Early work at Boeing (4) showed that thin films of insulation probably would not be effective in reducing plasma current leakage, due to intense flow of currents through even a small number of pinholes in the insulation (Fig. 5). Examination of the sheath model of plasma interaction in Fig. 6 shows that, above a threshold voltage where sheath dimensions equal or exceed the spacing between exposed conductors (bare interconnects or pinholes), very little reduction of total current collected should be expected from insulation of even most of the panel surface. This is probably related to the ''snap-over11 phenomena reported by NASA-LeRC (5). Tests were done at JSC using a IXIOm stainless steel panel, first operated at voltages to -3,000V with no insulation, then operated in identical (105/cc) plasma densities with >90% of the total surface area insulated by application of rnylar tape. Results are shown in Fig 7. Not only was the insulation not effective in reducing current leakage leakage at voltages over 100V, but it also caused increased currents and transient "arcing" to the plasma that prevented measurement of currents for voltages in excess of 200V. Such transient increases in current above the equilibrium space charge limiting values have frequently been observed. These "arcs” have been observed as bright flash points near solar cell interconnects at LeRC and from most dielectric surfaces within the high voltage plasma sheath volume surrounding the 10 meter panels tested at JSC. Current densities greatly in excess of even bipolar space charge limited values are observed. The arcing from large panels has been reduced by making exposed surfaces conductive, while adding large areas of insulation was observed to reduce the on-set voltage for arcing from -3,000V to -250V for the otherwise unaltered panel used in Fig. 7 test. The "arcing" mechanism is not understood at this time. It is clearly of importance to determine reliable criteria to avoid this phenomena on operational space systems. The needed solution may well come from existing plasma and materials investigations directed toward the GEO spacecraft charging problem. Although LEO plasma densities eliminate charging for most passive spacecraft surfaces, in the case of high voltage sheaths exclusion of the repelled species and accelleration of the attracted current carriers results in a local environment within the sheath similar to GEO during an intense storm. These sheaths may occur around known high voltage surfaces such as solar arrays, or even passive surfaces of large structures which acquire magnetically induced voltages due to orbital velocity. An interesting point is implied regarding the high voltage (Klystron, etc) vs. low voltage (solidstate microwave transmitter) options for SPS. Although avoidance of the "arcing" problem may appear to be a point in favor of selecting the low voltage option, just the opposite could be true. Physical damage to the "arcing" surface is very rare. The major design problems posed would seem to be increased average power loss and induced electrical transients. The low voltage devices may be much more susceptable to these transients than high voltage tubes, etc. Since arcing may occur due to other causes (induced voltages) than actual high operating voltage of the solar arrays, a high voltage system could well be less vulnerable to arcing problems. Development of adequate computational tools (similar to the NASCAP program now available for GEO spacecraft charging effects) for use in design calculations is needed in order to proceed with reasonable confidence in the design of higher voltage power systems for operation in LEO. Criteria are also needed to define ground and flight test requirements to validate the proposed design cal-
RkJQdWJsaXNoZXIy MTU5NjU0Mg==