CA2832816C - Accelerator - generator - Google Patents

Abstract

It is well known that the converting of a continuous relativistic electron beam to the sort of the modulated electron beam is the difficult task because all electrons of these beams have identical velocities. And neither a modulation of the beam current by a grid nor a grouping of electrons of the beam like in a klystron can practically solve these problems. So the offered invention makes possible the effective converting of the continuous relativistic electron beam to the sort of the modulated relativistic electron beam by means of use of the differences of kinetic energies of beam's electrons, that crossed resonant cavity, for the further dividing of the initial beam at least into two flows in the technical means with the separating apart of low-energy electrons into the first flow from high-energy electrons of the second flow. On the base of these conversions of the initial relativistic electron beam we offer the relativistic high-frequency generator with efficiency about 80%.

Description

Accelerator ¨ generator Technical field The invention is related to a technical area of supplying the high frequency power to an accelerator of the type that is described in patent US 5440211 (or US 5095486) and the offered invention describes a device for providing the above purpose.
Background of the invention The well-known accelerator Rhodotron TM is the most powerful type of the electron accelerator among all microwave electron accelerators. The accelerators of this type cover 80% of the total power of all accelerators being used in the world. Rhodotrons' radio-frequency generator usually is based on the most powerful radio-frequency tubes such as tetrodes or diacrodes.
Since each tube connects to the cavity of the accelerator by means of an intermediate cavity (as in article H. Tanaka -BEAM DYNAMICS IN A CW MICROTRON FOR INDUSTRIAL
APPLICATIONS", Proceedings of EPAC 2000, Vienna, Austria) that does not allow connection of two or more tubes using parallel scheme to the accelerator cavity without substantially complicating the accelerator. This fact limits the total power of the electron beam at the output of the accelerator and, in turn, limits the efficiency of the applications where Rhodotron TM is used.
A new device was suggested for supplying the Rhodotron TM with the additional high frequency power some time ago in the patent WO-2008-138998. This device does not contain any additional radio-frequency tubes and it is the closest to the device that will he described in the offered invention. More precisely, these two devices are similar only by their structure intended for transferring the kinetic energy of an high current electron beam, that is accelerated outside the accelerator cavity, into the energy of the electromagnetic field inside this cavity; however, the methods of their operations are different and there are several distinctive features of their constructions.
As shown in the Fig. 1 (copy of the patent WO-2008-138998), this structure comprises a DC
HV power supply (0,5-1,0 MV and 1-3 A) - (86), a cathode of the injector -(66) with a grid to control the beam current - (68), an accelerating tube for the distribution of the potential of electric accelerating field - (70), a high frequency power supply - (80) for delivering the electromagnetic power into the coaxial cavity - (50, 52) of the accelerator through the loop -(78) and all of these devices are placed in the tank - (88).

2 This aggregate of elements is executing the method where under the high potential from the cathode (66), the electron beam (60) is injected into the accelerating tube (70) through the grid (68).
The grid is modulating a current of this beam with the resonant frequency of the accelerator cavity (50, 52). Then the modulated beam is accelerated by going through the accelerating tube and then the maximums of the beam current come into the cavity in phase where the electrons of the beam return their kinetic energy to the electromagnetic field during the pass by electrons through the cavity. This energy is added to the energy that is delivered by the HF
generator (80) into the cavity through the loop (78).
In this method the accelerating tube of the straight action accelerator usually comprises a sequence of conductive disks interleaved by the isolated rings and has the resistive divider usually used for distribution of the potential of the acceleration along the accelerating tube.
Each disk is contacting with the definite element of this divider for distribution of the potential according to the preset gradient of the electric field. The distance between neighboring disks is usually equal to a few centimeters and the electric current of the divider is no more than 1-2 mil liamps in order to decrease the divider losses.
When a continuous unmodulated beam goes through the accelerating tube, its own escorted electromagnetic field is spreading from the beam to the walls of the vacuum tank without of any interaction with the accelerating tube since this field has a constant character and does not depend on the time. When the modulated beam goes through the accelerating tube its own escorted electromagnetic field has the alternative character and it is interacting with the accelerating tube, inducing back-current through the resistive elements of the divider because the distance between the disks is much less than the length-wave of the beam modulation and also the field cannot spread across the accelerating tube. As the result, under the accelerating, the electron current of the modulated beam cannot exceed a few percents of the divider current and this device cannot provide a lot of additional high-frequency power into the accelerator cavity.
The current about 1-3 ampere in the beam cannot be accelerated by the method of above patent WO-2008-138998. But this restriction does not exist if the low-voltage injector delivers unmodulated beam with such current into the accelerating tube. In other words the relativistic injector of continuous unmodulated electron beam can be formed from mentioned low-voltage injector and accelerating tube. The beam with such currents can be accelerated in the straight action accelerator when the additional focusing solenoid is placed on the outer side of the tank (80) to focus the beam of the relativistic injector.

3 The generator on the base of the offered invention provides up to 2-3 MW to the load in comparison with the most powerful diacrode TII-628 from Thales used in Rhodotron's generator that provides up to 1 MW. The efficiency of the said generator reaches 80% in comparison to 56%
of TH-628.
Summary of the Invention Accelerator-generator comprises at least one high-frequency cavity and a relativistic electron injector having an exit for transmitting the continuous unmodulated electron beam from this injector into this cavity for action upon all electrons of this beam to change their energies. The cavity provides accretion of energy in one part of electrons and reduction of energy of other part of electrons of this beam under going through of this cavity, providing concentration of all initial energy of electron beam in that part of electron beam that corresponds to the accelerating phase of field in the cavity and which next energizes the same or another cavity, transforming the energy of electron beam into the energy of the electromagnetic field in the cavity with great effectiveness.
Brief Description of the Drawings The invention contains following drawings:
FIG. 1- is copied Fig. 4 from patent WO-2008-138998 FIG. 2- demonstrates scheme of the first part of the offered device.
FIG. 3- demonstrates the dependence between the integral of the beam's kinetic energy from the increment of impulse of electrons in cavity after their accelerating or decelerating.
FIG. 4- demonstrates the scheme of the second part of the offered device for dividing the initial flow into two parts.
FIG. 5- shows the scheme of the accelerator-generator as a powerful I-IF
generator.
FIG. 6- shows the scheme of the accelerator-generator as a powerful HF
generator or as a fIV
injector for the accelerator with several cavities.
Detailed Description The continuous unmodulatcd flow of electrons accelerated previously up to the intended initial energy is acted on by the electric component of high frequency electromagnetic field of the cavity which is energized by the external generator. This generator compensates the power losses in the cavity's walls and supports the preset definite level of electromagnetic field in the cavity when there are different changes of the initial electron beam's current and the cavity's load. Under this action the

4 said field in the cavity accelerates the first part of beam electrons that meet the accelerating phase of the electromagnetic field in the cavity and decelerates the other part of beam electrons that meet decelerating phase of the electromagnetic field in the cavity. And then this continuous beam with electrons that have the different and periodic modulated energy is transported to the next area where the flow is acted by the force of transverse magnetic field which deflects electrons of these parts of the beam variously. The said flow after the crossing of this area filled up by magnetic field becomes divided up into two flows each of which already has the density of the modulated current.
After that the first flow with increased kinetic energy of electrons, is transported to the next area for further using and the second flow with minimum kinetic energy of electrons, is transported to the dump for elimination.
The first part of the offered device is shown in Fig.2. It comprises the outer and inner cylinders of the accelerator's coaxial cavity - (1), a high frequency generator ¨(2) for energizing the said cavity through the loop ¨(3), and a relativistic injector (10) that comprises a DC HV
power supply - (7) connected to the injector's cathode ¨(4), a DC power supply ¨(6) for varying the current of the injected unmodulated electron beam by means of the grid ¨ (5), an accelerating tube ¨ (8) to accelerate the electron beam from the cathode (4) and to transport this relativistic unmodulated continuous uniform beam of electrons ¨(9) into the coaxial cavity (1). All these devices are placed in a tank ¨ (11a) and the coil of the focusing solenoid ¨ (11) is positioned outside of this tank.
This part of device works as follows. The injector's cathode (4) is connected to the high-voltage output of the power supply (7). The current of the injector is controlled by the output voltage of power supply (6) connected to the cathode (4) and the grid (5). After that the continuous electron beam from the injector goes through the accelerating tube (8) and electrons acquire kinetic energy that is equal to the voltage of the power supply (7), before entering the coaxial cavity (1) that is energized by the high-frequency generator (2).
All electrons in the beam while crossing the cavity acquire additional impulse in accordance to the phase of each electron when entering the cavity and in accordance to the amplitude of the electromagnetic field in the cavity. This increment of the impulse can be calculated by the following formula (1):

5 A p = e0fEp(p, t, yO)dt = 2(e0/c)Asin (90+ RoVci(1/Q) cosQdQ (1) Where:
c - velocity of the light and u ¨velocity of the electron eo, mo ¨ charge, mass of the electron p = moyu, = (1- pz) , p = uic r, R ¨ radius inner and outer conductors of the coaxial cavity = 2nf, f = resonant frequency, To ¨ input electron phase Ep (p, t. (p0) = (1/p) Asin (ot + (p0) ¨ radial component of electric field in the cavity.
As the total impulse of electrons after crossing the cavity is equal to the sum po (initial impulse at the entrance to the cavity) and Ap (the increment of impulse at the exit from the cavity), the integral of the total impulse along the phase 9 = 90+Rw/c from 0 up to 2n is equal to 2np0.
The right part of the formula (1) was obtained under the assumption that all electrons of the beam have velocities which are approximately equal to the velocity of light.
It is met if Ap<<po and the beam's kinetic energy matching po is equal to 0.7-1.0 Mev before the entrance to the cavity. The results of the precise simulation for the actual Rhodotron sizes (r and R) give similar magnitude of the impulse compared with the formula (1) even for Ap¨po. The shift of phase changes a slightly for the electrons which are decelerated by the electromagnetic field in the cavity in comparison to (Rifle). Hence this expression (])can he used for analysis of the behavior of the electrons in the beam after crossing the cavity.
All electrons disposed in each transverse section of the beam have the same the input phase 90 before the cavity and also they have the same the output phase after the cavity. The transverse motion was not taken into account since all electrons had relativistic velocities. In this case the integration along the phase is the same as the integration along the length of the beam. The kinetic energy of the electrons corresponds to their impulses by means of the relativistic invariant and hence this energy can be calculated by means of the invariant using the total impulse of electrons after crossing the cavity, if Ap¨po. The integral of the beam's kinetic energy along the length of the beam for the interval 0 < 9 < 22 after the crossing of the cavity is not equal to the same integral of the kinetic energy when the beam has not yet crossed the accelerated cavity. The calculation of this integral can be done using numerical integrating. The character of dependence of this integral from

6 Ap is shown in Fig. 3 by the line ¨ a, which claims that the HF generator, energizing the cavity, supplies nonzero electromagnetic power during the process of the crossing of the cavity by the continuous electron beam and increases the total kinetic energy of the beam, if Ap approaches to po.
Furthermore the part of the beam that is decelerated (for TE <(p <21t) returns back its energy to the electromagnetic field and hence the remaining part of the beam (for 0 <p it) it) accumulates the power from the HF generator and from the first part of the beam too. The parts of the said integral of the beam's kinetic energy for the intervals 0 < < it and it <p < In are shown in Fig.3 by the line ¨
b and by the line ¨ c accordingly. The energy from the HF generator is given by the line - d. If Ap<po, the similar results (lines a, b, c and d) can be reached using the simulation of the electrons dynamics when the difference of the velocity of electrons for the phases it <
< 2rc from the velocity of the relativistic electrons is taken into account. And these results will not differ from lines in Fig. 3 more than on 5-10%. However, it is not important because this part of the electron beam for it < < 27c is not used in the subsequent processes. The part of beam for 0 < < it has just accrued already the all that might be possible. The results of these integrals are given in Fig.3 in relative forms, assuming WO is equal to the integral of the kinetic energy of the initial beam for the interval 0 < < 231 before crossing of the cavity.
FIG.3 demonstrates that line b exceeds the total kinetic energy of the initial beam if 0.95<Ap/po<1 though decelerated electrons (line c) still keep positive velocities when they leave the cavity. It means that all decelerated electrons can be removed of the beam without significant losses for deal. This function can be achieved by the device that works similarly to a magnetic mass-spectrometer and is shown in FIG.4. This device comprises: vacuum chamber (15) joined to the first exit hole and the next input hole of the coaxial cavity, two deflecting magnets (12a, 12b, 14) and the dump-box (13) for the waste electrons. All these elements work as follows:
accelerated and decelerated electrons of the beam that crossed the coaxial cavity for the first time are transported through the exit hole of the cavity into the vacuum chamber (15) and go through the first deflecting magnet (12 a, 12 b). The magnetic field of this magnet is chosen in such way that it takes out the decelerated electrons of the beam and transports them into the dump-box. It is possible if the core of the magnet consists of two parts (12a, 12b) placed next to each other. The first part of the core (12a) has a form close to a rectangle and its width is equal to the maximum radius of the electrons orbit (16) that leads decelerated electrons to dump-box. For instance, this radius can correspond to the kinetic energy of the initial beam before the cavity so all electrons after crossing the cavity that have the phase TC <

7 <27t will be diverted from the beam to the dump-box (13). Because the remaining electrons in the beam that have the phase 0 < < it have the energy higher than the energy of the initial beam, they will go through the first part of the core of magnet (12a) and then will go further along the orbits (17) through the second part of the core of magnet (12b) that has a form close to an asymmetric trapeze.
These electrons have a big dispersion at the exit from the first deflecting magnet and the second deflecting magnet (14) compensates this dispersion, directing these electrons along the definite line before the next entry of the electrons into the cavity. As mentioned above, if 0.95<dp/p0<1, these electrons have the total kinetic energy that is more than the initial beam before the first crossing of the cavity and they have the structure of the periodically modulated electron beam because all electrons with phases between it and 271 have been disposed into the dump-box.
After that the modulated electron beam is transported through the cavity for the second time.
If electrons with phases 0 < 9 < TE were accelerated by the electromagnetic field during their first pass through the cavity, they have to return their kinetic energy into the energy of the field during the second pass.
This condition can be met, if the length of electrons trajectory from the first input hole of the cavity to the second input hole of the cavity is equal to odd number of halves of wavelength of the electromagnetic field in the cavity. Since the level of the electromagnetic field in the cavity is the same as in the first crossing of the cavity by these electrons and the phase of the field is shifted by1800, these electrons of the beam will lose the part of their energy and will have their initial energy at the injection moment. As differences in the lengths of trajectories of electrons with different phases are small in the vacuum chamber (15) percentagcwise to the wave-length, they have negligible effects on the decelerating that can be counted during the next crossing of the cavity by these electrons. When the electrons complete the second pass (18) through the cavity (1) they are transported through the vacuum chamber (19) as shown in FIG. 5.
As all electrons of the beam here have the same impulse of motion that is equal to the initial impulse at the injection moment, they all move through the deflecting magnet (20) along the same trajectory (21). After that the beam is transported into the cavity the third time. By choosing the magnetic field magnitude in magnet (20) under the condition that the total time of electrons travelling from the second input hole of the cavity to the third input hole of the cavity is equal to an whole number of the periods of the high-frequency field in the cavity the second act of the decrease of the kinetic energy of electrons can be met and hence all electrons will be decelerated one more time along the trajectory (22).This portion of electrons precisely corresponds to the part of the initial beam that was

8 decelerated during the first crossing of the cavity by the beam because this part has the same impulse p=p0 before the last crossing of the cavity and has the same magnitude of integral of the kinetic energy in interval (x, 2n). Therefore the beam is also transported into the second dump-box (23) for eliminating. This means that the sum of the losses of the beam's energy corresponds to the double magnitude that is shown by the line ¨ c in FIG.3 and it also means that the transforming efficiency of the kinetic energy of the initial beam into the energy of the electromagnetic field is very high and achieves 75-80% when 0.95<Ap/po<1.
The coaxial cavity of the Rhodotron is not unique type of the cavity that provides reusability of the cavity for beam accelerating. There are several other types of cavities designed for the same purpose, for instance, like cavities for FANTRON or RIDGETRON that were described in patent CA 1306075 or patent US 5107221 and patent US 5376893.
In more common events of several cavities, connected in series, or a single cavity of any type that are crossed by the beam (9) only once can be used as the above cavity (1) for the purpose of periodical modulation of the electrons' energy in the beam. Also the exit of these cavities can be connected to the entry of the magnetic inverter directly or through other devices (such as focusing quadrupole magnets etc.) and the exit of magnetic inverter can be connected to another cavity, different from the cavity (1), where modulated electron beam will be accelerated or decelerated as shown in the Fig. 6.
In the Fig 6 items from (1) up to (18) and (23) have the same meaning as in Fig.5, but items from (19) up to (22) arc absent. The item (24) designates the second cavity of the generator for converting kinetic energy of the modulated electron beam into the energy of electromagnetic field which goes to the load (26) through an exit loop (25). This device functions similarly to a klystron hut where the drift tube is changed to the magnetic inverter. The cavities (1) and (24) can be independent or dependent. In latter case the feedback will decrease the requirement for power of generator (2), but in both cases the requirements for the time synchronization of the electrons exiting from cavity (1) and travelling to the cavity (24) are weaker or are absent completely if the device works as a generator. The cavities (1) and (24) may comprise sequences of several cavities connected in series and, if the device is used as high-voltage injector and accelerator, the item (26) means the second external generator for the cavity (24). Requirements of the time synchronization must be kept in accordance to the synchronization between external generators (2) and (26) in this case.

9 The invention is characterized by two main features. Firstly, the cavity (1) provides the increase of the energy for a portion of the electrons in the beam (9) that goes through the said cavity during an acceleration phase of the electromagnetic field in said cavity, and also provides the decrease of the energy for the rest of electrons in the beam (9) that goes through the said cavity during the deceleration phase of the electromagnetic field in said cavity, and has an electron beam exit connected to the magnetic inverter.
Secondly, this magnetic inverter has a volumetric area, filled by the magnetic field dividing the said beam (9) at least in two flows, separating low-energy electrons from high-energy electrons of the said beam (9) under the transportation of the said beam (9) through said area. The first flow, containing only the part with low-energy electrons of the said beam (9), will be dispatched further to a dump (13) located inside or outside of the said inverter, and, at the same time, the accelerated part of the said beam (9) will be converted into the periodically modulated electron beam (18).

Claims

Claim 1 1. An accelerator-generator, comprising an relativistic injector, a relativistic unmodulated continuous uniform beam of electrons and at least one resonant cavity filled by an electromagnetic field, is being characterized by that fact that an input of the resonant cavity joins to an exit of the relativistic injector injecting into said resonant cavity the relativistic unmodulated continuous uniform beam of electrons acquiring additional energy or losing almost all their initial energy at an exit of said cavity in accordance to the phase of the electromagnetic field in said cavity.