sim and exp of rh and sand in fbg

8/3/2019 Sim and Exp of Rh and Sand in Fbg

http://slidepdf.com/reader/full/sim-and-exp-of-rh-and-sand-in-fbg 1/10

Simulation and experiment of segregating/mixing of rice husk–sandmixture in a bubbling fluidized bed

Sun Qiaoquna, Lu Huilina,*, Liu Wentiea, He Yuronga, Yang Lidana, Dimitri Gidaspowb

a Department of Power Engineering, Harbin Institute of Technology, School of Energy Science and Engineering, Harbin 150001, Chinab Department of Chemical and Environmental Engineering, Illinois Institute of Technology, IL 60616, USA

Received 28 May 2003; accepted 30 September 2004

Available online 8 December 2004

Abstract

The fluidization behavior of rice husk–sand mixture in the gas bubbling fluidized bed is experimentally and theoretically studied. The

relevancy of the pressure drop profile of rice husk–sand mixture to the definition of its minimum fluidization velocity is discussed, and the

minimum fluidization velocity of rice husk–sand binary mixture is determined. The distributions of mass fraction of rice husk particles along

the bed height are measured, and the profiles of the mean particle diameter of mixture are determined. A multi-fluid gas–solid flow model is

presented where equations are derived from the kinetic theory of granular flow. Separate transport equations are constructed for each of the

particle classes, allowing for the interaction between particle classes, as well as the momentum and energy are exchanged between the

respective classes and the carrier gas. The distributions of the mass fraction of rice husk particles and the mean particle diameter of binary

mixture are predicted. The numerical results are analyzed, and compared with experimental data.

q 2005 Elsevier Ltd. All rights reserved.

Keywords: Sand-rice husk mixture; Kinetic theory of granular flow; Segregation; Fluidization

1. Introduction

Biomass is an important renewable energy resource. It

not only has a wide distribution, but also abounds in

quantity [1,2]. Gasification of biomass-agriculture and

forest residues in fluidized bed-reactors is widely used for

obtaining producer gas, synthesis gas and chemicals like

methanol, etc. [3–5]. The rice husk is the outer cover of the

rice and on average it accounts for 20% of the paddy

produced, on weight basis. Experimental results indicate

that fluidized bed combustion technology seems to be the

suitable technology for converting a wide range of agricultural residues into energy due to its inherent

advantages of fuel flexibility, low operating temperature

and isothermal operating condition [6]. The fluidization

characteristics of biomass materials are very important for

the modeling and design of the reactors. However, biomass

cannot be easily fluidized alone due to their peculiar shapes,

sizes and densities. For proper fluidization and processing in

the reactor, a second solid, usually an inert material like

silica sand, alumina, calcite, etc. is used to facilitate

fluidization of biomass. It also acts as a heat transfer

medium in the reactor. The fluidization of sand and biomass

mixtures is characterized by particles of different shapes,

sizes, densities and compositions. Rao et al. [7] studied on

the fluidization of mixtures of sands and biomass of rice

husk, sawdust and groundnut shell powder to determine the

minimum fluidization velocity in a fluidized bed. These

experimental results show that, in general, it is difficult tofluidize rice husk, and its fluidization behavior improves

when it is mixed with other solid particles.

Mixtures of solid particles of different size and/or

different density tend to separate in vertical direction

under fluidized conditions. The nonuniform distribution of

the different solid components is caused by a competitive

action of mixing and segregation mechanisms. The

component that tends to sink at the air distributor is referred

to as a jetsam, while the component that tends to float on

0016-2361/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fuel.2004.09.026

Fuel 84 (2005) 1739–1748

www.fuelfirst.com

* Corresponding author. Tel.: C86 10 0451 8641 2258; fax: C86 10

0451 8622 1048.

E-mail address: [email protected] (L. Huilin).

http://www.fuelfirst.com/

http://www.fuelfirst.com/



the fluidized bed surface is referred to as a flotsam. Hence,

in typical biomass combustion systems with a small amount

of biomass fuel particles in a bed of sand particles, the sand

will be the jetsam component, and the biomass fuel particles

is the flotsam component. Segregation behavior of biomass

fuel is of practical importance because the vertical location

of biomass fuel influences the in-bed combustion efficiency

of volatile matter.

Segregation behavior and minimum fluidization velocity

of binary mixture were experimentally studied in bubbling

fluidized beds [8–10]. Nienow et al. [11] and Rowe et al.

[12] studied the segregation in the bubbling fluidized bed

consisting of binary mixtures for both different particle size

and density. Ekinci et al. [13] experimented the density and

size segregation behavior determined from temperature

distributions. Pilar et al. [14] have thoroughly reviewed

several investigations reported on the fluidization of

mixtures of solids with different particle sizes as well as

mixtures of particles of different sizes and densities.

Hoffmann et al. [15] experimented the segregation of the

different particle sizes and densities of binary mixture in the

bubbling fluidized bed. Wang et al. [16] investigated

the particle concentration profiles and minimum fluidizing

velocity of ternary mixtures. Wu et al. [17] studied the

behavior of segregation of particles consisting of equal

density, but different sizes. Mohammad et al. [18] reported

expermental results of different binary mixtures in a gas

bubbling fluidized bed. Marzocchella et al. [19] tested the

particle size distribution in the equal density and dissimilar

size of binary mixture in a bubbling fluidized bed. Manfred

et al. [20] studied the mixing and segregation behavior of

spherical solids in a bubbling fluidized bed of silica sand,

and the time average segregation patterns of the solid

mixtures were obtained from single particle trajectories

measured by the particle detection system based on an

electromagnetic principle. Formisani et al. [21] reported an

experimental study of the fluidization behavior of mixtures

of glass beads particles differing in size at various average

compositions.

Theoretical analyses of multicomponent particles are

available based on extensions of kinetic theory of dense

gases, appropriately modified to include the effect of energy

dissipations due to inelasticity [22,23]. In all of the

aforementioned models, the equipartition of granular energy

(the mean kinetic energy due to particle velocity fluctu-

ations) of the respective particle classes is assumed in the

derivation of kinetic energy equation of particles. However,

this assumption is hold for molecular systems where

dissipative effects are absent, and when the mass ratio of

the respective particles is moderate. For granular flow of

particle mixture, this assumption is inappropriate due to the

dissipation associated with the inelasticity of particle

collisions. Gidaspow et al. [24] extended the kinetic theory

of dense gases to binary granular mixture with unequal

granular temperature between the particle phases. The

hydrodynamics of binary mixture with different sizes were

studied by Mathiesen et al. using a CFD model, and

predicted the axial and radial velocity and particle

concentrations in a riser [25]. Goldschmidt et al. [26]

studied the influence of the restitution coefficient on the

segregation behavior of dense gas-fluidized beds based on a

multi-fluid Eulerian model. Wachem et al. [27] simulated

the flow behavior of gas-fluidized bed with a bimodal

particle mixture using a computational fluid dynamics

Nomenclature

C d drag coefficient

d particle diameter

e restitution coefficient

g gravity

gsn binary radial distribution function H bed height

I unit tensor

L height

m mass of a particle

n normal direction

p fluid pressure

ps solid pressure

q fluctuating energy flux

Re Reynolds number

t time

u velocity

Greek letterstg gas stress tensor

ts particle stress tensor

Dh segment height

q granular temperature

mg gas viscosity

ms shear viscosity

xs bulk viscosity3g porosity

3s volume fraction of particles

rs particle density

gs energy dissipation

b drag coefficient

Subscripts

av average

dil dilute

g gas phase

lam laminar flow

m solid phase

max maximum packingr rice husk particles

s sand particles, silica sand particles

S. Qiaoqun et al. / Fuel 84 (2005) 1739–17481740



model. Lathouwers et al. [28] presented a multi-fluid

approach where macroscopic equations are derived from

the kinetic theory of granular flows using inelastic rigid-

sphere models accounting for collisional transfer in high-

density regions. Huilin et al. [29] gave an extension to

binary mixtures of particles using kinetic theory of dense

gases, and simulated flow behavior of particles of binarymixture in the bubbling fluidized bed [30].

Moving from these considerations, and out of the

empirical approach generally followed in most of the

available literature, this paper tries to identify and outline

the role of some of the particle properties and of the

operative variables that have a major influence on the

fluidization behavior of size segregating in the bubbling

fluidized bed with biomass particles. To this purpose,

experiments have been conducted with mixtures of sand and

rice husk particles in a fluidized bed. Based on Huilin et al.

[29] study, a generalized multiphase gas–solid flow model

was presented. The gas-particle drag and particle–particle

interactions are considered in the model. The model was

applied to simulate the flow behavior of rice husk–sand

mixture in the gas bubbling fluidized bed. The simulated

results are compared with experimental data of binary

mixture in the bubbling fluidized bed.

2. Experimental equipment, materials and procedure

2.1. Apparatus and bed materials

All the experiments of this study have been performed in

a cross-section area of 245!450 mm, and height of 2000 mm bubbling fluidization apparatus made of plexiglas

as shown in Fig. 1. The fluidized bed equipped with a high-

pressure drop perforated distributor of gas. Fluidizing air

flow rates were regulated by a set of rotameters. The

pressure drops across the distributor and the bed are

measured by U-tube manometers. Bed height was evaluated

by averaging the values read on two graduated scales at the

wall, and then used for determining bed porosity. A solenoid

valve on the feed line was employed to cut the air flux off

instantaneously. Several windows at the front wall are

arranged to take out the bed materials in the fixed bed

condition. Each window height is 50 mm.

Biomass material used in the present work is rice husk.

The other solid materials used are sand and silica sand

particles. The density and average diameter of sand particlesare 2600 kg/m3 and 360 and 440 mm, and the density and

averaged diameter of silica sand particles are 2700 kg/m3

and 360 and 710 mm, respectively. The average dimensions

of the rice husk particles are 2 mm wide, 1 mm thick and

10 mm long. The averaged density of rice husk particles is

950.6 kg/m3.

2.2. Procedure

The binary mixture materials of rice husk and sand

particles, indicated as R-S360 and R-S440, and rice husk

and silica sand particles, indicated as R-Q360 and R-Q710,

are used. The binary mixture is initially thoroughly mixed

by fluidization condition at the given superficial gas velocity

that is higher than the minimum fluidization velocity. The

pressure drop along bed height is measured by U-tube

manometers. The porosity, 3g,f , between two measured

points Dh can be calculated, and the porosity distribution

along bed height is determined at the fluidization state. The

gas flowrate is simultaneously shut off. Thus the freezing

particles in the fixed bed of mixing state associated to a

given steady fluidization condition. The solid was gently

drawn from each window. Each of these layers was then

sieved to measure by weighing. The mass fraction of solid

component and the porosity were determined. Both porosityand mass fraction were then referred to the average height of

the relevant layers and used to trace the respective profiles

as a function of height. According to measured particle

weight within a layer height Dho, the averaged porosity, 3g,o,

was determined in the fixed bed. The height at fluidization

condition corresponding to height Dho of the fixed bed is

DhZDhoð1K3g;oÞ = ð1K3g;f Þ. The technique employed is

used by many research groups [11,17,31], and many others

and widely accepted in the field of fluidization.

3. Mathematical model

We consider a binary mixture of smooth, nearly elastic

sphere of two different particle classes A and B. These

particles have mass mk , density rk and velocity uk , where

k is either classes A or B. The granular temperature of

particle classes k is defined as: qk Z hC 2k i = 3, where C k is

the fluctuating velocity of classes k. The laws of

conservation of mass, momentum and granular tempera-

ture are satisfied for gas phase and particle classes

individually. Section 3.1 gives the governing equations for

gas–solids two-phase flow model [29,32]. In the particle

kinetic energy conservation Eq. (4), the first term onFig. 1. Scheme of experimental system of a bubbling fluidized bed.

S. Qiaoqun et al. / Fuel 84 (2005) 1739–1748 1741



the left-hand side denotes the time dependency of the

granular energy, the second term is the convection of the

granular energy. On the right-hand side, the first term

denotes the creation of the granular energy, the second

term is the diffusion of the granular temperature and the

third term is the dissipation of granular energy due to

inelastic particle–particle collisions, and the last term isthe dissipation due to fluid friction. Conservation

equations of mass, momentum and granular temperature

is solved for each solid classes.

In order to describe solid-phase stress in the momentum

and granular temperature equations, the constitutive

relations are needed. These constitutive laws specify how

the physical parameters of the phases interact with each

others. The constitutive equations come from the inter-

actions of the fluctuating and the mean motions of the

particles. The couple between the various particle classes is

through particle pressure, radial distribution function at

contact, viscosities, particle collision dissipation and

conductivities.

3.1. Equations of gas–solid two-phase flow

(1) Continuity equation (mZgas phase, solid classes):

v

vt ð3mrmÞCV$ð3mrmumÞZ0 (1)

(2) Gas phase momentum equation:

v

vt ð3grgugÞCV$ð3grgugugÞ

ZKV pCV$tgC3grg gC

XsZA;B

fgsðusKugÞ (2)

(3) Momentum equation for solid classes s (sZA, B):

v

vt ð3srsusÞCV$ð3srsususÞ

ZV$tsC3srs gCfgsðugKusÞ

C

XmZA;B;ssm

fmsðumKusÞ (3)

(4) Kinetic energy equation for solid classes s (sZA, B):

3

2

v

vt ð3srsqsÞCV$ð3srsqsusÞ

Z ðts : V$usÞCV$qsKgsK3fsgqs (4)

(5) Constitutive relations of gas–solid two-phase flow

(a) Gas phase stress tensor:

tgZ 3gmg½VugCVuT g �K 2

33gmgV$ug (5)

(b) Stress tensor of solid phase s (sZA, B):

tsZ ðK psCxsV$usÞICms½VusCVuT s �

K2

3msV$us I (6)

(c) Solid pressure ps (sZA, B):

psZ 3srsqsC

XmZA;B

pc;smZ 3srsqs

C

XmZA;B

pð1CesmÞd 3smgsmnsnmmsmmmoqsqm

3ðm2sqsCm2

mqmÞ

!m2

oqsqm

ðm2sqsCm2

mqmÞðqsCqmÞ 3 = 2

!ð1K3DC6D2K10D3

/Þ

DZðmsqsKmmqmÞ

½ðm2sq2s Cm2mq2mÞCqsqmðm2s Cm2mÞ�1 = 2

;

esmZesCem

2; d smZ

d sCd m

2;

moZ ðmsCmmÞ ð7Þ(d) Radial distribution function at contact for mixtures:

gsmZ1

1K3p

3s;max

C6d sd m

d sCd m

d

1K3p

3s;max

2

C8d sd m

d sCd m

2d2

1K3p

3s;max

3(8)

dZ2pðnsd 2s Cnmd

2mÞ = 3 and 3pZ 3sC3m

(e) Solid phase bulk viscosity (sZA, B):

xsZ

XmZA;B

d sm

3

2ðmsqsCmmqmÞ2

pqsqmðm2sqsCm2

mqmÞ 1 = 2

pc;sm

(9)

(f) Solid phase viscosity (sZA, B):

msZ

XmZA;B

pc;sm

d sm

5

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi2ðmsqsCmmqmÞ2

pqsqmðm2sqsCm2

mqmÞ

s

C2mdil;s

12

PmZA;Bð1CesmÞgsm

! 1C4

5

XmZA;B

ð1CesmÞ3mgsm

" #2

ð10Þ

mdil;sZ5

ffiffiffiffip

p 96

d srsq1 = 2s;av (11)




qs;avZ2qsP

MZA;Bnm

ns

d sm

d s

2m2

oqm

ðm2s qsCm2

mqmÞh i1 = 2

m2oqsqm

ðm2s qsCm2

mqmÞðqsCqmÞh i3 = 2ð1K3uC6u2

K/Þ 2

ð12Þ

(g) Drag coefficient between the gas and the solid

phases:

fgsZ4fgsjErgunC ð1K4ÞfgsjWen & Yu (13)

fgsjErgunZ150ð1K3gÞ3smg

32gd 3s

C1:75rg3sjugKusj

3gd s3g%0:8 (14)

fgsjWen & Yu

Z3

4C d

3g3sjugKusjd s

3K2:65g 3gO0:8 (15)

4Z arctan150!1:75ð0:2K3sÞ

p

C0:5 (16)

C dZ

24

Reð1C0:15 Re0:68Þ Re%1000

0:44 ReO1000

((17)

ReZjugKusj3grgd s

mg

(18)

(h) Drag coefficient of particle–particle:

fsmZ

pc:sm

3

d sm

2

ðm2

sqsCm2mqm

Þpm2oqsqm 1 = 2(

C1

jusKumj V ln3s

3m

C3VlnðmmqmÞlnðmsqsÞ

Cqsqm

qsCqm

Vqm

q2m

KVqs

q2s

ð19Þ

(i) Collisional heat flux for solid phase (sZA, B):

qsZ 3sk sVqsC

XmZA;B

pc;smð1CesmÞ 9mm

5mo

ðumKusÞ

Cd sm

2m2mqm

pðm2sqsCm2mqmÞ 1 = 2

"! V ln

3s

3m

C3VlnðmmqmÞlnðmsqsÞ

C32m3

s m3mqsqm

pðm2sqsCm2

mqmÞ 1 = 2

mmqsqm

qsCqm

!Vqs

q2s

KVqm

q2m

C6mm

2m3s m3

mqsqm

m2sqsCm2

mqm

3 = 2

!Vqs

msq2s

KVqm

mmq2m

ð20Þ

k sZ2k dil;s

12 PmZA;B

ð1Cesm

Þgsm

! 1C6

5

XmZA;B

gsm3mð1CesmÞm

" #2

C23srsd s

ffiffiffiffiffiqs

p

r XmZA;B

3mgsmð1CesmÞ (21)

k dil;sZ75 ffiffiffiffip

p 384

d srsq1 = 2s;av (22)

(j) Dissipation of the turbulent kinetic energy due to

particle collisions (sZA, B):

gsZ

XmZA;B

3

d sm

2m2oqsqm

pðm2sqsCm2

mqmÞ 1 = 2

(

K3moðmsqsCmmqmÞ

4ðm2sqsCm2

mqmÞ V$usgð1KesmÞ pc;sm

(23)

All simulations were carried out in a two-dimensional

Cartesian space. The boundary condition of walls is treated

as no-slip boundaries for the gas phase. The partial slip

condition applied to the particle classes is given by Sinclairet al. [33]:

Kprs3sgo

ffiffiffiffiqs

p

2 ffiffiffi

3p

3s;max

usZ ð3sts$ nÞ (24)

3sðqs$ nÞZKusð3sts$ nÞC ffiffiffi

3p pð1KewÞrs3sgoq

3 = 2s

43s;max

(25)

The boundary condition at the top of the bed is a pressure

boundary. The pressure at this boundary is fixed to a

reference value. Neumann boundary conditions are applied

to gas flow. At the bottom of the bed, gas velocity is given.

The bottom is assumed impenetratable for the solid classes

by setting the solids axial velocity to zero.

The modified K -FIX program, which was previously used

in the fluidization [32], is carried out in the simulations of

this study [30]. The K-FIX code employs a staggered finite

difference mesh system. Phase velocities are centered on

cell boundaries, whereas all other quantities are located at

the center of the mesh. The equations for the solid-phase

granular temperature, solid-phase stress and the drag on the

particle mixture were implemented into this code. The gas

phase is assumed to be compressible and the calculated

pressure is used to determine the gas density. The values of




restitution coefficients of sand and silica sand particles of

esZ0.9, and rice husk of erZ0.6 are used. Initially, the two

classes particles of uniform-mixed are filled with in the bed

with a given solid mass fraction. All simulations are

continued for 50 s of real simulation time, which require up

to 1 or 2 weeks of computational time on a PC (40 GB hard

disk, 128 Mb RAM, and 600 MHz CPU). All presentedtime-averaged distributions were taken from 10 to 50 s from

simulation results.

4. Results and discussions

At the beginning of a fluidization experiment, the

particles of binary mixture can be charged into the bed in

different ways: the fixed bed arrangement may be that of a

well-mixed assembly of particles, of two completely

segregated layers of each component or may represent any

other intermediate situation. In all experiments, the well-mixed arrangement of particles is used as the initial particle

bed. Fig. 2a shows the profiles of measured bed pressure

drop for R-S440 and R-S360 mixture, and Fig. 2b for

R-Q710 and R-Q360 mixtures as a function of superficial

gas velocity. As the superficial gas velocity increases, the

bed pressure drop increases gradually in the fixed bed where

all particles are stabilization. Until at the point A, the total

pressure drop of bed goes to a constant, and equals to the

particle weight of mixture per area of the bed. The bed

pressure drop versus superficial gas velocities is plotted for

determining the minimum fluidization velocity of mixture.

The minimum fluidization velocity is obtained from the

intersecting point of the curve of fixed bed at defluidization

with the constant pressure line at the fluidization condition.

Fig. 3 shows the measured minimum fluidization velocity as

a function of the averaged mass fraction of rice husk

particles. It is observed that the measured minimum

fluidization velocity increases with the increase of the

averaged mass fraction of rice husk, and decreases with the

decrease of the sand particle size.

Given the weight of rice husk particles in the bed, the

averaged porosity, 3g, can be obtained from the measured

bed height of rice husk particles alone in the fixed bed. From

the measured bed pressure drop of rice husk particles alone

in the fixed bed, the equivalent diameter of a sphere rice

husk particle, d r,av, can be obtained by solving Ergunequation at the given superficial gas velocity [32]:

D p

H Z 150

ð1K3gÞ2mgug

33gd 2r;av

C1:751K3g

33g

rg

d r;av

u2g (26)

The calculated equivalent diameter of rice husk particles is

1.54 mm.

Fig. 4 shows the simulated segregation patterns for the

R-S360 binary mixture with the averaged mass fraction of

rice husk particles xav,r of 5.82% at the superficial gas

velocity of 0.58 m/s. At the state t Z0 the sand particles and

rice husk particles are assumed to be well-mixed. As the

computational time proceeds, the sand particles aregradually accumulated into the bottom of bed, and the

mass fraction of sand particles increases at the bottom and

decreases in the upper regime of bed. The rice husk

particles, however, are floated in the upper regime of bed.

The mass fraction of rice husk particles decreases in the

bottom and increases at the upper regime of bed. Within

10 s almost complete segregation of rice husk and sand

particles is predicted in the bed.

Fig. 5 shows the experimented and calculated mass

fraction of rice husk particles as a function of bed height for

the binary mixtures of R-S440 at the averaged mass fraction

of rice husk particles of 5.82% and at the superficial gas

velocity of 0.58 and 0.79 m/s, respectively. The mass

fraction of rice husk particles is small in the bottom, and

high at the upper regime of bed. The mass fraction of sand

particles, however, is high in the bottom and low at the top

of the bed. This means that the sand particles and rice husk

particles are segregated along the bed height. We see that

with the increase of superficial gas velocity the distribution

of mass fraction of rice husk particles tends to uniform along

the bed height. From experimental observations, we see that

the sand particle will carried out from bottom to top of the

bed by bubbles with the increase of superficial gas velocity.

At the same time the rice husk particles are carried fromFig. 2. Pressure drop of rice husk–sand and rice husk–silica sand binary

mixtures as a function of superficial gas velocity.

Fig. 3. Measured minimum fluidization velocity as a function of mass

fraction of rice husk particles.




the upper regime of bed to the bottom through particles

circulation in the bed. The higher the superficial gas

velocity, the stronger the mixing between the rice husk

particles and sand particles. Hence, the high gas velocity

gives a more uniform distribution of sand and rice husk

particles in the bed.

The effect of the mean diameter of silica sand particles

on the distribution of mass fraction of rice husk particles

is shown in Fig. 6. The mass fraction of rice husk

particles increases at the upper regime, and decreases in

the bottom in the bed with the diameter of silica sand

particles from 360 to 710 mm. The minimum fluidization

velocity of mixture particles increases with the increase of

the mean diameter of silica sand particles, seeing Fig. 2.

The excess gas velocity, (ugKumf ), decreases with the

increase of the diameter of silica sand particles at the

given superficial gas velocity. The bubble number will be

decreased with the increase of the diameter of silica sand

particles. This causes the mixing between silica sand and

rice husk particles becomes weak. Due to the mass

difference between sand particles and rice husk particles,

the silica sand particles will tend to sink the bed bottom,

and the rice husk particles will float at the upper regime

of bed. Hence, the diameter of jetsam particles effects on

the distribution of mass fraction of rice husk particles in

the bed.

Fig. 7 shows the effects of restitution coefficients of

particles on the mass fraction distribution of rice husk

particles for the binary mixture of R-Q360 at the superficial

gas velocity of 0.61 m/s. We see that as the restitution

coefficient of rice husk particles increases from 0.5 to 0.7

the mass fraction of rice husk particles decreases at the

lower region, and increases at the upper region of bed. A

variation of the restitution coefficient influences on the

momentum and energy exchange between sand particles

and rice husk particles since the dissipation of fluctuating

energy will change considerably by inelastic particle

collisions. It can be seen that the value of restitution

coefficient of rice husk particles effects on the distribution of

mass fraction of rice husk particles in the bed.

Fig. 4. Segregation patterns of binary mixture of R-S360 at the superficial gas velocity of 0.58 m/s.

Fig. 5. Profile of mass fraction of rice husk for R-S440 at the averaged mass

fraction of 5.82%.

Fig. 6. Distribution of mass fraction of rice husk for R-Q710 and R-Q360

binary mixtures.




From measured and computed mass fraction of

particles, the mean particle diameter is calculated. Fig. 8

shows the mean particle diameter distributions as afunction of height for the binary mixtures of R-S440 at

the averaged mass fraction of rice husk particles of 5.82%.

The mean particle diameter increases with the decrease of

bed height. The mean particle diameter is larger in the

bottom region than that at the top of bed. At the averaged

mass fraction of rice husk particles of 5.82%, the mean

particle diameter for the binary mixtures of R-S440 is

459.1 mm. From the simulation results of R-S440 binary

mixture, the computed mean particle diameter increases

from 448.9 mm in the bottom to 476.2 mm at the bed

surface at the superficial gas velocity of 0.58 m/s, and

452.9 mm in the bottom to 468.8 mm at the bed surface at

the superficial gas velocity of 0.79 m/s. We see that thedistribution of mean diameter of binary mixture of R-S440

along the bed height trends to uniform with the increase of

superficial gas velocity.

Fig. 9 shows the mean particle diameter distributions as a

function of bed height for the binary mixtures of R-Q360

and R-Q710 at the averaged mass fraction of rice husk

particle of 8.86%. The mean particle diameters for the

binary mixtures of R-Q360 and R-Q710 are 383.1 and

741.4 mm, respectively. For R-Q360 binary mixture, the

simulated mean particle diameter increases from 366.4 mm

in the bottom to 385.7 mm at the bed surface. The simulatedmean particle diameter increases from 715.4 mm in the

bottom to 749.2 mm at the bed surface for R-Q710 binary

mixture. This indicates that the particle diameter of jetsam

effects on the segregation/mixing behavior of particles in

the bed.

The simulated time-averaged vertical and lateral particle

velocity distributions are shown in Fig. 10 for the binary

mixture of R-S360 at the averaged mass fraction of rice husk

particles and the superficial gas velocity of 5.82% and

0.93 m/s, respectively. The simulated results show that in

the center part of bed the sand particles and rice husk

particles flow upward with a high velocity of particles. In

the wall region, the vertical velocities of sand and rice husk

particles are negative. This indicates that the particles flow

down near the walls. The circulation of sand and rice husk

particles is formed in the bed. From Fig. 10, we see that the

vertical velocity of the rice husk particles is larger than that

of the sand particles. The relative velocity between the sand

and rice husk particles is larger in the center than that at the

walls. The time-averaged lateral particle velocities are

Fig. 7. Effect of restitution coefficients on mass fraction distribution of

biomass particles.

Fig. 8. Distribution of mean diameter of binary mixture of R-S440.

Fig. 9. Profiles of mean diameter of binary mixture of R-Q360.

Fig. 10. Distribution of vertical particle velocity of sand and rice husk

particles.




negative in the left side and positive in the right side. This

means that the sand and rice husk particles flow from walls

into center of bed.

Fig. 11 shows the simulated granular temperature of the

sand particles and rice husk particles for the binary mixtures

of R-S360 at the averaged mass fraction of rice husk particles and the superficial gas velocity of 5.82% and

0.93 m/s, respectively. The granular temperature of particle

classes is high wherever there is a great deal of motion of

particles. It can be seen that the rice husk particles give a

lower fluctuation energy, while the sand particles have a

higher kinetic energy offluctuations. The simulated granular

temperatures of rice husk and sand particles are lower in the

low portion than that at the top of the bed. We see that the

sand particles and rice husk particles have unequal granular

temperature in the bed.

5. Conclusions

The fluidization behavior of a binary mixture of sand and

rice husk particles with the different diameter and density is

strongly influenced by the variations of superficial gas

velocity and density and diameter of jetsam particles in the

bed. The initial fluidization state of a binary mixture is

characterized by the minimum fluidization velocity at which

the total pressure drop equals to the particles weight per unit

area of bed, which depends upon the averaged mass fraction

of rice husk particles. The minimum fluidization velocities

of binary mixture for the different sands and rice husk

particles are experimentally determined in a bubbling

fluidized bed. The measured minimum fluidization velocity

increases with the increase of mass fraction of rice husk

particles in the bed.

A computational fluid dynamics model has been

presented where the kinetic theory of granular flow forms

the basis for the turbulence modeling of the solid classes.

Separate transport equations are used for each particle

classes leading to momentum and energy exchange between

respective classes, and between particles and gas phase. The

model has been applied to simulate the binary mixtures flow

in a bubbling fluidized bed. A parametric study has been

performed. The mass fraction distributions of rice husk

particles are predicted. The mean particle diameter

distributions along bed height are computed. Simulated

results indicate the particle size, mass fraction of sand

particles and superficial gas velocity have considerable

impacted on the segregating behavior of rice husk particles.

The simulated results are in agreement with measured massfraction of rice husk particles. The simulation results

indicate that the model captures the key features of binary

mixture fluidization of biomass. Since the validation of

fundamental hydrodynamic models can only be made when

the collision parameters of particles are set to the correct

values, there is a great need for experiments for which the

rheology of particles has been accurately determined.

Acknowledgements

This work was supported by the National ScienceFoundation through Grant No. 50376013; and NSFC-

PetroChina Company Limited under the cooperative project

No. 20490200.

References

[1] Tillman DA. Biomass co-firing: the technology, the experience, the

combustion consequences. Biomass Bioenergy 2000;19:365–84.

[2] Hanzade H. Combustion characteristics of different biomass

materials. Energy Conversion Mgmt 2003;44:155–62.

[3] Natarajan E, Nordin A, Rao N. Overview of combustion and

gasification of rice husk in fluidized bed reactors. Biomass Bioenergy1998;14:533–46.

[4] Werther J, Saenger M, Hartge EU, Ogada T, Siagi Z. Combustion of

agricultural residues. Prog Energy Combust Sci 1998;26:1–27.

[5] Armesto L, Bahillo A, Veijonen K, Cabanillas A, Otero J. Combustion

behavior of rice husk in a bubbling fluidized bed. Biomass Bioenergy

2002;23:171–9.

[6] Raskin N, Palonen J, Nieminen J. Power boiler fuel augmentation

with a biomass fired atmospheric circulating fluid-bed gasifier.

Biomass Bioenergy 2001;20:471–81.

[7] Rao TR, Ram JV. Minimum fluidization velocities of mixtures of

biomass and sands. Energy 2001;26:633–44.

[8] Chenug L, Nienow AW, Rowe PN. Minimum fluidization velocity of

a binary mixtutre of differnet sized particles. Chem Eng Sci 1974;29:

1301–3.

[9] Chiba S, Chiba T, Nienow AW, Kobayashi H. The minimumfluidization velocity, bed expansion and pressure drop profile of

binary particle mixtures. Powder Technol 1979;22:255–69.

[10] Noda K, Uchida S, Makinno T, Kamo H. Minimum fluidization

velocity of binary mixture of particles with large size ratio. Powder

Technol 1986;46:149–54.

[11] Nienow AW, Naimer NS, Chiba T. Studies of segregation/mixing in

fluidized beds of different size particles. Chem Eng Sci 1987;62:

52–63.

[12] Rowe PN, Nienow AW, Agbim AJ. The mechanism by which particle

segregate in gas fluidized beds-binary systems of near spherical

particles. Trans Inst Chem Eng 1989;50:310–23.

[13] Ekinci E, Atakul H, Tolay M. Detection of segregation tendencies in a

fluidized bed using temperature profiles. Powder Technol 1990;61:

185–92.

Fig. 11. Distributions of simulated granular temperature of sand and rice

husk particles.




[14] Pilar AM, Gracia-Gorria FA, Corella J. Minimum and maximum

velocities for fluidization for mixtures of agricultural and forest

residues with second fluidized solid. Int Chem Eng 1992;32:95–102.

[15] Hoffmann AC, Janssen LPBM, Prins J. Particle segregation in

fluidized binary mixtures. Chem Eng Sci 1993;48:1583–94.

[16] Wang RC, Chou CC. Particle mixing/segregation in gas–solid

fluidized bed of ternary mixtures. Can J Chem Eng 1995;73:793–9.

[17] Wu SY, Baeyens J. Segregation by size difference in gas fluidized

beds. Powder Technol 1998;98:139–50.

[18] Mohammad GR, Victor R, Milan C. Segregation potential in binary

gas fluidized beds. Powder Technol 1999;103:175–81.

[19] Marzocchella A, Salatino P, Di Pastena V, Lirer L. Transient

fluidization and segregation of binary mixture of particles. AIChE J

2000;46:2175–82.

[20] Manfred W, Franz F, Natalia I, Genadij L. Particle mixing in bubbling

fluidized beds of binary particle systems. Powder Technol 2001;120:

63–9.

[21] Formisani B, De Cristofaro G, Girimonte R. A fundamental approach

to the phenomenology of fluidization of size segregating binary

mixtures of solids. Chem Eng Sci 2001;56:109–19.

[22] Jenkins JT, Mancini F. Kinetic theory for binary mixtures of smooth,

nearly elastic spheres. Phys Fluids 1989;31:2050–7.

[23] Zamankhan P. Kinetic theory of multicomponent dense mixtures of

slightly inelastic spherical particles. Phys Rev E 1995;52:4877–91.

[24] Gidaspow D, Huilin L, Mangar E. Kinetic theory of multiphase flow

and fluidization: validation and extension to binaries. XIXth

International Congress of Theoretical and Applied Mechanics,

Japan 1996.

[25] Mathiesen V, Solberg T, Arastoopour H, Hjertager BH. Experimental

and computational study of multiphase gas/particle flow in a CFB

riser. AIChE J 1999;45:2503–18.

[26] Goldschmidt MJV, Kuipers JAM, van Swaaij WPM. Hydrodynamic

modelling of dense gas-fluidized bed using the kinetic theory of

granular flow. Chem Eng Sci 2001;56:571–8.

[27] Wachem van BG, Schouten JC, van den Bleek CM, Krishna R,

Sinclair JL. Comparative analysis of CFD models of dense gas–solid

systems. AIChE J 2001;47:1035–51.

[28] Lathouwers D, Bellan J. Modeling of dense gas–solid reactive

mixtures applied to biomass pyrolysis in a fluidized bed. Int

J Multiphase Flow 2001;27:2155–87.

[29] Huilin L, Gidaspow D, Manger E. Kinetic theory of fluidized binary

granular mixtures. Phys Rev E 2001;64:61301–9.

[30] Huilin L, Yurong H, Gidaspow D. Hydrodynamic modelling of binary

mixture in a gas bubbling fluidized bed using the kinetic theory of

granular flow. Chem Eng Sci 2003;58:1197–205.

[31] Geldart D, Baeyans J, Pope DJ, Van de Wijer P. Segregation in beds

of large particles at high velocoites. Powder Technol 1981;30:

195–205.

[32] Gidaspow D. Multiphase flow and fluidization: continuum and kinetic

theory description. San Diego: Academic Press; 1994.

[33] Sinclair JL, Jackson R. Gas-particle flow in a vertical pipe with

particle–particle interactions. AIChE J 1989;35:1473–86.


sim and exp of rh and sand in fbg

Documents