Summary form only given. The Hydrus project will provide an enhancement of the flash radiographic diagnostics available in the Hydrodynamics Department at AWE as part of new facilities for hydrodynamics research. The X-ray sources for the experiments carried out at AWE are provided by focused electron beam diodes driven by pulsed power accelerators. The most attenuating objects to be radiographed require an X-ray spectrum and dose that can only be provided by pulsed power machines operating at 7 to 14 MV. A design for a 14 MV IVA (induction voltage adder) has been completed by L3 Communications Pulse Sciences for AWE and three such machines are about to be manufactured to provide the main radiographic capability in the new facility. Research currently being conducted by AWE and international collaborators aims to also develop improved electron beam diodes to meet the requirements for higher radiographic resolution being demanded. Other experiments performed at AWE require a softer X-ray spectrum and hence a lower voltage pulsed power driver, two of which will be provided in a second area of the same facility. It has been determined that the optimum radiographic source will have a sub-100 ns pulse , a diameter of 1 mm or less, an X-ray dose of at least 15R at 1 metre and an X-ray spectrum end point energy of 1-1.5 MeV. An electron beam diode that has demonstrated performance o f this order is the plasma filled rod pinch driven by the Gamble II accelerator at NRL. After a concept study and discuss ions with possible vendors it has been decided procure two machines similar in performance to Gamble II. They will feature a Marx generator charging a single water pulse forming line which will in turn drive the diode via a 3 ohm water insulated output line.

Summary form only given. Two switch designs were tested in the prototype PFL test bed at L-3 as part of the recently completed IVA design effort for flash radiographic sources in the new Hydrus hydrodynamics research facility at AWE. One design, called the slab switch, was scaled from the Phermex Laser Triggered Gas Switch (LTGS) and the other, called the diaphragm switch, was scaled from the Nike LTGS. Both switches have a single ~8 cm, laser triggered gap pressurized to 150 psig with SF 6 , are 42 inches in diameter, and designed to fit within the same 7-ohm PFL which is nominally charged to 2.5 MV, in 1.5 mus. The PFL and switch were tested up to 3 MV in the prototype facility. Test results for the LTGS performance based on >2730 shots are presented in this paper along with the switch type down-selection that was based on electrical, mechanical, and manufacturing criteria. The performance and reliability test results for the overall PFL are presented separately. The LTGS test results include measurements of jitter, rate of prefire, and the throughput risetime (due to inductance and resistive phase combined). The PFL design was based on a relatively low inductance prediction of between 220 to 240 nH. The effect of laser focal position in the switch gap, flow rate of the continuously flowing SF 6 , gas density, laser power, the mechanical deformation of the switch, and laser optic alignment on the switch jitter, risetime, and arc formation were investigated and are presented. Two dimensional EM simulations of the switch were also performed to estimate the contribution of two dimensional effects on the risetime. Circuit models that match the measured performance are presented.

PIM is a water filled Blumlein connected, in parallel, to one or two 1.5 MV, cavities which inductively add voltage on a high impedance vacuum transmission line to a load diode operating at 1-3 MV. The radiographic performance of this IVA pulsed power machine depends on the inductive rise-time of the Blumlein switches. Selection of either one or two simultaneously closing switches on PIM was used to investigate the effects of rise time, in addition to voltage on the diode. Two Blumlein switches, designed and built by Titan PSD were initially tested in a self-break mode at two charging rates, from the Marx, the time to peak voltage being 0.7 and 1.5 mus in the two cases. The breakdown curves are matched with the predicted curves from 0.5 to 1.7 MV. Under this self-break mode it was found that there was a tendency for the second switch to close 7 mus later. A model for this is proposed and its effect on the PFL output pulse is illustrated. Two types of lasers have been installed, the results from closing both switches within 2 ns of each other was investigated.

Summary form only given. PIM is a Blumlein driven IVA (Inductive Voltage Adder) accelerator with each induction cell operating at up to 1.5 MV. Each cell is fed at a single point and hence has some asymmetry (greater than 20%) in the current flow into the bore where a MITL (Magnetically Insulated Transmission Line) feeds the output pulse to an electron beam diode load. The LSP 3D particle in cell code has been used to model the PIM induction cell and its predictions for the current asymmetry were compared with experimental data obtained from monitors in the outer (anode) of the MITL. The code also predicts what the associated asymmetry should be for the MITL and load impedances that have been used on the accelerator with results that contrast significantly with those for the smaller bore cells (35 rather than 67 cm bore) used in Sandia National Laboratory's RITS machine. The modelling was also used to determine a possible, and relatively simple, method of reducing the asymmetry to only a few percent that will be the subject of further design work.

In this paper we use three different designs for induction voltage adders to be used in radiographic X-ray sources to illustrate the differences between possible IVA architectures, and between their electrical characteristics as these affect the radiographic diodes that the IVAs drive. The architectures are chosen considering the spaces in which the FVAs must fit, and the electrical characteristics are determined by the approaches chosen for the IVA electrical subsystems. We describe the status of the various IVA technologies. The IVAs, (RITS, the AWE HRF and URSA) are large systems with similar output parameters in the 12-16 MV range.

An IVA (inductive voltage adder) research programme at AWE began with the construction of a small scale IVA test bed named LINX and progressed to building PIM (Prototype IVA Module). The work on PIM is geared towards furnishing AWE with a range of machines operating at 1 to 4 MV that may eventually supersede, with an upgrade in performance, existing machines operating in that voltage range. PIM has a water dielectric Blumlein of 10 ohms charged by a Marx generator. This has been used to drive either one or two 1.5 MV inductive cavities and fitting a third cavity may be attempted in the future. The latest two cavity configuration is shown which requires a split oil coax to connect the two cavities in parallel. It also has a laser triggering system for initiating the Blumlein and the prepulse reduction system fitted to the output of the Blumlein. A short MITL (magnetically insulated transmission line) connects the cavities, via a vacuum pumping section, to a chamber containing an e-beam diode test load. (4 pages)

High intensity pulsed electron beams are used to create bremsstrahlung x-ray sources for flash radiographic interrogation of dynamic experiments. Typical industrial sources operate below 200 GW/cm2 intensities, while experimental requirements can demand above 50 TW/cm2. Recent developments in pulsed power-driven high intensity electron beam systems have significantly increased these operating regimes, demonstrating 20 TW/cm2, and computations predict successful extrapolation to higher intensities. Detailed studies of electron beam configurations, both theoretical and experimental, and the prognosis for each to increase to the required levels is discussed.

Summary form only given. The radiographic source for a new Hydrodynamic Research Facility (HRF) at AWE, Aldermaston is a 14 MV, 100-kA to 150kA, Induction Voltage Adder (IVA) designed by AWE and PSD. At the end of the accelerator the inner and outer conductor of the magnetically insulated transmission line (MITL) are flared into a larger radii dustbin region. The inner conductor (cathode) terminates in a shaped non-emitting dome attached to an emitting sphere on rod, which makes up the cathode of an intense paraxial-focussed diode. This paper reports on electromagnetic particle-in-cell (PIC) simulations of the power flow from the end of the accelerator through the diode. Since the ideal current for the focussed diode is 40-60 kA the non-emitting dome is intended to shunt the remaining current to the outer conductor without disrupting the focus. The LSP PIC code was used to calculate the voltage and current delivered to the paraxial-focussed diode for different A-K spacings, accelerator output impedances, and MITL geometries (length and shape of taper as well as different radii). We also investigated the impact on power flow of emission from a portion of the dome, which is impacted by the MITL vacuum flow electrons. The ability of the non-emitting dome to re-trap a portion of the vacuum flow electrons in the cathode prior to the focussed diode is also addressed. The voltage input to the simulation was taken from the transmission line code predictions for the voltage at the entrance to the dustbin. Using a realistic voltage waveform was important since the shape and position of the shunting vacuum flow electrons depended strongly on the rate of change of the voltage. Interpretation of the simulation results also requires that the large capacitance of the non-emitting dome be considered. Initial results imply that the output section of the AWE-IVA can be designed to function as desired.

The Atomic Weapons Establishment at Aldermaston, U.K., has a number of pulsed-power-driven flash X-ray machines for diagnosing the hydrodynamics of explosively-driven high-Z materials. The most powerful of these machines is Mogul-E, which operates at about 10 MV and 30 kA, delivering about 450 R @ 1 m in a 5 mm spot. Longer-term plans envisage the upgrading of existing facilities through the construction of a multi-axis Hydrodynamic Research Facility (HRF). It is proposed that the HRF will be furnished initially with three inductive voltage adder (IVA) machines operating at /spl sim/14 MV, each giving 600 R @ 1 m in a 5 mm spot. It is envisaged that, following further research, the outputs will be increased towards 1000 R, with, it is hoped, a reduction in X-ray spot size. More speculative proposals involve increasing the number of machines from 3 to 5, and/or the splitting of the output end of one or more machines to drive more than one X-ray source per machine. An overview of the research programme necessary to achieve these aims is presented.

RTFS (Radiographic Integrated Test Stand) is an induction voltage adder designed by Sandia and PSD to provide 16-MV, 150-kA electron beams and other capabilities. Previous publications have reported on tests of a single pulse forming line and adder cell, including initial results of the effects of various degrees of non-uniform injection of current into the adder bore on magnetic insulation and power flow in the downstream MITL. Now RITS-3 has been constructed, consisting of three pfls driven by a common intermediate store; three induction cells, one driven by each pfl; a three-stage, 4-MV, 150-kA vacuum voltage adder; and an output MITL and diode. Here we report on (1) simulations of the three-stage adder using the MRC 3-D particle-in-cell code LSP that address the effects of injected current non-uniformities on magnetic insulation and power-flow both upstream and downstream in a multi-cell adder; (2) experimental results compared with simulations; and (3) initial performance of the RITS-3 pulse power.

As a result of the Pulsed Power Group's programme to gain expertise in IVA (inductive voltage adder) technology we have designed a prospective module for a future large IVA machine that would be suitable for full scale radiography at AWE. This machine would operate at 13 to 15 MV. The PIM machine has been built at AWE to test a single such module and we have been collecting performance data from it before fixing the design for the full size machine. PIM also provides the basis for the PPG's continued research programme in this area and hence possible other machines for the future. The PIM machine consists of a single inductive cavity pulsed by a water dielectric Blumlein pulse forming line which is charged by a Marx generator. The Marx generator is based on the Sandia hanging Marx design. It is constructed using 32 1.35 uF 100 kV DC capacitors. The Marx pulse charges the 10 ohm water Blumlein. which acts as a fast pulse forming section. The Blumlein switch initially installed was a co-axial two stage rimfire and trigatron switch. A laser triggering configuration utilising two radial switches, designed by Titan PSI, is also to be tested. The inductive cavity incorporates ferromagnetic cores formed from Metglas alloy tape and the fast pulse appears across a perspex (lucite) insulated accelerating gap. The pulse travels along a short vacuum transmission line to a ball/plane e-beam diode test load. Work has been carried out to test the module design. Initially the Marx was characterised into a resistive load to establish its operating regime. The Blumlein PFL and its Rimfire/Trigatron switch were then tested, also using a copper sulphate load. A peak output voltage of 1.7 MV was achieved from the PFL. The inductive cavity has been added and we are in the process of testing it. A pulse of 1.4 MV peak voltage has been successfully applied to it. The latest details of the system's characterisation are presented.