US2842705A - Particle accelerator - Google Patents

Description

July 8, 1958 M. CHODOROW PARTICLE ACCELERATOR 2 Sheets-Sheet 1 Filed June 13, 1955 Mirna/m Power 4424 INVENTOR. WWW/v (A/0004 04 PARTlQLIE AKJCELERATGR Marvin Chorlorow, hlenlo Park, Calih, assignor to The Board of Trustees of the Leland Stanford r. University, Stanford, Calif.

Application lune 13, @555, Serial No. 514,88

13 Claims. (Ci. 315 542} This invention relates generally to apparatus providing an interaction between charged particles and electromagnetic waves, and more particularly to linear accelerators for increasing the kinetic energy of electron beams and the like.

In linear accelerators, traveling wave tubes and certain other devices, a beam of charged particles, usually electrons, interacts with an electromagnetic wave having a phase velocity substantially equal to the beam velocity. Any substantial difference between the beam and the phase velocities is generally unfavorable; for example, in a linear accelerator such a difference may cause an electron to move relative to the electric field from a favorable to a less favorable accelerating position, or even to a decelerating position. With bandpass microwave transmission circuits commonly used in linear accelerators and the like, small changes in microwave frequency may produce relatively large changes in phase velocity so that the useful band width of such devices may be much smaller than the circuit band width. This is especially objectionable when narrow band transmission circuits are used, since the useful frequency range may then become quite narrow and require that the microwave frequency be very carefully regulated. One object of this invention is to provide a bandpass microwave circuit for linear accelerators and the like in which an average phase velocity remains relatively unchanged over a range of frequencies, so that the useful band width of the accelerator is a larger portion of the circuit band width.

A useful microwave circuit for devices of the type described is essentially an array of cavities coupled in series by capacitive irises which also serve as passages for a beam of electrons or other charged particles. The amount of coupling between adjacent cavities, and hence the band width of the transmission circuit, increases rapidly with increases in the iris diameter. In general, greater interaction between the-microwave and the charged particles, for a given length of the microwave transmission array, is obtained by making the group velocity of the microwaves relatively low, which requires a narrow band width and relatively small amounts of coupling between ad acent cavities. However, to provide suificiently large passages for the beam of charged particles, capacitive irises may be required which are substantially larger than is desirable for the degree of coupling needed to achieve a desired band width and group velocity. Accordingly, another object of this invention is to provide a narrow band structure for transmitting microwaves at low group velocities, but having relatively large capacitive irises. Other objects and advantages will appear as the description proceeds.

Briefly stated, in accordance with one aspect of this invention, a microwave transmission circuit for linear accelerators and the like is an array of iterated circuit elements coupled in series. Microwave energy is applied to a center one of the circuit elements, and is transmitted outward in opposite directions from the center element toward the two ends of the a On one side 2,342,705 Patented July 8, 1958 of the center element, successive elements of the array are capacitively coupled so that a fundamental component of a transmitted microwave has a phase velocity in the same direction as the group velocity: on the other side of the center element, successive elements of the array are inductively coupled so that: the fundamental component has a phase velocity oppositely directed to the group velocity. Consequently, phase velocities on both sides of the center element are in the same direction along the length of the array, and when the phase velocities are equal, a charged particle can travel from one end of the array to the other with a constant phase relation to the transmitted microwave. if the microwave frequency changes slightly, the phase velocities on opposite sides of the center element will not remain equal, but as one phase velocity increases, the other decreases, so that an average of the two velocities remains relatively unchanged. As a result, the charged particles may move forward with respect to the phase of a microwave transmitted in one half of the array, but they drop back with respect to the phase of a microwave transmitted in the other half of the array, so that the particles tend to regain their former phase position. In this way the useful frequency band width of the device is considerably extended.

According to another aspect of this invention, the circult band width of capacitively coupled circuit elements is reduced by providing inductive coupling which neutralizes only a part of the capacitive coupling and produces electrical characteristics similar to those that would be obtained with a smaller capacitive iris. In this way narrow band widths and low group velocities are obtained, with increased interaction between charged particlcs and the transmitted microwaves, while a relatively large passage is provided for the beam of particles. This inductive coupling may be obtained from a relatively small inductive iris between adjacent cavities in an array.

The invention will be better understood from the following description taken in connection with the accompanying drawings, and its scope will be pointed out in the appended claims. In the drawings:

Fig. 1 is a schematic longitudinal section of linear accelerator apparatus embodying principles of this invention and including a discloaded wave guide type of microwave transmission circuit;

Fig. 2 is a transverse section, taken along line 2-2 of Fig. 1, showing one of several disc-shaped partitions which divide the wage guide into an array of cavities;

Fig. 3 is a transverse section, taken along line 3-3 of Fig. 1, showing another of the partitions;

Fig. 4 is a transverse section showing an alternative wave guide partition which may be used in place of the partition shown in Fig. 2;

Fig. 5 is a transverse section showing an alternative wave guide partition which may be used in place of the partition shown in Fig. 3;

Fig. 6 is a fragmentary longitudinal section showing an alternative microwave circuit construction;

Fig. 7 is a diagram showing phase angle versus frequency characteristics of a microwave transmission circuit; and

Fig. 8 is a diagram illustrating movement of a charged particle relative to the electric field when the particle velocity is slightly different from the microwave phase velocity.

Referring now to Fig. l of the drawing, a hollow cylindrical wave .guide 1 has a plurality of disc-shaped transverse interior partitions, two of which are identified in the drawings by reference numerals 2 and 3, equally spaced along the wave guide axis. These partitions within the wave guide define a linear array of cavities which act as a microwave transmission circuit for transmitting microwave energy outward from a center cavity l toward the two ends of the array. The partitions to the left of center cavity 4 preferably are identical, and may be of the form best shown in Fig. 2. The partitions to the right of center cavity 4 preferably are identical, and may be of the form best shown in Fig. 3. Partitions 2 and 3 have central apertures 5 and 6 alined with the wave guide axis and capacitively coupling the cavities in series. Partitions 2 and 3 also have peripheral apertures 7 and 8 inductively coupling the cavities in series. As is hereinafter more fully explained, the capacitive coupling predominates between cavities to the right of center cavity 4, and the inductive coupling predominates between cavities to the left of center cavity 4.

Microwave electromagnetic energy is provided by any suitable microwave power supply 9 and is coupled into center cavity 4 by an inductive loop 10 or any other suitable coupling means. Microwave energy is transmitted from cavity to cavity through the coupling apertures from center cavity 4 toward each end of the array. Consequently, the transmitted microwave has oppositely directed group velocities on opposite sides of center cavity 4. In the electric field of the wave, the phase angle change between adjacent cavities is positive in the direction of energy flow when the coupling between the cavities is predominantly capacitive, and is negative in the direction of energy fiow when the coupling between the cavities is predominantly inductive. Thus the phase angle between cavities is positive in the direction of energy flow to the right of center cavity 4, and is negative in the direction of energy flow to the left of center cavity 4. However, since the energy flows in opposite directions on opposite sides of center cavity 4, all of the phase angles are positive in the left to right direction along the length of the wave guide.

The electric field associated with iterated structures does not vaiy sinusoidally with distance, but it can be analyzed into sinusoidal components. If 0 is the phase angle per section, and l is the section length, the phase constant B of a typical component is given by the equation where m is a positive or negative integer. The field component for which m=0 is called the fundamental; for other values ofm the components are called spatial harmonies. Some of these components have negative phase velocities and some have positive phase velocities. The phase velocity V of any component is related to the phase constant by the equation V ='w/B where w is 21r times the microwave frequency. The phase velocity V of the fundamental component is related to the phase angle 6 by the equation In a narrow band circuit where variations in w are small in comparison to w and l is a geometrical constant of the circuit, V is approximately proportional to the reciprocal of 6. In the microwave circuit shown in Fig. l, where 0 is positive in the left to right direction, the fundamental component of the microwave has phase velocities directed from left to right along the length of the array on both sides of center cavity 4, and the phase velocities on both sides of the center cavity are equal when both sides of the array have equal values of OH.

4 accelerating the electrons to the energy at which they enter the left-hand end of wave guide 1, and may include additional sections similar to wave guide 1 as well as prior art accelerating and bunching apparatus. Electron beam receiver 13 may be any apparatus for utilizing the high energy electrons leaving the right-hand end of wave guide 1, and many include additional sections similar to wave guide 1 or other accelerator apparatus for further increasing the kinetic energy of the beam. Many types of apparatus for generating and utilizing high velocity electron beams are known to those skilled in the art, and need not be described for an understanding of the present invention. Hollow wave guide 1 is evacuated in the usual manner to permit substantially unimpeded elec tron flow along the wave guide axis. If desired, magnctic or other focusing means may be employed to limit the radial dispersion of the beam.

For electrons 12 to interact favorably with the microwave and gain energy from the microwave electric field, the electron velocity must be reasonably close to one of the microwave phase velocities, which may be the phase velocity of the fundamental component or may be the phase velocity of one of the spatial harmonics. In linear accelerators the fundamental component is generally used, and the electron beam travels in the same direction as the phase velocity of the fundamental component--that is from left to right in Fig. l of the drawings. If a special harmonic having a phase velocity directed from right to left were to be used, the positions of electron beam generator 11 and electron beam receiver 13 would be interchanged so that the electron beam would travel from right to left. In either event, it is desirable that the electron beam velocity and the phase velocity being used should be substantially equal and in the same direction, so that electrons having a favorable phase position relative to the miocrowave will retain that favorable position as they travel along the length of the wave guide. Consequently, it is essential that unavoidable small changes in the microwave frequency must not produce large changes in the microwave phase velocities.

The microwave circuit shown in Fig. 1 is a bandpass circuit capable of transimtting only a limited range of frequencies. That portion of the circuit to the right of center cavity 4 can transmit frequencies within a pass band bounded by a lower cut-off radian frequency 00 and an upper cut-off radian frequency (.0 That portion of the circuit to the left of center cavity 4 can transmit frequencies within a pass band bounded by a lower cut-off radian frequency .w' and an upper cut-01f radian frequency w' These two pass bands overlap so that both pass bands include the operating frequency of the accelerator; preferably the two pass bands are identical, so that (0 equal (0' and (v equals w' In common with other iterated wave-transmission structures, this discloaded wave guide has characteristics analogous to those of a lumped-impedance filter circuit in that a frequencydependent phase shift in the transmitted wave occurs between adjacent circuit sections. In general, the phase shift per section varies throughout the pass band from O radians at one cut-off frequency to 1r radians at the other cut-off frequency.

Refer now to Fig. 7, which shows the phase shifting characteristics of the Fig. 1 microwave transmission circuit when the left and right halves of the circuit have identical pass bands. In Fig. 7 the phase angle 0 between electric fields in adjacent cavities is plotted as a function of the radian frequency w of the microwave. The lower and upper cut-off frequencies are represented by vertical lines at 0: and 1.0 respectively. The preferred operating frequency w of the accelerator is represented by a vertical line 14 midway between o and (v The magnitude of the phase angle between adjacent capacitively coupled cavities is represented by a solid-line curve 15, and the magnitude of the phase angle 0 between adjacent inductively coupled cavities 1s represented enclaves by the broken-line curve 16. It will be noted that curve increases from 0:0 at the lower cut-off frequency :0 to 6:11- radians at the upper cut-off frequency w while curve 16 decreases from 0=1r radians at the lower cut-olf frequency (11 to 0:0 at the upper cut-off frequency 01 In a narrow band circuit, where (o -o /w is small, the phase velocity of the microwave fundamental component is approximately proportional to the reciprocal of 0 and changes by a large amount with a relatively small change in w. In prior accelerators, these lar e phase velocity changes restrict the useful frequency range of the apparatus to a small fraction of the circuit bandwidth since favorable interaction with the electron beam requires a phase velocity nearly equal to the beam velocity. According to the present invention, while the phase velocity in one portion of the microwave circuit is increasing the phase velocity in another portion of the microwave circuit is decreasing so that an average of the two phase velocities changes by a relatively small amount and the usable frequency range of the apparatus is extended to a much larger proportion of the circuit band width.

Assume, for example, that the apparatus is designed to operate at point if], Fig. 7, where curves 15 and 16 cross. Operation at point 17 will occur whenever power supply 9 provides microwave energy having the radian frequency w represented by line 14, which intersects curves 15 and 16 at point 17. At this point the magnitude of the phase angle 0 is the same on both sides of the microwave transmission circuit and, assuming equal spacings 1 between the partitions 2 and 3, the phase velocities on both sides of the circuit are equal to each other and to the electron velocities. Now, assume that the microwave frequency inadvertently changes by an amount Aw, so that operation occurs at point 18 in the right-hand half of the circuit and at point 19 in the lefthand half of the circuit. This frequency change causes an increase M in the magnitude of the phase angle between adjacent capacitively coupled cavities, and consequently a decrease in the phase velocity in the righthand half of the circuit; and causes a decrease M in the magnitude of the phase angle between adjacent inductively coupled cavities, and consequently an increase in the phase velocity in the left-hand half of the circuit. These two opposite changes in phase velocity tend to compensate each other, so that the average phase velocity over the entire length of the microwave circuit changes by a relatively small amount.

The advantages of this arrangement can be better understood by reference to Fig. 8. ln Pig. 8, sine wave curve 29 represents a voltage wave which travels from left to right along the length of waveguide 1 at the phase velocity of the microwave component under consideration. Assume that an electron enters the left-hand. side of the waveguide at a phase position represented by point 211., in which position the electron encounters. a strong electric field which accelerates the electron or adds to its kinetic energy. In a linear accelerator, the electron may already be traveling at an extremely high velocity when it enters waveguide ll, so that the acceleration produced by the electric field is largely a relativistic increase in mass, rather than a substantial change in the electrons speed of travel. If the electron velocity is exactly equa to the phase velocity of wave 20, the electron retains its favorable position at point 21 and continues to gain energy during the entire period of its travel through waveguide 1.

Now assume that an increase in frequency A9 has increased the phase velocity in the left-hand portion of the microwave circuit to a value greater than the electron velocity. The electron progressively falls behind the travelling wave, falling back to points 22, 23, 24 and 25 as the electron and wave travel toward the center of the microwave circuit. The same frequency change which produced an increase in phase velocity on the left-hand side of the microwave circuit produces a decrease in phase velocity on the right-hand side of the microwave circuit, so that after the electron passes the center cavity 4, it begins to advance on the travelling wave and passes successively through points 24, 23, 22 and arrives back substantially at point 21 at the time when the electron leaves the right-hand end of waveguide 1. Although the electron, during its travel through waveguide 1, oscillated somewhat in its phase relation to the microwave, it remained in a relatively favorable phase relation for acceleration, and it had substantially the same phase position at the exit end of the waveguide that it had upon its entrance into the waveguide. The latter point is especially important when several accelerating sections are connected in cascade, since electrons arc delivered to the next accelerating section in a favorable phase position.

By way of comparison, consider what happened in prior art accelerators when an inadvertent frequency change caused the phase velocity to exceed the electron velocity by a comparable amount. An electron entering at the phase relation represented by point 21 would continually fall back relative to the wave through points 22, 23, 24, 25, 26, 27, 28 and 29, so that it would be in a relatively unfavorable phase position by the time it reached the exit end of the waveguide. Furthermore, if several accelerator sections were connected in tandem, the electron at point 29 would not be delivered to the following section in a favorable position for further acceleration. In fact, it would probably fall further behind into a decelerating position, where it would lose energy to the electric field.

From the foregoing, it is evident that this invention substantially reduces the change in phase position of an electron, or other charged particle interacting with a travelling microwave, in consequence of a change in the microwave frequency, so that the useful frequency bandwidth, which is limited by the extent to which changes in the electron phase position are permissible, becomes a substantially larger part of the circuit band width. This leads to several important advantages. For a given circuit band width, greater variation in the microwave supply frequency is permissible, and less careful frequency regulation is required. Conversely, for a given frequency stability, the circuit bandwidth can be made narrower by reducing the amount of coupling between adjacent cavities, as is hereinafter more fully explained, so that the group velocity of the microwaves is lower and a greater degree of interaction between the charged pa ticles and the microwave occurs within a given waveguide length. in other words, by decreasing the circuit band width, the same amount of electron acceleration can be obtained in a shorter, more compact accelerator.

In each of the cavities within waveguide 1, the electric field is strongest near the waveguide axis and decreases to substantially zero near the side walls of the cylindrical waveguide. The magnetic field, on the other hand, is maximum near the side walls of the waveguide and is substantially zero at the waveguide axis. Consequently, the central apertures 5 and 6 couple the electric field between adjacent cavities, and this provides capacitive coupling between the cavities. On the other hand, the peripheral apertures 7 and 8 couple the magnetic fields of adjacent cavities, and this provides inductive coupling between the cavities.

In general, the amount of coupling provided by each aperture increases with increases in aperture size. For example, circular aperture in a region of uniform electnic field provides an amount of capacitive coupling which varies substantially as the sixth power of the aperture diameter, or the third power of the area. For convenience, the capacitive apertures 5 and 6 are generally made circular; but the inductive apertures are preferably elongated when a large apenture is desired, since an increase in radial width would carry the aperture into 7 regions of low magnetic field strength, and hence provide less inductive coupling than a corresponding increase in the circumferential length of the aperture. The small inductive aperture 8 may be circular for convenience,

as best shown in Fig. 3, while the large inductive aperture 7 is preferably made areuate in shape, as shown in Fig. 2, to provide a larger amount of inductive coupling. Of course, the inductive coupling could be increased by providing a plurality of smaller apertures, spaced circumferentially around the periphery of partition 2, rather than by using a single, large aperture. Various shapes and forms of apertures may be used without departing from the principles of this invention.

The net coupling between adjacent cavities is equal to the difference between the capacitive coupling provided by the central aperture and the inductive coupling provided by the peripheral aperture. If the two apertures provided exactly equal amounts of coupling, a resonant condition would be obtained in which there would be substantially no transfer of electromagnetic energy between adjacent cavities. This would correspond to zero group velocity, and zero circuit band width. In partitions 2, located to the left of center cavity 4, the inductive aperture 7 is somewhat larger than necessary to neutralize the capacitive coupling provided through capacitive aperture so that the predominant coupling between cavities on the left-hand side of the array is inductive. The partitions 3, located to the right of center cavity 4, have inductive apertures 8 somewhat smaller than would be required to neutralize the capacitive coupling provided through capacitive aperture 6, so that the predominant coupling between cavities on the right-hand side of the array is capacitive. In fact, the principal purpose of aperture 8 is to reduce the effective size of aperture 6 for purposes hereinafter more fully explained.

if the diameter of aperture 5 is increased, the same amount of net coupling can be obtained by making a corresponding increase in the size of inductive aperture 7. This is best illustrated in Fig. 4, where partition 2 has a relatively large central aperture 5' and two inductive apertures '7 and 7 which can be proportioned to give the same amount of inductive net coupling as was obtained with partition 2 of Fig. 2. Conversely, if the size of the central aperture is reduced, the same amount of net coupling can be obtained by reducing the size of the inductive aperture. This is shown in Fig. 5, where the partition 3 has no inductive aperture at all, but has a small central aperture 6' which may be proportioned to give the same amount of capacitive net coupling as was obtained with the partition 3 shown in Fig. 3.

Whenever the net coupling is capacitive, the magnitude of plotted as a function of or has a positive slope, as is shown by curve 15 of Fig. 7. Whenever the net coupling is inductive the magnitude of 0 plotted as a function of to has a negative slope, as is indicated by curve 16 of Fig. 7. The circuitband width (w -w increases as the amount of net coupling increases and so does the magnitude of the slope of curves 15 and 16. The group velocity V is given by the differential equation Consequently the group velocity corresponding to any point on curves l and 16 increases as the circuit band width increases. To obtain a small band width and a correspondingly low group velocity, the net coupling between adjacent cavities, whether inductive or capacitive, is made small in magnitude.

To get the greatest amount of electron acceleration in a given length of the waveguide, the circuit band width should be as narrow as is permitted by the frequency stability of the microwave power supply. To obtain a narrower circuit band width, the net coupling between adjacent cavities must be small. In the case of inductively-coupled cavities, a small amount of net coupling is obtained by making the inductive coupling apertures 7 just'large enough to provide slightly more inductive coupling than is needed to neutralize the capacitive coupling provided by the central apertures 5. The size of central apertures 5 can be chosen to provide a passage of' best size for the electron beam. In the case of capacitively coupled cavities, the desired amount of net coupling can be obtained simply by choosing the proper size for the central aperture 6 without using any inductive coupling at all, as is shown in Fig. 5. Very often, however, this procedure would lead to a central aperture of insufiicient size to accommodate the electron beam properly. One principle of this invention is that the central: aperture can be increased to the size which best accommodates the electron beam, and that a small amount of net coupling can still be obtained by adding a relatively small inductive aperture 8, which provides'slightly less inductive coupling than is needed to neutralize the capacitive coupling provided by the central aperture so that the net coupling remains capacitive but is reduced in amount. With respect to the net coupling between cavities, aperture 8 reduces the effective size of aperture 6.

Coupling between adjacent cavities may be provided by means other than apertures. For example, Fig. 6 shows a waveguide section with partitions 2" having cylindrical collars 30 about the central apertures to form reentrant cavities between adjacent partitions, as shown. Capacitive coupling is provided through the central apertures. Inductive coupling is provide by wire coupling loops 31, which extend through openings near the periphery of partitions 2 and are joined at each end to the cylindrical wall of waveguide 1. Coupling loop 31 links the magnetic fields on each side of partition 2" and acts a an inductive coupling between the two cavities.

Again referring to Fig. 7, in a preferred embodiment of this invention curves 15 and 16 cover identical frequency rangesthat is, the cut-off frequencies w, and 012 of the microwave circuit to the right of center cavity 4, Fig. l, are the same as the cut-off frequencies w, and :0 of the microwave circuit to the left of center cavity 4. To obtain this relation, the amount of net capacitive coupling between cavities to the right of center cavity 4 and the amount of net inductive coupling between cavities to the left of center cavity 4 must be so related that the circuits on both sides of the center cavity have equal circuit band widths. Also, the two bandpass regions must be located at identical places in the frequency spectrum. The amount of net coupling, and hence the band width, can be adjusted by varying the size of the coupling apertures, as hereinbefore explained. The location of the passband in the frequency spectrum can be adjusted by varying the internal diameter of waveguide 1. In general, to obtain overlapping passbands for the inductively coupled and capacitively coupled circuits, the internal diameter of waveguide 1 must be smaller on the inductively-coupled side of the waveguide than it is on the capacitively coupled side of the waveguide, as is shown in Fig. 1. By calculation or experimental cut-andtry techniques, or both, an appropriate waveguide diameter and appropriate aperture sizes can be chosen for each side of the microwave circuit, so that the two passbands occupy overlapping portions of the frequency spectrum and so that the desired operating frequency lies within both passbands.

Although the two passbands preferably occupy identical positions in the frequency s ectrum, as shown in Fi 7, this is not absolutely essential, and in some cases it may be desirable to shift one passband relative to the other: provided, however, that the bands must overlap and both bands must include the desired operating frequency. Assume, for example, that the desired radian frequency w is represented by line 14, Fig. 7, which intersects curves 15 and 16 at point 17. With the circuit characteristics shown in Fig. 7, this gives a phase angle 0 between adjacent cavities of approximately 110. Now

assume that the desired phase angle 0 is corresponding to points 32 and 33 of Fig. 7. If the portion of hollow waveguide l. to the right of center cavity 4 has its interior diameter decreased, curve 15 of Fig. 7 will be shifted toward the right, and if the portion to the left of center cavity 4 has its interior diameter increased, curve 16 of Fig. 7 will be shifted toward the left. By a suitable choice of the waveguide diameters, points 32 and 33 can be brought into coincidence on line 14 at the desired operating frequency, and the desired phase angle of 90 per section can be obtained.

It will also be understood that the phase angle between cavities to the left of center section a need not be equal to the phase angle 6 between cavities to the light of center cavity 4, provided suitable adjustments are made in the spacings 1 between partitions to obtain equal phase velocities on both sides. For example, again referring to Fig. 7, assume that it is desired to operate at the frequency of points 13 and 19. In this case, the phase angle 0 between capacitively coupled cavities will be greater than the phase angle 0;, between inductively coupled capacities. However, in each case the phase velocity is inversely proportional to the value of 0/1, where l is the spacing between adjacent partitions. By spacing the partitions 2 somewhat closer together than the spacing between partitions 3, the same phase velocity can be provided on both sides of center cavity 4, despite differences in the phase angles.

As another alternative, the number of cavities on the righthand side of center cavity 4 need not be exactly the same as the number of cavities on the left-hand side of center cavity Assume, for example, that there are twice as many capacitively coupled cavities as there are inductively coupled cavities, and that the operating frequency is represented by point 17 of Fig. 7. Upon a change Aw of the operating frequency, there will be opposing changes M 'and M of the phase shifts between capacitively coupled cavities and inductively coupled cavities, respectively. For these opposite phase changes to compensate each other fully A0 nowshould be approximately twice as large as M since under the assumed conditions there are twice as many of the 0 phase shifts as there are 6 phase shifts. This can be accomplished by making the slope of curve 16 substantially twice as great in magnitude as the slope of curve 15 which, in turn, is accomplished by making the band Width of the circuit containing inductively-coupled cavities only half as large as the band width of the circuit containing capacitively-coupled cavities. Conseqently, the term center cavity, as used here, does not necessarily refer to the cavity in the precise center of the array; rather, it refers to that cavity, located in a generally central portion of the array and not at either extremity, into which microwave energy isintroduced from the supply 9.

Although this invention has been particularly described in connection with linear accelerators for increasing the kinetic energy of electron beams, it will be. appreciated that other charged particles, such as protons and other atomic nuclei, may be used in place of electrons, and that principles of the invention are likewise applicable to traveling wave tubes and other devices providing an interaction between charged particles and electromagnetic waves. It will be further understood that the invention is not limited to specific embodiments herein illustrated and described, and that the following claims are intended to cover all changes and modifications which do not depart from the true spirit and scope of the invention.

What is claimed is:

1. In combination, an array of circuit elements coupled in series, means supplying electromagnetic energy to a center one only of said circuit elements so that said array transmits electromagnetic energy from said center element in two opposite directions, coupling means providing positive phase angles less than 180 degrees in the direction of energy transmission between successive ones adjacent to said array.

2. Apparatus providing interactions between charged particles and an electromagnetic wave, comprising an an ray of iterated circuit elements forming a band-pass wavetransmission circuit, means providing in said array an electromagnetic wave having frequency-dependent phase angles of less than degrees between successive ones of said elements, means coupling a first plurality of said circuit elements so that the phase angle between successive elements increases in magnitude with increasing frequency over the pass band of said circuit, mean coupling a second plurality of said circuit elements so that the phase angle between successive elements decreases in magnitude with increasing frequency over the pass band of said circuit, whereby said wave has fundamental and spatial harmonic components with average phase velocities which are relatively independent of frequency, and means providing charged particles moving adjacent to said array at a velocity substantially equal to one of said average phase velocities.

3. Apparatus providing interaction between charged particles and electromagentic microwaves, comprising a linear array of band-pass circuit elements coupled in series to form a microwave transmission circuit, means supplying microwave energy to an intermediate one only of said circuit elements so that microwaves are transmitted in lengthwise directions along said array with oppositely directed group velocities on opposite sides of said intermediate element, coupling means between successive circuit elements on one side of said intermediate element providing a phase velocity of the fundamental microwave component directed similarly to the group velocity, coupling means between successive circuit elements on the other side of said intermediate element providing a phase velocity of the fundamental microwave component directed oppositely to the group velocity, whereby said phase velocities on both sides of said intermediate element are similarly directed lengthwise along said array, and means providing a beam of charged particles traveling in a lengthwise direction adjacent to said array and interacting with said microwaves.

4. Apparatus providing interaction between charged particles and electromagnetic microwaves, comprising means defining a linear array of cavities capacitively coupled in series by a plurality of linearly alined apertures to form a microwave transmission circuit, means inductively coupling a plurality of said cavities in series so that the predominant coupling between adjacent cavities atone end of said array is inductive and the predominant coupling between adjacent cavities at the other end of said array is capacitive, means for supplying microwave energy to a center part only of said array, and means directing charged particles through said linearly alined apertures.

5. In combination, a hollow waveguide having a plurality of interior partitions defining a linear array of cavities, said partitions having central apertures capacitively coupling said cavities in series, means for supplying microwave energy to a center one only of said cavities, and inductive coupling means providing predominantly inductive series coupling between the cavities on one side of said center cavity.

6. A linear accelerator for increasing the kinetic energy of moving charged particles, comprising a hollow cylindrical waveguide having a plurality of disc-shaped transverse interior partitions defining a linear array of cavities, said partitions having axially alined central apertures capacitively coupling said cavities in series, means supplying microwave energy to a center one only of said cavities, said hollow waveguide having a smaller uniform inside diameter on one side of said center cavity than on the other side, inductive coupling means providing predominantly inductive series coupling between the cavities on the smaller-diameter side of said waveguide, the series coupling between cavities on the larger-diameter side of the waveguide being predominantly capacitive, and means directing charged particles along the waveguide axis through said central apertures.

7. A linear accelerator for increasing the kinetic energy of an electron beam, comprising an evacuated hollow waveguide having a plurality of transverse interior partitions equally spaced along the waveguide axis, means supplying microwave energy to a center part only of said waveguide so that microwave energy is transmitted from said center part toward each end of the waveguide, said partitions having capacitive coupling apertures located at positions of strong electric field in the transmitted microwave, a plurality of said partitions also having inductive coupling apertures located at positions of strong magnetic field in the transmitted microwave, the predominant coupling being capacitive through partitions on one side of said center part and being inductive through partitions on the other side of said center part, and means directing an electron beam through said capacitive coupling apertures.

8. A linear accelerator for increasing the kinetic energy of an electron beam, comprising a linear array of bandpass circuit elements coupled together in series to form a microwave transmission circuit, substantially one-half of said circuit elements at one end of said array being inductively coupled in series and substantially one-half of said circuit elements at the other end of said array being capacitively coupled in series, means supplying microwave energy to a center portion only of said array so that microwave energy is transmitted in one direction through said inductively coupled elements and in the opposite direction through said capacitively coupled elements, all of said elements having band-pass filter characteristics within substantially the same range of frequencies, and means providing a beam of electrons moving in a lengthwise direction adjacent to said array.

9. A microwave transmission circuit comprising a hollow waveguide having a plurality of interior partitions defining a linear array of cavities, said partitions having central apertures capacitively coupling said cavities in series and having peripheral apertures inductively coupling said cavities in series, said peripheral apertures being sufficiently small relative to said central apertures that the coupling between said cavities is predominantly capacitive;

10. Apparatus providing interaction between charged particles and an electromagnetic wave, comprising a hollow cylindrical waveguide having a plurality of discshaped interior partitions defining an array of cavities, means supplying microwave energy to said waveguide, said partitions having linearly alined central apertures capacitively coupling said cavities in series, whereby said waveguide transmits microwaves having a fundamental component with a phase velocity and a group velocity in the same direction, said partitions having peripheral apertures inductively coupling said cavities in series, said inductive coupling being smaller than said capacitive coupling, whereby said group velocity is reduced, and means providing charged particles traveling through said central apertures.

11. A linear accelerator for increasing the kinetic energy of an electron beam, comprising an array of iterated circuit elements forming a band-pass microwave transmission circuit, means supplying microwave energy to said transmission circuit, means capacitively coupling said circuit elements in series and providing a predominantly capacitive coupling between adjacent elements, means inductively coupling said circuit elements in series, said inductive coupling being less than said capacitive coupling, whereby the band-width of said transmission circuit is reduced, and means providing an electron beam adjacent to said array.

12 Apparatus providing interaction between charged particles and an electromagnetic wave, comprising a linear array of iterated circuit elements forming a band-pass microwave transmission circuit, means supplyingmicrowave energy to a center element only of said array so that microwave energy is transmitted outward from said center element toward both ends of said array, means capacitively coupling said circuit elements in series, means inductively coupling the same circuit elements in series, said capacitive coupling being predominant over said inductive coupling between elements on one side of said center element, said inductive coupling being predominant over said capacitive coupling between elements on the other side of said center element, and means providing charged particles moving adjacent to said array.

13. A linear accelerator for increasing the kinetic energy of a beam of charged particles, comprising an evacuated hollow cylindrical waveguide having a plurality of disc-shaped transverse interior partitions equally spaced along the waveguide axis to define a linear array of cavities, means supplying microwave electromagnetic energy to the center one of said cavities, said partitions having central apertures linearly alined along the waveguide axis and capacitively coupling said cavities in series, said partitions also having peripheral apertures inductively coupling the same cavities in series, the partitions on one side of said center cavity having peripheral apertures which are small relative to the central apertures so that the predominant coupling between cavities is capacitive, the partitions on the other side of said center cavity having peripheral apertures which are large relative to the central apertures so that the predominant coupling between cavities is inductive, the interior diameter of said. waveguide being smaller on the inductively-coupled side than on the capacitively-coupled side, whereby said. waveguide transmits on respective sides of said center cavity microwaves with low group velocities oppositely directed on opposite sides and phase velocities similarly directed on opposite sides, and means providing a beam of charged particles traveling through said central apertures and along the waveguide axis.

References Cited in the file of this patent UNITED STATES PATENTS 2,439,401 Smith Apr. 13, 1948 2,636,948 Pierce Apr. 28, 1953 2,653,271 Woodyard Sept. 22, 1953

Images