(751a) Investigation of Particle Cluster Morphology in a Downflow Reactive System Via Large Eddy Simulations | AIChE

(751a) Investigation of Particle Cluster Morphology in a Downflow Reactive System Via Large Eddy Simulations


Gómez, N. A. - Presenter, Universidad Nacional de Colombia - Sede Medellín, Facultad de Minas, Bioprocesos y Flujos Reactivos
Vandewalle, L., Ghent University, Laboratory For Chemical Technolo
Reyniers, P. A., Ghent University
Van Cauwenberge, D. J., Ghent University
Molina, A., Universidad Nacional de Colombia – Sede Medellín, Facultad de Minas, Bioprocesos y Flujos reactivos
Fox, R., Iowa State University
Marin, G. B., Ghent University
Euler-Lagrange simulations were conducted in the open-source finite volume code software package OpenFoam to study the cluster formations in downers, and how these cluster structures are affected by the change of the physical properties such as density and viscosity of the fluid phase of the catalytic cracking (FCC) process.

Particle clusters in pneumatic conveying processes, such as FCC, typically do not allow an ideal contact between phases in the reactor and hence induce heat and mass transfer limitations, and as such, may reduce the efficiency of the reactions (low-conversion zones). Cluster formation has been the focus of several studies in fast fluidization processes. However the bulk of these studies were carried out for gas-solid reactors with for upward flowing fluid, so-called risers. In reactors with downward flowing fluid, so-called downers, some characteristics of clusters are still to be described, such as size, shape, and spatial distribution under FCC’s reactive conditions. The positive effect of gravity in the stream-wise direction is the main reason for the different behavior between risers and downers because it prevents particle back-mixing. One way to study these phenomena is to use computational fluid dynamics (CFD) under an Euler-Lagrange approach, which allows to assess gas and particle motion independently, the local interactions between the two phases, and collisions between particles.

A one-meter-long cylindrical downer was used as a canonical case to analyze the dynamics of particles using CFD. The simulation included 4.8 x 106 particles as inventory in steady state, which corresponds to a mean particle volume fraction of 0.01 over the simulated volume, a characteristic value of fast fluidization processes. The Smagorinsky LES model was implemented to calculate the subgrid scale turbulent values over the 7 million cells in which the volume was divided. The LES model allows to take into account the main fluctuations in the velocity fields of the gas phase, and the influence of these fluctuations on the particle motion. A four-lump mechanism was selected to describe the chemical behavior of the process. This mechanism includes gasoil, gasoline, light gases and coke as lumped species of the process.

The identification of clusters in the downer reactor was carried out using the instantaneous position of the particles in the domain and the inter-particle distance. The criterion to select the particles that are part of a cluster is based on the particle volume fraction: regions of the domain where the particle volume fraction is 3% or higher are considered to be clusters. Knowledge of the positions of the particles in a cluster was used to determine the shape, size and distribution of the clusters.

The detected clusters tend to be shaped like strands or sticks in both reactive and non-reactive simulations; particles naturally align in elongated structures to minimize the drag force and hence the shear stresses. In fully developed flow, the probability of finding clusters composed of 3 particles is about 40%. However, it is also possible to find clusters with over 20 particles but the probability for this is below 5%. The larger clusters are found near the walls and mainly towards the bottom of the reactor. The shear forces are reduced at the bottom of the reactor when the rate of reaction and, hence, the acceleration accompanying the expansion is low.

Additionally, the reactions have an influence on properties such as gas density and viscosity, which influence the particle’s response time. At the entrance where the gas is composed of gasoil only, the particle response time is around 0.01 s; however, once the reactions occur, the gas density and the viscosity decrease, which consequently increases the particle response time up to 0.05 s. Towards the bottom of the reactor, where the conversion is high (40% -typical value in lab scale reactors-), particles respond slower to external disturbances from the gas phase and, therefore, larger and more stable clusters can be formed. This effect reduces the efficiency in the contact time between phases, reducing, as well, the efficiency of the reactor.

This study helps to gain insight in the behavior of particles in downers, and also, the influence of clusters in the chemical performance of the reactor, particularly, in the fluid catalytic cracking process.


This paper has an Extended Abstract file available; you must purchase the conference proceedings to access it.


Do you already own this?



AIChE Pro Members $150.00
AIChE Graduate Student Members Free
AIChE Undergraduate Student Members Free
AIChE Explorer Members $225.00
Non-Members $225.00