US2738421A

US2738421A - Means for preventing the loss of charged particles injected into accelerator apparatus

Info

Publication number: US2738421A
Application number: US309007A
Authority: US
Inventors: Willem F Westendorp
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 1952-09-11
Filing date: 1952-09-11
Publication date: 1956-03-13
Anticipated expiration: 1973-03-13

Description

March 13, 1956 w. F. WESTENDOR'P MEANS FOR PREVENTING THE LOSS OF CHARGED PARTICL v INJECTED INTO ACCELERATOR APPARATUS Filed Sept. 11, 1952 2 Sheets-Sheet l Figl.

My: n in wzw on Inventor: Willem F Westendorp,

zlvz a. v

His Attorney.

March 1956 w. F. WESTENDORP 2,738,421

MEANS FOR PREVENTING THE LOSS OF CHARGED PARTICLES INJECTED INTO ACCELERATOR APPARATUS Filed Sept. 11, 1952 2 Sheets-Sheet 2 22 H22 X x,

SOURCE OF TIME mRY/NG WL77465 Inventor: Willem F We stendorp,

1 y W 4. m

' His Attorney.

United States Patent Willem F. Westendorp, Schenectady, genfiral Electric Company, a

Application September 11, 1952, Serial No. 309,007 7 Claims. (Cl. 250-27) N. Y., assignor to corporation of New The present invention relates to charged particle accelerator apparatus and, more particularly, to methods and means for preventing the loss of charged particles injected into accelerator apparatus.

Apparatus for accelerating charged particles by means of magnetic induction effects is shown and described in my prior United States Patents Nos. 2,394,071, 2,394,072 and 2,394,073, all of which were issued February 5, 1946, and assigned to the assignee of the present invention. Such apparatus can comprise a core of magnetic material including a pair of opposed, rotationally symmetrical pole pieces which define a toroidal gap wherein an evacuated annular container is positioned. The core is excited by means of windings that are energized by a source of timevarying voltage to produce a time-varying magnetic flux which links an equilibrium orbit within the evacuated container and a time-varying magnetic guide field which traverses the equilibrium orbit. Charged particles, e. g. electrons injected along the equilibrium orbit from an electron gun positioned adjacent to the orbit within the region of influence of the time-varying magnetic guide field, are accelerated to high energy levels by the timevarying magnetic flux during a great number of revolutions while the time-varying magnetic guide field constrains the patricles to follow paths along the equilibrium orbit. After acceleration to a desired energy level, the charged particles can be directed from the equilibrium orbit to a target for the generation of X-rays.

As is denoted in my above-mentioned patents and more particularly in United States Patent No. 2,297,305, patented September 29, 1942, by D. W. Kerst and assigned to the assignee of the present invention, it is necessary that the time-varying magnetic guide field existing between the faces of the rotationally symmetrical pole pieces be spatially distributed to provide a field intensity essentially inversely proportional to the radius to the power 12 in order that both radial and axial focusing forces will be present to constrain the particles to paths along the equilibrium orbit. Specifically, this power or exponent 11 must have a value lying between 0 and 1 if both radial and axial focusing forces are to exist. The combined effect of the radial and axial focusing forces contributed by the time-varying magnetic guide field is to produce for a very large proportion of the charged particles oscillatory motions along and about the equilibrium orbit, the amplitudes and frequencies of which are determined by the exponent n" of the time-varying magnetic guide field. In some instances, these oscillatory motions of the charged particles cause many of them to deviate sutficiently from the equilibrium orbit after one or several revolutions following injection that they strike the walls of the evacuted container or return to impinge upon the rear of the injector electrode structure. Thus, many charged particles are lost and the ultimate maximum output is severely limited.

Accordingly, a principal object of the present invention is to provide methods and means for preventing the 2,738,421 Patented Mar. 13, 1956 loss of injected charged particles in charged particle accelerator apparatus.

A further object of the present invention is to provide methods and means for reducing the amplitudes of the oscillatory motions of charged particles within magnetic induction accelerator apparatus.

A still further object of this invention is to provide methods and means for reducing the amplitudes of the oscillatory motions of charged particles without deleteriously affecting the time-varying magnetic accelerating flux in magnetic induction accelerator apparatus.

Yet another object of the present invention is to provide methods and means for obtaining maximum reduction of the amplitudes of the oscillatory motions of charged particles during initial periods including the periods of injection of charged particles into magnetic induction accelerator apparatus.

The above objectives and others of the present invention are accomplished, according to one but not necessarily the broadest of its aspects, by changing the distribution of the time-varying magnetic guide field in magnetic induction accelerator apparatus with supplementary magnetic guide fields which can have both time-varying. and time-constant components, the time-varying component being present only during an initial period including at least the period of injection of the chargedparticles. By thus changing the distribution of the timevarying magnetic guide field, the amplitudes of the oscillations of the charged particles are reduced and substantial loss of the charged particles is prevented.

The features which are desired to be protected herein are pointed out with particularity in the appended claims. The invention itself, together with further objects and advantages thereof, can best be understood by reference to the following description taken in connection with the accompanying drawings in which:

' Fig. l is'a partly sectionalized view of accelerator apparatus suitably embodying the invention; Fig. 2 is a view taken along line 22 of Fig. 1;

Figs. 3, 3A and 4 are graphical representations helpful in explaining the invention;

Fig. 5 is a schematic circuit diagram illustrating a means of energizing the apparatus of the invention; and

Fig. 6 is another graphical representation useful in explaining the invention.

Referring particularly now to Figs. 1 and 2, thereis shown in exemplary fashion magnetic induction accelerator apparatus suitably embodying the invention. The apparatus comprises a magnetic core 1, which can be laminated to minimize the generation of eddy currents therein. Core 1 includes laminated, rotationally symmetrical, opposed

pole pieces

2 and 3 having generallyoutwardly tapered pole faces 4 and 5 for the provision of a magnetic guide field traversing an equilibrium orbit X, as will be more fully described hereinafter. Coaxialv with

pole pieces

2, 3 and disposed between pole faces 4, Sis an evacuated annular container or vessel 6 of dielectric material which provides within its interior an annular chamber 7 wherein charged particles can be accelerated. The central portions of

pole pieces

2, 3 are terminated respectively by flat surfaces 8, 9 between which are dis-. posed laminated metallic disks 10, 11 and dielectric support spacers 12. Metallic disks 10, 11 serve the purpose of reducing the reluctance of the magnetic path in the region between surfaces 8 and 9.

Magnetic core 1 can be excited from a suitablesource of time-varying voltage 13 connected as indicated to series-connected energizing windings 14, 15 surrounding

pole pieces

2, 3. To minimize the current drawn from source 13, energizing windings 14 and 15 can be resonated by power-factor-correcting capacitors 16. Within chamber 7 adjacent equilibrium orbit X and also within the region of influence of the time-varying magnetic guide field existing between pole faces 4, 5 during operation of the apparatus, there is provided a charged particle source 17 which is supported from a hermetically-sealed side arm 18 of container 6. More detailed illustration and description of electron gun structure suitable for present purposes can be found by reference to the abovementioned patents or by reference to the United States Patent No. 2,484,549 of I. P. Blewett, patented October 11, 1949 and assigned to the assignee of the present invention.

It is well understood by those familiar with magnetic induction accelerator apparatus that energization of windings 14, by the source of time-varying voltage 13 results in a time-varying magnetic flux which traverses magnetic core 1 and

pole pieces

2, 3 to provide a timevarying magnetic flux that links equilibrium orbit X and a time-varying magnetic guide field that traverses the locus of equilibrium orbit X and the vicinity thereof between pole faces 4, 5. Electrons emitted by a gun 17 at a desired timed instant near zero in the cycle of magnetic flux and field variations are continuously accelerated during the acceleration portion of the cycle as they execute repeated revolutions along and about equilibrium orbit X. As a consequence, the injected electrons can be caused to assume energies of many millions of electron volts and then can be automatically directed to impinge upon a target (not shown) to produce X-rays for utilization exteriorly of the apparatus. Means including circuits for arranging the proper timed injection and subsequent direction of the charged particles to a target at or near the end of the acceleration cycle are disclosed in my aforementioned patents and additionally in United States Patent No. 2,394,070 of D. W. Kerst, patented February 5, 1946, and assigned to the assignee of the present invention.

.As has been explained in D. W. Kerst Patent No. 2,297,305, patented September 29, 1942, and assigned to the assignee of the present invention, the time-varying magnetic flux linking the equilibrium orbit may be caused to produce centripetal forces which balance the cen trifugal forces upon the charged particles undergoing acceleration providing the following relationship is satisfied:

Where A is the total change in the flux linking the equilibrium orbit, R is the radius of the orbit and Bo is the flux density of the time-varying magnetic guide field at the equilibrium orbit. The condition specified by this relationship may be realized by making the reluctance for one unit area of cross section of the magnetic path of the time-varying fiux greater by an appropriate amount at the equilibrium orbit than its average reluctance for one unit area of cross section within the orbit.

The fulfillment of the foregoing condition, however, only assures stable acceleration for those charged particles which are injected tangentially to their instantaneous circles or orbits. The instantaneous circle or orbit is the circular orbit along which a charged particle started at the proper position with the right energy will travel in a time-constant, radially symmetric magnetic field. With a time-varying magnetic flux as above specified, the loci of the instantaneous circles of all the charged particles approach and eventually essentially coincide with the equilibrium orbit. Consequently, meeting the foregoing condition does not take into consideration the requirements for stable acceleration of charged particles which tend for one reason or another to deviate from their respective instantaneous circles or to deviate from the equilibrium orbit when their respective instantaneous circles coincide therewith. Nevertheless, by arranging the spatial distribution of the time-varying magnetic guide field 'in the vicinity of the equilibrium orbit as specified by the following relationship, both radial Cir and axial focusing forces which tend to constrain deviating particles to their respective instantaneous circles or to the equilibrium orbit can be provided:

where H is the intensity of the time-varying magnetic field in the vicinity of the equilibrium orbit, H0 is the intensity of the guide field at the equilibrium orbit, R0 is the radius of the equilibrium orbit, R is the radius of a particular point under consideration and n is an exponent having a value lying between zero and one. The outwardly directed taper of pole faces 4 and 5 as illustrated in Fig. 1 enables the utilization of the conditions set forth in Equation 2.

From the foregoing it is apparent that the exponent n is a measure of the rate of decrease of the time-varying magnetic guide field with radius. Both radial and axial focusing forces exist if 0 n 1. For a uniform field, n equals zero and no axial focusing of the particles can take place. In a field inversely proportional to the radius, there are no radial focusing forces. Within the limits prescribed, the net radial focusing force which is directed toward the particular instantaneous circular orbit corresponding to the energy of an injected charged particle is at one unit length from this orbit:

bF B ev where F1 is the net radial focusing force above-defined, F is the total focusing force at any particular point under consideration, e is the charge upon the particle and v is the velocity of the particle. This force corresponds to the spring constant or the restoring force of an oscillating mass and gives rise to an oscillatory motion of each charged particle with respect to its respective instantaneous circular orbit. Upon injection the amplitude of these radial oscillations of a particular charged particle is initially a distance equal to or comparable to the distance from the injector or gun structure to the particles instantaneous circular orbit, depending upon the direction with which the particle leaves the injector. The stiffer the spring the faster the oscillations; if the spring is increased in stiffness during the oscillations, the frequency of the oscillations increases and the amplitude decreases. Therefore, as can be seen from Equation 3, if the value of :n is decreased during the time of charged particle injection, the spring constant increases and the amplitude of the oscillations of the injected charged particle decreases.

The period of one radial oscillation Tr is expressed by where To is the time for one passage of the particle around its instantaneous orbit. If n has a value of zero, Equation 4 shows that Tr equals To, and injected particles will strike the injector structure upon their first revolution. If n equals /1, the injected particles will strike the gun structure after two revolutions. However, if 0 n there are certain values of n and Tr for whichthe total oscillation pattern of the injected charged particles precesses, whereby the charged particles miss the injector structure for a relatively great number of revolutions. Nevertheless, this same precession eventually causes the points of maximum deviation of the charged particles to coincide with the injector structure or the wall of the container 6 whereby the particles are lost. Of course, some charged particles manage to avoid the injector or gun structure and the wall of the container for a suflicient number of revolutions such that space charge interactions and the approach of the particles respective instantaneous circles toward the equilibrium orbit cause the particles to be relatively closely confined to the equilibrium orbit whereupon their successful accelerationto number of the injected charged particles, however, the critical period is the initial period including the period of injection during which the focusing forces are relatively weak and the amplitude of the particles oscillations about their instantaneous circles relatively great.

According to the present invention, the loss of charged particles due to impingement upon the injector structure or the wall of the container is prevented by altering the exponent n of the time-varying magnetic guide field for an initial period including the period of injection. Thus, for radial injection as is utilized with gun 17positio-ned as illustrated in Fig. l, n is decreased during the build-up of the time-varying magnetic guide field, whereby the amplitudes of the charged particle of the oscillations are decrease to cause the particles to avoid the injector or gun structure even though precession of the oscillation pattern has brought themclose to the injector structure, as will be more fully explained hereinafter.

With particular reference again to Figs. 1 and 2, the desired variation in n is obtained by means of a system of auxiliary coils or windings 19, 2t), 21 and 22 which can be attached with a suitable adhesive to envelope 6. Windings 19, 29 comprise a pair of windings which can be uniformly axially spaced on opposite sides of the plane of the equilibrium orbit X; and windings 21, 22 comprise a pair or" windings which can be uniformly radially spaced on opposite sides of the equilibrium orbit such that they lie essentially in the plane of the orbit. It will be observed from Fig. 2 that these windings are serially connected and that the direction of current flow therein when they are energized is such that essentially no change in the flux linking the equilibrium orbit will occur as a result thereof. Thus, if the windings are energized in a manner to be described presently such that current enters through conductor 23, the direction of current flow in winding 21 will be counterclockwise, in winding 19 clockwise, in winding 22 counterclockwise, and in winding 20 clockwise. Consequently, since the windings are serially connected and the current flow through each of them will always be the same, they will have essentially no effect upon the time-varying magnetic flux linking the equilibrium orbit.

The effect of current flowing in conductors 1922 upon n, or the spatial distribution of the time-varying magnetic guide field, may best be understood by reference to Figs. 3, 3A and 4. Figs. 3 and 3A show the Wellknown field pattern for a straight conductor in free space and are particularly related to coil or winding 22, the cross section represented in Fig. 3 being taken on the right hand side of Fig. 1. With the current assumed to be fiowing in a direction into the paper, the field H22 Varies inversely proportional to the distance along XX, a line which is assumed to lie essentially in the plane of the equilibrium orbit. It is apparent that with current flowing in the same direction in both of the windings 21 and 22, the combination the fields therefrom results in cancellation near the center of container 6 or at equilibrium orbit X, an upward field to the right and a downward field to the left. The effect of windings l9 and 20 is illustrated in Fig. 4 by the vectors H19 and H20 resulting independent- 1y from currents in

windings

19 and 20 respectively, the cross sections being taken also at the right hand portion of Fig. 1. As will be observed, the resultant field HT is directed upwardly at the right of the conductors and downwardly at the left with an essentially zero field being present in the center between the conductors. Again the field in the right hand portion of Fig. 1 is strengthened on the outside beyond the equilibrium orbit and weakened for shorter radii, but this time by the action of windings 19' and 20. Thus, if windings 1922 are energized such that currents flow in each of the windings in the' directions above stated, the net field produced thereby is of such a distribution as to make the original field (see Equation 2) more nearly uniform; hence, the

utilization of auxiliary windings 1922 decreases n in the desired fashion. Moreover, if the windings are energized by a current pulse for an initial period including the period of injection of the charged particles the aforementioned loss of charged particles will be prevented.

Although the energization of windings 1922 With a pulse of current during an initial period including the period of injection substantially decreases the loss of charged particles, the decrease in n for radial injection should be as large as possible in order to make it most effective. In any given accelerating apparatus, however, it is obvious that the maximum change in n is from the value determined from the taper of the pole faces to nearly zero, where axial focusing is lost. In order to make the rate of change of n greater, the present invention contemplates superimposing upon the pulsating change in n, a steady state change which makes n larger than that for which the pole pieces of a given accelerator apparatus are shaped. This extension of the range to which n can be decreased is accomplished by passing a direct current through the windings 19-22, whereby the total current through the windings is a series of pulses superimposed upon a direct current component, as will be more fully explained presently. To provide the desired steady state increase in the value of n, the direct current must flow through the respective windings in a direction opposite to the direction of the pulsed current flow.

The aforementioned superimposed current pulses and direct current can be supplied to windings 1922 by means of the circuit schematically illustrated in Fig. 5, wherein reference characters employed heretofore are used to identify like elements. The circuit can be energized from the source of time-varying voltage 13 through an autotransformer 25 and a step-up transformer 26 which comprises a primary winding 27 and a secondary winding 28. During the portion of the cycle of timevarying voltage source 13 when charged particles are not being accelerated in the apparatus of Fig. 1, a pulseforming network 29 which can include capacitors 30 and inductances 31, is charged through a half-wave rectifier tube 32 from the secondary winding 28 of transformer 26. At a desired time during the portion of the cycle of time-varying voltage source 13 when charged particles are being accelerated in the apparatus of Fig. l and a relatively short time before the charged particles are injected, a grid-controlled, gaseous discharge device 33 is rendered conductive by a suitable timed firing circuit 34, whereupon the pulse-forming network 29 is discharged through a non-inductive resistance attenuator 35 and windings 19--22. Resistance attenuator 35 includes

resistors

36, 37 and 38 which, in conjunction with the inductance of windings 19-22, can be employed to control the rate of current built-up in the windings. The direction of the discharge of pulse-forming circuit 29 through windings 1922 is such that conductor 23 is positive with respect to conductor 24, whereby the current pulses flow into the windings through conductor 23 and out through conductor 24.

The direct current component is supplied in a direction opposite to the direction of the pulsed-current flow in windings 1922 by means of a source of direct voltage 39 connected to conductor 24 through a current-responsive meter at an adjustable resistor 41 and a reactor 42. Resistor 41 is employed to adjust the magnitude of the direct current component detected by meter 40, and reactor 42 isolates the direct current circuit from the pulse circuit to prevent interference of the former with the latter.

In Fig. 6 there is illustrated a waveform of the current through windings Ii -22, which can be obtained with the circuit of Fig. 5. At a desired instant, e. g., a fraction of a microsecond before the period Ti of injection of charged particles into the accelerator apparatus, discharge device .33 is rendered conductive by firing circuit 34 to initiate a current pulse represented by curve 43. The direct current component represented by line 44 is superimposed upon the current pulse in a direction opposite that of the pulses. This has the effect of causing the current pulse to start from an assumed negative value as illustrated. From the foregoing explanation, it is apparent that during the rise of the current pulse the value of n is decreased from the valuedetermined by pole piece design and the direct current component to a smaller value. During the relatively small time interval Ti charged particles are injected and accepted by the accelerator apparatus for acceleration. The amplitudes of the above-described oscillations of the charged particles about their instantaneous circles and the equilibrium orbit are, however, decreased by the action of the current pulse, whereby they are successfully accelerated to high energy levels during the acceleration cycle. Atter the pulse has lasted for approximately 2 microseconds, space with respect to the time-varying magnetic guide field and,

therefore, it does not appreciably alter the value of 11 determined by pole piece design except during the desired initial period when the time-varying magnetic guide field is relatively Weak. Nevertheless, it is within contemplation of the present invention that the direct current componentcan be supplied to windings 19-22 only for the duration of the pulse by means of a conventional timing and switching circuit. The timed firing circuit 34 can be one of several conventional circuits or, more particularly, may boot the saturable-strip-initiated type disclosed in my aforementioned Patent No. 2,394,071.

Those familiar with the art to which the present invention is related will readily realize that only one of the pairs of windings (-19, 20 and 21, 22) can be employed, providing the resulting effect uponthe time-varying magnetic flux linking the equilibrium orbit does not aifect the operation of the accelerator apparatus seriously. Moreover, the present invention is not limited to utilization with accelerator apparatus which employs magnetic induction phenomena alone, but can also be used in con r junction with synchrontron apparatus such as that disclosed in United States Patent No. 2,485,409, patented October 18, 1949, by myself and H. C. Pollock and assignedto the'assignee of the present invention. Further application of the present invention can be made in con- :7

nection with non-ferromagnetic accelerating apparatus, e. g. the apparatus disclosed in United States Patent No. 2,465',7 86, patentedMarch 29,1949, by J. P. Blewett and assigned to the assignee of the present invention.

While this invention has been described by reference to particular embodiments thereof, alternative constructions will readily occur to those skilled in the art. It is therefore intended in the appended claims to cover all such equivalent embodiments as may be within the true spirit-and scope of the foregoing description.

What I claim as new and desire to secure by Letters Patent'of the United States is:

1. 'ln apparatus for the acceleration of charged particles along an equilibrium orbit wherein injected. charged particles are accelerated at least partly by the action of a time varying magnetic flux that links the equilibrium orbit and-are constrained to follows paths along the equilibrium orbit-by a time-varying magnetic guide'field inthe vicinity of the equilibrium orbit, the improvement which comprises-.meansjfor varyingthedistribution of thetime-var-yngm g tic gu d fiel i u e tin the t m cry n magnetic flux that links the equilibrium orbit including a first pair of windings axially spaced on opposite sides of the plane of the equilibrium orbit and having essentially the same radius as the equilibrium orbit, a second pair of windings radially spaced on opposite sides of the equilbrium orbit and lying essentially in the plane of the equilibrium orbit, and means for energizing the windings during an initial period including the period when the charged particles are injected with time-varying currents that provide along the equilibrium orbit nearly equal but oppositely directed magnetic fields from the windings in each of said pairs of windings, whereby substantial deviation of the charged particles from the equilibrium orbit is prevented.

2. In apparatus for the acceleration of charged particles wherein injected charged particles are accelerated in oscillatory paths about and along an equilibrium orbit at least partly by the combined action of a time-varying magnetic fiux that links the equilibrium orbit and a time-varying magnetic guide field in the vicinity of the equilibrium orbit, the improvement which comprises means for reducing the amplitudes of the oscillatory paths of the charged particles including a first pair of windings uniformly axially spaced .on opposite sides of the plane of the equilibrium orbit and having essentially the same radius as the equilibrium orbit, a second pair of windings radially spaced on opposite sides of the equilibrium orbit and lying essentially in the plane of the equilibrium orbit, and means for energizing said pairs of windings during at least an initial period including the period of injection of the charged particles with currents that provide along the equilibrium orbit nearly equal but oppositely directed magnetic fields from the windings in each of said pairs of windings, whereby the distribution of the time-varying magnetic guide field is varied during said initial period and the amplitudes of the oscillatory paths of the charged particles are reduced.

3. In apparatus for the acceleration of charged particles wherein injected charged particles are accelerated in oscillatory paths along an equilibrium orbit at least partly by the combined action of a time-varying magnetic flux that links the equilibrium orbit and a time-varying magnetic guide field H that varies with an exponent of n in the vicinity of the equilbrium orbit as set forth in equanon where H is the intensity of the time varying magnetic field in the vicnity of the equilibrium orbit, Ho is the intensity of the guide field at the equilibrium. orbit, R0 is the radius of the equilibrium orbit, R is the radius of a particular point under consideration, and n is an exponent having a value lying between 0 and l, the improvement which comprises means for reducing the amplitudes of the-oscillatory paths of the charged particles including a first pair of windings axially spaced on opposite sides of the plane of the equilibrium orbit and having essentially the same radius as the equilibrium orbit, a second pair of windings radially spaced on opposite sides of the equilibrium orbit and'lying essentially in the plane of the equilibrium orbit, and means for energizing said pairs of windings during at least an initial period including the period of injection of the charged particles with currents that provide along the equilibrium orbit nearly equal but'oppositely directed magnetic fields from the windings in each of said pairs of windings, whereby the exponent n is varied during said at least initial period without appreciably afiFecting the magnitude of the flux and the amplitudes of the oscillatory paths of the charged particles are reduced.

4. Apparatus as in claim 3 in which said means for energizing said pairs of windings comprises a pulse generatin g means connected to said windings.

5. Apparatus as in claim 3 in which said meansfor energizing said pairs of windingscornprises a source of direct current and pulse generating means connected to rent continuously to said windings during operation of said windings to supply respectively direct current and the apparatus. pulsating current of opposite polarities to said windings.

6. Apparatus as in claim 3 in which all of said windings References Cited in the file of this Patent are connected in series and said energizing means com- 5 UNITED STATES PATENTS prises both a source of direct current and a pulse generat- 2 394 070 Kerst Feb 5 1946 ing means connected to said windings.

7. Apparatus as in claim 6 in which said pulse generating means is operative only during said initial period and A said source of direct current is connected to supply cur- 10 McMlnan 1953