Agglomeration study in the inlet section of a large scale spray dryer using stochastic Euler-Lagrange modelling

Spray drying is an essential unit operation for the production of powders from liquid slurry. Main events occurring inside the spray dryer are droplet drying and interactions between droplets and fine particles, leading to bouncing collisions, coalescence, agglomeration, and satellite droplet formation. In recent years, there has been growing interest to use computational fluid dynamics (CFD) for exploring such phenomena inside spray dryer systems. Researchers have extensively investigated single droplet-droplet and single particle-particle interactions using numerical and experimental methods. However, the literature on droplet-particle interactions with a quantitative description of agglomeration in spray drying is scarce, and mainly qualitative. In this paper, we explore the development of an Euler-Lagrange model with a stochastic approach for the prediction of collision, coalescence and agglomeration of partially wet particles in a spray dryer. In this approach, the dynamics of the gas phase is solved by an Eulerian equation as a continuum and the dynamics of the solid phase is solved by a Lagrangian equation as a dispersed phase, with conventional gas-solid coupling. Inside the spray chamber, the turbulent gas flow has an effect on the particle interactions. In a spray dryer, the number density of droplets is usually more than 1011 per cubic meter, effectively ruling out a deterministic approach in which each particle is individually tracked. We introduce a stochastic Direct Simulation Monte Carlo (DSMC) approach in which each particle searches randomly, in a local and spherical searching scope, for another particle to collide with during a particle time step. In a spray dryer, different outcomes can occur when a pair of particles collide. We use elementary models for collision, coalescence, break-up, drying and agglomeration, validated by experimental results from the literature and industrial data. In this paper we present details of the modelling approach using sub-models and preliminary simulation results. The ultimate aim of this project is to develop a simulation tool that can provide the particle size, velocity and flux distribution for a section of a large-scale spray dryer. These results ought to be used as boundary conditions for even coarser simulations, which will be used to design more efficient spray dryers that can produce higher throughputs.