PARTICLE AGGLOMERATION IN FLOW MODELLED WITH MOLECULAR DYNAMICS COUPLED TO A THERMAL LATTICE BOLTZMANN CODE

Abstract

Particle agglomeration can arise naturally (e.g. dust, salt) or as a result of industrial activities and/or combustion processes (e.g. spray drying, particle flame synthesis). The process itself and its mechanisms are important for many applications since the physical properties of the final structures are mainly determined by the composition, number, diameter and geometric arrangement of their constituent primary particles. Thus, knowing and controlling the extent of agglomeration meets a growing interest in environmental and industrial concerns.

The objective of the paper is to develop a simulation model of particles suspended in a flowing fluid using MD simulations coupled to a Lattice Boltzmann (LB) solver. These simulations allowed determining the agglomerate transport and deposition rates depending on the flow conditions and agglomerate structure and understanding the relationship between agglomerate characteristics (i.e. growth kinetics and morphology) and their behavior in a flow field.

Two systems of 2000 and 1000 particles were simulated at 300 K and 600 K both of them in a known fluid. Simulations using a Langevin thermostat were also performed to compare with the LB thermostat. This allowed quantifying the influence of the fluid flow on the agglomeration process and agglomerate properties. In further applications, this will help to a priori tailor the flow conditions to achieve a desired aggregate morphology.

As a result, reasonable aggregate morphologies were achieved. One of the main conclusions is that taking into account the fluid flow (LB solver) the agglomeration process of the particles is notably accelerated in comparison to the Langevin simulations. One of the main implications of this work could be the possibility of using a known fluid to accelerate an agglomeration process given a suitable fluid and to find a desirable configuration of agglomerates. Another conclusion is that the agglomeration process is sensitive to the temperature variation and that the number of particles in the system influences the final configuration of agglomerates in LB simulations.