Mark L. Kiefer

Simulation codes for light-ion diode modeling

The computational tools used in the investigation of light-ion diode physics at Sandia National Laboratories are described. Applied-B ion diodes are used to generate intense beams of ions and focus these beams onto targets as part of Sandias inertial confinement fusion program. Computer codes are used to simulate the energy storage and pulse forming sections of the accelerator and the power flow and coupling into the diode where the ion beam is generated. Other codes are used to calculate the applied magnetic field diffusion in the diode region, the electromagnetic fluctuations in the anode-cathode gap, the subsequent beam divergence, the beam propagation, and response of various beam diagnostics. These codes are described and some typical results are shown.

Three‐dimensional particle‐in‐cell simulations of applied‐B ion diodes

The three‐dimensional particle‐in‐cell code quicksilver [Seidel et al., Computational Physics, edited by A. Tenner (World Scientific, Singapore, 1991), p. 475] has been used to study applied‐B ion diodes. The impedance behavior of the diode in these simulations is in good agreement with both analytic theory and experiments at peak power. The simulations also demonstrate the existence of electromagnetic instabilities which induce divergence in the ion beam. Early in time, there is an instability at high frequency relative to the ion transit time τi, and the resulting beam divergence is low. However, later in time, the system makes a transition to an instability with a frequency close to 1/τi, and the ion beam divergence rises to an unacceptably high value. The transition is associated with the build‐up of electron space charge in the diode, and the resulting increase in the beam current density enhancement (J/JCL). Using different schemes to inhibit the electron evolution, the transition has both been post...

Theory of instability-generated divergence of intense ion beams from applied-B ion diodes

Over the course of the past few years, rapid progress has been made in the development of a theoretical understanding of the physics of applied-B ion diodes. Success in predicting diode current and voltage operating characteristics has been followed by new insight into the effects of electromagnetic instabilities on ion beam divergence

Recent developments in the 3D, electromagnetic, particle-in-cell code, QUICKSILVER

Summary form only given, QUICKSILVER code has been developed for performing charged-particle simulations in three dimensions using finite-difference, electromagnetic, particle-in-cell (PIC) techniques. QUICKSILVER has been designed to make advantageous use of vector, multiple-CPU computers with large central memory and fast, out-of-memory storage. Integral to QUICKSILVERs use are a user-friendly preprocessor, MERCURY, and several postprocessors for displaying various types of data. The MERCURY preprocessor is designed to allow the rapid setup of complex 3-D simulations while minimizing the errors that might be associated with that setup. MERCURY supports all the modeling capabilities of QUICKSILVER, as well as giving limited support (e.g., mesh generation) for other codes. Several postprocessors are used to analyze the wealth of data that can be produced by large 3-D codes. QUICKSILVER and its associated pre- and postprocessors reside on several different machines. This has necessitated the development of a portable file format which allows data to be transported among the various machines in an efficient, machine-independent binary file format

Contribution of the backstreaming ions to the self-magnetic pinch (SMP) diode current

Summary form only given. The results presented here were obtained with an SMP diode mounted at the front high voltage end of the RITS accelerator. RITS is a Self-Magnetically Insulated Transmission Line (MITL) voltage adder that adds the voltage pulses of six 1.3 MV inductively insulated cavities. Our experiments had two objectives: first to measure the contribution of the back-streaming ion currents emitted from the anode target to the diode beam current, and second to try to evaluate the energy of those ions and hence the actual Anode-Cathode (A-K) gap actual voltage. In any very high voltage inductive voltage adder (IVA) utilizing MITLs to transmit the power to the diode load, the precise knowledge of the accelerating voltage applied on the anode-cathode (A-K) gap is problematic. The accelerating voltage quoted in the literature is from estimates based on measurements of the anode and cathode currents of the MITL far upstream from the diode and utilizing the para-potential flow theories and inductive corrections. Thus it would be interesting to have another independent measurement to evaluate the A-K voltage. The diodes anode is made of a number of high Z metals in order to produce copious and energetic flash x-rays. The backstreaming currents are a strong fraction of the anode materials and their stage of cleanness and gas adsorption. We have measured the back-streaming ion currents emitted from the anode and propagating through a hollow cathode tip for various diode configurations and different techniques of target cleaning treatments, such as heating to very high temperatures with DC and pulsed current, with RF plasma cleaning and with both plasma cleaning and heating. We have also evaluated the A-K gap voltage by ion filtering techniques.

Numerical simulation of cathode plasma dynamics in magnetically insulated vacuum transmission lines

A novel algorithm for the simulation of cathode plasmas in particle-in-cell codes is described and applied to investigate cathode plasma evolution in magnetically insulated transmission lines (MITLs). The MITL electron sheath is modeled by a fully kinetic electron species. Electron and ion macroparticles, both modeled as fluid species, form a dense plasma which is initially localized at the cathode surface. Energetic plasma electron particles can be converted to kinetic electrons to resupply the electron flux at the plasma edge (the “effective” cathode). Using this model, we compare results for the time evolution of the cathode plasma and MITL electron flow with a simplified (isothermal) diffusion model. Simulations in 1D show a slow diffusive expansion of the plasma from the cathode surface. But in multiple dimensions, the plasma can expand much more rapidly due to anomalous diffusion caused by an instability due to the strong coupling of a transverse magnetic mode in the electron sheath with the expanding resistive plasma layer.

A monte-carlo code for computing transport coefficients in weakly ionized gas

Summary form only given. A Monte-Carlo code (MCSwarm) has been developed to provide transport coefficients in weakly ionized gases. The code can generate transport coefficients in crossed electric and magnetic fields for a wide varietv of gases including N2, O2, SF6 , H2, H2O, Ar, Ne, CO2, and He. The code follows many different interactions including rotational and vibrational modes, electronic excitation, ionization, and momentum transfer. The MC Swarm code provides an accurate method for computing electron mobility, ionization, and attachment rates suitable for particle-in-cell modeling of electron-beam-induced conductivity. The transport coefficients generated by MCSwarm are suitable at pressures above 1 torr. These coefficients have been incorporated into the fully electro-magnetic 3D PIC code Quicksilver (QS). To benchmark the QS model, we have modified an existing Febetron electron-beam pulser to produce an 100 ns electron beam pulses with a peak energy of 80 keV and a current density ranging from 1 A/cm2 to 1 kA/cm2. The net transported current 10 cm from the beam injection location agrees well with the QS simulations. The line-integrated electron density is measured by laser interferometry and also agrees well with the code calculations. The simulations show that emission boundary conditions are very important in determining where currents flow and that the effects of beam scattering in air are negligible for pressures below 20 torr. The code calculations further show that, for the parameters considered here, the electric field at high pressure is too small to cause avalanche and plasma production is dominated by beam-impact ionization.

Zeeman spectroscopy as a method for determining the magnetic field distribution in self-magnetic-pinch diodes (invited)

In the self-magnetic-pinch diode, the electron beam, produced through explosive field emission, focuses on the anode surface due to its own magnetic field. This process results in dense plasma formation on the anode surface, consisting primarily of hydrocarbons. Direct measurements of the beams current profile are necessary in order to understand the pinch dynamics and to determine x-ray source sizes, which should be minimized in radiographic applications. In this paper, the analysis of the C IV doublet (580.1 and 581.2 nm) line shapes will be discussed. The technique yields estimates of the electron density and electron temperature profiles, and the method can be highly beneficial in providing the current density distribution in such diodes.

Contribution of the backstreaming ions to the Self-Magnetic pinch (SMP) diode current

The results presented here were obtained with an SMP diode mounted at the front high voltage end of the RITS accelerator. RITS is a Self-Magnetically Insulated Transmission Line (MITL) voltage adder that adds the voltage pulses of six 1.3 MV inductively insulated cavities.

In-situ anode heating and plasma glow discharge cleaning and its effects on atomic constituents in the A-K gap in Self-Magnetic Pinch (SMP) experiments

Summary form only given. The RITS-6 inductive voltage adder (IVA) accelerator (3.5-8.5 MeV) at Sandia National Laboratories produces highpower (TW) focused electron beams (<; 3mm diameter) for flash x-ray radiography applications. The Self-Magnetic Pinch (SMP) diode utilizes a hollowed metal cathode to produce a pinched focus onto a high-Z metal anode converter. There is not a clear understanding as to the effects various contaminants such as: C, CO, H, H<;sub>2<;/sub>O, H<;sub>m<;/sub>C<;sub>n<;/sub>, O<;sub>2<;/sub>, and N<;sub>2<;/sub>, on the anode surface or in the bulk, may have on impedance dynamics, beam stability, beam spot size, and reproducibility.Various cleaning/outgassing methods have been explored such as: heating bulk Ta to temperatures of 1,000 °C for ~2000 s or more with and without thin films of pure Al or Au, pulse heating Ta foils (~100μm thick) to ~2,400 oC for ~1 s, and plasma glow discharge cleaning using an Argon-Oxygen 80/20 gas mixture for ~2000 s. The effects of in-situ cleaning were characterized via in-situ residual gas analysis, separate Temperature Programmed Desorption of witness samples, thermal modeling, and ultimately through radiographic and pulsed power performance of the diode. Initial experiments indicate a significant reduction in H and C as indicated by high-speed spectral analysis of plasmas at the converter and a reduction of back-streaming proton currents. Experiments are ongoing, and latest results will be reported.