For investigating the mechanism of high power microwave flashover and breakdown on dielectric surface with outgassing, firstly, the theoretical modeling is put forward, including dynamic equations, particle-in-cell (PIC) method, secondary emission, Monte-Carlo collision (MCC) method and outgassing model. Secondly, based on the theoretical modeling, the 1D3V PIC-MCC code is programmed by authors. By using this code, the flashover and breakdown on dielectric surface with weak and strong outgassing course under different gas moving velocities are studied numerically. The numerical results are concluded in the following. The flashover and breakdown on dielectric surface are caused by continuous increase of deposited power. For weak outgassing, multipacting is dominant. As outgassing coefficient increases, multipacting is promoted by ionization collision. The typical phenomena are the increases of space-charge field, average energy of surface-collision electrons and the number of surface-collision electrons. Here, the surface-collision electrons are caused by multipacting mostly. With the increase of gas molecule velocity, ionization course is suppressed by gas pressure decreasing near to the dielectric surface. For strong outgassing, ionization collision is dominant. As outgassing coefficient increases, the number of ions increases exponentially with ionization frequency increasing, multipacting is suppressed by ionization collision. The typical phenomena are the negative value of space-charge field on dielectric surface, the decrease of average energy of surface-collision electrons, and the exponential increase of surface-collision electrons caused by ionization collision near to dielectric surface. Here, the surface-collision electrons are caused by ionization mostly. With the increase of gas molecule velocity, the depth of gas is enlarged, thereby promoting the ionization collision.

The effects of compressibility on the Rayleigh-Taylor instability (RTI) are investigated. It is shown that the controversy over compressibility effects in the previous studies is due to improper comparison, in which the density varying effect obscures the real role of compressibility. After eliminating the density varying effect, it is found that the compressibility destabilizes RTI in both the cases of constant density and exponentially varying density when M 1. This destabilizing effect is more important at smaller values of the Atwood number AT or greater values of gravity g, and the increment in the growth rate produced by compressibility depends inversely on the pressure p or the ratio of specific heat Γ .

In this paper, an electromagnetic-field FDTD method coupled with plasma fluid model is put forward to investigate the different physical phenomena of high power microwave (HPM) flashover and breakdown on inner and outer surface of output-window. Based on the above theoretical models, a one-dimensional (1D) electromagnetic field and plasma interaction code is programmed by authors. By using the code, the HPM flashover and breakdown on inner and outer surface of output-window are simulated. The numerical results could be concluded as follows. For flashover and breakdown on outer surface, output microwave pulse is shortened without cut-off; there is a standing-wave distribution of electric field RMS (Root-Mean-Square) value before the window with fixed-positions of wave nodes and antinodes; there is a ultra-high-density (～1021 m-3) and ultra-thin (～mm) plasma shell with slow diffusion, microwave could penetrate the plasma-shell partly; the shortening of output microwave is caused by plasma absorption mostly. The output pulse of microwave could be lengthened by reducing the initial density or depth of plasmas; the other way is to shorten incident microwave pulse or reduce the value of incident microwave power. For flashover and breakdown on inner surface, there is also a standing-wave distribution of electric field RMS value before the window but the positions of wave nodes and antinodes vary with time; the plasma region moves toward the microwave source; with strong-outgassing, output microwave pulse is shortened without cut-off, there are “thread-like” ultra-high-density (～ 1021 m-3) and ultra-thin (～mm) plasma regions with slow diffusion, the distance between two “thread-like” regions is about a quarter of microwave wavelength, the shortening of output microwave is caused by plasma absorption mostly; with weak-outgassing and low electric field value, the output pulse of microwave is lengthened but cut-off finally, there are “belt-like” high-density (～ 1018 m-3) and thin (mm-cm) plasma regions with fast diffusion, the distance between two “belt-like” region is about a quarter of microwave wavelength, the shortening of output microwave is caused by plasma absorption mostly; with weak-outgassing and high electric field value, output pulse of microwave is cut-off quickly, “block-like” diffuse ultra-high-density (～1021 m-3) and deep (～ cm) plasma regions are formed with very fast diffusion, and the shortening of output microwave is caused by plasma reflection mostly.

The veracity of a low temperature plasma model is limited by the accuracy of the electron transport coefficient, which can be solved by simulating the electron transport process. When simulating the transport properties of electrons, there are a variety of approaches to dealing with the scattering of electrons and energy partition between the primary-electrons and secondary-electrons after electron-neutral particles’ collision. In this paper used is a model based on the Monte Carlo collision method to investigate the influence of scattering method and energy partition method on the electron transport coefficient. The electron energy distribution function, electron mean energy, flux mobility and diffusion coefficients, as well as the Townsend ionization coefficients are calculated in the hydrogen atom gas under a reduced electric field from 10 to 1000 Td. The calculation results show that the influence of the isotropic scattering assumption on the electron transport coefficients increases with reduced electric field increasing. However, even under a relatively low reduced electric field (10 Td), the calculated mean energy, flux mobility, and flux diffusion coefficient of electrons under the assumption of anisotropic scattering are 39.68%, 17.38% and 119.18% higher than those under the assumption of the isotropic scattering. The different energy partition methods have a significant influence on the electron transport coefficient under a medium-to-high reduced electric field (> 200 Td). Under a high electric field, the mean energy, flux mobility and flux diffusion coefficient calculated by the equal-partition method (the primary and secondary electrons equally share the available energy) are all less than the values from the zero-partition method (the energy of secondary-electrons is assigned to zero). While the change of Townsend ionization coefficient with reduced electric fields shows a different trend. The electron transport coefficient obtained by the Opal method lies between the values from the equal-partition method and the zero-partition method. In addition, considering the anisotropic scattering, the influence of energy partition method on the transport coefficient is higher than that under the assumption of isotropic scattering. This study shows the necessity of considering the anisotropic electron scattering for calculating the electron transport coefficient, and special attention should be paid to the choice of energy partition method under a high reduced electric field.

By comparing two diffierent torch geometries, it was found that the shielding flow has no significant effect on plasma velocity and temperature,except in the shock wave region. The shielding flow decreases the shock wave, and increases the arc voltage due to cooling. In the impinging geometry, shielding flow will crash the plasma jet after the nozzle exit and slightly increase the pressure in the torch. It was also shown that the component of shielding gas has no significant effect on plasma cuttingarc. The mole fraction of oxygen decreases very slowly along the axis and is still more than 90% at 10 mm downstream the nozzle exit.