Mesoscopic Numerical Method For Particle Capturing And Applications

Currently atmospheric pollution,especially particulate pollution has seriously affected the society.Particularly for the main source such as plant boilers,industrial boiler and vehicle exhaust emissions,the traditional collecter hardly meet the increasingly stringent emission targets.To improve the performance of the collecters,a better fundamental investigation on the mechanism is urgent.Due to the comolexity of the mutiscale and mutiphase fluid-particle intereaction problem involved,numerical simulations can be employed to reveal the detail of the interaction between the fluid and the particles.So In the Article,We study the mesoscopic numerical methods for particle capturing.Firstly,we use the lattice Boltzmann method with Lagrange tracking method to study the Collecter Chasing Mechanism,Krick thinks the movement in flow direction can't affect the efficiency,while our result indicates that it can improve the efficiency.And for the MPPS(the Most Penetrating Particle Size)particle,the improment is even larger.Also the hit angle and initial position of the collected particle is analysed.Secondly,by comparing the existing immersed boundary method(IBM),the IBM based on the discrete unified gas kinetic sheme(DUGKS)was put forward.We introduced the Strang-splitting method to implement external force term.the IBM-DUGKS method was tested with curve boundary and moving boundary case,the performance reaches the level of other similar approaches.In addition,the key problem about volume fraction calculation of immersed moving boundary method(IMBM)was studied,a method of region decomposition based on dichotomy and a look-up table method were proposed.Compared with the traditional region decomposition method.Both can maintain the calculation accuracy and greatly improve the calculation efficiency at the same time,the look-up table method even reach hundred or thousand times to accelerate.This method improves the the practicability of the IMBM very much.Finally,critical problems in the discrete element method(DEM)such as time step selection,bidirectional coupling algorithm,calculation framework,collision detection algorithm and parallelization were studyed.We compared different bidirectional coupling algorithms,espacilly the instability of them.The coupling calculation between LBM and DEM was realized,then particle sedimention problem with the Vogit model,single fiber efficiency verification with the JKR model was test.