Download files
Abstract
The majority of industrial particulate operations are highly energy intensive, which
leads to expensive operational costs. The fundamental particulate behavior mechanisms,
which determine the system collective dynamical behavior, is not fully
understood. Hence, it is difficult to optimize and control particulate processes.
This thesis aims to develop a systematic approach to modeling and analyzing
the overall/collective dynamical features of particulate systems. The dynamics
of particulate systems are modeled based on a stochastic approach in the form
of Markov chains. The models can be developed using particle behavior data
obtained from either experimental or numerical based approaches. A numerical
approach, in particular Discrete Element Method, is used in this work.
The collective dynamics of particle movement influences the effectiveness of
particulate operations. The Markov chains approach is used to model the collective
movement of monodisperse particulate systems under constant operating
conditions. The key operator represents the probability of particle movement from
one location to another, which can estimate particle trajectory. In addition, an
approach to analyzing the collective dynamics of particle movement is also developed,
in particular the oscillatory behavior and spatial distribution of particle
movements. The proposed model is then extended for systems with time-varying
operating conditions. This provides a way to optimize and control the system behavior
by manipulating the operating conditions. The Markov chains models for polydisperse particulate systems under both constant and time-varying operating
conditions are also developed. The model performs a parallel analysis of each type
of particle. This opens a pathway to monitor and analyze polydisperse particulate
systems. Additionally, the model has the potential to aid the implementation of
process control of polydisperse systems.
The development of a Markov chains model for a non-spatial distribution analysis
is also introduced. The operator represents the probability of non-locational
movements of a particle property between or within arbitrary intervals. This can
be used to model the collective dynamics of particle energy distributions. Additionally,
a measure to relate particle impact energy (which is unmeasurable during
operation) to kinetic energy (which can be estimated during operation) is proposed.
This provides a foundation for the development of an indirect impact
energy sensor which is useful for real-time monitoring and process control.