and the two burned areas are substantially one-fourth and v three-fourths the distance between the two'plates.

As the phase angle approaches zero and maximum energy absorption is effective; the two heated areas merge to form'a single area of ahigher temperature. Then the heating" effect becomes.

the practical equivalent of E I or current squared.

It should be noted that the pla'cingof a di' electric in the capacitance or a conductonin'the' inductance does not'alter the fact that'but one single standing'wave resides in the fundamental circuit. The circuitcanbe so adjusted bythe variable condenser C in Figure 1, that the plac-- ingof an energy absorbing dielectricin'the capacitance W'Will so change the frequency of the circuit that without further adjustment maxi: mum power output will become efiective. The circuit can also be adjusted by the variable con-' denser C so that by placing a conductor 'inthe combined inductances portion will so change the frequency of the circuit that withoutfurther ad'-' the j charge.

Ei h r e the eeneciiane i s i..e t

. heating doesnot necessitate the materialcom tacting the electrodes; making possible heating very high diel ctric constant materials without shorting the circuit. I V Following. is a'description. of the method used ior recording.'the heat pattern produced bythe standing wave pattern of the electric fieldof force existing between the condenser plates P and 1?" of a high frequency dielectric heater:

A one-eighth inch thick sheet of asbestos proved i to be a material which had the qualities best 2O suited for the purpose in hand for the following reasons, namely; V g

1. Asbestos withstands a high temperature. 2. It may be dried down to a bone dry condition readily. I v

3. It is very absorbent to moisture. I f 4. It is highly resistant to any rapid heat trans- 5. Its white color-is advantageous to contrast- 6. It. loads the R. F. generator to a stable condition..

. The chemicalus ed on the asbestos was cobaltfous nitrate which decomposes at a lowerv temper'atu're' thantheboiling point of watergand upon hightemperature, yields nitrogen tetraox lde, oxy en and cobalt oxide, the latter being a brownsolid; ..Hence, whereverthe electric field releases energy atthe greatest rate there will appear a brown area on the White asbestos.

. The procedura -Due to the release of free oxygen; when the cobaltous nitrate decomposes, we find .that paper, wood or any combustible substancewill start burning fiercely in the field even though the substance is very wet or damp.

en e, the use of asbestos paper. Apply the cobaltous nitrate solution to one side; of the as- Y bestos'sheet by v.means of an atomizer or ordinary garden'spray'hand pump.. Theareas which absorb heat at the greatest rate of course rise rapidlytohigher and.higher temperatures and these areas first; appear as pinkish blue due to the'crystalliz'ation'of the cobaltous nitrate as the sprayed-on coating becomes dry. Soon, in a matter ofiseconds the salt decomposes releasing'tox'yg'en and the brown .fumes of nitrogen tetraoxide; Asl'a con sequence, red hot areas may even result on the asbestos surface, registering temperaturesgreater than 6001 while a fractio'n of an'inch away thetemperature may be w'ellbdlovv' match ng point of water. I As the fieldisikept on, these heated spots spread, leaving theubrow'n oxide of cobalt deposited upon the surface of thefasbestospaper as a. permanent record of thel nori-uniform. generation of heat. created by ,m'eansof a half wave of electric field between the plates P, P. of the condenser. I

,Thej term.".chargef used inthe subsequent attempt to explain the phenomena observed, refers to a, quantity'ofv electricity,.i. e., electrons. A

.chargefis associatd'wit h an electric field, the

latters intensity changing with the change in I Current flows at any point in a conductor because there isJan electric field as the initiator cf thecurrent, Dueto. achanging electric field occupying an evacuated space .be-

tween the plates P, P of a condenser, a changing magnetic field is set up just as if a real electron disturbance existed between the spaced electrodes. Such a fictitious current is known as :a displacement current.

If any material substance is placed between the above electrodes P, P', then the changing electric field will initiate a displacement of electrons whether the electrons are free or bound. From measurements taken of standing waves on helical lines a coil) as compared to the standing waves set up by the same frequency on a similar conductor of approximately the same length, but stretched out straight, it is found that, in a sense, the speed of propagation may appear less in traversing the length of the axis of the helix as compared to traversing the length of the straight conductor. If the helix is terminated by condenser plates P and P and it is admitted that a half standing wave exists across the coil terminals, then it is reasonable to expect a half standing wave of electric field to exist across the gap between the condenser plates, thus completing a full standing wave system.

Traveling waves are said to obtain when maximum values of charge and of axial current occur simultaneously at the same cross-section of the conductor; a condition of zero reflection. Standing waves are said to obtain when nodes and antinodes form at fixed points along the conductor, these nodal points being due to the nullification and reinforcement of the reflected wave meeting the incident wave. In this case, the maxima of current density occur at instants differing in time by a quarter period and at locations differing in position by the equivalent of a quarter wave length from those of the maximum density of charge. Stated another way, standing wave motion obtains when all the displacement currents and electron shifts arrive at their maximum and minimum deviations from normal simultaneously. If in a traveling wave there are spots in the wave form where energy undergoes conversion at maximum rate as compared to other portions of the wave, then it is reasonable to expect these spots to remain in one position with respect to an oscillating system undergoing standing wave motion for the reason that current and voltage are also standing in the system.

At low frequencies, the current distribution in everyday conductors is quite uniform, both with respect to depth under the skin of the conductor and with respect to its axial length, be-. ing practically the same at all cross-sections. As the frequency is increased to the point where the electric circuit components become comparable in dimensions to that wave length corresponding to the frequency, standing waves result due to reflection or partial reflection and then charges and currents sit and flow on the conductor surfaces. Potential energy is associated with the charges that sit and kinetic energy is associated with currents that flow. A measure of the potential state is expressed as a ratio of unit work to unit charge, that special ratio of joules to coulombs being designated as volts. In the case of the kinetic state, it is more convenient to introduce time into the comparison and the resultant ratio also expresses volts as the ratio between the time rate of doing work and the time rate of flow of charge, namely, watts per ampere. From this relationship the product of volts times amperes expresses power in watts.

This review leads us to another. In the case of low frequency systems, for instance 60 cycles,

the applied voltage may be the resultant of as many as three component voltages, inductive reactance, capacitive reactance and the voltage in phase with the current. As a consequence, the applied voltage is usually out of phase with the current. At any instant of time the instantaneous power may be calculated by taking the product of the instantaneous applied voltage and the instantaneous current. As a consequence, also during one-half cycle with respect to time, there are established a positive envelope of energy and a negative envelope of ener y, the negative being the smaller and representing the energy returned to the generator from the dominant reactive sink. As this slow frequency traveling wave of current progresses through the circuit at nearly the speed of light these positive and negative crests of the envelopes of energy speed with the wave and the positive diiferential would of course produce uniform heating, say, in a resistance heating coil. The power expended in the resistor is then measured in watts by the relationship watts=I R, or E /R or E x I where E and I are given as eifective values.

Contrast the above situation with the one associated with frequencies high enough to produce wave lengths comparable in size with the physical dimensions of the circuit boundaries with which they are associated. Also be reminded of the fact that due to the apparent reduced speed of wave propagation along a helix and across the gap of a condenser, it is possible. in the case of a standing wave condition, to study the results of the aforementioned power crests with respect to space as well as time. As a consequence, any dielectric placed in the condenser gap will be heated throughout more or less non-uniformly provided a standing wave spans the gap. If the dielectric hashigh loss characteristics, such as high moisture content wood, then a high center heat obtains with resultant low amplitude of swing of the high frequency generator. On the other hand, if the dielectric is very low loss material, the amplitude of the generator will rise toward a maximum, the standing power peaks will separate and produce two relatively highly heated areas or sections each side of center between the condenser plates P, P. The energy is decidedly not uniformly placed and the total energy expended may be calculated from E X I cos 0 where E and I are effective values. It leads to absurdity to use the formulae PR and E /R with respect to the standing state described because, using 1 R at the voltage antinode where I approaches zero, the product approaches zero, whereas, if E /R is used where E is maximum, the ratio yields a substantial amount.

The curves shown in Figures 6, 7, 8 and 9 are quite descriptive of this phenomena. If Figures 2, 3, 4 and 5 are folded upon themselves about the vertical axis P, the result shows graphically the standing conditions of reactive voltage, applied voltage, current and positive and negative energy as they exist between the potential sinks of an oscillatory circuit in which the displacement currents and electron shifts are executing standing wave motion. The nonuniform placement of energy and the control of its placement as described herein assumes perfectly uniform physical and electrical properties of the material subjected to the electric and magnetic fields of force. It is also assumed that the components of the oscillating system are symmetrically disposed with respect to some 9 neutral axis and with respect to; some; neutral axis and with respect to adjacent objects and insulating supports.

The practice of electrode shaping, spacing, the use of auxiliary heat and of air gaps to control more or less the non-uniformity of energy placement has long been known. Variable physical and electrical properties in materials such as, for instance, knots, pitchpockets, variable den sities, cracks, voids and differencesinmoisture contents, in woods are disturbing but fortunately lie within practical safe limits. We combine known practices of field placement with thediscovery of the natural non-.uniformtwo' spot concentrations of energy: inthe standing wave elec trio and magnetic: fields of forceto. improve the control of heat placement in: substances so processed.

As an example: Tomakea'solid material suchas a panel from resin impregnated sawdust, the electrodes must be spaced away from the finished surface and center section of the compressed sawdust by meansv of suitably thick, low loss cauls, as shown in Figure 8. The impregnated sawdust being high loss material tends to merge the two spots, thus resulting in a center cook with no cook at the face of the live electrodes. Thus; the need for the cauls to position the crosssection of; the sawdust at the, center section where they energy is concentrated.

The procedures f r, showing the non-=uniform heating patternshavebeendescribed. The, exact methods of securing these-non-uniform heating resul s. a e; a follqws:

F r uniform, heating fan ckase 6" Wide, incheslong by A thick asbestos, bone'dry (or very low moisture content) at 27 inegacycles i, use 55 long inductance leads either side of center, condenser electrodes 2, P, 16 long x 8" wide spaced 6 7 apartwith; the dielectric;- therebetween, the short dimension (6") being across the electrodes.

2. For center heating, see Figure 9, of the package described in paragraph 1 above, all conditions are exactly the same except moisture content should be raised to a point sufficient to produce a large kinetic sink in the dielectric.

3. For heating the outside edges of the material and leaving the center cold to produce two spot heating as shown in Figure 7 and described above; at 27 megacycles i, use 86" long inductance leads either side of center. condenser plates 2 /4" x spaced l0" apart with a dielectric of asbestos at room humidity of a size 6" x 10" x with the long dimension placed between the electrodes. Thus, inductance has been increased and capacity reduced. This will produce extremely high temperatures in the edges of the dielectric nearest the electrodes and leave the material in the center relatively cool in comparison.

4;. To produce the same two spot heating as described in paragraph 3 above, but draw the two spots closer together, the only change required from the conditions in paragraph 3 would be to increase the moisture content so as to create a greater kinetic sink of a higher loss. This brings the spots closer together and produces a condition where the center and the outside edges of the dielectric are relatively cool, and the two areas therebetween are hot.

To give a further illustration: if taking Figures 7, 8 and 9 at 2'7 megacycles, the two spot heating of Figure '7 would be produced by having an inductance longer and a capacity smaller than the' samecomponents of either Figure 8 or 9-. Also. at the same 27 megacycl'e frequency, the relatively uniform distribution of heat obtained in Figure 8 across the dielectric B, the inductance would be shorterthan that in Figure 7 and longer thanthat of Figure 9"; while the capacity would be greater than that of th'ecapacity' of Figure 7 and less than the capacityof Figure 9; Also: at the same Z I- megacycle frequency, the relatively high peaked center heating obtained in thedielectric as per Figure 9 is obtained with an inductance length less than that ofeither Figure 7 or 8 butacapacity betweentheelectrodes? andP' greater than that between the corresponding elements of" either Figure 'l or 8. 7

Assuming the same dielectric is heated, greater capacity can be had by increasing the area of the electrodes on reducing the spacebetween the same-sized ones; lesser capacity is-the reverse; greater inductance increase can be hadby increasing the physical length-of therelati veinductance component and a. reduction can be had by reducing its length.

In all: such adjustment to: the field of" force by arranging relations-between inductance and capacity, the material beingheated may or may not touch the electrodes of the radio circuit.

This will depend on the capacity desiredand other considerations, such 1 as moisturecontent, dielectric constant, electrostatic strain inthe R; F. field, and many-other factors.

We; claim:

LTh'e herein described process: of" heating; dielectric material. substantially uniformly by a radio frequency field of. force: which comprises: placing thematerial between the electrodes of a condenser in a radio. circuit which includes inductance and: capacitance, the mar inductive" and: capa'citative components of which limit the residence of radio frequency energy to but one single full standing wave comprising current and voltage curves; substantially the entire length of one-half of the wave residing in the condenser; regulating the phase angle between the current and voltage curves in the condenser so that they will overlap sufiiciently for creating two power distribution curves whose crests will substantially coincide with the sides of the material and whose overlapping portions will create sufficient heat to equal substantially the heat generated at the two crests; whereby a more uniform internal heating of the material throughout its mass results.

2. The herein described process of heating a dielectric material by a radio frequency field of force to a higher degree in the center of the material than at its sides which comprises: placing the material between the electrodes of a condenser in a radio circuit which includes inductance and capacity, producing a radio frequency energy wave, regulating the phase angle between the current and voltage curves of said wave so that they will be in phase sufficiently for creating two power distribution curves whose crests will substantially merge and whose overlapping portions will create a combined heating effect greater than the heat of each separate crest; whereby a peaked heating effect is created in the interior of the material, between the said power crests.

3. The herein described process of heating dielectric material substantially uniformly by a radio frequency field of force which comprises: placing the material between the electrodes of a condenser in' a radio circuit which includes inductance and capacitance, the total inductive and capacitative components of which limit the residence of radio frequency energy to at least one single full standing wave having voltage and current curves; substantially the entire length of one-half of a single standing wave residing in the condenser; regulating the phase angle between the current and voltage curves in the condenser so that they will overlap sufiiciently for creating two power distributing curves whose crests will lie within the material and whose overlapping portions will create sufficient heat to equal substantially the heat generated at the two crests; whereby a more uniform internal heating of the material throughout its mass results.

4. The herein described process of heating a dielectric material substantially uniformly by radio frequency field of force which comprises; placing the material between the electrodes of a condenser in a radio circuit which includes inductance and capacitance; producing a radio frequency energy wave having voltage and current curves; substantially the entire length of one-half of a single standing wave residing in the condenser; regulating the phase angle between the current and voltage curves of said wave in the condenser so that they will be in phase sufiiciently for creating two power distribution curves whose crests will lie within the material and whose overlapping portion will create sufficient heat to equal substantially the heat generated at the two crests; whereby a more uniform internal heating of the material throughout its mass results.

5. The herein described process of heating a dielectric material by a radio frequency field of force to a higher degree at the opposite sides of the material than at the center which com- 12 prises: placing the material between a pair of cauls and arranging the electrodes of a condenser on the outer surfaces of the cauls; the condenser forming a part of a radio circuit which includes inductance and capacitance; producing a radio frequency energy wave having voltage and current curves; substantially the entire length of one-half of a single standing wave residing in the condenser; regulating the phase angle between the current and voltage curves of said wave in the condenser so that they will be in phase sufliciently for creating two power distribution curves whose crests will substantially coincide with the sides of the material and whose overlapping portions will create little heating effect within that portion of the material lying between the power curve crests; whereby the material will be heated to a greater extent at its opposite sides than at the center.

JULIUS W. MANN. GEORGE F. RUSSELL.

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

UNITED STATES PATENTS Number Name Date 2,068,799 Guyer Jan. 26, 1937 2,304,958 Rouy Dec. 15, 1942 2,308,043 Bierwirth Jan. 12, 1943 2,308,204 Parry Jan. 12, 1943 2,370,624 Gillespie Mar. 6, 1945 2,415,025 Grell et a1. Jan. 28, 1947 2,433,067 Russell Dec. 23, 1947 2,434,573 Mann et a1. Jan. 13, 1948 OTHER REFERENCES Taylor, Transactions of the A. S. M. E., April 1943, pages 201-212.

Images