Powder Technology 354 (2019) 64–70

Contents lists available at ScienceDirect

Powder Technology

journal homepage: www.elsevier.com/locate/powtec

On the two-phase theory of fluidization for Geldart B and D particles

Zhijie Fu a,b, Jesse Zhu a,b,⁎, Shahzad Barghi b, Yuemin Zhao a,⁎⁎, Zhenfu Luo a, Chenlong Duan a
a
School of Chemical Engineering & Technology, China University of Mining and Technology, Xuzhou 221116, China
b
Department of Chemical & Biochemical Engineering, Western University, London N6A 5B9, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Knowledge of the division of gas flow between the bubble and dense phases is important for modelling and
Received 3 November 2018 operation of bubbling fluidized beds. The two-phase theory of fluidization, which suggested that the bubble
Received in revised form 11 February 2019 flow rate being equal to the excess gas flow above the incipient fluidization, has been proved to be an overesti-
Accepted 22 May 2019
mation in most cases. While the two-phase theory has been modified by introducing a correction factor (Y), most
Available online 24 May 2019
previous studies were conducted for Geldart Group A powders. In the present work, the contribution to predict
Keywords:
the parameter Y for Geldart Group B and D particles has been formulated based on almost all the available exper-
Bubbling fluidized bed imental data. The experimental results demonstrated that the Y value increases with decreasing particle size or
Two-phase theory of fluidization density and increasing excess gas velocity. A new correlation has been developed to estimate the Y value for
Bubble phase Geldart Group B and D particles

0:024
Y ¼ 1:72Ar−0:133 U g −U mf

with an overall standard deviation of 19%. It only requires the knowledge of Archimedes number and excess
gas velocity. This correlation is in reasonable agreement with almost all the available data in literature and the
present work.
© 2019 Elsevier B.V. All rights reserved.

1. Introduction the distribution of gas flow. It is usually formulated as

Bubbling gas-solid fluidized bed is commonly operated at rela- Gb =A ¼ U g −U mf ð1Þ

tively lower gas flow rate, characterized by the solid particles
becoming individually suspended with interstitial gas flow and
gas bubbles rising with coalescence [1,2]. As the gas-solid
contacting and gas residence time are usually different between
the interstitial gas flow and bubble flow, the distribution of gas
flow will play a critical role in the modelling and design of fluid-
ized bed operations, especially for the gas-solid chemical reactions
[3–5], combustion and gasification [6–9], solids mixing and drying
[10–12], fluidized bed separation [13,14], etc. In general, it is con-
sidered that the bubbling fluidized bed is composed of the dense
(emulsion) phase and bubble phase [15,16], and the comprehen-
sive knowledge of the division of gas flow between these two
phases is therefore crucial for the fluidized bed operations. The
two-phase theory of Fluidization [17] suggested that all the gas
flow in excess of that required for incipient fluidization is in the
form of gas bubbles, which provides a possible way to analyze
⁎ Corresponding author at: Department of Chemical & Biochemical Engineering,
Western University, London N6A 5B9, Canada.
⁎⁎ Corresponding author.
E-mail addresses: jzhu@uwo.ca (J. Zhu), ymzhao_paper@126.com (Y. Zhao).
where Gb is the volumetric bubble flow rate, A is the cross-
sectional area of fluidized bed, U g is the superficial gas velocity,
and Umf is the minimum fluidization velocity. The two-phase the-
ory [17] implies that the bed vodiage and the interstitial gas veloc-
ity in the dense phase remain almost the same as in the incipient
fluidization state, which is of great importance for the modelling
and operation purposes. Unfortunately, most of the experimental
evidences [18–23] have demonstrated that the original two-phase
theory is only approximately true and tends to overestimate the
visible bubble flow in most cases.
There was considerable controversy over the reasons for the unreli-
able prediction of the original two-phase theory. Some authors
[21,24–30] have attributed the deficit of bubble flow to an increase in
the interstitial gas flow in dense phase above that required for mini-
mum fluidization. At the other extreme, a number of workers [31–33]
have ascribed this discrepancy to the through-flow of gas inside the
bubble phase. Other investigators [22,23] claimed that the original
two-phase theory postulate, even including through-flow in isolated
bubbles, substantially over-predicted the visible bubble flow rate.
Thus, many modifications to the two-phase theory of fluidization have

https://doi.org/10.1016/j.powtec.2019.05.051
0032-5910/© 2019 Elsevier B.V. All rights reserved.
 Z. Fu et al. / Powder Technology 354 (2019) 64–70
Fig. 1. The schematic diagram of the two-phase theory of fluidization.
been proposed in the literature, aiming to improve the accuracy and re-
liability for modelling and operation purposes.
The earlier form was known as n-type two-phase theory [27], in the
form of
Gb =A ¼ U g −U mf ð1 þ nδÞ ð2Þ
where n is the through-flow coefficient, and δ is the fraction of the cross-
sectional area occupied by gas bubbles. After then, they summarized the
available data in the literature and gave an extensive compilation of the
experimental value of n [21]. Their results indicate that the factor n were
reported to vary in the range of −8 ~ 140, respectively, which were
shown to be too difficult to be estimated.
Another form was developed by several workers [33–36] as


Gb =A ¼ Y U g −U mf ð3Þ
where Y is the correction factor. The parameter Y indicates the deviation
of the visible bubble flow rate from the original two-phase theory,
which was found to be usually below unity. The Y value for different
types of powders of Geldart's classification have been described as [37].
Fig. 2. Schematic diagram of the experimental apparatus: 1. Air filter; 2. Roots blower; 3. Tank; 4. Pressure gangue; 5. Gas valve; 6. Rotameter; 7. Air chamber; 8. Perforated distributor; 9.
Fluidized bed column; 10. Ruler; 11. U-shaped manometer; 12. Dust cover; 13. Dust collector.
65
0:8 b Y b 1:0 Group A powders
0:6 b Y b 0:8 Group B powders
0:25 b Y b 0:6 Group D powders
It is noteworthy that normal fluidization is extremely difficult for
Geldart Group C powders, and thus the corresponding Y value is com-
monly excluded. For Geldart Group A powders, a number of works
[38–40] have been conducted to investigate the gas flow distribution
due to its importance for the chemical reactions. The results concluded
that the range of the corresponding Y values is relatively narrow (0.8–
1.0), and the numerical value of 0.85 was usually recommended [37].
However, for Geldart Group B and D particles, there is no reasonable
and suitable predicting equation.
Comprehensive analysis of almost all experimental data shows
that the division of gas flow between the dense phase and bubble
phase is highly dependent on the particle size, density and superfi-
cial gas velocity. An attempt has been made to develop a correla-
tion for predicting the correction factor (Y) of two-phase theory
for bubbling fluidized bed with Geldart Group B/D particles. Almost
all available data on gas-solid systems have been correlated to val-
idate this correlation, and the calculated results have been com-
pared with the experimental data in literature and the present
work (Fig. 1).
66 Z. Fu et al. / Powder Technology 354 (2019) 64–70

2. Theory 3. Experimental

According to the modified two-phase theory [36], the parameter Y
indicates the deviation of the visible bubble flow rate from the original
two-phase model, which can be obtained from the estimation of the flu-
idized bed expansion. Accordingly, the volume occupied by gas bubbles
per bed cross-sectional area can be expressed as
dV b ¼ Gb dh=U b
where Vb is the volume occupied by bubbles, dh is the differential height
of fluidized bed, and Ub is the visible bubble flow rate. Thus, the total
volume occupied by gas bubbles in the bed is
Z H
Vb ¼ Gb dh=U b ¼ Gb H=U b
0
where U b is the average bubble flow rate, H is the fluidized bed height.
The volumetric bubble flow rate (Gb) should be calculated from the
modified two-phase theory,
3.1. Experimental setup
Experiments were conducted in a fluidized bed at ambient condi-
tions, as shown in Fig. 2. The experimental setup consists of mainly
four parts: (1) air supply including an air filter, a roots blower, and a
tank; (2) fluidized bed column with inner diameter of 152.4 mm.
ð4Þ (3) U-shaped monometers for pressure-drop measurement; (4) dust
collection device. After being filtered, the ambient air was sent to fluid-
ize the particles in the column through the air chamber and a perforated
distributor. The distributor is made of two plastic perforated plates with
filter cloth in between, and the orifice diameter is 1.5 mm with the total
open area of 11%. To investigate the expansion of fluidized bed, a ruler is
attached on the side of the column and the U-shaped piezometric pipes
ð5Þ are connected to the axial pressure taps on the other side of the column.
Fine dust generated during fluidization was collected by the dust collec-
tor device.
3.2. Experimental materials


Gb ¼ Y U g −U mf A ð6Þ Four types of solid materials have been employed in this work:
magnetite, sand, gangue and glass beads. These samples were
sieved into the following size fractions: 74–150 μm, 150–300 μm,
The important thing is to determine the average bubble flow rate, 300–425 μm, 425–590 μm, 590–710 μm. The particle properties of
and it can be given by the experimental materials of each size fraction are shown in
Z Tables 1, 2, 3, and 4. It is noted that the solid particles with the
H
Ub ¼ U b dh=H ð7Þ angle of repose above 38 is considered cohesive. The gangue sample,
0 which is the heavy product (N1.85 g/cm3 ) of the coal separation

For freely bubbling beds, the bubble rise velocity is usually estimated Table 1
The particle properties of magnetite samples.
fro an equation proposed by Davidson et al. [41]
Particle size fraction (μm) 74–150 150–300 300–425 425–590
pffiffiffiffiffiffiffiffi

U b ¼ 0:71 gDe þ U g −U mf ð8Þ Mean particle size (μm) 121 213 348 457
Particle true density (kg/m3) 4480 4650 4570 4540
Aerated bulk density (kg/m3) 2460 2667 2687 2652
where De is the diameter of an isolated bubble. Various correlations Archimedes number (Ar) 303 1715 7349 16,533
have been proposed for the estimation of mean bubble size in bubbling Angle of repose (°) 35.7 36.1 37.4 38.3
fluidized beds, among which Darton equation [42] is one of the most
commonly used correlations considering the effects of bed height, gas Table 2
distributor and gas velocity, and is defined as The particle properties of glass bead samples.

Particle size fraction (μm) 74–150 150–300 300–425 425–590 590–710

0:4  0:8
De ¼ 0:54 U g −U mf h þ 4A0:5
D =g 0:2 ð9Þ Mean particle size (μm) 101 209 356 469 648
Particle true density (kg/m3) 2550 2620 2650 2680 2640
Aerated bulk density (kg/m3) 1560 1603 1611 1605 1612
Expansion of bubbling fluidized beds for Geldart Group B and D par- Archimedes number (Ar) 100 912 4561 10,547 27,403
Angle of repose (°) 32.5 33.1 34.3 34.6 35.2
ticles in general results from the volume occupied by gas bubbles, and
total volume of bubbles can be written
Table 3

 The particle properties of sand samples.
V b ¼ H−Hmf A ð10Þ
Particle size fraction (μm) 74–150 150–300 300–425 425–590 590–710

Submitting Eqs. (5) and (6) into Eq. (10), the expansion of fluidized Mean particle size (μm) 113 224 368 475 636
Particle true density (kg/m3) 2430 2530 2410 2510 2500
beds leads to
Particle bulk density (kg/m3) 1493 1544 1610 1602 1593
Archimedes number (Ar) 134 1085 4581 10,261 24,534



H−Hmf =H ¼ Y U g −U mf =U b ð11Þ Angle of repose (°) 33.6 34.5 37.4 38.1 38.5

Combination of Eqs. (7), (8), (9) and (11), the parameter Y of the Table 4
The particle properties of gangue samples.
modified two-phase theory can be calculated from.
Particle size fraction (μm) 74–150 150–300 300–425 425–590 590–710
2 1:4  1:4 3
0:5 0:5 Mean particle size (μm) 118 215 372 486 625
H−Hmf 6 H þ 4A D − 4A D 7 Particle true density (kg/m3) 2010 2050 2160 2120 2090
Y ¼ 0:93 4
0:8 þ 15 ð12Þ
H H U g −U mf Particle bulk density (kg/m3) 1180 1240 1290 1360 1330
Archimedes number (Ar) 126 777 4241 9282 19,462
Angle of repose (°) 41.8 41.3 41.5 42.8 43.3
 Z. Fu et al. / Powder Technology 354 (2019) 64–70 67

1.6 1.6
74 150 um Magnetite particles 74 um Glass bead
150 300 um Magnetite particles 150 um Glass bead
300 425 um Magnetite particles 300 425 um Glass bead
1.2 425 590 um Magnetite particles 1.2 425 590 um Glass bead
590 710 um Glass bead
Y

Y
0.8 0.8

0.4 0.4

0.0 0.0
0 1 2 3 4 5 0 1 2 3 4 5
Ug Umf (cm/s) Ug Umf (cm/s)

1.6 1.6
74 150 um Sand particles 74 150 um Gangue particles
150 300 um Sand particles 150 um Gangue particles
300 425 um Sand particles 300 425 um Gangue particles
1.2 425 590 um Sand particles 1.2 425 590 um Gangue particles
590 710 um Sand particles 590 710 um Gangue particles
Y

0.8
Y

0.8

0.4 0.4

0.0 0.0
0 1 2 3 4 5 0 1 2 3 4 5
Ug Umf (cm/s) Ug Umf (cm/s)

Fig. 3. Plot of Y value against the excess gas velocity for different types of solid particles.

process, is collected from HuaiBei Coal Mine, Ltd., Anhui, China. It
may be noteworthy that the magnetite particles of 590–710 μm
was excluded in the present work due to its limited flowability.
4. Results and discussion
4.1. The effects of particle property and excess gas velocity on correction
factor Y
Measurements of the fluidized bed expansion have been carried
out to determine the correction factor Y by using Eq. (12). The pres-
sure drop against distance above the distributor graph was
employed to obtain the fluidized bed height at a certain superficial
gas velocity. In the present work, the initial bed height was 20 cm
and the excess gas velocity ranged from 0.5 cm/s to 4.5 cm/s. The Y
value of magnetite, glass bead, river sand and gangue particles
with the size range from 74 um to 710 um is reported as a function
of the excess gas velocity in Fig. 3.
As can be observed from Fig. 3, Y value was found to decrease
with the increasing particle size for the same material. This trend
may be explained that the greater the particle size, the greater
the interstitial gas velocity in the dense phase, which may results
in more deficit of bubble flow in the fluidized bed [20,21]. It can
also be observed from Fig. 3 that Y value of different types of solids
with same particle size decreases slightly with increasing particle
density. This can also be attributed to the increasing gas velocity
required to fluidize solid particles in the dense phase. Moreover,
it can be seen that there is an increase in Y value with the increas-
ing of excess gas velocity, which can be attributed to the tendency
of the gas flow moves more into the bubble phase. For the model-
ling and operation purposes, it is important to predict the Y value
for the modified two-phase theory, thus avoiding the experimental
measurements.
4.2. The correlation for estimating the correction factor Y
Experimental Y values of the present investigation have been
summarized and plotted against Archimedes number and excess
gas velocity in Fig. 4. To consider the effect of particle properties in
gas-solid fluidization systems, Archimedes number is employed
68 Z. Fu et al. / Powder Technology 354 (2019) 64–70

Ar (100-2000)
Ar (2000-10000) 1.0
Ar (10000-15000)
Ar (15000-30000)
0.8

0.6

0.10
0.4

Y
0.08

0.06
0.
U g - U mf (m/s)

2
0.04

0.
0.02
0
25
200

00
150

0
00
1000

0.00
00
5000
0

Ar

Fig. 4. Effect of Archimedes number and excess gas velocity on parameter Y.

and is defined by Fig. 5. The summary of Y values of all available data in literature and the present work.

 
3
Ar ¼ ρg ρp −ρg gdp =μ 2 ð13Þ
number increases with the increasing of particle size or density, which
may lead to an increase in interstitial gas velocity in the dense phase.
As can be observed, the Y value was found to decrease with the in- In recent years, many investigators [22,23,30,34,35,43–46] have car-
creasing of Archimedes number. It can be explained that Archimedes ried out experiments which allow to estimate the parameter Y from the

Table 5
Literature summary of experimental data on the correction factor (Y).

Reference Bed cross-section (cm) Bed height (cm) Particles Umf (cm/s) Ar Remf Y
3
Type ρp (kg/m ) dp (um)

R. D. Morse, [43] 6.35 71.12 Glass bead 2355 569 25.26 16,566 10.36 0.29–~0.60
6.35 26.92 569 27.01 11.08 0.10–~0.80
11.43 39.12 569 25.51 10.46 0.09–~0.33
6.35 63.5 452 18.29 8304 5.96 0.38–~0.62
6.35 63.5 452 18.59 6.06 0.16–~0.83
11.43 60.33 452 17.68 5.76 0.36–~0.75
6.35 63.5 285 7.92 2082 1.63 0.54–~1.22
11.43 29.54 285 8.08 1.66 0.12–~0.44
6.35 61.98 155 2.13 335 0.24 0.56–~1.54
6.35 75.57 155 0.76 0.08 1.13–~1.82
6.35 26.67 155 2.84 0.32 0.81–~1.22
11.43 61.45 155 2.59 0.29 0.10–~0.63
11.43 30.45 155 2.44 0.27 0.58–~0.82
6.35 60.96 101 0.43 93 0.03 1.14–~1.55
6.35 79.76 101 1.52 0.11 0.35–~1.75
11.43 56.52 101 1.52 0.11 0.90–~1.18
D. Geldart, [23] 30.8 21.95 Fine sand (narrow) 2600 101 1.37 102 0.1 0.58–~0.97
30.8 19.4 Fine sand (wide) 2600 128 1.4 208 0.13 0.59–~0.82
30.8 20 Coarse sand 2600 275 5.6 2063 1.11 0.38–~0.59
C. Flyer et al., [34] 22.9 11 Sand with iron oxide 2650 117 1.7 162 0.14 0.68–~0.80
22.9 23 2650 117 1.7 0.73–~0.80
22.9 40 2650 117 1.7 0.74–~0.80
22.9 65 2650 117 1.7 0.73–~0.79
A. M. Xavier et al., [35] 34 Commercial silica base catalyst 2500 84 2.6 376 0.16 0.17–~0.73
61 32 2500 158 2.6 0.3 0.28–~0.98
25 84 2.6 0.16 0.69–~0.90
P. N. Rowe et al., [30] 32.7 Silica catalyst 2600 57 2.9 18 0.12 0.17–~0.48
28 33.5 2.9 0.12 0.22–~0.54
34.8 2.9 0.12 0.17–~0.53
F. Johnsson et al., [44] 68 × 7 40 Silica sand 2600 150 2 17,418 0.22 0.55–~0.61
40 460 18 5.97 0.33–~0.45
40 790 40 22.77 0.31–~0.45
K. T. Tannous et al., [45] 43.5 19.2 Polystyrene 1016 1840 60 241,251 79.56 0.20–~0.57
D. J. Gunn et al., [46] 30 40 Diakon 1228 290 3.8 536 0.79 0.69–~0.85
A. Hepbasli, [22] 61 × 61 17.17 Raw perlite 1836 593 20.33 14,602 8.69 0.33–~0.52
10 20.33 8.69 0.16–~0.52
17.17 Sand 2486 1233 62.8 177,762 55.8 0.24–~0.36
10 62.8 55.8 0.23–~0.39
 Z. Fu et al. / Powder Technology 354 (2019) 64–70 69

measurements of fluidized bed expansion, and the experimental details

are shown in Table 5. The experimental data in literature and the pres-
0.3000 1.0
ent work, which covers the excess gas velocity below 1 m/s and Archi-
medes number ranges from 100 to 30,000, are presented in Fig. 5. An 0.4500
examination of almost all the available data reveals that the Y value reg- 0.5000
ularly increases with decreasing Archimedes number and increasing ex- 0.8
0.6000
cess gas velocity, and these published data precisely allow to develop a
correlation for evaluating the parameter Y. Accordingly, the experimen- 0.7000
tal results obtained have been fitted to the following expression 0.6
0.8000

Y

0:024 0.9500
Y ¼ 1:72Ar−0:133 U g −U mf ð14Þ
0.4
1.000

1.0
where Ar is Archimedes number which is given in Eq. 13 and the term
(Ug − Umf) represents the excess gas velocity. It should be mentioned 0.8
50
0.2
00
that the proposed correlation can be only used for Geldart B and D par- 0.6 10
00
ticles in conventional fluidized beds at ambient operating conditions. 0

Ug
0.4

15
Based on a similar approach by employing the original two-phase the-

00
U mf

0
20
0.2

(m/
ory assumption [17], the modified two-phase theory for Geldart B and

00

Ar
0
s)

25
D particles, then, can be written in the form of

00
0.0

0
30
00
−0:133
1:024

0
Gb ¼ 1:72Ar U g −U mf A ð15Þ
Fig. 7. A generalized description of the proposed correlation for predicting the correction
where Gb is the volumetric bubble flow rate and A is the cross-sectional factor (Y).
area of fluidized bed.
between the dense phase and bubble phase in a bubbling fluidized
4.3. Comparison with experimental data bed with Geldart Group B/D particles.

A comparison of the Y values calculated by using Eq. (14) with the
available experimental data in literature and the present work is il-
lustrated in Fig. 6. As can be observed, this correlation gives an over-
all standard deviation of 19% based on 156 data points in the
literature and 133 data points in the present work, and the corre-
sponding overall R-squared is 0.86, which is in reasonable agree-
ment with the experimental data. The proposed correlation only
requires the knowledge of Archimedes number and the excess gas
velocity, and it covers the widest Archimedes number range from
20 to 240,000. Therefore, this correlation has the advantages of
being considerably simpler with greater accuracy, and a more accu-
rate and reliable method for estimating the parameter Y for the mod-
ified two-phase theory has been obtained. As a result, it can be
widely used to accurately predict the distribution of gas flow
4.4. Further discussion of the correlation for correction factor Y
By plotting the calculated Y values using Eq. (14) against Archimedes
number and the excess gas velocity, a convenient graphical form for the
proposed correlation has been constructed and shown in Fig. 7. In this
graph, Archimedes number varied from 100 to 30,000 and the excess
gas velocity ranged between 0.01 and 1 m/s. As can be observed, there
is a sharp decrease in Y value from 0.9 to 0.5 with increasing Archimedes
number from 100 to 5000. However, above 5000, the Y value was found
to decrease slowly from 0.5 to 0.35 with further increasing Archimedes
number from 5000 to 30,000. Moreover, the excess gas velocity also af-
fects the correction factor Y. To be exact, the Y value, in relatively low
flow rates, increases slightly with increasing excess gas velocity. With
this knowledge, the correction factor Y can be quickly estimated for
Geldart Group B and D particles.

5. Conclusions
1.0 Literature Y values
Experiment Y values The correction factor Y for the two-phase theory of fluidization was
extensively studied for Geldart Group B and D particles. Experimental
evidences indicate that the Y value increases with decreasing particle
0.8 size or density and with increasing excess gas velocity. An equation
has been derived to predict the parameter Y for Geldart Group B and D
Predt. (Y)

particles, and can be expressed as

0.6
0:024
Y ¼ 1:72Ar−0:133 U g −U mf

It only requires the knowledge of Archimedes number and the ex-

0.4 - 0.133 0.024 cess gas velocity, and gives an overall standard deviation of 19% for
Y=1.72 Ar (Ug-Umf)
almost all available experimental data. Therefore, the proposed correla-
R-squared = 0.86 tion has the advantages of being considerably simpler and more accu-
Standard deviation = 19% rate. Furthermore, this correlation leads to a modified two-phase
0.2 theory for Geldart Group B and D particles, and can be given by
0.2 0.4 0.6 0.8 1.0

1:024
Expt. (Y) Gb ¼ 1:72Ar−0:133 U g −U mf A

Fig. 6. Comparison of the Y values calculated using Eq. (14) with all available experimental The above two correlations are shown to be as satisfactory for prac-
data. tical purposes, which can be used to accurately estimate the distribution
70 Z. Fu et al. / Powder Technology 354 (2019) 64–70

of gas flow between the dense and bubble phases in the bubbling fluid- [12] Q.Q. Sun, H.L. Lu, W.T. Liu, Y.R. He, L.D. Yang, D. Gidaspow, Simulation and experi-
ment of segregating/mixing of rice husk-sand mixture in a bubbling fluidized bed,
ized bed with Geldart B/D particles. Fuel. 84 (2005) 1739–1748.
[13] Q.G. Wang, J.F. Lu, W.D. Yin, H.R. Yang, L.B. Wei, Numerical study of gas-solid flow in
a coal beneficiation fluidized bed using kinetic theory of granular flow, Fuel Process.
Nomenclature Technol. 111 (2013) 29–41.
[14] Y. Zhang, Y.M. Zhao, L.Q. Lu, W. Ge, J.W. Wang, C.L. Duan, Assessment of polydis-
perse drag model for the size segregation in a bubbling fluidized bed using discrete
A Cross-sectional area of the fluidized bed, m2
particle method, Chem. Eng. Sci. 160 (2017) 106–112.
AD Area of single hole on the plate, m2 [15] J.F. Davidson, R. Clift, D. Harrison, Fluidization, 2nd edition, 1985 London.
Ar Archimedes number, dimensionless [16] D. Geldart, Gas Fluidization Technology, Chichester, Toronto, 1986.
dp Diameter of solid particle, m [17] R.D. Toomey, H.F. Johnstone, Gaseous fluidization of solid particles, Chem. Eng. Pro-
cess. 48 (1952) 220–226.
De Diameter of isolated bubble, m [18] D.J. Nicklin, Two-phase bubble flow, Chem. Eng. Sci. 17 (1962) 693–702.
g Gravitational acceleration, m/s2 [19] J.C.R. Turner, On bubble flow in liquids and fluidized beds, Chem. Eng. Sci. 21 (1966)
Gb Volumetric bubble flow rate, m3/s 971–974.
[20] J.F. Davidson, D. Harrison, The behavior of a continuously bubbling fluidized bed,
H Fluidized bed height at operation condition, m Chem. Eng. Sci. 21 (1966) 731–738.
Hmf Fluidized bed height at minimum fluidization state, m [21] J. Grace, R. Clift, On the two-phase theory of fluidization, Chem. Eng. Sci. 29 (1974)
n Through-flow coefficient, dimensionless 327–334.
[22] A. Hepbasli, Estimation of bed expansion in a freely-bubbling three-dimensional
Remf Reynolds number at minimum fluidization state, dimensionless gas-fluidized bed, Int. J. Energy Res. 22 (1998) 1365–1380.
Ub Visible bubble flow rate, m/s [23] D. Geldart, Expansion of gas fluidized beds, Ind. Eng. Chem. Res. 43 (2004)
Ub Average bubble flow rate, m/s 5802–5809.
[24] J.S.M. Botterill, J.S. George, H. Bespord, Bubble chains in gas fluidized beds, Chem.
Ug Superficial gas velocity, m/s
Eng. Prog. Symp. Ser. 62 (1966) 7–15.
Umf Minimum fluidization velocity, m/s [25] M.J. Lockett, J.F. Davidson, D. Harrison, On the two-phase theory of fluidization,
Vb Volume occupied by gas bubbles in fluidized bed, m3 Chem. Eng. Sci. 22 (1967) 1059–1066.
[26] D. Geldart, The expansion of bubbling fluidized beds, Powder Technol. 1 (1968)
Y Correction factor, dimensionless
355–368.
δ Fraction of the cross-sectional area occupied by gas bubbles, [27] J. Grace, D. Harrison, The behavior of freely bubbling fluidized beds, Chem. Eng. Sci.
dimensionless 24 (1969) 497–508.
ρg Density of the gas flow, kg/m3 [28] D. Geldart, The size and frequency of bubble in two- and three- dimensional gas-
fluidised beds, Powder Technol. 4 (1970) 41–55.
ρp Density of the solid particle, kg/m3 [29] D. Geldart, R.R. Granfield, Large particle fluidization, Chem. Eng. Sci. 29 (1974)
μ Viscosity of the gas flow, Pa.s 935–947.
[30] P.N. Rowe, L. Santoro, J.G. Yates, The division of gas between bubble and interstitial
phase in fluidized beds of fine powders, Chem. Eng. Sci. 33 (1978) 133–140.
Acknowledgement [31] J. Werther, Effect of gas distributor on the hydrodynamics of gas fluidized beds, Ger.
Chem. Eng. 1 (1978) 166–174.
[32] H.P. Michael, F. Liangshih, L.S. Thomas, Reactant dynamics in catalytic fluidized bed
The authors are grateful to the financial support by National Science reactors with flow reversal of gas in the emulsion phase, Chem. Eng. Sci. 37 (1982)
and Engineering Research Council of Canada, and National Natural Sci- 553–565.
ence Foundation of China (No. 51620105001). [33] K. Hilligardt, J. Werther, Local bubble gas hold-up and expansion of gas-solid fluid-
ized beds, Ger. Chem. Eng. 9 (1986) 215–221.
[34] C. Fryer, O.W. Potter, Experimental investigation of models for fluidized bed cata-
References lytic reactors, AIChE J 22 (1976) 38–47.
[35] A.M. Xavier, D.A. Lewis, J.F. Davidson, The expansion of bubbling fluidized beds,
[1] D. Kunnii, O. Levenspiel, Fluidization Engineering, 2nd edition, 1991 United States. Trans. Inst. Chem. Eng. 56 (1978) 274–280.
[2] C.Y. Wen, Handbook of Fluidization and Fluid-Particle System, Basel, New York, [36] D. Geldart, D.L. Keairns, Fluidization Technology, vol. 1, McGraw-Hill, New York,
2003. 1975 237.
[3] A.M. Botero, J.R. Grace, C.J. Lim, S.S.E.H. Elnashaie, T. Boyd, Pure hydrogen generation [37] R. Martin, Introduction to Particle Technology, Second Edition Wiley, West Sussex,
in a fluidized bed membrane reactor: application of the generalized comprehensive England, 2008.
reactor model, Chem. Eng. Sci. 64 (2009) 3826–3846. [38] R.J. Dry, M.R. Judd, T. Shingles, Two-phase theory and fine powders, Powder
[4] S. Modekurti, D. Bhattacharyya, S.E. Zitney, Dynamic modeling and control studies of Technol. 34 (1983) 213–223.
a two-stage bubbling fluidized bed adsorber-reactor for solid-sorbent CO2 capture, [39] J.W. Wang, M.A. Van der Hoef, J.A.M. Kuipers, Why the two-fluid model fails to pre-
Ind. Eng. Chem. Res. 52 (2013) 10250–10260. dict the bed expansion characteristics of Geldart a particles in gas-fluidized beds: a
[5] A. Bakshi, C. Altantzis, L.R. Glicksman, A.F. Ghoniem, Gas-flow distribution in bub- tentative answer, Chem. Eng. Sci. 64 (2009) 622–625.
bling fluidized beds: CFD-based analysis and impact of operating conditions, Pow- [40] K. Hong, S. Chen, W. Wang, J.H. Li, Fine grid two-fluid modeling of fluidization of
der Technol. 316 (2017) 500–511. Geldart a particles, Powder Technol. 296 (2016) 2–16.
[6] R. Radmanesh, J. Chaouki, C. Guy, Biomass gasification in a bubbling fluidized bed re- [41] J.F. Davidson, D. Harrison, Fluidised Particles, Cambridge Univ. Press, 1963.
actor: experiments and modeling, AIChE J 52 (2006) 4258–4272. [42] R.C. Darton, Bubble growth due to coalescence in fluidized beds, Chem. Eng. Res.
[7] A. Goyal, S. Pushpavanam, R.K. Voolapalli, Modeling and simulation of co- Des. 55 (1977) 274–280.
gasification of coal and petcoke in a bubbling fluidized bed coal gasifier, Fuel Pro- [43] R.D. Morse, Fluidization of granular solids-fluid mechanics and quality, Ind. Eng.
cess. Technol. 91 (2010) 1296–1307. Chem. 41 (1949) 1104–1117.
[8] Y.M. Geng, D.F. Che, An extended DEM-CFD model for char combustion in a bub- [44] F. Johnson, S. Andersson, B. Leckner, Expansion of a freely bubbling fluidized bed,
bling fluidized bed combustor of inert sand, Chem. Eng. Sci. 66 (2011) 207–219. Powder Technol. 68 (1991) 117–123.
[9] P. Basu, Combustion and gasification in fluidized beds. Boca, Ration, 2006. [45] K. Tannous, M. Hemati, C. Laguerie, Caracteristiques au minimum de fluidization et
[10] A. Tahmasebi, J.L. Yu, Y.N. Han, X.C. Li, A study of chemical structure changes of Chi- expansion des couches fluidisees de particules de la categorie D de Geldart, Powder
nese lignite during fluidized bed drying in nitrogen and air, Fuel Process. Technol. Technol. 80 (1994) 55–72.
101 (2012) 85–93. [46] D.J. Gunn, N. Hilal, The expansion of gas-fluidised beds in bubbling fluidization,
[11] C.S. Kannan, P.P. Thomas, Y.B.G. Varma, Drying of solids in fluidized beds, Ind. Eng. Chem. Eng. Sci. 52 (1997) 2811–2822.
Chem. Res. 34 (1995) 3068–3077.