Search

Spatially dispersive nonlinear transmission lines (NLTLs) have attracted interest as pulsed RF sources. The characteristics of these sources need to be further evaluated and understood to optimize their design. This paper presents the performance of a spatially dispersive nickel-zinc (NiZn) NLTL for two experiments. The first experiment evaluated the RF power and frequency generated as a function of the primary dielectric medium for SF 6 compared to the previously implemented Royco 602 transformer oil. The second experiment evaluated the RF power and frequency generated as a function of the ferrites' loss tangents, initial permeabilities, and dimensions. The NLTL's performance is presented for each experiment across a range of operational input voltage levels and ferrite bias conditions. Experimental primary dielectric performance results demonstrate approximately 4X increase in RF power at frequencies above 1.2 GHz for replacement of oil with SF 6 . Experimental ferrite performance results show output frequencies from 0.9 to 1.39 GHz and peak RF powers >2X higher for the higher permeability and larger volume ferrite.

A six cavity inductive voltage adder with pulsed power incorporated into a single assembly, linear transformer driver (LTD), has been developed. Each LTD cavity is driven by a single pulse forming network and spark gap switch. The output pulse is approximately square in shape due to the pulse forming networks (PFNs) used for the drivers. In some cases, when unreliable switch firing occurred, multiple output pulses were observed. The pulses stem from each cavity firing at a different time. A circuit model which explains the multi-pulse behavior and details the current path is developed. This observation shows the potential for an LTD to discharge each cavity individually to the load thus providing extremely high repetition rates on a common cathode stalk.

The nonlinear transmission line electron beam driver at the Air Force Research Laboratory produces a frequency agile modulated annular electron beam [1]. A linear kinetic bunching analysis is performed based on transmission line theory [2]. This analysis reveals that the axial length of the conical Friedman cathode lowers the maximum bunch amplitude, analogous to the effect of axial distance on the opening of a klystron bunching cavity. Additionally, the somewhat broadband nature of the NLTL pulse translates into a saw-tooth profile for the peak frequency of the electron beam as a function of axial distance down the drift tube. Most importantly, the direct relation between applied voltage and cathode current emission leads to in phase bunching in both current and velocity, which enhances the maximum bunching amplitude when compared to just voltage modulation. The purpose of this exploration is to inform the design of a relativistic klystron amplifier that uses the NLTL e-beam driver as a gun.

A gigawatt-class nonlinear transmission line based electron beam driver is experimentally demonstrated. Four experimental series, each with a different Marx bank charge voltage (15, 20, 25, and 30 kV), were completed. Within each experimental series, shots at peak frequencies ranging from 950 MHz to 1.45 GHz were performed. Peak amplitude modulations of the NLTL output voltage signal were found to range between 18% and 35% for the lowest frequency shots and between 5% and 20% for the highest frequency shots (higher modulation at higher Marx charge voltage). Peak amplitude modulations of the electron beam current were found to range between 10% and 20% for the lowest frequency shots and between 2% and 7% for the highest frequency shots (higher modulation at higher Marx charge voltage).

Non-linear ferro-electric materials exhibit electromagnetic shock-forming properties which can be a challenge for electromagnetic modeling algorithms such as finite-difference time-domain (FDTD). If the material response time is sufficiently fast, the shock fronts steepen until the Nyquist sampling criteria is violated, and the resulting aliasing leads to spurious grid oscillations. A flux corrected transport (FCT) algorithm is adapted for use on the staggered FDTD mesh and effectively removes the spurious oscillations. Numerical results demonstrate the effectiveness of the algorithm.

Many variables can affect the dielectric strength of coolants and insulating fluids including particulates, water content, temperature, viscosity, and pressure. The present paper evaluates the effects of dissolved water content, adjusted in a controlled manner, on the dielectric strength and the variability of breakdown voltages in synthetic polyalphaolefin (PAO) MIL-PRF-87252 coolants obtained from three different suppliers. Additionally, the effect on dielectric strength of using nitrogen gas versus air for the removal of water is evaluated. The results indicate that the dielectric strength of the fluids decreases by up to 80% over a range of 0 to 70 ppm water content. The variability in breakdown voltage generally increases with increasing water content for all of the samples tested. These results indicate the notable effects of water at concentrations below the MIL-PRF-87252 specified acceptable limit of 50 ppm. The effect on the dielectric strength of using two different sparging gases for the removal of water, dry air or dry nitrogen, was irrelevant when compared at similar water contents.

Capacitor energy density for submicrosecond discharge applications was investigated for capacitors based on the following: 1) polymer-ceramic nanocomposite (PCNC); 2) antiferroelectric (AFE); and 3) paraelectric (PE) ceramic dielectrics. The developmental PCNC dielectric enabled design, fabrication, and testing iterations to be completed relatively rapidly. The PCNC capacitors were nominally 4 nF and were tested to dc potentials of at least 75 kV. The capacitors were then charged from 20 to 48 kV with a dc high-voltage power supply and discharged into a nearly critically damped test circuit of up to 5 pulses/s (pps) repetition rate for lifetime testing. The discharge time was 65 ns. Shot life as a function of the charge voltage was compared for three design iterations. Changes in the manufacturing of the PCNC capacitors have yielded up to 100× improvements in pulse discharge life. The 1-2-kV prototype, nonlinear (antiferroelectric and paraelectric) multilayer ceramic capacitors had zero-voltage capacitance ratings of between 60 and 300 nF. They were charged to their operating voltage and discharged into a nearly critically damped load in 2-6 μs, depending on their capacitance, at repetition rates of up to 75 pps. Their operating voltage for fast, repetitive discharge was determined for lifetimes consistently over 105 shots. Discharge energy densities of 0.27-1.80 J/cc and energy losses of 7.9-36.8% were obtained for the packaged multilayer capacitors with different formulations of nonlinear dielectrics. Increased field-induced strain was correlated with increased permittivity and contributed to the limitations on the operating voltage. Multilayer ceramic capacitors fabricated from AFE and PE ceramic dielectrics have the potential to achieve high energy density owing to their high relative permittivities that vary with applied electric field, assuming they can be scaled up to sufficiently high voltages.

Over the last decade, the Air Force Research Laboratory, Directed Energy Directorate (AFRL/DE) has engaged in a high current density field emission cathode research program. This program explored the aspects of cathode materials as well as the details of cathode geometries and emission physics. This paper summarizes the results of this ongoing research effort to date. We review the history and motivation for the program, which provide insight into the physics issues of concern for various vacuum electronic sources. One important aspect of the program consists of the investigation of new cathode materials. For many high power microwave (HPM) sources, neutral out-gassing, which ties critically with cathode materials, plays a key role in the effective operation of the source. These material properties influence plasma formation, which in turn dictates the operation of an HPM device. For a cathode material, AFRL chose to focus on cesium-iodide-coated carbon fiber cathodes, which we discuss in detail here. A second important aspect of the program consists of understanding emission physics and the optimum geometries for a cathode. This aspect couples closely with electron beam quality, which in turns effects the electron beam interaction with microwaves in the HPM structure. This paper concludes with a discussion of the implementation of the cathode material on both a Magnetically Insulated transmission Line Oscillator and a relativistic magnetron.

A high-power liquid dielectric switch is being developed to satisfy the requirements for future directed energy applications. A flowing, high-pressure liquid dielectric was chosen for the design of a megavolt class switch operating at 100 pps. This paper reports on the design philosophy, modeling, and experimental results of a full size, single-shot prototype 250-300 kV concept validation test (CVT) switch which can transfer kilojoules per pulse. Analysis of design criteria and scaling for a compact, 100-pps, kilojoule, high-voltage switch are presented. Optimization studies indicate that a pressure range of 6.9-13.8 MPa (1000-2000 psi) appears to be ideally suited to a flowing dielectric rep-rate switch.

A compact pulse forming line (CPFL) concept based on a folded transmission line and high-breakdown strength dielectric was explored through an effort combining proof-of-principle experiments with electromagnetic modeling. A small-scale folded CPFL was fabricated using surface-mount ceramic multilayer capacitors. The line consisted of 150 capacitors close-packed in parallel and delivered a 300 ns flat-top pulse. The concept was carried to a 10 kV class device using a polymer-ceramic nanocomposite dielectric with a permittivity of 37.6. The line was designed for a 161 ns FWHM length pulse into a matched load. The line delivered a 110 ns FWHM pulse, and the pulse peak amplitude exceeded the matched load ideal. Transient electromagnetic analysis using the particle-in-cell code ICEPIC was conducted to examine the nature of the unexpected pulse shortening and distortion. Two-dimensional analysis failed to capture the anomalous behavior. Three-dimensional analysis replicated the pulse shape and revealed that the bends were largely responsible for the pulse shortening. The bends not only create the expected reflection of the incident TEM wave but also produce a non-zero component of the Poynting vector perpendicular to the propagation direction of the dominant electromagnetic wave, resulting in power flow largely external to the PFL. This analysis explains both the pulse shortening and the amplitude of the pulse.

The finite-difference time-domain (FDTD) method is a simple, robust tool for a variety of electromagnetic modeling scenarios. In order to model magnetic ferrite materials, the FDTD method is coupled with the Landau-Lifshitz-Gilbert (LLG) equation. The LLG equation is discretized by defining a magnetization vector at the center of each cell. The update at each time step then proceeds by (1) update the electric field with Ampere's law, (2) update the magnetic flux density with Faraday's law, (3) interpolate the magnetic field to the cell centers, (4) update the magnetization vector with the LLG equation, (5) interpolate the magnetization to the magnetic field locations on the Yee FDTD staggered grid, (6) update the magnetic field using B = μ 0 (H + M) where B is the magnetic flux density, μ 0 is the permeability of free space, H is the magnetic field, and M is the magnetization.

This paper documents recent work developing solid dielectric transmission lines for sub-microsecond, 100 kV class compact pulsed power systems. Polymer-ceramic nanocomposite materials have demonstrated sub-microsecond discharge capability in parallel plate capacitors and transmission lines [1, 2]. With a dielectric constant of approximately 50, the propagation velocity is 2.5 cm/ns, necessitating lines of several meters length to achieve > 100 ns pulse lengths. By folding the line in a fashion analogous to ceramic multilayer capacitors, the physical length of the line can be significantly shorter than the electrical length. We present the results of an experimental effort to develop a folded transmission line using a polymer-ceramic nanocomposite dielectric. The pulse length was somewhat shorter than expected based on a simple calculation using the geometry and the dielectric constant. Fully 3-D electromagnetic calculations were used to examine the role of the edges in curtailing the pulse length. Dielectric breakdown in this device occurred below the electric field threshold demonstrated in the prior work [1]. Improvements in the large scale fabrication of TiO 2 beginning with nanoscale grains have opened the possibility for producing single layer high voltage devices. Given a dielectric constant approaching 140, transmission lines using nano-TiO 2 can be considerably shorter than with other materials. Relatively thick, flat sheets of TiO 2 have been fabricated for testing up to 50 kV. Several transmission lines, employing a serpentine electrode geometry, have been manufactured and tested. Testing up to several 10's of kV has confirmed the operation of the lines according to the design. As expected, the triple point between the TiO 2 , electrode, and insulating medium has proven difficult to manage for high voltage operation. Several techniques to mitigate the effects of the triple point, including resistive grading at the edges of the electrodes, are discussed. Fully 3-D electromagnetic modeling is used to examine the effects of electrode geometry and composition on the performance of the lines.

Experiments and simulations were performed to investigate the electric field screening effects with two and four cathodes configuration. The experiments are run using carbon fiber field emitters with various spacing between fibers. Experiments and simulations show the current to be a strong function of the separation between the field emitters.

Thin a-C, a-C:N, a-C:H and a-C:N:H films were deposited onto silicon, glass and graphite substrates using ion beam techniques. Rutherford backscattering spectroscopy and elastic recoil detection analysis were performed in order to determine their atomic composition. The hardness of these films was measured using a Vickers hardness diamond indenter. The as-deposited a-C:H and a-C:N:H films were characterized using optical absorption spectroscopy, while the a-C and a-C:N films were studied by Raman spectroscopy. A shift of the absorption edge toward the lower energy region and a lower gap of a-C:N:H films was observed. The Raman spectra of a-C:N films show an increase in the I(D)/I(G) ration, as compared to those of non-nitrogenated diamond-like carbon films. A possible structure is suggested.

A parallel plate nonlinear transmission line (NLTL) was constructed. Periodic loading of nonlinear dielectric slabs provides the nonlinear capacitance and the gaps between provide linear inductive interconnects, this is essentially the same design used by Ikezi [1],[2]. The NLTL was modeled in a circuit simulation code using the experimentally measured form of the nonlinear capacitance. Dielectric loss included in the model as an equivalent series resistance derived from the measured loss tangent data affects the formation of RF oscillations. The diagnostics used on the experimental system are Bdots along the line and a current viewing resistor at the load. The diagnostics provide a measurement of the pulse evolution as it travels down the line. The waveforms from the experimental line qualitatively agree with the circuit model, showing no strong RF formation as a result of the loss.

Summary form only given. Csl coated carbon fiber cathodes have shown promise as a cold cathode for microwave and x-ray devices. In particular, the cathodes have demonstrated over 1 million shots lifetime at operating voltages at or in excess of 165 kV and current densities greater than 50 A/cm2. Further, the materials have also operated in a DC mode. While the vacuum emission characteristics have been well-studied1, the materials characteristics of the cathodes themselves, particularly after operation have not received great attention. Furthermore, while researchers at University of Wisconsin" have demonstrated a reduction in work function due to the Csl coating, the emission mechanism remains poorly understood. This paper gives results of a series of materials diagnostics investigating the cathode surface morphology as well as the changes in the carbon fiber structure with cathode shot history. We rely principally upon SEM images and Raman microscopy for this data. Finally, we present a basic model for electron emission from Csl coated carbon fibers that begins to satisfy some of the experimental results.

High voltage switches are fundamental components of high voltage pulsed power modulators. The goal of this research consists of improving the lifetime, and reducing the size of traditional high voltage switches and the associated ancillary equipment, while optimizing performance. A high power, high pressure, flowing liquid dielectric switch has been developed by university, government and contractor researchers for directed energy applications. Dielectric fluid (poly alpha-olefin) flows through apertures concentric around the switch housing and through the cathode axis into central aperture in a contoured anode. The fluid then distributes through the switch at flow rates from 0.13 L/s to 0.72 L/s (2 -11.4 gpm) and pressures from 3.45 MPa to 17.24 MPa (500 -2500 psi) using a compact pump cart. The switch has been designed and tested for charge voltages of -250 to -300 kV and operated with discharge energies to 280 J per pulse at pulse rates of 1 to 10 pps. Switch jitter has been tested in self break mode, with enhanced electrode configurations and with a trigatron trigger. Operation, jitter, and erosion were examined as a function of oil pressure, and flow rate and test results are reported.

A repetitive oil switch for directed energy applications has been developed in a joint effort between teams at the University of Missouri - Columbia, Alpha Omega Power Technologies and the Boeing Company. The switch is operated at test pressures to 17.24 MPa (2500 psi), flow rates to 0.72 L-s-1 (11.4 gpm), charge voltages to -300 kV and discharge energies to 275 J per pulse at 20 pps. An examination of the electrodes after 250,000 shots with the original design led to the design of an insert device which resulted in higher performance fluid flow within the switch. The flow shaper-enhanced switch was tested for 150,000 shots, the results of which are presented in the following paper. Electrode lifetime has been evaluated for stainless steel under the original and enhanced fluid flow conditions and is reported.

The University of Missouri-Columbia (UMC) is developing high power liquid dielectric switches intended to address future high power microwave (HPM) applications. Although requirements encompass a broad parameter space, the initial switch concept focuses on a 250-300kV output switch operated at 100 pps that will be scaled to 1MV. Failure to clear high electric field regions prior to the next charge cycle results in prefires, thereby limiting the maximum achievable repetition rate. Elevating the operating pressure, hence minimizing the bubble size and temporal properties, has alleviated this problem. This paper presents the design philosophy, modeling, and experimental results obtained from a single shot prototype operated in oil at pressures ranging from atmospheric pressure to greater than 13.8 MPa (2000 psi).

Gated field emission cathodes can provide short electron pulses without the requirement of laser systems or cathode heating required by photoemission or thermionic cathodes. The large electric field requirement for field emission to take place can be achieved by using a high aspect ratio cathode with a large field enhancement factor which reduces the voltage requirement for emission. In this paper, a cathode gate driver based on the output pulse train from a nonlinear transmission line is experimentally demonstrated. The application of the pulse train to a tufted carbon fiber field emission cathode generates short electron pulses. The pulses are approximately 2 ns in duration with emission currents of several mA, and the train contains up to 6 pulses at a frequency of 100 MHz. Particle-in-cell simulation is used to predict the characteristic of the current pulse train generated from a single carbon fiber field emission cathode using the same technique.

Various configurations of rare earth permanent magnet solenoids with a large bore were simulated using finite element computer modeling. The configuration effect on the axial field, Bz, and its uniformity, ΔBz, was analyzed. The magnetic orientation of individual segments and the aspect-ratio of the solenoid have a significant effect on Bz and ΔBz. Two-layered configurations assembled with trapezoidal or arc segments result in a more uniform Bz compared to configurations assembled with five-faced polyhedrons. The two-layered configuration also has the advantage of easier fabrication. A permanent magnet solenoid prototype, based on the two-layered arc segment configuration and built using Nd-Fe-B magnets, was experimentally tested. The solenoid was designed to have a length to inner diameter ratio of approximately 5:3 and a peak axial field magnitude of 1.8 kGauss. Experimental measurements agree with simulation predictions to within

Magnetostatic finite element modeling of permanent magnetic devices can provide accurate results only if the underlying physics is well captured in the model. A single device surrounded by an air padding region is the focus of this research. In the ideal modeling case, this padding region would be infinite in extent. Since an infinite padding region is impossible to implement, the padding sizes in modeling a permanent magnetic solenoid that can ensure accurate results were analyzed in our study. The relationship between accuracy of the simulation result and the surrounding air padding percentage in terms of the device size is demonstrated in this paper, which shows that physically impossible results can occur for a permanent magnetic solenoid if an insufficient padding percentage is used. Design engineers typically use padding percentages from 40 – 400%, and our results show that the required padding percentage depends on the openness of the magnetic circuit. A permanent magnetic solenoid may have more or less openness of the magnetic circuit when the magnetic vector of the side magnets has a larger or smaller angle β1. For a more open magnetic circuit with β1 near 90°, a higher padding percentage, up to 600%, is required. For a less open magnetic circuit with β1 ≤ 30°, 200% padding is recommended for better accuracy.

Cesium iodide (CsI) coated carbon fiber cathodes have shown promise as a cold cathode for microwave and x-ray devices. In particular, the cathodes have demonstrated over 1 000 000 shots lifetime at operating voltages at or in excess of 165 kV and current densities greater than 25A/cm2. While the vacuum emission characteristics have been well studied, the materials characteristics of the cathodes themselves, particularly after operation, have received little attention. Furthermore, while researchers at University of Wisconsin have demonstrated a reduction in work function of carbon due to the CsI coating, the details of the emission mechanism remain poorly understood. This article gives results of a series of materials diagnostics investigating the cathode surface morphology as well as the changes in the carbon fiber structure with cathode shot history. We demonstrate that the cathode surface undergoes several changes in relation to the bond line along the fiber-substrate interface as well as at the fiber ...