Simulation of particle agglomeration using dissipative particle dynamics
Attachment of particles to one another due to action of certain inter-particle forces is called as particle agglomeration. It has applications ranging from efficient capture of ultra-fine particles generated in coal-burning boilers to effective discharge of aerosol sprays. Aerosol sprays have their application in asthma relievers, coatings, cleaning agents, air fresheners, personal care products and insecticides. There are several factors that cause particle agglomeration and based on the application, agglomeration or de-agglomeration is desired. These various factors associated with agglomeration include van derWaals forces, capillary forces, electrostatic double-layer forces, effects of turbulence, gravity and brownian motion. It is therefore essential to understand the underlying agglomeration mechanisms involved. It is difficult to perform experiments to quantify certain effects of the inter-particle forces and hence we turn to numerical simulations as an alternative. Simulations can be performed using the various numerical simulation techniques such as molecular dynamics, discrete element method, dissipative particle dynamics or other probabilistic simulation techniques. The main objective of this thesis is to study the geometric characteristics of particle agglomerates using dissipative particle dynamics. In this thesis, agglomeration is simulated using the features of dissipative particle dynamics as the simulation technique. Forces of attraction from the literature are used to modify the form of the conservative force. Agglomeration is simulated and the characteristics of the result ing agglomerates are quantified. Simulations were performed on a sizeable number of particles and we observe agglomeration behavior. A study of the agglomerates resulting from the different types of attractive forces is performed to characterize them methodically. Also as a part of this thesis, a novel, dynamic particle simulation technique was developed by interfacing MATLAB and our computational C program.