EUROSIM 2016 & SIMS 2016

CFD Approaches for Modeling Gas-Solids Multiphase Flows

– A Review

W.K. Hiromi Ariyaratne 1 E.V.P.J. Manjula 1 Chandana Ratnayake 2 Morten C. Melaaen 1

1
Faculty of Technology, Natural Sciences and Maritime Sciences, University College of Southeast Norway, Post box 235,
N-3603 Kongsberg, Norway,
hiromi.ariyaratne, jagath.m.edirisinghe, morten.c.melaae @usn.no
2 Department of POSTEC, Tel-Tek, Kjølnes ring 30, N-3918, Porsgrunn, Norway, chandana.ratnayake@tel-

tek.no

Abstract Achievement of this understanding involves the

This review study focuses on the application of development of experimental measurement techniques,
Computational Fluid Dynamics (CFD) in the experimentally verified multiphase flow equations and
investigation of gas-solids multiphase flow systems. numerical simulation tools (Arastoopour, 2001).
The applicability and limitations of conventional Significant effort has been devoted to improving
models and recent developments of existing multiphase numerical tools, such as Computational Fluid
models for the prediction of gas-solids flows are Dynamics (CFD) tool, to predict such complex flows.
thoroughly overviewed. Use of conventional Eulerian- However, it has been identified that systems containing
Eulerian model for granular flows and Lagrangian one or more particulate phases are the most complex
approach incorporated with Discrete Element Method and challenging in the field of multiphase flow
(CFD-DEM) are quite well proven, however some modeling. To accurately predict the solids behavior, it
limitations restrict the use of these models in wide is necessary to choose a numerical method capable of
range of applications. Therefore, some new models accounting not only particle-fluid interactions but also
have been introduced to model gas-solids flows, as for particle-wall and particle-particle interactions in
example Dense Discrete Phase Model incorporated three dimensions and across any particle size
with Kinetic Theory of Granular Flow (DDPM- distribution (Parker et al., 2013).
KTGF), Dense Discrete Phase Model incorporated Different types of CFD models are available for the
with Discrete Element Method (DDPM-DEM) and prediction of gas-solids flows. Each model has
Computational Particle Fluid Dynamics (CPFD) inherent merits and disadvantages. Therefore, a certain
numerical scheme incorporated with the MultiPhase- model can be appropriate over another depending on
Particle-In-Cell (MP-PIC) method. These models have the factors prioritized by the user e.g. accuracy of the
been validated for certain applications under certain results, computational time, applicability in large-scale
conditions, however, further validation of these models systems, etc. Moreover, the models are still far from
is still a necessity. perfect and the available models are undergone many
improvements within the time. In this review paper,
Keywords: models, CFD-DEM, DDPM-KTGF, some modeling approaches available for the modeling
DDPM-DEM, MP-PIC of gas-solids flow systems are analyzed including their
applications and limitations. First, an overview of the
1 Introduction models is presented. Then, the two basic approaches
Applications involving gas-solids multiphase flows are and the different models available under basic
very common in numerous industrial processes and approaches are discussed.
also in various natural phenomena, such as sand storms
and cosmic dusts (Li et al., 2012). Pneumatic 2 Basic CFD Approaches for
conveying units, hoppers, solids separation units such Modeling of Gas-Solids Flows
as cyclones, bubbling and circulating fluidized beds
used in gasification, carbon capture, etc. can be A brief summary of the discussed approaches and
identified as some of the industrial process units models are presented in Figure 1. In dealing with
involved in gas-solids flows. To optimize the design modeling of gas-solids flows, the Eulerian-Eulerian
and operation of industrial processes and also to and the Eulerian-Lagrangian methods are the
understand natural phenomena which involve gas- frequently used approaches (Chen and Wang, 2014). In
solids flows, a thorough understanding of gas-solids Eulerian-Eulerian approach, all the phases are treated
flows is needed. as continuous phases while in Eulerian-

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 680
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

CFD modeling of gas-solids

multiphase flows

Eulerian-Eulerian approach Eulerian-Lagrangian approach

Gas phase – Eulerian treatment Gas phase – Eulerian treatment
Solid phase - Eulerian treatment Solid phase – Lagrangian treatment

Eulerian–Eulerian DPM DDPM- CFD-DEM MP-PIC

model for granular KTGF
flows
p-p p-p p-p p-p p-p
inetractions: interactions: interactions: interactions: interactions:
KTGF ignored KTGF soft sphere particle normal
model stress model

Figure 1. Summary of model approaches for gas-solids multiphase flow modelling.

Lagrangian approach, the fluid phase is treated as a Volume fractions of phases are assumed to be
continuous phase, but the solid phase is treated as a continuous functions of space and time. Since the
discrete phase. Eulerian-Eulerian model for granular volume of a phase cannot be occupied by the other
flows is described under Eulerian-Eulerian approach phases, the sum of volume fractions is equal to one.
and there are four main models under Eulerian- This is the concept of the phasic volume fraction
Lagrangian approach, namely, Lagrangian Discrete (Abbasi et al., 2011). The conservation equations of
Phase Model (DPM), Dense Discrete Phase Model mass, momentum and energy for the phases are then
incorporated with Kinetic Theory of Granular Flow obtained through an appropriate averaging process
(DDPM-KTGF), CFD-Discrete Element Method (typically ensemble-averaging) (Chen and Wang,
(CFD-DEM) and Computational Particle Fluid 2014). The averaging procedure leads to many
Dynamics (CPFD) numerical scheme incorporated unclosed terms, which must be modeled (Snider et
with the MultiPhase-Particle-In-Cell (MP-PIC) al., 2011). Constitutive relationships that are obtained
method. from empirical information and/or kinetic theory are
In addition to the difference in the way of solid used for this purpose (Abbasi et al., 2013). Eulerian-
phase treatment, another basic difference of the models Eulerian model for granular flows (Euler-granular
under both approaches is, the way of treating particle- model) is an example for Eulerian-Eulerian approach.
particle (p-p) interactions. DPM neglects the p-p As mentioned (Garg et al., 2012), commercially
interactions, and other models consider the p-p available codes like ANSYS Fluent, and open source
interactions through different approaches such as codes like CFDlib, OpenFOAM® and MFiX are all
kinetic theory of granular flow, particle normal stress capable of performing Eulerian-Eulerian simulations.
model, soft sphere model, etc. Similar forms of governing equations are solved in all
Much information about the model approaches are these codes and main difference can be found in
discussed in the following sections. closures for various sub models (such as solids
stresses, interphase drag, etc.) and in numerical
3 Eulerian-Eulerian approach treatment.
The Eulerian-Eulerian approach normally requires
In the Eulerian-Eulerian approach, fluid and particles less computational resources compared to Eulerian-
both are considered as continuous phases which are Lagrangian approaches (Chen and Wang, 2014). And
fully inter-penetrating (Zhang et al., 2012) i.e. solid this approach is quite traditional and has played a very
phase is treated as a pseudo-fluid (Abbasi et al., 2013).

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 681
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

important role in determining the fluid dynamic of particles are involved, the calculation of particle–
characteristics of gas–solids flow (Chen and Wang, particle interactions is very complex. It is not possible,
2014). Therefore, Eulerian-Eulerian approach has a even with super computers, to simulate a large-scale
wide application in gas–solids flows (Weber et al., system due to the extensive computational cost of
2013). However, this approach has major limitations in tracking each particle (Li et al., 2012). Because of this
considering variations of particle properties, as complexity of calculating particle-particle collisions
example, wide particle size distribution, density and the high collision frequency for volume fractions
diversification and sphericity consideration. above 5%, these calculations have been limited to the
Nevertheless, the particle size differences and/or order of 2105 particles and are often restricted to two-
density variations can affect the gas–solids flow dimensional solutions without a fluid phase (Snider et
behaviors such as solid segregation (Wang et al., al., 2011). To avoid this restriction, some methods
2014a), hence cannot be neglected in certain situations. have been developed with improvements in calculating
In that case, many separate continuity and momentum particle-particle and particle-wall interactions and also
equations are required to accurately represent the with concept of parcels. The concept of parcel is used
different particle types and sizes in this model to reduce the numbers of particles involved in
(Andrews and O'Rourke, 1996). However, the computations, resulting in a significant acceleration of
computational cost of inclusion of many phases cannot the speed of simulations (Chen and Wang, 2014).
then be overlooked, in fact it depends on the According to (Garg et al., 2012), all publicly available
computational capacity available. Some researchers codes except for MFiX-DEM employ a parcel-based
have quoted that the Euler-granular model cannot approach for the discrete phase. In the parcel approach,
easily account for some characteristics of realistic a finite number of parcels are tracked rather than using
particles such as shear stresses and inter-particle actual individual particles. Each parcel may represent a
cohesive forces for Geldart A particles when treated as fractional number of real particles. Typically, several
a pseudo fluid (Chen et al., 2013). Moreover, many particles with same properties (species, size, density,
researchers emphasize that incorporating of dissipation temperature, etc.) are grouped and put into a parcel.
in the Kinetic Theory of Granular Flow (KTGF) model This parcel is also called as a
considering the effects of wall roughness is an computational/numerical/notional/nominal particle in
important factor for the accurate prediction of results in different literature. However, as ANSYS Fluent
Eulerian-Eulerian model for granular flows (Chen and mentioned, convergence issues can arise, if fluid
Wang, 2014). volume fraction becomes zero due to either when
parcel size is bigger than cell size or too many parcels
4 Eulerian-Lagrangian Approach are squeezed into a cell due to softness of particles.
Larger parcel size reduces the number of parcels for a
In Eulerian–Lagrangian approach, the fluid phase is
still modeled with time-averaged Navier–Stokes certain mass flow hence lower computational cost.
equations and other conservation equations (Yin et al., However, the smallest cell should be larger than the
largest parcel size (as explained above). Therefore,
2014). The dispersed (solid) phase is treated by
tracking a large number of particles through the finding the balance for the optimum mesh is important
calculated flow field (Abbasi et al., 2012). Each when using parcel concept. Brief overview of some of
Eulerian-Lagrangian models are presented in next
particle is affected in its trajectory by three-
dimensional forces and Newtonian equations of motion sections.
are used for the calculations (Yin et al., 2014). 4.1 Lagrangian Discrete Phase Model
Commercially available codes like ANSYS Fluent and
Barracuda®, and open sources codes like MFiX-DEM,
(DPM)
KIVA, OpenFOAM® are capable of performing For low and intermediate solids loading, the inter-
Eulerian-Lagrangian simulations (Garg et al., 2012). particle spacing is high and hence the negligence of
The way of treating particle-particle interactions and particle-particle interactions might be justifiable. The
the numerical method used to solve the equations are commercial code; ANSYS Fluent has Lagrangian
the main differences in different Eulerian-Lagrangian Discrete Phase Model (DPM) with such a treatment for
codes. the flows with solids volume fraction less than 10%. In
Compared to Eulerian-Eulerian approach, Eulerian- that model, the volume occupied by solids is not taken
Lagrangian approach can provide analysis of flows into account when assembling the continuous phase
with a wide range of particle types, sizes, shapes and equations and particle pressure and viscous stresses
velocities (Lu et al., 2014). However, if details of due to particles are neglected. The fluid carrier
particle–particle and particle-wall collisions are influences the particulate phase via drag and turbulence
explicitly tracked, the traditional Lagrangian models and if the interaction with continuous phase is enabled,
also have some major limitations (Chen and Wang, additionally the particles in turn influence the carrier
2014). For the dense systems in where a large number fluid via reduction in mean momentum and turbulence.

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 682
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

So, this method has either one-way or two-way available CFD platforms through user defined
coupling between the phases, but not four-way functions are quite common practices. Standalone
coupling where the particle-particle interactions are DEM simulation codes (codes for pure particulate
considered (Fluent, 2013). The particle-wall collisions flows without carrier fluid) include open source codes,
are modeled through relatively simple models, often such as LAMMPS and YADE, and commercial codes,
based on a simple reflection coefficients of restitution. such as EDEM® and ITASCA. Efforts to couple such
standalone DEM codes to existing computational fluid
4.2 Dense Discrete Phase Model dynamic solvers have recently been undertaken. For
Incorporated with Kinetic Theory of example, the EDEM code provides users the ability to
Granular Flow (DDPM-KTGF) couple its DEM modules with other CFD codes such as
Dense Discrete Phase Model incorporated with Kinetic ANSYS Fluent. Recently, OpenFOAM® has been
Theory of Granular Flow (DDPM-KTGF) for modeling coupled to YADE and LAMMPS (Garg et al., 2012).
particle-particle and particle-wall interactions are a In DEM, the whole process of collision or contact is
quite recently developed model. This model is solved by numerical integration of the equations of
available in commercial code ANSYS Fluent and open motion. A collision is treated as a continuous process
source code OpenFOAM®. This is a hybrid model that occurs over a finite time wherein the contact force
composed with Eulerian-Eulerian and Eulerian- is calculated as a continuous function of the distance
Lagrangian approaches. In low solids volume fractions, between colliding particles. These are based on
the particles are treated in a Lagrangian manner, while physically realistic interaction laws; as example spring,
in high solids volume fractions, the particles are treated spring dashpot and Coulomb’s law of friction.
using Eulerian treatment. The solids stress acting on Empirical values for the spring stiffness coefficient,
particles resulting from inter-particle interactions is damping constant and friction coefficient are required.
computed from the stress tensor given by the KTGF Compared to Lagrangian DPM, this model gives more
which is similar to Eulerian-Eulerian approach for accurate predictions for dense and near-packing limit,
granular flows (Euler-granular model). Compared to however at the cost of slower computations. As many
Lagrangian DPM, this model extends the applicability other Eulerian-Lagrangian models, CFD-DEM
from dilute to dense phase since this accounts for the incorporates with parcel concept in some codes, since
effect of volume fraction of solid phase and particle- recently. The parcel concept reduces an inherent
particle interactions. Still the preciseness of treating limitation of using DEM in large-scale and dense
particle-particle interactions with KTGF is doubtful. particle systems. Explicitly tracking collisions of all
Despite, having benefits of Lagrangian methods and is real particles demands very high computational cost
applicable to large systems, it demands further tests compared to tracking parcels which consist of group of
and validations. Some predictions for coal gasification real particles. Billions of real particles in large
and coal oxy-fuel combustion in circulating fluidized commercial systems can be analyzed using millions of
beds (Adamczyk et al., 2014a; Klimanek et al., 2015), parcels (Snider, 2007). As example, in-built DEM
circulating fluidized bed boiler (Adamczyk et al., capability including parcel concept is now available in
2014b), impinging particle jet in a channel (Chen and CFD solver, ANSYS Fluent. It is called Dense Discrete
Wang, 2014), solid sorbent carbon capture reactor Phase Model incorporated with Discrete Element
(Ryan et al., 2013) and ceramic dispersion in liquid Method (DDPM-DEM) and this is quite a new feature
pool (Zhang and Nastac, 2014) are made using in ANSYS Fluent. Published data for the application of
DDPM-KTGF model. (Ryan et al., 2013) have DDPM-DEM are rare and some information can be
experienced less stability of DDPM-KTGF solution found for modeling of micron-particle transport,
compared to Euler-granular model and MP-PIC interactions and deposition in triple lung-airways (Feng
method for a given reactor design and (Chen and and Kleinstreuer, 2014) and coal-direct chemical-
Wang, 2014) highlights the requirement of further looping combustion (Zhang et al., 2014). Another
improvements for DDPM-KTGF model. CFD-DEM code; MFiX-DEM is limited to small
problem sizes due to high computational cost incurred
4.3 CFD-Discrete Element Method (CFD- in the particle neighbor search algorithm in where real
DEM) particles are considered (Garg et al., 2012). The CFD-
DEM has been extensively proven to be effective in
Soft sphere model based on Cundall and Strack, also many gas-solids applications (Chen and Wang, 2014).
called “Discrete Element Method (DEM)” or “Distinct
Element Method” can be used to explicitly track the
particle-particle and particle-wall interaction terms in
typical Eulerian-Lagrangian approach (Crowe et al.,
2012). This model approach is often referred to as
“CFD-DEM” in most of the literature. In-house
developed CFD-DEM codes or DEM codes coupled to

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 683
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

4.4 Computational Particle Fluid Dynamics for any number of solid species and to model particle
(CPFD) Numerical Scheme volume fraction from dilute (<0.1%) upto dense
Incorporated with the Multiphase- (>60%). Some of the applications of MP-PIC method
are bubbling and circulating fluidized beds (Chen et al.,
Particle-in-Cell (MP-PIC) Method 2013; Jiang et al., 2014; Karimipour and Pugsley,
The Computational Particle Fluid Dynamics (CPFD) 2012; Lan et al., 2013; Liang et al., 2014; Parker et al.,
numerical scheme incorporated with the MultiPhase- 2013; Wang et al., 2014b; Weber et al., 2013; Yin et
Particle-In-Cell (MP-PIC) method to describe the solid al., 2014; Zhang et al., 2012), fluidized bed gasifiers
phase is quite new Eulerian-Lagrangian approach for (Abbasi et al., 2011; Loha et al., 2014; Singh et al.,
calculating gas-solids flows. This is a version after 2013; Snider et al., 2011; Thapa et al., 2014), fluidized
several significant improvements of Particle-In-Cell beds for carbon capture (Breault and Huckaby, 2013;
(PIC) method used for single-phase flows since 1960s Clark et al., 2013; Parker, 2014; Ryan et al., 2013),
(Snider, 2001). As Snider, Clark and O'rourke gas/liquid/solid fluidized beds (O’Rourke et al., 2009;
mentioned, the MP-PIC method is, in turn, an Vivacqua et al., 2013; Zhao et al., 2009), Rayleigh-
extension of the stochastic particle method of the Taylor mixing layers (Snider, 2001), sedimentation
KIVA code (Snider et al., 2011). In the CPFD method, (Andrews and O'Rourke, 1996; Snider, 2001), downer
the real particles are grouped into parcels as in many reactors (Abbasi et al., 2012, 2013), dryer (Bigda,
other Eulerian-Lagrangian methods (Zhang et al., 2014) , 3-D particle jet (Snider, 2001), hopper flow (Lu
2012). The dynamics of the particle phase is predicted et al., 2014; Snider, 2007), particle flow in U-tube
in the MP-PIC method by solving a transport equation (Snider, 2007).
which is called Liouville equation for the particle In addition to these models, Sommerfeld has
distribution function. The particle distribution function developed a stochastic collision model to model the
contains particle properties as example, particle spatial inter-particle collisions (Laín and Sommerfeld, 2012).
location, particle velocity, particle mass, time, etc. Furthermore, a brief comparison of results obtained
(Karimipour and Pugsley, 2012). Unlike DEM models using above mentioned models can be found in
which calculate particle-to-particle force by a spring– elsewhere (Chen and Wang, 2014).
damper model and direct particle contact, the CPFD
methodology models particles' collision force on each 5 Conclusions
particle as a spatial gradient. A particle normal stress
model is developed from this concept to describe the A general overview of some of the available gas-solids
particle collisions (Wang et al., 2014b). In the flow modeling approaches is made in the current
computation, the stress gradient on the grid is first review paper. Eulerian-Eulerian and Eulerian-
calculated and then interpolated to discrete particles Lagrangian are the approaches in use. Further,
(Abbasi et al., 2013). The model has been undergone Lagrangian Discrete Phase Model (DPM), Dense
through many improvements such as including Discrete Phase Model incorporated with Kinetic
Bhatnager, Gross and Krook (BKG) collision model Theory of Granular Flow (DDPM-KTGF), CFD-
for gas/liquid/solids flows (O’Rourke et al., 2009), Discrete Element Method (CFD-DEM) and
including collision damping fluctuations due to Computational Particle Fluid Dynamics (CPFD)
inelastic collisions (O’Rourke and Snider, 2010), numerical scheme incorporated with the MultiPhase-
including return-to-isotropy term in collision source Particle-In-Cell (MP-PIC) method are the models
term (O'Rourke and Snider, 2012), including the discussed under Eulerian-Lagrangian approach.
effects of the contact force variations caused by The conventional Eulerian-Eulerian model for
inhibition of relative motions due to different particle granular flows and CFD-DEM models have widely
sizes and densities (O'Rourke and Snider, 2014), etc. been used for many applications and validated quite
Arena-Flow®, Barracuda® and OpenFOAM® are well. Despite this, both models still have major
some examples for the software/codes which have limitations with respect to accuracy and computational
CPFD implementation. Compared to Lagrangian DPM, cost, hence applying to large scale systems and to
this model can accurately model gas-solids flows of model flows with different particle properties are not
dense and close-pack limits. Solution cost is reduced very straightforward. Therefore, these models are being
since the collisions are not directly solved as in DEM under improvements and some new models have been
and also due to implementation of the parcel concept. introduced to model gas-solids flows, as example
Furthermore, MP-PIC method does not need to take the DDPM-KTGF, DDPM-DEM and MP-PIC. In addition
particle collisions implicitly, therefore a much larger to getting advantage of Lagrangian treatment of the
time step can be adopted (Yin et al., 2014). As particles, these models are said to be efficient
mentioned (Lu et al., 2014), this method can be used to compared to the conventional models. This might be
model systems with physical particle counts over due to the use of parcel concept and/or due to use of
1×1015 particles. In addition, the CPFD method has empirical approaches for modeling particle-particle
shown the ability to model full particle size distribution interactions, alternative algorithms and grid. Few

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 684
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

publications related to use of MP-PIC method are Powder Technology, 235(Supplement C): 238-247, 2013.
available mainly in fluidized bed applications, however doi:10.1016/j.powtec.2012.10.014.
published information for the applications of other X. Chen and J. Wang. A comparison of two-fluid model,
models are not very abundant. Therefore, the dense discrete particle model and CFD-DEM method for
applicability and validity of these quite recent models modeling impinging gas–solid flows. Powder Technology,
for the accurate predictions of gas-solids multiphase 254(Supplement C): 94-102, 2014.
doi:10.1016/j.powtec.2013.12.056.
flow modeling should be investigated. Moreover, all
the models need further improvements in order to S. Clark, D. M. Snider, and J. Spenik. CO2 Adsorption loop
apply for wide range of applications and scales. experiment with Eulerian–Lagrangian simulation. Powder
Technology, 242(Supplement C): 100-107, 2013.
doi:10.1016/j.powtec.2013.01.011.
Acknowledgements
C. T. Crowe, J. D. Schwarzkopf, M. Sommerfeld, and Y.
The authors would like to acknowledge the financial Tsuji. Multiphase flows with droplets and particles, Taylor
support provided by the Research Council of Norway & Francis Group, LLC. 2012.
under PETROMAKS II program and Det Norske
Y. Feng and C. Kleinstreuer. Micron-particle transport,
oljeselskape ASA. interactions and deposition in triple lung-airway
bifurcations using a novel modeling approach. Journal of
References Aerosol Science, 71(Supplement C): 1-15, 2014.
A. Abbasi, P. E. Ege, and H. I. de Lasa. CPFD simulation of doi:10.1016/j.jaerosci.2014.01.003.
a fast fluidized bed steam coal gasifier feeding section. ANSYS fluent theory guide 15.0. Canonsburg, PA, ANSYS
Chemical Engineering Journal, 174(1): 341-350, 2011. Inc. 2013.
doi:10.1016/j.cej.2011.07.085. R. Garg, J. Galvin, T. Li, and S. Pannala. Open-source
A. Abbasi, M. A. Islam, P. E. Ege, and H. I. de Lasa. MFIX-DEM software for gas–solids flows: Part I—
Downer reactor flow measurements using CREC-GS- Verification studies. Powder Technology, 220(Supplement
Optiprobes. Powder Technology, 224(Supplement C): 1- C): 122-137, 2012. doi:10.1016/j.powtec.2011.09.019.
11, 2012. doi:10.1016/j.powtec.2012.02.005. Y. Jiang, G. Qiu, and H. Wang. Modelling and experimental
A. Abbasi, M. A. Islam, P. E. Ege, and H. I. de Lasa. CPFD investigation of the full-loop gas–solid flow in a
flow pattern simulation in downer reactors. AIChE circulating fluidized bed with six cyclone separators.
Journal, 59(5): 1635-1647, 2013. doi:10.1002/aic.13956. Chemical Engineering Science, 109(Supplement C): 85-
W. P. Adamczyk, P. Kozołub, G. Węcel, A. Klimanek, R. A. 97, 2014. doi:10.1016/j.ces.2014.01.029.
Białecki, and T. Czakiert. Modeling oxy-fuel combustion S. Karimipour and T. Pugsley. Application of the particle in
in a 3D circulating fluidized bed using the hybrid Euler– cell approach for the simulation of bubbling fluidized beds
Lagrange approach. Applied Thermal Engineering, 71(1): of Geldart A particles. Powder Technology,
266-275, 2014a. 220(Supplement C): 63-69, 2012.
doi:10.1016/j.applthermaleng.2014.06.063. doi:10.1016/j.powtec.2011.09.026.
W. P. Adamczyk, G. Węcel, M. Klajny, P. Kozołub, A. A. Klimanek, W. Adamczyk, A. Katelbach-Woźniak, G.
Klimanek, and R. A. Białecki. Modeling of particle Węcel, and A. Szlęk. Towards a hybrid Eulerian–
transport and combustion phenomena in a large-scale Lagrangian CFD modeling of coal gasification in a
circulating fluidized bed boiler using a hybrid Euler– circulating fluidized bed reactor. Fuel, 152(Supplement
Lagrange approach. Particuology, 16(Supplement C): 29- C): 131-137, 2015. doi:10.1016/j.fuel.2014.10.058.
40, 2014b. doi:10.1016/j.partic.2013.10.007. S. Laín and M. Sommerfeld. Numerical calculation of
M. J. Andrews and P. J. O'Rourke. The multiphase particle- pneumatic conveying in horizontal channels and pipes:
in-cell (MP-PIC) method for dense particulate flows. Detailed analysis of conveying behaviour. International
International Journal of Multiphase Flow, 22(2): 379-402, Journal of Multiphase Flow, 39(Supplement C): 105-120,
1996. doi:10.1016/0301-9322(95)00072-0. 2012. doi:10.1016/j.ijmultiphaseflow.2011.09.006.
H. Arastoopour. Numerical simulation and experimental X. Lan, X. Shi, Y. Zhang, Y. Wang, C. Xu, and J. Gao.
analysis of gas/solid flow systems: 1999 Fluor-Daniel Solids Back-mixing Behavior and Effect of the Mesoscale
Plenary lecture. Powder Technology, 119(2): 59-67, 2001. Structure in CFB Risers. Industrial & Engineering
doi:10.1016/S0032-5910(00)00417-4. Chemistry Research, 52(34): 11888-11896, 2013.
J. Bigda. CPFD Numerical Study of Impact Dryer doi:10.1021/ie3034448.
Performance. Drying Technology, 32(11): 1277-1288, T. Li, R. Garg, J. Galvin, and S. Pannala. Open-source
2014. doi:10.1080/07373937.2014.929586. MFIX-DEM software for gas-solids flows: Part II —
R. W. Breault and E. D. Huckaby. Parametric behavior of a Validation studies. Powder Technology, 220(Supplement
CO2 capture process: CFD simulation of solid-sorbent CO2 C): 138-150, 2012. doi:10.1016/j.powtec.2011.09.020.
absorption in a riser reactor. Applied Energy, Y. Liang, Y. Zhang, T. Li, and C. Lu. A critical validation
112(Supplement C): 224-234, 2013. study on CPFD model in simulating gas–solid bubbling
doi:10.1016/j.apenergy.2013.06.008. fluidized beds. Powder Technology, 263(Supplement C):
C. Chen, J. Werther, S. Heinrich, H.-Y. Qi, and E.-U. Hartge. 121-134, 2014. doi:10.1016/j.powtec.2014.05.003.
CPFD simulation of circulating fluidized bed risers. C. Loha, H. Chattopadhyay, and P. K. Chatterjee. Three
dimensional kinetic modeling of fluidized bed biomass

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 685
September 12th-16th, 2016, Oulu, Finland
EUROSIM 2016 & SIMS 2016

gasification. Chemical Engineering Science, and radial liquid mixing in water-fluidized beds of two
109(Supplement C): 53-64, 2014. solids exhibiting layer inversion. Chemical Engineering
doi:10.1016/j.ces.2014.01.017. Science, 95(Supplement C): 119-127, 2013.
H. Lu, X. Guo, W. Zhao, X. Gong, and J. Lu. Experimental doi:10.1016/j.ces.2013.03.011.
and CPFD Numerical Study on Hopper Discharge. Q. Wang, H. Yang, P. Wang, J. Lu, Q. Liu, H. Zhang, L.
Industrial & Engineering Chemistry Research, 53(30): Wei, and M. Zhang. Application of CPFD method in the
12160-12169, 2014. doi:10.1021/ie403862f. simulation of a circulating fluidized bed with a loop seal
P. J. O'Rourke and D. M. Snider. Inclusion of collisional Part II—Investigation of solids circulation. Powder
return-to-isotropy in the MP-PIC method. Chemical Technology, 253(Supplement C): 822-828, 2014a.
Engineering Science, 80(Supplement C): 39-54, 2012. doi:10.1016/j.powtec.2013.11.040
doi:10.1016/j.ces.2012.05.047. Q. Wang, H. Yang, P. Wang, J. Lu, Q. Liu, H. Zhang, L.
P. J. O'Rourke and D. M. Snider. A new blended acceleration Wei, and M. Zhang. Application of CPFD method in the
model for the particle contact forces induced by an simulation of a circulating fluidized bed with a loop seal,
interstitial fluid in dense particle/fluid flows. Powder part I—Determination of modeling parameters. Powder
Technology, 256(Supplement C): 39-51, 2014. Technology, 253(Supplement C): 814-821, 2014b.
doi:10.1016/j.powtec.2014.01.084. doi:10.1016/j.powtec.2013.11.041.
P. J. O’Rourke and D. M. Snider. An improved collision J. M. Weber, K. J. Layfield, D. T. Van Essendelft, and J. S.
damping time for MP-PIC calculations of dense particle Mei. Fluid bed characterization using Electrical
flows with applications to polydisperse sedimenting beds Capacitance Volume Tomography (ECVT), compared to
and colliding particle jets. Chemical Engineering Science, CPFD Software's Barracuda. Powder Technology,
65(22): 6014-6028, 2010. doi:10.1016/j.ces.2010.08.032. 250(Supplement C): 138-146, 2013.
doi:10.1016/j.powtec.2013.10.005.
P. J. O’Rourke, P. Zhao, and D. Snider. A model for
collisional exchange in gas/liquid/solid fluidized beds. S. Yin, W. Zhong, B. Jin, and J. Fan. Modeling on the
Chemical Engineering Science, 64(8): 1784-1797, 2009. hydrodynamics of pressurized high-flux circulating
doi:10.1016/j.ces.2008.12.014. fluidized beds (PHFCFBs) by Eulerian–Lagrangian
approach. Powder Technology, 259(Supplement C): 52-64,
J. Parker, K. LaMarche, W. Chen, K. Williams, H. Stamato,
2014. doi:10.1016/j.powtec.2014.03.059.
and S. Thibault. CFD simulations for prediction of scaling
effects in pharmaceutical fluidized bed processors at three D. Zhang and L. Nastac. Numerical modeling of the
scales. Powder Technology, 235(Supplement C): 115-120, dispersion of ceramic nanoparticles during ultrasonic
2013. doi:10.1016/j.powtec.2012.09.021. processing of aluminum-based nanocomposites. Journal of
Materials Research and Technology, 3(4): 296-302, 2014.
J. M. Parker. CFD model for the simulation of chemical
doi:10.1016/j.jmrt.2014.09.001.
looping combustion. Powder Technology, 265(Supplement
C): 47-53, 2014. doi:10.1016/j.powtec.2014.01.027. Y. Zhang, X. Lan, and J. Gao. Modeling of gas-solid flow in
a CFB riser based on computational particle fluid
E. M. Ryan, D. DeCroix, R. Breault, W. Xu, E. D. Huckaby,
dynamics. Petroleum Science, 9(4): 535-543, 2012.
K. Saha, S. Dartevelle, and X. Sun. Multi-phase CFD
doi:10.1007/s12182-012-0240-7.
modeling of solid sorbent carbon capture system. Powder
Technology, 242( Supplement C): 117-134, 2013. Z. Zhang, L. Zhou, and R. Agarwal. Transient Simulations of
doi:10.1016/j.powtec.2013.01.009. Spouted Fluidized Bed for Coal-Direct Chemical Looping
Combustion. Energy & Fuels, 28(2): 1548-1560, 2014.
R. I. Singh, A. Brink, and M. Hupa. CFD modeling to study
doi:10.1021/ef402521x.
fluidized bed combustion and gasification. Applied
Thermal Engineering, 52(2): 585-614, 2013. P. Zhao, P. J. O’Rourke, and D. Snider. Three-dimensional
doi:10.1016/j.applthermaleng.2012.12.017. simulation of liquid injection, film formation and
transport, in fluidized beds. Particuology, 7(5): 337-346,
D. M. Snider. An Incompressible Three-Dimensional
2009. doi:10.1016/j.partic.2009.07.002.
Multiphase Particle-in-Cell Model for Dense Particle
Flows. Journal of Computational Physics, 170(2): 523-
549, 2001. doi:10.1006/jcph.2001.6747.
D. M. Snider. Three fundamental granular flow experiments
and CPFD predictions. Powder Technology, 176(1): 36-46,
2007. doi:10.1016/j.powtec.2007.01.032.
D. M. Snider, S. M. Clark, and P. J. O'Rourke. Eulerian–
Lagrangian method for three-dimensional thermal reacting
flow with application to coal gasifiers. Chemical
Engineering Science, 66(6): 1285-1295, 2011.
doi:10.1016/j.ces.2010.12.042.
R. Thapa, C. Pfeifer, and B. Halvorsen. Modeling of reaction
kinetics in bubbling fluidized bed biomass gasification
reactor. Internal Journal of Energy and Environment, 5(1):
35-44, 2014.
V. Vivacqua, S. Vashisth, A. Prams, G. Hébrard, N. Epstein,
and J. R. Grace. Experimental and CPFD study of axial

DOI: 10.3384/ecp17142680 Proceedings of the 9th EUROSIM & the 57th SIMS 686
September 12th-16th, 2016, Oulu, Finland