A MODEL FOR DC INTERRUPTION IN DIFFUSE VACUUM ARCS

A theoretical model for current interruption in a diffuse vacuum arc is described. Before current zero, the interelectrode plasma is modeled as an ion-neutral fluid through which the electrons are flowing. Immediately after current zero there is a short period during which the plasma resistance remains low while the direction of the electron flow is reversed. Upon reversal of the electron flow, a positive ion sheath grows into the plasma from the fomer anode driven by the transient recovery voltage.; The decay of the interelectrode plasma during interruption is described mathematically by the one-dimensional unsteady compressible flow equations, which are solved numerically using a modified method of characteristics, whereby the characteristic directions are calculated at points on a predetermined grid.; The growth of the positive ion sheath into the plasma after current zero is calculated using the unsteady sheath equations, enabling the post-arc current, the electric field at the former anode, and the power input to that same electrode to be evaluated using a fourth order Runge Kutta technique.; The post-arc data generated with the interruption model can be used to study reignition mechanisms in vacuum arcs. It is postulated that the electric field at the former anode and the power input to that electrode are critical factors in the establishment of a cathode spot which can lead to reignition during the first microseconds after current zero.; The theoretical model has been applied to practical interrupters with copper and beryllium electrodes. In order to check the theoretical analysis a DC commutation circuit has been constructed to test vacuum interrupters under severe conditions. The initial experimental post-arc data for interrupters with copper electrodes is in good agreement with the theoretical predictions. A comparison of the conditions under which failures occur experimentally with the model results, indicates that the interrupters reignite when the electric field at the former anode is greater than 10('7) V/m and the power input to that same surface exceeds 3 x 10('7) W/m('2).; The model predictions of the post-arc current for interrupters with beryllium electrodes are somewhat lower than the data of other experimenters would indicate. The validity of the ion-neutral model, used successfully for copper vacuum arcs, is in some doubt for beryllium vacuum arcs, due to the longer mean free paths between collisions in the beryllium metal vapor. Some suggestions for improving the plasma model under these conditions are given.