Experiments in modulating an electron beam of power greater than 30 GW revealed a mechanism that limits the output RF power of the NRL relativistic klystron amplifier. This mechanism is beam loading of high voltage gaps. It is found that the presence of charge at the gaps is analogous to connecting a capacitor in parallel with the HV gap. This capacitor has a component that depends only on the electron density and on the geometry. Its presence causes detuning of the cavity, resulting in a slow build-up time for the electron beam modulation, saturation of the RF current amplitude and occasionally erratic fluctuations in the RF current. The nonlinear components of this capacitor cause the spectrum of the current modulation to be shifted toward the high frequencies.

The results of measurements on the gain and efficiency of a high power travelling wave tube amplifier powered by a 850 keV, 0.8-1.6 kA, 100 nsec pencil electron beam are reported. The amplifier operates in the TM01 mode and is driven by a 250 kW magnetron. Gains of up to 35 dB and efficiencies of 11 percent have been achieved in a single stage amplifier with a power output of about 100 MW. Increases in the gain beyond 35 dB result in a driven oscillation due to positive feedback from a small impedance mismatch at the output. A two stage severed amplifier has also been developed to prevent the driven oscillation. Preliminary results indicate that the severed amplifier has a peak power output of 400 MW and an efficiency of over 40 percent.

An experimental study of a pulsed relativistic electron-beam O-type oscillator driving two O-type amplifiers and operating at levels of a few hundred megawatts in the X-band is reported. Two configurations are examined, in one of which a backward-wave oscillator (BWO) drives two independent traveling-wave amplifiers (TWAs), and in the other of which the same electron beam passes through both the BWO and the TWA structures. Results on power output, pulse distortion, gain, and output matching are presented. The pulse-shortening syndrome and its apparent absence in TWA are briefly discussed.

Successful injection-locking of relativistic magnetrons requires frequency stable microwave output with pulse lengths longer than the phase-locking time. We report the major experimental results from our stable and long-pulse relativistic magnetron system designed to address important issues relevant to the realization of injection-locking. Long voltage pulse operation 1js) with good stability was achieved and relatively long (250 ns) RF pulses at 5 MW output power level were generated. These pulses exceed the theoretical locking time for a 1 MW injected power source. The magnetron operates in S-band (3. 66 GHz) and is driven by a modulator which operates either at 700 kV/ 700 A or 250 kV/ 1. 7 kA. The stability of the narrow band microwave frequency is better than 3 parts in 1000. Possible RF pulse-shortening mechanisms are studied. 1.

A module that can accommodate seven magnetrons in a common vacuum vessel energized by a single pulsed power driver with a magnetic field provided by a single Helmholtz pair has been constructed. The goal of the program is to phase-lock the magnetrons producing up to 10 GW of coherent radiation at 2. 8 GHz. This paper reports the initial operation of the module 4 magnetrons operating without the interconnection bridges that produce phase-locking have produced 100 to 200 MW in each of 12 powerextraction arms. 1.

A magnetron-specific phase-locking model has been developed using the standard equivalent-circuit approach which takes into account the unconventional magnetron growth characteristics as well as the frequency pushing effect. These effects owe their origin to the highly nonlinear electron-wave interaction thus are believed to be more pronounced in relativistic magnetrons. The model predicts a wider locking-bandwidth and a shorter locking time than those in conventional locking theory. The phase-locked amplitude resonance occurs as the results indicate at an injection frequency different from that of the free-running oscillator. 1.

We discuss design of a multiple magnetron module and issues related to it including modelling of the phase-locking process with various geometries via a Van der Pol analysis. We also describe two important " confidence building" experiments one determined if the couplers load the magnetron output which would then restrict the number of couplers we would use. Another determined the effect of operating a magnetron with the axis of the cathode displaced from the axis of the magnet coils and showed we could use a single coil set for the entire module rather than individual coil sets for each magnetron. 1.

Stability properties of the extraordinary mode perturbations in relativistic electron flow in a cylindrical magnetron are investigated within the framework of the macroscopic cold fluid model. The eigenvalue equations for the extraordinary mode waves are obtained. In the tenuous beam limit the eigenvalue equation is considerably simplified and a closed algebraic dispersion relation is obtained. Numerical investigation of this dispersion relation over a broad range of system parameters has been carried out. It is concluded that the extraordinary mode perturbations in a tenuous electron flow in a cylindrical magnetron are absolutely stable. The full eigendifferential equation is solved numerically for the stability of the extraordinary modes for intense electron flow (Brillouin flow). The investigation is concentrated on low frequency perturbations (w Wc) and the A6 anode geometry. For this case all the lowest modes are found to be stable. 1.

The magnetron has attracted much attention lately as a means of efficiently generating high-power electromagnetic radiation. However there is at present no complete theoretical description of this device. As a step in the development of such a theoretical description simulations of A6 magnetron operation have been performed with the 2-D fully electromagnetic codes MAGIC and MASK. Simulations with MAGIC have revealed mode competition between the ir and the 2irmodes with other operating parameters an mode (5. 0 GHz) is observed. Also density profiles are seen to depart markedly from Brillouin flow during preoscillation and large RF fields are observed at saturation accompanied by energetic bombardment of the anode. Simulations with MASK exhibit the ir-mode and the 2ir-mode separately. The dominant mode is seen to depend on the cathode radius in agreement with experiment. 1 .

The relativistic multiresonator magnetron is analyzed extensively in cylindrical geometry using the two-dimensional self-consistent particle-in-cell simulation code MAGIC. In particular detailed comparisons are made between the simulation results and the experiments by Palevsky and Bekefi [Phys. Fluids 22 986 (1979)] using the A6 magnetron configuration. Within a constant scale factor the computer simulations show a similar dependence of microwave power on magnetic field with dominant excitations in the ir and 2ir modes. In the preoscillation regime it is found that the electron flow in the simulations differs substantially from the ideal Brillouin flow model. In the nonlinear regime the gain is dominated by the formation of spokes. The effects of the power supply impedance on magnetron performance are also investigated. 1.

The driven van der Pol equation is widely used to model the behavior of regenerative electronic and microwave oscillators driven by an external locking signal. In the case of high-power microwave oscillators such as relativistic magnetrons nonlinear frequency-shift effects are believed to be important. They have been modeled by inclusion in the van der Pol equation of an additional cubic restoring force (Duffing) term. Use of the slowlyvarying amplitude and phase approximation to study the behavior of the driven van der Pol-Duffing equation has predicted stable single-valued locked behavior within a skewed range of values of the frequency mismatch. For parameter values consistent with the slowly-varying amplitude and phase approximation numerical solutions of the van der Pol-Duffing equation confirm this prediction. For oscillators with high growth rate such as the relativistic magnetron however the slowly-varying amplitude and phase approximation may be unjustified. Furthermore regardless of the satisfaction of the assumptions on which the approximation is based it may fail to predict the occurrence of complicated dynamical behavior of coupled oscillators with important implications for phase locking. Numerical solutions of the driven van der Pol equation show that the time of transient evolution to phase-locked states instead of depending solely on the frequency mismatch as conventionally assumed is also a function of oscillator growth rate and injection power level. Recent work suggests that the form of model oscillator equation appropriate for the magnetron may differ in both nonlinear growth-saturation and frequency terms from the van der Pol-Duffing equation. 1.

Phase locking of a vircator array driven by a relativistic magnetron in a pure master-slave configuration has been achieved. A single high voltage pulse drives the relativistic magnetron and the vircator array in series. Phase locking occurs for 40 ns. The time required to lock is z. 10 ns. The peak array power is . 1 GW. The injection ratio p and allowable frequency difference 1M relationship required for phase locking is examined. The results suggest that the Adler''s inequality p (OlAf) f is a necessary but not sufficient condition to predict phase locking. A hybrid configuration in which the driven vircators are also mutually connected together has also been attempted. Phase locking is observed with better phase stability than in the pure master-slave configuration. 1.

The stability of the final equilibrium states achieved between two mutually coupled microwave oscillators is assessed as a function of the initial frequency detuning, the length of the interconnection between the two sources, the coupled power, the oscillator growth rate, and the loaded Q of the cavities enclosing the sources. The model is compared to measurements from an experiment in which two high power, short pulse cavity vicators are phase-locked. The observed, final, phase-locked states are all predicted to be stable. No states that are predicted to be unstable exhibit frequency locking or phase locking.

Work begun in 1986 is continuing on the reflex-diode oscillator mounted on one of the Aurora 10-MV 22- ci pulsers. The most recent work extracted microwaves radially into six rectangular WR-975 waveguides at the virtual cathode and into two additional arms a few meters downstream. Two of the waveguides were open-ended and radiated into an anechoic chamber. Diagnostics for radially extracted microwave energy included directional couplers card calorimeters and free-field sensors. The peak output is ''-50 3 per arm for the frequencies below 1. 0 (3Hz. Parameter variations have included the anode/cathode spacing and the positioning of a second anode foil. Dosimetry and electric field measurements were made to study beani dynamics in the drift region. Thus far the measurements suggest that our virtual cathode is weak. 1.

High power microwave (11PM) experiments with Aurora reflex diodes (ARDs) are analyzed. In particular the multi-wave guide logistics 11PM peak envelope power (PEP) and waveform consistency problems associated with multi-guide radial extraction are reviewed. A novel design involving an outer coaxial cavity to extract power from the ARD cavity via 2535 appropriately placed ports or an equivalently loaded peripheral opening (sideshoot aperture) is described. Tmproved extraction mode selection phase locking frequency stabilization and waveform smoothing via this coaxial ARD (or CARD) is predicted. Staged physics studies and scaling experiments to validate the 11PM operation of a full-scale CARD on a Pocobeam and Aurora are recommended. 1.

Computer simulations have been carried out to investigate the effects of a dense background plasma on the dispersion characteristics of backward wave oscillators. It is observed that the background plasma tends to reduce the phase velocity of the most unstable mode. As a result, a strong efficiency enhancement is found. On the other hand, the relatively weak beam-backward plasma wave instability introduced by the rippled waveguide wall is found to exert negligible impact on the interaction efficiency.

The recently developed concept of scalarized photons (formally photons of any polarization) is used to analyze the spontaneous emission in the uniform magnetic field free-electron laser in the microwave spectral region. With the electron beam energy of up to 10 MeV and the uniform magnetic field of up to 4 Tesla, the radiation (occurring with the fundamental and higher harmonic frequencies) can easily cover a 10- to 10,000 GHz spectral region.

The utility of the magnetically insulated transmission line oscillator (MILO) as a high power microwave source depends on how efficiently power can be extracted from it. We have designed a slow-wave stepped transformer for the purpose of axially extracting microwave power from a 3. 6 GHz coaxial MILO. The slow-wave transformer design was optimized using particle-in-cell simulation and tested in experiments performed on the HPM Simulation Division''s GEMINI and GYPSY water Blumlein pulse power sources. in this paper we summarize the slow-wave stepped transformer design and describe MILO axial power extraction experiments which yielded up to 300 MW of radiated power. 1.

Studies of a cyclotron autoresonance maser (CARM) are presented. The measurements are carried out at a frequency of 35 GHz using a mildly relativistic electron beam (1. 5 MeV 300 A 40 ns) generated by a field emission electron gun. Perpendicular energy is imparted to the electrons by means of a bifilar helical wiggler. Amplification measurements give a small signal growth rate of approximately 65 dB/m. The saturated power output is ''-''-33 MW and the corresponding electronic efficiency is ''-7.

Experiments have been performed in which multi-MW, frequency tunable (15-20 GHz) radiation has been generated by cyclotron autoresonance maser (CARM) oscillators driven by intense microsecond electron beams. Experiments have been performed utilizing two long-pulse intense electron beam generators. Initial CARM oscillator experiments were performed on a pulseline generator with peak parameters V

A CARM amplifier can produce the power required to feed a high gradient 1 TeV linear collider. The CARM is shown to be capable of extremely high power highly efficient and requires a relatively weak magnetic field. The designs of two proposed CARM amplifiers and a proof of principle experiment are presented. 1.

The theory of a Bragg reflector composed of periodic cylindrically symmetric corrugations is presented. For a wave satisfying the Bragg condition the corrugations can provide a reflection of nearly unity through the principles of constructive interference resulting in a frequency selective resonator. It is found that a maximum allowable value exists for the corrugation amplitude. 1.

Results are reported on the experimental and theoretical performance of a dielectrjc Cherenkov maser amplifier. The amplifier is designed to work in the TM01 mode of a cylindrical dielectric lined waveguide. In a low voltage low current (200 kV 100A ) experiment results indicate that the device has operated as an amplifier at a frequency of 9. 8 GHz. The rf input of the amplifier was provided by a tunable (9-10 GHz) 10 kW magnetron. An average power gain of approximately 0. 6 db/cm over a 21 cm interaction length was measured using electron beam parameters of 190 kV 90 amperes and a beam pulse length of 1 . Ls. Preliminary results of the scale up of the experiment to a system driven by a 700 kV to 1 MV 3 kA 100 ns electron beam with rf input provided by a tunable (8. 4-9. 6 GHz) 200 kW magnetron indicate a power output of 280 MW at 8. 6 GHz. However due to the large amount of feedback in the system the device might more appropriately be described as a frequency locked oscillator. 1.

It is shown that a planar-array feed has excellent potential as a solution to paraboloidal reflector distortion problems and beam-steering requirements. Numerical results from an algorithmic procedure are presented which show that, for a range of distortion models, appreciable on-axis gain restoration can be achieved with as few as seven elements. For beam-steering to + or - 1 MW, 19 elements are required. For arrays with either seven or 19 elements, high effective aperture elements give higher system gain than elements having lower effective apertures. With 37 elements, excellent gain and beam-steering performance to + or - 1.5 BW is obtained independently of assumed effective aperture of the array element. A few simple rules of thumb are presented for the design of the planar-array feed configuration.

The electromagnetic pulse radiated from a nonuniform circular aperture field has been studied comprehensively. The results show that, like the uniform field, the nonuniform aperture field has the ability to launch an electromagnetic missile. If the spectrum of the exciting pulse is unlimited, the nonuniformity of the distribution does not affect the rate of decay of the energy, but only its magnitude. On the other hand, if the frequency spectrum is limited, i.e., has a cutoff, the nonuniformity significantly affects the rate of decay. The more uniform the distribution of the aperture field, the slower the energy decreases. The electromagnetic pulse radiated from a special distribution - the point-source array - has been studied. With practical excitation, the energy from such an array also exhibits a slow decay.

Conventional radar pulses spread so that the energy reaching a reflecting target at range r decreases as r exp -2. The energy echoed from the target also decreases as r exp -2, so that the energy backscattered to the transmitter is proportional to the product, i.e., to r exp -4. Pulses of finite total energy that satisfy Maxwell''s equations need not decrease as r exp -2, but can instead decrease much more slowly, for instance as r exp -epsilon where epsilon is a positive number that can be chosen as small as desired. Such pulses are referred to as electromagnetic missiles. Here, the echo backscattered when an electromagnetic missile encounters a plate target is studied. By analogy with conventional radar, it might be expected that the energy of the echo would be proportional to r exp -2(epsilon). Instead, the reflected energy decreases even slower, proportional to r exp -epsilon, so that no squaring occurs, resulting in a larger than expected radar return.

Electromagnetic pulses of finite total radiated energy can deliver energy to a distant receiver that decreases with distance much more slowly than the usual r exp -2. Such electromagnetic missiles can be generated by any of a class of current distributions over one or another set of points contained in a bounded region of a two-dimensional or three-dimensional space. For a given currrent-bearing set of points as a transmitter, one can explore various time and space dependencies of the transmitting current. Over some classes of dependencies, the energy reaching the receiver decreases as r exp -epsilon, where epsilon is a positive number that can vary from one current to another. Current-bearing sets so far explored have dimension 0, 1 or 2. For these, the requirement of finite total radiated energy establishes a greatest lower bound on epsilon, dependent on the dimensions of both the set and the space in which it resides. This infimum defines what can be called the pulse index of a set relative to the space containing it. One might conjecture that the pulse index can be defined for sets of fractional Hausdorff dimension. It will be shown that this is so, and that for certain sets of fractional dimension, the index is related to the dimension just as it is for integral dimensions.

The cause of the phenomenon of self-focusing of intense laser beams is the nonlinear intensity dependent refractive index of optical materials. This property which leads to the self-focusing can be interpreted as an attractive force in between the photons thereby increasing the photon density. For large enough densities the photon gas may undergo Bose-Einstein condensation and if the attractive force is strong enough may even become a superfluid through a second order phase transition. This mechanism not possible for free photons is possible inside optical materials with a large refractive index because there the photons can be viewed as massive quasiparticles having a velocity less than the velocity of light. With the large dielectric constants of certain ferroelectric substances valid in the range of microwave frequencies the envisioned process might even work with microwave photons. This would make possible the generation of very intense highly coherent microwave beams. I .

The recoil force on the third-order nonlinear dynamic polarization charge of a relativistic test particle participating in collective bremsstrahlung in a nonequilibrium beam-plasma system is calculated in the Born approximation. A contribution arises from a term involving direct interaction of the field of the test particle in second order with its own third-order dynamic polarization.

The effects of high-power RF fields generated in the context of a strategic defense system on the atmosphere and ionosphere are discussed. The significance of density perturbations, electron accelerations, IR emissions, optical emissions, UV emissions, generation of RF noise field, and air breakdown due to the fields are discussed. The impact of these physical changes on communication, jamming, surveillance, and tracking are noted.

Exact time-dependent solutions of the source- and current-free acoustic equation and Maxwell''s equations are given, which, in the wave zone, are packets of energy shot through a cone. Such solutions are called " bullets." Any exact solution of the source- and current-free acoustic equation and Maxwell''s equations can be obtained in many ways by pulsing suitable sources and currents. By combining both previous results, explicit examples of pulsed sources and currents which generate the bullets are obtained. They appear to be physically realizable.

A new technique is described that is applicable to the synthesis of large aperture antennas for directed energy and conventional applications. The two principal technical components of this new approach are (1) the extensive application of digital techniques to control and to provide customized waveform synthesis at each transmitting element in the array and (2) the use of a transponding scheme via an RF link to calibrate individual elements and to cohere the entire array. The approach utilizes all-digital electronics except for the analog transponder receiver and power output amplifiers. The transponding RF link avoids the need for providing coaxial cabling to possibly thousands of individual antenna elements in a large two-dimensional aperture.

The lifetime of the circulating electron beam in the NRL modified betatron has been substantially increased from a few sec to over 700 sec by the addition of strong focusing windings to the device. The injected beam consistently spirals from the injector to the minor axis and is trapped. The s0. 5 kA trapped electron ring has been accelerated to 12 MeV from the injection energy of MeV. The beam acceleration has been confirmed not only by the x-ray attenuation technique but also with the detection. of photoneutrons from the reaction D(''y n)H. The experimental observations show that during acceleration the electrons cross initially higher order cyclotron resonances. Eventually when the ratio of the vertical B to toroidal magnetic field B0 is 1/9 i. e. at I 9 a substantial fraction of the ring electrons is lost. The remaining electrons are lost at higher B that is at lower values. Experiments are in progress to locate and eliminate the field error(s) that excite the resonances and thus to accelerate the electron ring to even higher energy. 1.

The UCI (University of California Irvine) stellatron project started in 1984 as an extension of researches on a modified betatron. In addition to the purpose of proof ofprinciple a goal was set at accelerating a 1 kA heair to 10 NeV in a 41 cm najor radius and 4 cm n[nor radius torus. Formation of a 200 A beam and acceleration up to 4 MeV were dezionstrated using an electron injector. The current was increased by u''eans of plas startup. In this scherre plasniolds injected into the torus were confined in a rising stellarator field and a nultikA runaway current was induced by applying the hetatron field. As predicted theoretically two distinct types of instabilities created loss of electrons. One of their was the highnode orbital resonance that caused gradual decay and terwination of the current. The other was the negative ness instability which disrupted the beair at the early phase of accel eration. The loss due to the orbital ''Instability was suppressed by reducing the aziiruthal ripple of the toroidal ivagnetic field to less than 1 within the whole voluire of the torus. The abrupt loss due to the negative nss instability was replaced by a gradual one when the beam orbit was inintained close to the torus wall. TTder these conditions beams of over 1 kA and with a density of approximately 1 kA/cm'' were accelerated to several MeV

We have measured the motion of electron beams of energies from 850 to 900 keY and currents from 300 to 1400 A as they pass through a 180-degree toroidal focussing channel configured with stellarator strong focussing fields as proposed by Roberson et aL for the stellatron variant of the betatron. Good agreement is shown between the measurements and calculations using the beam centroid equations of motion. The data give confidence that the low-dispersion stellarator bends under design in the Spiral Line Induction Accelerator development program will perform well.

This paper addresses the development of a new high voltage-gradient linear accelerator. This accelerator was energized by a modulated intense relativistic electron beam of power in the multigigawatt range. A 0. 2 cm diameter electron beam emerged from the accelerator with a peak current of 200 Amps and peak kinetic energy 60 1!eV. The length of the accelerating structure was 1 meter. Simple scaling laws suggest that similar accelerators with much higher average electric field can be built. 1.

Temperatures up to 50000 have been reached in water vortex stabilized Gerdien arcs. In arcs confined within the cores of supersonic hydrogen vortices much higher temperatures should be possible. Furthermore if these arcs are thermally insulated by a strong magnetic field temperatures up to a 106 K may be attainable. At these temperatures and in passing through a Laval nozzle the arc plasma can reach jet velocities of 100km/sec. If small quantities of heavy elements are entrained by this high velocity plasma jet these heavy elements are carried along reaching the same speed and upon condensation can form beams of clusters and microparticles.

The non-zero net dc force acting on relativistic beam electrons traveling a laser wave and isotropic turbulent plasma waves is calculated by using quantum-kinetics. The force is caused by net inverse bremsstrahlung (the absorption by inverse bremsstrahlung minus the emission by stimulated bremsstrahlung) of the electromagnetic wave in the isotropic turbulent plasma field and is not the Lorentz force. It is found that the average of this force is independent of the spread of the electron beam energy increases with the electron beam energy can be stronger than the Lorentz force of the laser wave for very strong laser intensities and electron-energies far beyond the energy region where a single-photon process can take place. 2.

An ultimate linac structure is realized by an appropriate crystal lattice (superlattice) that serves as a "soft" irised wave guide for x-rays.2 As the energy of accelerated particles increases, the radius of the beam has to decrease inversely proportional to the energy squared if we try to keep the number of events per time constant. This condition becomes very severe beyond the energy in excess of 100 GeV for a single-pass collider. It is in fact, so severe that smallest possible structures, i.e. crystal, have to be utilized for accelerator components. In this sense such an accelerator is the ultimate accelerator. High-energy (40keY) x-rays are injected into the crystal at the Bragg angle to cause Borrmann anomalous transmission, yielding slow-wave accelerating fields. Particles (such as muons) are channeled along the crystal axis. An alternative to injecting x-rays is exciting plasma waves as a wake of particle beams (such as proton or electron beams) injected in the crystal axis. The crystal is cryogenically cooled to avoid the lattice interference and damping. Accelerating fields of x-rays are most likely provided by (i) radiating electrons propagating along the crystal channels (with much lower energies than that of particles to be accelerated), or (ii) preactivated semiconductor or other active solid state medium that is triggered by an incoming pulse of x-ray laser. It is much desired to accomplish self-induced transparency similar to the triple soliton structure.3

Proper inclusion of self-field effects and careful design has led to development of induction accelerator gap geometries which afford a factor of 5000 or higher improvement in beam brightness over conventional gap designs form multi-kiloampere, multi-MeV electron accelerators. Because of the resulting TEM-like power feed to the accelerator gap region in these designs, the usual low frequency beam breakup instability associa ted with the induction cavity is suppressed. However , this geometry exhibits a significant , high frequency beam breakup problem associated with cavity resonances in the gap region itself. Analysis of this mode and the current status of simulation studies and methods of suppression of this instability are presented.

We have studied the magnetization of a high-3 hydrogen-plasma beam injected into a vacuum transverse magnetic field (f3 plasma energy density/magnetic field energy density 1). Nominal parameters were: T1 1 eV Te 5 eV n 3 x io13 cm3 v 7 x io6 cm/s. tpulse 70 us B : 300 0. Plasma beam characteristics were measured for a wide beam a/p 35 and a downstream distance x 300 p1 where a is the beam radius x is the downstream distance and p1 is the ion gyroradius. We observe a brief initial state of diamagnetic propagation followed by (magnetized) propagation propagation is accompanied by beam compression iransverse to with as much as a factor of four increase in density and a slight drift of the beam in the direction of the ion Lorentz force. At high fields B 200. . 300 G the observed magnetic field diffusion time is much faster than calculated from classical Spitzer conductivity and is more of the order of the diffusion time based on Hall conductivity. 1.

Gas puff cathodes offer a flexible systerr to study the nchanfsr' of the negative ion foriration In a pulse power—driven cathode. A gnetica1iy Insulated gas puff diode has been built at UCI and the forntion of H has been studied. }Tydrogen an'tonia, butane and oxygen were Investigated. Very low current less than 100 rrA/cri$ of was n'esured when hydrogen or airironia were puffed into the diode. Futane and oxygen produced at least ten tiv'es higher current density. Probably the negative ion pro— duction in the butane gas puff sovrce and in the solid polyethylene cathode source is due to the interaction of electrons with long hydrocarbon chains. This irechanisw doesnt require the formation of highly excited J4', which is needed if one assuns the source of the H is tiolecular hydrogen interacting with electrons. Because oxygen is an electro—negative gas, it was expected that negative Ions would he produced. These ions were used to test the negative ion diagnostic. The divergence of the ion bean' coniing froir the gas puff cathode was about 10 wr. This is the lowest divergence that has ever been rreasured at UT. Trproved nozzles that will produce higher gas gradient between the cathode and the anode will he tested in the future.

The possibility of using a dielectric-covered cathode in a pulsed magnetically insulated ion diode to produce high H ion fluxes is investigated. In this experiment an array of resistor-backed holes in a polypropylene substrate provides uniform prepulsed discharges over the cathode of an annular diode. Since the diode operates in a high vacuum differential pumping of the system is not needed and collisional losses of the ions are minimized. The ions evolve from the dielectric material on the cathode following prepulsed discharges and are extracted by applying a high-voltage pulse to the diode. The goal is to identify and understand the processes involved in generating high H ion fluxes. Preliminary measurements indicate that H ions can be produced with this method although the cathode is strongly affected by the presence of water. 1

Existing methods of matching the low-energy Hbeam from the ion source into the Radio. Frequency- Quadrupole (RFQ) accelerator using gas focusing and magnetic lenses suffer from beam loss emittance growth and inadequate control capability. It is shown that an electrostatic quadrupole (ESQ) system is an attractive alternative providing strong focusing and good control as well as preventing plasma build*up and associated problems. A new conceptual design of an ESQ system for matching a 120 keV 120 mA beam with a normalized 4xrms emittance of 3 x i0 m-rad into a 425-MHz RFQ is developed using laminarflow and linear-optics codes and a modified PARMILA particle-in-cell simulation code which includes the nonlinear effects of the external fields and space charge. Simulation results show a 100 transmission and low emittance growth. 1.

The wakefield anplitude and frequency of a Dielectric Wakefield Accelerator (DWA) is calculated taking into account the transverse dimensions of the wakefield cavity including the ratio of inner to outer radii of the dielectric liner S. The results indicate that the total wakefield an''plitude can have a significant contribution frori'' the first two transverse magnetic ntdes (TM01 and Th02) of the dielectric lined cavity for S . 7. AddItionally the wakefield amplitude of the fundamental node (Th01) decreases as the ratio of inner to outer radii S increases. The theoretical results are shown to be in good agreenent when predicting the wakefield frequencies with recent results of an experinental test of the PWA concept.

Phased array antenna systems are capable of radiating very intense microwave beams. However, while propagating through the atmosphere and the ionosphere the beam intensity can be significantly reduced by several mechanisms which are not of concern to less intense radiation. This paper discusses qualitatively the propagation energy loss associated with powerful phased arrays.

This paper demonstrates the frequency versatility of the developed Mk-6 electron cyclotron maser. In a 100-ns pulse, this system produced more than 100 kW power in both the Ka-band (26-40 GHz) and the W-band (75-110 GHz) with the excitation of the TE(01) mode at 35 GHz and the TE(03) mode at 96 GHz. This system was crudely step tunable in the W-band. Due to the grossly overmoded regime in the G-band (140-220 GHz), the maser proved to be quasi-continuously tunable over the entire range. This work culminated in 200 GHz oscillation. Fast digitizers, with single event 750 MHz bandwidth, were used to study the rise and fall times of the mm-wave output pulse. These experiments revealed structure in the pulse and showed that the maser was operating in a burst mode.

Recent experimental results have indicated that the use of a background plasma in high power microwave devices can increase the space charge limiting current in the device and also increase the microwave power output. Here, the effect of a background plasma on a dielectric Cherenkov maser is examined by solving the linearized, beam-plasma, dielectric lined waveguide dispersion relation. A discussion of the output frequency and growth rate dependence of the device on the background plasma density and dielectric lined waveguide parameters will be presented.

The properties of a plasma waveguide consisting of a cylindrical vacuum core surrounded by a plasma medium are studied. The frequency dependence of the guided modes in a plasma waveguide is analyzed, along with single mode propagation in the waveguide. It is found that the waveguide is less dependent on frequency than the dielectric waveguide, i.e., it is a wide-frequency-band transport system. The excitation of the plasma waveguide is discussed, and a survey is made of laser-created plasmas, the mechanism of photoionization, and the stability of the plasma.