Search

The numerical solution of very large three-dimensional electromagnetic field problems are challenging for various applications in the industry. In this paper, we propose a nonoverlapping domain decomposition approach for solving the three-dimensional Maxwell equations on MIMD computers, based on a mixed variational formulation. It is especially well adapted for the solution of the Vlasov-Maxwell equations, widely used to simulate complex devices like particle injectors or accelerators. This approach has the important property that it leads to reuse without modification most of an existing sequential code. The continuity at the interfaces is imposed by duality using Lagrange multipliers. Hence, the resulting parallel algorithm requires only to add an external preconditioned Uzawa solver. We present the results of some numerical experiments on a parallel distributed memory machine.

In this paper, a new method for coalescing particles in PIC codes is proposed. This coalescing process conserves the particle and mesh charge and current densities as well, and also the particle energy. Particular attention is devoted to the derivation of a method as general as possible, to be easily extended to other problems. Numerical examples are shown to illustrate the efficiency of the method.

The aim of this paper is to present a new explicit time scheme for electromagnetic particle simulations. The main property of this new scheme, which depends on a parameter, is to reduce and in some cases to suppress numerical instabilities that can appear in this context and are widely described in the literature. Other numerical properties are also investigated, and a numerical example is finally given to illustrate our purpose.

The aim of this paper is to present an extension of theP1conforming finite element method for the time-dependent Maxwell equations that has been previously exposed. We shall consider inhomogeneous media with the piecewise constant dielectric and magnetic parameters, and precisely, the interface of two such media. By analogy with the idea developed, we propose a method based on the dualization and then the approximation of the interface conditions in a way consistent with the one derived for the fields and the Lagrange multipliers of the divergence constraints inside the domain.

The aim of this paper is to present a method for solving the time-domain three-dimensional Maxwell equations, which can be coupled with a particle solver. For this purpose, Maxwell's equations are reformulated as a constrained wave equation system, with Lagrange multipliers associated to the conditions ∇ · β - 0 and ∇ · E - ρ/e 0 . We approximate both the fields and the Lagrange multipliers with a finite element method using a Taylor Hood element.

This paper proposed a linear analysis to overcome the numerical Cherenkov instability. Basically, it is based on a explicit time scheme for solving electromagnetic particle simulations. This scheme depends on a parameter, that allows us to reduce and in some cases to suppress the numerical Cherenkov instability that can appear in this context, and is widely described in the literature. Some properties of the scheme are also investigated. Numerical examples are finally given to illustrate our purpose.

In the general framework of charged particle beam modelling, we present a numerical method for solving the 3D Vlasov equation on unstructured meshes. We pay a particular attention to the vectorization of the resulting algorithms. The analysis of the execution CPU time and the comparison with the 2D case indicate the effectiveness of the proposed methods.

The numerical solution of very large 3D electromagnetic field problems are challenging for various applications in the industry. In this paper, we propose a nonoverlapping domain decomposition approach for solving the 3D Maxwell equations on MIMD computers, based on a mixed variational formulation. It is especially well adapted for the solution of the Vlasov-Maxwell equations, widely used to simulate complex devices like particle injectors or accelerators. This approach in particular leads to reuse without modification most of an existing sequential code.

A coupled particle finite-element method is developed for the numerical resolution of the time-dependent relativistic Vlasov-Maxwell equations in axisymmetric geometries. The use of a conforming finite-element method for the Maxwell solver supplemented with a mass-lumping technique leads to an explicit system for the computation of the electromagnetic fields components. For the Vlasov solver, a fully vectorized algorithm is used for the location of the particles in an unstructured mesh made up with triangles. The numerical values of the electromagnetic fields radiated by a dipole are compared with the analytical solution. The whole Vlasov-Maxwell solver is tested on a simplified free-electron laser electron injector, designed for an RF-based free electron laser and the results are compared with those obtained with a finite-difference code. >

There is a great interest in developing simulation methods that are more appropriate to the geometrical complexity of real devices. To approach geometric realism, we have investigated two different numerical methods to solve Maxwell's equations in r − z geometries: (1) a conforming finite-element formulation, (2) a finite-volume formulation on Delaunay-Voronoi meshes. We describe the design of two prototype PIC codes called VLAMINCK and MATISSE which apply these methods on unstructured meshes to fully electromagnetic particle simulation. Both codes are tested on a simplified injector geometry and the results are compared to those obtain with the MASK code using finite-difference formulation.

A picosecond gated MCP framing camera system is presented. The camera with four parallel microstripline cathode haslonger measuring time range and better gain uniformity. The camera is equipped with a pinhole-array adjustment systemand other practical accessories for laser plasma experiment. The measured spatial resolution ofthe camera is l5lp/mm with10% modulation and the exposure time ranges from 6Ops to iOOps with different gain. The whole system was used at the1 l# laser facility and got some results of several target type during laser plasma experiment.Keywords: MCP gating, framing camera, picosecond 1. INTRODUCTION Picosecond gated MCP framing camera has been developing rapidly during the last ten years and become an important diagnostic equipment for laser plasma experiment'71. We had reported a gated MCP framing camera using a single meander microstripline cathode with exposure time of iOOps and 6Ops"21. In this paper, a more functional gated MCPframing camera system is presented. With four parallel microstripline cathode, this camera has a longer measuring timerange and better gain uniformity. And some factors of operating during laser-plasma experiment was taken into account insystem designing and constructing. The camera is also equipped with a pinhole-array adjustment system and other practicalaccessories for laser plasma experiment.

MicroChannel Plates (MCPs) are used in high speed cinematography systems such as MCP framing cameras and streak camera readouts. In order to know the dynamic range or the signal to noise ratio that are available in these devices, a good knowledge of the performances of the MCP is essential. The point of interest of our simulation is the working mode of the microchannel plate--that is light pulsed mode--, in which the signal level is relatively high and its duration can be shorter than the time needed to replenish the wall of the channel, when other papers mainly studied night vision applications with weak continuous and nearly single electron input signal. Also our method allows the simulation of saturation phenomena due to the large number of electrons involved, whereas the discrete models previously used for simulating pulsed mode might not be properly adapted. Here are presented the choices made in modeling the microchannel, more specifically as for the physics laws, the secondary emission parameters and the 3D- geometry. In a last part first results are shown.

High precision and high dynamic range measurements are required to properly characterize the pulseshape in ICF laser systems. The dynamic range and precision of the measurements that can be made with streak tubes is determined by the number of photoelectrons that can be transported to the recording screen per channel and per temporal sample of the signal. They are limited by the overall current that the tube can deliver without distorting the signal. In order to build tubes with the large dynamic ranges required by this application, we need to understand at what current density level the tube response becomes nonlinear. We present results of experiments made by using a laser illuminating several streak tubes at various intensities. We show that charge depletion in the photocathode can occur in pulsed mode and limit the signal to levels well below that where space charge induced nonlinearities appear. The use of scientific grade CCD's for recording the streak traces has allowed the introduction of a new method to measure the absolute current in the tube. A 2D PIC electron optics code has been used to simulate the propagation of the beam in the input diode (photocathode/accelerating electrode) and in the drift region (accelerating electrode/screen) of a bilamellar streak tube. We compare the numerical results with the experiments. We conclude by comparing the bilamellar and classical electron optics tubes and show the advantages of the former tube design for this application.© (1997) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.