Development of high power relativistic klystron amplifiers (RKAs) at the naval research laboratory is going to be discussed. The present RKA is of a cylindrical geometry and is currently operating in the multi GW range. We are using the present geometry as a test bed for new concepts of high power RKAs. For example, a new geometry of the Relativistic Klystron Amplifier (RKA) was designed. The new RKA will use a thin 45 cm diameter electron beam of 50 GW power. This electron beam will be confined by an external magnetic field between two coaxially grounded metallic tubes. High voltage gaps feeding coaxial cavities will be inserted in the annular drift tube. A 5 MW RF source feeding the cavities will be used to modulate the electron beam. The frequency of modulation can be extended with this geometry to the 10 Ghz region. We predict that 30% - 40% of the power can be extracted as an RF pulse.

We have operated a high-current relativistic klystron (5 kA, 500 kV) at a repetition rate of 200 Hz for 200 shots in a single burst. The modulated beam current was approximately 80% of the DC current. With 60 ns duration electrical pulses, the microwave pulses at 1.3 GHz were of 50 ns duration. The peak power during a pulse was 150 MW; the average power during the burst was 1.5 kW. We anticipate reaching 1 GW peak power and 9 kW average power with improvements to the extraction cavity.

High-power microwave pulses can destroy electronics of targets at altitudes of 100 km or higher, and preliminary designs of microwave antennas driven by Relativistic Klystron Amplifiers have been sketched. This paper discusses: (1) the susceptibility of the atmosphere to microwave breakdown, and (2) the constraint on the design of a microwave weapon imposed by the need to avoid breakdown.

We report progress in experimental and theoretical development of the MIT 3.3 GHz relativistic klystron amplifier. An annular electron beam (approximately 400 kV, approximately 5 kA, 2 inch diameter) will be confined by an axial magnetic field (22.5 kG) provided by a superconducting magnet. A novel, compact first cavity embeds the electron gun. The RF power driving the cavity will be generated by a 1.5 MW tunable frequency (3.1 - 3.5 GHz) magnetron. Current bunching in the drift tube will be monitored with a series of electric and magnetic dipoles. Initial experiments will focus on the first cavity and the beam bunching it produces. The experimental set-up is designed to be as flexible as possible to facilitate optimization. Results will be compared with theory and simulations using MAGIC.

We present results of simulation studies of several novel RKA configurations each of which is designed to increase the amplification of the stimulated beam modulation. The main configuration under consideration is the two-beam RKA. Results presented here, in which the two beams have the same injection energy, do not show any increase in modulation over a single-beam RKA, even when the total beam current is close to the limiting current. Summarized are simulation studies of several novel RKA configurations, including the two- beam RKA and the sequential interaction of a single-beam RKA with a chain of passive cavities. We have also examined the interaction of an intense beam with a corrugated slow- wave structure that is contemplated for RF extraction. Other areas of research include the effect of a resistive wire inside the coaxial RKA near the beam formation region to damp unwanted TEM or TM modes and some relativistic klystron oscillator (RKO) configurations.

A self-consistent nonlinear theory of the current modulation in a relativistic electron beam propagating through a klystron amplifier is developed. The modulation amplitude increases, reaching peak and decreases slowly, as the beam propagates through the amplifier. A simple expression of scaling law for maximum current modulation is obtained. This scaling law could be useful in the design and fabrication of high-performance klystron. Nonlinear mode evolution in current profile is also investigated by Fourier-decomposing the current modulation obtained from the integro-differential equation. The mode evolution in a long-range propagation of electron beam exhibits various interesting physical properties. For example, the maximum amplitude of the fundamental mode (l equals 1) occurs at the propagation distance, where all other modes vanish or have a very small amplitude. This property ensures a monochromatic frequency.

In this paper, the design and properties of a multi-fin coaxial waveguide converter are examined in detail. The converter has been designed as a means of transferring power from a high-power Relativistic Klystron Amplifier (RKA) to an antenna system. The output from an RKA is a coaxial waveguide, but the radially polarized TEM mode propagating in a coaxial line is not suitable for radiation. In order to transmit high electromagnetic (EM) power to the elements of a transmitting antenna, the polarization of the electric fields in the feeding waveguide should be parallel. A rectangular waveguide excited in the H_1,0-mode has such a linear polarization. Therefore, it is necessary to provide a suitable matching transformer between the coaxial and rectangular waveguides. The purpose of the transformer is to convert the TEM mode in the coaxial line into the H_1,0-mode in a number of rectangular waveguides. These are then connected to the antennas.

Based on the equivalent circuit representation of a cavity impedance, a theoretical model of the cavity excitation by a modulated electron beam is developed. The beam current modulation is initiated by the first cavity and amplified in drift section between the first and second cavities. Properties of the second cavity excitation by the beam are described in terms of the phase shift (Psi) and amplitude (phi) ₂ of induced voltage in the cavity. The phase shift (Psi) and amplitude (phi) ₂ are determined in terms of the frequency difference ((eta) ), the cavity Q-value, the voltage multiplication factor ((chi) ), and magnitude ((phi) ₁) of the first cavity voltage. The voltage multiplication factor is proportional to the beam intensity, Q- value, and distance between the first and second cavities. On the other hand, it is inversely proportional to frequency and capacitance of the cavity, and beam electron velocity. Experimental study is carried out by making use of a klystron amplifier operable from 4.4 GHz to 5 GHz. Experimental data from the klystron agree remarkably well with amplifier performance profiles predicted by the theory.

A high-current relativistic klystron amplifier (RKA) is being developed with the goal of producing 1 kJ per pulse with a 1 microsecond(s) pulsewidth and a peak power of 1 GW. The three cavity tube is fully assembled and is undergoing high power testing. Peak power levels as high as 400 MW have been produced so far. Current experimental results are presented.

Experimental results to-date will be presented from a developmental effort to a produce a J- band (5.85 - 8.2 GHz) relativistic klystron amplifier (RKA) of the high current Naval Research Laboratory (NRL) genealogy. The nominal experimental parameters of this RKA are: V₀ approximately equals 600 kV; I₀ approximately equals 2 - 4 kA; B_z approximately equals 1.5 T; (tau) _beam approximately equals 300 ns; v_in approximately equals 6.6 GHz; P_in <EQ 500 kW. Because of the smaller component sizes which accompany this frequency ((lambda) approximately equals 4.5 cm as compared with (lambda) approximately equals 30 cm for the bulk of other RKA research efforts), much of the discussion will concentrate on the physical principles, fabrication issues, and experimental pitfalls associated with scaling the RKA design.

Experimental observation of wall plasma produced by a relativistic electron beam propagating through a 2.2 cm diameter stainless steel or copper plated stainless steel drift tube has been made at background pressures as low as 3 X 10^-6 Torr. An annular electron beam of thickness approximately 1 mm and outer diameter of approximately 1.8 cm was generated using a carbon fiber or graphite cathode and guided down the drift tube by a 15 kG magnetic field. The electron beam energy varied between 500 to 630 kV with total beam currents between 1 to 2 kA and pulse duration of approximately 750 ns. Two viewing ports and lens systems placed approximately 35 cm apart were used to collect light from plasma produced in the drift tube. The light was transmitted to two SPEX 1702/04 monochrometers using quartz optical fibers. Detection was done with photomultiplier tubes sensitive from 300 to 700 nm and lines from hydrogen, oxygen, iron and copper were observed. At the lowest background pressures no light emission was observed until approximately 1.5 microsecond(s) after the beam pulse for the unheated stainless steel drift tube. After baking and pumping the tube for two days light emission was observed approximately 1.8 microsecond(s) after the start of the beam pulse.

A two-beam configuration of the relativistic klystron amplifier (RKA) was developed. This configuration employs two annular coaxial electron beams instead of a single one. The interaction of this double beam with the 2-cavity RKA structure, operating at a frequency of 3.5 GHz, led to an energy feedback mechanism between the cavities via reflexing electrons, resulting in a large and stable amplitude current modulation.

A two-stream relativistic klystron amplifier (RKA) is proposed in which the amplification of stimulated beam modulation is achieved via the unstable two-stream interaction rather than the use of passive cavities. After a calculation of the limiting current, the amplification and saturation of the stimulated beam modulation are analyzed using a cold-fluid model and particle-in-cell simulation. Good agreement is found between theory and simulation in the linear regime. Fully modulated intense relativistic electron beams are obtained at saturation. In the proposed scheme, substantial beam energy difference is required to achieve large instability growth.

We are testing an enhanced version of the Choppertron, a high-power rf generator which shows great promise of achieving greater than 400 MW of output power at 11.4 GHz with stable phase and amplitude. This version of the Choppertron is driven by a 5-MeV, 1-kA induction accelerator beam. Modifications to the original Choppertron included aggressive suppression of high order modes in the two output structures, lengthening of the modulation section to match for higher beam energy, and improved efficiency. Final results of the original Choppertron experiment, status of the ongoing experiment and planned experiments for the next year are presented.

The Microwave Source Facility at the Lawrence Livermore National Laboratory is commencing a series of experiments involving reacceleration of a modulated beam alternating with extraction of energy in the form of X-band microwaves. The Choppertron, a high-power microwave generator, is used to modulate a 5-MV, 1-kA induction accelerator beam. The modulated beam is then passed through a series of traveling-wave output structures separated by induction cells. In this paper we report on computer simulations used in the design of these experiments. Simulations include analysis of beam transport, modulation, power extraction and transverse instabilities.

We have modified a two-dimensional relativistic klystron code, developed by Ryne and Yu, to simulate both the standing-wave free-electron laser two-beam accelerator and the relativistic klystron two-beam accelerator. In this paper, the code is used to study a standing-wave free- electron laser with three cavities. The effect of the radius of the electron beam on the RF output power; namely, a three-dimensional effect is examined.

Recent analysis of the Two-Beam Accelerator (TBA) by Wurtele, Whittum and Sessler has shown that the transfer cavities, both in the relativistic klystron version (RK/TBA) and the standing-wave free-electron laser version (SWFEL/TBA), can be characterized by a simple coupling impedance. In the two cases the radiation process is very similar: only the modes that couple to the electron beam are different. As a result, computer programs that are able to handle realistic cavities (with beam ports and coupling ports, etc.) can be employed to evaluate the performance of either version of the TBA. We have employed the code URMEL² to study the proper coupling impedance for a number of realistic cavities for a SWFEL.

The current state of proposals to use high power microwaves in the atmosphere is reviewed. HPM has been proposed to aid in the conversion of stratospheric ozone by partial breakdown, facilitating chemistry to eliminate the chlorine. Another proposal is over-the-horizon radar using a partial breakdown in ionosphere. A key to any such effort is rapid diagnosis of the state of the atmosphere before, during and after intervention. The technology requirements of these proposals is reviewed. The elements of an atmospherical modification program are identified and non-technical factors are discussed.

This paper introduces a modification to the bidirectional traveling plane wave decomposition, whereby one can select new basis functions resulting in different representations for a solution. This freedom of choice facilitates the solving of the homogeneous as well as the nonhomogeneous scalar wave equation. We shall also demonstrate the bidirectional method by solving two source problems: point charge moving uniformly in a lossless dielectric, and the Green's function for the Poisson equation.

This paper examines the interrelationship between the size of an antenna structure and its ability to produce narrow beam radiation when excited by signals of arbitrary time dependence. By using the well known spherical mode expansions for the scalar wave equation, a network theoretic interpretation is developed which is valid for all frequency. This allows the decomposition of the problem of studying the time domain behavior of a radiating structures into two distinct parts. The first deals exclusively with the angular dependence of the radiation, while the second deals only with the frequency content or corresponding temporal behavior of the excitation. The two parts are connected by the number of modes that are excited and their relative excitation strengths. The inter-relationship between the size of a radiating structure, the minimum beamwidth that it can efficiently produce and the resulting limitations in creating packets of radiation that are localized in both time and space for a variety of wideband excitations including baseband waveforms is considered.

This work presents the reduction of the collective bremsstrahlung recoil force on the bare charge of an unperturbed relativistic test particle in a nonequilibrium beam-plasma system. In this part of the bremsstrahlung process, the field of the test particle is scattered by the second- and third-order dynamic polarization current that is induced by the test particle and acts back on the bare charge of the unperturbed test particle. The expressions obtained here are to be used in determining the conditions for occurrence of a possible collective bremsstrahlung radiative instability in a nonequilibrium relativistic beam-plasma system.

We consider problems of the electrical strength in vacuum cavities for high-power microwave generators. A discussion is provided of the effect that the field emission and explosive emission of electrons have on breakdown.

Important features of high-current dielectric Cherenkov masers (DCMs) connected with their possible future development are considered theoretically. The start current problem is considered for the DCM oscillator in the intense beam (two-wave) operation regime which is inevitable for low millimeter and submillimeter ranges, when the beam-dielectric gap is comparable with the wavelength. The set of equations for start current finding has been obtained taking into account the frequency dependence of the beam-wave coupling. The calculations show that the values of start current are several times greater than ones obtained under the assumption that the beam-wave coupling is constant determined at the synchronous frequency.

Microwave sources based on backward-wave oscillators (BWOs) driven by relativistic electron beams are capable of producing high power coherent radiation in the cm and mm wavelength region. When the axial magnetic field is used in these devices to confine the electron beam satisfies the condition of cyclotron resonance there is a significant modification in the behavior of BWO due to beam coupling to cyclotron modes. A time-dependent, self-consistent theory of BWOs is developed taking into account a possible cyclotron interaction. The analysis of the system near the cyclotron resonance yields a number of physical effects including the power drop due to the cyclotron absorption observed in many BWO experiments. A series of experiments has been conducted to compare the measured BWO characteristics with the theoretically predicted ones.

We have increased the energy from the Varian-HDL, conventional-emission, high-power magnetron to greater than 125 Joules/pulse by operating at 5.4-microsecond(s) pulse widths. We have also demonstrated 250-Hz repetition rate operation for 200-pulse bursts, resulting in 8-kW average power. In addition, a pair of these magnetrons has been phase-locked to an RF- isolated, low power driver source. Excellent phase stability of several degrees has been achieved. We report on the results of these recent experiments.

The electrons in a gyropeniotron transfer their rotational energy through the transverse migration toward the ridges or the vanes of the waveguide without any orbital bunching. A slotted waveguide (magnetron type) is used for enhanced interaction impedance at high cyclotron harmonics even with low-energy beam and for better mode control. The number of vanes required is 2(s+1) where s is the cyclotron harmonic number of the gyropeniotron interaction. We have calculated the critical current, oscillation frequency, and characteristic wave number for the excitation of absolute instability in gyropeniotron amplifiers. We found that the threshold current decreases as (alpha) increases (at fixed beam voltage) where (alpha) is the ratio of the perpendicular velocity to the parallel velocity of an electron.

A Ka-band second-harmonic TE₂₁ gyro-TWT amplifier, capable of generating extremely high power, has been designed with a self-consistent nonlinear simulation code. To demonstrate that a harmonic gyro-TWT can generate significantly higher output power with better stability than fundamental gyro-TWT's due to the higher electron beam currents allowed for the weaker harmonic interaction, a Ku-band proof-of-principle experiment is being built at UCLA and described in this paper. An output power of 400 kW, with an efficiency of 20% and a constant-drive bandwidth of 6% have been predicted for this device. A single anode 100 kV, 20 A MIG is being built to generate the (alpha) equals (upsilon) _{(perpendicular})/(upsilon) _(parallel) equals 1 electron beam with (Delta) (upsilon) _PLL/(upsilon) _(parallel) equals 8%. A wide- bandwidth 0 dB two-mode phase-velocity coupler and a TE₂₁/TE₁₁ mode converter have been designed to couple the rf power into and out of the amplifier.

The design of a 95 GHz, 70 kV, 5 A, (alpha) equals 1.5, (Delta) (upsilon) _z/(upsilon) _z equals 7%, six-vane, (pi) mode, third-harmonic gyro-TWT is presented which is predicted by a nonlinear self-consistent simulation code to yield a peak output power of 90 kW and 26% efficiency, a saturated gain of 61 dB and 3% constant-drive bandwidth. The stability of the device is examined using linear theory and a multimode time dependent PIC code. In addition, the wideband input coupler and output mode converter are also described.

A two-dimensional ((partial)/(partial)(theta) equals 0), nonlinear theory of double-stream cyclotron masers is presented in a cylindrical configuration. A dispersion relation, which describes small-amplitude, space-charge cyclotron waves on two co-propagating gyrating electron beams in a finite axial magnetic field, is derived and used to analyze the double-stream cyclotron maser instability. The sensitivity of the instability growth rate to axial momentum spread of the electrons within each beam is investigated. The saturation levels of the instability are obtained from simulations for parameter regimes of experimental interest.

The traveling wave cyclotron (TWC) is defined in this paper as a cyclot ron-type interaction of a nonrelativistic electron beam with traveling waves in an inductive periodic waveguide. Experimental results and a theoretical model of the TWC interaction are presented in this paper. The TWC experiment is conducted in the microwave regime ("" 10 GHz) with a low-energy ("" 10 keV), low-current (< 1 A) electron beam. Considerable RF amplification has been observed in the TWC experiment. A TWC theoretical model presented in the paper shows that the TWC amplification observed is the result of the inductive impedance of the periodic waveguide near cutoff.

Numerical studies of the high efficiency results recently reported by Conde and Bekefi for a reversed field FEL are presented. These studies indicate that the predicted saturated power is rather sensitive to the detailed model used of the space-charge forces. The reported results for Group I FEL operation can be reproduced by increasing the emittance by 45% compared with the reversed field case. In this regime particle loss, rather than trapping, is the saturation mechanism. The Group II results are not consistent with the single-frequency model. A two- frequency model has been developed to study the influence of competition between the modes of the upper and lower FEL intersections. Initial results from this model indicates that the low power in the Group II experiment is not a consequence of mode competition.

A simplified 1-dimensional (1-D) nonlinear model has been developed to study the nonlinear electron trajectories in the virtual cathode oscillator. This model is based on a modified time- dependent soft-spring Duffing potential. The force equation contains the properties of a chaotic system and is believed to be a contributing factor to the random RF output observed in many vircator experiments. A low-voltage (100-kV) test bed has been designed and built for the study of the characteristic nonlinear trajectories of electrons in the device. The system can inject small amplitude microwaves into the potential well as a means of controlling the chaotic dynamics (sensitivity to initial conditions) of the reflexing electrons. In this case, controlling the chaos means confining electrons in the potential well for a longer time, and with more coherent bunching, which has relevance to microwave production.

A brief review of past multiwave Cherenkov generator experiments is presented, together with a summary of progress in an ongoing investigation.

Considering either a quantum effect or an electron spin resonance to formulate the interaction of an Electromagnetic Missile with matter, we compared the localized absorption of electromagnetic energy. We took into account temporal shapes of E-M pulses (a gaussian double derivative function: D.S.G) typically generated by new types of picosecond optoelectronic devices (Photo Conducting Pulsed Switching). For ultrashort time of interaction, the electronic transitions undergo a periodic perturbation governed by a non- sinusoid time relationship introducing a square-time function in agreement with the perturbation non-relativistic theory. We attempted to compute numerically the material E-M properties in the case of ultra-wideband pulses with coherence limit considerations. We studied the influence of the frequency spectrum on the ultrashort photoconductivity phenomena induced into semi-conductor materials by such type EM pulses train.

A set of self-consistent equations are used to study the nonlinear interaction of high power microwave pulse with the atmosphere in this paper. Both long and short pulses are investigated in our research. Electrons are treated as a fluid and their effects on the passing microwave pulse are determined by means of the simultaneous solution of Maxwell's equations and the electron fluid equations. Our one-dimensional fluid-model calculation shows that when the field strength of high power microwave pulse is above the threshold for air breakdown, large portions of the initial energy are absorbed by the electrons in the plasma region which are created by the avalanche process. Nevertheless, a significant amount of energy is still able to propagate very long distance. The amount of energy transmitted through the air to the targets depends very strongly on the initial energy of the pulse, its frequency, shape and length as well. In addition, we also calculate the electron density, conductivity and the characteristic frequency of plasma.

Motion of beam electrons in a spiral undulator with an axial guide magnetic field is investigated analitically. It is shown that the projections of electrons trajectories in the cross-section plane are in general case ellipses. Such motion can be represented as the superposition of two, clockwise and counter-clockwise, circular girations. Because of this the cyclotron resonance arises for both directions of the guide magnetic field orientations, conventional and inverse. The "anticyclotron" resonance is conditioned by the difference between the amplitudes of radial and azimuthal components of the undulator magnetic field for off-axis electrons.

We investigate the parameters of millimeter-range back-wave oscillator (BWO) developed with the use of a compact RADAN SEF-303A repetitive accelerator. The maximal repetition rate of this oscillator reaches 10 p.p.s. and is determined by thermal operation mode of the pulsed transporting solenoid. With the electron beam power ranging between 350 - 400 MW, the BWO which operated at the frequency 70 GHz had an output microwave power of 60 MW and an efficiency of 15%. For the described millimeter-range BWO the calculated starting current was 0.6 kA, the operating current 1.5 kA, and the required electron energy 270 keV. These parameters were close to those obtained experimentally.

The linear theory of the relativistic superdimensional Cherenkov devices is considered. The simulation results of Dielectric Cherenkov Maser (DCM) and relativistic Multiwave Cherenkov Generator (MWCG) are presented.

Utilization of the parametric down—conversion provides a unique possibility of absolute standardless brightness calibration ofhigh—temperature visible, infrared and near UV radiation. The calibration method has been suggested in Moscow State University in 1977 and then was applied there to experimental measurements of radiation spectral brightness for laser and luminiscence sources of visible range and filament lamps radiating in IR range. The method is based on the comparison between two signals: the signal p, corresponding to the measured radiation parametric up-conversion and the "noise" signal P corresponding to parametric scattering of the pump; this "noise" signal is observed both in the presence of the measured radiation at the converter's idler channel entrance and in the absence of it. The noise signal intensity is determined by the effective brightness of zero vacuum fluctuations at the idler frequency which equals to one photon per mode, or, in the energetic units, Bvac = hc2/A, where h is the Plank constant, c —the light velocity, )2 —the idler wavelength. Measuring the ratio of signals P" P' + P and P at the exit of the crystal—converter at frequency w' = wo — w (w0 being the pump frequency, and w —the frequency of the flux being calibrated), one can calculate the absolute value N of energetic brightness spectral density for the radiation which fills the idler channel of the converter: N = k[(P"/P )— 1] in the units "photon/mode" or B = BvacN in the energetic units (here icis a correction coefficient describing the non-ideal transparency of the converting crystal). No additional standard sources or detectors are required here; the method uses as a reference the effective value of zero vacuum fluctuations brightness. The reference effective brightness temperature depends on the radiation wavelegth: T0 = hc/Ak in 2, k being the Boltzmann constant, so that for A2 10 p Tvac 2000 K, and for ' 0.5 Tvac 40 000 K. In this work the accuracy of the method is analysed and the resolution for different schemes of the parametric photometer is calculated.

In this paper a new frequency scaling law that greatly reduces computational requirements and at the same time incorporates the nonlinear effects inherent to HPM propagation is discussed. Results of a comparison between predictions of air breakdown thresholds made using the frequency scaling law and experimental data taken at various frequencies are presented. The scaling law is implemented in an existing HPM propagation code and has been used recently to develop a new predictive capability that calculates the optimum energy, power, and antenna requirements necessary to transmit a desired fluence. These capabilities provide both the accuracy and rapid computational turnaround necessary for system studies that assess the effects of HPM propagation for particular HPM devices and that attempt to optimize device parameters for maximum air transmission. Samples of both forward propagation, predictive calculations and inverse, optimization calculations are presented.