Mechanism of high-frequency noise in particle-in-cell(PIC) simulation is analysed.An effective finite-difference time-domain(FDTD) method called High_Q FDTD method is developed in which high-frequency noise is filtrated and high quality factor device is calculated with PIC simulation.Formulation of the method is derived on the basis of time-biased and centered-difference FDTD method.Stability and convergence of this method are well analysed.Filter efficiency and calculation precion of this method are compared with those of the time-biased and centered-difference FDTD method in computation of a relativistic backward wave oscillator(RBWO).

A method is presented to deal with high-dimensional scalar PDF(probability density function) transport equation for two-phase flow in phase space (X, V), in which the PDF transport equation is reduced to velocity phase space.The particle velocity PDF f^v(V) is solved analytically and the particle position PDF is obtained statistically with Monte Carlo techniques. It shows that the method is better than the particle Reynolds stress trajectory model(PRST Model) in simulating a turbulent wall jet loaded with solid particles.

An effective finite-difference time-domain (FDTD) method,timebiased FDTD method,is developed to filtrate high-frequency noise for particle-in-cell (PIC) simulation.The formula is derived using time bias and relaxation iterations on the basis of centered-difference FDTD method.Stability and convergence of this method are analysed. In the computation of a relativistic magnetron,compared with the centered-difference FDTD method,filter efficiency of this method is shown.

A Lagrangian equation of v'_giv'_pi on the particles trajectory is obtained with stochastic process theory, and then Lagrangian equation of particle Reynolds stresses is exactly set up. Wang Stock experiment on heavy particle dispersion considered scale ratio and flow decay is simulated by both this method and Zhang's model^[1]. Simulation results show the new method is better than Zhang's model. A backward-facing step two-phase flow is calculated by the new model and the results show good agreement with the experiment results.