NUMERICAL STUDY OF SOOT FORMATION AND OXIDATION PROCESSES

Abstract

This thesis presents a new numerical approach for the simulation of soot formation and oxidation processes. A moment projection method (MPM) is developed. This method is simple, robust and easy to implement. More importantly, it can be used to handle the particle shrinkage and fragmentation problems with high accuracy. The MPM has been implemented into a Stochastic Reactor Model engine code to simulate the soot formation processes inside diesel engines. A modified Maximum Entropy method is adopted to generate the detailed particle size distribution. With this new approach, a comprehensive study on the soot particle dynamics is achieved. In addition, a new fragmentation kernel is proposed which is more accurate in predicting the soot fragmentation process than the previously accepted kernel. Finally, the MPM has been expanded into a bivariate method which has been shown to be able to effectively handle the two-dimensional population balance equations.