This paper reviews the last twenty years of work concerning surface flashover of insulators in vacuum. It complements an earlier review paper [H.C. Miller, Flashover of Insulators in Vacuum, IEEE Trans. Electr .Insul., Vol. 28, pp. 512-527, 1993]. The surface flashover voltage of insulators in vacuum depends upon many parameters, such as material, geometry, surface finish, and attachments to electrodes, but also on the applied voltage waveform, duration, single pulse or repetitive, and on the process history of the insulator, operating environment, and previous applications of voltage. Suggestions are made for choosing the material, geometry, and processing when selecting an insulator for a particular application. Some specific techniques for improving the holdoff voltage of insulators are recommended.

Flashover of insulators in vacuum: the last twenty years

Research on dynamics of streamer propagation along different dielectric materials under a uniform electric field is presented. In the three-electrode arrangement, the variation of the streamer propagation probability, velocity and light intensity with electric field in the air and along the different dielectric materials was measured through three photomultipliers. The photographs of the streamer propagation in the air and along the different dielectric materials were taken with an ultraviolet camera. The test results have shown that the electric field required for streamer propagation in air alone is less than along the insulation surface. Furthermore, for electric fields higher than streamer ‘stability’ propagation fields, the streamer propagates along insulation surface with a ‘fast’ and a ‘slow’ component. However, the streamer propagates in air alone without a ‘fast’ component, only with a ‘slow’ component. The velocity of the streamer propagating in air alone is significantly less than the velocity of the ‘fast’ component, and is higher than the velocity of the ‘slow’ component. The streamer ‘stability’ propagation field and velocity depend upon the nature of the dielectric material. Specifically, higher electric fields are required for streamers to propagate along a dielectric material with larger permittivity. The velocity of streamer propagation along the dielectric material is also affected by attachment of charge to the surface and photoemission of secondary electrons from the surface so much. Consequently, under the same electric field, the relationship between streamer propagation velocity and permittivity of dielectric material is not very evident, roughly inverse proportion.

Characteristics of streamer propagation along the insulation surface: influence of dielectric material

By using an electrostatic probe located close to the triple junction on the cathode surface, we have conducted real-time observations of the change in electric field due to surface charging of an insulator in vacuum. A cylindrical sample made of polymethylmethacrylate (PMMA) or Al/sub 2/O/sub 3/ has been exposed to HVDC. An axisymmetric simulation based on the secondary electron emission electron avalanche (SEEA) theory has been performed. The measured results agreed with the simulation concerning the polarity of the accumulated charge as well as the field strength. The simulation also predicts the inception of charging at a voltage well below the measured flashover voltage. These results clearly support the SEEA theory as a macroscopic mechanism of surface charging of an insulator in vacuum.

Real-time observation of surface charging on a cylindrical insulator in vacuum

This paper describes charging characteristics on a dielectric surface in vacuum by electron irradiation and field emission by a triple junction under negative and positive DC HV applications. The authors measured the 2-D distribution of electrostatic charging on a dielectric surface in situ. Experimental results revealed that the negative charge distribution caused by the electron beam had a conical shape over the whole surface. On the other hand, for a triple junction, it was an acute distribution around the triple junction. Moreover, they quantitatively investigated the difference of the 2-D charging distribution, using certain shape parameters, between the two charging processes mentioned above. In addition, they examined the time decay characteristics of the surface potential on the dielectric in vacuum.

Charging characteristics on dielectric surface by different charging processes in vacuum

Analysis, Prototyping, and Design of an Ionization Profile Monitor for the Spallation Neutron Source Accumulator Ring

Surface charge distributions of disk type aluminas held between a backside electrode (alumina holder) and a needle electrode to be excited by an impulse voltage (rise time 64 /spl mu/s, wave tail 700 /spl mu/s) were measured. The measurement of surface charge distribution on the whole surface area of alumina YSA998 and UHA99 after impulse voltage application revealed that the surface charging can be initiated either from the anode or from the cathode triple junction. The charging initiated from the anode triple junction (for positive polarity) produced positive charge at the anode region and the density is dependent on the applied voltage, while the charging initiated from the cathode triple junction region (for negative polarity) produced negative charge around the cathode region. For positive polarity, the critical values of charge density to the flashover for alumina YSA998 and UHA98 are 5.70 and 17.2 /spl mu/C/m/sup 2/, respectively.

H. Imada

Papers

Flashover from surface charge distribution on alumina insulators in vacuum