We have developed an empirical electrical-breakdown relation that can be used to design large-area water-insulated pulsed-power systems. Such systems often form an integral part of multiterawatt pulsedpower accelerators, and may be incorporated in future petawatt-class machines. We find that complete dielectric failure is likely to occur in water between a significantly field-enhanced anode and a lessenhanced cathode when E p 0:3300:026 eff 0:135 0:009. In this expression E p V p =d is the peak value in time of the spatially averaged electric field between the anode and cathode (in MV=cm), V p is the peak voltage across the electrodes, d is the distance between the anode and cathode, and eff is the temporal width (in s) of the voltage pulse at 63% of peak. This relation is based on 25 measurements for which 1 V p 4:10 MV, 1:25 d 22 cm, and 0:011 eff 0:6 s. The normalized standard deviation of the differences between these measurements and the associated predictions of the relation is 12%.

We have conducted a series of experiments designed to measure the flashover strength of various azimuthally symmetric 45 vacuum-insulator configurations. The principal objective of the experiments was to identify a configuration with a flashover strength greater than that of the standard design, which consists of a 45 polymethyl-methacrylate (PMMA) insulator between flat electrodes. The thickness d and circumference C of the insulators tested were held constant at 4.318 and 95.74 cm, respectively. The peak voltage applied to the insulators ranged from 0.8 to 2.2 MV. The rise time of the voltage pulse was 40 -60 ns; the effective pulse width [as defined in Phys. Rev. ST Accel. Beams 7, 070401 (2004)] was on the order of 10 ns. Experiments conducted with flat aluminum electrodes demonstrate that the flashover strength of a crosslinked polystyrene (Rexolite) insulator is 18 7% higher than that of PMMA. Experiments conducted with a Rexolite insulator and an anode plug, i.e., an extension of the anode into the insulator, demonstrate that a plug can increase the flashover strength by an additional 44 11%. The results are consistent with the Anderson model of anode-initiated flashover, and confirm previous measurements. It appears that a Rexolite insulator with an anode plug can, in principle, increase the peak electromagnetic power that can be transmitted across a vacuum interface by a factor of 1:18 1:44 2 2:9 over that which can be achieved with the standard design.

A 6.1-MV, 0.79-MA laser-triggered gas switch (LTGS) is used to synchronize the 36 modules of the Z machine at Sandia National Laboratories. Each module includes one switch, which serves as the last command-fired switch of the module, and hence is used to determine the time at which each module electrically closes relative to the other modules. The switch is $81-cm in length, $45-cm in diameter, and is immersed in mineral oil. The outer switch envelope consists of six corrugated monomer-cast acrylic insulators and five contoured stainless-steel rings. The trigger electrodes are fabricated from copperinfused tungsten. The switch is pressurized with several atmospheres of sulfur hexafluoride (SF 6 ), which is turbulently purged within 2 seconds after every shot. Each switch is powered from a 6-MV, 0.78-MJ Marx generator which pulse charges a 24-nF intermediate-store water capacitor in 1:4-s. Closure of the switch allows power to flow into pulse-forming transmission lines. The power pulse is subsequently compressed by water switches, which results in a total accelerator output power in excess of 70-TW. A previous version of the LTGS performed exceptionally at a 5.4-MV, 0.7-MA level on an engineering test module used for switch development. It exhibited a 1-jitter of $5 ns, a prefire and flashover rate less than 0.1%, and a lifetime in excess of 150 shots. When installed on the Z accelerator, however, the switch exhibited a prefire probability of $3%, a flashover probability of $7%, and a 15-ns jitter. The difference in performance is attributed to several factors such as higher total charge transfer, exposure to more debris, and more stressful dynamic mechanical loading upon machine discharge. Under these conditions, the replacement lifetime was less than ten shots. Since refurbishment of Z in October 2007, there have been three LTGS design iterations to improve the performance at 6.1-MV. The most recent design exhibits a prefire rate of less than 0.1%, a flashover rate of $0:2%, a single switch jitter of $6-ns, and a lifetime of greater than 75 shots. Modifications to achieve the performance improvement are detailed in this article.

Laser triggered, megavolt, megampere gas switches are frequently utilized to synchronize multiple pulsed power driver modules for inertial-confinement fusion, isentropic compression, and radiation physics experiments. The device developed to synchronize the 36 modules of the refurbished Z accelerator at Sandia National Laboratories is a 5.4 MV, 700 kA, sulfur-hexafluoride (SF 6 ) filled, laser triggered gas switch. At this operating level, switch jitter is 5 ns, the prefire rate is less than 0.1%, the average optic lifetime is greater than 200 shots, and the flashover rate is less than 1%. Over 1000 shots on a single-module test facility were conducted while iterating several potential design improvements, including utilizing low-erosion electrode material, varying SF 6 pressure, and modifying internal switch geometry all while keeping the basic switch architecture and footprint constant. Results of this development effort are presented herein.

We have developed a statistical model for the flashover of a 45 vacuum-insulator interface (such as would be found in an accelerator) subject to a pulsed electric field. The model assumes that the initiation of a flashover plasma is a stochastic process, that the characteristic statistical component of the flashover delay time is much greater than the plasma formative time, and that the average rate at which flashovers occur is a power-law function of the instantaneous value of the electric field. Under these conditions, we find that the flashover probability is given by 1 ÿ expÿE p t eff C=k , where E p is the peak value in time of the spatially averaged electric field Et, t eff R Et=E p dt is the effective pulse width, C is the insulator circumference, k / exp=d, and and are constants. We define Et as Vt=d, where Vt is the voltage across the insulator and d is the insulator thickness. Since the model assumes that flashovers occur at random azimuthal locations along the insulator, it does not apply to systems that have a significant defect, i.e., a location contaminated with debris or compromised by an imperfection at which flashovers repeatedly take place, and which prevents a random spatial distribution. The model is consistent with flashover measurements to within 7% for pulse widths between 0.5 ns and 10 s, and to within a factor of 2 between 0.5 ns and 90 s (a span of over 11 orders of magnitude). For these measurements, E p ranges from 64 to 651 kV=cm, d from 0.50 to 4.32 cm, and C from 4.96 to 95.74 cm. The model is significantly more accurate, and is valid over a wider range of parameters, than the J. C. Martin flashover relation that has been in use since 1971 [J. C. Martin on Pulsed Power, edited by T. H. Martin, A. H. Guenther, and M. Kristiansen (Plenum, New York, 1996)]. We have generalized the statistical model to estimate the total-flashover probability of an insulator stack (i.e., an assembly of insulator-electrode systems connected in series). The expression obtained is consistent with the measured flashover performance of a stack of five 5.72-cm-thick, 1003-cm-circumference insulators operated at 100 and 158 kV=cm. The expression predicts that the total-flashover probability is a strong function of the ratio E p =k, and that under certain conditions, the performance improves as the capacitance between the stack grading rings is increased. In addition, the expression suggests that given a fixed stack height, there exists an optimum number of insulator rings that maximizes the voltage at which the stack can be operated. The results presented can be applied to any system (or any set of systems connected in series) subject to random failures, when the characteristic statistical delay time of a failure is much greater than its formative time.