(493e) Coupled CFD and DEM Simulation of Fluid-Granular Systems

Coupled CFD and DEM Simulation of
Fluid-Granular Systems

Dalibor
Jajcevic¹, Charles Radeke¹, Georg Scharrer¹, Johannes
G. Khinast²

1: Research Center Pharmaceutical
Engineering, Graz, Austria

2: Graz University of Technology,
Institute for Process and Particles Technology, Austria

Granular flows are extremely important
for the pharmaceutical and chemical industry. The understanding of the impact
of particle size and related effects on the mean, as well as on the fluctuating
flow field, in granular flows is critical for design and optimization of powder
processing operations. Computational Fluid Dynamics (CFD) has become essential
to the design and development of many engineering devices. Although initially
limited to specialized engineering fields, in the last decades it became a
commonly-used tool for analyzing many complex technical problems that involves
fluid flows, heat transfer, chemical reactions etc. In the last few years
Discrete Element Model (DEM) simulations have been increasingly used to study and
analyze flows of granular systems with the main emphasis on granular flows
where the interaction between particles and a fluid phase may be neglected.
Instead only particle-particle interactions are considered. In pharmaceutical
industry, fine granular materials (usually smaller than 3 mm) are generally used,
but large enough that the gas drag overcomes the gravity. The particles are
moved and mixed leading to a non-uniform and unsteady flow situation, which is
beneficial for processes such as coating, granulation, draying etc. Thus, a
deep understanding of the fluid-particle interaction is important to overcome
severe difficulties in the design and scale-up of the industrial devices.

Applying above mentioned simulation
techniques a simulation of the fluid-granular systems in a multiphase setup is possible.
In the modeling, the drag force in only accelerating force acting on a particle
and thus plays an important role in the coupling between gas and solid phases,
see Van der Hoef et al. [1] and Wei Du et al. [2]. In literature, several
widely used models are available, such as Gidaspow et al. [3], Syamlal and
O'Brien [4], Di Felice [5] etc. Authors, such as Wei Du et al.
[2], reviewed the models and showed that the Gidaspow
model gives the best agreement with the experimental observation both
qualitatively and quantitatively. Gidaspow et al. [3] derivate the model
combining Ergun [7] equation for dense regimes and a correlation proposed by
Wen and Yu [8] for the more dilute regimes. In the last few years common used
model is a model proposed by Beetstra et al. [6]. Using
a similar approach as Hill et al. [9], Beetstra et al. [6] derived the model
based on a wide range of data for Reynolds numbers up to 1000 showing proper
limiting behaviour and is therefore believed to be more practical, see Deen et
al. [10].

In the first steep of the presented work
the model proposed by Gidaspow et al. [3] and the relative new model derivate
by Beetstra et al. [6] are investigated. The models are implemented via user
function in the commercial CFD code AVL-Fire in an Eulerian two-phase setup (a
gas and a solid phase). The user function allows defining the momentum
interfacial exchange between the phases, whereas the solid phase is treated as
non-moved, therefore a coupling between CFD and DEM code in this steep was not
required. Taking into account Ergun [7] equation the pressure drop is
calculated and compared with results of the CFD simulation for the Reynolds
numbers up to 2000, four particles diameters and three different void fractions.
The results show that both models can successfully predict the pressure drop
through a porous media composed of monodisperse particles. Nevertheless, the
model introduced by Beetstra et al. [6] shows increasing pressure drop
behaviour with a decreasing particle diameter, but still in an acceptable range
and is further used in this work.

In the second step, a coupled simulation
between in-house developed DEM code and the commercial CFD code AVL-Fire was
realized. In order to be able to simulate real-life granular systems, a new
technology was implemented and a high-performance DEM code was developed based
on the newly available ?Compute United Device Architecture? (CUDA) technology,
see Radeke et al [11]. The DEM code can simulate millions of particles on
readily available hardware. The simulations can be directly visualized by a GUI
using OpenGL display features. Based on the computational efficiency of our
method the development of large-scale simulations is highly convenient. The
data exchange between the codes was realized applying AVL Code Coupling
Interface (ACCI). ACCI is a software component to perform a co-simulation with
an arbitrary number of instances of different simulation programs. The codes
concurrently simulate the same time interval and pass information to each other
continuously. The results obtained in the simulation were verified using
experimental data of Link et al. [12]. The measurement is obtained in an experimental
study of the flow in a pseudo-2D spout-fluid bed taking into account different
operating conditions. A good agreement with the experimental data confirms the
accurateness of used coupling methodology.

1.
References

1.      
Van der Hoef  M. A., Ye M., van Sint Annaland
M., Andrews A. T., Sundaresan S., Kuipers J. A. M., ?Multi-Scale Modeling of
Gas-Fluidized Beds?, In: Advances in Chemical Engineering, vol. 31, pp. 65

2.      
Wei Du, Xiaojun Bao, Jian Xu, Weisheng Wei ?Computational
fluid Dynamics (CFD) modeling of spouted bed: Assessment of drag coefficient
correlations?, Chemical Engineering Science 61, pp. 1401-1420, Elsevier 2005

3.      
Gidaspow D, ?Multiphase Flow and Fluidization?, Continuum
and Kinetic Theory Description, 1994

4.      
Syamlal M. and O'Brien, T.J., ?Simulation of
granular layer inversion in liquid fluidized beds?, International Journal of
Multiphase Flow 14, pp. 473-481, 1988

5.      
Di Felice, R., ?The voidage functions for
fluid-particle interaction system?, International Journal of Multiphase Flow
20, pp. 153-159, 1994

6.      
Beetstra R., Van der Hoef M. A., Kuipers J. A.
M., ?Numerical study of segregation using a new drag force correlation for
polydisperse systems derived from Lattice-Boltzmann simulations?, Chem. Eng.
Sci. 62, pp. 246-255, 2007

7.      
Ergun S., ?Fluid flow through packed columns?,
Chemical Engineering

Progress 48, pp. 89?94, 1952

8.      
Wen Y.C. and Yu Y.H., ?Mechanics of
Fluidization?, Chemical Engineering

Progress Symposium Series 62, pp. 100?111, 1966

9.      
Hill R.J., Koch D.L., Ladd J.C., ?
Moderate-Reynolds-numbers flows in ordered and random arrays of spheres?,
Journal of Fluid Mechanics 448, pp. 243?278, 2001

10.   Deen N.G., Van Sint Annaland M., Van der Hoef  M. A., Kuipers
J.A.M., ?Reviewof discrete particle modeling of
fluidized beds?, Chemical Engineering Science 62, pp. 28-44, 2007

11.   Radeke C, Glasser B., Khinast J., ?Large-scale Mixer Simulations
Using Massively Parallel GPU Architectures?, Chemical Engineering Science,
Volume 65, Issue 24, pp. 6435?6442, 2010

12.   Link J., Zeilstra C., Deen N., Kuipers H., ?Validation of a Discrete
Particle Model in a 2D Spout-Fluid Bed Using Non-Intrusive Optical Measuring
Techniques?, Canadian Journal of Chemical Engineering 82, pp. 30-36, 2004