Aces20120100013 82255519 PDF
Aces20120100013 82255519 PDF
Aces20120100013 82255519 PDF
Received November 23, 2011; revised December 20, 2011; accepted December 30, 2011
ABSTRACT
In this study the effect of initial parameters such as inlet gas temperature, initial particles temperature and gas velocity
on temperature changes of solid particles and outlet gas temperature in a fluidized bed dryer was studied. For testing, an
experimental setup was established. With combination of air and Colza seeds belonging to D groups of the Geldart
classification (Geldart, 1986) fluidization regime was carried out. With five test series with maintaining the inlet gas
temperature, solid particle temperature and outlet gas temperature during time were carefully measured. To analyze
these data by using regression analysis to predict solid particle and outlet gas temperature, 2 correlations on initial parameters were presented. The result has shown that temperature gradients in the beginning of fluidization, is very high
and therefore the exponential functions in the regression model is used to predict the temperature changes.
Keywords: Fluidized Bed; Drying Process; Colza Seeds; Heat Transfer; Regression Model
1. Introduction
Fluidization is the phenomenon in which solid particles
in a gas or a liquid type are suspended and it has many
applications in many physical, chemical industries. One
of the most prevalent implementation of this phenomenon is to dry granular seed. Fluidized bed dryers have
many usages in chemical, agricultural and medical industries. The Major reason to use such dryers in those
industries is:
1) Height heat and mass transfer coefficients due to
gas-solid contact;
2) High quality in produced products because of harmony and solid-gas proper mixture;
3) They are suitable for operations in great scale;
4) They have low service and maintenance cost;
5) Gas flow voids particles crack and fraction.
High application of such dryers has led to many researches in this field which mostly are depended an experimental equations and today many experimental equations are available to predict heat and mass transfer coefficients provided from these researches. The significant
point is that in all of these researches, every equation has
been presented in specific condition limit of fluidization
regime type so conditions dominated on problem have
significant importance to use these equations because
every equation has validity on specific domain of particles type, fluidization regime, specific pressure and temCopyright 2012 SciRes.
perature.
Geldart categorized for the time fluidization regimes
of solid-gas in to 4 groups A, B, C, and D by running
precise tests [1]. These classifications are based on density, solid particle diameters, gas density. In Figure 1
you can see related diagram about this classification.
Botteril et al. analyzed pressure effect on heat transfer
coefficient between bed and a suspended surface [2]. The
results showed that temperance transference coefficient
between a bed and a suspended surface increased as bed
pressure enhanced. Also results proved that pressure effect on heat transfer coefficient reduced as particle size is
decreased.
Hariprasad et al. evaluated temperature effect on
minimum fluidization velocity (Umf) in their study [3].
They ran the tests for 9 kinds of different particles belonged to group B of Geldart classification at thermal
range of 298 to 973 Kelvin and calculated minimum fluidization velocity and presented equations for Umf. In
performed tests, experimental data belonged to Umf
compared to other equations which were obtained by
other researchers.
Rizzi et al. [4] used a laboratory device to evaluate
heat transfer in a fluidized bed containing grass seeds
which belonged to group D of Geldart classification and
finally presented an equation to predict heat transfer coefficient based on Reyrold number. Following to Khorshidi et al. reformed modeling and applied more proper
ACES
130
ET AL.
J. KHORSHIDI
equations and used the same data to analyze heat transference phenomenon in such dryers [5].
Our purpose in this research is to evaluate variations of
temperature in one fluidized bed dryer contains Colza
seeds belonged to group D of Geldart classification by
experimental tests in order to present temperature variation which Manifests high rate of heat transfer in such
dryer.
To analyze provides data we applied statistical method
based on Regression model which is used to point out to
studies related to variables relations and it was expanded
for the first time by Karl Pearson for statistical context.
Oursize acid and Glucoseinolat as the first Kahola variations were introduced. Kanola some was registered in
1978 by Canadian oil extracting institute. Several methods are used to dry such oily seeds but Gazor et al. (2008)
proved in their study that fluidization method for this
grain drying has had developmental effect on some of
quality specifications of extracted oil like color and acidity in addition to drying time meaningful reduction.
Proportional-Integral-Derivative.
ACES
J. KHORSHIDI
ET AL.
131
0.03 m
Bed height
0.2 m
Particle sphericity
0.91
Particle diameter
1.78 mm
Solid density
1145.77 kg/m3
u (m/s)
L (m)
Tso (C)
Tgo (C)
M (kg)
1.57
0.040
23
47.7
0.012
1.57
0.040
22.5
42.5
0.012
1.57
0.040
24.1
38.5
0.012
1.96
0.050
24.1
56.5
0.012
2.36
0.057
24.1
57.5
0.012
Mean
Median
Min
Std.
Deviation
Mox
Ts
42.43
40.95
22.5
10.24
54.9
Tgl
45.67
45.7
33
7.23
55.4
J. KHORSHIDI
132
ET AL.
Table 4. Pearson Correlation between solid particles and outlet gas temperature, initial parameters and time.
Pearson
Correlation Coefficient
Ts
Tgl
Ts
**0.84
**0.57
0.000
1
Tg
u0
Time
**0.57
**0.65
*0.37
0.000
0.000
0.000
0.020
**0.83
**0.84
**0.95
0.17
0.000
0.000
0.000
0.296
**0.63
**0.86
0.000
0.000
1.000
**0.87
0.000
1.000
Ts
P-Value. 2-tailed
Correlation Coefficient
**0.84
P-Value. 2-tailed
0.000
Correlation Coefficient
**0.57
**0.83
P-Value. 2-tailed
0.000
0.000
Correlation Coefficient
**0.57
**0.84
**0.63
P-Value. 2-tailed
0.000
0.000
0.000
Correlation Coefficient
**0.65
**0.95
**0.86
**0.87
P-Value. 2-tailed
0.000
0.000
0.000
0.000
Correlation Coefficient
*0.37
0. 17
P-Value. 2-tailed
0.020
0.296
1.000
1.000
1.000
*0.37
Tgl
Ts
Tg
u0
1.000
Time
**P < 0.01, *P < 0.05.
Table 5. Regression correlation between solid particles temperature with initial parameter and time.
Step
Variable
Coefficients
Statistic t
P-Value
constant
6866.3
1.9
0.062
0.871
**32.8
0.000
constant
3
s0
149728.9
**7.3
0.000
3
g0
0.871
**15.5
0.000
Tanh time
145901.4
**7.1
0.000
constant
149136.9
**7.7
0.000
Tg3
0.812
**25.9
0.000
Tanh time
145901.4
**7.4
0.000
u3
1037.6
*2.2
0.036
constant
402397.2
**5.7
0.000
3
g0
0.566
**7.9
0.000
Tanh time
145901.4
**8.9
0.000
2846.6
**4.5
0.000
11713.1
**3.7
0.001
T
4
Ts
Statistic F
P-Value
**1072.9
0.000
0.970
**1350.1
0.000
0.988
**1009.1
0.000
0.990
**1072.9
0.000
0.993
1.55
accept
final step t certain number has been calculated for Regression coefficients in = 0.01 which had been meaningful and meaninglessness assumption about Regression
coefficients are rejected, so Regression model could be
written as below:
Copyright 2012 SciRes.
(1)
11713.1Ts0
J. KHORSHIDI
are provided by the implementation of primary parameters and time to analyze residual independency, DurbinWatson statistic is used on a way that if this statistic rate
is more than 1.5 and lower than 2.5 we can accept residuals independence. Because Durbin-Watson statistic is
obtained as 1.55 we can conclude that residuals independency assumption is acceptable. To analyze residual
normalization, Kolmogrov-Smirnov test has been used
meaningful level was 0.274 about this test which shows
accepted assumption about residuals normalization. In
Figure 3 you can see results of simulation for solid particles temperature beside experimental data which are
drawn for tests 1, 4.
Note Mentioned figure is made by stepwise method
application. In first step of Regression, inlet gas primary
temperature variable which has the highest meaningful
correlation which solid articles temperature is entered in
to the model, in second step time variable, in third step
velocity variable and in fourth step temperature variable
for solid particles are added to Regression model. As it is
obvious in Table 6 F rate is calculated in fourth step for
(a)
ET AL.
133
(2)
11332.1Ts0
(b)
Figure 3. Comparison between experimental and simulated data for solid particle temperature: (a) Test #1; (b) Test #4.
(a)
(b)
Figure 4. Comparison between experimental and simulated data for outlet gas temperature: (a) Test #1; (b) Test #4.
ACES
J. KHORSHIDI
134
ET AL.
Table 6. Regression correlation between outlet gas temperature with initial parameter and time.
Step
Variation
Coefficients
Statistic t
P-Value
Constant
179416.2
**10.5
0.000
Tgl
5803.5
**16.6
0.000
Constant
213106.6
**19.6
0.000
Tgl
5803.5
**28.1
0.000
Tanh(time)
40078.2
**8.5
0.000
Constant
214596.4
**21.1
0.000
Tgl
4974.6
**13.1
0.000
Tanh(time)
40078.2
**9.1
0.000
23128.7
*2.5
0.016
Constant
430520.6
**5.5
0.000
Tgl
3459.8
**5.7
0.000
Tanh(time)
40078.2
**9.9
0.000
35422.4
**3.7
0.001
Ts
11332.1
**2.8
0.009
Statistic F
P-Value
R Square
**276.7
0.000
0.879
**430.8
0.000
0.959
**331.1
0.000
0.965
**296.2
0.000
0.971
Durbin-Watson Accept/Reject
1.53
Accept
4. Conclusions
In this research we have used an experimental setup to
evaluate initial parameters effects on temperature variation process in solid-gas fluidized bed dryers in which a
fluidization regime is made by air and solid combination
and 5 series of test performed to analyze bed internal
temperature variation and in every test inlet gas temperature is kept steady with high precision and solid particles and outlet gas temperature were registered during
time, then Regression method applied to analyze experimental data. Results obtained from this study shows: 1)
The maximum solid particles temperature variation and
outlet gas temperature variation occur at the beginning of
fluidization which show high heat transfer in these kinds
of dryers. 2) Because of high temperature variations
curve declivity, exponential functions are used to predict
temperature variation in Regression model. Equations
precision is very high to the point Regression equation
given for solid particles temperature variation bas 99%
conformity with experimental data precisely and presented equation accuracy for outlet gas temperature is
about 97%. 3) Solid particles temperature variation velocity during time is more than variations related to outlet
gas. 4) After passage from unsteady condition, solid particles temperature and outlet gas temperature from bed
inclined to each other. 5) The most effective parameter
on heat transfer, is the inlet gas temperature to bed and
Copyright 2012 SciRes.
REFERENCES
[1]
[2]
[3]
[4]
[5]
J. Khorshidi, H. Davari and F. Dehbozorgi, Model Making for Heat Transfer in a Fluidized Bed Dryer, Journal
of Basic & Applied Sciences, Vol. 1, No. 10, 2011, pp.
1732-1738.
[6]
S. Minaei and E. Hazbavi, Determination and Investigation of Some Physical Properties of Seven Variety Rapeseed, Iranian Journal of Food Science and Technology,
Vol. 5, No. 4, 2008, pp. 21-28.
ACES
J. KHORSHIDI
ET AL.
135
Nomenclature
Inlet gas temperature
Tg 0
Tgl
Ts
Ts0
Particle diameter
dp
mm
Particle density
kg/m3
Gas density
kg/m3
u0
m/s
umf
m/s
kg
Bed high
ACES