The development of fundamental understanding through modelling and simulation of important phenomena related to turbulent multiphase flows, such as particle deposition and agglomeration, is of interest to a wide range of industrial and natural processes. These include chemical engineering, mineral processing, agriculture, oil and gas, and nuclear waste management among many others. In the present study, particle-laden turbulent pipe flows are studied using direct numerical simulation of the fluid phase in combination with Lagrangian particle tracking, with a particular focus on predicting and elucidating the dynamics of particle-particle interactions. The models used, which have been developed and validated in the present study, enhance our understanding of these flows, particularly surrounding the processes which lead to particle collisions and agglomeration, as well as deposition and bed-formation. An energy-balance based agglomeration determination technique is used along with four-way coupling to predict particle aggregation due to collision interactions within the flow. The impact of behavioural modification effects is investigated by varying influential parameters such as the reduced surface potential, temperature, inverse Debye length, Hamaker constant, coefficient of restitution, and Reynolds number. It is discovered that the electric double layer repulsion dynamics exerts little effect on collision and agglomeration behaviour, attributed to the large particle diameters examined in comparison to the effective range of those forces. However, this study demonstrated that the restitution coefficient has a significant influence on the behaviour of particle-particle agglomeration, with a decrease in the coefficient resulting in higher aggregation rates. The Hamaker constant and Reynolds number variations both lead to major impacts on particle-particle interaction. It is determined that collisions and agglomeration events occur more frequently for increased Hamaker constants and Reynolds numbers.
Finally, a novel empirical correlation has been developed to predict the critical deposition velocity, at which the onset of particle deposition in particle-laden turbulent pipe occurs. The results indicate that at high Stokes numbers, the particle dispersion function and mean vertical displacement values fall quickly, and the formation of beds is observed. However, at lower Stokes numbers thinner dune-like structures are formed on the lower pipe wall. Using a volume-coverage percentage in the near-wall region of the lower half of the pipe as an identifier to define of the onset of deposition provides strong agreement with experimental predictions as well as the novel empirical correlation for prediction the critical deposition velocity.
