[en] The second axis of the DARHT flash X-ray facility at Los Alamos National Laboratory (''DARHT-II'') is a multiple-pulse, 18.4 MeV, 2 kA induction electron linear accelerator [1]. A train of short (∼50 ns) pulses are converted via bremsstrahlung to X-rays, which are then used to make radiographic images at various times (nominally four) during a ''hydrotest'' experiment. The train of pulses is created by carving them out of a two microsecond long macropulse, using a fast switching element called a kicker [2]. The unused portion of the macropulse is absorbed in a beam dump. Thus, upstream of the kicker, two microseconds of beam are transported through a vacuum system roughly sixty meters long. These conditions involve length and, specifically, time scales which are new to the transport of high-current beams. A concern under such conditions are the macroscopic interactions between the electron beam and positive ions created by impact ionization of the residual gas in the vacuum system. Over two microseconds, the ion density can develop to a hundredth or even a tenth of a percent of the beam density--small, to be sure, but large enough to have cumulative effects over such a long transport distance. Two such effects will be considered here: the ion hose instability, where transverse forces conspire to pull the electron beam farther and farther off axis, and background gas focusing, where radial forces (with respect to the beam) change the beam envelope during the course of the macropulse. The former effect can cause beam emittance growth (affecting the ability to focus the beam on the target) and eventually catastrophic beam loss; the latter can cause either serious degradation of the statically tuned final focus on the converter target, or a pinching of the beam on the surface of the main dump to the point where the heat flux causes damage. The beam transport upstream of the kicker has two distinct phases. First, the beam is created and accelerated up to 18.4 MeV over a distance of about fifty meters. Then the true downstream transport begins: the beam drifts through a matching section in preparation for the kicker, over some ten meters; the long-pulse beam then travels about four more meters from the kicker to the main dump. In the accelerator, the beam energy is obviously not constant; the transport is emittance-dominated and done through nearly continuous solenoidal focusing. In the downstream section, there are only two discrete solenoids over the entire fourteen meters and the transport is largely ballistic. Since ion hose has been studied in the accelerator [3] and since the lack of continuous focusing is considered a concern with respect to ion hose in the downstream section, the focus of this study is only from the exit of the accelerator to the main dump. A more in-depth description of the baseline (ion-free) DARHT-II downstream transport, including description of the actual transport elements and their use, will not be presented in this document; such details can be found in the documents cited in the References. The study of these effects will be done in stages. In the next section, the nature of the residual gas in the vacuum system will be considered, along with the various assumptions made in characterizing the creation of ions. Then the ion hose instability will be described in its simplest form. In the fourth section, additional features of ion hose will then be presented which attempt to capture some of the key behavior. Then a much more complete model using particle-in-cell (PIC) numerical techniques will be described, followed by details of the specific implementation used here. In section seven, the code will be benchmarked against results published in the literature. Section eight has the most relevant material: the actual study of the effects of ion hose and background gas focusing in the DARHT-II downstream transport region. In section nine, a simple experiment which can be tacked on to existing experiments is proposed in order to verify the modeling. Finally, the results are summarized and the very last section lists references

[en] Neutralized transport of relativistic electron beams can achieved in various circumstances. In one form, the beam is transported through a plasma, either pre-formed or beam generated, where the plasma electrons are ejected due to the space charge influence of the beam. The beam can be fully neutralized this way if the plasma is sufficiently dense. Typically, the transport physics of concern in this case are the various macro- and micro-instabilities that can develop due to interactions of the beam with the plasma; charge and current neutralization are certainly important but tend to be just one set of concerns among many. The study of beam/plasma interactions has been active for many years [e.g. 1]. In a different scenario, the beam impinges on a plasma with a sharp boundary (as maintained on the timescale of a beam pulse) and, via space charge, extracts ions from the plasma; extraction energies can be hundreds of kilovolts in the case of tightly focused, high current beams. In this case, the ions have a lower density than the beam and are not accompanied by a plasma electron population; the main transport issue is charge neutralization. Such a sharply bounded plasma can occur via ionization of surface impurities from a solid target; the transport of the beam through this thin layer is typically not of interest relative to the transport upstream of the surface and the beam/target interactions beyond the surface. Since the partial neutralization of the beam changes its focusing characteristics on the target, and since the high extraction energy means the ion column is moving rapidly into the beam and introducing strong time variation, this 'backstreaming ion' phenomenon has been an area of active study in the transport of the high-intensity electron beams used in radiographic accelerators (see [2] for an example of such machines). However, much of the work has been experimental [3] and numerical [4]. The conceptual understanding provided by pencil-and-paper analysis thus far [5,6] has covered the important topics (such as disruption length and neutralization fraction, to be defined later) but generally in the context of idealized beam envelopes that lack enough detail to examine the 'control knobs' in a focusing system, such as focal length, time-varying magnetic lenses, or time-averaged focusing. The obvious reason for the lack of analytic results is that they are hard to come by, even in highly simplified scenarios. This study will make quite a few simplifying assumptions and still will not produce much in the way of simple closed-form answers. However, it is possible to provide limiting expressions for some of the quantities of interest, and for others there are simple algebraic systems that can studied via spreadsheet rather than a more time-consuming analysis package. Furthermore, the reduced expressions also provide a fast model that can be embedded in more complex parametric or optimization studies, where repeated calls to something like a PIC numerical model would be impractical

[en] The bremsstrahlung converter target in radiographic accelerators is not, in general, considered a high-technology piece of equipment. In its essential form it is merely a solid plate of high-Z metal, usually tungsten (W) or tantalum (Ta); electrons go in, X-rays come out [1]. However, there are some important factors to keep in mind for this kind of target system. One is a constraint on the target itself: the proper thickness of material. Too little material reduces the probability that an electron will have a significant nuclear collision before exiting the plate. Too much material has a number of effects: small-angle scattering will occur to such an extent that bremsstrahlung photons will not be pointed in the forward direction. Electrons which small-angle scatter away and then back to the forward direction will have moved to larger radii as they traverse the target, increasing the effective source size. Electrons ''backscattered'' from the target --primaries or secondaries ejected from the upstream surface after sufficient angular scatter--exert a defocusing force on the incoming beam due to increased space charge at fixed (or even slightly reduced) current. Finally, a sufficiently thick target will begin to self-attenuate the X-ray photons produced in the upstream portion of the plate. A second constraint is obvious but is harder to accommodate when designing a radiographic accelerator system. The angular distribution of the incoming electron beam will change the forward dose. Just as electrons which have undergone small-angle scatter will no longer produce forward dose, electrons which have large angles before they ever enter the target cannot produce forward dose. Accurate prediction of dose requires incorporating the effect of the initial angle of the electron coming into the target material. The further step of controlling the angular distribution--which means keeping it as close to zero as possible--is difficult since it tends to drive important beam parameters in directions we do not want (large spot size) or cannot achieve (very low emittance). In this report we characterize the bremsstrahlung performance of Ta and W converter targets over a range of electron energies (2-20 MeV) and angles. The studies are all performed with the MCNP radiation transport code [2], version 4b. A number of steps are involved in the process. First, we must construct the absorption properties of air given photons of various energies, so that the distribution of photons produced by the incoming electrons can be converted into a dose rate as commonly used in radiography. Then we study the dose rate of various electron energies in Ta and W, as a function of target thickness, and find that there is an optimum for a given energy. Following that, we describe how to incorporate angular dependence in the (rather obtuse) MCNP interface, and then study how the dose rate varies with the distribution of incoming angles. The major results and useful curve fits are summarized at the end. The appendix contains listings of useful scripts and MCNP templates. This report does not address the backscattered electron issue [3]. That problem is complex because it requires self-consistent treatment of the electrons in the external electric fields once they are ejected from the target. Suffice it to say that a thicker target produces more backscatter and therefore more defocusing. When choosing a target thickness, rather than selecting from the peak output given below, one may prefer to choose a (usually much thinner) target corresponding to the 95% output level, or even lower as desired. Of course, a target which must withstand multiple beam pulses has constraints driving the thickness in the opposite direction, in order to maintain sufficient line density during the hydrodynamic evolution of the material

[en] Using dielectric wall accelerator technology, we are developing a compact induction accelerator system primarily intended for pulsed radiography. The accelerator would provide a 2-kA beam with an energy of 8 MeV, for a 20-30 ns flat-top. The design goal is to generate a 2-mm diameter, 10-rad x-ray source. We have a physics design of the system from injector to the x-ray converter. We present the results of injector modeling and PIC simulations of beam transport. We also discuss the predicted spot size and the on-axis x-ray dose

[en] Using the dielectric wall accelerator technology, they are developing a compact induction accelerator system primarily intended for pulsed radiography. Unlike the typical induction accelerator cell that is long compared with its accelerating gap width, the proposed dielectric wall induction accelerator cell is short and its accelerating gap width is comparable with the cell length. In this geometry, the RF modes may be coupled from one cell to the next. They will present recent results of RF modeling of the cells and a prediction of the transverse beam instability on a 2-kA, 8-MeV beam

[en] High current electron beams have been used as x-ray drivers for x-ray radiography. Typically, several thousand amperes of electron beam current at 20 MeV is focused to a millimeter spot size on a x-ray converter. Within a single pulse, the heating of the target by the electron beam will lead to rapid desorption of surface contaminants. The space charge potential of the electron beam will pull ions out of this plasma layer upstream into the beam. These backstreaming ions can act as a focusing lens which cause the beam to be overfocused at a waist upstream. The final beam spot size on the target would then be larger than intended, and the x-ray radiography resolution is reduced. We have designed a self-biased ion trap for the Experimental Test Accelerator (ETA-II) beam by using an Advanced Test Accelerator (ATA) inductive cell to prevent the backstreaming ions from moving upstream and forming a long ion focusing channel. We have studied the effects of this type of ion trap on the final focusing of the electron beam with the ETA-II beam parameters. Simulation results will be presented

[en] Ions extracted from a solid surface or plasma by impact of an high intensity and high current electron beam can partially neutralize the beam space charge and change the focusing system. We have investigated ion emission computationally and experimentally. By matching PIC simulation results with available experimental data, our finding suggests that if a mix of ion species is available at the emitting surface, protons dominate the backstreaming ion effects, and that, unless there is surface flashover, ion emission is source limited. We have also investigated mitigation, such as e-beam cleaning, laser cleaning and ion trapping with a foil barrier. The temporal behavior of beam spot size with a foil barrier and a focusing scheme to improve foil barrier performance are discussed

[en] Four short current pulses with various pulse widths and spacing will be delivered to the x-ray converter target on the second-axis of the Dual-Axis Radiographic Hydrodynamic Test (DARHT-II) facility. To ensure that the DARHT-II multi-pulse target will provide enough target material for x-ray production for all four pulses, the target needs either to survive the strike of four electron pulses or to accommodate target replenishment. A distributed target may survive hitting of four electron pulses. For target replenishment, two types of target configurations are being considered: stationary target systems with beam repositioning and dynamic moving target systems. They compare these three target systems and their radiographic performance

[en] The second axis of the Dual Axial radiography Hydrodynamic Test (DARHT-II) facility at LANL is currently in the commissioning phase[1]. The beam parameters for the DARHT-II machine will be nominally 18 MeV, 2 kA and 1.6 (micro)s. This makes the DARHT-II downstream system the first system ever designed to transport a high current, high energy and long pulse beam [2]. We will test these physics issues of the downstream transport system on a scaled DARHT-II accelerator with a 7.8-MeV and 660-A beam at LANL before commissioning the machine at its full energy and current. The scaling laws for various physics concerns and the beam parameters selection is discussed in this paper

[en] A flashover arc source that delivered up to 200 mJ on the 100s-of-ns time-scale to the arc and a user-selected dielectric surface was characterized for studying high-explosive kinetics under plasma conditions. The flashover was driven over thin pentaerythritol tetranitrate (PETN) and poly(methyl methacrylate) (PMMA) dielectric films and the resultant plasma was characterized in detail. Time- and space-resolved temperatures and electron densities of the plasma were obtained using atomic emission spectroscopy. The hydrodynamics of the plasma was captured through fast, visible imaging. Fourier transform infrared spectroscopy (FTIR) was used to characterize the films pre- and post-shot for any chemical alterations. Time-resolved infrared spectroscopy (TRIR) provided PETN depletion data during the plasma discharge. For both types of films, temperatures of 1.6-1.7 eV and electron densities of ∼7-8 x 10¹⁷/cm³∼570 ns after the start of the discharge were observed with temperatures of 0.6-0.7 eV persisting out to 15 μs. At 1.2 μs, spatial characterization showed flat temperature and density profiles of 1.1-1.3 eV and 2-2.8 x 10¹⁷/cm³ for PETN and PMMA films, respectively. Images of the plasma showed an expanding hot kernel starting from radii of ∼0.2 mm at ∼50 ns and reaching ∼1.1 mm at ∼600 ns. The thin films ablated or reacted several hundred nm of material in response to the discharge. First TRIR data showing the in situ reaction or depletion of PETN in response to the flashover arc were successfully obtained, and a 2-μs, 1/e decay constant was measured. Preliminary 1 D simulations compared reasonably well with the experimentally determined plasma radii and temperatures. These results complete the first steps to resolving arc-driven PETN reaction pathways and their associated kinetic rates using in situ spectroscopy techniques.