Surface flashover is a vital problem that causes pulse transmission failure in high power facilities, such as the Z machine, PTS and Dragon linear induction accelerators(LIA). Once flashover grows across insulator surface in vacuum, the process chasing high voltage and intense current is limited as a result of insulation capability reduces. When working with rep-rate or burst mode, vacuum surface flashover happens with much higher probability. Therefore, understanding the mechanism for flashover under multi-pulse is very important for designing the next generation high power facilities.
For the purpose of studying the insulating performance bottleneck of pulsed power technology, we have designed several experiments on two frequently-used insulators, nylon and cross-linked polystyrene(XLPS) under mono-pulses and triple-pulses with different pulse intervals based on the existing Multiple Pulse Platform in Institute of Fluid Physics. Using three sets of conventional Marx-Blumlein devices, our platform generates three single rectangular pulses(i.e., pulse A, B and C) separately which meet in the confluent/blocking network and that converge into a triple-pulse with continuously adjustable intervals since the time delay before trigger is under control.
[1] A Rogowski coil is placed at the exit of the vacuum chamber to ground to measure the flashover current across insulator surface. Compared to previous experiments
[2], we take advantage of the current limiting resistance to prevent the flashover from damaging insulators and so make results statistic. A 50Ω

matching resistance of high voltage cables is also used as a divider to improve the measurement range of insulators from 180-300kV to 30-300kV. Paralleling with the insulator is a resistance voltage probe(RVP) used to measure the sample voltage which would evaporate while flashover is forming.
We have tested two intervals of 500ns and 2μs

on both nylon and XLPS cylindrical insulation materials with the thickness of 10mm. Hundreds of data is to be processed and analysed. XLPS insulators show stabler properties under triple-pulses. Nervertheless, laid aside for a period of time, mostly one night, nylon insulators could undergo a better voltage. Generally, if the preceding pulse causes flashover then subsequent pulses would definitely cause the same breakdown, and the conditioned voltage(U
co) of each three pulse differs a lot and decreases progressively, which is not noticed in former experiment results. We suppose it is connected to the build time of penetrating conducting pathway and charge transport mechanism. It is significant to studying insulating technology in pulsed power.