Vacuum insulator development for the dielectric wall accelerator

Abstract

At Lawrence Livermore National Laboratory, we are developing a new type of accelerator, known as a Dielectric Wall Accelerator, in which compact pulse forming lines directly apply an accelerating field to the beam through an insulating vacuum boundary. The electrical strength of this insulator may define the maximum gradient achievable in these machines. To increase the system gradient, we are using "High Gradient Insulators" composed of alternating layers of dielectric and metal for the vacuum insulator. In this paper, we present our recent results from experiment and simulation, including the first test of a High Gradient Insulator in a functioning Dielectric Wall Accelerator cell. a) Electronic mail: harris89@llnl.gov 2

Vacuum Surface Flashover of High Gradient Insulators Under Nanosecond Pulse

“…As the number of the dielectric layers increases, the coupling effect will gradually dominate the performance of the HGI because more ICJTs are involved in the process. Using thicker metal layers may reduce the coupling effect among layers and hence improve the performance of the HGI, while the thickness ratio I /M should stay high enough, otherwise the effective insulator length will reduce significantly and the performance of the HGI will decline conversely [7]. For instance, the thickness ratio I /M of the HGI with eight dielectric layers reported in this paper is ∼60.…”

“…In addition, the experimental data shown in Fig. 8 as well as those given in [2] and [7] are all rather scattered. The reason could be that the surfaces of the thin metal layers at the CTJ/ICTJs are difficult to be polished to the ideal condition.…”

“…The test procedure is similar to what was used in [7]. The difference is that Harris increased and reduced the output voltage by the same small step.…”

“…The flashover strength of the best HGI sample is ∼75% higher than that of the conventional alumina insulator. The HGIs assembled from submillimeter stainless steel foils with fixed thickness M and dielectric sheets with different thickness I have been tested under the voltage pulsewidth of 20-120 ns in [2], [7], and [8]; the best HGIs with different materials and sizes performed 20%-60% better than corresponding conventional insulators. The HGI samples generally showed improved performance as I increased for a fixed thickness of M in most of these experiments [2], [6]- [8], which is in contradiction with the scaling law.…”

Measurements of an expanding surface flashover plasma

“…Because vacuum surface flashover presents one of the primary limits on the achievable electric field in vacuum devices, an improved understanding of this process and the methods to control it is critical for the development of advanced, compact devices. [2][3][4] However, the ability of these discharges to produce copious quantities of light and plasma has also been put to good use in systems as diverse as closing 5 and opening 6 switches, as well as electron, 7-9 ion, 10,11 x-ray, 12 UV, 13 and microwave 14 sources. In many of these applications, the velocity at which the plasma expands away from the discharge site is of great importance.…”