(77e) Modeling Cohesive Gas-Particle Flows with a Parcel-Based Lagrangian Approach

Lagrangian approaches are often advantageous
to model processes in chemical engineering applications since particle-related
phenomena (e.g., the deposition and evaporation of liquids, liquid absorption
in pores and asperities, or chemical reactions) can be studied in a straight
forward fashion. Parcels-based approaches form a sub-group of such approaches, and
track packages of particles instead of the full particle population. Consequently,
parcel-based approaches can be seen as a conceptually straight forward
extension of approaches that track each individual particle. However, it is
clear that as parcel size (i.e., the set of particles that represents a parcel)
increases, parcel-based approaches will lose accuracy, especially for cohesive
powders.

Increasing parcel size can be
seen as “particle coarsening”, and there have been previous attempts [1,2] to probe
the effect of such a coarsening strategy on the fluid-particle drag force. Still,
there is no systematic investigation on how parcel interactions should be
scaled with respect to parcel size to yield an identical rheological behavior of
a gas-particle mixture. This is especially true in case liquid bridge-induced
cohesion is considered. Our present work focusses on closing this gap by considering
statistics of (i) particle velocities, (ii) stresses, and (iii) local void
fractions in gas-particle flows that contain wet particles. Specifically, we present
results of CFD-DEM simulations that track all involved particles, and that are
then filtered (i.e., spatially averaged) considering various (particle) filter
sizes (see Figure for an illustration). These statistics are then compared with
results of parcel-based simulations that utilize different approaches to scale
parcel interactions. These parcel-based simulations were performed with the
original fluid grid resolution (i.e., that of our CFD-DEM simulations) in order
to avoid effects due to under-resolved gas motion.

Our results allow us to draw
conclusions on (i) the maximum permissible parcel size, the (ii) optimal particle-to-fluid
mapping strategy, as well as (iii) the optimal strategy to scale cohesive
interactions when performing parcel-based simulations of fluidization. We conclude
our study with a comparison of our findings with that of a recent study that
investigated the rheology of fluidized particles that exhibit van der Waals
interactions [3].

Figure: Ilustration of the
simulation domain, as well as the particle-based filtering strategy to obtain
parcel statistics (Bo is the Bond number that quantifies the degree of
cohesion).

[1] A. Ozel, J. Kolehmainen, S.
Radl, S. Sundaresan, Fluid and particle coarsening of drag force for
discrete-parcel approach, Chemical engineering science 155:258-267, 2016.

[2] A. Ozel, Y. Gu, C. C.
Milioli, J. Kolehmainen, S. Sundaresan, Towards filtered drag force model for
non-cohesive and cohesive particle-gas flows, Physics of Fluids 29:103308, 2017.

[3] Y. Gu, A. Ozel, J.
Kolehmainen, S. Sundaresan, Computationally generated constitutive models for
particle phase rheology in gas-fluidized suspensions, Journal of Fluid
Mechanics 860, 318-349, 2019.